Opinion
18-cv-864
01-21-2022
IN RE DEALER MANAGEMENT SYSTEMS ANTITRUST LITIGATION, MDL 2817 This document relates to ALL CASES
MEMORANDUM OPINION AND ORDER
Robert M. Dow, Jr., United States District Judge.
Pursuant to Federal Rule of Evidence 702 and the Supreme Court's decision in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), the parties have moved to exclude or limit the testimony of ten proposed expert witnesses: Edward M. Stroz [859]; Gordon Klein [863]; Daniel L. Rubinfeld [867]; Dr. Kevin Murphy [873]; Dr. Mark Israel [877]; Catharine Lawton [879]; Dr. Michael A. Williams [881]; Allan Stejskal [883]; Nancy Miracle [885]; and Brian Halpin [887]. For the following reasons, the motions to exclude Williams [881] and Stejskal [883] are granted in part and denied in part. The motions to exclude Klein [863], Rubinfeld [867], Murphy [873], Israel [877], Lawton [879], and Halpin [887] are denied. The motion to exclude Miracle [885] is granted in part, denied in part, and deferred in part. The motion to exclude Stroz [859] is denied in part and deferred in part. As noted in the minute entry issued today, the parties will be given an opportunity to (a) review this comprehensive opinion, (b) submit supplemental briefs on how, if at all, these rulings affect the briefing and proper disposition of the pending summary judgment motions, and (c) advise the Court on any other pertinent developments of which they believe the Court should be aware. The Court will then prepare and issue a second, comprehensive opinion disposing of the summary judgment motions in due course.
I. Background
The background of this MDL has been set forth in multiple opinions, knowledge of which is assumed here. See, e.g., [176], [425], [504], [505], [506], [507], [749], [857]. The Court briefly recounts this background here, to provide context for its analysis of the parties' Daubert motions.
A. The DMS Market
A data management system (“DMS”) is software that automobile dealers use to help manage their businesses, including accounting, sales, inventory, service, customer information, and human resources functions. Dealers typically store a substantial portion of their data within their DMSs. Virtually every franchised new car dealership in the United States now uses a DMS. Defendants CDK Global, Inc. (“CDK”) and The Reynolds and Reynolds Company (“Reynolds”) both provide DMSs to auto dealerships throughout the United States. CDK and Reynolds each have significant market power in the DMS market. Together, they control approximately 75% of the United States market by number of dealers and approximately 90% when measured by number of vehicles sold.
B. The DIS Market
To effectively run their dealerships, dealers engage third-party software application providers-also known as “vendors”-to provide services such as inventory management, customer relationship management, warranty services, repair orders, and electronic vehicle registration and titling. Vendors need to access and use data located on dealers' DMSs in order to provide their services to dealers. Vendors generally obtain access by engaging data integrators, which provide data integration services (“DIS”) to vendors for a fee. Data integrators (also called “DIS providers”) collect and standardize data stored on DMSs. Their DIS services allow vendors to access dealer data and extract, organize, and integrate it into a usable format. They may also provide value-enhancing services, such as putting data from different DMSs into a uniform format and correcting data errors.
CDK and Reynolds each offer data integration to vendors. CDK's DIS is called Third Party Access (“3PA”) and Reynolds' is called the Reynolds Certified Interface (“RCI”). 3PA and RCI only provide DIS for Defendants' respective DMSs. There are also independent third-party data integrators-including Plaintiffs Authenticom, Inc. (“Authenticom”) and Loop LLC (“Autoloop”), among others-that offer DIS to vendors. CDK owns two third-party data integrators: Digital Motorworks, Inc. (“DMI”) and IntegraLink.
C. The Challenged Agreements
Historically, CDK and Reynolds had “open” DMSs, meaning that they did not take steps to prevent their dealer clients from authorizing third-party access to the dealers' data. Over time, Reynolds began to “close” its DMS by selectively blocking third-party data integrators from accessing dealer data stored on the Reynolds DMS. As Reynolds allegedly reduced competition for DIS through its blocking activities, Reynolds increased the fees it charged for data integration through its own RCI product. CDK, by contrast, marketed its DMS as an “open” system.
Various Plaintiffs (which are specifically identified where relevant in this opinion) allege that by the summer of 2013, Reynolds was taking aggressive action to prevent independent DIS providers from serving dealers who obtained their DMSs from Reynolds. This hampered CDK's ability to provide DIS to those dealers. Around September 2013, CDK allegedly abruptly stopped trying to compete with Reynolds by emphasizing CDK's “openness”-i.e., the ease with which independent App vendors can access data in the dealer's DMS using any DIS service they choose (subject to dealer approval), and the cost of doing so. Plaintiffs posit that this was due to an agreement between Reynolds and CDK to drive independent DIS providers from the market by, among other things, agreeing to a common marketing message on “data security” to condition the market to accept that pretext for excluding independent DIS providers. In return, Plaintiffs allege, Reynolds allowed CDK access to Reynolds' platform.
In February 2015, CDK and Reynolds entered into three written agreements: (1) the Data Exchange Agreement or “Wind Down” Agreement; (2) the 3PA Agreement; and (3) the RCI Agreement. In the Data Exchange Agreement, CDK agreed to wind down its data integration business on the Reynolds DMS, with Reynolds promising not to block CDK's access to the Reynolds system during the wind-down period (approximately 5 years). During that period, Reynolds agreed that CDK could continue to extract dealer data just as it had before, using login credentials provided by dealers. As for other independent data integrators, CDK and Reynolds each agreed not to take any steps to assist any person that it reasonably believed to have plans to access or integrate with the other party's DMS.
The other two written agreements, the 3PA Agreement and the RCI Agreement (collectively referred to as the Data Integration Agreements) provided CDK and Reynolds with reciprocal access to each other's DIS programs. Pursuant to the agreements, Reynolds received five free years of 3PA access. Reynolds also agreed to access CDK's DMS exclusively through 3PA and not to contract with any third parties to access the system. The agreements also provided that CDK and Reynolds would deny data integrators access to each other's DMSs.
In addition to written agreements, senior executives from the two Defendants also allegedly agreed to restrict access to dealer data and “destroy” data integrators like Authenticom. Plaintiffs allege that, by eliminating competition for data integration services, CDK and Reynolds have seized control over dealer data and thwarted dealers' ability to control access to and the usage of their data. As a result, vendors have no choice but to access dealer data through CDK's and Reynolds' own DISs, at much greater cost than they would incur if they purchased services from independent data integrators. Plaintiffs allege that shortly after CDK and Reynolds entered into the Data Exchange Agreement, CDK began renegotiating contracts with vendors for 3PA access and requiring vendors to use 3PA exclusively for all of their dealer-customers on CDK DMSs.
D. The EVR Market
Since 1992, CDK and Reynolds have also been parties to a joint venture that wholly owns Defendant Computerized Vehicle Registration, Inc. a/k/a CDK Vehicle Registration, Inc. (“CVR”). CDK owns 80% of CVR, while Reynolds owns 20%. CVR, like Plaintiff Motor Vehicle Software Corporation (“MVSC”), is a provider of electronic vehicle registration and titling services (“EVR”). EVR providers partner with state governments to issue the physical registration, license plates, and titles for vehicles sold at car dealerships. In order to process the registration and title with a state, EVR providers require basic information about the car sale, including the vehicle, buyer, and financing details. That data is stored on a dealership's DMS.
Plaintiff MVSC alleges that in January 2014, and possibly earlier, CDK and Reynolds entered into a horizontal agreement to block MVSC from participating in their third-party access programs (3PA and RCA). In addition to blocking MVSC from participating in their 3PA and RCI programs, CDK and Reynolds have allegedly cut off MVSC's ability to access car sales data through intermediaries. According to MVSC, CDK and Reynolds have intimidated dealers by threatening that they will be in breach of their DMS contracts if they provide MVSC with data from their DMSs.
E. Procedural Matters
Based on these facts, the parties have brought a variety of claims and counterclaims, which have been consolidated in this MDL. Some of the claims have been dismissed, while others have been resolved through settlement. The parties have engaged in extensive discovery and filed crossmotions for summary judgment, which are currently pending. See [777], [949], [954], [963], [964], [967], [970], [976]. Also pending are several motions by Defendants to bar certain claims and theories that, according to Defendants, were not timely disclosed. See [773], [787].
The claims that remain active include: Authenticom's claims against CDK and Reynolds for horizontal conspiracy and exclusive dealing in violation of Sherman Act § 1, monopolization in violation of Sherman Act § 2, and tortious interference; Autoloop's claims against CDK for horizontal conspiracy and exclusive dealing in violation of Sherman Act § 1, monopolization in violation of Sherman Act § 2, and parallel Florida state law claims; the proposed Dealership Class's claims against CDK for conspiracy and exclusive dealing in violation of Sherman Act §§ 1 and 2 and related state antitrust and consumer protection laws; MVSC's claims against Reynolds for conspiracy in violation of Sherman Act §§ 1 and 2 and related California and Illinois state law laws; and Reynolds and CDK's counterclaims against various Plaintiffs for breach of contract, violations of the Computer Fraud and Abuse Act (“CFAA”), the Digital Millennium Copyright Act (“DMCA”), and the Defend Trade Secrets Act (“DTSA”), and related Wisconsin and California state law claims.
II. Daubert Standard
Rule 702 and Daubert provide the legal framework for the admissibility of expert testimony. See Bielskis v. Louisville Ladder, Inc., 663 F.3d 887, 893-94 (7th Cir. 2011); United States v. Pansier, 576 F.3d 726, 737 (7th Cir. 2009). Rule 702 requires that the district judge act as a “‘gate-keeper' who determines whether proffered expert testimony is reliable and relevant before accepting a witness as an expert.” Winters v. Fru-Con Inc., 498 F.3d 734, 741 (7th Cir. 2007) (quoting Autotech Tech. Ltd. Partnership v. Automationdirect.com, 471 F.3d 745, 749 (7th Cir. 2006) (internal quotation marks omitted)); see also Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 147-49 (1999); Daubert, 509 U.S. at 589.
Under Federal Rule of Evidence 702, “[a] witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if: (a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; and (d) the expert has reliably applied the principles and methods to the facts of the case.” Fed.R.Evid. 702; see also Ortiz v. City of Chicago, 656 F.3d 523, 536 (7th Cir. 2011). Rule 702 and Daubert require the district court to engage in a three-step analysis before admitting expert testimony. The court must determine (1) “whether the witness is qualified”; (2) “whether the expert's methodology is scientifically reliable”; and (3) “whether the testimony will ‘assist the trier of fact to understand the evidence or to determine a fact in issue.'” Myers v. Ill. Cent. R.R. Co., 629 F.3d 639, 644 (7th Cir. 2010) (quoting Ervin v. Johnson & Johnson, Inc., 492 F.3d 901, 904 (7th Cir. 2007)). The proponent of the expert bears the burden of demonstrating, by a preponderance of the evidence, that “the expert's testimony would satisfy the Daubert standard.” Lewis v. CITGO Petroleum Corp., 561 F.3d 698, 705 (7th Cir. 2009) (citing Fed.R.Evid. 702 advisory committee's note (2000 Amends.)). District judges possess considerable discretion in dealing with expert testimony. Carroll v. Otis Elevator Co., 896 F.2d 210, 212 (7th Cir. 1990); see also General Electric Co. v. Joiner, 522 U.S. 136, 141-43 (1997) (holding that abuse of discretion standard applies in reviewing district court rulings on admissibility of proposed Rule 702 opinion testimony).
A witness is qualified as an expert if she has sufficient “knowledge, skill, experience, training, or education” relating to the subject matter of her testimony. Demouchette v. Dart, 2012 WL 6568232, at *3 (N.D. Ill.Dec. 14, 2012) (quoting Fed.R.Evid. 702). “Whether a witness is qualified as an expert can only be determined by comparing the area in which the witness has superior knowledge, skill, experience, or education with the subject matter of the witness's testimony.” Gayton v. McCoy, 593 F.3d 610, 616 (7th Cir. 2010) (quoting Carroll, 896 F.2d at 212).
As to reliability, the Court must ensure that a proposed expert's methodology is “scientifically valid, ” Daubert, 509 U.S. at 592-93, and that her conclusions are “based on sufficient facts or data, ” Fed.R.Evid. 702(b). “The non-exclusive list ofDaubert reliability factors for scientific evidence includes whether or not the theory or technique has been (1) tested, (2) subjected to peer review and publication, (3) analyzed for known or potential error rate, and/or is (4) generally accepted within the specific scientific field.” Lapsley v. Xtek, Inc., 689 F.3d 802, 810 (7th Cir. 2012) (citing Daubert, 509 U.S. at 593-94). But this “list of specific factors neither necessarily nor exclusively applies to all experts or in every case.” Kumho Tire, 526 U.S. at 141 (internal quotation marks omitted).
“Rule 702's requirement that expert opinions be supported by ‘sufficient facts or data' means ‘that the expert considered sufficient data to employ the methodology.'” Manpower, Inc. v. Insurance Co. of Pennsylvania, 732 F.3d 796, 808 (7th Cir. 2013) (quoting Stollings v. Ryobi Technologies, Inc., 725 F.3d 753, 766 (7th Cir. 2013)). “For example, if an expert seeks to testify about an average gross sales price but is going to base the testimony on sales to only a single customer, a court would appropriately exclude the testimony because a single observation does not provide a sufficient basis for calculating an average.” Stollings, 725 F.3d at 766. Whether an expert “selected the best data set to use, however, is a question for the jury, not the judge.” Manpower, 732 F.3d at 809; see also A.H. v. Illinois High School Ass'n, 263 F.Supp.3d 705, 714 (N.D. Ill. 2017) (observing that “the ‘sufficient facts or data' requirement is a somewhat Gordian one-an expert must take into account all of those facts which are necessary to support [her] findings, not just some of them, and [she] cannot conveniently disregard contrary evidence, but a party cannot disqualify the other side's expert simply by disputing the facts upon which [she] relied” (internal quotation marks and citation omitted)). Assuming there is “a rational connection between the data and the opinion, ” the “expert's reliance on faulty information is a matter to be explored on cross-examination; it does not go to admissibility.” Manpower, 732 F.3d at 809 (citing Walker v. Soo Line R.R. Co., 208 F.3d 581, 589 (7th Cir. 2000)). “Our system relies on crossexamination to alert the jury to the difference between good data and speculation.” Schultz v. Akzo Nobel Paints, LLC, 721 F.3d 426, 432 (7th Cir. 2013).
“[T]he law grants a district court the same broad latitude when it decides how to determine reliability as it enjoys in respect to its ultimate reliability determination.” Kumho Tire, 526 U.S. at 142; see also Pansier, 576 F.3d at 737; Lewis, 561 F.3d at 704-05. At bottom, the Court's goal is “to assure that experts employ the same ‘intellectual rigor' in their courtroom testimony as would be employed by an expert in the relevant field.” Jenkins v. Bartlett, 487 F.3d 482, 489 (7th Cir. 2007) (quoting Kumho Tire, 526 U.S. at 152). Its “focus, of course, must be solely on principles and methodology, not on the conclusions that they generate.” Daubert, 509 U.S. at 595. Nonetheless, as the Supreme Court has recognized, “conclusions and methodology are not entirely distinct from one another, ” and while “[t]rained experts commonly extrapolate from existing data[, ] ... nothing in either Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert.” General Electric, 522 U.S. at 146. “‘An expert who supplies nothing but a bottom line supplies nothing of value to the judicial process.'” Wendler & Ezra, P.C. v. Am. Int'l Grp., Inc., 521 F.3d 790, 791-92 (7th Cir. 2008) (quoting Mid-State Fertilizer Co. v. Exch. Nat'l Bank, 877 F.2d 1333, 1339 (7th Cir. 1989)). In short, “[i]t is critical under Rule 702 that there be a link between the facts or data the expert has worked with and the conclusion the expert's testimony is intended to support.” United States v. Mamah, 332 F.3d 475, 478 (7th Cir. 2003). Where that link is missing, “[a] court may conclude that there is simply too great an analytical gap between the data and the opinion proffered.” General Electric, 522 U.S. at 146.
Finally, the requirement that evidence or testimony assist the trier of fact to understand the evidence or to determine a fact in issue “goes primarily to relevance.” Daubert, 509 U.S. at 591. “Expert testimony which does not relate to any issue in the case is not relevant and, ergo, nonhelpful.” Id. (internal quotation marks and citation omitted); see also Cage v. City of Chicago, 979 F.Supp.2d 787, 804 (N.D. Ill. 2013).
III. Plaintiffs' Antitrust Claims
A. Legal Framework
The primary claims in this case arise under Sections 1 and 2 of the Sherman Act, 15 U.S.C. §§ 1 and 2. Before turning to the parties' specific challenges to the experts who address the Sherman Act claims, the Court finds it useful to outline the legal requirements for such claims, the unique way that summary judgment is applied to them, and the interaction and differences between summary judgment and Daubert analyses in the antitrust context.
Section 1 of the Sherman Act “is designed to prevent businesses from entering into collusive agreements.” Omnicare, Inc. v. UnitedHealth Group, Inc., 629 F.3d 697, 705 (7th Cir. 2011). It prohibits “[e]very contract, combination ... or conspiracy, in restraint of trade or commerce, ” 15 U.S.C. § 1, “though courts have long restricted its reach to agreements that unreasonably restrain trade.” Omnicare, 629 F.3d at 705 (citing State Oil Co. v. Khan, 522 U.S. 3, 10 (1997)). “To prevail under § 1 under any theory, plaintiffs generally must prove three things: (1) that defendants had a contract, combination, or conspiracy (‘an agreement'); (2) that as a result, trade in the relevant market was unreasonably restrained; and (3) that they were injured.” Id.; see also Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 585-86 (1986); Denny's Marina, Inc. v. Renfro Prods., Inc., 8 F.3d 1217, 1220 (7th Cir. 1993).
Allegations concerning the existence of an agreement “‘usually take one of two forms: (1) direct allegations of an agreement, like an admission by a defendant that the parties conspired; or (2) more often, circumstantial allegations of an agreement, which are claimed facts that collectively give rise to a plausible inference that an agreement existed.'” Always Towing & Recovery, Inc. v. City of Milwaukee, 2 F.4th 695, 703 (7th Cir. 2021) (quoting Alarm Detection Sys., Inc. v. Village of Schaumburg, 930 F.3d 812, 827 (7th Cir. 2019)). “The task before any plaintiff is thus to find and produce evidence that reveals coordination or agreement[.]” Kleen Prods. LLC v. Ga.-Pac. LLC, 910 F.3d 927, 936 (7th Cir. 2018). “To show concerted action, antitrust plaintiffs must produce evidence that would allow a jury to infer that the alleged conspirators ‘had a conscious commitment to a common scheme designed to achieve an unlawful objective.'” Omnicare, 629 F.3d at 705 (quoting Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 764 (1984)). Put another way, “the circumstances of the case must reveal ‘a unity of purpose or a common design and understanding, or a meeting of minds in an unlawful arrangement.'” Id. (quoting American Tobacco Co. v. United States, 328 U.S. 781, 810, (1946)). “More evidence is required the less plausible the charge of collusive conduct.” In re High Fructose Corn Syrup Antitrust Litigation, 295 F.3d 651, 661 (7th Cir. 2002).
At the summary judgment stage-the next step in this case-the plaintiff's “‘offer of conspiracy evidence must tend to rule out the possibility that the defendants were acting independently.'” Kleen Products, 910 F.3d at 934 (quoting Bell Atlantic Corp. v. Twombley, 550 U.S. 544, 554 (2007)); see also Matsushita, 475 U.S. at 588; Miles Distributors, Inc. v. Specialty Constr. Brands, Inc., 476 F.3d 442, 449 (7th Cir. 2007) (“When a plaintiff attempts to defeat summary judgment by highlighting circumstantial evidence of a conspiracy, some of the evidence must tend to exclude the possibility that the alleged conspirators acted independently rather than in concert.”). “[A]n antitrust plaintiff, like all others, is entitled to try to meet that burden with either direct or circumstantial evidence.” Kleen Products, 910 F.3d at 934; see also In re Text Messaging Antitrust Litigation, 630 F.3d 622, 628-29 (7th Cir. 2010); Miles Distributors, 476 F.3d at 449). Of course, determining whether an unlawful arrangement exists is “much easier if there [is] a smoking gun, ” such as an admission by an insider, “buried in [the often] voluminous record.” Omnicare, 629 F.3d at 706; see also In re Plasma-Derivative Therapies Antitrust Litigation, 764 F.Supp.2d 991, 998 (N.D. Ill. 2011); cf. High Fructose Corn Syrup Antitrust Litigation, 295 F.3d at 654 (“admission by the defendants that they agreed to fix their prices is all the proof a plaintiff needs”). But more typically, the plaintiff's evidence will be inferential and “of two types- economic evidence suggesting that the defendants were not in fact competing, and noneconomic evidence suggesting that they were not competing because they had agreed not to compete.” Id. at 655. “While no single piece of information may win the day, the whole may be greater than the sum of its parts in tending to exclude the possibility of conscious parallelism.” Kleen Products, 910 F.3d at 934.
In contrast to Section 1 of the Sherman Act, Section 2 “covers both concerted and independent action, but only if that action ‘monopolize[s],' or ‘threatens actual monopolization,' a category that is narrower than restraint of trade.” American Needle, Inc. v. National Football League, 560 U.S. 183, 190 (2010) (quoting 15 U.S.C. § 2 and Copperweld Corp. v. Independence Tube Corp., 467 U.S. 752, 767 (1984)). “‘Monopoly power is the power to control prices or exclude competition.'” Braman v. The CME Group, Inc., 149 F.Supp.3d 874, 896 (N.D. Ill. 2015) (quoting United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 391 (1956)). “To state a claim for monopolization under Section 2, a plaintiff must allege that that defendant engaged in predatory or anticompetitive conduct.” In re Dealer Management Systems Antitrust Litigation, 362 F.Supp.3d 477, 497-98 (N.D. Ill. 2019) (citing Mercatus Grp., LLC v. Lake Forest Hosp., 641 F.3d 834, 854 (7th Cir. 2011)). “Where defendant has engaged in unlawful restraint of trade that would independently violate Section 1 ..., it is well established that it also violates Section 2 if it acquires or maintains a monopoly by means of that restraint of trade.” Id. at 498 (quoting Gumwood HP Shopping Partners, L.P. v Simon Prop. Grp., Inc., 2013 WL 3214983, at *7 (N.D. Ind. Mar. 13, 2013)).
To recover under the antitrust laws, a plaintiff “must prove that [its] damages were caused by the unlawful acts of the defendant.” MCI Communications Corp. v. American Tel. & Tel. Co., 708 F.2d 1081, 1161 (7th Cir. 1983) (citing 15 U.S.C. § 15; emphasis in MCI); cf. Omnicare, 629 F.3d at 705 (“section 4 of the Clayton Act, 15 U.S.C. § 15, provides a private cause of action for the enforcement of § 1 [of the Sherman Act]”). “This is the essence of ‘antitrust injury'”-i.e., “‘injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants' actions unlawful.'” MCI, 708 F.2d at 1161 (quoting Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 489 (1977)). In other words, “[i]f a plaintiff has suffered financial loss from the lawful activities of a competitor, then no damages may be recovered under the antitrust laws.” Id.
