1 2 3 4 5 6 7 8 9 Robert A. Mittelstaedt #60359 ramittelstaedt@jonesday.com Craig E. Stewart #129530 cestewart@jonesday.com David C. Kiernan #215335 dkiernan@jonesday.com Michael T. Scott #255282 michaelscott@jonesday.com JONES DAY 555 California Street, 26th Floor San Francisco, CA 94104 Telephone: (415) 626-3939 Facsimile: (415) 875-5700 Attorneys for Defendant APPLE INC. 10 UNITED STATES DISTRICT COURT 11 NORTHERN DISTRICT OF CALIFORNIA 12 SAN JOSE DIVISION 13 14 15 THE APPLE iPOD iTUNES ANTI-TRUST LITIGATION 16 _____________________________________ 17 This Document Relates To: 18 ALL ACTIONS 19 20 Lead Case No. [Class Action] C 05-00037 JW (HRL) APPLE’S SUPPLEMENTAL OBJECTIONS TO REPLY DECLARATION OF ROGER C. NOLL AND SUPPLEMENTAL OPPOSITION TO CLASS CERTIFICATION MOTION Date: April 18, 2011 Time: 9:00 a.m. Courtroom: 8, 4th Floor 21 22 23 24 25 26 27 28 Apple's Supp. Obj. To Noll Reply Decl. C 05-00037JW (HRL) 1 On April 4, 2011, Apple filed its objections (Doc. 572) to the Reply Declaration of Roger 2 C. Noll (Doc. 551) filed on March 28, 2001 with Plaintiffs’ class certification reply papers. In 3 those reply papers, Plaintiffs argued that Professor Noll had run a “working regression analysis” 4 that “demonstrates that impact and damages can be proven by relying on common proof.” Doc. 5 550, pp. 2, 8. Apple objected to Professor Noll’s purported analysis on the ground, among others, 6 that Professor Noll’s regression lacks foundation and any indicia of reliability. Doc. 572, pp. 2-3. 7 Apple now submits these supplemental objections and opposition in light of Professor Noll’s 8 testimony at his deposition on April 7, 2011. At that deposition, Professor Noll confirmed the 9 validity of Apple’s objection by admitting that his regression model does not work and is not 10 reliable. 11 Professor Noll’s testimony is addressed more fully in Apple’s opposition to Plaintiffs’ 12 motion to exclude the opinion of Michelle Burtis and in the supporting declaration of Michelle 13 Burtis, which Apple is filing today. As those documents describe, Professor Noll admitted: • 14 His regression is not a workable, valid damages model: “I do not regard this as a 15 damage model.” Noll Dep. at 90.1 “I am not here to defend as the final damages 16 equation that which I have done because it’s obviously incomplete.” Id. at 110. • 17 He does not know whether his model contains specification errors—i.e., errors in 18 constructing the model that could bias the results. When asked whether there is a 19 “factor that’s been excluded from this regression that leads to the wrong answer,” he 20 admitted “I don’t know that. That’s why I wouldn’t offer this as a damages model.” 21 Id. at 91. He conceded that his model is “not proof that this is the right specification.” 22 Id. at 92. • 23 His model does not account for such things as Apple’s pricing strategy. Id. at 185 24 (“Q. Does your current regression in your reply report account for the fact that Apple 25 changed its retail prices and wholesale prices, list prices infrequently? A. No, that’s 26 why it’s not a damage model.”). 27 28 1 The cited pages from the Noll deposition are found in Exhibit 1 to the Kiernan Declaration submitted in support of Apple’s opposition to the motion to exclude. Apple's Supp. Obj. To Noll Reply Decl. C 05-00037JW (HRL) 1 • The results of his model may reflect a “spurious correlation”—i.e., attributing to the 2 launch of Harmony a price change that was actually caused by another factor. Id. at 3 87-89. 4 • He thus cannot draw any “causal inferences from that regression” (id. at 90), including 5 any inference on the pivotal question for which Plaintiffs offer his testimony—i.e., 6 whether Apple’s software update that shut down Harmony caused any change in the 7 price of iPods. As he summarized, “I cannot rule out anything based on the regression 8 I have because I’m not relying on it for anything.” Id. at 130-131. 9 In short, Professor Noll has admitted that, rather than producing a reliable model, he has 10 simply taken a set of data, input it into an incorrectly specified computer model and produced 11 unreliable results that are not probative on any relevant issue. For this reason, as well as the other 12 reasons stated in Apple’s previously filed objections, the Court should exclude Noll’s reply 13 declaration and give his purported regression no weight. 14 Alternatively, should the Court not exclude Noll’s reply declaration, Apple respectfully 15 requests that the Court consider in connection with Plaintiffs’ motion for class certification the 16 memorandum and supporting Burtis Declaration that Apple is filing today in opposition to 17 Plaintiffs’ motion to exclude the opinion of Dr. Burtis. 18 Dated: April 11, 2011 JONES DAY 19 20 By: 21 Attorneys for Defendant APPLE INC. 22 23 /s/ Robert A. Mittelstaedt Robert A. Mittelstaedt SFI-672476v1 24 25 26 27 28 Apple's Supp. Obj. To Noll Reply Decl. C 05-00037JW (HRL)