Mr. Dee's Inc. v. Inmar, Inc.

Full title: MR. DEE'S INC., et al. Plaintiffs, v. INMAR, INC., et al. Defendants.

Court: United States District Court, Middle District of North Carolina

Date published: Sep 16, 2021

1:19cv141 (M.D.N.C. Sep. 16, 2021)

Opinion

1:19cv141

09-16-2021

MR. DEE'S INC., et al. Plaintiffs, v. INMAR, INC., et al. Defendants.

L. Patrick Auld United States Magistrate Judge

MEMORANDUM OPINION AND ORDER

L. Patrick Auld United States Magistrate Judge

This matter comes before the Court on “Defendants' Motion to Exclude the Supplemental Report and Testimony of Plaintiffs' Expert Witness Pursuant to FRE 702 and Daubert” (Docket Entry 212) (the “Expert Motion”). For the reasons that follow, the Court will deny the Expert Motion, such that the Court may consider the proposed expert testimony in deciding whether class certification remains appropriate.¹

1 The undersigned United States Magistrate Judge enters an order rather than a recommendation because “[a] Daubert motion to exclude testimony presents a non-dispositive matter, ” Masimo Corp. v. Philips Elec. N. Am. Corp., 62 F.Supp.3d 368, 388 (D. Del. 2014).

BACKGROUND

I. Procedural History

Alleging anticompetitive behavior in violation of Section 1 of the Sherman Act, Mr. Dee's Inc., Retail Marketing Services, Inc., and Connecticut Food Association (the “Plaintiffs”) seek relief in this action against Inmar, Inc. (“Inmar”), Carolina Manufacturer's Services (“CMS”), ² Carolina Services, ³ and Carolina Coupon Clearing, Inc. (“CCC”)⁴ (the “Defendants”). (Docket Entry 145 (the “Operative Complaint”), ¶¶ 1, 114.) According to the Operative Complaint, Defendants conspired with International Outsourcing Services, LLC (“IOS”) “to allocate customers and markets and to fix prices” (id., ¶ 1), which scheme allegedly resulted in increased coupon-processing fees for entities that transacted with Defendants or IOS. (See id., ¶¶ 1-7.)

2 CMS, a subsidiary of Inmar, “sells coupon[-]processing services to manufacturers.” (Docket Entry 145, ¶ 7.)

3 “Purported Defendant ‘Carolina Services' is not a separate entity, but a d/b/a of [Carolina Coupon Clearing, Inc.].” (Docket Entry 141 at 3.)

4 CCC, likewise a subsidiary of Inmar, “sell[s] retail processing services to retailers, retail co-operatives/wholesalers, and state associations.” (Docket Entry 145, ¶ 7.)

Plaintiffs originally sought to bring this action, pursuant to [Federal Rule of Civil Procedure] 23 [(“Rule 23”)], on behalf of . . . [a]ll entities in the United States that have been overcharged by IOS or [Defendants] for coupon[-]processing services from January 1, 1997 through the present (and continuing until the effects of IOS's and [Defendants]'s fraudulent and anticompetitive scheme ceases).

(Id., ¶ 96.) To that end, Plaintiffs moved to certify two classes: (1) “[a]ll manufacturers in the United States that have purchased coupon[-]processing services directly from IOS, Defendants, or their subsidiaries or affiliates, from April 11, 2001 through March 28, 2007” (Docket Entry 151 at 15); and (2) “[a]ll retailers or retailer associations in the United States that have purchased coupon[-]processing services directly from IOS, Defendants, or their subsidiaries or affiliates, from April 11, 2001 through March 28, 2007” (id. at 15-16). (See Docket Entry 150 (the “Class Certification Motion”).)

In connection with the Class Certification Motion, Plaintiffs submitted a report from Kathleen Grace (“Grace”), whom Plaintiffs tendered as an expert (see Docket Entry 151 at 15). (Docket Entry 151-12 (the “Original Grace Report”).) Defendants opposed the Class Certification Motion (see Docket Entry 158), attaching their own expert report from Michael Kheyfets (“Kheyfets”) (Docket Entry 159-1 (the “Original Kheyfets Report”)). Plaintiffs replied (Docket Entry 172), submitting a rebuttal report from Grace (Docket Entry 172-1 (the “Rebuttal Grace Report”)).⁵ Concurrent with their opposition to the Class Certification Motion, Defendants moved to exclude Grace's testimony and the Original Grace Report. (Docket Entry 160 (the “Old Expert Motion”); see also Docket Entries 161 (Declaration), 162 (supporting memorandum).) Plaintiffs responded in opposition (Docket Entry 175), ⁶ and Defendants replied (Docket Entry 188).

5 Plaintiffs later filed a corrected version of the Rebuttal Grace Report (Docket Entry 205-5).

6 Plaintiffs later filed a corrected version of their memorandum opposing the Old Expert Motion (Docket Entry 211-2).

During the pendency of the Class Certification Motion and the Old Expert Motion, Plaintiffs sought to compel Defendants to produce certain data that Defendants had withheld during discovery. (Docket Entry 166; see also Docket Entry 167 (supporting memorandum).) The Court (per the undersigned United States Magistrate Judge) granted that request, ordering Defendants to “produce to Plaintiffs, for the period from 2000-09, ‘[a]ll coupon processing transactional data'” (Docket Entry 187 (the “Discovery Order”) at 28 (quoting Docket Entry 141-2 at 11)). Together with that relief, the Court authorized Plaintiffs to file a supplement regarding the Class Certification Motion and/or the Old Expert Motion, “limited to discussion of the [above-referenced] coupon processing transactional data produced by Defendants” (id. at 28-29).

On December 23, 2020, Plaintiffs filed two such supplements (Docket Entries 193 (the “First Supplement”), 196 (the “Second Supplement”)), along with another report from Grace (Docket Entry 193-2 (the “Supplemental Grace Report”)). Via the First Supplement, Plaintiffs proposed the following narrowed class definitions:

(1) manufacturers listed in “Appendix A” [of the Supplemental Grace Report], which are those that paid observably higher CCC or IOS shipping fees during the conspiracy period (April 11, 2001 to March 28, 2007); and

(2) retailers or retailer associations listed in “Appendix B” [of the Supplemental Grace Report], which are those that paid observably higher CCC shipping fees during the conspiracy period.

(Docket Entry 193 at 7 (internal commas omitted).) The Second Supplement (i) explained how Grace analyzed the newly produced transactional data (Docket Entry 196 at 4-7, 9-12), (ii) disclaimed reliance on the analysis in the Original Grace Report (id. at 4), (iii) argued that Plaintiffs have mooted Defendants' challenges to the Original Grace Report (id.), and (iv) asked the Court to deny the Old Expert Motion (id. at 12). Because Plaintiffs abandoned the analysis in the Original Grace Report, the Court (per the undersigned) terminated the Old Expert Motion as moot, but noted that Defendants could “properly challenge the new opinions [in the Supplemental Grace Report].” (Text Order dated Dec. 29, 2020.)

On May 12, 2021, Defendants filed the Expert Motion, thereby renewing their effort to exclude Grace's opinions, to include the Supplemental Grace Report. (Docket Entry 212; see also Docket Entry 213 (supporting memorandum).) Another expert report from Kheyfets accompanied the Expert Motion. (Docket Entry 212-2 (the “Supplemental Kheyfets Report”).) Plaintiffs opposed the Expert Motion (see Docket Entry 221), tendering numerous exhibits, to include a declaration from Grace (Docket Entry 221-8 (“Grace Declaration”)).⁷ Defendants replied. (Docket Entry 227.)

7 Plaintiffs simultaneously filed sealed, unredacted versions of their opposition brief and an accompanying exhibit. (See Docket Entries 223, 223-4.)

II. Factual Background

A. Coupon Industry Overview

Manufacturers of consumer goods issue coupons to consumers, who can redeem the coupons at retailers. (Docket Entry 151-12, ¶ 8; accord Docket Entry 159-1, ¶¶ 21-22.) Consumers receive a discount off the purchase price from the retailer, which then seeks reimbursement from the manufacturer that issued the coupon. (Docket Entry 151-12, ¶ 8; accord Docket Entry 159-1, ¶ 22.) Two types of coupon processors facilitate those transactions between retailers and manufacturers: retail coupon processors (“Retail Processors”) and manufacturer coupon processors (“Manufacturer Processors”). (Docket Entry 151-12, ¶ 8; accord Docket Entry 159-1, ¶ 22.)

“Retailers collect coupons redeemed at their stores and send them to Retail Processors. Retail Processors then sort and count the coupons, and send the coupons and invoices for the coupon values to manufacturers or to [M]anufacturer[ P]rocessors.” (Docket Entry 145, ¶ 13 (internal parenthetical omitted); accord Docket Entry 151-12, ¶ 8; Docket Entry 159-1, ¶ 22.) Manufacturers who rely on counts by Retail Processors utilize “one-count” processing, in contrast to “two-count” processing, by which manufacturers direct Manufacturer Processors to conduct a second count. (Docket Entry 159-1, ¶ 22 n.16.) Under two-count processing, Manufacturer Processors audit coupons “to ensure that the Retail Processor has submitted an accurate count of coupons, that the coupons were properly redeemed, and that coupons were not otherwise fraudulently submitted.” (Docket Entry 145, ¶ 13; accord Docket Entry 159-1, ¶ 22 n.16.) After Manufacturer Processors determine the value of coupons redeemed, manufacturers reimburse retailers. (Docket Entry 151-12, ¶ 8; accord Docket Entry 159-1, ¶ 22.)

Regarding the pricing of coupon-processing services, retailers typically pay processing fees to Retail Processors. (Docket Entry 151-12, ¶ 10; accord Docket Entry 159-1, ¶ 134 (Exhibit 28).) In addition, manufacturers often pay incremental fees, including special handling fees and shipping fees. (Docket Entry 151-12, ¶ 8; accord Docket Entry 159-1, ¶ 42 (Exhibit 1).) Under a full-service agreement, manufacturers pay such fees to the Retail Processor; under a pay-direct arrangement, manufacturers remit payment directly to the retailer. (Docket Entry 151-12, ¶ 40; accord Docket Entry 159-1, ¶ 42 (Exhibit 1 Notes)). However, manufacturers may “refuse to pay [incremental] fees and ‘chargeback' the value of the fees to the Retail Processor.” (Docket Entry 145, ¶ 15; accord Docket Entry 151-12, ¶ 10; Docket Entry 159-1, ¶ 42.) In that circumstance, a Retail Processor may “extract the incremental fee from a [retailer] by subtracting the value of the manufacturer's incremental fee chargeback from the amount the Retail Client receives for coupons submitted by the [retailer] to the Retail Processor.” (Docket Entry 145, ¶ 15; accord Docket Entry 159-1, ¶¶ 38, 41.) When Retail Processors deduct the chargeback in that manner, retailers “that purchase directly from a manufacturer may simply deduct the value of the incremental fee chargeback from the amount the retailer owes the manufacturer or wholesaler in connection with the purchase of product.” (Docket Entry 145, ¶ 17; accord Docket Entry 159-1, ¶¶ 38, 41.)

