Opinion
MDL No. 00-1311
August 22, 2003
MEMORANDUM AND ORDER
Plaintiff West Bend Elevator, Inc. ("West Bend"), a Wisconsin corporation, brings this price-fixing lawsuit under Wisconsin's antitrust statute. The Court previously certified the lawsuit as a class action. Now before the Court is defendants' motion to decertify the class. After carefully considering the papers filed by the parties and all the pleadings in this action, and having had the benefit of oral argument, defendants' motion to decertify the class is GRANTED.
BACKGROUND
This lawsuit arises out of an alleged 15-year conspiracy by defendants to fix the price of methionine. Defendants sell synthetic methionine in variety of forms to direct purchasers, that is, those who purchase the methionine directly from defendants. West Bend is a Wisconsin grain elevator and feed mill and also operates a hog farm. During portions of the alleged conspiracy, it purchased methionine to resell and for its own use in its hog farm operation. West Bend filed this putative class action under Wisconsin's antitrust laws in Wisconsin state court on behalf of all indirect methionine purchasers; that is, those who purchased methionine from someone other than defendants.
Defendants successfully removed the action to Wisconsin federal court and the Multi-District Litigation Panel transferred the action to this Court. See 28 U.S.C. § 1407. The Multi-District Litigation ("MDL") matters that were pending before this Court included a consolidated nation-wide class action on behalf of direct purchasers of methionine, several actions by individual opt-out direct purchasers, and this matter. The Court certified the direct purchaser class action and in August 2002 approved a $107 million settlement. All of the direct purchaser opt-out actions settled earlier this year. West Bend is the only methionine price-fixing case that is still pending.
A. West Bend's First Motion to Certify the Class
The Court denied West Bend's initial motion for class certification on the ground that plaintiff had not come forward with a "colorable method" of proving antitrust impact, otherwise known as injury in fact, on a class-wide basis. In re Methionine Antitrust Litigation, 204 F.R.D. 161, 165 (N.D. Cal. 2001). In particular, the Court held that plaintiff's expert, John M. Connor, did not consider that the proposed class included both resellers and ultimate users and that the overpriced product — methionine — is incorporated into other products. Id. at 164-65.
B. West Bend's Second Motion to Certify the Class
West Bend subsequently amended its complaint to bring its claims on behalf of a much narrower class, namely, those who indirectly purchased methionine from defendants in Wisconsin for end use as an animal feed additive; in other words, only those Wisconsin indirect purchasers who did not resell the methionine. West Bend defined the proposed class as follows:
All persons and entities . . . who purchased methionine in the State of Wisconsin for end use as an animal feed additive (excluding pet feeds for dogs, cats, birds, and fish) indirectly from any of the defendants at any time during the period January 1, 1985 to the end of 1998.
After the parties engaged in further class certification discovery, West Bend moved for certification of this narrower class.
The motion's primary issue was again whether West Bend would be able to prove on a class-wide basis that the class members were injured by the alleged price-fixing conspiracy. To prove that each class member was actually injured, West Bend must demonstrate that the "overcharge" was passed on to each class member, that is, that the methionine reseller who purchased methionine (either directly or indirectly) from defendants paid an inflated price for the methionine and then "passed on" that overcharge to the class member. The Court assumed, for the purpose of the motion, that West Bend could prove the direct purchasers were overcharged for methionine, as well the approximate amount of the overcharge over time. The issue presented was whether West Bend could prove on a class-wide basis that the overcharge was passed on through the various distribution channels to the indirect purchaser end users.
In support of its renewed motion to certify, West Bend relied on a new expert, Dr. Keith Leffler. Dr. Leffler's theory of antitrust impact was, in brief, that given the longevity of the conspiracy, it is common sense that the methionine end users were injured. He therefore proposed comparing the price of methionine products before the conspiracy (pre-1985) to shortly after the conspiracy began to measure the impact of the conspiracy. Any end user during the 15-year class period would be impacted by the conspiracy regardless of whether at any given time a reseller might raise the price of the methionine-containing product for some reason rather than an overcharge; the price paid by the end user would still be higher than it would have been absent the conspiracy because the whole price structure would have been raised because of the conspiracy.
Dr. Leffler proposed to calculate the overcharge pass-on rate for the class using multiple regression analysis. He testified that he would start with the prices paid by end users for methionine-containing products, "regress out" most of the independent factors that might have influenced the price to end users (labor costs, price of other ingredients, etc.), and be left with how much of an overcharge was passed on to end users.
