"The Apple iPod iTunes Anti-Trust Litigation"

Filing 285

Reply in Support of 240 Motion for Decertification of Rule 23(B)(3) Class filed by Apple Inc.. (Mittelstaedt, Robert) (Filed on 11/9/2009) Modified on 11/13/2009,(counsel failed to properly link to motion.) (cv, COURT STAFF).

"The Apple iPod iTunes Anti-Trust Litigation"

Doc. 285

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page1 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Robert A. Mittelstaedt #60359 [e-mail] Craig E. Stewart #129530 [e-mail] Michael Scott #255282 [e-mail] JONES DAY 555 California Street, 26th Floor San Francisco, CA 94104 Telephone: [phone] Facsimile: [phone] Attorneys for Defendant APPLE INC. UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA SAN JOSE DIVISION

THE APPLE iPOD iTUNES ANTI-TRUST LITIGATION.

Case No. C 05-00037 JW C 06-04457 JW DEFENDANT'S REPLY IN SUPPORT OF MOTION FOR DECERTIFICATION OF RULE 23(B)(3) CLASS Date: Time: Place: November 23, 2009 9:00 A.M. Courtroom 8, 4th floor

Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW
Dockets.Justia.com

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page2 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 IV. V. VI. VII. B. C. 3. III. II.

TABLE OF CONTENTS Page INTRODUCTION .......................................................................................................................... 1 I. PLAINTIFFS' PROPOSED STANDARD FOR MOTIONS TO DECERTIFY IS CONTRARY TO THE CASES THEY CITE..................................................................... 2 PLAINTIFFS CANNOT BLAME APPLE FOR THEIR TACTICAL DECISION NOT TO ATTEMPT TO SHOW THAT THEIR ECONOMIST'S PROPOSED METHODS WILL WORK. ................................................................................................ 3 DR. NOLL PROPOSES THE SAME UNRELIABLE METHODS AS DR. FRENCH. .................................................................................................................... 4 A. Temporal Benchmark/Before-After. ....................................................................... 5 1. 2. Rapid technological innovation coupled with an insufficient number of price changes in the before period make this method unworkable......... 5 Dr. Noll's dismissive attitude towards "coolness" underscores another central problem with all of his methods......................................... 7 Dr. Noll's approach to the "start date" reflects another fundamental problem he shares with Dr. French. ............................................................ 9

Yardstick Method.................................................................................................. 11 Markup Method..................................................................................................... 11

PLAINTIFFS' ATTACK ON DR. BURTIS IS UNFOUNDED AND IRRELEVANT.................................................................................................................. 12 PLAINTIFFS HAVE ALSO FAILED TO OFFER A RELIABLE METHOD TO SHOW IMPACT............................................................................................................... 13 NET OVERCHARGE....................................................................................................... 14 RESELLERS..................................................................................................................... 15

CONCLUSION ............................................................................................................................. 15

-i-

Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page3 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

TABLE OF AUTHORITIES Page Cases Ballard v. Equifax Check Servs., Inc., 186 F.R.D. 589 (E.D. Cal. 1999) ................................................................................................. 2 Bell Atl. Corp. v. AT&T Corp., 339 F.3d 294 (5th Cir. 2003)...................................................................................................... 13 In re Graphics Processing Units Antitrust Litig., 253 F.R.D. 478 (N.D. Cal. 2008)............................................................................................. 3, 4 In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305 (3d. Cir. 2008).............................................................................................. 2, 4, 13 In re New Motor Vehicles Canadian Export Antitrust Litig., 522 F.3d 6 (1st Cir. 2008) .......................................................................................................... 13 Los Angeles Mem'l Coliseum v. Nat'l Football League, 791 F.2d 1356, 1367 (9th Cir. 1986).......................................................................................... 14 Reed v. Advocate Health Care, 2009 WL 3146999 (N.D. Ill. Sept. 28, 2009) .............................................................................. 4 Slaven v. BP Am., Inc. 190 F.R.D. 649 (C.D. Cal. 2000) ................................................................................................. 2 Rules Federal Rules of Civil Procedure Rule 23 ................................................................................................................................ passim Federal Rules of Civil Procedure Rule 23(b)(3).................................................................................................................... 1, 13, 15 Treatises Fed. R. Civ. P. 23(c) advisory committee's note ......................................................................... 3, 4

- ii -

Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page4 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

