Rail Freight Fuel Surcharge Antitrust Litigation (No. Ii)- Mdl No. 2925

Opinion

UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA

IN RE RAIL FREIGHT FUEL SURCHARGE MDL Dkt. No. 1869 ANTITRUST LITIGATION (NO. I) Miscellaneous No. 07-489 (MDL I)

Judge Beryl A. Howell

IN RE RAIL FREIGHT FUEL SURCHARGE ANTITRUST LITIGATION (NO. II) MDL Dkt. No. 2925 Miscellaneous No. 20-008

(MDL II)

Judge Beryl A. Howell

OXBOW CARBON & MINERALS LLC, et al., Civil Action No. 11-1049

Plaintiffs, Judge Beryl A. Howell

v. FILED UNDER SEAL PENDING COUNSEL REVIEW FOR

UNION PACIFIC RAILROAD CO., et al., REDACTION OF CONFIDENTIAL BUSINESS INFORMATION

Defendants. UNSEALED FOLLOWING COUNSEL REVIEW (NO REDACTIONS)

MEMORANDUM OPINION

This antitrust litigation challenges fuel surcharges imposed by the four largest freight railroad companies in the United States as they responded to volatile oil prices in the early 2000s. Plaintiffs, hundreds of customers shipping freight on these railroads, brought antitrust claims against BNSF Railway Co., CSX Transportation, Inc., Norfolk Southern Railway Co., and Union Pacific Railroad Co. (“defendants”), alleging that they conspired to impose

2

coordinated, aggressive, and universal fuel surcharges (“FSCs”), in violation of section 1 of the Sherman Act, 15 U.S.C. § 1. See, e.g., Direct Purchaser Pls.’ Consolidated Am. Class Action Compl., In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL I”), MDL No. 1869, 7-mc- 489, ECF No. 91-1; Indirect Purchaser Pls.’ Consolidated Am. Compl., id., ECF No. 93; Am. Compl., In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL II”), MDL No. 2925, 20-mc- 008, ECF No. 5; Compl., Oxbow Carbon & Minerals, LLC v. Union Pac. R.R. Co. (“Oxbow”), 11-cv-1049, ECF No. 1.

Some of those claims have been pending since as early as 2007. See Transfer Order, MDL I, ECF No. 1 (Nov. 6, 2007, U.S. Judicial Panel on Multi-District Litigation (“MDL Panel”) transferring seven cases from various districts for consolidated pretrial proceedings, in MDL I, in the District Court for the District of Columbia). After passing multiple procedural steps—including the grant and then denial of class certification, prompting two trips to the D.C. Circuit, several intervening motions and a third trip to the D.C. Circuit, actions by the MDL Panel to create MDL II and subsequently to transfer a total of 114 cases for consolidated pretrial proceedings in MDL I and MDL II before this Court, and several years of discovery in each of these two related MDLs—briefing on dispositive and related motions were completed at the end of 2024, with defendants’ pending motions for summary judgment addressed at a lengthy hearing held on June 18, 2025. See Defs.’ Common Mem. in Supp. of Defs.’ Joint & Individ. Mots. for Summ. J. (“Defs.’ Mem.”), MDL I, ECF No. 1163 (sealed); MDL II, ECF No. 982 (sealed); Oxbow, ECF No. 303 (sealed). For the reasons explained below, defendants’ motions for

3

summary judgment are granted, and the related pending motions from both plaintiffs and defendants are denied as moot. 1

* * *

To aid in review of this Memorandum Opinion, given the breadth and complexity of the alleged antitrust conspiracy and the length of this decision, an overview is provided. Part I provides the relevant factual and procedural background in this case, including an overview of the freight rail industry and FSCs (section A), the timing and terms of defendants’ FSCs and pricing behavior before and during the alleged conspiracy period (section B), and the extensive procedural history of this case (section C).

Part II provides the legal standard governing defendants’ motions for summary judgment on the Section I Sherman Act claim under Federal Rule of Civil Procedure 56.

Part III addresses the merits and disposition of the summary judgment motions. In sum, plaintiffs theorize that the four defendants conspired, starting in 2003, to increase their profits by raising FSCs and imposing them universally across all of their unregulated traffic, using the prevailing high fuel costs in the early 2000s as an excuse. Section A explains in detail the legal framework governing consideration of such an alleged antitrust conspiracy in the context of an oligopolistic market, with only a few competitors, and what plaintiffs must prove through either direct or circumstantial evidence. Section B concludes, after noting plaintiffs’ admitted lack of direct proof positive of the alleged conspiracy, that they must rely on circumstantial evidence to prove the alleged conspiracy, with evidence demonstrating parallel conduct and further “plus” factors that tend to disprove the possibility of independent action. Section C considers

1 With a few exceptions, the parties filed identical briefs across all three dockets in MDL I, MDL II and Oxbow. Unless otherwise specified, ECF numbers will refer to the numbers in the MDL I docket, MDL No. 1869, 7-mc-489.

4

plaintiffs’ attempt to demonstrate that defendants acted in a parallel way to raise FSCs and expand application of FSCs to shippers for carload and coal traffic and concludes that plaintiffs’ evidence does not so demonstrate, because defendants’ FSC formulas differed more during than prior to the alleged conspiracy period (subsection 1), the timing of defendants’ changes to their FSCs was not aligned (subsection 2), and defendants did not expand their application of FSCs in a uniform way on the same timeline (subsection 3). Given that plaintiffs cannot demonstrate that defendants acted in a parallel fashion, their circumstantial evidence is insufficient to allow for an inference of conspiracy, and defendants are entitled to summary judgment.

Sections D and E nonetheless go on to consider, if plaintiffs were able to show that defendants’ behavior was sufficiently parallel, whether plaintiffs’ evidence could tend to exclude the alternative inference of independent conduct and to conclude that plaintiffs would still fail to survive summary judgment. Section D, in particular, considers the evidence defendants have advanced to support an inference of independent action, demonstrating that they had a rational business motivation for their imposition of FSCs during the alleged conspiracy period (subsection 1) and that their decisions were the result of independent, internal decision-making (subsection 2). Section E further considers plaintiffs’ evidence to exclude the inference of independent conduct, examining plaintiffs’ arguments regarding defendants’ purported pretextual justifications for the imposition of FSCs (subsection 1), defendants’ common motive to conspire (subsection 2), defendants’ opportunities to conspire (i.e., meetings where defendants purportedly discussed FSCs) (subsection 3), defendants’ sharing of information related to FSCs (subsection 4), defendants’ related actions through their trade industry association allegedly to facilitate the conspiracy (subsection 5), defendants’ changes in business practices during the alleged conspiracy period (subsection 6), defendants’ perceived acts against

5

their self-interest (subsection 7), the oligopolistic nature of the industry as conducive to collusion (subsection 8), and defendants’ actions to adjust their base rates near the end of the conspiracy period to bake-in the higher prices that FSCs produced (subsection 9). Subsection 10 puts all these pieces of evidence together to evaluate holistically plaintiffs’ showing and concludes that the evidence fails to tend to exclude an inference of independent but consciously parallel conduct regarding carload and coal traffic and thus cannot sustain a claim for a Sherman Act conspiracy.

Section F evaluates plaintiffs’ claim that defendants also conspired with respect to their intermodal traffic. Subsection 1 explains why intermodal traffic is evaluated separately, looking to the unique qualities and pricing formulas applied to intermodal customers. Subsection 2 concludes that defendants’ conduct was not parallel, and subsection 3 explains that even if defendants’ approach to intermodal FSCs were parallel, plaintiffs’ evidence does not tend to exclude an inference of independent, though consciously parallel action. Plaintiffs thus fail in demonstrating any genuine dispute of fact on a claim of conspiracy warranting proceeding to a jury.

Section G explains why that conclusion obviates the need to consider the parties’

arguments about damages and other pending motions.

Part IV provides a brief conclusion summarizing the disposition of defendants’ motions for summary judgment. I. BACKGROUND The factual background and procedural history relevant to the pending motions are described below.

6

A. Background on Railroad Industry and Fuel Surcharges The four defendants in this case control most of this country’s freight railroads, operating as an oligopoly. See Pls.’ Counterstatement of Facts ¶ 1 (“Pls.’ COF”), ECF No. 1198-2 (sealed); Defs.’ Mem. CSX Transportation, Inc. (“CSX”) and Norfolk Southern Railway Co. (“NS”) operate primarily in the East, while Union Pacific Railroad Co. (“UP”) and BNSF Railway Co. (“BNSF”) operate primarily in the West. See Defs.’ Statement of Undisputed Material Facts (“Defs.’ SUMF”) ¶ 5.1, ECF No. 1163-1; DA12504 at 12508, Defs.’ App’x Vol. 24, ECF No. 1172-24 (sealed) (Sep. 17, 2008, presentation on National Rail Freight Infrastructure Capacity and Investment Study to the Railroad Energy Transportation Advisory Committee); Defs.’ Mem. at 6. 2 Together, defendants account for around 90% of the freight railroad industry revenue. See In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL I Class Cert. II”), 725 F.3d 244, 247 (D.C. Cir. 2013); Defs.’ Mem. at 5-6; PX583 at 19, Pls.’ App’x Vol. 9B, ECF No. 1221-1 (sealed) (Oct. 2006, GAO Report on Freight Railroads).

Most of the traffic on these railroads is referred to as “carload traffic,” Defs.’ Mem. at 4-

5, with the bulk movement of coal at times referred to separately as “coal traffic,” id.; see Pls.’ Opp’n to Defs.’ Mot. for Summ. J. (“Pls.’ Opp’n”) at 115, ECF No. 1218 (sealed). The industry also refers separately to “intermodal traffic,” which is containerized freight requiring more than one transportation mode to arrive at its destination (e.g., railroad to truck). See Defs.’ Mem. at

2 “DA” citations refer to materials in defendants’ appendices filed in support of their joint and individual motions for summary judgment, see ECF No. 1172, and the supplemental appendices filed in support of their replies, ECF No. 1245. Pin cites refer to the DA page number at the bottom center of each page. The exhibit locations are indicated by the appendix volume number (which is also identified via ECF number for each volume’s first appearance).

“PX” citations refer to materials in plaintiffs’ appendices filed in support of their opposition, see ECF Nos.

1219-25. Pin cites refer either to the last three digits of the Bates number of the cited page (if preceded by an apostrophe) or otherwise to the page number of clearly paginated reports, depositions, or pleadings. The exhibit locations are likewise indicated by the appendix volume number (again identified via ECF number for each volume’s first appearance).

7

xi, 4-5; Defs.’ SUMF ¶ 36; Pls.’ Objections and Responses to Defs.’ SUMF (“Pls.’ Objs. & Resp.”) ¶ 36, ECF No. 1198-1 (sealed) (not contesting that fact). Another way to categorize railroad traffic is distinguishing rate-regulated traffic, which is subject to rules enacted and enforced by the Surface Transportation Board (“STB”), and private contract or rate-unregulated traffic. See Defs.’ Mem. at 11. This dispute only involves the latter, rate-unregulated traffic. Id.

Given the limited number of railroads operating across the United States, freight railroad companies must sometimes engage in joint ventures, known as “interline partnerships.” See Defs.’ SUMF ¶ 5; Pls.’ COF ¶ 36; Defs.’ Mem. at 7. “[I]nterline movements are shipments carried along two or more railroads’ tracks under a common arrangement,” in contrast to “single- line shipments [which] are moved by one carrier on its own tracks.” In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL I Section 10706 Decision II”), 34 F.4th 1, 6 (D.C. Cir. 2022). Interline shipments, by their nature, require competing railroads to “coordinat[e] . . . over logistics and shipping rates.” Id. (quoting MDL I Class Cert. II, 725 F.3d at 247 n.1); Defs.’ SUMF ¶ 5.6; Pls.’ Objs. & Resp. ¶ 5 at 26 (not contesting the need to sometimes coordinate over interline rates while maintaining that defendants are still horizontal competitors); Pls.’ COF ¶ 37 (recognizing situations where limited coordination may occur for interline traffic). Such interline arrangements are encouraged because they promote efficiency while avoiding the need for further railroad mergers in an already concentrated industry. See 49 C.F.R. § 1180.1(c) (in policy statement for mergers between Class I railroads cautioning against “further consolidation of the few remaining Class I carriers” as “likely to result in a number of anticompetitive effects” and noting “that other private-sector initiatives, such as joint marketing agreements and interline partnerships, can produce many of the efficiencies of a merger while risking less potential harm to the public”). To facilitate such interline arrangements, discussions to coordinate interline

8

shipments are statutorily protected: Under 49 U.S.C. § 10706(a)(3)(B)(ii), Congress barred the use of discussions about interline shipments to infer an antitrust conspiracy and precluded their admission in proceedings where antitrust violations are alleged. See MDL I Section 10706 Decision II, 34 F.4th at 6.

Operating fuel-intensive enterprises, railroads are sensitive to fluctuations in fuel pricing.

Fuel costs generally increased from 2000 to 2008, after a sharp increase in 1999. Defs.’ SUMF ¶¶ 3.1-3.2; Pls.’ Objs. & Resp. ¶ 3 (admitting that defendants’ graph accurately reflects fuel prices during the period); Defs.’ SUMF ¶ 2 (describing the percentage increase in fuel costs out of total expenses for railroads during the 2000s); Pls.’ Objs. & Resp. ¶ 2 at 10 (not disputing the self-reported statistics but disputing that the increase was substantial or growing throughout the conspiracy period); Defs.’ Mem. at 32 (describing massive increases in price in 1999 and early 2000 and increases continuing through early 2003). In addition to steep and volatile fuel costs, railroads also faced declining rates in the early 2000s such that inflation-adjusted railroad rates declined every year from 1985 to 2004 (with a single year exception). See PX593 at 1, Pls.’ App’x Vol. 10A, ECF No. 1202-3 (sealed) (Jan. 16, 2009, STB Study of Railroad Rates: 1985- 2007).

In response to the fluctuating and increasing fuel prices, by the early 2000s, railroads commonly keyed their prices to indices that reflected fuel costs, using price escalators like the Rail Cost Adjustment Factor (“RCAF”) or imposing “fuel surcharges” (“FSCs”), which were variable charges on top of railroads’ base rates dependent on the contemporaneous price of fuel. See Defs.’ Mem. at xi; MDL I Section 10706 Decision II, 34 F.4th at 6-7. FSCs were also used by other transportation providers, including non-defendant railroads and trucking companies, around this period to recover increasing fuel costs. See Defs.’ SUMF ¶ 1; Pls.’ Objs. & Resp. ¶ 1

9

at 3 (agreeing that other transportation providers used FSCs); DA11527, Defs.’ App’x Vol. 23, ECF No. 1172-23 (sealed) (July 2009, Department of Transportation report, describing various surcharge and price adjustment mechanisms used to control for fuel price volatility); DA4034 at 4034, 4040, Defs.’ App’x Vol. 9, ECF No. 1172-9 (sealed) (Sep. 17, 2002, internal BNSF email from Thomas Hund to David Burr et al., forwarding Bear Sterns analysis mentioning trucking companies’ FSCs); DA4289 at 4300, Defs.’ App’x Vol. 11, ECF No. 1172-11 (sealed) (Apr. 3, 2003, internal BNSF email from Sam Kyei, BNSF Chief Economist, to Gary Shoop et al., attaching presentation on BNSF Fuel Surcharges Cost Recovery Improvements including table of FSC methodologies for trucking companies); e.g., DA6654, Defs.’ App’x Vol. 16, ECF No. 1172-16 (sealed) (Sep. 18, 2002, letter from Canadian National (“CN”), a non-defendant railroad, about FSCs); DA6650 at 6652, id. (June 12, 2001, copy of contract attachments describing FSCs at Canadian Pacific Railway (“CPR”), a non-defendant railroad); infra Part III.D.1.

FSCs may be either “rate-based,” which impose a surcharge “as a function of the base rate” once “fuel prices exceed [a] trigger price,” or “mileage-based,” which “raise total rates in proportion to shipping distances.” MDL I Section 10706 Decision II, 34 F.4th at 6-7. In application, as defendants explain, a rate-based FSC may use a “formula” that “might specify a 2% surcharge when oil is $29/bbl [barrel], thus adding $20 to the price of a shipment with a base rate of $1,000 per carload.” Defs.’ Mem. at 10. The oil price would be calculated according to an average or minimum of oil prices over a certain time period according to a widely accepted index, either the WTI (West Texas Intermediate crude oil price) or the HDF (U.S. National Average On-Highway Diesel Fuel price, published by the U.S. Department of Energy). See id. at x-xi. For example, the formula might specify that an oil price would be considered $29/bbl if the

10

HDF measured the price of oil at or above $29/bbl every day for the past 30 days. By the mid- 2000s, “fuel surcharge provisions became ubiquitous, governing the vast majority of the defendants’ shipments.” MDL I Class Cert. II, 725 F.3d at 248; see infra Parts I.B, III.C.3.

In 2007, the practice came under scrutiny, however, by the STB. See MDL I Class Cert.

II, 725 F.3d at 248. The STB issued a decision that prohibited FSCs for rate-regulated traffic, citing concern about the “disconnect between the purported rationale for the fuel surcharges— fuel cost recovery—and the formula’s dependence on base rates, which need not reflect the marginal fuel costs of a particular shipment.” Id. While the STB decision did not affect rates for unregulated freight outside of the STB’s regulatory authority, this decision triggered questioning by shippers of such unregulated traffic about the propriety of FSCs in their contracts too.

This case arises out of such questioning and alleges conspiratorial price-fixing in violation of the Sherman Act, section 1, between March 2003 and December 2008. In re Freight Fuel Surcharge Antitrust Litig. (“MDL I MTD Order (Direct Purchasers)”), 587 F. Supp. 2d 27, 30 (D.D.C. 2008); Pls.’ Opp’n at 1. 3 B. Factual Summary The following facts are undisputed by the parties. 4

3 The parties dispute the end date of the alleged conspiracy. See Defs.’ Mem. at 99-106 (arguing that, at a minimum, no triable issue exists for a conspiracy beyond January 2007); Pls.’ Opp’n at 140-53 (arguing that the conspiracy extended beyond January 2007 to at least December 2008). As explained infra Part III.G, the precise end date of the conspiracy is immaterial because plaintiffs have not put forth evidence allowing a plausible inference of conspiracy. The period from March 2003 to December 2008 will be referred to as the “alleged conspiracy period.” As explained infra Part I.C.1, this Court determined in MDL II that claims predicated on alleged acts occurring after December 2008 were barred by the statute of limitations, which was tolled only as to the claims initially filed in MDL I during the pendency of the class certification issue. See MDL II MTD Order on Time-Barred Claims, No. 20-mc-8 (BAH), 2020 WL 5016922, at *29 (D.D.C. Aug. 25, 2020). 4 The identification of undisputed material facts has been unnecessarily difficult in this case. Between defendants’ Statement of Undisputed Material Facts, ECF No. 1163-1, plaintiffs’ Objections and Responses, ECF No. 1198-1, plaintiffs’ Counterstatement of Facts, ECF No. 1198-2, and defendants’ Rebuttal, ECF No. 1242, over 660 pages have been submitted on the stated facts alone—550 of these from plaintiffs. Both sides accuse the other of failing to comply with local rules. See Pls.’ Objs. & Resp. at 1-2; Defs.’ Rebuttal at 1-2. Plaintiffs insist that defendants submitted legal conclusions, inferences, and immaterial matters as “facts,” Pls.’ Objs. & Resp. at 1-2,

11

1. Carload Traffic FSCs In the early 2000s, prior to the alleged start of the conspiracy in 2003, all four defendants independently and unilaterally adopted rate-based FSCs for their carload traffic in response to unpredictable and rapidly rising fuel prices. See Pls.’ COF ¶¶ 22-32, 46; id. Part II.C (titled, “Prior to 2003, the Defendants did Not Coordinate to Impose Uniform Fuel Surcharges”); Defs.’ SUMF ¶¶ 3, 6; Pls.’ Objs. & Resp. ¶¶ 3, 6 (agreeing with these points); see also In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL I Class Cert. III”), 292 F. Supp. 3d 14, 122-23 (D.D.C. 2017). The formulas used by three of the defendants to calculate their FSCs were largely identical, while BNSF’s formula was different.

a) FSCs Prior to 2003 When Alleged Conspiracy Began

requiring a response to defendants’ facts consisting of broad, narrative counterarguments, without indicating whether each supporting fact listed by defendants is admitted or denied, see generally Pls.’ Objs. & Resp. Defendants critique plaintiffs for not providing a concise counterstatement of facts and instead dumping hundreds of pages of argument on the Court, Defs.’ Rebuttal 1-11; see generally Pls.’ COF, with the result that defendants did not respond seriatim to the plaintiffs’ counterstatement of facts, see generally Defs.’ Rebuttal.

Both Federal Rule of Civil Procedure 56(e)(2) and D.D.C. Local Civil Rule 7(h) permit a district court to assume as undisputed any facts that are identified by the moving party in its statement of material facts but are not properly addressed or controverted by the opposing party. See FED. R. CIV. P. 56(e)(2) (providing that “[i]f a party fails to properly address another party’s assertion of fact . . . the court may . . . consider the fact undisputed for purposes of the motion [for summary judgment]”); D.D.C. LCvR 7(h)(1) (“In determining a motion for summary judgment, the Court may assume that facts identified by the moving party in its statement of material facts are admitted, unless such a fact is controverted in the statement of genuine issues filed in opposition to the motion.”); see also Grimes v. District of Columbia, 794 F.3d 83, 96 (D.C. Cir. 2015) (Griffith, J., concurring) (noting that the 2010 amendment to Rule 56 “reflects the ‘deemed admitted’ provisions in many local rules” (quoting FED. R. CIV. P. 56 advisory committee’s note (2010)). The D.C. Circuit “has long upheld strict compliance with the district court’s local rules on summary judgment when invoked by the district court,” Burke v. Gould, 286 F.3d 513, 517 (D.C. Cir. 2002), recognizing that such compliance “is justified both by the nature of summary judgment and by the rule’s purposes,” which include “isolat[ing] the facts that the parties assert are material, distinguish[ing] disputed from undisputed facts, and identif[ying] the pertinent parts of the record,” Gardels v. CIA, 637 F.2d 770, 773 (D.C. Cir. 1980).

Regardless of whether either side precisely complied with the local rules or would have done a better job if directed to re-do their respective statements and counterstatements of material facts, defendants are correct that plaintiffs’ unwieldy factual submissions appear manufactured to generate a genuine dispute of fact. See Defs.’ Rebuttal at 2. Plaintiffs’ Objections and Responses make more difficult the task of determining those facts on which the parties agree, and their Counterstatement of Facts is a morass of briefing that goes far beyond identification of facts. In an effort to be fair to both sides’ factual contentions while preserving some judicial efficiency, the following approach has been taken. First, given that plaintiffs’ Objections and Responses, ECF No. 1198-1, were so lengthy, imprecise, and unyielding, any of defendants’ facts not clearly contested by plaintiffs therein will be deemed admitted. Second, because plaintiffs’ Counterstatement of Facts, ECF No. 1198-2 was so lengthy, repetitive, and argumentative, defendants’ refusal to respond piecemeal was not unreasonable, so only facts advanced by plaintiffs that are consistent with defendants’ statement of facts will be deemed undisputed.

12

BNSF adopted its first FSC in the 1990s. Defs.’ SUMF ¶ 11.1. In 2000, BNSF adopted a formula similar to that used by the other defendants at the time: a 2.0% surcharge triggered at a price of $28/bbl and increasing by 1.0% for every $2.50 increase in price. Pls.’ COF ¶ 25; Defs.’ Mem. at 33. The trigger price was determined by the monthly average price of fuel on the WTI index. Defs.’ SUMF ¶ 11.3; Pls.’ Objs. & Resp. ¶ 11 (not disputing this specific fact); Pls.’ COF ¶¶ 25, 68. BNSF modified its formula again, effective July 1, 2001, to impose a surcharge that moved away from that used by the other defendants’: 1.0% at a $1.30/gal trigger price with 0.5% surcharge increase for every $0.05/gal increase in price. Defs.’ Mem. at 33; Pls.’ Objs. & Resp. ¶ 11 (not disputing these figures); Pls.’ COF ¶ 26. The new formula calculated fuel price based on an average of the last month’s prices on the HDF index instead of the WTI index. Pls.’ COF ¶ 69; Defs.’ Mem. at 33; Defs.’ SUMF ¶ 11.5.

In the period of 2000 through 2002, the other three defendants used the same formula, which was virtually identical to the one BNSF also used until its July 2001 modification: a 2.0% surcharge applicable if the WTI index was above $28/bbl for the past thirty consecutive days and increasing by 2% for every $5/bbl increase in price. NS adopted that FSC in June 2000, see Defs.’ Mem. at 33; Pls.’ COF ¶ 28; Defs.’ SUMF ¶¶ 23-24; Pls.’ Objs. & Resp. ¶ 23 (not disputing the date or formula), CSX on October 25, 2000, see Defs.’ Mem. at 33; Pls.’ COF ¶ 23; Defs.’ SUMF ¶¶ 17.3-17.4, and UP on December 1, 2002, see Defs.’ Mem. at 33; Pls.’ COF ¶ 27; Defs.’ SUMF ¶¶ 30.2, 31.1; Pls. Objs. & Resp. ¶ 31 at 152 (not disputing the accuracy of the formula itself).

During these initial years, FSCs were not universally imposed on all unregulated freight or collected from some such freight where they were imposed. For instance, in February 2003, 39% of CSX’s carload traffic revenue was covered by FSCs. Defs.’ SUMF ¶ 19.4; Pls.’ Objs. &

13

Resp. ¶ 19 at 101 (contesting that “CSX successfully imposed” (emphasis added) and collected FSCs but not contesting that they were imposed on 39% of revenue). Some plaintiffs in this litigation were nevertheless subject to contracts with rate-based FSCs even during this period prior to the alleged conspiracy. Defs.’ SUMF ¶¶ 7, 47.7; Pls.’ Objs. & Resp. ¶ 7 at 31, ¶ 47 at 239 (not disputing those points).

b) FSCs During Alleged Conspiracy Beginning in 2003 As fuel prices continued to rise in 2002 and early 2003, defendants reconsidered their FSC formulas. See Defs.’ SUMF ¶ 3 (demonstrating fuel costs); Pls.’ Objs. & Resp. ¶ 3 (not contesting statistics or graph of fuel prices presented by defendants); infra Part III.D.2. Each defendant made an FSC formula change in 2003 or 2004. Plaintiffs contend that these changes were conspiratorial.

On March 20, 2003, CSX announced a new formula to take effect June 1, 2003. See Defs.’ Mem. at 34; Defs.’ SUMF ¶ 20.8; Pls.’ COF ¶ 111 (not disputing the formula but suggesting the announcement was March 19, not March 20, based on PX163 at ‘357, which difference is inconsequential); PX163 at ‘357, Pls.’ App’x Vol. 3, ECF No. 1200-3 (sealed) (March 19, 2003, internal CSX email, stating merely that the press release would go out that day); DA2616, Defs.’ App’x Vol. 4, ECF No. 1188-4 (sealed) (Mar. 20, 2003, press release). A surcharge of 0.4% would be imposed at a lower trigger of $23.01/bbl with a 0.4% surcharge increase for every $1/bbl increase in price. See PX163 at ‘357, Pls.’ App’x Vol. 3; DA2616, Defs.’ App’x Vol. 4; DA2745, id. (Mar. 31, 2003, letter from Pricing Director, stating that announcement was made on March 20, 2003). Fuel price remained tied to the WTI index and was calculated based on average prices from the prior month, not the minimum of the last thirty consecutive days. See PX163 at ‘357, Pls.’ App’x Vol. 3; DA2745 at 2745-46, Defs.’ App’x

14

Vol. 4. A few weeks later, on April 4, 2003, UP announced that it would match CSX’s formula, also to take effect June 1, 2003. Defs.’ SUMF ¶ 33.4; Pls.’ Objs. & Resp. ¶ 33 at 164-65 (not disputing these facts).

BNSF, on the other hand, announced a separate new formula of its own on May 7, 2003.

Having already adopted the HDF index in 2001, BNSF maintained the index but changed the trigger to $1.25/gallon, imposing a 0.5% surcharge at that price and then increasing the surcharge by 0.5% for every $0.05/gal increase in fuel price. See Defs.’ SUMF ¶ 13.9; Pls.’ SUMF ¶ 13 at 74-75 (not contesting this formula). BNSF kept its calculation of fuel price based on the average monthly price. See Defs.’ SUMF ¶ 13.9 (citing DA3979, Defs.’ App’x Vol. 9 (June 1, 2003, BNSF Rules Book 6100-A, Item 3375C)); Pls.’ SUMF ¶ 13 at 74-75 (not contesting this formula).

On May 9, 2003, despite UP’s prior announcement about following CSX, UP changed course. UP announced a different surcharge that, like BNSF’s, relied on the HDF benchmark. See Defs.’ SUMF ¶ 33.6; Pls.’ Objs. & Resp. ¶ 33 at 165-66 (not disputing that fact). UP’s trigger was ten cents higher than BNSF’s, at $1.35/gal, though its formula used a mathematically equivalent surcharge of 1.5% with the same increment, a 0.5% increase for every $0.05/gal increase in fuel price. Fuel prices under that formula would likewise be calculated based on the average of the prior month’s HDF prices. Defs.’ SUMF ¶¶ 33.6, 33.9 (citing DA1563, Defs.’ App’x Vol. 2 (May 9, 2003, fuel surcharge revision notice)); Defs.’ Mem. at 34; Pls.’ Objs. & Resp. ¶ 33 at 165-66 (not disputing the formula).

NS made no changes that spring, instead waiting until the next year to make an adjustment. On January 9, 2004, NS announced a new formula, effective March 1, 2004, that mimicked CSX’s. See Defs.’ SUMF ¶ 26.6; Pls.’ Objs. & Resp. ¶ 26 at 139-40 (not contesting

15

this fact). NS imposed a 0.4% surcharge at the trigger price of $23.01/bbl based on the prior month’s average of the WTI and increased that surcharge 0.4% for every $1/bbl increase in price. Defs.’ SUMF ¶ 26.6 (citing DA5936, Defs.’ App’x Vol. 16 (Jan. 8, 2004, internal NS email from Seale to David Goode et al., describing the new FSC)); Defs.’ Mem. at 34; Pls.’ Objs. & Resp. ¶ 26 at 139-40 (not contesting the formula itself). NS maintained that formula until 2006 when NS “rebased,” in other words, transferred much of the fuel recovery from the FSC into its base price. Defs.’ SUMF ¶¶ 27.1, 51.4; Pls.’ Objs. & Resp. ¶¶ 27, 51 (not contesting the change but only its implications). At that point, NS increased its trigger point to $64, decreased its increment to 0.3% per $1 changes in the WTI price, and increased its base price by 16.4%. Defs.’ SUMF ¶ 27.1.

Defendants increased their FSC coverage over time, applying FSCs more broadly across their contracts for carload traffic throughout the mid-2000s. At the highest point in 2006, more than half of each defendants’ carload traffic transactions by revenue were covered by an FSC, with over 90% of such transactions by NS covered. See DA21943 at 22036, Defs.’ App’x Vol. 34, ECF No. 1172-34 (sealed) (Report of plaintiffs’ experts Dr. Douglas Bernheim and Dr. Leslie Marx (“Bernheim & Marx, Pls.’ Expert, Report”), Fig. IV-14, cited at Defs.’ SUMF ¶ 19.4 n.102); Pls.’ Objs. & Resp. at 33 (citing same table).

2. Coal Traffic FSCs Prior to the alleged conspiracy period, all defendants applied their carload formulas to their coal traffic. During the alleged conspiracy period, UP and BNSF adopted coal-specific FSCs, while, in contrast, CSX and NS continued to apply their general carload formulas to coal- specific traffic too. See Defs.’ SUMF ¶¶ 21, 28; Pls.’ Objs. & Resp. ¶ 21 (not disputing the formulas); Pls.’ Objs. & Resp. ¶ 28 (admitting that NS applied carload formula to coal shipments during the conspiracy). In June 2004, BNSF announced a different formula for coal traffic,

16

taking effect July 1, 2004. The surcharge was 0.5% at a $1.24/gal trigger with an increase of 0.5% for every $0.04/gal increase and with fuel prices calculated based on the average of the last month’s prices in the HDF index. Defs.’ SUMF ¶ 14.4; Pls.’ Objs. & Resp. ¶ 14 (not disputing the formula itself); Pls.’ COF ¶ 290.

In October 2004, UP also announced a different FSC formula for coal traffic than its carload formula. In contrast to BNSF’s coal-traffic formula, UP’s new coal traffic formula used a tiered scheme with a trigger of $1.35/gal calculated based on the past month’s average of the HDF. See Defs.’ SUMF ¶ 34.3; Pls.’ COF ¶ 291. Between $1.35 and $1.60/gal, UP’s surcharge increments were the same as its carload (0.5% for every $0.05 increase); for prices above $1.60/gal, there was a 0.75% increase for each $0.05 increase in the HDF price. See Defs.’ SUMF ¶ 34.3; Pls.’ Objs. & Resp. ¶ 34 (not disputing formula); Pls.’ COF ¶ 291.

BNSF announced a change in its formula again less than a year later. In March 2005, BNSF announced a shift to a mileage-based, as opposed to a rate-based, FSC to take effect January 1, 2006, for all traffic, i.e., both carload and coal traffic. Defs.’ SUMF ¶¶ 15.1, 15.6; Pls.’ Objs. & Resp. ¶ 15 (not disputing the basic facts). BNSF had previously considered but never before adopted a mileage-based surcharge. Defs.’ SUMF ¶¶ 15.1, 15.2; Pls.’ Objs. & Resp. ¶ 15 at 89 (discussing BSNF’s earlier considerations of FSCs). This FSC imposed a $0.01/mile charge triggered at $1.25/gal with a $0.01/mile increase for every $0.06/gal price increase; prices were determined by the prior month’s average on the HDF index. Defs.’ Mem. at 35; Pls.’ Objs. & Resp. ¶ 15 (not disputing formula). In October 2005, however, BNSF announced that the mileage-based surcharge would be applied only to coal traffic in January 2006, noting customers’ need for more time to adjust to the new FSC. Defs.’ SUMF ¶¶ 15.4- 15.6; DA4635, Defs.’ App’x Vol. 12, ECF No. 1172-12 (sealed) (Oct. 20, 2005, internal BNSF

17

email from Linda Wertz to listserv, attaching advisory shared with customers); Pls.’ Objs. & Resp. ¶ 15 (not disputing the announcement); Pls.’ COF ¶¶ 556-57. BNSF’s mileage-based formula, UP’s coal-specific formula, and CSX and NS’s carload formulas applied to coal traffic for the remainder of the alleged conspiracy period.

3. Intermodal Traffic FSCs All four defendants also adopted FSCs on intermodal traffic prior to 2003 and then continued to impose FSCs throughout the alleged conspiracy period. BNSF’s first intermodal FSC took effect on March 1, 2000. At the price of $1.10/gal an FSC of 0.5% applied with increases of 0.5% for every $0.05/gal increase in fuel price. Fuel prices were calculated based on the weekly average of the HDF index. Defs.’ Mem. at 72; Defs.’ SUMF ¶ 38.1 (citing DA3969 at 3969-70, Defs.’ App’x Vol. 9); Pls.’ Objs. & Resp. ¶ 38 (not disputing these figures). BNSF changed its formula in September 2001 to have a 1.0% surcharge at $1.24/gal, increasing by 0.5% for every $0.04/gal price increase. See Defs.’ SUMF ¶ 41.2; Pls.’ Objs. & Resp. ¶ 41 at 202 (not contesting the formula); Defs.’ Mem. at 72. BNSF did not change its formula again during the alleged conspiracy period. Defs.’ SUMF ¶ 41.2; Pls.’ Objs. & Resp. ¶ 41 at 202 (not contesting that point); Defs.’ Mem. at 72.

NS also adopted intermodal FSCs in the early 2000s, though NS had separate domestic and international formulas. Effective March 20, 2000, NS’s Domestic division adopted a flat 4% FSC on all intermodal traffic. Defs.’ Mem. at 72; Defs.’ SUMF ¶ 38.2; Pls.’ Objs. & Resp. ¶ 38 (not disputing this formula). In April 2002, NS Domestic adopted a formula that matched BNSF’s 2001 formula: a 1.0% surcharge at $1.24/gal, increasing by 0.5% for every $0.04/gal price increase. Defs.’ Mem. at 72; Defs.’ SUMF ¶ 38.2; Pls.’ Objs. & Resp. ¶ 38 (not disputing the formula). NS Domestic did not change its published intermodal formula during the alleged

18

conspiracy period. Defs.’ SUMF ¶ 42.2; Pls.’ Objs. & Resp. ¶ 42 at 208 (not disputing that modifications to the published formula were not made).

Effective March 1, 2001, NS’s International division also imposed an FSC on intermodal traffic: 1.0% starting at $1.30/gal, increasing 1.0% for every $0.01/gal increase in price, with the price based on the average of the HDF index for the prior three weeks. See Defs.’ Mem. at 72. NS International changed the FSC formula twice during the alleged conspiracy period, adopting NS’s Domestic formula in August 2006 and adjusting to a price calculated by the HDF monthly average in August 2007. Defs.’ SUMF ¶¶ 42.3, 42.4; Pls.’ Objs. & Resp. ¶ 42 (not disputing the FSC formulas).

CSX Intermodal (“CSXI”), CSX’s Intermodal division, established an FSC for intermodal traffic starting April 1, 2000. Defs.’ SUMF ¶ 38.3; Pls.’ Objs. & Resp. ¶ 38 (not disputing the adoption of FSC formula); Defs.’ Mem. at 72. 5 The FSC was calculated using the following formula: ((HDF on last Monday of the prior month - $1.10)/$1.10)*0.10*100. See Defs.’ Mem. at 72. In October 2004, CSXI changed its formula to match BNSF’s intermodal formula with a surcharge of 1.0% at $1.24/gal with an increase of 0.5% for every $0.04/gal increase in price, though CSXI calculated the price differently—using the HDF of the last Monday of the previous month. See Defs.’ Mem. at 72; see also Defs.’ SUMF ¶¶ 14.4, 44; Pls.’ Objs. & Resp. ¶ 44 (not disputing the formula).

UP adopted an intermodal FSC on November 1, 2000 with a unique formula: (HDF price/$1.292 -1) * 20.0% [the fuel weight]. See Defs.’ Mem. at 72; Defs.’ SUMF ¶ 43.1. The

5 The parties dispute whether the entity that provided intermodal rail service during the alleged conspiracy period, CSXI, is a defendant in MDL I, see infra Part I.C (Procedural History), because CSXI was not specifically named and is a distinct subsidiary of CSX Corporation, separate from the named defendant, CSX Transportation, another CSX Corporation subsidiary. See Pls.’ Objs. & Resp. ¶ 37 at 190; Defs.’ SUMF ¶ 37. Plaintiffs argue, however, that CSXI merged into the named defendant CSX (i.e., CSX Transportation) in 2010 so CSX can be held liable for the corporations’ intermodal shipments. Pls.’ Objs. & Resp. at 190-91. Plaintiffs cannot survive summary judgment on intermodal shipments regardless, so this dispute need not be resolved.

19

HDF price was calculated based on the prior month’s average. See Defs.’ SUMF ¶ 38.4 (citing DA19037, Defs.’ App’x Vol. 30, ECF No. 1172-30 (sealed) (Nov. 8, 2000, customer letter)); Pls.’ Objs. & Resp. ¶ 43 (not disputing any of these figures but only whether they were actually collected or broadly imposed pre-conspiracy period). UP adjusted the figures in this formula in October 2001, see Defs.’ Mem. at 72; Defs.’ SUMF ¶ 43.2, and then made three more adjustments during the alleged conspiracy period. In 2005, UP adjusted the formula to: (HDF price/$1.253 – 1) * 18.5% [the fuel weight]. See Defs.’ Mem. at 72; Defs.’ SUMF ¶ 43.2. Then, in 2006, UP used the formula: (HDF price/$1.253 – 1) * 16.5 [the fuel weight]. See Defs.’ Mem. at 72; Defs.’ SUMF ¶ 43.3. In 2007, UP made the formula dependent on a weekly rather than monthly HDF measure. See Defs.’ SUMF ¶ 43.5; Pls.’ Objs. & Resp. ¶ 43 (not disputing the formulas).

In short, no defendant altered its published intermodal formula from the start of 2003 through October 2004—the period when all four defendants altered their carload formulas and the alleged conspiracy supposedly formed. See Defs.’ SUMF ¶ 40; Pls.’ Objs. & Resp. ¶ 40 (not insisting otherwise). Defendants’ intermodal formulas only overlapped in the respect that BNSF’s formula effective in late 2001 was adopted by NS Domestic in April 2002 and NS International in 2006.

