Supreme Court of the United States - SCOTUSblogNo. _____ IN THE Supreme Court of the United States FEESERS, INC., Petitioner, —v.— MICHAEL FOODS, INC., and SODEXHO, INC., Respondents

No. ______

IN THE

Supreme Court of the United States

FEESERS, INC.,Petitioner,

—v.—

MICHAEL FOODS, INC., and SODEXHO, INC.,

Respondents.

ON PETITION FOR A WRIT OF CERTIORARI TO THE UNITED STATESCOURT OF APPEALS FOR THE THIRD CIRCUIT

PETITION FOR A WRIT OF CERTIORARI

KANNON K. SHANMUGAM

WILLIAMS & CONNOLLY LLP725 Twelfth Street, N.W.Washington, D.C. 20005(202) 434-5000

d

JEFFREY L. KESSLER

Counsel of RecordJOHN F. COLLINS

EAMON O’KELLY

SUSANNAH P. TORPEY

DEWEY & LEBOEUF LLP1301 Avenue of the AmericasNew York, New York 10019(212) [email protected]

Counsel for Petitioner

June 2, 2010

QUESTION PRESENTED

Under Section 2(a) of the Robinson-Patman Act,15 U.S.C. § 13(a), a plaintiff in a secondary-line ortertiary-line price discrimination case may estab-lish competitive injury by proving that �“a favoredcompetitor received a significant price reductionover a substantial period of time.�” Volvo Trucks N.Am., Inc. v. Reeder-Simco GMC, Inc., 546 U.S.164, 177 (2006).

The question presented is:Whether, in order to support a finding of com-

petitive injury under the Robinson-Patman Act, aplaintiff must also prove that the favored and dis-favored purchasers bought the discriminatorilypriced products at the exact same moment atwhich they or their customers competed to resellthose products.

i

26922 • Dewey: Fessers • USSC (revised 12-1-08) • LJB 5/28/10 12:00

CORPORATE DISCLOSURE STATEMENT

Feesers, Inc., has no parent corporation, and nopublicly held corporation owns 10% or more of itsstock.

ii

26922 • Dewey: Fessers • USSC (revised 12-1-08) • LJB 5/28/10 12:00; KW 6/1 7:15 pm

iii

26922 • FDewey: Fessers • USSC TCA (revised 12/1/08) • LJB • 12:00 • 5-28-2010; CRS LJB 5/28

QUESTION PRESENTED . . . . . . . . . . . . . . . . . . . i

CORPORATE DISCLOSURE STATEMENT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii

TABLE OF AUTHORITIES . . . . . . . . . . . . . . . . . . v

OPINIONS BELOW . . . . . . . . . . . . . . . . . . . . . . . . . . 1

JURISDICTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

STATUTORY PROVISION INVOLVED . . . . 1

STATEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

REASONS FOR GRANTING THE PETITION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

A. The Third Circuit�’s Timing Requirement Creates A Fundamental Conflict With The Case Law Of Other Circuits . . . . . . . . 15

B. The Third Circuit�’s Novel TimingRequirement Is Contrary To Sixty Years Of This Court�’s PrecedentsConcerning The Requirement OfCompetitive Injury In A PriceDiscrimination Case . . . . . . . . . . . . . . . . 21

TABLE OF CONTENTSPAGE

ivPAGE

C. The Question Presented Is AnExceptionally Important One ThatWarrants This Court�’s Review . . . . . 29

CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

APPENDIX AFeesers, Inc. v. Michael Foods, Inc. and Sodexho, Inc., 498 F.3d 206 (3d Cir. 2007) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1a

APPENDIX BFeesers, Inc. v. Michael Foods, Inc. andSodexho, Inc., 632 F. Supp. 2d 414 (M.D. Pa. 2009). . . . . . . . . . . . . . . . . . . . . . . . . . . 34a

APPENDIX CFeesers, Inc. v. Michael Foods, Inc. andSodexho, Inc., Civ. No. 1:CV-04-0576, 2009 WL 5226916 (M.D. Pa. June 30, 2009) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141a

APPENDIX DFeesers, Inc. v. Michael Foods, Inc. andSodexho, Inc., 591 F.3d 191 (3d Cir. 2010) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152a

APPENDIX EFeesers, Inc. v. Michael Foods, Inc. andSodexho, Inc., Nos. 09-2548, 09-2952, 09-2993, slip op. (3d Cir. Mar. 4, 2010)... 191a


v

Caribe BMW, Inc. v. Bayerische Motoren Werke Aktiengesellschaft, 19 F.3d 745 (1st Cir. 1994) . . . . . . . . . . . . . . 30-31

Chroma Lighting v. GTE Prods. Corp., 111 F.3d 653 (9th Cir. 1997) . . . . . . . . . . . . 15, 18

Coastal Fuels v. Caribbean Petroleum Corp., 79 F.3d 182 (1st Cir. 1996) . . . . . . . . . . 15-16, 18

DeLong Equip. Co. v. Washington Mills Electro Minerals Corp., 990 F.2d 1186 (11th Cir. 1993) . . . . . . . . . 16-18

Falls City Indus., Inc. v. Vanco Beverage, Inc., 460 U.S. 428 (1983) . . . . . . . . . . . . . . . . . . 23, 24, 29

Feesers, Inc. v. Michael Foods, Inc. and Sodexho, Inc., 591 F.3d 191 (3d Cir. 2010)(App. 152a-190a) . . . . . . . . . . . . . . . . . . . . . . . . passim

Feesers, Inc. v. Michael Foods, Inc. and Sodexho, Inc., 498 F.3d 206 (3d Cir. 2007)(App. 1a-33a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . passim

Feesers, Inc. v. Michael Foods, Inc. and Sodexho, Inc., Civ. No. 1:CV-04-0576, 2009 WL 5226916 (M.D. Pa. June 30, 2009)(App. 141a-151a) . . . . . . . . . . . . . . . . . . . . . . . . . 11, 25

TABLE OF AUTHORITIESPAGECases:


viPAGE

Feesers, Inc. v. Michael Foods, Inc. and Sodexho, Inc., 632 F. Supp. 2d 414 (M.D. Pa. 2009)(App. 34a-140a) . . . . . . . . . . . . . . . . . . . . . . . . . passim

Flood v. Kuhn, 407 U.S. 258 (1972) . . . . . . . . . . . . . . . . . . . . . . 29, 32

FTC v. Morton Salt Co., 334 U.S. 37 (1948) . . . . . . . . . . . . . . . . . . . . . . passim

Goldfarb v. Virginia State Bar, 421 U.S. 773 (1975)) . . . . . . . . . . . . . . . . . . . . . 30

Hartley & Parker, Inc. v. Fl. Beverage Corp., 307 F.2d 916 (5th Cir. 1962) . . . . . . . . . . . . 18

Hasbrouck v. Texaco, Inc., 842 F.2d 1034 (9th Cir. 1987) . . . 16, 18-19, 20

Innomed Labs, LLC v. Alza Corp., 368 F.3d 148 (2d Cir. 2004) . . . . . . . . . . . . . 19

Jefferson County Pharm. Ass�’n v. Abbott Labs., 460 U.S. 150 (1983). . . . . . . . . . . . . . . 29-30, 31-32

Mid-South Distribs. v. FTC, 287 F.2d 512 (5th Cir. 1961) . . . . . . . . . . . . 20

Moog Indus., Inc. v. FTC, 238 F.2d 43 (8th Cir. 1956) . . . . . . . . . . . . . 20

Perkins v. Standard Oil Co., 395 U.S. 642 (1969) . . . . . . . . . . . . . . . . . . 24-25, 31

Rose Confections, Inc. v. Ambrosia Chocolate Co., 816 F.2d 381 (8th Cir. 1987) . . . . . . . . 16, 19-20


viiPAGE

Stelwagon Mfg. Co. v. Tarmac Roofing Sys., Inc., 63 F.3d 1267 (3d Cir. 1995) . . . . . . . . . . . . . 8

Texaco Inc. v. Hasbrouck, 496 U.S. 543 (1990) . . . . . . . . . . . . . . . . . . . . . passim

Toledo Mack Sales & Serv., Inc. v. Mack Trucks, Inc., 530 F.3d 204 (3d Cir. 2008) . . . . . . . . . . . . . 11, 31

U.S. v. Cooper Corp., 312 U.S. 600 (1941) . . . . . . . . . . . . . . . . . . . . . . 32

U.S. v. Int�’l Boxing Club, 348 U.S. 236 (1955) . . . . . . . . . . . . . . . . . . . . . . 32

Volvo Trucks N. Am., Inc. v. Reeder-Simco GMC, Inc., 546 U.S. 164 (2006) . . . . . . . . . . . . . . . . . . . . . passim

Statutes:

28 U.S.C. § 1254(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1Robinson-Patman Act, 15 U.S.C. § 13 . . . . . passim

Other Authorities:

ABA Antitrust Section, Monograph No. 4, The Robinson-Patman Act: Policy and Law Volume I (1980) . . . . . . . . 14

26922 • FDewey: Fessers • USSC TCA (revised 12/1/08) • LJB • 12:00 • 5-28-2010; CRS LJB 5/28;KW 6/1 12 pm

PETITION FOR A WRIT OF CERTIORARI

Feesers, Inc., respectfully petitions for a writ ofcertiorari to review the judgment of the UnitedStates Court of Appeals for the Third Circuit inthis case.

OPINIONS BELOW

The opinion of the court of appeals reversing thedistrict court�’s summary judgment is reported at498 F.3d 206 (App., infra, 1a-33a) and the opinionof the court of appeals reversing the districtcourt�’s trial verdict is reported at 591 F.3d 191(App., infra, 152a-190a). The district court�’s judg-ment and order following trial is reported at 632F. Supp. 2d 414 (App., infra, 34a-140a). The dis-trict court�’s decision and order denying defen-dants�’ motions for reconsideration (App., infra,141a-151a) is unreported.

JURISDICTION

The court of appeals entered judgment on Jan-uary 7, 2010 (App., infra, 152a), and denied peti-tioner�’s petition for rehearing on March 4, 2010(App., infra, 191a-192a). The jurisdiction of thisCourt is invoked pursuant to 28 U.S.C. § 1254(1).

STATUTORY PROVISION INVOLVED

Section 2(a) of the Robinson-Patman Act (RPAor Act), 15 U.S.C. § 13(a), provides in pertinentpart:

It shall be unlawful for any personengaged in commerce, in the course of


such commerce, either directly or indi-rectly, to discriminate in price betweendifferent purchasers of commodities of likegrade and quality, where either or any ofthe purchases involved in such discrimi-nation are in commerce, where such com-modities are sold for use, consumption, orresale within the United States or anyTerritory thereof or the District ofColumbia or any insular possession orother place under the jurisdiction of theUnited States, and where the effect ofsuch discrimination may be substantiallyto lessen competition or tend to create amonopoly in any line of commerce, or toinjure, destroy, or prevent competitionwith any person who either grants orknowingly receives the benefit of such dis-crimination, or with customers of either ofthem: . . . .

STATEMENT

Petitioner, Feesers, Inc., filed a price discrimi-nation lawsuit seeking: (1) a declaration thatrespondent Michael Foods, Inc., unlawfully dis-criminated as to price in favor of respondentSodexho, Inc., and against Feesers, in violation ofSection 2(a) of the Robinson-Patman Act, and thatSodexho knowingly induced or received suchunlawful price discriminations, in violation of Sec-tion 2(f), and (2) injunctive relief against the con-tinuation of that discrimination. Following aten-day trial, the district court entered judgmentin favor of Feesers and issued an injunctionagainst respondents. Respondents moved for

2


reconsideration, but the district court denied themotion. The court of appeals reversed, holdingthat Feesers and Sodexho �“were not competingpurchasers�” as a matter of law because the com-petition between Feesers and Sodexho for any spe-cific customer occurred before the exact momentat which Sodexho purchased Michael Foods prod-ucts at discriminatory prices for resale to that cus-tomer (i.e., what the court of appeals called the�“timing of the competition�” requirement). App.,infra, 156a-157a, 165a, 178a-179a. The court ofappeals held that Feesers therefore could notshow that it suffered competitive injury under theRPA, and instructed the district court to enterjudgment as a matter of law for respondents. Id.at 165a, 178-179a, 190a.

1. This case presents a substantial circuit con-flict on an issue that has broad practical implica-tions for all non-retail businesses, including those,like Feesers, in the food distribution industry. Thecourt of appeals�’ �“timing of the competition�”requirement amounts to a judicial repeal of theRPA for non-retail businesses, including the fooddistribution industry. Other circuits have refusedto read a �“timing of the competition�” requirementinto the RPA, and have instead consistently ana-lyzed competitive injury for purposes of the RPAin accordance with this Court�’s decision in FTC v.Morton Salt Co., 334 U.S. 37 (1948), which did notinclude any such timing requirement. This Court,moreover, has consistently applied the MortonSalt competitive-injury test for more than sixtyyears, holding that a plaintiff �“need only provethat a seller had charged one purchaser a higherprice for like goods than he had charged one ormore of the purchaser�’s competitors.�” Morton Salt,

3


334 U.S. at 45 (emphasis added). The practicaleffect of the court of appeals�’ decision is to createa judicial exemption from the RPA for certainindustries�—an exemption that it is the role ofCongress, not the courts, to create.

2. Petitioner Feesers is a broadline distributorof food products, servicing institutional food cus-tomers within 200 miles of Harrisburg, Pennsyl-vania. App., infra, 42a. Respondent Sodexho is afood-service management, procurement, and dis-tribution company, and the world�’s largest pur-chaser of food. Id. at 42a-43a. Sodexho competeswith Feesers to procure food products for sale toinstitutional customers in the same geographicarea. Id. at 43a-44a. Respondent Michael Foods isthe largest supplier of processed eggs and potatoesin the United States. Michael Foods supplies eggand potato products to intermediaries, such asFeesers and Sodexho, for resale to institutionalfood customers. Id. at 42a.

The institutional food-service business involvesthe sale of food and food-related products and ser-vices to institutions such as schools, colleges anduniversities, and healthcare facilities. App., infra,40a, 43a. Product suppliers, such as MichaelFoods, manufacture food products. Id. at 42a.Broadline distributors, such as Feesers and SyscoCorporation (a contract distributor for Sodexho),procure and distribute food products sold by prod-uct suppliers and resell them to institutional cus-tomers. Id. at 42a-43a. Food-service managementcompanies, such as Sodexho, provide various ser-vices and products to institutional customers,including the same kind of food procurement anddistribution services that are provided to insti-

4


tutional customers by broadline distributors likeFeesers. Id. at 42a-44a, 62a, 86a-87a.

Institutions that serve meals have several com-petitive options for purchasing food products.They can either perform food management services internally (known as �“self-operation�” or�“self-op�”) or outsource those functions to a foodmanagement company. App., infra, 41a. One com-petitive option is for an institution to have mealsprepared and served by its own employees, and topurchase the food products used in those mealsfrom distributors, such as Feesers. Id. at 41a, 43a.Distributors generally purchase food productsfrom product suppliers at standard list prices, anddeliver and resell those products to the institu-tions at prices agreed upon by the distributors andtheir institutional customers. Id. at 42a-44a, 67a.

Another competitive option is for institutions tocontract with a food management company, suchas Sodexho, which provides both food managementand food procurement and distribution services toits customers. App., infra, 42a-44a. To supplyfood, Sodexho generally enters into contracts withproduct suppliers like Michael Foods whereby itprocures food at special, highly discounted pricesand arranges for its resale and delivery to itsinstitutional customers. Id. at 44a.

A third competitive option is a hybrid betweenthe first two: namely, for the institutions to pre-pare and serve meals themselves while outsourc-ing the food procurement function to a grouppurchasing organization (�“GPO�”). App., infra, 44a.GPOs negotiate the prices at which food productsare acquired from product suppliers, and then pro-vide such food products to institutions in compe-

5


tition with both food management companies anddistributors. Id. at 43a-44a, 51a-52a. Sodexhoowns and operates its own GPO, Entegra. Id. at56a.

Food management companies and GPOs (suchas Sodexho and its Entegra division) do not typi-cally warehouse and deliver products using theirown employees, but instead subcontract thesefunctions to chosen distributors. App., infra, 62a-63a. Specifically, Sodexho contracts with Sysco topurchase and deliver food products to Sodexho�’sinstitutional customers at prices that Sodexhonegotiates. Id. at 6a-7a, 43a-44a.

Institutional customers sometimes negotiatedirectly with product suppliers like Michael Foodsfor lower prices that �“deviate�” from the standardlist prices. See App., infra, 67a-69a. Distributorsresell those products to the institutions at these�“deviated�” prices plus a distribution fee, and �“billback�” to the supplier the difference between thelist prices and the deviated prices. Id. at 6a, 67a,71a n.6. Such customer-specific, deviated pricesare not discriminatory because they are availableto all companies that distribute food products tothese customers. Id. at 68a-69a.

Michael Foods, however, also negotiated muchlower deviated prices with Sodexho, whichSodexho then used to gain a competitive advan-tage over distributors like Feesers. App., infra,69a, 85a; see also id. at 43a-66a. Specifically,Sodexho contracted with product suppliers such asMichael Foods for the purchase of food products atuniquely low prices that were available for resaleto all of Sodexho�’s customers. Id. at 6a-7a, 69a,85a. Michael Foods (and other product suppliers)

6


sell these food products at the special deviatedprices to Sodexho�’s contract distributor, Sysco,which delivers the products to Sodexho�’s institu-tional customers. Id. Sysco generally invoicesSodexho for the cost of the food at the Sodexho-negotiated, deviated prices plus Sysco�’s distribu-tion fee, and Sodexho in turn generally bills thecost of the food to its institutional customers. Id.

3. In 2004, Feesers filed a price discriminationlawsuit seeking: (1) a declaration that MichaelFoods unlawfully discriminated as to price infavor of Sodexho and against Feesers, in violationof Section 2(a) of the Robinson-Patman Act, andthat Sodexho knowingly induced or received suchunlawful price discriminations, in violation of Sec-tion 2(f) of the Act, and (2) injunctive reliefagainst the continuation of that discrimination.App., infra, 162a. Feesers alleged that, becauseSodexho was able to use its massive purchasingpower to extract discriminatory prices fromMichael Foods and other suppliers, Feesers was ata severe disadvantage when competing withSodexho to sell food to institutional customers. Seeid. at 87a.

In 2006, the district court granted summaryjudgment in favor of respondents. The court deter-mined that Feesers had established three of thefour elements of its Section 2(a) claim: (1) MichaelFoods sold food products in interstate commerce totwo different purchasers, Feesers and Sysco(Sodexho�’s contract distributor); (2) the productssold were of the same grade and quality; and (3)Michael Foods discriminated as to price againstFeesers and in favor of Sodexho. See App., infra,9a. The district court concluded, however, that

7


Feesers could not establish the fourth element ofits RPA claim, competitive injury. See id. at 11a,15a-22a.

The court of appeals reversed. It did not disturbthe district court�’s determination that Feesers hadestablished the first three elements of its Section2(a) claim. App., infra, 9a-10a. The court ofappeals held, however, that with respect to thefourth element, competitive injury, �“[t]he DistrictCourt required Feesers to prove too much.�” Id. at15a. The court reasoned that, in order to establishcompetitive injury, �“Feesers need only prove that(a) it competed with Sodexho to sell food and (b)there was price discrimination over time byMichael Foods.�” Id. (citing Volvo Trucks N. Am.,Inc. v. Reeder-Simco GMC, Inc., 546 U.S. 164, 177(2006)).

The majority rejected the view of a dissentingjudge that Feesers and Sodexho could not beshown to be in competition, as a matter of law,because Sodexho�’s business is of a �“different char-acter�” than Feesers�’. App., infra, 17a n.9. The rel-evant question, the majority held, �“is whether twocompanies are �‘in economic reality acting on thesame distribution level,�’ rather than whether theyare both labeled as �‘wholesalers�’ or �‘retailers.�’�” Id.at 16a (quoting Stelwagon Mfg. Co. v. TarmacRoofing Sys., Inc., 63 F.3d 1267, 1272 (3d Cir.1995)). The issue of fact to be determined by thedistrict court at trial, therefore, was whetherFeesers and Sodexho �“are each directly after thesame dollar.�” Id. at 17a.

4. The case proceeded to trial. During the ten-day trial, the district court was presented withextensive evidence, including the testimony of

8


respondents�’ executives, Feesers�’ employees, andten customers hand-picked by Sodexho; some11,000 pages of exhibits, including thousands ofpages of Sodexho documents; and the unrebuttedexpert testimony of Feesers�’ economic expert. See,e.g., App., infra, 39a-62a, 67a-85a, 89a-112a. Thedistrict court subsequently issued a detailed opin-ion setting forth the factual basis for its findingsthat, among other things, (1) Feesers competedwith Sodexho to sell Michael Foods products, and(2) there was �“stunning�” price discrimination overtime by Michael Foods in favor of Sodexho andagainst Feesers. Id. at 43a-66a, 67a-88a. As aresult, the district court held that Feesers wasentitled to the inference of competitive injuryestablished in FTC v. Morton Salt Co., 334 U.S. 37,46-47 (1948). App., infra, 88a. The district courtalso held that respondents had failed to rebut thisinference of competitive injury or to prove any affir-mative defenses. Id. at 89a-132a. The court thusentered judgment in favor of Feesers and issued aninjunction. Id. at 139a-140a.

Specifically, the district court found that at theheart of the competition between Feesers andSodexho were Sodexho�’s continuous efforts to con-vert self-operated institutions to food-service man-agement and the corresponding efforts bydistributors, such as Feesers, to hold on to theirself-op customers and to win the business ofSodexho-managed institutions. See, e.g., App.,infra, 43a-48a. In particular, the court found that:

when an institution switches from self-opto management, the incumbent distributoris displaced. Conversely when a managedinstitution switches to self-op, the func-tions previously performed by a food ser-

9


vice management company, including thesale and delivery of food, are once againperformed by a distributor. Accordingly,Feesers and Sodexho compete for thesame portion of an institution�’s food ser-vice budget.

Id. at 48a.Moreover, the district court specifically exam-

ined the �“timing of the competition�” and foundthat �“[f]ood service management companies, dis-tributors and GPOs all compete formally andinformally for the sale of food to institutions�” andthat this competition �“was ongoing and not limitedto the formal RFP [request for proposal] process.�”App., infra, 59a, 66a. The court found that insti-tutional customers sometimes �“use the RFP pro-cess to gain ideas, but remain�” with a distributor,instead of hiring a food management company. Id.at 59a. Institutional customers do not always staywith a food management company for long, butrather �“use [them] to �‘fix�’ current problems andthen return to self-op.�” Id. Sodexho itself had pre-pared numerous �“churn reports�” that tracked theback-and-forth competition for institutional cus-tomers between food management companies anddistributors. Id. at 46a-47a.

There was also overwhelming evidence sup-porting the district court�’s finding that MichaelFoods had engaged in what the district court char-acterized as �“stunning�” price discrimination in favor of Sodexho and against Feesers. App.,infra, 73a; see id. at 67a-85a. As the court found,Sodexho used its massive purchasing power toobtain uniquely low prices from suppliers likeMichael Foods, which it then expressly promoted

10


when competing for institutional customers withdistributors such as Feesers. Id. at 102a-112a.This strategy to use its massive purchasing poweras a competitive advantage against distributorslike Feesers was set forth in numerous Sodexhodocuments, and was confirmed by the testimony ofSodexho executives. Id. at 98a-112a. MichaelFoods executives similarly testified that Sodexho,as �“the big dog[ ] in contract management andhealthcare,�” used its purchasing leverage toextract the lowest prices and highest rebates fromMichael Foods. Id. at 128a.

Respondents moved for reconsideration, arguingthat the institutional food business is a non-retail�“bid market�” akin to the markets for customizedtrucks in Volvo Trucks, 546 U.S. 164, and ToledoMack Sales & Serv., Inc. v. Mack Trucks, Inc., 530F.3d 204 (3d Cir. 2008). The district court deniedthis motion, finding that institutional food dis-tribution is not a �“bid market.�” App., infra, 141a-151a. As the court explained, in Volvo Trucks, �“thelosing bidder would never actually purchase theitem which was the subject of the competition.�” Id.at 146a. There were thus never two purchasers ofthe discriminatorily priced products. Id. Respon-dents provided no basis to extend this holding ofVolvo Trucks to �“cases such as this, where thegoods in question are perishable commodities thattwo competitors regularly purchase and keep instock for resale to customers.�” Id.

5. In the decision under review, the court ofappeals reversed, holding that Feesers andSodexho �“were not competing purchasers�” as amatter of law. App., infra, 155a-156a. The court ofappeals reasoned that, despite the fact that both

11


Feesers and Sodexho purchased Michael Foodsproducts and competed to resell those products tothe same customers, there could be no RPA claimin a non-retail industry like wholesale food dis-tribution because Sodexho did not purchaseMichael Foods products for resale to a particularinstitutional customer until after that institutionhad decided whether to purchase the productsfrom Feesers or from Sodexho. Id. at 156a-157a,178a-179a. According to the court of appeals, sucha non-retail industry could be classified as a �“bidmarket,�” in which, according to the court ofappeals, this Court�’s decision in Volvo Trucks sup-ported the imposition of a �“timing of the competi-tion�” requirement. Id. at 181a-187a. Based uponthis unprecedented hurdle for proving competitiveinjury, the court of appeals held that Feesers andSodexho were not �“competing purchasers�” for pur-poses of the RPA. Id. at 156a-157a, 165a, 178a-179a.

Although the court of appeals acknowledgedthat such a narrow construction of the RPAresulted in �“elevat[ing] form over substance,�” itstated that it must �“dutifully follow[ ] theSupreme Court�’s lead by narrowly construing theRPA.�” App., infra, 168a. The court of appeals didnot explain how its new timing requirementrelated to proving competitive injury in the whole-sale food industry, especially in the face of the dis-trict court�’s factual findings that Sodexho hadpromoted and used the discriminatory prices itnegotiated with Michael Foods and other suppliersto compete with Feesers and other distributors forinstitutional customers that were promised bySodexho that they would receive the benefits ofthese discriminatory prices only if they acquired

12


the food products through Sodexho. Nor did thecourt of appeals explain why non-retail industriesshould be treated differently under the RPA fromother industries when neither Congress nor thecourts interpreting the Act have ever made such adistinction.

6. Petitioner filed a petition for rehearing,which was denied without recorded dissent. App.,infra, 191a-192a.

REASONS FOR GRANTING THE PETITION

In the decision under review, the Third Circuitadopted a novel timing requirement for provingcompetitive injury that does not appear in the textof the Robinson-Patman Act. In so doing, theThird Circuit created an irreconcilable conflictwith the law in other circuits, which haveexpressly rejected the argument that the RPA�’scompetitive-injury element has a timing require-ment. The Third Circuit�’s new barrier to recoveryin RPA cases is inconsistent not only with thedecisions of other circuits, but also with more thansixty years of this Court�’s precedents for provingcompetitive injury in a secondary-line or tertiary-line price discrimination case, as most recentlyreaffirmed in Volvo Trucks. Moreover, by creatinga new insurmountable hurdle for proving com-petitive injury in wholesale food distribution andother non-retail industries or what it termed �“bidmarkets,�” the Third Circuit�’s decision subverts thecongressional policies embodied in the RPA.Indeed, the Third Circuit�’s timing requirementhas made it impossible for plaintiffs, like Feesers,to prosecute price discrimination claims in thefood distribution industry�—the very industry,

13


ironically, in which concerns about price dis-crimination sparked the passage of the RPA in thefirst place. See ABA Antitrust Section, MonographNo. 4, The Robinson-Patman Act: Policy and LawVolume I, 9 (1980). Nothing in this Court�’s prece-dents permits, let alone directs, the Third Circuitto create such a barrier to recovery under theRPA.

The need for the Court to provide additionalguidance in this case is particularly compelling inthis context; indeed, the Third Circuit itselfadmitted that it found the RPA to be confusing,and that other lower courts have experienced dif-ficulties in applying the competitive-injuryrequirement of the RPA. App., infra, 184a n.17.That confusion will now be substantially com-pounded by the circuit conflict created by theThird Circuit�’s creation of a timing requirement.While the Third Circuit now requires that plain-tiffs in an RPA case prove that the favored com-pany purchased the discriminatorily pricedproducts at precisely the same time at which thecompetition occurred, other circuits have notdeparted from the traditional test for competitiveinjury, which does not focus on the precisemoment when the discriminatory sales are engi-neered, whether or not non-retail purchases or�“bid markets�” are involved. In short, this case sat-isfies all of the traditional criteria for certiorari,and the petition should therefore be granted.

14


A. The Third Circuit�’s Timing Require-ment Creates A Fundamental ConflictWith The Case Law Of Other Circuits

Until now, no court of appeals has held that aplaintiff satisfying the Morton Salt test cannotprove competitive injury under the RPA based onthe timing of competition. The Third Circuit�’s newtiming requirement thus represents a radicaldeparture from the established case law of theother circuits, and the resulting conflict warrantsthis Court�’s review.

1. Before the decision below, the federal courtshad uniformly held that, in order to prove a vio-lation of Section 2(a) of the RPA, a secondary-lineor tertiary-line plaintiff need only establish: (1)that there were sales to two different purchasersin interstate commerce; (2) that the products soldwere of the same grade and quality; (3) that therewas discrimination in price; and (4) that the pricediscrimination had a prohibited effect on compe-tition. See, e.g., Texaco Inc. v. Hasbrouck, 496 U.S.543, 556 (1990). In FTC v. Morton Salt Co., 334U.S. 37, 46-47 (1948), this Court established thatthe fourth element, competitive injury, could beproven either through evidence of lost sales orprofits, or by showing that (1) the plaintiff and thebeneficiary of the price discrimination were incompetition, and (2) there was substantial pricediscrimination over time. See, e.g., Volvo Trucks,546 U.S. at 177; Texaco, 496 U.S. at 556.

The courts of appeals have consistently appliedthe Morton Salt test without imposing any addi-tional timing requirement. See Chroma Lightingv. GTE Prods. Corp., 111 F.3d 653, 654 (9th Cir.1997); Coastal Fuels v. Caribbean Petroleum

15


Corp., 79 F.3d 182, 191, 193 (1st Cir. 1996);DeLong Equipment Co. v. Washington Mills Elec-tro Minerals Corp., 990 F.2d 1186, 1201-02 (11thCir. 1993); Hasbrouck v. Texaco, Inc., 842 F.2d1034, 1041 (9th Cir. 1987), aff�’d, Texaco, 496 U.S.543; Rose Confections, Inc. v. Ambrosia ChocolateCo., 816 F.2d 381, 385, 387 (8th Cir. 1987).

2. The Third Circuit�’s decision holding that aplaintiff in a secondary-line or tertiary-line casecannot prove competitive injury under the RPA ina non-retail industry �“where the competition forsales to prospective customers occurs before thesale of the product for which the RPA violation isalleged�” creates a conflict with all of those deci-sions. App., infra, 156a-157a (emphasis in origi-nal).

Other courts of appeals have considered thequestion of whether the RPA�’s competitive injuryelement requires a plaintiff to prove that the pur-chases of the discriminatorily priced goodsoccurred at the same time as the competition.Those courts have refused to import a timingrequirement into the RPA. The Eleventh Circuit�’sdecision in DeLong Equipment, supra�—a true bidmarket case�—is particularly instructive. In thatcase, the Eleventh Circuit expressly rejected, as amatter of law, the type of timing requirement thatthe Third Circuit imposed here. The price dis-crimination in DeLong was related to a largerscheme between a distributor (BCS) and a sup-plier of media used to polish jet engine parts(Washington Mills) involving a single customer,Pratt & Whitney. See 990 F.2d at 1191. Washing-ton Mills provided secret discounts to BCS thatwere not available to other distributors, such as

16


DeLong, that competed with BCS to supply mediato Pratt. Id. Because Washington Mills alsodelayed in providing DeLong with specificationsnecessary for its approval as a distributor byPratt, DeLong was unable to bid on an October1983 Pratt purchase order. BCS won the bid, andsupplied Pratt�’s anticipated media requirementsfor the following year. See id. at 1192-93, 1202.Thereafter, DeLong was hindered in its ability tocompete with BCS because it could not matchBCS�’s prices due to Washington Mills�’ price dis-crimination. Nonetheless, in 1984, DeLong suc-ceeded in selling to Pratt small quantities ofWashington Mills-manufactured media. See id. at1192-93.

Washington Mills argued that DeLong and BCSwere not �“competing purchasers�” and were not inactual competition for Pratt�’s business becauseBCS had won the October 1983 bid and filled thatorder with media that it purchased from Wash-ington Mills before DeLong either (i) was qualifiedto bid for sales to Pratt or (ii) actually purchasedany media from Washington Mills. See id. at 1201-02. The Eleventh Circuit rejected that argument,concluding that �“there is no doubt but that bothBCS and DeLong were after the same Pratt dollar�”and were in �“head-to-head competition�” for Pratt�’sbusiness. Id. The Eleventh Circuit further heldthat �“[w]hile there must be two sales made by thesame seller to at least two different purchasers attwo different prices, there is no requirement thatthe two sales be made at precisely the same time orplace.�” Id. (emphasis added) (citations omitted).

The facts of DeLong are materially indistin-guishable from the facts here. Just as the disfa-

17


vored purchaser in DeLong sold some of the prod-ucts at issue to Pratt after the favored purchaserwon the bid, Feesers also sold some Michael Foodsproducts to institutions after they had chosenSodexho for food management and purchasing.See, e.g., App., infra, 84a (�“[S]ome Sodexho cus-tomers have chosen to utilize Feesers for [certainMichael Foods] purchases.�”); id. at 91a (A Feesersself-op customer �“solicited proposals in an RFPprocess, and ultimately chose Sodexho,�” but con-tinued to purchase certain Michael Foods productsfrom Feesers.); see also id. at 60a-61a (findingthat Sodexho�’s documents �“demonstrate directcompetition with distributors�” to sell food evenafter a customer has awarded Sodexho a contractfor food management services).

Other courts of appeals also have held thatthere is no �“timing of the competition�” require-ment for proving competitive injury. For example,the Fifth Circuit rejected a timing requirement inHartley & Parker, Inc. v. Fl. Beverage Corp., 307F.2d 916 (5th Cir. 1962). In Hartley, the disfa-vored purchaser ceased purchasing discriminato-rily priced products from the supplier, butcontinued to sell such products out of inventory.Id. at 921. The Court held that �“[t]he purpose ofthe Act would be defeated . . . if it were given sostrict a construction as to require two actual pur-chases at precisely the same time.�” Id. And othercircuits have routinely upheld findings of com-petitive injury under the Morton Salt test withoutimposing a timing requirement. See ChromaLighting v. GTE Prods. Corp., 111 F.3d 653, 654(9th Cir. 1997); Coastal Fuels v. CaribbeanPetroleum Corp., 79 F.3d 182, 191, 193 (1st Cir.1996); Hasbrouck v. Texaco, Inc., 842 F.2d 1034,

18

26922 • Dewey: Fessers • USSC (revised 12-1-08) • LJB 5/28/10 2:00; KW 6/1 11:40 am

1041 (9th Cir. 1987), aff�’d, Texaco, 496 U.S. 543;Rose Confections, Inc. v. Ambrosia Chocolate Co.,816 F.2d 381, 385, 387 (8th Cir. 1987). Those deci-sions, moreover, involved both retail and non-retail industries. Accordingly, the Third Circuit�’sholding that this Court�’s precedents �“prevent theapplication of the RPA�” to non-retail industries,see App., infra, 165a, conflicts with the holdings ofother circuits that have declined to limit the appli-cation of the RPA in non-retail markets.

For example, in Innomed Labs, LLC v. AlzaCorp., 368 F.3d 148 (2d Cir. 2004) (Sotomayor, J.),the Second Circuit held that non-retail distribu-tion contracts are �“unquestionably covered by theAct.�” Id. at 159. As then-Judge Sotomayorexplained: �“The Robinson-Patman Act was [ ]intended to apply to all commodities distributioncontracts, in all positions in the distributionchain.�” Id. at 161. Rejecting the argument thatexclusive distribution contracts should beexcluded from the RPA, Judge Sotomayor statedthat �“[s]uch an arbitrary result would contravenethe Act�’s purpose of creating even competitionthroughout the market, at all levels of the distri-bution chain.�” Id. (emphasis added). JudgeSotomayor further explained that the �“Robinson-Patman Act was drafted amid fears that the riseof integrated chain stores would overpower thetraditional distribution model and its reliance onwholesale distributors and retailers to move prod-ucts from manufacturers to consumers.�” Id. at 160(emphasis added).

Similarly, in Rose Confections, Inc. v. AmbrosiaChocolate Co., 816 F.2d 381 (8th Cir. 1987), where

19


the plaintiff was a repackager that sold to retail-ers, the Eighth Circuit upheld an inference of com-petitive injury under the Morton Salt test withoutimposing an additional timing requirement in anon-retail context. The Eighth Circuit held that�“Morton Salt requires a plaintiff to show only asubstantial price difference�” among competitors inorder to prove competitive injury. 816 F.2d at388.1

In sum, the Third Circuit has created a clearconflict among the courts of appeals on how toapply the competitive injury requirement of theRobinson-Patman Act. If the Third Circuit�’s tim-ing requirement is allowed to stand, it will createmassive confusion among litigants and the lowercourts trying to interpret an already �“complicatedarea of law�” that has, by the Third Circuit�’s ownadmission, �“flummoxed the federal courts�” on mul-tiple occasions. App., infra, 184a n.17. It alsothreatens to cause substantial confusion fornational companies trying to price their productsacross different jurisdictions. The resulting circuitconflict, on an important issue concerning theinterpretation of the Robinson-Patman Act, mer-its review by this Court.

20


1 See also Hasbrouck, 842 F.2d at 1040 (�“To hold thatprice discrimination between a wholesaler and a retailercould never violate the Robinson-Patman Act would leaveimmune from antitrust scrutiny a discriminatory pricingprocedure that can effectively serve to harm competition�”and would be �“contrary to the objectives of the Robinson-Pat-man Act.�”); Mid-South Distribs. v. FTC, 287 F.2d 512 (5thCir. 1961) (upholding RPA claim where distributor buyinggroups were favored over independent distributors); MoogIndus., Inc. v. FTC, 238 F.2d 43 (8th Cir. 1956) (same).

