AMENDED REDACTED1

MEMORANDUM OPINION AND ORDER

ANN D. MONTGOMERY, District Judge.

I. INTRODUCTION

On July 19, 2016, the undersigned United States District Judge heard oral argument on Plaintiffs' Motion for Class Certification [Docket No. 607]. Defendants SuperValu, Inc. ("SuperValu") and C&S Wholesale Grocers, Inc. ("C&S") (collectively, "Defendants") oppose the Motion. For the reasons set forth below, Plaintiffs' Motion is granted in part and denied in part.

II. BACKGROUND2

This multi-district litigation consolidates the antitrust lawsuits of retail grocers against SuperValu and C&S, the two largest full-line grocery wholesalers in the United States. See Second Consol. Am. Class Action Compl. [Docket No. 99] ("Second Am. Compl.") ¶ 1. Plaintiffs allege Defendants conspired to allocate customers and territories through an Asset Exchange Agreement ("AEA"), and that Defendants used the allocations to charge retailers supra-competitive prices, in violation of Section 1 of the Sherman Act, 15 U.S.C. § 1. See id. ¶¶ 34-44, 77-83. Plaintiffs assert their claims as a class action. Id. ¶¶ 67-75.

A. Parties

Plaintiffs are retail grocery stores who purchased wholesale grocery products and related services directly from SuperValu and C&S. Id. ¶¶ 5-10. SuperValu's business is primarily in the Midwest, and C&S's business is largely concentrated in New England. Id. ¶ 1.

Prior to June 2002, SuperValu was C&S's largest competitor in New England, where C&S was the primary wholesaler. In re Wholesale Grocery Prods. Antitrust Litig., 752 F.3d 728, 729 (8th Cir. 2014), cert. denied 135 S.Ct. 2805 (2015). C&S did not compete with SuperValu in the Midwest, where SuperValu's main competitor was Fleming Companies ("Fleming"). Id. at 729-30; Expert Report of Jeffrey J. Leitzinger, Ph.D., March 1, 2016 [Docket No. 621] ("Leitzinger Report") ¶ 17.

B. Wholesale Grocery Industry

Grocery wholesalers such as SuperValu and C&S act as "middlemen" in the grocery supply chain, purchasing products from manufacturers, storing those products at distribution centers, and later reselling those products to retail grocers. Leitzinger Report ¶ 11; Expert Report of Dr. John H. Johnson, IV, May 6, 2016 [Docket No. 624] ("Johnson Report") ¶ 26. "Full-line" wholesalers such as SuperValu and C&S distribute tens of thousands of products in every product category, whereas "partial line" wholesalers offer a more limited set of product categories. Leitzinger Report ¶¶ 40-42; Wholesale Grocery, 752 F.3d at 729. The large volume purchases by full-line wholesalers result in cost benefits that are passed from the wholesaler to the retailer in the form of lower prices. Leitzinger Report ¶ 11. The Plaintiff grocers have testified that they require access to a full-line wholesaler to maintain their competitiveness because full-line wholesalers generate economies of scale and also carry the full line of products that retailers need to sell to their customers. Id. ¶ 46.

C. The Asset Exchange Agreement

In 2003, Fleming declared bankruptcy, and C&S announced its intention to acquire Fleming's Midwest wholesale grocery distribution business, which would have resulted in C&S becoming SuperValu's major competitor in the Midwest. Id. ¶ 17; Wholesale Grocery, 752 F.3d at 730. In September 2003, C&S and SuperValu entered into an Asset Exchange Agreement (the "AEA"), in which C&S agreed to designate SuperValu to receive Fleming's Midwestern assets and SuperValu agreed to transfer its New England assets to C&S. Wholesale Grocery, 752 F.3d at 730. The AEA included reciprocal non-compete provisions with each Defendant agreeing not to supply former customers served from a distribution center exchanged in the agreement for two years, and each Defendant agreeing not to solicit those customers for a period of five years. Id.

Plaintiffs allege the purpose of the AEA was to allocate customers and territory and to agree not to compete with each other for the customers and territories they exchanged. Second Am. Compl. ¶¶ 1-3, 33, 76-83. Plaintiffs aver the elimination of competition between SuperValu and C&S in regional markets allowed each Defendant to charge supra-competitive prices to their retail customers. Id. ¶ 39.

D. SuperValu's ABS Pricing

In the Midwest market, SuperValu uses a pricing formula known as "activity based sell" ("ABS") pricing for its dry grocery, general merchandise, dairy, and frozen product categories. See In re Wholesale Grocery Prods. Antitrust Litig., No. 09-MD-2090, 2012 WL 3031085, *3 (D. Minn. July 25, 2012) [Docket No. 352] ("Class Certification Order"). The ABS pricing system sells products to retail grocers at each product's actual average cost to SuperValu plus two types of fees—operating fees and service fees. Id.

Operating fees, which are at issue in this Motion, are based on retailers' ordering behavior, the handling costs of the items they purchase, and SuperValu's requisite margins. Odette Decl. [Docket 636] Ex. J at SV00129381-83, SV00129398. Operating fees are a function of a wide range of factors: some are specific to customer's ordering behavior, some are specific to the distribution center ("DC"), and some are specific to the product being purchased. Johnson Report ¶ 30 n.28. As a result, a customer's operating fees can vary based on item mix, pallet quantity, average number of cases received, cube space, length of time at the warehouse, number of minutes on a picking ticket, number of cases on a reference number, cost to load a retailer's order, administrative costs, and warehouse expenses. Id.