4. Adjustments in Response to STB Action Shippers of rate-regulated traffic complained to the STB about rate-based FSCs, and the STB, in 2007, determined that the “practice [of] comput[ing] fuel surcharges as a percentage of the base rates” was “unreasonable” because they have “no prospect of reflecting the actual increase in fuel costs for handling [that] particular traffic”—i.e., the railroads’ variable costs— given that base-rate pricing differs by route and commodity. PX653 at 6, Pls.’ App’x Vol. 10B, ECF No. 1202-4 (sealed) (Jan. 25, 2007, STB decision); Rail Fuel Surcharges (“STB

20

Decision”), Ex Parte No. 661, 2007 WL 201205, at *1, 4 (S.T.B. Jan. 25, 2007); see also Defs.’ SUMF ¶ 52.1; MDL I Class Cert. II, 725 F.3d at 248. 6 As the STB explained, because the base rates for regulated traffic are not tied to fuel but rather other factors such as the number of transportation alternatives, “[t]wo shippers may have traffic with identical fuel costs, but if one starts out with a higher base rate . . ., it will pay dramatically more in fuel surcharges.” STB Decision, 2007 WL 201205, at *4; see id. (“Because railroads rely on differential pricing, under which rates are dependent on factors other than costs, [for rate-regulated traffic] a surcharge that is tied to the level of the base rate, rather than to fuel consumption for the movement to which the surcharge is applied, cannot fairly be described as merely a cost recovery mechanism.”). The STB directed rate-regulated carriers to “change this practice,” id. at *1, but explicitly differentiated other FSC options, such as mileage-based FSCs, which they suggested were a reasonable alternative, id. at *6. The STB also noted particular concern with the practice of “double dipping,” wherein railroads would sometimes “apply[] to the same traffic both a fuel surcharge and a rate increase that is based on a cost index that includes a fuel component, such as the [RCAF].” Id. at *1. Railroads were instructed to “change th[at] practice as well.” Id.

Although the shippers had requested both a prospective rule and retroactive damages, the STB rejected the damages request, making clear it was “not awarding any remedies for rate- based fuel surcharges that may have been used in the past.” Id. at *5. In fact, the STB recognized that “[i]n view of the long history of rate-based fuel surcharges in the rail industry,” railroads could not “be faulted for assuming that fuel surcharges calculated as a percentage of the

6 Generally, “captive shippers,” those who do not have transportation alternatives to rail, are subject to regulated rates. See U.S. GOV’T ACCOUNTABILITY OFF., GAO-17-166, Freight Rail Pricing: Contracts Provide Shippers and Railroads Flexibility, but High Rates Concern Some Shippers at i (Dec. 2016), https://www.gao.gov/assets/gao-17-166.pdf. The largest industries for regulated traffic are agriculture, chemicals, coal, food, and nonmetallic minerals. Id. at 2.

21

base rate were permissible.” Id. at *7. The STB also specified that its ruling had no bearing on unregulated traffic, which the STB has “no authority to regulate” and thus no basis upon which to opine that pricing practices are “unreasonable.” Id. at *10. For rail pricing “governed by a contract,” railroads and shippers could continue to negotiate mutually agreeable arrangements to allocate fuel costs and manage fluctuations without STB interference and regardless of whether such agreements reflected the railroads’ variable costs. Id. Despite the limits of the STB’s decision to the specific context of rate-regulated traffic and its rejection of retroactive damages, shippers of unregulated freight traffic were nonetheless motivated to make a similar challenge to the use of FSCs on unregulated traffic in a judicial forum, seeking damages in suits eventually consolidated in the instant multi-district litigation.

Defendants around that time responded to the STB decision by adjusting their FSC formulas for both rate-regulated and unregulated traffic. Specifically, in March 2007, CSX announced a mileage-based surcharge for its rate-regulated traffic and also applied this surcharge to new private contracts for non-intermodal traffic. Defs.’ SUMF ¶ 52.2; Pls.’ Objs. & Resp. ¶ 52 at 248-49 (disputing only how quickly and effectively the mileage-based FSCs were actually applied to unregulated traffic). UP announced the same approach as CSX the same month. Defs.’ SUMF ¶ 52.3; Pls.’ Objs. & Resp. ¶ 52 at 248-49 (not disputing that fact). UP “rebased”—increasing its base rate while raising the trigger price so that more of the fuel cost would be recovered by the base rate. Defs.’ SUMF ¶ 52.3; Pls.’ Objs. & Resp. ¶ 52 at 249-50 (disputing only that this marked a shift away from conspiratorial rates and emphasizing this shift was revenue neutral). BNSF expanded its existing mileage-based FSC to cover regulated traffic. Defs.’ SUMF ¶¶ 52.5-52.6; Pls.’ Objs. & Resp. ¶ 52 at 249 (not disputing the expansion). NS eliminated FSCs on regulated shipments but continued with its 2006 “rebased” formula for

22

unregulated shipments. Defs.’ SUMF ¶ 52.8; Pls.’ Objs. & Resp. ¶ 52 at 251-52 (not disputing NS’s formula).

5. Summary Charts

The below tables summarize defendants’ FSCs for carload, coal, and intermodal traffic from 2000 until early 2007 (the time of the STB decision), with comparative terms outlined. See Defs.’ Mem. at 33-35, 72. Plaintiffs confirm that they do not dispute the formulas themselves or the dates as represented in these charts. See Hearing (June 18, 2025) on Mots. for Summ. J.

Transcript (“MSJ Hr’g Tr.”) at 65:21-24, ECF No. 1263.

Carload and Coal FSCs

BNSF CSX NS UP

2000- Oct. 1, 2000 Oct. 25, 2000 Jun. 30, 2000 Dec. 1, 2002 2002 WTI benchmark SAME SAME SAME $28/bbl trigger SAME SAME SAME 2.0% surcharge at trigger SAME SAME SAME $2.50/bbl increment $5/bbl increment $5/bbl increment $5/bbl increment 1.0% surcharge increment 2.0% surcharge increment 2.0% surcharge increment 2.0% surcharge increment Monthly average index Index per prior 30 consec. Index per prior 30 consec. Index per prior 30 consec. days days days Jul. 1, 2001 HDF Benchmark $1.30/gal 1.0% surcharge at trigger $0.05/gal increment 0.5% surcharge increment Monthly average index

2003 May 7, 2003 (ann.); Mar. 20, 2003 (ann.); Dec. 2003 (ann.); May 9, 2003 (ann.);

Jun. 1, 2003 (eff.) Jun. 1, 2003 (eff.) Mar. 1, 2004 (eff.) Jun. 1, 2003 (eff.)

HDF benchmark WTI benchmark WTI benchmark HDF benchmark $1.25/gal trigger $23.01/bbl trigger $23.01/bbl trigger $1.35/gal 0.5% surcharge at trigger 0.4% surcharge at trigger 0.4% surcharge at trigger 1.5% surcharge at trigger

23

$0.05/gal increment $1/bbl increment $1/bbl increment $0.05/gal increment 0.5% surcharge increment 0.4% surcharge increment 0.4% surcharge increment 0.5% surcharge increment Monthly average index Monthly average index Monthly average index Monthly average index

Note: Apr. 4, 2003, announced matching CSX.

2004 Coal-specific surcharge Coal-specific surcharge Jun. 2004 (ann.); July 1, Oct. 2004 (ann.); Nov. 2004 2004 (eff.) (eff.)

HDF benchmark HDF benchmark $1.24/gal trigger $1.35/gal trigger 0.5% surcharge at trigger 1.5% surcharge at trigger $0.04/gal increment $0.05/gal increment 0.5% surcharge increment 0.5% surcharge increment Monthly average index Monthly average index Second tier:

$1.60/gal trigger

4.25% surcharge at trigger $0.05/gal increment

0.75% surcharge increment Monthly average index

2005 MILEAGE-based surcharge applied to Coal Mar. 2005 (ann.); Jan. 1, 2006 (eff.)

HDF benchmark $1.25/gal trigger $0.01/mile surcharge at trigger $0.06/gal increment $0.01/mile surcharge increment Monthly average index

2006 Jul. 2006 WTI benchmark

24

$64/bbl trigger

1.0% surcharge at trigger $1/bbl increment

0.3% surcharge increment

Monthly average index

Increase in base rate of

16.4%

Intermodal FSCs

BNSF CSX NS UP NS Domestic NS Int’l 2000- Mar. 1, 2000 Apr. 1, 2000 Mar. 20, 2000 Mar. 1, 2001 Nov. 1, 2000 2002 HDF benchmark HDF benchmark HDF benchmark HDF benchmark HDF benchmark $1.10/gal trigger ((HDF on last Mon of 4% flat rate $1.30/gal trigger (HDF price/$1.292 – month - 1)*20.0% [the fuel 0.5% surcharge at trigger $1.10)/$1.10)*0.10*100 1.0% surcharge weight]

$0.05/gal increment $0.10/gal increment 0.5% surcharge increment 1.0% surcharge incr.

Weekly average index Daily fuel price Weekly average index Previous 3 weeks avg. index Monthly avg. index

Late 2001 Apr. 1, 2002 Oct. 9, 2001 HDF Benchmark HDF benchmark HDF benchmark $1.24/gal $1.24/gal (HDF price/$1.253 – 1)*20.0% [the fuel

1.0% surcharge at trigger 1.0% surcharge weight]

$0.04/gal increment $0.04/gal increment 0.5% surcharge increment 0.5% surcharge incre.

Weekly average index Weekly average index Monthly avg. index 2003

2004 October 4, 2004 HDF benchmark

$1.24/gal trigger

1.0% surcharge at trigger $0.04/gal increment

0.5% surcharge increment

25

Last Mon. of previous

month

2005 Apr. 1, 2005 HDF benchmark

(HDF price/$1.593 –

1)*18.5% [the fuel

weight]

Monthly avg. index

2006 Aug. 21, 2006 Oct. 1, 2006 HDF benchmark HDF benchmark $1.24/gal trigger (HDF price/$1.253 – 1)*16.5 [the fuel

1.0% surcharge weight]

$0.04/gal increment

0.5% surcharge incr.

Weekly avg. index Monthly avg. index

C. Procedural Background Given the nearly eighteen-year-long pendency of this litigation, the procedural history is extensive. Only the key elements are outlined here, with more detail available in the cited prior decisions.

1. Procedural History The first plaintiff shippers challenged the four defendant railroad companies’ FSCs in a lawsuit filed in the District of New Jersey on May 14, 2007, a few months after the STB decision. See Defs.’ SUMF ¶ 53; Pls.’ Objs. & Resp. ¶ 53 at 254 (in agreement). Plaintiffs alleged that defendants’ executives agreed to impose artificially high and uniform FSCs on shippers and achieved the express goal of 100% FSC coverage and virtually identical fuel surcharges, in part, by settling on common trigger points and pressuring the Association of American Railroads (“AAR”) to create a new cost index, the All Inclusive Index Less Fuel (“AIILF”) to replace the preexisting index. MDL I MTD Order (Direct Purchasers), 587 F.

26

Supp. 2d at 30-31; In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL I Class Cert. I”), 287 F.R.D. 1, 47 (D.D.C. 2012). Eighteen separate class actions were filed in six districts that year before they were consolidated by the MDL Panel into MDL I, MDL No. 1869, D.D.C. Misc. No. 7-489, in November 2007. That MDL was assigned to Judge Paul L. Friedman in the District Court for the District of Columbia. See MDL I MTD Order (Direct Purchasers), 587 F. Supp. 2d at 29. These claims involved only carload and coal traffic, not intermodal shipments. See id.; Consolidated Am. Class Action Compl., ECF No. 91-1; Indirect Purchasers Consolidated Am. Compl., ECF No. 93. Plaintiffs were divided into two putative classes—direct purchasers and indirect purchasers—and the claims of both groups survived motions to dismiss. See MDL I MTD Order (Direct Purchasers), 587 F. Supp. 2d at 29; In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL I MTD Order (Indirect Purchasers)”), 593 F. Supp. 2d 29, 32 (D.D.C. 2008) (dismissing only state-law claims due to preemption), aff’d sub. nom. Fayus Enters. v. BNSF Ry. Co., 602 F.3d 444 (D.C. Cir. 2010).

In 2011, another antitrust suit alleging price-fixing conspiracy claims against only UP and BNSF was filed in this Court by certain plaintiffs, and that case was transferred to Judge Friedman as related to MDL I. See Order Transferring Case, Oxbow, 11-cv-1049, ECF No. 37. The claims of those plaintiffs also survived a motion to dismiss. See Oxbow Carbon & Minerals LLC v. Union Pac. R.R. Co., 81 F. Supp. 3d 1, 16 (D.D.C. 2015).

Plaintiffs did not succeed in certifying a class action. Initially, the district court granted class certification for direct purchaser plaintiffs in MDL I, but that decision was vacated on appeal. See MDL I Class Cert. I, 287 F.R.D. 1 (granting class certification for direct purchasers and appointing Quinn Emmanuel and Hausfeld LLP as co-lead class counsel); MDL I Class Cert. II, 725 F.3d at 255 (vacating and remanding class certification because the damages model put

27

forth by plaintiffs failed to show that all plaintiffs were harmed by the alleged conspiracy—13% were uninjured in the experts’ model). On remand, the district court denied class certification as to direct purchasers, finding that plaintiffs failed to establish that common questions of law or fact predominated as to this putative class and that plaintiffs’ damages regression model was flawed as a means of demonstrating such predominance, a decision affirmed on appeal. See MDL I Class Cert. III, 292 F. Supp. 3d 14, aff’d 934 F.3d 619 (D.C. Cir. 2019).

Following denial of class certification, additional claims were filed in 31 actions in 16 districts, alleging the same price-fixing conspiracy by the four defendant freight railroads to coordinate their FSC programs. These new plaintiffs alleged a conspiracy period of 2003 to 2008 and encompassing carload, coal, and intermodal shipments. See Preliminary Rep. at 1, MDL II, 20-mc-8, ECF No. 78. The parties’ positions diverged on how best to address these new actions: Defendants wanted these cases transferred to a new MDL, but plaintiffs wanted them consolidated with the existing MDL I, before Judge Friedman. The MDL Panel determined that, since the newly filed actions were in an earlier procedural posture, they should not be consolidated, but that centralization in the same district could have same benefits because of efficiencies and the need to avoid inconsistent rulings. See MDL Panel Transfer Order, MDL II, 20-mc-8, ECF No. 1 (Feb. 6, 2020). This second MDL, MDL II, MDL No. 2925, 20-mc-8, was assigned to this Court, while the original MDL I proceeded before Judge Friedman.

On motions to dismiss in MDL II, defendants contended that certain of the complaints filed in that litigation were time-barred because they did not fall within the four-year statute of limitations for antitrust claims. See MDL II MTD Order on Time-Barred Claims, No. 20-mc-8 (BAH), 2020 WL 5016922, at *10 (D.D.C. Aug. 25, 2020). Plaintiffs argued that their claims were tolled based on claims brought in the putative class in MDL I, which put defendants on

28

notice and fell within the American Pipe tolling doctrine. See id. at *10-11; Am. Pipe & Constr. Co. v. Utah, 414 U.S. 538 (1974). The Court agreed generally with plaintiffs, as defendants conceded, that the core of these plaintiffs’ claims could proceed. See MDL II MTD Order on Time-Barred Claims, 2020 WL 5016922, at *12. Some plaintiffs, however, brought “novel” allegations beyond that core, id., namely that “three additional railroads participated,” that “defendants coordinated the uniform implementation of mileage-based surcharges,” and “defendants’ conspiratorial conduct” extended “outside of the 2003-2008 period defining the MDL I putative class,” id. at *1. These additional allegations about new defendants, mileage- based surcharges, and new conspiratorial conduct after 2008 were found to be untimely and not tolled by the original allegations of the putative classes. See id. at *17-27. At the same time, to the extent that post-2008 allegations “simply set forth a more complete picture of how the conspiracy alleged by the putative class was able to operate, without altering the scope of defendants’ potential liability,” such allegations were held to be permissible, as were allegations of harm that occurred before 2008 and simply had lingering effects into later years. See id. at *26. This decision was reaffirmed in a denial of a motion for reconsideration. See In re Rail Freight Surcharge Antitrust Litig. (No. II) (“MDL II Reconsideration Time-Barred Claims”), No. 20-mc-8 (BAH), 2020 WL 6198487 (D.D.C. Oct. 22, 2020).

Those three matters—MDL I, MDL II, and Oxbow—encompass all of the related antitrust claims against these four defendants. After the initial motions to dismiss were resolved, the parties in all three matters engaged in extensive discovery, resulting in multiple disputes resolved by both the district court judges and magistrate judges in numerous opinions after thorough consideration, including in several in-person hearings. In MDL I, discovery began in 2009 and extended through 2021. See Scheduling Order, ECF No. 296 (commencing discovery summer of

29

2009); Stipulated Order, ECF No. 1049 (ordering expert discovery and reports completed by December 2021). Judicial input was obtained on over a dozen discovery disputes in MDL I. 7 In MDL II, discovery opened in 2020 and closed in 2023 and required two hearings over disputes before this Court. See Scheduling Order, MDL II, ECF No. 102; Min. Entry (3/26/2021); Min. Entry (9/13/2022); Min. Order (11/8/2022) (issuing scheduling order extending the end of discovery through 2023); Min. Order (8/8/2023) (modifying schedule for finishing fact and expert discovery by end of 2023). In Oxbow, discovery opened in 2016 and eventually extended through 2020 and likewise involved a discovery hearing. See Order, Oxbow, ECF No. 97 (setting deadline for discovery as end of 2016); Scheduling Order, ECF No. 165 (extending discovery until end of 2020); Min. Entry (8/24/2017) (discovery hearing before M.J. Harvey). In MDL II alone, the parties conducted 373 depositions. Joint Status Report at 2, MDL II, ECF No. 925. The parties thus had ample opportunity to obtain a wealth of documents, amounting to millions of pages, see Mem. Op., MDL I, ECF No. 534 (noting that as of only late 2011, millions of documents had already been exchanged), with over 50,000 pages produced as exhibits for the pending motions—20,613 on the part of defendants and 29,538 on the part of plaintiffs.

In March 2024, MDL I and Oxbow were transferred from Judge Friedman to this Court.

See Order Reassigning Litigation, MDL I, 7-mc-489, ECF No. 1129; Order Reassigning Litigation, Oxbow, 11-cv-1049, ECF No. 278.

7 See, e.g., Mem. Op. Denying Defs.’ Mot. for Phased Discovery, MDL I, ECF No. 291 (Facciola, M.J.);

Mem. Op. Grant’g in Part and Denying in Part Pls.’ Mot. to Compel, ECF No. 313 (Facciola, M.J.); Mem. Op. Grant’g in Part and Denying in Part Pls.’ Second Mot. to Compel, ECF No. 371 (Facciola, M.J.); Mem. Order Grant’g in Part and Denying in Part Pls.’ Mot. for In Camera Rev. of Withheld Docs, ECF No. 374 (Facciola, M.J.); Mem. Order Denying Third-Party Mot. to Quash Subpoena, ECF No. 418 (Facciola, M.J.); Mem. Op. Re: In Camera Rev., ECF No. 436 (Facciola, M.J.); Mem. Order Re: Scope of 30(b) Depositions, ECF No. 487 (Facciola, M.J.); Mem. Op. Grant’g in Part and Denying in Part Defs.’ Mot. to Compel, ECF No. 534 (Facciola, M.J.); Mem. Op. & Order Grant’g Defs.’ Mot. to Compel, ECF No. 699 (Facciola, M.J.); Op. & Order Grant’g Pls.’ Leave to File Supp’l Expert Rep., ECF No. 756 (Friedman, J.); Mem. Op. & Order Denying Defs.’ Mot. for Leave to File Late Dec. of Dr. Eakin, ECF No. 786 (Friedman, J.).

30

While many other motions have been decided over the lengthy pendency of these cases, as referenced, one additional motion is notable. In preparation for dispositive motions briefing in MDL I, defendants revived, in 2020, an issue that had previously been the subject of defendants’ motion raised at the class certification stage but never decided, namely, to exclude interline- related communications from any prohibited purpose in accord with 49 U.S.C. § 10706. See Defs.’ Mot. to Exclude Interline-Related Communications, MDL I, ECF No. 927; see also Defs.’ Mot. to Exclude Interline-Related Communications from Consideration for Class Certification, ECF No. 417; In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL I Section 10706 Decision I”), 520 F. Supp. 3d 1, 9 (D.D.C. 2021) (explaining that the court “did not reach defendants’ [prior] Section 10706 motion”). In 2021, the district court denied defendants’ motion to exclude such evidence and bar certain inferences. See MDL I Section 10706 Decision I, 520 F. Supp. 3d at 38. That decision was appealed to the D.C. Circuit, which vacated and remanded the district court’s opinion and gave further guidance on application of section 10706. See MDL I Section 10706 Decision II, 34 F.4th at 5.

Judge Friedman then ordered the parties to meet and confer, considering that the “D.C.

Circuit ha[d] made clear how to read and apply Section 10706 in this case” so the “parties should [have] be[en] able to determine by mutual agreement—without any involvement of the Court— which documents relating to the pending summary judgment motions should be excluded, in whole or in part.” See Mem. Op. & Order at 2, MDL I, ECF No. 1088. Despite significant agreement, the parties still needed judicial guidance as to some documents. See Joint Status Rep., ECF No. 1094. After reviewing the disputed documents and ordering further briefing in 2023, Judge Freidman granted in part and denied in part defendants’ request to exclude certain evidence early the next year. See In re Rail Freight Fuel Surcharge Antitrust Litig. (“MDL I

31

Section 10706 Decision III”), 721 F. Supp. 3d 16 (D.D.C. 2024) (reviewing all of the evidence submitted by the parties in joint status report, ECF No. 1094, which outlined the documents relevant to the summary judgment motions). In general, that opinion excluded records about freight movements constituting solely shared interline traffic, any discussions of interline traffic, and any documents about such discussions. Id. Any deviations from interline discussions must have been fleeting or inconsequential to be excluded under § 10706. See MDL I Section 10706 Decision, 721 F. Supp. 3d at 22-23 (D.D.C. 2024). The parties have affirmed that although Section 10706-excluded material may remain in expert reports or elsewhere in the record, they have not relied on such material in their briefs in support of their pending dispositive and related motions. See Defs.’ Mem. at 8 n.2; Pls.’ Opp’n at 140.

2. Pending Motions

Pursuant to a scheduling order issued by this Court shortly after transfer of MDL I and Oxbow, defendants moved for summary judgment in 2024. See Scheduling Order, MDL I, ECF No. 1148; Min. Order (June 28, 2024) (amending that schedule). The parties briefed the issues, as well as related motions across the three dockets, MDL I, MDL II, and Oxbow. See supra n.1. Defendants filed individual motions for summary judgment, see NS’s Mot. for Summ. J., ECF No. 1155 (sealed); UP’s Mot. for Summ. J., ECF No. 1157 (sealed); BNSF’s Mot. for Summ. J., No. 1158 (sealed); CSX’s Mot. for Summ. J., ECF No. 1160 (sealed), with a common memorandum in support, ECF No. 1163. Plaintiffs filed a common opposition, ECF No. 1218, and defendants filed a common reply, Defs.’ Reply in Supp. of Summ. J. (“Defs.’ Reply”), ECF No. 1048 (sealed). Defendants also filed in MDL II a partial motion for summary judgment based on allegedly time-barred damages claims. See MDL II, ECF No. 983.

In addition to defendants’ combined and separate motions for summary judgment, seven additional motions are pending. Defendants filed three motions to exclude certain opinions or

32

evidence, including (1) to exclude plaintiffs’ expert’s average damages opinions, under Federal Rules of Evidence 702 and 403, see Defs.’ Mot. to Exclude Avg. Dam. Ops., ECF No. 1169 (sealed); Pls.’ Opp’n to Defs.’ Mot. to Exclude Avg. Dam. Ops., ECF No. 1196 (sealed); Defs.’ Reply in Supp. Mot. to Exclude Avg. Dam. Ops., ECF No. 1234 (sealed); (2) to exclude plaintiffs’ positive econometric overcharge models, under Rule 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), see Defs.’ Mot. to Exclude Overcharge Models, ECF No. 1165 (sealed); Pls.’ Opp’n to Defs.’ Mot. to Exclude Overcharge Models, ECF No. 1195 (sealed); Defs.’ Reply in Supp. Mot. to Exclude Overcharge Models, ECF No. 1240 (sealed); and (3) to exclude expert reliance on improper evidence, invoking Federal Rule of Evidence 702 and the statutory use bar in 49 U.S.C. § 10706, see Defs.’ Mot. to Exclude Expert Reliance, ECF No. 1167 (sealed). The last defense motion concerning purportedly “improper evidence” focuses on expert testimony about the existence of the alleged conspiracy, testimony on defendants’ alleged state of mind and intentions, testimony providing improper factual narratives, and testimony covering inadmissible material under § 10706. See id.; see also Pls.’ Opp’n to Defs.’ Mot. to Exclude Expert Reliance, ECF No. 1196 (sealed); Defs.’ Reply in Supp. Mot. to Exclude Expert Reliance, ECF No. 1236 (sealed).

For their part, plaintiffs filed four motions to exclude certain opinions or declarations, including (1) to exclude, under Rules 702 and Daubert, the opinion of defendants’ expert, Joseph Dettmar, who they argue improperly opines on the STB’s encouragement and endorsement of cooperative arrangements like interline alliances, see Pls.’ Mot. to Exclude Dettmar, ECF No. 1171 (sealed); Defs.’ Opp’n to Pls.’ Mot. to Exclude Dettmar, ECF No. 1192 (sealed); Pls.’ Reply in Supp. Mot. to Exclude Dettmar, ECF No. 1244 (sealed); (2) to exclude, under Rules 702 and Daubert, defendants’ rebuttal expert, Avram Tucker, who was retained to analyze the

33

reliability and reasonableness of plaintiffs’ experts’ methods and opinions, see Pls.’ Mot. to Exclude Tucker, ECF No. No. 1175 (sealed); Defs.’ Opp’n to Pls.’ Mot. to Exclude Tucker, ECF No. 1189 (sealed); Pls.’ Reply in Supp. Mot. to Exclude Tucker, ECF No. 1248 (sealed); (3) to strike, as improper, under Federal Rule of Evidence 1006, two declarations—those of Eric Fanchiang and Kelsey Bryan, which summarize information about defendants’ price authorities, see Pls.’ Mot. to Strike, ECF No. 1197 (sealed); Defs.’ Opp’n to Pls.’ Mot. to Strike, ECF No. 1227 (sealed); Pls.’ Reply in Supp. of Pls.’ Mot. to Strike, ECF No. 1231 (sealed); and (4) in the Oxbow case, to exclude, under Federal Rules of Evidence 403 and 702, the opinions of defendants’ experts, Dr. Edward Snyder and Dr. Lauren Stiroh, whose opinions pertained at least in part to claims that were dismissed pursuant to a stipulation and, in plaintiffs’ view, express opinions that are now irrelevant, see Pls.’ Mot. to Exclude Snyder & Stiroh, Oxbow, ECF No. 309 (sealed); Defs.’ Opp’n to Pls.’ Mot. to Exclude Snyder & Stiroh, ECF No. 324 (sealed); Pls.’ Reply in Supp. Mot. to Exclude Snyder & Stiroh, ECF No. 367 (sealed). 8 II. LEGAL STANDARD In general, a court must “grant summary judgment if the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” FED. R. CIV. P. 56(a). “As to materiality, the substantive law will identify which facts are material. Only disputes over facts that might affect the outcome of the suit under the governing law will properly preclude the entry of summary judgment. Factual disputes that are irrelevant or unnecessary will not be counted.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986). As to a genuine dispute, the “evidence [must be] such that a reasonable jury could return a verdict for the nonmoving party.” Id. At the summary judgment stage, “courts are required to

8 The expert reports challenged by the parties are not relied upon in reaching the findings and conclusions on the defendants’ summary judgment motions addressed in this Memorandum Opinion.

34

view the facts and draw reasonable inferences ‘in the light most favorable to the party opposing the [summary judgment] motion.’” Scott v. Harris, 550 U.S. 372, 378 (2007) (alterations in original) (quoting United States v. Diebold, Inc., 369 U.S. 654, 655 (1962)).

Section 1 of the Sherman Act forbids any “contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce.” 15 U.S.C. § 1. To survive a motion for summary judgment in a section 1 Sherman Act case, plaintiffs “must establish that there is a genuine issue of material fact as to whether [defendants’] entered into an illegal conspiracy that caused [plaintiffs] to suffer a cognizable injury.” Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 585-86 (1986); see also Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 764 (1984) (plaintiffs must prove a “conscious commitment to a common scheme designed to achieve an unlawful objective”).

Although “‘on summary judgment the inferences to be drawn from the underlying facts . . . must be viewed in the light most favorable to the party opposing the motion[,]’ . . . antitrust law limits the range of permissible inferences from ambiguous evidence in a § 1 case.” Matsushita, 475 U.S. at 587-88 (quoting Diebold, Inc., 369 U.S. at 655). For instance, “if the claim is one that simply makes no economic sense[,] [plaintiffs] must come forward with more persuasive evidence to support their claim than would otherwise be necessary.” Id. at 587.

Moreover, “conduct as consistent with permissible competition as with illegal conspiracy does not, standing alone, support an inference of antitrust conspiracy.” Id. at 588. This means that plaintiffs’ evidence must support an inference of conspiracy that is “reasonable in light of the competing inferences of independent action or collusive action,” by “‘tend[ing] to exclude the possibility’ that the alleged conspirators acted independently.” Id. (quoting Monsanto Co., 465 U.S. at 764). By tending to exclude the possibility of independent action, plaintiffs’

35

evidence need not render such an inference completely impossible, since “[r]equiring a plaintiff to ‘exclude’ or ‘dispel’ the possibility of independent action places too heavy a burden on the plaintiff.” In re Publ’n Paper Antitrust Litig., 690 F.3d 51, 63 (2d Cir. 2012) (further instructing that “if a plaintiff relies on ambiguous evidence to prove its claim, the existence of a conspiracy must be a reasonable inference that the jury could draw from that evidence; it need not be the sole inference”).

Still, plaintiffs’ evidence must make the inference of a conspiracy more “attractive” than the alternative inference of independent action. See Fed. Prescription Serv., Inc. v. Am. Pharm. Ass’n, 663 F.2d 253, 267 (D.C. Cir. 1981) (“Although parallel behavior may support an inference of conspiracy when the alleged co-conspirators have acted in a way inconsistent with independent pursuit of economic self-interest, that inference is warranted only when a theory of rational, independent action is less attractive than that of concerted action.”); see also Blomkest Fertilizer, Inc. v. Potash Corp. of Saskatchewan, 203 F.3d 1028, 1032 (8th Cir. 2000) (en banc) (“[I]f it is as reasonable to infer from the evidence a price-fixing conspiracy as it is to infer permissible activity, then the plaintiff’s claim, without more, fails on summary judgment.”); Kleen Prods. LLC v. Ga.-Pac. LLC, 910 F.3d 927, 934 (7th Cir. 2018) (“The [plaintiffs] need evidence that would allow a trier of fact to nudge the ball over the 50-yard line and rationally to say that the existence of an agreement is more likely than not.”); Anderson News, L.L.C. v. Am. Media, Inc., 899 F.3d 87, 98 (2d Cir. 2018) (“[I]f the evidence is in equipoise, then summary judgment must be granted against the plaintiff.”); In re Text Messaging Antitrust Litig. (“Text Messaging”), 782 F.3d 867, 879 (7th Cir. 2015) (Posner, J.) (“The plaintiffs have presented circumstantial evidence consistent with an inference of collusion, but that evidence is equally

36

consistent with independent parallel behavior. . . . The district court had therefore no alternative to granting summary judgment in favor of the defendants.”).

This requirement does not establish a “special burden” on plaintiffs in antitrust cases.

Eastman Kodak Co. v. Image Tech. Servs., Inc., 504 U.S. 451, 468 (1992). Matsushita simply demands that the nonmoving party’s inferences be reasonable in order to reach a jury, in light of the alternative inferences and economic realities. See id. at 468-69.

The reason why plaintiffs’ evidence must tend to exclude an inference of independent action, or show that an inference of conspiracy from a pattern of behavior could be found to be more likely than not, is that “conscious parallelism,” where firms “in a concentrated market . . . recognize their shared economic interests and their interdependence with respect to price and output decisions,” but act without an agreement, is not unlawful. Bell Atl. v. Twombly, 550 U.S. 544, 553-54 (2007). “[P]arallel conduct or interdependence, without more, . . . [is] just as much in line with a wide swath of rational and competitive business strategy” as with “conspiracy.” Id. at 554. Thus, the Supreme Court has instructed that even “[t]acit collusion, sometimes called oligopolistic price coordination or conscious parallelism . . . by which firms in a concentrated market might in effect share monopoly power, setting their prices at a profit-maximizing, supracompetitive level by recognizing their shared economic interests and their interdependence with respect to price and output decisions,” is a “process, not in itself unlawful.” Brooke Grp. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 227 (1993). Instead, some agreement, whether “tacit or express,” must be evinced to prove a conspiracy. Twombly, 550 U.S. at 553 (quoting Theatre Enters., Inc. v. Paramount Film Distrib. Corp., 346 U.S. 537, 541 (1954)) (explaining that something evincing a meeting of the minds is necessary); see also United States v. Gen. Motors Corp., 284 U.S. 127, 142-43 (1966) (explaining that a showing of an express

37

agreement is not necessary); In re Chocolate Confectionary Antitrust Litig. (“Chocolate”), 801 F.3d 383, 395 (3d Cir. 2015) (“An important corollary to the agreement requirement is that § 1 liability cannot be predicated on a defendant’s unilateral actions, no matter its anticompetitive motivations.”). III. DISCUSSION Plaintiffs have not provided evidence tending to exclude the possibility of independent action or indicating that an inference of an agreement is more plausible than that of unilateral decision-making. Plaintiffs rely on circumstantial evidence yet cannot demonstrate that defendants even acted in a parallel fashion, let alone that their decisions were motivated by an illicit agreement as opposed to self-interested decision-making in an interdependent oligopoly. Defendants provide ample evidence to suggest instead that independent decision-making drove defendants’ choices and that elevated FSCs and expanded FSC application were economically rational. Plaintiffs point to several meetings and communications among defendants, actions taken through their trade association (the AAR), and industry and market conditions conducive to collusion, but even all of these circumstances together do not suggest that defendants conspired rather than acted independently. Just as for carload and coal traffic, plaintiffs likewise cannot point to evidence tending to exclude independent action for intermodal traffic.

Defendants’ motions for summary judgment are therefore granted, as a reasonable jury could not conclude that defendants’ approaches to FSCs in the early to mid-2000s were the result of a conspiracy. The parties’ seven related motions to exclude evidence, opinions, and declarations are denied as moot.

A. Proving a Conspiracy To meet their burden of demonstrating that an inference of conspiracy is more likely than one of independent action, plaintiffs may present either direct or circumstantial evidence.

38

Monsanto, 465 U.S. at 764. “Direct evidence explicitly refers to an understanding between the alleged conspirators, while circumstantial evidence requires additional inferences.” Golden Bridge Tech., Inc. v. Motorola, Inc., 547 F.3d 266, 271 (5th Cir. 2008). The strongest evidence would be a manifest agreement not to compete, such as proof that defendants got together and exchanged assurances of common action. See In re Flat Glass Antitrust Litig. (“Flat Glass”), 385 F.3d 350, 361 (3d Cir. 2004). Alternatively, plaintiffs may use circumstantial evidence to show an inference of independent action is less persuasive than one of conspiracy by demonstrating parallel conduct buttressed by certain “plus factors,” such as a motive to conspire, actions against self-interest, and other evidence implying a traditional conspiracy. Chocolate, 801 F.3d at 398; In re Baby Food Antitrust Litig. (“Baby Food”), 166 F.3d 112, 132-33 (3d Cir. 1999) (“[A] conspiracy based on consciously parallel behavior requires a plaintiff not only to show parallel behavior and awareness in their decision making, but also certain plus factors.”).

Some of those factors, however, may be equally consistent with lawful interdependence and thus not add anything to the parallel conduct in aid of an inference of conspiracy. See Flat Glass, 385 F.3d at 360-61 (noting that motive to conspire and actions against self-interest “largely restate the phenomenon of interdependence” and thus “may not suffice—by themselves—to defeat summary judgment on a claim of horizontal price-fixing among oligopolists”). Accordingly, a factor is only considered a true “plus factor,” something more beyond the parallel conduct, if it “tends to exclude the possibility that the alleged conspirators acted independently.” In re Domestic Airline Travel Antitrust Litig. (“Domestic Airline Travel”), 691 F. Supp. 3d 175, 192 (D.D.C. 2023); City of Moundridge v. Exxon Mobil Corp., No. 4-cv- 940 (RWR), 2009 WL 5385975, at *4 (D.D.C. 2009) (“[P]lus factors” are meant to “indicate the existence of a conspiracy by ruling out the legitimate explanation for parallel behavior” and

39

“suggest[ing] that an alleged conspirator’s actions are ‘inconsistent with independent pursuit of economic self-interest.’” (quoting Fed. Prescription Serv., 663 F.2d at 267)), aff’d 409 F. App’x 362 (D.C. Cir. 2011). “[I]f a benign explanation for the action is equally or more plausible than a collusive explanation, the action cannot constitute a plus factor.” Williamson Oil Co., Inc. v. Philip Morris USA, 346 F.3d 1287, 1310 (11th Cir. 2003).

The “economic theory of interdependence . . . recognizes the differences between competitive markets (markets with many smaller firms) and oligopolistic markets (concentrated markets with only a few firms)” and explains why parallel conduct, without more, is not probative of conspiracy. Chocolate, 801 F.3d at 397. In an oligopolistic market, “a single firm’s change in output or price ‘will have a noticeable impact on the market and on its rivals.’” Id. (quoting Flat Glass, 385 F.3d at 359). “Any rational decision” made, therefore, by an oligopolist, “must take into account the anticipated reaction of the other firms.” Id. (alterations in original) (quoting Flat Glass, 385 F.3d at 359). As the Third Circuit has explained, the “upshot is oligopolists may maintain supracompetitive prices through rational, interdependent decision-making as opposed to unlawful concerted action, if the oligopolists independently conclude that the industry as a whole would be better off by raising prices. Even though this practice of parallel pricing, known as ‘conscious parallelism,’ produces anticompetitive outcomes, it is lawful under the Sherman Act.” Id. Conscious parallelism is lawful because rather than “an agreement,” this is a “necessary fact of life in oligopolies,” id. (second passage quoting Baby Food, 166 F.3d at 122), for which “courts have no effective remedy,” id. at 397-98. So, plaintiffs must provide evidence—characterized as “plus factors” or otherwise—that goes beyond conscious parallelism, which “cannot alone create a reasonable inference of conspiracy.” Id. at 398; see also Brooke Grp., 509 U.S. at 227 (describing the lawful process of conscious

40

parallelism); PHILLIP E. AREEDA & HERBERT HOVENKAMP, ANTITRUST LAW: AN ANALYSIS OF ANTITRUST PRINCIPLES AND THEIR APPLICATION ¶ 1433a (2025) (“The courts are nearly unanimous in saying that mere interdependent parallelism does not establish the contract, combination, or conspiracy required by Sherman Act § 1.”).

In short, “[t]he crucial question is whether each defendant’s conduct stemmed from an independent decision or from an agreement, though such agreement need not be express, and it may be inferred from all of the circumstances.” Domestic Airline Travel, 691 F. Supp. 3d at 194 (quoting Standard Iron Works v. ArcelorMittal, 639 F. Supp. 2d 877, 894-95 (N.D. Ill. 2009)). On this question, defendants insist that “tacit collusion” does not constitute a Sherman Act conspiracy, Defs.’ Reply at 4 (quoting Text Messaging, 782 F.3d at 872); see also Brooke Grp., 509 U.S. at 227, while plaintiffs posit that the conspiracy may be tacit or express, Pls.’ Opp’n at 9 (citing Twombly, 550 U.S. at 553). Both sides are correct, as the parties and their respective preferred cases recognize in focusing on the critical factual issue of the existence of an agreement, a meeting of the minds. Defs.’ Reply at 4; Pls.’ Opp’n at 9; Text Messaging, 782 F.3d at 879 (“Collusion is illegal only when based on agreement. Agreement can be proved by circumstantial evidence.”). Twombly instructs that the agreement may be the result of an exchange of words or other circumstances leading to a meeting of the minds, 550 U.S. at 553, while Text Messaging reminds that conscious parallelism, without a meeting of the minds, does not constitute such an agreement, 782 F.3d at 879.

When evaluating whether plaintiffs have made such a showing, a “court must look to the evidence as a whole and consider any single piece of evidence in the context of other evidence.” Flat Glass, 385 F.3d at 369; Williamson Oil, 346 F.3d at 1301 (“[T]he district court is obligated to give the price fixing plaintiff(s) ‘the full benefit of their proof without tightly

41

compartmentalizing the various factual components and wiping the slate clean” after scrutinizing each. (quoting Continental Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699 (1962))). Though a court should not weigh conflicting evidence, which is “the job of the jury,” if no single item of undisputed evidence presented by plaintiffs points unequivocally to conspiracy, the court must also avoid the cursory conclusion that the evidence as a whole cannot defeat summary judgment. In re High Fructose Corn Syrup, 295 F.3d 651, 655 (7th Cir. 2002). 9 Rather, discrete evidentiary facts must be viewed collectively to ensure the whole picture is fairly assessed.

B. Plaintiffs Have No Direct Evidence of Conspiracy.

Plaintiffs have not pointed to any admission or “smoking gun” kind of evidence here that proves a conspiracy without the necessity of any inferences. Nor do plaintiffs even squarely assert that they have direct evidence. See Pls.’ Opp’n at 14 (stating that plaintiffs’ evidence is sufficient, however characterized); Defs.’ Reply at 6; see also MSJ Hr’g Tr. at 73:21-74:9 (plaintiffs’ counsel stating, “We are proceeding under Matsushita. We are not claiming the Court needs to look at one document . . .”); id. at 21:13-22:2 (defendants’ counsel expressing that plaintiffs have not “responded” to defendants’ contention that plaintiffs have only put forth circumstantial evidence “by saying no, no, no, this is direct evidence of the agreement, everybody wrote down what they were going to charge or everybody got together and said this is

9 Defendants have cited the Ninth Circuit’s two-part framework to explain the test for summary judgment in a section 1 Sherman Act case. Defs.’ Mem. at 16-17 (citing In re Citric Acid Litig., 191 F.3d 1090, 1094 (9th Cir. 1999) (describing the inquiry as (1) whether the defendant can “rebut an allegation of conspiracy by showing a plausible and justifiable reason for its conduct that is consistent with proper business practice” and (2) whether plaintiff can provide “specific evidence tending to show that the defendant was not engaging in permissible competitive behavior” (first passage quoting Richards v. Neilsen Freight Lines, 810 F.2d 898, 902 (9th Cir. 1987))). While that test does not appear to be inconsistent with the framework articulated by the Supreme Court or by other circuits, the Ninth Circuit’s approach has not been universally adopted, including not by the D.C. Circuit, cf. generally Kreuzer v. Am. Acad. of Periodontology, 735 F.2d 1479 (D.C. Cir. 1984), so that two-step analysis is not applied here.