B. Nature of the Alleged Conspiracy

According to the Operative Complaint:

“Inmar and another processor named NCH Marketing Services Inc. (‘NCH') each have roughly 50 percent share of th[e] market [for

manufacturer coupon-processing services]. NCH focuses its business on serving the processing needs of large manufacturers, while Inmar [through its subsidiary CMS] serves small and large manufacturers.” (Id., ¶ 22.) Historically, Inmar and NCH likewise offered retail coupon-processing services (id., ¶ 26), with Inmar (through its subsidiary CCC doing business as Carolina Services) serving mostly non-deducting retailers and NCH “perform[ing] retail coupon[-]processing services for large, deducting retailers (such as mass merchandisers)” (id.). IOS used to participate in both markets (for manufacturer and retail coupon-processing services). (Id., ¶¶ 25, 30.)

As the Court previously explained, the Operative Complaint alleges that:

“On July 17, 2000, Chris Balsiger, CEO of IOS, sent Robert Carter, President of CMS, a letter concerning ‘the large increases that [IOS has] seen in charge back rates to coupon submitters that do not have deduct capability[.]'” The letter blamed Inmar and its subsidiaries for the increase in chargebacks and threatened those entities with litigation.

Following that threat, “Balsiger entered a series of related agreements with Inmar on behalf of IOS intended to restrain competition and increase prices through: (a) market and customer allocation; (b) price fixing; and (c) transfer of confidential retailer client data to Inmar.” Such agreements “eliminated IOS as a competitive threat to Inmar in the manufacturer coupon[-]processing market; eliminated Inmar as a competitive threat to IOS in the retail coupon[-]processing market; and assured substantial profit increases for both IOS and Inmar by fixing and raising coupon[-]processing fees.”

In 2001, IOS exited the manufacturer coupon[-]processing market and turned over its clients to CMS, in exchange for CMS sharing its revenues with IOS and engaging in a broader conspiracy (involving Inmar and IOS). That alleged conspiracy involved (i) IOS selling Retail Client

data to Inmar and Inmar refraining from competition with IOS (to include participating in a program under which Inmar accepted, without auditing, IOS's coupon count), (ii) IOS subprocessing coupons for CCC's retail customers and not providing coupon services to CCC's customers, (iii) IOS and Inmar jointly marketing services to mass merchandise retailers, sharing revenue, and not competing for the business of such retailers, and (iv) IOS transferring its manufacturer customers to Inmar and refraining from providing services to such customers. Those “agreements had the purpose and effect of fixing the prices of retail coupon[-]processing fees.” Via the anti-competitive scheme, IOS and Inmar generated substantial profits by increasing incremental fees (to include freight fees), charging additional types of fees, and invoicing fewer coupons at a time (to further increase fees).

(Docket Entry 239 at 3-5 (internal citations and brackets omitted).)

III. Expert Submissions

By way of background, the Court summarizes the principles and methodologies underlying the proffered expert analyses. As detailed above, Plaintiffs have alleged that Defendants engaged in anticompetitive behavior in violation of the Sherman Act, a federal antitrust statute. Anticompetitive behavior, such as “[c]ollusive or monopolistic practices[, ] may increase the price that consumers or businesses pay for what they buy.” American Bar Association, Proving Antitrust Damages: Legal and Economic Issues pt. I ch. 1 (3d ed. 2017). To determine the impact of an antitrust violation, courts use “the familiar ‘but-for' proposition” and ask whether “the plaintiff would be in a better financial position but for the violation, ” Id. The violation caused injury “[i]f the plaintiff's actual condition is worse than what it would have been in the hypothetical but-for world, ” Id. Assuming proof of injury and causation, “a damages award measures the extent of the difference in outcomes between the actual world and the but-for world to compensate the plaintiff for this injury.” Id. Courts “accept a degree of uncertainty in . . . ascertaining business damages . . . [because t]he vagaries of the marketplace usually deny [courts] sure knowledge of what [the] plaintiff's situation would have been in the absence of the defendant's antitrust violation.” J. Truett Payne Co. v. Chrysler Motors Corp., 451 U.S. 557, 566 (1981). However, a damages computation may not involve “speculation or guesswork.” Bigelow v. RKO Radio Pictures, Inc., 327 U.S. 251, 264 (1946).

The before-during-after model presents one way of quantifying damages. “Th[is] model[] seek[s] to isolate harm flowing from the challenged anticompetitive conduct by comparing the plaintiff's experience in the same market before, or after, the misconduct with its experience during the time the conduct was ongoing, which is sometimes referred to as the ‘damages period.'” American Bar Association, supra, pt. II ch. 4 subsec. C. Another option, the “yardstick” model (alternately, the “benchmark” model), “seek[s] to isolate harm flowing from the challenged anticompetitive conduct by comparing the experience of the plaintiff in a market subject to anticompetitive conduct to the experience of the plaintiff, or another firm, in a comparable market unaffected by the defendant's violation.” Id. A third model, “the difference-in-differences approach[, ] combines the before-during and benchmark approaches [by] compar[ing] the differences in outcomes before and after the alleged anticompetitive conduct across the benchmark and affected markets.” Id., pt. II ch. 6 subsec. G.

Of further significance:

Because the before-during-after approach looks outside the damages period in the same market, and the yardstick approach looks to a different market, one may need to examine changes in economic conditions over time or across markets, the plaintiff's circumstances, or other variables which could affect prices, sales, and profits. One tool widely used by economists to account for such changes in economic conditions is a statistical technique known as multiple regression analysis. In the context of antitrust damages, multiple regression analysis is useful because it allows an expert to control for variables other than the anticompetitive conduct that may change between the damages period, or the market in question, and the benchmark or yardstick used to model the but-for world.

Id., pt. II ch. 4 subsec. C (internal footnote omitted). In a simple regression analysis, one can model the relationship between an explanatory variable (like cost) and a dependent variable (like price). Id., pt. II ch. 6 subsec. C(1). Multiple regression analysis reflects the reality of “multiple influences on the dependent variable.” Id., pt. II ch. 6 subsec. C(2). “[W]hen used properly[, ] multiple regression analysis is one of the mainstream tools in economic study and . . . an accepted method of determining damages in antitrust litigation.” In re Flat Glass Antitrust Litig., 191 F.R.D. 472, 486 (W.D. Pa. 1999) (collecting cases).

A. Grace

Grace received a B.A. in economics and finance from Wofford College and obtained a master's degree and PhD in quantitative economics from Clemson University. (Docket Entry 151-12 at 2.) She has worked as an investment banking analyst and professor. (Id.) First at Furman University and later at Wofford College, Grace taught “in the areas of economic theory, financial economics, corporate finance, and corporate valuation.” (Id.) She previously “ha[s] provided expert opinions and support on corporate valuation techniques, partnership valuation, cash flow and liquidity for multi-national corporations as well as partnerships.” (Id.) She possesses “training and experience in analyzing economic incentives, analyzing economic actions, and reviewing corporate financial records” (id.), and she has contributed to several publications, some of which involved her estimation of overcharges to electricity customers (id. at 3).

Turning to her analyses in this action, in November 2020, as a result of the Discovery Order, Grace received an additional dataset from Defendants that revealed ancillary fees charged to manufacturers by CCC, IOS, and NCH. (Docket Entry 193-2, ¶ 5.) That dataset included “the fee amount invoiced, the fee amount actually paid, the amount of any fee unpaid, and the identity of the organization that charged the fee.” (Id., ¶ 8.) Grace aggregated each coupon program and determined, for each year between 2000 and 2007, a “mean shipping fee payment per 1, 000 coupons for Inmar, NCH, and IOS” (id., ¶ 11). Grace viewed the allegedly anticompetitive agreements between Defendants and IOS, which remained in place from April 11, 2001 until March 28, 2007 (Docket Entry 151-12, ¶ 19) as defining the duration of the alleged conspiracy (Docket Entry 193-2, ¶ 18). For brevity, the Court refers to that time period as the “conspiracy period.”

Using the data from January 2000 through 2007, Grace “conduct[ed] three separate analyses to estimate shipping fee overcharges to manufacturers: (i) a before-during analysis [(“Before-During Analysis”)], (ii) a benchmark analysis [(“Benchmark Analysis”)], and (iii) a difference-in-differences analysis [(“DiD Analysis”)].” (Id., ¶ 17.) To determine the effect of the alleged conspiracy on retailers, Grace “estimate[d] how much retailers paid in shipping fee overcharges” (id.) as a result of manufacturers refusing to pay higher shipping fees and CCC extracting the unpaid shipping fees from retailers (id., ¶ 43). Based on those analyses, Grace compiled “a list of manufacturers and retailers that paid observably higher shipping fees [during the conspiracy period].” (Id., ¶ 18.)

First, via the Before-During Analysis, Grace “model[ed] CCC's paid shipping fee per coupon in the pre-conspiracy period, controlling for volume, and then use[d] that relationship to forecast what a competitive shipping fee would have been absent a conspiracy.” (Id., ¶ 20.) Grace calculated as damages the difference between the amounts that manufacturers paid to CCC in shipping fees and the forecasted competitive shipping fee. (Id.) Grace concluded that some manufacturers who transacted with CCC overpaid by $28.3 million during the conspiracy period. (Id., ¶ 23.) Employing that same methodology for manufacturers served by IOS, Grace calculated overcharge damages as $2.4 million. (Id., ¶¶ 24-27.)

Second, Grace employed the Benchmark Analysis to “model NCH's paid shipping fees per coupon over time, controlling for volume, and then use[d] that relationship to forecast what a competitive shipping fee would have been absent a conspiracy.” (Id., ¶ 29.) According to Grace, such approach “controls for extraneous non-conspiratorial factors that may have affected shipping fees during the [conspiracy] period.” (Id.) Grace calculated overcharges by both CCC and IOS by comparing increases in their shipping fees to the model that Grace used to “establish the relationship between NCH's shipping fees with volume and time.” (Id., ¶ 31.) The Benchmark Analysis yields a damages estimate of $41.7 million. (Id., ¶ 32.)

Third, in connection with the DiD Analysis, Grace “quantif[ied] the relationship between NCH shipping fees (the control) and CCC shipping fees (the treatment group) in the pre-conspiracy period” (id., ¶ 34) by determining “how NCH fees changed over time compared to how CCC fees changed over the same time period, controlling for volume” (id.). As with the Benchmark Analysis, Grace asserted that “[t]he inclusion of NCH . . . controls for extraneous non-conspiratorial factors that may have affected shipping fees during the [conspiracy] period, such as changes in postage rates.” (Id.) Grace found that, before the alleged conspiracy, “for 2-count coupons, CCC's prices were 4.22 times NCH's prices for similar-volume customers, and for 1-count coupons, CCC's prices were 2.67 times NCH's prices.” (Id., ¶ 39.) Grace “calculated as overcharges instances in which CCC's shipping fees increased above NCH shipping fees in excess of [the pre-conspiracy relationship]” (id., ¶ 40), resulting in “a damages estimate of $37.1 million” (id.). However, that estimate pertains only to manufacturers served by CCC, as Grace concluded that “IOS's shipping fee changes do not correlate in any statistically meaningful way to NCH shipping fee changes, making [the DiD A]nalysis inapplicable to IOS's fees.” (Id., ¶ 41.)