To calculate operating fees, SuperValu uses nine ABS fee factors called "fee drivers" which represent the core supply chain activities: (1) receiving, (2) storage, (3) finance, (4) selecting, (5) loading, (6) customer minimum, (7) item minimum, (8) overhead, and (9) percent of purchase. Odette Decl. Ex. J at SV00129400. Because the factors are influenced by conditions specific to each DC, such as local wage rates or average cost of storing an item at a particular warehouse, the exact fee amounts vary by DC. Each DC has its own unique set of default fee driver calculations, as does each product line within a DC. See, e.g., Johnson Report Ex. 86 (setting ABS fee driver rates among product lines at the Champaign DC).

The ABS pricing system also includes more than 80 "component IDs," each of which corresponds to a type of activity (such as loading or selecting) and has an associated formula. Johnson Report ¶ 31. The component IDs can be unique to a customer or groups of customers depending on the circumstances. For example, SuperValu customers who choose to pick up their orders from the SuperValu DC instead of paying for SuperValu to transport the product to the retail store are given a "will call loading credit" to offset a portion of the ABS loading fee that is typically charged to other retailers. Id. ¶¶ 61-62. This credit is applied using the component ID "LDG." As another example, SuperValu charges a fixed ABS operating fee or caps the ABS fees for some customers based on customer negotiations or other situations. Id. ¶ 57. For these customers, component ID "FFP" is used. Id.

The ABS pricing system also allows for customer-specific or situation-specific modifications to the ABS pricing system using "authorization codes" and "flow codes." Johnson Report ¶¶ 65-69. Authorization codes enable SuperValu to apply "specific fees, freight, or other programs to specific customers." Id. ¶ 66. Similarly, flow codes allow SuperValu to "determine fees and apply customer-specific pricing." Id. ¶ 68.

In addition to the ABS pricing system, SuperValu individually negotiates rebates, market subsidies, and other financial incentives with some of its customers. See, e.g., Defs.' Mem. Opp'n Mot. Class Certification [Docket No. 623] at 27.

Although the unique component IDs, authorization codes, and flow codes within the ABS pricing system allow for customer-specific modifications, SuperValu's transactional data shows that for purchases made by class members during the class period, customer-specific modifications applied to only a small portion of a given customer's overall purchases. Expert Rebuttal of Jeffrey J. Leitzinger, Ph.D., June 17, 2016 [Docket No. 637] ("Leitzinger Rebuttal") ¶¶ 12, 44, 47, Attach. 4. For each proposed class, the large majority of every member's purchases were charged under an ABS formula that was identical to other members of that class. See id. Attach. 4.

E. Class Certification History

On October 31, 2011, Plaintiffs moved to certify two broad putative classes: (1) retailers who purchased products or related services from Defendants in the Midwest market (the "Midwest Class"); and (2) retailers who purchased products or services in the New England market (the "New England Class").3 The Midwest Class was represented by Plaintiff D&G Inc. ("D&G"), and the New England Class was represented by Plaintiff DeLuca's Corporation ("DeLuca's"). See Pls.' Mot. Class Cert. [Docket No. 202] 1-3.

This Court denied class certification of both proposed classes on July 16, 2012, after determining that Plaintiffs could not show class-wide impact through common evidence. See Class Certification Order, 2012 WL 3031085, at *9-17. With respect to the New England Class, the Court rejected Plaintiffs' attempts to show impact through three methods—the contrary hypothesis test, the variance test, and list prices—because none of the methods could establish that upcharges had uniformly increased for all class members. Id. at *10-14.

Addressing the Midwest Class, the Court noted that the "formulaic nature" of the ABS pricing method applied by SuperValu to charge customers in the Midwest market made a better case for certification than the New England Class. Id. at *16. Nevertheless, the Midwest Class could not prove impact by common evidence because each of SuperValu's DCs across the Midwest inserted different values into the ABS formula, requiring an analysis of each DC's fees and competitive conditions. Id.

On August 31, 2012, D&G requested leave to move for certification of a narrower Midwest Class of grocers who were charged using ABS pricing and who were supplied by SuperValu's Champaign, Illinois DC. See Letter, Aug. 31, 2012 [Docket No. 362]. D&G argued that the Champaign DC-based Class would "entirely avoid[] the problems created by different ABS formulas for different distribution centers." Id. at 1. DeLuca's made no request to narrow the New England Class.

On January 11, 2013, the Court granted summary judgment to Defendants. In re Wholesale Grocery Prods. Antitrust Litig., No. 09-MD-2090, 2013 WL 140285 (D. Minn. Jan. 11, 2013) [Docket No. 427]. Based on the grant of summary judgment to Defendants, the Court denied as moot D&G's request for renewed class certification of the narrower Midwest Class. Id. at *15. D&G appealed both the summary judgment ruling and the denial of certification of the Midwest Class. Notice of Appeal [Docket No. 429].

On May 21, 2014, the Eighth Circuit reversed the grant of summary judgment against D&G, but affirmed the denial of certification of the Midwest Class. Wholesale Grocery, 752 F.3d at 733-36. Additionally, the Eighth Circuit vacated the denial of D&G's request for leave to pursue certification of a narrower class of SuperValu customers in the Midwest who were supplied by the Champaign, Illinois DC and charged using the ABS pricing formula. Id. at 736. The Eighth Circuit stated that "the evidence suggests the ABS fee inputs would be standardized for this narrow class," and that this Court should "consider, in light of our holding that the wholesalers are not entitled to summary judgment, whether to certify this class." Id.

F. Present Motion

Midwest Plaintiffs now seek certification of five proposed DC based classes:

Hr'g Ex. [Docket No. 642] at 5.

