"The Apple iPod iTunes Anti-Trust Litigation"
Declaration of Helen Zeldes in Support of 214 Reply to Response to Motion to Certify Class and Appointment of Co-Lead Class Counsel filed byStacie Somers. (Attachments: # 1 Exhibit, # 2 Exhibit, # 3 Exhibit)(Related document(s) 214 ) (Haeggquist, Alreen) (Filed on 5/19/2009)
UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA SAN JOSE DIVISION
THE APPLE IPOD ITUNES ANTITRUST LITIGATION ___________________________________ This Document Relates To: ALL ACTIONS
) ) ) ) ) ) ) )
Case No. C 07-6507 JW
Reply Affidavit of Gary L. French, Ph.D., Regarding Class Certification
TABLE OF CONTENTS
Table of Exhibits................................................................................................................. ii I. Introduction............................................................................................................. 1 A. Qualifications.............................................................................................. 1 B. Background ................................................................................................. 2 C. Retention ..................................................................................................... 3 D. Assumptions and Additional Materials Considered ................................... 3 II. Common Proof of Impact and Liability.................................................................. 5 A. Dispersion in iPod Retail Prices Does Not Preclude Common Proof of Impact ......................................................................................................... 5 1. Economic Theory Predicts and Explains Price Dispersion................... 5 2. Price Dispersion Would Exist Regardless of Apple's Alleged Misconduct............................................................................................ 7 3. iPod Prices at Major Retailers Exhibit Less Variation than Defendant's Anecdotal Evidence Implies ................................................................. 7 4. Price Dispersion Does Not Preclude Econometric Analysis of Overcharges and Pass-Through to Indirect Purchasers ........................ 8 B. Contrary to Defendant's Portrayal, the iPod Reseller Market is Highly Concentrated ............................................................................................... 9 C. Numerous Other Defendant Criticisms Are Inaccurate or Irrelevant ....... 11 1. Defendant's Criticism That Use of Averages Masks Variation Is Unwarranted........................................................................................ 11 2. Common Impact Does Not Require Universal Lock-In ..................... 14 3. Defendant Mischaracterizes My Testimony Regarding the Relevance of Minimum Advertised Prices........................................................... 16 III. Feasible Methods to Estimate Class-Wide Damages on a Common Basis .......... 18 A. The French Affidavit Proposed Appropriate Benchmarks ....................... 18 1. The Yardstick Method Was Considered ............................................. 18 2. Economic Analysis to Determine the Impact Period Is Routine ........ 19 B. Third-Party Retail Price Data to Perform the Proposed Analyses Are Available ................................................................................................... 19 IV. Conclusions........................................................................................................... 20
TABLE OF EXHIBITS
Recent Advertised Retail Prices of Portable Digital Music Players by Apple's Top Nine Retailer Direct Purchasers Apple's Direct Sales of iPod Units to Reseller Direct Purchasers, April 2003 to March 2008
UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA SAN JOSE DIVISION
THE APPLE IPOD ITUNES ANTITRUST LITIGATION ___________________________________ This Document Relates To: ALL ACTIONS
) ) ) ) ) ) ) )
Case No. C 07-6507 JW
Reply Affidavit of Gary L. French, Ph.D., Regarding Class Certification
STATE OF VIRGINIA
) ) SS.: COUNTY OF ARLINGTON ) Gary L. French, Ph.D., being duly sworn, deposes and says,
I. A. 1.
I am senior vice president and consulting economist at Nathan Associates Inc., a consulting firm established in 1946 that provides economic and financial research and analysis to public and private clients in the United States and abroad. My qualifications and experience, as well as a list of the matters in which I have testified over the last four years, are contained in the text and an appendix of my
expert report concerning class certification that Plaintiff1 submitted on February 23, 2009 ("French Affidavit").2
The French Affidavit described economic evidence and analysis showing that any impact on and damages incurred by members of the proposed class of indirect purchasers of the Apple3 iPod due to Apple's alleged misconduct would be established using common proof and common methodologies. Defendant's counsel deposed me on April 3, 2009.4 On April 20, 2009, Defendant filed its memorandum opposing Plaintiff's motion for class certification ("Opposition Memorandum").5
Since the French Affidavit was filed, Apple implemented a policy affecting the Apple iTunes Music Store ("iTMS") originally announced in January 2009.6 Under the changes that took effect on April 7, 2009, Apple removed the digital rights management ("DRM") from digital music sold through iTMS, charged iTMS customers a fee to upgrade previously purchased DRM-protected digital audio files
Stacie Somers is referred to herein as "Plaintiff."
