Opinion
No. M 07-1827 SI MDL No. 1827
02-21-2012
In re: TFT-LCD (FLAT PANEL) ANTITRUST LITIGATION This Order Relates to: All Indirect-Purchaser Plaintiff Class Actions
ORDER DENYING DEFENDANTS'
MOTION TO EXCLUDE EXPERT
TESTIMONY OF JANET S. NETZ AND
WILLIAM S. COMANOR
On February 10, 2012, the Court heard argument on defendants' motion to exclude the testimony of plaintiffs' economic experts, Dr. Janet S. Netz and Dr. William S. Comanor. Having considered the moving papers and the arguments of the parties, and for good cause appearing, the Court hereby DENIES defendants' motion.
BACKGROUND
This antitrust class action involves allegations of price-fixing among manufacturers of thin-film transistor liquid-crystal display ("TFT-LCD") panels. TFT-LCD panels are electronic components found in a number of products, including computer monitors, notebook computers, and televisions. See generally Order Granting Indirect Purchaser Plaintiffs' Motion for Class Certification (March 28, 2010), Master Docket No. 1642, at 1-3 ("Class Certification Order"); Indirect Purchaser Plaintiffs' Third Consolidated Amended Complaint, Master Docket No. 2694, at ¶¶16-54 ("TAC"). The panels have no independent utility, and are not available to the average consumer. Instead, they are incorporated into finished products as discrete, physical objects within the products. Class Certification Order at 2.
Defendants in this case are the major global manufacturers of TFT-LCD panels. See TAC at ¶121 (alleging that "the top six companies (Samsung, LGD, Chi Mei, AU Optronics, Sharp and Chunghwa) currently control in excess of 80% of the LCD panels market"). Plaintiffs, consumers who purchased televisions, monitors, and notebook computers that contain defendants' TFT-LCD panels ("indirect-purchaser plaintiffs" or "IPPs"), contend that defendants conspired to raise the prices of the panels between 1999 and 2006. Plaintiffs claim that this conspiracy caused them to pay artificially high prices for the TFT-LCD products they purchased. They seek injunctive relief under federal law and damages under the antitrust and consumer-protection laws of 23 states and the District of Columbia.
In order to prevail on their claims for damages, plaintiffs must establish that the price-fixing conspiracy affected the prices they paid for the TFT-LCD products they purchased. Because plaintiffs did not purchase those products directly from the defendants, they propose to prove this effect through a two-step process. First, plaintiffs will prove that the conspiracy resulted in higher prices for defendants' customers - those companies that directly purchased TFT-LCD panels from defendants. Second, plaintiffs will show that this initial overcharge was "passed through" the manufacturing and retail chains and was included in the final price they paid for their TFT-LCD goods.
To help make this showing, plaintiffs have retained the services of two experts. The first, Dr. Janet S. Netz, holds a Ph.D. in Economics and is "a founding partner of ApplEcon, LLC." Chen Decl., Exh. 2 ("Netz Report") at 1. The second, Dr. William S. Comanor, is an "economist and Professor of Economics at the University of California, Santa Barbara." Chen Decl., Exh. 13 ("Comanor Report") at 2.
The experts' reports cover slightly different topics, but overlap to a significant degree. The plaintiffs' primary expert is Dr. Netz, whose expert report reaches the following conclusions:
1. The cartel formed by Defendants and Co-conspirators was effective at raising the price above the level that would have been obtained had Defendants and Co-conspirators [] behaved independently.Netz Report at 4.
2. The cartel increased the price to direct purchasers by 12.1%.
3. The overcharge to direct purchasers was passed through to indirect purchaser class members.
4. [IPP] [c]lass members paid a price 12.1% higher than they would have had there
not been a cartel.