Fourth, because “some [manufacturers] simply refused to pay higher shipping fees” (id., ¶ 43) via chargebacks and because CCC could extract those chargebacks from retailers, “[Grace] perform[ed] a before-after regression analysis to assess chargeback changes in the pre-conspiracy period compared to the conspiracy period.” (Id. (describing the “Chargeback Analysis”).) More specifically, Grace relied on a “regression to estimate the relationship between chargebacks and gross shipping for each invoice in the pre-conspiracy period, where gross shipping is equal to shipping dollars paid by the manufacturer plus shipping dollars charged back to retailers” (id., ¶ 44). Using that equation “to estimate what chargebacks should be in the conspiracy period[, Grace] then calculate[d] any difference between observed chargebacks v. forecasted chargebacks in the conspiracy period to estimate damages due to ‘excess chargebacks' to retailers.” (Id.) Grace concluded that certain retailers paid $8.8 million in excess chargebacks during the conspiracy period. (Id.)

B. Kheyfets

Kheyfets obtained a B.A. and M.A. in economics from Boston University. (Docket Entry 212-2 at 65.) His work experience includes analyst positions at Volpe National Transportation Systems Center, NERA Economic Consulting, and Edgeworth Economics, LLC, where he currently holds the position of partner. (Id.) He possesses substantial experience as a consulting expert in antitrust, cybersecurity, and data privacy matters and has contributed to numerous publications and presentations. (Id. at 66-69.)

Kheyfets identified the purpose of the Supplemental Kheyfets Report as assessing whether the analyses in the Supplemental Grace Report can

• [s]erve as common proof of economic impact to all (or nearly all) members of the newly[ ]defined proposed manufacturer class;

• [s]erve as common proof of economic impact to all (or nearly all) members of the newly[ ]defined proposed retailer class; and

• [s]erve as a formulaic approach to reliably estimate damages from the alleged conduct to proposed manufacturer and retailer class members.

(Id., ¶ 4.) Kheyfets concluded that Grace had failed to offer a theory linking the alleged anticompetitive conduct to the damages she purported to calculate (id., ¶¶ 4-7) and that, in any event, her methodologies lacked reliability (as to both the proposed manufacturer class (id., ¶¶ 8-9) and proposed retailer class (id., ¶¶ 10-13)). In the following subsections, the Court describes in more detail Kheyfets's critiques on each of those subjects.

1. Absence of Link Between Alleged Conspiracy and Shipping Fee Increases

The Supplemental Grace Report differs from the Original Grace Report, which purported to identify damages to a broader class of manufacturers and retailers based on several types of fees. (Id., ¶¶ 14-16.) More specifically, the newly proposed classes omit more than 2, 500 manufacturer accounts and 81 retailer accounts for which Defendants produced data and focus solely on shipping fees. (Id., ¶¶ 17-18.) However, in Kheyfets's view, both the data and the Supplemental Grace Report undermine Plaintiffs' assertion that shipping fees constitute “standardized, undifferentiated commodity services, the prices of which are not negotiated” (id., ¶ 19 (quoting Docket Entry 193 at 13)). (See id., ¶¶ 19-21 (explaining that manufacturers paid vastly different rates for shipping).)

As far as Grace's economic theory for how the alleged conspiracy affected shipping fees, Kheyfets opined that Grace failed to account for the fact that manufacturers could (and did) charge back shipping fees to CCC during the conspiracy period. (Id., ¶ 22.) Moreover, some manufacturers dictated their own reimbursement rates for shipping, and chargeback practices varied over time and among manufacturers. (See id.) Given those circumstances, Kheyfets faulted Grace for not explaining why the alleged conspiracy caused only certain entities to pay higher shipping fees. (Id., ¶ 23; see also Id. ¶¶ 23-24 (asserting that size of the manufacturer (i.e., average monthly coupon volume) cannot explain differences in chargeback rates, despite Grace's contrary suggestion).) According to “t-tests”⁸ that Kheyfets conducted, among manufacturer accounts that paid shipping fees for two-count coupon processing before and during the alleged conspiracy, only 25 percent of those accounts paid statistically significant increased shipping fees to CCC, and 41 percent paid such fees to IOS. For one-count coupon processing, 41 percent of manufacturer accounts paid statistically significant increased shipping fees to CCC. (Id., ¶ 26 n.48.)

8 The Original Kheyfets Report described t-tests as “standard statistical test[s]” of “the ‘natural null hypothesis'” that an individual customer's shipping fees did not increase (to a statistically significant degree) during the conspiracy period. (Docket Entry 159-1, ¶ 82.)

Per Kheyfets, the Supplemental Grace Report reflects several other shortcomings related to the causation of economic harm. First, Grace erred in positing that competition among retail processors would forestall shipping fee increases (id., ¶ 25 (discussing Docket Entry 193-2, ¶ 15)), given that manufacturers lacked contractual relationships with retail processors or control over which retailer processor a retailer employed. (Id., ¶ 25 & n.46.) Second, a document upon which Grace relied to demonstrate Defendants' acknowledgment of an anticompetitive shipping fee strategy does not show that the relationship between CCC and IOS (and resulting coordination on other matters) caused an increase in shipping fees. (Id., ¶ 27.) Third, Grace failed to (i) analyze whether Defendants inflated shipping fees by including fewer coupons per invoice (as Plaintiffs have asserted)⁹ and (ii) explain how such change in invoicing practices could affect manufacturers that paid a fixed per-coupon shipping fee. (Id., ¶ 28.) Fourth, Grace ignored the existence of individualized rebate agreements by which some manufacturers nominally paid shipping fees (instead of charging them back) in exchange for a negotiated shipping fee rebate. (Id., ¶¶ 29-33.)

9 Kheyfets conducted another t-test to assess whether average No. of coupons per invoice changed during the conspiracy period, which revealed that only 22 percent of manufacturer accounts experienced a statistically significant decrease in the No. of coupons per invoice. (Id., ¶ 28 n.54.)

2. Problems with Estimated Damages for Manufacturer Class At the threshold, Kheyfets flagged a No. of issues with Grace's calculation of “aggregate shipping fee trends” (id., ¶ 35 (quoting Docket Entry 193-2, ¶¶ 11-16)). For example, Grace's analysis shows that, in 2000 (prior to the alleged conspiracy), Inmar's average per-coupon shipping fee tripled that of NCH, and IOS's doubled Inmar's. (See Id. (discussing Supplemental Grace Report, Figure 2).) As a result, Kheyfets maintained that a comparison of shipping fees across processors should account for such pre-conspiracy differences. (Id.) In addition, each processor's average per-coupon shipping fee varied during the conspiracy period: IOS charged lower shipping fees in 2003 and 2004 (relative to 2000), and, in 2005, Inmar's shipping fees declined (relative to 2004) while IOS's increased (see id.), but Grace has failed to reconcile those variations with the alleged conspiracy to elevate shipping fees (see id.). Finally, Kheyfets stated that (contrary to Grace's conclusion) a comparison of the shipping fee paid by the median-volume manufacturer during each year between 2000 and 2007 fails to illustrate “‘market-wide' pricing behavior.” (Id.)

Turning to the regressions Grace used to calculate classwide damages, Kheyfets identified the following general problems:

• Grace's analysis obscured important variations between manufacturers, like differences in their willingness to pay shipping fees (id., ¶ 40);

• Grace “control[led] for volume” (id.) but failed to appreciate that manufacturers that processed similar volumes paid “vastly different shipping fees” (id.); and

• Grace purported to control for non-conspiratorial “economic factors like postage rates and labor costs” but declined to explain how her models accounted for such variables (id., ¶ 41).

Regarding Grace's application of the three methodologies, Kheyfets characterized the results as internally inconsistent and economically “nonsensical.” (Id., ¶ 42.) More specifically, for approximately sixty percent of the “manufacturer accounts in the proposed class that paid shipping fees to CCC” (id., ¶ 44), the three methodologies produce divergent results as to overcharge damages. (See id., ¶¶ 43-45.) In other words, for the same manufacturer account, two methodologies identify damages whereas the other methodology finds none. (See id., ¶ 44.) Morever, each methodology predicts that many manufacturer accounts that paid no shipping fees during the proposed class period would have paid shipping fees in the but-for world. (See id., ¶¶ 48, 55, 62.) Each methodology also forecasts a negative shipping fee for some coupons, which suggests that, in the but-for world, the processor would have paid, rather than charged, the manufacturer for shipping its coupons. (See id., ¶¶ 48, 56, 63.)¹⁰ In one case, Kheyfets reported that Grace “manually overwr[o]te any negative but-for price generated by her [Benchmark Analysis] for 2-count coupons with a price of $1 per 1, 000 coupons.” (Id., ¶ 56.)

10 Kheyfets further identified concerns specific to each methodology. (See id., ¶¶ 49 (explaining that Before-During Analysis produced $9.6 million in negative damages or undercharges), 52 (challenging assumption underlying Benchmark Analysis that CCC's fees, absent the alleged conspiracy, would mirror NCH's), 61 (noting Grace's concession that DiD Analysis cannot calculate damages for accounts served by IOS).)

Finally, “[t]o assess the[] reliability” of Grace's three methodologies, [Kheyfets] appl[ied them] to transactions on which no anticompetitive effects are alleged.” (Id., ¶ 67.) In particular, Defendants produced data regarding at least five categories of shipping fees as to which Plaintiffs have alleged no conspiracy-related increase: (1) “CCC during the ‘benchmark' period preceding the alleged conspiracy” (id.); (2) “IOS during the ‘benchmark' period preceding the alleged conspiracy” (id.); (3) “non-Defendant NCH during the proposed class period” (id.); (4) “‘pay direct' retailers during the proposed class period” (id. (internal footnote omitted));¹¹ and (5) “‘direct submitter' retailers during the proposed class period” (id. (internal footnote omitted)).¹² In every case, Kheyfets found that each methodology calculated damages as to some supposedly conspiracy-free transactions, ranging from 18 percent of accounts to 97 percent of accounts. (Id., ¶ 68.) In Kheyfets's view, those “false positives” demonstrate that Grace's analyses cannot “distinguish[] transactions affected by the alleged conduct from those that were not.” (Id.)

11 According to Kheyfets, such retailers “used retailer processors like IOS and CCC but set their own shipping fees” (id., ¶ 67 n.130).

12 Kheyfets reported that those retailers “bypassed retailer processors like IOS and CCC altogether, submitting their invoices directly to CMS (a manufacturer processor) and set their own shipping fees” (id., ¶ 67 n.131).