United States District Court for the Northern District of California, San Jose Division, The Apple iPod iTunes Antitrust Litigation., Case No. 07-6507 (JW), Affidavit of Gary L. French, Ph.D., February 23, 2009. Apple Inc. is referred to herein as "Apple" or "Defendant."
United States District Court for the Northern District of California, San Jose Division, Stacie Somers v. Apple Inc., Case No. 07-6507 (JW), Deposition of Gary L. French, Ph.D., April 3, 2009 (hereinafter, "French Deposition Transcript").
United States District Court for the Northern District of California, San Jose Division, Stacie Somers v. Apple Inc., Case No. 07-6507 (JW), Defendant's Memorandum in Opposition to Motion for Class Certification, April 20, 2009.
French Aff. at 16, 34 and 42.
to DRM-free versions, and introduced a three-tier pricing scheme (with each title priced at $0.69, $0.99 or $1.29).7
Counsel for Plaintiff has asked me to consider and evaluate the economic arguments propounded in Defendant's Opposition Memorandum, specifically: To comment on the issues of iPod retail price dispersion and reseller heterogeneity, among others, raised in the Opposition Memorandum; and To address Defendant's criticisms regarding the damage methodology described in the French Affidavit, particularly Defendant's comments pertaining to potential competitive benchmarks, the commencement of the impact period, and the availability of data.
Nathan Associates continues to bill Plaintiff at a rate of $450 per hour for my efforts and at the usual and customary rates for the time of other economists and analysts employed at Nathan Associates who have assisted in this analysis under my direction. As before, the compensation that Nathan Associates receives for these efforts is not contingent on the outcome of this matter.
Assumptions and Additional Materials Considered
The formulation of this reply continues to rely on the assumption that, since April 2003, when Apple began to sell Fairplay DRM-restricted music content, Defendant tied sales of the Apple iPod to audio downloads sold through Apple's iTMS and
See, for example, Chmielewski, Dawn. "Hottest Tracks to Cost $1.29 at iTunes Starting April 7," Los Angeles Times, March 26, 2009; Stone, Brad. "Making Sense of New Prices on Apple's iTunes," The New York Times, April 7, 2009; and The Associated Press. "Changes Take Effect in Apple's iTunes Prices," MSNBC.com, April 7, 2009.
monopolized and attempted to monopolize sales of portable digital audio players, as Plaintiff alleges. It is not merely assumed that the alleged violations impacted members of the proposed Class. As described herein and in the French Affidavit, economic analysis establishes that such impact would be shown by common proof.
In preparing this reply affidavit, I have reviewed or relied upon, in addition to the items listed in Appendix B in the French Affidavit, Defendant's Opposition Memorandum, the Court's recent Order in the direct purchaser action,8 relevant industry press, and other materials cited herein. The opinions expressed in this affidavit may change if additional information warrants.
This reply affidavit is organized as follows: Section II addresses issues that Defendant raised in its Opposition Memorandum concerning the existence of common proof of impact and liability. Section III revisits the questions of feasible methods and available data to estimate class-wide damages on a common basis. Section IV contains my conclusions.
United States District Court for the Northern District of California, San Jose Division, The Apple iPod iTunes Antitrust Litigation., Case No. C 05-00037 (JW), Order Granting in Part Defendant's Motion for Judgment on the Pleadings; Ordering Supplemental Briefing, May 15, 2009.
COMMON PROOF OF IMPACT AND LIABILITY
In its Opposition Memorandum, Defendant mistakenly claims that dispersion in the prices of Apple iPods and the number and variety of iPod retail outlets preclude common proof of impact and damages. In addition, Defendant makes erroneous arguments regarding the use of averages in the estimation of damages and the relevance of lock-in. Furthermore, Defendant mischaracterizes my testimony regarding the relevance of Apple's "minimum advertised prices."