B. The Third Circuit�’s Novel TimingRequirement Is Contrary To SixtyYears Of This Court�’s PrecedentsConcerning The Requirement OfCompetitive Injury In A Price Dis-crimination Case

1. In its seminal Morton Salt decision, thisCourt rejected an argument that an FTC cease-and-desist order should be set aside because theFTC failed to show that the discriminatory dis-counts had in fact caused injury to competition.Morton Salt, 334 U.S. at 45-46. As the Courtobserved, the RPA�’s legislative history �“makes itabundantly clear that Congress considered it to bean evil that a large buyer could secure a compet-itive advantage over a small buyer solely becauseof the large buyer�’s quantity purchasing ability.�”Id. at 43. Thus, the Court explained, a plaintiffneed prove only substantial price discriminationamong competing purchasers over time in order tosatisfy the competitive injury requirement:

It would greatly handicap effectiveenforcement of the Act to require testi-mony to show that which we believe to beself-evident, namely, that there is a �‘rea-sonable possibility�’ that competition maybe adversely affected by a practice underwhich manufacturers and producers selltheir goods to some customers substan-tially cheaper than they sell like goods tothe competitors of these customers. Thisshowing in itself is sufficient to justify ourconclusion that the Commission�’s findingsof injury to competition were adequatelysupported by evidence.

Id. at 50-51.

21


There was no suggestion in Morton Salt of atiming exception to the general test for provingcompetitive injury, whereby a favored non-retailpurchaser could evade liability under the RPA bydelaying its purchases of goods to be resold to aparticular customer until after it had won thecompetition for that customer�’s business. On thecontrary, the Court stated that �“the language ofthe Act, and the legislative history�” make it clearthat a plaintiff �“need only prove that a seller hadcharged one purchaser a higher price for likegoods than he had charged one or more of the pur-chaser�’s competitors.�” Id. at 45 (emphasis added).

This Court has applied the Morton Salt com-petitive-injury test for more than sixty years. See,e.g., Volvo Trucks, 546 U.S. at 177. Similarly, asexplained above, all of the courts of appeals pre-viously followed Morton Salt without imposingany requirement that a non-retail plaintiff provethat its competitor purchased the products atissue at favorable, discriminatory prices at exactlythe same time that the competition for the resaleof those products occurred.

In its radical departure from this settled law,the Third Circuit reasoned that this Court�’s prece-dents required the �“narrow[ ] constru[ction] [of]the RPA�” in non-retail markets where the�“�‘allegedly favored purchasers [bear] little resem-blance to [the] large independent departmentstores or chain operations�’ that the RPA wasintended to target.�” App., infra, 165a-166a (citingVolvo Trucks, 546 U.S. at 181); 181a. But thisCourt has rejected precisely the distinction thatthe Third Circuit drew between retail and non-retail markets for proving competitive injury

22


under the RPA. Indeed, Morton Salt itselfinvolved both retail and non-retail segments of thefood distribution industry. Although the case prin-cipally involved price discrimination betweenfavored chain stores and disfavored retailers, thedefendant Morton Salt Co. also provided discrim-inatory prices to a wholesaler that passed on theprice discriminations. 334 U.S. at 40-41, 42 n.5.The Court expressly rejected the proposition thatcompetitive injury arising from the discounts pro-vided to the favored wholesaler had not beenproven, stating that such discrimination �“need notbe separately treated�” from the discrimination infavor of the retail chain stores. Id. at 42 n.5.

Over the last six decades, this Court has reaf-firmed the viability of the Morton Salt test forproving competitive injury, as well as its appli-cation to non-retail markets. See, e.g., Texaco,Inc., 496 U.S. at 559 (applying the Morton Salttest to gasoline wholesalers); Falls City Indus.,Inc. v. Vanco Beverage, Inc., 460 U.S. 428, 435-36(1983) (applying Morton Salt to beer distributors);see also Volvo Trucks, 546 U.S. at 179-81 (declin-ing to adopt a broad rule that the RPA �“does notreach markets characterized by competitive bid-ding and special-order sales�”). This Court hasnever held that there is a �“timing�” exception toMorton Salt, to be invoked when the favored pur-chaser in a non-retail market does not purchasethe discriminatorily priced products at the sametime at which the competition to resell the prod-ucts occurs.

2. Moreover, there is no way to reconcile theThird Circuit�’s timing requirement with the deci-sions of this Court that have confirmed the appli-

23


cability of the RPA to situations in which the com-petitive injury occurs at the tertiary level or evenlower down the distribution chain. See Texaco, 496U.S. at 554-57; Falls City, 460 U.S at 436; Perkinsv. Standard Oil Co., 395 U.S. 642, 647 (1969). Incases such as these, in which the competitiveinjury is remote from the point of discrimination,neither this Court nor the lower federal courtshave ever applied a requirement that the timing ofthe price discrimination be simultaneous with thecompetitive injury that it causes.

Indeed, this Court has warned against narrow-ing the reach of the RPA beyond that intended byCongress, through �“artificial�” readings of thestatutory language and �“strictly literal�” applica-tions of the �“competing purchasers�” requirement.For example, in Falls City, the Court rejected theargument that the RPA does not permit a findingof competitive injury where the favored and dis-favored purchasers did not compete for the samecustomers, but their customers competed with oneanother. 460 U.S at 436. Although the Court notedthat, �“[i]n a strictly literal sense,�” the favored anddisfavored distributors were not �“competing pur-chasers�” as in Morton Salt, it unanimously heldthat �“the competitive injury component of aRobinson-Patman Act violation is not limited toinjury to competition between the favored and dis-favored purchaser; it also encompasses the injuryto competition between their customers.�” Id.; seealso Texaco, 496 U.S. at 554-57 (holding that pricediscrimination against a disfavored buyer, whichpurchased directly from the supplier and com-peted with later customers of the favored buyer,was a violation of Section 2(a)); Perkins, 395 U.S.at 647 (rejecting Ninth Circuit�’s reading of Section

24


2(a) that held that there could not be competitiveinjury at the �“fourth line�” as �“wholly . . . artificial�”and �“unwarranted by the language or purposes ofthe Act�”).

The Third Circuit�’s novel timing requirementrests on the proposition that Sodexho, as thefavored purchaser, does not purchase MichaelFoods products for resale to a particular customeruntil after it has won the competition for that cus-tomer�’s business, and that Feesers and Sodexho,therefore, are not �“competing purchasers.�” App.,infra, 178a-179a. That reasoning, however, simplyignores the role of Sysco, Sodexho�’s contract dis-tributor and�—as the Third Circuit itself found inits earlier opinion�—the �“second purchaser�” forRPA purposes. App., infra, 10a. Regardless of thetiming of Sodexho�’s purchases of Michael Foodsproducts from Sysco for a particular customer, itis indisputable that Feesers and Sysco continu-ously purchase the same Michael Foods productson an ongoing basis and �“keep [them] in stock forresale to customers.�” App., infra, 146a.

Ironically, as the Third Circuit recognized in itsearlier opinion, �“[t]his case is most analogous toTexaco v. Hasbrouck.�” See App., infra, 10a-11an.5. As the court stated:

This is a difficult case to categorizebecause the discrimination allegedlyimpacts competition between the disfa-vored purchaser (Feesers) and the cus-tomer of the favored purchaser (Sodexho).. . . Regardless, categorizing the discrimi-nation in this case as second- or third-lineis not essential, as long as there is a pro-hibited effect on competition.

25


Id. n.5 (citations omitted). By ignoring the con-temporaneous purchases by Feesers and Syscoand focusing instead on the timing of purchasesfor specific customers by Sodexho (Sysco�’s �“cus-tomer�” under Section 2(a)), the Third Circuit hascreated a timing rule that is flatly incompatiblewith Texaco.

3. In imposing its new timing requirement, theThird Circuit stated that it was following thisCourt�’s direction in Volvo Trucks. That interpre-tation of Volvo Trucks is misguided. In VolvoTrucks, the Court overturned an RPA verdict forthe plaintiff on the ground that plaintiff�’s evi-dence, which �“paired occasions on which it com-peted with non-Volvo dealers for a sale toCustomer A with instances in which other Volvodealers competed with non-Volvo dealers for a saleto Customer B,�” failed to show that the plaintiff�“compete[d] with beneficiaries of the alleged dis-crimination for the same customer.�” 546 U.S. at178 (emphases in original). As a result, the Courtheld that the plaintiff had failed to provide suffi-cient evidence of competitive injury in a cus-tomized bid market in which only one competitorwould make purchases as the winning bidder. Id.at 178-80.

Nothing in Volvo Trucks supports the timingrequirement created in Feesers II.2 Volvo Trucks

26


2 Feesers respectfully suggests that the Court maybenefit from soliciting views from the United States, andparticularly from the Federal Trade Commission, the expertantitrust agency charged with enforcing the RPA, on theimportance of the question presented in this petition to theenforcement of the Act. The Court solicited the views of theUnited States in Volvo Trucks and may benefit from doing sohere as well.

reiterated Morton Salt�’s holding that �“a permis-sible inference of competitive injury may arisefrom evidence that a favored competitor receiveda significant price reduction over a substantialperiod of time.�” Volvo Trucks, 546 U.S. at 177; seealso id. at 179 (citing with approval Falls City,which applied the Morton Salt rule). Indeed, evenif the wholesale food distribution industry were a�“bid market�” industry, the Court in Volvo Trucksspecifically declined to adopt a broad rule that theRPA �“does not reach markets characterized bycompetitive bidding and special-order sales.�” 546U.S. at 180-81.

Further, in holding that the plaintiff in VolvoTrucks had not proven a Section 2(a) violation, theSupreme Court distinguished the bid market forcustomized trucks in that case from the situationmore commonly encountered in RPA cases, wherecompeting purchasers resell commodity products.Volvo Trucks, 546 U.S. at 178-79. As the Courtstated: �“The [RPA] centrally addresses price dis-crimination in cases involving competitionbetween different purchasers for resale of the pur-chased product. Competition of that characterordinarily is not involved when a product subjectto special order is sold through a customer-specificcompetitive bidding process.�” 546 U.S. at 169-70(emphasis added). That observation has no appli-cation to the Feesers case because the wholesalefood distribution industry fits the traditional RPAparadigm. As the trial court found: �“the goods inquestion are perishable commodities that twocompetitors regularly purchase and keep in stockfor resale to customers.�” App., infra, 146a.

In fact, if the institutional food business couldbe characterized as a �“bid market,�” then virtually

27


every non-retail industry could be so character-ized. As the district court found (in a factual find-ing that the Third Circuit did not disturb):

[C]ompetition for the sale of MichaelFoods egg and potato products to institu-tional food service customers was ongoingand not limited to the formal RFP [bid]process. Food service management com-panies, distributors, and GPOs all com-pete formally and informally for the saleof food to institutions. Even when a com-pany initiates the RFP bidding process,which includes only food service manage-ment companies, that choice is not final orlimited to the companies submitting a bid.

App., infra, 66a; see also id. at 59a (competitionbetween Sodexho and Feesers �“occurs not just inthe formal request for proposal (�‘RFP�’) process,but also on an informal basis all the time�”).3

28


3 The Volvo Trucks Court noted a need for caution inapplying the RPA where �“there is no evidence that anyfavored purchaser possesses market power.�” 546 U.S. at 181.Here, however, the evidence established that Sodexho wasthe world�’s largest purchaser of food and was able to use itsvast power as the �“big dog�” in food purchasing to obtain�“stunning�” discounts not available to any other competitor.App., infra, 128a-129a; 73a. Indeed, Sodexho�’s own docu-ments and trial testimony left no doubt that it was Sodexho�’sspecific strategy to use its massive purchasing power toextract uniquely favorable price discriminations from sup-pliers of food to gain a competitive advantage over distrib-utors like Feesers. Id. at 98a-112a; see also id. at 85a(Sodexho obtained the �“lowest�” prices and large bonuses anddiscounts unavailable to others). This is the classic situationin which the RPA was intended to apply. See Morton Salt,334 U.S. at 43.

C. The Question Presented Is An Excep-tionally Important One That War-rants This Court�’s Review

Finally, in holding that �“the RPA was not meantto cover the type of competition present in theinstant case,�” App., infra, 185a, the Third Circuiteffectively created a judicial exemption from theRPA that is contrary to this Court�’s long line ofauthority holding that new antitrust exemptionsshould not be created by the courts.

Specifically, the Third Circuit concluded thatthe RPA should not apply because (1) Feesers andSodexho are not �“competing retail stores,�” and (2)Sodexho �“operates in a bid market.�” App., infra,181a. Not only is this holding contrary to theunderlying purpose of the RPA, which was passedin part to protect �“wholesale distributors,�” seesupra at 19, but it also creates an impermissiblejudicial exemption from the RPA. This Court hasconsistently held that the courts should not createantitrust exemptions. See Falls City, 460 U.S. at436 (�“The determination whether to alter thescope of the Act must be made by Congress, notthis Court.�”); Jefferson County Pharm. Ass�’n v.Abbott Labs., 460 U.S. 150, 169-71 (1983) (refus-ing to find an exemption from the RPA because�“no unambiguous evidence of congressional intent�”to create that exemption existed); Flood v. Kuhn,407 U.S. 258, 276-82 (1972) (stating that exemp-tions from the antitrust laws should be made �“bylegislation and not by court decision�”).

Furthermore, this Court has specifically ana-lyzed the question of what exemptions exist underthe RPA and held that the �“only express exemp-tion [under the RPA] is that for nonprofit insti-

29


tutions contained in 15 U.S.C. § 13c.�” JeffersonCounty, 460 U.S. at 154-55. In Jefferson County,the Court specifically declined to create a newexemption to the RPA, observing that �“�‘our caseshave repeatedly established that there is a heavypresumption against implicit exemptions�’ from theantitrust laws.�” Id. at 154-55, 158 (quoting Gold-farb v. Virginia State Bar, 421 U.S. 773, 787(1975)). As the Court has repeatedly instructed,�“�‘the antitrust laws, and [the RPA] in particular,are to be construed liberally, and [ ] the exceptionsfrom their application are to be construedstrictly.�’�” Id. at 159 (quoting and citing SupremeCourt cases).

The exception that the court of appeals createdhere is a broad one that undermines the objectivesof the RPA. The Third Circuit�’s new timing testmakes it impossible for plaintiffs to prosecuteprice discrimination claims in non-retail indus-tries like food distribution. By manipulating thetiming of its purchases through an intermediarysuch as Sysco, a power buyer like Sodexho couldevade liability under the RPA simply by (1) usingits purchasing power to obtain lower prices fromsuppliers of commodities, (2) using the competitiveadvantage arising from those lower prices to wincustomers, and (3) routing the products though anintermediary that would hold them in inventory,but (4) delaying the purchase of the goods fromthe intermediary for a particular customer untilafter it has already won that customer�’s businesswith promises of lower, discriminatory prices.That is exactly what happened in this case. SeeApp., infra, 44a; 63a. The Third Circuit�’s new tim-ing test, by �“elevat[ing] form over substance,�” id.at 168a, would permit a purchaser �“to defeat the

30


statute�’s clear objectives by transforming unlaw-ful, into lawful, price discrimination.�” See CaribeBMW, Inc. v. Bayerische Motoren Werke Aktienge-sellschaft, 19 F.3d 745, 750 (1st Cir. 1994)(Breyer, C.J.). This Court, however, has repeat-edly instructed that the RPA should not be con-strued in a way that �“would allow pricediscriminators to avoid the sanctions of the Act bythe simple expedient of adding an additional linkin the supply chain.�” Perkins, 395 U.S. at 647;accord Texaco, 496 U.S. at 567 n.26.

Moreover, this Court has specifically rejectedthe court of appeals�’ reasoning that its grafting ofnew limitations onto the RPA is justified becausesome commentators have determined that theRPA is antithetical to competition policy. CompareJefferson County, 460 U.S. at 170, with App.,infra, 166a, 167a n.12, and Toledo Mack, 530 F.3dat 228 n.17. Judicial policy disagreements thathave arisen subsequent to passage of the Act can-not justify reading new limitations into the RPA.Jefferson County, 460 U.S. at 163-69. As the Courtexplained:

The Robinson-Patman Act has beenwidely criticized, both for its effects andfor the policies that it seeks to promote.Although Congress is well aware of thesecriticisms, the Act has remained in effectfor almost half a century. And it certainlyis �“not for [this Court] to indulge in thebusiness of policy-making in the field ofantitrust legislation. . . . Our function endswith the endeavor to ascertain from thewords used, construed in the light of therelevant material, what was in fact theintent of Congress.�”

31


Id. at 170 (quoting United States v. Cooper Corp.,312 U.S. 600, 606 (1941)).

The determination of �“whether an exemptionshould be granted�” �“is for Congress to resolve, notthis Court.�” Flood, 407 U.S. at 277; U.S. v. Int�’lBoxing Club, 348 U.S. 236, 243 (1955). �“There isno reason, in the absence of an explicit exemption,to think that Congressmen . . . intended to denysmall businesses . . . protection from the compe-tition of the strongest competitor of them all.�” Jef-ferson County, 460 U.S. at 170-71. Accordingly,�“[t]o create an exemption here clearly would becontrary to the intent of Congress.�” Id. at 171.Certiorari should be granted to prevent the ThirdCircuit�’s policy disagreement with the RPA fromrendering the Act inoperative in non-retail indus-tries�—including the very industry that motivatedCongress to enact the RPA in the first place.

32


CONCLUSION

The petition for a writ of certiorari should begranted.

Respectfully submitted,

JEFFREY L. KESSLERCounsel of Record

JOHN F. COLLINSEAMON O�’KELLYSUSANNAH P. TORPEYDEWEY & LEBOEUF LLP1301 Avenue of the AmericasNew York, New York 10019(212) [email protected] K. SHANMUGAMWILLIAMS & CONNOLLY LLP725 Twelfth Street, N.W.Washington, D.C. 20005(202) 434-5000Counsel for Petitioner

June 2, 2010

33


APPENDIX

APPENDIX APRECEDENTIAL

UNITED STATES COURT OF APPEALSFOR THE THIRD CIRCUIT

__________No. 06-2661

__________FEESERS, INC.,

Appellantv.

MICHAEL FOODS, INC.;SODEXHO, INC.__________

Appeal from the United States District Courtfor the Middle District of Pennsylvania

(D.C. Civil No. 04-cv-00576)District Judge: Honorable Sylvia H. Rambo

__________Argued May 10, 2007

Before: RENDELL, JORDAN andALDISERT, Circuit Judges.

(Filed: August 14, 2007)__________

1a

26922 • Dewey • SUPREME COURT APPENDIX part: 1 LJB 5-13-10; crs LJB 5/20

Jeffrey L. Kessler [ARGUED]Johns F. CollinsEamon O�’KellyDewey Ballantine1301 Avenue of the AmericasNew York, NY 10019

Steven M. WilliamsCohen, Seglias, Pallas, Greenhall & Furman240 North Third Street, 8th FloorHarrisburg, PA 17101

Counsel for Appellant

Brett L. MessingerDuane Morris30 South 17 StreetUnited PlazaPhiladelphia, PA 19103-4196

Margaret M. ZwislerLatham & Watkins555 11th Street, N.W., Suite 1000Washington, DC 20004

Counsel for Appellee Michael Foods Inc.

Matthew M. HaarSaul EwingTwo North Second StreetPenn National Insurance Tower7th FloorHarrisburg, PA 17101

2a


Christopher M. SheehanMartin F. Gaynor, III [ARGUED]Cooley, Manion & Jones21 Custom House StreetBoston, MA 02110

Counsel for Appellee Sodexho Inc.

__________

OPINION OF THE COURT

__________

RENDELL, Circuit Judge.

I.

Plaintiff Feesers, Inc. appeals the DistrictCourt�’s grant of summary judgment in favor ofdefendants Michael Foods and Sodexho in thisantitrust action. In its complaint, Feesers allegedthat Michael Foods violated section 2(a) of theRobinson-Patman Act, 15 U.S.C. § 13(a), by sellingits potato and egg products at lower, and thus dis-criminatory, prices to Sodexho. It further allegedthat Sodexho violated section 2(f) of the Act, 15U.S.C. § 13(f), by knowingly inducing the dis-criminatory pricing.

The District Court found that Feesers failed toprove the fourth element of its prima facie caseunder section 2(a), namely that the alleged dis-crimination had a prohibited effect on competi-tion, because Feesers failed to show that it was in�“actual competition�” with Sodexho. See Feesers,

3a


Inc. v. Michael Foods, Inc. & Sodexho, Inc., No.04-Civ-576, 2006 WL 1274088, slip op. at 23 (M.D.Pa. May 4, 2006). We will reverse because the Dis-trict Court used the wrong standard in makingthis determination and we conclude that Feesershas proffered sufficient evidence of competitionbetween itself and Sodexho for sales of food prod-ucts to food service facilities to allow a reasonablefactfinder to conclude that these companies are in�“actual competition.�” Moreover, the District Courterroneously put the burden on Feesers to provenot only �“actual competition,�” but also thatMichael Foods�’ discriminatory pricing causedFeesers to lose sales to Sodexho, rather than plac-ing the burden on Michael Foods to rebut theinference of injury to competition that arises from proof of a substantial price discriminationbetween competing purchasers over time.

II.

Most of the underlying facts are undisputed.Where there is a dispute, we view the facts in thelight most favorable to Feesers. Andreoli v. Gates,482 F.3d 641, 644 (3d Cir. 2007).

The customers of Sodexho and Feesers are foodservice facilities that sell meals, snacks, and bev-erages, such as school, hospital, and nursing homecafeterias. Both Sodexho and Feesers sell foodproducts to food service facilities in the States ofPennsylvania, New Jersey, Maryland, Delaware,and Virginia. Feesers is a full-line distributor offood and food-related products (�“products�”) thatdistributes these products to institutional cus-tomers. Sodexho is a food service management

4a


company that provides facility management andoperation services to its clients and, in most cases,also sells products to the facilities. Sodexho doesnot warehouse and deliver products directly to itsclients, but rather contracts with its clients to pro-cure products for them and then subcontracts withdistributors who distribute the products to thefacilities. Both Feesers and Sodexho contract withfood service facilities to provide them with prod-ucts from Michael Foods. Michael Foods is a sup-plier of egg and potato products.

A food service facility will contract with eitherSodexho or Feesers, but not both,1 to buy food andfood-related products. A food service facility mayeither contract with Sodexho for Sodexho to oper-ate the facility and procure products,2 or contract

5a


1 On a few occasions, both Sodexho and Feesers haveserved the same facility at the same time, but the facilitycontracted directly only with the food service managementcompany, which in turn contracted with Feesers. Feeserswas, at one time, the prime distributor for a food servicemanagement company called The Wood Company. Wood con-tracted for Feesers to be its �“primary non-exclusive dis-tributor.�” However, Sodexho purchased Wood part-waythrough the term of the Feesers-Wood primary distributorcontract. The Feesers contract for the facilities previouslyserviced by Wood expired at the end of 2002 and was notrenewed by Sodexho. Instead, Sodexho chose Sysco as itsprime distributor for the region. App. 8127.

2 When Sodexho takes on a facility as a client, Sodexhousually contracts with the facility both to procure food onbehalf of the facility and to operate the facility. However, ascounsel for Sodexho acknowledged at oral argument, thereare a limited number of Sodexho-operated facilities forwhich Sodexho does not provide procurement services. Forsome healthcare facilities, Sodexho provides food manage-ment services, but the facility will handle its own food pro-curement. App. 1175, 1415 1255.

with Feesers for Feesers to procure products andthe facility will self-operate or hire a third-partyoperator. To procure products for a facility,Feesers purchases products directly from MichaelFoods and then resells the products to food servicefacilities.

Sodexho�’s process to procure products fromMichael Foods for resale to food service facilitiesis a bit more complicated. Sodexho itself does notpurchase products from Michael Foods, butemploys a distributor, such as Sysco Corporation.3

Although product suppliers like Michael Foodsgenerate price lists that set forth the prices atwhich they sell food to distributors, Sodexho hasnegotiated lower deviated pricing with MichaelFoods. The transaction proceeds as follows:Michael Foods sells products to Sodexho�’s desig-nated distributor at list prices and the distributor,which is usually Sysco, then resells the productsto Sodexho and provides Michael Foods with proofof delivery of products to Sodexho; Sysco invoicesMichael Foods for the difference between the listprice and the Sodexho-negotiated deviated price;Sodexho then purchases these products from Syscopursuant to a �“prime distributor agreement,�”which specifies the price that Sodexho will paySysco for each product. Under the agreement,Sysco sells the Michael Foods products to Sodexhofor the Sodexho-negotiated price plus an agreed-upon markup. App. 9706. Sysco�’s resale price of

6a


3 Sysco is the designated �“prime distributor�” forSodexho in 48 states. App. 2535. Sodexho usually deter-mines which company will distribute the products to itsfacility clients, although in rare instances the facility maychoose the distributor. App. 7920.

Michael Foods�’ products to Sodexho reflects thelower prices in the deviated pricing agreementbetween Sodexho and Michael Foods. See Feesers,Inc., slip op. at 5.

After Sodexho purchases the Michael Foodsproducts from Sysco at the agreed-upon prices, itresells the products to a food service facility cus-tomer and charges the cost of the products to thecustomer as an �“operating expense.�” The food ser-vice facility generally does not interact directlywith Sysco or any other Sodexho-designated dis-tributor. Instead, the facility pays Sodexho for theinvoiced cost of the food-plus, in most cases, a�“procurement expense�” of 0.9% of the invoicedamounts-as part of the facility�’s reimbursement ofSodexho for �“operating expenses.�” Thus, becauseMichael Foods charges Sysco less for productsresold to Sodexho than it charges Feesers for thesame products, Sodexho�’s customers pay less thanFeesers�’ customers for these products.

Feesers�’ customers are, in general, self-operatedfacilities, while none of Sodexho�’s customers areself-operated.4 However, food service facilities mayswitch from being self-operated to being operatedby a management company like Sodexho. When aself-operated facility that previously bought prod-

7a


4 Entegra, a group purchasing organization (�“GPO�”)affiliated with Sodexho, does serve self-operated facilities.Entegra provides its clients with access to a portfolio of con-tracts negotiated by Sodexho with suppliers of food and food-related products. A facility employing Entegra�’s servicesmay use a Sodexho-contracted distributor or its own con-tracted distributor to distribute foods that the facility pur-chases pursuant to Entegra-negotiated price lists. Entegra,however, is a separate legal entity from Sodexho and is nota party to this action. App. 9100.

ucts from Feesers is converted to a Sodexho-oper-ated facility, Sodexho operates the facility andgenerally also procures the new client�’s food prod-ucts, thereby displacing Feesers. For example, theJewish Home of Greater Harrisburg was self-oper-ated and bought its products from Feesers. It thenbecame a Sodexho-managed facility and stoppedbuying products from Feesers. St. Mary�’s CatholicSchool was also a Feesers customer and self-oper-ated facility, which then switched to being oper-ated by Sodexho and no longer buys products fromFeesers. Sodexho will approach self-operated(�“self-op�”) facilities to convert them to Sodexho-operated facilities. App. 1425 (Deposition ofChristophe Rochette of Sodexho) (�“[Y]ou asked merepeatedly, are we interested in converting self-op? That is what we are. So, I mean, I think thatwe should [be] clear that for the record, that yes,we convert self-op. That is what we do.�”). Sodexhohas solicited at least five facilities served byFeesers to become Sodexho customers. Sodexhocustomers end up paying less for products fromMichael Foods than they would pay if they wereself-operated and purchased the same productsfrom Feesers.

On the other hand, facilities also switch frombeing operated by Sodexho to being self-operated.In these cases, Sodexho will no longer procure foodfor the facility and the facility will seek outanother company, such as Feesers, from which tobuy its food products. The Meadows NursingHome was a Sodexho customer and switched tobeing a self-operated facility and a Feesers cus-tomer, in part because Michael Foods agreed togive Feesers the same product pricing given to

8a


Sodexho. In 1998, Sodexho lost nine accounts toself-operation. App. 1426. In 1999, eight Sodexhocustomers switched to being self-operated. App.1427.

Feesers sued Michael Foods and Sodexho in theUnited States District Court for the Middle Dis-trict of Pennsylvania, alleging that Michael Foodsviolated section 2(a) of the Robinson-Patman Act,15 U.S.C. § 13(a), by selling products at discrimi-natory prices to Sodexho and that Sodexho vio-lated section 2(f) of the Act, 15 U.S.C. § 13(f), byknowingly inducing the discriminatory pricing.Defendants moved to dismiss the complaint on thegrounds that Feesers had not adequately pled thatit was in actual competition with Sodexho. TheDistrict Court denied the motion and allowed theparties to proceed to discovery. After discovery,the parties all moved for summary judgment.

The District Court found that Feesers hadestablished three out of the four elements of itssection 2(a) claim against Michael Foods: thatsales were made to two different purchasers ininterstate commerce; that the product sold was ofthe same grade and quality; and that defendantdiscriminated in price as between the two pur-chasers. Feesers, Inc. v. Michael Foods, Inc. &Sodexho, Inc., No. 04-Civ-576, 2006 WL 1274088,slip op. at 10-18 (M.D. Pa. May 4, 2006). First, theCourt noted that there was no dispute that thegoods purchased from Michael Foods were of thesame grade and quality. Feesers, slip op. at 10.The Court also found that �“because the facts thatestablish that Michael Foods sold products at dif-ferent prices are not in dispute . . . price discrim-ination exists within the context of the Act.�” Id. at

9a


11. Finally, as to the requirement that there betwo purchasers in interstate commerce, the Courtconcluded that the facts show that Michael Foodssold to two purchasers, Feesers and Sysco. TheCourt concluded that this is a case of �“third-line�”discrimination, i.e., when a seller�’s price dis-crimination harms competition between customersof the favored and disfavored purchasers. Id. at 12n. 8. The Court did not reach the issue of whetherSodexho is a direct �“economic�” purchaser fromMichael Foods, which would, presumably, makethis a second-line discrimination case (i.e., dis-crimination that harms competition between twopurchasers). Defendants do not challenge thesefindings on appeal.5

10a


5 The parties do not debate whether this is a second-or third-line discrimination case, but we note that the Dis-trict Court�’s conclusion that this is a case of third-line dis-crimination appears to be incorrect. This is not clearly eithera second-line or third-line case, but falls somewhere inbetween these categories. See George Haug Co. v. RollsRoyce Motor Cars, Inc., 148 F.3d 136, 141 n. 2 (2d Cir. 1998)(noting that �“secondary-line price discrimination[ ] occurswhen a seller�’s discrimination impacts competition amongthe seller�’s customers; i.e. the favored purchasers and dis-favored purchasers . . . tertiary-line [discrimination] occurswhen the seller�’s price discrimination harms competitionbetween customers of the favored and disfavored purchasers,even though the favored and disfavored purchasers do notcompete directly against another�”). This is a difficult case tocategorize because the discrimination allegedly impacts com-petition between the disfavored purchaser (Feesers) and thecustomer of the favored purchaser (Sodexho).

This case is most analogous to Texaco v. Hasbrouck, 496U.S. 543, 110 S.Ct. 2535, 110 L.Ed.2d 492 (1990), in whichseveral gas retail stations brought suit against Texaco forselling gas to two distributors at discounted prices. The dis-

However, the District Court found that Feesersfailed to proffer sufficient evidence to prove thefourth element of its prima facie case: that thediscrimination had a prohibited effect on compe-tition. Id. at 24. The District Court found thatFeesers did not meet its burden to show that itwas in �“actual competition�” with Sodexho as of thetime of the price differential. Id. at 23. The Dis-trict Court noted that Feesers had failed to provethat it competes with Sodexho �“at the same func-tional level.�” Id. at 21. The Court also found thatFeesers failed to proffer evidence that it lost cus-tomers to Sodexho because of food prices, ratherthan for other reasons relating to the managementservices Sodexho provides. Without this evidence,the Court found that Feesers could not prove thatit competes with Sodexho. Accordingly, becauseFeesers failed to establish a prima facie caseunder section 2(a) of the Act, the Court grantedsummary judgment in favor of defendants and

11a


tributors then resold gas to retail stations that competeddirectly with plaintiffs and they also operated their ownretail stations that competed directly with plaintiffs. Id. at549-51, 110 S.Ct. 2535. The Supreme Court did not catego-rize the case as a second- or third-line case, but insteadobserved that �“[t]he additional link in the distribution chaindoes not insulate Texaco from liability if Texaco�’s excessivediscount otherwise violated the Act.�” Id. at 567, 110 S.Ct.2535; see also Perkins v. Standard Oil Co., 395 U.S. 642, 89S.Ct. 1871, 23 L.Ed.2d 599 (1969) (finding actionable pricediscrimination resulting in injury to competition betweendisfavored purchaser and customer of customer of favoredpurchaser, but not categorizing the case as third-line orfourth-line discrimination). Regardless, categorizing the dis-crimination at issue in this case as second- or third-line isnot essential, so long as there is a prohibited effect on com-petition.

denied Feesers�’ motion for summary judgment.6

Id. at 24. Feesers now appeals.

III.

We exercise plenary review over the DistrictCourt�’s grant of summary judgment in favor ofdefendants, and we apply the same standard thatthe District Court should have applied. Andreoli,482 F.3d at 647. Summary judgment is appropri-ate when �“the pleadings, depositions, answers tointerrogatories, and admissions on file, togetherwith the affidavits, if any, show that there is nogenuine issue as to any material fact and that themoving party is entitled to a judgment as a matterof law.�” Fed.R.Civ.P. 56(c). We �“must view thefacts in the light most favorable to the nonmovingparty and draw all inferences in that party�’sfavor.�” Farrell v. Planters Lifesavers Co., 206 F.3d271, 278 (3d Cir. 2000).

Section 2(a) of the Robinson-Patman Act, 15U.S.C. § 13(a), provides in relevant part that:

It shall be unlawful for any personengaged in commerce, in the course ofsuch commerce, either directly or indi-rectly, to discriminate in price betweendifferent purchasers of commodities of likegrade and quality, where either or any ofthe purchases involved in such discrimi-nation are in commerce, where such com-

12a


6 The District Court noted that, given that Feeserswas unable to establish a section 2(a) claim, its section 2(f)claim against Sodexho �“necessarily fails.�” Feesers, slip op. at24.

modities are sold for use, consumption, orresale within the United States or anyTerritory thereof or the District ofColumbia or any insular possession orother place under the jurisdiction of theUnited States, and where the effect ofsuch discrimination may be substantiallyto lessen competition or tend to create amonopoly in any line of commerce, or toinjure, destroy, or prevent competitionwith any person who either grants orknowingly receives the benefit of such dis-crimination, or with customers of either ofthem.

As the District Court correctly stated, in orderto prove a violation of section 2(a) of the Robinson-Patman Act, a plaintiff must show (1) that saleswere made to two different purchasers in inter-state commerce; (2) that the product sold was ofthe same grade and quality; (3) that defendantdiscriminated in price as between the two pur-chasers; and (4) that the discrimination had a pro-hibited effect on competition. See Texaco Inc. v.Hasbrouck, 496 U.S. 543, 556, 110 S.Ct. 2535, 110L.Ed.2d 492 (1990). This appeal concerns thefourth element of Feeser�’s claim under section2(a): competitive injury. Specifically, we mustdecide whether the District Court applied the cor-rect legal standard to determine whether Sodexhoand Feesers are in actual competition andwhether it erred in holding that Feesers did notproffer sufficient evidence to allow a reasonablefactfinder to conclude that it is in actual compe-tition with Sodexho.

13a


To establish the fourth element of its primafacie case against Michael Foods, Feesers wasrequired to show that there is �“a reasonable pos-sibility that [the] price difference may harm com-petition,�” i.e., �“competitive injury.�” Falls CityIndus., Inc. v. Vanco Beverage, Inc., 460 U.S. 428,434-35, 103 S.Ct. 1282, 75 L.Ed.2d 174 (1983)(emphasis added). As we stated in J.F. Feeser, Inc.v. Serv-A-Portion, Inc., �“[i]n keeping with the Act�’sprophylactic purpose, designed to prevent theoccurrence of price discrimination rather than toprovide a remedy for its effects, section 2(a) doesnot require that the discrimination must in facthave harmed competition. Instead, a reasonablepossibility of harm, often referred to as competi-tive injury, must be shown.�” 909 F.2d 1524, 1531(3d Cir. 1990) (internal citations and bracketsomitted).

�“Competitive injury�” is established prima facieby proof of �“a substantial price discriminationbetween competing purchasers over time.�”7 FallsCity Indus., 460 U.S. at 435, 103 S.Ct. 1282 (citingFTC v. Morton Salt Co., 334 U.S. 37, 46, 50-51, 68S.Ct. 822, 92 L.Ed. 1196 (1948); id. at 60, 68 S.Ct.822 (Jackson, J., dissenting in part)) (emphasisadded). In order to establish a prima facie viola-

14a


7 Although the purchasers in Falls City Industrieswere in actual competition with one another, injury to com-petition between a purchaser and a customer of a purchaseris also actionable under the Act. As the Supreme Courtmade clear in Perkins v. Standard Oil Co., 395 U.S. 642, 89S.Ct. 1871, 23 L.Ed.2d 599 (1969), there is no basis in theAct for immunizing price discrimination �“simply because theproduct in question passed through an additional formalexchange before reaching the level of [the plaintiff�’s] actualcompetitor.�” Id. at 648, 89 S.Ct. 1871.

tion of section 2(a), Feesers does not need to provethat Michael Foods�’ price discrimination actuallyharmed competition, i.e., that the discriminatorypricing caused Feesers to lose customers toSodexho. Rather, Feesers need only prove that (a)it competed with Sodexho to sell food and (b) therewas price discrimination over time by MichaelFoods.8 This evidence gives rise to a rebuttableinference of �“competitive injury�” under § 2(a). SeeMorton Salt, 334 U.S. at 46, 68 S.Ct. 822. Theinference, if it is found to exist, would then haveto be rebutted by defendants�’ proof that the pricedifferential was not the reason that Feesers lostsales or profits. See Falls City Indus., 460 U.S. at435, 103 S.Ct. 1282.