INTRODUCTION Plaintiffs' opposition confirms why a class of past iPod purchasers does not satisfy Rule 23(b)(3). Plaintiffs have failed to establish any reliable method to prove impact and damages with common proof. Like Somers' economist, plaintiffs' economist Dr. Noll has failed to demonstrate that the prices of the original 2001-2002 white iPod can be used to divine the "but for" prices of the iPod touch, shuffle, nano and mini from 2003 to present. Indeed, it defies common sense for them to suggest that they could take, for example, the $399 price of the original iPod in 2001 and determine whether, if Apple had licensed its anti-piracy software, the price of the iPod shuffle in 2009 would have been lower than $79. These circumstances precluded certification in Somers, and the result should be the same here. As the Court found in Somers, this is a failure at the initial step of "demonstrat[ing] supracompetitive pricing for iPods" even before reaching the issue of whether any purported overcharges were passed on to indirect purchasers. Dkt. 80, p. 12. Although both economists proposed "potentially available" methods to prove impact and damages, neither could say whether they would actually work here. They did not gather the necessary data, identify the specific variables that affect iPod pricing, or actually run any model. Although plaintiffs tout Dr. Noll's "accomplishments" in other areas, he has no "clue" how iPods work, cannot tell an iPod touch from the first iPod, and thinks the notion that iPods are "cool" is a "complete joke." In these circumstances, identifying "potential methods," which is all that he has done, is not enough to satisfy Rule 23, particularly given the obstacles identified by the Court in Somers and Dr. Burtis' showing that the proposed methods will not work. This conclusion is buttressed by Dr. Noll's admission in another case that his preferred method does not work for products with "rapid technological change" that experienced a "boom in sales." That description fits iPods to a tee. As Dr. Noll put it, in these circumstances, this method is "unlikely to produce reliable results." Plaintiffs' brief confirms two other reasons for decertification. Dr. Noll concedes that, under plaintiffs' theory, the price of iTS music may have been higher without the alleged violation. This means that, to prove impact, plaintiffs must show that a consumer paid more in -1Def's Reply I/S/O Decert. Of 23(B)(3) Class 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page5 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

alleged iPod overcharges than the consumer saved by any iTS undercharges. That is an individualized inquiry, and plaintiffs have failed to show any reliable classwide method of proof. Moreover, plaintiffs' argument that resellers like Best Buy and Target, which do not buy iTS music, do not have a net overcharge problem only underscores that resellers are differently situated and do not belong in a consumer class action. Resellers buy huge volumes of iPods, through different channels and at different prices and terms from consumers. Consumers do not adequately represent these differently situated resellers, and a class action by consumers is not a necessary or superior way to pursue a reseller claim, if any valid claim existed. I. PLAINTIFFS' PROPOSED STANDARD FOR MOTIONS TO DECERTIFY IS CONTRARY TO THE CASES THEY CITE. As shown in Apple's motion (p. 6), the Court has broad discretion to revisit certification decisions because they are "inherently tentative," and the Court applies the same standard used in deciding whether to certify the class in the first place. Plaintiffs' argument (p. 5) that decertification requires "new controlling law or facts" is contrary to their own case. Slaven v. BP Am., Inc. expressly rejected that argument, recognizing that a motion to decertify "assists the Court in performing its role as gatekeeper, or manager, of the class action." 190 F.R.D. 649, 652 (C.D. Cal. 2000). As that court stated, "Plaintiffs fail to recognize ... that this Court's usual reluctance to entertain motions for reconsideration simply does not apply in the class certification context." Id. at 651; see also Ballard v. Equifax Check Servs., Inc., 186 F.R.D. 589, 593 n.6 (E.D. Cal. 1999) (due to the court's power to revisit certification rulings, it "need not consider whether `reconsideration' is also warranted"). Slaven's reference to a defendant's "heavy burden" was based on the now-widely discredited view that "doubts regarding the propriety of class certification should be resolved in favor of certification." 190 F.R.D. at 651 (citation omitted). As the more recent cases hold based on recent amendments to Rule 23, "the court should not suppress `doubt' as to whether a Rule 23 requirement is met." In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 321-22 (3d. Cir. 2008). Instead, "the trial court [must be] satisfied, after a rigorous analysis, that the prerequisites of Rule 23 have been satisfied." In re Graphics Processing Units Antitrust Litig., 253 F.R.D. -2Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page6 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