3. Problems with Estimated Damages for Retailer Class Kheyfets challenged “Grace's assertion that any shipping fee a manufacturer chose to charge back would necessarily be ‘paid by the CCC retail client'” (id., ¶ 70), which “obscures th[e] fact” that CCC did not automatically pass through the full amount of every chargeback (id.). Illustrating that point, Kheyfets determined that more than 80 percent of the proposed retailer class paid no shipping fee chargebacks during the conspiracy period. (Id., ¶ 71.) Furthermore, even among the chargebacks that CCC attempted to pass through, Kheyfets noted that retailers retained the ability to deduct payment from the manufacturer who charged back the shipping fee. (Id., ¶ 72.) According to Kheyfets, the impact on retailers that Grace attempted to calculate would require her to: (1) “[s]how that CCC charged higher shipping fees to manufacturers than it would have in the but-for world” (id., ¶ 75); (2) “[s]how that manufacturers charged back more in total shipping fees than they would have in the but-for world” (id.); (3) “[s]how that the amounts of shipping fee chargebacks CCC charged to all (or nearly all) proposed retailer class members were greater than in the but-for world” (id. (internal footnote omitted)); and (4) “[s]how that proposed retailer class members actually paid more in shipping fee chargebacks and did not increase their deductions to manufacturers or receive higher rebates to offset any increase in these charges” (id. (internal footnote omitted)).

Regarding the specifics of Grace's Chargeback Analysis, she allocated shipping fees and chargebacks to retailers appearing on a CCC invoice and concluded that, before the conspiracy period, CCC passed on 24.5 cents in chargebacks for each additional dollar in shipping fees. (Id., ¶ 76.) Grace calculated as damages any instance when, during the conspiracy period, charges to a retailer exceeded that rate. (Id.) However, Grace ignored the $16.5 million in undercharges that her model produced (i.e., instances when actual chargebacks during the conspiracy period fell below her model's prediction). (Id., ¶ 77.) Accounting for such negative damages, only 20 members of the proposed 304-member retailer class sustained any “positive” damages during the conspiracy period. (Id., ¶ 78.) Finally, like the other analyses, the Chargeback Analysis discerns impact (in the form of “excess chargebacks”) to pay-direct retailers that set their own shipping fees (rather than deferring to Defendants on that front). (Id., ¶ 80.)

DISCUSSION

I. Relevant Standards

A. Sherman Act

Section 1 of the Sherman Act prohibits “[e]very contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade, ” 15 U.S.C. § 1. “To prove a violation of [that provision], a plaintiff must show the existence of an agreement in the form of a contract, combination, or conspiracy that imposes an unreasonable restraint on trade.” Oksanen v. Page Mem'l Hosp., 945 F.2d 696, 702 (4th Cir. 1991) (emphasis added); see also Wilder Enters., Inc. v. Allied Artists Pictures Corp., 632 F.2d 1135, 1139 n.1 (4th Cir. 1980) (“The necessary elements for recovery under [Section] 1 of the [Sherman] Act are: (1) an agreement, conspiracy, or combination among the defendants in restraint of trade; (2) injury to the plaintiff's business and property as a direct result; (3) damages that are capable of reasonable ascertainment and are not speculative or conjectural.”). The second element, injury, involves two distinct inquiries: (1) “whether the plaintiff has indeed suffered harm, or ‘injury-in-fact, '” Cordes & Co. Fin. Servs. v. A.G. Edwards & Sons, Inc., 502 F.3d 91, 106 (2d Cir. 2007); and (2) “whether any such injury is ‘injury of the type the antitrust laws were intended to prevent and that flows from that which makes [the] defendants' acts unlawful, '” Id. (quoting Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 489 (1977)). As to that second prong, “[t]o establish an antitrust injury, [a p]laintiff['s] losses must stem from anticompetitive acts . . . that reduce output or raise prices to consumers.” R. J. Reynolds Tobacco Co. v. Philip Morris, Inc., 199 F.Supp.2d 362, 395 (M.D. N.C. 2002) (internal quotation marks omitted).

It bears emphasizing that “injury-in-fact” remains distinct from the third element, damages. In that regard, “injury-in-fact is whether the plaintiffs were harmed and damages quantify by how much.” In re Ethylene Propylene Diene Monomer (EPDM) Antitrust Litig., 256 F.R.D. 82, 88 (D. Conn. 2009). However, a single calculation may show both injury-in-fact and damages: “if [a plaintiff] paid a higher actual price than the but-for price, he has shown [injury-in-fact]; by calculating the extent of that difference[, ] he has shown damages.” Id. Proving damages necessarily proves injury, but anticompetitive conduct may cause injury without damages. Id. For example, “a plaintiff [may] suffer antitrust injury-in-fact and yet have no damages because it has taken steps to mitigate the actual price paid through rebates, discounts, and other non-price factors such as lowered shipping costs, technical services, or any other type of purchase incentive.” Id. at 88-89.

B. Class Actions

Rule 23 authorizes “[o]ne or more members of a class” to sue as representative parties on behalf of all members only if:

(1) the class is so numerous that joinder of all members is impracticable;

(2) there are questions of law or fact common to the class;

(3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and (4) the representative parties will fairly and adequately protect the interests of the class.

Fed. R. Civ. P. 23(a). Courts commonly refer to those “threshold requirements applicable to all class actions . . . as ‘numerosity,' ‘commonality,' ‘typicality,' and ‘adequacy.'” Krakauer v. Dish Network, L.L.C., 925 F.3d 643, 654 (4th Cir. 2019). Rule 23 also “contains an implicit threshold requirement that the members of a proposed class be ‘readily identifiable.'” EQT Prod. Co. v. Adair, 764 F.3d 347, 358 (4th Cir. 2014) (describing what other courts have referred to as “ascertainability” requirement).

“After satisfying these threshold requirements, Plaintiffs must ‘demonstrat[e] that the proposed class fits into one of the specific forms of class adjudication provided by Rule 23(b).'” Koepplinger v. Seterus, Inc., No. 1:17CV995, 2020 WL 2063416, at *28 (M.D. N.C. Apr. 29, 2020) (unpublished) (quoting Krakauer, 925 F.3d at 655), recommendation adopted, 2020 WL 5705915 (M.D. N.C. Aug. 26, 2020) (unpublished). One option, Rule 23(b)(3), requires “the plaintiff [to] show both that ‘[1] questions of law and fact common to class members predominate over any questions affecting only individual class members, and [2] that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.'” Krakauer, 925 F.3d at 655 (quoting Fed.R.Civ.P. 23(b)(3) for “predominance” and “superiority” requirements). That type of class action “requires notice to class members, who are afforded an opportunity to opt-out of the class at the certification stage.” Id.

Notwithstanding the foregoing requirements, “Rule 23 explicitly envisions class actions with . . . individualized damage determinations.” Gunnells v. Healthplan Servs., Inc., 348 F.3d 417, 428 (4th Cir. 2003) (discussing Fed.R.Civ.P. 23(c)(4), which allows class certification as to particular issues). Accordingly, “the need for individualized proof of damages alone will not defeat class certification.” Id. at 429. The United States Court of Appeals for the Fourth Circuit has noted that,

“in actions for money damages under Rule 23(b)(3), courts usually require individual proof of the amount of damages each member incurred.” When such individualized inquiries are necessary, if “common questions predominate over individual questions as to liability, courts generally find the predominance standard of Rule 23(b)(3) to be satisfied.”

Id. at 428 (internal citation omitted) (quoting 5 Moore's Federal Practice § 23.46 [2][a] (1997)).

C. Expert Testimony

Federal Rule of Evidence 702 (“Rule 702”) governs the admissibility of expert testimony in federal courts. In re Lipitor (Atorvastatin Calcium) Mktg. v. Pfizer, Inc., 892 F.3d 624, 631 (4th Cir. 2018). Although the Federal Rules of Evidence otherwise limit the admissibility of opinion testimony, Rule 702 authorizes “experts qualified by ‘knowledge, skill, experience, training, or education' [to] submit an opinion.” Thomas J. Kline, Inc. v. Lorillard, Inc., 878 F.2d 791, 799 (4th Cir. 1989) (quoting Fed.R.Evid. 702). In that regard, “to decide whether the expert has sufficient specialized knowledge to assist the jurors in deciding the particular issues in the case, ” Belk, Inc. v. Meyer Corp., 679 F.3d 146, 162 (4th Cir. 2012) (internal quotation marks omitted), courts must “consider the proposed expert's full range of experience and training, ” Id. (internal quotation marks omitted).

Once a court determines that a proposed expert possesses the requisite qualifications, it must make “a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue.” Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 592-93 (1993) (citing Fed.R.Evid. 104(a)). In particular,

[Rule 702] permits an expert to testify where the expert's “scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue, ” so long as the expert's opinion is “based on sufficient facts or data, ” “is the product of reliable principles and methods, ” and the expert “has reliably applied the principles and methods to the facts of the case.”

Pfizer, Inc., 892 F.3d at 631 (quoting Fed.R.Evid. 702). “Implicit in the text of Rule 702 . . . is a district court's gatekeeping responsibility to ‘ensur[e] that an expert's testimony both rests on a reliable foundation and is relevant to the task at hand.'” Nease v. Ford Motor Co., 848 F.3d 219, 229 (4th Cir. 2017) (quoting with emphasis Daubert, 509 U.S. at 597).

In other words, Rule 702 effectively imposes a relevance threshold, insofar as it “requires a valid scientific connection to the pertinent inquiry as a precondition to admissibility.” Daubert, 509 U.S. at 591-92. As concerns the reliability of a proposed expert's testimony, factors that may bear on that determination

include: (1) whether a theory or technique can be or has been tested; (2) whether it has been subjected to peer review and publication; (3) whether a technique has a high known or potential rate of error and whether there are standards controlling its operation; and (4) whether the theory or technique enjoys general acceptance within a relevant scientific community.

Cooper v. Smith & Nephew, Inc., 259 F.3d 194, 199 (4th Cir. 2001). However, such factors “[a]re neither definitive[] nor exhaustive.” Id.; see also Daubert, 509 U.S. at 594-95 (emphasizing flexible nature of Rule 702 inquiry). The Rule 702 inquiry focuses “solely on principles and methodology, not on the conclusions that they generate.” Daubert, 509 U.S. at 595. The United States Supreme Court has acknowledged, however, that “conclusions and methodology are not entirely distinct from one another.” General Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997). For that reason, exclusion may result when a court “conclude[s] that there is simply too great an analytical gap between the data and the opinion proffered.” Id.

As far as the applicable procedure, “[t]he trial court must have the same kind of latitude in deciding how to test an expert's reliability, and to decide whether or when special briefing or other proceedings are needed to investigate reliability, as it enjoys when it decides whether that expert's relevant testimony is reliable.” Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999); see also Cooper, 259 F.3d at 199 (“The proponent of the testimony must establish its admissibility by a preponderance of proof.”). “[A] court is not required to hold a hearing simply because a party has raised a Daubert issue.” United States v. Davis, 602 F.Supp.2d 658, 663 (D. Md. 2009). Nonetheless, “the parties must ‘have an adequate opportunity to be heard' before the trial court makes its decision.” Dealers Supply Co., Inc. v. Cheil Indus., Inc., No. 1:03CV654, 2007 WL 4790800, at *2 (M.D. N.C. Mar. 8, 2007) (unpublished) (quoting Group Health Plan, Inc. v. Philip Morris USA, Inc., 344 F.3d 753, 761 n.3 (8th Cir. 2003)). Importantly, “‘the trial court's role as a gatekeeper is not intended to serve as a replacement for the adversary system,' and consequently, ‘the rejection of expert testimony is the exception rather than the rule.'” United States v. Stanley, 533 Fed.Appx. 325, 327 (4th Cir. 2013) (quoting Fed.R.Evid. 702 advisory committee's notes, 2000 Amendment).¹³

13 “As in all questions of admissibility, the proferring party must come forward with evidence from which the court can determine that the proffered testimony is properly admissible. However, there is no requirement in Daubert, or any other controlling authority, that the proffering party must ‘prove' anything to the court before the testimony in question can be admitted.” Maryland Cas. Co. v. Therm-O-Disc, Inc., 137 F.3d 780, 783 (4th Cir. 1998).