Once causation of damages has been established, “[c]alculating damages presents a separate challenge.” In re Sulfuric Acid Antitrust Litigation, 743 F.Supp.2d 827, 866 (N.D. Ill. 2010). “‘Damages calculations in antitrust cases seek to compare plaintiffs' actual experience in the real world with what the plaintiffs' experience would have been, ‘but for' the antitrust violation.'” In re Rail Freight Fuel Surcharge Antitrust Litigation, 292 F.Supp.3d 14, 56 (D.D.C. 2017) (quoting In re Pool Prods. Distribution Mkt. Antitrust Litigation, 166 F.Supp.3d 654, 678 (E.D. La. 2016)); see also Greater Rockford Energy & Tech. Corp. v. Shell Oil Co., 998 F.2d 391, 395 (7th Cir. 1993). “[T]he amount of damages may be determined by a just and reasonable estimate as long as the jury verdict is not the product of speculation or guess work.” MCI, 708 F.2d at 1161 (emphasis in MCI); see also Sulfuric Acid Antitrust Litigation, 446 F.Supp.2d at 924; Fontana Aviation, Inc. v. Beech Aircraft Corp., 432 F.2d 1080, 1087 (7th Cir. 1970) (“Once there is sufficient evidence of the fact of damage to go to the jury, the assessment of the amount is largely a matter of the jury's consideration of all the evidence.” (citing Bigelow v. RKO Radio Pictures, 327 U.S. 251, 264 (1946))).
The Supreme Court “has been willing to accept a degree of uncertainty in the calculation of damages” in antitrust cases and, therefore, “strict proof of what damages have been caused by which acts has not been required.” MCI, 708 F.2d at 1161 (citing Continental Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699 (1962); Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555 (1931)). “Not requiring strict disaggregation of damages among the various unlawful acts of the defendant serves to prevent a defendant from profiting from his own wrongdoing and makes sense when damages arise from a series of unlawful acts intertwined with one another.” Id. (citing Bigelow, 327 U.S. at 264-65); see also J. Truett Payne Co. v. Chrysler Motors Corp., 451 U.S. 557, 565-67 (1981); Ohio-Sealy Mattress Mfg. Co. v. Kaplan, 745 F.2d 441, 448 (7th Cir. 1984); Sulfuric Acid Antitrust Litigation, 446 F.Supp.2d at 924. “It is essential, however, that damages reflect only the losses directly attributable to unlawful competition.” MCI, 708 F.2d at 1161 (emphasis in MCI). Plaintiffs must “prove in a reasonable manner the link between the injury suffered and the illegal practices of the defendant.” Id.
Assessing the reliability of experts' economic testimony in the antitrust context presents some unique challenges, as “[t]he Daubert rules were not developed with either antitrust or economics in mind.” Phillip E. Areeda & Herbert Hovenkamp, ANTITRUST LAW: AN ANALYSIS OF ANTITRUST PRINCIPLES AND THEIR APPLICATION ¶ 309a, at 4 (4th & 5th eds. 2015-2021) (“Areeda & Hovenkamp”). “Daubert itself is based on a fairly traditional concept of the scientific method in which investigation proceeds by the formation of a hypothesis, the development of experiments that ‘test' the hypothesis, with the durability of the hypothesis increasing and decreasing as tests are repeated”-a concept that “works well for natural and physical sciences, ” but “is not particularly well suited to economics.” Id.; see also id. at 2 (“economics' criteria for testability ... sometimes fall short of the criteria stated for other sciences”). Antitrust courts have generally “responded to questionable or novel economic testimony by being fairly reluctant to exclude it under Daubert rules, ” while at the same time “scrutiniz[ing] admitted testimony carefully and often harshly under summary judgment rules.” Id.
“The cumulative nature of expert and non-expert testimony generally makes summary judgment rather than Daubert exclusion a much more practical vehicle in antitrust cases, at least in the majority of circumstances where the basic competency of the expert or her methodology is not in issue.” Areeda & Hovenkamp ¶ 309c2, at 16. For instance, when an expert opines on whether certain conduct is indicative of collusion, if the challenged testimony “is left in, then the judge considering a motion for summary judgment can look at all parts of the record: the economist's testimony showing that the structural conditions, the incentives, or the conduct make a certain agreement plausible or likely; and the testimony of other witnesses concerning whether such an agreement actually occurred.” Id. “In this process, the court almost always takes into account doubts about the credibility, relevance, or robustness of what the expert had to say.” Id. Another topic that may be more appropriately left for summary judgment is disaggregation of damages-for instance, where a defendant challenges a damages expert on the basis that she is unable “to show what portion of [the] plaintiff's loss resulted from unlawful conduct.” Id. at 17. “Assuming that the economist is using acceptable criteria, ” in many cases “the correct resolution is not exclusion of the testimony under Daubert but rather a summary judgment grant because the expert's testimony has not been able to isolate those elements of the plaintiff's damage that were caused by the antitrust violation.” Id. With these principles in mind, the Court turns to the parties' Daubert challenges concerning expert testimony on the Sherman Act claims.
B. Economics Experts
1. Israel [877] - Authenticom, Autoloop, and MVSC's expert on antitrust collusion, damages
Defendants move to exclude certain opinions of Dr. Mark Israel. See [889-1] (report). Israel has a Ph.D. in economics from Stanford University. He specializes in the economics of industrial organization. He has taught courses at Stanford and the Kellogg School of Management, Northwestern University, concerning business strategy, industrial organization economics, and econometrics. Since 2006, Israel has worked at Compass Lexecon, an economic consulting firm, where his work has focused on the application of economic theory and econometric methods to competitive analysis of the impact of mergers, antitrust issues including a wide variety of singlefirm and multi-firm conduct, class certification, and damages estimation.
Israel has been retained by Plaintiffs Authenticom, AutoLoop, and MVSC to analyze the economic impact of Defendants' alleged unlawful conduct. More particularly, Israel was asked to: (1) analyze the relevant markets in which the alleged conduct occurred; (2) assess the extent of Defendants' market power in those relevant markets; (3) provide an economic analysis of whether and how Defendants' alleged conduct was jointly profitable for Defendants and harmful to competition in the relevant markets; and (4) quantify the impact of that alleged conduct on competition, competitors, and customers in the relevant markets, including a calculation of the damages suffered by vendors. Israel also responds to certain arguments raised by Defendants' economic experts. His “fundamental conclusion, ” which he elaborates in a 195-page report [8891], a 160-page reply report [889-2], and deposition testimony [889-3], is as follows: “CDK's and Reynolds' actions have harmed competitors, competition, and consumers in DIS markets and App markets and have harmed competition and consumers in the DMS market. ... [T]he mechanism of this harm was the joint blocking of independent DIS providers from accessing the CDK and Reynolds DMSs, which effectively monopolized CDK's and Reynolds' respective DIS markets, thus leading to higher DIS prices and higher DIS costs to App vendors, thereby also harming competition in App markets.” [889-1] at 10. Israel estimates damages from overcharges due to DIS price elevation, forced switching to CDK or Reynolds' DIS services, and in total. See id. at 129-33.
a. Collusion
Defendants first challenge Israel's opinions concerning collusion by criticizing what they label his testimony on “openness.” Israel takes the position that Defendants' actions are indicative of collusion. He explains that for many years Defendants competed on DMS “openness, ” which is “the ease with which independent App vendors can access data in the dealer's DMS using any DIS service they choose (subject to dealer approval), and the cost of doing so.” [1001] at 9. Before Defendants' alleged collusive agreements, Reynolds was less “open” than CDK, because it more vigorously blocked independent DIS providers. CDK competed for customers on the basis that its dealers were free to use independent DIS providers on its platform, which caused Reynolds to lose sales to CDK. Israel recounts that in July 2013, Reynolds began taking more aggressive action to prevent independent DIS providers from serving Reynolds dealers, which hampered CDK's data integration services on Reynolds' DMSs. In September 2013, CDK abruptly stopped competing with Reynolds on the basis of “openness.” Among other things, CDK and Reynolds allegedly agreed to convey a common message on data security. In return, Reynolds allowed CDK access to its platform. Israel explains how, based on the framework to which Defendants agreed in 2013, they reached formal agreements aligning their data access policies by February 2015, “setting the stage for CDK's blocking of independent DIS providers.” Id. at 10 (citing [889-1] at 98). Israel summarizes Defendants' shared strategic goal of eliminating competition from independent DIS providers and the evidence that they explicitly agreed to cooperate in doing so. See [889-1] at 9697.
Defendants move to exclude Israel's opinion that CDK and Reynolds agreed to eliminate competition on “openness” on the basis that the opinion is “hopelessly vague” and therefore unreliable and unhelpful to the factfinder. [878] at 7. Defendants maintain that “[b]y refusing to provide any firm sense of what an agreement as to ‘openness' entails, Dr. Israel can (and does) count any fact he wants as evidence in support of his theory.” Id. at 11. Defendants further argue that the vagueness of Israel's opinion makes it impossible to test its accuracy.
The Court is not convinced that Israel's opinions are too vague to be reliable or helpful to a jury. As Plaintiffs emphasize, even informal agreements can violate the Sherman Act. “[A] conspiracy” that violates Section 1 of the Sherman Act can be found not only from “a meeting of minds” in the sense contract law might require, but also if “the conspirators had a unity of purpose or a common design and understanding.” American Tobacco, 328 U.S. at 810. “[E]ven a wink and a nod” may reveal coordination or agreement. Kleen Products, 910 F.3d at 936. Plaintiffs' and Israel's theory is that the evidence shows that CDK agreed to stop criticizing Reynolds' blocking policies in the DMS market; Reynolds agreed to favor CDK's DIS provider; and the parties agreed to adopt a common marketing message with respect to data security and to cooperate in targeting independent DIS providers. See [889-1] at 92-101. Defendants' own experts also use the term “openness” and have been able to describe and respond to Israel's opinions. To the extent that Defendants question the sufficiency of the basis for Israel's opinion, that can be “explore[d] on cross examination and argument for the benefit of the trier of fact.” In re Disposable Contact Lens Antitrust Litigation, 329 F.R.D. 336, 372 (M.D. Fla. 2018).
Defendants' next argument is that Israel applied no reliable methodology to conclude that, absent a conspiracy, CDK and Reynolds acting unilaterally would not have secured their systems in the same manner in which they allegedly colluded. While Israel theorized that CDK would not have acted unilaterally to reduce “openness” because doing so would have risked losing too many DMS customers, he did not test that theory by evaluating the anticipated losses to CDK from unilateral closure or comparing them to the gains, including unilateral increases in 3PA prices or increased security. Defendants further maintain that Israel's interpretation of documents on this topic is not expert analysis because he did not apply any specialized economic knowledge to his interpretation and is no more competent that then average juror in interpreting the evidence.
The Court agrees with Plaintiffs that this is an insufficient basis to exclude Israel's testimony. The general consensus in the case law is that “economic experts may testify as to ‘whether certain conduct is indicative of collusion,' or ‘consistent' with a conspiracy.” Rail Freight Fuel Surcharge Antitrust Litigation, 292 F.Supp.3d at 52 (quoting In re Processed Egg Products Antitrust Litigation, 81 F.Supp.3d 412, 420 (E.D. Pa. 2015), and In re Urethane Antitrust Litigation, 2012 WL 6681783, at *3 (D. Kan. Dec. 21, 2012)); see also Disposable Contact Lens Antitrust Litigation, 329 F.R.D. at 372-73 (recognizing that courts “regularly admit expert testimony that certain conduct or evidence is ‘consistent with a finding that Defendants engaged in a conspiracy to fix prices'” and collecting cases (quoting In re Delta/Airtrain Baggage Fee Antitrust Litigation, 245 F.Supp.3d 1343, 1359 (N.D.Ga. 2017)); In re Titanium Dioxide Antitrust Litigation, 2013 WL 1855980, at *4 (D. Md. May 1, 2013) (same). “Such testimony necessarily involves consideration of the defendants' conduct in the case.” Rail Freight Fuel Surcharge Antitrust Litigation, 292 F.Supp.3d at 52. For instance, an expert might testify that an alleged arrangement “furthered [Defendants'] collusive interest, but was inconsistent with Defendants pursuing their independent, unilateral self-interest, ” id. as that is “precisely the type of inquiry an economist can be expected to make using his expertise.” Processed Egg Products Antitrust Litigation, 81 F.Supp.3d at 424; see also Urethane Antitrust Litigation, 2012 WL 6681783, at *3 (rejecting argument that expert's opinion “that particular events, assuming they occurred, are consistent with a conspiracy” was “not helpful to a jury because it is not based on an economic analysis and thus is not sufficiently scientific or technical as required under Fed.R.Evid. 702, ” because “structure-conduct-performance analysis by an economist is well-accepted in this field” and it may “be helpful to a jury for an expert to put events into an economic context”). In short, to the extent that an expert, “after reviewing the documentary record, opines as to whether defendants' behavior is consistent with collusion, this is permissible expert opinion.” Rail Freight Fuel Surcharge Antitrust Litigation, 292 F.Supp.3d at 52.
To avoid summary judgment, Plaintiffs will need to come forward with evidence of conspiracy that “tend[s] to rule out the possibility that the defendants were acting independently.'” Kleen Products, 910 F.3d at 934. Plaintiffs are “entitled to try to meet that burden with either direct or circumstantial evidence, ” or both. Id. As courts have found sufficient in other antitrust cases, Israel's testimony synthesizes the economic and non-economic evidence and opines that it supports the conclusion that Defendants' conduct was not equally consistent with their permissible independent interests as it was with improper activity. See Rail Freight Fuel Surcharge Antitrust Litigation, 292 F.Supp.3d at 52. Defendants' own experts (Bresnahan and Whinston) performed similar evaluations of the record and offer opinions that the documents do not support a hypothesis of unlawful collusion. Defendants do not identify any case law requiring Israel to use a formal model quantifying the relative gains from unilateral versus conspiratorial conduct; nor do they identify any model that Israel could have, but did not, use instead. The sufficiency of the evidence relied on by Israel will be left to a battle of the experts.
In a related argument, Defendants contend that Israel's conclusion that their behavior was more consistent with conspiratorial rather that unilateral conduct fails to consider obvious alternative explanations. According to Defendants, Israel failed to consider that DMS switching patterns may have changed (i.e. Reynolds' loss of customers to CDK decreased from 2013 on) due to: 1) other changes in the competitive landscape, like the rise of Dealertrack and Auto/Mate as substantial DMS competitors; or 2) the possibility that dealers who preferred open system had already left Reynolds, which had been engaging in “blocking” for years by 2013. Defendants also argue that Israel failed to consider other reasons why Defendants agreed to a “wind-down” concerning DMI's access to Reynolds' DMS. Defendants emphasize that the wind-down avoided disruptions to their customers and that Reynolds used wind-downs with other parties, too.
Plaintiffs respond that Israel did not improperly ignore competition from DMS competitors in evaluating switching data. They explain that the fact that dealers continued to leave Reynolds for other DMS providers-which continued to compete on the basis of greater openness-at undiminished rates but not for CDK “reinforces the conclusion that CDK stepped back its efforts to win that business.” [1001] at 20. Plaintiffs also point out that Defendants have no hard evidence that dealers that wanted independent DIS services had already switched from Reynolds to CDK before 2013. As to alternative reasons for Defendants' wind-down agreement, Plaintiff respond that the possibility that Reynolds had non-collusive reasons to negotiate a wind-down with CDK does nothing to undermine Dr. Israel's point: doing so also served the objectives Defendants pursued through collusion. As the foregoing discussion indicates, Israel and Defendants both make arguments that provide “a rational connection between the data and the opinion.” Manpower, 732 F.3d at 809. Therefore, they are properly addressed through cross-examination, not Daubert motions.
b. Damages
In the portion of his analysis measuring the impact of Defendants' alleged collusion on DIS prices, Israel examined DIS prices from January 1, 2013 to the first part of 2019, measuring price for each combination of application, DIS, and DMS on an average revenue per unit basis. Each data point reflects what a vendor pays for DIS, from a given DIS provider servicing dealers using a given DMS, in a given month.
In their Daubert motion, Defendants argue that Israel's pricing analysis and damages regression cannot help the jury because they do not model actual RCI prices or pricing conduct. Rather, Israel relies on 3PA and RCI “average monthly recurring DIS prices per rooftop.” Defendants contend that this results in a “composition effects” error because, by using average revenue per unit, Israel shows RCI price increases in situations where prices remained the same but demand for more expensive RCI products increased. Citing Blue Cross & Blue Shield of Wisconsin v. Marshfield Clinic, 152 F.3d 588, 594 (7th Cir. 1998), Defendants assert that “[t]he Seventh Circuit has rejected an expert's pricing analysis based on similar methodological errors.” [878] at 24.
In his reply report, Israel explains why, in his opinion, it would be inappropriate to use RCI's listed prices-i.e., “panel data”-rather than data reflecting actual sales, as Reynolds' expert Bresnahan proposed in part of his report. According to Israel, Bresnahan's proposal was “highly unusual, ” because in Israel's experience, “the goal where possible is always to base prices on actual invoice data, which tell you how many units (here of a given package) are sold.” [889-2] at 95. Israel posits that “[t]he reason this is superior can be seen in all of Professor Bresnahan's figures that use the panel data”: “Each of these figures are based on unweighted average prices (i.e., per rooftop per month) across packages, meaning that a package that was sold to just a few rooftops for just a few months would count the same as the largest Apps in the analysis.” Id. at 95-96. Israel also points out that Bresnahan and CDK's damages expert, Murphy, used the same RCI sales data that Israel used when they “turn[ed] to actual regression analysis of the DIS prices.” Id. at 96. Plaintiffs acknowledge that while a “composition effects” error may hypothetically be possible, Defendants' experts have not shown that using different data would have produced a materially different result.
Defendants' criticisms of Israel's pricing analysis and damages regression concern “how the selection of data inputs affect the merits of the conclusions produced by an accepted methodology, ” and thus go to the weight rather than the admissibility of Israel's damages opinion. Manpower, 732 F.3d at 808. The parties have advanced competing expert opinions concerning the best inputs to use to calculate damages attributable to DIS price increases. At this gatekeeping stage of the case, it would be inappropriate for the Court to “evaluate the quality of an expert's data, inputs, or conclusions.” In re Zimmer Nexgen Knee Implant Products Liability Litigation, 2015 WL 3669933, at *25 (N.D. Ill. June 12, 2015). Defendants' citation to Marshfield Clinic does not convince the Court otherwise. In that case, the Seventh Circuit was reviewing the district court's grant of summary judgment, not a ruling on a Daubert challenge. The court of appeals held that damages in a market division case against a medical clinic could not be based on an expert report comparing per-patient (rather than per-service) prices that the clinic charged during the period it engaged in market division with prices charged during the same period by medical service providers in different regions. 152 F.3d at 594. The court reasoned that the calculation failed to correct for other factors that affected the clinic's prices-most notably, the fact that the “record disclose[d]” that “all there is to the higher average price per patient” charged by the clinic was that the clinic had “referred to it patients who are sicker than average and so require longer treatment.” Id. Here, by contrast, Defendants have not developed a record suggesting that a hypothetical composition effects error has skewed Israel's calculations in any material way.
Defendants' next criticism of Israel's damages analysis is that he did not assume the proper “but-for” world-that is, what Plaintiffs' experience would have been, ‘but for' the antitrust violation.'” Rail Freight Fuel Surcharge Antitrust Litigation, 292 F.Supp.3d at 56. In particular, Defendants argue that Israel improperly attributed all of 3PA's and RCI's price increases after September 2013 to unlawful conduct even though, as he opines in his report, CDK and Reynolds had the unilateral power to raise their prices at least to some extent. Defendants maintain that Israel's model needed to, but did not, account for CDK and Reynolds' lawful ability to raise prices unilaterally and that this failure renders his damages model inadmissible. After all, Defendants contend, “Israel cannot rest his conspiracy opinions on the premise that firms will always maximize profits and simultaneously account for no possibility of unilateral price increases by either CDK or Reynolds over a several year period.” [878] at 27.
Plaintiffs respond that because Israel's model “identifies as overcharges the departure from the trends predicted by Defendants' historical pricing behavior-which was oligopolistic, not conspiratorial-facilitated by the conspiratorial exercise of joint market power, the model effectively grants Defendants the estimated expected effects of individual market power.” [1001] at 26. In particular, Israel used a “difference-in-differences” (or “DiD”) regression econometric model designed to “measure the effect of CDK's and Reynolds' conduct by comparing their DIS prices [to] the prices of a control group, consisting of independent DIS providers, over the same time period.” [889-1] at 120-21. Israel also used “time and cross-sectional fixed effects to control for factors that affect DIS prices.” Id. at 121, n.289. He then “test[ed] whether CDK's and Reynolds' conduct increased DIS prices for 3PA and RCI by more than the increase in prices in the benchmark group during the initial-conspiracy and post-February 2015 periods.” Id. at 12122. According to Israel, “[b]y comparing the change in prices charged in the affected group to the changes in prices charged in the benchmark group, ” he “isolate[d] the effect of the CDK and Reynolds' agreement to block their DMSs from other factors affecting DIS prices over time or across DMS=App combinations.” Id. at 122. Israel also considered other methods of testing collusion's impact, including one that substituted CDK's and Reynolds' own pre-conduct prices as the relevant baseline and one that used industry control variables. According to Israel, all of the models he used showed the “same fundamental trends: first Reynolds, and then CDK, were able to sustain massive price increases.” [1001] at 25. As for Defendants' argument that Israel cannot both assume that firms always act to maximize their profits and fail to account for Defendants' lawful ability to raise prices unilaterally, Plaintiffs respond that this logic just “confirms that, pre-collusion, [Defendants] were exercising what market power they each had, all else equal.” Id. at 25-26.
Both sides offer seemingly reasonable arguments, supported by testimony from well qualified and respected economists, concerning why their construction of the but-for world is appropriate. (The opinions of CDK's rebuttal expert, Dr. Murphy, are discussed below.) Their competing arguments highlight the extent to which they disagree on the underlying facts, their relevance, and the inferences that should be drawn from them. The experts on both sides have asserted at least some factual basis for their views of the appropriate “but for” world. And both offer a “rational connection between the data and the opinion.” Manpower, 732 F.3d at 809. Daubert does not empower the Court to decide whether either expert “selected the best data set to use.” Id. This is a question for the factfinder-unless the material facts are undisputed and there is only one reasonable inference to be drawn from them, entitling one party to summary judgment. Cf. Story Parchment, 282 U.S. at 562 (where the evidence was sufficient to allow “the jury to find that [defendants'] price cutting and the resulting lower prices were directly attributable to the[ir] unlawful combination, ” it was error for the lower court to assume that defendants' “acts would have been the same if they had been acting independently of one another, with the same resulting curtailment of prices”). Perhaps the framework of summary judgment will allow the Court to resolve the parties' arguments concerning the appropriate but-for world; perhaps it will not. Regardless, all of the parties' arguments go beyond the proper scope of Daubert. See Syufy Enterprises v. American Multicinema, Inc., 793 F.2d 990, 1003 (9th Cir. 1986) (concluding that “[i]t was for the jury to consider [plaintiff's] arguments about the [differences between the yardstick and the but-for world] on the validity of comparison and to adjust its damage award accordingly”). In re Prograf Antitrust Litigation, 2014 WL 7641156, at *3 (D. Mass. Dec. 23, 2014) (recognizing that “[c]ourts have been reluctant to determine comparability before trial”); Cason-Merenda v. Detroit Med. Ctr., No. 06-15601, 2013 WL 1721651, at *12 (E.D. Mich. Apr. 22, 2013) (concluding that comparability issues were “another ‘battle of the experts' that is not appropriate for resolution in the context of a Rule 702 motion to exclude an expert's testimony”).