Dispersion in iPod Retail Prices Does Not Preclude Common Proof of Impact The regression models proposed in the French Affidavit account for price dispersion, which is typical of many markets. Economic theory predicts and explains price dispersion, even for homogeneous products. Such price dispersion would exist whether or not Apple engaged in the misconduct alleged in Plantiff's complaint. Moreover, the dispersion that Defendant identifies in the retail prices of the Apple iPod does not preclude reliable estimation of class-wide damages. Defendant has therefore given undue attention to the magnitude and relevance of dispersion in the prices of the Apple iPod.9
1. Economic Theory Predicts and Explains Price Dispersion 11. Economic theory predicts and numerous empirical economic studies have identified price dispersion in a variety of products. As the late Professor George Stigler
For example, Defendant states, "Their retail prices have varied by as much as $50 or more for the same model iPod, resulting in thousands of different prices on the 42 different iPod models sold during the alleged class period." (Opp. Memo., p. 1). Yet elsewhere, Defendant claims that prices do not vary enough in order to execute a regression analysis. (Opp. Memo., pp. 16-17).
theorized, price dispersion can arise as a consequence of the combination of search costs and imperfect information: When buyers incur search costs, the sellers of that particular product may price discriminate against consumers unwilling to incur the costs of searching for a lower price. Such price discrimination by sellers is one explanation that accounts for variation in the retail price of a given product, at a given moment in time.10
Imperfect information is a reality of most markets, including those for products sold both online and in "brick-and mortar" retail establishments. Empirical studies have found price dispersion among products and services as varied as airline tickets,11 CDs and books,12 digital cameras and flatbed scanners sold online,13 and miscellaneous consumer goods.14
13. While the existence of search costs and imperfect information can explain differences in the prices of a specific iPod model at a point in time, differences in
Stigler, George (1961). "Economics of Information," Journal of Political Economy, Vol. 69, pp. 213225. See also Salop, Steven and Joseph Stiglitz (1977). "Bargains and Ripoffs: A Model of Monopolistically Competitive Price Dispersion," Review of Economic Studies, Vol. 44, pp. 493-510; and Pratt, John W., David A. Wise and Richard Zeckhauser, (1979). "Price Differences in Almost Competitive Markets," Quarterly Journal of Economics, Vol. 93, pp. 189-211. Clemons, Eric K., Il-Horn Hann and Lorin M. Hitt (1999). "The Nature of Competition in Electronic Markets: An Empirical Investigation of Online Travel Agent Offering," working paper, The Wharton School, University of Pennsylvania. Brynjolfsson, Erik and Michael D. Smith (2000). "Frictionless Commerce? A Comparison of Internet and Conventional Retailers," Management Science, Vol. 46, pp. 563-585. Baylis, Kathy and Jeffrey M. Perloff (2002). "Price Dispersion on the Internet: Good Firms and Bad Firms," Review of Industrial Organization, Vol. 21, pp. 305-324. Pratt, John W., David A. Wise and Richard Zeckhauser, (1979). "Price Differences in Almost Competitive Markets," Quarterly Journal of Economics, Vol. 93, pp. 189-211.
features, characteristics and capabilities also contribute to price differences across various iPod models.
2. Price Dispersion Would Exist Regardless of Apple's Alleged Misconduct 14. Given the differences in iPod models and the market realities of imperfect information in the market for portable digital music players, and given the differences in the costs that prospective purchasers of the Apple iPod would incur and are willing to incur to seek out the lowest price, it is axiomatic that dispersion would exist in the retail prices of Apple iPods, whether or not Apple engaged in the tying and monopolization conduct as alleged. The monopoly premium imposed by Apple's alleged misconduct would simply raise the range or distribution of iPod prices across the board.
3. iPod Prices at Major Retailers Exhibit Less Variation than Defendant's Anecdotal Evidence Implies 15. Empirical evidence collected from the websites or stores of Apple's largest retailer direct purchasers of iPods shows that the variation in the prices that the major retailers of iPods charge in the United States is typically nominal. As detailed in Exhibit 1, among Best Buy, Wal-Mart, Target, Circuit City, Costco Wholesale, RadioShack, Sam's Club, CompUSA and Amazon.com, there is modest variation in the advertised retail prices of four different models of Apple iPods (Classic, Nano, Touch and Shuffle) after accounting for memory size and generation of the technology. As Exhibit 1 shows, of these nine top retailers, seven appear to have essentially the same regular prices for most of the iPod models listed. Only Costco
Wholesale and Sam's Club, two warehouse clubs, have slightly lower prices. From April 2003 to March 2008, the nine retailers listed in Exhibit 1 purchased about 80 percent of the iPod units that Apple sold directly to resellers.15
4. Price Dispersion Does Not Preclude Econometric Analysis of Overcharges and Pass-Through to Indirect Purchasers 16. The dispersion in iPod prices that Defendant has pointed to in the marketplace is amenable to regression analysis to estimate a misconduct overcharge and passthrough to indirect purchasers. Accounting for potential systematic retail price differences among retailers is one reason the French Affidavit proposes to rely on retailer-specific data in these regression models. These data include Apple's direct purchaser iPod transaction data as well as monthly retailer-level MP3 player price data by brand and SKU.16 The existence and availability of the latter were verified through direct inquiries with a data vendor, the NPD Group. The level of detail captured in these point-of-sale data permits granular analysis of the determinants of Apple iPod prices over time.