II. Analysis

A. Preliminary Matters

As an initial matter, the parties have offered divergent views on the relationship between the Expert Motion and the Class Certification Motion. Defendants have contended that the Court must resolve the Expert Motion before ruling on the Class Certification Motion because their objections to Grace's opinions bear on issues underlying class certification. (Docket Entry 213 at 8.) In contrast, Plaintiffs effectively deny that the Expert Motion carries implications “critical to resolution of the [Class Certification M]otion” (Docket Entry 221 at 25), by suggesting that the Court may grant certification “on the direct evidence that [Defendants] raised shipping fees because of their agreements with IOS, without reaching [] Grace's expert opinions” (id. at 26).

“No controlling precedent dictates whether a district court must conduct a Daubert analysis at the class[-]certification stage or how focused or full that analysis should be.” Williams v. Estates LLC, No. 1: 19CV1076, 2021 WL 1581239, at *5 n.3 (M.D. N.C. Apr. 22, 2021) (unpublished) (collecting conflicting cases). The Court (per the undersigned) endorses the view that “a full or thorough Daubert analysis” remains appropriate at this stage “when ‘expert testimony is in fact critical to class certification.'” Baxley v. Jividen, 504 F.Supp.3d 539, 543 (S.D. W.Va. 2020) (quoting Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 813 (7th Cir. 2012)); accord Robinson v. Nationstar Mortg. LLC, Civ. Action No. 14-3667, 2019 WL 4261696, at *13 (D. Md. Sept. 9, 2019) (unpublished); Childress v. JPMorgan Chase & Co., No. 5: 16CV298, 2019 WL 2865848, at *2 (E.D. N.C. July 2, 2019) (unpublished); Krakauer v. Dish Network, L.L.C., No. 1: 14CV333, 2015 WL 5227693, at *5 (M.D. N.C. Sept. 8, 2015) (unpublished); Coleman v. Union Carbide Corp., Civ. Action No. 2: 11-0366, 2013 WL 5461855, at *22-23 (S.D. W.Va. Sept. 30, 2013) (unpublished).

Whether testimony qualifies as “critical” to class certification depends (at least in part) on the proper scope of the class-certification inquiry. “Rule 23 grants courts no license to engage in free-ranging merits inquiries at the certification stage, ” Amgen Inc. v. Connecticut Ret. Plans & Tr. Funds, 568 U.S. 455, 466 (2013), and “[m]erits questions may be considered to the extent - but only to the extent - that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied, ” id.; however, “a court's class-certification analysis must be ‘rigorous' and may ‘entail some overlap with the merits of the plaintiff's underlying claim, '” Id. at 465-66 (quoting Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 351 (2011)). Highlighting the tension between those principles, the Supreme Court has deemed certification improper under Rule 23(b)(3) when a court “refus[ed] to entertain arguments against [an expert's] damages model that bore on the propriety of class certification, simply because those arguments would also be pertinent to the merits determination, ” Comcast Corp. v. Behrend, 569 U.S. 27, 34 (2013). According to the Supreme Court, to accept at the class-certification stage “any method of measurement” susceptible to classwide application, “no matter how arbitrary[, ] . . . would reduce Rule 23(b)(3)'s predominance requirement to a nullity.” Id. at 36.

Here, the dispute surrounds the extent to which Grace's opinions impact the Class Certification Motion. (Compare Docket Entry 213 at 8, with Docket Entry 221 at 25-26.) On the one hand, in their briefing in support of the Class Certification Motion, Plaintiffs have referenced “direct evidence, including testimony from IOS's CEO” regarding the allegedly anticompetitive agreements between Defendants and IOS (Docket Entry 172 at 6), as well as “Defendants' internal documents” (id.) that reflect (i) trends in Defendants' processing volume and financial performance and (ii) Defendants' supposedly admitted strategy of increasing shipping fees to enhance financial performance (id. at 6-7). On the other hand, Plaintiffs' briefing suggests considerable reliance on Grace's opinions in advocating for class certification. (See Docket Entry 172 at 7-8, 10, 12-15 (citing Grace as authority).) At bottom, although Plaintiffs may not “rely exclusively upon [Grace's] opinions to support the assertion that the[ir] claim can be commonly proven and therefore satisfies [Rule 23] requirements, ” Rhodes v. E.I. DuPont de Nemours & Co., No. 6:06CV530, 2008 WL 2400944, at *5 (S.D. W.Va. June 11, 2008) (unpublished) (emphasis added), the Supplemental Grace Report appears to weigh heavily on the certification issue. Accordingly, the Court will engage in the traditional Daubert analysis to determine whether the Supplemental Grace Report should inform the class-certification decision, bearing in mind the importance of avoiding unnecessary inquiry into the merits of Plaintiffs' claim.

Turning to another threshold issue, Plaintiffs have argued that the Court should deny the Expert Motion as untimely, pointing to the passage of time since the parties finished briefing the Class Certification Motion. (Docket Entry 221 at 25 (noting that five months elapsed before Defendants filed Expert Motion).) For their part, Defendants have indicated that their receipt, in February 2021, of “all of Grace's programs needed to analyze [the] Supplemental [Grace] Report” delayed their preparation of the Expert Motion (and accompanying Supplemental Kheyfets Report). (Docket Entry 213 at 3 n.2.) In addition, Defendants have highlighted that (i) the Court denied the Old Expert Motion without prejudice (and without specifying a deadline to challenge the Supplemental Grace Report) (Docket Entry 227 at 11), (ii) Plaintiffs declined to engage with Defendants concerning the “timetable for filing” the Expert Motion (id.), and (iii) Defendants filed the Expert Motion “before the close of expert [and fact] discovery” (id.).

The record reflects that, on June 11, 2019, the parties jointly agreed to a proposed discovery plan and schedule, including deadlines for the close of fact and expert discovery and for dispositive motions. (Docket Entry 128 at 1-2 (proposing March 1, 2021, as close of fact discovery and August 2, 2021, as close of expert discovery and deadline for dispositive motions).) That agreement did not set a deadline for motions in limine, to include Daubert challenges. (See Id. at 1-3.) The Court (per the undersigned) adopted the parties' proposed schedule (Text Order dated June 14, 2019) and later granted a joint motion to extend fact discovery until June 1, 2021 (Text Order dated Feb. 3, 2021).

Given that Defendants filed the Expert Motion before the close of discovery (see Docket Entry 212 (filed May 12, 2021)), and the record reflects no deadline for motions in limine or Daubert motions (see Docket Entry 128 at 1-2), timeliness concerns provide no basis for dismissal. See generally Virginia Vermiculite, Ltd. v. W.R. Grace & Co.-Conn., 98 F.Supp.2d 729, 731 n.1 (W.D. Va. 2000) (noting that close of discovery may bear on timeliness of request for Daubert hearing). Plaintiffs' contrary contention falls short, especially because they have cited to a single case involving a post-deadline Daubert motion filed less than two months before trial. (See Docket Entry 221 at 25 (citing Rybas v. Riverview Hotel Corp., Civ. Action No. 12-3103, 2015 WL 10096189, at *7 (D. Md. Jan. 15, 2015) (unpublished) (declining to consider Daubert motion filed more than two weeks after court-imposed deadline)).) That authority bears no weight under the circumstances presented here.

B. Grace's Qualifications

Defendants have sought to exclude Grace's opinions on the grounds that she lacks qualifications in antitrust economics. (Docket Entry 213 at 4-5 (highlighting absence of prior testimony as antitrust expert and of publications in relevant fields); see also Docket Entry 227 at 3 (emphasizing that Grace developed opinions solely for purpose of testifying).) In response, Plaintiffs have asserted that Grace possesses sufficient knowledge and experience to qualify as an expert in this action. (Docket Entry 221 at 9-10 (describing three degrees in economics, to include coursework in antitrust economics, as well as experience as professor and contributor to publications).)

As the Fourth Circuit has explained:

[T]he test for exclusion is a strict one, and the purported expert must have neither satisfactory knowledge, skill, experience, training nor education on the issue for which the opinion is proffered. One knowledgeable about a particular subject need not be precisely informed about all details of the issues raised in order to offer an opinion.

Thomas J. Kline, Inc., 878 F.2d at 799 (deeming proposed expert unqualified to opine regarding unjustified credit discrimination and price discrimination when she had obtained a degree in business administration, id., “[her] general business education included [no] training in the area of antitrust or credit, ” Id. at 800, her sole publication “had nothing to do with price discrimination, credit, or antitrust generally, ” Id. at 799, and her work experience involved analyzing companies' financial health and offering expert testimony in complex business cases, not making credit decisions, see id.).

Here, to the extent Defendants have presented Grace's qualifications as independently disqualifying (see Docket Entry 213 at 3-6), ¹⁴ exclusion remains unwarranted on that ground. Grace possesses adequate qualifications “to analyze the issues of impact, damages, and economic conditions in [] Plaintiffs' case against Defendants” (Docket Entry 151-12 at 3). First, Grace has obtained three degrees in pertinent fields - a Bachelor's degree in economics and finance, as well as a Master's degree and doctorate in quantitative economics. (Id. at 2.) Second, her former employment as a professor “in the areas of economic theory, financial economics, corporate finance, and corporate valuation” (id.) suggests relevant knowledge and experience. Third, at least some of her publications signal her ability to calculate overcharges, an issue pertinent to this action. (See Id. at 3.)

14 At times, Defendants have blended their criticisms of Grace's methodology (which the Recommendation addresses in the following subsection) with their concerns about her qualifications. (See, e.g., Id. at 5 (“Grace lacks knowledge in the basic subject matter of her reports in this case: the common impact of alleged anticompetitive conduct. Instead, Grace employs a results-driven methodology that discards any facts inconsistent with her theory.” (internal citation omitted)).)

Even assuming that Grace lacks significant training in antitrust economics or pre-suit familiarity with the coupon-processing industry (Docket Entry 213 at 4-5), an expert “need not be precisely informed about all details of the issues raised in order to offer an opinion, ” Thomas J. Kline, Inc., 878 F.2d at 799.¹⁵ Instead of requiring demonstrated expertise in antitrust specifically (as opposed to quantitative economics more generally), the Court will consider Grace's qualifications as a factor in the broader reliability inquiry. See Elcock v. Kmart Corp., 233 F.3d 734, 749 (3d Cir. 2000) (“[A]n expert's level of expertise may affect the reliability of the expert's opinion.” (internal quotation marks omitted)).