Defendants also criticize how Israel defined the relevant antitrust market for dealer data integration services. Defining the market involves identifying which firms/products compete most closely with one another. Israel explains that once a market is defined, “the market shares of the identified participants and a measure of concentration (often the Herfindahl-Hirschman Index or “HHI”) can be calculated, which can be a helpful piece in assessing the existence and extent of market power.” [889-1] at 30. According to Israel, economists generally define relevant antitrust markets using the “hypothetical monopolist test” to identify the competing products that constrain a firm's behavior. This test works as follows: It “starts with a proposed relevant market-a small set of products that are particularly close competitors-and considers whether a hypothetical monopolist over all those products could raise prices without losing so many customers as to make that price increase unprofitable. If a hypothetical monopolist over all of the products in the proposed relevant market would find it profitable to impose a small but significant and non-transitory increase in price (‘SSNIP,' often taken to be a five percent increase in price over the competitive level)-because too few customers would turn to other products outside of the proposed market to defeat the price increase by making it unprofitable-then that set of products constitutes a relevant product market. If a hypothetical monopolist would not find it profitable to impose a SSNIP-because too many customers would turn to other products outside of the proposed market-then the proposed relevant market is too narrow, and a broader set of substitute products is considered. The market definition exercise proceeds by adding products until a large enough set of products is included so that a SSNIP by the hypothetical monopolist would be profitable.” Id. at 31. Applying the hypothetical monopolist test, Israel arrived at one market for the sale of a DMS to franchise car dealers within the United States; two distinct aftermarkets for the sale of DIS-one to vendors serving CDK dealers, and a second to vendors serving Reynolds' dealers; and distinct aftermarkets for the sale of applications (or “apps”) to dealers, which are distinguished both by DMS and by functionality.
Defendants maintain that Israel's definition of the DIS market is flawed because it excluded Dynamic Reporting-a DMS reporting tool offered by Reynolds and used by third parties to obtain data from a DMS. In arguing that Dynamic Reporting should have been included, Defendants rely on testimony from one of their own experts, Stejskal, and on Israel's acknowledgment that vendors such as MVSC have used Dynamic Reporting “to ‘[e]ffectively' build their own integration.” [878] at 30. Defendants also criticize Israel for failing to distinguish between DIS providers that offer “basic extraction, ” rather than “complex integration.” Id. In Defendants' view, these are not functional equivalents because, for example, no third-party data extractor (like Authenticom) could reliably provide “real-time” or “write-back” to vendors, especially on the Reynolds DMS. Id.
The Court agrees with Plaintiffs that the proper treatment of particular DIS providers is a battle of the experts that should be left to the factfinder, as there is evidence in the record supporting both sides' positions. Israel explains that he excluded Dynamic Reporting from his DIS market definition based on testimony by vendors and Authenticom's CEO that, in his view, make clear that “automated extraction” of data from a DMS “is far superior to tools” like Dynamic Reporting “that rely on manual actions.” [1001] at 29. As to the basic/complex distinction, Plaintiffs point out that the evidence is mixed on independent DIS providers' ability to provide real-time and writeback services. Both parties' opinions are based on data, which can be interpreted in different ways. It is up to the factfinder to evaluate the quality and relative persuasiveness of that data. See Zimmer Nexgen Knee Implant Products Liability Litigation, 2015 WL 3669933, at *25; Sumotext Corp. v. Zoove, Inc., 2020 WL 533006, at *11-12 (N.D. Cal. Feb. 3, 2020) (denying exclusion of expert's market definition as unreliable for “fail[ing] to substantiate his conclusions under the SSNIP test using empirical analysis” where his “opinions ha[d] a foundation in qualitative data such as employee testimony, third-party industry analyses, and media coverage”).
Finally, Reynolds seeks to exclude Israel's purported opinion concerning the economic feasibility of Reynolds' RCI pricing proposal to MVSC. However, MVSC does not offer Israel to testify in that manner, see [1001] at 30, making this portion of the Daubert motion moot.
2. Williams [881] - Dealer Plaintiffs' expert on antitrust collusion, damages
In motion [881], CDK moves to exclude the opinions of Dr. Michael Williams. See [889 7] (report). Williams has a B.A. degree in economics from the University of California, Santa Barbara, and received his M.A. and Ph.D. degrees in economics from the University of Chicago. He is currently a Director at Competition Economics, LLC, specializing in analyses involving antitrust, industrial organization, and regulation. Williams has published articles in a number of academic journals and provided testimony in numerous federal and state courts and administrative bodies.
The Dealership Plaintiffs retained Williams to analyze whether “well-accepted methodologies” and “common evidence” exist from which a fact-finder could (1) determine that Defendants conspired to restrain competition and (2) quantify damages. [889-7] at 7-8.
a. Collusion
Defendants argue that Williams' opinions concerning whether the evidence is consistent with the existence of a conspiracy should be excluded because he fails to analyze “the question at the heart of the case”-which, according to Defendants, is “whether Plaintiffs can muster evidence that ‘tends to exclude the possibility that' CDK and Reynolds ‘acted independently' in deciding to secure their respective DMSs.” [882] at 16 (quoting Matsushita, 475 U.S. at 588). Defendants contend that Williams did not attempt to determine whether the conduct he observed could be explained by CDK's and Reynolds' independent, non-conspiratorial incentives, particularly enhanced system security; an ability to charge higher DIS prices; and reduced legal/reputational risk from system breaches. Id. at 17-18.
The Court is not convinced that this these arguments present grounds for excluding Williams' opinion. As Plaintiffs point out, Matsushita sets a summary judgment standard for cases in which an antitrust plaintiff relies solely on circumstantial evidence to prove the existence of a collusive arrangement in violation of Sherman Act § 1. The parties dispute whether Defendants rely solely on circumstantial evidence or also have evidence that would qualify as “direct” under relevant precedent. See, e.g., Tichy v. Hyatt Hotels Corp., 376 F.Supp.3d 821, 834 (N.D. Ill. 2019) (explaining that “[d]irect evidence is ‘explicit and requires no inferences to establish the proposition or conclusion being asserted'”) (quoting High Fructose Corn Syrup Antitrust Litigation, 295 F.3d at 661). The Court finds it unnecessary and premature to resolve that dispute in the context of a Daubert motion, as the arguments are not developed and the parties crossreference their summary judgment motions, which remain pending and will be decided in a subsequent opinion. For purposes of Daubert, Williams' testimony need not, standing alone, be sufficient to satisfy Matsushita. To be admissible, it suffices that the expert's testimony would “assist the trier of fact to understand the evidence or to determine a fact in issue.” Daubert, 509 U.S. at 591 (quoting Fed.R.Evid. 702(a)). Thus, an economist may testify about factors supporting a finding of conspiracy, regardless of whether any of those factors (alone or in combination) “tend to exclude” the possibility of independent action sufficient to defeat summary judgment. Kleen Products LLC v. International Paper, 2017 WL 2362567, at *14 (N.D. Ill. May 31, 2017) (“Whether [expert] has drawn a reasonable conclusion based on his analysis and actually made a prima facie case for conspiracy is a matter to be examined at summary judgment, and whether he has made a persuasive case is a matter left to the jury.”).
In any event, Williams' report does consider whether there were independent, non-conspiratorial incentives for Defendants' behavior. But instead of viewing the evidence in the same manner as Defendants, Williams concludes that Defendants engaged in a number of actions contrary to their independent self-interests but-for the existence of a conspiracy. See [889-7] at 65 et seq. For instance, Williams opines that Defendants' explanation that their exclusionary conduct and price increases were necessary to protect data security was pretextual. See id. at 74-76.
Defendants also move to exclude Williams on the basis that he “is not ‘any more competent than the average juror' at the ‘interpretation of correspondence and other evidence'” going to the existence of a conspiracy. [882] at 18. The Court rejects this argument for the same reasons explained in relation to Israel. Testimony that an alleged arrangement “furthered [Defendants'] collusive interest, but was inconsistent with Defendants pursuing their independent, unilateral selfinterest, ” Rail Freight Fuel Surcharge Antitrust Litigation, 292 F.Supp.3d at 52, is “precisely the type of inquiry an economist can be expected to make using his expertise.” Processed Egg Products Antitrust Litigation, 81 F.Supp.3d at 424.
Defendants also challenge as unreliable Williams' analysis of seven “plus factors” that allegedly tend to exclude the inference that CDK acted unilaterally. Where, as here, defendants deny a conspiracy on the basis that they were acting independently, “Courts have allowed plaintiffs to proceed by presenting evidence that there is some ‘plus factor' which reduces the probability of independent action.” Alexander v. Phoenix Bond & Indem. Co., 149 F.Supp.2d 989, 1000 (N.D. Ill. 2001). As an initial matter, Defendants question the use of this approach at all, emphasizing that “[i]t has been nearly three decades since the Seventh Circuit has used the phrase “plus factors” in an antitrust case.” [882] at 18 (citing Reserve Supply Corp. v. OwensCorning Fiberglas Corp., 971 F.2d 37, 51 (7th Cir. 1992)). However, the Court finds nothing inherently wrong with using a “plus factors” analysis. It appears to be fairly common in this district to refer to “plus factors” in analyzing Sherman Act § 1 claims. See, e.g., Tichy, 376 F.Supp.3d at 834 (“In considering whether Plaintiff's allegations are sufficient [to support finding of an agreement based on circumstantial evidence], the court asks whether Plaintiff has sufficiently alleged (1) parallel conduct and (2) additional factual circumstances, or ‘plus factors,' indicating an agreement.”); Plasma-Derivative Protein Therapies Antitrust Litigation, 764 F.Supp.2d at 998 (“A court faced with a dispositive motion will examine the evidence as follows. First, the court will look for evidence of interdependence or, put another way, that the defendants have an economic motive to behave in concert. Second, the court looks to all the evidence suggesting agreement, or ‘plus factors,' which can be in the form of economic or non-economic evidence.”); In re Broiler Chicken Antitrust Litigation, 290 F.Supp.3d 772, 797-801 (N.D. Ill. 2017) (discussing “plus factors” identified by plaintiffs in support of Sherman Act claim); Washington County Health Care Authority, Inc. v. Baxter Int'l Inc., 328 F.Supp.3d 824, 840-44 (N.D. Ill. 2018) (same); In re Local TV Advertising Antitrust Litigation, 2020 WL 6557665, at *9-10 (N.D. Ill. Nov. 6, 2020) (same); Standard Iron Works v. ArcelorMittal, 639 F.Supp.2d 877, 896 (N.D. Ill. 2009) (Sherman Act § 1 case using term “plus factor”).
The Seventh Circuit also continues to analyze “plus factors” that may support an inference of conspiracy, even if it has not recently referred to them as such. See, e.g., Omnicare, 629 F.3d at 709 (“Information exchange can help support an inference of a price-fixing agreement[.]”); Kleen Products, 910 F.3d at 936 (a firm's “abrupt change in business practices ... might support an inference of conspiracy”). A plus-factor analysis provides a way to organize circumstantial evidence that, in the plaintiff's view, reduces the probability that defendants were acting independently. Williams does not claim that this is a “rigid framework” that must be applied in a particular way, as Defendants suggest. [1030] at 11; see also [889-7] at 76 & n.242 (“I understand that ‘courts emphasize that these plus factors should not be viewed in a vacuum but rather should be considered in their entirety as the backdrop against which the alleged behavior takes place.'” (quoting ABA Section of Antitrust Law, ANTITRUST LAW DEVELOPMENTS, 7th ed., p. 11 (2012))).
Defendants also challenge the reliability of Williams' opinions concerning the seven plus factors that he identifies. Defendants characterize the opinions, overall, as a “superficial and unreliable analysis” that is “fundamentally unhelpful to a jury.” [1030] at 11. Citing Valspar Corp. v. E.I. Du Pont De Nemours & Co., 152 F.Supp.3d 234, 244 (D. Del. 2016), aff'd, 873 F.3d 185, 201 (3d Cir. 2017), Defendants claim that “Courts have rejected Dr. Williams's liability opinions in cases where he relied on similarly conclusory statements and ‘plus factor' analysis rather than genuine economic expertise.” [1030] at 11. Valspar, however, was decided on a summary judgment motion; it did not involve Daubert issues or suggest that Williams' “plus factor” analysis was not sufficiently relevant or helpful to satisfy Rule 702. Rather, the Valspar court concluded based on the full summary judgment record in a price-fixing case that evidence the defendants increased the frequency of their price increase announcements was insufficient to create an inference of conspiracy. 152 F.Supp.3d at 245-53. More generally, courts considering “plus-factor” analyses in the context of Daubert have allowed such testimony, as “[a]n expert's analysis of plus factors focusing on market conditions and firm behavior finds its roots in the extensive antitrust literature.” Titanium Dioxide Antitrust Litigation, 2013 WL 1855980, at *12 (citing scholarly literature, “including the work of Richard A. Posner, Joseph E. Harrington, and W. Kip Viscusi, as well as the Handbook of Antitrust Economics”); see also, e.g., Fleischman v. Albany Medical Center, 728 F.Supp.2d 130, 159-60 (N.D.N.Y. 2010) (identifying a number of “plus factors” as “relevant to the determination of whether circumstantial evidence points to a conspiracy, ” including “1) a motive to conspire, which can be evidence that the industry is susceptible to price-fixing; (2) noncompetitive behavior, i.e., evidence that the defendants acted contrary to their economic self interest; and (3) evidence of a traditional conspiracy, such as a high level of inter-firm communications that would suggest that the defendants consciously agreed not to compete”).
The Court now turns to the seven specific “plus factors” identified by Williams. Overall, Defendants' criticisms of Williams' report-including whether he characterizes particular emails and other pieces of evidence in the same manner as Defendants, or takes into consideration each and every justification Defendants assert for taking particular actions, see [1030] at 11-12-are ones suited for cross-examination, not reasons to exclude Williams' “plus factor” analysis either in full or in part.
In his first plus-factor, Williams opines based on his price analysis that Defendants' actual DMS and DIS prices exceed but-for prices. Defendants argue that the fact that prices are elevated says nothing about whether Reynolds and CDK actually conspired, because Williams recognizes that each Defendant had unilateral power to increase prices without a conspiracy. Plaintiffs respond that Williams' use of a pre-conspiracy benchmark period controls for the impact of non-conspiratorial supply-and-demand factors, including any unilateral market power that CDK or Reynolds may have had. This is a battle of the experts issue that should be explored in crossexamination. Defendants also criticize Williams for quoting from an article that uses the term “super-plus factor” to refer to such pricing evidence. However, Williams does not himself use the term “super-plus factor” and Plaintiffs express no intention of eliciting the use of this term from their witnesses at trial. To the extent that Defendants are worried about this use of terminology, they are free to raise the issue before trial in a motion in limine.
Defendants group Williams' second and third plus factors, which are that Defendants' communications and monitoring of one another reflect actions against their self-interests but-for the existence of a conspiracy, and that these information exchanges occurred at high levels in the structural hierarchy of Defendants' firms. For instance, in September 2013, Defendants allegedly spoke of “‘forg[ing] a common perspective' on market messaging regarding data security to ‘speak with similar voices to the industry, '” [889-7] at 68, and in December 2014 strategized to “collaborate” on market messaging promoting that it is a “positive outcome” to wind down “unapproved integration, ” id. at 70.
Without going into all of the specific examples provided by Williams, see [889-7] at 6771, Defendants contend that their communications “have a clear and obvious legitimate business purpose-namely, the negotiation and execution of the February 2015 agreements, ” [882] at 20. Defendants also maintain that “simply identifying communications between ‘high level executives' of two firms does not support an inference of conspiracy.” Id. at 22. However, “a high level of inter-firm communications that would suggest that the defendants consciously agreed not to compete” has been recognized as a relevant plus-factor. Fleischman, 728 F.Supp.2d at 159-60; see also Re/Max Int'l v. Realty One, Inc., 173 F.3d 995, 1009 (6th Cir. 1999) (one “important factor[] to evaluate” in determining whether circumstantial evidence tends to exclude the possibility of independent conduct is “whether the defendants have exchanged or have had the opportunity to exchange information relative to the alleged conspiracy”); In re Plywood Antitrust Litigation, 655 F.2d 627, 634 (5th Cir. 1981) (concluding that defendants' parallel pricing conduct plus numerous items of direct evidence of communication between high-level personnel on pricing policy adequately support the jury's verdict for plaintiffs on Sherman Act § 1 claim).
Defendants group Williams' fourth and fifth plus factors, which are the Data Exchange (or “Wind Down”) Agreement and the Reciprocal Data Integration Agreements. Defendants' criticism boils down to disagreement with Williams' opinion that, absent a conspiracy, the agreements would be contrary to Defendants' independent self-interests and Defendants would have competing in the market for DIS service used to access data on Reynolds' DMS. Defendants also argue that Williams is not qualified to offer opinions about the agreements because at his deposition he did not know specifically what the agreements prohibit or allow. See [882] at 23. Williams appropriately did not opine on the legal effect of the agreements. But as an expert in economics, he is qualified to assess whether the level of cooperation embodied in the agreements was anticompetitive and against Defendants' unilateral self-interests. See Allen v. Dairy Farmers of America, Inc., 2013 WL 6909953, at *11 (D. Vt. Dec. 31, 2013) (expert may testify about “economic incentives of the various participants” in the conspiracy based on “the contractual relationships between the participants, Defendants' internal documents, ” pricing structure, and other evidence).
Williams' sixth plus factor is “actions indicating efforts to enforce” the conspiracy-in particular, Defendants' alleged agreement that all written communications about the Data Exchange Agreement, including to vendors, must follow pre-approved templates and should be limited. See [889-7] at 72-73. Defendants argue that these communications were not “enforcement” activities but rather were necessary to effectuate the wind-down. This is another matter of interpretation that should be explored during cross-examination.
The final plus-factor Williams identifies is that Defendants allegedly offered pretextual explanations for their conduct and price increases. Williams focuses on CDK's announcement in June 2015 that it was launching a “SecurityFirst” initiative. Williams points to evidence that, internally, CDK “acknowledged that the ‘objective' of the Secure the DMS initiative was ‘to prevent and/or severely inhibit non-approved access to DMS, forcing vendors into the 3PA or other CDK approved access programs and capture additional revenue opportunities.'” [889-7] at 74. Defendants, citing Lamb's Patio Theater, Inc. v. Universal Film Exchanges, Inc., 582 F.2d 1068, 1070 (7th Cir. 1978), claim that pretextual statements are not plus-factors. But Lamb was decided at summary judgment and merely held that pretextual reasons for the defendants' allegedly collusive conduct-standing alone-would not suffice to prove the existence of a conspiracy. See id. (“even if Lamb's could show that Universal's announced business reasons were not legitimate, such showing would not satisfy Lamb's burden of establishing the existence of the conspiracy element essential to its prima facie antitrust case”). Defendants' alleged pretextual reasons for their agreement is just one piece of evidence going to whether Defendants conspired.
Finally, Defendants challenge Williams' opinions that the DMS and DIS markets are “highly concentrated” and that there are “substantial antitrust barriers to entry” in the DMS market. Defendants argue that this analysis is unhelpful to a jury because “structural evidence does not ‘tend to exclude the possibility of independent action.” [882] at 25 (quoting Kleen Products, 910 F.3d at 935). The Court agrees with Plaintiffs that this opinion should not be excluded because it may “help the jury to understand the economic characteristics of the DMS market and why they tend[] to make that market relatively more conducive to the formation and maintenance of a conspiracy even if, standing alone, they do not answer the ultimate question of whether Defendants conspired.” [992] at 28 (internal quotation marks and citation omitted).
b. Damages
i. Illinois Brick
Williams assumes for purposes of his damages calculation that Defendants' alleged conspiracy caused the Dealership Class both “direct” and “indirect” damages. He calculates both using the same methodology-multiplying an “Overcharge to Vendors” by a “Pass-Through Rate.” The “Vendors” in this scenario are DIS providers that (1) purchase certified integration from CDK and RCI through their 3PA and RCI programs, respectively; and (2) sell DIS services to members of the Dealership Class. Williams concludes that direct and indirect damages are exactly the same, slightly over $339 million. See [889-7] at 91, 99. This total is composed of approximately $202 million in “overcharges of fees for Reynolds' RCI Program” and $137 million in “overcharges of fees for CDK's 3PA Program.” Id. at 99.
Defendants argue that Williams' opinion on “direct” damages must be excluded because it ignores the Court's ruling at the motion to dismiss stage that, under Illinois Brick Co. v. Illinois, 431 U.S. 720 (1977), the Dealership Class could not recover their portion of Defendants' alleged overcharges for DIS. In that ruling [507], the Court explained that under Illinois Brick, “a federal antitrust plaintiff may not seek damages based on alleged supracompetitive prices passed through by a purchaser earlier in the distribution chain.” Id. at 19. The Court concluded that Plaintiffs' exclusive dealing claim based on vendor contracts and their Section 2 monopoly claim were barred by Illinois Brick, because those claims sought to “recover alleged supracompetitive prices passed through to vendors” that paid those prices in the DIS market. Id. at 21. “Furthermore, ” the Court held, “to the extent that Plaintiffs seek to bring a horizontal conspiracy claim based on agreements to restrain competition in the DIS market, Plaintiffs' horizontal conspiracy claim fails under Illinois Brick and Plaintiffs may not proceed under that theory.” Id. That claim would be “derivative of more direct injuries to participants or direct consumers in the DIS market, ” which the Dealer Plaintiffs “implicitly recognize[d] would be barred by Illinois Brick.” Id. at 20. However, the Court declined to dismiss the horizontal conspiracy claim to the extent that it was “based on alleged anticompetitive conduct in the DMS market, ” rather than the DIS market, “as Plaintiffs allege[d] that its DMS lost functionality and is worth less as a result of CDK's agreements with Reynolds.” Id.
Defendants argue that Williams' direct damages analysis runs afoul of the Court's motion to dismiss ruling and must be excluded because, “where plaintiffs seek ‘direct' damages using the same formula employed ‘to calculate their “overcharge” damages,' Illinois Brick bars their claims ‘notwithstanding the fact that the plaintiffs label these damages' as something else.” [882] at 8 (quoting Drug Mart Pharmacy Corp. v. American Home Products Corp., 2002 WL 31528625, at *8 (E.D.N.Y. Aug. 21, 2002)). The Dealership Plaintiffs respond that Illinois Brick does not apply to them because they are direct purchasers of Defendants' DMSs. They contend that it simply does not matter that Williams “referenced passed-on DIS charges” when calculating direct damages. [992] at 11. “By manipulating DMS interfaces to block data integrators, ” the Dealers maintain, “Defendants caused dealers' injury-including reductions in the value and functionality of their DMSs.” Id. at 10. The Dealers explain that they care about “the total value of use in their DMS choice” and “consider data integrator fees when choosing among DMSs.” Id.