478, 492 (N.D. Cal. 2008) (citation omitted); see also Fed. R. Civ. P. 23(c) advisory committee's note ("A court that is not satisfied that the requirements of Rule 23 have been met should refuse certification until they have been met.") II. PLAINTIFFS CANNOT BLAME APPLE FOR THEIR TACTICAL DECISION NOT TO ATTEMPT TO SHOW THAT THEIR ECONOMIST'S PROPOSED METHODS WILL WORK. Plaintiffs recognize (p. 15, n.10) that economists in other cases have actually performed regression analyses at the class certification stage in an attempt to establish the reliability of their methods of proving classwide impact and damages. As this Court made plain in Somers, Rule 23 requires plaintiffs to show that their proposed methods are reliable as applied. Here, everyone agrees that Dr. Noll has not performed any regression analysis. And he admits that until he does so, he has no way to determine if his methods will work.1 Plaintiffs decided that Dr. Noll should not perform regression analyses. When he asked them whether to do so, they said: "No, that's not going to happen." Noll Dep. II 139:2-19.2 Plaintiffs suggest (p. 4) that Apple agreed that no regression analysis was needed at this stage. That is false, as shown by the e-mail quoted by plaintiffs. Plaintiffs put the last sentence of the email in bold font in the hope that the previous sentence will be skipped over. That sentence makes clear that it was plaintiffs who took the position that Dr. Noll did not need to produce an impact or damage study at this stage. Indeed, plaintiffs never claimed that they wanted or needed anything more for what they had asked Dr. Noll to do. Moreover, even after the Court's ruling on
1

Noll Reply p. 32 ("[N]o one can yet know that an omitted variable problem exists in this matter."); p. 36 ("[T]he existence, nature and extent of the omitted variable problem in a price regression that would be used to calculate damages in this case can not be known, much less proved, until data are discovered and regressions are run."); p. 45 ("[N]o conclusion can be drawn from these price cuts without discovery and analysis."); p. 52 ("Information about wholesale prices and sales requires discovery."); p. 53-54 ("I do not yet possess enough facts about Apple's business strategies to know how best to model its price policies."); p. 54 ("[E]conomists must select a model on the basis of discovery and the success of each model in fitting the data."); see also Noll Dep. II 14:20-15:22, 23:19-24:8, 30:21-31:9. All cited deposition pages are attached to the accompanying Kiernan Declaration. Unless otherwise noted, "Ex." references in this memorandum are to exhibits to the Kiernan Declaration. -3Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW
2

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page7 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

plaintiffs' class motion, plaintiffs have not asked for most of the information that Dr. Noll now claims he would need. Kiernan Decl., ¶¶ 2-8. And Dr. Noll did not obtain much of the thirdparty data that he says he might use. See, e.g., Noll Dep. II 234:25-239:25. Plaintiffs' actions belie any revisionist claim that they wanted Dr. Noll to do anything more than what he has done. Indeed, despite the false impression they seek to create, plaintiffs do not assert that they would have asked Dr. Noll to do additional work if Apple had produced more data. Plaintiffs alternatively argue that they are not required to show that their models will work. As this Court ruled in Somers, however, at least in cases like this in which "significant challenges" exist to implementing the proposed methods (Dkt. 80 (Somers), p. 12), a plaintiff must do more than simply offer potentially "available" or "plausible" methods. Indeed, even in cases in which a plaintiff's expert has run regression analyses, certification has been denied because the analysis was inadequate. Graphics Processing, 253 F.R.D. at 495-97; Reed v. Advocate Health Care, 2009 WL 3146999, at *20-21 (N.D. Ill. Sept. 28, 2009). That result applies a fortiori here where plaintiffs have not yet attempted to construct their model. In short, plaintiffs cannot obtain class certification where they have failed to gather the necessary data and to ask their expert to perform the necessary analyses. Whether motivated by an attempt to save money or to avoid the risk of demonstrating the unreliability of their methods, plaintiffs' effort is insufficient under Rule 23 in the circumstances of this case.3 III. DR. NOLL PROPOSES THE SAME UNRELIABLE METHODS AS DR. FRENCH. The thrust of plaintiffs' opposition is their mistaken assertion (pp. 2, 11) that their economist's approach bears "only minimal resemblance" to Dr. French's approach and that

Requiring plaintiffs to do the work necessary to show that their proposed models are reliable is supported by the 2003 amendments to Rule 23. As the Advisory Committee Note explains, this amendment makes clear that plaintiffs should be required before seeking class certification to undertake the discovery necessary to "identify the nature of the issues that actually will be presented at trial," so that the court can "mak[e] the certification decision on an informed basis." Fed. R. Civ. P. 23(c) advisory committee's note; see also Hydrogen Peroxide, 552 F.3d at 318-19 n.20, 23 (discussing significance of this change and relying on it to conclude that "[e]xpert opinion with respect to class certification, like any matter relevant to a Rule 23 requirement, calls for rigorous analysis"). -4Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