Defendant asserts that anecdotal variation in iPod retail prices is fatal to common proof of class-wide impact and common damage methodology. But regression analysis can control for differences in the level of retailer-specific pricing through the introduction of indicator or "dummy" variables.17 In a regression model in
French Aff., Exhibit 5. French Aff. at 70-71.
As described in note 65 of the French Affidavit, an indicator variable takes a value of one or zero based on whether the attribute in question, such as a particular retailer, is applicable to the data observation.
which the overcharge to members of the indirect purchaser class is directly measured,18 such an indicator variable for each retailer can be included as an explanatory variable. Such an indicator variable would control for systematic differences in the level of retail prices at different retail outlets, and the regression result would show which retailers tend to charge higher or lower prices than their competitors. Similarly, various product features, characteristics, and capabilities can be included as explanatory variables in the regression model. The resulting regression model would yield an estimate of the anticompetitive overcharge percentage across the retailers and iPod models.
Contrary to Defendant's Portrayal, Apple's iPod Sales to Resellers Are Highly Concentrated Of Apple's iPod sales to resellers, nearly 90 percent are sold to twelve "mega" resellers.19 In its Opposition Memorandum, Defendant misleadingly overstates the impact and consequence of the number and heterogeneity of resellers of the Apple iPod.20 First, the "thousands" of resellers that Defendant claims belies the concentration of Apple iPod sales in a handful of resellers. From Apple's own data, over 1,000 of its 1,183 reseller direct purchasers bought fewer than 2,000 iPods from Apple during April 2003 to March 2008; in contrast, the top twelve direct purchasers together account for nearly 50 million units, or almost 90 percent of these Apple iPod unit sales over the same period (Exhibit 2). Nine of these top
This regression model is described in the French Aff. at 69. French Aff., Exhibit 5. Opp. Memo., pp. 1, 4.
twelve direct purchasers are retailers. After excluding the sales to the three nonretailers, Ingram Micro Inc., AF Services21 and Tech Data Product, the top nine retailers accounted for approximately 80 percent of Apple's iPod sales to resellers during this five-year span.22
Second, as described above, in the direct approach to estimating the overcharge to indirect purchasers, the use of retailer indicator variables in regression analysis can account for systematic differences in the level of pricing of individual retailers. In the two-step approach to estimating the indirect purchaser overcharge, the direct purchaser price data can be matched to indirect purchaser data "by iPod model, month or quarter of purchase, and retail outlet" (emphasis added), as described in the French Affidavit.23
Ultimately, Defendant exaggerates the implications of the number of iPod resellers by ignoring the concentration of iPods handled by the dozen largest resellers and by failing to consider the flexibility of multiple regression analysis, such as that proposed in the French Affidavit, to take such factors into account in a model of class-wide damages.
AF Services, LLC operates as a wholly-owned subsidiary of PC Mall, Inc. (PC Mall, Inc., 10-K, filed March 16, 2009, Exhibit 21.1). AF Services was previously characterized as a retailer in the French Affidavit. (French Aff. at 24). French Aff. at 24 and at Exhibit 5. French Aff. at 67-68.
Numerous Other Defendant Criticisms Are Inaccurate or Irrelevant
21. In its Opposition Memorandum, Defendant offers additional criticisms of my analysis. As described in the first section below, Defendant mischaracterizes the proposed use of averages to estimate overcharges and overstates the relevance of the court's decision in the Pioneer Valley matter (in which an entirely distinct damage methodology was proposed). The second section addresses and responds to Defendant's contention that whether and to what extent an iPod consumer is locked in to the iPod due to iTMS is idiosyncratic. Additionally, Defendant mischaracterizes my testimony regarding Apple's minimum advertised prices, as explained in the third section below.
1. Defendant's Criticism That Use of Averages Masks Variation Is Unwarranted 22. Defendant misleadingly claims in its Opposition Memorandum that, "Although [Dr. French] has not done any work yet on a damage model, he says that he plans to deal with all these variations using `averages'.... As the Pioneer Valley court found, his purported use of `averages' is meaningless" and averages mask variation due to reseller differences, among other characteristics.24 Defendant criticizes me for purportedly proposing an impact and damage methodology that relies on broadly averaged data and, in support of this contention, cites a court's decision in the Pioneer Valley matter,25 even though the two proposed damage methodologies are entirely different. Defendant exhibits an inability or unwillingness to distinguish
Opp. Memo., pp. 2, 7, 11-14.