5. The total nominal amount of harm borne by class members is $3.23 billion over the class period.
Dr. Comanor's report contains similar conclusions. It is primarily directed at the overcharge to direct purchasers, which Dr. Comanor calculates to be approximately 9.6%. Based on economic theory and the characteristics of the TFT-LCD market - but little transactional analysis - Dr. Comanor hypothesizes that the expected pass-through rate to consumers would fall between 100% and 200%.Comanor Report at 26. Dr. Comanor relies on Dr. Netz's analysis - which calculated a pass-through rate of slightly above 100%, depending upon the general product category - to confirm this hypothesis, and employs a "conservative" pass-through rate of 100%. Id. at 29. He ultimately estimates that the resulting damages to indirect purchasers amounted to $2.2 billion. Id. at 20, 29.
100% pass-through signifies that an intermediary charges its customers for the entire amount of an increase in its costs. Pass-through rates may be higher than 100%. For example, if a retailer were always to set the price of a product at twice the price it purchased the product for, any increase in the price paid by the retailer would be doubled when the retailer sets its price, resulting in a 200% pass-through rate.
Defendants now move to exclude the testimony of Dr. Netz and Dr. Comanor. They contend that neither expert can satisfy the requirements of the Federal Rules of Evidence or Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993).
LEGAL STANDARD
Federal Rule of Evidence 702 provides that expert testimony is admissible if "scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue." Fed. R. Evid. 702. Expert testimony under Rule 702 must be both relevant and reliable. Daubert, 509 U.S. at 589. When considering evidence proffered under Rule 702, the trial court must act as a "gatekeeper" by making a preliminary determination that the expert's proposed testimony is reliable. Elsayed Mukhtar v. Cal. State Univ., 299 F.3d 1053, 1063 (9th Cir. 2002), amended by 319 F.3d 1073 (9th Cir. 2003). As a guide for assessing the scientific validity of expert testimony, the Supreme Court provided a nonexhaustive list of factors that courts may consider: (1) whether the theory or technique is generally accepted within a relevant scientific community, (2) whether the theory or technique has been subjected to peer review and publication, (3) the known or potential rate of error, and (4) whether the theory or technique can be tested. Daubert, 509 U.S. at 593-94; see also Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137 (1999).
DISCUSSION
Defendants' motion is based, in large part, on their continued insistence that this case is not amenable to class treatment. As they have argued in the past, defendants contend that plaintiffs cannot prove their case unless they produce highly specific, individualized evidence of classwide injury. Underlying defendants' argument is their assertion that plaintiffs must establish impact "with certainty." Motion at 13; see also Motion at 15 (stating that plaintiffs "overcharge models are useless as evidence that . . . all direct purchasers paid an overcharge on all relevant LCD panels."). Based on this standard, defendants assert that plaintiffs must identify the overcharge that was placed on every LCD panel sold during the conspiracy period and trace that overcharge through the manufacturing and retail distribution chains until plaintiffs can identify the overcharge paid by a particular class member for a particular LCD product. Motion at 13 (stating that it is "necessary to trace the progress of the individual component whose price was allegedly fixed through the specific distribution chain at issue in order to establish with certainty that any increase in retail price of the finished product was not actually caused by other components or configurations unrelated to the supposedly inflated cost fo the LCD panel").
Of course, if defendants' standard were correct then common issues would not predominate in this case. Other courts have concluded as much. See, e.g., In re Graphics Processing Units Antitrust Litig., 253 F.R.D. 478, 505 (N.D. Cal. 2008) ("[T]he only way to fully assess pass-through in this action would be to conduct a wholesaler-by-wholesaler and re-seller-by-re-seller investigation, which would essentially result in 'thousands of mini-trials, rendering this case unmanageable and unsuitable for class action treatment.'"); In re Flash Memory Antitrust Litig., 2010 WL 2332081, at *11 (N.D. Cal., June 9, 2010) ("Determining the pass-through for an individual class member first requires the identification of the particular channel applicable to the class member's purchase, and tracing the overcharge through the various intermediaries that lie between the particular Defendant and the member."). Defendants' position, however, is inconsistent with the nature of this case.