3

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page8 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

comparing the two approaches is "specious at best." In fact, and as plaintiffs eventually admit (p. 11), both economists propose the same three methods and both of them focus on the "before-after method" as the method that they will likely try to use.4 Both economists claim that they will analyze prices paid by direct purchasers like Wal-Mart, Amazon, or Target as well as retail prices. While Somers' economist also proposed to analyze whether direct purchasers passed on any overcharge, this Court found that he faltered on this initial step of "demonstrat[ing] supracompetitive pricing for iPods," even before reaching the separate question of pass-on to consumers. Dkt. 80 (Somers), p. 12. As the following discussion shows, Dr. Noll is no more persuasive than Dr. French in asserting that the three proposed methods are plausible, and he certainly has not demonstrated that any of them would be reliable here. A. Temporal Benchmark/Before-After. 1. Rapid technological innovation coupled with an insufficient number of price changes in the before period make this method unworkable. The problem is not, as plaintiffs suggest (p. 14), simply that "multiple models of iPods exist." Rather, the obstacle is that many of the key innovative features of the post-2003 iPod models did not exist before 2003--innovations such as dramatically reduced sizes with even greater capacity, touch screens, wireless internet access, photo and video capabilities, and the availability of tens of thousands of applications from the App Store. As Dr. Noll admitted, the original iPods were "thick and clunky" relative to the new iPods, like the nano, which are "really tiny and sleek." Noll Dep. II 45:9-13. Reflecting these dramatic changes, the prices of the new iPods bear no resemblance to earlier iPods--$399 for an original iPod compared to $79 for a current iPod shuffle. On these facts, Dr. Noll has failed to demonstrate that he can use the price formulation process in 2001 for the original iPod to determine what the "but for" price of, for example, an iPod touch with new functionality should have been in 2009. This is a particularly impossible

4

Dr. Noll stated that, given the difficulties inherent in the other methods, he "guess[es]" that the before-after method is the one he will try to use. Noll Dep. II 238:6-23. -5Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page9 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

task because, as Dr. Noll admitted, no theoretical basis exists to conclude that the alleged violation--Apple's use of proprietary software--would have had any impact on iPod prices, let alone an impact large enough to measure with a regression analysis with the multitude of factors affecting iPod prices in the marketplace. Noll Dep. II 85:20 - 86:5. Although Dr. Noll now claims that he can account for all of this, he was more realistic in his expert report in the Flash Memory case. He admitted there that the before-after method is "unlikely to produce reliable results" in the face of rapid technological innovation leading to a "boom" in sales: "To implement the before-after method, transactions data are needed for a couple of years before the instigation of price collusion and a substantial period after the end of the class period. . . . Without such data, the before-after test can not be implemented. Prior to 1999, NAND flash memory was not a particularly important product. Innovations in both consumer electronics products and NAND flash memory in the early 2000s completely transformed the NAND flash memory market, increasing sales by two orders of magnitude. As a result, trying to compare the price formulation process in 1997-1998 with prices after the boom in small consumer electronic appliances is unlikely to produce reliable results." Ex. 5 (Decl. of Roger G. Noll, in In Re Flash Memory Antitrust Litigation, p. 54.).5 The very same obstacle exists here. Dr. Noll admits that portable devices for playing digital music have undergone tremendous change during this decade. Noll Dep. II 94:17-95:8. And like Flash Memory, the sales of portable digital players have increased by at least two orders of magnitude since 1999.6

Dr. Noll denied reaching this conclusion, perhaps because he was mistakenly thought his Flash Memory report was under seal and not available publicly. Noll Dep. II 102:7-103:3. He claimed that a reply declaration overcame the obstacle. But at least the public redacted version makes no mention of it or says anything about whether a before-after method to determine impact or damages would work. Ex. 13. In fact, in Flash Memory, Dr Noll was simply trying to show that prices charged by different vendors responded to the same factors; he was not attempting to prove any overcharge caused by alleged misconduct. Thus, even if he would eventually succeed, he would not establish that he could do what he proposes here. Noll Dep. II 59:3-9.
6

5

Compare Ex. 6 (reporting worldwide 1999 MP3 player sales of 1.3 million) to Ex. 7 (estimating worldwide 2009 MP3 player sales of 225 million, more than two orders of magnitude higher than in 1999). -6Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page10 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