United States District Court for the Southern District of Texas, Houston Division, Pioneer Valley Casket Company, Inc., et al. v. Service Corporation International, et al., Civil Action No. H-05-3399.
between the estimation of an average overcharge or average pass-through (both of which the proposed regression models will yield) from the use of broadly averaged data in the regression models, data that I have not proposed using in the manner that Defendant alleges.
There is no similarity in the impact and damage methodologies employed in Pioneer Valley and those proposed here. Pioneer Valley involved the refusal of the top three manufacturers of caskets to sell to independent casket distributors ("ICDs") pursuant to an alleged conspiracy among the manufacturers and funeral homes. For reasons explained in my four expert reports in that case, not having access to the top manufacturers' caskets put the ICDs at a competitive disadvantage to funeral homes and prevented their sales from being as large as they would have been otherwise. Thus, the measure of impact and damages to the proposed ICD class was the lost profits on lost casket sales by ICDs.
24. The judge relied on the magistrate judge's Memorandum and Recommendation of November 24, 2008, to deny class certification in his March 26, 2009 order. The magistrate argued against class certification, in part, because my proposed sample survey of ICDs could not, in his view, be used to "prove the fact of damages as to every class member."26 Nor did the magistrate believe that a lost profit damage methodology using an average ICD profit margin could take account of other
United States District Court for the Southern District of Texas, Houston Division, Pioneer Valley Casket Company, Inc., et al. v. Service Corporation International, et al., Civil Action No. H-05-3399, Memorandum and Recommendation, November 24, 2008 ("Memorandum and Recommendation"), p. 27.
factors affecting an ICD's profitability.27 However, no survey methodology is being proposed to estimate damages in the present case. Rather, a multiple regression model would be employed that would indeed separate the influence of Defendant's conduct on iPod prices from other factors that cause price variations, such as the market variables measuring demand and supply conditions, features and characteristics of the various iPod models, and the retailers selling the iPod. Thus, no average price of iPods analogous to the average profit margin proposed in Pioneer Valley would be employed. Rather, the regression model would measure the effect of Defendant's conduct on iPod prices separately from the influence of other factors influencing iPod prices.
Furthermore, at no point have I proposed to "run [the] analysis using only monthly averages across all resellers..." as Defendant claims.28 In support of this allegation, Defendant misrepresents my deposition testimony regarding the nature of retail data collected by the NPD Group. As described in the French Affidavit, the NPD Group's retail price data is collected at the individual retailer level for specific models of MP3 players (at the SKU level).29 During the deposition, I testified that the NPD Group "collect[s] retail data from most retailers...on MP3 players"30 and publishes the "monthly average [prices] for each product" on a monthly basis.31 The
27 28 29 30 31
Memorandum and Recommendation, pp. 13, 23-24. Opp. Memo., p. 7. French Aff. at 70-71. French Dep. Tr., p. 16:2-15. French Dep. Tr., p. 16:16-24.
data are thus a panel of point-of-sale prices, by retailer, by month, and by SKU. Through multiple telephone conversations conducted with representatives of the NPD Group prior to the filing of the French Affidavit, it was verified that the NPD Group reports monthly retail data at the SKU and retailer level. Defendant's concern regarding the use of prices averaged across resellers and its criticism that I do "not actually know if [the retail data to run the proposed analysis] exists"32 are therefore unwarranted.
2. Common Impact Does Not Require Universal Lock-In 26. In its Opposition Memorandum, Defendant claims that the question of consumer lock-in to the iPod depends on individual circumstances, such as the size of the iTMS library, willingness and ability to "burn and rip," and income.33 In advancing this argument, Defendant disregards the proposed methodology and Plaintiff's allegation, which states that lock in of a sufficient number of iPod purchasers can yield market-wide effects in the price of the iPod, effects that would impact all indirect purchasers. Whether such lock-in occurred sufficiently to affect the market price for the iPod is a question common to all indirect purchasers that, as proposed in the French Affidavit, is answerable by common evidence and analysis. Thus, even if the question of whether an individual were locked in to the iPod were to be answered according to idiosyncratic facts, the issue would be irrelevant to the determination of common proof of class-wide impact.
Opp. Memo., p. 7. Opp. Memo., pp. 8-9.