Class members in each of the five DC based classes would include all customers that paid ABS fees on wholesale grocery products in all four SuperValu product categories (grocery, frozen, dairy, and general merchandise) purchased directly from an above-named DC from December 31, 2004 through September 13, 2008 (the "Class Period") and that are located in a relevant geographic market. See Pls.' Mot. Class Certification at 1-2.4

III. DISCUSSION

A. Class Certification Standard

The requirements for class certification are set forth in Rule 23 of the Federal Rules of Civil Procedure ("Rule 23"). Rule 23 implicitly requires that a defined class exists and that the class representatives fall within the class. Mathers v. Northshore Mining Co., 217 F.R.D. 474, 484 (D. Minn. 2003). The proposed class must also satisfy the four explicit requirements of Rule 23(a), which are numerosity, commonality, typicality, and adequacy. Fed. R. Civ. P. 23(a); Donaldson v. Pillsbury Co., 554 F.2d 825, 829 (8th Cir. 1977). Additionally, a plaintiff must fulfill one of the three subsections of Rule 23(b) to obtain certification for the proposed class. Fed. R. Civ. P. 23(b); Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 163 (1974); In re St. Jude Med., Inc., 425 F.3d 1116, 1119 (8th Cir. 2005). Here, Plaintiffs invoke Rule 23(b)(3), which permits a class action to be maintained if two requirements are met: "Common questions must `predominate over any questions affecting only individual members'; and class resolution must be `superior to other available methods for the fair and efficient adjudication of the controversy.'" Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 615 (1997) (quoting Fed. R. Civ. P. 23(b)(3)).

Plaintiffs bear the burden of proving both the threshold Rule 23(a) factors and the two requirements of Rule 23(b)(3). See Coleman v. Watt, 40 F.3d 255, 258 (8th Cir. 1994). When considering a motion for class certification, the court takes the substantive allegations in the plaintiff's complaint as true. Lockwood Motors, Inc. v. Gen. Motors Corp., 162 F.R.D. 569, 573 (D. Minn. 1995). Nonetheless, "sometimes it may be necessary for the court to probe behind the pleadings before coming to rest on the certification question." Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350 (2011) (quoting Gen. Tel. Co. v. Falcon, 457 U.S. 147, 160 (1982)). This inquiry is necessary because "the class determination generally involves considerations that are `enmeshed in the factual and legal issues comprising the plaintiff's cause of action.'" Coopers & Lybrand v. Livesay, 437 U.S. 463, 469 (1978) (citation omitted). Only after a "rigorous analysis" of the proposed class and the requirements of Rule 23 may a court grant class certification. Falcon, 457 U.S. at 161; Dukes, 564 U.S. at 350-51. A district court has broad discretion in deciding whether class certification is appropriate. Ebert v. Gen. Mills, Inc., 823 F.3d 472, 477 (8th Cir. 2016).

B. Implicit Rule 23 Criteria

Rule 23 implicitly requires that a defined class exists and that the class representatives fall within the class. Mathers, 217 F.R.D. at 484. Defendants argue that the proposed classes fail the requirement that a defined class exists. Specifically, Defendants contend that the proposed classes are unascertainable because the class definitions make class membership dependent upon whether a customer is "located in a relevant geographic market." The relevant geographic markets will not be determined until a finding of fact is made. Until that determination is made, argue Defendants, it is impossible to determine which customers are in each class and would be bound by the judgment.

"It is elementary that in order to maintain a class action, the class sought to be represented must be adequately defined and clearly ascertainable." Sandusky Wellness Ctr., LLC v. Medtox Sci., Inc., 821 F.3d 992, 996 (8th Cir. 2016). A class must be defined using objective, rather than subjective, criteria that will allow for class members to be identified in an administratively feasible manner. In re Delta/Airtran Baggage Fee Antitrust Litig., No. 09-MD-2089, 2016 WL 3770957, at *3 (N.D. Ga. July 12, 2016). "The precise contours of the class need not be ascertained before certification so long as the class members can be identified at some stage of the proceeding." Id. (quotation omitted).

Here, the proposed class definitions allow for class members to be identified using objective criteria during the litigation. Once the geographic market is determined, SuperValu's transaction data can be used to ascertain which class members fall within the geographic market. Because objective criteria will define the precise contours of the classes at some stage in the proceeding, the proposed classes are not infirm due to ascertainability. Thus, the classes are adequately defined and clearly ascertainable. Additionally, the class representatives fall within the defined classes. Therefore, the implicit requirements of Rule 23(a) are satisfied.

C. Expanded Class Definition

Turning to another issue related to class definitions, Defendants argue that Plaintiffs have impermissibly expanded the class definition by including customers located in a "relevant geographic market" which, according to Plaintiffs' expert, includes parts of Missouri. The Court agrees that including customers in Missouri would expand, rather than narrow, the Midwest Class definition that Plaintiffs proffered earlier in this suit.

"District courts typically . . . hold a plaintiff seeking class certification to the definition espoused in the relevant complaint." West v. Nissan N. Am., Inc., No. 13-4070, 2014 WL 825217, at *3 (W.D. Ark. March 4, 2014) (quoting Savanna Grp., Inc. v. Trynex, Inc., No. 10-7995, 2013 WL 66181, at *2 (N.D. Ill. Jan 4, 2013)); see also Costelo v. Chertoff, 258 F.R.D. 600, 604-05 (C.D. Cal. 2009) (The Court is bound to class definitions provided in the complaint and, absent an amended complaint, will not consider certification beyond it."); but see Chapman v. First Index, Inc., 796 F.3d 783, 785 (7th Cir. 2015) (stating revised class definition does not require amendment to complaint because "the obligation to define the class falls on the judge's shoulders under Fed. R. Civ. P. 23(c)(1)(B)") .