42

what the formula is going to be or something. And they couldn’t have . . .”). Plaintiffs therefore must demonstrate through circumstantial evidence that an inference of conspiracy is more reasonable than that of independent action.

C. Plaintiffs Cannot Demonstrate that Defendants’ Conduct was Parallel as to Carload and Coal Traffic.

When relying on “circumstantial evidence of conscious parallelism” to prove a section 1 claim, a plaintiff must “first demonstrate that the defendants’ actions were parallel,” as all parties here agree. Domestic Airline Travel, 691 F. Supp. 3d at 194 (second passage quoting In re Beef Indus. Antitrust Litig., 907 F.2d 510, 514 (5th Cir. 1990)); Williamson Oil Co., 346 F.3d at 1301 (“First, the court must determine whether the plaintiff has established a pattern of parallel behavior.”); MSJ Hr’g Tr. at 34:6-8 (defendants’ representation that circumstantial evidence likely cannot overcome lack of parallelism); id. at 106:14-19 (plaintiffs’ representation that they are not arguing the law does not require parallel conduct). In part because parallel conduct alone is not sufficient to establish a conspiracy—only parallel conduct in the presence of additional plus factors suffices—caselaw explicating the minimum degree of parallelism required for a conspiracy is not easily distilled, as may be expected with such a fact intensive inquiry. See, e.g., Williamson, 346 F.3d at 1304-05 (only considering the plus factors because “no party contests the satisfaction of the first prong of our inquiry, viz., the existence of parallel pricing behavior”); Tera Grp., Inc. v. Citigroup, Inc., No. 24-135, 2024 WL 4501967, at *3 (2d Cir. Oct. 16, 2024) (affirming a dismissal of a complaint for failure to plead a section 1 Sherman Act claim but “assuming arguendo that [plaintiff] . . . sufficiently pled one or more acts of parallel conduct”). Generally, parallelism is shown by the same or very similar behavior over the same general timeframe. See Anderson News, 899 F.3d at 104.

43

Pricing and strategy changes and other policy and practice adjustments that align can all qualify as parallel behavior. See In re RealPage, Inc., Rental Software Antitrust Litig. (No. II), 709 F. Supp. 3d 478, 505 (M.D. Tenn. 2023) (“Courts accept allegations of parallel pricing and parallel changes in policies or strategies as evidence of parallel conduct.”); see, e.g., Domestic Airline Travel, 691 F. Supp. 3d at 202-03 (discussing defendants’ capacity changes as part of the alleged conspiracy). With respect to pricing, “[d]iscernible parallel conduct requires more than simply indefinite ‘higher’ prices . . . alone.” Domestic Airline Travel, 691 F. Supp. 3d at 194 (quoting Baby Food, 166 F.3d at 129-130); Baby Food, 166 F.3d at 129-130 (“[R]ising prices do not themselves permit an inference of a collusive market dynamic.” (alteration in original) (quoting Brooke Grp., 509 U.S. at 237)). Though “parallel pricing does not require ‘uniform prices,’” the pricing must at least be “within an agreed upon range.” Baby Food, 166 F.3d at 132 (quoting United States v. Socony-Vacuum Oil Co., 310 U.S. 150, 222 (1940)). The pricing changes should also be more similar than the general alignment that is expected among competitors. See In re Pork Antitrust Litig. (“Pork”), -- F. Supp. 3d --, 2025 WL 1224694, at *27 (D. Minn. 2025) (suggesting that converging around certain prices may not be enough to demonstrate parallelism because “given the competitive nature of the pork market, one would expect competitors’ prices to generally move in parallel” but still finding parallelism based on other evidence); see also Baby Food, 166 F.3d at 135 (“Parallel pricefixing must be so unusual that in the absence of an advance agreement, no reasonable firm would have engaged in it.”). 10 Multiple price changes over time that defendants are shown to be uniformly tracking one another further help demonstrate parallelism. See, e.g., In re Titanium Dioxide Antitrust Litig., 959 F.

10 As this cited remark made in the context of discussing “the plus factors” in Baby Food, 166 F.3d at 135, illustrates, when examining the degree of parallelism among defendants, the threshold parallel conduct inquiry may bleed into the plus factors because particularly parallel conduct—of the kind that would not be expected from mere interdependence alone—may also tend to exclude an inference of independent action.

44

Supp. 2d 799, 825 (D. Md. 2013) (noting the “pervasive[ness]” of the “parallel price increases,” i.e., 25 over the course of eight years).

Regarding timing, parallel conduct generally occurs over a “time period suggestive of prearrangement.” Anderson News, 899 F.3d at 104. That time period may differ depending on the context. See, e.g., Park Irmat Drug Corp. v. Express Scripts Holding Co., 911 F.3d 505, 517 (8th Cir. 2018) (“We do not hold that actions taken within six months of each other can never constitute parallel conduct, but only that the terminations here, executed under dissimilar circumstances and separated by six months, did not constitute parallel conduct.”). While simultaneity of price increases is not necessary to showing parallelism, simultaneity is considered a plus factor that leans in favor of inferring a conspiracy. See In re Musical Instruments & Equip. Antitrust Litig., 798 F.3d 1186, 1195-96 (9th Cir. 2015); Christopher R. Leslie, The Factor/Element Distinction in Antitrust Litigation, 64 WM. & MARY L. REV. 585, 625-26 (2023).

Applying these lessons regarding pricing, strategy, or practice changes over time leads inexorably to the conclusion that defendants did not take parallel action during the alleged conspiracy period in changing their FSC programs with respect to either their formulas or coverage policies. Their formulas prior to the start of the alleged conspiracy period were more uniform than their formulas after defendants made changes in 2003, and the timing of defendants’ decisions does not demonstrate coordinated action. Likewise, defendants’ coverage policies, i.e., their approach to how frequently and uniformly FSCs must be imposed in unregulated freight contracts, were adopted at different times and lacked uniformity throughout the alleged conspiracy period.

45

1. Defendants’ FSC Formulas Differed More During the Conspiracy Period than Before.

Plaintiffs argue that defendants engaged in coordinated action with respect to their unregulated carload and coal freight traffic by lowering their trigger prices, escalating FSC percentages at smaller increments on the HDF and WTI indices, resulting in higher FSCs at every interval, and changing the method of determining whether the trigger price was met (e.g., based on the average of the prior month’s prices or other calculation). Pls.’ Opp’n at 36-39, 114. While defendants did make changes to their FSC formulas in 2003 and 2004, they do not evince parallel action. Prior to the alleged conspiracy, three of the four defendants (UP, CSX, and NS) had identical formulas for carload and coal freight traffic. See supra Parts I.B.1-I.B.2, I.B.5. Plaintiffs, indeed, admit that the defendants’ FSC formulas were closely aligned before the alleged conspiracy period. See MSJ Hr’g Tr. at 70:21-71:2 (Court: “Q. . . . I think in the pre- conspiracy period, based on the charts that are in the record, there was a fairly close alignment among the defendants in their FSC formulas. Wouldn’t you agree on that?” to which plaintiffs’ counsel responded, “I would agree . . . I would agree on that.”). After 2003, only CSX and NS had the same formula. See id. Defendants’ changes during the alleged conspiracy period therefore made their formulas more dissimilar. See PX989 ¶ 75, Pls.’ App’x Vol. 19B, ECF No. 1224-2 (sealed) (revised reply report of plaintiffs’ expert, Dr. Allen E. Jacobs, Fig. II.10, demonstrating that average FSC rates pre-conspiracy period were 1.2% for all defendants except BNSF, for which they were 1.6%, and during the alleged-conspiracy period, FSCs were more variable: 13.4% for BNSF and UP carload, 16.9% for BNSF coal, 18.5% for UP coal, and 16.3% for CSX and NS).

After initially announcing that UP would follow CSX, UP went in a different direction entirely, separating from CSX and NS by adopting the HDF index. See supra Parts I.B.1, I.B.5.

46

Although BNSF also had the HDF index, UP adopted a different formula from BNSF’s: a higher $1.35/gal trigger point for application of a fuel surcharge instead of BNSF’s $1.25/gal trigger point—meaning the UP surcharge would kick in when fuel costs were higher than for shippers using BNSF. Compare DA4603, Defs.’ App’x Vol. 12 (May 7, 2003, BNSF Rules Book 6100- A), with DA1573 at 1574-75, Defs.’ App’x Vol. 2, ECF No. 1172-2 (sealed) (May 12, 2003, internal UP email from Christine Stenger-Bleske to UP listservs, with attached UP announcement of modifications to FSC). Leadership at BNSF observed this difference, noting in July 2003 that UP would be “very competitive at lower Highway Diesel Fuel (HDF) prices.” DA4415, Defs.’ App’x Vol. 11, ECF No. 1172-11 (sealed) (July 1, 2003, memorandum from Kyei to Steve Bobb).

Plaintiffs argue that the formulas during the conspiracy period were functionally “equivalent,” resulting in the same charges, even if superficially different in how they produced those outcomes. Pls.’ Opp’n at 91. In other words, despite the differences, plaintiffs essentially argue the formulas were similar enough to reflect parallel conduct. For example, they portray UP’s change as “match[ing]” BNSF’s formula, explaining that the terms were identical over the $1.35 threshold, and they generated the same FSC percentages throughout the conspiracy period. Id. at 26, 97. Plaintiffs cite deposition testimony and communications where BNSF and UP leaders recognized the similarity. See, e.g., PX610 at 119, Pls.’ App’x Vol. 10A (Deposition of John Lanigan, Executive Vice President and Chief Marketing Officer of BNSF (“Lanigan, BNSF, Dep.”) at 119:14-24, noting that BNSF and UP’s programs are identical above $1.35); PX401, Pls.’ App’x Vol. 6B, ECF No. 1220-2 (sealed) (May 9, 2003, internal UP email from Larry Nowakowski to listservs, stating, “UP’s carload surcharge . . . mirrors BNSF’s except that UP does not apply a 1% surcharge when the price is between $1.30 and $1.349/gallon”); PX502,

47

Pls.’ App’x Vol. 7, ECF No. 1201-4 (sealed) (Oct. 20, 2004, internal UP email from Kurt Schroeder to Chuck Adams, recognizing that UP changed its surcharge to “match BNSF’s ‘standard’ HDF based surcharge” in June 2003). Plaintiffs even suggest that BNSF and UP’s formulas were also functionally equivalent to CSX and NS’s, despite the use of different fuel indices, see Pls.’ Opp’n at 37, 107, further citing comments recognizing rough conversions between those fuel indices (WTI and HDF) trigger prices, see PX610 at 97, Pls.’ App’x Vol. 10A (sealed) (Lanigan, BNSF, Dep. at 97:11-21, indicating $1.25 HDF is “roughly equivalent” to $23 WTI); PX84 at 1, Pls.’ App’x Vol. 2, ECF No. 1219-2 (sealed) (undated BNSF email text, explaining that the new trigger price “reflect[s] the $23 WTI crude strike price being used by UP in their new carload fuel surcharge table”); PX474, Pls.’ App’x Vol. 7 (June 3, 2003, handwritten notes from CSX-UP alliance meeting, noting that both roads start at a $23 trigger, despite UP having adopted a trigger on the HDF scale at that point). Plaintiffs discount the differences in the formulas as “distinctions without a difference,” MSJ Hr’g Tr. at 72:5-9, reflecting, perhaps, defendants’ shrewdness in avoiding antitrust scrutiny when forming the conspiracy, id. at 85:19- 25, 86:4-13.

Even assuming that defendants’ pricing formulas were functionally equivalent in terms of the fuel surcharges produced, defendants’ conduct would not be sufficiently parallel. A conspiracy requires not just similar price outcomes among defendants, as plaintiffs insist defendants reached, but parallel price changes—or parallel behavior of some other kind—during the conspiracy period. See Anderson News, 899 F.3d at 104 (using the term “parallel action”). Defendants’ formulas already contained the same elements (i.e., trigger price, use of index, percentage of base price increasing at a given rate, etc.) prior to the alleged conspiracy period and were also aligned on their substantive terms. During the alleged conspiracy period,

48

defendants at most maintained some general alignment while making some changes to specific terms that ultimately made the formulas more dissimilar. That does not signify parallel behavior, especially given that “one would expect competitors’ prices to generally move in parallel” due to the “competitive nature of the [rail freight] market.” Pork, 2025 WL 1224694, at *27. Competitors may try to attract customers by offering a pricing regime structured differently than its peers’ or by offering different discounts, but they will often maintain overall similar prices, driven by common market factors, similar costs, and comparable profit margin goals. In this case, because all the formulas were tied in some way to fuel pricing, they generally followed the price of oil, resulting in a similar trend line. See MSJ Hr’g Tr. at 120:6-19 (defendants’ counsel describing graph shown at hearing that defendants’ surcharges generally followed the oil price trend line). 11 Parallel conduct, however, requires pricing decisions so “unusual,” they could not be expected from an ordinary competitive market. Baby Food, 166 F.3d at 135. That competitors’ prices were generally on par with one another revenue-wise, despite being distinguishable facially, is not unusual.

In any event, despite what plaintiffs perceive as functional equivalency in the formulas, defendants’ pricing mechanisms were fundamentally different. As one exhibit cited by plaintiffs put it, “WTI increases do not directly translate into HDF market increases. When oil prices finally begin to decline, WTI provision will probably fall faster than our HDF.” PX502, Pls.’ App’x Vol. 7 (Oct. 20, 2004, internal UP email from Schroeder to Adams). Further, although the trigger prices for BNSF and UP did not result in significantly different FSCs during the conspiracy period because fuel costs exceeded UP’s higher trigger point, that result was not obvious at the time the parties decided their FSC formulas, as they did not know how fuel prices

11 Defendants’ graph mapping their surcharges against oil prices was not in the record as an exhibit and was not introduced into the record at the hearing.

49

would fluctuate. By adopting different formulas, defendants made different bets on and took different risks regarding how fuel prices would vary.

Plaintiffs also try to represent UP’s change of course in shifting to a different HDF-based formula—after announcing that it would match CSX exactly—as “advanc[ing] the conspiracy” by bringing UP and BNSF, the two Western railroads, in line with one another. Pls.’ Opp’n at 92. Yet, UP did not “match” BNSF at all—UP’s formula had a different and higher trigger price than BNSF, giving UP customers more of a cushion to avoid paying a fuel surcharge in a fluctuating fuel price market. UP’s decision thus made defendants’ FSC formulas more distinct from one another and does not reflect an effort to make its formula the “same” as others’ “in the eyes of customers,” as plaintiffs contend. Id. (second passage quoting PX241 at ‘361, Pls.’ App’x Vol. 4, ECF No. 1200-4 (sealed) (May 14, 2003, NS/UP Alliance Meeting Notes)). 12 Additionally, plaintiffs’ argument that UP and BNSF’s formulas are fundamentally the same deteriorates when this same logic is applied more broadly. If UP and BNSF’s formulas should be considered equivalent with a $0.10 difference in trigger price, BNSF cannot be considered to have made any meaningful change to its FSC during the alleged conspiracy period at all—BNSF only changed its trigger price by $0.05 and adjusted its starting surcharge by 0.5%. See supra Parts I.B.1, I.B.5; Defs.’ Reply at 35-36; contra Pls.’ Opp’n at 25 (insisting that BNSF

12 Plaintiffs explicitly cite a comment in PX241 from the NS-UP Alliance Meeting minutes that “it would be a positive outcome if all roads had the same process [for FSC application] in the eyes of our customers,” id. at ‘361, three times in their brief and seven times in their Counterstatement of Facts. See Pls.’ Opp’n at 26, 92, 101; Pls.’ COF at 71, 111, 118, 190, 208, 243-44. Plaintiffs, contending the exhibit serves as practically direct evidence of conspiracy, see MSJ Hr’g Tr. at 74:4-18, overuse yet under contextualize that statement. The minutes refer to a general discussion between NS and UP about “fuel surcharges and various application processes as used by different roads,” and general “consensus” “that it would be” “positive” “if all roads had the same process in the eyes of customers.” PX241 at ‘361 (emphasis added). Not only does that comment not reflect a desire harbored by defendants for uniform prices, as plaintiffs suggest—instead referring to customers’ desire for more uniformity in, and concomitant ease in shippers’ understanding of, the logistics for imposition of FSCs—but it also does not refer to UP’s recent FSC change and its relation to BNSF’s formula at all. Plaintiffs’ persistent effort to distort this vague language to their liking is unavailing.

50

made a meaningful change to have its FSC “more closely align[]” with its competition (quoting PX914 at ‘387, Pls.’ App’x Vol. 15, ECF No. 1204-1 (sealed) (May 1, 2003, BNSF internal draft of Fuel Surcharge Script)). In fact, this change did not make any difference to FSCs charged at all during the alleged conspiracy period, given the then-prevailing fuel prices. See BNSF Mem. in Supp. of Mot. for Summ. J. (“BNSF Mem.”) at 7, ECF No. 1158-1 (sealed) (“In practical terms, therefore, BNSF’s only change to its published carload rate-based FSCs during the supposed conspiracy was to add a 0.5% surcharge between $1.25 and $1.30 HDF. And because HDF never dipped below $1.30 through 2008, BNSF’s old and new programs proved to have identical effect.”). Plaintiffs’ effort to argue, on the one hand, that small differences are inconsequential when comparing UP and BNSF surcharge formulas and, on the other, that small differences are significant when examining BNSF’s pricing alone, is self-defeating and ultimately unpersuasive in showing parallelism.

BNSF and UP further diverged from other defendants during the alleged conspiracy period by applying separate formulas to coal traffic, while NS and CSX continued applying their carload formulas. See supra Parts I.B.2, I.B.5; DA15709 at 15709-10, Defs.’ App’x Vol. 26, ECF No. 1172-26 (sealed) (June 8, 2004, BNSF Pricing Update, BNSF announcing coal-specific formula); DA1624 at 1624-25, Defs.’ App’x Vol. 3, ECF No. 1172-3 (sealed) (Oct. 1, 2004, Letter from Lance Fritz to shippers, UP announcing coal-specific formula). BNSF and UP’s coal formulas also did not align with one another. BNSF first announced its coal FSC and then UP conducted its own analysis before also deciding to adopt one. See infra Part III.D.2. UP selected a different formula, and the two formulas never generated the same FSC percentage during the alleged conspiracy period. See supra Parts I.B.2, I.B.5; DA24600 at 24601, Defs.’ App’x Vol. 38, ECF No. 1172-38 (sealed) (Report of defendants’ expert, Kevin Murphy, at Exh. 52, graph

51

demonstrating percentage difference between the companies’ coal FSC rates); Pls.’ Opp’n at 117 (asserting that the “average difference between the carriers’ monthly FSCs . . . was just over two percentage points”). NS considered but never adopted a coal-specific formula during the conspiracy period. See DA18074 at 18074-77, Defs.’ App’x Vol. 29, ECF No. 1172-29 (sealed) (July 14, 2004, NS internal memo on FSC Strategies and Options, considering changes to carload/coal FSCs); DA18072, id. (July 16, 2004, email from Pat Glennon, NS Dir. of Marketing Systems, to Don Seale, NS EVP and Chief Marketing Officer, et al., discussing potential changes and recommendation to keep coal FSC the same).

BNSF departed even further from UP and the other defendants when it began applying a mileage-based formula to its coal traffic in March 2005. See supra Parts I.B.2, I.B.5; DA3099 at 3146, Defs.’ App’x Vol. 5, ECF No. 1172-5 (sealed) (Lanigan, BNSF, Dep. at 241:2-11, testifying that the announcement for a mileage-based FSC was made March 29, 2005). In fact, BNSF had already been considering this change to a mileage-based system by the time UP announced its rate-based coal-specific formula, undercutting any inference that the two firms were coordinating on coal-specific FSCs. See DA4481, Defs.’ App’x Vol. 12 (Aug. 30, 2004, internal BNSF email from Kyei to George Pacochoa et al., indicating that Lanigan requested a new review of a mileage-based surcharge). UP did not follow BNSF and held off on adopting a mileage-based formula until after the STB decision in 2007. See DA19086 at 19093, Defs.’ App’x Vol. 30 (Mar. 21, 2007, press release).

As defendants point out, NS also shifted away from the other defendants by moving away from FSCs in 2006, announcing a new formula that incorporated more of the fuel cost into its base rate and reduced reliance on FSCs. See NS Mem. in Supp. of Mot. for Summ. J. (“NS Mem.”) at 1, 12, ECF No. 1155-1 (sealed); DA5191, Defs.’ App’x Vol. 14, ECF No. 1172-14

52

(sealed) (Apr. 24, 2006, NS press release, describing this change with respect to regulated traffic); DA18281, Defs.’ App’x Vol. 29 (May 8, 2006, internal NS email from Glennon to Michael Lambert, describing how the initial focus of the change will be on public rates but it will subsequently be applied to private contracts); PX260, Pls.’ App’x Vol. 4 (May 30, 2006, internal NS email from Dedra Smith to MKT Industrial Prods. listserv, describing how it should be implemented in private contracts). Specifically, in mid-2006, NS increased its base rate by 16.4% and adopted a $64 surcharge based on the WTI index with a 0.3% increment for every $1 price increase. PX260, Pls.’ App’x Vol. 4. Plaintiffs discount this change as merely a perpetuation of the conspiracy because NS locked into its base pricing its over-recovery on fuel cost, protecting its future profits. Pls.’ Opp’n at 56-57; see also id. at 147-48. Plaintiffs miss the point. Regardless of whether changes to shift more of the fuel cost into the base rate could be indicative of some past conspiracy, as discussed infra Part III.E.9, the fact that such a change occurred during the apparent conspiracy demonstrates, at a minimum, a lack of parallelism among defendants and independent decision-making on the part of NS, undermining the very existence of said conspiracy.

Further, plaintiffs discount the differences in formulas—such as BNSF and UP adopting separate formulas for coal traffic but CSX and NS sticking to their carload formulas for coal traffic—contending such differences should be disregarded as logical. For instance, BNSF and UP shipped more coal and shipped longer distances, thus making coal fuel costs more significant for their business. See Pls.’ Opp’n at 118-19. Ironically, plaintiffs are arguing defendants’ own point: Defendants took different approaches to their FSC formulas based on their unique business needs and considerations. See infra Part III.D.1. They did not pursue economically irrational choices simply to further a conspiracy.

53

The evidence on defendants’ FSCs and the FSC formulas’ increasing dissimilarity, rather than convergence, during the conspiracy period is virtually indisputable.

2. The Timing of Defendants’ Changes to FSC Formulas Undercuts Any Parallelism.

Time lags in pricing decisions undermine a finding of parallel behavior and an inference of conspiracy. See Baby Food, 166 F.3d at 131-32. Where time elapses between defendants’ decisions, they have had a “good deal of time to react to competitors’ moves and consider their own,” suggesting they were acting unilaterally. See In re Text Messaging Antitrust Litig., 46 F. Supp. 3d 788, 807-08 (N.D. Ill. 2014) (“[T]he time elapsed between the [price] increases,” seven months, three months, and two months, among other evidence, was “as equally suggestive of independent action as collusion,” if not “more suggestive of independent action.”), aff’d, 782 F.3d 867 (7th Cir. 2015). The Third Circuit in Baby Food found three-to-seven-month gaps between defendants’ implementation of prices to undercut an inference of conspiracy. See 166 F.3d at 131-132. Similarly, the Seventh Circuit in Kleen Products held plaintiffs lacked evidence of a conspiracy where one defendant did not follow until over a month later. 910 F.3d at 936.

Here, the timing of defendants’ changes does not allow for a finding of parallel conduct.

Defendants did not rapidly make changes at once or even make them in an efficient fashion. They adopted their alleged conspiratorial FSCs over the course of nine months, see supra Parts I.B.1-I.B.2, I.B.5, with NS conducting three internal evaluations before following CSX’s formula three quarters of a year later, see infra Part III.D.2(d). During this time, NS acted competitively and incompatibly with a conspiracy, intending to gain market share from CSX by offering lower FSCs. See DA5574, Defs.’ App’x Vol. 15, ECF No. 1172-15 (sealed) (Apr. 29, 2003, internal NS email from Glennon to Seale, conveying, “The CSXT formula faces continued opposition

54

. . . . NS can use this as leverage in negotiating larger rate increases into base rates, or additional traffic.” (emphasis added)). Even CSX, UP, and BNSF, which all made changes to their FSCs in spring of 2003, made them months apart: CSX in March, UP in April but then changing course in May, and BNSF in May. See Kleen Prods., 910 F.3d at 936 (noting no evidence of conspiracy where one defendant followed another over a month later and noting that “[e]ven the attempts that saw quick turnaround times do little to raise suspicions. If it is in a company’s self-interest to imitate a price leader’s increase, why wait to enjoy the benefit?”). BNSF and UP also adopted their coal-specific FSCs four months apart. See supra Parts I.B.2, I.B.5. So even if those FSCs were sufficiently similar to be deemed parallel, their adoption was far from it.

Plaintiffs rightly point out that “simultaneous action is [] not a requirement” to the finding of a cartel. Pls.’ Opp’n at 91 (alteration in original) (quoting In re Broiler Chicken, 290 F. Supp. 3d 772, 791 (N.D. Ill. 2017)) (also citing other out-of-circuit district court cases and a single Fourth Circuit case, SD3, LLC v. Black & Decker (U.S.) Inc., 801 F.3d 412, 429 (4th Cir. 2015)). They emphasize that a few months’ gap between announcements in pricing changes does not preclude a conspiracy. Indeed, no hard cutoff exists. Nonetheless, gaps in time between defendants’ decisions do undermine any pattern of parallel behavior, and longer gaps do so more severely.

Plaintiffs also contend that NS agreed early on to follow CSX but simply faced technical difficulties that, despite efforts to overcome them, alone resulted in the delay. Pls.’ Opp’n at 100-05 (citing PX320 at ‘632, Pls.’ App’x Vol. 5, ECF No. 1201-1 (sealed); PX366 at ‘216, Pls.’ App’x Vol. 6A, ECF No. 1220-1 (sealed)). They insist that although NS could not bring itself into alignment with CSX immediately, NS “spent the better part of the year laying the groundwork to do so as soon as possible.” Id. at 103 (citing PX364 at ‘119, Pls.’ App’x Vol. 6A;

55

PX347 at ‘654-56, Pls.’ App’x Vol. 5; PX270 at ‘523, Pls.’ App’x Vol. 4; and PX325 at ‘466, Pls.’ App’x Vol. 5). Plaintiffs misrepresent the record. The emails they cite for this contention clearly indicate that NS did not in fact decide early on to adopt CSX’s formula, nor were technical concerns dispositive in any such choice. Moreover, NS never resolved the technical issues it did have, demonstrating that such issues were inconsequential. NS simply used the same phased-in approach when it implemented the formula in early 2004 that NS would have used had it adopted the formula in March 2003, without technical systems changes.

Examining plaintiffs’ contentions more closely, the first email plaintiffs cite, sent by NS’s Director of Marketing Systems, Pat Glennon, on March 20, 2003, lists three pros and five cons for adopting CSX’s formula—none of which relate to technical considerations. See PX320 at ‘632, Pls.’ App’x Vol. 5. Glennon only mentions technical considerations as an aside at the bottom of the email. See id. In the second email cited by plaintiffs, sent just a few days later, Chief Marketing Officer (“CMO”) Don Seale lets Glennon know that he asked someone to look into whether “accounting/billing for the CSX format would be doable,” but he does not give any indication that any such challenges would be insurmountable. See PX366 at ‘216, Pls.’ App’x Vol. 6A.

None of the documents provide any indication that NS actually made a decision—or was likely to make a decision—to follow CSX until November 2003 and cannot be considered to have “la[id] the groundwork,” Pls.’ Opp’n at 103, to align with CSX until then. Plaintiffs cite an email from Glennon reviewing feedback from the Group Vice Presidents and concluding that the consensus was to delay the switch until 2004 while still laying the groundwork. PX364 at ‘119, Pls.’ App’x Vol. 6A. Plaintiffs overlook the important context, however, that Glennon’s email was sent in November and that Glennon’s earlier email (in the same exhibit), asking for feedback

56

on the best course of action, indicates clearly that NS had made no decision yet about switching to CSX’s formula but merely held it open as one possibility. See id. at ’121 (Nov. 4, 2003, internal NS email from Glennon to Gary Wendorf et al., expressing, “Should NS adopt the CSXT FSC? Please provide your feedback by Friday (11/7) on the options listed below,” including “Retain the current NS standard” and “Adopt UP or BNSF standard”). Another email cited by plaintiffs from Glennon, telling Seale that the review teams endorsed the option of “proceed[ing] with the program to remove the full text FSC note from private NS authorities and position ourselves for a conversion”—in other words, setting NS up to switch to the CSX formula—likewise was not sent until November 2003 and was merely a recommendation, not a final decision. PX325 at ‘466, Pls.’ App’x Vol. 5. Finally, plaintiffs cite PX347 at ‘654-66, Pls.’ App’x Vol. 5, an email conveying the actual announcement of NS’s FSC change in January 2004, which plaintiffs describe as “anticipated in April 2003” but that exhibit nowhere suggests that NS made a decision then, see id. In fact, in April 2003, an NS email following a note from Glennon states explicitly that NS “seem[ed] to be headed in the direction of not making a change to the FSC provisions like the new CSX provisions, at least not right away.” PX270 at ‘522, Pls.’ App’x Vol. 4 (Apr. 30, 2003, internal NS email from unknown to Tom Rappold et al.); id. (Apr. 29, 2003, email from Glennon to Seale, sharing that “[t]he consensus recommendation is that we do not adopt the CSXT formula at this time”).

Moreover, plaintiffs’ assumption that NS was delayed only by a technical problem is not supported by the evidence. NS had contracts that either hard-coded the FSC or that referenced the conditions of carriage, in a separate terms and conditions document. NS Mem. at 9. NS’s initial FSC was put in the conditions of carriage. See DA18019 at 18026, Defs.’ App’x Vol. 29 (June 1, 2000, NS Conditions of Carriage # 1-B, Rule 255- Fuel Surcharge). The perceived

57

technical difficulty was that NS would have to update individually the contracts with the hard- coded FSCs, requiring a phased-in approach to an FSC change. See NS Mem. at 9-10. If NS had wanted to change its FSC sooner, however, NS could have changed its conditions of carriage so at least the contracts importing those conditions would have the new formula. NS did not do so, despite making other changes to its conditions in the summer of 2003. See id.; DA18317 at 18325, Defs.’ App’x Vol. 29 (June 1, 2003, NS Conditions of Carriage #1-E Rule 255); DA18336 at 18341-42, id. (Aug. 4, 2003, NS Conditions of Carriage #2-D: Coal, Coke and Iron Ore, Condition 55.2).

In fact, when NS did change its FSC, NS did so without making any changes to its approach of hard-coding formulas—demonstrating that this structure was not a barrier after all. NS embraced the phased-in approach: NS announced the new rate, updated carrier conditions, updated public rate formulas, and then individually changed the hard-coded contracts. NS Mem. at 10-11; DA18419 at 18423, Defs.’ App’x Vol. 29 (Deposition of Pat Glennon, NS (“Glennon, NS, Dep.”) at 102:9-23, testifying that “we applied the fuel surcharge on public authorities right away and then we applied them on private authorities as they came up for renegotiation. That would have been the approach we would have taken had the condition of carriage issue never existed”); Defs.’ Reply at 32 (“NS did not overcome th[e] technical obstacles.” (emphasis omitted)). NS did not even allow contract formulas to change automatically for those that imported the conditions of carriage, which would have been the most expeditious approach. It waited until renegotiations. See NS Br. at 10; DA5576 at 5576-77, Defs.’ App’x Vol. 15 (Jan. 8, 2004, internal NS memo from Glennon to Seale et al., conveying, “In connection with private authorities (contracts and private quotes), NS will opt to delay application of the new FSC until the current private publications expire or are renegotiated”); DA18284, Defs.’ App’x Vol. 29

58

(Jan. 29, 2004, email from NS Marketing explaining that approach to customers); DA18286, id. (Mar. 12, 2004, NS letter to customers expressing the same). NS simply did not wish to follow CSX until its existing formula proved not to be winning greater market share, nine months later. See infra Part III.D.1(d).

In sum, even if plaintiffs’ contention were accepted that defendants’ formulas were functionally equivalent in terms of resulting in similar charges and that could be suggestive of collusive action, the fact that none of the changes occurred in sharp succession, but rather unfolded over an extensive period of time demonstrates that defendants’ behavior was not parallel.

3. Defendants’ Approaches to Increasing Coverage Do Not Show Parallel or Concerted Action.

Plaintiffs, perhaps aware of the lack of uniformity in defendants’ FSC formulas, also devote significant attention to defendants’ other changes in their approach to their FSC programs—particularly their efforts to make FSCs “non-negotiable and imposed across-the- board, with the goal of 100% FSC coverage.” Pls.’ Opp’n at 36. Plaintiffs argue that defendants engaged in parallel conduct to expand the application of their FSCs, for instance by requiring employees to add FSCs to all contracts and requiring them to obtain approval by high-level executives for any exceptions. See Pls.’ Opp’n at 39-45. According to plaintiffs, all four defendants adopted the same internal policies to that effect during the conspiracy period, which achieved the desired goal of massive increases in FSC coverage and exorbitant profits. See id. at 42, 46, 48. Whereas before the conspiracy period, defendants did not always impose FSCs and were amenable to negotiation, plaintiffs contend that defendants were far more stringent during the alleged conspiracy period. See id. at 43 (arguing that defendants conveyed to plaintiffs that their FSCs were nonnegotiable).

59

As an initial matter, defendants argue that plaintiffs’ theory of an independent “coverage”

conspiracy must fail. Defs.’ Reply at 39, 74. If, as the above discussion illustrates, plaintiffs did not conspire to coordinate FSCs, then “what did they agree to seek to impose widely?” Id. at 74. Defendants are correct that without an underlying conspiracy over the FSCs themselves, defendants’ changes in internal policies—designed to maximize profit by applying their own unique pricing mechanisms more broadly—cannot constitute a conspiracy. At most, plaintiffs’ arguments about coverage serve as evidence, even if unpersuasive, that defendants were internally enforcing their commitment to impose the supposedly coordinated FSCs, which were not in fact parallel at all.

Even if defendants’ coverage policies themselves could constitute a conspiracy in restraint of trade under the Sherman Act or could constitute one in combination with somewhat parallel pricing decisions, defendants’ policies during the conspiracy period to expand application of FSCs do not illustrate parallel conduct suggestive of a conspiracy. By the end of the conspiracy period, all four defendants maintained a policy of including FSCs in every contract and requiring high-level approval for exceptions, but their adoption of such policies occurred over the course of years, with some defendants adopting such policies even before the alleged conspiracy period.

BNSF and UP had broad, universal coverage policies prior to 2003 and made efforts to further implement those policies to expand their coverage during the alleged conspiracy period. BNSF stated in 2002 that all contracts should include an FSC and began tracking its progress toward 100% coverage in 2001. See DA4019 at 4019-20, Defs.’ App’x Vol. 9 (Feb. 15, 2002, internal BNSF email from Chuck Schultz to Greg Patient describing recommendations from Kyei, approved by “planners,” and instructing that “every contract should include a fuel

60

surcharge clause. . . . Any deviation from the attached policy should require the approval of the Business Unit VP”); DA4083, Defs.’ App’x Vol. 10, ECF No. 1172-10 (sealed) (Nov. 4, 2002, internal BNSF email from George Gillen to Sara Thornsedt, stating, “As you are probably aware, we are making a concerted effort to increase the application of fuel surcharges to Carload documents. Effective Oct 1 we started to apply the fuel surcharge to 293 rate documents that were formerly exempt”); DA3868 at 3874, Defs.’ App’x Vol. 9 (Apr. 9, 2001, Review of BNSF Fuel Surcharges presentation, showing tracking of percentage of business covered by FSCs); DA4104 at 4105-07, Defs.’ App’x Vol. 10 (Jan. 21, 2003, Q4 2002 BNSF Earnings Conference Call, with public reporting of FSC coverage and describing goal of getting more aggressive to increase coverage). 13 UP, as of late 2002, also adopted a policy of universal coverage and tried to implement FSCs for all traffic (unless prevented by preexisting contracts) prior to the alleged conspiracy period. See DA1362 at 1366, Defs.’ App’x Vol. 2 (Oct. 24, 2002, internal UP email from Greg Sovey to Jack Koraleski, UP CMO et al., attaching procedures stating, “Effective immediately, all new pricing documents and amendments created must now contain the Fuel Surcharge language”); id. at 1363 (instructing inclusion of FSC language in “the appropriate rules circulars and/or tariffs” and “private price documents”). Plaintiffs suggest otherwise, arguing that UP changed course in 2003, but they focus on the earlier part of the 2000 to 2002 period, not the time in late 2002 when UP unilaterally decided to focus on increasing coverage. See Pls.’ Opp’n at 16 (citing PX611 at 23-24, Pls.’ App’x Vol. 10A (Deposition of Robert Knight, UP (“Knight,

13 BNSF’s exhibits reveal further indications that they focused on FSC expansion prior to the alleged start of the conspiracy, even after the initial announcements of such coverage policies in 2002. See DA7107, Defs.’ App’x Vol. 17, ECF No. 1172-17 (sealed) (Feb. 13, 2003, internal BNSF email from Kathleen Regan to Todd Olsen stating, “Focus on Fuel Cost Recovery will be immense in the coming months”); DA3099 at 3106-07, Defs.’ App’x Vol. 5 (Lanigan, BNSF, Dep. at 26:11-29:5, describing how he made it a goal, upon assuming the role of CMO, by early 2003, to expand FSC coverage).

61

UP, Dep.”) at 23:14-24:9, describing the lack of an FSC policy but speaking only generally to the 2000-2002 period); COF ¶ 65 (citing documents from 2000-2001 and the earlier part of 2002, predating the change in policy)). Plaintiffs’ single example where a shipper negotiated the FSC out of its contract in December 2002 but was not able to do so a year later is too thin a reed to find that UP made a significant policy change in Spring 2003. See Pls.’ Opp’n at 42-43; COF ¶ 66; PX600 at 38, Pls.’ App’x Vol. 10A (Deposition of Howard Kaplan, shipper, at 38:11-17); PX546 at ‘467, Pls.’ App’x Vol. 8, ECF No. 1201-5 (sealed) (Dec. 19, 2002, emails between Kaplan and UP); PX547, id. (Dec. 22, 2003, emails between Kaplan and UP, noting “all contracts without fuel language will have fuel language upon renewal” per “a mandate by UP management”). Instead of what plaintiffs perceive as a significant policy shift during the alleged conspiracy period, the fact that exceptions to FSC coverage became rarer appears to be due both to timing, as long-term contracts were renegotiated, and more rigorous enforcement of the policies already in place.

While both UP and BNSF articulated and doubled down on these FSC coverage policies during the alleged conspiracy period by adopting requirements for VP approval of exceptions to applying FSCs, such expressions were not significant changes in course, given the preexisting policies, and they did not occur in parallel, or in close succession to one another or any other defendant, at the start of the alleged conspiracy. BNSF adopted the requirement that all exceptions be approved by VPs in April 2003. See PX50, Pls.’ App’x Vol. 1, ECF No. 1219-1 (sealed) (Apr. 9, 2003, internal BNSF email from Denis Smith to James Obermiller, stating, “Effective immediately and urgently per John Lanigan. Authority to omit FSC provision is to be granted to VP’s only (who will also clear with John). This needs to be communicated immediately as well”); DA4340, Defs.’ App’x Vol. 11 (same); PX122 at ‘919, Pls.’ App’x Vol.

62

2 (Apr. 24, 2003, internal BNSF email from Gillen to William Brown et al., describing the change in approval level required). BNSF teams had originally approved a similar policy in February 2002, when evaluating how to achieve the goal of greater coverage. See DA4019, Defs.’ App’x Vol. 9 (Feb. 15, 2002, internal BNSF email from Patient to Schultz, “Any deviation from the attached policy should require the approval of the Business Unit VP”); DA4048 at 4066, 4071, 4080, Defs.’ App’x Vol. 10 (Nov. 1, 2002, presentation on contract audit action items, emphasizing that FSCs need to be applied in more contracts and including as an action item the creation of an approval process for not including FSCs); Defs.’ Reply at 43. UP did not do the same until months later, in June 2003—again, merely facilitating its pre- conspiracy period policy of universal coverage. See Pls.’ Opp’n at 40-41; PX1111 at ‘922, Pls.’ App’x Vol. 23, ECF No. 1206-1 (sealed) (June 19, 2003, internal UP email from Todd Whitman to Bob Toy, forwarding messaging stating, “Friendly reminder—surcharge exceptions now require VP approval”); see also PX414 at ‘597, Pls.’ App’x Vol. 7 (Jan. 5, 2004, email from Toy to Schroeder, stating, “The Highway Diesel Fuel surcharge policy is required for any new and renewed contracts . . . . ANY deviation from that policy requires advance sign-off by Bob Toy”).