27. Under Plaintiff's allegations, common proof of impact does not require that all iPod purchasers were locked in. Common proof derives from a subset of purchasers being sufficiently locked in to the iPod due to Apple's alleged market power in the downloadable digital music segment to have elevated their willingness to pay for a given iPod or to prefer an iPod over another brand of portable digital music player. If the preferences of a sufficiently large proportion of iPod purchasers were distorted in this fashion by Defendant's conduct, then this change in preferences and behavior could elevate the demand and therefore the market prices for iPods above the prices that would have prevailed but for Apple's conduct.
28. The questions relevant to common impact in this matter therefore rely on evidence common to members of the proposed class. Such questions for the tying claim may include: (i) does the Defendant possess market power in downloadable digital music, (ii) whether iPod purchasers changed their preferences and behavior in response to the tie (or does a tie exist for some iPod consumers), and (iii) whether and to what degree these changes in consumer preferences and behavior inflated the prevailing retail price of the iPod. For the monopolization claim, questions that would be examined with common evidence include: (i) whether the DRM that protected music purchased from iTMS locked in consumers to iPods and excluded these consumers from considering other brands of portable digital music players, (ii) whether lock-in resulted in higher demand for and prices of Apple iPods, and (iii) whether the lock-in and the elevation in demand were sufficient to enable
Apple to achieve monopoly power in the portable digital music player market. Thus, Apple's claim that impact must be proven at the individual level and cannot be proven at the market level is inaccurate, given that the consequence of the alleged tie was to distort the prices that Apple could command for iPods.
3. Defendant Mischaracterizes My Testimony Regarding the Relevance of Minimum Advertised Prices 29. Defendant misleadingly claims that, "As an essential part of his `pass on' analysis, plaintiff's economist posits...that Apple requires sellers to adhere to Apple's `minimum advertised price.'"34 On the contrary, at no point have I asserted that pass-through from direct purchasers to indirect purchasers requires adherence to Apple's minimum advertised prices. As stated in the French Affidavit, a condition for Apple's minimum advertised prices to be common proof of impact on indirect purchases is that the minimum advertised prices were binding.35 Contrary to Defendant's interpretation, the question of whether these minimum advertised prices were binding does not depend solely on a legal constraint whereby Apple explicitly requires resellers to use minimum advertised prices as minimum transaction prices (which, as Defendant points out, Apple's sample contracts with resellers explicitly state was not the case). Minimum advertised prices can be binding in the marketplace if advertised prices shape consumer expectations of the price of a particular product. In other words, the minimum advertised prices do not
Opp. Memo., p. 5. See also Opp. Memo., p. 15. French Aff. at 28.
need to be legally binding on transaction prices in order to be influential or practically binding on selling prices.
30. Implicit in many economic and marketing studies is that advertised prices convey information to consumers and influence the price that consumers expect to pay for a product.36 Long-standing empirical economic research also shows that price advertisements tend to lower the prices that consumers pay for numerous products.37 The combination of these two phenomena suggests that setting minimum advertised prices may limit the reduction in prices associated with advertising. Furthermore, these minimum advertised prices are not an "essential" element of pass-through analysis for indirect purchasers. Rather, under the conditions described in the French Affidavit, the minimum advertised prices may be informative and suggestive of the conclusion that impact on indirect purchasers was common. However, the French Affidavit proposes econometric methods that explicitly rely on transaction or point-of-sale data--not on these minimum advertised prices--to establish common impact and damages for the proposed Class of indirect purchasers.
See, for example, Carlton, Dennis W. and Jeffrey M. Perloff. Modern Industrial Organization: Fourth Edition. Pearson: Addison-Wesley, Boston, 2005 (hereinafter, "Carlton and Perloff"), p. 474; Compeau, Larry D, Dhruv Grewal and Rajesh Chandrashekaran (2002). "Comparative Price Advertising: Believe It or Not," Journal of Consumer Affairs, Vol. 36, No. 2, pp. 284-294; Buehler, Stefan and Dennis L. Gärtner (2009). "Making Sense of Non-Binding Retail-Price Recommendations," SSRN Working Paper (available at http://ssrn.com/abstract=1351138).
For a brief survey of this research, see Carlton and Perloff, p. 481.