The Court finds Defendants' interpretation of Illinois Brick more persuasive and therefore grants their Daubert motion to the extent that Williams includes in his analysis “supracompetitive prices passed through vendors” to the Dealers as direct damages incurred by the Dealers on their Sherman Act § 1 conspiracy claim. [507] at 19. This opinion is contrary to Illinois Brick and to the Court's motion to dismiss ruling. “Expert opinions that are contrary to law are inadmissible.” Loeffel Steel Products, Inc. v. Delta Brands, Inc., 387 F.Supp.2d 794, 806 (N.D. Ill. 2005); see also Luxco, Inc. v. Jim Beam Brands Co., 2016 WL 4611385, at *5 (N.D. Ill. Sept. 6, 2016); System Development Integration, LLC v. Computer Sciences Corp., 886 F.Supp.2d 873 (N.D. Ill. 2012) (expert testimony on subcontractor's economic loss was inadmissible where it included lost opportunity costs that were not a proper component of damages under Illinois law).
In Illinois Brick, the plaintiffs (the State of Illinois and local governmental entities) purchased finished buildings made with “concrete block [priced] $3 million higher by reason of [a] price-fixing conspiracy” between concrete block manufacturers. 431 U.S. at 727. (The manufacturers sold to masonry contractors, which in turn sold to general contractors, from whom plaintiffs purchased the block in the form of masonry structures. Id. at 726.) The Supreme Court held that only the overcharged direct purchaser, and not others in the chain of manufacture or distribution, were entitled to sue for antitrust damages under the Clayton Act. See id. at 728-29. The Court applied the rule equally to plaintiffs and defendants: defendants could not use a pass-on theory as a defense to damages, just as plaintiffs could not use a pass-on theory to obtain damages. See id. at 724-26. The Court recognized that a “one-sided” rule would “substantially increases the possibility of inconsistent adjudications and therefore of unwarranted multiple liability for the defendant by presuming that one plaintiff (the direct purchaser) is entitled to full recovery while preventing the defendant from using that presumption against the other plaintiff.” Id. at 730. The Court was “unwilling to ‘open the door to duplicative recoveries.'” Id. at 731 (quoting Hawaii v. Standard Oil Co., 405 U.S. 251, 264 (1972)).
The Court recognizes that the Dealers, unlike the purchases in Illinois Brick, have a direct contractual relationship with Defendants, from whom they purchase DMSs. But the bottom line remains that the Dealers may not recover under federal law for the overcharges that vendors allegedly passed on to them after purchasing services from Defendants' RCI and 3PA programs in the DIS market. If the vendors had not passed on Defendants' alleged RCI/3PA overcharges to the Dealers, then-according to Williams' report, at least-Defendants' alleged conspiracy would have had no measurable impact on the value or functionality of the DMSs that the Dealers purchased from Defendants. Illustrating this point, the overcharges allegedly passed through to the Dealers are the same injury claimed by vendors, including Plaintiffs Authenticom and Autoloop. Their expert, Israel, calculates damages in the same manner as Williams, estimating damages from overcharges due to DIS price elevation. Authorizing the Dealers to pursue the same overcharges as the vendors runs the same risk of duplicate recoveries that concerned the Illinois Brick Court. 431 U.S. at 731. It would also apply in a one-sided manner, allowing both the Dealers and the vendors to pursue overcharges against Defendants but not allowing Defendants to raise vendors' pass-through of overcharges as a defense to their claims.
The Seventh Circuit's analysis in In re Brand Name Prescription Drugs Antitrust Litigation, 123 F.3d 599 (7th Cir. 1997), abrograted on other grounds by Rivet v. Regions Bank of La., 522 U.S. 470 (1998), is instructive. In that case, pharmacies brought suit against drug manufacturers for an alleged price-fixing conspiracy in violation of § 1 of the Sherman Act, based on the manufacturers' practice of giving discounts only to hospitals and other favored classes of customers. The pharmacies purchased the drugs through wholesalers, which alleged passed on overcharges that they incurred “by virtue of the price-fixing conspiracy.” Id. at 606. The court held that Illinois Brick barred the pharmacies from complaining about the manufacturers overcharging wholesalers, even if “every cent” of the overcharge was passed on to the pharmacies. See id. The Court noted that it could “imagine the ... case reconfigured in a way” that took it “out of the orbit” of Illinois Brick, if the manufacturers' consistent rejection of the pharmacies' request to purchase discounted drugs as “buying groups” constituted “a refusal to enter into direct contractual relations with certain retailers” that could be “successfully challenged as a boycott.” Id. Then, “the Illinois Brick rule, which is a rule concerning overcharges, would fall away” and the “plaintiffs would be permitted to prove up whatever damages they could show had flowed from the boycott, provided they weren't seeking to recover overcharges, for that would entail to very incidence analysis that Illinois Brick bars.” Id. (emphasis added). However, that was “precisely what th[e] [pharmacies were] seeking, ” which the court predicted “may be why the plaintiffs have not cast their case in the boycott mold.” Id. The Dealers, like the pharmacies, seek through Williams' opinion to recover overcharges that vendors paid for DIS-not damages flowing from any allegedly anticompetitive conduct in the DMS market. This is the same “incidence analysis that Illinois Brick bars.” Id.; see also Drug Mart, 2002 WL 31528625, at *7-8 (retail purchasers of prescription drugs from wholesalers were barred by Illinois Brick from asserting price discrimination claim against drug manufacturers; although claim was characterized as one for lost profits as result of manufacturers' refusal to sell to them directly; pharmacies were in fact indirect purchasers seeking to recover price overcharges).
The Supreme Court's recent decision in Apple, Inc. v. Pepper, 139 S.Ct. 1514 (U.S. 2019), does not affect this conclusion. In Apple, the Court held that “iPhone owners [who] purchase apps directly from the retailer Apple, who is the alleged antitrust violator, ” and “pay the alleged overcharges directly to Apple, ” were the “proper plaintiffs” under Illinois Brick “to maintain [an] antitrust suit.” Id. at 1521. The Court rejected Apple's theory that Illinois Brick “allows consumers to sue only the party who sets the retail price, whether or not that party sells the good or service directly to the complaining party.” Id. Unlike the Apple plaintiffs, the Dealers do not purchase DIS directly from, or pay alleged overcharges directly to, CDK or Reynolds. While “Apple confirms that Illinois Brick is a bright-line rule allocating the right to sue to direct purchasers alone, ” Marion Healthcare, LLC v. Becton Dickinson & Company, 952 F.3d 832, 840 (7th Cir. 2020), the Dealers are direct purchasers of Defendants' DMSs-not DISs. Thus, the arrangement through which the Dealer Plaintiffs participate in the DIS market mirrors Judge Wood's classic explanation of the rule of Illinois Brick, quoted with approval by Justice Kavanaugh in Apple: “‘if manufacturer A sells to retailer B, and retailer B sells to consumer C, then C may not sue A.'” Apple, 139 S.Ct. at 1521 (quoting Loeb Industries, Inc. v. Sumimoto Corp., 306 F.3d 469, 481-82 (7th Cir. 2002)). Here, Defendants are in the “A” position; non-party vendors are in the “B” position; Dealers are in the “C” position. Under Illinois Brick and Apple, Dealers can sue Defendants and seek to recover damages as direct purchasers of DMS but cannot- at least under federal law-claim damages stemming from an “overcharge” passed “down a chain of distribution.” Apple, 139 S.Ct. at 1525. While it is true that “CDK may not gerrymander its way out of this lawsuit merely because of the damage calculation method, ” [992] at 11-12 (emphasis added), it is equally true that Dealers cannot gerrymander into their damages claim overcharges that they did not directly pay for services they did not directly purchase, see [1030] at 6-7. The upshot is that the Court grants Defendants' motion to exclude Williams' opinion to the extent that it includes overcharges passed on to the Dealers as an element of damages under the Sherman Act. However, the Court expresses no opinion at this time as to whether Williams can offer the same opinion in support of a claim for damages under the antitrust laws of one or more states, as the parties have not adequately briefed that issue. Therefore, the Court grants Defendants' motion to exclude Williams' opinion concerning the direct damages that the Dealers are allegedly owed under the Sherman Act.
Defendants do not claim that a ruling in their favor on the Illinois Brick argument makes a ruling on their other damages arguments unnecessary. The Court assumes that it does not; at a minimum, there are still state-law antitrust claims pending that may proceed on an indirect damages theory. See [507] at 26 (“many states expressly have rejected Illinois Brick's bar on indirect purchaser claims by enacting ‘repealer statutes' abrogating the Supreme Court's prohibition on indirect-purchaser actions as articulated in Illinois Brick” (quoting In re Dairy Farmers of Am., Inc. Cheese Antitrust Litigation, 2015 WL 3988488, at *6 (N.D. Ill. June 29, 2015)).
ii. Other damages arguments
Apart from their Illinois Brick challenge to Williams' direct damages model, Defendants challenge both his direct and indirect damages models on the basis of reliability. To calculate damages, Williams (like Israel) uses a difference-in-differences regression model. According to the Dealers, the model controls for supply and demand factors, including legal, unilateral conduct on the part of suppliers. The model is intended to estimate the effects of anticompetitive conduct using changes between the benchmark period and the damages period in an outcome variable (in this case, monthly DIS fees) for a treatment group (CDK and Reynolds) and control group (Authenticom and SIS). According to Williams, changes in Authenticom's and SIS's monthly DIS fees between the benchmark and damages periods should provide the counterfactual estimate of how CDK's and Reynolds' monthly DIS fees would have changed but-for the alleged conspiracy.
Defendants argue that the damages model is unreliable for multiple reasons. First, Defendants contend that Williams fails to justify his damages model's “key” assumption-that Defendants' DIS prices moved in “parallel” with an assumed competitive benchmark prior to the start of the alleged conspiracy period in September 2013. Williams made this assumption by “eyeballing” price charts, even though granular price data was available. Defendants assert that this does not satisfy Daubert, especially since Williams admits that he failed to consider a significant non-parallel increase in Reynolds' pricing in August 2013, before the alleged conspiracy period began. In response, the Dealers maintain that Williams used the standard, accepted, peer-reviewed method for determining whether prices are parallel in the benchmark period in a DiD regression analysis-a visual examination of the data. They explain that the “common trends assumption can be investigated using data on multiple periods, where a visual examination of the trends between the treatment group and the control group (to test the parallel paths assumption) is used.” [992] at 16.
The Dealers provide several examples from the economics literature, but very little analysis. Their first example, an excerpt from an econometrics book, does not discuss the subject; rather, it is about the relationship between clustering (correlation within a group of observations) and a model's statistical precision. See [989-4]. The other examples consist of, as Defendants characterize them, “dense technical papers with no pincites.” [1030] at 7, n.2. Pincites, searchable PDFs, and additional analysis from the Dealers all would have been helpful. Nonetheless, these sources do suggest that in a DiD regression analysis, visual examination is used to compare the paths of the test group and control group prior to and after the event/activity being studied. See [989-6] at 7 (study with charts showing DiD estimates for opioid overdose mortality for nonHispanic white adults in study examining association between U.S. automotive assembly plant closures and opioid overdose mortality); [989-7] at 4, 13 (study using DiD analysis to examine expansion of a mine's demand for local inputs on local living standards and compare households located close to the mine to households farther away; chart shows until 2001 real income followed similar trends in both locations, then diverged and showed a relative increase in areas located closer to the mine); [987-8] at 19-21 (using DiD analysis to study whether trade liberalization may have fostered per capital income convergence among liberalizers relative to the control group of the rest of the world and drawing “overall message” from visual tables that trade liberalization did not generate any significant, systematic convergence). Regardless, Defendants fail to explain what more they believe is required or how the examination of more granular price data would have affected Williams' analysis. Further, Williams adequately took into consideration Reynolds' “non-parallel” increase in pricing in August 2013, the month before the “beginning” of the alleged conspiracy period in September 2013. Williams points to evidence that CDK and Reynolds “engaged in anticompetitive communications in the period June 2013 through August 2013, and knew the cooperation was in place.” [992] at 16 & n.4. To the extent that Williams relied on “faulty information, ” this is “a matter to be explored on cross-examination, it does not go to admissibility.” Manpower, 732 F.3d at 809.
Defendants also argue that Williams' damages opinion should be excluded because it does not assume the proper “but-for” world. As they argued in their motion to exclude Israel, Defendants take the position that Williams erred by attributing all of CDK and Reynolds' 3PA and RCI price increases to unlawful conduct even though, in his opinion, CDK and Reynolds “had the power to raise prices to some extent in the but-for world with no unlawful conduct.” [992] at 13. Since Williams “necessarily implies that in the absence of conspiracy a rational firm would have exercised its power to raise prices by some amount, ” he should not have assumed that the entirety of any measured price increase after September 2013 was attributable to the alleged conspiracy. Id. Accordingly, Defendants conclude, “his damages model does not attempt to isolate damages attributable to the alleged conspiracy, meaning his opinions cannot assist the jury.” Id. at 14 (citing MCI, 708 F.2d at 1162).
The Dealers respond that Williams' analysis did consider whether Defendants' conduct could be explained by unilateral motivations. His DiD regression analysis controls for the impact of non-conspiratorial supply-and-demand factors, including any unilateral market power that CDK or Reynolds may have had. Williams also analyzed and disagreed with the arguments offered by CDK's experts (Murphy and Whinston) that even in the absence of a conspiracy, CDK would have closed its DMS and raised its 3PA prices. The Court finds this argument inappropriate for resolution on Daubert motions, for the same reasons explained above in relation to Israel.
Finally, Defendants argue that Williams's damages model is unreliable because it incorporates damages from alleged exclusive-dealing provisions that the Court dismissed under Illinois Brick. According to Defendants, “[a] damages model agnostic to the defendant's liability- which produces the same result whether or not some of the plaintiff's claims fail on the merits- is inherently unreliable and consequently inadmissible.” [882] at 15 (citing Marshfield Clinic, 152 F.3d at 593). The Dealers respond that “a claim, as opposed to the conduct, may disappear for reasons other than a determination that such an antitrust claim constitutes legal conduct.” [992] at 15. “Simply because a claim in a conspiracy has been dismissed at the initial litigation phase does not convert defendants' conduct (in this case, exclusive dealing contracts) into lawful action.” Id. (citing Spray-Rite Service Corp. v. Monsanto Co., 684 F.2d 1226, 1243 (7th Cir. 1982)). The Court declines to exclude Williams on this basis. If Defendants are able to make a convincing showing at summary judgment that the Dealers will be unable “isolate those elements of the plaintiff's damage that were caused by” unlawful conduct, then they may be entitled to judgment in their favor at that time. See Areeda & Hovenkamp ¶ 309c2, at 17. The Court also notes that, even if it accepts Defendants' argument, there are many state law antitrust claims still pending, Illinois Brick does not apply in many states, and it is therefore unclear what relevance Williams' opinion may have despite the dismissal of the federal exclusive dealings claims.
3. Murphy [873] - CDK's expert rebutting Israel and Williams
In motion [873], the Dealer Class and AutoLoop seek to exclude certain opinions of Dr. Kevin Murphy. Murphy holds a doctorate in economics from the University of Chicago and is the George J. Stigler Professor of Economics in the Booth School of Business and the Department of Economics at the University of Chicago, teaching graduate level courses in microeconomics, price theory, empirical labor economics, and the economics of public policy issues. He has authored or co-authored more than 65 articles in a variety of areas in economics and is a Fellow of the Econometric Society and a member of the American Academy of Arts and Sciences. Murphy is also a Senior Consultant at Charles River Associates (“CRA”), a consulting firm that specializes in the application of economics to law and regulatory matters. He has submitted expert reports in numerous cases and testified in various federal courts and before the U.S. Senate and state regulatory agencies.
Murphy was retained by CDK to review damages analyses submitted on behalf of the Dealer Class and the Vendor Class. Murphy does “not specifically address liability” in his report. [875-11] at 8. Instead, he opines on whether Israel and Williams “have demonstrated with proper economic analysis and empirical evidence that the two putative classes suffered antitrust injury and damages.” Id. He summarizes his opinions as follows, id. at 8-9:
• Plaintiffs' experts improperly attribute essentially all the changes in CDK's and Reynolds' prices for integration with their respective DMSs during the claimed damages periods to the alleged conspiracy, because they ignore the unilateral reasons for changes in policies and pricing by these DMS providers.
• Neither Israel nor Williams provides a proper analysis of economic damages because both assume an improper but-for world and consequent pricing that does not reflect how Defendants would have acted independent of any alleged conspiracy.
• CDK had business reasons, unrelated to reducing competition, for securing its DMS against unauthorized access by third-party data integrators that were unrelated to reducing competition. The impact of doing so on prices charged by 3PA program without the challenged agreements is the proper but-for world from which to calculate economic damages, if any, suffered by Plaintiffs due to the challenged 2015 agreements.
• Neither Israel nor Williams offers a damages analysis specific to claimed anticompetitive injury resulting from CDK's unilateral conduct, even though both claim that CDK possessed “unilateral” market power that enabled it to increase prices for its 3PA program by securing its DMS from unauthorized third-party access. Their implicit assumption that CDK could have accomplished the same goals unilaterally and had an incentive to do so means that damages associated with the alleged conspiracy are either zero or substantially smaller than they have estimated, even accepting their methodologies.
Murphy also opines that Israel and Williams' empirical quantifications of damages contain numerous flaws that lead to substantial overestimates of damages, as well as that Williams' passthrough analysis is conceptually and empirically flawed.
Plaintiffs argue that Murphy's opinions should be excluded for three primary reasons. First, Plaintiffs argue that Murphy's opinions are prohibited by the Supreme Court's decision in Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555 (1931). According to Plaintiffs, Murphy's report “assumes the existence of an unlawful conspiracy that harmed MDL Plaintiffs” and, “[f]rom that starting assumption, ” opines that “CDK would have engaged in the same conduct - blocking third-party integrators and raising 3PA prices - absent that conspiracy.” [875] at 15. Plaintiffs assert that Murphy's argument is “no different from the reasoning Story Parchment held ‘must be rejected as unsound.'” [875] at 15 (quoting Story Parchment, 282 U.S. at 562).
Story Parchment involved an alleged conspiracy between three companies to monopolize trade in vegetable parchment, exclude petitioner from the market, and destroy petitioner's business. The jury returned a verdict in petitioner's favor, which was vacated by the appellate court and remanded to the trial court “with directions to enter judgment for respondents upon the ground that petitioner had not sustained the burden of proving that it had suffered recoverable damages.” 282 U.S. at 560. The Supreme Court “rejected as unsound” the appellate court's assumption that respondents' “acts would have been the same if they had been acting independently of one another, with the same resulting curtailment of prices, ” which the appellate court used to justify vacating the jury's verdict. 282 U.S. at 562. The appellate court's assumption was “unsound” because the trial record contained sufficient evidence for the jury to conclude that respondents' conspiracy caused petitioner's antitrust injury. In particular, the record contained “evidence from which the jury reasonably could have found that, in pursuance of the conspiracy, respondents sold their goods below the point of fair profit, and finally below the cost of production; that petitioner had an efficient plant and sales organization, and was producing a quality of paper superior to that produced by either of the three [defendant] companies; and that current prices ... were higher during a period antedating the unlawful combination and price cutting in pursuance of it than afterward.” Id. at 561. The Supreme Court recognized that “[i]t does not necessarily follow, of course, that these higher prices would have continued except for the conspiracy, but it is fair to say that the natural and probable effect of the combination and price cutting would be to destroy normal prices; and there was evidence of the prices received by petitioner before the cut prices were put into operation, and those received after, showing actual and substantial reductions, and evidence from which the probable amount of the loss could be approximated.” Id. In sum, “[u]pon a consideration of the evidence, ” the Court was of the “opinion that it was open to the jury to find that the price cutting and the resulting lower prices were directly attributable to the unlawful combination.” Id.
As to damages, the Court rejected the appellate court's view “that the verdict of the jury ... was based upon mere speculation and conjecture” “in so far as it included damages for” the “difference between the amounts actually realized by petitioner and what would have been realized by it from sales at reasonable prices except for the unlawful acts of the respondents.” Id. at 562. While it was “true that there was uncertainty as to the extent of the damage, ” “there was none as to the fact of damage; and there is a clear distinction between the measure of proof necessary to establish the fact that petitioner had sustained some damage and the measure of proof necessary to enable the jury to fix the amount.” Id. The Court explained that “[t]he rule which precludes the recovery of uncertain damages applies to such as are not the certain result of the wrong, not to those damages which are definitely attributable to the wrong and only uncertain in respect of their amount.” Id.
Story Parchment does not preclude Murphy's criticism of Israel and Williams' opinions concerning the appropriate “but-for” world. Nor does the line of cases which, according to Plaintiffs, has “uniformly ... held that where an antitrust plaintiff ‘can show damages caused by [a defendant's] antitrust violation, the fact that [the defendant] might have caused the same damages by a lawful' means ‘is irrelevant.'” [875] at 14 (quoting Lee-Moore Oil Co. v. Union Oil Co., 599 F.2d 1299, 1302 (4th Cir. 1979); Gelboim v. Bank of Am. Corp., 823 F.3d 759, 777 n.15 (2d Cir. 2016)). Rather, Story Parchment cautions against taking away from the factfinder the determination of how defendants would have acted but-for the alleged conspiracy and what damages are attributable to the conspiracy. Contrary to Plaintiffs' argument otherwise, Murphy does not assume the alleged conspiracy “harmed MDL Plaintiffs, ” [875] at 15, or concede that Plaintiffs “can show [any] damages, ” id. at 14. He assumes “that the Court finds that [the 2015] agreements are illegal, by which [he] mean[s] that the Court concludes that it was illegal for CDK and Reynolds to coordinate and enter into formal contracts memorializing the obligations and benefits of those agreements.” [875-11] at 32. But he disputes that Israel's or Williams' analyses prove that the alleged conspiracy caused Plaintiffs any antitrust injury. See generally Greater Rockford Energy and Technology Corp. v. Shell Oil Co., 998 F.2d 391, 395 (7th Cir. 1993) (“In establishing antitrust injury, courts must first delineate the type of interests protected by the antitrust laws, and second, must determine whether the violation was the cause-in-fact of the injury: that ‘but for' the violation, the injury would not have occurred” (quotingMCI, 708 F.2d at 1161). As Defendants put it, “Murphy does not posit a hypothetical world where Plaintiffs' damages might have or could have been caused by lawful action; he simply points out that Plaintiffs' experts' own affirmative opinions”-which according to Murphy “conflate unilateral and conspiracy damages-infer a but-for world that tends to negate any damages on Plaintiffs' conspiracy claim.” [991] at 16-17; see also [875-11] at 29 (Murphy explaining that “[a] more reasonable assumption in specifying a but-for world where the assumed coordination with Reynolds did not occur is that CDK would have undertaken the same conduct with substantially the same impact on integration prices as it did in the actual world”). For the same reasons explained above in relation to Israel's opinion, it is up to the factfinder to decide which expert's “but for” world is the more convincing one in consideration of the facts in the record.