Most fundamentally, there is no requirement that plaintiffs prove impact "with certainty." To the contrary, impact, like the other elements of plaintiffs' case, must be established by a preponderance of the evidence. In fact, antitrust plaintiffs benefit from an especially lenient burden when it comes to proof of impact. See Knutson v. Daily Review, Inc., 548 F.2d 795, 811 (9th Cir. 1976) ("Even as to this minimal quantum of injury, the standard is relaxed; otherwise, it would defeat the loose standard applied even to the amount of damages in antitrust cases."). Thus, plaintiffs need not be able to articulate the precise degree to which every individual class member was injured; it suffices to show that it was more likely that not that classwide impact occurred.
In addition, as this Court has previously noted, the nature of the industry renders defendants' standard inappropriately strict. See Class Certification Order at 30. Plaintiffs assert that TFT-LCD panels are fungible commodities, and were largely treated as such by defendants and their customers. Id. (distinguishing Graphics Processing because "TFT-LCD panels are fungible, interchangeable, and largely homogenous"); Netz Report at 6-7 (noting that a panel purchaser could obtain panels from different manufacturers, provided they met the customer's specification). It is therefore unnecessary for plaintiffs to provide evidence of panel-by-panel impact. Rather, plaintiffs may resort to generalized methods of proof.
This does not, as defendants argue, violate the Rules Enabling Act. See Ortiz, 527 U.S. 815, 845 (1999) ("[N]o reading of [Rule 23] can ignore the [Rules Enabling Act's] mandate that rules of procedure 'shall not abridge, enlarge or modify any substantive right . . . .'" (internal quotation marks omitted)). An individual plaintiff would not be required to meet the stringent standard defendants propose. Rather, such a plaintiff would be entitled to rely on evidence much like the evidence plaintiffs here have set forth: evidence that defendants illegally raised the prices of their TFT-LCD panels, and that retail prices increased as a result.
Thus, the Court rejects the standard articulated by defendants. Plaintiffs need not identify the overcharge on each and every panel sold to direct purchasers, and they need not trace that specific overcharge through the manufacturing and retail chains to the ultimate purchaser. The fact that plaintiffs lack perfect proof does not mean that plaintiffs lack any proof at all.
Turning to the merits of defendants' motion, the Court does not agree that the opinions of Dr. Netz and Dr. Comanor are inadmissible. Before addressing defendants' arguments for wholesale exclusion of plaintiffs' experts, however, the Court will address a narrower form of relief that defendants proposed at oral argument: that, at a minimum, the Court must preclude plaintiffs' experts from arguing that their regression models establish causation or impact.
I. Admissibility of Regression Analyses
As part of their analyses of the cartel's impact on indirect purchasers, Dr. Netz and Dr. Comanor each rely on regression models. To determine the initial overcharge to direct purchasers, both Dr. Netz and Dr. Comanor constructed "before-and-after" models, which "use[] determinants of price during competitive (non-cartel) periods to forecast the price that would have obtained but for the existence of the cartel." Netz Report at 87; see also Comanor Report at 17 ("The method employed is a variation on the widely-accepted Before/After approach."). As stated above, Dr. Netz's model estimates that the conspiracy resulted in a weighted average overcharge of 12.1%. Netz Report at 93. Dr. Comanor's model estimates an average overcharge of approximately 9.6%. Comanor Report at 20, 30. In addition, Dr. Netz constructed a second regression model to help determine the rate at which this overcharge was passed through to plaintiffs. Netz Report at 93-110.
Defendants contend that these models, while relevant to damages, may not be used to establish either impact to direct purchasers or pass-through to indirect purchasers. See, e.g., Motion at 2 ("[The average pass-through rate] might be useful for calculating the aggregate amount of damages [but] it say[s] nothing about whether every class member, most class members, or any particular class member, paid an illegal overcharge and therefore suffered antitrust injury."). Defendants' overarching point is that regression models necessarily use averaging, and that the models therefore cannot establish that every class member was affected by the conspiracy. See, e.g., Graphics Processing, 253 F.R.D. at 504 ("Dr. Netz's regressions would . . . be overly reliant on averages and would thus sweep in an unacceptable number of uninjured plaintiffs . . . .").