III. FEASIBLE METHODS TO ESTIMATE CLASS-WIDE DAMAGES ON A COMMON BASIS 31. The damage methodologies proposed in the French Affidavit are consistent with long-standing and conventional empirical methods that economists have used to compute damages on a class-wide basis in a litigation setting.38 Nevertheless, Defendant chooses to propound criticisms that are in turn inapplicable (that the proposed model fails to account for the prices of competing portable digital music players), irrelevant (that the impact period is unspecified) and untrue (that I failed to confirm the existence and format of third-party retail price data).
The French Affidavit Proposed Appropriate Benchmarks
1. The Yardstick Method Was Considered 32. Defendant claims that the damage methodology described in the French Affidavit fails to consider the price of competing portable digital music players as a potential benchmark.39 This claim reveals an incomplete reading of the French Affidavit, which specifically states, "The yardstick approach may compare iPod pricing in the United States with the pricing of another product (such as other MP3 players)...." (emphasis added).40 Defendant's criticism is therefore invalid.
Multiple regression analysis is a statistically tool widely used in litigation settings, especially in class action antitrust cases. See, for example, Rubinfeld, Daniel L. (2000). "Reference Guide on Multiple Regression," Reference Manual on Scientific Evidence, 2nd edition, Federal Judicial Center, pp. 179227; and Baker, Jonathan B. and Daniel L. Rubinfeld (1999). "Empirical Methods in Antitrust Litigation: Review and Critique," American Law and Economics Review, Vol. 1, pp. 386-435 at 391-393. Opp. Memo., pp. 23-24. French Aff. at 65.
2. Economic Analysis to Determine the Impact Period Is Routine 33. Defendant further criticizes the lack of specificity regarding the onset of the impact period of Apple's alleged violations.41 However, it is customary for econometricians constructing damage models to test for the beginning of the impact period by examining statistical evidence of the behavior of prices with guidance from textual evidence (such as that which further discovery may yield in this matter) about the timing of the alleged misconduct. As described in deposition testimony, possible start dates for the impact period include April 2003, when iTMS became available for Mac users and October 2003, when iTMS became available for Windows users. As I further testified, it is more relevant to determine at what point after iTMS was introduced was Apple able to achieve market power in downloadable digital music by establishing iTMS and seeing demand for downloadable digital music grow by a "quantum leap."42 This question can be rigorously addressed using regression analysis such as that described in the French Affidavit.43
Third-Party Retail Price Data to Perform the Proposed Analyses Are Available
The French Affidavit notes that the price data to construct and execute the proposed damage models included Apple's direct purchaser transaction prices for the iPod from October 2001 until the present and retail price data for the iPod, such as that
41 42 43
Opp. Memo., p. 17. French Dep. Tr., pp.80:7-85:7. French Aff. at 65-69.
collected by the NPD Group. Notwithstanding Defendant's accusation to the contrary, the availability of retail price data for MP3 players from the NPD Group was verified during several phone calls with NPD Group employees prior to the French Affidavit's filing, as described above. Defendant's claim that I failed to confirm the existence of iPod retail data is therefore inaccurate.
35. After considering the arguments presented in Defendant's Opposition Memorandum, I maintain the conclusions I reached in my initial report that: a. Proof of whether the proposed Classes have been economically injured by Apple's alleged misconduct would be common to Class members, and that such common proof would predominate over any individual questions; b. There are feasible methodologies and data to calculate aggregate damages to the Indirect Purchaser Damages Class on a common basis; and c. Economic analysis of liability issues would be common to all Class members.
Exhibit 1. Recent Advertised Retail Prices of Portable Digital Music Players by Apple's Top Nine Retailer Direct Purchasers[a]
Player iPod Classic - 120GB iPod Classic - 160GB iPod Nano - 8GB - 4th Gen iPod Nano - 16GB - 4th Gen iPod Touch - 8GB - 2nd Gen iPod Touch - 16GB - 1st Gen iPod Touch - 16GB - 2nd Gen iPod Touch - 32GB - 2nd Gen iPod Shuffle - 1GB iPod Shuffle - 2GB Best Buy Price Sale Price 249.99 149.99 199.99 229.99 279.99 299.99 399.99 49.99 69.99 229.99 Wal-Mart Price 247.88 147.88 197.88 227.88 297.88 397.88 48.72 68.72 Target Price 249.99 149.99 199.99 229.99 299.99 399.99 49.99 69.99 Circuit City Price Sale Price 249.99 299.99 227.99 149.99 199.99 299.99 299.99 399.99 146.99 226.99 296.99 Costco Wholesale Price Sale Price 239.99[b] 224.99[b] 139.99[b] 219.99[b] 289.99[b] 389.99[b] 133.99[b] 214.99[b] 279.99[b] RadioShack Price 249.99 279.97 149.99 199.99 229.99 299.99 399.99 49.99 Sam's Club Price 236.87 139.63 377.00 46.78 CompUSA Price 249.99[c] 139.99, 149.99[c] 199.99[c] 229.99[c] 299.99[c] 399.99[c] Amazon.com Price Sale Price 249.99 229.99 349.00 332.15 149.99 199.99 229.99 299.99 319.99 399.99 49.00 69.99 144.99 182.99 224.99 289.99 369.88 45.00 64.00
[a] Unless otherwise noted, data reflect online prices advertised on retailers' respective websites during December 2008. Prices of refurbished models and bundled products are not included. [b] Data reflect in-store prices posted on May 16, 2009 at Costco Wholesale location in Frederick, Maryland. [c] Data reflect online prices advertised on retailer's website during May 15, 2009. Prices of refurbished models and bundled products are not included. Sources: www.bestbuy.com, www.walmart.com, www.target.com, www.circuitcity.com, Costco Wholesale, www.radioshack.com, www.samsclub.com, www.compusa.com, www.amazon.com.