The other two defendants did not adopt any kind of similar policies until years after both UP and BNSF’s original pre-conspiracy-period policies and their reinforcement in the Spring of 2003. NS did not adopt any broad coverage policies or requirements for high-level approval for exceptions to exclude FSCs until at least early 2004, coinciding with the change to its FSC formula. See PX326 at ‘841, Pls.’ App’x Vol. 5 (Jan. 8, 2004, email from Glennon to Seale et al., stating, “Any deviations from the new standard must have GVP approval and must be reviewed by Pricing Services for applicability”); PX272 at ‘412, Pls.’ App’x Vol. 4 (June 11,

63

2004, email from Tom Bayrer to Danny Smith, stating, “All deals must have the new FSC”); PX293 at ‘040, Pls.’ App’x Vol. 5 (Mar. 8, 2004, email from Joe Osborne to Seale, stating, “I assume that in all cases there is to be no deviation from NS’s published FSC without your approval[,]” with response, “Correct”). 14 CSX did not adopt similar coverage policies until at least late 2005. To the contrary, CSX continued to allow the inclusion of the RCAF price escalator index as an alternative to FSCs until then. See DA2802 at 2807, Defs.’ App’x Vol. 5 (June 2003, CSX presentation on Contract Escalators, describing as the “current policy” that “[a]ll price vehicles must contain fuel surcharge language or RCAF(u)”); DA2295 at 2300, Defs.’ App’x Vol. 4 (Oct. 31, 2003, internal CSX email from Frank Onimus to Sales & Marketing team, sharing an attachment entitled “Policy: Fuel Surcharge Application—Contracts, Quotes, and Price Lists” and dated Apr. 25, 2003, reading “All private prices are subject to either Fuel Surcharge or 100% of the RCAF(u)”); DA6925 at 6930, Defs.’ App’x Vol. 17, ECF No. 1172-17 (sealed) (June 1, 2005, CSXT Industrial & Agricultural Products Midyear Review Presentation, stating that RCAF(u) remained “acceptable as an alternative” to fuel surcharges); DA2672, Defs.’ App’x Vol. 4 (Dec. 21, 2005, internal CSX email from Clyde McIntyre to Gooden et al., announcing switch to using AIILF

14 Both sides cite ample evidence about NS’s pre-conspiracy period approach to coverage, and this evidence, on balance, indicates that while NS wanted to implement FSCs broadly, the company was also amenable to exceptions and discounts. Compare, e.g., DA5612, Defs.’ App’x Vol. 16 (Apr. 21, 2000, internal NS email from Bob Scharpf to Thomas Brugman stating, “Product Managers should continue to add the fuel surcharge note to all new/reissued contracts and private quotes (NSSQs) and all new open quotes and tariffs”); DA5613, id. (May 11, 2000, internal NS email from Bob Scharpf to Thomas Brugman, “[The FSC] should rarely be negotiable”); DA6050, id. (Sep. 27, 2000, internal NS email from D.V. Corcoran to Earl Pinkston et al., stating, “In our conference call this morning . . ., we were given our charge to work with our customers and secure fuel surcharge increases all [sic] ALL business”), with DA18017, Defs.’ App’x Vol. 29 (May 30, 2000, email from Brugman to Seale et al., apparently running an FSC exception by Seale); DA8351 at 8352, Defs.’ App’x Vol. 20, ECF No. 1172-20 (sealed) (June 22, 2000, email from Brugman, NS, to Dale Cox of Cargill, stating, “With Don Seale’s approval we have offered to waive the surcharge language in exchange for an additional 1% increase across the board”); PX276 at ‘420, Pls.’ App’x Vol. 4 (Apr. 14, 2000, email from Pat Lewis to Brugman instructing to take out the FSC if it is a “deal buster” but emphasizing that “[t]he fuel surcharge should always be a part of the contract negotiation process”). NS’s approach in 2004 and after more clearly demonstrated firmer efforts to impose FSCs broadly.

64

index and FSC instead of RCAF(u)). CSX regularly negotiated discounts on its FSCs or offered RCAF instead of surcharges altogether. See, e.g., CSX Mem. in Supp. of Summ. J. (“CSX Mem.”) at 13-14 & n.15, ECF No. 1160-1 (sealed) (providing examples of CSX negotiating an RCAF arrangement instead of a surcharge in 2003, 2004, and 2005); id. at App’x A (asserting that 33 of MDL I and MDL II plaintiffs’ contracts with CSX had RCAF clauses instead of FSCs); DA18068, Defs.’ App’x Vol. 29 (Jan. 15, 2004, emails between NS and shipper, demonstrating that NS gave a discount to a shipper to match CSX’s offer, which was inclusion of RCAF(u) escalator without an FSC); DA2898, Defs.’ App’x Vol. 5 (July 28, 2006, internal CSX email from Wayne Flynn to Kyle Hancock, discussing discounting FSC in a contract with a shipper in order to compete with NS, even after the change in NS’s FSC application policy). CSX likewise did not require senior management approval for exceptions to FSCs until the end of 2005. See DA16514, Defs.’ App’x Vol. 27, ECF No. 1172-27 (sealed) (Dec. 12, 2005, internal CSX email from Dean Piacente to Rich Karn et al., stating, “Any exceptions such as Rcafu must be renewed with Clarence [Gooden]”); PX193, Pls.’ App’x Vol. 4 (Feb. 12, 2006, internal CSX email from Hancock to Kathy Halper et al., stating, “NO ONE is authorized to approve a deal WITHOUT Fuel Surcharge” and articulating process for seeking exceptions).

Plaintiffs argue that CSX adopted a requirement that VPs review any exceptions from FSC application in May 2004, Pls.’ Opp’n at 42, but their citations are inapposite. One of their exhibits cited in support of that CSX timing, PX193, Pls.’ App’x Vol. 4 (Feb. 12, 2006, email from Hancock to Halper et al.), is from 2006. Two others are from 2003 but do not reflect adoption of this policy. One is merely a recommendation for a change in their program. PX181 at ‘493, Pls.’ App’x Vol. 4 (Mar. 11, 2003, internal CSX email from John Couch to Onimus, with attached PowerPoint reflecting that they “[n]eed a written policy that Fuel Surcharge will

65

apply on all new moves – even those with RCAF(U) [with] exceptions to SVP”). The other suggests that FSCs “must apply to all rate structures” as of November 2003, but emails in the same chain confirm that use of the escalator, RCAF(u), at 100% was still an alternative to the surcharge. PX1159 at ‘623-24, Pls.’ App’x Vol. 24, ECF No. 1225-1 (sealed) (Oct. 31, 2003, email from Onimus to Gooden et al.). The one exhibit from 2004 is similarly misleading: A presentation from May of that year merely offers for an internal discussion a change to the policy such that “[e]verything is subject to fuel surcharge. Long term contracts can have RCAF(u), AIILF, or fixed increases of at least 5%/year – plus fuel surcharge. Exceptions elevated.” DA25960 at 25966, Defs.’ App’x Vol. 41, ECF No. 1245-1 (sealed) (May 17, 2004, email from Couch to Onimus, with attached presentation on Escalators 2004 Discussion); PX170 at ‘574 (same). That same presentation confirms that the “current Fuel surcharge policy” is that “[t]he fuel cost recovery charge does not apply on contracts that have 100% of the RCAF(u) or CPI or other index.” DA25960 at 25963, Defs.’ App’x Vol. 41. Later discussion also confirms that no special exception was needed to omit a surcharge and use RCAF instead in 2004, consistent with CSX’s longstanding position in the market. See DA7967 at 7968, Defs.’ App’x Vol. 19, ECF No. 1172-19 (sealed) (May 19, 2004, email from Dwight Price to Bill Hofmann et al.).

Plaintiffs also try to demonstrate uniformity in defendants’ policies for FSC coverage by pointing to the significant expansion in coverage that occurred during the alleged conspiracy period, see Pls.’ Opp’n at 17-19, 48-49, 111, 115-16, 126, suggesting that such notable growth must indicate a conspiratorial scheme. Defendants were unable to impose FSCs effectively with any regularity prior to conspiring, plaintiffs argue. See id. Then, FSC coverage exploded, with all four defendants covering the vast majority of their traffic with FSCs by 2007 (between 60 and

66

90% by revenue). See DA21943 at 22036, Defs.’ App’x Vol. 34 (Bernheim & Marx, Pls.’ Expert, Report at Fig. IV-14).

The expansion in coverage itself, however, is not probative of a conspiratorial scheme nor even parallel conduct for the simple reason that a common effect does not prove a common cause. The expansion in coverage is explained, at least in part, by the relatively mundane fact that as term contracts reached their end-dates and renegotiations for renewal terms came due over the course of the early 2000s, defendants continued to add FSCs, consistent with their self- interest. See DA25322 at 25333, Defs.’ App’x Vol. 39, ECF No. 1172-39 (sealed) (Deposition of plaintiffs’ expert Dr. Jeffrey Leitzinger (“Leitzinger, Pls.’ Expert, Dep.”) at 441:20-43:19, explaining that longer rollout time would be needed for contracts that covered longer time periods); DA24873 at 24884, id. (Deposition of plaintiffs’ expert Dr. Jacobs (“Jacobs, Pls.’ Expert, Dep.”) at 574:17-75:14, explaining that contracts coming up for renewal may influence rates of coverage among the defendants); e.g., DA2144 at 2158, Defs.’ App’x Vol. 3 (Oct. 24, 2000, internal CSX email from Jim Mitchell to listservs, attaching presentation instructing employees to put FSC language in new and renewing contracts). Coverage thus naturally expanded.

Additionally, the increases in defendants’ FSC coverage were not wholly uniform and the differences are at least partly explainable by the defendants’ different policies over time. UP’s coverage increased greatly in late 2002—reaching close to 50% of traffic—just before the start of the alleged pre-conspiracy period, consistent with its adoption of a universal coverage policy around that time. See DA713 at 810-11, Defs.’ App’x Vol. 1 (Report of defendants’ expert Dr. Joseph Kalt (“Kalt, Defs.’ Expert, Report”) ¶ 148, Fig. IV-7A, demonstrating sharp increase in FSC coverage as percentage of revenue just before start of alleged conspiracy and showing about

67

55% coverage for private transactions); DA21943 at 22036, Defs.’ App’x Vol. 34 (Bernheim & Marx, Pls.’ Expert, Report at Fig. IV-14, estimating around 45% of FSC coverage on overall traffic revenue just before the start of the conspiracy); DA20487 at 20673, Defs.’ App’x Vol. 32, ECF No. 1172-32 (sealed) (Reply Report of plaintiffs’ expert Dr. Gordon Rausser (“Rausser, Pls.’ Expert, Reply Report”), Fig. 76, demonstrating the same significant increase in coverage in late 2002 and early 2003 but reaching only about 35% pre-conspiracy period).

CSX had a slower increase in coverage over time than other defendants, peaking at only 60% in 2007, in contrast to NS which hit 90% in 2006 and 2007, and remaining consistently 30- 40 percentage points below NS in 2005 and 2006. See DA21943 at 22036, Defs.’ App’x Vol. 34 (Bernheim & Marx, Pls.’ Expert, Report at Fig. IV-14); DA21312 at 21380, Defs.’ App’x Vol. 33 (Report of Dr. Bernheim, Pls.’ Expert, Fig. 22); see also CSX Mem. at 14-15; DA6450 at 6462, Defs.’ App’x Vol. 16 (Deposition of Dr. Rausser (“Rausser, Pls.’ Expert, Dep.”) at 290:23- 293:12, not disputing the coverage levels provided by defendants’ experts, only the inferences drawn from them). CSX, in fact, remained below all the other defendants since late 2004. See DA21943 at 22036, Defs.’ App’x Vol. 34 (Bernheim & Marx, Pls.’ Expert, Report at Fig. IV- 14). That is consistent with CSX’s delay in adopting a universal coverage policy and an exceptions-approval policy until late 2005, and the other defendants’ adoption of such policies by 2004.

Finally, plaintiffs emphasize that defendants all began tracking their FSC coverage and their progress toward 100% during the alleged conspiracy period, Pls.’ Opp’n at 40-41, 45-46, but such efforts are not probative of conspiratorial or even noteworthy parallel action. Defendants had to keep track of FSC application for ordinary business (financial, accounting, etc.) purposes. The record also shows at least some defendants started tracking coverage before

68

the conspiracy period when they first started implementing FSCs. For instance, while plaintiffs suggest that BNSF began tracking with earnest its FSC coverage in 2003 and 2004, see Pls.’ Opp’n at 40; PX68 at ‘532-33, Pls.’ App’x Vol. 1 (Sep. 10, 2004, internal BNSF email from Lanigan to David Garin et al., sharing that 92% of contracts signed in 2004 have FSC), the record indicates that BNSF started monitoring its coverage many years prior, when this company started expanding FSC coverage, see DA3982 at 3984, Defs.’ App’x Vol. 9 (July 12, 2001, internal BNSF letter from Paul Anderson to Tom Hund, explaining expectation that FSCs will double pursuant to “a more aggressive policy” adopted in July 2001); see also, e.g., DA5948, Defs.’ App’x Vol. 16 (Aug. 4, 2000, internal NS email from Stacia Minton to Seale et al., attaching spreadsheets with status of fuel surcharge billing). Moreover, as defendants decided whether to change their FSCs in 2003 and 2004, they likely paid more attention to statistics like coverage levels. That is not, however, the kind of pattern of parallel behavior that could serve as the foundation of a conspiracy.

In sum, defendants’ policy changes over the course of several years, as they increased their attention to FSCs, do not illustrate a pattern of parallel behavior consistent with coordinated action. Plaintiffs have insisted that the relevant conspiracy consists of both defendants’ FSC changes and changes to their policies—an overall agreement to increase all-in freight shipping rates for all customers—but even taking all of defendants’ actions together, the supposed parallelism is elusive. CSX announced a change to its FSC formula in March 2003 but did not apply it universally and continued to allow RCAF escalators as an alternative until 2005. UP announced it would follow CSX in April 2003 but then changed its mind and adopted a different formula to compete with BNSF in May 2003. In June, UP decided to change approval policies to enforce a previously articulated goal of universal coverage. BNSF, tracking these

69

announcements, decided in April 2003 to establish and enforce a universal coverage policy and high-level approval requirement for exceptions and changed its formula around the same time as UP the next month—although BNSF’s change was marginally different from its previous formula and not the same as UP’s. Almost a year later, in early 2004, NS adopted the same formula as CSX and decided it would apply universally with exceptions requiring high-level approval.

These “varying courses of action” over 2003, 2004, and 2005—with some defendants deciding early on to adopt unique formulas and to enforce preexisting policies of universal coverage and others waiting to make any changes at all—do not constitute parallel conduct. See Anderson News, 899 F.3d at 105 (holding that defendants’ different approaches to ceasing to do business with plaintiff over the course of several weeks was not “parallel”). That defendants all increased prices in some fashion over the course of a year and took efforts to enforce those higher prices is not enough. See Baby Food, 166 F.3d at 132 (“Although we recognize that parallel pricing does not require ‘uniform prices,’ . . . defendants’ prices were neither uniform nor within any agreed upon range of each other.” (quoting Socony-Vacuum, 310 U.S. at 222)). Without being able to show parallel conduct, plaintiffs cannot circumstantially prove a conspiracy, and summary judgment is warranted for defendants. See Williamson Oil Co., 346 F.3d at 1301.

D. Even if Defendants’ Conduct was Parallel, Defendants’ Conduct is Consistent with Lawful Interdependence.

Even if defendants’ conduct were sufficiently parallel with respect to either (or both) the pricing formulas for FSCs themselves and coverage policies, their conduct would still be more

70

consistent with conscious parallelism than the existence of an unlawful agreement. 15 Defendants’ decisions to raise surcharges and expand FSC coverage are representative of rational business interests to promote price stability and profit maximization. Defendants have also put forth extensive evidence of each defendant’s internal analyses and justifications for its decisions. An inference of unilateral action is thus, at least at the outset, an “attractive” one. See Fed. Prescription Serv., Inc., 663 F.2d at 267.

1. Defendants Had a Rational Business Motivation for Aggressively Pursuing Higher FSCs.

The existence of an independent business justification for the defendants’ behavior makes inferring the requisite illegal conspiratorial agreement for a Sherman Act section 1 violation more difficult. If an “innocent explanation of the questioned conduct . . . is plausible and more logical than a theory of concerted action, then a conspiracy may not be found.” Kreuzer v. Am. Acad. of Periodontology, 735 F.2d 1479, 1488 (D.C. Cir. 1984); Moundridge, 2009 WL 5385975, at *6 (“[W]here there is an independent business justification for the defendant’s behavior, no inference of conspiracy can be drawn.” (quoting Blomkest, 203 F.3d at 1037)); Golden Bridge Tech., 547 F.3d at 272-73 (“[I]t is not sufficient under Matsushita for [plaintiffs] to simply propose conceivable motives for conspiratorial conduct” where an “independent

15 Plaintiffs seize on this alternative argument posited by defendants—that if defendants’ conduct was parallel, such conduct was merely conscious parallelism and not conspiratorial, Defs.’ Mem. at 60-64—as a concession that defendants’ conduct was indeed parallel, Pls.’ Opp’n at 90, but this characterization is a leap too far. Contrary to plaintiffs’ position, defendants assume for the sake of argument, as does the Court here, that if they were acting in a parallel enough fashion, such behavior was due to independent awareness of what was best for the industry and their individual profits, not due to any illegal agreement.

Moreover, consideration of this alternative argument is prudent. Although plaintiffs have not argued and no binding caselaw has held that parallel conduct is not a prerequisite to showing an antitrust conspiracy with circumstantial evidence, some courts have commented that showing parallel conduct may not be necessary. See Pork, 2025 WL 1224694, at *13 (“The Court is willing to entertain the theory that ‘[i]t is possible to prove a conspiracy without evidence of parallel noncompetitive conduct,’ as especially as more courts have come to find that parallel conduct merely ‘provides one means by which a plaintiff in a § 1 case may make a circumstantial showing of concerted activity.’” (first quoting In re Broiler Chicken Antitrust Litig. II, 702 F. Supp. 3d 635, 658 (N.D. Ill. 2023), and then quoting Cason-Merenda v. Detroit Med. Ctr., 862 F. Supp. 3d 627-28 (E.D. Mich. 2012))).

71

financial motive” exists to explain defendants’ decisions.). Where there is an “obvious alternative explanation,” there is “no reason to infer that [defendants] had agreed among themselves to do what was only natural anyway.” Twombly, 550 U.S. at 566-68.

Rising input costs is a common, rational explanation for increased prices. See Supreme Auto Transp., LLC v. Arcelor Mittal USA, Inc., 902 F.3d 735, 738-39 (7th Cir. 2018) (noting that “[t]he rising cost of inputs would provide an obvious innocent explanation for the increase in steel prices” but not considering that argument because it was waived); see also Williamson Oil Co., 346 F.3d at 1311 (explaining that where the “only viable route back to profitability was to increase prices,” doing so was in their economic interest and “that they did so in a parallel manner does not establish collusion”). As explained previously, “[f]ollowing a competitor’s price increases can be consistent with rational self-interest in oligopolies. Each firm in a tight oligopoly might think that it will reap greater profits if it imitates, rather than undermines, its peers’ price hikes. And it might reach that conclusion without any conscious coordination with its competitors.” Kleen Prods., 910 F.3d at 935 (internal citations omitted). Defendants’ behavior here is reflective of such self-interested decision-making.

Defendants’ focus on FSCs was driven by erratic and rapidly rising fuel prices in the early 2000s, due to instability in the Middle East. Fuel prices shot up in 1999 and were expected to continue to rise over time, and these increased and increasing fuel costs were not expected to be recoverable under defendants’ then-extant pricing. See DA3558 at 3558-59, Defs.’ App’x Vol. 5 (Nov. 3, 2000, internal BNSF letter from Paul Anderson to Matt Rose, BNSF CEO, noting significant increases in fuel costs in 2000 and projecting increases for the next year over $216M with fuel surcharges only covering $24.5M); DA7152 at 7165, Defs.’ App’x Vol. 17 (SEC Form 10-Q for BNSF, stating, “Fuel expenses of $257 million for the first quarter of 2001 were . . . 22

72

percent[] higher than the first quarter of 2000, primarily as a result of a . . . 20 percent[] increase in the average all-in cost per gallon of diesel fuel”). Defendants had also experienced declining shipping rates (when adjusted for inflation) every year from 1985 (with one exception). Pls.’ COF ¶¶ 13-15, 524. That motivated defendants to adopt FSCs to better recover fuel costs and improve their profitability.

Both parties agree that in this pre-2003 period, defendants made that choice unilaterally—each determining FSCs were in their own self-interest. See Pls.’ COF ¶¶ 21-32, 46; id. Part II.C (titled, “Prior to 2003, the Defendants did Not Coordinate to Impose Uniform Fuel Surcharges”). Both parties also agree that defendants operated as an oligopoly during this time, making their choices not completely independent but rather interdependent. See MDL I Class Cert. I, 287 F.R.D. at 14 (quoting plaintiffs’ briefing that “defendants ‘operated as a price- discriminating, interdependent, but non-collusive oligopoly’”); Defs.’ Mem. at 6 (“Defendants accept [the contention] for present purposes.”). Defendants were able to raise prices knowing that their choices would affect other defendants’ decisions and the industry as a whole. As plaintiffs point out, defendants perhaps recognized they had been “fighting for revenue share as opposed to rate adequacy,” Pls.’ Opp’n at 15, and decided they would be all better off with higher prices.

The same business rationale—maximizing fuel cost recovery to enhance profits—that motivated defendants unilaterally to adopt FSCs in 2000-2002 persisted throughout the beginning of the alleged conspiracy period. Fuel prices continued to be unstable. See, e.g., DA2507 at 2509-10, Defs.’ App’x Vol. 4 (Jan. 31, 2003, Bear Stearns CSX Corporation analysis, describing concern about rising fuel prices and uncertain political situation in the Middle East and projecting fuel expense as a risk for growth in 2003); DA2785 at 2785, 2789,

73

Defs.’ App’x Vol. 4 (Mar. 11, 2003, internal NS email from Gooden to Onimus, asking for forecasts with war scenarios and noting “we’re taking our forecast down until the war situation plays out and with fuel cost both oil and gas nearing historical highs”). In particular, just before March 2003 when defendants began changing their FSCs, fuel prices increased significantly. See UP Mem. in Supp. of Mot. for Summ. J. (“UP Mem.”) at 4, ECF No. 1157-1 (sealed) (citing U.S. Energy Information Administration statistics revealing that oil prices increased 23% from October 14, 2002 to March 10, 2003); PX786, Pls.’ App’x Vol. 13B, ECF No. 1202-3 (sealed) (May 2003, UP analysis, reflecting shipper comment that “it is well understood that prices of fuel have skyrocketed since last fall”); DA25975, Defs.’ App’x Vol. 41, ECF No. 1245-1 (sealed) (Feb. 12, 2003, Morgan Stanley report “Lowering 1Q03 Estimates Due to Rising Fuel Prices,” stating that a “[r]ecent spike in fuel prices causes us the revisit our fuel price assumptions”); DA2772 at 2723, Defs.’ App’x Vol. 4 (Feb. 27, 2003, Credit Suisse report “The impact of rising fuel prices,” stating, “As fuel prices have continued to rise throughout 1Q:03, we felt it prudent to revisit the sensitivity of the various Class 1 carriers”); DA25987, Defs.’ App’x Vol. 41 (Feb. 18, 2003, Morgan Stanley report “2003 Railroad Outlook and 4Q02 Review,” stating that “most railroads are looking at 20% to 50% higher fuel expense in 1Q03 due to unfavorable comparisons with a year ago and near record prices today”). Not surprisingly, as a result of these fuel price increases, defendants were preoccupied with fuel costs and the effects on their bottom line. See DA2785 at 2785, 2789, Defs.’ App’x Vol. 4; DA25969, Defs.’ App’x Vol. 41, ECF No. 1245-1 (sealed) (Jan. 6, 2003, internal CSX email from Greg Stephens to Dan Gregson et al., stating, “There is an indication that fuel price may substantially exceed plan this year”); DA4451 at 4466, Defs.’ App’x Vol. 12 (Apr. 14, 2003, BNSF Fuel Surcharges: Fuel

74

Surcharge Review Team Recommendations, indicating anticipated recovery of fuel costs only at 100% FSC participation rates for non-coal traffic).

Concerns about maintaining profit margins and fuel cost recovery drove defendants to reconsider their FSC formulas, and they made no secret about this. In fact, all four defendants expressly cited fuel costs as the reason for their changes. See, e.g., PX185, Pls.’ App’x Vol. 4 (Mar. 20, 2003, CSX announcement stating, “Given the continued instability in oil prices we must, like most American companies, implement better ways to manage fuel prices changes and market volatility.”); DA1449 at 1450, Defs.’ App’x Vol. 2 (Apr. 4, 2003, UP announcement); DA4451 at 57, Defs.’ App’x Vol. 12 (Apr. 14, 2003, BNSF internal analysis stating, “Fuel surcharges must . . . reduce fuel cost volatility.”); DA5576, Defs.’ App’x Vol. 15 (Jan. 2004 internal NS memo, citing price volatility in the energy market). At the same time, demand for rail freight services was increasing, giving defendants greater power to extract higher prices and thus enhancing their economic motivations to do so. See DA6895 at 6896, Defs.’ App’x Vol. 17 (Feb. 15, 2011, Report by Federal Railroad Administration describing a surge in demand in 2004); DA6899 at 6903, id. (May 2005 CBO Report “Freight Rail Transportation: A Review of the 2004 Experience,” describing “unprecedented surge in demand” facing the railroads by mid- 2004); see also Brooke Grp., 509 U.S. at 237 (“Where . . . output is expanding at the same time prices are increasing, rising prices are equally consistent with growing product demand.”).

That FSCs are a logical way of combatting rising fuel costs is evinced by their adoption across other industries during this period as well. Trucking companies adopted FSCs, as did airlines. See DA25148 at 25157-60, Defs.’ App’x Vol. 39 (Report of defendants’ expert Dr. Mark Israel (“Israel, Defs.’ Expert, Report”), describing trucking companies’ adoption of FSCs). In one case, a federal court in the Eastern District of New York rejected an antitrust claim against

75

airlines that had imposed surcharges from 2004 to 2006, describing “rapidly rising jet fuel prices” as an “obvious potential ‘stimuli’ and ‘discernible reason’ aside from collusion that plausibly could have instigated independent decisions by defendants to impose surcharges.” LaFlamme v. Societe Air France, 702 F. Supp. 2d 136, 152 (E.D.N.Y. 2010). Non-defendant railroads also adopted FSCs. See Defs.’ Reply at 45 n.32 (listing non-defendant railroads with FSCs); e.g., PX846 at ‘499-503, Pls.’ App’x Vol. 15 (Apr. 27, 2006, statement of non-defendant railroad CN before STB, describing the surcharge they adopted in 2001 on a rate-basis, using WTI with a 0.25% increase for every $1 increase above $25). To be sure, as plaintiffs argue, practices in other industries do not establish whether defendants conspired here, Pls.’ Opp’n at 88-89 (also pointing out that 21 air freight carriers “pleaded guilty to a price-fixing conspiracy involving FSCs” (emphasis omitted) around the same time as the start of this litigation); MSJ Hr’g Tr. at 68:7-13 (plaintiffs’ counsel repeating same point), but the general movement across industries toward FSCs is highly probative of whether defendants had legitimate business motives for adopting similar practices.

Plaintiffs resist the justification of higher fuel prices for defendants’ increased FSCs, countering that at the time when defendants supposedly entered the conspiracy, fuel costs were “low” and “flat” or declining. Pls.’ Opp’n at 80 (citing NS Mem. at 5; COF ¶¶ 85-91). 16 Plaintiffs’ point is muddied, however, by the fact that defendants made disparate decisions about their FSCs at different points in time. In any case, while mid-2003 did see declining fuel prices, see PX38 at ‘559, Pls.’ App’x Vol. 1 (cited by plaintiffs) (BNSF July 2003 Analyst Presentation reporting on fuel prices dropping in the second quarter of 2003), the higher prices in very early

16 Plaintiffs cite to one portion of an extended report for the fact that fuel costs were “low and flat” at this time but, on examination, that report does not precisely specify the time period to which that characterization of those costs refers. See PX648 ¶¶ 61, 148, Pls.’ App’x Vol. 10B.

76

2003 are the ones that drove at least defendants CSX, BNSF, and UP to reconsider their FSCs. Further, plaintiffs overlook that defendants sought to increase their protection from fuel cost instability in the face of both temporarily declining but potentially imminently increasing costs. To the extent that defendants anticipated fuel prices to decrease later in 2003, that would be consistent with their choice to lower their trigger prices—so FSCs would operate at lower prices—to maximize profits and protect against fuel price instability. See PX595 at ‘817, Pls.’ App’x Vol. 10A (Deposition of Michael Giftos, CSX (“Giftos, CSX, Dep.”) at 109, explaining that he “thought fuel prices were going down” and wanted to ensure they would recover fuel costs even at lower prices). Again, defendants nonetheless made different choices with the same information. NS, seeing potentially declining prices, decided to hold off on making any changes to its FSC. See DA5574, Defs.’ App’x Vol. 15 (Apr. 29, 2003, internal NS email from Glennon to Seale, noting that “[f]uel costs are dropping . . ., which would make the current NS FSC a non- event for our customers”). 17 Defendants thus had a rational business motive for raising their FSCs and for expanding their coverage. The more widely defendants imposed FSCs across unregulated freight contracts, the greater the protection of their profits against possible increased fuel prices. Again, operating as oligopolists, defendants may have realized that they could all benefit by more stringently adhering to their higher prices, giving out more limited discounts and exceptions. See DA6412 at 6417, Defs.’ App’x Vol. 16 (Rausser, Pls.’ Expert, Dep. at 281:8-15, recognizing that defendants would be expected to “make every attempt to implement” FSC programs “as broadly

17 For these reasons, plaintiffs cannot successfully distinguish the conscious parallelism exhibited in Chocolate, 801 F.3d at 399, from the facts here on the basis that defendants there were responding to an increase in costs while fuel prices were “low and flat” here. See Pls.’ Opp’n at 84.

77

as possible absent a conspiracy” in their “unilateral self-interest,” though naturally limited by competitive interactions with other railroads).

Plaintiffs argue that offering an “alternative business justification” does not “exonerate an antitrust defendant” and, correctly, insist that the Court must still “carefully evaluat[e], and find[] insufficient, plaintiffs’ other evidence of conspiracy.” Pls.’ Opp’n at 79. As discussed infra in Part III.E, this Court goes on to do just that. The existence of a logical business rationale for defendants’ actions, however, provides a strong basis for the inference of independent action, effectively raising the bar for what plaintiffs’ evidence must tend to exclude.

2. Defendants Have Demonstrated their Decisions Were the Result of Independent Decision-Making and Internal Analyses.

Aside from the cross-industry business rationale contributing to the motivation for all defendants’ allegedly conspiratorial actions, each individual defendant has demonstrated its decisions were the result of extensive internal analysis and independent decision-making. Evidence of independent analyses demonstrating defendants’ reasoning for a decision, while not dispositive, may undercut an inference of collusion or even demonstrate a lack of conscious parallelism. See Text Messaging, 46 F. Supp. 3d at 808 (explaining that evidence of reports and “internal debates within the carriers about the increases[] indicates” they “had a good deal of time to react to competitors’ moves and consider their own,” suggesting independent action); Baby Food, 166 F.3d at 132 (“These memoranda reveal that defendants engaged in independent pricing determined by market conditions . . . . They provide a striking insight into the defendants’ marketing strategy which negates the plaintiffs’ inference of conscious parallelism.”). Here, detailed evidence of each defendant’s internal debates and analyses as each one considered when and how to change its FSCs demonstrates independent decision-making and logical, self-interested business justifications for the choices each defendant made.

78

a) CSX

CSX decided to change its formula after fuel prices dropped below the trigger point for a single day on March 18, 2003, after 29 consecutive days above that price. See DA2935, Defs.’ App’x Vol. 5 (Mar. 19, 2003, internal CSX email from Clyde McIntyre to Gooden et al., stating that the higher FSC would not be triggered due to the dip in price); DA7080, Defs.’ App’x Vol. 17 (Mar. 19, 2003, internal CSX email from Dan Murphy to McIntyre, indicating the same). CSX’s formula at the time required 30 consecutive days above the trigger price for the FSC to apply, with the result that, due to this single day fuel price dip and despite CSX and its customers’ expectations, the surcharge was inapplicable for the whole month. See supra Parts I.B.1, 1.B.5. CSX had already been considering adjusting its FSC formula by dropping its trigger price and changing the 30-consecutive-day rule, but this frustrating occurrence spurred immediate change. See DA2286 at 2290, Defs.’ App’x Vol. 4 (Mar. 11, 2003, internal CSX email from Couch to Onimus with attached presentation, comparing surcharges to BNSF and considering options to change it); DA2185, Defs.’ App’x Vol. 3 (Mar. 19, 2003, internal CSX memo from Giftos to Fred Favorite et al. stating, “[T]he current program and formula is simply unresponsive to the dramatic increases in fuel prices. I have spoken with Michael Ward this morning about a new program we want to announce immediately”); DA17582 at 17611-13, 17619-21, Defs.’ App’x Vol. 28, ECF No. 1172-28 (sealed) (Giftos, CSX, Dep. at 99:3-101:9, 110:11-112:6, explaining that the dip in prices that occurred foreclosed fuel cost recovery and caused financial concern, leading them to act quickly to change their policy).

CSX moved to a 30-day average calculation for the trigger price, which BNSF had already adopted, see supra Part I.B.5, to avoid a future repeat of the one-day-dip problem. At the same time, CSX also adjusted its trigger price and, having considered BNSF’s formula use of the

79

HDF index, CSX decided to stick with using the WTI. See DA2639, Defs.’ App’x Vol. 4 (Mar. 19, 2003, internal CSX memo from Couch to Giftos, comparing BNSF’s formula to a proposed new CSX formula and explaining that the proposed new CSX formula is superior because it is more similar to their existing formula and easier to pitch to customers). The change was announced two days later. See DA2640, id. (Mar. 20, 2003, CSX draft press release, explaining the new formula, the change to the 30-consecutive day approach, and that increased fuel costs are a big threat to the company and FSCs have been adopted in the trucking industry and others); DA2616, id. (Mar. 20, 2003, CSX press release, explaining the same). CSX executives testified explicitly that their decision was made completely independently, without having related discussions with any other railroads. DA1927 at 1940, Defs.’ App’x Vol. 3 (Deposition of Michael Ward, Chairman and CEO of CSX, at 222:16-223:16); DA17582 at 17630-31, Defs.’ App’x Vol. 28 (Giftos, CSX, Dep. at 124:25-125:4); see also CSX Mem. at 8 (noting that plaintiffs only asked these two leaders and no evidence contradicts this).

Plaintiffs argue that CSX had been planning to raise its trigger price but adopted a softer escalation formula, and then, suddenly without any explanation, CSX adopted a more aggressive escalation formula that only UP had considered internally (i.e., 0.4% for $1.00/barrel increase), suggesting that CSX did not act independently. Pls.’ Opp’n at 19, 91, 93-94; see PX501 at ‘386, Pls.’ App’x Vol. 7 (Mar. 11, 2003, internal UP email from Adams to Schroeder); PX181 at ‘495- 96, Pls.’ App’x Vol. 4 (Mar. 11, 2003, internal CSX email from Couch to Onimus with attached presentation, originally proposing new FSC formula with less aggressive slope). Contrary to plaintiffs’ suggestion, however, CSX did consider the 0.4% escalation rate before announcing its decision and in fact weighed the pros and cons of that rate against the less aggressive 0.2% rate also under consideration. See DA2639, Defs.’ App’x Vol. 4 (Mar. 19, 2003, internal CSX memo

80

from Couch to Giftos, outlining comparisons between two proposed CSX programs and other competitors’ programs the day before the announcement but clearly indicating that earlier analysis of those options predated the memo); PX163, Pls.’ App’x Vol. 3 (Mar. 19, 2003, internal CSX email from Couch to Hancock, attaching presentation depicting current fuel cost recovery and the new change). 18 Plaintiffs do not point to any document or communication suggesting CSX was influenced by or aware of UP’s internal considerations. Instead, the internal CSX analyses clearly indicate the makings of a rational, independent business decision.

CSX’s decisions regarding application of its FSC were also guided by internal discussions and unilateral business strategies. Recall that, unlike the other defendants, CSX decided to allow the use of a price escalator index, the RCAF, as an alternative to FSCs for several more years. See supra Part III.C.3. CSX considered making its program more universal as early as 2003 but decided to retain RCAF as an alternative. See, e.g., id.; PX181 at ‘493, Pls.’ App’x Vol. 4 (Mar. 11, 2003, internal CSX email from Couch to Onimus, with attached PowerPoint reflecting that they “need a written policy that Fuel Surcharge will apply on all new moves – even those with RCAF(U) [with [e]xceptions to SVP”); PX1159 at ‘623, Pls.’ App’x Vol. 24 (Oct. 31, 2003, internal CSX email from Onimus to Gooden et al.); DA25960 at 25966, Defs.’ App’x Vol. 41 (May 17, 2004, internal CSX email from Couch to Onimus, with attached presentation on Escalators 2004 Discussion suggesting as a “change[] to the policy” that “[e]verything” be “subject to fuel surcharge” with “[e]xceptions elevated”); DA7967 at 7968,

18 Plaintiffs contest the thoroughness of this deliberation, pointing to testimony from Michael Giftos, CSX’s Chief Commercial Officer, that he didn’t “ask[] for analysis” and that the new FSC program was “essentially designed . . . that night,” DA17582 at 17624-25, Defs.’ App’x Vol. 28 (Giftos, CSX, Dep. at 115:7-116:16). See MSJ Hr’g Tr. at 83:5-13. Acting quickly with limited analysis, however, is not the same as conspiring. Giftos testified that they did not have much time as they moved swiftly to resolve the one-day dip problem experienced days before, but he asked his colleague to look at other railroads’ publicly posted surcharges and did not believe any others had the formula they adopted. See DA17582 at 17624-25, Defs.’ App’x Vol. 28 (Giftos, CSX, Dep. at 115:19-116:12).

81

Defs.’ App’x Vol. 19 (May 19, 2004, internal CSX email from Dwight Price to Bill Hofmann et al., affirming that no special exception was necessary to use RCAF instead of an FSC at that time). Retaining the RCAF option allowed CSX to give shippers competitive offers that differentiated CSX from the other railroads. CSX, nonetheless, recognized that FSCs were generally preferable to the RCAF escalator. See DA6925 at 6930, Defs.’ App’x Vol. 17 (June 20, 2005, IAP 2005 MidYear Review presentation, stating that “in most cases we prefer fixed increases with Fuel Surcharge. RCAF(U) is acceptable as an alternative”). Thus, in late 2005, CSX decided to implement FSCs as the standard, see DA2672, Defs.’ App’x Vol. 4 (Dec. 21, 2005, internal CSX email from Clyde McIntyre to Gooden et al., announcing switch to using AIILF index and FSC instead of RCAF(u)); DA16514, Defs.’ App’x Vol. 27 (Dec. 20, 2005, internal CSX email from Lube to Evans et al., announcing requirement of VP level review of exceptions). CSX’s approach to coverage is thus likewise charted through internal discussions and analyses reflecting rational, independent decision-making.

b) UP

The same one-day dip in fuel prices on March 18, 2003, that prompted CSX concern also increased UP’s motivation to consider changes to its FSC formula. See UP Mem. at 4-5; DA1417 at 1417-18, Defs.’ App’x Vol. 2 (Mar. 11, 2003, internal UP email from Adams to Schroeder, with attached proposal considering changing the FSC formula to calculate the trigger price based on a monthly average even before the one-day dip that March); DA1254 at 1255, id. (Deposition of John Koraleski, UP CMO (“Koraleski, UP, Dep.”), at 77:14-78:25, explaining that in February/March, they analyzed alternatives to their FSC because they were very frustrated with the lost fuel cost recovery from the single day price drop); DA1554 at 1555, id. (May 5, 2003, UP Revised FSC communication sheet describing what motivated the FSC changes:

82

“During the First Quarter, two flaws in the program became evident. First, the 30-day reporting criteria was not responsive to market conditions, for rising or falling oil prices. Second, the immediate implementation of a surcharge after the threshold criteria was met did not allow adequate time for customers to adjust freight bill rating and payment mechanisms”). UP paid close attention to CSX’s announced change in formula and decided to follow its competitor, making its own announcement on April 4, 2003. See DA1432, id. (Mar. 24, 2003, internal UP email from Greg Sovey to Adams et al., asking for an assessment of “what is going on” with CSX pricing); DA1449 at 1449-52, id. (Apr. 4, 2003, UP press release announcing change). Upon receiving backlash from customers, however, UP rapidly changed course. See DA1453, id. (Apr. 11, 2003, internal UP email from Randy Evans to Greg Barbe, listing companies that had expressed resentment about their FSC formula); DA1459, id. (Apr. 18, 2003, internal UP email from Joel Blalock to Brian McDonald, stating, “I received strong push back on our plans to implement the new fuel surcharge June 1”). 19 UP conducted extensive internal analysis and decided instead to adopt a formula structurally similar to BNSF’s that used a monthly average to calculate the trigger price and adopted both the HDF index and same incremental percentage increase. See DA1497 at 1497- 98, id. (Apr. 27, 2003, internal UP email from Greg Barbe to Brad Thrasher et al., describing a meeting planned to discuss the FSC and including customer questions and complaints, including concern about the fuel price variability the FSC would create); DA1494, id. (Apr. 26, 2003,

19 See also, e.g., DA1454, Defs.’ App’x Vol. 2 (Apr. 16, 2003, internal UP email from Greg Barbe to Koraleski, mentioning a discussion “in light of recent customer push-back”); DA1455, id. (Apr. 10, 2003, internal UP email from Brian McDonald to Koraleski, attaching an email from a shipper expressing discontent and stating, “I am hearing that shippers are talking about petitioning the STB on this issue. In my 20 years with the railroad, I have never heard such anger and bitterness than I have heard since the announcement of this action.”); DA1461 at 1462, id. (Apr. 21, 2003, email from customer to UP stating, “This change is both unreasonable and unfair in view of the expected price of oil”); DA1466, id. (Apr. 22, 2003, memo from customer National Grain and Feed Association to UP CEO raising concerns about the newly proposed FSC).