Plaintiffs' second argument for excluding Murphy's opinion is that there are multiple “improper, speculative assumptions underlying his but-for world.” [875] at 22. As with Defendants' similar criticisms of Plaintiffs' experts, the Court is not convinced that any of the alleged flaws are so serious as to warrant exclusion of Murphy's opinion. For instance, Plaintiffs argue that Murphy failed to engage in economic analysis to show that it would have been in CDK's unilateral interest to engage in the same blocking conduct absent the alleged conspiracy, or to analyze CDK's profitability. However, according to Defendants, Murphy did not ignore these issues but simply found them insignificant in comparison to the strong financial incentive that CDK had to secure its DMS, which he summarizes in his report and discussed at his deposition. See [991] at 12-13. Plaintiffs have not shown that their experts performed such an analysis, either. Plaintiffs also argue that Murphy ignores the evidence that CDK only decided to “secure” its DMS in July 2014 after extensive discussions with Reynolds, and only ‘locked down” its DMS in February 2015 after entering into written contracts with Reynolds. But Murphy does acknowledge the agreements, and concludes that there is nothing in them that “would prevent CDK and Reynolds from competing strongly with each other, or that commits either one to prevent third-party DIS providers from accessing its own or the other's DMS.” [875-11] at 35. Murphy and Plaintiffs' experts simply take different views on the significance of the written agreements.
Plaintiffs further contend that Murphy had no basis for opining that in the but-for world, CDK would have “secured” its DMS and doubled its 3PA prices in September 2013. But according to Defendants, this mischaracterizes Murphy's opinion on the timing of the hypothetical price increase. To the extent that there is any inconsistency between the timing set out in his report and in his deposition testimony, this is a matter to explore in cross-examination.
The third and final major criticism of Murphy's opinions-which is advanced by the Dealers only-is that his pass-through regression analysis and criticism of Williams' pass-through regression analysis should be excluded because he assumes facts not present in this case. Williams calculated an average pass-through of the alleged 3PA and RCI overcharges based on a sample of pricing data from 13 apps for which vendors included a line item for “DMS fees” (or similar) on their invoices to dealers. Many vendors did not include such line-items, and were not considered by Williams. Williams found a high level of variation in the pass-through rates in the itemized invoices, with some vendors passing on just a small fraction of the alleged overcharge and other “passing on” DMS fees that were substantially larger than the alleged overcharge. Murphy criticized this approach on the basis that dealers care about the overall price for their apps, rather than how the prices are broken down on their invoices, and have an incentive to portray price increases as increases in their underlying input costs. Murphy therefore found it appropriate to calculate an average pass-through rate differently, looking to how overall app prices changed when vendors' DIS costs changed. However, as Plaintiffs point out, this method has its flaws as well, because app prices are affected by demand and cost factors other than data integration fees charged by data integrators to vendors. It is also possible, and Defendants point to at least some evidence in the record supporting this theory, that vendors may have reduced app prices to offset increases in DIS fees passed on to dealers. See [991] at 24. Neither expert's methodology is flawless, but both at minimum offer a “rational connection between the data and the opinion.” Manpower, 732 F.3d at 809. This topic is also better left to the trier of fact. See generally Wipf v. Kowalski, 519 F.3d 380, 385 (7th Cir. 2008) (“Especially in a case of dueling experts, ... it is left to the trier of fact, not the reviewing court, to decide how to weigh the competing expert testimony.”).
4. Lawton [879] and Rubinfeld [867] - Authenticom's lost profits damages expert and Defendants' rebuttal expert, respectively
Motions [879] and [867] both seek to exclude expert opinions concerning Authenticom's lost profits damages; the latter motion also addresses a number of opinions relating to Defendants' counterclaims, which are addressed below. In motion [879], Defendants move to exclude the opinion of Authenticom's damages expert, Catharine Lawton. See [889-4] (Lawton's expert report). In motion [867], Plaintiffs move to exclude the opinion of the defense expert who rebuts Lawton's opinion, Professor Rubinfeld. See [871-1] (Rubinfeld's rebuttal expert report).
Lawton is a Managing Director at Berkeley Research Group, which provides consulting services on a variety of financial and economic issues. She holds a bachelor of science degree in finance and economics. In her report, Lawton assumes that the factfinder will determine liability in Authenticom's favor. She describes in detail the evidence in the record concerning the background of the DMS industry and Defendants' positions within it; Authenticom's financial performance before and during Defendants' alleged anticompetitive conduct; and the link between Defendants' alleged anticompetitive conduct and harm to Authenticom.
Lawton then attempts to quantify Authenticom's damages using a “yardstick” methodology. Her “yardstick” compares Authenticom's volume of sales to dealerships that use CDK and Reynolds DMSs (its “other connections”) with its volume of sales to dealerships that use DMSs other than CDK or Reynolds (its “CDK/Reynolds connections”). It presumes, in essence, that but-for Defendants' alleged illegal conduct, Authenticom's sales to the two groups of dealers would have moved in tandem. She explains that this is an appropriate yardstick because: 1) Authenticom sells the same products at the same (or similar) prices to both categories of dealers; 2) both groups of dealers use DIS providers like Authenticom in a similar manner; and 3) there is a correlation between her yardstick and Authenticom's CDK/Reynolds connections from the period prior to the alleged conspiracy. Lawton also considered whether any adjustments to the yardstick should be made to account for market factors that influenced the yardstick but would not have occurred in a but-for world. She adjusted her estimate of damages downward to account for the possibility that the alleged conspiracy caused some dealers to switch away from CDK and Reynolds to other DMSs. She adjusted the yardstick by using Authenticom's historical growth rates rather than its observed growth rates after February 2018 to account for the fact that Defendants' conduct allegedly forced Authenticom to lay off about 60 percent of its staff in February 2018. Finally, Lawton performed several “reasonableness checks” to ensure that her damages estimate aligned with available contemporaneous economic data. She concluded that her estimates were in line with the internal valuations of Authenticom made by CDK and CDK's investment consultant. She concludes based on her yardstick model that Authenticom's damages from the alleged anticompetitive conduct range between $88.9 million and $107.1 million based on damages starting in March 2015.
Rubinfeld is the Robert L. Bridges Professor of Law and Professor of Economics Emeritus at the University of California, Berkeley and Professor of Law at New York University. Among other qualifications, he has served as Deputy Assistant Attorney General in the Antitrust Division at the U.S. Department of Justice from June 1997 through December 1998; authored two textbooks, Microeconomics and Econometric Models and Economic Forecasts (both with Robert Pindyck); and written several academic articles that focus on damages methodology. Rubinfeld was retained by Defendants to review and assess Lawton's damages analysis and, to the extent necessary, provide where feasible an alternative estimation of damages.
Rubinfeld opines that Lawton's methodology is flawed because she does not clearly articulate a “but-for” world. Rubinfeld contends that as a result of Lawton's commingling of distinct claims of misconduct-(1) an alleged conspiracy between CDK and Reynolds to block hostile integrators; (2) the allegedly unlawful 2015 Agreements themselves; and (3) the allegedly unlawful exclusive dealing provisions in CDK's and Reynolds' contracts with vendors-her model is incapable of disaggregating damages should a jury find some, but not all, of these alleged practices to be unlawful. Rubinfeld further argues that even if a jury concludes that all of these acts are unlawful, Lawton's damages model is flawed because she fails to identify a realistic but-for world that takes into account the alleged economic reality that, absent any collusive behavior, Reynolds and CDK would have had strong unilateral incentives to secure their DMSs. In addition, Rubinfeld identifies a number of alleged conceptual flaws and methodological errors that, he contends, make Lawton's damages estimates unreliable and unsupportable.
Defendants move to exclude Lawton, while Authenticom moves to exclude Rubinfeld. Several of the parties' criticisms are overlapping and therefore the Court considers them together where possible.
Defendants' first argument for excluding Lawton is that the “yardstick” she chose is not reasonably comparable to the impacted market because it includes thousands of connections that were improperly classified as “other connections. More particularly, Defendants maintain that thousands of connections that Lawton classified as “other connections” should have been classified as “CDK/Reynolds connections”; that Lawton included connections to software application vendors that are not DMSs; and that Lawton should not have included connections to its own inventory management application, Carpod, or connections to “independent” dealers. Lawton and Authenticom respond to each of these criticisms in some detail. See [999] at 17-23. Authenticom, in turn, moves to exclude Rubinfeld's criticism of Lawton on the basis that his critiques of the connections that she included in her yardstick “rely on an insufficient factual basis, ” [871] at 19, as demonstrated by Rubinfeld's alleged inability to adequately explain his criticisms when he was deposed, see id. at 20-22. The Court does not find it useful to go into the details of these arguments because the parties' criticisms all are “really about the correctness of [the experts'] conclusions, not the reliability of [their] methodology.” In re Fluidmaster, Inc., Water Connector Components Products Liability Litigation, 2017 WL 1196990, at *22 (N.D. Ill. Mar. 31, 2017). As both sides acknowledge, the yardstick approach is a well-established methodology; the parties therefore “cannot complain that the methodology itself is unreliable.” Areeda & Hovenkamp, ¶ 399b, at 2. Rather, their criticisms concern “how the selection of data inputs affect the merits of the conclusions produced by an accepted methodology, ” and therefore are properly handled through cross-examination and “left to the jury” for ultimately resolution. Manpower, 732 F.3d at 808; see also Zimmer Nexgen Knee Implant Products Liability Litigation, 2015 WL 3669933, at *25 (“Rule 702 does not permit ‘the district court to choose between * * * two [competing] studies at the gatekeeping stage' or evaluate the quality of an expert's data, inputs, or conclusions.”). As Rubinfeld's alleged correction of Lawton's inputs demonstrates, these are not issues that bear on the reliability of Lawton's methodology or the sufficiency of the available facts or data.
Defendants' next argument is that Lawton's “yardstick” was inflated by the alleged anticompetitive conduct because, according to Authenticom's theory of the case, “competition for DMS polling connections has decreased as a result of Defendants' alleged conduct” by forcing other competitors out of business. [880] at 22. Defendants claim that “[c]ourts frequently exclude damages models ... where a but-for-world assumes little competition in a ‘supposedly much more attractive environment.'” Id. (quoting S. Pac. Commc'ns Co. v. AT&T Co., 556 F.Supp. 825, 1077-78 (D.D.C. 1982)). Assuming this is correct, Defendants have not pointed to concrete evidence that competition to Authenticom for “other connections” has, in fact, decreased substantially since the beginning of the alleged conspiracy. Defendants do not concede this and according to Lawton's report, while a number of DIS providers have exited the market, several others continue to offer DIS today-especially to customers that do not depend on CDK and Reynolds DMSs. Defendants and Rubinfeld also do not suggest what adjustments should have been, but were not, made to Lawton's yardstick. More fundamentally, “[a]rguments about what factors an expert should have controlled for in conducting a yardstick analysis generally go to the weight, rather than the admissibility, of the expert's testimony.” Tawfilis v. Allergan, Inc., 2017 WL 3084275, at *6 (C.D. Cal. June 26, 2017); see also In re Prograf Antitrust Litigation, 2014 WL 7641156, at *3 (“Courts have been reluctant to determine comparability before trial.” (collecting cases)).
Defendants also make several arguments concerning Lawton's alleged failure to account for the impact that Defendants' lawful, independent actions had on Authenticom's ability to compete in the DIS market. Defendants contend that Lawton failed to adjust her yardstick to take into account certain blocking activities that Reynolds engaged in prior to the start of the alleged conspiracy in September 2013. Relatedly, Defendants argue that Lawton's opinions must be excluded because her damages model fails to isolate damages caused by alleged illegal conduct. According to Defendants, “Lawton concedes that she has included damages in her calculation (which she cannot disaggregate) that were caused by actions Defendants took unilaterally and lawfully before the conspiracy and that they would have continued to take regardless of the conspiracy.” [880] at 25. Authenticom disputes, however, whether Reynolds' blocking activities would have continued after September 2013 in the absence of a conspiracy. According to Authenticom, the costs that Reynolds' unilateral blocking efforts from June to August 2013 imposed on Reynolds were among the reasons that Reynolds sought to conspire with CDK. That is, competition from CDK constrained Reynolds' efforts to close its DMS; the conspiracy allegedly eliminated this competitive constraint. On the flip side, Authenticom argues that Rubinfeld should be excluded because at his deposition he “conceded basic facts showing that the conduct of Defendants prior to the conspiracy differed significantly from their conduct after they conspired- and admitted that he had done nothing to account for them.” [871] at 17.
As the Court determined above with regard to Israel, the parties' competing arguments on what the “but for” world should look like go beyond the scope of Daubert. As Rubinfeld put it at his deposition, “any business decision has pluses and minuses, ” [871] at 17-leaving ample room for disagreement concerning what actions Defendants could or would have undertaken independently “but for” the alleged conspiracy. These arguments may be explored during cross- examination (assuming neither side is able to convince the court that it is entitled to summary judgment).
Defendants' next criticism of Lawton is that she allegedly manipulated Authenticom's profit margins to increase damages. In particular, Defendants contend that Lawton inappropriately categorized Authenticom's development costs as “fixed” rather than “variable.” Lawton explains that she classified development costs as she did after speaking with Authenticom's controller, who told her that development costs related to the development of Authenticom products such as DealerVault and did not vary with incremental volume. Defendants also criticize Lawton for cherry-picking the profit margin her analysis assumed based solely on what would produce a higher damages figure. However, Lawton explains why the margin she projected was equal to whichever of two figures was higher-the weighted average of the variable margin for the prior 12 months, or the margins Authenticom actually achieved. Lawton reasons that, “if Authenticom was able to increase its margins despite the negative impacts of the Defendants' alleged anticompetitive conduct on its margins, then those actual margins represent the best estimate of the but-for margins rather than the pre-damages period margins.” [999] at 29. Both of Defendants' criticisms raise “battle of the experts” issues that should be resolved through cross-examination, not Daubert motions.
Finally, Defendants argue that “three hundred pages of ‘factual' advocacy” in Lawton's opinion should be excluded” because they constitute an “improper advocacy-inspired interpretation of evidence.” [880] at 29. Defendants do not point to any particular portions of the report that they find to be objectionable. The Court will not exclude any portion of Lawton's report on this ground as it was appropriate-and indeed necessary-for Lawton to disclose the factual basis for her opinions. Lapsley v. Xtek, Inc., 689 F.3d 802, 810 (7th Cir. 2012) (“of course, an expert who plans to testify to an opinion must make the basis of that opinion available for evaluation by the court and opposing parties”); Fed.R.Civ.P. 26(a)(2)(B)(iii) (requiring expert report to include “the facts or data considered by the witness” in forming her opinions).
The Court now turns to Authenticom's remaining grounds for excluding Rubinfeld's opinions rebutting Lawton. As alluded to above, Authenticom criticizes Rubinfeld for being unable to recall the details of all his opinions during his deposition, which Authenticom suggests is due to Rubinfeld's overreliance on his staff. The Court does not find these sufficient reasons to exclude his opinions. Defendants point out that “Rubinfeld was transparent about his process, pursuant to which he wrote much of the first draft of his reports, ” supervised his staff, and “thoroughly reviewed portions drafted by his assistants.” [994] at 14; see also Dura Automotive Systems of Indiana, Inc. v. CTS Corp., 285 F.3d 609, 612 (7th Cir. 2002) (“An expert witness is permitted to use assistants in formulating his expert opinion, and normally they need not themselves testify.”); Zollicoffer v. Gold Standard Baking, Inc., 335 F.R.D. 126, 149 (N.D. Ill. 2020) (“Experts can rely on assistants to formulate an opinion, and can rely on an assistant's work within a certain discipline without being an expert in that discipline themselves[.]” (internal citations omitted)). There is no suggestion that Rubinfeld's assistants “exercise[d] professional judgment that is beyond [his] ken.” Dura Automotive, 285 F.3d at 613. To the extent that Rubinfeld is unable recall particular details of his opinions, or needs to reference his reports before answering, this might hurt the effectiveness of his testimony at trial but is not a sufficient reason to exclude him under Daubert.
Authenticom also seeks to exclude Rubinfeld's criticism of Lawton's treatment of connections associated with two dealer groups, Lithia and Group 1, on the basis that he failed to disclose this opinion in his rebuttal report, even though at that time he had the data necessary to do so. Authenticom asserts that this testimony should be barred as untimely under Federal Rules of Civil Procedure 26 and 37. In response, Defendants represent that Rubinfeld's criticisms were included in his report, in which he showed that the Lithia and Group 1 connections should be categorized as “Non-‘Franchise New Car DMSs'” (which should be included in Lawton's yardstick) rather than “Franchise New Car DMSs” (which should not be included). [994] at 21. Even if Rubinfeld had not disclosed this criticism in his report, the Court would not be inclined to exclude it, but rather would provide Authenticom with an opportunity to supplement Lawton's report. To the extent that Authenticom believes that supplementation is necessary, it is free to file an appropriate motion requesting leave to supplement.
For these reasons, the Court denies Defendants' motion to exclude Lawton's opinions [879] and also denies Plaintiff's motion to exclude the opinions of Rubinfeld that rebut Lawton [867]. To the extent that motion [867] addresses other topics and/or rebuts other experts, those issues are considered in the appropriate sections below.
5. Klein [863] - MVSC's expert on lost profits
In motion [863], Reynolds seeks to exclude the testimony of Plaintiff MVSC's damages expert, Gordon Klein (“Klein”). Klein is a certified public accountant (“CPA”) and an attorney. He has been a faculty member of UCLA's Anderson School of Management since 1981, teaching a variety of courses, and was a faculty member of UCLA's school of law from 1987 to 2000. He has authored several books on managerial accounting, cost accounting, introductory accounting, and managerial finance. He has served as an expert witness in numerous cases concerning the measurement of lost profits; the measurement of variable, fixed, incremental and non-incremental costs; enterprise valuation, damages, and unjust enrichment. He also has experience as a derivatives trader and consultant in analyzing the business models, financial performance, and valuation of publicly traded and privately held companies and as an advisor to closely held businesses and entrepreneurs.
As recounted in the background section above, MVSC provides EVR (electronic vehicle titling and registration) software to auto dealers. CDK and Reynolds also provide EVR software to dealers through CVR-a joint venture that is 80% owned by CDK and 20% owned by Reynolds. MVSC alleges that CDK and Reynolds blocked MVSC's integration with their DMSs to prevent MVSC from expanding its business in four states: California, Illinois, Virginia, and Wisconsin. More particularly, MVSC alleges that Defendants (1) blocked its membership in CDK and Reynolds' respective third-party access programs; (2) blocked its automated access to data through third-party integrators such as Authenticom; and (3) marketed CVR's status as the EVR provider with “certified” integration to the CDK and Reynolds DMSs through third-party access programs (whereas MVSC was not “certified”). MVSC refers to this generally as Reynolds' and CDK's “interference.”
According to MVSC, without DMS integration, its EVR services were slower, required manual entry of data or another workaround, and inhibited MVSC's ability to offer “real-time” registration and titling services, which allow consumers to receive their registration and license plates at the time of sale. The absence of automated real-time data access allegedly put MVSC at a competitive disadvantage, particularly in states such as Illinois and Virginia that require EVR transactions to be completed at the point of sale. While MVSC has developed certain manual “workarounds, ” it remains at a competitive disadvantage because the workarounds require dealership employees to expend time and effort in performing manual tasks and learning workaround procedures. Manual workarounds also increase the risk of errors and developing them imposes substantial financial costs of MVSC. MVSC and CDK reached a settlement in October 2019. Since that time, MVSC has had access to CKD's DMS as an approved partner in the 3PA program.
MVSC retained Klein to determine the damages that MVSC has allegedly sustained due to Defendants' interference. See [865-7] at 6 (Klein's report). Klein opines that, as a result of the interference, MVSC has sustained lost profits in four geographic markets: the states of California, Illinois and Virginia (where MVSC currently does business) and Wisconsin (where MVSC has taken steps to enter the EVR market, but has not been authorized to provide service). Klein asserts that his focus on four states only is conservative because, absent Defendants' interference, “it is quite likely that MVSC would have entered additional states, not just Wisconsin.” Id. at 7.
MVSC began competing in California in 2008 and currently processes approximately 60% of all new vehicle dealership EVR transactions in that state. MVSC began efforts to enter the Illinois market in 2013 and in 2015 was formally authorized by the state of Illinois to “develop an EVR solution.” [865-7] at 12. In Virginia, MVSC became an approved EVR vendor in 2014. MVSC has taken steps to enter the Wisconsin EVR market since 2014, but as yet has not been approved by the state to offer EVR services. Klein identifies two primary reasons for attributing this failure to Defendants' interference: First, the interference allegedly has diverted financial and management resources away from expansion efforts toward the costly and time-consuming projects of responding to the interference and developing workarounds to gain access to data. Second, Klein infers that the interference has reduced state regulatory agencies' confidence in MVSC's ability to provide services to the market.
Klein estimates that, but for Defendants' interference, MVSC would have obtained between 60% and 70% of the market share in each of the four states. Klein also considered the possibility that MVSC's damages were attributable to factors unrelated to Defendants' interference (including inadequacy of MVSC's products or services, regulatory barriers in Wisconsin, inability to roll out a product quickly enough, and lack of current availability beyond a few states) and concluded that these factors were insignificant. Based “principally on [his] analysis of MVSC's rapid growth in California prior to the onset of the [i]nterference and based on MVSC's success in Oregon, a state in which MVSC was not impeded by” interference, Klein concludes that interference “has been, at the very least, a substantial factor in causing MVSC's damages.” [8657] at 7. Taking into account incremental costs saved, Klein calculated lost profits as the difference between (a) profits that would have been earned in the absence of interference; and (b) actual profits earned.
Reynolds seeks to exclude Klein's testimony on several grounds. First, Reynolds argues that Klein “fails to disaggregate harm to MVSC caused by an alleged unlawful conspiracy between Reynolds, CDK and CVR from harm caused by each defendant's unilateral conduct which would still exist in the but-for world.” [865] at 13. However, in its opening brief, Reynolds does not identify any particular lawful “unilateral conduct” that it believes caused MVSC's damages. In its reply, Reynolds asserts that Klein “fails to take into account, for example, state budget constraints that impact MVSC's prospects of being approved to provide EVR services; the nature of EVR competition in a particular state; and the price and quality of MVSC's EVR services in a particular state.” [1039] at 7. Generally, arguments raised for the first time in a reply brief are waived, due to the lack of opportunity for the other side to respond. See O'Neal v. Reilly, 961 F.3d 973, 974 (7th Cir. 2020) (“[W]e have repeatedly recognized that district courts are entitled to treat an argument raised for the first time in a reply brief as waived.”); Narducci v. Moore, 572 F.3d 313, 324 (7th Cir. 2009) (“[T]he district court is entitled to find that an argument raised for the first time in a reply brief is forfeited.”). Regardless, Reynolds does not identify anything in the record suggesting that the factors it mentions in passing do, in fact, contribute to MVSC's alleged lack of ability to compete in the four identified markets. Reynolds' cursory reference to other market forces that may be at place is not enough to convince the Court that exclusion under Daubert is required. Cf. Mitsui Sumitomo Ins. Co. v. Moore Transp., Inc., 500 F.Supp.2d 942, 950-51 (N.D. Ill. 2007) (perfunctory, undeveloped, and skeletal arguments are waived).