The Second Amended Complaint states that "[t]he Midwest market includes the states of Illinois, Indiana, Iowa, Michigan, Minnesota, Ohio, and Wisconsin (`the Midwest')" and defines the proposed "Midwest Class" as "all entities in the Midwest that purchased full-line wholesale grocery products or related services directly from Defendants in the Midwest market." Second Am. Compl. ¶¶ 61, 67. Plaintiffs respond that including parts of Missouri in the proposed Champaign DC Classes does not expand the class alleged in the Second Amended Complaint because the Complaint did not allege an exhaustive list of states in the Midwest. Rather, the Complaint stated that the Midwest "includes" certain states.

Characterizing a list of Midwest states in the operative Complaint as not exhaustive does not necessarily lead to a conclusion that Missouri should be included in the list of Midwest states. Indeed, one cannot determine solely from reading the Second Amended Complaint whether Missouri falls within the Plaintiffs' definition of "Midwest." However, Plaintiffs' first motion for class certification provided clarification by expressly defining the "Midwest" Class as "entities located in Illinois, Indiana, Iowa, Michigan, Minnesota, Ohio, and Wisconsin." Pls.' [First] Mot. Class Cert. [Docket No. 202] at 1. Missouri is notably absent from the list of states comprising the Midwest.

After the first motion for class certification was denied, D&G sought the Court's permission to file a revised class certification motion seeking to certify a "narrowed Class consisting of all Midwest customers" who purchased products in all four ABS product lines from SuperValu's Champaign DC. Mem. Supp. Pls.' Request Leave File Second Am. Class Certification Mot. [Docket No. 420] at 1 (emphasis added). D&G's request did not state that the "narrowed Class" would include entities located in Missouri or that the definition of "Midwest" in the first class certification motion would be enlarged to include an additional state. See generally, id. Although this Court denied D&G's request for permission to file a revised class certification motion, the Eighth Circuit vacated that denial and asked this Court to consider the whether to certify the "narrower class of SuperValu customers" described in D&G's request. Wholesale Grocery, 752 F.3d at 736 (emphasis added).

The present motion for class certification now defines the proposed DC-based classes to include customers "located in a relevant geographic market," which Plaintiffs' expert asserts includes parts of Missouri. Pls.' Mot. Class Certification at 1-2; Leitzinger Report ¶ 54, n.135. Rather than being a "narrower class of SuperValu customers," the newly-proposed class has been expanded to include customers in Missouri. If Missouri retailers are included in the revised classes, the ABS sales to the proposed Champaign Non-Arbitration Class increase by more than $400,000,000, causing the size of the ABS sales volume to be more than double what it would be without the Missouri retailers. See Johnson Report ¶ 98, n.227. Thus, Plaintiffs have impermissibly expanded the class definition.

Further, even if Plaintiffs were not constrained by the class definition in the Second Amended Complaint, a district judge has discretion to insist "that representative plaintiffs put their class definitions on the table `[a]t an early practicable time.'" Chapman, 796 F.3d at 785 (quoting Fed. R. Civ. P. 23(c)(1)(A)) (alteration in original). Here, Plaintiffs had the opportunity to include the Missouri retailers in their class definition when they first moved for class certification, and again when they briefed their request for permission to file a second class certification motion. Plaintiffs' effort to now add Missouri to the proposed class at this stage of the litigation is rejected. See, e.g., id. (affirming district court's refusal to certify revised class because "a district judge has discretion to reject an attempt to remake a suit more than four years after it began").

Therefore, any class to be certified will not include Missouri grocery retailers.

D. Rule 23(a) Requirements

1. Numerosity

Rule 23(a)(1) requires that the class must be so numerous that joinder of all members individually is "impracticable." Fed. R. Civ. P. 23(a)(1). Impracticability of joinder is a question to be determined by the court "based upon all the circumstances surrounding a case." Boyd v. Ozark Air Lines, Inc., 568 F.2d 50, 55 (8th Cir. 1977). In determining whether numerosity is satisfied, the court considers "the size of the class, . . . the nature of the action, the size of the individual claims, the inconvenience of trying individual suits, and any other factor relevant to the practicability of joining all the putative class members." Paxton v. Union Nat'l Bank, 688 F.2d 552, 559-60 (8th Cir. 1982). Courts generally recognize that the existence of 40 or more class members raises a presumption that joinder is impracticable. Lockwood Motors, 162 F.R.D. at 574; 1 William B. Rubenstein, Newberg on Class Actions § 3.12 (5th ed. 2011). The Eighth Circuit has affirmed the certification of classes with as few as 20 members. See, e.g., Ark. Educ. Ass'n v. Bd. of Educ. of Portland, Ark. Sch. Dist., 446 F.2d 763, 765-66 (8th Cir. 1971) (upholding class of 17-20 members).

Defendants contend that joinder is practical because the class members are easily identifiable in the transaction data and are business entities asserting large claims. However, each of the five proposed classes includes 40 or more members, making joinder of all their claims impracticable. The Court concludes that numerosity is satisfied.

2. Commonality

Rule 23(a)(2) requires that there be "questions of law or fact common to the class." Fed. R. Civ. P. 23(a)(2). "Commonality requires a plaintiff to show that the class members `have suffered the same injury.'" Dukes, 564 U.S. at 349-50 (quoting Falcon, 457 U.S. at 157). "[F]or purposes of Rule 23(a)(2) even a single common question will do." Id. at 359 (quotation marks and alterations omitted).

This case presents several questions common to each of the proposed classes, including: whether Defendants agreed to allocate territories or customers for full-line grocery wholesale products and services; whether Defendants' agreement caused prices for full-line wholesale products and services to be higher for class members than they otherwise would have been; whether all class members were injured by Defendants' conduct; and whether Defendants' conduct violates anti-trust laws. Commonality is satisfied.