As discussed above, this argument is largely premised on the notion that plaintiffs must establish, to a certainty, class members' injuries on a panel-by-panel basis. This is not the correct standard. The very source defendants rely on confirms this point. Although it warns against using the "reduced-form pricing equation" because that model "assumes that a conspiracy has the same effect on every purchaser," the treatise endorses the use of "somewhat more complex models that do not make such an assumption." ABA Section of Antitrust Law, Econometrics: Legal, Practical and Technical Issues 222 (2005). The "somewhat more complex" models the treatise suggests would break the class into categories and sub-classes but would continue to regress data. Thus, these models would also necessarily rely on averaging. See id. at 221 ("By their very nature, regressions summarize data."). They would therefore be unable to provide the panel-by-panel proof that defendants implicitly demand. Cf. id. at 223 (noting "practical and theoretical issues" of regression analyses that "present a formulaic estimate of [a conspiracy's] impact and damages" on every member of the class).
Defendants' concerns, while stated too broadly, are not wholly without merit. Obviously, a model cannot be used to prove of one of its basic assumptions. As defendants point out, plaintiffs appear to recognize this limitation. See, e.g., Opp'n at 2 (stating that, only after determining that there was "class-wide harm" did the experts "construct[] before-and-after regression models to quantify the amount of harm in the form of an average percentage overcharge."); Opp'n. at 7 ("[The experts'] before and after [regression model] results, while also evidence of common impact, are not the basis for their conclusion that there was common impact."); Opp'n at 19 ("The pass-through studies are not designed to [distinguish between cartel overcharges and other cost changes], nor could they, nor need they.").
The Court finds that defendants' proposed remedy is inappropriate, and premature, at this juncture. Even if regression models are not enough, standing alone, to establish classwide impact, they may nevertheless be relevant to the issue. A large average overcharge, for example, might make it more likely that every direct purchaser was overcharged to some degree. And, of course, if defendants are correct that plaintiffs acknowledge the limits of their experts' models, they will have little trouble establishing those points on cross-examination.
Accordingly, the Court DENIES defendants' request to preclude plaintiffs' experts from testifying that their models establish either impact to direct purchasers or pass-through to indirect purchasers. When the experts' specific testimony is before the Court, defendants may renew their objections.
II. Exclusion of Plaintiffs' Experts
Defendants' motion seeks exclusion of Dr. Netz's and Dr. Comanor's opinions on three grounds. Defendants claim: (1) both experts' opinions regarding the overcharge paid by direct purchasers are not scientifically based; (2) Dr. Netz's calculation of the pass-through rate is unreliable; and (3) Dr. Netz's expert report suffers from miscellaneous econometric and mathematical errors.
A. Overcharge to Direct Purchasers
As mentioned above, plaintiffs' first step in establishing classwide impact is proving that direct purchasers were overcharged as a result of the conspiracy. Defendants contend that neither Dr. Netz nor Dr. Comanor used a reliable methodology to make this connection. They claim that plaintiffs have failed to come up with a "common, formulaic method for showing that direct purchasers paid an overcharge on every LCD panel they placed in the distribution chain for eventual resale to indirect purchasers." Motion at 14.