Neither of two main cases on which Reynolds' argument relies, MCI, 708 F.2d 1081, or Marshfield Clinic, 152 F.3d 588, was decided in the context of a Daubert motion. In MCI, the court of appeals reversed, in part, a jury verdict in the plaintiff's favor in an antitrust action brought under § 4 of the Clayton Act. In that case, the complaint contained twenty-two counts. The district court granted summary judgment for the defendant on seven counts, the jury found for the defendant on five of the remaining fifteen, and the court of appeals set aside the jury's verdict in the plaintiff's favor on several others. The plaintiff's damages study, however, assumed that all twenty-two of the defendant's charged acts were illegal, and its lost profits study did “not establish any variation in the outcome depending on which acts of [the defendant] were held to be legal and which illegal.” Id. at 1163. This left the jury “with no way to adjust the amount of damages to reflect lawful competition, ” id., and required the case to be “remanded for a proper determination of damages, ” id. at 1160. Here, there has been no determination that any of the conduct on which Klein's damages analysis relies is “lawful.” While MCI suggests reasons why Defendants could potentially be entitled to summary judgment or a directed verdict-or even that additional work might needed to make “a proper determination of damages”-it does not warrant excluding Klein on the basis of failure to comply with Daubert.
In Marshfield Clinic, a jury found in the plaintiff's favor in an antitrust case arising out of allegations that the defendant, the dominant provider of medical services in part of Wisconsin, illegally monopolized the provision of certain medical services in its region, enabling it to increase its prices for those services above competitive levels to the detriment of the insurer plaintiff. The appellate court threw out the jury's verdict on all but one claim, for division of markets, and remanded for a new trial limited to damages on that claims. Following remand and additional discovery, the district court granted summary judgment to the defendant on the basis that the plaintiff had produced no evidence that it suffered any injury as a result of the sole remaining claim. See Marshfield Clinic, 152 F.3d at 590-91. In affirming the district court's judgment, the appellate court reviewed multiple reports by the plaintiff's two economic experts, as well as “some nonexpert evidence of damages, ” and also found that the plaintiff “waived the right to seek damages on the basis of its opponent's expert reports”-which showed that the plaintiff suffered some damages, but much less than the plaintiff's experts calculated. Id. at 593-95. Writing for the panel, Judge Posner characterized plaintiffs' two experts reports as “worthless” because they “fail[ed] to correct for salient factors, not attributable to the defendant's misconduct, that may have caused the harm of which the plaintiff is complaining do not provide a rational basis for a judgment.” Id. at 593. Significantly, “[a]s far as the record disclose[d], ” the defendant charged a “higher average price per patient” than other Wisconsin providers of medical services due to “differences in the treatment mix, ” rather than the defendant's unlawful division of markets. Id. at 593-94. In this case, by contrast, Defendants' Daubert briefing does not identify, let alone provide a comprehensive assessment of the full record concerning, any lawful causes of MVSC's alleged damages. It would be premature to exclude Klein without having a fuller picture of the EVR market and the panoply of relevant evidence.
In a related argument, Reynolds contends that Klein “applied no expert methodology at all” to opine that Defendants' alleged interference was the primary factor that caused MVSC's harm. [865] at 16. Of course, an expert may not rely solely on his ipse dixit; however, that is not what Klein has done, even if there is room for debate about whether he considered all potential causes of MVSC's antitrust injury or gave them their appropriate weight. Reynolds' criticisms suggest that Klein's assumptions are “vulnerable, ” but not so speculative or lacking in basis as to be inadmissible. Stollings, 725 F.3d at 768. To the extent that Reynolds believes that Klein “makes unfounded or incorrect assumptions, [it] remains free to address those deficiencies through vigorous cross-examination.” Africano v. Atrium Medical Corp., __ F.Supp.3d __, 2021 WL 4264237, at *4 (N.D. Ill. Sept. 20, 2021).
For the California part of his analysis, Klein analyzed MVSC's growth rate “before and after” the interference, which is a “generally accepted method[] for proving antitrust damages.” Conwood Co. v. U.S. Tobacco Co., 290 F.3d 768, 793 (6th Cir. 2002) (citing Eleven Line, Inc. v. N. Texas State Soccer Ass'n, Inc., 213 F.3d 198, 207 (5th Cir. 2000)). He also relied on direct record evidence that California dealers had declined to purchase MVSC's services because of interference. Reynolds criticizes this part of his analysis, on the basis that Klein relies on account information provided by MVSC's COO, Joe Nemelka, concerning why certain dealers declined to do business with MVSC. Nemelka provided a sworn declaration explaining that certain dealers had cited MVSC's lack of “certified” access to data or automated access as a primary or significant reason for leaving MVSC or deciding not to use MVSC. Klein also reviewed other records, such as emails, that were consistent with Nemelka's declaration. Despite Reynolds' claim that this was improper, an expert does not run afoul of Daubert by relying on representations provided by a client. “[I]t is proper for counsel to furnish factual assumptions to experts as long as the factual assumptions are supported by the record.” Artunduaga v. University of Chicago Medical Center, 2016 WL 7384432, at *5 (N.D. Ill.Dec. 21, 2016); see also Williams v. Illinois, 567 U.S. 50, 69 (2012) (“Modern rules of evidence continue to permit experts to express opinions based on facts about which they lack personal knowledge[.]”); Tuf Racing Products, Inc. v. American Suzuki Motor Corp., 223 F.3d 585, 591 (7th Cir. 2000) (CPA could calculate discounted present value of client's lost future earnings using financial information furnished by client and assumptions given to him by counsel). That Reynolds may be able to point to some contrary evidence does not change the Court's conclusion, as “[e]xperts routinely base their opinions on assumptions that are necessarily at odds with their adversary's view of the evidence.” Richman v. Sheahan, 415 F.Supp.2d 929, 942 (N.D. Ill. 2006). To the extent that Klein allegedly “reli[ed] on faulty information, ” this is “a matter to be explored on cross-examination; it does not go to admissibility.” Manpower, 732 F.3d at 809.
As to all four states, Klein's conclusion that interference was a substantial factor in causing MVSC's damages relied in part on MVSC's rapid growth in California prior to the onset of the interference and MVSC's success in Oregon-a state in which MVSC was not impeded by interference. [865-7] at 7. Klein also explained why he rejected other potential causal factors based on his experience in business plan development and related areas. First, he considered that MVSC may have offered an inferior product but discounted that possibility due to evidence of dealers' satisfaction with MVSC's product and MVSC's track record of success in other states. Second, Klein explained that MVSC's damages could not be attributed to the length of time taken to develop manual workarounds because that development was resource-intensive, still left MVSC with a subpar method for accessing data on the DMS, and in any case, was only necessary because of the interference. Third, he rejected the possibility that MVSC's low market share in some states could be attributed to its lack of national scale because MVSC was able to obtain 60 percent market share in California without national scale. There is “a link between the facts [and] data” Klein has “worked with and the conclusions [his] testimony is intended to support, ” which satisfies Daubert. Mamah, 332 F.3d at 478. “Our system relies on cross-examination to alert the jury to the difference between good data and speculation.” Schultz, 721 F.3d at 432.
Finally, the Court considers whether Klein adequately considered other potential causes of MVSC's failure to obtain any market share in Wisconsin. In his report, Klein reasoned that MVSC's delayed entry into Wisconsin could not be attributed to regulatory barriers unrelated to the conspiracy both because MVSC had been able to receive approval from other state regulators and because the difficulty that MVSC has had in receiving regulatory approval in Wisconsin is due in substantial part to MVSC's lack of automated or certified data integration to Defendants' DMS. At his deposition, Klein was confronted with correspondence from the Wisconsin Department of Transportation, dated February 2017, indicating that MVSC's application had been rejected because it was not accepting any new vendors into its EVR program. MVSC had not provided this correspondence to Klein (or Reynolds) and Klein did not consider it in his report. See [865] at 22-23 & n.11; [865-2] at 38-41. “Ignoring relevant data is not a scientifically valid method.” Cates v. Whirlpool Corp., 2017 WL 1862640, at *15 (N.D. Ill. May 9, 2017). However, the Court cannot say that is what Klein has done here, even though it is a close call. As Klein pointed out during his deposition, the February 2017 letter does not state that Wisconsin was closed to other EVR providers at all times relevant to Klein's analysis and is only part of the record concerning this topic. Klein also considered evidence in the record indicating that Wisconsin was “very interested” in MVSC's 2014 application to be an EVR provider, and considered but rejected that application. [1009] at 24-25; see also [1009-7] at 3; [866-12]. MVSC maintains that it will prove at trial that “the rejection was due in substantial part to the interference as demonstrated by CVR using the fruits of that interference-MVSC's lack of integration-in presentations to Wisconsin regulators.” [1009] at 25. Reynolds offers its own experts, Rubinfeld and Bresnahan, to criticize Klein's opinion. The Court finds it appropriate to leave the resolution of the competing experts' testimony to the factfinder. See Wipf, 519 F.3d at 385 (“Especially in a case of dueling experts, ... it is left to the trier of fact, not the reviewing court, to decide how to weigh the competing expert testimony.”).
C. Other Experts
1. Stejskal [883] - Authenticom, Autoloop, MVSC, and the Dealers' expert on DMS switching costs and third-party data integrator security
In motion [883], CDK and Reynolds seek to exclude certain opinions of Allan Stejskal (“Stejskal”), who was retained by Authenticom, MVSC, the Vendor Class, and the Dealership Class (“Plaintiffs”) to opine on several facets of the automotive industry, including the DMS market, third-party vendor applications, and access to dealer data and data integration service providers. See [889-12] (Stejskal's report).
Stejskal has over 25 years of experience in the automotive retail industry. From 1995 to 2000, Stejskal was a Vice President of CDK, which was then called ADP Dealer Services, where he ran CDK's nascent e-commerce division. From 2007 to 2014, Stejskal was an executive at Dealertrack, which is the nation's third-largest DMS provider behind CDK and Reynolds. Among other roles, Stejskal was the General Manager running Dealertrack's DMS business from 2010 to 2012. He created Dealertrack's OpenTrack certified data integration program, which allows third parties to access dealer data stored on the Dealertrack DMS. From 2017 to 2018, Stejskal was CEO for an international DMS company, Incadea, which is CDK's primary competitor in Europe and South America. On the dealer side of the automotive industry, Stejskal has served two stints as a senior executive at AutoNation, the largest auto dealership in the United States, most recently from 2006 to 2007. In that role, he was responsible for AutoNation's relationship with DMS companies and oversaw the DMS conversion process when AutoNation acquired new stores. Also from 2006 to 2007, Stejskal founded and was President of Secure Access (“OSA”), an industry coalition that advocated for dealer control over dealer data. During that time, he developed data security guidelines to help dealers safeguard their data.
Motion [883] concerns two of the topics covered by Stejskal's expert report-DMS switching and third-party data integrator security. In the section of his opinion concerning DMS switching, Stejskal identifies reasons why a dealership's decision to switch DMSs is “not taken lightly, ” which is a hurdle for DMS vendors to overcome when seeking new customers. See [88912] at 19-24, ¶¶ 40-48. A decision to switch to a new DMS comes with both “hard” and “soft” costs. Hard costs include things like hardware, setup, training, and data conversion from the old system. Soft costs include time devoted to system configuration, training, “go-live” data conversion, impact to customer satisfaction, employee turnover, and lost sales and productivity during the ramp-up time after the dealership “goes live” on the new DMS.
Defendants do not dispute Stejskal's qualifications or his opinions concerning the DMS market, third party software applications, or data integration services.
Defendants argue that Stejskal's opinions on DMS switching should be excluded because they are based on anecdotes and speculation, rather than any expert methodology. He has not analyzed data concerning switching costs and admits he performed no quantitative analysis concerning DMS switching frequency. At bottom, Defendants argue, Stejskal's opinion is just an observation that switching entails some costs, which would be obvious to a layperson.
Under Rule 702, an expert may be deemed qualified “by ... experience, ” as well as “knowledge, skill, ... training or education.” Fed.R.Evid. 702; see also Africano, 2021 WL 4264237, at *3. “This is a “liberal standard; the expert need only have some ‘specialized knowledge that would assist the trier of fact.'” Id. (quoting Wilson v. City of Chicago, 6 F.3d 1233, 1238 (7th Cir. 1993)). Expert witness testimony need not “be ‘scientific' (natural scientific or social scientific) in character.” Tuf Racing Products, 223 F.3d at 591 (quoting Kumho Tire, 526 U.S. at 150). “The principle of Daubert is merely that if an expert witness is to offer an opinion based on science, it must be real science, not junk science.” Id. In short, “[a]nyone with relevant expertise enabling him to offer responsible opinion testimony helpful to judge or jury may qualify as an expert witness.” Id. “An expert's testimony is not unreliable simply because it is founded on his experience rather than on data[.]” Metavante Corp. v. Emigrant Sav. Bank, 619 F.3d 748, 761 (7th Cir. 2010); see also Viamedia, Inc. v. Comcast Corp., 951 F.3d 429, 484 (7th Cir. 2020) (district court erred by excluding expert opinion that drew significantly on expertise to add context and supporting information); Toney v. Quality Resources, Inc., 323 F.R.D. 567, 578-79 (N.D. Ill. 2018).
Applying these principles, the Court concludes that Defendants' motion to exclude Stejskal's opinions on DMS switching should be granted in part and denied in part. The Court agrees with Plaintiffs that Stejskal's testimony concerning various ways that DMS switching may be difficult and disruptive for dealers will be helpful to lay jurors. Although they might already understand at a very high level that switching comes with some costs, they would lack familiarity with the inner workers of automotive dealerships and specific disruptions that may come with switching DMS providers. Stejskal's experience has given him familiarity with the types of hard and soft “costs” that may be associated with switching DMSs. On the dealer side, his role at AutoNation required him to oversee the DMS conversion process when AutoNation acquired new stores. He also spent more than a decade working for DMS providers, including CDK (then ADP Dealer Services) and Dealertrack, where he was responsible for running the nation's third largest DMS business. This experience “enabl[es] him to offer responsible opinion testimony” about the general types of hard and soft costs that dealers may face in the DMS conversion process. Tuf Racing Products, 223 F.3d at 591; see also, e.g., National Jockey Club v. Ganassi, 740 F.Supp.2d 950, 964 (N.D. Ill. 2010) (banking industry expert was qualified to testify as to banking practices in lawsuit involving breach of a personal guarantee obligation under a lease, where he had extensive experience in relevant aspects of the banking industry); James Hamlin & Co. PC v. Czarnecki & Schlenker, LLC, 2021 WL 4845780, at *10-11 (E.D. Wis. Oct. 18, 2021) (person with 25 years of experience in acquiring 20 accounting firms could testify as expert in action alleging breach of asset purchase agreement by describing for jury what he knew about successfully acquiring accounting firms, such as retaining clients and billing; although views of witness were not “methodology, ” he did not propose to testify to scientific principles that could be tested empirically and his ideas could be grounded in accepted experience in his field).
However, Stejskal will not be allowed to opine as to how frequently switching costs dissuade dealers from changing DMS vendors. Plaintiffs acknowledge that Stejskal did not analyze data or perform a quantitative analysis on switching frequency. Therefore, any opinion he might provide on switching frequency would be “connected to existing data only by the ipse dixit of the expert, ” which Daubert and the Rules of Evidence do not allow. General Electric, 522 U.S. at 146. Plaintiffs suggest that this is not the intent of Stejskal's testimony and point out that “Stejskal never said switching DMS providers is impossible or ‘not a viable option.'” [996] at 11. But at least one portion of his report could be read as supplying “a bottom line” on how switching costs affect switching frequency. Wendler & Ezra, 521 F.3d at 791-92. In particular, Stejskal writes that “[t]he prospect of going through the ‘trough of despair'” (his colorful term for the DMS conversion process) “is frequently enough to dissuade a dealership from making a switch or even going through the competitive bid process.” [889-12] at 23, ¶ 47 (emphasis added). In the accompanying footnote, Stejskal includes a parenthetical citing to deposition testimony of Paul Whitworth that “CDK only lost 1.4% of its dealership customers in 2018, despite ‘massive cuts to their customer service' and a marketing push by Dealertrack.” Id., n.67. Without performing his own quantitative analysis or analyzing the data provided by others in the litigation, Stejskal has no basis for opining whether dealerships “frequently” or “infrequently” switch DMSs or for associating any particular switching percentages with the generalized “costs” that he outlines. Therefore, at trial, he will be prohibited from drawing such connections.
The second topic at issue in motion [883] is third-party data integrator security. Defendants challenge two of Stejskal's opinions on this topic. First, Stejskal states that in his experience, dealerships take the protection of personally identifiable information (“PII”) in written and electronic form “seriously” and “make it a point to train their staffs accordingly.” [889-12] at 39. Apart from his own experience in the industry he cites to the deposition of Mark Johnson, a dealership IT director who explained online training modules that employees must complete, and the deposition of Whitworth, who opined that “I think our industry, as with so many industries, by and large does a good job in providing suitable protections for data.” Id., n.106.
The Court agrees with Defendants that this opinion fails to satisfy Daubert and, therefore, it will be excluded. Stejskal is not an expert in cybersecurity and he has worked for only one dealer group, Auto Nation. His report provides no basis for concluding that he has knowledge concerning how seriously automotive dealers, as an industry, are about cybersecurity or what training, if any, they provide for their staff. Even if Stejskal were qualified to offer his opinion, his report provides “nothing but a bottom line, ” which “supplies nothing of value to the judicial process.” Wendler & Ezra, 521 F.3d at 791-92. Stejskal's citation to Whitworth's equally conclusory testimony and Johnson's testimony about the training provided at one dealership does not save his opinion; there is “simply too great an analytical gap between the data and the opinion proffered.” General Electric, 522 U.S. at 146.
Stejskal's second opinion concerning third-party data integrator security is that “[t]here is nothing inherently insecure about dealers choosing to work with third-party data integrators.” [889-12] at 39, ¶ 81. In support, he states that “the OSA ‘Automotive Retail Data Security Guidelines' explicitly included the ability of dealers to provide third parties of their choosing with User IDs and passwords.” Id. He concludes that, “[i]n my 20-plus years in the automotive retail market, I am not aware of any data breaches caused by companies like Authenticom and DMI providing data integration services using this method.” Id.
The Court will exclude this opinion under Daubert, as well. Beyond being quite vague (what does “not inherently insecure” mean?; is the PII actually secure, or not?), it does not appear to be tied to any relevant data or experience. The OSA “Guidelines” on which Stejskal relies were developed nearly 15 years ago, during the two-year period when Stejskal was President of OSA. The Guidelines were a “set of operating principles, ” [889-12] at 36, ¶ 75, which, as Defendants point out, do not discuss the technical aspects of data integration and were never ratified by any formal body. Plaintiffs, claiming that this “misses the point, ” argue that “the Guidelines provide an appropriate basis of support for Mr. Stejskal's expert opinion because they reflect that a prominent coalition of over 50 industry stakeholders, including CDK ..., studied this issue and concluded that dealers should be entitled to use independent data integrators.” [996] at 17. This may be, but it tells the Court nothing about whether the use of third-party data integrators presents any cybersecurity risks, what those risks might be, whether dealers or their third-party data integrators typically take steps to ensure the protection of data, or what steps they take. Finally, without any evidence that a significant proportion of dealers have adopted the Guidelines' suggested “method” of providing third parties with User IDs and passwords, it is simply irrelevant that Stejskal is “not aware of any data breaches caused by companies like Authenticom and DMI providing data integration services using this method.” [889-12] at 39, ¶ 81.
2. Stroz [859] - CDK's expert on cybersecurity and APIs
CDK offers Edward Stroz as an expert in cybersecurity. He has more than 35 years of investigative experience, first as a Special Agent for the Federal Bureau of Investigation and then in the private sector. He is the founder and Co-President of Stroz Friedberg, which is now a division of Aon. Stroz Friedberg is an international investigations, intelligence, security, and risk management firm, which assists clients with matters including digital forensic preservation and analysis, security consulting, data breach investigations, fraud investigations, theft of trade secrets, and monitorships. Stroz offers opinions on security concerns faced by DMS providers and security issues posed by hostile DMS access. He also attempts to quantity the number of times that CDK's security measures have been evaded-a topic that is relevant to the calculation of CDK and Reynolds' damages for their DCMA counterclaims, which are discussed further below.
Stroz opines, among other things, that the development of custom application programming interfaces (“APIs”) that third-party integrators could use to access the DMS “[would] not adequately address the security concerns from hostile access.” [998-2] at 40. Plaintiffs argue that this opinion should be excluded because Stroz's report offers no evidence to support his opinion; he performed no analysis to determine whether APIs could address Defendants' security concerns or what burden APIs would place on Defendants; and he did not assess the use of APIs by other DMS providers. All of these criticisms go to the weight, not the admissibility, of Stroz's opinions. In his reply report, Stroz lays out a number of reasons why, from Defendants' perspective, APIs would not be an adequate solution. He explains that creating APIs for data brokers like Authenticom would not provide any transparency or accountability as to the actual end-users of the data, because neither CDK nor Reynolds have visibility into which vendors receive which data from Authenticom. This would undermine the reporting tools and information that CDK and Reynolds provide to dealers that are designed to provide transparency into who actually has access to data maintained in their licensed DMS. Stroz also points out that developing custom APIs would require Defendants to expend substantial resources. These efforts would be superfluous because they would duplicate functionality that third-party software vendors are already offered through 3PA and RCI. If Plaintiffs disagree with Stroz's analysis, “[v]igorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means” of presenting their position to the jury. Daubert, 509 U.S. at 596.
3. Halpin [887] - Authenticom, Autoloop, and MVSC's expert concerning Cottrell's
In motion [887], CDK and Reynolds seek to exclude the expert report and testimony of Brian Halpin, who was retained by Plaintiffs to opine on the authenticity of a Microsoft Word document authored by Authenticom's CEO, Stephen Cottrell. The document on which Halpin opines consists of notes Cottrell made on April 4, 2016, allegedly documenting a conversation he had a day earlier with CDK executive Dan McCray (“McCray”). According to Cottrell and his notes, McCray referenced an agreement between the major DMS companies to effectively lock Authenticom and other third parties out of the market. McCray also allegedly made a profane threat to “destroy” Authenticom. In his expert report, Halpin concludes that the notes' “internal file metadata” reflect that the document “was created on April 4, 2016 at 10:41:00 AM” and “has not been changed since it was last saved on April 4, 2016 at 11:20:00 AM.” [889-15] at 13.
Defendants seek to exclude Halpin's opinion on the ground that it is not relevant to any fact at issue and therefore would not be helpful to the jury. Although Defendants dispute that the notes accurately reflect what Cottrell and McCray discussed, they do not dispute when the notes were created or whether they were subsequently modified. Defendants maintain that Halpin's report is not relevant and therefore would not be helpful to a jury because there “are no claims or defenses by any of the parties that rely upon the timing or modification of” the notes. [888] at 5. In addition, Defendants assert, their offer to stipulate that the notes were created on April 4, 2016 and have not since been edited “renders ... Halpin's proposed testimony undeniably improper.” Id. at 6. Defendants further argue that Halpin's testimony should be excluded because its potential prejudicial effect “far outweighs” its probative value. Id. at 8.