This is not the only obstacle Dr. Noll faces. As Dr. Burtis explains, there is an insufficient number of required price changes to use a regression to establish the impact of the key variables on price. Burtis Report, ¶ 17. Challenging that conclusion, plaintiffs assert (p. 7) that Dr. Burtis has not collected or analyzed price data. That is incorrect. Dr. Burtis' report shows the small number of price changes before April 2003. Burtis Report, Ex. E. Plaintiffs' criticism that she could not, sitting at her deposition, remember cases with a before period longer than 18 months is irrelevant. The issue is whether sufficient price changes exist, not simply the length of the before period. Tellingly, plaintiffs fail to identify any case in which the baseline period contained so few price changes. With no good answer, Dr. Noll avoided the issue in his supplemental report. At deposition, Dr. Noll eventually conceded that he did not know if he would have enough price changes to produce reliable results. Noll Dep. II 23:19-24:2. 2. Dr. Noll's dismissive attitude towards "coolness" underscores another central problem with all of his methods. It is fundamental that a regression analysis cannot produce reliable results where factors affecting price are not identifiable and quantifiable. Burtis Report, ¶ 18; Noll Reply, p. 32. In that circumstance, the influence of an unidentified or unquantified factor will be mistakenly attributed to the dummy variable that measures the supposed effect of the alleged misconduct. If, for example, consumers' perception of the iPod as cool accounts for 20% of the iPod price, and if Apple's use of proprietary software accounts for zero %, a regression analysis that fails to account for coolness will find a 20% overcharge where none exists. See Burtis Report, ¶ 18; Burtis Reply Report, ¶ 5. Dr. Noll made clear at deposition that he rejects the concept that iPods are "cool" and that this perception could affect demand. This from an individual who has never bought or used an iPod (Noll Dep. II 10:8-9), cannot tell an iPod touch from the first iPod (id. 8:11-25), thinks a shuffle is called a "clip,"7 has "no clue" how they work (id. 119:1-15), and has no interest in why

Dr. Noll claimed that the shuffle is known as a "clip" "in common parlance" and in unidentified "trade literature." Noll Dep. II 103:19-104:5. -7Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

7

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page11 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

consumers buy them.8 Given his utter lack of personal familiarity with the product, Dr. Noll must rely on scholarly and other literature rather than his personal expertise. But then Dr. Noll ignores even that. The literature he cites shows that the consumers who buy iPods do so in part because they consider them "cool." "[S]ome people are undoubtedly drawn to the iPod by the buzz around it." Ex. 8, p. 87. "Firms that possess "the cool factor" have a powerful advantage over their competitors." Ex. 9, pp. 14-17. Apple has an "obvious `cool factor'" as described in an article entitled The iPod Phenomenon: Identifying a Market Leader's Secrets through Qualitative Market Research. Ex. 10, p. 239. Dr. Noll, however, cannot account for this qualitative attribute, so he dismisses all of this as "joke." Noll Dep. II 144:10-20. As a fallback, Dr. Noll claims that if coolness affects iPod demand, it can be ignored because it has been a constant influence from the very first day that iPods were sold. But Dr. Noll has no basis to conclude that coolness has had a constant, unchanging impact on iPod demand from the launch of the first iPod model in 2001 through the successive launches of the 42 different models over nearly a decade. Indeed, his position defies common sense and is contradicted by his own cited authority. See e.g., Ex. 8, p. 88 (iPod "has jumped from minority coolness to mass coolness."). Dr. Noll's final fallback argument (Noll Reply p. 22) is that coolness results from "functionality and ease of use" and that those attributes can be objectively measured. However, neither common sense nor the literature cited by Dr. Noll suggests that a product is cool simply because it is functional and easy to use. Otherwise, a toaster or Kleenex would be cool, and one pair of blue jeans would be as cool as any other. As described in the sources he cites but then ignores, coolness is much more ephemeral and difficult to measure. "Of course `coolness' is a hard factor to define." Ex. 10, p. 249.9

"What their reasons are [for buying iPods] in some sort of psychological sense is irrelevant." Noll Dep I 83:20-21. Capturing the essence of Dr. Noll's response, another study cited by him explains: "To people grounded in rationality...the whole subject [of coolness] can be incredibly frustrating." Ex. 8, p. 76. -8Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW
9

8

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page12 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