15 Defendants' attempt to compare Grace to “a far more qualified individual” (Docket Entry 227 at 3) who nonetheless faced exclusion under Daubert misses the mark. (See Id. (citing Doe v. Ortho-Clinical Diagnostics, Inc., 440 F.Supp.2d 465, 471 (M.D. N.C. 2006)).) In the case cited by Defendants, this Court (per United States District Judge James A. Beaty, Jr.) assessed the qualifications of a proposed expert (a medical doctor) and his testimony regarding the cause of autism in a minor child. Doe, 440 F.Supp.2d at 468-78. After acknowledging the doctor's credentials (board certification in two fields, as well as “more than 50 peer-reviewed medical papers, ” Id. at 471), the Court noted the pertinent fields (pediatrics, pediatric neurology, epidemiology, and biostatistics) in which the doctor lacked certification, Id. The Court further referenced “more than ten . . . cases, ” id., when the doctor's proposed “testimony ha[d] either been excluded or accorded little or no weight based upon a determination that he was testifying beyond his expertise, Id. at 471-72 & n.5 (collecting cases). Notwithstanding those serious concerns, the Court declined to exclude the doctor on his qualifications alone and proceeded to evaluate his methodology in arriving at the opinions he sought to offer. See Id. at 472-78. Especially given the differences between Grace and the proposed expert in Doe (whose track record evinced concerns about his qualifications), a proper Daubert inquiry requires the Court to “focus, principally, . . . on the acceptability of the methodology by which [Grace] reached h[er] ultimate conclusion[s], ” Id. at 472, rather than exclude her testimony and the Supplemental Grace Report on the basis of her qualifications.

C. Reliability of Grace's Opinions

Defendants have argued that numerous deficiencies in the Supplemental Grace Report warrant its exclusion. Their criticisms fall into three general categories: the economic theory underlying the Supplemental Grace Report, the implementation of each methodology, and the results of those analyses. Plaintiffs have defended the Supplemental Grace Report on each front, as explained in more detail below.

1. Economic Theory

Defendants have sought exclusion of the Supplemental Grace Report because it offers no “coherent economic explanation of the causal mechanisms by which the alleged conduct elevated shipping fees” (Docket Entry 213 at 9), particularly in light of the fact that manufacturers could refuse to pay such fees ( id.). Defendants have raised a similar concern about Grace's Chargeback Analysis. (See Id. at 16-17 (explaining how ultimate effect of chargeback depends on decisions by manufacturers, retailers, and coupon processors).) In response, Plaintiffs have asserted that Grace need not prove causation independent of direct evidence suggesting that collusive behavior enabled increased shipping fees (which evidence she cited in the Supplemental Grace Report). (Docket Entry 221 at 12-13 (citing Docket Entry 193-2, ¶¶ 14-15).) To the extent Defendants have challenged the significance of such direct evidence, Plaintiffs have categorized that dispute as going to the weight, rather than the admissibility, of Grace's opinions. (Id. at 13.) Finally, Plaintiffs have noted that even manufacturers that charged back shipping fees could end up paying such fees indirectly if the Retailer Processor extracted the fees from retailers that in turn deducted the fees from moneys owed to the manufacturers. (Id. at 14)

Grace's opinions pass muster under Daubert, notwithstanding Defendants' theory-based critique. Defendants have relied on two cases in support of their argument for exclusion, both of which applied the Supreme Court's decision in Comcast. (See Docket Entry 213 at 10 (citing Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 993 F.3d 774 (9th Cir. 2021), and In re Rail Freight Fuel Surcharge Antitrust Litig. - Mdl No. 1869, 725 F.3d 244 (D.C. Cir. 2013)).) “The plaintiffs in [Comcast] had proposed four theories of antitrust impact, all but one of which the district court rejected. The plaintiffs' sole ground, however, for asserting that damages could be calculated on a classwide basis was a model that assumed the validity of all four theories.” In re Rail Freight, 725 F.3d at 253 (internal citation omitted). Because the expert's model failed to identify damages linked to the only viable theory of antitrust liability, the Supreme Court deemed predominance lacking and class certification improper. Comcast, 569 U.S. at 34; see also Id. at 36-37 (“At the evidentiary hearing, [the plaintiffs' expert] expressly admitted that the model calculated damages resulting from ‘the alleged anticompetitive conduct as a whole' and did not attribute damages to any one particular theory of anticompetitive impact.”). Here, Grace's “regression models test only one theory of liability, ” Olean Wholesale Grocery Coop., 993 F.3d at 789. Accordingly, Grace's analyses do not generate the same problem that existed in Comcast, and the Court thus will not exclude her opinions on that basis.¹⁶

16 Given that determination, the Court need not consider whether Comcast should inform the Daubert inquiry at all. See In re Urethane Antitrust Litig., 166 F.Supp.3d 501, 510 (D.N.J. 2016) (“Comcast is of dubious relevance because it involves predominance issues arising in the Rule 23 context, whereas the instant matter concerns whether expert testimony is reliable and relevant under Rule 702.”); see also Dzielak v. Whirlpool Corp., Civ. No. 2: 12-0089, 2017 WL 1034197, at *14 (D.N.J. Mar. 17, 2017) (unpublished) (“Comcast presents questions for the class[-]certification stage that are distinct from, and logically posterior to, the Daubert analysis.”).

2. Implementation

In advocating for exclusion, Defendants have identified the following supposed flaws with Grace's approach: the record contradicts the assumptions underlying her analyses (Docket Entry 213 at 17-18); she failed to account for factors unrelated to the alleged conspiracy (id. at 18-23); and the Daubert reliability factors otherwise weigh in favor of exclusion (id. at 23-24). Plaintiffs have resisted exclusion, contending that (i) record evidence supports Grace's conclusions (Docket Entry 221 at 13), (ii) Grace controlled for non-conspiratorial factors and properly used NCH for that purpose (id. at 15-17, 22-23), and (iii) Grace has “satisfie[d] the disputed Daubert reliability factors” (id. at 23 (standard capitalization applied and emphasis omitted)). For the following reasons, the Court declines to grant the Expert Motion on any of Defendants' proffered bases.

First, Defendants have challenged Grace's apparent assumption that only anticompetitive behavior could explain increased shipping fees during the alleged conspiracy, pointing to record evidence that shipping fees began increasing in 1997 (before the alleged conspiracy), and noting that another document upon which Grace relied reflects increased shipping fees in 1999. (Docket Entry 213 at 17-18.) By way of response, Plaintiffs have maintained that (i) “admissions and other evidence” support the notion that reduced competition caused increased shipping fees (Docket Entry 221 at 13-14), (ii) IOS allegedly “committed fraud beginning in 1997” (id. at 18), and (iii) “[Defendants] entered a ‘strategic alliance'” before signing the allegedly anticompetitive agreements, such that decreased competition may have occurred before April 2001 (id.).

“When facts are in dispute, experts sometimes reach different conclusions based on competing versions of the facts.” Fed.R.Evid. 702 advisory committee's notes, 2000 Amendment. Accordingly, courts should not “exclude an expert's testimony on the ground that the court believes one version of the facts and not the other.” Id. Consistent with that reasoning, “questions regarding the factual underpinnings of the [expert witness's] opinion affect the weight and credibility” of the witness'[s] assessment, ‘not its admissibility.'” Bresler v. Wilmington Tr. Co., 855 F.3d 178, 195 (4th Cir. 2017) (quoting Structural Polymer Grp. v. Zoltek Corp., 543 F.3d 987, 997 (8th Cir. 2008)). However, “[e]xpert opinion evidence based on assumptions not supported by the record should be excluded.” Tyger Constr. Co. v. Pensacola Constr. Co., 29 F.3d 137, 143 (4th Cir. 1994); see also Bresler, 855 F.3d at 196 (noting that erroneously admitted expert opinion in Tyger “conflicted directly with uncontroverted record evidence”).

Here, Defendants have stopped short of demonstrating a direct conflict between Grace's opinions and “uncontroverted record evidence, ” Bresler, 855 F.3d at 196. A dispute about the significance of documents or admissions reflecting Defendants' shipping fee strategy fails to justify exclusion. See Id. Furthermore, Grace's opinions may coexist with record evidence of increased shipping fees in 1997 and 1999; in that regard, she opined that Defendants' anticompetitive conduct caused an increase in shipping fees during the conspiracy period, not that Defendants never increased shipping fees outside the alleged conspiracy or that any such increase necessarily resulted from anticompetitive conduct. Defendants' critique merely highlights Grace's failure to attribute the 1997 or 1999 increases in shipping fees to some non- conspiratorial cause. That failure, though grounds for potential criticism, does not render Grace's opinions inadmissible.¹⁷

17 Nonetheless, Plaintiffs' response on the foregoing subject draws out a potential weakness in Grace's methodologies. In that regard, Plaintiffs have introduced the possibility of anticompetitive conduct before April 2001 (see Docket Entry 221 at 18-19 (“Because anticompetitive behavior occurred prior to the [conspiracy] period, and a price war would have resulted absent a conspiracy, [] Grace's model understates damages.”)), whereas Grace deemed that same period free from anticompetitive effects and thus a suitable benchmark period (see Docket Entry 193-2, ¶ 20). (See Docket Entry 227 at 7 (Defendants accusing Plaintiff of “trying to have it both ways”).) Stated somewhat differently, Plaintiffs and Grace have taken inconsistent positions as to whether anticompetitive conduct occurred before April 2001, thus casting doubt on the accuracy of Grace's Benchmark Analysis and DiD Analysis. Because that issue does not concern a disputed fact and more closely tracks Defendants' false-positive critique, the Court addresses it in the “Results” subsection.

Second, Defendants have contended that Grace considered no factors other than volume in modeling what non-conspiratorial factors impacted shipping fees during the alleged conspiracy period. (Docket Entry 213 at 19 (identifying lawful behavior, individualized contract negotiations, and cost increases as potentially relevant factors).) Moreover, according to Defendants, the data indicates that volume “fails to explain [price] differences.” (Id. at 20.) Relatedly, Defendants have contested the suitability of NCH as a benchmark. (Id. at 14 (pointing out that, before alleged conspiracy, CCC and IOS already charged substantially higher shipping fees than NCH); see also Docket Entry 227 at 6 (observing that Grace, during her deposition, declined to identify NCH as direct competitor of Inmar).) Plaintiffs have disputed those assertions, first noting that the omission of a variable pertains to probativeness, not admissibility, and further arguing that Grace's comparison to NCH (in the Benchmark Analysis and DiD Analysis) accounts for non-conspiratorial changes in shipping fees (such as increases in postage rates or labor costs). (Docket Entry 221 at 15-16; see also Id. at 22 (asserting that “NCH faced the same labor market, postage rates, and other market conditions that affected [Defendants and IOS], and its fees show an alternative estimate of cost”)). Furthermore, Plaintiffs have insisted that “Grace accounted for differences in bargaining power between small and large customers” (id. at 16) and “for the results of contract negotiations . . . by factoring in volume and actual prices charged” (id.).