The Court denies motion [887] without prejudice to Defendants' ability to challenge Halpin's testimony through motions in limine in advance of trial. As noted above, step three of the Daubert analysis requires the Court to determine “whether the testimony will ‘assist the trier of fact to understand the evidence or to determine a fact in issue.'” Myers, 629 F.3d at 644 (quoting Ervin, 492 F.3d at 904). “Expert testimony must be relevant and factually linked to the case in order to meet Rule 702's ‘helpfulness' requirement.” Mason v. City of Chicago, 631 F.Supp.2d 1052, 1059 (N.D. Ill. 2009) (quoting United States v. Gallardo, 497 F.3d 727, 733 (7th Cir. 2007)). Halpin's proposed testimony satisfies that standard. Despite Defendants' assertion that none of Plaintiffs' claims rely on when Halpin's notes were created or whether they were later modified, those issues could be relevant to the jury's assessment of whether Cottrell's recollection of the April 3, 2016 conversation is accurate. A jury may find it probative that the notes were taken less than a day after the conversation when, arguably, it was fresh in Halpin's mind and litigation was not yet contemplated.
Defendants argue, based on Brand Name Prescription Drugs Antitrust Litigation, 186 F.3d at 786, that “expert testimony should be excluded where it ‘mainly concern[s] a matter not in issue' and can ‘be[] removed as an issue at trial by a stipulation of the parties'-precisely as Defendants tried to do here.” [888] at 6. But the Court does not read Brand Name Prescription Drugs to encourage using Daubert in the manner Defendants propose. During the eight-week trial in that case, “[t]he defendants spent days cross-examining” one of the plaintiffs' principal expert witnesses and “ultimately persuaded the district judge to exclude most of his testimony under ... Daubert” and grant judgment as a matter of law in their favor. Brand Name Prescription Drugs, 186 F.3d at 786. But “what was objectionable about [the expert's] evidence, ” Judge Posner observed, “actually had nothing to do with Daubert; it was that the evidence mainly concerned a matter not in issue.” Id. The subject of the expert's opinion “should have been removed as an issue at trial by a stipulation of the parties.” Id.
Here, the Court agrees with Defendants that it would make little sense to waste a jury's time on expert testimony concerning matters to which Defendants are willing to stipulate. Of course, Plaintiffs cannot be forced to stipulate, as a stipulation is “a contract between two parties to agree that a certain fact is true.” United States v. Barnes, 602 F.3d 790, 796 (7th Cir. 2010); see also United States v. Swiatek, 819 F.2d 721, 731 n.4 (7th Cir. 1987) (“It may be true that the introduction of this evidence was not essential, but . . . [a] party may insist on proving a fact even when there has been an offer to stipulate.”). But that does not mean that Defendants are without recourse if Plaintiffs refuse to accept a reasonable stipulation that will streamline and shorten the trial. “Concerns about trial efficiency may fit under various Rule 403 headings: confusion of the issues, undue delay, waste of time, and needless presentation of cumulative evidence.” Thompson v. City of Chicago, 722 F.3d 963, 971 (7th Cir. 2013); see also MCI, 708 F.2d at 1171 (“Litigants are not entitled to burden the court with an unending stream of cumulative evidence.”). This rule gives the trial judge “considerable latitude even with admittedly relevant evidence in rejecting that which is cumulative, and in requiring that which is to be brought to the jury's attention to be done so in a manner least likely to confuse that body, ” so long as a party is not “effectively prevented from presenting his or her case.” Thompson, 722 F.3d at 971.
Defendants make a number of Rule 403 arguments, which the Court finds unnecessary to resolve until closer to trial. At that point, the parties will have a better understanding of the full universe of witnesses, evidence, and issues that they intend to introduce. The parties will not have limitless time to present their cases and therefore will need to make strategic decisions about how to try the case most efficiently. To the extent they still cannot agree on whether Halpin will testify, the trial judge will be in position to employ “the balancing process contemplated by” Rule 403, which “is best undertaken at the trial itself.” Adams v. Ameritech Servs., Inc., 231 F.3d 414, 428 (7th Cir. 2000).
IV. Defendants' Counterclaims
For ease of reference, in discussing CDK and Reynolds' counterclaims, the Court will continue to refer to CDK and Reynolds as “Defendants” and the Counter-Defendants as “Plaintiffs.”
A. Stroz [859] - CDK's expert on damages for DMCA Violations
Congress enacted the DMCA in 1998 “to strengthen copyright protection in the digital age.” Universal City Studios, Inc. v. Corley, 273 F.3d 429, 435 (2d Cir. 2001); see also Navistar, Inc. v. New Baltimore Garage, Inc., 2012 WL 4338816, at *3 (N.D. Ill. Sept. 20, 2012). Section 1201 of the DMCA addresses liability for circumventing systems that protect copyrights. See 17 U.S.C. § 1201. It provides that “[n]o person shall circumvent a technological measure that effectively controls access to a work protected under this title.” Id. § 1201(a)(1)(A). The DMCA defines “circumvent a technological measure” to mean “descramble a scrambled work, to decrypt an encrypted work, or otherwise to avoid, bypass, remove, deactivate, or impair a technological measure, without the authority of the copyright owner.” Id. at § 1201(a)(3)(A). The term to “circumvent protection afforded by a technological measure” is defined to mean “avoiding, bypassing, removing, deactivating, or otherwise impairing a technological measure.” Id. at § 1201(b)(2)(A).
In their DMCA counterclaims against Authenticom, CDK and Reynolds allege that dealers are contractually prohibited from granting data integrators like Authenticom access to their DMS. See generally [225] (Reynolds' counterclaims); [299] (CDK's counterclaims). Neither CDK nor Reynolds authorizes such access. Instead, they have instituted various technological measures designed to block data integrators, such as CAPTCHA prompts and a “YES/NO” prompt, which are described in more detail below. They also have disabled user accounts that they believe to be associated with allegedly unauthorized third parties, such as Authenticom. Reynolds alleges that since at least 2009, Authenticom has repeatedly violated the DMCA's anti-circumvention provision, 17 U.S.C. § 1201(a)(1)(A), by bypassing these technological measures to access Reynolds' DMS.
As discussed above, CDK offers Edward Stroz as an expert in cybersecurity. In addition to assessing the extent to which “hostile” access to DMSs poses security risks-a topic addressed in the previous section of this order-Stroz was retained to quantify the number of times that certain Plaintiffs engaged in allegedly unlawful conduct related to accessing DMSs. More particularly, Stroz opines on how often, allegedly in violation of the DCMA, Authenticom bypassed three CDK security measures: (1) the “YES/NO” prompt; (2) CAPTCHA prompts; and (3) the disabling of user accounts associated with unauthorized third parties.
1. The Yes/No prompt
In June 2016, CDK deployed a prompt on Authenticom accounts that required users to “enter YES to confirm that you are an authorized dealer employee” or to “[e]nter NO to exit this program.” [995] at 11. Beginning on June 15, 2016, Authenticom deployed an automated script to respond “YES” and access the DMS. CDK withdrew the prompt on most Authenticom accounts by July 19, 2016. Id. Stroz attempted to calculate the number of times that Authenticom used automated means to answer the Yes/No prompt in the 34 days between June 15 and July 19, 2016. CDK had data logs for 98 accounts across 42 DMS servers for the full 34-day period at issue. Stroz identified these accounts as Authenticom accounts because they met the following criteria: (1) they were used by Authenticom; (2) they used automated means to access the CDK DMS; (3) they were active during the same 34-day window; and (4) CDK decided to deploy the YES/NO prompt against them. Stroz looked at the 98 accounts for the full 34-day period at issue and determined that a total of 22, 901 successful Authenticom responses (100% of the successful responses) were automated. There were an average of 7.4 successful responses per day per account.
CDK did not have data for the full set of 2, 750 accounts, over 1, 300 servers, which it determined to be Authenticom accounts meeting the same four criteria just described. Authenticom also did not provide Defendants with any more information about its YES/NO responses. See [995] at 16. Stroz therefore attempted to extrapolate his findings concerning the 98 accounts to the full set of 2, 750 accounts. He determined the average number of accounts that answered a login attempt on each day during the 34-day period (464.5). He multiplied this number by the 98-account subset's average number of successful responses per day per account (7.4), yielding a total of 116, 868 automated responses for the 34-day period.
Plaintiffs do not challenge the process Stroz used to determine number of YES/NO responses for the 98 accounts. However, they maintain that Stroz's extrapolation to the full set of accounts was unreliable because he has not demonstrated that the 98 accounts are representative of the full set of 2, 750 accounts. Plaintiffs argue that there is reason to question the representativeness of Stroz's sample because “90% of the 91 accounts in his sample were active on 13 of the 14 days in the sample, whereas 90% of the over 2, 000 accounts to which he extrapolates the sample were only active on (at most) a single day.” [861] at 13. CDK responds that “the data showing the broader population of users who received the prompt did not even purport to show the intensity of access by those users. And to the extent there was any variation in that intensity, the 42-server sample also reflected it.” [995] at 17 & n.8. The sample contained a broad distribution of users with a variety of user intensities: from those that averaged less than 1 access per day, to those that accessed on average about 51 times per day. See [998-2] at 54 & n.232.
The Court finds it a close call as to whether the uncertainty of Stroz's extrapolation warrants exclusion under Daubert. Although “damages cannot be based on pure speculation or guesswork, ” they “also need not be proven with the certainty of calculus.” BE &K Const. Co. v. Will & Grundy Counties Bldg. Trades Council, 156 F.3d 756, 770 (7th Cir. 1998); see also Phoenix Bond & Indem. Co. v. Bridge, 911 F.Supp.2d 661, 675 (N.D. Ill. 2012). And “[w]hen a defendant's wrong makes it difficult for the plaintiff to prove damages, all reasonable doubts about the amount of damages are resolved in the plaintiff's favor.” Id. CDK's alleged damages stem from Plaintiffs' surreptitious use of technology to avoid the “YES/NO” prompts. The Court therefore begins with the proposition that Plaintiffs should “not benefit from” any “uncertainty [they] created.” Id.; see also Mid-America Tablewares, Inc. v. Mogi Trading Co., Ltd., 100 F.3d 1353, 1365 (7th Cir. 1996) (“Speculation has its place in estimating damages, and doubts should be resolved against the wrongdoer.”).
That said, the Seventh Circuit has cautioned that a “non-random sample might undermine the reliability of the statistics.” Chavez v. Ill. State Police, 251 F.3d 612, 643 (7th Cir. 2001) (civil rights case in which plaintiffs' expert used statistical sampling to determine racial breakdown of motorists stopped by police). Nonetheless, the Court is not convinced that Stroz's extrapolation must be excluded based on the case law cited by Plaintiffs, in which experts could have examined, but apparently ignored, fuller, more detailed data in favor of “convenience” sampling or cherrypicking to achieve more favorable results. See, e.g., Espenscheid v. DirectSat USA, LLC, 705 F.3d 770, 774-75 (7th Cir. 2013) (in Fair Labor Standards Act (“FLSA”) case, rejecting plaintiffs class counsel's proposal to extrapolate from experience of small, unrepresentative sample of 42 proffered class members to that of 2341 to avoid “2341 separate hearings” on damages for hours plaintiffs worked but did not report on time sheets, where “[t]he unreported time for each employee could be reconstructed from memory, inferred from the particulars of the jobs the technicians did, or estimated in other ways”); American Honda Motor Co., Inc. v. Allen, 600 F.3d 813, 818-19 (7th Cir. 2010) (expert who tested one vehicle and extrapolated findings to whole fleet of vehicles produced over seven-year period did not offer reliable testimony; he “was not being as thorough as he might otherwise be due to Plaintiffs' reluctance to pay for more testing”); Farmer v. DirectSat USA, LLC, 2013 WL 1195651, at *4-7 (N.D. Ill. Mar. 22, 2013) (in FLSA case, excluding expert opinion concerning number of unpaid hours worked by class members where expert relied on 3 months of GPS data from 30 class members' vans, 7 depositions, and timesheets from 4 class members to extrapolate to 500 employees over 16 months-rather than GPS data for 500 employees over the whole 16-month period-where record showed there were large seasonal fluctuations in hours worked, GPS data was questionable, and sample may have been cherry-picked by plaintiffs' counsel); Allgood v. Gen. Motors Corp., 2006 WL 2669337, at *10-11 (S.D. Ind. Sept.18, 2006) (excluding proposed expert who estimated plaintiffs' risk of developing cancer, where his risk assessment was “performed by using only a limited number of the available samples, and those that would tend to magnify greatly the risk calculation” and he “failed to offer any scientific justification for his sample selection choices”). Rather, Stroz used all of the data available to him, which was confined to 98 accounts, to extrapolate to the full set of 2, 750 accounts. He proceeded in this manner because this was the data that CDK had, and Plaintiffs never provided any additional data. Plaintiffs do not explain what additional analysis Stroz could or should have done with the limited data available to him to ensure that the accounts he examined in detail were representative of the larger universe of accounts. See [861-7] at 54 (Stroz explaining in reply report that Authenticom's rebuttal witness, Dr. O'Brien, “points to no available data, and I am aware of none, that would allow any further assessment of representativeness”).
“Rule 702 [ ] does not condition admissibility on ... a complete and flaw-free set of data.” Lees v. Carthage Coll., 714 F.3d 516, 524-25 (7th Cir. 2013) (internal quotation marks omitted). “The critical inquiry” in assessing an expert's extrapolation from existing data “is whether there is a connection between the data employed and the opinion offered.” General Electric, 522 U.S. at 146. This is not an instance where there is “no rational connection between the data used and the conclusion arrived at.” Kleen Products, 2017 WL 2362567, at *9. Stroz explains the basis for his belief that the subset of accounts he examined was representative of the larger group of accounts, including that all of the accounts met the same four criteria for being identified as Authenticom accounts (used by Authenticom; used automated means to access the CDK DMS; were active during the 34-day window; and CDK decided to deploy the YES/NO prompt against them). Stroz's method of extrapolation also apparently took into consideration and excluded accounts for which no access attempts were made on a particular day, since his calculation was based on the average number of accounts that answered a login attempt on each day during the 34-day period (464.5 of the 2, 750 total accounts).
What is not clear, however, is whether it was reasonable to assume 7.4 logins per day for the active accounts in the larger group. Neither side assesses this factual issue in any detail. Plaintiffs claim that “[o]ther evidence indicates CDK targeted accounts with more frequent DMS access for more frequent interruption, ” [1048] at 6. But the larger group of accounts that Stroz considered, like the smaller subset, included only accounts that CDK decided to deploy the YES/NO prompt against. In other words, all of the accounts that Stroz considered were ones that Defendants targeted for interruption with the YES/NO prompts. Plaintiffs' rebuttal expert, Daniel O'Brien, also acknowledged that it is possible that the full set of users who received the prompt accessed the DMS with the same frequency as those in the smaller set. Given the limited information available to Stroz, and the rational connection between the available data and the conclusions he reached, the Court finds on balance that it should be left to a jury to assess whether it was reasonable for Stroz to assume 7.4 logins per day for the active accounts in the larger group. Plaintiffs are “free to use cross-examination to attack the assumption” of 7.4 logins per day and ask Stroz “how altering the assumption would affect his analysis.” Stollings, 725 F.3d at 766. “A jury should be capable of understanding how the value of the estimate affected [Stroz's] conclusions.” Id. at 767; see also Fluidmaster Water Connector Components Products Liability Litigation, 2017 WL 1196990, at *30 (concluding that most of defendant's arguments concerning expert survey-including “the survey's small sample size, unrepresentative sample composition, [and] whether the results of conjoint studies can be adequately extrapolated to the retail setting.-go to weight, not admissibility”).
In addition to challenging the reliability of Stroz's extrapolation, Plaintiffs argue that Stroz was not qualified to perform this analysis, which “required statistical expertise-including experience in designing statistical models, selecting samples, and measuring the standard error of estimates-that Mr. Stroz does not have.” [861] at 10 (citing [861-5], Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE 241 (3d ed. 2011)). CDK responds that Stroz does not need to be a statistician to make the calculations he did. They explain that “[h]e was retained to assess the cybersecurity implications of hostile integration, and the more specific practices in question here (CAPTCHA and YES/NO prompt evasion; and use of the Profile Manager tool), which are at the heart of his general expertise in cybersecurity and forensic investigations, which in turn often require analysis of voluminous data.” [995] at 15 n.6.
The reference manual on which Plaintiffs rely provides: “In some instances, the expert is faced with more information than is possible to process. ... If the expert elects to study a sample of the data, the expert needs to have carefully considered how the information will be used to ensure that the data sample is large enough and contains sufficient information. Usually, this requires that the expert has constructed a model of damages and a related sampling plan that includes an estimate of the sampling error. Unless the expert is a trained statistician, the expert should seek outside help in designing the sampling plan.” [861-5] at 25-26. This guidance does not appear to fit what Stroz did, as he was not “sampling” data that was too voluminous to process. He was looking at all the available data, and assessing whether any conclusions could be drawn about the accounts for which fuller information was not available. Plaintiffs do not question whether Stroz was qualified to analyze the 98 accounts and opine on how many times Authenticom evaded Defendants' YES/NO prompts. Plaintiffs focus only on Stroz's qualifications to extrapolate to the 2, 750 accounts. They do not explain how an expert in statistics would be better suited to perform that analysis, given the limited data available and given the need for an expert to understand the relevant cybersecurity technology.
Plaintiffs argue further that Stroz's method of calculating Yes/No responses within the sample from the 42 servers does not fit the facts in the record. [861] at 17. But the argument about whether and when certain accounts should be counted is the kind of input to a calculation that goes to an opinion's weight, not its admissibility. Manpower, 732 F.3d at 807. Plaintiffs also challenge Stroz's reliance on CDK's tagging accounts as being associated with Authenticom, especially where CDK tagged an account as being associated with multiple vendors or with “unknown vendors.” [861] at 18-19. But Stroz also relied on Authenticom's own records and admissions to identify Authenticom accounts, see [861-1] at 87-88, which supports a conclusion that Stroz's input was reliable. If Plaintiffs believe that CDK and Stroz have failed to provide sufficient evidence that certain accounts belonged to Authenticom, they can make those arguments to the jury and, if persuaded, the jury can discount or exclude altogether those accounts from its calculations.
2. Profile Manager
In late 2016, CDK began disabling Authenticom accounts. In response, Authenticom developed a software called Profile Manager, which accessed the DMS's Update User Profile (“UUP”) function and automatically reactivated disabled accounts. Authenticom began using Profile Manager at dealerships around November 4, 2016. CDK deployed a security patch to stop Profile Manager on April 25, 2017. In discovery, Authenticom did not produce records of the number of times Profile Manager was launched or successfully re-enabled accounts between November 4, 2016 and April 25, 2017. Authenticom's 30(b)(6) witness also could not answer how often the program ran.
In his report, Stroz attempted to calculate damages based on the number of times that Authenticom launched Profile Manager during the relevant period. Stroz relied on emails from Authenticom indicating that it provided Profile Manager to 4 dealers by November 4, 2016, to at least 100 by January 26, 2017, to more than 200 by February 8, 2017, and to more than 350 by April 17, 2017. Stroz assumed that between each of those dates, the number of dealers with Profile Manager remained constant (e.g., he assumed that between November 4, 2016 and January 26, 2017, only 4 dealers had Profile Manager). He also reduced the number of dealers for a period in February based on information that Profile Manager service was interrupted, and he subtracted a number of dealers to whom the security patch applied before April 14, 2016.
To determine the number of times Profile Manager launched, Stroz looked for automated access to the UUP function and developed three factors to identify it: the account had a high ratio of access per hour; at least 70% of the accesses took less than 10 seconds; and at least 70% occurred at 20 minutes after the hour (when Stroz observed that the automated accounts typically accessed the UUP function). He identified 52 accounts that met these three criteria. Stroz then reviewed 16 full days (and 2 partial days) of data regarding accounts accessing the UUP function from March 19-April 5, 2017; this data was provided by Defendants and was the only data available. The 52 accounts that Stroz identified logged in 12, 449 times over the 16 full days, an average of 15 times per day. This was consistent with an Authenticom document indicating that Profile Manager was designed to run every hour from 7 a.m. to 10 p.m.-i.e., 15 times per day. Stroz concluded that Profile Manager ran an average of 15 times per day. To calculate the number of times Profile Manager launched between November 4, 2016 and April 24, 2017, Stroz multiplied 15 times the number of dealers with Profile Manager for each day of the relevant period for a total of 64, 770 launches.
Stroz also tried to calculate the number of times that Profile Manager re-enabled a disabled account. As of April 2016, CDK was disabling targeted accounts at least once per day, though sometimes more often. Stroz assumed that Profile Manager successfully re-enabled only one user account per day per dealer group, for a total of 4, 318 re-enabled accounts over the relevant period.
In addition, Stroz ran his analysis for a particular dealer group, the Warrensburg CounterDefendants, for a period of March 20, 2017 (the earliest date for which he had data) to August 21, 2019. He used a daily launch number of 11, rather than 15, because that was the dealer group's average during the 2-week sample period. Stroz calculated that the Warrensburg CounterDefendants ran Profile Manager 9, 735 times, which re-enabled a user account at least 36 times.
Plaintiffs challenge the reliability of Stroz's extrapolation from a 16-day sample in March and April 2017 to the entire four-and-a-half-month period (and longer for the Warrensburg Counter-Defendants). The core of the challenges to his extrapolation is the same as described in the prior section on the YES/NO prompts, and CDK's response is the same. The Court's resolution of this challenge is also essentially the same; in this section, the Court focuses on the arguments that are specific to Stroz's Profile Manager analysis.
As with his opinion concerning YES/NO prompts, Stroz was working with limited data. Authenticom allegedly developed the Profile Manager software to evade Defendants' disabling of accounts due to alleged unauthorized access, and it did not produce any records or a 30(b)(6) witness identifying the number of times it launched profile Manager or reactivated disabled accounts. As Authenticom's alleged actions (or inactions) have made it difficult for Defendants to prove damages, reasonable doubts about the amount of damages are properly resolved in Defendants' favor. See BE&K, 156 F.3d at 770; Phoenix Bond, 911 F.Supp.2d at 675.
According to Plaintiffs, Stroz's calculations are unreliable because he has not demonstrated that the 16-day period he analyzed (March 19 to April 5, 2017) is representative of the full period (November 4, 2016 to April 25, 2017). In particular, Plaintiffs argue that it is not a reasonable assumption that Profile Manager was launched fifteen times per day per account for the whole period because the earliest there is even anecdotal evidence that the program was running automatically was February 13, 2017-over half-way through the whole period at issue. However, this argument relies on a single, unclear email that Plaintiffs do not adequately explain or support with an affidavit or other evidence of when Profile Manager became capable of being run automatically. See [861-11] at 2-3. Plaintiffs also do not contest that when Profile Manager was running automatically, it was designed to do so 15 times per day. Nor do Plaintiffs dispute that Profile Manager was running 15 times per day during the 2-week period for which Defendants were able to obtain detailed data. While he did not have “a complete and flaw-free set of data” from which to work, Lees, 714 F.3d at 524-25, there was a rational connection between the data on which Stroz relied and his calculations for the full time period. See Kleen Products, 2017 WL 2362567, at *9. And while Plaintiffs criticize Stroz for allegedly failing to use reliable statistical methods like testing the representativeness of his “sample” or calculating a standard error, it does not explain how Stroz could have done so given the limitations in the data Authenticom provided.