The District Court required Feesers to prove toomuch. It placed the burden on Feesers to show notonly that it �“actually competes�” with Sodexho, but

15a


8 In Volvo Trucks North America, Inc. v. Reeder-SimcoGMC, Inc., the Supreme Court reiterated that �“a permissi-ble inference of competitive injury may arise from evidencethat a favored competitor received a significant price reduc-tion over a substantial period of time.�” 546 U.S. 164, 126S.Ct. 860, 870, 163 L.Ed.2d 663 (2006). In that case, the juryverdict for the plaintiff was overturned because the plain-tiff�’s evidence, which compared �“occasions on which it com-peted with non-Volvo dealers for a sale to Customer A withinstances in which other Volvo dealers competed with non-Volvo dealers for a sale to Customer B�” failed to show that,over time, the defendant consistently favored other Volvodealers over the plaintiff. Id. at 871. The plaintiff�’s evidenceshowed that the plaintiff competed directly with other Volvodealers for a sale to a particular customer on only two occa-sions and failed to show that the price discrimination onthose two occasions was significant. Id. at 872. Thus, theplaintiff�’s evidence was insufficient to raise an inference ofcompetitive injury.

also that �“food costs and distribution are the deter-mining factors�” in a consumer�’s choice betweenhiring Sodexho or Feesers, i.e., that Sodexho�’slower food prices are why customers switch frombuying products from Feesers to buying productsand management services from Sodexho. Feesers,slip op. at 45 (emphasis added). In the absence ofsuch evidence, the District Court concluded thatFeesers failed to establish �“actual competition.�”

The District Court was concerned that Sodexhoand Feesers are not at the same �“functional level�”and are therefore not in �“actual competition�” inthe same market. This concern is understandablegiven that the facts of this case are somewhatunusual. First, the involvement of Sysco createsan additional link in the chain of distributionbetween Michael Foods and Sodexho, which doesnot exist in Feesers�’ distribution chain. Second,most alleged violations of section 2(a) of theRobinson-Patman Act involve competition betweentwo traditional resellers, such as two food dis-tributors or two retail gas stations, that buy com-modities from a seller and then resell thecommodities to customers. Here, however, Feesersis a traditional commodity reseller, while Sodexhoresells commodities to clients only in conjunctionwith the sale of services, such as food preparationand facility management services.

However, as we observed in Stelwagon Manu-facturing Co. v. Tarmac Roofing Systems, Inc., therelevant question is whether two companies are�“in economic reality acting on the same distribu-tion level,�” rather than whether they are bothlabeled as �“wholesalers�” or �“retailers.�” 63 F.3d1267, 1272 (3d Cir. 1995). To determine whether

16a


Sodexho and Feesers compete to resell food prod-ucts to the same group of customers, we must con-duct a �“careful analysis of each party�’s customers.Only if they are each directly after the same dol-lar are they competing.�” M.C. Mfg. Co. v. Tex.Foundries, Inc., 517 F.2d 1059, 1068 n. 20 (5thCir. 1975); see also George Haug Co. v. Rolls RoyceMotor Cars, Inc., 148 F.3d 136, 141-42 (2d Cir.1998) (noting that determining �“the presence orabsence of functional competition between pur-chasers of a commodity is simply a factual processwhich focuses on whether these purchasers weredirectly competing for resales among the samegroup of customers�”).9 The District Court did notview the evidence, as it should have, in the lightmost favorable to Feesers, and instead found thatFeesers and Sodexho do not compete, without giv-

17a


9 With due respect to our dissenting colleague, thatSodexho�’s business is of a �“different character�” thanFeesers�’, Dis. Op. at 24, is beside the point when we areevaluating whether Feesers has established that it is in�“actual competition�” with Sodexho. The threshold questionis whether a reasonable factfinder could conclude thatSodexho and Feesers directly compete for resales of MichaelFoods products among the same group of customers. The dif-ference in the character of these two businesses might verywell be determinative at the next stage of the analysis dis-cussed below, namely, in evaluating defendants�’ evidencethat facilities choose to buy from Sodexho rather thanFeesers for reasons unrelated to Sodexho�’s lower food prices.It may well be found, based on defendants�’ evidence, thatthe different character of Sodexho�’s business, rather than itslower food prices, causes customers to buy food fromSodexho rather than Feesers. If this is the case, thenFeesers�’ claim under the Robinson-Patman Act fails. How-ever, this is not the same as finding that they are not in�“actual competition.�”

ing due consideration to the evidence of actualcompetition proffered by Feesers.

The evidence here could lead to a different con-clusion than that reached by the District Court.Although Sodexho resells Michael Foods productsto food service facilities that it operates, whileFeesers resells Michael Foods products to self-operated food service facilities, the evidence,viewed in the light most favorable to Feesers,shows that Feesers�’ customers and Sodexho�’s cus-tomers are not two separate and discrete groups offood service facilities. Feesers proffered evidencethat customers may be self-operated for sometime, then switch to Sodexho, or, alternatively,may be operated by Sodexho and then switch toself-operation. Two food service facilities, St.Mary�’s Catholic School and the Jewish Home ofGreater Harrisburg, were Feesers customers andself-operated facilities, but then switched to beingoperated by Sodexho and no longer buy food fromFeesers. App. 7072, 7139. Feesers also profferedevidence that the Meadows Nursing Home was aSodexho customer and switched to being a self-operated facility and a Feesers customer, in partbecause Michael Foods agreed to give Feesers thesame pricing as it gives to Sodexho. App. 7073.There is also evidence that Sodexho actively solic-its self-operated facilities to become Sodexho-oper-ated, and also loses some customers each yearthat decide to self-operate instead of usingSodexho�’s operation services.

Our dissenting colleague attributes customers�’decisions to switch from buying products fromFeesers to buying products from Sodexho to thefact that �“clients may choose to switch between

18a


the market for unprepared food stuffs and themarket for prepared meals.�” Dis. Op. at 33. Hesuggests that Sodexho does not sell unpreparedfood, but rather �“prepared meals,�” and that we areconfusing �“cost accounting with actual businesstransactions�” by concluding otherwise. Id. at 25.However, the record in this case belies that asser-tion. To the contrary, a factfinder could concludethat Sodexho sells unprepared food to its cus-tomers. The record is replete with agreementsbetween facilities and Sodexho wherein the facil-ities are not charged for �“prepared meals,�” butrather for the cost of unprepared food and sup-plies, the cost of labor, and a management fee.Sodexho in fact promotes its ability to get lowerprices for the food products that its customers usein their facilities. Sodexho notes in its promotionalmaterials that �“food and supplies are a major por-tion of the cost of a food service program.�” App.5121. It goes on to boast that its �“extensive net-work of purchasing resources can lower the pricesof food and supplies . . . while actually improvingthe quality of the products you use.�” Id. In its pro-motional materials and proposals to potentialclients, Sodexho could not be more clear that itsells food products to its clients and passes alongthe price discounts that it is able to secure fromits product suppliers in the price that it chargesits clients for the products.10 In fact, Sodexho�’s

19a


10 See App. 5121 (Proposal for Northern BurlingtonCounty Regional School District) (�“Our reputation and sizegive us advantages over smaller food service managementorganizations. In turn, the savings in which [sic] we obtainwill be passed on to your District. You will be charged thesame prices as Sodexho Marriott Services pays for all prod-

superior product prices are touted as resultingfrom Sodexho�’s �“leveraging [its] procurementpower as the industry�’s largest purchaser of food.�”App. 3806 (Proposal for Abington Friends School).This is a major thrust of its sales pitch.11

Sodexho�’s charging its customers for the cost offood products cannot be characterized as mere�“cost accounting�” any more so than any other busi-ness�’ charging a customer for invoiced goods isjust �“cost accounting.�” At minimum, Feesers hasproffered sufficient evidence to create a genuinefactual dispute as to whether Sodexho and Feesersboth resell food products to the �“same group ofcustomers.�”

If substantial price discrimination between com-peting purchasers over time is established, thenthe inference of competitive injury arises. See

20a


ucts. Your District will receive all the benefits of our volumeand trade discounts, except for cash discounts.�”).

11 See App. 3622 (Proposal for Beth Sholom Home ofEastern Virginia) (�“Utilization of the Sodexho purchasingprogram provides great financial benefits to our partnerfacilities. As the industry leader in food procurement withpurchasing responsibility for approximately 5,300 facilitiesthroughout the United States, Sodexho is able to purchasefood at pricing that is not able to be realized by smallerorganizations.�”); App. 3650 (Proposal for Lancaster RegionalMedical Center) (�“Sodexho Marriott Services clients benefitfrom the combined purchasing power of our company withMarriott International, Inc. and Sodexho Alliance. Our foodand supply prices are exceptional, as are the quality andsystems used to support the purchasing function. In additionto those savings, you enjoy discounts on many other itemsyou buy, such as food service equipment, laboratory sinks,uniforms for front desk or security personnel, light bulbs,carpet, etc. Our prices for most items range from 5 to 25%lower than the next best price.�”).

Morton Salt, 334 U.S. at 46, 68 S.Ct. 822. How-ever, this inference is not irrebuttable. As theSupreme Court stated in Falls City Industries,Inc. v. Vanco Beverage, Inc., 460 U.S. 428, 103S.Ct. 1282, 75 L.Ed.2d 174 (1983):

In Morton Salt this Court held that, forthe purposes of § 2(a), injury to competi-tion is established prima facie by proof of a substantial price discriminationbetween competing purchasers over time.334 U.S. at 46, 50-51, 68 S.Ct. at 828,830-831; see id. at 60, 68 S.Ct. at 835(Jackson, J., dissenting). In the absence ofdirect evidence of displaced sales, thisinference may be overcome by evidencebreaking the causal connection between aprice differential and lost sales or profits.F. Rowe, Price Discrimination Under theRobinson-Patman Act 182 (1962); seeChrysler Credit Corp. v. J. Truett PayneCo., 670 F.2d 575, 581 (CA5 1982).

Id. at 435, 103 S.Ct. 1282 (emphasis added). Theinference could be rebutted with evidence prof-fered by defendants that the price discriminationdoes not cause food service facilities to decide tobuy food from Sodexho rather than Feesers. How-ever, the District Court improperly put the burdenon Feesers to prove that a difference in the price ofproducts causes facilities to switch from buyingfrom Feesers to buying from Sodexho. Feesers, slipop. at 22 (�“It is undisputed that Michael Foodsoffered Feesers pricing that matched its pricing toSodexho because the Meadows was a Sodexhoclient, however, Feesers failed to establish that

21a


the availability of that pricing was the determin-ing factor for the Meadows in making theswitch.�”). This was error. The burden is on defen-dants to show the absence of the causal link.

Our dissenting colleague takes issue with theRobinson-Patman Act on policy grounds and urgesthat we are applying it too broadly, so as to renderprice discrimination between non-competitors aviolation of the Act. We reject this characteriza-tion of the record before us, and suggest thatCongress has written the law, and courts haveconstrued it, to apply to situations where dis-criminatory pricing poses a threat to competition.Viewing the evidence in the light most favorableto Feesers, a factfinder could conclude that this issuch a situation. Therefore, it is for the factfinder,here the District Court, to decide whether defen-dants�’ actions fit within the contours of whatCongress has proscribed. We will remand for it todo so.

IV.

Accordingly, for the reasons set forth, we willreverse the grant of summary judgment in favor ofdefendants and remand for further proceedingsconsistent with this opinion.

JORDAN, Circuit Judge, dissenting.To prove its case under the Robinson-Patman

Act, Feesers has tried to show that it is in actualcompetition with Sodexho. Feesers has argued atlength about customers switching from self-oper-ation to outsourcing and back again. Those argu-

22a


ments, however, start with the premise thatFeesers and Sodexho sell the same products. Theevidence is to the contrary, and, in my view,Feesers has failed to raise a genuine issue ofmaterial fact on this crucial point. Because sum-mary judgment for Michael Foods and Sodexho isproper on that basis alone, I respectfully dissent.

I

The undisputed evidence in this case demon-strates that Sodexho�’s business is of a very dif-ferent character than Feesers�’s. Feesers buysunprepared food from suppliers, such as MichaelFoods, and resells that unprepared food to itsinstitutional clients. Feesers�’s involvement endsthere. Its clients then take the unprepared foodsand prepare meals for their individual customers.Sodexho, on the other hand, is a food managementcompany that contracts with institutions to man-age food service operations. Its institutionalclients do not themselves provide food service.Instead, Sodexho buys the unprepared food, pre-pares meals, and sells the prepared meals to indi-vidual customers. Unlike Feesers, Sodexho doesnot sell unprepared food.12

23a


12 I agree with the majority, Maj. Op. at note 4, that weshould not consider the activities of Entegra ProcurementServices, LLC. While it is a wholly-owned subsidiary ofSodexho, Entegra is a separate legal entity. Feesers has notpresented a sufficient basis for piercing the corporate veiland holding Sodexho liable for Entegra�’s actions. Thatleaves the question of whether Michael Foods could be liablefor discriminating in favor of Entegra rather than Sodexho.However, Feesers has not made out a prima facie case ofprice discrimination based on sales made to Entegra,

Feesers inaccurately claims the contrary is true.Relying on a contortion of terms in Sodexho�’s con-tracts with some of its institutional clients,Feesers says that Sodexho does distribute unpre-pared foods. More specifically, because Sodexho issometimes reimbursed by its customers for certainoperating expenses, including the cost of food,Feesers contends that Sodexho is selling unpre-pared food products to its clients. The DistrictCourt agreed, stating that Sodexho sells food to itsinstitutional clients, because �“[t]he Sodexho pro-posals and contracts that Feesers has provided asevidence establish that Sodexho, at least in somecases, accounts for food costs as a separate lineitem within operating costs when billingaccounts.�” Feesers�’s argument and the DistrictCourt�’s conclusion, which, I regret, my colleaguesin the majority have accepted, confuses costaccounting with actual business transactions.There is a world of difference between the two. Cf.Creque v. Texaco Antilles Ltd., 409 F.3d 150, 154(3d Cir. 2005) (holding that a conveyance of prop-erty was not actually a sale despite the use ofaccounting formalities, because �“we must lookbeyond formalities and accounting entries to thetrue nature of the conveyance�”).

Sodexho and its clients agree to allocate costsand profits in various ways. For some of itsclients, Sodexho operates the food service andassumes all responsibility for either making aprofit or losing money. (Appx. at A1545, 12:5-8.) Ifsales are less than costs for those accounts,Sodexho bears the loss. (Id. at A1545, 12:9-11.)

24a


because Feesers has failed to present any evidence of suchsales.

For other clients, Sodexho is reimbursed for oper-ating costs and charges a management fee, withthe remaining profit or loss either going to theclient or being shared between the client andSodexho. (Id. at A1546-48, 13:10-15:21.) In thosecases, Sodexho invoices the client for specifiedoperating expenses, including software, informa-tion systems, decorations, delivery services,unprepared food stuffs, and salaries for Sodexhoemployees. (Id. at A2160-61, A2177-78, A2195-96,A2215-16, A2233-34; see also id. at A1256-66.)

Sodexho�’s receiving reimbursement of suchexpenses according to these contracts is nothingmore than an accounting method that allowsSodexho and its clients to allocate potential prof-its or losses. The accounting method does notmean that Sodexho is in the business of sellingunprepared food, any more than it means Sodexhois a seller of computer software, or of accountingservices, or decorations, or any other specificallylisted operating expenses. If Microsoft tried toclaim Sodexho was competing with it for softwaresales, it would be only marginally more of astretch than Feesers�’s claim. There is no evidencesupporting the notion that any Sodexho clientcalls and asks for a hundred bags of frozen pota-toes, as they might when calling Feesers. Theycall Sodexho when they want prepared french friesand other ready-to-eat food for their customers.The cost accounting provisions in the Sodexho con-tracts simply do not support the conclusion thatSodexho sells unprepared food products in com-petition with Feesers.13

25a


13 Likewise, Sodexho�’s promotional materials, whichtout its ability to negotiate low acquisition prices for unpre-

We are left, then, with the following facts.Feesers buys and resells unprepared food.Sodexho buys unprepared food, prepares meals,and then sells the prepared meals. The preciselegal issue presented is whether those facts raisea genuine issue as to �“actual competition�” betweenFeesers and Sodexho, as that requirement is prop-erly understood under the Robinson-Patman Act.As discussed below, I do not believe they do.

II

Some historical perspective is in order. TheRobinson-Patman Act has been called the �“WrongWay Corrigan�” of antitrust, because it �“often oper-ates to harm consumers for the benefit of weakeror less efficient dealers. It moves antitrust policyin precisely the wrong direction.�” Herbert Hov-enkamp, The Antitrust Enterprise 192 (2005). Thatthis case is now moving forward for trial high-lights both the misguided policy behind the Robin-son-Patman Act and the blunt mechanisms used

26a


pared foods, do not demonstrate that it sells unpreparedfood to its clients. Those materials do not change the factthat Sodexho buys unprepared food and, instead of resellingit, uses it in a business that changes it into a different prod-uct, namely prepared meals. That Sodexho is able to operateat lower cost is important to its institutional clients notbecause those clients have any interest in repurchasingunprepared food. They do not, since they are not self-oper-ating cafeterias. It is important because lower operatingcosts translate into more profit to be shared by Sodexho andthe clients. Thus, the majority opinion is, I believe, mistakento rely on those promotional materials as showing thatSodexho is in the business of reselling the unprepared foodstuffs it acquires from Michael Foods.

to enforce it. Summing up the virtually uniformdisdain which antitrust experts have long had forthe Act, Judge Robert Bork wrote almost thirtyyears ago that �“[a]lthough [the Act] does not pre-vent much price discrimination, at least it has sti-fled a great deal of competition.�” Robert H. Bork,The Antitrust Paradox 382 (1978). This casedemonstrates the Act�’s exceedingly counter-pro-ductive character.

First of all, as a theoretical matter, there is noreason to presume that price discrimination posesa threat to competition. Price discounts are gen-erally good for consumers. The theory behind theAct is that one competitor may use a price differ-ence to drive its (presumably smaller and weaker)competitors out of the market. In the absence ofmarket power, however, such a scheme is highlyunlikely to succeed. A manufacturer like MichaelFoods generally has no interest in shutting downefficient distribution channels for its products,because it is locked in competition with other foodsuppliers. Distributors like Feesers that areunhappy with the prices charged by Michael Foodshave the option, in a competitive market, to geteggs and potatoes elsewhere. Thus, any realthreat to competition requires monopolistic mar-ket power and could be dealt with under the Sher-man Act, with the accompanying requirement forproof of such power.

That difference in required proof is crucial, andhighlights why, even if price discrimination werea real threat to competition, the Robinson-PatmanAct is not a good means to stop it. The Morton Saltinference discussed by the majority, Maj. Op. atSec. III, allows plaintiffs to proceed to trial in a

27a


Robinson-Patman case without any proof thatcompetition has been or will be harmed. Instead,such plaintiffs rely on the threat of harm to them-selves as a proxy for threatened harm to compe-tition. The difficulty is that a competitor will alsobe harmed by vigorous competition, if that com-petitor cannot adjust by becoming more efficient.The Act provides no way of distinguishingbetween an inefficient competitor and one that isharmed by an actual threat to competition itself.

These logical flaws in the Act have led to con-siderable academic criticism of it and haverecently prompted the Antitrust ModernizationCommission, which was created by statute andappointed by the President and the leadership ofCongress, to recommend that Congress repeal theAct in its entirety. Antitrust Modernization Com-mission, Report and Recommendations, April2007, at iii, 317-26. According to the Commission,the Act is �“antithetical to core antitrust princi-ples,�” because it �“protects competitors over com-petition and punishes the very price discountingand innovation in distribution methods that theantitrust laws otherwise encourage.�” Id. at iii.

Now, I readily acknowledge that these policyconcerns cannot override the will of Congress, andI do not suggest that this Court should attempt torepeal the Act by construing it into the oblivion itso richly deserves. But, given the threat that anoverly broad reading of the Act poses to desirablecompetition, this Court certainly should not readthe Act to cover factual situations where only atenuous argument can support its application.

That the Act should be construed relatively nar-rowly is not a radical approach. On the contrary,

28a


the Supreme Court has recently emphasized thatthe Act should be construed �“consistently withbroader policies of the antitrust laws.�” VolvoTrucks N. Am., Inc. v. Reeder-Simco GMC, Inc.,546 U.S. 164, 126 S.Ct. 860, 873, 163 L.Ed.2d 663(2006) (internal quotation marks omitted).Because lower prices are generally good for con-sumers, applying the Act broadly threatens todampen desirable price competition, forcing con-sumers to pay higher prices for goods. To avoidthat threat, the Supreme Court has stated that itwill �“resist interpretation [of the Act] geared moreto the protection of existing competitors than tothe stimulation of competition.�” Id. at 872 (empha-sis in original). In particular, an interpretation ofthe Act that protects individual distributorsrather than competition between brands ignoresthe �“primary concern�” of the antitrust laws withinterbrand, rather than intrabrand, competition.Id.; see also Leegin Creative Leather Prods., Inc. v.PSKS, Inc., 551 U.S. 877, 127 S.Ct. 2705, 2715,168 L.Ed.2d 623 (2007) (�“[T]he primary purpose ofthe antitrust laws is to protect this type of [inter-brand] competition.�” (internal citation and quo-tation marks omitted)). We should be following theSupreme Court�’s lead in resisting such an inter-pretation. Instead, the decision today goes beyondeven the protection of competitors to the protec-tion of non-competitors.

The requirement that a claimant show actualcompetition limits the Act to its proper scope.�“Mindful of the purposes of the Act and of theantitrust laws generally,�” the Supreme Court hasexplained that the Act �“does not ban all price dif-ferences charged to different purchasers of com-

29a


modities of like grade and quality.�” Volvo Trucks,126 S.Ct. at 870 (internal quotation marks omit-ted). �“[R]ather, the Act proscribes price discrimi-nation only to the extent that it threatens toinjure competition.�” Id. Therefore, while �“a per-missible inference of competitive injury may arisefrom evidence that a favored competitor receiveda significant price reduction over a substantialperiod of time,�” such an inference only arises ifthe two purchasers are in �“actual competition.�”Id.; see also Stelwagon Mfg. Co. v. Tarmac RoofingSys., Inc., 63 F.3d 1267, 1271 (3d Cir. 1995).

The competitive injury inference was first dis-cussed some sixty years ago in the Morton Saltcase. 334 U.S. 37, 50-51, 68 S.Ct. 822, 92 L.Ed.1196 (1948). There, small grocery stores wereallegedly harmed by volume discounts on Mortonbrand salt that were given to large chain grocerystores. Id. at 41, 68 S.Ct. 822. That situation pre-sented the paradigmatic set of facts that Congresswas attempting to address with the Robinson-Pat-man Act. Congress sought to address the per-ceived evil of large chain stores securing volumediscounts not available to small independently-owned stores. Volvo Trucks, 126 S.Ct. at 869(�“Congress responded to the advent of large chainstores. . . .�”); see also Richard A. Posner, TheRobinson-Patman Act 25-26 (1976) (calling the Act�“the high-water mark of the anti-chain-storemovement�”). In Morton Salt, the competing storespurchased and resold the same commodity, tablesalt, to the same group of customers.

Last year, in Volvo Trucks, the Supreme Courtdeclined to apply the Morton Salt inference,because the plaintiff, a Volvo dealer, had failed to

30a


show that it actually competed with the otherdealers who allegedly received more favorableprices on trucks made by Volvo. 126 S.Ct. at 870-72. In a market that operates by bidding, theplaintiff could not show that it had ever directlycompeted on a bid with a favored dealer. Id. at871. The Court compared the situation to the Mor-ton Salt paradigm, stating that �“there [was] nodiscrete �‘favored�’ dealer comparable to a chainstore or a large independent department store.�”Id. Thus, the Act did not prohibit the differentprices offered to the Volvo dealers.

Until now, we too have limited the Morton Saltcompetitive injury inference to cases like MortonSalt. In J.F. Feeser, Inc. v. Serv-A-Portion, Inc.,we held that the plaintiffs, including the sameFeesers we see here,14 competed with other dis-tributors to buy and resell the same portion-con-trolled food products. 909 F.2d 1524, 1526-27 (3dCir. 1990). More recently, in Stelwagon Mfg. Co. v.Tarmac Roofing Sys., Inc., we held that the plain-tiff, a distributor of roofing products, could be inactual competition with a company that, althoughit was known as a manufacturer, actually pur-chased the identical roofing products and resoldthem to the same group of customers as did theplaintiff. 63 F.3d 1267, 1271-72 (3d Cir. 1995). Inboth cases, the �“actual competition�” arose from theresale of identical products to the same group ofcustomers, just as in Morton Salt.

Similarly, the Court of Appeals for the SecondCircuit has stated that �“[d]etermining the pres-

31a


14 By the time the J.F. Feeser case reached this Court,J.F. Feeser, Inc. had been renamed Feesers, Inc. 909 F.2d at1526.

ence or absence of functional competition betweenpurchasers of a commodity is simply a factual pro-cess which focuses on whether these purchaserswere directly competing for resales among thesame group of customers.�” George Haug Co. v.Rolls Royce Motor Cars Inc., 148 F.3d 136, 141-42(2d Cir. 1998) (citing FTC v. Fred Meyer, Inc., 390U.S. 341, 349, 88 S.Ct. 904, 19 L.Ed.2d 1222(1968)). In the George Haug case, a service stationpurchased and resold the same Rolls Royce auto-mobile parts as a Rolls Royce dealer that allegedlyreceived more favorable prices from the manu-facturer. Id. at 141. Such direct competition forthe resale of the same product to the same cus-tomers qualifies as �“actual competition�” under theAct.

In this case, Feesers has succeeded in removingthe concept of �“actual competition�” from its foun-dations in Morton Salt. The undisputed facts showthat Sodexho and Feesers do not sell the sameproducts, not even some of the time. Feesers sellsunprepared foodstuffs, while Sodexho preparesand sells meals. Sodexho does not provide unpre-pared food in addition to other services; it oper-ates strictly in the separate market for preparedmeals. The fact that clients may choose to switchbetween the market for unprepared food stuffsand the market for prepared meals does not makethe markets the same and is, therefore, beside thepoint. To conclude that Feesers and Sodexho arein actual competition to sell to the same market,we would also have to conclude that grocery storesare in actual competition with restaurantsbecause both types of businesses sell food. Even if,in the abstract, that could be called competition,

32a


the situation is far removed from the one in Mor-ton Salt and should not be held to satisfy therequirement for �“actual competition�” under theAct. By sending this case back for trial, wewrongly give credence to a theory of �“actual com-petition�” so broad as to effectively read therequirement out of the Act.

Because the facts here fail to show actual com-petition, as required for Feesers to prove its case,I would affirm the grant of summary judgment forthe defendants, and I therefore dissent.

33a


APPENDIX B

UNITED STATES DISTRICT COURTFOR THE MIDDLE DISTRICT OF PENNSYLVANIA

__________Civil No. 1:CV-04-0576Judge Sylvia H. Rambo


Plaintiff,v.

MICHAEL FOODS, INC. and SODEXHO, INC.,Defendants.__________

MEMORANDUM

I. IntroductionIn this Robinson-Patman Act case, Plaintiff

Feesers, Inc., a broad line food distributor, claimsthat Defendant Michael Foods discriminatedagainst Plaintiff by offering lower prices on its eggand potato products to Defendant Sodexho, a foodservice management company. Feesers chargesMichael Foods with a violation of Section 2(a) ofthe Robinson-Patman Act for price discrimination,and Sodexho with a violation of Section 2(f) forinducing the price discrimination.

34a


This lawsuit was initiated by Feesers on March17, 2004. Thereafter, the parties moved for sum-mary judgment. In May 2006, this court foundthat Feesers had established the first three elements of the prima facie case of price discrim-ination, but granted summary judgment to Defen-dants on the ground that Plaintiff had failed tooffer sufficient evidence to establish competitiveinjury. Plaintiff appealed the adverse holding tothe Third Circuit Court of Appeals, whichreversed this court�’s holding that there was noevidence of competitive injury and remanded thecase for trial. See Feesers, Inc. v. Michael Foods,Inc., 498 F.3d 206 (3d Cir. 2007).

Over the course of three weeks in January 2008,this court held a trial and received evidence on thefive issues remaining to be decided. First, whetherPlaintiff is entitled to an inference of competitiveinjury, the fourth element of the prima facie caseof price discrimination. Second, if Plaintiffreceives an inference of competitive injury,whether Defendants have rebutted that inferenceby breaking the causal connection between thelower prices and any competitive injury. Third,whether Defendants have established the affir-mative defense of �“meeting competition�” by show-ing that the lower prices to Sodexho were offeredto meet the equally low prices offered by MichaelFoods�’ competitors. Fourth, whether Plaintiff hasproven that Sodexho induced price discrimination.Finally, whether Plaintiff is entitled to equitablerelief. The following are the court�’s findings offacts and conclusions of law.

35a


II. Prima Facie Case of Price DiscriminationThe Robinson-Patman Act was passed by

Congress to respond to the issue of large chainstores utilizing their great purchasing volume tosecure lower prices than their smaller competi-tors. Section 2 of the Clayton Act, as amended bythe Robinson-Patman Act, provides in pertinentpart:

It shall be unlawful for any personengaged in commerce . . . to discriminatein price between different purchasers ofcommodities of like grade and quality, . . .where the effect of such discriminationmay be substantially to lessen competitionor tend to create a monopoly in any line ofcommerce, or to injure, destroy, or preventcompetition with any person who eithergrants or knowingly receives the benefit ofsuch discrimination, or with customers ofeither of them.

15 U.S.C. § 13(a). Where a disfavored purchaserestablishes that a price discrimination has causedlost sales or profits, competitive injury is estab-lished. However, in Morton Salt, the SupremeCourt held that because the Robinson-Patman Actaims to prevent such injury, proof of lost sales orprofits is not necessary to seek injunctive reliefunder the Act. F.T.C. v. Morton Salt, 334 U.S. 37,49-51, 68 S. Ct. 822, 92 L. Ed. 1196, 44 F.T.C.1499 (1948). Specifically, the Court recognizedthat a permissible inference of competitive injurymay arise where a favored purchaser receives asignificant discount from the price received by its

36a


competitors that endures over a substantial periodof time. Id.

Accordingly, to establish an inference of pricediscrimination under the Robinson-Patman Act inthe absence of lost sales or profits, a plaintiff mustshow: �“(1) that sales were made to two differentpurchasers in interstate commerce; (2) that theproduct sold was of the same grade and quality;(3) that defendant discriminated in price asbetween the two purchasers; and (4) that the dis-crimination had a prohibited effect on competi-tion.�” Feesers, Inc. v. Michael Foods, Inc., 498 F.3d206, 212 (3d Cir. 2007) (citing Texaco Inc. v. Has-brouck, 496 U.S. 543, 556, 110 S. Ct. 2535, 110 L.Ed. 2d 492 (1990)).

Here, the court has already found that Plaintiffhas established the first three elements of theprima facie case of price discrimination.1 FeesersInc. v. Michael Foods, Inc., 2006 U.S. Dist. LEXIS29695, 2006 WL 1274088, at *5-8 (M.D. Pa. May 4,2006). The only element at issue is the final one�—competitive injury. The Third Circuit hasinstructed that:

�“Competitive injury�” is established primafacie by proof of �“a substantial price dis-crimination between competing pur-chasers over time.�” In order to establish aprima facie violation of section 2(a),Feesers does not need to prove thatMichael Foods�’ price discrimination actu-ally harmed competition, i.e., that the dis-criminatory pricing caused Feesers to lose

37a


1 This holding was not challenged by any party onappeal. Accordingly, this ruling remains the law of the case.(See Doc. 230.)

customers to Sodexho. Rather, Feesersneed only prove that (a) it competed withSodexho to sell food and (b) there wasprice discrimination over time by MichaelFoods. This evidence gives rise to a rebut-table inference of �“competitive injury�”under § 2(a). The inference, if it is foundto exist, would then have to be rebuttedby defendants�’ proof that the price differ-ential was not the reason that Feesers lostsales or profits.

Feesers, 498 F.3d at 213 (citing Falls City Indus.,Inc. v. Vanco Beverage, Inc., 460 U.S. 428, 434-35,103 S. Ct. 1282, 75 L. Ed. 2d 174 (1983)).

Accordingly, in order to establish a prima facieshowing of price discrimination, Feesers mustdemonstrate by a preponderance of the evidencethat (1) it was in actual competition for the samedollar with Sodexho for the sale of food to insti-tutional customers, and (2) Michael Foods�’ dis-crimination in price between Sodexho and Feeserswas substantial over time. If this burden is met,then Feesers is entitled to an inference of com-petitive injury. These two elements will be dis-cussed in turn.

A. Actual Competition

1. Legal StandardThe Third Circuit has instructed that �“[t]o

determine whether Sodexho and Feesers competeto resell food products to the same group of cus-tomers �‘[the court] must conduct a careful analy-sis of each party�’s customers. Only if they are eachdirectly after the same dollar are they compet-

38a


ing.�’�” Feesers, 498 F.3d at 214 (quoting M.C. Mfg.Co. v. Tex. Foundries, Inc., 517 F.2d 1059, 1068n.20 (5th Cir. 1975)). However, Feesers need notprove that the price of Michael Foods�’ productswas the determinative factor in any customer�’sdecision to choose Sodexho or Feesers. Feesers,498 F.3d at 213-216. Instead, in order to demon-strate actual competition for the same dollar,Feesers must show that it competes with Sodexhofor the same portion of an institution�’s food ser-vice budget. Id.

1. Findings of FactWith this standard in mind, the court turns to

the evidence presented at trial in this case.Feesers claims that the evidence demonstratesthat it competes with Sodexho to sell MichaelFoods egg and potato products to certain institu-tional food service customers within a 200 mileradius of Harrisburg, Pennsylvania. Defendantsargue that the evidence demonstrates thatFeesers and Sodexho offer two completely differ-ent services�—Feesers food distribution andSodexho food management�—and that customersdo not perceive the two companies to be in com-petition with each other. In evaluating the evi-dence, the court will first discuss the distributionchain in the food service industry generally, andthe roles played by the parties to this suit withinthat system. Thereafter, the court will examinethe customers for whose business Feesers claimsthat both it and Sodexho compete. In particular,the court will focus on the decision making processthat occurs when a customer decides how to pur-chase food, including the timing of the decision

39a


and the factors informing it.2 The following arethe court�’s factual findings, established by a pre-ponderance of the evidence presented at trial.

The food service industry generallyThis case concerns competition for the sale of

Michael Foods�’ egg and refrigerated potato prod-ucts to institutions providing dining services. Inorder to put this competition in context, a generaldescription of food manufacturing, distribution,and institutional dining services is required. Insti-tutional dining services consists of meals preparedaway from home. Institutions providing diningservices include colleges and universities; educa-tional institutions such as elementary and highschools; hospitals, nursing homes, and other groupliving institutions; and corporations.

An institutional dining services program encom-passes a number of discrete functions. These

40a


2 In this analysis, the court looks to the customers�’decision making process not to see if price of food is a deter-minative factor�—an inquiry the Third Circuit has held isinappropriate in assessing whether the parties are in actualcompetition, see Feesers, Inc., 498 F.3d at 213-216�—butrather to determine whether and when a customer mayswitch from self-op (in which food is purchased and deliv-ered from distributors such as Feesers) to food service man-agement (in which food is purchased from food servicemanagement companies such as Sodexho who in turnarrange for purchase and delivery by distributors). Feesersand Sodexho are only in competition if customers have achoice between one to the other. The customers�’ motivationfor choosing one company over the other is relevant to deter-mine whether the inference of price discrimination has beenrebutted, see id. at 216, and will be discussed at lengthbelow.

include menu planning, procurement of food,delivery of the food, hiring and supervision ofemployees to prepare and serve the food, andmaintenance of a kitchen and cafeteria. Institu-tions have a wide range of options to accomplishthese tasks. An institution may choose to self-operate (�“self-op�”) its dining services program andperform all dining services tasks internally. Onthe other hand, an institution may choose to out-source part or all of these functions. When aninstitution makes the decision to outsource, it hasessentially two options. The first is that the insti-tution can act as a general contractor, and hireother companies to perform one or more food ser-vices tasks. The other option is that the institu-tion may choose to hire a food servicemanagement company to act as a general con-tractor. In turn, the food service managementcompany subcontracts the responsibility of pro-viding one or more food service functions to othercompanies.

Regardless of how institutions choose to orga-nize their dining services, all must purchase food(procurement) and have it delivered to the loadingdock of the institution (distribution). Distributionis always performed by a distributor. However, forprocurement�—which in this context includes bothbargaining for the price of the food, and purchas-ing the food�—an institution has a number ofoptions. It may procure food from distributors, orcontract with a group purchasing organization(GPO) or food service management company toprocure food. GPOs generally bargain for a lowerprice, but do not actually purchase the food forresale to institutions. On the other hand, food ser-

41a


vice management companies bargain with manu-facturers for lower food prices and arrange for thepurchase and delivery of that food for resale to theinstitution.

The PartiesThe court now turns to the particular roles

played by the parties to this case within the foodservice industry. Plaintiff Feesers is a broadlinefood distributor, carrying thousands of lines offood and food-related products from many differ-ent manufacturers. Feesers is a regional distrib-utor based in Harrisburg, Pennsylvania, and itdoes most of its business within a 200 mile radiusof Harrisburg, an area encompassing parts ofPennsylvania, New Jersey, Maryland, Virginia,and Delaware. Feesers�’ customers include all seg-ments of the food service market; however, thislitigation concerns Feesers�’ institutional food ser-vice customers such as hospitals, schools, nursinghomes, colleges, and corporations. In addition todirectly distributing food to self-op food servicesproviders, Feesers also contracts with food servicemanagement companies to deliver food to insti-tutional food service companies that they manage.

Defendant Michael Foods is a national manu-facturer of egg and refrigerated potato products.Its egg products are sold under the brand names�“Papetti�’s�” and �“M.G. Waldbaum�” and its refrig-erated potato products under the brand name�“Northern Star.�” Michael Foods is the largest pro-ducer of liquid eggs in the nation.