Defendants raised several challenges to West Bend's proposed methodology. With respect to whether the amount of the overcharge could be determined on a class-wide basis, they argued that the data Dr. Leffler proposed using would necessarily be inaccurate and therefore the amount of the overcharge would necessarily be inaccurate. The Court rejected this argument, noting that many courts have accepted the use of multiple regression analysis and that there is no requirement that a particular method result in a wholly accurate number, and, in any event, the accuracy of plaintiff's figure would be related to the sample size input into Dr. Leffler's formula; the larger the sample the more accurate, as an average, would be the resulting number. Sep. 20, 2002 Memorandum and Order at 7.
Defendants also argued that the data necessary for performing Dr. Leffler's proposed multiple regression analysis was not available. The Court held that it was premature to deny West Bend's motion on that ground. Dr. Leffler had stated in his declaration that sufficient data could be collected and the Court could not find, on the record before it at the time, that it could not. Id.
Defendants argued further that the price of methionine in the animal feed sold to end users is so minuscule compared to the other feed ingredients and the product as a whole that Dr. Leffler could not compute a reasonably accurate overcharge rate on a class-wide basis. Dr. Leffler rebutted this argument by noting that there was no reason to believe that the pass-on rate was any different for methionine than for the other ingredients and therefore the rate could be determined by calculating the rate for other ingredients which consist of a larger percentage of the price of the final product Id. at 7-8.
Since West Bend's expert had proposed a "colorable" method of proving the amount of the illegal overcharge on a class-wide basis, the Court concluded that questions common to the class predominated and that class action status was otherwise appropriate. Id. At 11-12.
D. Defendants' Motion to Decertify the Class
As the MDL pretrial proceedings have come to a close, including expert discovery, defendants now move to decertify the class. They argue that Dr. Leffler has not performed any of the data gathering or analysis that was necessary to make his theory of proving the amount of the overcharge on a class-wide basis "colorable."
DISCUSSION
Once a district court certifies a class, it continues to possess the authority and discretion to determine whether the certification should be amended or vacated in its entirety. See Fed.R.Civ.P. 23(c)(4)(B);Armstrong v. Davis, 275 F.3d 849, 872 n. 28 (9th Cir. 2001). As courts have noted,
"[o]nce a class is certified, the parties can be expected to rely on it and conduct discovery, prepare for trial, and engage in settlement discussions on the assumption that in the normal course of events it will not be altered except for good cause. Sometimes, however, developments in the litigation, such as the discovery of new facts or changes in the parties or in the substantive or procedural law, will necessitate reconsideration of the earlier order and the granting or denial of certification or redefinition of the class."O'Connor v. Boeing North American, Inc., 197 F.R.D. 404, 409-10 (C.D. Cal. 2000), quoting Cook v. Rockwell Int'l Corp., 181 F.R.D. 473, 477 (D. Colo. 1998) (internal quotation and citation omitted). While plaintiff carried the initial burden of proving that class certification was appropriate, defendants shoulder the burden of proving that certification is no longer warranted. It is worth noting that plaintiff has not sent notice to the class, even though the Court certified this class action one year ago.
Defendants argue that changes in the litigation, namely, West Bend's abandonment of its proposed methodology for determining the amount of the overcharge on a class-wide basis, warrant vacating the order certifying the class. See Monaco v. Stone, 187 F.R.D. 50, 59 (E.D.N.Y. 1999) ("A class may be decertified if later events demonstrate that the reasons for granting class certification no longer exist or never existed."); Cook, 181 F.R.D. at 478 ("Decertification is warranted where materially changed or clarified circumstances have been shown that would make the continuation of the class action improper.").
The Court agrees that West Bend no longer proposes to prove the amount of the overcharge by the methodology Dr. Leffler originally proposed. Dr. Leffler did not perform a multiple regression analysis as he promised he would do. He did not gather any of the data which he promised he would do. And he did not examine any of the multiple variables that could affect the price of methionine in the products sold to end users. He did not even analyze the price of other ingredients in the animals feeds as a way of determining the pass-on rate for methionine. In short, he did none of the proposed analysis upon which the Court relied in granting West Bend's motion for class certification.