Plaintiffs also challenge the reliability of Stroz's calculations on the grounds that he used a different number of average daily launches for the Warrensburg Counter-Defendants (11) than he used for other dealers who deployed Profile Manager (15). However, it was rational to use a daily launch number of 11, rather than 15, for this dealer group because that was the group's average during the 2-week sample period. Any relevant differences between the Warrensburg Counter-Defendants and other dealers can be explored through cross-examination.
3. CAPTCHA
In October 2017, CDK began using CAPTCHA prompts to “combat accounts associated with unauthorized third-parties” attempting to access the CDK DMS. [995] at 6. By requiring the user to enter obscured numbers and letters, a CAPTCHA prompt is designed to prevent automated access (i.e., access by a computer rather than a human). Authenticom developed software to overcome the CAPTCHA prompts. Stroz attempted to determine the number of times that Authenticom circumvented CDK's CAPTCHA prompts. First, he developed four criteria to identify accounts that submitted automated responses: (1) the account submitted a blank answer to at least 80% of the prompts; (2) the account submitted a blank answer at least 95% of the times it faced particular fonts (with a minimum of 20 attempts); (3) the account successfully answered a prompt twice in under one second, or twice in under two seconds at least ten times; and (4) the account attempted to log in more than 300 times in a 24-hour period.
Stroz then attempted to identify accounts that were related to Authenticom. He began with a set of accounts that CDK tagged as belonging to Authenticom. He also tried to identify accounts based on the queries they sent to the DMS. Authenticom ran queries to the DMS in a particular database query language (“ENG”), and Stroz used documents that Authenticom produced to find ENG queries that it commonly ran on CDK's DMS. Stroz reviewed all ENG queries on particular dealer servers from March 13, 2016 to July 24, 2016 to determine which accounts ran ENG queries that Authenticom used. Stroz excluded from consideration any queries that appeared in Authenticom's documents but contained fewer than six output fields, on the grounds that they were too generic to identify an Authenticom-related account. That eliminated from consideration 8 of 60 queries. Stroz concluded that the other 52 queries were sufficiently distinctive to identify Authenticom accounts.
CDK's records of access to the DMS show 1, 614 accounts that ran an ENG query associated with Authenticom, 98% of which CDK had also tagged as belonging to Authenticom. Stroz concluded that accounts running Authenticom-related queries and tagged by CDK as belonging to Authenticom were Authenticom accounts, and he found 126 that had responded to CAPTCHA queries between October 2017 and January 5, 2018. Stroz found that those accounts gave 7, 628 successful automated response to CAPTCHA prompts, including 1, 256 times on behalf of the Continental Counter-Defendants.
Plaintiffs raise three objections to these opinions. First, Plaintiffs argue that Stroz lacks the expertise to give these opinions because he is not an expert in reading software code. But Plaintiffs do not explain why he needs to be one or why issues like automated responses to database security measures or review of database query logs are not within the scope of his cybersecurity and digital forensic analysis expertise.
Second, Plaintiffs challenge Stroz's methodology on the basis that he purportedly fails to explain how he concluded that the 52 ENG queries with six output fields or more were so distinctive that he could identify them as queries from Authenticom. However, his report explains how he came to this conclusion. He began with an Authenticom document that contained ENG queries that Authenticom commonly used. He then looked at CKD's logs of ENG queries and found the accounts that ran the ENG queries that Authenticom commonly used. He eliminated from consideration simpler ENG queries, a set of 8 with fewer than six output fields, because they were too generic to be used to identify Authenticom. That left 52 ENG queries. If that was an error, it was an error in Authenticom's favor, because it meant that Stroz's analysis pulled in fewer queries and fewer dealer accounts. U.S. ex rel. Tyson v. Amerigroup Illinois, Inc., 488 F.Supp.2d 719, 733-34 (N.D. Ill. 2007) (expert's method was not unreliable when his choices in calculating damages provided a more conservative estimate, which benefited the objecting party); cf. Kleen Products, 2017 WL 2362567, at *10 (declining to exclude expert opinion because expert's choices were reasonable, “even if not the most conservative”). From those 52 queries, Stroz then identified the number of times those queries were run and the accounts that ran those queries, which Stroz concluded were Authenticom accounts.
He cross-referenced those accounts with a list of accounts that CDK had tagged as Authenticom-related, another move that Plaintiffs contest. But the overlap in the separately-compiled lists-98%-is a suggestion of reliability. Furthermore, CDK used the names of some accounts to tag them as Authenticom (such as “authcom, ” “dvault, ” and “dvault1”). [998-4] at 2. None of this is say that Stroz's conclusions are right or wrong, only that his report lays out a method and is supported by the evidence that CDK cites, and therefore should not be excluded on the basis Plaintiffs urge.
4. Grouping of DMCA violations by dealership group
Finally, Plaintiffs argue in several places in their motion that Stroz improperly groups together alleged DMCA violations by dealership, rather than breaking them out by individual dealers. CDK takes the position that the dealers in a group have joint and several liability and therefore it was unnecessary to break out the violations by individual dealer. The Court reserves ruling on this point until summary judgment.
B. Miracle [885] - Authenticom's expert on various issues
In motion [885], Defendants seek to exclude the testimony of Nancy Miracle, who was retained by Authenticom to opine on a number of subjects relating to Authenticom's efforts to obtain data from Defendants' DMSs and Defendants' attempts to use technological means (including CAPTCHA and challenge questions) to hinder those efforts. Miracle is the president and senior partner of Digital Miracles, LLC, a company she founded that specializes in the evaluation and development of software products with particular emphasis on design-arounds, unusual problem solutions, software reverse engineering, code analysis, prior art/prior use evaluation, and litigation support. She has more than 50 years of expertise in the areas of systems design and analysis and has worked as a programmer and technical manager for a variety of small companies and start-ups. She has served as a litigation consultant for over 20 years.
Miracle's report contains a summary of her opinions. First, she opines that CAPTCHA is not a technological measure to “control access” as that term would be understood in the software design and cybersecurity field because, by design, all information needed to satisfy a CAPTCHA is provided on the face of the prompt itself; that providing the information called for by a prompt does not “circumvent” the prompt; and that the outdated and simplistic CAPTCHA technology used by Defendants was not “effective” even at preventing automated answers to the prompt. Second, she opines that Defendants' challenge questions are not considered technological access control measures in the software design and cybersecurity field for essentially the same reasons CAPTCHA is not such a measure. Third, Miracle opines that Defendants' measures to detect and disable passwords that dealers provided to Authenticom were not measures to “control access” as that term is understood in the field, because they do not require a user to provide information or perform some act that is in the user's possession and that indicates that the uses has authorization to obtain access. Fourth, Miracle opines that Authenticom's use of “profile manager” software to re-enable usernames that had been disabled did not involve “bypassing” or “decrypting” any technological measure that controlled access to CDK's DMS. Fifth, Miracle summarizes that Defendants' CAPTCHA and challenge questions did not “control access” to Defendants' executable program files; those prompts just appeared in the terminal displayed by Defendants' DMSs, which were not part of the executable file or otherwise committed to any fixed form.
Miracle's sixth opinion goes to whether Defendants' CAPTCHA mechanisms protect access to a work of “creative expression” in the ERA-IGNITE program. Miracle states that it is her understanding that the elements in the DMS display or user interface that appear after the CATCHA is solved are common, standard, and functional design elements (specifically, a progress indicator and popup window) that do not, in her opinion, constitute creative expression that is subject to copyright protection. Finally, Miracle provides a rebuttal to part of Stroz's report. She opines that Stroz's estimate of Authenticom's DMCA damages is not reliable because he incorrectly assumes that all queries designed by Authenticom were run by Authenticom, while in fact Authenticom shared its queries with third parties.
Defendants move to exclude Miracle on a number of bases, with their lead, threshold argument being that Miracle's opinions invade the province of the Court by purporting to define statutory terms-including “effectively controls access” and “circumvent a technological measure”-that are already defined in the DMCA. Defendants maintain that Miracle's definitions contradict the plain text of the DMCA and case law interpreting it, in addition to well-accepted technical standards.
In response, Authenticom claims that Defendants misunderstand Miracle's report; she is not planning to tell the jury how to construe the DMCA, but rather “will offer testimony on factual issues that will aid the jury in determining whether the DMCA has been violated.” [1007] at 11. More particularly, Authenticom asserts that Miracle may offer testimony on “four factual issues that the trier of fact may need to resolve if this Court does not grant Authenticom's motion for summary judgment.” Id. at 10. These “factual issues” include: (1) whether Defendants' technological measures are capable of distinguishing between would-be users who have authority from Defendants and those who do not, which may be relevant to the jury's determination of whether those measures “effectively control access” within the meaning of the DMCA; (2) the technical means by which Authenticom responded to the Defendants' measures, which may be relevant to the jury's determination of whether Authenticom “circumvented” those measures within the meaning of the DMCA; (3) the process by which Authenticom's software “polled” (that is, accessed) Defendants' DMS, including when and whether that software encountered Defendants' technological measures, which may be relevant to the jury's determination of whether any copyrighted work was protected by the technological measures; and (4) Miracle's response to Stroz's efforts to count alleged DMCA violations.
The Court agrees with Defendants that Authenticom appears, at least in part, to be “backpedal[ing] from [Miracle's] opinions and recast[ing] them as ‘testimony regarding technical facts.” [1032] at 5. In her report, Miracle offers her own definitions of key terms that are already defined in the DMCA, asserts that her definitions are consistent with the DMCA's definitions, and then applies her definitions to conclude that Authenticom's actions did not violate the DMCA. In particular, Miracle asserts that, “[i]n the field of software design and cybersecurity, a technological measure is considered to ‘control access' if it requires a user to provide information (or perform some act) that is in the user's possession and indicates that the user has the computer or network administrator's authorization to obtain access. The simplest and most common example is an administrator-provided or administrator-authorized password.” [889-10] at 10-11. She claims that the “field's” understanding of “control access” is “consistent with” the DMCA's definition of ‘effectively controls access to a work, ” id. at 11, and then goes on to opine why Defendants' CAPTCHA prompts, challenge questions, disabling of passwords, and activity monitoring are not measures to control access. Similarly, Miracle states that “[i]n the field of software design and cybersecurity, “[t]he terms used in the DMCA to define “circumvention” of a technological access measure also have well-accepted meanings in the software design and cybersecurity field.” Id. at 11-12. That term is already defined by the DMCA. See 17 U.S.C. § 1201(a)(3)(B) (“a technological measure ‘effectively controls access to a work' if the measure, in the ordinary course of its operation, requires the application of information, or a process or a treatment, with the authority of the copyright owner, to gain access to the work”). Similarly, Miracle expands on the DMCA's definition of “circumvent a technological measure, ” 17 U.S.C. 1201(a)(3)(A), by offering her definitions of what “bypass” and “avoid” mean in the field of software design and cybersecurity. [889-10] at 12. Using these definitions, she states that “it is my opinion that Authenticom did not ‘circumvent' the technological barriers that Defendants rely on for their DMCA claim.” Id. at 12-13.
“Although experts may provide opinions as to the ultimate factual issues in a case, they may not testify “as to legal conclusions that will determine the outcome of the case” under Rule 702.” Scottsdale Ins. Co. v. City of Waukegan, 689 F.Supp.2d 1018, 1022 (N.D. Ill. 2010) (citing Pansier, 576 F.3d at 738, Good Shepherd Manor Found. v. City of Momence, 323 F.3d 557, 564 (7th Cir. 2003)). Miracle's opinions, outlined above, “amount[] to an instruction on the law and application of the law to the facts, ” which “invades the province of both the jury and this Court.” Anderson v. City of Chicago, 454 F.Supp.3d 808, 819 (N.D. Ill. 2020); see also Securities & Exchange Comm'n v. Ferrone, 163 F.Supp.3d 549, 563 (N.D. Ill. 2016) “(an expert usually cannot testify about how a law should be interpreted or what it means”); George v. Kraft Foods Global, Inc., 800 F.Supp.2d 928, 932 (N.D. Ill. 2011) (“Put simply, expert opinions that seek to define the meaning of statutes are disallowed under the Federal Rules of Evidence.”); Davis v. Duran, 277 F.R.D. 362, 371 (N.D. Ill. 2011) (rejecting “plaintiff's position ... that since evidence concerning the use of deadly force is relevant, and because [plaintiff's proposed expert] is an expert in the use of force, he is qualified to give the opinion”; that “opinion would not be helpful, since under the Federal Rules of Evidence, it is the role of the judge, not the witness, to instruct the jury on the applicable principles of law”). Therefore, Defendants' motion to exclude Miracle is granted to the extent that her report “set[s] forth [what] she believes the relevant law to be” and “what the legal conclusion should be based on the law and facts that [s]he describes.” Old Republic Ins. Co. v. Ness, Motley, Loadholt, Richardson & Poole, 2006 WL 3782994, at *2 (N.D. Ill.Dec. 21, 2006).
The Court still must consider, however, whether Miracle should be allowed to testify about the four “factual issues” that Authenticom identifies in its response brief. The Court will defer its ruling on the first two issues- (1) whether Defendants' technological measures are capable of distinguishing between would-be users who have authority from Defendants and those who do not, and (2) the technical means by which Authenticom responded to the Defendants' measures-until after it has ruled on the relevant summary judgment motions. The Court's decision may moot the proposed testimony in any number of ways. For instance, if the Court construes the DMCA in a manner that conflicts with Miracle's industry definitions of the same terms used in the DMCA (i.e., “effectively control access” and “circumvented”), her factual distinctions between what Authenticom did and what is allegedly allowed under the DMCA may have no relevance.
The Court will also defer ruling on the third issue-the process by which Authenticom's software “polled” Defendants' DMS and its relevance to whether any copyrighted work was protected by the technological measures-and instead give it due consideration when ruling on the parties' summary judgment motions. Miracle's proposed testimony is relevant solely to the issue of copyrightability of the work that Defendants allegedly protected by technological measures. As Authenticom concedes, see [1007] at 21, in the Seventh Circuit, “‘copyrightability is always an issue of law, '” Janky v. Lake County Convention & Visitors Bureau, 576 F.3d 356, 363 (7th Cir. 2009) (quoting Gaiman v. McFarlane, 360 F.3d 644, 648 (7th Cir. 2009)), “which the court will decide.” Pivot Point Int'l, Inc. v. Charlene Products, Inc., 932 F.Supp. 220, 225 (N.D. Ill. 1996). In making this determination, “district courts may and do rely on expert testimony to distinguish between protected and unprotected material in a work for which copyright is claimed.” Team Play, Inc. v. Boyer, 391 F.Supp.2d 695, 699 (N.D. Ill. 2005). When resolving the summary judgment motions, Miracle's testimony will be “merely a factor to be weighed” by the Court, id. at 700, and the Court will disregard the testimony to the extent “it is unreliable” or otherwise fails to satisfy Daubert. Robb v. Burlington Northern & Santa Fe. Ry. Co., 100 F.Supp.2d 867, 872 (N.D. Ill. 2000) (explaining that evidence that is inadmissible under Daubert cannot be used to oppose summary judgment).
Finally, the Court denies Defendants' motion to preclude Miracle from offering testimony to respond to Stroz's efforts to count alleged DMCA violations. As explained in the prior section, Stroz attempted to determine the number of times that Authenticom circumvented CDK's CAPTCHA prompts. Starting with a set of accounts that CDK tagged as belonging to Authenticom, Stroz used documents that Authenticom produced to find ENG queries that Authenticom commonly ran on CDK's DMS. Based on her experience in the industry, Miracle opines that the mere fact that a query was executed on CDK's DMS that resembled a query found in Authenticom documents does not mean that Authenticom was the entity that executed the query. She explains that software programmers share code with third parties for a wide variety of reasons, including efficiency. Defendants emphasize that Miracle did not investigate how often Authenticom shared queries, either generally or in regard to the specific queries at issue here. They argue that, without a way to quantify Authenticom's query sharing, Miracle cannot say whether the potential impact on Stroz's results was inconsequential or significant.
While the Court agrees with Defendants that Miracle cannot go beyond her report by attempting to quantify the impact that code-sharing may have had on the accuracy of Stroz's calculations, her testimony may nonetheless be helpful to the jury in understanding the potential shortcomings of Stroz's calculations. Her testimony on common practices in the industry would be a piece of Authenticom's larger defense, which might also include testimony from and documents authenticated by fact witnesses that have personal knowledge of Authenticom sharing queries. Defendants will be free during cross-examination to bring out the limitations of Miracle's opinion.
C. Rubinfeld [867] - CDK and Reynolds' expert on various issues
1. Trade secrets
CDK brings counterclaims against Authenticom for violation of the Defend Trade Secrets Act, 18 U.S.C. § 1836 and the Wisconsin Uniform Trade Secrets Act. Rubinfeld opines on the reasonable royalty CDK should allegedly be awarded on the trade secrets claim. In his report, he notes that there are several methods for determining reasonable royalties, including established royalties, hypothetical negotiations, the Association of International Certified Professional Accountants (“AICPA”) analytical method, and discounted cash flow analysis. See [871-5] at 19-20. The method Rubinfeld chooses is to evaluate the hypothetical negotiation between a willing licensee and a willing licensor at the time of the infringement, which allows him “to derive a measure of the value of the intellectual property obtained from the intellectual property holder.” Id.; see also id. at 32. He opines that the price CDK charges third-party vendors under its 3PA program is a reasonable estimate of the “royalty” or price CDK would have charged Authenticom for unauthorized access to its DMS. According to Rubinfeld, this price most closely approximates the results of a hypothetical negotiation between a willing licensee and willing licensor, since it is the result of actual transactions between CDK and its 3PA customers.
Authenticom argues that this opinion should be excluded as “incoherent” because Rubinfeld does not clearly define the applicable “trade secret.” Authenticom also contends that Rubinfeld's choice of 3PA prices as a benchmark “makes little sense” because he does not do a “rigorous comparison of CDK's 3PA customers to Authenticom, or of 3PA to the DMS access Authenticom required to provide independent integration services” or “address CDK's historical policy of permitting independent integration, or Authenticom's allegations that CDK's 3PA prices were inflated by the anticompetitive conduct at issue.” [871] at 29.
None of these criticisms warrant exclusion of Rubinfeld's opinion, but rather are proper topics for cross-examination. A reasonable royalty analysis “necessarily involves an element of approximation and uncertainty.” Lucent Technologies, Inc. v. Gateway, Inc., 580 F.3d 1301, 1325 (Fed. Cir. 2009) (internal quotation marks and citation omitted). The “hypothetical negotiation” approach chosen by Rubinfeld is a commonly used methodology for determining a reasonable royalty. See Sunoco Partnership Marketing & Terminals L.P. v. U.S. Venture, Inc., 436 F.Supp.3d 1099, 1128-29 (N.D. Ill. 2020). And, as CDK points out, when actual pricing data is available, it is “usually the best measure of a ‘reasonable' royalty . . . because it removes the need to guess at the terms to which parties would hypothetically agree.” Monsanto Co. v. McFarling, 488 F.3d 973, 978-79 (Fed. Cir. 2007). In addition to applying a reasonable methodology, Rubinfeld took into account differences in quality between the 3PA integration provided by CDK, which includes real-time and bi-directional (writeback) access to the DMS, and the screen-scraping techniques used by Authenticom. Rubinfeld used the first decile of 3PA prices during the damages period- between $25 and $51 per month-to estimate a reasonable royalty. He explains that this price range aligns with basic, extract-only integration packages available through 3PA, which most closely resemble Authenticom's unauthorized data extractions. To the extent that Authenticom disputes Rubinfeld's assumptions, it can raise its criticisms through cross-examination.
2. CFAA
Finally, Rubinfeld opines on the damages Authenticom should owe CDK on its CFAA claim. As one alternative form of damages, he assessed the impairment to CDK's systems caused by Authenticom's unauthorized access and querying. He relied on testing by CDK performed in 2017 to measure the percentage of its server resources (“reality”) consumed by Authenticom at four dealership groups, comprising 78 rooftops. This yielded 96 data points. He calculated an average consumption rate of 3.4%, which he apportioned as a percentage of CDK's server hosting fees for the DMSs regularly accessed by Authenticom during the damages period to calculate minimum impairment damages.
Rubinfeld also calculated system impairment costs incurred by Reynolds, but used a different methodology based on available information. Reynolds provided summaries of direct labor costs incurred during the damages period that were associated with detecting and disrupting unauthorized access to its DMS, which totaled $6,677,169. Rubinfeld allocated 17.3% of these costs ($1,155,550) to Authenticom as damages, representing the percentage of hostile access on the CDK DMS that CDK identified as Authenticom between July 2016 and November 2017. He used the percentage ascertained from CDK data because a Reynolds-specific number was not available.
Plaintiffs argue that Rubinfeld's damages opinion should be excluded because it relies on “plainly unjustifiable extrapolation.” [871] at 30. First, Plaintiffs maintain that it was unreasonable for Rubinfeld to use 17.3% as the multiplier for Reynolds' damages, since that percentage was based on data concerning Authenticom's access to CDK's DMS. Defendants counter that this was reasonable because “Reynolds did not have visibility as to what percentage of unauthorized access Authenticom represented, in part because Authenticom concealed its access from Reynolds by attempting to trick Reynolds' DMS and deleting evidence of the reports it ran off of Reynolds DMSs.” [994] at 32. Defendants also assert that the CDK numbers are likely conservative for Reynolds because Reynolds had already stopped certain hostile integrators from accessing its DMSs. Therefore, Defendants posit, Authenticom likely represented a higher percentage of unauthorized access on Reynolds' DMS during the damages period. Second, Plaintiffs argue that Rubinfeld's calculation of CDK's avoidance damage should be excluded because Rubinfeld cannot and does not justify his “facially arbitrary sampling methodologies, ” which looked at a single day of data for four dealership groups. [871] at 30-31.
As Stroz did, Rubinstein used the limited available data to make a rough estimate of damages, since Authenticom's alleged actions at concealment made it impossible to obtain better data. “Speculation has its place in estimating damages, and doubts should be resolved against the wrongdoer.” America Tablewares, 100 F.3d at 1365. Rubinstein explains why his estimate is likely conservative, which also weighs against finding that his opinion is too unreliable to be admissible. Tyson, 488 F.Supp.2d at 733-34. Plaintiffs are “free to use cross-examination to attack the assumption” of 17.3% and ask Rubinfeld “how altering the assumption would affect his analysis.” Stollings, 725 F.3d at 766. A jury should be capable of understanding this. Finally, Plaintiffs do not offer any reason to believe that the four dealership groups tested by CDK were significantly different from Authenticom's dealer customers as a whole or that Authenticom's queries were more frequent during the tested days than during other parts of the damages period. This is another topic that can be explored in cross-examination.
V. Conclusion
For these reasons, the motions to exclude Williams [881] and Stejskal [883] are granted in part and denied in part. The motions to exclude Klein [863], Rubinfeld [867], Murphy [873], Israel [877], Lawton [879], and Halpin [887] are denied. The motion to exclude Miracle [885] is granted in part, denied in part, and deferred in part. The motion to exclude Stroz [859] is denied in part and deferred in part. As indicated in today's minute entry, the parties will be given an opportunity to (a) review this comprehensive opinion, (b) submit supplemental briefs on how, if at all, these rulings affect the briefing and proper disposition of the pending summary judgment motions, and (c) advise the Court on any other pertinent developments of which they believe the Court should be aware. The Court will then prepare and issue a second, comprehensive opinion disposing of the summary judgment motions in due course.