83

internal UP email from Diane Knutson to Koraleski et al., describing that at the meeting they will provide analysis, including of other railroads’ and trucking companies’ FSCs). 20 UP determined, after much discussion, that the HDF index more closely tracked its fuel costs, was easier to administer, and would make them competitive to BNSF. See DA1534 at 1537-38, 1542, id. (May 5, 2003, internal UP email from Toy to Craig Collins et al.). UP chose a higher trigger price than used by BSNF, however: one that was halfway between UP’s previous trigger price and the one UP had announced in early April, and that was less aggressive than BNSF’s at moderate fuel prices. See id.; id. at 1539; DA1457 at 1457-58, id. (Apr. 14, 16, 2003, internal UP email chain with Toy and Schroeder, expressing change in plan). That trigger price was chosen by April 30, 2003, without any knowledge that BNSF was about to change its trigger price on May 7, 2003. See DA4374 at 4375, Defs.’ App’x Vol. 11 (Apr. 30, 2003, internal UP email from Kyei to Nick Murray, attaching new FSC); DA1534 at 1536, 1539, Defs.’ App’x Vol. 2 (May 5, 2003, UP presentation, showing comparison to other FSCs using BNSF’s FSC formula in effect prior to the change made in May 2003); DA1554 at 1556, id. (May 7, 2003, info sheet with UP trigger price released the same day as BNSF announcement). UP announced its changes a few days later on May 9, 2003.

This display of independent decision-making and internal analysis strongly indicates that UP was not acting in concert with the other defendants. Contesting this conclusion, plaintiffs argue, first, that UP does not point to any internal analysis explaining why UP decided to follow

20 See also DA1518, Defs.’ App’x Vol. 2 (UP analysis for Apr. 28, 2003, meeting); DA1492, id. (Apr. 25, 2003, internal UP email from Adams to Chandra Henley, providing BNSF’s carload surcharge history for consideration); DA1524-25, id. (Apr. 28, 2003, internal UP email from Joe O’Connor to Henley et al., listing various items for analysis in considering new FSC, including benchmarking studies, cost recovery for certain products, and how alternatives compare to the status quo); DA1468, id. (Apr. 24, 2003, internal UP email from Michael Drelicharz to Harvey Wood, with attached statistical and revenue analysis including analysis of fuel prices); DA1531 at 1532, id. (May 2, 2003, internal UP email from Edna Oakman to Vern Davis et al., forwarding meeting agenda including discussion of FSC).

84

CSX. Pls.’ Opp’n at 95-96. Surely then, plaintiffs reason, UP must have coordinated with CSX. Yet, CSX’s pricing plan that UP initially decided to follow was public, requiring no secret coordination between UP and CSX. A corporation may rationally decide to copy a competitor on the assumption that it will fare better with identical prices (whatever they are) rather than going its own course and differentiating itself. In any event, UP reversed course and did not follow CSX, so if UP did coordinate with CSX, UP broke that agreement and made its own decision almost instantly. Had a conspiratorial agreement been in effect, that act would have been expected to have been met with “disciplinary measures” to punish UP as a “cheater[]” who “did not go along with the price increases. But that type of evidence is conspicuously absent.” Kleen Prods., 910 F.3d at 937.

Plaintiffs argue, second, that UP has not put forth evidence showing why they decided to abandon CSX and instead follow BNSF. Pls.’ Opp’n at 96-97 (suggesting BNSF’s formula was viewed as even worse by customers). To the contrary, the citations above indicate ample evidence demonstrating why UP changed its formula in the way it did, even if some customers would have been opposed to the alternative.

UP again acted of its own accord when adopting a coal-specific surcharge in October 2004. UP did so months after BNSF did and only after its own internal analyses revealed that its fuel costs for coal were also not adequately recovered by carload FSCs. See DA19317 at 19319, Defs.’ App’x Vol. 30 (Knight, UP, Dep. at 92:3-11, describing how their coal surcharge program grew out of internal analyses determining that coal shipments burned more fuel); DA1599, Defs.’ App’x Vol. 2 (June 10, 2004, internal UP email from Linda Brandi to Michael Drelicharz, asking for analysis of BNSF’s coal program); DA1612, Defs.’ App’x Vol. 3 (Aug. 24, 2004, internal UP email from Drelicharz to Knight, with analysis of fuel cost recovery for different

85

sectors); DA1602 at 1602-11, id. (Aug. 24, 2004, internal UP emails between Knight, Koraleski, and Drelicharz, describing recommendation of following BNSF’s lead after analysis). 21 UP continued evaluating its FSC programs over the years. In 2005, UP considered changing its FSC but decided against it. See DA19038, Defs.’ App’x Vol. 30 (Dec. 7, 2005, internal UP email from Lee Grimes to Toy); DA19052, id. (Dec. 19, 2005, internal UP email from Toy to Eric Butler, noting, “Attached is the status of the second fuel surcharge team”); DA19053, id. (Dec. 16, 2005, internal UP presentation analyzing alternatives). When BNSF adopted a mileage-based coal surcharge, UP analyzed whether that program would be beneficial but decided to stick to its rate-based FSCs. See DA1648, Defs.’ App’x Vol. 3 (Aug. 18, 2005, internal UP email from Adams to Henley, noting BNSF’s FSC change and asking for a meeting to discuss); DA1661, id. (Sep. 1, 2005, UP Fuel Surcharge Analysis, comparing mileage based to rate based and concluding mileage-based surcharge would negatively impact UP). 22 All of these decisions are backed by extensive internal discussion and analysis that lack any indication of coordination with any other railroad. In fact, internal emails from that period actually include warnings not to communicate with BNSF about their FSC change. See DA1633 at 1633-34, id. (Aug. 16, 2005, internal UP emails from Schroeder and Toy, sharing public information about

21 See also DA15756 at 15762, Defs.’ App’x Vol. 27 (Deposition of Paul Anderson, UP, at 150:15-151:9, explaining the rationale behind a separate FSC for coal traffic); DA19309 at 19313, Defs.’ App’x Vol. 30 (Koraleski, UP, Dep. at 27:10-17, describing how UP reached conclusion, from its own analysis, to have a separate coal surcharge).

22 A later analysis revealed that BNSF’s mileage-based program might recover more revenue than UP’s program applied to the same traffic base, but UP still did not follow BNSF. See DA1664, Defs.’ App’x Vol. 3 (Oct. 14, 2005, UP analysis).

86

BNSF’s change and including legal guidance that no one should communicate with BNSF about this action). 23 c) BNSF

BNSF, likewise, made its decisions after thorough consideration of the business impact, with an extensive analytical approach. BNSF put together an internal Fuel Surcharge Review Team of seventeen people in early March 2003. See DA4112, Defs.’ App’x Vol. 10 (Mar. 6, 2003, internal BNSF email invite from Sam Kyei, BNSF Chief Economist, with agenda for pre- meeting on FSC presentation on in advance of staff meeting on Mar. 10, 2003); DA4114 at 4114, 4114-15, id. (Mar. 13, 2003, internal BNSF email from Kyei to Gary Shoop et al., initiating internal review team, attaching a presentation for consideration, and telling the team he will schedule two back-to-back meetings to discuss); DA15859 at 15864, Defs.’ App’x Vol. 27 (Deposition of Sam Kyei, BNSF (“Kyei, BNSF, Dep.”) at 27:23-28:16, describing putting together an internal review team to look into mileage-based FSCs, at Lanigan’s behest); DA4451 at 4461-62, Defs.’ App’x Vol. 12 (Apr. 14, 2003, presentation on BNSF Fuel Surcharges: Fuel Surcharge Review Team Recommendations). This team met multiple times and considered several alternative trigger prices. BNSF considered a $1.10 trigger price, which would have covered its anticipated fuel cost shortfall, as well as the CSX price, which, at the time, UP had followed, but BNSF, in the end, chose not to go with either. See DA4114 at 4114, 4117, Defs.’ App’x Vol. 10 (Mar. 13, 2003, internal BNSF email from Kyei, with attached presentation “BNSF Fuel Surcharges Cost Recovery Improvements – Part 2,” dated Mar. 10, 2003, considering the $1.10 price); DA4451 at 4461, 4463-66, 4468, Defs.’ App’x Vol. 12 (Apr. 14,

23 Regarding coverage, as previously explained, UP made the decision to adopt a universal FSC application policy prior to the alleged conspiracy period, so its decision (at least for purposes of this litigation) must have been independent. See supra Part III.C.3.

87

2003, internal BNSF presentation “BNSF Fuel Surcharges: Fuel Surcharge Review Team Recommendations,” considering the $1.10 price and CSX and UP’s rates). BNSF selected a $1.25 rate, see DA3978 at 3979, Defs.’ App’x Vol. 9 (BNSF Rules Book 6100-A), and announced that change on May 7, 2003, before UP made its announcement to switch to HDF and adopt a more competitive trigger price. See DA1565, Defs.’ App’x Vol. 2 (UP press release); DA4603, Defs.’ App’x Vol. 12 (BNSF Rules Book 6100-A) (new formula “issued May 7, 2003 – Effective June 1, 2003”). UP’s switch was unexpected and led to concern at BNSF about whether they would be sufficiently competitive. See DA4415 at 4415-16, id. (July 1, 2003, memo from Kyei to Bobb, stating, “BNSF needs to have a position on this UP action” and noting that UP will be competitive at lower prices).

BNSF continued analyzing its fuel cost recovery and, about a year later, determined that carload FSCs did not cover fuel costs for coal shipments. See DA4426 at 4426, 4428, Defs.’ App’x Vol. 11 (May 28, 2004, internal memo from Kyei to Rose, describing the fuel cost recovery shortfall for coal over the last five years). In April 2004, BNSF began considering a separate coal FSC program. See DA15645, Defs.’ App’x Vol. 26 (Apr. 21, 2004, internal BNSF email from Patrick Moynihan to Gillen et al., stating that “the current carload fuel surcharge doesn’t recover coal fuel expense. How might we do better? I’d like to discuss with both of you”). BNSF then determined that a separate program would better recover costs and would be fairer to all customers. See DA15756 at 15759, 15761, Defs.’ App’x Vol. 27 (Deposition of Paul Anderson, Asst. VP of Finance, BNSF, at 22:1-24, 147:3-17, testifying that “through various different analyses, the decision was made that it would be more accurate to be fair to all of the customers, to break coal out and make it associated with its own specific table”).

88

BNSF was clearly a first-mover again when adopting a mileage-based FSC. BNSF conducted several rounds of extensive analysis and finally announced its new program in March 2005, with an effective date of January 1, 2006. See infra Part III.E.7; DA4481, Defs.’ App’x Vol. 12 (Aug. 30, 2004, internal BNSF email from Kyei to FSC review group, stating, “John Lanigan has suggested that we reconstitute the team that looked at options for a miles per gallon approach in surcharging for fuel”); DA4521 at 4532-35, 4541-46, Defs.’ App’x Vol. 12 (Oct. 18, 2004, internal BNSF email from David Burr to Thomas Hund, identifying a mileage-based plan but also pointing out issues with adopting it for all traffic, especially interline); DA7781 at 7806, Defs.’ App’x Vol. 18, ECF No. 1172-18 (sealed) (Jan. 20, 2005, BNSF presentation, identifying mileage-based program as requiring 6-12 months of implementation to remove systems barriers, notify customers and interline customers, etc.); DA4573, Defs.’ App’x Vol. 12 (Mar. 28, 2005, internal BNSF email from Lanigan, announcing adoption of mileage-based program); DA4575, id. (Mar. 29, 2005, press release announcing application to most of BNSF’s freight traffic). In response to customer backlash, BNSF decided to hold off on a broad implementation of a mileage-based FSC and instead pilot the mileage-based program with coal and agricultural traffic only. See DA4626 at 4628, id. (Oct. 17, 2005, Mileage-Based Fuel Surcharge Update, Marketing Staff Meeting, sharing customer feedback results from a survey, showing significant opposition); DA4605 at 4605-06, id. (May 2005 BNSF Mileage-Based Fuel Surcharge Customer Feedback); DA4581 at 4581-82, id. (Apr. 4, 2005, BNSF Mileage-Based Fuel Surcharge Customer Feedback); DA4635 at 4636, id. (Oct. 20, 2005, BNSF memo to customers, describing the mileage-based program effective Jan. 1, 2006, for coal and agricultural products only, with an effective date for other products to follow later). No other defendant adopted a mileage-based program at that time.

89

Finally, BNSF’s decisions to expand coverage and require VP-level approval for exceptions to the inclusion of FSCs resulted from independent analysis and decision-making. Recommendations for such policies were voiced and discussed internally in 2002, significantly before the start of the alleged conspiracy period. See supra Part III.C.3; DA4019 at 4019-20, Defs.’ App’x Vol. 9 (Feb. 15, 2002, internal BNSF email from Chuck Schultz to Greg Patient, describing recommendations, approved by Kyei, and instructing that “[e]very contract should include a fuel surcharge clause. . . . Any deviation from the attached policy should require the approval of the Business Unit VP”). BNSF’s Chief Marketing Officer also testified that he personally had a goal by early 2003, upon his arrival at BNSF, to expand FSC coverage. See DA3099 at 3106-07, Defs.’ App’x Vol. 5 (Lanigan, BNSF, Dep. at 26:11-29:5).

d) NS

Lastly, NS’s decision-making was also marked by ample internal analyses. NS conducted three rounds of review of its FSC program before finally deciding to follow CSX. The first round of review in April 2003 led to no change. See DA5087 at 5089, Defs.’ App’x Vol. 14 (Apr. 7, 2003, internal NS memo, comparing pros and cons of adopting CSX standard); DA5820, Defs.’ App’x Vol. 16 (Apr. 7, 2003, internal NS email from Glennon to Seale et al., with attached memo presenting analysis on whether to change NS formula in light of CSX change); DA5557 at 5559, Defs.’ App’x Vol. 15 (Apr. 14, 2003, internal NS email from Tom Bayrer to Bill Fox, explaining that they have analyzed CSX’s changes and decided not to follow and instead to focus on expanding FSCs to more traffic); DA5709 at 5709-16, Defs.’ App’x Vol. 16 (Apr. 15, 2003, internal NS email from Charlie Brenner to Joe Osborne et al., requesting the team make a recommendation about whether to change the FSC based on the attached analyses); DA5574, Defs.’ App’x Vol. 15 (Apr. 29, 2003, internal NS email from Glennon to Seale, stating,

90

“The consensus [from the group that conducted the analysis] was that NS should not adopt the CSXT formula and should, for the time being, retain the current NS formula”). 24 NS believed that prices would go down, and NS could maintain a competitive advantage over CSX. As NS’s Director of Marketing Systems put it, “The CSX formula faces continued opposition . . . . NS can use this as leverage in negotiating larger increases into base rates, or additional traffic.” DA5574, Defs.’ App’x Vol. 15 (Apr. 29, 2003, internal NS email from Glennon to Seale et al.); id. (“[I]t [also] appears we may be entering an extended period of reasonable oil costs which would make the current NS FSC a non-event for our customers.”).

In June, NS once again considered following CSX but decided against this action. See DA5843, Defs.’ App’x Vol. 16 (June 30, 2003, internal NS email from Brenner to Seale et al., stating, “Mr. Seale wants to discuss the Fuel Surcharge and possible changes to the format and calculation”); DA5717 at 5717-25, Defs.’ App’x Vol. 16 (July 3, 2003, internal NS email from Glennon to Wendorf et al., regarding a comparison study between NS FSC and competitors’ FSCs to consider whether to make a change and attaching the analysis from April 2003). NS again predicted that fuel prices would be dropping, and their internal analyses recommended maintaining their current formula at that time. See DA5095 at 5095-106, Defs.’ App’x Vol. 14 (July 23, 2003, internal NS memo from Seale to L.I. Prillaman, then-NS CMO, sharing recommendation that NS not change its FSC at that time based on internal analysis and predicted fuel prices, which were expected to decline); DA5726, Defs.’ App’x Vol. 16 (July 10, 2003, internal NS email from Glennon to Wendorf et al., providing fuel price forecasts, anticipating a drop through the end of 2003); DA5673, id. (Aug. 4, 2003, email from Seale to Prillaman, stating, “Pat Glennon makes some very good points here in favor of not modifying our trigger

24 See also DA5008 at 5050, Defs.’ App’x Vol. 14 (Deposition of Don Seale, NS EVP and CMO, at 236:10-

21, explaining that NS looked at the CSX program in March and April 2003 but decided not to adopt it).

91

for the FSC at this time. I am inclined to agree—do you have a major objection to holding for now?” (emphasis in original)).

NS revisited the issue of changing its FSC again in October 2003. See DA5125, Defs.’

App’x Vol. 14 (Nov. 4, 2003, internal NS email from Glennon to Wendorf et al., noting, “Regarding the NS Fuel Surcharge, [Don Seale] has requested that we take a fresh look at the case for and against NS adopting the new CSXT FSC standard” and including several analyses for consideration); DA5647, Defs.’ App’x Vol. 16 (Oct. 31, 2003, internal NS email from Glennon to Seale, sharing what will be included in the FSC review, summarizing Seale’s request for “[p]ro/con recap of each carrier’s formula, [u]pdate on what % each is producing YTD, [a] forecast of what each will produce based on current projections, [a] general update on energy projections for next year,” and comments from the GVPs on whether the FSC should change).

In December 2003, NS decided to adopt CSX’s formula because NS’s lower FSCs were not generating the greater volume for which the company had hoped, and fuel prices had not declined but rather remained volatile. See DA5008 at 5030, Defs.’ App’x Vol. 14 (Deposition of Don Seale, NS EVP and CMO (“Seale, NS, Dep.”) at 127:13-128:12, explaining decision made in December); 25 DA5844, Defs.’ App’x Vol. 16 (Nov. 6, 2003, internal NS email from Joe Osborne to Glennon, stating, “So far, I have seen no evidence that our customers are rewarding NS versus CSX for maintaining a more reasonable FSC program”). Volatile fuel prices were particularly frustrating given NS’s existing requirement that the fuel price remain above the

25 Seale testified that throughout 2003, “[W]e had two things in mind: One, we hoped that oil would come back down and it may become less of an issue and we would maintain the original formula; Number 2, we had hoped that customers would award more business to us by having a lower fuel surcharge. We did not see that taking place over the course of 2003. We did not see oil prices and the volatility make us anymore comfortable. It looked like it was getting worse. So, therefore, we were trying to make a determination of the best approach and the decision was made to match the competition in the marketplace. And that was the crux of the decision made in December and then for publication in March.” DA5008 at 5030, Defs.’ App’x Vol. 14 (Seale, NS, Dep. at 127:13- 128:12).

92

trigger for 30 consecutive days, which made application of the FSC uncertain and unpredictable. See DA5582 at 5582-84, Defs.’ App’x Vol. 15 (undated internal NS memo from Marketing Services to Merchandise Marketing Commodity Teams, explaining, as reasons for the FSC change, wild fluctuations in fuel pricing and the problems with last minute changes when a FSC is predicted but then the price is not maintained above the trigger level for the prior 30 consecutive days); DA5803, Defs.’ App’x Vol. 16 (Jan. 8, 2004, internal NS email from Glennon to Seale et al., citing as a rationale for the FSC change: “[o]ngoing price volatility in the energy market” and “[t]he need for a formula that is more responsive to incremental cost fluctuations”); see also DA5008 at 5036-37, 5062, Defs.’ App’x Vol. 14 (Seale, NS, Dep. at 164:23-165:8, 166:22, 286:1-22). NS understood that customers, too, benefitted from predictability in pricing, and even appreciated some uniformity among different railroads. See DA6034, Defs.’ App’x Vol. 16 (Aug. 14, 2001, internal NS email from Glennon to Brenner et al., suggesting customer confusion and complaints over how the 30-consecutive-day period is calculated for triggering the FSC); DA6046 at 6047, id. (Oct. 11, 2002, email from Tom Pellington, customer, to Seale, stating, “My issue isn’t the surcharge as much as it is the timing of it going into effect. A mass change in rates during the month, like a fuel surcharge, is very costly and time consuming for DJJ to administer”); DA5803, id. (Jan 8, 2004, internal NS email from Glennon to Seale et al., stating, “Since June of 2003, the UP and CSXT have opted to drop our current standard and adopt unique FSC formulas. As a result, the BNSF, UP, CSXT, CN, CP, and NS all maintain unique FSC programs. This fragmentation has led to customer complaints about having to track multiple FSCs and rate maintenance complexity”). 26 NS announced their change January 9,

26 In fact, part of what NS cited in its announcement of the FSC change as a motivating factor was that UP and CSX dropped “our current standard”—the standard that all three had prior to the alleged conspiracy period— resulting in dis-uniformity by the Spring of 2003. By following CSX, NS would enhance uniformity, appeasing some customer concerns. See PX347 at ‘654, Pls.’ App’x Vol. 5 (same as DA5803).

93

2004, which was effective March 1, 2004. DA5936, id. (Jan. 8, 2004, internal NS email from Seale to Goode et al., announcing publicly posting a modification to the FSC the next day).

At the same time NS decided to adopt a new FSC formula, the company also decided to adopt the standard across-the-board coverage, requiring any exceptions be approved by Group Vice Presidents. See supra Part III.C.3. This policy was internally announced in the same email as the FSC change, see DA5803 at 5804, Defs.’ App’x Vol. 16 (Jan. 8, 2004, internal email from Glennon to Seale et al.), demonstrating that NS’s decision-making on these policy changes resulted from the same internal evaluation process as for the FSC formula itself. NS made these choices after thorough consideration of how to further its self-interest in the competitive market.

Plaintiffs try to discount NS’s internal analyses by insisting that Seale had the only say in the matter, with his decision driven by the conspiracy formed with leaders at the other railroads, and that others at NS did not agree with his choice. See Pls.’ Opp’n at 104-05. The supposed internal opposition that plaintiffs cite, however, see id., derives from emails in April 2003. Indeed, everyone seemed to agree not to change NS’s FSC then. Given all that occurred between April and December in terms of other railroads’ decisions, NS’s own data on fuel cost recovery and volume, and the evolving cost of fuel, that decisionmakers at NS changed their mind is not surprising. In any case, even if plaintiffs were correct that everyone around Seale disagreed with his choice, there is nothing unusual about a leader considering the pros and cons presented by his subordinates but coming to a different conclusion and having the final say.

Plaintiffs further suggest that NS’s analyses were influenced by improper private information NS obtained from CSX. See Pls.’ Opp’n at 21-22 (citing COF ¶ 136, which cites DA5709 at 5711, Defs.’ App’x Vol. 16 (Apr. 7, 2003, internal NS memo from Glennon to Seale)). The April 7, 2003, NS memo cited by plaintiffs stated that “CSXT is taking the position

94

that they are simply referring to the FSC clause in their T&C [terms and conditions] and that by changing the FSC clause in the T&C will implement that new FSC in all publications, public and private. That assumption is questionable for private authorities, particularly contracts, and they privately admit that.” DA5709 at 5711, Defs.’ App’x Vol. 16 (emphasis added). To the extent plaintiffs point to this NS memo as indicating conferral between the two companies on an agreement, the results do not bear this out. Specifically, NS did not act on that information to come to an agreement with CSX and, if anything, that comment discouraged NS to change its formula since NS retained its own formula for nine more months. While plaintiffs seize on this comment about some people at CSX “privately admit[ting]” to questions about the effectiveness of the method CSX was employing or planned to employ to ensure coverage, as an example of improper use of competitively sensitive information, see Pls.’ Opp’n at 21-22, that is a stretch. This reference did not appear to have anything to do with the specific pricing formula adopted by NS and, in any event, this memo has no relevance to proving a section 1 Sherman Act conspiracy.

As previously outlined, plaintiffs also contend that NS actually decided much earlier to follow CSX and simply had to delay implementation due to technical problems. See supra Part III.C.3. As explained, the record belies the contention that NS wanted to change its FSC earlier on, and in any case, the Sherman Act does not prohibit merely possible conspiracies that would have occurred but for technical disruptions.

Finally, while not disputing that internal analyses occurred, plaintiffs insist that internal meetings and analyses are not inconsistent with collusive conduct. Pls.’ Opp’n at 92-93 (citing Text Messaging, 782 F.3d at 876, for the unremarkable observation that purported independent evaluations may be present even in express collusion cases). Plaintiffs are correct. Defendants’

95

evidence of internal analyses and a logical business rationale for pursuing higher FSCs and broader coverage simply enables an inference of unilateral action. Plaintiffs’ burden in order to defeat summary judgment, however, is to show not only that this same conduct could be consistent with collusion but rather that such an inference of collusion is more persuasive in light of the alternative possibility. See Fed. Prescription Serv., 663 F.2d at 267.

E. Plaintiffs’ Evidence Does Not Tend to Exclude the Inference of Conscious Parallelism.

Plaintiffs’ burden on summary judgment is to provide evidence demonstrating that the inference of a conspiracy can be drawn “in light of the competing inferences of independent action or collusive action.” Matsushita, 475 U.S. at 587. This decision could end here: “[A]s discussed above, [there is] a factual flaw: defendants did not in fact engage in parallel conduct. Without ‘parallel acts’ to be reviewed ‘in conjunction with’ the circumstantial evidence, . . . evidence supporting the presence of certain plus factors . . . can provide little support for a finding of unlawful conspiracy.” Anderson News, 899 F.3d at 112 (quoting United States v. Apple, Inc., 791 F.3d 290, 315 (2d Cir. 2015)). In other words, plaintiffs have not provided any direct evidence of conspiracy, and their circumstantial evidence consisting of apparent parallel conduct, combined with certain “plus factors,” fails because defendants’ conduct was not actually parallel.

Assuming, however, a sufficient showing of parallel conduct—which falls short here as not parallel enough—plaintiffs still could not meet their burden in light of the strong inference of independent action available from defendants’ evidence and the lack of evidence from plaintiffs as to the presence of “plus factors” or otherwise, demonstrating collusion. Id. Plaintiffs’ evidence is sorted into nine somewhat overlapping categories of proposed plus factors for ease of review and then, finally, considered as a whole. See Flat Glass, 385 F.3d at 369. As the

96

following discussion demonstrates, “[w]hile some of the evidence” plaintiffs proffer “may have a negative tinge, . . . the evidence does not, taken together, tend to exclude the possibility of permissible independent behavior.” Stanislaus Food Prods. Co. v. USS-POSCO Indus., 803 F.3d 1084, 1092 (9th Cir. 2015). 27 1. Evidence of Pretextual Justifications As discussed supra in Part III.D, defendants put forth evidence demonstrating that their decisions to adopt higher FSCs and expand their coverage were driven by logical business justifications: Defendants sought to insulate their fuel cost recovery against rising and volatile fuel prices to protect and maximize their profitability. Plaintiffs emphasize that this was the common rationale they shared with customers. See, e.g., PX255 at ‘754, Pls.’ App’x Vol. 4 (Apr. 28, 2003, NS letter to customer expressing that FSCs were designed to be fair and share the costs of increasing oil prices); PX190 at ‘453, id. (May 13, 2003, CSX letter to customers expressing that FSCs were intended to “recoup some of the huge increases in our costs”); PX163 at ‘357, Pls.’ App’x Vol. 3 (Mar. 19, 2003 internal CSX email from Couch to Hancock et al., noting that a press release will go out stating that “CSXT will be changing the Fuel Surcharge program to better capture the changes to our cost of fuel”); PX866 at ‘190, Pls.’ App’x Vol. 15 (Mar. 21, 2007, email from UP to Thomas McMillen, GM, with customer press release, stating, “From the beginning of our fuel surcharge program in 2002, Union Pacific's goal has been to

27 Plaintiffs highlight that Judge Friedman previously noted that “the documentary evidence is strong evidence of conspiracy and class-wide injury to so-called carload traffic.” MDL I Class Cert. III, 272 F. Supp. 2d at 32; Pls.’ Opp’n at 4. That comment, however, was made in the context of class certification, which although occurring after some discovery, did not concern whether plaintiffs could “prove their case.” MDL I Class Cert. III, 272 F. Supp. 2d at 89 (“At the class certification stage, . . . plaintiffs need not prove their case.”). In determining merely whether plaintiffs’ “elements of their claim are ‘capable of proof at trial through evidence that is common to the class rather than individual to its members,’” id. (quoting In re Hydrogen Peroxide Antitrust Litig., 422 F.3d 305, 311-12 (3d Cir. 2008)), the court did not scrutinize the specifics of plaintiffs’ evidence in the context of conflicting evidence proffered by defendants, which is the analysis now required at the summary judgment stage. In short, that observation during consideration of class certification is irrelevant.

97

have a fuel surcharge that recovers incremental fuel costs as a result of the extraordinary increase in energy prices. Our program was designed to be responsive to our customers; fair and equitable; and simple to administer”). 28 Plaintiffs contend that such a rationale was purely pretextual because fuel costs were not a threat to defendants at the time they changed their FSCs. See supra Part III.D.1; Pls.’ Opp’n at 51-52. 29 According to plaintiffs, defendants also were not merely recovering fuel costs; they reaped exorbitant profits from FSCs. See Pls.’ Opp’n at 46-47, 49 n.22, 81-82, 111; PX395 at 13, Pls.’ App’x Vol. 6B (Oct. 11, 2005, UP Fuel Surcharge Analysis presentation, showing an overall fuel recovery percentage of 119%). Plaintiffs point to multiple communications where defendants recognized as much—that FSCs were a “profit-center” for the railroads. PX20 at ‘138, Pls.’ App’x Vol. 1 (Mar. 27, 2006, internal BNSF presentation); see also PX93 at ‘944, Pls.’ App’x Vol. 2 (Jan. 2, 2004, internal BNSF email from David Burr to Matthew Feldman, stating, “We believe the FS program is a revenue maximization program, not protection against fuel prices”); PX630 at 49, Pls.’ App’x Vol. 10B (Deposition of Charles Wickliff Moorman, NS CEO (“Moorman, NS, Dep.”), at 49:14-16, in response to “Q. So it is not a pure cost recovery

28 See also, e.g., PX347 at ‘659, Pls.’ App’x Vol. 5 (Jan. 8, 2004, Fuel Surcharge Issues and Answers document circulated to Merchandise Marketing Commodity Teams when announcing NS’s new surcharge, stating, “Fuel prices have reached extraordinary levels and continue to fluctuate widely. In order to maintain consistent price structures we need a more flexible and reactive means to cover broad fluctuations in operating costs due to fuel prices”). 29 For this conclusion, plaintiffs cite, e.g.: PX776 ¶ 77, Pls.’ App’x Vol. 13B (Israel, Defs.’ Expert, Report, noting that future prices for WTI pointed to falling in early 2000 to early 2003, but suggesting that meant that a “conspiracy to raise prices via fuel surcharges would not be expected to be effective among the putative conspirators”); PX270 at ‘522, Pls.’ App’x Vol. 4 (Apr. 29, 2003, internal NS email from Glennon to Seale, predicting entering a period of reasonable oil costs and therefore not keen to change its FSC); PX344 at ‘447, ‘449, Pls.’ App’x Vol. 5 (Apr. 27, 2006, statement of Seale, NS, before STB, testifying that a “relatively stable and predictable period of fuel prices ended in mid-2003,” and NS thought and hoped that the oil price increases experienced at that point were temporary). As previously explained, the volatile and uncertain nature of fuel costs drove defendants’ focus on FSCs, even if pricing was temporarily declining at some point in 2003. See supra Part III.D.1.

98

mechanism? A. It is certainly not”). 30 Plaintiffs argue that the rationales for BNSF and UP adopting coal-specific formulas were pretextual too, because the cost of carrying coal, plaintiffs contend, is actually less than traditional carloads. Pls.’ Opp’n at 118 & n.45. Even the STB recognized that FSCs were not just about recovering fuel costs. See PX653, Pls.’ App’x Vol. 10B (STB decision, Jan. 25, 2007) (“A fuel surcharge is . . . purportedly [intended] to recoup increases in the carrier’s fuel costs.”), STB Decision, 2007 WL 201205, at *1; MDL I MTD Order (Direct Purchasers), 587 F. Supp. 2d at 36 (describing the STB’s finding that the defendants’ FSCs for rate-regulated traffic was “misleading and ultimately unreasonable” because “there [was] no real correlation between the rate increase and the increase in fuel costs for that to which the surcharge is applied” (quoting STB Decision, 2007 WL 201205, at *4)). 31 Plaintiffs’ point is that the false explanation for why defendants changed their FSCs was an “artifice perpetuated jointly,” Pls.’ Opp’n at 82, and such pretextual excuses “can disprove the likelihood of independent action,” id. at 51 (quoting Rossi v. Standard Roofing, Inc., 156 F.3d 452, 478-79 (3d Cir. 1998)). “Pretext alone,” however “does not create a reasonable inference of conspiracy.” Chocolate, 801 F.3d at 411. “That rising costs may not have been the full or even

30 See also PX187 at ‘525, Pls.’ App’x Vol. 4 (June 3, 2004, internal CSX email from Fredrik Eliasson to Dean Piacente, stating, “[P]lease don’t communicate to anybody outside this company that a customer who pays [the FSC] is paying twice his fair share”); PX110 at ‘452, Pls.’ App’x Vol. 2 (May 5, 2006, internal BNSF email from Kevin Kaufman to Rose, responding to concern that the other railroads are not happy with BNSF’s approach to fuel with comment that “fact is that [surcharges] make money. Big money. That is what it is about,” suggesting a competitive money-making approach to FSCs); PX449 at ‘385, Pls.’ App’x Vol. 7 (Mar. 9, 2006, internal UP email from Mark Kelehan to Toy, expressing need to protect company from appearance of increasing rates via both base rate increases and FSCs and from embarrassment of regulatory investigation); PX864 (May 15, 2006, US Rail – Steering Committee Update, reflecting customer observation that “[a]ll [railroads] claim that intent with FSC is to recover incremental costs, whereas it is quite evident that it is used as an additional revenue stream. All however open to further discussion on this point.”). 31 Judge Friedman went on in this opinion to note that “[w]hile the Court is not, of course, bound by the STB opinion, the STB’s findings, and the logic underlying them, reinforces the Court’s view that plaintiff’s claims are plausible.” MDL I MTD Order (Direct Purchasers), 587 F. Supp. 2d at 36 n.5. The STB Decision does not, however, lend plaintiffs support at this later stage of litigation, where plaintiffs must put forth actual evidence of conspiracy, not merely a plausible theory, especially considering the differences in rate-regulated and unregulated traffic. See supra Parts I.A, I.B.4.

99

real reason for increasing prices does not show whether the real reason was interdependence or a conspiracy.” Id. at 411-12. Rising prices in the face of stagnant or falling costs “may be not because [defendants have] agreed not to compete but because all of them have determined independently that they may be better off with a higher price. That higher price, moreover—the consequence of parallel but independent decisions to raise prices—may generate even greater prices (compared to competitive pricing) if costs are falling, provided that consumers do not have attractive alternatives.” Text Messaging, 782 F.3d at 871-72. 32 As a result, plaintiffs’ argument that defendants’ justifications about fuel cost recovery were pretextual, even if true, would not show that defendants conspired here.

Moreover, defendants’ justifications were not merely pretextual. Some of the defendants failed to recover their fuel costs in the early 2000s and adopted FSCs as a result. See, e.g., DA3558 at 3558-59, Defs.’ App’x Vol. 5 (Nov. 3, 2000, internal BNSF letter from Anderson to Rose, noting significant increases in fuel costs in 2000 and projecting increases for the next year of over $216 million, with fuel surcharges only covering $24.5 million); DA25969, Defs.’ App’x Vol. 14 (Jan. 6, 2003, internal CSX email from Greg Stephens to Dan Gregson, stating, “There is an indication that fuel price may substantially exceed plan this year”); see supra Part III.D. In 2003, defendants remained concerned about their ability to recover fuel costs effectively in the face of rising or unpredictable prices and accordingly reconsidered their formulas. See supra Part III.D. Indeed, at least one firm did grapple with under-recovery of fuel costs throughout the conspiracy period. See DA7929 at 7931, Defs.’ App’x Vol. 19 (July 2006, CSX Fuel Surcharge

32 Text Messaging goes on to explain how defendants could land on the same elevated prices independently.

“Competitors in concentrated markets watch each other like hawks. Think of what happens in the airline industry, where costs are to a significant degree a function of fuel prices, when those prices rise. . . . The airline’s competitors will monitor carefully the effects of the airline’s response to the higher fuel prices afflicting the industry and may well decide to copy the response should the responder’s response turn out to have increased its profits.” 782 F.3d at 875.

100

Recovery Status, stating “CSX baseline fuel surcharge economics show under-recovery in every year since the beginning of our fuel surcharge program”).

Plaintiffs may be correct that defendants did not risk a net loss due to rising fuel prices, but defendants, as for-profit corporations, did not merely want to break even. Defendants wanted to maintain and grow (as much as possible) their profit margins—without having fuel costs chip away at that. See PX630 at 57-58, 126, Pls.’ App’x Vol. 10B (Moorman, NS, Dep. at 57:7-58:8, explaining that NS didn’t view profits as “overrecovery,” rather NS tried to maintain or increase its profit margin and wanted its prices to facilitate that in a way that reasonably reflected what was happening in the marketplace). 33 That defendants emphasized to customers the “cost recovery” portion of this goal and not the concomitant motive to protect profit margin does not make their concern with fuel costs disingenuous. 34

33 Plaintiffs, in particular, attack defendants’ explanations about changing their FSC pricing formulas due in part to the one-day dip on March 18, 2003, in fuel prices by showing that CSX recognized it would not actually suffer any losses because the fuel price decreases would make up for it. See Pls.’ Opp’n at 52-53 (citing DX2893 at 2895, Defs.’ App’x Vol. 5 (Mar. 25, 2003, internal CSX memo from Giftos to Ward, CSX, stating “[T]he fuel surcharge will be about $9 million less than previously projected but this revenue reduction should be more than offset by declining prices”)). Yet, again, defendants are not in the business of avoiding losses; they saw an opportunity to protect and promote their profits and took it. 34 Plaintiffs also suggest that defendants’ reticence to make public their “revenue earned from FSCs” stems from concern about the “appearance” of high FSCs and “embarrassment . . . should a regulatory agency investigate this pricing mechanism,” in the absence of a legitimate justification. See Pls.’ Opp’n at 55 (quoting PX449 at ‘385, Pls.’ App’x Vol. 7 (Mar. 9, 2006, internal UP email from Mark Kelehan to Kate Betsworth, stating, “Education is appropriate but so is fool-proofing systems to protect the company (1) from any appearance of building rates with high fuel costs and then charging fuel surcharge and (2) from any embarrassment this appearance would bring should a regulatory agency investigate this pricing mechanism”)); see also PX187 at ‘525, Pls.’ App’x Vol. 4 (June 3, 2004, internal CSX email from Eliasson to Piacente, stating, “[P]lease don’t communicate to anybody outside this company that a customer who pays [the FSC] is paying twice his fair share”); PX449 at ‘385, PX286 at ‘582, Pls.’ App’x Vol. 5 (May 18, 2004, internal NS email from Kitty Vollbrecht to Rob Martinez, stating, “Glennon told me today that based on the high level numbers from accounting it seems like all the Merchandise rev/car increases are coming from fuel surcharges. I’m not sure if Don [Seale] would be in agreement with having this widely known, or be subject to questions from Finance”). These emails, which express different concerns at different points in time, represent nothing more than defendants’ real-world recognition that profit maximization may at times look exploitative to customers and invite regulatory attention. Efforts to maximize profit, however, do not imply an agreement in restraint of trade.

101

Likewise, plaintiffs’ contention that defendants not only preserved but increased their profitability is of no matter. FSCs were not meant to recover fuel costs dollar-for-dollar—and in fact did not. See, e.g., DA7929 at 7931, Defs.’ App’x Vol. 19. Whatever success the FSCs had for defendants in navigating the unpredictable fuel market does not suggest a conspiracy any more than shrewd business decisions or conscious parallelism. In any event, defendants’ FSCs prior to the conspiracy period had the same earning power—and thus would be responsible for the same perceived over-recovery—as defendants’ FSCs during the alleged conspiracy. See Defs.’ Mem. at 42; DA7623 at 7626-27, Defs.’ App’x Vol. 18 (Class Certification Hr’g Tr. (Sep. 27, 2016), testimony of plaintiffs’ expert, Dr. Rausser at 396:22-397:24, analyzing one of defendants’ expert’s graphs depicting a possible “large over-recovery in the but-for world”). 35 Plaintiffs therefore cannot link the general profitability of FSCs to defendants’ challenged conduct to support an inference of conspiracy.