Millions of iPod Units
10 12 14 16 18 0 2 4 6 8
[a] Only 129 largest direct purchasers are shown. 1,054 direct purchasers buying fewer than 2,000 units each during April 2003 to March 2008 are not depicted. Source: Defendant's Responses to Plaintiff's First Set of Interrogatories, August 28, 2008, Attachment A.
Top 12 Direct Purchasers
Exhibit 2. Apple's Direct Sales of iPod Units to Reseller Direct Purchasers[a] April 2003 to March 2008
Best Buy Wal-Mart Target Corporation Circuit City Stores Inc Costco Wholesale Ingram Micro Inc RadioShack AF Services Sam's CompUSA Amazon.com Tech Data Product CDW Logistics Inc PC Connection Frys Electronics J and R Computer World B and H Photo Video Corp Fred Meyer Target.com P C Richard and Son Incentive Concepts Meijer Inc Sharper Image Corporation ABT Electronics And Appliances Micro Center Crutchfield Corporation Zones Inc The Mac Store New Zoom Inc Tweeter Home Entertainment Datavision Computer Video American TV and Appliance RC Willey Home Furnishings MacMall Retail Stores Nebraska Furniture Mart LA Computer Company Conn's Tekserve Corporation Sixth Avenue Electronics City Inc Audible Com US Merchants Bose Corporation Ultimate Electronics Good Guys Virgin Megastore RCS Experience Brandsmart USA Electronic Express Inc Costco Wholesale Corporation Robinson's May La Curacao Filene's Vanns Inc Bloomingdales Nike Retail Guitar Center Stores Inc Small Dog Electronics Ecost Com Unitek Computer Stores Fry's Electronics Hecht's Shopko Stores Inc HH Gregg Musicland Group Inc Video Hi-Tech Corp Foley ' s Powermax First Tech Computer Bell Microproducts Inc Mobile Planet Inc Borders Inc Pro Tape Systems Smart Machines Quad H Technology Corp AMA Systems Famous Barr Unique Photo Inc Soundtrack and Ultimate Business Agents Computer Supercenter Of Altitunes Partners LP Magnolia Audio Video Musical Fulfillment Services Boscovs Department Store Mac Accessory Center BDI Laguna Inc Simply Mac Inc Advance Computer Source Corp The Alaska MacStore Microwarehouse Inc Of Ohio Musicians Friend MacNet Computer Center Adorama Inc Mac Resource Flint Audio Video TD Curran Inc Macy's.com MacAuthority Foxwoods Resort Casino Marshall Fields Department Kompugard Inc Ultimate Electronics Inc D I P Sam Ash Music Corp Bloomingdales By Mail Electronics City Network Doctor Computer Advantage Inc Connecting Point Cmptr Samys Camera Power Mac Pac Sterling Technology Group Inc Advanced Micro Systs Inc Pro Video Sales Inc Connecting Pt Cmptr Ctr Net Systems Inc MacOutfitters Mac Made Easy Weingarten Gallery MacSuperstore Electronics Expo, LLC Bundy Computer Unitek Digital Expercom Lord & Taylor Studio D MacLife Springboard Media Three Star Tech Inc Impact Merchandising LLC
Disclaimer: Justia Dockets & Filings provides public litigation records from the federal appellate and district courts. These filings and docket sheets should not be considered findings of fact or liability, nor do they necessarily reflect the view of Justia.
Why Is My Information Online?