Defendant Sodexho is a multinational food ser-vice corporation headquartered in France andwhich does business around the world, including

42a


the five state area in which Feesers operates. As aresult of recent mergers with other food servicemanagement companies including Marriott andthe Wood Company,3 Sodexho is currently both thelargest food service management company and thelargest private purchaser of food in the world.Sodexho�’s customers are large institutions acrossthe country, including many schools, hospitals,nursing homes, corporations, and universitieswithin Feesers�’ geographical zone of operation.

Competition between Feesers and SodexhoAs a general rule, institutions with dining ser-

vices fall into two categories: self-op and managed.Feesers only sells food to self-op institutions, andSodexho only sells food to managed institutions inconjunction with its food management services.When a self-op institution procures its food fromFeesers, it generally bargains with Feesers overboth the price of the food and the distribution feefor delivery of that food.4 Alternatively, a self-opmay choose to utilize the services of a GPO to bar-gain for a lower price. In this case, the institutionpurchases food from a distributor at the GPO-

43a


3 At the time of the acquisition, Feesers was the pri-mary distributor for the Wood Company. After the acquisi-tion, Feesers continued to provide distribution for someformer Wood accounts which were being managed bySodexho. Feesers subsequently competed to be the primarydistributor for Sodexho, but lost the bid to Sysco, a nationaldistribution company which remains Sodexho�’s primary dis-tributor.

4 Sometimes an institution bargains directly withMichael Foods for a discount on the food, in a transactionreferred to as a deviated billback. This transaction will bediscussed in detail later.

negotiated price plus a payment to the distributorfor delivery. On the other hand, an institutionmanaged by Sodexho contracts with Sodexho toarrange for the procurement and delivery of rawfood to the institution. Sodexho then bargains fora lower food price from the manufacturer andhires a distributor to purchase food at theSodexho-negotiated price for resale to Sodexho atthat price plus a distribution fee. Sodexho in turnbills the customer for the cost of food at Sodexho�’snegotiated price, plus the distribution fee. Thus,a self-op institutional food service customer pur-chases food from a distributor such as Feesers,while a managed institution purchases the samefood from a food service management companysuch as Sodexho.

Though it would appear that Feesers andSodexho serve two discrete groups of customers, infact institutional customers regularly switch fromself-op to management and vice versa. SomeFeesers customers have switched to Sodexho,including the Jewish Home of Greater Harrisburgand St. Mary�’s Catholic School. More rarely, someSodexho customers such as the Meadows haveswitched to self-op and become Feesers customers.Both Feesers and Sodexho actively seek the busi-ness of self-op institutions. The testimony at trialfrom both Feesers and Sodexho employees, as wellas Sodexho�’s securities filings and strategic plan-ning documents demonstrate that Sodexho seeksto convert self-op institutional customers to foodservice management. Moreover, the same docu-ments demonstrate that Sodexho has been suc-cessful in this goal.

44a


Sodexho DocumentsSodexho�’s internal documents also demonstrate

competition with distributors such as Feesers tosell food to institutions such as schools, hospitals,nursing homes, and colleges. For instance, inSodexho�’s Form 20-F filed with the U.S. Securitiesand Exchange Commission Sodexho describes itscompetition for the business of self-ops as follows:

Our success depends on our ability toretain and renew existing client contractsand to obtain and successfully negotiatenew client contracts.. . .Our business and growth strategiesdepend in large part on the continuationof a trend in business, education, health-care and government markets toward out-sourcing services. The decision tooutsource depends upon customer per-ceptions that outsourcing may providehigher quality services at a lower overallcost and permit customers to focus on corebusiness activities. We cannot be certainthat this trend will continue or not bereversed or that customers that have out-sourced functions will not decide to per-form these functions themselves.Management has also identified a trendamong some of our customers towards theretention of a limited number of preferredvendors to provide all or a large part oftheir required services.

45a


(Sodexho Form 20-F, P303 at 8.) Additionally, indescribing opportunities for future growth,Sodexho notes the following:

Healthcare represents the largest poten-tial market for Food and Facilities Man-agement Services with outsourcing ratescomparatively low. We estimate that morethan half of this market is in short-staycenters (public and private hospitals) andthe remainder in long-term care facilitiesfor the elderly and the dependent. Onaverage, we estimate that about one thirdof this food service market is currentlyoutsourced, with short-stay facilities gen-erally more likely to outsource than long-stay facilities by a ratio of almosttwo-to-one. We estimate that the educa-tion market is about one-third outsourcedin food service, with about one quarter ofprivate sector institutions and about threequarters of public institutions outsourcingfood service.

(P303 at 22.)Sodexho tracked both new self-op conversions

and accounts lost back to self op in documents itreferred to as �“churn reports.�” One such churnreport demonstrates that from 2000 until 2003,Sodexho gained approximately $330 million dol-lars in market share from self-op conversions inthe hospital market, while losing approximately$142 million in accounts that converted back toself-op. (Health Care Services, Hospitals StrategicPlan, FY05-07, P160 at 40.) By contrast, compet-ing food service management companies gained

46a


$408 million of the market in new self-op conver-sions while losing $253 million back to self-op.(Id.) Accordingly, Sodexho concludes that it �“isconverting self-op faster than any other singlecompetitor, and is closing 45% of all Self-Op con-versions in the market.�” (Id.) Churn reportsdemonstrate that institutions regularly switchback and forth between self-op and management,but that management has been gaining marketshare in recent years.

In addition to describing competition for self-ops, Sodexho repeatedly refers to competition withself-ops. For example, in its securities filing,Sodexho describes competition with other food ser-vice management companies, and then goes on todescribe self-ops as a source of competition:

Competition in the industryThere is significant competition in thefood and facilities management servicesbusiness from local, regional, national andinternational companies of varying sizes,a number of which have substantial finan-cial resources. . . . Existing or potentialclients may also elect to self-operate theirfood or other services, or to utilize otherpurchasing arrangements, thereby reduc-ing or eliminating the opportunity for usto serve them or compete for the account.

(P190 at 9.) This characterization of self-ops as acompetitor also appears in Sodexho strategic plan-ning documents assessing competition in varioussegments of the institutional food services indus-try. (See, e.g., Sodexho Health Care Division ThreeYear Plan FY01-03, P189 at 25 (describing self-op

47a


acute care institutional customers as �“an increas-ingly formidable competitor able to effectivelyreplicate contractor offerings�” and noting that�“market dynamics and internal cultural actorscause many self-op institutions to �‘think twice�’before outsourcing.�”); Sodexho Competitive Intel-ligence Overview Findings, Health Care Services:Hospitals Phase I, 2003, P170 at 2 (identifyingself-op as Sodexho�’s number one competitor).)

As noted above, Sodexho, like other food servicemanagement companies, essentially performs thefunction of a general contractor. When a self-opbecomes a Sodexho-managed institution, it reliesupon Sodexho to perform all dining services func-tions for which it was previously responsible,including procurement and distribution of food.Though this function is still performed by a dis-tributor, the institutional customer does notselect, contract with, or pay the distributorselected by Sodexho, as it would if the institutionwere self-op. Thus, when an institution switchesfrom self-op to management, the incumbent dis-tributor is displaced. Conversely when a managedinstitution switches to self-op, the functions pre-viously performed by a food service managementcompany, including the sale and delivery of food,are once again performed by a distributor. Accord-ingly, Feesers and Sodexho compete for the sameportion of an institutions�’ food service budget.

As Sodexho recognized in a number of strategicplanning documents,

Self-operated businesses can be expectedto continue to seek the appropriate bal-ance of cost savings, operational quality(including regulatory compliance), and

48a


patient satisfaction that meet the orga-nization�’s needs and culture.The balance described above will beachieved with or without contractorsdepending upon an organization�’s needsand capabilities at any given time�—therewill be limited loyalty to outsourcing, ingeneral, or any one contractor.Self-operated facilities can be expected touse contractors opportunistically�—themoment value in excess of cost is not per-ceived (facility is now clean/in regulatorycompliance, obvious cost improvementshave been made, step increase in patientsatisfaction has been achieved, etc.) thecontracting relationship will be at risk.The �‘Battle for Value�’ will intensify

(Competitive Intelligence Overview Findings,Health Care Services: Hospitals Phase I, 2003,P170 at 3; P182 at 2.) Similar assessmentsappears in other Sodexho strategic planning doc-uments for the Senior Services sector, (see, e.g.,P178 at 15; P166 at 4), in the Education sector,(P190 at 103 (describing self-ops as Sodexho�’snumber one competitor).) In its Senior ServicesExecutive Abstract Phase I: FY 03-05, Sodexhooffered the following �“Summary of CompetitorStrategy for Self-operated institutions:

Support services, especially food and din-ing services, is considered a �‘core compe-tency�’ and an integral part of the residentoffer provided by Senior�’s facilities. Con-tractors, therefore, are perceived as taking away administrators�’ value and

49a


expertise. Until contractors prove thatthey can provide needed value to respondto new market issues, there will be littlemovement towards outsourcing.Facilities will continue to receive much ofthe critical expertise they require throughgovernment resources, consultants, orvendors (e.g. menus through Sysco).

(P178 at 15; see also Senior Services StrategicPlan FY04-06, Competitive Intelligence, P166 at4.)

Defendants argue that the fact that the strate-gic planning documents do not specifically mention distributors as a competitor categorydemonstrates that distributors such as Feesers donot compete with Sodexho. To the contrary, it isclear that when Sodexho refers to self-op as a�“competitor,�” it means that self-ops are able toreplicate the same functions that Sodexho itselfprovides. This includes procurement of food fromdistributors such as Feesers. Thus, when Sodexhorefers to self-ops as �“competitors�” this includes allother companies providing functions that Sodexhoseeks to contract to perform, including distribu-tors.

In response to the trend towards segmentationof functions, Sodexho explored the possibility ofunbundling its services to win self-op accounts. Inthe Executive Abstract, Senior Services Phase IStrategic Plan FY03-05, Sodexho took note of thefollowing �“Future Opportunit[y]�”

Co-Sourcing: This concept has been indevelopment and some testing over the

50a


past 18 months in both the former Woodand Sodexho companies. A contractualapproach to �‘consulting�’, this offer mayallow us to sell our services in an �‘unbun-dled�’ portfolio to Systems, as well assmaller facilities. We need to continue toenergetically pursue this offer as a possi-ble solution to penetrating the self-opmarket.

(P178 at 29.)

GPOs as CompetitorsThough Sodexho�’s strategic planning documents

do not specifically mention distributors as com-petitors, they do discuss the competitive threatposed by GPOs, whose functions overlap with dis-tributors. These documents shed further light oncompetition between Sodexho and distributors,including Feesers. GPOs negotiate with manu-facturers for lower prices on behalf of their mem-bers. Additionally, GPOs may arrange for the saleand distribution of food at a discount, generally bycontracting with a distributor. Thus, like distrib-utors such as Feesers, GPOs perform some, butnot all of the functions provided by food servicemanagement companies such as Sodexho.

Distributors and food service management com-panies compete with GPOs for the business ofinstitutions providing dining services. In itsstrategic planning documents, Sodexho noted that�“GPOs are an increasingly aggressive competitor�”and warned to �“[l]ook for individual GPOs toemerge as competitors in future years.�” (Compet-itive Intelligence Overview Findings, Health Care

51a


Services: Hospitals Phase I, 2003, P170 at 2.)Sodexho found that self-ops had an �“[i]ncreasingperception that comparable or better purchasingeconomies can be obtained through GPOs. Accord-ingly, there is an increasing number of facilitiesseeking/joining GPOs.�” (Id. at 4.)

The trend continued in the following years. InPhase I: FY 03-05 Health Care Services StrategicPlan, Sodexho describes the health care industryas �“highly dynamic, challenged, and cost focused,�”in part due to �“GPOs and e-commerce producing acommodity approach.�” (P1 82 at 2.) Accordingly,�“as a result of these trends, health care executiveswill look to [Sodexho] to . . . [p]rovide products andservices that deliver cost savings at acceptablequality levels or predictable cost with higher sat-isfaction levels.�” (Id. at 3.)

In the Hospitals 04-05 Strategic Plan�—Phase II,Strategy and Ambition, Sodexho defined the Bat-tle for Value as follows:

Part of our business dilemma and in fact,the whole industry�’s business dilemma, isthat our value proposition has lost its�‘edge.�’ This challenge is clearly indicatedby low penetration rates that haven�’tmaterially moved in years. Under currentmarket dynamics, our forecast is that thehospital market will not produce enough�“actionable�” outsourcing decisions to sus-tain growth for the major competitors (seesales plan). In general, the lines between[Sodexho], contractor capabilities, andself-operated capabilities have blurred.Contractors, therefore, are competing forthe same small slice of a churning market

52a


resulting in a commodity mentality and astrained market. This phenomenon hasalso, in effect, raised the basic �‘price ofadmission�’ higher�—features previouslyviewed by contractors as value added arenow necessary base components of anexpected offer. This is the �“Battle ForValue.

(P167 at 12 (emphasis added).)Later, in its Health Care Services Hospitals

FY05-07 Strategic Plan, Phase II, Sodexhoobserved the trend of increased competition withGPOs continuing:

GPOs /E-Commerce: The growing numberand complexity of GPOs combined withthe emergence of e-commerce businessapplications in the procurement arenahave, in effect, equalized the playing fieldenabling all entities to gain procurementleverage. Even a non-GPO aligned stand-alone facility can aggregate its buy withother facilities through E-commerce. Ourhistorical pricing advantage is dramati-cally minimized. Additionally, GPOsincreasingly dictate many or all aspects ofthe procurement process including prod-uct selection, distribution, etc.

(P160 at 16.) The implication of this trend forSodexho Health Care Services is that �“[h]istoricaland clear point of differentiation will not be therefor us in the future�” �“[i]ncreasing amount of timedefending and explaining (market baskets, meet-ings, etc.) our prices�” �“[w]e are on the defensive

53a


and our credibility suffers�” and �“[c]ompetitors maygain access to our accounts through �‘back door�’purchasing relationships.�” (P160 at 16.)

In a summary of competitor strategy for Sum-mary of Competitor Strategy for GPOs, Sodexhoobserved the following:

�• GPOs will become even more aggres-sive and sophisticated in the analysis,pricing, and level of detail requiredfrom their preferred partners pro-curement programs

�• Increasingly setting pricing structuresto compete with and/or win marketbaskets

�• Emergence of more GPOS that willtarget our high volume accounts . . .

�• Capture all volume discountallowances possible for their members

�• Require preferred partners to run allproduct procurement through theGPO rather than their own programs

�• GPOS will increasingly act as forprofit entities as they seek to add ser-vices and products for their members

�• The line between alliance partner andcompetitor will continue to blur, andlikely disappear, as GPO services andproducts overlap with and contradict[Sodexho�’s] products and services(consulting, etc.)

�• Continue to improve food procurementskills

54a


�• May even begin to bid services via e-commerce

�• Require partners to adhere to theirprogram specifications (distributor,purchasing program, product specs,etc.) and impose financial penaltiesper contract terms for non-compliance

(P160 at 10.) Sodexho concluded: �“Our primaryresponse should be to a) re-evaluate our pricingstrategy so we are competitive at the �‘loadingdock�’, . . . c) explore ways to expand Entegra�’s roleand presence in the healthcare industry.�” (P160 at10.) Sodexho�’s strategy to achieve dominant mar-ket share includes �“Establish competitive foodpricing so [Sodexho] is competitive at the loadingdock versus GPO pricing structure.�” (P160 at 31.)In its Health Care Services Hospitals FY05-07,Strategic Plan Phase I, Sodexho acknowledgedthat it needed to reconsider its relationships withGPOs because �“[t]he line between alliance partnerand competitor will continue to blur, and likelydisappear, as GPO services and products overlapwith and contradict [Sodexho�’s] products and ser-vices (consulting, etc.).�” (P161 at 50.)

This evidence demonstrates that GPOs, like dis-tributors such as Feesers, compete with Sodexhoto perform some, but not all, of the functionsSodexho offers to its customers Institutional cus-tomers may choose to procure food from Sodexhoin conjunction with its management services, orthey may choose to self-op and procure food fromeither a GPO or a distributor such as Feesers. Evi-dence of Sodexho�’s increasingly fierce competitionwith GPOs indicates that Feesers, Sodexho, and

55a


GPOs are all competing for business of the samegroup of institutional customers.

EntegraOne response by Sodexho to increasing seg-

mentation of functions in the food service industrywas to develop its own GPO, Entegra, to providefood service procurement unbundled from the dis-tribution and management services typically pro-vided by Sodexho. Entegra is a wholly ownedsubsidiary of Sodexho, and it has access toSodexho deviated pricing. Entegra is not a defen-dant to this action, but its strategic planning doc-uments shed light on competition between Feesersand Sodexho by offering further evidence of theincreasing segmentation in the institutional din-ing services market and the increasing interest ofinstitutions in saving money by lowering foodprices.

In a strategic planning document, Entegradescribes the Health Care and Senior Servicesmarket as follows:

Heath Care and Senior Services will bedominated by a few key GPO�’s . . . thesehealth care GPO giants will lose marketshare as large systems opt to self-con-tract. In response, the GPO�’s will seek tomaintain their profits by spilling over intothe schools and campus markets wherethey will take hold over the next two tothree years. They will find a responsivecustomer due to strong plays by Sysco andUS Foodservice [both broadline distribu-tors] to increase their margins in thewake of waning national competition.

56a


(Entegra�’s Three-Year Plan, FY2005-2007, PhaseI, P163 at 4.) Further in the same document,Entegra describes a trend towards segmentationof procurement and food service management: �“Itis becoming increasingly more common thathealth care systems will look at procurement andmanagement as two different functions . . . Somecustomers are going a step further and looking atdistributor and manufacturer relationships as twodifferent functions to be outsourced.�” (Id. at 19.)Entegra summarized the risks and opportunitiesof this trend as follows:

There is an increasing tendency of sys-tems in health care to look at procurementas a separate function from the manage-ment services Sodexho provides. . . Thiscan lead to further utilization of entegrawith a resulting erosion in margins. Wewill need to compete for this procurementbusiness in order to continue to haveaccess to our contracted products. . .As systems pull back from GPO relation-ships they will look for providers such asSodexho to offer services that address theneeds of all of their facilities�—manage-ment for some, procurement for others, co-sourcing for yet others. This can be a riskand an opportunity. . . It is likely that cur-rent entegra clients will bid distributionand procurement services as two differentactivities in the future. . . . We expect thatmany will negotiate their own distribution

57a


agreements and look for a third party toprovide manufacturer agreements.

(P163 at 20.)The Entegra Procurement Services Three Year

Plan FY 2005-2007, Phase II supports this trend:Opportunities, Risks & New Business�—Risks continue to be in the area of pricingcompetitiveness and the spread of grouppurchasing organizations. While pricingcompetitiveness is a challenge for entegrafrom a sales and retention perspective itis increasingly becoming an issue for theretention of purchasing in Sodexho-man-aged business. Due to this, entegra isbeing considered more frequently as anoption for Sodexho business.

(P159 at 6.) Entegra continued to observe the mar-ket trend towards segmentation:

It is becoming increasingly more commonthat health care systems will look at pro-curement and management as two differ-ent functions . . . Some customers aregoing a step further and looking at dis-tributor and manufacturer relationshipsas two different functions to be out-sourced.

(Id. at 12.) �“A strong systems offer that includesmanagement services for select sites and pro-curement services for others can be a strong reten-tion activity with existing clients or a lead-in witha potential client.�” (Id. at 13.)

58a


Timing of CompetitionDefendants argue that regardless of the evi-

dence that Feesers and Sodexho compete for thesame institutional customers, they are not inactual competition for any one customer becauseSodexho typically competes only in a formal bid-ding process in which distributors such as Feesersdo not participate. However, contrary to Defen-dant�’s assertions, the evidence presented at trialdemonstrates that competition occurs not just inthe formal request for proposal (�“RFP�”) process,but also on an informal basis all the time. Sodexhosales employees regularly make informal contactswith targeted institutions, some of which areFeesers customers.

As a general matter, Defendants are correct thatthe evidence demonstrates that most self-op insti-tutions engage in a formal RFP process prior tosigning a contract with a food service managementcompany. The RFP process generally only includesfood service management companies such asSodexho, Compass, and Aramark. Depending onthe size and type of institution, the RFP processmay be extremely detailed and lengthy andinclude bids on a wide range of services not pro-vided by distributors. However, not every RFPprocess concludes with a contract with a food ser-vice management company. As Sodexho noted in astrategic planning document, sometimes �“Self-Opsuse contractors to �‘fix�’ current problems and thenreturn to self-op or use the RFP process to gainideas but remain in-house.�” (P190 at 19.)

Additionally, Sodexho competes with distribu-tors such as Feesers outside of the formal RFPprocess. Jay Marvin, Sodexho�’s Senior Vice Pres-

59a


ident for Sales and Marketing for the Health CareDivision, described how Sodexho seeks to acquirea new client. For self-ops, Sodexho first identifiesinstitutions that meet its client profile (generallylarger institutions). Sodexho sales team thenmakes contacts and forms relationships at theinstitution. The team gauges the institution�’sinterest in management and determines whetherthere are any particular problems to be solved. Ifthe institution is interested in management, thenit puts out an RFP. The RFP process for hospitalsis usually quite lengthy and formal, and generallyonly involves other food service management com-panies such as Aramark and Compass. The pro-cess can be less formal with senior servicesinstitutions, but a proposal is still generallyrequired.

Moreover, in some cases, Sodexho�’s proposalsthemselves demonstrate direct competition withdistributors. This can occur when a self-op insti-tution chooses to utilize Sodexho management ser-vices while retaining distributors for some or allfood procurement. For instance, in a proposal tothe Beth Sholom Home of Eastern Virginia,Sodexho urged the institution, which was alreadya Sodexho customer, to utilize Sodexho�’s pro-curement program as well:

Purchasing ProgramUtilization of the Sodexho purchasing pro-gram provides great financial benefits toour partner facilities. As the industryleader in food procurement with purchas-ing responsibility for approximately 5,300facilities throughout the United States,

60a


Sodexho is able to purchase food at pric-ing that is not able to be realized bysmaller organizations.Currently the full resources of theSodexho procurement program are notbeing utilized. A significant portion of thedepartment purchasing is done utilizingUS Foodservice[a broadline food distrib-utor]. It is our recommendation that BethSholom Home of Eastern Virginia takefull advantage of the kosher procurementprogram and pricing afforded by its part-nership with Sodexho. Use of the Sodexhokosher vendors would not only streamlinethe ordering process but would substan-tially reduce pricing, specifically in thekosher poultry market.

(P77 at 2.)

Customer TestimonyDefendants also called ten customer witnesses

who testified unanimously that they did not per-ceive Feesers to be a competitor with Sodexho.5

However, the court does not infer from this testi-mony that Feesers is not in competition withSodexho for the sale of food to institutional cus-tomers for a number of reasons. It appeared thatthe witnesses believed that two companies arecompetitors only if they offer precisely the sameset of services. At least one witness (Shippens-burg) was under the impression that Feesers was

61a


5 The customer testimony will be addressed in greaterdetail when addressing Defendants�’ attempt to rebut theinference of competitive injury.

seeking the opportunity to bid against Sodexho inthe RFP process. Many other institutions hadalready committed to either self-op or manage-ment at the time they were considering eitherFeesers or Sodexho, and accordingly their choiceswere limited to either distributors or managementcompanies.

No testimony was presented by any witness foran institution that was considering a switch fromself-op to management, or vice versa. However, itis clear from Sodexho�’s own internal documentsthat many institutions do regularly switch fromself-op to management. Thus, the court cannotinfer from the testimony of the ten customer wit-nesses who were not considering such a switchthat no customer ever does consider such a switch.Accordingly, the court gives no weight to the con-clusions of the ten customer witnesses thatFeesers does not compete with Sodexho.

3. Conclusions of LawBased on the above findings of fact, the court

concludes that Feesers and Sodexho are in actualcompetition for the same dollar in the sale ofMichael Foods products to institutional customerswithin Feeser�’s geographical zone of operation.Feesers and Sodexho both compete to sell MichaelFoods egg and refrigerated potato products to thesame institutional customers. The court will takethis opportunity to address a few of Defendants�’arguments on this point.

First, Defendants argue that Feesers competesonly with other distributors such as Sysco, ratherthan food service management companies such asSodexho. In support of this argument, Defendants

62a


point out that Sodexho has no delivery trucks orwarehouses, and does not directly perform distri-bution for its clients. However, the fact thatSodexho chooses to subcontract the physical deliv-ery of food to a distributor (generally Sysco) ratherthan perform this function itself is of no signifi-cance in determining whether Sodexho competeswith Feesers. From the perspective of the insti-tutional customer, that customer contracts withSodexho to provide food and distribution, notwith-standing the fact that Sodexho in turn contractswith a distributor to perform the function. Thus,Sodexho competes with distributors in the sale offood and distribution to institutional food servicecustomers. The Supreme Court has repeatedlystated that the Robinson-Patman Act should notbe construed in a way that �“would allow price dis-criminators to avoid the sanctions of the Act bythe simple expedient of adding an additional linkin the supply chain.�” Perkins v. Standard Oil Co.of Cal., 395 U.S. 642, 647, 89 S. Ct. 1871, 23 L.Ed. 2d 599 (1969); accord Texaco Inc. v. Has-brouck, 496 U.S. 543, 567 n.26, 110 S. Ct. 2535,110 L. Ed. 2d 492 (1990) (applying same principleto similar facts).

It is also not significant that Feesers andSodexho have never simultaneously submitted anRFP to the same customer at the same time. Asnoted above, the evidence presented at trialdemonstrates that competition in the industry isnot confined to the formal RFP bidding process.For instance, Sodexho sales employees testifiedthat they solicit the business of self-op facilitiesthat have never before considered food servicemanagement. Additionally, the evidence estab-

63a


lished that some customers initiate the RFP pro-cess in order to gain ideas for how to improvetheir food service programs, but ultimately decideto remain self-op. Other customers utilize foodservice consultants to submit a self-op bid in theRFP process. Finally, some customers hire a foodservice management company to correct problemswith their food service programs, only to returnwhen the problems are fixed. In this competitiveenvironment, the fact that no distributor has eversubmitted a proposal in a food service manage-ment RFP process does not establish that distri-bution companies are not in competition with foodservice management companies for those accounts.Instead the evidence demonstrates that competi-tion occurs when a customer considers switchingfrom self-op to food service management, or viceversa. Although the customer�’s decision may beinfluenced by the RFP process, competition for anaccount is not confined to the process.

Defendants cite Volvo Trucks North Am., Inc. v.Reeder-Simco GMC, Inc., 546 U.S. 164, 126 S. Ct.860, 163 L. Ed. 2d 663 (2006), in support of theirclaim that Feesers was not in competition withSodexho for the sale of Michael Foods products. InVolvo Trucks, the Court held that a truck manu-facturer did not violate the Robinson-Patman Actby offering different discounts to dealers submit-ting bids for custom-made trucks to customersbecause there was no actual competition betweenthe dealers for the same customers. In reachingthis conclusion, the Court carefully examined thecompetitive process for the sale of custom-madetrucks. The record revealed that customers pur-chased custom-made trucks through a competitive

64a


bidding process. Customers solicited bids fromnumerous dealerships selling different brands oftrucks, but because Volvo dealers generally oper-ated in separate geographic zones, customersrarely submitted simultaneous bids from two dif-ferent Volvo dealers at the same time. In prepar-ing a bid, Volvo dealers often requested a discountfrom Volvo. In the rare instance that two Volvodealers were submitting a bid to the same cus-tomer, Volvo�’s policy was to offer the same dis-count to both dealers. Reeder-Simco complainedthat other dealers received steeper discounts fromVolvo, but the Court held that this was an inap-propriate comparison to make because those salesinvolved only other non-Volvo dealers. In otherwords, because the competitive process for thesale of custom-made trucks was formal and lim-ited to only a handful of dealers, Reeder-Simcowas not in actual competition with other Volvodealers in transactions in which it did not submita bid. The Court acknowledged that Reeder-Simcocompeted with other Volvo dealers for the oppor-tunity to bid on sales, but noted that at this stagein the competition, no dealership had secured anydiscount from Volvo that they could use to gain acompetitive advantage over other Volvo dealers.Id. at 178-179. Accordingly, the Court concludedthat the discounts received by other Volvo dealersduring the course of any bidding process in whichReeder-Simco did not bid did not result in com-petitive injury to Reeder-Simco. Id. at 179.

Volvo Trucks teaches that in order to determinewhether two purchasers are in actual competition,a court must carefully examine both the competi-tive process and the customers. Defendants

65a


attempt to draw an analogy between Volvo Trucksand the facts of this case, claiming that the com-petitive process for the sale of Michael Foods prod-ucts is narrowly confined to the RFP process, inwhich Feesers and other distributors do notdirectly participate. However, as noted above, thecourt rejects Defendants�’ narrow view of compe-tition in this case. The evidence presented at trialestablishes that competition for the sale ofMichael Foods egg and potato products to insti-tutional food service customers was ongoing andnot limited to the formal RFP process. Food ser-vice management companies, distributors, andGPOs all compete formally and informally for thesale of food to institutions. Even when a companyinitiates the RFP bidding process, which includesonly food service management companies, thatchoice is not final or limited to the companies sub-mitting a bid. Volvo Trucks is not controllingbecause competition in this case is much broaderthan that at issue in that case. Unlike VolvoTrucks, the fact that Feesers does not participatein head-to-head bidding against Sodexho in a for-mal RFP process does not demonstrate thatFeesers does not compete with Sodexho for thesale of Michael Foods products to those customers.

B. Substantial Price DiscriminationOver Time

1. Legal StandardIn order to pose a risk of injury to competition,

there must be substantial price discriminationover time. �“The presumption of the requisiteadverse competitive effects contemplated by sec-

66a


tion 2(a) is most likely to arise when the price dif-ferential is (1) substantial enough to influence adisfavored customer�’s resale price; or (2) occurs ina market with low profit margins and intensivecompetitive conditions.�” J.F. Feeser v. Serv-A-Por-tion, 909 F.2d 1524, 1538 (3d Cir. 1990) (internalcitations omitted). Temporary or trivial price dif-ferences are not normally enough to risk an injuryto a disfavored competitor. However, the longerthe discount is offered and the greater magnitudethe price difference, the more likely it is that theprice discrimination will cause an injury to com-petition.

2. Findings of FactThe evidence presented at trial establishes that

Michael Foods discriminated in price in favor ofSodexho and against Feesers in the sale of egg andpotato products to a sufficient degree and durationto entitle Feesers to an inference of competitiveinjury. The court will now review the evidence ofprice discrimination presented at trial and brieflyrespond to Defendants�’ criticisms of that evidence.

Michael Foods PricingMichael Foods generally sells its products to dis-

tributors at list price. For egg products, MichaelFoods�’ list price applies nationally. Because com-petition for potato products varies by region due tocompetition by relatively small regional potatoproduct manufacturers, Michael Foods has dif-ferent pricing arrangements based on region forits potato products. However, one list price appliesto distributors competing within the same geo-

67a


graphical region for the sale of Michael Foodspotato products.

Until the mid-1990s, all distributors purchasedfood at the national list price from manufacturers,including Michael Foods. In the 1990s, some largeinstitutions began negotiating directly with man-ufacturers for discounts, or deviations from thelist price, in exchange for guarantees of minimumpurchases. Such transactions proceed as follows:First, a distributor purchases the product at listprice from the manufacturer. The product is pur-chased at list price because at the time of pur-chase, the product is destined for the distributor�’swarehouse, and has not yet been allocated to aparticular customer. Next, the distributor sells theproduct to the designated customer at the devi-ated price plus a distribution markup separatelynegotiated by the customer. Finally, the distrib-utor bills the manufacturer for the differencebetween the list price and the deviated price afterpresenting proof that the product was delivered tothe customer who negotiated the deviated price.

Today, sixty percent of purchases from MichaelFoods are made at deviated, rather than list price,according to Mark Westphal, CFO for MichaelFoods. More large self-op institutions are negoti-ating directly with manufacturers. Institutionsmay contact manufacturers directly, or ask a dis-tributor to assist it in securing deviated pricing.Such deviated pricing is institution-specific. Inother words, a distributor may only acquire goodsat the deviated price that will be sold to institu-tions which have secured a deviated price. Thismeans that the institution may choose anotherdistributor or even food service management, and

68a


still enjoy the lower price it negotiated. Thoughinstitution-specific deviated pricing results inlower pricing to some institutions, the lower pric-ing is unlikely to cause competitive injury becausemost institutions such as schools, hospitals, andnursing homes do not compete with other insti-tutions on the basis of food costs.

In recent years, Michael Foods has extendeddeviated pricing to food service management com-panies, including Sodexho. Unlike the deviatedpricing described above, Sodexho deviated pricingis not institution-specific. Instead, it applies toevery institution that Sodexho manages. Accord-ingly, Sodexho can use its low deviated price bothto win new accounts and to keep current cus-tomers. However, if a customer switches fromSodexho to either self-op or another food servicemanagement company, it loses the benefit of thelower Sodexho prices for food purchases.

In addition to deviated pricing, Michael Foodsgrants a number of other types of discounts. Theseinclude volume discount allowances, preferredsupplier allowances, marketing allowances, blan-ket bid allowances, local allowances, and UniProrebates. These discounts may be granted to eitherdistributors like Feesers or food service manage-ment companies such as Sodexho. A volume dis-count allowance is a price reduction based uponthe volume of a customer�’s purchases. A preferredsupplier allowance is essentially a signing bonusfor designating Michael Foods as a primary sup-plier of a food product. Marketing allowances area type of discount offered for certain products tem-porarily in exchange for the promotion of thoseproducts for resale to customers. For instance,

69a


Michael Foods grants marketing allowances tocustomers who attend a food show promotingMichael Foods food products. Blanket bidallowances are discounts that apply to certain cat-egories of customers, such as nonprofits orschools. Michael Foods also grants discounts tomembers of UniPro, a GPO. Essentially there arethree types of UniPro discounts: (1) product-spe-cific, volume-based rebates granted to UniProbased on the purchases of all UniPro members; (2)annual UniPro rebates to support marketingefforts; and (3) the UniPro tier growth program.

Magnitude of Price DiscriminationHaving surveyed and briefly described Michael

Foods�’ general pricing arrangements, the courtnow turns to the specific discounts received bySodexho and Feesers respectively, and the com-petitive impact of these discounts. To establish themagnitude of the price discrimination at issue inthis case, Feesers presented expert testimony andother evidence from economist Dr. Robert Larner.Defendants did not present any rebuttal experttestimony, choosing instead to rely on its crossexamination of Dr. Larner and testimony byMichael Foods CFO Mark Westphal. After hearingDr. Larner�’s testimony and reviewing the evi-dence, the court finds Dr. Larner�’s methodologyappropriate and reliable, and his findings wellsupported by the evidence. The court will nowbriefly describe Dr. Larner�’s study and his con-clusions, and address Defendants�’ criticisms ofthat study.

Dr. Larner examined sales data from MichaelFoods, Sysco (Sodexho�’s primary distributor for

70a


much of the time in question), Sodexho, andFeesers from 2000 until 2004. Using that data, hecompared the Sodexho deviated price (referred toby Dr. Larner as the �“Sodexho delivered price�”)6

for Michael Foods products with that paid byFeesers. His findings are presented in a series ofcharts attached to his June 13, 2005 expert report(hereinafter referred to as the �“Larner Report�”),and demonstrative exhibits, which were all admit-ted into evidence at trial.

First, Dr. Larner determined the baseline pricessecured by Sodexho and Feesers respectively forthe eleven Michael Foods egg and potato productsmost commonly purchased by Feesers. Collectivelythese nine egg and two potato products constituteover eighty percent of Michael Foods�’ total sales toFeesers. Dr. Larner then determined the Sodexhodeviated price and the price paid by Feesers forthe sale of Michael Foods products. To establishthe Sodexho deviated price for each item, Larnerstarted by examining the supply agreementsbetween Sodexho and Michael Foods. For Sodexho,Larner found that the prices paid were the same

71a


6 The Sodexho deviated price is the price bargained forby Sodexho and received by Sodexho�’s approved distributor(Sysco) for Michael Foods products that are resold toSodexho. As discussed above, in this deviated billbacktransaction, Sysco initially pays the Michael Foods list pricefor these products, then sells them to Sodexho for the devi-ated price plus a separately negotiated distribution markup.Sysco then bills back Michael Foods for the differencebetween the list price and the Sodexho deviated price. Bycontrast, Feesers pays the distributor list price. Defendantsdispute whether the Sodexho delivered price is the appro-priate price for comparison. This argument will be discussedbelow.

as described in the supply agreements. BecauseFeesers has no supply agreements with MichaelFoods, Larner looked at Michael Foods�’ list pricesto determine the price paid by Feesers for thesame products. As a member of the distributorGPO UniPro,7 Feesers received the prices set forthon Michael Food�’s distributor buying group pricelist. Finally, Larner examined the transactionprices at which Michael Foods sold products toFeesers and Sodexho. For Sodexho, Larner foundthis information in Sysco�’s database of deviatedsales data. For Feesers, this information was con-tained in Feeser�’s accounts payable database.Larner found that the prices reflected in thesethree sources of data were consistent, and accord-ingly in his analysis he compared the pricesreflected in Sodexho�’s contract pricing documentswith those in Feeser�’s accounts payable database.

Next Dr. Larner compared the Sodexho deviatedprice to the price paid by Feesers to MichaelFoods. Dr. Larner�’s findings are presented in aseries of tables attached as exhibits to his report,and in demonstrative exhibits presented at trial.In these tables, for each of the eleven top-sellingMichael Foods products Larner compares the aver-age monthly purchase price paid by Feesers withthe Sodexho delivered price for the same product.These charts account for product-specific deviated

72a


7 Defendants introduced evidence of Feesers�’ receipt ofdiscounts from Michael Foods for two purposes: (1) todemonstrate that the discrimination in favor of Sodexho wasnot substantial, and (2) to show that Feesers should notreceive the injunctive relief it seeks because Feesers has�“unclean hands�” due to its receipt of customer-specific devi-ated pricing and UniPro discounts. The former argument isaddressed here, and the latter is discussed below.

prices only, and do not include the non-product-specific preferred supplier and volume allowances,which are presented separately in Exhibit D-13.The prices listed are the average prices paid byboth Feesers and Sodexho weighted by volume foreach month. The tables present the price differ-ence both in dollar and percentage terms.