Plaintiffs respond that in the absence of the conspiracy, defendants would have cut their prices to be more competitive and would not have sustained such over-recoveries. See Pls.’ Opp’n at 81-82. That assumption is completely speculative. Plaintiffs make four points in support that do little to advance their position. First, plaintiffs note the fact that BNSF reduced its FSC in early 2003 as part of a temporary discount but not again once the alleged conspiracy period started. See Pls.’ Opp’n at 81. Yet, plaintiffs have no reason to suggest any different outcome would have occurred in their but-for universe (i.e., but-for the alleged conspiracy). Second, plaintiffs cite handwritten notes from an unidentified author at a CSX-UP meeting in June 2003, PX474 at ‘088, Pls.’ App’x Vol. 7, noting that both companies have a $23 trigger and

35 Defendants admit that the coal-specific surcharges adopted by BNSF and UP did have greater earning power, see Defs.’ Mem. at 42 n.10, but because those were adopted only by BNSF and UP—four months apart with different formulas—these changes neither reflect parallel conduct nor demonstrate a conspiracy among the defendants. See supra Parts III.C.1-III.C.2.

102

“may have to revisit $23 if prices stay high.” Pls.’ Opp’n at 81. Whatever that unknown author meant, the note cannot suggest defendants would have changed their prices in the absence of an agreement; if anything, this comment suggests defendants were open to still changing their prices during the alleged conspiracy period. Regardless, as a general principle of economics, oligopolists will adhere to profitable pricing decisions if they can do so without a cost, as defendants appeared to do here, and that is possible without any extant conspiratorial agreement. See Text Messaging, 782 F.3d at 871-72, 874-85; Defs.’ Reply at 37-38.

Third, plaintiffs allude to the difference in defendants’ FSC coverage rates before and after the conspiracy, based on the presumption that their supposed inability to impose surcharges more widely prior to the conspiracy period evaporated due to an agreement. See Pls.’ Opp’n at 82 n.31. As explained, however, defendants’ approach to widespread FSC coverage was quite nuanced, with some defendants greatly increasing coverage before the conspiracy and others’ coverage expanding naturally over time as more contracts came up for renewal and hence renegotiation. See supra Part III.C.3.

Fourth, plaintiffs reference a comment made in a BNSF report—“I am concerned about our reliance on fuel surcharge as a mechanism to control prices. While it is working well now, it would only take one railroad to abandon this in an attempt to gain market share to cause this to fail. We need to closely monitor UP’s hedging activities for indications they might place less reliance on fuel surcharge.” PX126, Pls.’ App’x Vol. 2 (Sep. 12, 2005, internal BNSF report “Enterprise-Wide Risk Assessment: BNSF Railway Risk Profile”); Pls.’ Opp’n at 82 n.31, 48, 97. To plaintiffs, such a comment signifies that, absent a conspiratorial agreement, one defendant’s prices would have dropped and the whole market would have shifted. This nefarious interpretation, however, is not borne out by the actual text. BNSF’s concern was that other

103

defendants might depart from the consciously parallel pattern more straightforwardly, and this indicates that there was not an agreement at that time, with all defendants simply acting independently as profit-maximizing oligopolists that could decide to depart and adopt a new strategy on their own whim.

Lastly, plaintiffs note that the Canadian railroads, which had also adopted FSCs, reduced their FSCs as fuel prices went up during the conspiracy period, suggesting these non-defendants’ FSC reductions show that defendants remained bound by an agreement to keep FSCs high. See Pls.’ Opp’n at 48 (citing PX984 at 66, Pls.’ App’x Vol. 19A, ECF No. 1205-2 (Reply Report of Dr. Leitzinger, Pls.’ Expert, ¶ 132)). As defendants respond, however, CPR and CN had previously higher FSCs than those of defendants, and defendants did, in any case, alter their FSCs around the same time period as those changes in 2005 and 2006. See Defs.’ Reply at 45 n.32; PX846 at ‘506, Pls.’ App’x Vol. 15 (Apr. 27, 2006, Statement of CN before the STB, attachment showing changes in FSCs); DA25967 at 25968, Defs.’ App’x Vol. 41 (Sep. 7, 2004, internal CSX email from McIntyre to Gooden et al., attaching a chart comparing FSCs of railroads in 2004). BNSF moved toward a mileage-based FSC and NS rebased. See supra Parts I.B.1(b), I.B.5, III.C.1, and infra Part III.E.9.

Even if defendants had misrepresented to customers the purpose of FSCs, that would not itself constitute a conspiracy in restraint of trade and would, at most, be circumstantial evidence that could—with other evidence, though not alone—support an inference of conspiracy. Nonetheless, as explained, because defendants’ stated rationales were not dishonest or pretextual, they do not advance plaintiffs’ inference of conspiracy even to that limited degree. Importantly, plaintiffs also cannot point to any evidence suggesting that defendants agreed with one another

104

on the purported justifications for FSCs or that they discussed their public statements with each other at all.

2. Evidence of Common Motive to Conspire Seeing defendants’ articulated justifications as pretextual, plaintiffs impute to defendants a common motive to conspire. Plaintiffs point out that defendants struggled with “vicious historical price wars,” stagnant revenues, and “[d]estructive [p]ricing” among competitors prior to the alleged conspiracy period, which supposedly gave defendants a motive to conspire. Pls.’ Opp’n at 61 (citing PX127 at ‘617, Pls.’ App’x Vol. 3 (May 28, 2002, Ernst & Young Enterprise Wide Risk Assessment for BNSF, compiling individual comments from executive team and indicating difficulty in imposing FSCs given competition), PX128 at 114-15, id. (comments collected for same report, expressing concern about competitive dynamics), and PX277 at ‘612, Pls.’ App’x Vol. 4 (June 9, 2000, internal NS email from Ken Yopp to Thomas Brugman, expressing that NS was worried about customer feedback from the FSC and was planning to track loss of business attributable to it)). Much like a pretextual justification, however, a mere motive to conspire does not alone tend to exclude the possibility of independent action. As Flat Glass explained, such a motive “largely restate[s] the phenomenon of interdependence” among oligopolists. 385 F.3d at 360. Plaintiffs’ evidence must tend to exclude the possibility that mere interdependence explains defendants’ actions. See id. at 360-61. “Motivation . . . therefore adds nothing to it.” AREEDA & HOVENKAMP, supra, ¶ 1434c1.

Moreover, the motive to conspire tends to be “negated” “when the defendant shows a ‘plausible and justifiable reason for its conduct that is consistent with proper business practice.’” Anderson News, 899 F.3d at 112 (quoting PHILLIP E. AREEDA & HERBERT HOVENKAMP, ANTITRUST LAW ¶ 308, at 175-76 (4th ed. 2011)). As explained supra Part III.D.1, defendants all pursued a logical business aim of recovering fuel cost in a volatile market while

105

maintaining or expanding profit margins. Notwithstanding any motive to conspire, defendant had a clear rational business motive to act in a consciously parallel fashion. Further, the apparent motive to conspire existed prior to the conspiracy period—according to plaintiffs’ citations, as early as 2000—yet defendants adopted FSCs independently then. As a result, any conspiratorial motive does not move the needle in plaintiffs’ favor.

3. Evidence of Opportunities to Conspire: Meetings about FSCs Beyond mere motive, plaintiffs emphasize frequent discussions and meetings among defendants, which meetings they contend provided opportunities to conspire. Pls.’ Opp’n at 62, 120. “[T]he mere opportunity to conspire, among antitrust defendants,” however, is also “insufficient evidence from which to infer an anticompetitive conspiracy.” Moundridge, 2009 WL 5385975, at *9 (quoting Cooper v. Forsyth Cnty. Hosp. Auth., Inc., 789 F.2d 278, 281 (4th Cir. 1986)). “Even if ‘opportunity’ were defined narrowly and limited to proven interfirm contacts, deeming opportunity to be prima facie proof of conspiracy would lead to widespread condemnation of procompetitive collaborations.” AVEEDA & HOVENKAMP, supra, ¶ 1417b. Therefore, “evidence of ‘opportunity’ should be accorded little, if any weight. Company personnel do not often operate in a vacuum, or ‘plastic bubble’; they sometimes engage in the longstanding tradition of social discourse.” Baby Food, 166 F.3d at 133.

Instead, what matters is the nature of specific conversations that occurred among defendants at particular meetings. Plaintiffs put forth evidence of several meetings where defendants allegedly discussed FSCs, suggesting that discussions involving pricing are especially probative. Pls.’ Opp’n at 49, 62. Though discussions of pricing may be more probative than mere opportunities to conspire, such discussions are still not themselves sufficient to establish conspiracy. See Baby Food, 166 F.3d at 125; Kleen Prods., 910 F.3d at 941 (holding that “economic and noneconomic evidence that could suggest suspicious activity” is not necessarily

106

enough to “find impermissible coordination”). To be probative of conspiracy, the evidence must suggest that the exchange of pricing information impacted defendants’ decisions. See Baby Food, 166 F.3d at 125. 36 Such impact could be demonstrated, for instance, by relevant discussions among defendants accompanied by near simultaneous price increases. As the Seventh Circuit explained, meetings among defendants and comments from one about “giv[ing] the group a heads up” and “learn[ing] valuable info from each other” “would be more compelling if the immediate sequel to any of these meetings had been a simultaneous or near- simultaneous price increase by the defendants[] [i]nstead [of] substantial lags,” especially given that “there [was] no evidence of what information was exchanged.” Text Messaging, 782 F.3d at 878.

That connection between the discussions about pricing and defendants’ conduct is particularly important in industries with frequent substantive communication among competitors—like the railroad defendants here, who must discuss prices, including FSCs, for interline shipments. See, e.g., PX360 at ‘003, Pls.’ App’x Vol. 6A (Jan. 11, 2001, Fuel Surcharge Processing Best Business Practices Document, explaining that interline partners must agree on FSCs and how they will be calculated); Defs.’ Reply at 21-22 & n.13; Pls.’ Opp’n at 109 (recognizing that some meetings may have had the legitimate purpose of communicating about interline traffic). In these circumstances, “when companies have legitimate business reasons for their contacts, plaintiffs must offer some evidence that moves beyond speculation about the content of what was conveyed.” Kleen Prods., 910 F.3d at 938. While the need to

36 As a leading treatise on the matter put it, “the existence of meetings or phone conversations among the defendants does not warrant the inference that they agreed about prices, terms of dealing with the plaintiff, or any other subject matter even if such subjects were discussed. . . . [S]uch meetings or conversations merely amount to the opportunity to conspire about the subject matter alleged by the plaintiff and that such opportunities do not warrant the inference of a conspiracy about the alleged subject matter.” AREEDA & HOVENKAMP, supra, ¶ 1406a.

107

discuss interline traffic does not immunize every discussion about FSCs, accord Pls.’ Opp’n at 136-37, not every meeting can be viewed as inherently suspect either. Cf. id. at 50 (citing Publication Paper, 690 F.3d at 65, where the Second Circuit found sufficient evidence supporting conspiracy, which included evidence of meetings and calls with no other social or professional purpose where defendants discussed pricing and their intentions to make pricing changes). Of course, where they exist, “[i]ncriminating remarks by defendants’ employees can support the inference that a conspiracy existed.” Kleen Prods., 310 F.3d at 939.

Here, plaintiffs focus on eight meetings that took place between April and July of 2003, all of which defendants describe as mere “opportunities” to conspire without any indication that an agreement was formed. Defendants therefore urge that most of the meetings should be excluded as irrelevant under Federal Rule of Evidence 403. Defs.’ Reply at 18. Putting aside the Rule 403 challenge and assuming arguendo that all the meetings are admissible, none can be reasonably tied to any action of defendants to support an inference that an agreement was formed. These meetings proffered by plaintiffs as circumstantial evidence supporting a finding of a conspiracy are reviewed chronologically—and all found to fall short of this proffer.

a) April 3-8, 2003, Five-Day National Freight Transportation Association (“NFTA”) Meeting

Plaintiffs contend that all four railroads’ executives attended a five-day National Freight Transportation Association (“NFTA”) meeting in early April 2003 and their conversations led to two internal decisions at BNSF. Pls.’ Opp’n at 23-24, 34-35. Right after the conference, on April 9, 2003, BNSF’s CMO, John Lanigan, internally announced a universal FSC coverage policy and a requirement for VP-level approval for exceptions. See PX50 at ‘935, Pls.’ App’x Vol. 1 (Apr. 9, 2003, internal BNSF email from Smith to Obermiller, stating, “Effective

108

immediately and urgently per John Lanigan. Authority to omit FSC provision is to be granted to VPs only (who will also clear with John)”).

Plaintiffs do not offer anything but the sequence in timing to suggest that an agreement related to FSC coverage was made at this conference. Drawing conclusions based solely on suspicious timing may be risky when events may simply be coincidental, which appears to be the case here in the context of other relevant evidence. Cf. AREEDA & HOVENKAMP, supra, ¶ 1417b (cautioning against inferring conspiracy from a mere opportunity to conspire). For example, by the time of this NFTA meeting, BNSF already had internally discussed a universal coverage policy with requirements for high-level approval of exceptions and had focused on expanding coverage prior to the alleged conspiracy period. See supra Part III.C.3. In addition, BNSF had a strong self-interest in expanding its FSC coverage. See supra Parts III.C.3, III.D.1. The conclusion that defendants must have made an agreement at that meeting also does not make sense given that no other competitor issued a similar policy around this time. See id. The meeting also cannot be reasonably connected to BNSF’s decision to stick to its rate-based, rather than mileage-based surcharge, as plaintiffs suggest. Pls.’ Opp’n at 35 (“Days later, BNSF formally abandoned its plans to switch to a competitively advantageous mileage-based FSC.”). Plaintiffs do not point to any additional evidence of discussions had there.

In short, while bumping elbows with his competitors may have inspired BNSF’s CEO to get more competitive by doubling down on its application of FSCs, no agreement can be inferred from plaintiffs’ limited evidence—after years of discovery—about the NFTA meeting.

b) April 15-17, 2003, CMO Meeting Plaintiffs next describe an “unsupervised getaway with CMOs and CMOs-in-waiting from all four Defendants, and with no other employees invited,” in mid-April 2003. Pls.’ Opp’n

109

at 22 (emphasis in original) (citing, e.g., PX356 at ‘220 (Feb. 18, 2003, email from Koraleski, UP, to individuals from CSX, NS, BNSF, CN, and CPR, inviting them to attend a retirement party trip for a BNSF employee, Schultz)). Plaintiffs suggest that discussions at this meeting drove NS to change its FSC, see id., describing this event as a “shindig” without any apparent purpose, MSJ Hr’g Tr. at 89:12-90:3. As plaintiffs put it, prior to the meeting, NS had not made any decision about whether to change its FSC. See Pls.’ Opp’n at 22; PX254 at ‘316-17, Pls.’ App’x Vol. 4 (Apr. 15, 2003, internal NS email from Glennon to Bayrer, stating, “Here is the analysis I sent [Seale]. I don’t have any indication what we might do”); PX913 at ‘541, Pls.’ App’x Vol. 15 (Apr. 15, 2003, internal NS email from Bayrer to Glennon, stating, “Fox said that he thought [Seale] was going to hold off on any changes for now consistent with your advice”). Then, on the day of the executives’ return, Seale, NS’s EVP and CMO, indicated via email that “at this point, we do plan to modify our FSC,” without having even received the full analysis he had requested. PX254 at ‘316, Pls.’ App’x Vol. 4 (Apr. 17, 2003, internal NS email from Seale to Kathryn McQuade). According to plaintiffs, Seale made the decision based on that CMO meeting alone, and NS then adopted “CSX’s FSC as quickly as it could, and ultimately announced it was adopting the same FSC program as CSX, with the intent of ‘standardiz[ing]’ FSCs.” Pls.’ Opp’n at 23 (alteration in original) (citing COF ¶¶ 206-23; PX347 at ‘654-56, Pls.’ App’x Vol. 5 (Jan. 8, 2004, internal NS email from Glennon to Seale et al., announcing FSC change)).

This interpretation of the facts about what occurred at this retirement party for the BNSF CMO makes no sense. Regardless of what Seale said in any email in April 2003, NS did not in fact change its FSC until January 2004 and did not internally settle on that decision until December 2003. See supra Part III.D.2(d). That complete disconnect in the timing is fatal to

110

any inference about NS’s decision-making that plaintiffs propose to tie to the mid-April 2003 CMO meeting.

A closer look at the relevant email chain presents a more logical chain of events. On April 15, 2003, at 9:01 AM, apparently around the same time the CMO conference began, Tom Bayrer, an NS employee, asked Charlie Brenner, another NS employee, and Pat Glennon, NS Director of Marketing Systems, whether NS was considering FSC changes. See PX913 at ‘541, Pls.’ App’x Vol. 15. Glennon wrote back 44 minutes later with the analysis he had previously sent to Don Seale, NS CMO, and said, “I don’t have any indication what we might do.” See id. Bayrer replied that he thought Seale was likely to hold off on any changes. See id. Just minutes later, at 10:07 AM, Brenner shared Glennon’s analysis with a broader internal NS group, stating that Seale had reviewed it in the morning meeting. PX254 at ‘316, Pls.’ App’x Vol. 4. Another NS colleague then emailed Seale early that afternoon to let him know that the Finance, Cost, and Fuel groups within NS had been discussing FSCs and asked him, “would your groups also like to join a review of alternative surcharge mechanisms?” Id. (Apr. 15, 2003, internal NS email from McQuade to Seale). Two days later, Seale responded, “we already have a review underway and Bill Fox and I are discussing it for both Coal & Merchandise. Note the attached input from Pat Glennon and the Team that is addressing it. Ike [Prillaman] and I had a conference call with all the Merchandise GVPs on 4/14 and at this point, we do plan to modify our FSC – but we are analyzing further action.” Id. (Apr. 17, 2003, internal NS email from Seale to McQuade).

This chain makes clear that Seale’s review of Glennon’s document and discussions with others internally at NS (including Ike Prillaman, then-CMO, later CEO) that occurred on April 14 and 15 led him to a tentative position on April 17. That Glennon did not know where Seale stood on April 15 and that Bayrer had a different sense of Seale’s position in no way suggests

111

that Seale’s attendance at a meeting with other defendants alone led him to change his mind. These emails rather indicate unequivocally that any position he took was based on meetings with and analyses provided by NS colleagues during the prior few days. Importantly, Seale’s email on April 17 was also far from a “foregone conclusion” on the issue. Pls.’ Opp’n at 101. Clearly, NS did not choose to change its FSC until eight months later after further analysis. See supra Parts III.C.2, III.D.2(d).

In short, plaintiffs’ assumption that the mid-April CMO meeting led to an agreement among defendants that NS implemented as rapidly as possible is simply not in accord with the facts.

c) April 21, 2003, NS-CSX Meeting about “Fuel Hedging”

Plaintiffs further contend that NS invited CSX for a meeting on April 21, 2003, to learn about its “fuel hedging practices.” Pls.’ Opp’n at 23; PX164 at ‘396, ‘399, Pls.’ App’x Vol. 3 (Apr. 17, 2003, email with agenda for meeting). Plaintiffs suggest such a meeting is necessarily suspect because “[t]ypically, a company does not teach its largest competitor lessons in ‘reducing fuel expense volatility,’ which is a competitively sensitive issue.” Pls.’ Opp’n at 23.

The meeting does not, however, advance plaintiffs’ inference of a conspiracy about FSCs. First, fuel hedging is a different issue from FSCs. Although both are ways to address fuel expense volatility and are thus related, they are different techniques altogether, as both parties seem to agree. See Pls.’ Opp’n at 23; Defs.’ Reply at 24 n.15. Second, plaintiffs have no evidence that FSCs or other competitively sensitive information were discussed. One attendee testified in detail about everything he could remember from the meeting—and did not include any mention of FSCs or other precise competitively sensitive information about fuel cost recovery. See PX608 at 24-25, 35-68, Pls.’ App’x Vol. 10A (Deposition of Chris Neikirk, Dir.

112

of Finance of NS, at 23:13-25:18, 35-68). Also, hedging activities were not a secret during this period, as defendants disclosed their hedging in public earnings calls and customer presentations around the same time. See DA8061 at 8162-64, Defs.’ App’x Vol. 19 (Feb. 24, 2003, NS Corp. Annual Report Form 10-K); DA8044 at 8047-48, id. (Apr. 23, 2003, NS Corp. Earnings Conference Call); PX711 at 26, Pls.’ App’x Vol. 11, ECF No. 1221-4 (sealed) (2003 CSX Corp. Form 10-K); DA8246 at 8247, Defs.’ App’x Vol. 20, ECF No. 1172-20 (sealed) (Apr. 1 through June 30, 2003, Activity Report for Trevor Pardee, showing a customer presentation on fuel hedging).

That NS and CSX somehow shared semi-public information on one subject in order to come to an agreement on another is implausible.

d) Alleged April 23, 2003, BNSF-CSX Meeting Plaintiffs reference an “April 23 meeting between BNSF and CSX senior executives, including Giftos, Gooden, Lanigan, BNSF CEO Matt Rose, and CSX CEO Mike Ward,” after which “BNSF formally abandoned its plans to switch to a mileage-based FSC.” Pls.’ Opp’n at 24; see also id. at 105-06. Plaintiffs’ only evidence of this meeting is an email invite that was sent on April 23, 2003, but another version of the email clarifies that the meeting occurred on October 1, 2003. PX33 at ‘200, Pls.’ App’x Vol. 1 (Apr. 23, 2003, email from Barbara Turner [affiliation not identifiable], sending invite); DA25923, Defs.’ App’x Vol. 41 (updated email invite with date and time of meeting, Oct. 1, 2003, from 3:00-6:15 PM). In any case, BNSF’s decision not to pursue a mileage-based surcharge in Spring 2003 was driven by internal analysis, including a meeting in March where the team identified specific obstacles at that time. See DA4261 at 4281, Defs.’ App’x Vol. 11 (Mar. 31, 2003, BNSF PowerPoint presentation on Proposed Fuel Surcharges Rate Methodology).

113

e) April 30, 2003, NEMC Meeting Defendants’ CMOs and CMOs-in-waiting also attended a late April 2003 meeting of the AAR’s Network Efficiency Management Committee (“NEMC”). See Pls.’ Opp’n at 24-25. According to plaintiffs, BNSF decided during this meeting to change its FSC to match up with UP and CSX. See id.; PX84 at 1, Pls.’ App’x Vol. 2 (Apr. 30, 2003, internal BNSF email from Kyei to Nick Murray). Plaintiffs also point out that BNSF and NS, at this time, both adopted policies to expand FSC application, requiring high level approval for exceptions. Pls.’ Opp’n at 24-25; PX84 at 1, Pls.’ App’x Vol. 2; PX270 at ‘523, Pls.’ App’x Vol. 4 (Apr. 29, 2003, internal NS email from Glennon to Seale).

Plaintiffs’ suggestion that this NEMC meeting influenced key decisions made by defendants is not supported by the record. As an initial matter, plaintiffs have no evidence of discussions or notes from those meetings showing that defendants engaged in relevant discussions there. Thus, plaintiffs rely on the coincidental timing alone. Cf. Text Messaging, 782 F.3d at 878 (where plaintiffs did not have evidence of “simultaneous or near-simultaneous price increase[s]” or what was exchanged at the relevant meetings, “there is no basis for an inference that they were using the meetings to plot price[] increases”).

Moreover, the emails plaintiffs cite do not evince the policy changes that plaintiffs contend occurred contemporaneous to the meeting. Regarding BNSF’s apparent policy change to expand its coverage of FSCs, the email cited merely states that “participation rates will be increased as opportunities present thems[elve]s by price authority.” PX84 at 1, Pls.’ App’x Vol. 2. In other words, BNSF stated that participation rates would naturally increase as pricing authorities were revised over time—not that BNSF changed its policies. See supra Part III.C.3 (describing BNSF’s approach to broad application of FSCs since 2002). While BNSF did

114

change both its FSC policies (with respect to requiring high level approvals for exceptions) and FSC formula in this timeframe, both of those decisions were supported by extensive internal analyses. See supra Parts III.C.3, III.D.2(c); DA4019 at 4019-20, Defs.’ App’x Vol. 9.

Regarding NS’s apparent policy changes, plaintiffs, at best, misconstrue or, at worst, misrepresent the email they cite as support to pin the prompt for the change on the NEMC meeting. The day before the NEMC meeting, Glennon sent an email to Seale recommending that NS not change its FSC at that time and concluded with a caveat that “the group suggests the adoption of the following changes to better position NS to accommodate a FSC change if and when the decision is made”—including only there the mention of a requirement that FSCs be imposed as a standard condition unless Seale approves otherwise. PX270 at ‘522-23 (emphasis added), Pls.’ App’x Vol. 4. Simply raising an idea to consider after a certain hypothetical occurs is far from “implement[ation of] policies to facilitate FSC adoption.” Pls.’ Opp’n at 25. NS made no changes to either its FSC or its policy in spring of 2003. Plaintiffs also do not point to any other change in FSC formula or policy that was decided around the time of the meeting. BNSF’s unilateral decision to adopt a unique FSC formula and enforce it broadly does not evince a conspiracy, despite it occurring close in time to one of defendants’ several trade association meetings.

f) May 14, 2003, NS-UP Alliance Meeting Plaintiffs repeatedly reference a May 14, 2003, meeting among two of the defendants, NS and UP, where the notes reflect comments plaintiffs characterize as suggestive of conspiracy. See Pls.’ Opp’n at 26, 102. Handwritten notes from that meeting from an unknown author state that “uniform application across the industry would be helpful,” PX460 at ‘652, Pls.’ App’x Vol. 7, and the minutes from that meeting contain the statement that “[c]onsensus was reached that it

115

would be a positive outcome if all roads had the same process in the eyes of our customers.” PX241 at ‘361, Pls.’ App’x Vol. 4 (Minutes: NS/UP Alliance Meeting); DA1594 at DA1594-96, Defs.’ App’x Vol. 2 (same); DA5216 at 5218, Defs.’ App’x Vol. 14 (same). The minutes also reflect that “[t]here was some question as to whether or not the industry can work together on this opportunity probably through the AAR, or do roads have to handle the opportunity on a bilateral basis. Follow up: Jack Koraleski/Mike Hemmer, NS – Don Seale/Greg Summy.” DA5216 at 5218, Defs.’ App’x Vol. 14; PX241 at ‘361, Pls.’ App’x Vol. 4. Koraleski and Hemmer were UP’s CMO and General Counsel, respectively, at the time. See Defs.’ Mem. at 50. Seale and Summy were NS’s SVP for Marketing and in-house counsel, respectively. Id. Defendants argue that, though these exhibits need not be excluded for irrelevance under Rule 403, they are after the period of relevant FSC changes and therefore not probative of any conspiracy. Defs.’ Reply at 20.

As previously explained, discussions of pricing alone—without evidence of an agreement or evidence of those discussions’ impact on defendants’ conduct—are not problematic, especially given that, here, defendants were discussing legitimate complaints from customers about the variation in formulas across the railroads. See Text Messaging, 782 F.3d at 878; see e.g., DA1454 at 1455, Defs.’ App’x Vol. 2 (Apr. 10, 2003, email from shipper to Travis Vanderpool, UP, expressing frustration with “the introduction of individual initiatives by each carrier, large and small alike, which not only applies different percentages but indexes as standards for applying changes in the surcharges”); see also supra Part III.D.1; Defs.’ Mem. at 50. The cited language itself does not evince defendants coming to an agreement on FSCs, at most reflecting acknowledgement of customer desire for defendants to take similar approaches in the “process” of applying FSCs, see supra note 12, and no defendant changed its FSC after this

116

meeting, so plaintiffs are unable to show how these pricing discussions led to any price increases or other conduct by defendants. UP had made its public announcement about its FSC formula five days prior to the meeting; BNSF made its public announcement one week before, and CSX was two months before. See supra Part III.C.2. NS, of course, did not make any change until the very end of 2003. See id. Even under plaintiffs’ theory of the alleged conspiracy, the bulk of the purported conspiratorial conduct occurred before this purportedly consequential meeting.

To the extent that plaintiffs try to suggest that this meeting encouraged NS to change its FSC, the substantial time gap between the meeting and NS’s decision undercuts any inference that this meeting had such an influence. See Text Messaging, 782 F.3d at 878 (stating that the “evidence would be more compelling if” price increases were “the immediate sequel to any of these meetings”). Moreover, because NS had already been evaluating a change to its FSC as of April 2003, plaintiffs cannot even demonstrate that the meeting instigated NS to look into adopting a formula more uniform to that of the other defendants. See supra Parts III.C.2; III.D.2(d); PX262 at ‘161, Pls.’ App’x Vol. 4 (Apr. 14, 2003, internal NS email from Neikirk to Bayrer and Fox, suggesting “it may be worthwhile to take a look at improving the effectiveness of our fuel surcharge”).

Plaintiffs next attempt to connect this meeting to defendants’ alleged conspiracy to “maintain[] and grow[] fuel-surcharge coverage . . . by agreeing to eliminate any discounts or waivers.” Pls.’ Opp’n at 99. No defendant, however, changed their FSC coverage policy around this time. BNSF had already imposed a universal coverage policy with high-level requirements for exceptions the month before, and UP, by plaintiffs’ own account, did not adopt something similar until June. See supra Part III.C.3. NS and CSX did not adopt similar coverage policies

117

until subsequent years. Id. Any connection to these decisions is too diffuse to aid any inference of conspiracy.

Additionally, plaintiffs point to statements made prior to the May 14, 2003, meeting to imply that NS intended to engage in conspiratorial discussions there. Pls.’ Opp’n at 102. Commenting on a recent discount employed by BNSF, Seale wrote, “regarding our conversation with UP this week, note below that BNSF has now announced a rollback of its FSC to 2% from 5% even though their formula calls for 5%. The industry is not looking good on this issue and a stable approach has been the best approach.” PX303 at ‘321, Pls.’ App’x Vol. 5 (May 12, 2003, internal NS email from Seale to Prillaman). Plaintiffs surmise that “not a single defendant ever did [such] a programmatic reduction again of the type that BNSF did” because “consensus was reached” at the subsequent meeting that such diversions from the conspiracy would be unwise. MSJ Hr’g Tr. at 95:9-97:15. Plaintiffs’ speculation appears unfounded. The internal NS email does not suggest any coordination but merely comments on the state of the industry. BNSF did not attend the subsequent meeting with NS, nor the next alliance meeting that plaintiffs cite in June. Importantly, NS took no action following the meeting themselves to suggest any conspiratorial discussion occurred or that any pricing discussion shaped NS’s strategy. Without anything more, that email and the occurrence of the meeting do not move the ball for plaintiffs.

Finally, plaintiffs suggest that the intermodal conspiracy may have been formed at this meeting because the leaders of intermodal divisions attended. Pls.’ Opp’n at 124 (citing PX460 at ‘652, Pls.’ App’x Vol. 7 (May 14, 2003, handwritten notes from NS-UP Alliance meeting). Plaintiffs have provided no evidence of any discussions of intermodal traffic, however. See also infra Part III.F. Again, the mere presence of defendants in the same room does not imply misconduct.

118

g) June 3, 2003, CSX-UP Alliance Meeting Plaintiffs also reference a June 3, 2003, meeting among CSX and UP for which handwritten notes by an unidentified author state, “Fuel surcharge- GIFTOS, CSX has not seen outcry that UP has,” PX474 at ‘088, Pls.’ App’x Vol. 7 (June 3, 2003, handwritten notes for alliance meeting); DX1589 at 1591, Defs.’ App’x Vol. 2 (same), and the meeting minutes include the statement, “Fuel surcharges: discussion of customer reaction to our current processes” under “State of the Industry,” PX167 at ‘417, Pls.’ App’x Vol. 3 (June 3, 2003, CSX- UP Alliance Meeting minutes). See Pls.’ Opp’n at 99-100. The handwritten notes also suggested that “both roads[, which] start at a $23 trigger[, m]ay have to revisit $23 if [prices] stay high.” PX474 at ‘088, Pls.’ App’x Vol. 7; DX1589 at 1591, Defs.’ App’x Vol. 2. Plaintiffs consequently conclude that CSX and UP must have discussed FSCs and that UP’s decision not to change its formula subsequently—despite acknowledging a need to revisit trigger prices if fuel prices stay high—“corroborates its participation in the conspiracy.” Pls.’ Opp’n at 100.

Defendants do not argue that evidence about this meeting needs to be excluded under Rule 403 but urge that because the meeting occurred after the relevant period of changes, no meaningful inferences about its effect on defendants’ conduct can be drawn. Defs.’ Reply at 20. Indeed, the meeting occurred three months after the supposed start of the conspiracy and a month after UP changed its formula, and no defendant changed their FSC after the meeting. Moreover, the meeting notes are unidentifiable, exhibiting no indicia of formality. Finally, the comment that both CSX and UP might have to consider raising their trigger price if prices stay high is a simple, straightforward observation that market forces may require defendants to adapt pricing over time. It does not suggest that the two firms agreed to move their prices in tandem or to

119

maintain the same formula. Plaintiffs strain too hard to infer a conspiracy from a stray comment about potentially changing trigger prices if fuel prices stay high or go higher.

h) July 30, 2003, BNSF-NS Meeting Lastly, plaintiffs point to a meeting on July 30, 2003, between BNSF and NS where the two companies proposed a “discussion of a potential industry position on fuel surcharge provisions at next N.E.M.C. meeting” that their CMOs, Lanigan and Seale, agreed to lead. DA4420 at 4424, Defs.’ App’x Vol. 11 (July 30, 2003, BNSF/NS Joint Marketing Meeting); PX243 at ‘424, Pls.’ App’x Vol. 4 (same). Plaintiffs interpret the statement as indicative of a conspiracy. Pls.’ Opp’n at 102-03. Defendants do not argue that this meeting is wholly irrelevant but merely that plaintiffs’ desired inference makes no sense. See Defs.’ Reply at 30- 31. Under scrutiny, little can be drawn from this comment about a proposed discussion topic at an upcoming NEMC meeting.

First, a statement about a “potential” future “industry position” suggests that none was extant already—four months into the conspiracy after three of the four defendants changed their FSC pricing. The statement in that sense undermines plaintiffs’ alleged conspiracy.

Second, BNSF CMO Lanigan’s testimony about the July 30, 2003, meeting forecloses any inference that conspiratorial discussion occurred there. Lanigan has testified that, as far as he could remember, the discussion at this meeting was “around customer feedback on ease or lack of ease of doing business with the industry with regard to fuel surcharges” because of defendants’ divergent approaches “and whether or not it was a topic that we might be able to discuss at NEMC.” PX610 at 140, Pls.’ App’x Vol. 10A (Lanigan, BNSF, Dep. at 140:7-19). Lanigan resisted that they reached any formal resolution on the matter, determining merely that “it was a topic worthy of introducing at the next meeting.” Id. at 142:1-20. He specified that

120

customers who brought up the desire for an industry standard may have been interline or local in nature but that it mainly “arose because of interline concerns” and “the discussions were all around interline and interline efficiencies and issues.” Id. at 143:21-144:20. Therefore, while the evidence may not be excludable under section 10706 (which only excludes evidence solely about interline traffic), the motivation for the discussion appears proper rather than suspicious.

Third, any discussion about FSCs that did occur at the July 30, 2003, meeting lacks probative value because plaintiffs cannot connect it to any probative follow-on discussion, evidence of an agreement, or even any pricing or policy decision made by defendants around the same time period. In fact, defendants offer affirmative evidence that the future conspiratorial discussions contemplated by plaintiffs’ interpretation of the ambiguous statement from the meeting notes did not occur. See DA5914 at 5914-5932, Defs.’ App’x Vol. 16 (Sep. 11, 2003, minutes from subsequent NEMC meeting lacking any indication of discussion about FSCs); DA3099 at 3132, Defs.’ App’x Vol. 5 (Lanigan, BNSF, Dep. at 147:16-25) (testifying that he and Seale did not lead a discussion on FSCs at the next NEMC meeting); PX596 at 226-29, Pls.’ App’x Vol. 10A (Seale, NS, Dep. at 226:1-229:22, explaining that he may have had discussions with other railroads about FSCs but they would have been about interline traffic).

* * *

In sum, none of the meetings relied upon by plaintiffs aid them in making plausible an inference of conspiracy. Regarding the first five meetings, all of which occurred in April 2003, plaintiffs have no evidence of what defendants discussed there. See supra Parts III.E.3(a)-(e). They rely solely on conjecture about apparently coincidental timing to link those meetings to defendants’ decisions, but as explained, these timelines do not add up. Regarding the remainder of the meetings, which occurred mid-May and later, those are “the opposite in every way,”

121

Defs.’ Reply at 20, of what is expected in a conspiracy—“conversations implying that later uniformity might prove desirable” followed by “uniform behavior among competitors,” id. (quoting Brown v. Pro Football, Inc., 518 U.S. 231, 241 (1996)). See supra Parts III.E.3(f)-(h). Instead, the comments from those meetings reflect a desire for future uniformity during a period when plaintiffs contend defendants’ practices were already uniform, not directly followed by any greater uniformity or meaningful changes. Such meetings do not evince conspiracy. 37 4. Evidence of Defendants’ Sharing of Information and Monitoring Plaintiffs point to additional communications among defendants where they shared pricing information or coverage rates and discussed FSCs, suggesting they were monitoring one another to ensure adherence to the conspiracy. See, e.g., Pls.’ Opp’n at 45. Indeed, monitoring of one another’s conduct may enable the imposition of “disciplinary measures on the ‘cheaters’” and discourage defendants from straying from an agreement. Kleen Prods., 910 F.3d at 937. Yet, again, “to survive summary judgment,” plaintiffs must provide “[e]vidence that the exchanges of information had an impact on pricing decisions,” so mere “evidence of sporadic exchanges of shop talk among field sales representatives who lack pricing authority is insufficient.” Baby Food, 166 F.3d at 125 (emphasis added). Likewise, “[a]lthough the possession of competitor price lists is consistent with conspiracy, it does not, at least in itself, tend to exclude legitimate competitive behavior.” In re Citric Acid Litig., 191 F.3d 1090, 1103 (9th Cir. 1999). That is because “it makes common sense to obtain as much information as

37 Plaintiffs also briefly reference a BNSF-UP meeting regarding a jointly owned coal line in August 2004, two months before UP adopted its coal-specific FSCs. Pls.’ Opp’n at 120. This meeting, however, had obviously legitimate purposes considering that the coal line was jointly owned, and as defendants point out, the meeting also included personnel from the AAR, coal shippers, and the STB—“an unimaginable locus for conspiracy.” Defs.’ Reply 19 n.11; PX823 at ‘625, Pls.’ App’x Vol. 14B, ECF No. 1203-5 (sealed) (Oct. 26, 2004, UP CEO expense report reflecting August 2, 2004, meeting with coal shipper Peabody and August 3, 2004, “Powder River Basin train trip with BNSF, STB, Peabody, AAR, and Westar representatives”); PX821 at ‘694-95, id. (undated agenda for trip lacking any reference to fuel surcharge discussion and reflecting mix of shippers attending); PX822 at ‘203, id. (July 20, 2004, UP internal email from Linda Brandi to Richard Vasy et al., identifying shipper attendees).

122

possible of the pricing policies and marketing strategy of one’s competitors.” Baby Food, 166 F.3d at 126. When the pricing information is obtained from a source other than the direct competitor itself, that is particularly suggestive of fair competition and not conspiracy. See id.; Chocolate, 801 F.3d at 408.

Plaintiffs cite several communications between BNSF and UP involving FSCs. Pls.’

Opp’n at 45 (citing COF ¶¶ 327, 364-65). One such communication is an internal UP email from Judith Blecha to Randy Davison and others on April 3, 2003, stating that “Jim Arnett from BNSF called and asked if I “had heard anyone discussing charging a fuel surcharge on miles.” PX469 at ‘852, Pls.’ App’x Vol. 7. Such a low-level query is quintessential “shop talk” that, unconnected from any specific subsequent action by defendants, lacks any inferential value. Baby Food, 166 F.3d at 125. Neither BNSF nor UP adopted a mileage-based fuel surcharge until two years later.