In any event, Dr. Noll's argument that coolness is the same as "functionality" and "ease of use" cannot be reconciled with his assertion that coolness remained unchanged over the last eight years. To state the obvious, and as Dr. Noll acknowledged, iPod functionality has increased dramatically, from a simple music player to a small computer capable of taking and displaying photos and videos, serving as a PDA, receiving and send e-mail, accessing the Internet, and running tens of thousands of application from the App Store, in addition to playing music. Noll Dep. II 94:17-95:11. 3. Dr. Noll's approach to the "start date" reflects another fundamental problem he shares with Dr. French. Nor do plaintiffs adequately address the additional major obstacle of identifying when, if ever, Apple's use of proprietary software for iTS would have had any impact on iPod prices, i.e., the start date for the impact period. This is important because if the wrong start date is used, the analysis may find impact and damages where none exist. For example, it may attribute an increase in iPod demand to the alleged violation when it should have been attributed to conduct not challenged in this case, like the launch of the music store itself, which plaintiffs concede was a pro-consumer, pro-competitive innovation. Despite the critical importance of this issue, plaintiffs say only (p. 20), in one short sentence, that Dr. Noll shows why the April 2003 launch of the music store is a "reasonable starting period for the class." The issue, however, is not the starting period for the class but when, if ever, a relative increase in iPod prices compared to "but for" prices can be attributed to Apple's decision not to license its proprietary software. Although plaintiffs' brief is silent on this, Dr. Noll's second report (p. 46) asserts that consumers first suffered damages in April 2003 because that is the start of the class period alleged in the complaint. But the complaint merely sets forth allegations; it is nothing for an expert to rely on. Dr. Noll further claims that April 2003 is the start date because that is when owners of competing portable digital music players would have been "exclud[ed]" from buying music from iTS. But that is not the damage theory of this case. Plaintiffs are not claiming, much less attempting to quantify, damages for owners of iPod rivals who were discouraged from buying iTS -9Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page13 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

music. The only damages sought in this case are an alleged iPod overcharge. Thus, the argument that owners of rival players were harmed by not being able to use iTS provides no basis for selecting April 2003 as the start date for an alleged iPod overcharge.10 In any event, even if plaintiffs had any basis for selecting April 2003 as the alleged start date of impact and damages, it would only compound their problems. The April 2003 launch of iTS increased demand for devices that could play digital music including iPods. Plaintiffs challenge only one aspect of iTS, i.e., that the music could be played with fewer intermediate steps on iPods than competing portable music players. Thus, as he acknowledged (Noll Dep. I 127:22-128:19), Dr. Noll needs to disentangle any effect of the decision not to license its antipiracy software from all the other aspects of the iTS launch that may have increased demand for iPods. He has not proposed let alone demonstrated any reliable method to do so. He asserts that he can examine effects of subsequent increases in online music. But doing so will not account for the numerous other innovations of the iTS introduction, such as a la carte song purchases, CD burning, and ownership, as opposed to monthly subscription, of online music. Burtis Reply Report ¶ 14. Nor has Dr. Noll proposed or demonstrated any way to disentangle the effects of the launch of new iPod models in April 2003 with new features including USB ports. Dr. Noll admitted that improvements in the way that iPods connected to computers could impact their demand. Noll Dep. II 127:25-129:21; Ex. 11, p. 3 (2001 article cited by Dr. Noll explaining that USB would enable Apple to sell iPods to "the Windows world"). But he erroneously claimed that iPods connect wirelessly to host computers, falsely implying that adding a USB port would be unimportant. Id.

10

Perhaps recognizing that this approach for picking the start date would not work, Dr. Noll tried to invent another theory at deposition. He claimed that, in the but for world, Apple would make decisions regarding the music store as if it did not sell iPods and would have licensed its antipiracy software to portable digital player manufacturers even before launching its music store. Noll Dep. II 74:23-76:3. For good reason, however, this theory has never been urged by plaintiffs. No antitrust duty exists that would require Apple to alert competitors in advance of its plans to launch a music store and ensure that they were ready and able to manufacture compatible players. This is just a stop-gap creation of Dr. Noll's with so little merit that he did not mention it in either of his reports and that plaintiffs do not adopt it. - 10 Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page14 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

In short, for these reasons and the others set forth previously, Dr. Noll has failed to establish that the before-after method will work here B. Yardstick Method.

As unreliable as the other methods are, plaintiffs acknowledge (pp. 22-23) that the yardstick method "poses more challenges" than the other methods. Indeed, Dr. Noll admitted that he had doubts that he could identify a product sufficiently comparable to iPods to permit him to determine what iPod prices or margins would have been in the but for world: the "hangup[] is identifying the appropriate benchmark products." Noll Dep. I 72:2-9. Dr. Noll contradicts that admission in his second report, in which he suggests that his only concern with the yardstick method is "the cost and availability of information about other products." Whatever his concerns, the fact is that, as he admitted at his recent deposition, he does not know if the results of any yardstick method will be robust until he identifies a comparable product and attempts to run a regression. Despite the time he has had, he has not done either. Noll Dep. II 223:12-23. C. Markup Method.