“[F]ailure to include variables” in a regression analysis normally “affect[s its] probativeness, not its admissibility.” Bazemore v. Friday, 478 U.S. 385, 400 (1986). “It is only the rare case where the regressions are so incomplete as to be irrelevant and the expert's decisions regarding control variables are the basis to exclude the analysis.” P&G Co. v. Stone Container Corp. (In re Linerboard Antitrust Litig.), 497 F.Supp.2d 666, 678 (E.D. Pa. 2007) (internal quotation marks omitted). For example, the Fourth Circuit reversed summary judgment premised on an expert report that had omitted “important variables . . . from the multiple regression analysis, ” where “the evidence presented . . . [showed] that th[o]se variables [we]re crucial.” Smith v. Virginia Commonwealth Univ., 84 F.3d 672, 677 (4th Cir. 1996). In contrast, “[m]erely pointing to economic conditions that may affect the dependant variable is not enough to call into question the reliability of an econometric model.” In re Polypropylene Carpet Antitrust Litig., 93 F.Supp.2d 1348, 1365 (N.D.Ga. 2000).

In regard to a proposed expert's choice of benchmark, as a general rule, “arguments regarding the comparability or appropriateness of a particular benchmark go to the weight, not the admissibility, of a benchmarking analysis.” AngioDynamics, Inc. v. C.R. Bard, Inc., __F.Supp.3d__, __, 2021 WL 1792394, at *47 (N.D.N.Y. May 5, 2021). However, that choice may weigh in favor of Daubert exclusion when the benchmark wholly fails to provide a sufficient comparison. For example, in a horizontal price-fixing conspiracy case involving the inflation of real estate broker commissions, the plaintiff's expert chose as a benchmark “the commission amount for lower[-]priced homes, ” Hyland v. HomeServices of Am., Inc., No. 3: 05CV612, 2012 WL 12995647, at *12 (W.D. Ky. July 3, 2012) (unpublished), based on the assumption that commissions from the sale of lower-priced homes allowed agents to cover their costs, Id. at *11-12. The proposed expert faced exclusion on the grounds that he failed to “identif[y] any market which actually employs such a commission structure” and instead contemplated “a fictional, perfectly competitive market, ” rendering his testimony speculative and irrelevant (and thus inadmissible). Id. at *12.

Here, although Defendants have marshaled evidence that shipping fees varied among manufacturers and that volume cannot fully explain those price differences, that criticism affects the weight rather than the admissibility of Grace's opinions. Individual contract negotiations could affect the shipping fees charged or paid, but (as Plaintiffs have suggested) the alleged conspiracy could influence that process (rather than stand apart from it). (See Docket Entry 221 at 16 (citing In re Polyurethane Foam Antitrust Litig., 314 F.R.D. 226, 258 (N.D. Ohio 2014) (“Contract negotiations thus took place in the context of artificially inflated baseline pricing, effects which likely became ‘baked into' the contracts.”)).)¹⁸ In addition, the analyses using non-conspirator NCH as a benchmark seem to model how presumably competitive shipping fees responded to changes in postage rates and labor costs during the conspiracy period. For those reasons, Grace's choice of variables does not invalidate, as a matter of law, her opinions under Daubert.

18 The Court addresses negotiations (and resulting contracts) that predated the alleged conspiracy in the “Results” subsection.

Similarly, Defendants' challenge based on the suitability of NCH as a benchmark goes to the weight, not the admissibility, of Grace's testimony and the Supplemental Grace Report.¹⁹ Although NCH may have served different customers than Defendants, NCH nonetheless provided coupon-processing services to manufacturers and retailers, and its costs (and corresponding prices, to include shipping fees) plausibly may inform a benchmark analysis. Importantly, “a benchmark need not be perfectly comparable, so long as it allows the jury to calculate a ‘reasonable estimate of damages' as required under the antitrust laws.” AngioDynamics, Inc., __F.Supp.3d at__, 2021 WL 1792394, at *47 (emphasis added). Therefore, Grace's use of NCH as a benchmark does not warrant exclusion.

19 In resisting that conclusion, Defendants have invoked a class-certification decision that mentioned neither Daubert nor Rule 702. (See Docket Entry 227 at 6 (citing Bell Atl. Corp. v. AT&T Corp., 339 F.3d 294, 304 (5th Cir. 2003)).)

Third, Defendants have asserted that Grace (during her deposition) failed to (i) “identify a published source for her methodology” (Docket Entry 213 at 23); (ii) “identify a publication by anyone else embodying the methodology used in [the Original Grace] Report” (id.); (iii) “identify any source in which her methodology was subjected to peer review” (id.); (iv) “calculate a potential rate of error associated with her methodology” (id.); and (v) “identify any “source outside of [herself] that would indicate that the methodology used in [the Original Grace R]eport enjoys general acceptance with the relevant community of experts pertinent to this case” (id.). Plaintiffs have pointed out that those supposed shortcomings relate to the Original Grace Report (about which Defendants deposed Grace in August 2020), not the Supplemental Grace Report (which did not exist at the time of Grace's deposition). (Docket Entry 221 at 23-24.) In any event, Plaintiffs have insisted that Grace's methodologies enjoy general acceptance in antitrust economics (id.) and that “regression analyses have known rates of error” (id. at 24).

Courts have deemed use of the methodologies at issue (before-during, benchmark, difference-in-differences) appropriate in antitrust cases. See, e.g., Messner, 669 F.3d at 826 (concluding that proposed expert could use “difference-in-differences methodology to estimate [] antitrust impact”); Conwood Co., L.P. v. United States Tobacco Co., 290 F.3d 768, 793 (6th Cir. 2002) (characterizing “yardstick test and [] before-and-after test” as “generally accepted methods for proving antitrust damages”); In re Disposable Contact Lens Antitrust, 329 F.R.D. 336, 394-97 (M.D. Fla. 2018) (admitting expert testimony relying on before-during model); Lumco Indus. v. Jeld-Wen, Inc., 171 F.R.D. 168, 174 (E.D. Pa. 1997) (approving of yardstick model). In contrast, courts have applied greater scrutiny to less established methodologies and excluded them based on insufficient testability, peer review, standardization, or general acceptance. See, e.g., Precision Fabrics Grp., Inc. v. TieTex Int'l, Ltd., Nos. 1: 13CV645, 1:14CV650, 2016 WL 6839394, at *8 (M.D. N.C. Nov. 21, 2016) (unpublished) (excluding proposed expert testimony derived from simple visual inspection of “near microscopic swelling of a 45-micron sized film”); United States v. Johnson, 122 F.Supp.3d 272, 330-36 (M.D. N.C. 2015) (excluding proposed expert testimony that purported to quantify, via observational techniques and surname analysis, pattern of race-based police practices).

Here, although Defendants correctly have cited the factors that may inform whether a methodology passes muster under Daubert (see Docket Entry 213 at 23 (listing factors discussed in Cooper, 259 F.3d at 199)), those factors fail to capture the nature of Defendants' criticisms. In that regard, Defendants have refrained from attacking the relevant methodologies as inherently unreliable or fundamentally flawed; instead, Defendants have focused on supposed errors in Grace's implementation of those methodologies (discussed above) and the soundness of her conclusions (discussed in the following subsection). (See Id. at 11-24.) Furthermore, Defendants have developed no argument to support their more generalized critique, offering only Grace's deposition testimony as evidence of inadequacy. (See Id. at 23-24.) As Plaintiffs have observed (Docket Entry 221 at 23-24), that deposition testimony pertained to analyses upon which Plaintiffs no longer rely (see Docket Entry 196 at 4). Therefore, the Court declines to exclude Grace's opinion testimony or the Supplemental Grace Report on this front.

3. Results

Echoing the Supplemental Kheyfets Report, Defendants have contended that the flaws in Grace's approach lead to contradictory and absurd results. (Docket Entry 213 at 11-17, 21-23.) For their part, Plaintiffs have characterized Defendants' results-based arguments as irrelevant to admissibility, because disagreement concerning an expert's conclusions cannot justify Daubert exclusion. (Docket Entry 221 at 21-22.)

As mentioned above, Rule 702 and Daubert instruct courts to evaluate a proposed expert's methodology, not the results generated or the conclusions reached. Daubert, 509 U.S. at 595. However, exclusion remains appropriate when “opinion evidence . . . is connected to existing data only by the ipse dixit of the expert.” Joiner, 522 U.S. at 143-46 (upholding exclusion on grounds that studies upon which expert relied did not support opinion on causation of lung cancer). Assessing a proposed expert's conclusions may reveal “an analytical gap between the data and the opinion proffered.” Id.

On balance, the Court declines to view Defendants' results-based challenges as categorically irrelevant to the Expert Motion, to the extent they could reveal a Joiner-like flaw. However, absent such defect, that type of challenge bears on the weight, rather than admissibility, of expert testimony. See Daubert, 509 U.S. at 596 (“Vigorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.”). Defendants' arguments fall into two general categories, internal consistency and damages estimates, each of which the Court addresses below.

a. Internal Consistency

Defendants have criticized Grace's three models for producing conflicting answers as to whether a particular entity suffered an injury (i.e., paid an overcharge) during the alleged conspiracy. (Docket Entry 213 at 12.) For approximately 60 percent of manufacturers, the before-during analysis yields a negative answer, whereas the other two methodologies discern impact. (Id. at 15-16.) Plaintiffs have defended the internal consistency of Grace's analyses, insisting (based on the Grace Declaration) that “the correlation between the three models is extremely high, ranging from 96% to 99%” (Docket Entry 221 at 22 (citing Docket Entry 221-8, ¶ 4)). Defendants offered no reply on that subject. (See Docket Entry 227 at 4 (suggesting that “Plaintiffs [have] fail[ed] to contend with” consistency-based critique but nowhere addressing Grace's correlation-based defense).)

By estimating damages among proposed class members, Grace purported to satisfy, in a single step, two elements of Plaintiffs' claim: injury and damages. As previously noted, those concepts remain distinct; an assessment of injury produces a binary (i.e., yes or no) answer, whereas calculation of damages approximates the degree of injury. Kheyfets's internal-consistency criticism focuses on the former, the instances in which Grace's three methodologies diverge on the question of injury, without substantial discussion about the degree of divergence (i.e., the differences in damages amounts). (See Docket Entry 212-2, ¶¶ 43-45 & n.98 (identifying more than 3000 manufacturers as affected by internal inconsistency and describing contradictory damages calculations for four manufacturers).)

Upon review of Appendix A of the Supplemental Grace Report, it appears that the Before-During Analysis most often departs from the Benchmark Analysis and DiD Analysis when the latter two approaches identify relatively low damages. For example, for one manufacturer, Owen Mumford, Inc., the Benchmark Analysis and DiD Analysis projected $1.80 and $1.82 in damages, respectively. The Before-During Analysis reflected none (via the notation “-”).²⁰ An error of that magnitude appears less troubling than, for example, alternative models that produce wildly different results ($100,000 in damages according to one model, versus zero according to another). See, e.g., In re LIBOR-Based Fin. Instruments Antitrust Litig., 299 F.Supp.3d 430, 477 (S.D.N.Y. 2018) (noting, in context of Daubert inquiry, importance of “extent to which [inconsistent] models disagree”). By focusing on the instances of divergence on injury rather than the degree of divergence on damages, Kheyfets may exaggerate the extent to which the results of the three models diverge. Taken together with Grace's sworn statement that her three methodologies remain highly correlated, the internal inconsistency that Kheyfets has identified does not provide an adequate basis for exclusion.