Another communication is an internal BNSF email, on July 27, 2005, that plaintiffs interpret as referencing a discussion with UP about their FSCs: “Tom [Hund] and I met with Matt this afternoon regarding FSC coverage (Nick and Sam presented our data). This was precipitated by the fact that UP reported higher FSC coverage than we did for Q2.” PX96 at ‘372, Pls.’ App’x Vol. 2 (July 27, 2005, email from Lanigan to David Garin et al.). This email does not suggest any suspicious conversation for at least two reasons: (1) the email and referenced meeting is dated in July 2005, over two years into the purported conspiracy and not near any relevant changes in FSC formula or policy for either BNSF or UP, see supra Parts I.B, III.C, and (2) the meeting Lanigan references appears to be an internal one, as Matt refers to BNSF CEO Matt Rose, and the information on UP coverage had been publicly reported on an earnings call the week before, DA 7835 at 7847, Defs.’ App’x Vol. 18 (July 21, 2005, Q2 2005

123

Union Pacific Earnings Conference Call). The other extraneous communications that plaintiffs cite are similarly inapplicable. 38 Plaintiffs next reference apparent communications between BNSF and another defendant, CSX. Pls.’ Opp’n at 45, 124. In particular, plaintiffs emphasize that defendants were monitoring one another’s FSC coverage to ensure everyone was working toward 100%, perpetuating the conspiracy. See id. at 45. First, plaintiffs reference an internal CSX email sharing what BNSF had told them earlier: “BNSF stated that 82% of their carload and [i]ntermodal business either had fuel surcharge or an escalator on the revenue.” PX169 at ‘476, Pls.’ App’x Vol. 3 (Oct. 6, 2003, internal CSX email from Couch to Eliasson); PX664 at ‘627, Pls.’ App’x Vol. 10B (same). This email is more probative than plaintiffs’ other emails because the substance reflects a communication between defendants about specific FSC-related information. Its value overall is nonetheless limited: Because BNSF aggregated its freight traffic subject to both FSCs and the RCAF escalators, CSX could not use this information to monitor BNSF’s commitment to the alleged conspiracy. Plaintiffs have not alleged that defendants conspired to use RCAF escalators as a pricing tool, only that they conspired to use FSCs, and this data point tells CSX little about how much BNSF is using FSCs specifically. Further, BNSF publicly shared the same

38 As part of their narrative that defendants were in constant communication with one another, plaintiffs refer to a letter, dated March 10, 2003, from BNSF to UP discussing potential line-sharing options. Pls.’ Opp’n at 27 n.11; PX522 at ‘267, Pls.’ App’x Vol. 7 (Mar. 10, 2003, letter from Rose, BNSF CEO, to Ike Evans, UP COO). Such invitation for a collaboration has nothing to do with FSCs, however, and plaintiffs do not even try to make that connection. Plaintiffs next cite an internal email reacting to an estimate of UP’s FSC coverage, but, again, that is hardly suspicious considering that the estimate appears to have been derived from a Bear Sterns analyst report. See Pls.’ COF ¶ 364; PX413 at ‘429, Pls.’ App’x Vol. 7 (Jan. 15, 2004, internal UP email from Adams to Koraleski); DA26028 at 26029-30, 26035, Defs.’ App’x Vol. 41 (Jan. 16, 2004, internal UP email from Koraleski to Sovey that Sovey then forwarded to Adams, attaching the report). Finally, plaintiffs cite a communication that is not between BNSF and UP at all but rather an email from a shipper to UP, referencing BNSF’s pricing plans, likely as competitive leverage, which is completely inapposite, suggesting only normal competitive market conditions. See Pls.’ COF ¶ 364; PX508 at ‘813-14, Pls.’ App’x Vol. 7 (Apr. 29, 2003, email from shipper representative, Wilma VanWetten, to Dennis Miller, UP). In fact, UP reacted, internally, to this customer’s email with concern about “a downward-pressure environment” and “fuel surcharge $ [] melting away at the competition,” which undercuts any inference of conspiracy. See id. at ‘813 (Apr. 29, 2003, internal UP email from Toy to Greg Barbe).

124

information—in fact, even more FSC-specific information—later that month, indicating that they did not view this information as competitively sensitive. See DA25924 at 25937, Defs.’ App’x Vol. 41 (Oct. 21, 2003, BNSF Q3 2003 Earnings Call Tr., stating, “In an over all [sic] basis 80% of revenue covered by some form of fuel charge which includes either R[CAF] or a straight contractual fuel surcharge and th[e]n that coverage gives us about 60% over all [sic] coverage on fuel surcharge because there is different puts and takes on contracts and different coverage levels and those sorts of things”); see also DA26296 at 26298, Defs.’ App’x Vol. 42, ECF No. 1245-2 (sealed) (Lanigan, BNSF, Dep. at 212:6-9, “Q. Can you give me any business reason why you would be sharing with CSX your internal coverage participation ratios for your fuel surcharge program? A. It’s public data.”).

Second, plaintiffs cite an internal UP email exchange about a discussion with NS in early 2005. Pls.’ Opp’n at 45. A Senior Business Director, Royce Fink, emailed another colleague that he “spoke to NS this morning and they said that 90% of all their agreements/contracts are subject to their fuel surcharge and is non-negotiable. Their fuel cost for 2004 was $450 million vs ours at $1.8 billion. There are some exceptions but none within their food and consumer group.” PX851 at ‘378, Pls.’ App’x Vol. 15 (Feb. 9, 2005, internal UP email from Royce Fisk to Kari Kirchhoefer). This email has probative value, as these statistics focus on FSCs exclusively, but exchanges about pricing—especially among those “who lack pricing authority”—are not enough if they cannot be connected to pricing decisions themselves. Baby Food, 166 F.3d at 125. Neither UP nor NS made any changes to their formulas or policies in 2005. Further, a single email reflecting an exchange of information that may largely be public already does not establish regular monitoring among defendants.

125

In short, plaintiffs lack two components to establish meaningful information sharing and monitoring behavior among defendants. First, they have not demonstrated that defendants possessed nonpublic, advanced pricing information about their competitors that was connected in any way to their decisions. See Baby Food, 166 F.3d at 126; Chocolate, 801 F.3d at 407-09. Plus, any communications that did occur were between middle-level managers, not executives. See Baby Food, 166 F.3d at 125; Chocolate, 801 F.3d at 407-09. In Chocolate, the Third Circuit held that defendants’ “gathering [of] price information of competitors” reflecting impending but not yet public price increases did not support an inference of conspiracy where there was no strong evidence it came directly from competitors, “much less their upper-level executives” and where that information did “not reveal pricing plans dependent on others following.” 801 F.3d at 407-09. The same is true here. Defendants’ gathering of pricing information is more consistent with normal competitive conduct than conspiracy.

Second, plaintiffs have not gestured to a single enforcement action—even implicit—

taken by any defendant to ensure a purported co-conspirator stayed in line or to punish a defendant for straying. They do not evince even subtle pressure or cajoling among defendants, despite their supposed monitoring of one another. “With no punishment, or even a mechanism to punish, the inference tends toward no agreement.” Kleen Prods., 910 F.3d at 937 (affirming grant of summary judgment and noting that plaintiffs “propose two possible mechanisms for enforcement, but they have not pointed to any evidence indicating that either one was used”); Domestic Airline Travel, 691 F. Supp. 3d at 205-07 (citing these same principles but noting a dispute of fact as to whether members of the investment community acted to discipline Delta for not maintaining low enough capacity levels).

126

In sum, plaintiffs’ evidence about communications among defendants does not tend to exclude an inference of independent action.

5. Evidence of Facilitating Actions Using AAR Plaintiffs argue that defendants used the trade association AAR to further their conspiracy. In particular, plaintiffs point to the AAR’s adoption in December 2003 of a new fuel cost index, the All Inclusive Index Less Fuel (“AIILF”), which replaced a prior cost recovery index that included fuel, i.e., the “Rail Cost Adjustment Factor” or “RCAF.” Pls.’ Opp’n at 2, 30-31. Such indices reflect the cost of certain inputs at any point in time, such that any pricing formula keyed to that index will automatically adjust to reflect market fluctuations, requiring fewer manual changes. See generally Defs.’ Mem. at 10. Plaintiffs contend that defendants pursued the AIILF for the “express purpose of facilitating their industry-wide application of FSCs.” Pls.’ Opp’n at 2. Plaintiffs also reference other discussions at AAR meetings apparently suggestive of collusion.

A “facilitating practice” is “an activity that makes it easier for parties to coordinate price, output, or other behavior in an anticompetitive way.” AREEDA & HOVENKAMP, supra, ¶ 1407b; Domestic Airline Travel, 691 F. Supp. 3d at 208. Plaintiffs argue that the adoption of AIILF was a facilitating practice because the use of an index without fuel was more compatible with the imposition of FSCs. See Pls.’ Opp’n at 2, 31-32. Use of the RCAF index along with FSCs in the pricing formula gave an impression of double recovery because RCAF itself also took fuel prices into account. See id. In contrast, use of the non-fuel AIILF index along with FSCs did not create such an impression and lent FSCs a “veneer of legitimacy” by making them seem ordinary and compatible with industry standards. Id. at 31. Since FSCs were more profitable for defendants than the use of RCAF without FSCs, plaintiffs reason that the AIILF was advantageous to defendants. See PX327 at ‘903, Pls.’ App’x Vol. 5 (Feb. 13, 2004, internal NS email from Alby

127

Pfeifer to Glennon et al., explaining that using the RCAF index without surcharges is less profitable for the railroads than having an FSC). Plaintiffs contend that to lend the AIILF even more credibility, defendants pressured the AAR to publish the index on their website in November 2004. Pls.’ Opp’n at 32-33; PX157 at ‘767, Pls.’ App’x Vol. 3 (Nov. 11, 2004, NEMC meeting agenda, stating “UP is recommending an agenda item to determine if there is additional carrier interest in having the index publish[ed]”); PX629 at 269, Pls.’ App’x Vol. 10B (Koraleski, UP CMO and former Chair of AAR, Dep. at 269:1-12, acknowledging that agenda item); PX415 at 6, Pls.’ App’x Vol. 7 (undated AAILF presentation, stating, “Recommendation – AAR publish the [AIILF] on their Web site. . . . Facilitates and supports carrier initiatives for cost recovery”).

The establishment of a new index more conducive to the use of FSCs may appear at first glance to be a facilitating practice suggestive of a conspiracy, but the circumstances around the adoption of this index foreclose any such inference. First, the establishment of the AIILF was not a joint project that all the defendants pursued together as would be expected if the purpose was to further a common conspiratorial goal. To the contrary, the evidence shows that BNSF was the only defendant that pushed for the change. As defendants explain, BNSF had been unsatisfied with RCAF since at least early 2003, finding that its fuel component was a “combination of a typically inaccurate forecast of fuel prices and the adjustment to account for the forecast error from two quarters prior. The forecast [could ]not be very accurate because fuel prices are so volatile by nature.” DA7765 at 7766, Defs.’ App’x Vol. 18 (Nov. 7, 2003, internal BNSF email from Feldman to Hund et al., identifying that RCAF is particularly problematic during periods of dramatic change such as in the early 2000s); see also Defs.’ Reply at 52-53; DA4113, Defs.’ App’x Vol. 10 (Mar. 11, 2003, internal BNSF memo from Cheri Carlson to

128

Greg Stengem, stating, “Marketing’s analysis shows that on contracts written with RCAF escalators, BNSF does not recover the costs of fuel in volatile times”); DA7917 at 7917-19, Defs.’ App’x Vol. 19 (July 21, 2003, internal BNSF memo from Kyei to Rose, expressing that costs are generally not fully recovered when using an RCAF escalator). BNSF therefore largely abandoned RCAF in favor of using FSCs but wanted a pricing index that could be reasonably applied in conjunction with the surcharges.

BNSF worked alone to get the AIILF adopted. BNSF asked AAR leadership to consider adopting a different index in October 2003 without discussing this request with any other railroads. See DA15859 at 15886-87, 15889-90, Defs.’ App’x Vol. 27 (Kyei, BNSF, Dep. at 135:8-138:4, 146:13-151:9). AAR leadership conversed solely with BNSF when developing the index. See DA7807 at 7809, Defs.’ App’x Vol. 18 (Dec. 2, 2003, email from Clyde Crimmel, AAR, to Kyei, BNSF, sharing with only BNSF a draft of the AIILF in advance of a conference call); DA7673 at 7674, id. (Nov. 24, 2003, email from Craig Rockey, AAR, to Kyei, BNSF, discussing the “Proposed New Cost Index”). 39 The index was discussed and approved at an AAR meeting in December 2003 with the defendants present, though no formal vote was ever taken. See DA15906 at 15934, Defs.’ App’x Vol. 27 (Deposition of Matthew Rose, BNSF CEO (“Rose, BNSF, Dep.”) at 205:25-206:19, explaining that the index was adopted and discussed at a meeting but there was no action item or vote at the AAR meeting—it was just letting “the Board know that there was going to be another index out there”); DA12531 at 12536-39, Defs.’ App’x Vol. 24 (Deposition of Craig Rockey, AAR, at 105:24-106:8, 109:7-111:7, 119:14-

39 BNSF certainly perceived its effort to secure the new index as unilateral. DA7776 at 7777, Defs.’ App’x Vol. 18 (Mar. 15, 2005, internal BNSF memo from Kyei to Rose, stating, “In December 2003, you single-handedly got the A.A.R. to establish a non-fuel RCAF index, now called the All-Inclusive Index Less Fuel”). Even plaintiffs admit that BNSF took the lead. See Pls.’ COF ¶ 393 (noting that BNSF’s CEO “led discussions with the AAR in support of the creation of what became the AIILF”).

129

120:18, testifying that no formal vote was taken on the index at the AAR meeting, and although the AAR members had to generally approve, the other railroads did not request the index).

Publication of the AIILF on the AAR’s website, despite plaintiffs’ emphasis of this fact, see Pls.’ Opp’n at 32-33, 59, 99, is ultimately irrelevant. After its adoption, the AIILF was initially printed in quarterly publications. PX7 at ‘209, Pls.’ App’x Vol. 1 (AAR Railroad Cost Indexes). For ease and efficiency, UP recommended instead publishing the AIILF on the AAR website, which recommendation was adopted at the November 2004 NEMC meeting. See PX415 at 6, Pls.’ App’x Vol. 7 (undated AIILF presentation). The publication decision therefore does not represent concerted action any more than the decision to adopt the AIILF itself.

Second, the circumstances around the adoption of the AIILF were not secretive in any way, contrary to what would be expected of a facilitating activity to further an illegal conspiracy. The AAR’s purpose is to allow defendants to collaborate lawfully on issues facing the industry as a whole. By adopting the AIILF, they were doing just that, out in the open and alongside non- defendant railroads and AAR leadership. See Text Messaging, 782 F.3d at 878 (noting an expert witness’s testimony admitting that the presence of non-defendants at meetings makes it less likely that defendants were engaging in illegal collusion there); see e.g., DA7684, Defs.’ App’x Vol. 18 (Dec. 11, 2003, AAR Board of Dirs. Meeting minutes, listing invitees present, including leaders of non-defendant railroads, AAR leadership, and other related companies’ leaders). In other words, adopting an index of this kind was a proper use of the AAR forum and accompanied by the usual course of business surrounding AAR actions.

Third, adoption of the AIILF index was rather inconsequential for defendants’ FSC practices. For one, the AIILF did not replace RCAF to facilitate universal adherence to FSCs. Indeed, CSX continued using RCAF as an alternative to FSCs until the end of 2005. See

130

DA6925 at 6930, Defs.’ App’x Vol. 17 (June 1, 2005, CSXT Industrial & Agricultural Products Midyear Review presentation, conveying that RCAF(U) remained “acceptable as an alternative” to fuel surcharges); DA2672, Defs.’ App’x Vol. 4 (Dec. 21, 2005, internal CSX email from McIntyre to Gooden et al., announcing switch to using AIILF); supra Part III.C.3. Mainly BNSF used AIILF, which is unsurprising considering BNSF unilaterally pushed for its adoption. See BNSF Mem. at 10. But see Defs.’ Reply at 56-57 (noting that BNSF’s use, though immediate, was not widespread).

Other defendants rarely used the new index or took several more years to embrace that option. Plaintiffs can only point to a single example of NS’s use of AIILF in a 2007 contract, see Pls.’ Opp’n at 33 (citing COF ¶ 417); see also PX919 at ‘566, Pls.’ App’x Vol. 16A, ECF No. 1204-2 (sealed), and sometimes when shippers mentioned AIILF to NS, NS employees were unaware of the new index, see, e.g., DA6068, Defs.’ App’x Vol. 16 (June 30, 2006, email from NS employee to shipper, stating, “I am not immediately familiar with the all inclusive index without fuel and cannot offer a comparison at the moment. I’ll see what I can find out”). 40 Similarly, although UP made AIILF its default index in September 2004, after an internal discussion, only 3.3% of its total shipments had an AIILF provision. See DA8263 at 8264-65, Defs.’ App’x Vol. 20 (Sep. 22, 2004, internal UP memo endorsing use of the AIILF); DA8261 at 8262, id. (Apr. 27, 2004, internal UP email forwarding an inquiry about AIILF for discussion); DA1101 at 1150, Defs.’ App’x Vol. 2 (Report of Dr. Carlton, Defs.’ Expert, Tbl. 10, showing less than 8% of UP’s total shipments by revenue used AIILF in any given year during the alleged

40 See also DA5827, Defs.’ App’x Vol. 16 (July 18, 2005, email from shipper to NS employees, asking about using AIILF, and NS employee forwarding a question to another NS employee about what AIILF is); DA8228 at 8229, Defs.’ App’x Vol. 20 (Sep. 7, 2006, internal NS email from Randy McFarland to Randy Carter et al., stating, “EC is requesting the application of the AIILF . . . vs. the 85% RCAF. I don’t believe our CBG has experience in this index; however, perhaps our Merchandise brethren have knowledge on this escalation”).

131

conspiracy period). 41 This all reinforces the finding that the adoption of AIILF was not pursued collectively by defendants, nor was this index facilitative of the alleged conspiracy.

Plaintiffs argue the degree of defendants’ use does not matter because they still conspired to adopt it, see Pls.’ Opp’n at 33—but plaintiffs are wrong on two fronts. First, defendants did not act collectively to urge the adoption of the AIILF; as the evidence summarized above reveals, BNSF largely acted alone. Second, they conflate the conspiracy about FSCs with a conspiracy about the price index alone. As defendants explain, even a concerted effort to create the AIILF would not constitute a conspiracy: “Nobody claims any railroad agreed to use AIILF or change[d] any contracts to use it. Nor is AIILF a price charged to anyone.” Defs.’ Reply at 51- 52. Rather, the AIILF is only useful to plaintiffs if they can connect the index to the use of conspiratorial FSCs, and the fact that the index lacked practical import to defendants makes it meaningless for proving the conspiracy claimed. The “creation of AIILF” is not “any more consistent with a conspiracy on fuel surcharges than with independent behavior with respect to those same fuel surcharges.” Id. at 57.

Beyond adoption of the AIILF, plaintiffs cite to other allegedly facilitative communications and meetings before the AAR, but such evidence is not suggestive of the claimed conspiracy either. As an initial matter, the meetings that plaintiffs reference occurred in October and December of 2003, long after most of the defendants changed their FSC formulas in supposed formation of the conspiracy in the Spring of 2003. 42 Moreover, plaintiffs are unable to point to any notes or discussions showing that defendants actually used the AAR to coordinate

41 The UP contract using AIILF that plaintiffs point to is from December 2008—at the end of the conspiracy.

See PX918 at ‘732-33, Pls.’ App’x Vol. 15; Pls.’ Opp’n at 33 (citing COF ¶ 417). 42 The April 30, 2003, NEMC meeting was a subset of the AAR, but for the reasons explained in Part III.E.3(e), that meeting is not suggestive of conspiracy either.

132

FSCs or otherwise collude. For instance, plaintiffs emphasize an internal CSX email sent from in-house counsel to the company’s executives warning that “it would likely be an anti-trust violation to discuss/suggest setting an AIILF industry fuel surcharge formula program at NEMC. The most we can discuss is the best practices for interline accounting/settlement of fuel surcharges,” given that “[f]uel surcharge programs are independently initiated by individual carriers.” See PX172 at ‘643, Pls.’ App’x Vol. 3 (Nov. 1, 2004, internal CSX email from McIntyre to Eliasson et al.); Pls.’ Opp’n at 32; cf. PX612 at 179-80, Pls.’ App’x Vol. 10A (Deposition of Clarence Gooden, CSX, at 179:17-180:5, testifying that, to his recollection, lawyers at the AAR did not allow discussion of publication of AIILF). CSX seemingly followed counsel’s advice: Plaintiffs do not point to any notes or discussions showing that AAR meetings were used to coordinate FSCs.

Similarly, plaintiffs cite a presentation given by BNSF at an AAR meeting in November 2005, which shared that BNSF intended to expand aggressively its FSC recovery rate. See PX869 at ‘067, Pls.’ App’x Vol. 15; Pls.’ Opp’n at 34. The presentation did not suggest, however, that the intention was part of some kind of agreement, and plaintiffs offer no related notes or discussions suggesting the presentation spurred the formation of one. In any case, by 2005, BNSF had already long been pursuing growth of its FSC coverage, and no pricing or coverage decision was made by any defendant around this time period that could be linked to the presentation.

Plaintiffs’ final two pieces of evidence likewise lack any connection to conspiratorial discussions about FSCs. See Pls.’ Opp’n at 34. Plaintiffs point to an email from BNSF to AAR suggesting consideration of approaches to FSC management as a topic for an NEMC meeting on organizational staffing, but staffing choices have nothing to do with the substantive pricing and

133

strategies of FSCs themselves. See PX13 at ‘391, Pls.’ App’x Vol. 1 (Aug. 8, 2005, email from Linda Hurt, BNSF, to Rockey, AAR, et al.). Plus, even if a relevant discussion about FSCs did occur there, plaintiffs once again cannot tie it to relevant pricing decisions made by defendants, given that little change occurred after late 2005. Lastly, plaintiffs note a discussion of fuel hedging that was recorded in the minutes from the AAR Treasury/Finance Division meeting on November 14, 2004. See Pls.’ Opp’n at 34; PX221 at ‘051, Pls.’ App’x Vol. 4. As explained previously, fuel hedging is not a synonym for FSCs, so the mere mention of fuel hedging discussions does not support plaintiffs’ inference of a conspiracy, and again, no pertinent changes occurred following that meeting. 43 In short, the evidence does not demonstrate that defendants used the AAR to form, facilitate, or perpetuate any conspiracy.

6. Evidence of a Major Change in Business Practices Plaintiffs also assert that defendants significantly changed their business practices during the alleged conspiracy period, which they argue indicates a conspiracy. See Pls.’ Opp’n at 2, 87- 88. Plaintiffs describe a shift from fierce price competition and limited application of FSCs prior to the alleged conspiracy period to vast over-recovery of fuel costs and widespread application of FSCs when defendants started meeting about, discussing, and sharing information on FSCs during the alleged conspiracy period. See id.; see also id. at 14-15 (citing PX127 at ‘617, Pls.’ App’x Vol. 3 (May 2002, Enterprise Risk Assessment by Ernst & Young, prepared for BNSF that includes individual comment that BNSF has a “general inability to mitigate fuel cost

43 Plaintiffs cite to Judge Friedman’s encouraging comment at the motion-to-dismiss stage that “[s]hort of being in the boardroom at the meeting, it is hard for the Court to imagine how plaintiffs could more fulsomely allege that defendants entered into an agreement at the AAR meetings.” MDL I MTD Order (Direct Purchasers), 587 F. Supp. 2d at 34; Pls.’ Opp’n at 31. Obviously, however, even a “fulsome[] alleg[ation]” says nothing about whether plaintiffs have proven such allegations, and the fuller context available at the summary judgment stage demonstrates that the evidence simply does not back up the inferences plaintiffs seek to draw.

134

increases through fuel surcharges”), to support plaintiffs’ argument about pre-conspiracy period competition).

Plaintiffs are correct that a “departure from pre-conspiracy conduct” can “create an inference of a conspiracy,” but the shift in the industry’s business practices must be “radical” or “abrupt.” Chocolate, 801 F.3d at 410 (last two excerpts quoting Toys “R” Us, Inc. v. FTC, 221 F.3d 928, 935 (7th Cir. 2000)); Valspar Corp. v. E.I. Du Pont De Nemours & Co., 873 F.3d 185, 196 (3d Cir. 2017) (clarifying that a change in industry practices must be “more than just an uptick”). In Chocolate, the Third Circuit affirmed a grant of summary judgment to defendants, in part because, given the historical parallel pricing among defendants, no “abrupt shift in behavior [to] support a reasonable inference of conspiracy” was apparent. 801 F.3d at 410.

Plaintiffs cannot meet that standard. First, as explained, defendants’ FSC formulas moved away from one another during the conspiracy period with greater degrees of differentiation. See supra Part III.C.1. Defendants, prior to the conspiratorial period, had almost identical formulas, and during the first nine months of the alleged conspiracy period, they made disparate, small changes to these formulas as they assessed their profitability, their competitors’ public announcements of changes, and the fluctuating state of fuel pricing. See id.; supra Parts III.C.2-III.C.3; Defs.’ Mem. at 41-43. Even assuming that defendants’ FSCs during the 2003- 2007 period had remained as parallel as pre-conspiracy, “public parallel price increase announcements . . . ‘consistent with how this industry has historically operated’” do not “raise an inference of conspiracy.” Valspar, 873 F.3d at 196 (quoting Valspar Corp. v. E.I. Du Pont De Nemours, 152 F. Supp. 3d 234, 252 (D. Del. 2016)). In other words, little may be gleaned from the continuation of parallel pricing that was not “historically . . . uncommon.” Chocolate, 801 F.3d at 410. If anything, defendants here, like those in Chocolate, continued their prior trend,

135

making pricing decisions with their competitors’ movements in mind. See, e.g., DA3546 at 3547, Defs.’ App’x Vol. 5 (Sep. 15, 2000, notes from BNSF Carload Fuel Surcharge meeting recommending following NS’s pricing approach prior to the alleged conspiracy period). As the court in Chocolate also points out, “it is generally unremarkable for the pendulum in oligopolistic markets to swing from less to more interdependent and cooperative.” 801 F.3d at 410. Even if defendants’ conduct appeared more interdependent and parallel during the alleged conspiracy period compared to before, that would not necessarily tend to exclude the possibility of conscious parallelism borne of unilateral action.

Plaintiffs attempt to distinguish the instant case from Kleen Products, 910 F.3d at 931, where the Seventh Circuit affirmed summary judgment for the defendants. Pls.’ Opp’n at 59-60. In that case, containerboard companies were sued for conspiring to increase prices and reduce output from 2004 to 2010. The Kleen Products Court acknowledged a “fine line between agreement and . . . conscious parallelism,” but, after noting that the oligopolistic market structure made both collusion or “follow the leader” pricing easier, concluded the continuation of historical practice would not allow for an inference of conspiracy since the evidence reflected “only changed market conditions,” not a “shift in firm behavior.” 910 F.3d at 931, 935-36. Plaintiffs argue that defendants are different because they did make changes to their practices. Pls.’ Opp’n at 60. In fact, however, defendants here increased prices during the alleged conspiracy period in response to market conditions, as did the containerboard companies in Kleen Products, and the changes made to defendants’ practices were a continuation of, though somewhat more differentiated than, pricing that occurred pre-conspiracy. See Kleen Prods., 910 F.3d at 937-38. This is even less suggestive of conspiracy than the more exacting parallelism found to be insufficient proof of conspiracy in Kleen Products.

136

Second, regarding FSC coverage, plaintiffs cannot demonstrate a major change in conduct by defendants at a specific point in time. The expansion of FSCs occurred as an evolution, as contracts came up for renewal and as defendants adjusted their policies to commit more broadly to a program that was working well for them. See supra Part III.C.3. As previously explained, defendants made changes to their policies over the course of several years, and the increase in coverage rates was not parallel in terms of absolute levels nor rate of growth. See id. Plaintiffs therefore cannot demonstrate an “abrupt” change in defendants’ approach to coverage.

Plaintiffs point to Judge Friedman’s finding in the class certification stage “by a preponderance of the evidence that the fuel surcharge programs applied before the class period were nothing like the widespread and uniform application of standardized fuel surcharges during the class period. . . . The fuel surcharges that defendants put in place in the spring of 2003 were of a different breed.” MDL I Class Cert. I, 287 F.R.D. at 50; MSJ Hr’g Tr. at 75:1-20; 108:10- 109:15 (plaintiffs’ counsel referencing slides provided to the Court with this quote). As previously noted, see supra n.27, however, that finding was made in the limited context of concluding that an injury-in-fact would be “capable of proof at trial with evidence common to the class.” MDL I Class Cert. I, 287 F.R.D. at 43. Not only was the decision containing that finding vacated by the D.C. Circuit, see MDL I Class Cert. II, 725 F.3d 244, but that finding was made before evidentiary exclusions based on Section 10706 and before defendants fully briefed and demonstrated the alternative evidence. Upon the fuller record now available, the kind of significant shift suggestive of conspiracy, as artfully framed by plaintiffs in prior stages of this litigation, has simply fallen apart with close analysis of the factual evidence properly contextualized.

137

Third, plaintiffs cannot demonstrate an abrupt change in business practices based on defendants’ meetings and communications about FSCs starting in 2003. Plaintiffs contend that prior to spring 2003, defendants’ senior executives never discussed FSCs at meetings. Pls.’ Opp’n at 17-28, 108-09 (citing, e.g., PX596 at 107-09, Pls.’ App’x Vol. 10A (Seale, NS, Dep. at 107:12-109:5, testifying only that NS did not discuss FSCs with other railroads before their original adoption in 2000); PX631 at 53, Pls.’ App’x Vol. 10B (Deposition of David Goode, NS, at 53:11-20, not recalling conversations with counterparts about FSCs prior to 2003); PX630 at 145-46, id. (Moorman, NS, Dep. at 145:20-146:9, not recalling any discussions or coordination with other railroads on FSCs prior to 2003). Plaintiffs cite supplemental interrogatory responses from each defendant apparently affirming this conclusion. See Pls.’ COF ¶¶ 93-94 (citing PX644-PX647 No. 2, Pls.’ App’x Vol. 10B (defendants’ supplemental interrogatory responses)).

Plaintiffs’ broad statement is misleading. Though defendants’ evidence of pre-

conspiracy period discussions of FSCs is somewhat limited, defendants did discuss pricing with one another prior to the conspiracy, including sometimes about interline FSCs. See PX595 at 194-98, Pls.’ App’x Vol. 10A (Giftos, CSX, Dep. at 194:12-198:23, describing pre-conspiracy discussions of FSCs related to interline traffic); DA26023 at 26023-25, Defs.’ App’x Vol. 41 (Mar. 28, 2001, agenda for NS-UP Alliance Meeting, including discussion of “[i]ntermodal [j]oint [p]ricing,” with an action item to “[f]ollow up that UP and NS are together in joint rates); DA26053 at 26057, id. (Oct. 31, 2001, agenda for CSX-UP Alliance Meeting, including topic “[r]ate [s]implification”); see also MSJ Hr’g Tr. at 125:23-126:11 (defendants’ counsel noting that pre-conspiracy communications about FSCs for interline traffic would be excluded from the record). Defendants’ supplemental interrogatory responses also do not clearly stand for the principle that plaintiffs contend. While those responses did identify discrete meetings and

138

conversations during 2003 and later, those meetings do not appear to be so widespread to represent a sea change in defendants’ approach. See PX644-647 at Response No. 2, Pls.’ App’x Vol. 10B. Regardless, any “uptick,” Valspar, 873 F.3d at 196, in frequency of discussions about FSCs does not constitute the kind of “abrupt” change, Chocolate, 801 F.3d at 410, that tends to exclude the possibility of independent action. Volatile fuel costs continued to concern defendants, and discussions may have come up more frequently as defendants focused on addressing this market condition through FSCs.

Plaintiffs go on to criticize aspects of defendants’ meetings during the alleged conspiracy period, describing these meetings as not publicized and not properly documented, with no lawyers in attendance. See Pls.’ Opp’n at 28-29 (citing PX474 at ‘088, Pls.’ App’x Vol. 7, and PX167 at ‘147, Pls.’ App’x Vol. 3, to suggest that meeting notes were sanitized). Though this description certainly sounds suspicious and warranting further examination, such scrutiny shows only that plaintiffs’ description is misleading. Regarding publicity, many of the meetings plaintiffs cite did not occur in secret at all but rather were regular alliance meetings, committee meetings, or AAR meetings attended regularly by defendants and sometimes non-defendants. See supra Part III.E.3. The presence of lawyers is irrelevant and equally suggestive of conspiracy as not. Lawyers are often not present at business meetings, and their presence does not immunize defendants from engaging in conspiracy. Regarding meeting notes, plaintiffs suggest that official minutes were sanitized to omit any discussion of FSCs. Plaintiffs point to PX474, which contains anonymous handwritten notes from the June 3, 2003, CSX-Alliance meeting and mentions FSCs. PX474 at ‘088, Pls.’ App’x Vol. 7 (noting that UP is facing public outcry and both roads start at a $23 trigger). Exhibit PX167, which appears to be the official minutes from the meeting, by contrast, does not mention FSCs. PX167, Pls.’ App’x Vol. 3. This

139

discrepancy does not, however, suggest that defendants were trying to hide misconduct. A brief, sidebar conversation about FSCs could have been of interest to a particular individual employee while being of no moment to the broader group. In any event, plaintiffs’ evidence of meetings and discussions evince neither a meaningful change in business practices nor bear up under scrutiny as sufficiently suspicious behavior indicative of unlawful conspiracy.

7. Evidence of Acts Against Self-Interest Next, plaintiffs contend that defendants acted against their self-interest in imposing FSCs during the alleged conspiracy period in several ways. See Pls.’ Opp’n at 34-36, 45-46, 63-66, 106. While evidence that antitrust defendants acted “against their economic interests” may support an inference of conspiracy, Moundridge, 2009 WL 5385975 at *6; Chocolate, 801 F.3d at 398, none of the choices plaintiffs identify turn out to be contrary to defendants’ interests, no matter how vigorously plaintiffs urge this point.

First, plaintiffs argue generally that defendants acted against their own profit interests by maintaining parallel pricing structures instead of exploiting their competitive advantages on fuel costs and efficiency by differentiating their pricing. Pls.’ Opp’n at 64. Plaintiffs cite a table produced by one of their experts, Dr. Jeffrey Leitzinger, illustrating different measures of fuel efficiency across defendants. See id.; PX769 at Exh. B06, Pls.’ App’x Vol. 12 (Report of Dr. Leitzinger, Pls.’ Expert (“Leitzinger, Pls.’ Expert, Report”), at Exh. B06). Defendants’ fuel costs differed yet they nonetheless adopted similar pricing mechanisms. See PX769 at Exh. B06, id.; PX819 ¶¶ 320, 325, Pls.’ App’x Vol. 14B, ECF No. 1203-5 (sealed) (Report of Pls.’ Expert, Dr. Allen Jacobs (“Jacobs, Pls.’ Expert, Report”) ¶¶ 320, 325, with graph showing disparate fuel consumption by share of revenue and concluding that defendants “chose not to exploit competitive advantages that might have been realized by railroads with superior fuel efficiencies either at a system level or for business segments (e.g., for specific routes, movements, etc.)”).

140

As a general matter, adopting FSCs and insisting on broader coverage is not at all inconsistent with defendants’ interests in recovering fuel costs, regardless of that cost, and maximizing profit. As explained supra Part III.D, FSCs allowed defendants to secure more stability in light of volatile fuel prices in the early 2000s. More specifically with respect to competitive advantages here, conscious parallelism explains defendants’ actions just as much as conspiracy. As is common among oligopolists, defendants may have realized they could earn more profits and secure more stability by maintaining high FSCs, like their competitors, rather than lowering their prices to try to increase volume and creating a race to the bottom. See Defs.’ Reply at 11 (noting the certainty of higher profit margins may be preferable to pursuing uncertain gains in market share); Kleen Prods., 910 F.3d at 935-36 (explaining that rivals might not compete for greater market share where one “might think that it will reap greater profits if it imitates, rather than undermines, its peers’ price hikes”). In any case, defendants did differentiate their pricing to reap competitive advantages. BNSF and UP adopted coal-specific formulas, and BNSF adopted a mileage-specific formula before other defendants. See supra Part III.C.1. Defendants did not maintain absolute uniformity at the expense of competitive differentiation.

Second, despite BNSF’s ultimate adoption of a mileage-based formula, plaintiffs impugn BNSF’s earlier choices to twice forego that formula (in 2003 and 2004) before deciding in 2005 to adopt one, asserting that BNSF did so only to adhere to the conspiracy. See Pls.’ Opp’n at 63, 106; PX132 at ‘03, 016, ‘027-28, Pls.’ App’x Vol. 3 (email invite for Mar. 18, 2003 meeting, from Kyei to internal BNSF Fuel Surcharge Review Team with attachments, explaining that Lanigan had the idea of adopting a mileage-based FSC and asking the group to look into it); PX610 at 235, Pls.’ App’x Vol. 10A (Lanigan, BNSF, Dep. at 235:4-20); PX769 ¶¶ 188-91, Pls.’

141

App’x Vol. 12 (Leitzinger, Pls.’ Expert, Report ¶¶ 188-91, describing Lanigan’s analysis and BNSF’s choice not to pursue a mileage-based system despite economic evidence and customer requests otherwise). Contrary to the inference plaintiffs seek to draw, BNSF made its decisions after extensive analysis and consideration of the business impacts. See supra Part III.D.2. In 2003, during the first round of analysis, BNSF anticipated logistical hurdles, such as integration of collection and payment systems, and customer blowback if a mileage-based option were adopted at that point. See PX132 at ‘023, Pls.’ App’x Vol. 3 (Feb. 24, 2003, presentation on BNSF Fuel Surcharges Cost Recovery Improvements, identifying “[c]ustomer [i]ssues,” “[s]ystems” problems, and “[e]xceptions [m]anagement” challenges); DA4240 at 4248-50, Defs.’ App’x Vol. 10 (materials for Fuel Surcharge Review Team meeting on Mar. 28, 2003, with attached Mar. 17, 2003, internal BNSF email from Renita Lee to Kyei, identifying cons of adopting a mileage-based program, including problems with “integration of customer accounts payables systems” and issues with “current[] rating/collection systems”); DA4261 at 4281, Defs.’ App’x Vol. 11 (Mar. 31, 2003, BNSF Fuel Surcharges: Fuel Surcharge Review Team Recommendations presentation, noting a need to fix “systems issues” before adopting mileage- based program). 44 BNSF decisionmakers decided to focus on increasing FSC participation and not try to make a major program change at the same time. DA4240 at 4248-50, Defs.’ App’x Vol. 10.

44 See also DA4451 at 4468, Defs.’ App’x Vol. 12 (Apr. 14, 2003, BNSF Fuel Surcharges: Fuel Surcharge Review Team Recommendations presentation, summarizing issues and recommendations, recommending increasing participation rates as soon as possible and warning of challenges to a mileage-based program); DA4258 at 4258-59, Defs.’ App’x Vol. 10 (Mar. 28, 2003, internal BNSF email from Loretta Long to Kyei, covering “risks with implementing fuel surcharge based on miles,” including current rating/collection systems that do not support such rates; customer ease of doing business because of integration of customer accounts; usage of mileage tables not part of system; and current internal revenue reporting and rating/billing policies would require conversion for a mileage- based system); Defs.’ Mem. at 38-39.

142

BNSF conducted further analysis in 2004 and decided to adopt a mileage-based program, though its approach was incremental. See DA4481, Defs.’ App’x Vol. 12 (Aug. 30, 2004, internal BNSF email from Kyei to FSC review group, stating, “John Lanigan has suggested that we reconstitute the team that looked at options for a miles per gallon approach in surcharging for fuel”); DA15859 at 15894, Defs.’ App’x Vol. 27 (Kyei, BNSF, Dep. at 236:15-24). BNSF could not immediately adopt such a program on interline traffic without further approval and adaptations, and the rollout for any mileage-based FSC was anticipated to take 6-12 months, given the modifications that BNSF and its customers would have to make. DA4521 at 4532-35, 4541-46, Defs.’ App’x Vol. 12 (Oct. 18, 2004, internal BNSF email from David Burr to Thomas Hund et al., identifying a mileage-based plan but also pointing out issues with adopting it for all traffic, especially interline, and concerns with customers having to modify systems); DA7781 at 7806, Defs.’ App’x Vol. 18 (Jan. 20, 2005, BNSF Mileage-Based Fuel Surcharge Programs: Fuel Surcharge Review Team Recommendations presentation, identifying mileage-based program as requiring 6-12 month implementation to remove systems barriers, notify customers, adapt for interline customers, etc.); Defs.’ Reply at 63 n.44 (identifying exhibits evincing concerns about customer acceptance with mileage-based system). BNSF announced the new program in March 2005 with an effective date of January 1, 2006. DA4573, Defs.’ App’x Vol. 12 (Mar. 28, 2005, email from Lanigan, announcing adoption of mileage-based FSC).

Plaintiffs emphasize that BNSF hardly applied its mileage-based formula until much later. See Pls.’ Opp’n at 36. Although the mileage-based formula was originally set out to apply to most of BNSF’s traffic—around 75%—including BNSF’s portion of some interline shipments (BNSF did not plan to apply it to joint-rate interline movements), see DA4573, Defs.’ App’x Vol. 12 (Mar. 28, 2005, internal BNSF email from Lanigan to Anderson et al.), BNSF later that

143

year decided to delay implementation for most customers. See DA4635 at 4636, id. (Oct. 20, 2005, memo to BNSF customers, describing the mileage-based program effective Jan. 1, 2006, for coal and agricultural products with an effective date for other products to follow later). Plaintiffs suggest this decision was driven by BNSF’s desire to adhere to the conspiracy.

BNSF, however, had perceived significant customer backlash—and the evidentiary record backs this up as what drove the BNSF decision to limit application of the mileage-based formula and slowly acclimate customers over time. See supra Part III.D.2; DA4626 at 4628, Defs.’ App’x Vol. 12 (Oct. 17, 2005, BNSF Mileage-Based Fuel Surcharge Update, Marketing Staff Meeting, sharing customer feedback results from a survey, showing significant opposition); DA4605 at 4605-06, id. (May 2005 BNSF Mileage-Based Fuel Surcharge Customer Feedback); DA4581 at 4581-82, id. (Apr. 4, 2005, BNSF Mileage-Based Fuel Surcharge Customer Feedback). As a result, BNSF did not apply its mileage-based formula across the board (i.e., including to intermodal and carload customers) until 2009. See DA4950, Defs.’ App’x Vol. 14 (Aug. 5, 2008, BNSF press release); DA4939 at 4941, id. (Aug. 2008, Pricing Update FAQs Draft). The anecdotal evidence cited by plaintiffs that some customers may have embraced a mileage-based system earlier does not undermine BNSF’s logical business decision to take a slower, more piecemeal approach to applying its new formula to allay customer backlash. See PX112 at ‘574, Pls.’ App’x Vol. 2 (May 5, 2006, internal BNSF email from Lanigan to Rose et al., stating, “I had three customers at NARS tell me they wanted us to go to mileage-based”); Pls.’ Opp’n at 35 (citing Pls.’ COF ¶ 230). Plaintiffs cannot attribute BNSF’s decision to walk back its mileage-based formula to pressure from other defendants or seemingly conspiratorial discussions—the evidentiary record simply does not support that theory. Evidence of the internal analysis and customer feedback underway at BNSF amply supports BNSF’s choice.