The price difference Dr. Larner found is stun-ning. For example, in Exhibit D-1, Larner com-pared the pricing for table ready eggs (MF 15222),which was Michael Foods�’ top selling product toFeesers, accounting for 28.4% of Michael Foods�’total sales to Feesers from 2000 until 2004. Theaverage weighted monthly average price paid byFeesers for this product was $12.04 higher thanthe price received by Sodexho, and on average,Feesers paid 67.8% more for this product thanSodexho paid for the same product at the sametime. Dr. Larner found similar differences for thenext ten top selling Michael Foods products. (SeeLarner Expert Report, Exs. D-2-D-11.) Altogether,Dr. Larner found that on average from 2000 until2004,8 Feesers paid $9.56, or 59.% more thanSodexho for the eleven top selling Michael Foodsproducts taken together. (Id. Ex. D-12.)

Moreover, this price disparity does not accountfor the non-product-specific discounts and allow-ances granted by Michael Foods, which Dr. Larnersummarizes in Exhibit D-13 of his report. Thesediscounts fall within three categories: marketingallowances, a preferred supplier allowance, and a

73a


8 It should be noted that Michael Foods did not supplypotatoes to Sodexho until 2002. Accordingly, in making thiscalculation, Dr. Larner used the weighted monthly averageprices for potato products from 2002 until 2004 only.

volume growth incentive allowance. The chartreflects that both Feesers and Sodexho receivedsome modest discounts for marketing allowancesduring the years in question. Feesers received aslightly higher discount than Sodexho for mar-keting allowances, at $.04/lb with a minimum pur-chase requirement, as compared with Sodexho�’s$.10/case ($.005/lb for a 15 lb case). However, themarketing allowance is not a pure volume dis-count because it requires the recipient to expendsome funds and effort in marketing the product inorder to be eligible.

On the other hand, Sodexho alone received thepreferred supplier and volume growth incentiveallowances. These discounts were in exchange fora minimum volume commitment of approximately76 million pounds of Michael Foods egg productsannually, and for designating Michael Foods asthe preferred supplier of eggs and potatoes forSodexho. (See 2002 Egg Contract between Sodexhoand Michael Foods, P7 ¶ 3.) For the preferred sup-plier allowance for eggs, Sodexho received lumpsum payments from Michael Foods totaling $2.2million for the 1999 and 2002 egg contracts, aswell as a retroactive allowance of $137,493 in2002. (See Larner Report Ex. D-13; 2002 Egg Supply Contract between Sodexho and MichaelFoods, P7.) For the 2002 potato contract, Sodexhoreceived a total payment of $75,000 paid in lumpsums of $25,000 a year from 2002 through 2004.(Id.) For the volume growth incentive allowance,Sodexho received a discount per pound for volumegrowth in sales over and above those of the pre-vious year. (Id.) For example, if Sodexho sold tenpercent more egg products than the previous year,

74a


it would receive a payment of $.01/lb for thosesales over and above the baseline. These twoallowances were paid directly to Sodexho, ratherthan its distributor Sysco, and Feesers received nosuch discounts from Michael Foods.

Although not included the weighted monthlyaverage prices, Dr. Larner also calculated Feesers�’product-specific UniPro rebates and Sodexho�’sproduct-specific allowances in comparing theprices for Michael Foods�’ egg products. (See id.Exs. D-1-D-9, columns 5 and 6.) The UniProrebates and allowances received by Feesers aredwarfed by Sodexho�’s price advantage for theseproducts. For instance, in the sale of MichaelFoods�’ table ready eggs, Feesers received a UniProrebate of $1.20/case from 2000 until October 2002,when the discount increased to $1.28/case. (Ex. D-1, columns 5 and 6.) On the other hand, Sodexhoreceived a product-specific allowance for the sameproduct in the amount of $.90/case from 2000 until2002, when the allowance increased to $1.20/case.However, although Feesers received a slightlylarger product-specific rebate than Sodexho forthis product (at most $.30/case more thanSodexho), the Sodexho delivered price for thesame product ranged from $ 8.70/case to$12.15/case less than the price paid by Feesersduring the same time period. (See id. Exs. D-2-D-9, columns 5 and 6.)

It must be noted that Dr. Larner�’s calculationsdo not take into account the deviated pricing thatFeesers receives in certain transactions. Feeserspurchases some Michael Foods products at devi-ated prices as distributor for certain institutionsand food service management companies. For

75a


instance, Feesers received deviated pricing asprime distributor for the Wood Company, andlater briefly for Sodexho after Sodexho acquiredWood. However, Feesers�’ receipt of deviated pric-ing restricted to a particular food service man-agement company cannot be used by Feesers tocompete with that company for customers. Accord-ingly, it would be inappropriate to consider theserestricted deviated prices when determiningwhether Sodexho�’s discount is substantial enoughto cause competitive injury to Feesers.

The Sodexho deviated pricing is passed directlyto customers, while the preferred supplier and vol-ume growth incentive allowances are retained bySodexho at the corporate level. However, there isevidence that Sodexho passes these benefits tocustomers indirectly, in the form of financial guar-antees, interest free renovation loans, or for somecustomers, cash signing bonuses. For instance, thefollowing language appears in a Sodexho proposalto Lehigh University:

While our purchasing practices assure allunits receive products and services atcompetitive pricing, Sodexho also negoti-ates corporate discounts and rebates thatare realized at the corporate level. Promptpayment discounts and other rebates orallowances obtained from vendors sup-plies, or distribution companies, includingthose obtained through our national andregional purchasing arrangements basedon Sodexho�’s total purchases, are retaineda the corporate level. For example, meet-ing volume commitments on a nationalbasis may result in a negotiated rebate for

76a


achieving this volume level, and thoserebates are paid to Sodexho on a nationalbasis, and are not realized at the accountlevel. Maximizing these opportunitiesallows Sodexho to offset corporate over-head expense and ultimately affords theopportunity to offer our customers, such asLehigh University, financially viable andcompetitive agreements.

(Sodexho Proposal to Lehigh University, P95 at62-63 (emphasis added).)

Defendants�’ Criticism of Dr. LarnerDefendants challenge Dr. Larner�’s calculations

on a number of grounds. Defendants chose not topresent any expert testimony in rebuttal by theirown expert economist. Instead, Defendantsattacked Dr. Larner�’s analysis on cross examina-tion and presented testimony by Michael FoodsCFO Mark Westphal concerning Feesers�’ use ofdeviated pricing for certain customers. These crit-icisms and evidence will be addressed in turn.

Defendants argue that Dr. Larner should havecompared the Sodexho deviated price plus Sysco�’smarkup with the price paid by Feesers ratherthan the Sodexho deviated price alone, which ispaid by Sysco. This argument fails for two rea-sons. First, Dr. Larner argues convincingly thatprice discrimination should be compared at thelevel at which it occurs, which in this case is atthe initial transaction between Michael Foods andSysco for the sale of goods at the Sodexho deviatedprice. While Sodexho negotiates directly withMichael Foods for the sale of food, Michael Foodsdoes not sell its products directly to Sodexho, but

77a


rather sells them to Sodexho�’s designated dis-tributor, usually Sysco. However, this additionallink in the supply chain is more form than func-tion. After all, in its contracts with customersSodexho assumes contractual responsibility forthe procurement and distribution of food, and itnegotiates directly with the manufacturer for alower price on that food. In this case, althoughSysco performs the role of a distributor�—a rolethat was formerly performed by Feesers under theWood contract�—it does not perform the same func-tion as a distributor serving a self-op. In servingSodexho, Sysco does not negotiate a price for thesale of food to Sodexho, or to Sodexho�’s customer.In any event, the deviated pricing is passed ondirectly to customers, who are invoiced by Sodexhofor the cost of food. The cost of food consists of theSodexho-delivered price plus a markup for distri-bution. The distribution markup is separatelynegotiated by Sodexho.

Second, Dr. Larner did an additional analysiscomparing the price paid by Feesers with theSodexho price plus the Sysco markup. This com-parison appears in the E-series of tables in Dr.Larner�’s expert report. While this comparisonslightly reduces the price disparity, it does noteliminate it. For example, when Sysco�’s distribu-tion markup is considered, the price disparity inthe sale of Michael Foods table-ready eggs isreduced from an average of 67.8% (Exhibit D-1) to44.6% (Exhibit E-1).9

78a


9 The court further notes that this is an inappropriatecomparison because it does not take into account the dis-tribution costs to Feesers associated with delivering thegoods to its customers.

The court also rejects Defendants�’ argumentthat Dr. Larner�’s calculations fail to account forthe allowances received by UniPro. This is incor-rect. Dr. Larner�’s analysis clearly compared theUniPro rebates with the allowances secured bySodexho in his calculations in Exhibits D-1through D-9. Furthermore, these modest discountsscarcely compare with the massive disparitybetween the Sodexho deviated price and the unre-stricted national distributor list price paid byFeesers.

Defendants also argue that Feesers was the ben-eficiary of deviated prices for specific customers,and that Dr. Larner inappropriately failed toinclude these deviated prices in his calculations.Mark Westphal testified that from 2000 until2003, 77% of Feesers purchases were made atdeviated prices. Michael Foods also submitted con-tracts for 44 separate customers who receiveddeviated pricing and designated Feesers as thesole distributor eligible to receive such deviatedpricing. However, each and every one of theseprice deviations was customer specific, and couldnot be used by Feesers to win or retain anaccount.

Moreover, even if the Sodexho deviated priceshad been compared with the deviated pricessecured by self-op institutions and other food ser-vice management companies served by Feesers,this would not alter the court�’s conclusion. Oncross examination, Westphal admitted that noneof these restricted deviated prices were nearly asgreat as the Sodexho deviated price. For instance,Westphal testified that the Swarthmore deviatedprice for Michael Foods egg products was $.725/lb

79a


as compared with Sodexho�’s deviated price of$.58/lb, a significant price disparity. When askedby Feesers�’ counsel, Westphal was unable to recallany institution that received a deviated pricewithin $.10/lb of the Sodexho deviated price.There is a good explanation for this. In 2002,Sodexho secured a �“most favored nation�” clause inits egg contract, guaranteeing that no one wouldget a lower price than Sodexho for Michael Foodsegg products. (2002 Egg Supply Contract betweenSodexho and Michael Foods, P7 ¶ 5(b).)

Finally, the court rejects Defendants�’ argumentthat Sodexho-level prices were made available toFeesers. Before commencing this litigation,Feesers requested Sodexho-level deviated pricingfrom Michael Foods. Michael Foods agreed to thelower price only where Feesers demonstrated thatit was seeking to win the account of a currentSodexho customer. Michael Foods�’ offer of Sodexhopricing to Feesers does not mean that the Sodexhodiscounts were practically available to Feesers inorder to compete with Sodexho. Michael Foodsoffered to provide Sodexho level pricing to Feesersonly upon proof that a customer was a currentSodexho customer. The problem is that MichaelFoods�’ offer only addressed one part of competi-tion�—Feesers�’ attempts to win over Sodexhoaccounts. It does not address Sodexho�’s attemptsto convert Feesers�’ self-op customers to Sodexhomanagement. Feesers has no way of knowingwhen Sodexho is trying to lure its customersaway�—as noted above, competition in this indus-try is not narrowly confined to the formal RFPprocess, and Sodexho may spend years cultivatingpotential customers before convincing them to con-

80a


sider making a switch. Accordingly, MichaelFoods�’ limited offer of deviated prices for Feesersto compete with Sodexho customers did not makeSodexho�’s deviated prices available to Feesers asa practical matter.

Substantiality of Price DiscriminationHaving determined the level of price discrimi-

nation, the court must decide whether the pricedisparity was substantial. Whether a price dis-crimination is substantial depends on the partic-ular industry and customers. The more pricesensitive the industry, the more likely it is thateven a small difference in price is substantialenough to cause competitive injury. Here, the evi-dence demonstrates that the food service industryis extremely price sensitive and populated byincreasingly sophisticated and budget-consciousinstitutional customers. Three categories of evi-dence are particularly relevant: Sodexho�’s strate-gic planning and marketing documents, customertestimony, and expert opinion testimony by Dr.Larner.

Sodexho�’s strategic planning documents empha-size the importance of price in winning over newcontracts. For example, in Phase I: FY 03-05Health Care Services Strategic Plan, Sodexho pre-dicted that �“[o]ur client will seek a series of prod-ucts and services which produce cost savings atacceptable quality levels . . . We will have to orga-nize and allocate resources to the development ofmoney saving products and services. . . Clientsmay demand to see pricing on a line item basis forease of price/vendor comparison.�” (P1 82 at 5.)

81a


Jay Marvin testified that GPOs typically serveself-op clients, and had a strong participationamong acute care hospitals. As noted above, GPOsare an alternative to the use of a distributor forself-op institutions. Because of GPO pressure,Sodexho was increasingly concerned about marketbaskets and other line by line price comparisonsrequested by customers. (Id. at 16.) The healthcare strategic plan noted the following GPOs / E-commerce industry trends:

GPOs /E-Commerce: The growing numberand complexity of GPOs combined withthe emergence of e-commerce businessapplications in the procurement arenahave, in effect, equalized the playing fieldenabling all entities to gain procurementleverage. Even a non-GPO aligned stand-alone facility can aggregate its buy withother facilities through E-commerce. Ourhistorical pricing advantage is dramati-cally minimized. Additionally, GPOsincreasingly dictate many or all aspects ofthe procurement process including prod-uct selection, distribution, etc.

(P182 at 16.) The implications of this trend forSodexho Senior Services are that: �“[h]istorical andclear point of differentiation will not be there forus in the future�” �“[i]ncreasing amount of timedefending and explaining (market baskets, meet-ings, etc.) our prices�” �“[w]e are on the defensiveand our credibility suffers�” �“Competitors may gainaccess to our accounts through �‘back door�’ pur-chasing relationships.�” (P182 at 16.) Likewise, in

82a


the Health Care Services Hospitals FY05-07Strategic Plan, Phase II, Sodexho observed that�“[t]he need to continually demonstrate incremen-tal value will intensify.�” (P160 at 6.) Sodexho con-cluded that �“[o]ur primary response should be toa) re-evaluate our pricing strategy so we are com-petitive at the �‘loading dock�’, . . . c) explore way toexpand Entegra�’s role and presence in the health-care industry.�” (Id. at 10.)

One of Sodexho�’s �“Key Initiatives�” was a �“Pur-chasing Improvement Plan�” with the objective �“tomaximize Gross Profit Margin and to maximizeour GPO & Strategic Business partnerships whileattempting to address the growing issues sur-rounding our uncompetitive market basket resultsin the field.�” (P160 at 23.) This initiative includedthe creation of a �“Competitiveness Task Force toaddress food cost competitiveness in the field,addressing issues of maintaining the balancebetween food cost and VDA [volume discountallowance] income. . .�” (P160 at 23.) In the HealthCare Division Three Year Plan for 2001-2003,Sodexho declared that [t]he Health Care Divisionwill focus on the following major initiatives for theAcute Care segment: Develop a low-cost food offer-ing designed to reduce costs by a minimum of10%.�” (P1 89 at 46.)

The significance of food costs is also demon-strated by the participation of some customers inthe U.S. Department of Agriculture�’s Net OffInvoice (�“NOI�”) commodity program. The NOI pro-gram allows school districts to obtain certain pro-cessed foods at a discount subsidized by thegovernment. However, the NOI program can onlybe utilized by authorized distributors, see 7 C.F.R.

83a


§§ 250.12, 250.3, and Sodexho�’s primary distrib-utor Sysco was ineligible. Accordingly, someSodexho customers have chosen to utilize Feesersfor NOI purchases. This evidence further supportsthe conclusion that schools are extremely sensitiveto small price differences.

Finally, at trial Dr. Larner offered his expertopinion that the price discrimination in favor ofSodexho was substantial enough to cause com-petitive injury to Feesers. Dr. Larner testifiedthat the food service industry is characterized byintense competition and tight profit margins, andhe concluded that the price discrimination in favorof Sodexho that he found in his expert report wassubstantial enough to cause competitive injury toFeesers.

Defendants dispute Dr. Larner�’s conclusion thatthe price differences noted above were substantial.Defendants argue that Dr. Larner should haveperformed a quantitative analysis to determinethe price sensitivity of customers in the food ser-vice industry, isolating price from other factorscustomers consider. On cross examination, Dr.Larner conceded that such an analysis could beperformed. The court is satisfied that such a studyis not necessary. Defendants further argue thatLarner failed to separate the effects of deviatedpricing for Michael Foods products from the devi-ated pricing Sodexho receives from other manu-facturers. However, it is well settled that wherethe price discrimination at issue affects only asmall number of articles sold, the Robinson-Pat-man Act still applies. See F.T.C. v. Morton Salt,334 U.S. 37, 49, 68 S. Ct. 822, 92 L. Ed. 1196, 44F.T.C. 1499 (1948). The court credits Dr. Larner�’s

84a


expert opinion that the price discrimination infavor of Sodexho was so substantial and sustainedthat Feesers is entitled to an inference of com-petitive injury.

In sum, the court finds that Michael Foods hasengaged in substantial price discrimination infavor of Sodexho. Sodexho has received significantand long-term price discounts from Michael Foods,including the lowest deviated prices offered byMichael Foods. Although Feesers also receivedcertain discounts by virtue of its membership inUniPro, these discounts are dwarfed by thosegranted to Sodexho. Additionally, Sodexho alonereceived large signing bonuses and volume baseddiscounts. Though these discounts were notpassed on to customers directly, they were used toprovide other benefits to customers, a selling pointto win and retain customers.

3. Conclusions of LawBased on the evidence presented at trial, the

court concludes that Feesers has established by apreponderance of the evidence that Michael Foodsdiscriminated in price in Sodexho�’s favor, and thatthis price discrimination was significant enoughin both magnitude and duration to cause compet-itive injury to Feesers.

Normally the quantum of proof necessary toprove substantial price discrimination depends onthe level at which the discrimination occurred.Accordingly, tertiary discrimination (injury to thecustomers of the disfavored purchaser) generallyrequires a greater magnitude of price difference toinjure competition than secondary discrimination(injury to the customers of the disfavored pur-

85a


chaser). This is because the effects of secondaryprice discrimination are generally less attenuated,and it is more likely the case that the discountwill be passed on to the benefit of the favored pur-chaser�’s customer. However, there are limits tothe usefulness of the analytical distinctionsamong the various levels of injury resulting fromprice discrimination. For instance, in cases wherean institution plays a dual role, such as a com-pany functioning as a dual distributor andretailer, courts have declined to label such dis-crimination as falling at a particular level. See,e.g., Texaco v. Hasbrouck, 496 U.S. 543, 110 S. Ct.2535, 110 L. Ed. 2d 492 (1990). The Robinson-Pat-man Act is concerned with the functions providedby a company, not the label that company choosesto apply.

The Third Circuit has noted that this is one ofthose cases that cannot be easily categorized aseither secondary or tertiary discrimination, butrather falls somewhere between the two. SeeFeesers, Inc. v. Michael Foods, 498 F.3d 206, 211n.5 (3d Cir. 2007). This is because Sodexho obtainsfood through its distributor Sysco, though itdirectly negotiates the price of that food from themanufacturer, Michael Foods. On the other hand,Feesers obtains food directly from manufacturersand sells it directly to institutional customers.Accordingly, the Sodexho transaction introducesan additional link in the supply chain.

Here, Sodexho offers two distinct functions to itscustomers: management (performed by Sodexhoitself) and distribution and procurement (forwhich Sodexho subcontracts with distributors). Inthe context of functional discounts, the Supreme

86a


Court has observed that �“[m]anufacturers will bemore likely to effectuate tertiary line price dis-crimination through functional discounts to a sec-ondary line buyer when the favored distributor isvertically integrated.�” Hasbrouck, 496 U.S. at 565-66. A loose analogy may be drawn between therole played by Sodexho in the distribution chain,and that occupied by the favored dual-functionwholesaler/retailers in Hasbrouck. Nevertheless,it is clear that the discounts secured by Sodexhofrom Michael Foods are not retained by Sodexho,but instead pass directly to Sodexho�’s customersat a level that is likely to cause those customers tochoose Sodexho. The evidence also establishesthat as a general matter, this is a price sensitiveindustry populated by increasingly sophisticatedcustomers. Moreover, the discounts secured bySodexho were not temporary, but rather a long-term arrangement that is precisely the type ofprice discrimination most likely to harm compe-tition. Accordingly, the court finds that the mag-nitude of price discrimination in this case issubstantial and sustained enough that customersmay be persuaded to switch from self-operation tofood service management in order to obtain dis-counts on food products and thereby lower theiroverall costs of food service operation.

Defendants�’ argument that there is no compet-itive injury because Michael Foods products con-stitute a small percentage of any one customer�’spurchases is foreclosed by the Supreme Court�’sdecision in F.T.C. v. Morton Salt. In that case, theSupreme Court rejected a similar claim by largegrocery store chains that price discrimination in

87a


the sale of table salt did not constitute competitiveinjury:

There are many articles in a grocery storethat, considered separately, are compar-atively small parts of a merchant�’s stock.Congress intended to protect a merchantfrom competitive injury attributable todiscriminatory prices on any or all goodssold in interstate commerce, whether theparticular goods constituted a major orminor portion of his stock. Since a grocerystore consists of many comparativelysmall articles, there is no possible wayeffectively to protect a grocer from dis-criminatory prices except by applying theprohibitions of the Act to each individualarticle in the store.

Id. at 49. Likewise here, egg and potato productsconstitute a small portion of any individual cus-tomer�’s food purchases from Feesers and Sodexho.

C. Competitive InjuryIn sum, the court concludes that Feesers has

established a prima facie case of price discrimi-nation. Feesers is in competition with Sodexho forthe sale of Michael Foods products to institutionalcustomers, and Michael Foods engaged in sub-stantial and sustained price discrimination infavor of Sodexho. Under the facts of this case,Feesers is entitled to the Morton Salt inference ofcompetitive injury.

88a


III. Rebuttal of Inference of PriceDiscrimination

A. Legal StandardOnce a plaintiff has established a prima facie

case of price discrimination, the burden shifts tothe defendant to rebut the inference of competitiveinjury. Falls City Indus., Inc. v. Vanco Beverage,Inc., 460 U.S. 428, 435, 103 S. Ct. 1282, 75 L. Ed.2d 174 (1983), citing F. Rowe, Price Discrimina-tion Under the Robinson-Patman Act 182 (1962).Here, in order to rebut the inference, MichaelFoods must show an absence of a causal linkbetween discrimination and lost sales or profits.Feesers, Inc., 498 F.3d at 216; see also In re BoiseCascade, 113 F.T.C. 956 (1990).

B. Findings of FactDefendants claim that the evidence presented at

trial demonstrates that the lower price Sodexhoreceives plays no role in a customer�’s choicebetween food service management or self-op ser-viced by a distributor such as Feesers, but insteadthat customers were motivated by other factorssuch as services. The court will examine the evi-dence presented at trial regarding the significanceof food costs as opposed to other factors in a cus-tomer�’s decision to switch from self-op to food ser-vice management or vice versa. In this inquiry,the court will review three categories of evidence:testimony from customer witnesses, Sodexho�’sstrategic planning documents, and Sodexho�’s mar-keting materials.

89a


Food service budget for institutional customers

Whether self-op or managed, an institution�’sfood service budget primarily consists of two fac-tors: raw food and labor, with overhead andadministrative costs make up the balance of a foodservice budget. The evidence presented at trialdemonstrates that the price of raw food may rangeanywhere from 20 to 50 percent of a facility�’s foodservice budget. These numbers are not static, butmay vary depending upon the efficiency of the foodpreparation, menus, and participation in a diningprogram. In other words, if food costs are higher,then food constitutes a higher percentage of thefood service budget, and the cost of the entire foodservice budget might be higher. Moreover, the foodservice industry is extremely price sensitive. Thisfinding is supported by customer testimony aboutthe importance of the bottom line, as well as Dr.Larner�’s testimony.

The price of raw food is more significant to cer-tain Sodexho customers due to the payment struc-ture set forth in their contracts with Sodexho.Sodexho has two different types of food servicemanagement contracts: a profit and loss (�“P&L�”)contract, and a management fee contract. In aP&L contract, which is usually offered only tolarger customers, Sodexho offers a financial guar-antee that the dining services will not lose money,and the institution shares in a certain percentageof the profits. On the other hand, in a manage-ment fee contract, Sodexho receives a set fee forits management services, and it bills the institu-tion for other costs including food. An institutionwith a management fee contract is more likely to

90a


be sensitive to food costs because it is directlyinvoiced for those costs. On the other hand, cus-tomers with a P&L contract also benefit fromlower food costs because that increases the prof-itability of the dining services.

Customer TestimonyMichael Foods relies primarily on testimony

from ten customer witnesses in an effort to rebutthe inference of price discrimination. According toDefendants, the testimony from these witnessesdemonstrates that food costs are not a significantfactor to customers. The court will first brieflyreview the testimony, and then explain why itfinds this evidence unpersuasive.

Defendants offered testimony from customerwitnesses responsible for food service operationsat three schools. Robert Bruchak is in charge ofdining services for Daniel Boone Area School Dis-trict. Daniel Boone was originally self-op butswitched to Sodexho in 2004. The change wasprompted by the sudden resignation of the school�’sfood service director. Daniel Boone solicited pro-posals in a formal RFP process, and ultimatelychose Sodexho because Sodexho identified poten-tial cost savings in labor management and pro-vided a financial guarantee of profitability.Throughout this process Bruckak did not performa line by line comparison of the cost of food,though a school board member once requestedinvoices. For a time, Daniel Boone continued toutilize Feesers for NOI purchasing, with savingsof about $60,000 annually. According to Bruchak,Sodexho told Daniel Boone that it offered lower

91a


prices for food in its proposal, but this was not thedeciding factor for Daniel Boone.

Stanley Majewski is responsible for food serviceoperations at Bethlehem Area School District. Theschool district was self-op until 1999, when itswitched to management under Wood and laterSodexho. At the time of the switch, the institutionconsidered both self-op and food service manage-ment. The district�’s primary consideration wasservice, not cost, and Sodexho won the contractbecause it had more management experience thanother bidders in the RFP process. Nevertheless,food costs constitute about forty percent of theinstitution�’s dining services budget paid toSodexho, and that portion of the budget previouslywent to Feesers. There is also evidence that costswere important to the school district: to win thecontract, Sodexho promised a large investment inthe kitchen, and the district participates in theNOI program to save money on food costs.

David Matyas oversees food service at CentralBucks School District. The institution was previ-ously self-op and switched to Sodexho manage-ment. The change was motivated primarily byfinancial considerations, particularly the financialguarantee of profitability offered by Sodexho.According to Matyas, food costs by themselves arenot an important concern for the school becausethe financial guarantee puts the pressure onSodexho to ensure that revenue equals expenses.Matyas believes that the financial guarantee isunrelated to Sodexho�’s discounts and reimburse-ments for food costs, but offered no explanation forthis belief.

92a


Defendants also offered testimony from two col-lege and university customers. Dr. George Harp-ster is responsible for overseeing the food serviceprogram for Shippensburg University. The uni-versity has utilized food service management foras long as the witness has been responsible fordining services, at least since 1987. To choose afood service management company, the universitysubmits requests for proposals to managementcompanies. Shippensburg chose to give greaterweight to qualitative factors than quantitative factors in the RFP process. According to Dr. Harp-ster, Shippensburg University has never consid-ered self-op, and Dr. Harpster has no conception ofhow the institution could switch to self-op, orwhat factors he would consider in making thedecision. Dr. Harpster testified that he was underthe impression that in the instant litigationFeesers was seeking the opportunity to bid forfood service contracts during the RFP process.

Wayne Clickner is a food service consultant forthe fourteen public universities that are membersof the Pennsylvania State System of Higher Edu-cation (PASSHE). Clickner has been involved inthe RFP process for dining services at thePASSHE member schools on at least thirteenoccasions. One such RFP process was SlipperyRock University, which switched from self-op tofood service management with Sodexho. Accordingto Clickner, Slippery Rock�’s decision to switchfrom self-op to management was motivated pri-marily by the university�’s desire to avoid workingwith unionized employees, rather than the cost offood. Clickner testified that in his time evaluatingproposals, he had never observed a university

93a


choose to contract with a management companybased solely on the cost of individual food items.However, total meal cost is an important consid-eration, and meal cost is determined in part bythe cost of raw food and Sodexho�’s financialinvestment in the dining services program. Click-ner also testified that all of the PASSHE schoolshave P&L contracts, rather than management feecontracts.

Defendants offered testimony from three retire-ment community customers. Michael Jacobs isresponsible for food service at Deer MeadowsRetirement Community. Deer Meadows Retire-ment Community is managed by Sodexho. Jacobstestified that he has never considered self-op inthe past and would never consider it in the future.Moreover, Jacobs testified that he had no ideawhat factors he would need to consider in decidingto make such a switch. The institution has a man-agement fee contract with Sodexho, so it reim-burses Sodexho directly for the cost of food.Although Jacobs has never compared the cost offood in the past, he testified that if food costsincreased substantially, he would investigate.

Michael Peck is responsible for dining servicesat York County Pleasant Acres RehabilitationCenter, a long term care facility. The facility iscurrently managed by Sodexho. In 2003 the facil-ity solicited proposals by other food service man-agement companies. Because the problems wererelated to service, not costs, financial considera-tions were given the lowest priority when the pro-posals were evaluated. Peck testified that hebelieved food costs were more fixed than laborcosts, but on cross examination acknowledged that

94a


if the facility were to switch to self-op, he wouldutilize a GPO or seek out other purchasing optionsto save money on food.

Seth Levy is responsible for overseeing food ser-vices for the Jewish Home of Greater Harrisburg.The facility was self-op until 2003, when itswitched to Sodexho management. The decision tooutsource was made after the food managementteam resigned, and Levy was unable to recruitexperienced replacements. The institution choseSodexho because it offered a lower overall costthan other management companies. Sodexho�’s pro-posed budget was also significantly lower than theinstitution�’s self-op food budget. Although costwas important, at the time the decision to out-source was made, Levy did not compare the cost ofspecific food items on a line by line basis. Levytestified that he understood that Sodexho�’spromise to drive down food costs in its proposalmeant that Sodexho could get a better price thanits competitors because of the volume of food itpurchased.

Defendants also presented testimony from twohospital customers. Joseph Gagliardo is respon-sible for food services at Lewistown Hospital. Thehospital was self-op until 2007, when it switchedto Sodexho management. The decision to switchwas motivated primarily because of service issues,rather than cost, and the institution performed noline-by-line comparison of the cost of food. Lewis-town Hospital has a management fee contractwith an investment for renovation.

Philip Guarnaschelli is responsible for over-seeing Pinnacle Health�’s food service operation.Pinnacle was self-op fifteen years ago but

95a


switched to Wood (later Sodexho) in 2001. Whenthe hospital made the switch, it was concernedabout labor costs and management expertise,rather than food costs. Guarnaschelli testifiedthat at this point, Pinnacle is committed to man-agement and would not consider returning to self-op. However, on cross examination he testifiedthat if he were to consider switching, he wouldcompare the total cost of management to the totalcost of self-op. Guarnaschelli also testified that hewould consider a savings of 25% on the cost of foodsignificant.

Finally, Defendants presented testimony fromClyde Harris, who oversees food service operationsat Air Products and Chemicals. Harris testifiedthat the company has outsourced its dining ser-vices for at least ten years, and has no intention ofever switching to self-op. The company chooses tooutsource because it does not have the skill ordesire to self-op, and Harris testified that he doesnot know what factors the company would need toconsider to make the decision to self-op.

According to Defendants, the customer testi-mony establishes that price is not an importantconsideration for institutional food service cus-tomers. Instead, Defendants claim that laborissues, management expertise, and other factorsare more important to customers than price. How-ever, the court is not satisfied that such a broadinference can be drawn from the experience andperception of these witnesses. Most of the wit-nesses were satisfied Sodexho customers. Sodexhopromised them a low price for food, and deliveredon that promise, in part by securing a lower costfor Michael Foods products than its competitors.

96a


Though none of the witnesses could recall theprice of Michael Foods products in particular, itwas apparent to the court that the witnesses hadnot felt the need to verify Sodexho�’s prices. How-ever, this does not mean that none of the cus-tomers for whom Feesers and Sodexho compete isconcerned about the cost of food. This apparentlack of concern about pricing may be due to thefact that many of the witnesses had P&L con-tracts, rather than management fee contracts.This means that the witnesses were more con-cerned with the bottom line than with the com-ponent prices of the services offered by Sodexho.

Defendants point out that the Third Circuitfound customer testimony pertinent to the issue ofcompetitive injury its opinion reversing thiscourt�’s grant of summary judgment in favor ofDefendants. See Feesers, Inc., 498 F.3d at 214-15.Indeed, if Feesers had produced customer wit-nesses who testified that the cost of food was thereason they switched from Feesers to Sodexho,that evidence would have been extremely per-suasive on the issue of competitive injury. How-ever, testimony presented by Defendants from afew customers who did not find price significantdoes not have the same weight, particularly whereas here, there is other evidence suggesting thatprice is quite important to other customers in thesame industry. Here, the testimony of Defendants�’customer witnesses cannot be reconciled withother persuasive and undisputed evidence of theimportance of price to other customers�—namelySodexho�’s strategic planning documents describ-ing the importance of price to win and retain cus-tomers, and the proposals and promotional

97a


material submitted by Sodexho to win customers.This evidence demonstrates that the cost of food isa significant part of the food service budgets forthese institutions. The court will briefly review anumber of these documents submitted by Plaintiffat trial.

Sodexho Strategic Planning documentsSodexho�’s strategic planning documents and tes-

timony from Sodexho�’s employees demonstrate theimportance of price for Sodexho to win and retaincustomers. Christopher Rochette, former SodexhoSenior Vice President of Strategic Planning, tes-tified that from 2000 until 2005, Sodexho�’s strate-gic plan was to position itself as the low costprovider in the education and health market. Thisgoal is reflected in Sodexho�’s strategic plans dur-ing that time. For instance, in noting increasedeconomic pressures on hospitals, Sodexho�’s HealthCare Services Strategic Plan Phase I: FY 03-05,predicted the following implications:

Our client will seek a series of productsand services which produce cost savingsat acceptable quality levels . . . We willhave to organize and allocate resources tothe development of money saving productsand services. . . Clients may demand to seepricing on a line item basis for ease ofprice/vendor comparison.

(P182 at 5.) Sodexho�’s strategy to achieve domi-nant market share included �“[e]stablish[ing] com-petitive food pricing so [Sodexho] is competitive atthe loading dock versus GPO pricing structure.�”

98a


(P160 at 31.) Sodexho offered the following sum-mary of competitor findings for self-op in its:

Summary of Competitor Findings, Self-OperatedFinding: �“Increasing perception that com-parable or better purchasing economiescan be obtained through GPOs. Accord-ingly, there is an increasing number offacilities seeking/joining GPOs.�”Implications: �“Economies of scale advan-tage will carry less weight than in thepast, purchasing will come underincreased scrutiny, more GPOs.�”Possible response: �“Quality products andofferings at defined price points. . . Proac-tive market basket analysis, Procurementmarketing.�”

(Competitive Intelligence Overview Findings,Health Care Services: Hospitals Phase I, 2003,P170 at 4.)

Likewise, Sodexho�’s Health Care ServicesStrategic Plan Phase I: FY 03-05 noted a trendtowards greater cost-consciousness among its cus-tomers, fueled in part by a wider variety of pro-curement options, such as GPOS:

GPOs /E-commerce industry trendsGPOs /E-Commerce: The growing numberand complexity of GPOs combined withthe emergence of e-commerce businessapplications in the procurement arenahave, in effect, equalized the playing fieldenabling all entities to gain procurement

99a


leverage. Even a non-GPO aligned stand-alone facility can aggregate its buy withother facilities through E-commerce. Ourhistorical pricing advantage is dramati-cally minimized. Additionally, GPOsincreasingly dictate many or all aspects ofthe procurement process including prod-uct selection, distribution, etc.

(P1 82 at 16.) The implication of this trend forSodexho Senior Services: �“Historical and clearpoint of differentiation will not be there for us inthe future�” �“Increasing amount of time defendingand explaining (market baskets, meetings, etc.)our prices�” �“We are on the defensive and our cred-ibility suffers�” �“Competitors may gain access toour accounts through �‘back door�’ purchasing rela-tionships.�” (P182 at 16.) �“The desire for costreductions is the number one impetus for SchoolDistricts to consider outsourcing.�” (P190 at 15.)

Strategic planning documents for other seg-ments of the institutional food service industryalso emphasize the importance of food costs inwinning and retaining customers. In its StrategicPlan for School Services Division, Sodexho notedthat �“[t]he desire for cost reductions is the numberone impetus for School Districts to consider out-sourcing.�” (P190 at 15.) The plan went on to notethat �“[m]anufacturers and vendors are providingvalue-added services that compete directly withthe support services provided by private man-agement companies,�” (id. at 18), and �“[c]ompeti-tors, like ourselves, have not found the lever toopen up demand for self-op conversion to out-sourcing. Self-Ops use contractors to �‘fix�’ current

100a


problems and then return to self-op or use theRFP process to gain ideas but remain in-house.�”(Id. at 19.)

Likewise, Sodexho�’s number one �“strategicimperative�” for its Senior Services division wasthe development of a �“Low Cost Food Model.�”(P179 at 4.) There is also evidence that lower foodcosts were important to customers in Sodexho�’scorporate services division. (See Sodexho Corpo-rate Services 3 Year Plan Phase I, P396 at 42(describing initiatives to lower food costs to ben-efit customers).

The strategic planning documents also indicatethat Sodexho utilized discounts to increase itsprofit margin. For instance in Sodexho�’s HealthCare Services Strategic Plan Phase I: FY 03-05, akey initiative was a �“gross profit improvementplan�” achieved by lowering food costs and cap-turing additional volume discount allowances frommanufacturers:

Key Initiative # 7:Purchasing Improvement PlanPerformance improvement objectiveTo maximize Gross Profit Margin and tomaximize our GPO & Strategic Businesspartnerships while attempting to addressthe growing issues surrounding ouruncompetitive market basket results inthe field.

(P160 at 23.) This initiative included the creationof a �“Competitiveness Task Force to address foodcost competitiveness in the field, addressingissues of maintaining the balance between food

101a


cost and VDA income. . .�” (Id.) Together, thesestrategic planning documents indicate that lowerfood costs are critical both to win and retain newcustomers and to improve Sodexho�’s profit margin.

Sodexho Promotional and Marketing Materials

Even more persuasive than its strategic plan-ning documents, however, are Sodexho�’s promo-tional and marketing materials. The marketinginformation admitted into evidence demonstratesthat Sodexho emphasized its procurement poweras a selling point when trying to win customers.

References to Sodexho�’s lower prices were ubiq-uitous in Sodexho�’s promotional materials to cus-tomers. For example, in a proposal to CharlestownRetirement Community, Sodexho emphasized thatit could provide customers with lower prices forfood than a self-op could obtain from distributors:

Sodexho�’s Procurement SystemsSodexho Health Care Services�’ clientshave the option of selecting the purchas-ing and distribution services that bestmeet their needs.Experience has shown that when ourclients evaluate procurement and distri-bution, they focus largely on the cost ofgoods. While virtually all of the analysesperformed indicate that Sodexho-procuredfood and supplies are less expensive thana market basket of goods of comparablespecification and quality purchased froma non-Sodexho source, it is important torecognize the many other relevant factors

102a


to consider when evaluating procurementand distribution services.

(P74 at 11.) The proposal goes on to compare theadvantages and disadvantages of food servicemanagement versus self-operation. The first twoadvantages of food management include �“[m]ini-mum food procurement costs through mass purchase�” and �“[f]ood contract management com-panies employ specialized staffs which focus onbidding food items, evaluating costs daily, anddeveloping purveyor relationships to minimizefood costs.�” (P74 at 14.) Conversely, the numberone disadvantage to self-operation is that �“[a] self-operated Food Service can only obtain the purchasing power in food procurement it can nego-tiate. Food procurement will depend largely on theprevailing prices extended by the local market.Group purchasing contracts for food are availablebut are only good for certain items. . .�” (P74 at 16.)

In a proposal to Messiah Village, Sodexho toldcustomers that �“you compete on costs, but you winon quality�” and noted that Sodexho had a �“[s]olidtrack record of consistently driving costs out yearover year.�” (P62 at 4.) This language appeared innumerous other proposals, including Villa TeresaNursing Home (P85 at 65), Jewish Home ofGreater Harrisburg (P79 at 4), and the Madlynand Leonard Abramson Center for Jewish Life (P118 at 3). In a proposal to Southern Ocean CountyHospital, Sodexho described itself as �“the industryleader in procurement�” �“offer[ing] vast purchasingvolumes�” with the outcome of reduced costs to thecustomer. (P142 at 28.) In another proposal sub-mitted to Lancaster Regional Medical Center,

103a


Sodexho boasted that �“[o]ur prices for most itemsrange from 5 to 25% lower than the next bestprice.�” (P89 at 17.)

An institution�’s request for a �“market basket�” isanother indication that an institution is moreprice-sensitive. A market basket is a list of priceson a variety of individual food items requested sothat the institution can perform a line by linecomparison of costs. According to Feesers�’ salesemployees, these are almost always requested byself-op customers seeking bids from distributors.However, some institutions also request marketbaskets from food service management companieswhen considering whether to switch from self-opto management, indicating that the cost of food isa significant factor in that determination. Accord-ing to Sodexho�’s Jay Marvin, market baskets arerequested by about 10% of potential customers inthe health care industry. For example, in a pro-posal to Lehigh University, Sodexho touts the volume based discounts it receives from manu-facturers and suggests that these discounts willresult in a more competitive market basket:

As one of the largest purchasers of food inthe nation, Sodexho uses this leverage toprovide our customers with access to highquality, name brand products at competi-tive pricing. Because of the volume of ourpurchasing, and our ability to providewin-win guarantees to our vendors,Sodexho is a very attractive and in-demand customer for the leading manu-facturers and distributors across thecountry.

104a


(P95 at 62.) According to Sodexho, these volumediscounts would have the following benefits toLehigh University: �“First, it assures that on amarket basket approach each of our units receivescompetitive pricing at the account level. . . Inmany instances, we are able to make purchasesfor, or on behalf of our customers, at substantialsavings.�” (Id.)

To streamline the process of preparing proposalsto customers, Sodexho created certain templates,many of which emphasize that Sodexho is able toobtain lower prices on food than self-ops as aresult of its larger purchasing volume. The courtwill now examine some of those templates in turn.

The �“Focus on Procurement�” template appearsin a number of Sodexho�’s proposals to colleges anduniversities, including Sodexho�’s proposal to Cam-den County College:

Sodexho is part of an international pur-chasing network, one of the largest pri-vate purchasing networks in the nation. Itincludes hotels and restaurants aroundthe world as well as thousands of diningservices partnerships around the country.Because of this volume�—and our ability toprovide win-win guarantees to our ven-dors�—Sodexho is a very attractive and in-demand customer for high quality man-ufacturers across the country. This factprovides a number of benefits for our col-lege and university dining service part-ners.First, it means the most competitiveprices on the widest selection of products.

105a


While on occasion a local overstock orother unique situation can �‘beat�’ us on asingle item, when it comes to all the prod-ucts consumed by a dining service opera-tion day in and day out, the overall priceswe are able to command on your behalfare the most competitive offered any-where.It should be noted that these volumeagreements are not with �‘generics,�’ butwith recognized leaders in every cate-gory. . . In addition, our national volumemeans a lot more �‘extras�’ for our clients,such as third party training materials,supplementary marketing materials andspecial promotions, the opportunity to testnew products, services, custom-createdproducts and equipment as well as moreresponsive service than ever before.

(P94 at 83.) This template has been used bySodexho in proposals to Alvernia College (P81 at59), Howard Community College (P137), St.Mary�’s Seminary and University (P1 30), CheyneyUniversity of Pennsylvania (P1 20), and Hood Col-lege (P1 14).

The �“Tremendous National Buying Power�” tem-plate is also used in Sodexho�’s proposals to colleges and universities, such as BloomsburgUniversity:

Sodexho Marriott Services obtains onlythe highest quality raw food products foruse in the production of all consumableitems. Under the direction of our SeniorVice President for Purchasing, the Pur-

106a


chasing Department with its professionalfood buyers, adheres to the highest speci-fications in the industry.At Sodexho Marriott Services, our pur-chasing program is designed to let ourprofessional buyers and purchasingagents apply their skills full-time to thejob of providing our clients with the finestavailable food products at the lowest pos-sible cost.. . .Sodexho Marriott Services�’ tremendousnational buying power will be utilizedwhenever possible for cost advantages. . . .Sodexho Marriott Services maintains acorporate purchasing department that isadvantageous for us in many ways. Thepurchasing department works with com-panies to obtain the lowest possibleprices. The lowest prices are then lockedin through long term contracts. . .

(P91 at 362.) This language also appeared in aproposal to Shippensburg University (P90 at 353).

Sodexho�’s �“Controlling Costs by Leveraging Pro-curement Power as the Largest Purchaser of Food�”template appears in a Sodexho proposal to theFriends School of Baltimore:

Maximize value. We deliver the highestlevels of quality and service, while stayingwithin the funding levels. We will controlcosts by employing proven financial con-trols, training programs, and operationalefficiencies, and by leveraging our pro-

107a


curement power as the industry�’s largestpurchaser of food.

(P78 at 1.) A few pages later, the template con-tinues:

As the largest private purchaser of food inthe country, Sodexho has partnershipswith some of the largest, most familiarand most beloved brands in the nation.You should expect us to apply our pur-chasing leverage to deliver superior qual-ity products at prices below thoseavailable to individual schools or smallerproviders.

(P78 at 14.) This template appears in proposals toother schools, including the Archdiocese ofPhiladelphia, (P75 at 14), the George School (P119), Waldron Mercy Academy (P117), AbingtonFriends School (P61), Haverford School (P107),Salesianum School (P105), Moravian Academy(P103), and Norfolk State University (P59).

Another template, �“Bringing in the Best at aPrice Point Unavailable to Smaller Providers,�”was used in a proposal to Warren County PublicSchools:

As the largest private purchaser of food inthe country, Sodexho Marriott Serviceshas partnerships with some of the largest,most familiar and most beloved brands inthe nation, bringing in �‘the best�’ at a pricepoint unavailable to individual schools orsmaller providers.

(P150 at 150.) This template also appears in pro-posals to the Blue Ridge School (P88 at 19), North-

108a


ern Burlington County Regional School District(P87 at 61), Piscataway Schools (P147 at 87),Carteret School District (P86 at 107), SpotswoodSchool District (P144 at 114), Penns Valley AreaSchool District (P141 at 67), School District of theChathams (P140 at 79), Queen Anne�’s PublicSchools (P80 at 115), and the Hazleton AreaSchool District (P136 at 70).

Another Sodexho template, �“Deliver pricesbelow those available to smaller providers,�” wasused in a Sodexho proposal to the FreeholdRegional High School District:

As the largest private purchaser of food inthe country, Sodexho has partnershipswith many of America�’s favorite brands. . . You should expect us to applyour purchasing leverage to deliver supe-rior quality products at prices below thoseavailable to individual schools or smallerproviders.

(P145 at 93 and 105.) The template goes on to toutthe advantages of Sodexho�’s procurement lever-age:

ProcurementFood and supplies are a major portion ofthe cost of a food service program. Sodex-hos purchasing plan for your district con-sists of the following primary elements:�• Buying power. . .Sodexho�’s extensive network of purchas-ing resources will continue to strive for

109a


lower prices of food and supplies for yourDistrict while improving the quality of theproducts you use. With Sodexho�’s buyingpower your District has the opportunity toexperience a significant decrease in thecost of the products you use.. . .Your District will continue to benefit fromSodexho�’s buying expertise. Our reputa-tion and size give us buying advantagesover smaller food service managementorganizations. In turn, the savings inwhich we obtain will be passed on to yourDistrict. You will be charged the sameprices as Sodexho pays for all products.Your District will receive all the benefitsof our volume and trade discounts, exceptfor cash discounts. Sodexho will utilize itstechnical support and its national andlocal buying power whenever possible toobtain the best value for your District.

(P145 at 109.) Finally, the template drives homethe message that Sodexho�’s lower prices for foodwill result in reduced costs for the customer:

�• We have successfully transitioned 300school districts from self-operation tooutsource management and consultingservices.

�• We are the high quality�—low costprovider. . .

When Sodexho assumes the managementof self-operated food services, everyone

110a


benefits. . . . The cost always goes down.Always.

(P145 at 43.) This template was used by Sodexhoin proposals to Franklin Township Schools (P127),Springfield School District (P76), Boonton PublicSchools (P73), East Penn School District (P115),South Plainfield School District (P116), BerlinTownship School District (P69), Manheim Town-ship School District (P70), Chesterfield SchoolDistrict (P71), Daniel Boone Area School District(P65), Carmichaels Area School District (P101),and Warren County Public Schools (P150 at 181).

A template touting �“reduced costs from vast pur-chasing volumes�” was utilized in proposals tohealth care institutions, including SouthernOcean County Hospital:

ProcurementOutcome: Reduced CostsAs the industry leader in food procure-ment, Sodexho offers vast purchasing vol-umes. Because of our many operationsacross the country, you�’re assured of safe,reliable products at competitive prices.

(P142 at 28.) This template also appeared inSodexho proposals to Clearfield Hospital (P84 at28), Holy Redeemer Health System (P82), St.Joseph Medical Center (P139 at 24), Village atMorrison�’s Cove (P129), St. Lawrence Rehabili-tation Center (P124), Marian Manor (P113),Brookline Village (P112), St. Luke�’s Hospital (P60at 151), Bridgewater Retirement Community(P104), and Lorien Nursing & Rehabilitation Cen-ter (P100).

111a


Sodexho�’s Corporate Services division also uti-lized templates touting its lower prices. Forinstance, in a proposal to Bassell USA, Inc.,Sodexho stated: �“Why Sodexho? One of theLargest Buyers of Food in the United States. . .Assists in reducing food costs through increasedpurchasing power.�” (P83 at 34.)

C. Conclusions of LawBased on these findings of fact, the court con-

cludes that Defendants have failed to meet theirburden of rebutting the inference of competitiveinjury by showing that there is no causal connec-tion between the price discrimination and com-petitive injury to Feesers. The evidence presentedat trial demonstrates that food costs constitute asignificant portion of institutional food servicebudgets, and that lower food costs were an impor-tant part of Sodexho�’s strategic plans to win andretain customers, and improve its profit margin.Most significant, Sodexho touts its lower prices inpromotional material to customers. In light of thisevidence, the court declines to draw any broadinference from the testimony of the customer wit-nesses called by Defendants. The court concludesthat Defendants have failed to rebut the inferenceof competitive injury.

IV. Meeting Competition Defense

A. Legal StandardSection 2(b) of the Robinson-Patman Act per-

mits a seller to rebut a prima facie case of dis-crimination by �“showing that his lower price or

112a


the furnishing of services or facilities to any pur-chaser or purchasers was made in good faith tomeet an equally low price of a competitor, or theservices or facilities furnished by a competitor.�” 15U.S.C. § 13(b). This is known as the �“meeting com-petition�” defense.

The purpose of the defense is to promote com-petition by permitting a seller to defend itselfagainst inroads by a competitor. Accordingly, aseller invoking this defense must establish thatthe price concession was granted in order tomeet�—not beat�—a lower price offered by a com-petitor. If a seller successfully asserts the meetingcompetition defense, then there can be no liabilityfor the buyer who induced the discriminatoryprices. Great Atlantic & Pacific Tea, Inc. v. F.T.C.,440 U.S. 69, 77-78, 99 S. Ct. 925, 59 L. Ed. 2d 153(1979). In order to successfully invoke the defense,a seller need not prove that it in fact met a lowerprice offered by a competitor. F.T.C. v. A.E. StaleyMfg. Co., 324 U.S. 746, 759-60, 65 S. Ct. 971, 89 L.Ed. 1338, 40 F.T.C. 906 (1945). However, theseller must prove that the lower price was offeredin good faith to meet its competitor�’s price. If aseller offers a lower price in bad faith, there canbe no defense, even if the price did not beat com-petition. Great Atlantic & Pacific Tea, Inc., 440U.S. at 83 (�“Since good faith, rather than absolutecertainty, is the touchstone of the meeting-com-petition defense, a seller can assert the defenseeven if it has unknowingly made a bid that in factnot only met but beat competition.�”) Conversely, ifthe seller acted in good faith, the defense may beinvoked even where the price offered fell belowthat of a competitor. See, e.g., id. at 83-84 (uphold-

113a


ing meeting-competition defense where seller�’soffer was lower than competitor�’s because seller�’soffer was reasonable and made in good faith).

Good faith is a �“flexible and pragmatic, not atechnical or doctrinaire, concept. The standard ofgood faith is simply the standard of the prudentbusinessman responding fairly to what he rea-sonably believes is a situation of competitivenecessity.�” Falls City Indus. v. Vanco Beverage,Inc., 460 U.S. 428, 441, 103 S. Ct. 1282, 75 L. Ed.2d 174 (1983) (quoting In re Continental Baking,63 F.T.C. 2071, 2163 (1963)). In order to satisfythis standard, the seller must �“show the existenceof facts which would lead a reasonable and pru-dent person to believe that the granting of a lowerprice would in fact meet the equally low price of acompetitor.�” A.E. Staley Mfg. Co., 324 U.S. at 759-60. Generally, to establish good faith the sellermust show that it engaged in at least some rea-sonable inquiry to evaluate a claim of a lowerprice by a competitor. Viviano Macaroni v. F.T.C.,411 F.2d 255 (3d Cir. 1969).

No particular method of verification is requiredin order to establish the meeting competitiondefense. However, the Supreme Court has identi-fied a number of factors which, though notexhaustive, may be useful in evaluating the goodfaith of the seller. These include evidence that (1)�“a seller had received reports of similar discountsfrom other customers;�” (2) a �“seller was threat-ened with the termination of purchases if the dis-count were not met;�” (3) the seller�’s �“efforts tocorroborate the reported discount by seeking doc-umentary evidence;�” (4) the reasonableness of thecompeting offer in light of available market data;

114a


and (5) �“the seller�’s past experience with the par-ticular buyer in question.�” U.S. Gypsum, 438 U.S.422, 455, 98 S. Ct. 2864, 57 L. Ed. 2d 854 (1978)(internal citations and quotations omitted).

Verifying a competitor�’s price can be difficult.Generally the only two options available are toask the buyer, or to ask the competitor, both ofwhich have their drawbacks. Buyers may be reluc-tant to share the specifics of a competing offer, inthe hope that a seller will offer an even lowerprice, or they might lie about the price offered.See, e.g., In re Beatrice Foods, 76 F.T.C. 719(1969). Another risk is that the buyer will refuseto answer, or threaten to terminate business deal-ings altogether. See, e.g., Great Atlantic & PacificTea Co., Inc., 440 U.S. at 69. On the other hand,direct communication with a competitor aboutprices, even for the purpose of verifying a com-peting offer, entails the risk of violating the price-fixing provisions of the Sherman Act. See, e.g.,United States v. U.S. Gypsum, 438 U.S. 422, 98 S.Ct. 2864, 57 L. Ed. 2d 854 (1978). Moreover, as theSupreme Court has observed, competitors have aninterest in keeping price concessions secret:�“[p]rice concessions by oligopolists generally yieldcompetitive advantages only if secrecy can bemaintained; when the terms of the concession aremade publicly known, other competitors are likelyto follow and any advantage to the initiator is lostin the process.�” Id. at 456.

Thus, in a situation where a seller has limitedinformation about the prices offered by his com-petitors, the meeting competition defense may beunavailable �“since unanswered questions aboutthe reliability of a buyer�’s representations may

115a


well be inconsistent with a good-faith belief that acompeting offer had in fact been made.�” U.S. Gyp-sum, 438 U.S. at 455-56.

B. Findings of FactMichael Foods�’ defense centers on the negotia-

tions for three contracts, each for a duration ofthree years: the 1999 egg contract, the 2002 eggcontract, and the 2002 potato contract. However,the discriminatory prices persisted during theentire period at issue in this case�—from 1999until 2004. Michael Foods presented testimonyfrom Vicky Wass, the main negotiator for MichaelFoods. Defendants chose not to call Sodexho�’snegotiator as a witness.

For the purpose of this litigation, the first sig-nificant contract between the parties was the 1999egg contract, which Wass negotiated on behalf ofMichael Foods. This contract included deeply dis-counted deviated prices on many Michael Foodsegg products, a million dollar signing bonus, andother rebates. Wass testified that she believedthese discounts were necessary to meet competi-tion. However, at the time this contract was nego-tiated, Wass did not know of any other offer by aparticular competitor, but she believed competi-tors would offer similar prices because Sodexhowas such a large and attractive customer. Accord-ing to Wass, during this negotiation Sodexho didnot mention any other competitor by name,describe any other offer it had received, or other-wise imply that it had received another offer. Nordid Wass do any investigation to determinewhether the discounts offered by Michael Foods to

116a


Sodexho matched similar discounts offered bycompeting food manufacturers:

Q. My question is, when you negotiatedthis �’99 agreement, do you recalldoing anything to check what thecompetitive prices were being offeredto Sodexho to make sure you weren�’tbeating those competitive prices?

A. I recall I was in a competitive situa-tion with my competitor on the 1999agreement as well as the 2002 agree-ment.

Q. In �’99 specifically, do you recall seeingany prices of a competitor fromSodexho?

A. I don�’t recall, but I remember conver-sations about such.

Q. Okay. Do you recall conversationsabout specific prices being offered bya competitor?

A. Not specific, but in scope, yes.Q. In scope?A. In scope.Q. What do you mean by in scope?A. Essentially, where I had to be, maybe

not showing me exactly a price tomatch a price, but looking at theentire portfolio of what I was offeringand asking me to, you know, do a bet-ter job with my pricing.

117a


Q. Well, do you recall in �‘99 if anyone,either orally or in writing, gave you aspecific price of a competitor beingoffered to Sodexho?

A. Specific price, no, sir.Q. Do you recall anyone telling you what

the duration was of the offer of a com-petitive price they had from someoneelse in 1999?

A. No, I do not.Q. So you didn�’t know the duration. You

didn�’t know the specific price. Cor-rect?

A. No, not the specific price.Q. So all you knew was, Sodexho said,

you�’ve got to do better? I need a betterprice. Correct? That�’s fair?

A. They would tell me�—yes. Generally,yes, sir.

(Trial Transcript, Jan. 18, 2008, 188:15-189:22.)Thus, when Sodexho requested price concessions,Wass simply assumed that the requested conces-sions matched an offer Sodexho had received froma competitor. However, Defendants presented noevidence at trial to support this assumption, andit seems more likely that Sodexho simply wantedbetter prices and felt that it was a big enough cus-tomer to push for it. Accordingly, the court findsthat Defendants have failed to meet their burdenof proving that the discounts on the 1999 egg con-tract were offered to meet competition.

118a


Likewise, Defendants presented no evidenceabout the terms or duration of any offer Sodexhomay have received from other egg or potato com-panies for the 2002 contracts. Wass was alsoresponsible for negotiating the 2002 egg andpotato contracts on behalf of Michael Foods. Dur-ing the course of the negotiation, Wass soughtadvice from her supervisors, Dean Sprinkle andMark Westphal about offering a new discount toSodexho on the egg contract. (November 13, 2001email from Vicky Wass to Dean Sprinkle, P210.)In this negotiation, Michael Foods was proposingto reduce the number of products on deviated pric-ing and instead offer larger rebates off of the listprice. In the email, Wass warned Sprinkle thatSodexho �“wants something significant,�” but theprice discount proposed was not linked to any spe-cific offer to Sodexho by a competitor:

Q. And when you were proposing, offer-ing to him, and we�’ll go through theother terms of the offer, the newterms of the deal�—

A. Yes.Q. �—at that point, in November 13, 2001,

Sodexho hadn�’t mentioned a wordabout any competitive offer yet, cor-rect?

A. This negotiation went on for quite awhile. There was back and forth. Andthere was some talk�—I mean, it wasalmost an expectation on my partthat my competition would be there,number one. Number two, you know,

119a


there was an inference that I wasn�’tsharp enough, that this wasn�’t work-ing, that, you know, he wanted some-thing better.

Q. I understand he told you he wantedsomething better, it wasn�’t working.He didn�’t tell you, here is a specificcompetitive proposal you have tomeet, right?

A. He didn�’t show me one.Q. He didn�’t orally tell you the details of

one?A. He did not orally tell me the details of

that proposal, no sir.Q. And, in fact, on November 13, he

didn�’t even yet mention the name of aspecific competitor as of the 13, cor-rect?

A. I can�’t recall. I mean these werelengthy. I mean, I can�’t recall.

(Trial Transcript, Jan. 18, 2008, 203:17-204:14.)Two weeks later, in another email to Dean

Sprinkle, Wass stated that Michael Foods needs todo better on its pricing and that Sodexho had toldher that if Michael Foods did not do better, itwould lose both the egg and the potato contract.(Email from Vicky Wass to Dean Sprinkle, P209.)When pressed for the details of this negotiationwith Sodexho, Wass testified as follows:

Q. And he told me that, if we hold withthis present proposal, that the egg

120a


contract will be awarded to Sunny-fresh and we will not be awarded thepotato contract, correct?

A. That�’s correct.Q. But he did not tell you any detail of

what would be in any contract withSunnyfresh, right, not one detail?

A. He did not give me any specific pric-ing, any specific rebates, no specific.

Q. About anything?A. He told me my deal was not as good as

theirs.Q. And you didn�’t ask him for any of

those details, right?A. I would not. I wouldn�’t think to ask

him for that.Q. No one told you in the company,

Vicky, if he wants us to give a betterprice, go out and find out the details.Ask him to give you the details. Noone told you that, right?

A. It�’s not done that way. I mean, youcan�’t�—you�’re not going to go and askthe customer to provide documents,because that�’s unethical and it�’sreally bad for them. The only way Iwould be able to confirm and 100 per-cent verify that would be going toSunnyfresh, and Sunnyfresh isn�’tgoing to give me that informationeither. So when you are looking at

121a


that, you�’ve got to look at all theresources and all of your experienceand come up with some kind of a sum-mation of what you think is real. I didknow at the time�—Sunnyfresh, tradi-tionally, is lower cost than we are,traditionally. I do know that otheragreements that we�’ve been negotiat-ing, not only myself, as well as myteam members, were getting like, youknow, like pricing. I was not surprisedby the things that was coming forthon his demands. But did I see some-thing? Did I see a document? No.

Q. You didn�’t hear any details either?You didn�’t see it or hear it?

A. I asked him how close I needed to be,you know, in order to meet the com-petition.

Q. Did you know how long any offer wasfrom Sunnyfresh, whether it was aone-year deal, a two-year deal, any-thing about any length?

A. No.Q. Did you know what the signing bonus

would be from Sunnyfresh?A. No.Q. Did you know what the rebates would

be from Sunnyfresh?A. No, sir.

122a


Q. Did you know what the specific pricewould be, the starting price from Sun-nyfresh?

A. No, I did not.Q. As you�’re sitting here today, when you

made this offer, you had no way ofknowing whether you were going to bebelow Sunnyfresh, at Sunnyfresh, orbetter than Sunnyfresh? You had noway to know, right?

A. There�’s no way to know exactly whathe had on his desk.

Q. But I�’m not talking about exactly, likeyou didn�’t know like to the penny. Youhad no way of knowing at all? Youcould have been 20 percent betterthan Sunnyfresh. You could have been20 percent worse than Sunnyfresh.You had no way of knowing at all,right?

A. Only what my experience with�—whattheir behavior was with other agree-ments, that�’s the only way I wouldhave a good concrete idea that wewere�—that what he was saying wasaccurate.

(Trial Transcript, Jan. 18, 2008, 209:8-211:21.)Wass further explained the difficulty in deter-

mining the prices offered by a competitor duringexamination by defense counsel:

Q. Now with such strong competition asto both eggs and potatoes, what does

123a


Michael Foods do to try to keepinformed about competitive condition,competitive pricing?

A. Well, we�’re limited on what we haveaccess to, but we certainly keep ourear to the ground and are constantlywatching what the markets are doing,what the egg markets and potato mar-kets are doing, as well as what themarketplace is doing, you know, howaggressively our competitors areapproaching these customers. And weget some competitive intelligencethrough that by just kind of knowingwhere they�’re trading or what they�’reproviding to other�—in other negotia-tions.

Q. Do you sometimes lose accounts toyour competitors?

A. Absolutely.Q. Do you sometimes find out competi-

tive information through that event?A. Very rarely. We hardly ever find out

after the fact.Q. So how hard is it to go specific infor-

mation about your competitors�’ pric-ing?

A. It�’s very difficult. The only way wewould get that, if the competitionwould provide that data to us.

Q. How often do your competitors tellyou?

124a


A. That would not happen at all. Also,we, our customers, it�’s bad form, andit�’s unprofessional, and it�’s not done,nor do we ask it very intently becausewe certainly don�’t want our customersto be sharing our pricing programswith our competition. So it�’s bad formto go there, and it�’s�—we�’re never pro-vided that information.

Q. Did you try to keep generally aware ofmarket pricing, market competition?

A. Certainly.(Trial Transcript, Jan. 28, 2008, 13:1-6.) Whenasked about other methods of competitive intelli-gence, Wass testified that Michael Foods couldroughly determine the prices offered by competi-tors when it lost contracts through competitivebidding:

Q. . . . Was there other general marketintelligence information you couldlook to, to see whether you thoughtthat Sunnyfresh terms you were beingtold about seemed plausible?

A. The only basis that I would have is, Iwould know what the egg marketswere doing at that time. And a lot oftimes, you kind of know where peopleare trading whenever the egg marketsare where they are. Then the otherpiece of it is, the other contracts thatwere being negotiated, those that wewere losing or those that we were win-ning, we pretty much could probably

125a


gather some general intelligence ifindeed what was being offered up bySodexho at that time was reasonable,not surprising.

(Trial Transcript, Jan. 28, 2008, 26:18-27:6.) How-ever, Wass later testified that in the competitivebidding process, Michael Foods does not typicallyknow how well its bid measured up to successfulcompetitors:

Q. Do you have any experience previ-ously with Sodexho where they madea demand, and then you found it wasclose to your competitor�’s price?

A. No, not afterwards, no.Q. And you don�’t have that experience

your whole time with Sodexho, right?A. No, you normally don�’t find that out.

It�’s not a public bid.(Trial Transcript, Jan. 18, 2008, 212:5-12.)

Shortly before the commencement of the 2002egg contract negotiations, Sodexho invitedMichael Foods to submit a bid to supply potatoproducts as well, which Michael Foods did not sellto Sodexho at that time. During these negotiationsMichael Foods was aware that Resers, a competi-tor, was the incumbent on the contract, but itknew nothing about the terms of any offer Reserswas making. In these negotiations, led by VickyWass, Michael Foods offered significant rebates onpotato products to Sodexho. However, Wass tes-tified at trial that Michael Foods had no way ofknowing the terms of any offer by a competitor

126a


such as Resers, and how close Michael Foods�’ offercame:

Q. I�’m asking, Ms. Wass, on the potatocontract, and we�’ll get to the contract,you offered, for example, sizablerebates on certain kinds of potatoes,correct?

A. We offered rebates, yes, sir.Q. And you have no way of knowing,

when you offered that, whether theincumbent was offering the samerebates, bigger rebates, or smallerrebates, correct?

A. No, sir, I don�’t.Q. You had no way of knowing whether

or not your starting prices for pota-toes were the same or different fromthe incumbent, correct?

A. Not unless I physically saw the actualoffer, no.

(Trial Transcript, Jan. 18, 2008, 213:19-214:5.)Towards the end, Sodexho threatened to walk

away from both the egg and the potato negotia-tions unless Michael Foods granted the conces-sions it was seeking. In an email to DeanSprinkle, Vicky Wass wrote the following:

Just got off the phone with Mitch. He isextremely unhappy with us because hedoes not see that we have addressed hisrequests. He told me that if we hold withthis present proposal that the egg contract

127a


will be awarded to Sunnyfresh effectiveSeptember 1, 2002 and we will not beawarded the potato contract. He wantsthe following changes:1) Take down the ceilings and floors to theoriginal contract.2) Readjust the formulas to the originalcontract.3) Increase the 1,000,000 payment to 1.2MM. $1,000,000 for the egg products pieceand $ 200,000 in lieu of the reduced BetterN�’ Eggs price.4) Begin program January 1st whichwould increase the Manufacturer�’s rebatefrom 3 cents to 5 cents at that time.5) Increase the all other products rebateat list pricing from $2.00/case to $3.00/case.6) Mitch also wants an additional $77,000as monetary adjustment for dollars lost byshifting the August end date to January.They are not messing around. These guysare the �“big dogs�” in contract managementand healthcare and they are pushing theirposition. Please advise. Mitch wants me torespond back to him today.

(Nov. 27, 2001 Email from Vicky Wass to DeanSprinkle, P209.) Ultimately Michael Foods wonboth the 2002 egg and potato contracts when itgranted each of Sodexho�’s requests outlined in theemail. However, Defendants offered no evidencethat these numerous additional discounts were

128a


calculated to meet rather than beat any competingoffer by Sunnyfresh. Rather, as the last paragraphof the email makes clear, Michael Foods accededto Sodexho�’s demands because it was �“pushing[its] position�” as �“�‘big dogs�’ in contract manage-ment.�” (Id.)

Moreover, even after the contracts were signed,Michael Foods offered further discounts toSodexho. In 2003, during a period of great volatil-ity in the egg market, Michael Foods voluntarilyoffered Sodexho a further concession on egg prices.(See November 13, 2003 letter to Sodexho, P248 at2.) According to Wass, this discount was offeredonly to Sodexho and was not in response to anycompetitive offer by another company, but rather,as Wass explained at trial, �“I offered this in thespirit of our strategic partnership with them.�”(Trial Transcript, Jan. 18, 2008, 245:13-14.)

For the most part, the court found Wass to be acredible and candid witness. Nevertheless thecourt does not accept her assertion that the dis-counts granted by Michael Foods to Sodexho wereoffered in good faith for the purpose of meetingcompetition. Wass�’s testimony establishes that shesimply did not have enough information aboutcompetitive offers from other egg and potato man-ufacturers to craft an offer calculated in good faithto meet, and not beat Michael Foods�’ competition.Indeed, for the 1999 egg negotiation and most ofthe 2002 egg and potato negotiation, Sodexhonever made reference to any other offer by a com-petitor. Essentially Wass cited two reasons in sup-port of her assumption that Sodexho had receivedoffers from Michael Foods�’ competitors: (1)Sodexho�’s demands for lower prices were in line

129a


with what Wass expected her competitors coulddeliver; and (2) Sodexho is a very attractive cus-tomer likely to receive other offers from MichaelFoods�’ competitors due to its large purchasing vol-ume.

The court finds both of these facts insufficient tojustify Wass�’s assumption that Sodexho hadreceived other offers from competitors. Wass�’sbelief that competitors could deliver the conces-sions Sodexho requested was based on her generalknowledge of the egg market and her negotiationsin other contracts. This knowledge is of littlevalue however, given that Wass testified that sherarely learns the prices her competitors are offer-ing. This leaves only the fact that Sodexho is alarge customer likely to receive other offers fromcompetitors due to its large purchasing volumes.However, there are two problems with inferringfrom this evidence that Sodexho�’s demands werebased on its receipt of a competing offer. First, isthe fact that it is in Sodexho�’s interest to secure alower price from Michael Foods, rather than amatching price. Accordingly, it is more likely thatSodexho�’s demands were calculated to beat ratherthan meet any other offers they may havereceived. Indeed, Wass testified that Sodexhorepeatedly told Michael Foods where it needed tobe to win the contract, which is not necessarily thesame as the price needed to meet competition.

However, the greater problem is that acceptanceof Wass�’s assumption would be contrary to the pri-mary purpose of the Robinson-Patman Act, whichis to prevent large buyers from utilizing their pur-chasing power to secure lower prices than theirsmaller competitors. If the meeting competition

130a


defense could be satisfied merely by showing thata particular customer was large and thereforelikely to receive lower prices from competitors,then the Act�’s purpose would be largely thwarted.This is particularly true where, as here, the largebuyer made no reference to any offer by any par-ticular competitor for most of the negotiations.Moreover, when Sodexho mentioned Sunnyfreshtowards the end of the negotiation, it providedMichael Foods with no information about thedetails or duration of that offer. Under these cir-cumstances, Michael Foods was not in a positionto make an offer reasonably calculated to meet,rather than beat the alleged offer by Sunnyfresh.In sum, the court is not persuaded that the dis-counts granted to Sodexho constituted a good faithoffer to meet competition, rather than a conces-sion to win the business of a large and powerfulbuyer.

C. Conclusions of LawBased on the factual findings, court concludes

that Michael Foods failed to demonstrate a goodfaith effort to meet competition by other egg andpotato suppliers. This case is analogous to the sit-uation referred to in U.S. Gypsum, in which theseller lacks sufficient information to make a goodfaith offer that meets, rather than beats that of acompetitor. Instead, the court concludes that likethe seller in Viviano Macaroni, Michael Foods�’discounts were �“made in an effort to obtain addi-tional business from [the buyer] and not to defenditself against the inroads of rapacious competi-tors.�” Viviano Macaroni Co., 411 F.2d at 258.

131a


Defendants urge a different conclusion, relyingupon Great Atlantic & Pacific Tea Company. Inthat case, the seller, Borden Dairy, was in nego-tiations with A&P, a grocery store chain, over amilk supply contract. During the course of thenegotiations, A&P, a longstanding Borden cus-tomer, informed Borden that it had received a bet-ter offer from a competitor. When asked, A&Prefused to provide any additional details about theoffer, other than to say that the �“�‘offer was noteven in the ballpark�’ and that a $50,000 improve-ment would not be a drop in the bucket.�” 440 U.S.at 84. Stating that it was making the offer inorder to meet competition, Borden submitted asecond bid that turned out to be lower than itscompetitor�’s bid. A&P did not inform Borden thatits bid was lower. On these facts, the SupremeCourt found that Borden was entitled to the meet-ing competition defense because it reasonablyrelied upon a credible threat of termination by anestablished customer.

The instant case is distinguishable from GreatAtlantic & Pacific Tea Company because unlikethe seller in that case, Michael Foods did not seekadditional information to verify Sodexho�’s claimthat Michael Foods offer was not good enough, nordid Michael Foods inform Sodexho that it wasgranting the concessions for the purpose of meet-ing competition, rather than simply to winSodexho�’s business. Accordingly, the court con-cludes that Defendants have failed to establishthe meeting competition defense.

132a


V. Inducement of Price Discrimination (Section 2(f))

A. Legal StandardSection 2(f) of the Robinson-Patman Act pro-

vides that �“[i]t shall be unlawful for any personengaged in commerce, in the course of such com-merce, knowingly to induce or receive a discrimi-nation in price which is prohibited by thissection.�” 15 U.S.C. § 13(f). �“[T]he buyer whomCongress in the main sought to reach was the onewho, knowing full well that there was little like-lihood of a defense for the seller, nevertheless pro-ceeded to exert pressure for lower prices.�”Automatic Canteen v. F.T.C., 346 U.S. 61, 79, 73S. Ct. 1017, 97 L. Ed. 1454, 49 F.T.C. 1763 (1953).

B. Findings of Fact and Conclusions ofLaw

Based on the court�’s factual findings above,which the court will not here repeat, Plaintiff hasdemonstrated that Sodexho knowingly induced orreceived the price discrimination detailed abovefrom Michael Foods. The most persuasive evidenceof knowing inducement is the promise Sodexhoextracted from Michael Foods known as the �“mostfavored nations�” clause. This clause requiredMichael Foods to provide Sodexho with the lowestprice on Michael Foods products. Although it istrue that the clause did not require that the pricebe lower than any other purchaser, it is clear thatthe purpose was to secure a price well below thelist price received by smaller purchasers such asFeesers. This was not just form language. The evi-dence demonstrates that Sodexho vigorously

133a


enforced this contractual promise. For instance, in2002, when Sodexho acquired the Wood Company,it learned that Wood had received better pricingthan Sodexho on certain products. Accordingly,Sodexho demanded and received compensation forthe breach of its contract.

The court�’s conclusion that Sodexho knowinglyinduced price discrimination is also supported bySodexho�’s strategic planning documents demon-strating that Sodexho intended to secure discountsfrom manufacturers such as Michael Foods inorder to increase its profit margin and gain moremarket share through conversion of self-op insti-tutions to food service management. Sodexho�’spromotional materials touting its ability to securelower food prices than its competitors also supportthis conclusion. Altogether, the evidence at trialoverwhelmingly establishes that Sodexho know-ingly induced Michael Foods to discriminate inprice. Thus the court concludes that Sodexho vio-lated Section 2(f) of the Robinson-Patman Act.

VI. Equitable ReliefAs the prevailing party in this action Feesers

seeks two forms of equitable relief from the court.First, Feesers seeks a declaratory judgment pur-suant to 28 U.S.C. § 2201 that Michael Foods hasunlawfully discriminated as to price againstFeesers and that Sodexho has unlawfully inducedor received such price discrimination. Second,Feesers seeks injunctive relief pursuant to Section16 of the Clayton Act, 15 U.S.C. § 26.10 Specifi-

134a


10 Feesers also seeks reasonable attorneys�’ fees andcourt costs, which will be considered in a separate opinionand order.

cally, Feesers seeks to enjoin Michael Foods fromdiscriminating in price for the sale of food toFeesers and Sodexho, and to enjoin Sodexho fromcontinuing to induce or receive unlawful price dis-crimination from Michael Foods or any other foodmanufacturer that sells food to both Feesers andSodexho. Defendants oppose equitable relief, argu-ing that Feesers has unclean hands due to itsreceipt of deviated prices, that Feesers is unlikelyto be injured by Sodexho�’s receipt of deviatedprices, and that deviated pricing is critical toMichael Foods�’ business. These arguments will bediscussed in turn.

First, the court rejects Defendants�’ argumentthat Feesers has unclean hands as a result of itsreceipt of deviated pricing. As discussed above,there is a significant difference between the devi-ated pricing received by Sodexho and Feesers.Deviated pricing received by Feesers is customer-specific.11 Any distributor or food service man-agement company servicing the account wouldhave access to the same food prices. By contrast,the deviated pricing received by Sodexho may beused by Sodexho to compete for new accounts orretain current customers. As noted above, this

135a


11 This customer-specific deviated pricing would onlyviolate the Robinson-Patman Act if it had the effect ofimpairing competition between favored and non-favored cus-tomers. But this scenario is unlikely. Students would notchoose a school or university because one cafeteria negoti-ated a lower price for raw food than the other. Likewise hos-pitals and nursing homes are unlikely to gain a competitiveadvantage over rival institutions based on the price of rawfood at the loading dock. However, if such a showing weremade, then that price discrimination would violate theRobinson-Patman Act.

pricing gives Sodexho a competitive advantageover smaller rivals competing to resell raw foodproducts to institutional dining services. If a cus-tomer switches from self-op to Sodexho, it benefitsfrom Sodexho�’s deviated prices, but if it revertsback to self-op or chooses another managementcompany, it loses access to Sodexho�’s deviatedprices.

Defendants further argue that an injunction isnot necessary to protect Feesers from competitiveinjury because the deviated pricing it complains ofhas been in place for many years, and Feesers hasbeen unable to identify any lost sales resultingfrom such pricing. The court rejects this argu-ment. The court is satisfied that the level of pricediscrimination by Michael Foods in favor ofSodexho is great enough, and the customers forwhom both Feesers and Sodexho compete aresophisticated enough, that it is a matter of timebefore that price disparity causes Feesers to losecustomers to Sodexho.

Finally, Defendants warn the court of theallegedly disastrous consequences of barring devi-ated pricing to Sodexho. According to MichaelFoods, the company would be forced into financialruin if it were required to lower its prices toFeesers for resale to all customers. However,Michael Foods would not necessarily be requiredto extend lower prices to Feesers in order to com-ply with the Robinson-Patman Act. For instance,Michael Foods could raise Sodexho�’s price tomatch the national list price, or it could remaincompetitive by lowering the national list price.The court offers no opinion as to which methodMichael Foods must adopt in order to comply with

136a


this order. However, it is clear that the Act pro-hibits Michael Foods from discriminating in priceagainst Feesers and in favor of Sodexho. Accord-ingly, Plaintiff is entitled to an injunction pro-hibiting Michael Foods from engaging in suchdiscrimination, and Sodexho from inducing suchdiscrimination.

One aspect of the injunctive relief Feesers seeksagainst Sodexho merits further discussion.Feesers seeks to enjoin Sodexho from inducing orreceiving discriminatory pricing not only fromMichael Foods, but also from other manufacturersfrom which both Feesers and Sodexho purchasegoods. The court declines to grant such broadinjunctive relief. Although the evidence presentedat trial suggests that Sodexho has negotiatedprice discounts with manufacturers other thanMichael Foods, those manufacturers are notnamed as parties to this suit, and the details oftheir pricing arrangements with Sodexho are notbefore the court. Thus the court is in no position todetermine whether Sodexho has knowinglyinduced or received price discriminatory pricingfrom other manufacturers. Accordingly, the courtwill enjoin Sodexho from inducing or receiving dis-criminatory pricing from Michael Foods, but theinjunctive relief shall not extend to other manu-facturers.

137a


VII. ConclusionFor the foregoing reasons, the court finds in

favor of Plaintiff. Defendant Michael Foods vio-lated Section 2(a) of the Robinson-Patman Act bydiscriminating in price between competing pur-chasers, and Sodexho violated Section 2(f) of theAct by inducing such discrimination. An appro-priate order will issue.

s/ Sylvia H. RamboSYLVIA H. RAMBOUnited States District Judge

Dated: April 27, 2009.

138a


UNITED STATES DISTRICT COURTFOR THE MIDDLE DISTRICT OF PENNSYLVANIA



Plaintiff,v.


ORDER

In accordance with the foregoing memorandumof law, IT IS HEREBY ORDERED AND ADJUDGEDTHAT:

(1) The Clerk of Court shall enter judgment asfollows:

(a) Michael Foods has unlawfully discrimi-nated as to price against Feesers and Sodexhohas unlawfully induced or received such pricediscrimination in violation of the Robinson-Patman Act.

(b) Michael Foods is hereby enjoined fromdiscriminating unlawfully in price in favor ofSodexho and against Feesers.

139a


(c) Sodexho is hereby enjoined from contin-uing to induce or receive unlawful price dis-crimination from Michael Foods.

(2) No later than 30 days from the date of thisorder, Plaintiff shall submit a petition for rea-sonable attorneys�’ fees and costs, supported byaffidavits describing the experience of the attor-neys and their standard hourly rates. ThereafterDefendants shall have 30 days to submit aresponse to the petition.

s/ Sylvia H. RamboSYLVIA H. RAMBOUnited States District Judge

Dated: April 27, 2009.

140a


APPENDIX C

IN THE UNITED STATES DISTRICT COURTFOR THE MIDDLE DISTRICT OF PENNSYLVANIA



Plaintiff,�—v.�—


MEMORANDUM

Before the court are motions by DefendantsMichael Foods and Sodexho to amend the court�’sApril 27, 2009 order granting judgment in favor ofPlaintiff Feesers, Inc. pursuant to Fed.R.Civ.P.59(e). For the reasons that follow, Michael Foods�’motion will be denied, and Sodexho�’s motion willbe granted in part and denied in part.

I. BackgroundOn April 27, 2009, following a three week bench

trial, this court issued an opinion and order grant-ing a declaratory judgment in Plaintiff�’s favor and

141a

26922 • Dewey • APPENDIX part: 3 (revised 12/1/08) LJB 1:00 5/12/10

issuing injunctive relief. On May 5, 2009, Plaintifffiled a motion for contempt and injunctive relief,which the court granted on May 26, 2009 (Doc.431). On May 6, 2009, Defendant Michael Foodsfiled a motion to alter or amend judgment pur-suant to Fed.R.Civ.P. 59(e). (Doc. 404.) A brief insupport thereof was filed the same day. (Doc. 405.)On May 11, 2009, Defendant Sodexho also filed amotion to alter or amend judgment pursuant toFed.R.Civ.P. 59(e), (Doc. 414), and a brief in sup-port thereof, (Doc. 415). On May 21, 2009, Feesersfiled briefs in opposition, (Docs. 427, 428), towhich Michael Foods and Sodexho submitted replybriefs, (Docs.439, 442). Accordingly, the motionsare ripe for disposition.

II. StandardThe purpose of a motion for reconsideration �“is

to allow a court to correct manifest errors of lawor fact, or in limited circumstances, to presentnewly discovered evidence, but not to relitigate oldissues, to advance new theories, or to secure arehearing on the merits.�” Gutierrez v. Ashcroft,289 F.Supp.2d 555, 561 (D.N.J.2003) (internalcitations omitted). Reconsideration of a judgmentis an extraordinary remedy and is generally onlygranted where �“(1) an intervening change in thelaw has occurred, (2) new evidence not previouslyavailable has emerged, or (3) the need to correct aclear error of law or prevent a manifest injusticearises.�” Id.

142a


III. DiscussionDefendants challenge the court�’s opinion and

order on a number of grounds. First, MichaelFoods and Sodexho argue that the Third Circuit�’sdecision in Toledo Mack Sales & Service, Inc. v.Mack Trucks, Inc., 530 F.3d 204 (3d Cir. 2008)entitles them to judgment in their favor. Addi-tionally, Sodexho argues that the order barring itfrom inducing price discrimination is impermis-sibly vague and overbroad. These arguments willbe addressed in turn.

A. Toledo MackBoth Defendants argue that Toledo Mack enti-

tles them to judgment in their favor, but for dif-ferent reasons. Michael Foods argues that ToledoMack represents an intervening change in lawthat entitles it to a reversal of the court�’s judg-ment. Sodexho argues that the case simply appliesthe principles set forth in the Supreme Court�’s2006 opinion in Volvo Trucks North America, Inc.v. Reeder-Simco GMC, Inc., 546 U.S. 164, 126S.Ct. 860, 163 L.Ed.2d 663 (2006), and that thiscourt committed a clear legal error by finding inFeesers�’ favor. However, both parties presentessentially the same legal arguments that thecourt considered and rejected in the April 27, 2009opinion and order.

Toledo Mack and Reeder-Simco concerned com-petition for the sale of custom made trucks amongcar dealerships operating in distinct geographiczones, and in both cases, courts found that com-petition was limited to a formal bidding process.In Reeder-Simco, which this court discussed at

143a


length in the April 27, 2009 opinion and order,(see Doc. 395 at 24-26), the Supreme Court heldthat a price disparity caused no competitive injurybecause the plaintiff had never submitted a bid inhead-to-head competition with favored dealers. InToledo Mack, the Third Circuit applied this hold-ing to almost identical facts, concluding that theRobinson-Patman Act does not apply to a caseinvolving �“a single sale of a customized good via acompetitive bidding process.�” 530 F.3d at 228.

According to Defendants, these cases entitlethem to judgment in their favor because Feesersnever proved that Michael Foods actually sold itany products that later became the basis for head-to-head competition with Sodexho. Defendantsargue that like the situation in Mack Trucks, atthe time a customer chooses between Sodexho andFeesers, the two companies possess nothing morethan an offer to sell.

The court rejects as baseless Michael Foods�’ con-tention that �“Feesers did not offer any evidencethat Michael Foods actually sold to Feesers andSodexho, the products that then become the basisof head-to-head competition for the same cus-tomer.�” (Doc. 405 at 4.) Unlike the seller in ToledoTrucks, Michael Foods does more than merelyoffer a lower price in a competitive bidding situ-ation. The court has already determined thatMichael Foods made sales to two purchasers, anelement of the prima facie case of price discrimi-nation, in its 2006 summary judgment opinion,and that holding was undisturbed by both partieson appeal, and remains the law of the case.

Defendants further argue that the timing ofcompetition in this case precludes a finding of

144a


competitive injury. In support of this argument,Defendants cite the following passage in ToledoMack:

Although Mack dealers may compete withone another by bidding against each otherfor the same deal, and the amount of salesassistance Mack offers to each dealer maywell determine whether a customerchooses to accept a bid from one Mackdealer or another, Mack does not sell atruck to the dealer until the customeractually selects a dealer�’s bid. Because nosale takes place until a customer acceptsa dealer�’s bid, the amount of sales assis-tance Mack is willing to provide to a par-ticular dealer is part of an offer by Mackto sell, not a sale. Regardless of any com-petition between the dealers during thebidding process, only a dealer whose bid isaccepted by a customer will actually buy atruck from Mack. Therefore, only one sale,not two actually results.

530 F.3d at 228. Here, Defendants correctly pointout that this court found that competition waslimited to the time at which an institution ischoosing between self-op and management. At thistime, Defendants argue, Michael Foods has eithernot actually sold the products destined for thosecustomers to Feesers and Sysco, or it has soldthem at list price. Thus, according to Defendants,this situation is identical to that in Toledo Mack,and requires a reversal of the judgment in favor ofFeesers.

145a


Essentially, Defendants urge a reading ofToledo Mack that would impose a new element inthe prima facie case of price discrimination underthe Robinson-Patman Act�—a sale of the commod-ity to two different sellers prior to the competitionfor the resale of those goods. However, such areading is not warranted by Toledo Mack, andwould be contrary to the purposes of the Robinson-Patman Act.

In Toledo Mack, the Third Circuit addressed theapplicability of the Robinson-Patman Act to casesinvolving closed bidding for custom-made goods.In that case, as well as Reeder-Simco, the losingbidder would never actually purchase the itemwhich was the subject of the competition. Defen-dants provide no support for their argument thatthe holding of Toledo Trucks should be extendedto cases such as this, where the goods in questionare perishable commodities that two competitorsregularly purchase and keep in stock for resale tocustomers. Moreover, a prior sale requirementwould render the Robinson-Patman Act inappli-cable to price discrimination in the sale of anyperishable commodity which is then resold pur-suant to a supply contract that exceeds the shelflife of that commodity. The most logical reading ofToledo Mack and Reeder-Simco is that the hold-ings of those cases apply only to competition forthe sale of custom-manufactured goods that isrestricted to bidding markets. Because those casesare inapplicable to the case at bar, the court willdeny Defendants�’ motions to amend judgment inlight of Toledo Mack.

146a


B. Vagueness and Overbreadth of OrderSodexho further argues that the court�’s order

enjoining it �“from continuing to induce or receiveunlawful price discrimination from MichaelFoods,�” (April 27, 2009 Order ¶ (1)(c).), should beamended because it is vague and overbroad.Sodexho characterizes the order as an impermis-sibly vague �“obey the law�” order, and furtherclaims that the order is insufficient to put it onnotice as to what conduct would violate the order.

Federal Rule of Civil Procedure 65(d) providesthat �“[e]very order granting an injunction andevery restraining order must: (A) state the rea-sons why it was issued; (B) state its terms specif-ically; and (C) describe in reasonable detail-andnot by referring to the complaint or other docu-ment-the act or acts restrained or required.�” Aninjunction simply commanding a defendant toobey the law does not satisfy this specificityrequirement. Belitskus v. Pizzingrilli, 343 F.3d632, 650 (3d Cir.2003); see also Meyer v. Brown &Root Const. Co., 661 F.2d 369, 373 (5th Cir. 1981).

Sodexho claims that the injunction does notdescribe in reasonable detail the acts restrained orrequired, but rather �“simply parrots the generalterms of the Robinson-Patman Act�” in violation ofthe specificity requirement of Fed.R.Civ.P. 65(d).(Doc. 415 at 7.) The court disagrees. Here, theorder Sodexho challenges provides that �“Sodexhois hereby enjoined from continuing to induce orreceive unlawful price discrimination fromMichael Foods.�” (April 27, 2009 Order ¶ (1)(c).)The order also includes a declaratory judgmentproviding that �“Michael Foods has unlawfully dis-

147a


criminated as to price against Feesers andSodexho has unlawfully induced or received suchprice discrimination in violation of the Robinson-Patman Act.�” (April 27, 2009 Order ¶ (1)(a).) Addi-tionally, the order is accompanied by a 83 pagetrial opinion setting forth this court�’s findings offacts from trial and legal conclusions. Thus, theinjunctive relief is far more specific than a simplecommand to obey the Robinson-Patman Act; itplaces Sodexho on notice of its conduct that vio-lated the Act and which is now prohibited by theorder.

Sodexho further argues that the order should beamended because �“it contains ambiguities thatgive rise to two possible instances of indisputableoverbreadth.�” (Doc. 415 at 8.) First, Sodexhoargues that the order omits the scienter require-ment because it does not specifically prohibitSodexho from knowingly inducing or receivingprice discrimination. The court does not believethe order is ambiguous, particularly in light of thethis court�’s trial court findings in the accompa-nying opinion that Sodexho actively sought toobtain food at lower prices than its competitors.Nevertheless, in order to clarify Sodexho�’s confu-sion, the court will grant its request and amendthe order to include the word �“knowingly.�”

Second, Sodexho argues that the order isambiguous and overbroad because it is notexpressly limited to price discrimination againstFeesers, but instead prohibits Sodexho frominducing or receiving any unlawful price dis-crimination from Michael Foods. Sodexho assertsthat it is uncertain whether it may continue toreceive its negotiated pricing if that same pricing

148a


is extended to Feesers, because that pricing wouldstill be lower than the national list price receivedby other distributors. There is evidence in therecord that Sodexho is in competition with dis-tributors other than Feesers for the sale ofMichael Foods products to institutional food ser-vice customers, namely Sodexho�’s strategic plan-ning documents. However, the only plaintiff inthis case is Feesers, and the court agrees thatthere is an insufficient record to support a broaderinjunction barring Sodexho from inducing orreceiving lower prices than distributors other thanFeesers.1 Accordingly, the court will amend theinjunction against Sodexho to specify that it mayno longer knowingly induce or receive price dis-crimination against Feesers from Michael Foods.

IV. ConclusionMichael Foods�’ motion to amend judgment will

be denied, and Sodexho�’s motion to amend judg-ment will be granted in part and denied in part.An appropriate order will issue.

s/Sylvia H. Rambo United States District Judge

Dated: June 30, 2009.

149a


1 The court expresses no opinion on the merits of anyclaims other distributors may have against Sodexho if itknowingly induces or receives price discrimination fromMichael Foods against other competitors.

IN THE UNITED STATES DISTRICT COURTFOR THE MIDDLE DISTRICT OF PENNSYLVANIA

__________

Civil No. 1:CV-04-0576JUDGE SYLVIA H. RAMBO

__________

FEESERS, INC.,Plaintiff,

�—v.�—


ORDER

For the reasons set forth in the foregoing mem-orandum of law, IT IS HEREBY ORDERED THAT:

(1) Michael Foods�’ motion to alter judgment(Doc. 404) is DENIED;

(2) Sodexho�’s motion to alter judgment (Doc.414) is GRANTED IN PART and DENIED IN PARTas follows:

(a) The motion is GRANTED with respect to¶ 1(c) of the April 27, 2009 opinion and order(Doc. 395), which shall be amended to read:�“Sodexho is hereby enjoined from continuing

150a


to knowingly induce or receive from MichaelFoods unlawful price discrimination againstFeesers.�”

(b) In all other respects, the motion isDENIED.

(3) The Clerk of Court is directed to issue anamended judgment in accordance with this order.

s/Sylvia H. Rambo United States District Judge

Dated: June 30, 2009.

151a


APPENDIX DPRECEDENTIAL


__________Nos. 09-2548, 09-2952, 09-2993

__________Feesers, Inc.,

Appellant in No. 09-2993v.

Michael Foods, Inc.; Sodexho, Inc.,Appellants in Nos. 09-2548 and 09-2952__________

On Appeal from the United States District Courtfor the Middle District of Pennsylvania

District Court No. 04-cv-00576District Judge: The Honorable Sylvia H. Rambo__________

Argued October 29, 2009

Before: SMITH, FISHER, and NYGAARD,Circuit Judges.

(Filed January 7, 2010)

152a


Lisa M. Card, Esq.John F. Collins, Esq.Jeffrey L. Kessler, Esq. (Argued) George E. Mastoris, Esq.Eamon O�’Kelly, Esq.Susannah P. Torpey, Esq.DEWEY & LEBOEUF1301 Avenue of the AmericasNew York, NY. 10019-0000

Steven M. Williams, Esq.COHEN, SEGLIAS, PALLAS, GREENHALL & FURMAN240 North Third Street7th FloorHarrisburg, PA 17101-0000Counsel for Feesers, Inc.

James S. Ballenger, Esq.Jennifer L. Giordano, Esq.Maureen E. Mahoney, Esq.Eric J. McCarthy, Esq.Margaret M. Zwisler, Esq.LATHAM & WATKINS555 11th Street, N.W.Suite 1000Washington, DC. 20004-0000

David A. Applebaum, Esq.Robert L. DeMay, Esq.David G. Parry, Esq.LEONARD, STREET & DEINARD150 South Fifth StreetSuite 2300Minneapolis, MN. 02300-0000

153a


Roy T. Englert, Jr., Esq. (Argued)ROBBINS, RUSSELL, ENGLERT, ORSECK

& UNTEREINER1801 K Street, N.W.Suite 411Washington, DC. 20006-0000

Bridget E. Montgomery, Esq.ECKERT, SEAMANS, CHERIN & MELLOTT213 Market Street8th FloorHarrisburg, PA. 17101-0000

Alfred C. Pfeiffer, Esq.LATHAM & WATKINS505 Montgomery StreetSuite 1900San Francisco, CA 94111-0000Counsel for Michael Foods, Inc.

Scott P. Fink, Esq.Jennifer B. Furey, Esq.Martin F. Gaynor, III, Esq.COOLEY, MANION & JONES21 Custom House StreetBoston, MA 02110-0000

Thomas Demitrack, Esq.JONES DAY901 Lakeside AvenueNorth PointCleveland, OH 44114-0000

154a


Hashim M. Mooppan, Esq.Glen D. Nager, Esq. (Argued)JONES DAY51 Louisiana Avenue, N.W.Washington, DC. 20001-0000

Matthew M. Haar, Esq.SAUL EWINGTwo North Second StreetPenn National Insurance Tower, 7th FloorHarrisburg, PA 17101-0000Counsel for Sodexho, Inc.

__________OPINION__________

SMITH, Circuit Judge.This appeal by Feesers, Inc. (�“Feesers�”), a food

distributor, arises out of a Robinson-Patman Actclaim for unlawful price discrimination, 15 U.S.C.§ 13 (the �“RPA�”), against Michael Foods, Inc.(�“Michaels�”), a food manufacturer, and Sodexo,Inc. (�“Sodexo�”),1 a food service management com-pany. Feesers claims that Sodexo was able to pur-chase egg and potato products from Michaels at adiscounted price that was unavailable to Feesers.Following a bench trial, the District Court enteredjudgment for Feesers. We will vacate that judg-ment and instruct the District Court to enter judg-

155a


1 Sodexho, Inc. changed its name to Sodexo, Inc. duringthe course of this litigation. We will refer to the company byits new name.

ment as a matter of law for Michaels and Sodexo.Feesers and Sodexo were not competing pur-chasers, and, therefore, Feesers cannot satisfy thecompetitive injury requirement of a prima faciecase of price discrimination under § 2(a) of theRPA.2 In doing so, we hold that, in a secondary-

156a


2 Section 2(a) of the RPA, in relevant part, states that:It shall be unlawful for any person engaged incommerce, in the course of such commerce, eitherdirectly or indirectly, to discriminate in pricebetween different purchasers of commodities oflike grade and quality, where either or any of thepurchases involved in such discrimination are incommerce, where such commodities are sold foruse, consumption, or resale within the UnitedStates or any Territory thereof or the District ofColumbia or any insular possession or other placeunder the jurisdiction of the United States, andwhere the effect of such discrimination may besubstantially to lessen competition or tend tocreate a monopoly in any line of commerce, or toinjure, destroy, or prevent competition with anyperson who either grants or knowingly receives thebenefit of such discrimination, or with customers ofeither of them: Provided, That nothing hereincontained shall prevent differentials which makeonly due allowance for differences in the cost ofmanufacture, sale, or delivery resulting from thediffering methods or quantities in which suchcommodities are to such purchasers sold ordelivered[.] . . . . And provided further , Thatnothing herein contained shall prevent personsengaged in selling goods, wares, or merchandise incommerce from selecting their own customers inbona fide transactions and not in restraint oftrade: And provided further, That nothing hereincontained shall prevent price changes from time totime where in response to changing conditions

line price discrimination case, parties competingin a bid market cannot be competing purchaserswhere the competition for sales to prospective cus-tomers occurs before the sale of the product forwhich the RPA violation is alleged.

When reviewing a judgment entered after abench trial, we exercise �“plenary review over [the][D]istrict [C]ourt�’s conclusions of law�” and its�“choice and interpretation of legal precepts.�” Am.Soc�’y for Testing & Materials v. Corrpro Cos., 478F.3d 557, 566 (3d Cir. 2007) (internal quotationsomitted). Findings of fact are reviewed for clearerror. Id. The District Court had subject matterjurisdiction over this case under 28 U.S.C. § 1331,and we exercise appellate jurisdiction under 28U.S.C. § 1291.

Michaels and Sodexo raise a host of issues inthis appeal, but in light of this Court�’s decision inToledo Mack Sales & Service, Inc. v. Mack Trucks,Inc., 530 F.3d 204 (3d Cir. 2008), and the SupremeCourt�’s decision in Volvo Trucks North America,Inc. v. Reeder-Simco GMC, Inc., 546 U.S. 164, 126S. Ct. 860, 163 L. Ed. 2d 663 (2006), we needaddress only the issue of whether Sodexo andFeesers were �“competing purchasers�” for purposesof the RPA. Feesers, Inc. v. Michael Foods, Inc.,498 F.3d 206, 213 (3d Cir. 2007) (quoting Falls

157a


affecting the market for or the marketability of thegoods concerned, such as but not limited to actualor imminent deterioration of perishable goods,obsolescence of seasonal goods, distress salesunder court process, or sales in good faith indiscontinuance of business in the goods concerned.

15 U.S.C. § 13(a).

City Indus. v. Vanco Beverage, 460 U.S. 428, 435,103 S. Ct. 1282, 75 L. Ed. 2d 174 (1983)).3

I.

The following facts were found by the DistrictCourt after a bench trial. Feesers, Inc. v. MichaelFoods, Inc., 632 F. Supp. 2d 414, 418 (M.D. Pa.2009).

Structure of the Food Service Industry

The food service industry consists of a three-tierdistribution system: manufacturers sell productsto distributors, who resell those products to oper-ators, including self-operators (�“self-ops�”) and foodservice management companies. Id. at 420-21.Self-ops are institutions that perform all diningservices internally. Food service managementcompanies perform institutions�’ dining services fora fee, id., and primarily target schools, hospitals,and nursing homes. Sometimes operators negoti-ate with manufacturers for discounted prices,known as �“deviated prices.�” Id. at 432. In those

158a


3 Three of the four requirements of a § 2(a) Robinson-Patman claim have already been established by Feesers andare not contested in this appeal. Feesers, Inc., 498 F.3d at208. Feesers has shown �“that sales were made to two dif-ferent purchasers[, Feesers and Sodexo,] in interstate com-merce; that the product sold was of the same grade andquality; and that [Michaels] discriminated in price asbetween the two purchasers.�” Id. at 211. What remains forresolution by this Court is the fourth requirement, a show-ing �“that the discrimination had a prohibited effect on com-petition.�” Id. at 212.

instances, the distributor purchases the productat list price from the manufacturer, sells the prod-uct to the operator at the deviated price, andreceives the difference between the list price andthe deviated price from the manufacturer. Id. Anoperator may also seek discounts from manufac-turers by joining a Group Purchasing Organiza-tion (�“GPO�”). A GPO is a collection of operatorswho negotiate food prices collectively to achievegreater bargaining power against manufacturersand distributors. Id. at 421. �“GPOs generally bar-gain for a lower price, but do not actually pur-chase the food for resale to institutions.�” Id.

The Parties in this Appeal

Michaels is a manufacturer of egg and potatoproducts that sells in bulk, nationwide. Id. It isthe largest producer of liquid eggs in the UnitedStates. Id. Feesers is a regional distributor thatdistributes Michaels�’s products, and others, tooperators within a 200-mile radius of Harrisburg,Pennsylvania. Id. Sodexo is a multinational foodservice management company that serves insti-tutions around the world. Id. Its services includeplanning menus, ordering food, preparing andserving meals, and overseeing labor issues. It isthe largest private purchaser of food in the world.Id. Sodexo owns Entegra, a GPO. Id. at 427.

Michaels�’s Pricing of Food Products

Michaels sells sixty percent of its products atdeviated prices. Id. at 432. It has offered deviatedpricing to self-ops since the mid-1990s and to foodservice management companies, like Sodexo, since

159a


at least 1999. �“[O]n average from 2000 until 2004,Feesers paid $9.56, or 59% more than [Sodexo] for[Michaels�’s] eleven top selling products.�” Id. at434. This pricing difference was described as�“stunning�” by Feesers�’s expert witness. Id. Thedeviated pricing Sodexo received from Michaelswas not institution-specific, so Sodexo could �“useits low deviated price . . . to win new accounts andto keep current customers.�” Id. at 432.

Competition between Feesers and Sodexo

Feesers sells food to self-op institutions and foodservice management companies.4 Id. at 421-22.Sodexo sells food in conjunction with its food service management services. Id. at 422. Institu-tional customers �“regularly switch [between] self-op [and] management,�” and at least threeinstitutions have switched between Feesers andSodexo. Id.5 Both companies regularly seek self-opbusiness. Id. Feesers tries to distribute for self-opswhile Sodexo tries to convert self-ops to food ser-vice management.

160a


4 The District Court found that Feesers sold only toself-op institutions, Feesers, Inc., 632 F. Supp. 2d at 421-22.This finding was clearly erroneous. The District Court�’s ownfact finding describing Feesers�’s business explains thatFeesers distributed food for Wood, a food service manage-ment company. Compare id. (�“Feesers only sells food to self-op institutions[.]�”), with id. at 421 n.3 (�“Feesers was theprimary distributor for the Wood Company,�” a food servicemanagement company.).

5 The Jewish Home of Greater Harrisburg and St.Mary�’s Catholic School both switched from Feesers toSodexo. Feesers, Inc., 632 F. Supp. 2d at 422. The Meadowsswitched from Sodexo to Feesers. Id.

When a self-op switches to Sodexo, it relies onSodexo to handle all dining services functions,such as procurement and distribution of food. Id.Sodexo itself is not a distributor, but it decideswhich distributors its customers will use. Id.Thus, when an institution switches from self-op toSodexo, the incumbent distributor who distributedfor the self-op may be replaced. Id. BecauseFeesers could be displaced by Sodexo�’s chosen dis-tributor if Sodexo wins a self-op�’s business, thetwo companies compete �“when a customer con-siders switching from self-op to food service man-agement, or vice versa.�” Id. at 430.6 Accordingly,Feesers and Sodexo �“compete[d] for the same por-tion of an institution�’s food service budget.�” Id. at420.

Competition between Feesers and Sodexooccurred informally prior to the request for pro-posal (�“RFP�”) process ordinarily required by largeinstitutions.7 Id. at 428. To grow its client base,Sodexo identifies institutions that meet its clientprofile and then builds relationships with thoseinstitutions. Id. at 428-29. During informal con-tacts with a prospective institutional customer,Sodexo �“gauges the institution�’s interest in man-

161a


6 We regard this inferred fact as highly questionable,but the finding does not rise to the level of clear error. In ourview, assuming that Sodexo replaced Feesers with anotherdistributor, Feesers�’s competitor would be the other dis-tributor, not Sodexo.

7 Food service management companies compete witheach other through a formal RFP process to win institutions�’business. Feesers, Inc., 632 F. Supp. 2d at 428. The RFP pro-cess is usually limited to food service management compa-nies. Id.

agement and determines whether there are anyparticular problems to be solved.�” Id. at 428. If theinstitution is interested in management, it willthen put out a RFP and Sodexo will follow throughin that process. Id. Aside from seeking newclients, Sodexo also touts its access to discountedfoods to its existing customers that utilize it forpreparation and ordering of food, but not for dis-tribution. Id. at 429. This is done, in part, toencourage those customers to switch to Sodexo�’schosen distributor. Id.8

Procedural History

On March 17, 2004, Feesers sought a declara-tory judgment stating that (1) Michaels unlaw-fully discriminated in price under § 2(a) of theRPA by selling egg and potato products to Sodexoat significantly lower prices than it did to Feesersand (2) Sodexo violated § 2(f) of the RPA by know-ingly inducing those discriminatory sales. 15U.S.C. § 13(a) and (f). Feesers also sought per-manent injunctive relief under § 16 of the ClaytonAct. 15 U.S.C. § 26. On May 4, 2006, the DistrictCourt granted summary judgment for the defen-dants, concluding that Feesers had satisfied the

162a


8 The District Court also identified other evidenceshowing competition between Feesers and Sodexo, includingSodexo�’s SEC filings, Feesers, Inc., 632 F. Supp. 2d at 422,and its internal strategic documents, id. at 423. None of thisevidence stated that Sodexo regarded any distributor as acompetitor. Id. at 422-23 (noting that Sodexo�’s SEC filingsidentified lower overall costs of food service management asa means of promoting itself over self operation); id. at 425(�“Sodex[o]�’s strategic planning documents do not specificallymention distributors as competitors[.]�”).

first three elements of a prima facie case of pricediscrimination, but not the fourth element, com-petitive injury. �“The District Court was concernedthat [Sodexo] and Feesers [we]re not at the same�‘functional level�’ and [we]re therefore not in�‘actual competition�’ in the same market.�” Feesers,Inc., 498 F.3d at 214.

Feesers appealed and this Court reversed.9 Weheld that the District Court had applied the wrongstandard in concluding that Feesers and Sodexowere not in competition. Id. at 208. The panelexplained the proper standard and remanded thecase to the District Court for further proceed-ings.10

On remand, after a bench trial, the DistrictCourt entered judgment for Feesers and enjoinedMichaels from engaging in unlawful price dis-crimination. Michaels then suspended all sales toFeesers. In response, Feesers sought an order ofcontempt and a permanent injunction forbidding

163a


9 We reversed in a 2-1 decision. The dissent concludedthat Sodexo and Feesers were not in actual competitionbecause they �“d[id] not sell the same products.�” Feesers, Inc.,498 F.3d at 220 (Jordan, J., dissenting). In reaching thatconclusion, the dissent explained that Feesers �“res[old] . . .unprepared foods to its institutional clients,�” whereasSodexo �“prepare[d] meals, and s[old] the prepared meals toindividual customers.�” Id. at 218. (Jordan, J., dissenting).The majority disagreed, noting that �“a factfinder could con-clude that Sodex[o] s[old] unprepared food to its customers�”because some of Sodexo�’s agreements with institutionalclients did not charge for �“�‘prepared meals,�’ but rather forthe cost of unprepared food and supplies, the cost of labor,and a management fee.�” Id. at 215.

10 The prior decision is explained in Section IV(A),infra.

Michaels from refusing to deal with Feesers. OnMay 26, 2009, the District Court held Michaels incontempt and enjoined it from refusing to �“sell itsproducts to Feesers on the same terms as they aresold to [Sodexo], so long as Feesers otherwisemeets its standards as a customer.�” Michaels andSodexo now appeal the District Court�’s judgmentand the permanent injunction.

II.

�“�‘Competitive injury�’ [under § 2(a) of the RPA] isestablished . . . by proof of �‘a substantial price dis-crimination between competing purchasers overtime.�’�” Feesers, Inc., 498 F.3d at 213 (quoting FallsCity Indus., 460 U.S. at 435) (footnote omitted)(emphasis in original). �“Feesers does not need toprove that [Michaels�’s] price discrimination actu-ally harmed competition, i.e., that the discrimi-natory pricing caused Feesers to lose customers toSodex[o]. Rather, Feesers need prove only that (a)it competed with Sodex[o] to sell food and (b)[that] there was price discrimination over time by[Michaels].�” Feesers, Inc., 498 F.3d at 213 (foot-note omitted) (emphasis in original).

To determine whether Feesers competed withSodexo to sell food, �“the relevant question iswhether [the] two companies �‘[we]re in economicreality acting on the same distribution level.�’�” Id.at 214 (quoting Stelwagon Mfg. Co. v. TarmacRoofing Sys., Inc., 63 F.3d 1267, 1272 (3d Cir.1995)). Recognizing that the phrase �“economicreality�” provides little guidance in how toapproach the competition inquiry, this Court, inthe prior appeal in this case, explained that two

164a


parties are in competition only where, after a�“careful analysis of each party�’s customers,�” wedetermine that the parties are �“each directly afterthe same dollar.�” Feesers, Inc., 498 F.3d at 214(quoting M.C. Mfg. Co. v. Tex. Foundries, Inc., 517F.2d 1059, 1068 n.20 (5th Cir. 1975)). We refer tothis dollar-for-dollar analysis as the competingpurchaser requirement. The Supreme Court�’sguidance in Volvo Trucks, 546 U.S. at 179-80, andthis Court�’s precedent in Toledo Mack, 530 F.3d at226-29, compel us to conclude that Feesers andSodexo were not competing purchasers. Thus,Feesers cannot satisfy the first element requiredto show competitive injury, and its RPA claimsmust fail as a matter of law.11

A.

In application, the competing purchaser require-ment will vary based on the nature of the marketand the timing of the competition. In a bid mar-ket, if the competition between the favored anddisfavored purchaser occurs before the purchase ofthe goods from the seller, then the disfavored pur-chaser cannot show that it and the favored pur-chaser were competing purchasers. Volvo Trucks,546 U.S. at 178-79. This rule prevents the appli-cation of the RPA to markets where the �“allegedlyfavored purchasers [bear] little resemblance to[the] large independent department stores orchain operations�” that the RPA was intended to

165a


11 Because Feesers cannot satisfy the first elementrequired to show competitive injury, we need not discusswhether it experienced price discrimination over time.

target, id. at 181, and helps �“construe the [RPA]�‘consistently with the broader policies of theantitrust laws,�’�” id. (quoting Brooke Group v.Brown & Williamson Tobacco Corp., 509 U.S. 209,220, 113 S. Ct. 2578, 125 L. Ed. 2d 168 (1993)).