3. Typicality

Rule 23(a)(3) requires that the claim or defense of the representative parties is typical of the claims or defenses of the proposed class. Fed. R. Civ. P. 23(a)(3). This requirement is met when there are "other members of the class who have the same or similar grievance as the plaintiff." Alpern v. UtiliCorp United, Inc., 84 F.3d 1525, 1540 (8th Cir. 1996). Rule 23(a)(3) is "intended to assure the claims of the class representative are similar enough to those of the class, so that the representatives will adequately represent the class." In re Baycol Prods. Litig., 218 F.R.D. 197, 205 (D. Minn. 2003). The burden of showing typicality is "fairly easily met so long as other class members have claims similar to the named plaintiff." DeBoer v. Mellon Mortg. Co., 64 F.3d 1171, 1174 (8th Cir.1995).

The typicality requirement is satisfied here. Each class representative is a member of the class it seeks to represent, and the claims brought by the class representatives are based on the same legal theory. Defendants' arguments to the contrary do not compel a different conclusion. Defendants contend that D&G, Blue Goose, and Millennium fail to satisfy the typicality requirement because: (1) unlike many class members, D&G and Blue Goose never negotiated long-term supply agreements with individually negotiated rebates, price concessions and alternative fee arrangements; and (2) D&G and Millennium switched to co-ops as their primary suppliers, which contradicts Plaintiffs' theory that co-ops were not competitive alternatives to full-line suppliers.

"When the claim arises out of the same legal or remedial theory, the presence of factual variations is normally not sufficient to preclude class action treatment." Donaldson, 554 F.2d at 831. Here, the factual differences identified by Defendants as distinguishing D&G and Blue Goose from their class do not alter the conclusion that their claims are sufficiently similar to those of their respective classes that they will adequately represent their classes. Although D&G and Millennium switched to co-ops during the class period, their claims also are sufficiently similar to those of their respective classes that they will adequately protect the interests of the classes, and the interests between D&G and Millennium and the classes do not conflict. See, e.g., In re Wirebound Boxes Antitrust Litig., 128 F.R.D. 268, 270-71 (D. Minn. 1989) (holding class representative satisfied Rule 23(a) requirements of typicality and adequacy even though the representative's conduct arguably undercut plaintiffs' claims in the litigation).

4. Adequacy

Rule 23(a)(4) focuses on whether: (1) the class representatives have common interests with the members of the class, and (2) whether the class representatives will vigorously prosecute the interest of the class through qualified counsel. Paxton, 688 F.2d at 562-63. Defendants do not dispute that Plaintiffs' counsel are qualified, but argue that D&G, Blue Goose, and Millennium are not adequate representatives for the same reasons asserted in their argument against typicality. The Court rejected those arguments as they pertained to typicality and rejects them as they pertain to adequacy as well. Additionally, the proposed representatives have demonstrated their commitment to vigorously prosecute the interests of the classes by responding to interrogatories, producing documents, and being deposed. The adequacy requirement is satisfied.

E. Rule 23(b)(3) Requirements

Rule 23(b)(3) requires: (1) that common questions of law or fact predominate over individual questions, and (2) that the class action mechanism is superior to any alternative form of litigation for the same claims. Fed. R. Civ. P. 23(b)(3).

1. Predominance

The predominance standard is "demanding." Amchem, 521 U.S. at 624. In examining this requirement, the Court "must conduct a limited preliminary inquiry, looking behind the pleadings . . . to determine whether, given the factual setting of the case, if the plaintiffs['] general allegations are true, common evidence could suffice to make out a prima facie case for the class." Blades v. Monsanto Co., 400 F.3d 562, 566 (8th Cir. 2005) (internal citation omitted). "The question at class certification is not whether the plaintiffs have already proven their claims through common evidence. Rather, it is whether questions of law or fact capable of resolution through common evidence predominate over individual questions." In re Zurn Pex Plumbing Prod. Liab. Litig., 644 F.3d 604, 619 (8th Cir. 2011) (citing Blades, 400 F.3d at 566-67).

Whether a question is common or individual depends on the nature of the evidence. Blades, 400 F.3d at 566. "If, to make a prima facie showing on a given question, the members of a proposed class will need to present evidence that varies from member to member, then it is an individual question." Id. Conversely, "[i]f the same evidence will suffice for each member to make a prima facie showing, then it becomes a common question." Id. Put another way, "a claim will meet the predominance requirement when generalized evidence proves or disproves the elements of the claim on a class-wide basis, because such proof obviates the need to examine each class member's individual position." Buetow v. A.L.S. Enters., Inc., 259 F.R.D. 187, 190 (D. Minn. 2009) (internal quotations and alterations omitted).

Plaintiffs allege Defendants allocated territory and customers in violation of Section 1 of the Sherman Act, 15 U.S.C. § 1, and bring suit for damages under Section 4 of the Clayton Act, 15 U.S.C. § 15. Second Am. Compl. ¶¶ 76-83. To prevail, Plaintiffs must prove (1) Defendants violated federal antitrust laws, (2) Plaintiffs suffered some resulting injury, and (3) a measure of damages. Blades, 400 F.3d at 566; In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 311 (3d Cir. 2009); In re Visa Check/Mastermoney Antitrust Litig., 280 F.3d 124, 136 (2d Cir. 2001). For the proposed classes to be certified, therefore, Plaintiffs must demonstrate that common evidence can establish each of these elements. Visa Check/Mastermoney, 280 F.3d at 136.

a. Antitrust Violation

The first element can be shown with generalized evidence because proof of an antitrust conspiracy in violation of Section 1 of the Sherman Act "involves primarily common issues of fact and law." Wirebound Boxes, 128 F.R.D. at 271; see also Blades, 400 F.3d at 572 ("Evidence that appellees entered into a conspiracy that would affect all class members would perforce be evidence common to all class members for proving the conspiracy.").

b. Impact

"Impact" means that the alleged antitrust violation caused injury to the plaintiff. Blades, 400 F.3d at 569. "[A]n increase in price resulting from a dampening of competitive market forces is assuredly one type of [antitrust] injury." Blue Shield of Va. v. McCready, 457 U.S. 465, 482 (1982). This is the type of injury alleged by Plaintiffs. See Second Am. Compl. ¶¶ 39, 70(a), 77, 83 (alleging Defendants' allocation of customers and territories allowed Defendants to charge supra-competitive prices in those territories). Therefore, to proceed as a class action, Plaintiffs must be able to show with common evidence that each class member was charged supra-competitive prices.