The Court disagrees. As discussed in its class certification order, the Court has found Dr. Netz's analysis appropriate for establishing common impact to direct purchasers:
In support of their argument that antitrust impact can be proven on a class-wide basis, plaintiffs have submitted the expert report of economist Dr. Janet S. Netz. Dr. Netz has examined the TFT-LCD industry and market to determine if the indirect purchaser plaintiffs would have suffered impact as a result of the alleged price-fixing conspiracy. Dr. Netz assumes that there was a conspiracy among TFT- LCD manufacturers as plaintiffs have alleged, and her report asserts that there are a number of characteristics that make the TFT-LCD industry highly susceptible to cartelization and price-fixing, including (1) the lack of an alternative source of supply of TFT-LCD panels and the lack of an alternative technology; (2) the existence of barriers to entry; (3) regular meetings and interactions that allowed defendants to exchange information, come to agreements, and police cheating; and (4) difficulty in cheating due to the use of most-favored customer clauses. Dr. Netz's report examines the price targets set by the cartel and actual market prices, and finds that these prices match well. Based on these analyses, Dr. Netz concludes that the cartel was successful in increasing prices.Class Certification Order at 24-25 (footnotes and citations omitted). Dr. Netz relies on much the same information in her expert report. See Netz Report at 28-87 (using economic theory and analysis of evidence in this case to conclude that the cartel effectively raised prices for direct purchasers).
Next, Dr. Netz considered the commonality of the impact on direct purchasers. Dr. Netz examined the pricing strategies used by defendants, including the use of most-favored customer clauses, the centralization of pricing decisions, and price negotiations starting from a common basis. Dr. Netz also analyzed the prices for TFT-LCD panels to determine whether the prices are related by market forces, and concludes that there is a common pricing structure for TFT-LCD panel prices. In reaching this conclusion, she examined the correlation between panel prices, and found that 76% of panel price correlations (approximately 50,000 of 66,000 correlations) are positive and statistically significant. Dr. Netz then used regression analysis to examine the determinants of panel prices, and found that 77% of price variation is determined by five variables that describe the panel characteristics (panel size, resolution, volume, application, and time-period). Based on these analyses, Dr. Netz concludes that most of the variation in transaction prices is driven by common factors rather than individual ones, and that these common influences on price are susceptible to being estimated using a formula.
Defendants do not engage this component of Dr. Netz's analysis. Instead, they contend that the above analysis, standing alone, is not sufficient to prove common impact. Reply at 12 ("'Economic theory,' without more, is not a reliable methodology for demonstrating class wide impact . . . ."). They appear to argue that some "formal statistical technique" is required to prove that every direct purchaser was overcharged for LCD panels.
The Court disagrees. It is certainly possible that "economic theory" or a high-level analysis of the characteristics of a particular market may be too general or otherwise insufficient to establish classwide impact. See, e.g., In re Aftermarket Automotive Lighting Products Antitrust Litig., 276 F.R.D. 364, 370 (C.D. Cal. 2011) (noting possibility that expert evidence could be "too general" to demonstrate classwide impact); Graphics Processing, 253 F.R.D. at 502 ("Dr. Meyendorff essentially asks this Court to presume impact to direct purchasers based on the characteristics of the market."). Here, however, plaintiffs have far more detailed information, such as defendants' transactional data, detailed notes of discussions held at the crystal meetings, and the prices and production restrictions that defendants purportedly agreed to. This evidence, and Dr. Netz's corresponding analysis, is sufficient to constitute a reliable method of establishing common impact. Cf. In re Commercial Tissue Products, 183 F.R.D. 589, 595 (N.D. Fla. 1998) ("[E]ven if there is considerable individual variety in pricing because of individual price negotiations, class plaintiffs may succeed in proving class-wide impact by showing that the minimum baseline for beginning negotiations, or the range of prices which resulted from negotiation, was artificially raised (or slowed in its descent) by the collusive actions of defendants."); ABA Section of Antitrust Law, Econometrics at 220-21 (noting that "formal statistical techniques . . . may be required to show common impact" (emphasis added)).