Plaintiffs' brief does little more than express Dr. Noll's "confiden[ce]" that he will be able to determine "typical mark-ups" in other consumer electronic industries and opine that, if iPods' profit margin is higher, it must be because Apple did not license its software to competitors. They have done nothing to show that the method will work here. Indeed, the published study referred to as an example shows why this approach will not work. The study shows that profit margins of companies selling consumer electronics ranges from a high of over 50% to negative figures, with the iPod's reported gross margin lower than at least three firms in its supply chain. Ex. 12, p. 9, Table 2, p. 12, Table 5. Dr. Noll has suggested no method to determine where in that wide range the iPod would have fallen but for the misconduct alleged here. Dr. Noll claims (p. 50) that this is a better approach than the yardstick method because it avoids the need to find a product comparable to iPods and "does not examine the technical details of products." But if Dr. Noll cannot find a comparable product, it is illogical to say that using data from dissimilar products sold in other industries becomes more reliable simply because it is - 11 Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page15 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

"more aggregate." See Burtis Reply Report ¶ 21. And if Dr. Noll does not intend to adjust for "technical details" of the products in the different industries he intends to survey, his results could not possibly be reliable. Id. Finally, neither plaintiffs nor Dr. Noll has produced any document showing that a gametheoretic method has ever been used to determine impact or damages. Moreover, this method does not depend on any Apple data. If there were any basis to think it would work here, nothing has stopped Dr. Noll from doing the work. In short, Dr. Burtis' "extensive explanation of how the types of regression models proposed by Dr. French cannot be reliably applied to the complex product and pricing dynamic underlying the claims in this case" applies with full force to the same regression models proposed by Dr. Noll. Dkt. 80 (Somers), p. 11. IV. PLAINTIFFS' ATTACK ON DR. BURTIS IS UNFOUNDED AND IRRELEVANT. None of plaintiffs' criticisms of Dr. Burtis is appropriate, and none excuses plaintiffs from their failure to carry their burden under Rule 23. As to relative accomplishments, as far as the record shows, Dr. Noll has never "accomplished" a regression analysis that attempts to do anything close to what he is proposing here. Dr. Burtis did not "crib" her report in Somers. Her report directly addresses Dr. Noll's report in this case. True, she identifies similar defects to those in Somers. That is because Dr. Noll is proposing the same methods as Dr. French. Plaintiffs criticize (p. 8) Dr. Burtis for being unable at deposition to "identify even a single difference between Dr. French and Professor Noll." To be more precise, the deposition question asked about differences in the economists' "three basic approaches"--which were the same. Dr. Burtis testified, accurately, that the three basic methods were the same and that neither economist "went very far along the road of implementing them." Burtis Dep. 85:14-21. Dr. Burtis added that, unlike Dr. French, Professor Noll was not even sure his regression analysis would use a "dummy variable" although he offered no hint as to how a regression without such variable could work. Id. 87:3-88:5. Contrary to plaintiffs' suggestion (p. 11), the defect in Dr. French's report was not its length. It was the lack of substance and failure to demonstrate that the proposed methods would - 12 Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page16 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

work here. Dr. Noll's supplemental report adds more pages, much of it generic filler. But as shown above it falls short of establishing that the three methods will work. Plaintiffs' assertion (p. 8) that Dr. Burtis and Apple initially viewed Noll's proposed methods as "unassailable" or conceded their validity is false and unsupported. To the contrary, Dr. Burtis correctly concluded that his first report provided little information about his methodologies. Burtis Dep. 81:4-17. And as shown in Apple's motion (Dkt. 240, pp. 6-7), Apple also relied on Dr. Noll's admission that he did not know whether his proposed methodology would work. Apple further argued that he had not offered any proposal on how to prove a net overcharge taking music prices into account. Contrary to the plaintiffs' implication (p. 9), the Court did not rule on either point. V. PLAINTIFFS HAVE ALSO FAILED TO OFFER A RELIABLE METHOD TO SHOW IMPACT. The issue here is not limited, as plaintiffs state (pp. 1, 9), to whether they have demonstrated a reliable method of determining the amount of damages. They also must show that their methods will show impact or fact of damage. The requirement of antitrust impact is an "element of the cause of action" and "often is critically important for the purpose of evaluating Rule 23(b)(3)'s predominance requirement." Hydrogen Peroxide, 552 F.3d at 311. "[W]here fact of damage cannot be established for every class member through proof common to the class, the need to establish antitrust liability for individual class members defeats Rule 23(b)(3) predominance." Bell Atl. Corp. v. AT&T Corp., 339 F.3d 294, 302 (5th Cir. 2003); In re New Motor Vehicles Canadian Export Antitrust Litig., 522 F.3d 6, 28 (1st Cir. 2008) (plaintiffs must present evidence "that each member of the class was in fact injured"). Although plaintiffs state (p. 9) that one can use Dr. Noll's regression methods to show impact and damages, they have not shown that those methods are reliable here. Indeed, at his first deposition, Dr. Noll acknowledged that he would not attempt to use his regression methods to show impact: "[I]t's too difficult, beyond the scope of a reasonable attainment by an econometrician." Noll Dep. I p. 248:10-249:2. And his second report does not assert that the regressions are intended to prove impact. Rather, his second report states (p. 10) that "harm to - 13 Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page17 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