Plaintiffs present a three-step analysis to argue that class-wide impact can be shown with common evidence: first, the elimination of competition due to the AEA enabled SuperValu to achieve supra-competitive net margins; second, the supra-competitive margins strongly support the conclusion that SuperValu charged supra-competitive ABS prices; and third, every class member paid supra-competitive ABS prices because every member (I) made purchases in all four ABS product categories and (ii) was charged under the same ABS formula for the vast majority of its purchases.

In step one of the impact analysis, Plaintiffs assert that after Defendants' agreement not to compete, SuperValu enjoyed net margins at its Class DCs in the Midwest that were nearly three times higher than they would have been but for Defendants' agreement.

According to a net margin comparison conducted by Plaintiffs' expert, Dr. Jeffrey Leitzinger, SuperValu's net margins for the Class DCs after the AEA (when SuperValu faced no competition from C&S in the Midwest) were over 5%, whereas its net margins in New England prior to the AEA (where SuperValu used to compete with C&S) were only 1.7%.5

Leitzinger Report ¶ 140. The amount by which the net margins of the Class DCs exceed those of pre-AEA New England is the amount by which SuperValu's ABS fees were allegedly inflated. Plaintiffs refer to this amount as the "overcharge amount." Id. ¶ 126.

According to economic theory as explained by Dr. Leitzinger, the net margin in a given market depends solely on two factors: market demand elasticity and the nature of competition. Id. ¶¶ 126, 137; Leitzinger Rebuttal ¶ 103. Dr. Leitzinger opines that because the demand for full-line wholesale groceries did not materially differ between the Class DCs and pre-AEA New England, the differences in net margins are the result of differences in competition. Leitzinger Report ¶ 137; Leitzinger Rebuttal ¶¶ 88-94.

Dr. Leitzinger further opines that pre-AEA New England is an "ideal" competitive benchmark because SuperValu was C&S's closest and largest competitor in that region prior to the AEA, causing the competitive dynamics in pre-AEA New England to be "very similar to those that would have existed in the Midwest but for Defendants' agreement." Leitzinger Report ¶ 129.

Defendants raise multiple challenges to Plaintiffs' net margin analysis. They argue that Dr. Leitzinger's net margin analysis is flawed because it fails to account for differences in cost structures between the Midwest and New England DCs. For example, SuperValu's Midwest DCs had significantly lower operating costs than its New England DCs due to higher volume and lower-cost labor force. Johnson Report ¶¶ 44-45, Fig. 2. Defendants contend that when the DCs' operating costs are analyzed as a percentage of sales, the differences in costs are large enough to completely offset the differences in margins. Thus, argue Defendants, the differences in net margins are not due to "overcharges" or anticompetitive effects. However, as noted earlier, if competition and demand are the same across two markets, differences in costs will be reflected in different prices, not different margins. Leitzinger Report ¶ 137; Leitzinger Rebuttal ¶ 103. In other words, lower costs would result in lower prices rather than higher margins. As Dr. Leitzinger notes, if C&S had been competing with SuperValu in the Midwest, it would have had the same volume and labor efficiency advantages and thus would have enjoyed similarly low operating costs in the Midwest. Leitzinger Rebuttal ¶ 104. As a result, SuperValu would have had to set its ABS fee structure lower to protect its customer base from the competition. Leitzinger Report ¶ 97.

Defendants also argue that the benchmark comparison with pre-AEA New England does not account for differences over time, between regions, and between customers. Such differences include: store size, presence and size of competitors, regional customer concentrations, overall mix of product categories, and differences in demand for various products. Plaintiffs respond that their expert has analyzed these factors and concluded that they did not pose significant differences that would warrant further controls to the benchmark analysis. Leitzinger Report ¶¶ 128-33; Leitzinger Rebuttal ¶¶ 87-94; 105-06. The Court finds this sufficient at this stage.

Defendants further argue that the differing margins between the "benchmark" New England DCs (the Portland DC and the Andover/Cranston DC) undermine Plaintiffs' argument that higher net margins are proof of supra-competitive pricing. The net margin for the Portland DC was 3.29%, which was more than twice that of Andover/Cranston DC's net margin of 1.48%. Johnson Report ¶ 43. Defendants contend that if the only distinction between the Midwest and New England margins is anti-competitive conduct, the two benchmark DCs in New England should be the same. Defendants thus conclude that net margins cannot be relied upon as evidence of supra-competitive pricing. However, Dr. Leitzinger attributes the differences in the net margins to the Portland DC's remote location in northern Maine with a low population and relatively few grocery retailers and few competing wholesalers. Leitzinger Rebuttal ¶ 99. Dr. Leitzinger thus concludes that "the differences in margin results . . . likely reflect the operation of exactly that which those margins are being used here to reflect—the role of competition." Id. Dr. Leitzinger further notes that had the Portland DC been excluded from the benchmark analysis, the pre-AEA New England margin benchmark would have been even lower and the estimated overcharge would have been even higher. Therefore, the differing net margins at the benchmark New England DCs are not fatal to Plaintiffs' net margin analysis.