144

Third, plaintiffs point to defendants’ sharing of pricing information with one another, which they argue evinces coordination because such sharing of competitive information is not in any individual defendant’s self-interest. Pls.’ Opp’n at 45-46, 65. According to plaintiffs, this information ranged from fuel costs to hedging practices to information about FSC policies and coverage, see supra Parts III.E.4-5, all of which information had no reason to be shared absent collusion. As previously explained, however, see supra Part III.E.4, “exchange of pricing information by itself is an insufficient basis upon which to allow an inference of agreement to fix prices.” Baby Food, 166 F.3d at 126. Exchanging such information may increase economic efficiency in certain circumstances. Id. at 118.

All of the communications reviewed piecemeal in supra Part III.E.4 revealed exchange of public, not private, pricing information. Such exchanges might in some way aid self-interest but certainly made no practical impact on defendants’ competitive positions and thus could not be against their self-interest. Defendants’ need to exchange information for interline pricing purposes may have reduced their desire to be unnecessarily obstructionist with one another in the contexts of requests for public non-interline pricing information. Further, plaintiffs are unable to explain how sharing information on fuel hedging, see id., even if not in defendants’ self-interest, would evince a conspiracy for FSCs, given that the two are different strategies for managing costs. Without more, such communications cannot tend to exclude the alternative inference of independent action.

8. Evidence of Industry Structure Conducive to Collusion Plaintiffs point out the uniquely concentrated nature of the railroad industry, the huge barriers to entry, highly inelastic demand, and interchangeable services among defendants to argue that railroads operate in an environment highly conducive to collusion, making the inference of conspiracy more likely. Pls.’ Opp’n at 66-67. Plaintiffs further emphasize the

145

regular communication that occurred among defendants for legitimate purposes (such as interline discussions) as well as the signaling that occurred through traditional earnings calls and private analyst events—which plaintiffs characterize as other facilitating practices, because they built an institutional understanding of one another’s pricing, lowering the costs of monitoring compliance. Id. at 68-69.

Kleen Products discussed some of these same factors that contribute to making a market conducive to collusion: a “small number of manufacturers, . . . inelastic demand, a standardized commodity product, and high barriers to entry.” 910 F.3d at 935. These same factors, however, make it both easier to form a “cartel or to follow the leader independently.” Id. (emphasis in original). “[T]he fewer the firms, the easier it is for them to engage in ‘follow the leader’ pricing[,] . . . which means coordinating their pricing without an actual agreement to do so.” Id. The district court in that case “properly found that the economic evidence did not tend to exclude the possibility of independent action” because competing inferences could have been drawn. Id.; see also Williamson Oil, 346 F.3d at 1317 (“[T]he majority of the market characteristics on which [plaintiffs] focus[],” namely, “concentration of sellers, inelastic demand at competitive prices, high barriers to entry, a fungible product, principal firms selling at the same level in the chain of distribution, prices that can be changed quickly, cooperative practices and a record of antitrust violations,” are “simply indicia that the tobacco industry is an oligopoly, which is perfectly legal.”). In other words, these characteristics are just as consistent with conscious parallelism as an oligopoly and do not push plaintiffs over the line toward an inference of conspiracy.

Defendants make several astute observations that further undercut plaintiffs’ argument here. First, defendants’ services are not actually “interchangeable.” See Defs.’ Reply at 11 n.6.

146

A shipper in the Eastern part of the United States cannot turn to a railroad with tracks only in the West, for instance, or ship on a path where a particular railroad does not have tracks. Second, these features of the railroad industry, such as high barriers to entry and highly inelastic demand, existed prior to the conspiracy period, yet did not result—according to plaintiffs—in conspiratorial conduct then. See Defs.’ Reply at 11. Third, that defendants frequently shared pricing and coverage information in legitimate fora, including public earnings calls, particularly supports conscious parallelism and indicates why defendants may have all focused more on growing their FSC revenue during the alleged conspiracy period absent any agreement.

9. Evidence of Defendants’ “Rebasing”

Finally, plaintiffs argue that defendants’ changes to their formulas after departing from the supposedly conspiratorial FSCs “support the inference of collusion.” Pls.’ Opp’n at 56-57, 147-48. Either during or after the STB’s investigation into and decision on FSCs for regulated traffic, defendants adjusted their rates to bake higher fuel costs into the base rate (which they describe as “rebasing”), moving away from rate-based FSCs while maintaining revenues. See id. at 4, 146-49. These changes, however, again do little to advance an inference of conspiracy.

To recount plaintiffs’ evidence of these purportedly collusive changes later in the conspiracy period, NS first made a significant change to its FSCs in Spring of 2006, initially for shipments on tariffs and public quotes and then implemented in private contracts over time, to end up revenue neutral, largely transferring the cost of FSCs into its base rate. See Pls.’ Opp’n at 56-57; DA5191, Defs.’ App’x Vol. 14 (Apr. 24, 2006, NS press release, stating “Norfolk Southern . . . today announced that effective July 1 2006 Norfolk Southern will revise its fuel surcharge program for local and joint line non intermodal traffic originating and moving on Norfolk Southern issued tariff and public quotes issued on or after July 1 2006”); PX318, Pls.’ App’x Vol. 5 (Mar. 21, 2006, internal NS email from Kathleen Smith to David Lawson,

147

analyzing FSCs and options to incorporate fuel costs into base rates); PX98 at ‘622, Pls.’ App’x Vol. 2 (Apr. 20, 2006, internal BNSF email from Kyei to Lanigan, providing BNSF leadership modeling of NS’s proposal if BNSF were to adopt it, which would “swap approximately $1.1B of FSC into your rates”); PX630 at 187, Pls.’ App’x Vol. 10B (Moorman, NS, Dep. at 187:10- 11, describing the 2006 FSC change as intended to be revenue neutral); PX529 at ‘312, Pls.’ App’x Vol. 7 (June 14, 2006, internal UP email from Doug Glass to James Lusk, describing NS’s FSC change and how they had no negative feedback on their re-basing); PX260, Pls.’ App’x Vol. 4 (May 30, 2006, internal NS email, describing how the new formula, consisting of a 16.4% base increase and a $64 surcharge based on the WTI index with a 0.3% increment increase, should be implemented in private contracts). Plaintiffs argue that NS concealed its increase to base rates, only highlighting that FSCs were reduced or eliminated. Pls.’ Opp’n at 56-57. From defendants’ perspective, however, this apparently unilateral change by NS instead shows a lack of parallelism, undermining the claimed conspiracy altogether. Defs.’ Reply at 64; NS Mem. at 12-13; see supra Part III.C.1.

After the STB’s decision in January 2007, two other defendants made changes to move toward mileage-based FSCs and shift fuel costs into base prices. 45 UP moved to a mileage-based FSC in March 2007, and to make this pricing shift mostly revenue neutral, put more of its fuel cost into base pricing. See DA1692, Defs.’ App’x Vol. 3 (Mar. 2007, Sales Talking Points – UPRR FSC Plans); DA1700, id. (Mar. 21, 2007, Customer News Announcement, “Union Pacific Mileage-based Fuel Surcharge Program Effective 4/26/2007”). CSX moved to a mileage-based FSC around the same time and also attempted to “convert a large portion of the expected 2007-

45 Defendants represented at the hearing that such changes to unregulated pricing were logical outcomes of the STB’s decision, given that regulated fares are an option for many shippers and thus form a backdrop to the negotiations for even shippers currently operating under private contracts. See MSJ Hr’g Tr. at 19:2-24.

148

2008 over-recovery into [base] price.” PX175 at ‘663, 666, 673, Pls.’ App’x Vol. 4 (July 24, 2006, CSX Draft Fuel Surcharge Strategy presentation); see also PX174, id. (Mar. 2007, CSX 2007 S&M Goals: Impact from mileage based FSC).

BNSF already had a mileage-based program and expanded it in 2007 and 2008, first to shipments covered by the STB ruling and then to carload and intermodal customers. BNSF did not, however, otherwise “rebase.” See DA4658 at 4660-61, Defs.’ App’x Vol. 12 (Mar. 7, 2007, BNSF email to BNSF Industrial Products and Coal Customers); DA4950, Defs.’ App’x Vol. 14 (Aug. 5, 2008, BNSF press release, regarding expansion of mileage-based FSC to carload and intermodal customers); DA1699, Defs.’ App’x Vol. 3 (Mar. 11, 2007, internal UP employees’ emails about BNSF’s approach).

According to plaintiffs, these actions were not taken independently but rather as part of an agreement to continue the conspiratorial pricing. See Pls.’ Opp’n at 149. In support of this inference, plaintiffs assert that defendants discussed FSCs with one another during this period, see id. at 150-51, pointing, in particular, to defendants continuing to meet through September 2009 under the auspices of AAR’s NEMC, which plaintiffs consistently depict as facilitating conspiratorial conversations, see id. at 150.

Like much of plaintiffs’ evidence, however, this last-gasp apparent conspiratorial movement toward rebasing and mileage-based FSCs does not show anything more than a normal desire for profit maximization in an interdependent industry, in a manner designed to forestall further regulatory concern—not the existence of any agreement or even unusual parallel conduct. Defendants’ changes in their FSCs after the STB decision were clearly motivated by that clear decision critical of FSCs and a concomitant desire to avoid further regulatory scrutiny, as well as their normal desire to maintain revenues and profit margins. That defendants may have had

149

discussions with one another about FSCs after such a big decision from the STB is not surprising, but plaintiffs do not point to any aspect of those discussions suggesting that defendants were trying to uphold a prior conspiratorial agreement. See Pls.’ Opp’n at 150-51 (citing PX511 and PX730); PX511 at ‘555, Pls.’ App’x Vol. 7 (Feb. 7, 2007, internal UP email from Koraleski to Tom Gehl, stating, without explicitly referencing discussions, that “CSX is going down the PC Miler approach. As well. . . . NS is planning to do away with the fuel surcharge altogether an[d] just change rates when they need to”); PX730 at ‘350, Pls.’ App’x Vol. 11 (Feb. 9, 2007, apparent email invite from CSX entitled “Discussion w/ BNSF re: Fuel Surcharge (+Paul Hitchcock)”).

In any case, to the extent that plaintiffs suggest that defendants perpetuated a conspiracy by changing their base rates, that is not the conspiracy that plaintiffs have alleged in this case. Plaintiffs allege a conspiracy about the use of FSCs starting in 2003, not base rates starting nearly five years later. See MDL I Class Cert. I, 287 F.R.D. at 44.

10. Holistic Evaluation of Plaintiffs’ Evidence As stated previously, plaintiffs’ claim fails at the outset because they lack direct evidence of conspiracy and have not shown parallel conduct by the defendants, whose FSC pricing and coverage policies became more, rather than less, dissimilar during the purported conspiracy period. Plaintiffs’ evidence of various “plus” factors and circumstantial evidence is, then, necessarily insufficient to show conspiracy. To ensure, however, the reliability of this conclusion that the outcome would not change even if defendants’ conduct were parallel enough, with full consideration given to all the nuances of plaintiffs’ evidence, all of plaintiffs’ evidence has been evaluated here.

By necessity, in a case as complex and with as voluminous a record as this one, the evidence of the plus factors identified by plaintiffs was evaluated piecemeal. The ultimate

150

inquiry, however, is a holistic one, and while “zero plus zero equals zero,” courts have warned against supposing “that if no single item of evidence presented by the plaintiff points unequivocally to conspiracy, the evidence as a whole cannot defeat summary judgment.” High Fructose Corn Syrup, 295 F.3d at 655. Nonetheless, putting all of plaintiffs’ evidence together, the evidence does not tend to exclude the inference of independent action to defeat summary judgment.

Most of plaintiffs’ evidence was misguided and of little to no probative value. For instance, plaintiffs attempted to show that defendants’ justification of fuel cost recovery for adopting and expanding their FSCs was pretextual, but such a conclusion is belied by the record, and the vast over-recovery plaintiffs decry is not solely attributable to defendants’ changes to their FSC formulas and policies during the alleged conspiracy period. See supra Part III.E.1. Likewise, plaintiffs put forth evidence of multiple meetings among defendants during the Spring 2003, but lacking evidence of what was discussed at those meetings, they are meaningless for purposes of inferring conspiracy. See supra Parts III.E.3(a)-(e). Plaintiffs also put forth evidence of a potential facilitating practice, the adoption of a new fuel cost index by the AAR, but the adoption of the AIILF index was driven primarily by one defendant and had little impact on defendants’ FSC practices. See supra Part III.E.5. Further, plaintiffs could not demonstrate a major change in defendants’ business practices, given their parallel FSC formulas prior to the alleged conspiracy period, nor could they show that defendants took specific actions that were clearly against their self-interest. See supra Parts III.E.6-7. Finally, plaintiffs could not show that defendants changed their base rates and reduced their imposition of FSCs around the time of the STB’s decision for the purpose of perpetuating the prior conspiratorial agreement to impose

151

elevated FSCs. See supra Part III.E.10. All of that evidence is “zero plus zero.” High Fructose Corn Syrup, 295 F.3d at 655.

Some of plaintiffs’ other evidence offers slightly more probative value, but even evaluated cumulatively, the evidence does not tend to exclude the possibility of mere conscious parallelism and independent conduct, especially given the strong showing made by defendants that they acted in their unilateral self-interest to raise prices in a way consistent with their business objectives. See supra Part III.D. Regarding meetings among defendants, plaintiffs can show three meetings where defendants potentially discussed FSCs or other fuel-cost-recovery mechanisms, and where two defendants perhaps contemplated future discussions about industry alignment on FSCs. See supra Parts III.E.3(f)-(g). Yet, these meetings occurred after the formation of the alleged conspiracy in early spring 2003 and cannot be connected to any of defendants’ pricing or coverage decisions. Further, given that some of these meetings were only bilateral, plaintiffs have to string several of these meetings together, each laced with speculation, in order to even create a narrative of conspiracy. See, e.g., MSJ Hr’g Tr. at 107:20-108:9 (plaintiffs’ counsel explaining that although the May 14, 2003, discussion only involved NS and UP, UP then met with CSX in June, effectively accruing consensus through a chain effect). Regarding FSC-related communications among defendants, plaintiffs have shown one low-level query about FSC structure, a single sharing of aggregate statistics by one defendant in October 2003, and a discussion of coverage figures among two defendants in February 2005. See supra Part III.E.4. Similarly, through AAR-related communications, plaintiffs have shown potentially one additional comment about FSC coverage, a potential discussion about staffing structure for FSC accounting groups, and a potential discussion on fuel hedging. See supra Part III.E.5. Lastly, plaintiffs have demonstrated that perhaps defendants communicated more about FSCs

152

after the start of the alleged conspiracy period than before, and that the railroad industry has certain attributes conducive to collusion. See supra Parts III.E.2, 6, 8.

In sum, plaintiffs can show that over the course of five years of the alleged conspiracy, defendants had a few meetings and a handful of sporadic communications where FSCs may have merely come up, though not contemporaneously to any change in defendants’ conduct or the context of creating an agreement, and they can point to some general market factors characteristic of oligopolies. That evidence confirms that defendants were aware of and interested in their competitors’ approaches to FSCs and that they were able to watch and follow their competitors’ moves to maintain higher prices for the market as a whole. That is merely conscious parallelism. Such evidence cannot exclude the possibility of independent action and allow an inference of conspiracy. The evidence is at least as consistent with—in fact, more consistent with—independent conduct than a conspiracy. See Matsushita, 475 U.S. at 587-88.

Two cases help illustrate why this conclusion is compelled on this evidentiary record. In Chocolate, the court determined that limited possession of advanced pricing information, conduct consistent with pre-conspiracy conduct, and a weak showing of pretextual justifications for decisions weighed in favor of granting summary judgment to the antitrust defendants. See 801 F.3d at 407-12. By contrast, in Flat Glass, evidence about coordinated price increases by competitors that were not economically justifiable, suspiciously close in time to meetings and “tightly linked” to exchanges of future pricing information involving high-level executives resulted in a reversal of summary judgment for the defendants. See 385 F.3d at 368-69. The facts here more closely resemble Chocolate: Defendants made disparate and economically rational changes to their FSCs that were not coincident with suspicious meetings or exchanges of private, future pricing information among executives.

153

Accordingly, defendants are entitled to summary judgment on plaintiffs’ carload (and coal) FSC claims.

F. No Dispute of Material Fact Exists for Intermodal Traffic.

Intermodal traffic, which is different in important respects from carload (and coal) traffic and operates separately from it, requires a separate inference of conspiracy. Plaintiffs do not point to any direct evidence of an intermodal traffic conspiracy, nor can they overcome summary judgment through indirect evidence. Plaintiffs cannot demonstrate that defendants’ intermodal traffic FSCs were parallel, nor can they demonstrate that even if those FSCs were parallel, they were not simply the result of conscious parallelism. Defendants are thus entitled to summary judgment on the alleged conspiracy over intermodal traffic as well.

1. Intermodal Traffic Would Require a Separate Conspiracy.

As defendants rightly observe, intermodal traffic operates separately from carload traffic at all of these railroads. Intermodal traffic uses different equipment, involves different customers, and creates different competitive dynamics. Defs.’ Mem. at 69-71. Regarding the equipment, defendants have specific intermodal “hubs” that allow for the shipments to switch seamlessly, via specialized gates and cranes, from trucks or boats to rail and vice versa. See id. at 70; DA11295, Defs.’ App’x Vol. 23 (article from UP webpage, “What is Intermodal Equipment?,” accessed June 7, 2024). Regarding the different competitive landscape, trucks, for instance, play at least some role in the market for intermodal traffic, whereas they are not substitutes for carload traffic. See, e.g., DA15134 at 15137, Defs.’ App’x Vol. 26 (Deposition of Steve Vicary of plaintiff ABF Freight at 67:18-68:3, describing choice between using just trucking or intermodal services for a shipment). 46

46 Defendants suggest that an intermodal conspiracy is economically implausible because defendants also compete with trucking companies for that freight, see Defs.’ Mem. at 80, while plaintiffs respond that “trucks are

154

As a result, defendants have structured their intermodal traffic to be managed by separate departments and employ separate intermodal-specific pricing formulas. See DA18264, Defs.’ App’x Vol. 29 (Dec. 18, 2002, internal NS memo from Prillaman to M.R. McClellan, reporting on updates for intermodal department); DA19117 at 19118, Defs.’ App’x Vol. 30 (showing departmental breakdown at UP and intermodal operating separately); supra Part I.B.3 (referencing CSX’s separate entity managing intermodal traffic, CSXI). Compare Defs.’ Mem. at 34 (showing carload FSC formulas) to id. at 72 (showing different intermodal formulas); see also supra Parts I.B.1-3. Separate pricing formulas and strategic decisions mean that any conspiracy must be separately inferred. As defendants put it, a conspiracy over one product or service cannot implicate a conspiracy over all of a company’s other products and services. See Defs.’ Mem. at 83-84; Defs.’ Reply at 71; see Matsushita, 475 U.S. at 584 n.7 (explaining that because plaintiffs “must show that the conspiracy caused them an injury for which the antitrust laws provide relief,” pricing practices outside of those that supposedly caused plaintiffs’ harm are not considered part of the same conspiracy).

Plaintiffs insist that the scope of the conspiracy is necessarily an issue for the jury. See Pls.’ Opp’n at 121-22. They cite no binding caselaw, however, to that effect and, at the motions hearing, expressed that it did not matter whether the Court considered the alleged intermodal conspiracy to be separate but overlapping with the carload conspiracy or part of a single greater one. See MSJ Hr’g Tr. at 110:3-9. In any case, the evidence here shows that defendants treated

not a meaningful substitute” for intermodal traffic. See Pls.’ Opp’n at 133-35. Regardless of the significance of trucking competition, the existence of any overlap between intermodal traffic and trucking differentiates intermodal from carload traffic and supports separately analyzing a potential intermodal conspiracy.

155

intermodal traffic and pricing completely differently from that for carload traffic. Plaintiffs’ intermodal evidence is thus properly analyzed as indicative of a separate conspiracy. 47 2. Defendants’ Conduct was Not Parallel.

While plaintiffs insist that changes to intermodal traffic FSCs and their increase in coverage are the result of an illegal conspiracy, plaintiffs point to no direct evidence of any agreement, nor can they even demonstrate parallel conduct among defendants with respect to intermodal traffic. Defendants’ approaches to intermodal FSCs during the conspiracy period is more disparate than even their approaches to carload FSCs.

As explained earlier—and as plaintiffs concede—BNSF did not change its intermodal formula at all during the conspiracy period after a revision in 2001. See DA3985, Defs.’ App’x Vol. 9 (July 25, 2001, internal BNSF email from Renita Lee to Ed Zajac et al., regarding publishing the new intermodal formula on the website); DA4664 at 4664-938, Defs.’ App’x Vols. 13 and 14 (BNSF Pricing Update memos over the years showing no change in the intermodal formula during the conspiracy period); Pls.’ Opp’n at 127 (admitting that BNSF did not make meaningful changes to its intermodal formula during the conspiracy period). BNSF considered modifying its intermodal formula in the spring of 2003, when modifying its carload formula, but decided against any such change due to customer retention considerations. See DA4451 at 4453-65, Defs.’ App’x Vol. 12 (Apr. 14, 2003, BNSF Fuel Surcharges: Fuel Surcharge Review Team Recommendations presentation); DA4415 at 4415-16, Defs.’ App’x Vol. 11 (July 1, 2003, internal BNSF memo from Kyei to Bobb, explaining they did not change the intermodal formula because they were concerned about customer reaction).

47 Whether the intermodal pricing changes are viewed as separate from or part of a broader scheme with the changes to the carload formulas makes little difference. Plaintiffs are unable to exclude the possibility of parallel action, even considering all of the evidence together.

156

Similarly, NS only changed its international intermodal formula near the end of the conspiracy period, keeping everything else the same. NS adopted the same intermodal formula as BNSF for its domestic freight in 2002, before any alleged conspiracy. See DA18001 at 18003, Defs.’ App’x Vol. 29 (Decl. of Jim Bolander, GVP for NS Domestic Intermodal, describing the HDF intermodal surcharge adopted in 2000); DA7564, Defs.’ App’x Vol. 18 (Nov. 13, 2003, internal NS email from Will Santangelo to Stacia Minton et al., describing the FSC in place since 2002: “Intermodal has been billing a fuel surcharge since April 2002 and is currently assessing a 3.5% for domestic traffic and 2% for international.”); DA7572 at 7573, id. (Nov. 4, 2002, internal NS email from Kimberly Thompson to Mike McClellan, showing tables with fuel surcharges for intermodal); DA7387, id. (Apr. 2, 2002, internal NS email from Chris Luebbers to Russ Kirchhoff et al., stating, “NS Intermodal is re-implementing the fuel surcharge based upon the DOE’s weekly Deisle [sic] Index”). That formula was not altered during the conspiracy period, as plaintiffs concede. See Pls.’ Opp’n at 127 (agreeing that NS did not meaningfully change its domestic formula during the conspiracy period). NS merely changed its international intermodal formula in 2006 to match its domestic formula. See DA7389, Defs.’ App’x Vol. 18 (Nov. 16, 2006, internal NS email from Augie Eckhardt to Robert Randolph, stating, “We put this scale into place for International traffic back in August. Prior to that we did have a different International scale, but now all of our Intermodal traffic is consolidated under this same scale”); DA18001 at 18004, Defs.’ App’x Vol. 29 (Decl. of Jim Bolander, NS GVP of Intermodal, at ¶ 11, stating, “On August 21, 2006, the international intermodal scale was brought in line with the domestic intermodal scale ($1.24/0.5%/$0.04) and was adjusted on a weekly basis”).

157

UP, for its part, changed its intermodal formula twice during the conspiracy period but not until 2005 and 2006, making its formula less aggressive each time. UP first adopted an intermodal FSC in 2000. See DA19047, Defs.’ App’x Vol. 30 (Nov. 22, 2004, internal UP email from Marty H. Coalson, Asst. VP of Product Dev. & Yield Mgmt. for UP Intermodal, to Jennifer Hamann, circulating the current policy, “Effective October 9, 2001, . . . all UPRR pricing documents making reference to this [Master Intermodal Transportation Agreement] will be subject to a Fuel Surcharge”); DA19037, id. (Nov. 8, 2000, memo from UP Intermodal Business Team, stating, “Effective October 9, 2000, except as otherwise provided in connection with the particular rates or changes, all UP pricing documents will be subject to a fuel surcharge”). In 2005, UP decided its intermodal FSC was too high compared to its competitors, so it lowered the company’s intermodal rates. See DA1666 at 1667, Defs.’ App’x Vol. 3 (Oct. 17, 2005, internal UP email from Trena Bryan to Ed Weber, showing the adjustment down in 2005); DA19001 at 19001-02, Defs.’ App’x Vol. 30 (Oct. 2005, internal UP email chain, explaining the change that occurred March 2005, to “lower the variance” with BNSF’s FSC to remain competitive, including with truckers). Despite the adjustment in 2005, UP continued to think that the intermodal FSC formula would not be competitive, so UP frequently manually adjusted the formula and pursued further changes in late 2005, eventually implementing them in late 2006. See DA19001, id. (Oct. 18, 2005, internal UP email from Michael Boone to Ed Weber, stating that “the formula driven surcharge amounts have experienced a compounding effect that puts the formula-driven surcharge out of synch with the competition. The competition includes other railroads as well as truckers. Consequently, the intermodal group has been forced to manually adjust the formula-driven amounts based on the competitive alternatives available to our shippers. This is not exact science and suggests we need to consider a different formula that

158

reflects today’s reality without sacrificing fuel cost coverage”); DA19045, id. (Sep. 12, 2006, internal UP email from Weber to Robert Knight, announcing the intermodal FSC change and describing it as “less aggressive (i.e., reduced slope)” but “more robust” and explaining the need for it after having to manually override the formula; also describing its goal as becoming more “market competitive”); DA19099 at 19099-100, id. (Oct. 30, 2006, internal UP email from Nobu Terasaki to Randy Davison et al., describing the 2006 formula change). Notably, UP’s formula also differed markedly from that of other defendants in its use of a unique fuel weight component. See supra Parts I.B.3, I.B.5.

Finally, CSX argues that the relevant defendant, CSX Transportation, did not provide intermodal services at all—those were provided by a different entity, CSXI, which is not a defendant in MDL I. See Defs.’ Mem. at 74-75; MDL I Class Cert. III, 292 F. Supp. 3d at 123 & n.37 (CSXI “is not a party to this case”); DA16501 at 16503, Defs.’ App’x Vol. 27 (fiscal year 2007, CSX Corporation Form 10-K, stating, “CSX Intermodal . . . is a stand-alone, integrated intermodal company”); DA16506 at 16509, id. (fiscal year 2010, CSX Corporation Form 10-K, stating, “CSX Intermodal . . . was a subsidiary of CSX until it merged with CSXT during 2010,” and describing intermodal services as now being provided by CSXT). Regardless, CSXI adopted a unique intermodal formula in 2000. See DA2137 at 2138, Defs.’ App’x Vol. 3 (Sep. 15, 2000, Merchandise Marketing and Sales: Fuel Surcharge presentation, stating, “On March 6, 2000 CSX Intermodal announced a fuel surcharge calculated on the DOE calculated price of on- highway price of diesel fuel on the last Monday of the month”); DA3007, Defs.’ App’x Vol. 5 (CSX Intermodal Fuel Surcharge Chronology (from 2000 to May 14, 2007)). CSXI changed its formula in 2004 to match that of BNSF and NS, but no other defendant made changes around that time.

159

In short, defendants did not exhibit parallel conduct in raising their intermodal FSCs during the conspiracy period. CSX, via affiliate CSXI, is the only defendant that changed its intermodal FSC during the conspiracy period to align with other defendants but did so several years after those defendants independently adopted that formula. The other defendants’ choices—maintaining the status quo, pursuing internal alignment, or lowering their prices—do not indicate parallel, concerted action. See Anderson News, 899 F.3d at 105 (holding that defendants’ “varying courses of action” were not “parallel”).

Plaintiffs argue that the lack of changes by some of the defendants during the conspiracy period is not dispositive because the conspiracy is demonstrated by coordinated efforts leading to overall higher rates and greater surcharge coverage. See Pls.’ Opp’n at 122, 127-28. Plaintiffs point to a chart from defendants’ expert, Dr. Joseph Kalt, which seems to suggest that defendants’ surcharge rates on intermodal traffic were actually quite uniform during the conspiracy period. See Pls.’ Opp’n at 127; PX637, Pls.’ App’x Vol. 10B (Kalt, Defs.’ Expert, Report ¶ 200 and Fig. IV-14B). Plaintiffs also point out an email from BNSF suggesting that the eastern and western railroad FSC rates on intermodal traffic were “virtually identical” at that point in time. PX799 at ‘598, Pls.’ App’x Vol. 13B (Mar. 27, 2006, internal BNSF email from Kevin Williams to Brian Large et al.). Even accepting that degree of uniformity, plaintiffs do not point to other evidence indicating this was the result of a conspiracy.

3. Plaintiffs Cannot Show Defendants’ Conduct Was More Consistent with Conspiracy than Conscious Parallelism.

Even if defendants’ conduct as to intermodal traffic were considered parallel, plaintiffs’

evidence does not tend to exclude the possibility of independent action. At the outset, the generally aligned formulas—among CSXI, BNSF, and NS from 2004 through 2006—do not seem to have been adopted in coordination with one another, given the long time gaps between

160

each of their adoptions. In fact, BNSF and NS adopted the same formulas in the pre-conspiracy period, so that change is not proof of a conspiracy forming during the period or in the way alleged by plaintiffs to support the conspiracy claim. CSXI did not change its formula until over two years after NS matched BNSF. Such a delay, allowing ample time for unilateral decision- making, is “as equally suggestive of independent action as collusion,” if not “more suggestive of independent action.” In re Text Messaging, 46 F. Supp. 3d at 808; see also Baby Food, 166 F.3d at 131-32.

Plaintiffs also point to UP’s changes—which moved toward less aggressive FSCs—

contending these changes were motivated in part by trying to match BNSF and that this is indicative of a concerted effort to bring all of their formulas into harmony. See Pls.’ Opp’n at 129; PX390 at ‘388, Pls.’ App’x Vol. 6B (Mar. 22, 2005, internal UP email from Nobu Terasaki to John Kaiser, stating, “As a matter of comparison, 18% fuel weight will be more aligned to BNSF surcharge index particularly in the higher bracket”). UP may have been following BNSF, but that is classic follow-the-leader parallelism, which, without more, is not indicative of a conspiracy, especially given the time lag between BNSF’s changes pre-conspiracy period and UP’s changes in 2005. Again, plaintiffs offer no evidence demonstrating that UP and BNSF “intentionally worked in a coordinated fashion.” Pls.’ Opp’n at 129.

Further, the same rational business motivation for adopting FSCs and increasing their coverage during this period, see supra Part III.D.1, also explain why defendants took actions to raise and expand their FSCs with respect to intermodal traffic. See Moundridge, 2009 WL 5385975, at *6 (“[W]here there is an independent business justification for the defendants’ behavior, no inference of conspiracy can be drawn.” (alteration in original) (quoting Blomkest Fertilizer, Inc., 203 F.3d at 1037)); cf. Pls.’ Opp’n at 133 (insisting that because defendants were

161

over-recovering, they were not responding to the need to recover fuel costs). 48 These considerations provide a strong basis for inferring unilateral action guided by conscious parallelism rather than illegal collusive conduct.

Plaintiffs point to meetings and communications among defendants as probative of an intermodal conspiracy, but this evidence, under scrutiny, is simply not persuasive. First, plaintiffs reference all of the meetings previously discussed, see supra Part III.E.3, and explain how the leaders overseeing intermodal traffic attended those meetings, to suggest that they also indicate an intermodal conspiracy. See Pls.’ Opp’n at 123-24. For the reasons already explained, however, those meetings are not probative of conspiracy, and because none were close in time to any of the changes made to intermodal formulas, they are particularly irrelevant here. Second, plaintiffs reference certain discussions already covered, such as PX664, see supra Part III.E.3, which is an email exchange from October 2003 among CSX employees inquiring about what had been gleaned from a meeting with BNSF about the latter’s coverage rates. See Pls.’ Opp’n at 124. As explained, that information was made public the same month and did not isolate FSC data, making difficult to see how this information could advance the FSC conspiracy. In any case, as before, plaintiffs do not connect this email to any intermodal pricing decision made by either company.

Plaintiffs also focus on an expansion in FSC coverage for intermodal traffic that occurred during the conspiracy period. See Pls.’ Opp’n at 124-26. As with carload traffic, many defendants sought to apply their FSCs widely during the conspiracy period. See PX629, Pls.’ App’x Vol. 10B (Koraleski, UP, Dep. at 93-95, explaining that until March 2003, “we were intermodal and we were contract specific in most cases” but after, “we would implement a fuel

48 Plaintiffs again suggest that defendants’ justifications for FSCs are pretextual because they over-recovered, Pls.’ Opp’n at 133, but this argument fails for the same reasons as it did with carload traffic, see supra Part III.E.1.

162

surcharge across a wide base of our business”); PX437 at 3, Pls.’ App’x Vol. 7 (2005 UP Analyst Meeting presentation, stating, “We have adopted a standard Intermodal fuel surcharge policy that is applied universally to all of our customers as their contracts expire”); PX224 at ‘471, Pls.’ App’x Vol. 4 (May 13, 2005, CSXI Fuel Surcharge presentation, stating, “Full Fuel Surcharge on all customers – No Exceptions!”); PX82, Pls.’ App’x Vol. 2 (June 7, 2006, internal CSX email from Kyei to Kevin Williams, describing an “interest” in “verify[ing] whether there was total FSC discipline on all our traffic, local and interline” but also asking for more reporting on “non-standard FSC programs and associared [sic] reasons”); see supra Part III.C.3.

Yet, such a curated selection of assorted communications fails to demonstrate either significant, uniform changes in policy during the conspiracy period or other concerted actions explaining the increase. As with carload FSC coverage, see supra Part III.C.3, intermodal coverage significantly increased in 2002, just prior to the alleged start of the conspiracy, and this fact indicates that the increase in coverage was the result of unilateral, not collusive, action. See, e.g., DA4086 at 90, Defs.’ App’x Vol. 10 (Jan. 15, 2003, internal BNSF memo from Schultz to Rose, demonstrating 40.7% of BNSF’s intermodal revenue was covered by FSCs by late 2002); Defs.’ Mem. at 76 (adopting a figure from Dr. Israel’s report, DA25148 at 25164, Defs.’ App’x Vol. 39, Fig. 33, demonstrating broad intermodal FSC coverage prior to 2003). Other than BNSF, which had around 40% coverage just prior to the alleged start of the conspiracy, the defendants’ intermodal FSC coverage was well over 80% before 2003. See Defs.’ Mem. at 76 (incorporating Fig. 33 of Dr. Israel’s report, DA25148 at 25164). Coverage also increased naturally over time throughout the conspiracy period as contracts came up for renewal and renegotiation, just as with carload coverage, see supra Part III.C.3. See DA23325 at 23428-29 (Leitzinger, Pls.’ Expert, Report ¶ 167); DA24749 at 24752-53 (Report of Defs.’ Expert, Dr.

163

Matthew Wright ¶ 41, describing how contracts are multi-year); PX437 at 3, Pls.’ App’x Vol. 7 (2005 UP Analyst Meeting presentation); DA5008 at 131, Defs.’ App’x Vol. 14 (Seale, NS, Dep. at 131:2-6, testifying, “We knew it was going to take a while to get coverage. It took a long time to get coverage. Because we had contracts that ran for multi years, and we had other rate authorities that ran for multiple years”). Indeed, plaintiffs do not point to any set of parallel policy changes for intermodal FSC coverage during the conspiracy period.

Plaintiffs contest the meaningfulness of defendants’ statistics given that, despite high coverage rates for intermodal freight traffic pre-conspiracy, many of the contracted FSCs were never collected during the pre-conspiracy period. See Pls.’ Opp’n at 126 (citing DA4086 at 4090, showing that while 40.7% of BNSF’s intermodal revenue was subject to FSCs, only 1.9% was billed). Even so, plaintiffs still lack evidence of any agreement to increase coverage, and as previously explained, an agreement to increase coverage of a non-coordinated, non-parallel set of FSCs would still not constitute the conspiracy plaintiffs have alleged—to impose coordinated, elevated FSCs. See supra Part III.C.3. Plaintiffs’ evidence simply does not tend to exclude the inference of independent action for intermodal traffic.

G. Other Issues and Pending Motions.

Given that plaintiffs have not put forth sufficient evidence to allow for any inference of conspiracy on any segment of defendants’ freight traffic at issue, summary judgment is granted to defendants, and the parties’ other arguments and partial motions for summary judgment need not be reached. In particular, for example, defendants’ argument that plaintiffs lack proof of impact, causation, and damages need not be considered. See Defs.’ Mem. at 84-99. Further, because plaintiffs cannot prove a conspiracy at all, they cannot prove that one extended beyond 2007, so the parties’ arguments about the proper end point for the alleged conspiracy (i.e., the impact of the STB decision and the initial filing of complaints) need not be evaluated. See Defs.’

164

Mem. at 99-107; Pls.’ Opp’n at 140-53. Likewise, defendants’ arguments for partial summary judgment on the allegedly time-barred damages claims in MDL II, ECF No. 983, also need not be reached, so that motion, too, is denied as moot.

The parties have also filed seven evidentiary motions that need not be resolved because the challenged evidence at issue in those motions and expert materials in question were not crucial to the findings and holdings made in denying summary judgment. For example, defendants’ Rule 702 and 403 Motion to Exclude Average Damages Opinions, ECF No. 1169, need not be decided as this motion pertains only to damages, an issue that need not be reached. Similarly, defendants’ Rule 702 Motion to Exclude False Positive Econometric Overcharge Models, ECF No. 1165, need not be decided for the same reason. These motions are therefore denied as moot.

Defendants also moved to exclude expert reliance on improper evidence, ECF No. 1167, arguing that plaintiffs’ experts improperly assume the alleged conspiracy, opine on defendants’ state of mind and intentions, provide improper factual narrative based on interpretation of documents, and repeat or rely on material inadmissible under section 10706. Such portions of plaintiffs’ expert reports were not necessary to rely upon in deciding summary judgment, so this motion of defendants is also denied as moot.

Plaintiffs filed four motions to exclude certain of defendants’ expert reports and opinions, none of which were relied on, so these motions are therefore all denied as moot. For example, plaintiffs moved to exclude defendants’ expert Joseph Dettmar, ECF No. 1171, who opined on the STB’s activities and interline traffic, which did not prove necessary to reach a decision on summary judgement. Plaintiffs also moved to exclude defendants’ expert Avram Tucker, ECF No. 1175, who used forensic analysis to assess the reliability and reasonableness of plaintiffs’

165

experts. Despite defendants’ citations to Mr. Tucker’s Report and plaintiffs’ claims of inaccuracy in his report, this report was also not necessary to reach a decision on summary judgment. Plaintiffs further moved to exclude defendants’ experts, Drs. Edward Snyder and Lauren Stiroh, in the Oxbow litigation, 11-cv-1049, ECF No. 309. These experts mainly offered opinions with respect to counts that the parties have since dismissed, but they also offered rebuttal to plaintiffs’ experts on the plus factors and the nature of competition during the alleged conspiracy period. Again, reliance on these opinions was unnecessary to reach a decision on summary judgment. Finally, plaintiffs moved to strike the declarations of attorneys Eric Fanchiang and Kelsey Bryan, ECF No. 1197, which summarized specific price authorities for which plaintiffs seek damages, but, again, given that no reliance on the declarations was necessary and plaintiffs are not entitled to damages, the motion is also denied as moot. IV. CONCLUSION Both sides in this case have been vigorously litigating the existence of a conspiracy among defendants regarding FSCs in the rail freight industry for nearly two decades. Plaintiffs amassed and presented to the Court 29,538 pages of documents (filed across 33 volumes of exhibits) about defendants’ activities from 2003 through 2008 and beyond—demonstrating countless meetings, communications, pricing decisions, negotiations, and more. From this evidence, plaintiffs put together a detailed narrative of apparent collusion that has, on the surface, some appeal. Their underlying evidence, however, does not support the conclusions they propound. “Here, what Plaintiffs claim are disputes of fact are in truth disputes about the reasonableness of inferences, which is precisely what Matsushita directs the Court to gauge at summary judgment.” Defs.’ Reply at 1. Plaintiffs’ inferences are not sustainable. Upon close inquiry, their evidence lacks the crucial thrust of a conspiracy claim: facts that tend to exclude

166

the possibility that defendants acted unilaterally and made their decisions independently, even if they acted in a way that was consciously parallel to one another in a tight oligopoly. See Matsushita, 475 U.S. at 587-88. Without such evidence, plaintiffs cannot survive summary judgment.

This Memorandum Opinion will be filed under seal for a brief period to allow the parties an opportunity to request redactions of confidential business information before the public docketing of the decision. Recognizing the “‘strong presumption in favor of public access to judicial proceedings,’ including judicial records,” In re Leopold, 964 F.3d 1121, 1127 (D.C. Cir. 2020) (emphasis added) (quoting United States v. Hubbard, 650 F.3d 293, 318 (D.C. Cir. 1980)), the parties are directed to submit jointly by 2:00 pm ET on June 27, 2025, any proposed redactions, supported by analysis applying the Hubbard factors. Given the long pendency of this case, much of the business information is dated, so the Court expects that any necessary redactions will be minimal.

An order consistent with this Memorandum Opinion will be entered contemporaneously.

Date: June 24, 2025

__________________________

BERYL A. HOWELL

United States District Judge