20 Given Kheyfets's representation that Grace's models produce negative damages (see, e.g., Docket Entry 212-2, ¶ 49), the Court recognizes the possibility that “-” may not actually represent zero, in any or all cases.

b. Damages Estimates

Defendants have taken issue with three aspects of Grace's calculation of damages to manufacturers: (i) the No. of manufacturers that would have paid higher shipping fees in the but-for world than they actually paid during the conspiracy period, (ii) the No. of coupons that Defendants would have shipped at a negative fee in the but-for world (as well as Grace's decision in some cases to assign a $1.00 value to those transactions, instead of the negative fee), and (iii) the extent to which Grace's models show antitrust impact in transactions supposedly outside the alleged conspiracy. (Docket Entry 213 at 12-16.) As concerns Grace's Chargeback Analysis, Defendants have asserted that (i) more than half of the proposed retailer class incurred damages totaling 20 dollars or less (with some retailer's damages amounting to less than one dollar) (id. at 16), (ii) Grace calculated but then ignored $16.5 million in undercharges to the proposed retailer class (id. at 17), and (iii) the analysis likewise generates false positives, to include labeling as damaged “retailers that set their own shipping fees” (id. (emphasis omitted)).

Plaintiffs have contested the relevance of undercharges (Docket Entry 221 at 19-20), asserting that antitrust injury occurs at the time of an overcharge and that indirect offset benefits should “not affect the ultimate measure of damages” (id. at 20 (quoting In re Airline Ticket Comm'n Antitrust Litig., 918 F.Supp. 283, 286-87 (D. Minn. 1996))). Regarding the calculation of negative shipping fees, Plaintiffs have argued that result merely “suggests that downward pricing pressure exceeded the amount of the shipping fee and would have affected other coupon[-]processing fees . . . .” (Id. at 21-22 (emphasis added).) Per Plaintiffs, Grace assigned a $1.00 fee to such transactions as a “conservative adjust[ment].” (Id. at 22.) Finally, as concerns Defendants' false-positive critique, Plaintiffs have suggested that collusion affected the categories of transactions Kheyfets purported to distinguish as outside the alleged conspiracy. (Id. at 17-19.)

The Court begins with Defendants' argument that some entities benefitted from the alleged conspiracy by paying lower shipping fees (or facing fewer chargebacks) during the conspiracy period than Grace predicted in the but-for world. In pressing that issue, Kheyfets aggregated undercharges, both on a class-wide basis and for one particular manufacturer (Mr. Dee's, a proposed class representative) to determine whether that manufacturer paid more, on the whole, during the conspiracy period than in the but-for world. (Docket Entry 213 at 13 (citing Docket Entry 212-2, ¶ 49).) According to his calculations, the entire class of manufacturers sustained $9.6 million in undercharges, with Mr. Dee's paying around $7,000 less in shipping fees as a result of the alleged conspiracy. (Id.) As concerns the proposed retailer class members, those entities collectively incurred $16.5 million in negative damages, in terms of the excess chargebacks that Grace purported to calculate. (Id. at 16-17.) In Defendants' view, Grace's failure to deal with that issue bears on the Daubert inquiry (id.), whereas Plaintiffs have labeled undercharges irrelevant to this action as a whole (Docket Entry 221 at 19-20).

As a general matter, the Court discerns nothing inherently unreliable about a methodology that identifies both overcharges and undercharges. For example, Grace's methodologies need only identify whether, in a particular transaction, a manufacturer paid more (or less) in shipping fees than the but-for rate. The existence of undercharges does not necessarily undermine the ability of Grace's methodologies to identify injury-in-fact (or lack thereof). Nor does the existence of undercharges suggest that no injury-in-fact occurred. “Paying an overcharge caused by the alleged anticompetitive conduct on a single purchase suffices to show - as a legal and factual matter - impact or fact of damage.” Astrazeneca AB v. UFCW (In re Nexium Antitrust Litig.), 777 F.3d 9, 27 (1st Cir. 2015) (quoting Davis & Cramer, Antitrust, Class Certification, and the Politics of Procedure, 17 Geo. Mason L. Rev. 969, 984-85 (2010)). An overcharge constitutes injury-in-fact regardless of whether a later transaction (involving the same entity) results in an undercharge. See Id. (“[I]f a class member is overcharged, there is an injury, even if that class member suffers no damages.” (emphasis added)). Therefore, aggregating transactions (as Kheyfets did) cannot demonstrate the absence of injury-in-fact.

However, consistent with the foregoing principles, undercharges may offset damages in an antitrust action. “In a horizontal price-fixing case, . . . mitigation and offset generally do not affect the ultimate measure of damages.” In re Airline Ticket Comm'n Antitrust Litig., 918 F.Supp. at 286 (emphasis added). For example, if an overcharged buyer responds by raising its own prices, it remains “entitled to damages” like an overcharged buyer who fully “absorbs the loss, ” Hanover Shoe, Inc. v. United Shoe Mach. Corp., 392 U.S. 481, 489 (1968). See also In re Airline Ticket Comm'n Antitrust Litig., 918 F.Supp. at 286 (“[W]here a direct purchaser passes on overcharges to its customers, ‘damages are established by the amount of the overcharge,' regardless of ‘whether the victim of the overcharge has partially recouped its loss in some other way.'” (quoting Hawaii v. Standard Oil Co. of Calif., 405 U.S. 251, 262 n.14 (1972)). Notably, the Supreme Court [has] crafted . . . exceptions [to that general rule] for situations in which ‘it [is] easy to prove that [the plaintiff] has not been damaged.'” Id. at 287 (discussing Hanover Shoe, 392 U.S. at 494, 504).

Here, the assessment of damages likely will permit consideration of overcharges and undercharges. As a threshold matter, it seems unlikely that undercharges qualify as the kind of “indirect offset benefits” (Docket Entry 221 at 19) that, in some cases, can frustrate the calculation of damages. At least as to manufacturers, Defendants have not sought to reduce Plaintiffs' damages by the amount that Plaintiffs recouped from another entity through other means; rather, Defendants have attempted to limit an entity's damages to the net overcharge amount, accounting for all transactions during the conspiracy period. Kheyfets's calculations suggest that the parties can determine, at the pertinent time, whether each proposed class member sustained net damages. However, that merits-phase determination neither invalidates Grace's methodologies nor nullifies the instances of overcharge injury that Grace identified.²¹

21 This case may present somewhat unique circumstances given that, via chargebacks, a would-be injured entity (a manufacturer) can pass along an allegedly illegal overcharge (a supracompetitive shipping fee) to another entity (a retailer), by means of an alleged conspirator (a retailer processor). It remains unclear whether the generally applicable principle, i.e., “[t]he first buyer from a conspirator is the right party to sue, ” Paper Sys., Inc. v. Nippon Paper Indus. Co., 281 F.3d 629, 631 (7th Cir. 2002), contemplates such a transaction.

Turning to Defendants' criticism that some of Grace's calculations produced a negative shipping fee in the but-for world, that issue goes to the accuracy of Grace's models, for purposes of determining the extent of damage, not whether a manufacturer suffered an injury. Although Defendants have deemed arbitrary the $1.00 fee Grace assigned in place of the negative fee, they have failed to explain how that decision impacts the reliability of her methodologies in identifying injury-in-fact among proposed class members. In other words, Grace's substitution of the $1.00 fee in place of the negative fee may have underestimated damages, but it did not overestimate injury. Because admissibility under Daubert (for class-certification purposes) depends on an expert's reliable estimation of injury, rather than an accurate calculation of damages, the Court declines to exclude Grace's testimony on those grounds.

Finally, the Court addresses Defendants' contention that false-positive results render Grace's opinions inadmissible. (See Docket Entry 213 at 22 (citing Rail Freight).) In Rail Freight, the United States Court of Appeals for the Third Circuit vacated a class-certification decision and remanded for further consideration, noting that the proposed expert's model had produced false-positive results and that the district court's failure to consider that issue fell short of the rigor required by Comcast (which opinion issued after the district court decided to certify the class). Rail Freight, 725 F.3d at 253, 255. On remand, the district court determined that the proposed expert satisfied Daubert, given that he had produced a testable model “using sufficient data and widely accepted methods, ” In re Rail Freight Fuel Surcharge Antitrust Litig., 292 F.Supp.3d 14, 58 (D.D.C. 2017); see also Id. at 59 (“find[ing] no reasons to exclude [] damages model as irrelevant or unreliable under Daubert”). That conclusion nonetheless permitted further scrutiny of the model for purposes of class certification. See Id. (“The [c]ourt will fully explore [the] defendants' criticisms regarding the damages model's alleged propensity for false positives, but that is an examination appropriately for class certification, not for Daubert.” (internal citation omitted)). The district court ultimately found predominance lacking and denied certification. Id. at 144-45.

Accordingly, the sole authority upon which Defendants have relied ultimately viewed the issue of false positives as relevant to class certification, not admissibility. Another court, considering the same issue and interpreting Rail Freight (before the above-referenced decision on remand), deemed the proposed expert testimony admissible and concluded that “the false[-]positive results . . . might call into question [the] models as ultimate proof of injury and damages, ” In re Mushroom Direct Purchaser Antitrust Litig., Master File No. 06-0620, 2015 WL 5775600, at *6 (E.D. Pa. Aug. 5, 2015) (unpublished). In any event, to the extent Grace's false-positive results bear on her reliability, Plaintiffs have identified means by which anticompetitive conduct could have tainted the transactions that Kheyfets deemed unaffected by the alleged conspiracy. (Docket Entry 221 at 21.) That circumstance distinguishes this case from Rail Freight, where “the false[-]positive analysis relied upon comparison to a group of shippers indisputably unaffected by the conspiracy.” In re Mushroom Direct Purchaser Antitrust Litig., 2015 WL 5775600, at *6 (emphasis added) (internal quotation marked omitted). If, as Plaintiffs have suggested, anticompetitive conduct influenced entities outside the alleged conspiracy or transactions before the conspiracy period, Grace's analyses would understate damages without overstating injury. For those reasons, the Court declines to treat the instances of potential false positives as grounds for exclusion.

CONCLUSION

The Expert Motion requires the Court to decide the question of whether Grace's opinions can reliably assist the resolution of the Class Certification Motion, not whether Plaintiffs have satisfied Rule 23 requirements, whether their proposed class definitions merit revision, or whether they can ultimately prevail in proving injury and damages. On balance, the Court deems her qualified to offer expert testimony and her opinions sufficiently reliable to survive Defendants' Daubert challenges.

IT IS THEREFORE ORDERED that the Expert Motion (Docket Entry 212) is DENIED.