Based on the foregoing, the Court concludes that Plaintiffs' net margin analysis is capable of showing that SuperValu's net margins were higher than they would have been but for Defendants' AEA.

Step two of Plaintiffs' impact analysis asserts that the supra-competitive margins at the Midwest Class DCs necessarily mean that SuperValu's ABS prices at those DCs were also supra-competitive. Collectively, sales of products in ABS categories accounted for over 70% of the Midwest DCs' sales. Leitzinger Report ¶¶ 52, 141. Based on this volume, Dr. Leitzinger opines that no other category or categories of non-ABS sales were large enough to account for the difference between SuperValu's margins in the Midwest and pre-AEA New England. He explains:

Id. ¶ 143; see also Leitzinger Rebuttal ¶¶ 117-18.

Additionally, the report references instances of SuperValu universally applying ABS fee adjustments to a DC or region with the goal of maintaining or improving margins. See Leitzinger Report Exs. 160-61, 169. In determining an appropriate pricing level for ABS fees, SuperValu considers the "competitive landscape" and whether the ABS fees will be "palatable." Id. Ex. 169 at SV00028676, Ex. 160 at SV00220298 (listing "competitive situation" as factor a in determining whether to implement an ABS fee increase).

Plaintiffs have demonstrated that common evidence is capable of showing that the allegedly supra-competitive margins at the Midwest Class DCs necessarily mean that SuperValu's ABS prices at those DCs were supra-competitive.

Step three of Plaintiffs' impact analysis asserts that every member of each DC Class paid supra-competitive prices because every member: (i) bought in all four product categories, and (ii) was charged under the same ABS formula for the vast majority of its purchases.

Defendants argue that by analyzing net margins rather than actual prices paid by each class member, Plaintiffs cannot show that supra-competitive profits were obtained from all class members. Defendants insist that individualized evidence is necessary to determine whether every class member was injured, because (i) even within a single DC, ABS fees were individualized and customer-specific, and (ii) SuperValu negotiated a host of unique pricing concessions such as rebates, market support payments, and other off-invoice discounts.

Defendants' argument as to customer-specific ABS fees fails to defeat class certification because unique ABS fee drivers were applied to only a small fraction of any given class member's purchases. Dr. Leitzinger analyzed all ABS purchases for each DC Class member and discovered that for four out of the five Class DCs, every class member was charged under an single, identical ABS formula for at least 91% of their purchases. See Leitzinger Rebuttal ¶¶ 28-30, Attach. 4. For the fifth class (Hopkins), every class member was charged under a single, identical ABS formula for over 78% of their purchases. Id. Attach 4. Thus, although the ABS pricing system is capable of applying thousands of unique fee drivers and fee driver combinations, only a small percentage of these customized fee drivers was actually applied to the ABS fees of any class member.

Defendants also argue that even in the absence of customer-specific modifications to the ABS system, ABS fees are inherently individualized because pricing depends on customer characteristics and ordering behavior. For each ABS product category within each Class DC, prices for some class members increased, some decreased, and some stayed the same. Johnson Report Figs. 9-12. Further, the ABS fees for the Class DCs actually declined after the AEA. Defendants thus contend that individual issues predominate because antitrust impact can only be determined by analyzing the actual price paid by each class member and determining whether that class member's price would have been lower in a hypothetical, "but for" world.

The Court disagrees. Although the ABS price a customer ultimately pays for a given product will vary based on the customer's ordering behavior, this variation is the result of algorithms embodied in the ABS pricing formula rather than customer-specific pricing. See, e.g., Odette Decl. Ex. J (SuperValu 2/27/2009 Power Point Presentation) at SV00129381 (describing ABS as "charging a fee based on retailer behavior and the supply chain costs associated with the products they buy"); Leitzinger Report Ex. 13 (Chew Dep.) ("[N]one of the fees that we have is associated with a who . . . . [T]he fees are associated with behavior."). Thus, if SuperValu's common pricing mechanism—the ABS system—was set above a competitive level, class members who made purchases under the ABS formula were charged more than if the pricing system had not been inflated. Additionally, Plaintiffs need not prove that prices increased to show impact. A price may be supra-competitive without actually increasing if a competitive price would be lower than the price charged.

Defendants also argue that Plaintiffs merely speculate that SuperValu would have responded to competition from C&S by reducing ABS fees across the board instead of using rebates, discounts, and other pricing concessions to secure business from competitors. See Johnson Report ¶ 88. Dr. Leitzinger responds that this approach of maintaining substantially higher ABS fees while offering large, selective rebates to particular customers would potentially lead to high customer dissatisfaction and would likely prove unworkable. Leitzinger Rebuttal ¶¶ 83-85. The Court finds this sufficient at this stage.

Defendants' argument regarding unique pricing concessions such as rebates and other individually-negotiated off-invoice discounts also lacks merit. No class member's off-invoice discounts were large enough to offset the amount of the overcharge, with the exception of [redacted].6 See id. Attach. 4. Moreover, courts in this district have held that class certification is warranted where individually negotiated discounts are made from a common pricing mechanism that was inflated from the start. See, e.g., In re Potash Antitrust Litig., 159 F.R.D. 682, 696 (D. Minn. 1995) (finding common proof of impact where inflated prices served as starting point for individually negotiated price).

Defendants further argue that injury from higher transportation costs for some retailers following the AEA requires individual inquiry. After the AEA: (1) some former Fleming customers' transportation fees increased, (2) other former Fleming customers' transportation fees decreased, and (3) existing SuperValu customers' transportation fees did not change. However, the lower transportation fees for some former Fleming customers were too small to eliminate those customers' overcharges. Leitzinger Report ¶ 94 n.220.