Instead of addressing the above, defendants' argument focuses on the experts' use of "before and after" regression models to demonstrate that direct purchasers were impacted by the conspiracy. Motion at 14 ("[T]hese regression models . . . attempt to calculate the average overcharge direct purchasers paid for LCD panels - they were never designed to prove, and in fact cannot prove, that direct purchasers paid an overcharge on every LCD panel placed in the distribution chain from 1999 to 2006."). Because the before-and-after models are not the only methodology plaintiffs rely on to prove impact to direct purchasers, the Court is not convinced that Dr. Netz's opinion is unreliable.
Indeed, it is unclear the extent to which Dr. Netz relies on the model for this purpose, given that it does not appear in the section of her expert report that addresses common impact to direct purchasers. See Netz Report at 28-87.
B. Pass Through to Indirect Purchasers
In her report, Dr. Netz concludes that the overcharge incurred by direct purchasers was passed through to indirect purchasers. Defendants contend that this opinion is unsound. Much of their argument focuses on their contention that Dr. Netz relies too heavily on averages to show that an overcharge was passed through to every individual class member. Motion at 8 ("Dr. Netz's averages . . . provide no data about whether each individual class member was injured."). As discussed above, the Court does not agree that such a specific purchaser-by-purchaser showing is required. See also Class Certification Order at 30-31.
Defendants' specific challenges to Dr. Netz's pass-through analysis also do not provide reason to exclude her opinion. First, defendants contend that Dr. Netz's opinion that pass-through occurred is "based on a mere assumption" because "[p]erfect pass-on is a theoretical construct that cannot be presumed to exist in the real world." Motion at 6. Defendants assert that Dr. Netz acknowledges as much. See Chen Decl., Exh. 8 (testimony of Dr. Netz that pass-through occurs "the vast majority of the time"). Defendants also claim that Dr. Netz's pass-through conclusions are contradicted by anecdotal evidence obtained during discovery. This evidence consists of deposition testimony or declarations by various OEMs and retailers that increases in costs of components or products were not always passed along to customers. See, e.g., Chen Decl. at Exhs. 3 (Dell), 4 (Best Buy), 5 (Newegg.com), 6 (Target), 7 (Radioshack), 24 (HP). Finally, defendants assert that transaction data obtained from Dell demonstrates that pass through did not always occur. They claim that Dell sold some of the panels it purchased to third parties and that in 30% of these sales, Dell's sales price "did not move in lockstep with changes in Dell's cost to purchase panels." Motion at 9.
None of these arguments is reason to exclude Dr. Netz's testimony. Defendants' disagreements with the bases for her opinion do not render that opinion so inherently suspect that it may not be admitted at trial. These arguments are precisely the type of matter that is appropriate for cross-examination.
As a corollary argument, defendants argue that Dr. Netz's methodology "sweep[s] in an unacceptable number of uninjured plaintiffs." Motion at 7 (quoting Graphics Processing, 253 F.R.D. at 504). They claim that: "Even assuming that by 'vast majority of the time' Dr. Netz meant 99% of the time, and assuming, conservatively, 50 million members of the classes here, that leaves 500,000 class members potentially eligible to collect treble damages although they suffered no injury whatsoever. If Dr. Netz meant '95% of the time,' that would amount to 2.5 million uninjured members of the class." Id.
Setting aside the implications of this argumentthat class treatment is inappropriate because the injury in this case was too widespreadthe Court declines to exclude Dr. Netz's opinion based upon one passing sentence in her deposition. Dr. Netz has opined that all class members were affected by the conspiracy. Netz Report at 87 ("I conclude that the higher prices borne by direct purchasers were passed on to class members . . ."). Defendants may explore her deposition testimony on cross examination.
Next, defendants assert that Dr. Netz's regression model "cannot distinguish between a retail price increase caused by the pass-through of a cartel overcharge and a retail price increase caused by the retailers' decision to raise prices for other, legitimate reasons." Motion at 9. Defendants contend that there are thousands of different products that contain LCD panels, and that many of these products contain thousands of different configurations. See Chen Decl., Exh 3 at ¶11 (noting that there are over 40,000 potential configurations for the Dell Inspiron 15). They claim that this circumstance renders it impossible to determine the effect an increase in panel prices had upon the price of a finished product. Defendants rely on Judge Armstrong's criticism of Dr. Netz from a case involving allegations of price fixing in among manufacturers of flash memory:
The tracing process requires more than the one-size-fits-all theoretical construct proposed by Dr. Netz. Within the three broad categories of products at issue (flash memory cards, USB flash drives and flash-based digital media players), there are thousands of differentiated products with diverse price points that have been purchased by putative class members over the course of the last decade. Such products are sold to class members by hundreds of different retailer suppliers. The price of any given product can vary across retailers, and differ even for a single retailer at a given point in time. Further, different retailers respond to cost changes in different ways, with some choosing not to pass-through cost changes in the form of higher prices for the end-user. The "economic theory" cited by Dr. Netz's accounts for none of these anomalies.Flash Memory, 2010 WL 2332081, at
The Court again finds that defendants' focus is too narrow. Plaintiffs need not reconstruct each LCD product sold to a class member and determine the price of its constituent parts to establish that the class member paid an overcharge attributable to the conspiracy. Rather, the Court agrees with plaintiffs that they may establish pass-through by showing that companies in the manufacturing and distribution chains passed along cost increases in general. Dr. Netz's report does exactly that. Plaintiffs' proof need not rise to the level of establishing how every intermediary treated every LCD panel or product:
A key concept in Dr. Leffler's analysis is the logical proposition that an increase in the cost of Microsoft OS affects each purchaser in the same manner as an increase in any other cost: "the impact of a change in Microsoft's charge to the distribution channelGordon v. Microsoft Corp., 2001 WL 366432, at *11 (Minn. Dist. Ct., March 30, 2001). As above, defendants' arguments presume that plaintiffs must prove these matters to an absolute certainty, when such is not the case.
would have the same impact as a comparable change in other costs faced by the distribution channel." Leff. Aff. ¶ 21. This concept permits the effect on consumer prices of a Microsoft OS overcharge to distributors to be determined by actual analysis of the effects of other cost increases.
Finally, defendants assert that Dr. Netz's models are "100-Percent-Pass-Through Machines." Motion at 17. They claim that Dr. Netz's regressions "virtually always generate pass-through of 100% or more," regardless of whether the data actually shows any pass-through whatsoever. Defendants contend that this occurs because Dr. Netz compares the cost of a product to its price, instead of comparing the change in cost to the resulting change in price, if any. Defendants provide two example of this, but neither convinces the Court that Dr. Netz's methodology is unsound. See Motion at 18. If, as defendants contend, their examples are "simple but accurate" then they will be appropriate material for cross-examination. They are not, however, sufficient to establish that Dr. Netz's opinions are unreliable.
III. Other Errors
Finally, defendants contend that Dr. Netz's report includes a number of fundamental errors. For example, they assert that Dr. Netz's model relies on only half the available data because it ignores data from Japanese firms. They contend that including this data would reduce the average overcharge computed by her model in half, from 12.1% to 5.8%. They also contend that Dr. Netz's model fails to account for the downward trend in LCD panel prices during and after the conspiracy period. Finally, defendants argue that Dr. Netz's reply model contains three "critical" errors: 1) it applies the wrong calculation to convert from the but-for price (produced as a logarithm) to the final prediction; 2) it uses a log-based weighting scheme when Dr. Netz meant to use a levels-based scheme; and 3) the model predicts a negative 6.3% overcharge in 2001.
None of these purported errors are serious enough to warrant exclusion of Dr. Netz's report. Defendants may confront Dr. Netz about her methodological choices on cross-examination.
CONCLUSION
For the foregoing reasons and for good cause shown, the Court hereby DENIES defendants' motion to exclude expert testimony of Drs. Netz and Comanor. Docket Nos. 4380, 4381.
IT IS SO ORDERED.
_______________
SUSAN ILLSTON
United States District Judge