competition," his term for impact, can be proved "by showing that the market share and profit margin of the iPod were higher after the launch of iTMS." At deposition, however, he admitted the obvious: an increasing market share and profit margin may result from superior products, unrelated to whether Apple licensed its proprietary software at issue in this case. Thus, merely showing an increase in market share or margins does not show harm to competition or impact of the alleged violation. Finally, as discussed above, Dr. Noll did not run any models to determine whether his methods would work for any aspect of the case, including impact. VI. NET OVERCHARGE. Plaintiffs attack (p. 24) Dr. Burtis for suggesting that, under plaintiffs' theory, the price of iTS music may have been lower as a result of the alleged violation. But Dr. Noll agrees with Dr. Burtis on this point, and he described it as an empirical inquiry because economic theory does not answer the question. Thus, in the but for world where Apple would have licensed its software, "[p]rices for music downloads could go up, down, or stay the same." Noll Reply p. 26. Plaintiffs' brief ignores this concession. Nor do plaintiffs refer to Dr. Noll's further concession that, if music was priced lower in the tying world, whether an individual consumer is harmed by the alleged violation will depend on the relative size of that consumer's purchases of music and iPods, and the relative size of the undercharge for music and overcharge for iPods. Noll Dep. II 205:23-206:19; see alsoBurtis Report ¶ 36; Dkt. 240, p. 9. This is an individual issue that defeats certification. Contrary to plaintiffs' suggestion (p. 23), this Court has not already decided the issue. And the cases they cite relate to an entirely different doctrine--i.e., whether direct purchasers can sue for the "full overcharge" even if they passed on some or all of it. Even in that context, plaintiffs must show that they paid a net overcharge. Nor is it relevant that plaintiffs' tying claim has been dismissed. The requirement of showing a net injury is not limited to section 1 tying cases, but is instead a general principle of antitrust law. See Los Angeles Mem'l Coliseum v. Nat'l Football League, 791 F.2d 1356, 1367 (9th Cir. 1986) ("[A]n antitrust plaintiff may recover only to the `net' extent of its injury; if benefits accrued to it because of an antitrust violation, those benefits must be deducted from the gross damages caused by the illegal conduct."). And the - 14 Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW

Case5:05-cv-00037-JW Document285

Filed11/09/09 Page18 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

requirement applies with particular force here, where plaintiffs' section 2 claims are based on exactly the same conduct as their section 1 tying claim. VII. RESELLERS. Trying to minimize the net overcharge obstacle, Dr. Noll observed that resellers like Best Buy and Target buy only iPods and not iTS content and, thus, do not face the same issue as consumers. But this argument simply underscores that resellers are differently situated and should not be included in a class with consumers. Resellers do not buy iPods in the same channels or at the same prices paid by consumers and do not have the same interests as consumers. Plaintiffs have never shown that they can adequately represent resellers or that permitting them to do so is a superior method for resellers to resolve any claim against Apple. Presumably, if Best Buy or Target think that they have paid a supracompetitive price for iPods, they have adequate resources to pursue such claims on their own. No reason exist for resellers to want individual consumers to represent their interests. In any event, a reseller class has the same deficiencies as a consumer class. As Dr. Noll observed, resellers account for the vast majority of damages that would be claimed here because their purchases are so much larger than individual direct consumers like Charoensak, Rosen and Tucker. Noll Dep. II 18:10-15. Yet, as with individual consumers, Dr. Noll has not gathered the data (indeed, plaintiffs have not requested the prices resellers paid for iPods), identified the relevant variables or run any models. CONCLUSION For the foregoing reasons, the Rule 23(b)(3) damages class should be decertified. Dated: November 9, 2009. Respectfully submitted, Jones Day By: /s/ Robert A. Mittelstaedt Robert A. Mittelstaedt Counsel for Defendant APPLE INC.
SFI-622543v7

- 15 -

Def's Reply I/S/O Decert. Of 23(B)(3) Class C 05-00037 JW; C 06-04457 JW