The question, then, is whether what Dr. Leffler did do is sufficient to support a finding that the class was damaged by the price-fixing conspiracy and is sufficient to support a finding on a class wide basis of the amount or amounts of the overcharge. If it is not, common questions of law or fact may not predominate, making maintenance of the class inappropriate. See Seligson v. Plum Tree, Inc. 61 F.R.D. 343, 344 (E.D. Pa. 1973) (stating that decertification is warranted where individual issues of proof predominate).
Dr. Leffler performed seven simply (rather than multiple) regressions. He concludes from those regressions that the pass-on rate for the entire class for the entire 15-year damage period is the unweighted average of five of the regressions, or, at a minimum, the lowest rate of the same five regressions. He multiplies those pass-on rates by the amount of the overcharge and by the estimated volume of methionine purchased by end users to calculate damage to the class as approximately $4 million to $6.5 million, depending on which pass-on rate is used.
Dr. Leffler's new — and much simplified — method of computing the extent of the antitrust injury is unreasonable for several reasons. First, the simple regression analysis does not take into account any of the myriad variables that could account for the price of the final product to the end user. For example, in one regression, he determined a pass-on rate by taking the average price of methionine and comparing it to the price of "FM Poultry Mix" sold to a single end user. The FM Poultry Mix is composed of only eight percent methionine. Although the relationship between the price of methionine to direct purchasers and the price of FM Poultry Mix to the end user will depend on many variables, including labor, transportation costs, and most importantly, changes in the price of the other 92 percent of the Mix's ingredients, Dr. Leffler's regression does not take into account any of these factors. He fails to do so even though the whole premise behind his original proposed methodology was that such factors had to be considered. See Declaration of Keith Leffler (dated February 14, 2002) at ¶ 18 (stating that to determine the pass-on rate he would use multiple regression analysis "since a proper estimate of the pass-on rate must control for the influence of other ingredient price changes or other cost changes) (emphasis added); id. at ¶ 20 (stating that the variables he intends to study in his regression analysis include "indices of labor costs over time, a variable for whether delivery is included, a variable measuring whether the payment terms vary among customers or over time, and a variable to account for changes in the economic conditions impacting livestock industries").
Second, the seven regressions are not representative samples of the Wisconsin methionine end user market. None of the samples involves finished feed product Three of the regressions involve sales to a single end user. Only three of the regressions actually link manufacturers' prices to prices paid by end users, and these regressions analyze products with a high percentage of methionine; they are not representative of the large volume of products sold to end users. All of the data involve only two intermediaries, and only one intermediary is involved in the three regressions that actually link manufacturers' prices with prices to end users.
Third, Dr. Leffler's regressions analyze data from an extremely limited time period. The damages period is 1984 through 1999, yet none of the regressions includes data from before 1995, and four of them do not include any data from before late 1996. Some of Dr. Leffler's regressions even include data from outside the damages period. Based on this data, Dr. Leffler opines as to the pass-on rate for the entire 15-year conspiracy. He has not offered any rational explanation as to how such analysis is valid.
Fourth, Dr. Leffler's methodology fails to weigh the various regressions. He does not account for the disparity in volume of methionine sold from month-to-month in any given regression, and he fails to take into account the different data involved in each regression. He simply gives each regression (other than the two he ignores) equal weight in computing an average pass-on rate. Nor does he take into account the different products involved in each regression
In sum, if West Bend had moved for class certification on the basis of the methodology Dr. Leffler ultimately employed, the Court would have denied plaintiff's motion. Dr. Leffler's method is "so insubstantial as to amount to no method at all." In re Potash Antitrust Litig.., 159 F.R.D. 682, 697 (D. Minn. 1995). As the record in these consolidated cases amply demonstrates, the synthetic methionine industry in the United States is complex. It involves multiple sellers, a myriad of distribution channels, and hundreds of different products. The complexity is increased when one considers the impact to end users, many of whom purchased products containing very little methionine. Dr. Leffler's original proposed methodology recognized and addressed this complexity. The methodology he ultimately employed, however, does not take any of it into account. Accordingly, West Bend does not have "colorable" method for proving antitrust injury, and the extent of that injury, on a class-wide basis.
Since the extent of the antitrust impact will have to be determined individually, the Court concludes that individual issues predominate and this indirect purchaser price-fixing case is not appropriate for resolution as a class action.
CONCLUSION
For the foregoing reasons, the Court GRANTS defendants' motion to decertify the class. Since West Bend never sent notice to prospective class members, the Court need not send notice of the dissolution to the prospective class.
IT IS SO ORDERED.