In practice, the rule, like other restrictions onthe reach of the RPA, prevents the unprincipledapplication of the statute. Indeed, because theRPA often has �“anticompetitive�” effects that �“promote rather than . . . prevent monopolisticpricing practices,�” Small Business and the Robin-son-Patman Act: Hearings before the Special Subcommittee on Small Business and the Robin-son-Patman Act of the House Select Committee onSmall Business, 91st Cong. 146-47 (1969) (testi-mony of Richard A. Posner), the Supreme Court,in seeking to construe the statute consistentlywith the broader policies of the antitrust laws, hasrepeatedly limited its reach by:

. Expanding the means through which RPAdefendants can attack the �“competition�”element of a prima facie case of price dis-crimination, Texaco v. Hasbrouck, 496U.S. 543, 561, 110 S. Ct. 2535, 110 L. Ed.2d 492 (1990) (�“A supplier need not satisfythe rigorous requirements of the cost jus-tification defense in order to prove that aparticular functional discount is reason-able and accordingly did not cause anysubstantial lessening of competitionbetween a wholesaler�’s customers and thesupplier�’s direct customers.�”) (footnoteomitted);

166a


. Focusing the competition inquiry on�“interbrand competition,�” Volvo Trucks,546 U.S. at 180;

. Explaining that the RPA does not �“ban allprice differences charged to different pur-chasers of commodities of like grade andquality,�” id. at 176 (quoting Brooke GroupLtd., 509 U.S. at 220);

. �“[R]esist[ing] interpretation[s] [of theRPA] geared more to the protection ofexisting competitors than to the stimula-tion of competition,�” Volvo Trucks, 546U.S. at 181 (emphasis omitted); and,

. �“[R]ecogni[zing] [that] the right of a sellerto meet a lower competitive price in goodfaith may be the primary means of recon-ciling the [RPA] with the more generalpurposes of the antitrust laws,�” Great Atl.& Pac. Tea Co. v. FTC, 440 U.S. 69, 82n.16, 99 S. Ct. 925, 59 L. Ed. 2d 153(1979) (interpreting RPA to provide robustmeeting competition defense).12

167a


12 Accord Falls City Indus., 460 U.S. at 451-52 (vacat-ing judgment that defendant did not have a meeting com-petition defense); Cass Sunstein, Interpreting Statutes in theRegulatory State, 103 Harv. L. Rev. 405, 487 (1989) (�“Courtsshould narrowly construe statutes that serve no plausiblepublic purpose, and amount merely to interest-group trans-fers . . . . Th[is] idea helps explain a number of decisions inareas of economic regulation, such as . . . the courts�’approach to the Robinson-Patman Act.�”) (citing Robert H.Bork, The Antitrust Paradox: A Policy at War with Itself 409-10 (1978)).

This Court has dutifully followed the SupremeCourt�’s lead by narrowly construing the RPA. InToledo Mack, we explained that we will �“narrowlyinterpret�” the RPA, even if doing so will result in�“elevat[ing] form over substance.�” Toledo Mack,530 F.3d at 228 n.17.

While the competing purchaser requirement hasits roots in FTC v. Morton Salt, 334 U.S. 37, 46-51, 68 S. Ct. 822, 92 L. Ed. 1196, 44 F.T.C. 1499(1948), the most recent decisions discussing thatrequirement are Volvo Trucks and Toledo Mack.Both decisions emphasized that proving �“sub-stantial price discrimination between competingpurchasers overtime,�” Volvo Trucks, 546 U.S. at179 (quoting Falls City Indus., 460 U.S. at 435)(emphasis omitted), requires accounting for thetiming of the alleged competition and the natureof the market. Volvo Trucks, 546 U.S. at 178-79;see Toledo Mack, 530 F.3d at 228.

In Volvo Trucks, the Supreme Court rejected aninference of competitive injury where the plaintiff,Reeder-Simco (�“Reeder�”), could not show that itwas a competing purchaser. Volvo Trucks, 546U.S. at 179-80. Reeder was a Volvo dealer whocompeted with other dealers (both Volvo brandand others) through a customer-specific biddingprocess for sales to individuals seeking custom-built trucks. Id. at 169. Reeder alleged that Volvosold trucks to other Volvo dealers at unlawfullydiscriminatory prices, giving those other dealersan unfair advantage in selling to prospective cus-tomers. The customer-specific bidding processbegan with the customer stating its specificationsand inviting bids from dealers it had selected. Id.at 170. The selected dealers would submit bids to

168a


the customer and the dealer that won the bidwould arrange for the manufacturer, in this case,Volvo, to build the truck for the customer. Id. Likethe deviated pricing system of food manufacturers,it was common for truck manufacturers to offer�“customer-specific discounts to their dealers.�” Id.Prior to submitting a bid to a customer, a Volvodealer would ask Volvo if it could get a discountfor the customer. Id. Volvo would then decide on acase-by-case basis what discount it would grant aparticular customer based on factors like indus-try-wide demand and whether the customer hadpreviously purchased from Volvo. Id. While thediscount varied based on many factors, the dealersalways knew what discounts they could offer acustomer before submitting their bids to the cus-tomer. See id.

The specific question presented in the case waswhether �“a manufacturer offering its dealers dif-ferent wholesale prices may be held liable forprice discriminations proscribed by Robinson-Pat-man, absent a showing that the manufacturer dis-criminated between dealers contemporaneouslycompeting to resell to the same retail customer.�”Id. at 169. In deciding that question in the nega-tive, the Supreme Court concluded that Reedercould not establish an inference of competitiveinjury based on the timing of the competitionbetween the dealers and the nature of the market,Reeder�’s evidence of competitive injury, and thegoals of the RPA.

The timing of the competition between the deal-ers and the nature of the market were critical tothe Supreme Court�’s reasoning. At the initialstage of competition in the bid market, where

169a


dealers were competing to win the right to submita bid to a customer, �“competition [wa]s notaffected by differential pricing [because] a dealerin the competitive bidding process approach[ed]Volvo for a price concession . . . only after it ha[d]been selected by a retail customer to submit abid.�” Id. at 178-79. Prospective customers chosewhich dealers could submit bids based on a vari-ety of factors �“including the existence vel non of arelationship between the potential bidder and the[prospective] customer, geography, and reputa-tion.�” Id. at 179. After the prospective customerchose who could submit bids, the relevant marketnarrowed to the few dealers who were chosen:�“Once a retail customer has chosen the particulardealers from which it will solicit bids, �‘the rele-vant market becomes limited to the needs anddemands of a particular end user, with only ahandful of dealers competing for the ultimatesale.�’�” Id. (quoting Reeder-Simco GMC, Inc. v.Volvo GM Heavy Truck Corp., 374 F.3d 701, 719(8th Cir. 2004) (Hansen, J., dissenting)).

The Supreme Court was also unimpressed withReeder�’s evidence purporting to show competitiveinjury. Reeder produced three types of evidence tosupport its allegations.13 The two types of evidence

170a


13 The first type, evidence of head-to-head comparisonsbetween Reeder and other Volvo dealers, is not relevant inthe instant case. Reeder�’s evidence showed only �“twoinstances over [a] five year course�” where it bid againstother Volvo dealers, so called head-to-head comparisons: Oneinstance where it and another Volvo dealer received thesame discount and Reeder lost to the other dealer becausethe customer had previously bought from the other dealer,and one instance where it and the opposing Volvo dealer

relevant to the instant case were Reeder�’s �“com-parisons of [discounts] [it] received for four suc-cessful bids against non-Volvo dealers, with larger[discounts] other successful Volvo dealers receivedfor different sales on which [it] did not bid (pur-chase-to-purchase comparisons),�” Volvo Trucks,546 U.S. at 177 (emphasis in original), and �“com-parisons of [discounts] offered to [it] in connectionwith several unsuccessful bids against non-Volvodealers, with greater concessions accorded otherVolvo dealers who competed successfully for dif-ferent sales on which [it] did not bid (offer-to-pur-chase comparisons),�” id. at 177-78 (emphasis inoriginal). These two types of evidence did not cre-ate an inference of competitive injury because (1)the alleged price discrimination did not occur forthe same customer and (2) Reeder did not attemptto show that other Volvo dealers were consistentlyfavored. Id. at 178.14

171a


received matching discounts from Volvo and neither won thebid. Volvo Trucks, 546 U.S. at 172. On this evidence, theSupreme Court noted that Reeder showed the �“loss of onlyone sale�” and that �“Reeder and the other dealer received thesame concession�” in that instance. Id. at 180. In the otherinstance of head-to-head competition, both Volvo dealersreceived the same concession and neither won the bid. Id.The Supreme Court concluded that �“if price discriminationbetween two purchasers existed at all, [a sale that wouldhave resulted in $30,000 more in gross profits for Reeder]was not of such magnitude as to affect substantially com-petition between Reeder and the �‘favored�’ Volvo dealer.�” Id.

14 Notably, in this case, Feesers produced evidenceshowing that Michaels consistently favored Sodexo. Feesers,Inc., 632 F. Supp. 2d at 434. This type of evidence was notproduced in Volvo Trucks, so the Supreme Court neverexplained whether both or only one of its reasons for reject-

The Supreme Court also signaled that it wasuninterested in permitting innovative applicationsof the RPA and would resist �“interpretation[s]geared more to the protection of existing com-petitors than to the stimulation of competition.�”Id. at 181 (emphasis in original). It also notedthat the custom truck market bore �“little resem-blance to [the] large independent departmentstores or chain operations�” that the RPA originallyintended to target. Id.

This Court used similar reasoning in ToledoMack. 530 F.3d at 226-29. That case had factssimilar to Volvo Trucks�—Toledo, a Mack truckdealer, would submit bids to prospective cus-tomers who wished to purchase customized Macktrucks. Id. at 209. In creating a bid, Toledo wouldseek out a �“transaction-specific discount [fromMack] known as �‘sales assistance.�’�” Id. �“Theamount of sales assistance [Mack offered] varie[d]according to the nature of the relationshipbetween the dealer and the customer, the numberof trucks ordered, potential competition, and otherfactors.�” Id. Toledo sued Mack under the RPA,claiming that it consistently received less salesassistance than other Mack dealers.

In affirming the district court�’s grant of sum-mary judgment for the defendant, Mack, on theRPA claim, this Court explained that the timing of

172a


ing the inference of competitive injury need be rectified inorder to infer competitive injury. 546 U.S. at 179 n.3. As welater explain, this Court, in Toledo Mack, rejected the argu-ment that evidence showing that a certain purchaser wasconsistently favored was sufficient to infer competitiveinjury in a bid market where the competition occurred priorto the actual sale. Toledo Mack, 530 F.3d at 228-29.

competition and the nature of the market are crit-ical factors to consider when determining whetherthe plaintiff can show that it was a competing pur-chaser of a favored purchaser. We concluded thatbecause the competition between Mack dealersoccurred during the bidding process, and not atthe time of the actual sale, Toledo could not sat-isfy the competing purchaser requirement or thetwo purchaser requirement:

Because no sale takes place until a cus-tomer accepts a dealer�’s bid, the amountof sales assistance Mack is willing to pro-vide to a particular dealer is part of anoffer by Mack to sell, not a sale. Regard-less of any competition between the deal-ers during the bidding process, only adealer whose bid is accepted by a cus-tomer will actually buy a truck fromMack. Therefore, only one sale, not two,actually results.

Id. at 228.Toledo, unlike the plaintiff in Volvo Trucks, did

not offer evidence of head-to-head competitionbetween it and other Mack dealers. Id. at 215. Butit did provide expert testimony regarding �“theaverage amounts of sales assistance Mack offeredto Toledo as compared with the average amount ofsales assistance Mack offered to other [Mack]dealers,�” i.e., evidence showing that Mack con-sistently favored other dealers as compared toToledo. Id. That evidence was rejected by thisCourt as irrelevant because even if the �“amount ofsales assistance Mack offer[ed] to each dealer . . .determine[d] whether a customer cho[se] to accept

173a


a bid from one Mack dealer or another, Mack doesnot sell a truck to the dealer until the customeractually selects a dealer�’s bid.�” Id. at 228. Thus,only one sale, not two, resulted from the compe-tition. Id. This was true in part because the salewas divorced from the competition and Toledocould not show that it was a competing purchaservis-a-vis other Mack dealers. See id.

Finally, the Toledo Mack Court noted that, likeVolvo Trucks, �“the alleged price discriminationd[id] not implicate the original purpose of the RPAbecause �‘the allegedly favored purchasers [we]redealers with little resemblance to large indepen-dent department stores or chain operations.�’�” Id.at 227 (quoting Volvo Trucks, 546 U.S. at 181).

B.

While this Court�’s conclusion in Toledo Mackundoubtedly turned on the fact that �“one sale, nottwo, actually result[ed],�” Toledo Mack, 530 F.3d at228, it was not reached by a simple application ofthe RPA�’s two purchaser requirement. It wasreached through the combined effect of the RPA�’stwo purchaser and competitive injury require-ments�—i.e., the competing purchaser require-ment. Id.; see Falls City Indus., 460 U.S. at 435(explaining competing purchaser requirement);Volvo Trucks, 546 U.S. at 179 (same).

In Toledo Mack we held that because the com-petition among dealers for prospective customerbusiness occurred before the purchase of the truckto be sold to the customer by the winning dealer,the relevant market for the sale to the customerwas already limited to one at the time the manu-

174a


facturer sold the dealer the truck. See ToledoMack, 530 F.3d at 228. Because the relevant mar-ket was only one dealer making one purchase fromthe manufacturer for resale to one customer, thetwo purchaser requirement could not be satisfied.See id. Thus, this Court rejected Toledo�’s RPAclaim for lack of two purchasers, which was basedon the lack of a competitive market, i.e. the lack ofa competing purchaser. See id. This conclusioncomports with M.C. Mfg., 517 F.2d at 1065, thedecision relied upon by this Court in Toledo Mack,530 F.3d at 228, and in this Court�’s prior decisionin this case, Feesers, Inc., 498 F.3d at 214(instructing the District Court to apply the FifthCircuit�’s test to �“determine whether Sodex[o] andFeesers compete to resell food products to thesame group of customers�”) (citing M.C. Mfg. Co.,517 F.2d at 1068 n.20). In addition, the SupremeCourt�’s reasoning in Volvo Trucks further con-firms our understanding of the competing pur-chaser requirement.

In M.C. Mfg., two companies, Universal andH/R, manufactured lifting plugs for sales to thegovernment. M.C. Mfg., 517 F.2d at 1061. Bothcompanies purchased �“unfinished plug castings�”from Texas Foundries and those castings wereused to create the lifting plugs. Id. Both compa-nies would purchase castings after they had wona contract with the government. Id. at 1067. In itscomplaint, Universal alleged that H/R and TexasFoundries violated the RPA because (1) TexasFoundries quoted a lower price to H/R than Uni-versal for their respective bids for a governmentcontract and H/R won that contract (the �“1971Contract�”), id. at 1061-62, 1066-67, and (2) Texas

175a


Foundries sold unfinished plug castings to Uni-versal at a higher price in a separate contract (the�“1970 Contract Extension�”). Id. at 1065-66. Uni-versal argued that the prices it received in the1970 Contract Extension were unlawfully dis-criminatory as compared to the prices H/Rreceived in the 1971 Contract. In doing so, Uni-versal�’s allegations appeared to satisfy the twopurchaser requirement because the two companieswere both purchasing the same type of unfinishedplug casting from Texas Foundries. This appear-ance, however, was misleading because the con-tracts from which Universal�’s purported injuriesflowed were distinct markets open only to a singleproducer. See id. at 1067. H/R and Universal werenot competing purchasers because the 1970 Con-tract Extension and the 1971 Contract each �“rep-resented a separate, distinct market open only toa single producer.�” Id. �“The very nature of th[o]semutually exclusive commitments in the respectivecontracts meant that Universal and H/R could nothave been �‘in competition�’ with respect to theirseparate purchases from Texas Foundries pur-suant to the government contracts.�” Id. �“There-fore, while the price discrepancy between [TexasFoundries�’s sales to H/R under the 1971 Contractand to Universal under the 1970 Contract Exten-sion] could have affected Universal�’s profits underthe [1970 Contract Extension], this discriminationin no way diminished Universal�’s competitive abil-ity in that plug market.�” Id. Thus, even though�“Universal and H/R were competitive bidders onthe 1971 [C]ontract[, t]hey could not be . . . com-petitive purchasers as required by the Act either

176a


under their respective separate contracts or underboth.�” Id.15

Similar reasoning was also invoked in VolvoTrucks. 546 U.S. at 178. There, the SupremeCourt discounted the purchase-to-purchase andoffer-to-purchase evidence offered by Reeder inpart because that evidence did not show thatReeder competed �“with beneficiaries of the allegeddiscrimination for the same customer.�” Id. (empha-sis in original). �“That Volvo dealers may bid forsales in the same geographic area�” was of noimport to the Supreme Court because that factwas not relevant to whether two dealers �“com-pet[ed] for the same customer-tailored sales.�” Id.at 179. �“Once a retail customer has chosen theparticular dealers from which it will solicit bids,�‘the relevant market becomes limited to the needsand demands of a particular end user, with only ahandful of dealers competing for the ultimatesale.�’�” Id. (quoting Reeder-Simco GMC, Inc., 374F.3d at 719 (Hansen, J., dissenting)).

Accordingly, we reject the argument that ToledoMack was a simple application of the two pur-chaser requirement. Implicit in the Toledo MackCourt�’s holding was the conclusion that Toledocould not show it was a competing purchaser ofother Mack truck dealers. In other words, the twopurchaser requirement could not be satisfied

177a


15 See M.C. Mfg., 517 F.2d at 1066 (�“Even if the sales atdifferent prices are contemporaneous, involve goods of likegrade and quality, the price distinction is not justified bygood business cause, and it causes injury to the disadvan-tage[d] purchaser, recovery under the Act is precludedabsent proof that the price variance detrimentally affectedcompetition.�”).

because the relevant market of competition waslimited to one dealer, one customer, and one truckmanufacturer at the time of the sale of the truck,i.e., there were no competing purchasers. ToledoMack, 530 F.3d at 228.

C.

Applying the teachings of Volvo Trucks andToledo Mack to the instant case, it is clear thatFeesers never experienced a competitive injuryfrom Sodexo�’s purchases and sales of Michaels�’sproducts because Feesers and Sodexo were notcompeting purchasers. See Volvo Trucks, 546 U.S.at 179. The competition between Feesers andSodexo for institutions�’ business occurred prior toMichaels�’s sales of food products to Feesers andSodexo, �“when a customer consider[ed] switchingfrom self-op to food service management, or viceversa.�” Feesers, Inc., 632 F. Supp. 2d at 430. Atthat time, Sodexo would not yet have secured anyproducts from Michaels for resale to the prospec-tive customer because the customer would only bedeciding whether it wished to begin the RFP pro-cess or, if it had already chosen to engage in theRFP process, whether to invite Sodexo to partici-pate in that process. Once the customer has cho-sen whether to self-operate or contract with a foodservice management company, �“the relevant mar-ket becomes limited to the needs and demands ofa particular end user, with only a handful of [dis-tributors or food service management companies]competing for the ultimate sale.�” Volvo Trucks,546 U.S. at 179; Toledo Mack, 530 F.3d at 228.Thus, Feesers and Sodexo�’s competition at that

178a


early stage was irrelevant to the sales made byMichaels after that competition was complete. Ifan institution chose to self-operate, Sodexo wouldbe eliminated from the competition, and if aninstitution chose to contract with a food servicemanagement company, Feesers would be elimi-nated from the competition. After making that ini-tial decision, the customer then has to choosewhich distributor or food service managementcompany it will hire. Only after that process iscomplete would the customer then actually pur-chase food from Michaels through the winning dis-tributor or food service management company.

At all events, assuming Feesers and Sodexoengaged in head-to-head competition, and the dis-counts granted by Michaels to the two companiesdetermined from which company an institutionwould purchase Michaels�’s products, the compet-ing purchaser requirement would still not be sat-isfied because Michaels does not make a sale untilthe institution chooses a particular distributor orfood service management company and thenbegins purchasing Michaels�’s products throughthat company. See Toledo Mack, 530 F.3d at 228.The relevant market at the time of the sale ofMichaels�’s products will have already been nar-rowed to one�—the company that won the institu-tion�’s business. See id.

While the timing of the competition and thenature of the market compel us to conclude thatFeesers and Sodexo were not competing pur-chasers, it is also relevant that the evidence pro-duced by Feesers was the same type of average

179a


discount evidence produced in Toledo Mack.16

Compare Toledo Mack, 530 F.3d at 215 (plaintiffproducing evidence comparing �“the averageamount of sales assistance�” received by the plain-tiff as compared to other Mack truck dealers),with Feesers, Inc., 632 F. Supp. 2d at 434 (plaintiffproducing evidence showing that Sodexo consis-tently received �“stunning�” price discounts thatamounted to a 59% difference in prices betweenFeesers and Sodexo over four years). The ToledoMack Court rejected such evidence as insufficientto prove injury to competition in part because�“merely offering lower prices to a customer doesnot give rise to a price discrimination claim.�”Toledo Mack, 530 F.3d at 227-28 (citing Cross-roads Cogeneration Corp. v. Orange & RocklandUtils., Inc., 159 F.3d 129, 142 (3d Cir. 1998)). Aplaintiff must also show that the effect of thelower prices was to injure competition. ToledoMack, 530 F.3d at 228 (citing Crossroads Cogen-eration Corp., 159 F.3d at 142); Volvo Trucks, 546U.S. at 181. Yet that showing is impossible where,

180a


16 The discount schemes in Volvo Trucks and ToledoMack were largely indistinguishable from the deviated pric-ing system used in the food manufacturer industry. SeeToledo Mack, 530 F.3d at 209-10 (explaining that requestsfor sales assistance to Mack occurred prior to submission ofbid to customer); Volvo Trucks, 546 U.S. at 170-71 (same).Food service management companies, self-ops, and GPOs,like the truck dealers in Toledo Mack and Volvo Trucks deal-ing with manufacturers, availed themselves of deviated pric-ing arrangements with food manufacturers. In general, theseentities know the discount they will receive before they purchase products from manufacturers. Food service man-agement companies can adjust their bid to a prospective cus-tomer to incorporate these deviated pricing arrangements.

as here, the case involves sales via a bidding pro-cess and the competition occurs before the biddingprocess even begins. See Toledo Mack, 530 F.3d at228.

In addition, the Supreme Court�’s directive tonarrowly construe the RPA to address the basicpurposes of the statute further informs our con-clusion that Feesers was not a competing pur-chaser of Sodexo. Volvo Trucks, 546 U.S. at180-81; see Toledo Mack, 530 F.3d at 227. Theprice discrimination identified by Feesers bears�“little resemblance to [the] large independentdepartment stores and chain operations�” thestatute was originally intended to target. ToledoMack, 530 F.3d at 227 (quoting Volvo Trucks, 546U.S. at 181). Here, like in Volvo Trucks, there is amyriad of differences between retail stores andfood service management companies and food dis-tributors.

First, in many respects, Sodexo and Feesers donot compete. Sodexo prepares and sells meals andhandles all dining service functions for its cus-tomers. Feesers only distributes food. Competingretail stores, in contrast, generally compete to sellfungible goods to the same group of customers.Second, Sodexo operates in a bid market withother food service management companies, andcompetes with Feesers only in a preliminary stagewhere a prospective customer is deciding whetherto self-operate or hire a food service managementcompany. Retail stores compete over prospectivecustomers every time a customer decides to pur-chase a product, and those purchases are notmade in a bid market. Third, Sodexo competes forcustomers with Feesers prior to purchasing food

181a


from Michaels. Retail stores generally purchaseproducts from manufacturers and then competewith other retailers based on pricing.

In sum, because any competition betweenFeesers and Sodexo occurred at the time an insti-tution was deciding whether to self-operate or hirea food service management company, and anyresulting sale of Michaels�’s products would haveto occur after that competition, Feesers cannotshow that it was a competing purchaser of Sodexo.The evidence produced by Feesers only furtherconfirms the futility of its RPA claims, becausesuch evidence�—evidence showing consistent favor-ing of another purchaser over the plaintiff overtime by a manufacturer in a bid market�—wasrejected in Toledo Mack. Such evidence cannotsupport an inference of competitive injury in a bidmarket. Finally, the Supreme Court�’s instructionsto narrowly construe the RPA also compel us toreject Feesers�’s RPA claims.

III.

The District Court, after thoughtful considera-tion of the Volvo Trucks and Toledo Mack deci-sions, determined that those decisions were notcontrolling for three reasons: (1) Volvo Trucksinvolved only formal competition whereas theinstant case involves formal and informal compe-tition; (2) application of Toledo Mack to theinstant case would misconstrue that decision�’sholding by imposing a new requirement under theRPA, divorced from the statutory text, that themanufacturer�’s sale of the commodity to two dif-ferent sellers occur prior to the competition for the

182a


resale of those goods; and (3) a logical reading ofToledo Mack limits that decision�’s applicability tocustom-manufactured goods. We reject each ofthese reasons in turn.

The District Court reasoned that because �“[f]oodservice management companies, distributors, andGPOs all compete formally and informally for thesale of food to institutions,�” the instant case wasdistinguishable from Volvo Trucks, which itbelieved involved only a formal bidding process.Feesers, Inc., 632 F. Supp. 2d at 431. Contrary tothe District Court�’s belief, the market in VolvoTrucks involved both formal and informal compe-tition. In that case, a customer�’s decision torequest a bid from a particular dealer was basedon informal competitive factors such as �“an exist-ing relationship, . . . reputation, and cold callingor other marketing strategies initiated by indi-vidual dealers.�” Volvo Trucks, 546 U.S. at 170(internal quotation omitted) (emphasis added).Sodexo�’s actions were indistinguishable from theactions of truck dealers in Volvo Trucks. Sodexocompeted for institutions�’ business through theformal RFP process, and through �“informal con-tacts with targeted institutions.�” Feesers, Inc., 632F. Supp. 2d at 428.

The District Court�’s second reason, that con-struing Toledo Mack to apply to the instant casewould require imposing a new requirement underthe RPA that the sale of the commodity by themanufacturer to two different sellers occur priorto the competition for resale of those goods, is amisunderstanding of the competing purchaser

183a


requirement.17 The rule the District Courtdescribes is not new�—it is simply the product ofthe competing purchaser requirement, which con-siders the relevant market, a bid market, and thetiming of the competition, before the sale to themanufacturer. The M.C. Mfg. Court explained thatthere is a �“competitive purchaser�” requirementinherent in the �“two purchaser�” and �“competitiveinjury�” elements. M.C. Mfg., 517 F.2d at 1067; seeVolvo Trucks, 546 U.S. at 179; Toledo Mack, 530F.3d at 228. In Feesers�’s prior appeal, weembraced that approach to the competing pur-chaser requirement by stating that Sodexo andFeesers compete only if �“they are each directlyafter the same dollar.�” Feesers, Inc., 498 F.3d at

184a


17 That being said, the District Court�’s desire to avoidmisapplying our precedent in this complicated area of law iscommendable. Indeed, this is not the first time the RPA hasflummoxed the federal courts, nor, barring a repeal of thelaw, will it be the last. Compare, e.g., Van Dyk ResearchCorp. v. Xerox Corp., 631 F.2d 251, 255 n.2 (3d Cir. 1980)(asserting in dicta that failure to prove the �“fact of injury�”can conclusively bar injunctive relief) (citing Merit Motors,Inc. v. Chrysler Corp., 569 F.2d 666, 668 n.2, 670 n.14, 187U.S. App. D.C. 11 (D.C. Cir. 1977)), with Feesers, Inc., 498F.3d at 213 (explaining that plaintiff need not prove actualharm to competition to receive injunctive relief) (citing FallsCity Indus., 460 U.S. at 435). The RPA places the federalcourts in an inescapable Catch-22. We are asked to apply theRPA, a statute that �“is fundamentally inconsistent with theantitrust laws,�” Antitrust Modernization Commission,Report and Recommendations 312 (2007), in a fashion thatis �“consistent[ ] with the broader policies of the antitrustlaws.�” Volvo Trucks, 546 U.S. at 181 (quoting Brooke Group,509 U.S. at 220). This conundrum is bound to create confu-sion for judges called upon to apply the RPA in a host of settings.

214 (quoting M.C. Mfg., 517 F.2d at 1068 n.20).We now hold that, simply put, Feesers and Sodexocannot compete for the same dollar because theirresales of Michaels�’s products to institutions, bytheir �“very nature[, were] mutually exclusivecommitments.�” M.C. Mfg., 517 F.2d at 1067.18 TheRPA does not ordinarily protect competition where�“a product subject to special order is sold througha customer-specific bidding process.�” VolvoTrucks, 546 U.S. at 170 (contrasting such compe-tition with �“competition between different pur-chasers for resale of [a] purchased product�”). Inother words, the RPA was not meant to cover thetype of competition present in the instant case.

Third, the District Court reasoned that a logicalreading of Toledo Mack limited that decision�’sapplicability to custom-manufactured goods. Thisconclusion is refuted by the Toledo Mack Court�’sreliance on the M.C. Mfg. decision, Toledo Mack,530 F.3d at 228 (citing M.C. Mfg., 517 F.2d at1065 (manufacturing generic product)), and thisCourt�’s explicit guidance to apply the principles ofM.C. Mfg. to this action. Feesers, 498 F.3d at 214(citing M.C. Mfg., 517 F.2d at 1068 n.20). More-over, there is no reason to limit the reach of theToledo Mack decision to customized goods becausethe underlying principles, pertaining to the timingof the competition and the nature of the market,remain the same whether applied to generic goods

185a


18 Notably, we do not hold that the sales of products bythe manufacturer to two purchasers must always occur priorto the competition between the two purchasers. Our holdingis limited to bid markets that closely resemble the marketsin this case, Volvo Trucks, and Toledo Mack.

or customized goods. This Court�’s directive to�“narrowly interpret the oft-questioned RPA�” alsosupports rejection of the District Court�’s view.Toledo Mack, 530 F.3d at 228 n. 17. A narrowinterpretation, one that limits the applicability ofthe Act, calls for taking an expansive view ofToledo Mack�’s holding and not limiting it to cus-tomized goods. See id. Finally, even if the ToledoMack decision was limited to customized goods,Sodexo offers Michaels�’s food products as part of acustomized service to customers. Feesers, Inc., 632F. Supp. 2d at 428 (finding that Sodexo sometimes�“determines whether there are any particularproblems to be solved [at an institution]�”).19 Pre-sumably, problems vary across institutions so theproposed solutions for any given institution wouldbe tailored to that institution�’s needs. In fact, themere existence of a formal RFP process shows that

186a


19 For example, in a proposal to the Beth Sholom Houseof Eastern Virginia, Sodexo urged the institution to utilizeits food procurement program to �“take full advantage of[Sodexo�’s] kosher vendors.�” Feesers, Inc., 632 F. Supp. 2d at429. The proposal states that using Sodexo�’s kosher vendorswould �“streamline the ordering process [and] substantiallyreduce pricing�” for the institution. Id. Kosher food pur-chasing is an institution-specific requirement and thus is acustomized offering. The same would be true for hospitals,which utilize lengthy RFP processes to confirm that all thespecial needs of the hospital are met by the food servicemanagement company. In fact, the foods ordered for any par-ticular institution would depend on the �“size and type ofinstitution�” and may include �“bids on a wide range of ser-vices.�” Id. at 428. Sodexo also enters into profit and loss con-tracts where �“[it] offers a financial guarantee that the diningservices will not lose money, and the institution shares in acertain percentage of the profits.�” Id. at 442.

institutions require customized contracts to servetheir specific needs.

IV.

Having determined that Feesers and Sodexowere not competitors, three outstanding issuesremain. First, whether this Court�’s holding isbarred by the law of the case. Second, how thisCourt�’s holding will affect the existing permanentinjunction ordered by the District Court. Third,the effect of concluding that Feesers cannot provea § 2(a) claim against Michaels on the § 2(f) claimagainst Sodexo. We discuss each of the issues inturn.

A.

The �“law of the case . . . doctrine posits thatwhen a court decides upon a rule of law, that deci-sion should continue to govern the same issues insubsequent stages in the same case.�” Arizona v.California, 460 U.S. 605, 618, 103 S. Ct. 1382, 75L. Ed. 2d 318 (1983). The �“doctrine does notrestrict a court�’s power but rather governs itsexercise of discretion.�” Pub. Interest ResearchGroup of N.J., Inc. v. Magnesium Elektron, 123F.3d 111, 116 (3d Cir. 1997) (citations omitted). �“Acourt has the power to revisit prior decisions of itsown or of a coordinate court in any circumstance,although as a rule courts should be loathe to do soin the absence of extraordinary circumstancessuch as where the initial decision was clearly erro-neous and would make a manifest injustice.�”Christianson v. Colt Indus. Operating Corp., 486

187a


U.S. 800, 816, 108 S. Ct. 2166, 100 L. Ed. 2d 811(1988) (citing Arizona, 460 U.S. at 618 n.8).

Feesers argues that this Court held, in its prioropinion, that the evidence of price discriminationin the record was sufficient to apply an inferenceof competitive injury. If this argument were true,it would be difficult for us now to conclude thatFeesers cannot show that it was a competing pur-chaser, as being a competing purchaser is a pre-requisite to the application of the inference.Feesers�’s argument, however, fails for several reasons.

First, this Court�’s prior opinion did not hold asFeesers now claims. This Court reversed the Dis-trict Court�’s summary judgment for the defen-dants explaining that the District Court used thewrong standard in concluding that Feesers andSodexo were not in actual competition. Feesers,Inc., 498 F.3d at 208. In doing so, we noted that �“ifsubstantial price discrimination between com-peting purchasers over time is established, thenthe inference of competitive injury arises.�” Id. at216 (emphasis added). At that early stage of thelitigation, this Court believed only that �“Feesersha[d] proffered sufficient evidence of competitionbetween itself and Sodex[o] . . . to allow a rea-sonable factfinder to conclude that [they] [we]re inactual competition.�” Id. (internal quotation omit-ted). This Court then remanded the case for fur-ther proceedings consistent with its opinion. Id. at216. Nowhere in the prior opinion did this Courthold that Feesers and Sodexo were competing pur-chasers or, more generally, that Feesers hadestablished an inference of competitive injury.Thus, the law of the case does not prevent us from

188a


holding that Feesers and Sodexo were not com-peting purchasers under the RPA.

Second, our present review of this case is con-ducted with the benefit of a full record establishedat trial. That record was not available to thisCourt when we decided Feesers�’s appeal fromsummary judgment. We now know that Feeserscannot show that it and Sodexo were competingpurchasers based on the timing of their competi-tion and the nature of the market�—issues thatwere never discussed in the prior opinion, pre-sumably, because a complete record had not beenestablished. Finally, even if this Court had pre-viously held otherwise, our holding in this casewould be a permissible reevaluation of precedentin light of intervening authority, Toledo Mack. SeeInstitutional Investors Group v. Avaya, 564 F.3d242, 276 n.50 (3d Cir. 2009) (citing Reich v. D.M.Sabia Co., 90 F.3d 854, 858 (3d Cir. 1996)).

B.

The permanent injunction issued by the DistrictCourt states: �“[Michaels] is enjoined from refusingto sell its products to Feesers on the same termsas they are sold to Sode[x]o, so long as Feesersotherwise meets its standards as a customer.�”This injunction was issued under Section 16 of theClayton Act (1) as a remedy for contempt and (2)to prevent future competitive injury to Feesers.Because we are reversing the District Court�’sjudgment as a matter of law, neither of its reasonsfor the injunction survive. An injunction issuedbased on civil contempt cannot stand where theunderlying order on which it is based is invalid.

189a


See Universal Athletic Sales Co. v. Salkeld, 511F.2d 904, 909-10 (3d Cir. 1975). Our holding todayrenders the need to protect Feesers from furtherinjury non-existent, because Feesers, as a matterof law, is not a competing purchaser vis-a-visSodexo.

C.

Feesers�’s claim against Sodexo arises under § 2(f) of the RPA. That provision states: �“It shallbe unlawful for any person engaged in commerce,in the course of such commerce, knowingly toinduce or receive a discrimination in price whichis prohibited by this section.�” 15 U.S.C. § 13(f).Because a prima facie case of price discriminationunder § 2(a) of the RPA cannot be establishedagainst Michaels, Sodexo cannot be held liable forinducement. Great Atl. & Pac. Tea Co., 440 U.S. at76 �“[A] buyer cannot be liable if a prima facie casecould not be established against a seller.�”).

V.

Feesers cannot show that it and Sodexo werecompeting purchasers, and therefore, cannot showthat it suffered competitive injury under theRobinson-Patman Act. Accordingly, we willreverse the District Court�’s judgment for Feesersand instruct the District Court to enter judgmentas a matter of law for Michaels and Sodexo.

190a


APPENDIX E


__________Nos. 09-2548, 09-2952, 09-2993

__________Feesers, Inc.

Plaintiff/Appellee�—v.�—

Michael Foods, Inc.; Sodexho, Inc.Defendants/Appellants

__________SUR PETITION FOR REHEARING

__________Present: SCIRICA, Chief Judge, SLOVITER,

MCKEE, RENDELL, BARRY, AMBRO, FUENTES,SMITH, FISHER, CHAGARES, JORDAN,

HARDIMAN, and Nygaard* Circuit Judges,

The petition for rehearing filed by appellee inthe above-entitled case having been submitted tothe judges who participated in the decision of thisCourt and to all the other available circuit judges

191a


* The vote of the Honorable Richard L. Nygaard,Senior Judge for the Third Circuit Court of Appeals, is lim-ited to panel rehearing.

of the circuit in regular active service, and nojudge who concurred in the decision having askedfor rehearing, and a majority of the circuit judgesof the circuit in regular service not having votedfor rehearing, the petition for rehearing by thepanel and the Court en banc, is denied.

By The Court,

/s/ D. Brooks SmithCircuit Judge

Dated: March 4, 2010

ARL/cc: All Counsel of Record

192a