Because the nature of the evidence is sufficient to show that ABS prices were supracompetitive, and that every member of each DC Class made purchases in all four ABS product categories and was charged under the same formula for the vast majority of its purchases, common evidence can suffice to prove class-wide impact.

c. Damages

Plaintiffs argue that damages can be proven with common evidence because each individual class member's overcharge damages can be calculated using a single formula: multiplying each class member's ABS purchases from their Class DC by the percentage of price inflation of that DC's ABS prices. Plaintiffs contend that even if some individualized issues such as rebates for some class members were presented, these issues do not overwhelm the common issues, thus class certification is proper. Defendants respond that individually negotiated prices prohibit damage calculation using common evidence.

In proving the amount of damage in an antitrust case, a plaintiff "need only present evidence from which the fact finder may make a just and reasonable estimate of the damage that is not based on speculation or guesswork." Nat'l Farmers' Org., Inc. v. Associated Milk Producers, Inc., 850 F.2d 1286, 1293 (8th Cir. 1988); Piggly Wiggly Clarksville, Inc. v. Interstate Brands, Corp., 215 F.R.D. 523, 530 (E.D. Tex. 2003) ("[D]amages in the antitrust context need not be proven with exact particularity, [but] may not be merely speculative."). Individual issues in calculating damages do not preclude class certification unless those individual issues overwhelm common issues. In re Target Corp. Customer Data Sec. Breach Litig., 309 F.R.D. 482, 489 (D. Minn. 2015); Potash, 159 F.R.D. at 697 ("[T]he fact that the damages calculation may involve individualized analysis is not by itself sufficient to preclude certification when liability can be determined on a class-wide basis."); Wirebound Boxes, 128 F.R.D. at 272 ("The amount of damages largely involves individualized questions. This is typically true in antitrust class actions, however, and does not preclude certification.").

Here, Dr. Leitzinger offers a generalized mathematical formula that can be applied to all class members to calculate a reasonable estimation of their damages. Although some individualized adjustments for customer-specific ABS pricing modifications or off-invoice discounts will be required for some class members, as discussed above, these individual issues do not overwhelm the common issues stemming from class members' purchases of products under SuperValu's largely universal ABS pricing formula.

Defendants also argue Dr. Leitzinger's damages model is insufficient under Comcast Corp. v. Behrend, 133 S.Ct. 1426 (2013), because the model fails to account for the numerous factors that can affect prices and margins. This argument lacks merit. In Comcast, plaintiffs' damages model was based on theories of impact that had been rejected by the district court. Id. at 1430-31. The plaintiffs' expert "expressly admitted" that the damages model was not capable of identifying the damages attributable to the single impact theory that remained in the case. Id. at 1434. The Supreme Court held that the Rule 23(b)(3) predominance requirement was lacking due to the damage model's "inability to bridge the differences between supra-competitive prices in general and supra-competitive prices attributable to [the remaining theory of anti-trust impact]." Id. at 1435. Unlike Comcast, Plaintiffs offer only a single theory of anti-trust impact—that each class member paid supra-competitive ABS fees as a result of Defendants' agreement not to compete.

Plaintiffs have sufficiently demonstrated that common evidence of damages predominates over individualized issues.

d. Relevant Product and Geographic Markets

Defendants argue that Plaintiffs will be unable to establish relevant product and geographic markets through common evidence.7 The Court disagrees. With respect to the relevant product market, determining whether co-ops and partial-line suppliers are reasonable substitutes for full-line wholesalers does not require individualized proof. As to the relevant geographic market, the market is defined as the area in which "a defendant supplier draws a sufficiently large percentage of its business" and "only a small percentage of purchasers have alternative suppliers to whom they could practicably turn in the event that a defendant supplier's anticompetitive actions result in a price increase." Little Rock Cardiology Clinic PA v. Baptist Health, 591 F.3d 591, 598 (8th Cir. 2009). This analysis largely involves common evidence, and individual issues do not predominate.

2. Superiority

Federal Rule of Civil Procedure 23(b)(3) also requires a court to find "that a class action is superior to other available methods for the fair and efficient adjudication of the controversy." Rule 23(b)(3) provides that the superiority analysis consider the following factors:

Here, the members of the proposed classes do not appear interested in individually controlling separate actions. The Court has not been advised of any other litigation commenced by members of the class. Concentrating this litigation is desirable to efficiently resolve the Midwest Plaintiffs' claims. Lastly, at this stage in the litigation, the Court does not believe that Plaintiffs' proposed class action will be unmanageable. Therefore, the class action mechanism is superior to any alternative form of litigation for the Midwest Plaintiffs' claims against Defendants.

IV. CONCLUSION

Based upon the foregoing, and all the files, records, and proceedings herein, IT IS HEREBY ORDERED that:

1. Plaintiffs' motion for class certification [Docket No. 607] is GRANTED IN PART and DENIED IN PART as follows:

2. The five certified Classes are defined as follows:

3. Excluded from each of the five Classes are:

4. The Court hereby appoints D&G, Inc. d/b/a Gary's Foods as the representative of the Champaign DC Non-Arbitration Class; Blue Goose Super Market, Inc. as the representative of the Champaign DC Arbitration Class; Nemecek Markets, Inc. as the representative of the Green Bay DC Class; Millennium Operations, Inc. d/b/a R.C. Dick's Market as the representative of the Hopkins DC Class; and Elkhorn-Lueptows, Inc., Jefferson Lueptow's, Inc., and East Troy Lueptows, Inc. as the representatives of the Pleasant Prairie Class; and

5. The law firms of Boies, Schiller & Flexner LLP, Kotchen & Low LLP and Lockridge Grindal Nauen PLLP are designated as Co-lead counsel for the five Classes. Lockridge Grindal Nauen PLLP is designated as liaison counsel for the five Classes.

BY THE COURT: