United States District Court, S.D. Florida.
*1336 Eugene Stearns, Miami, FL, Sidney Pertnoy, Gerald Bowen, McLean, VA, for plaintiffs.
Larry Stewart, Miami, FL, Robert Abrams, Robert Brookheiser, Stuart Harris, Darren B. Bernhard, Robert Wallis, Exxon Company, U.S.A., Houston, TX, for defendant.
GOLD, District Judge.
Exxon dealers [i.e. direct served branded dealers] have filed this class action against Exxon Corporation based on a claimed breach of Exxon's contractual obligation to its dealers, and the dealer class as a whole, to charge open wholesale gasoline prices in good faith after it implemented its Discount for Cash ("DFC") Program in 1982. By prior orders, the court has addressed various aspects of the controversy.[1] The current matter arises from numerous motions which essentially seek to exclude expert testimony on damages and related matters. To address these motions, the court has conducted an extensive Daubert[2] hearing over six days.[3] At the Daubert hearing; the court heard the direct testimony and cross-examination of the parties' expert witnesses. The Daubert hearing was the last step in a lengthy process established by prior court orders. The process sought to assure that each party would have a complete opportunity to establish, or not, the reliability of its expert witness' opinions on damages and related matters, as well as the lack of reliability of the opposing expert's opinions.
By order dated March 25, 1999, the parties were permitted until April 15, 1999 to file their expert reports. On that date, the Plaintiffs filed the expert report of Dr. Raymond P.J. Fishe. Exxon filed the expert report of Dr. Joseph P. Kalt. The parties then were permitted until May 3, 1999 to complete their final depositions of the opposing party's expert. Thereafter, each party was permitted to challenge the opposing party's expert by Daubert motions explaining: (i) why the expert is not qualified to testify competently regarding the matters he or she intends to address; (ii) which aspects of the expert's methodology is claimed to be unreliable as determined by the sort of inquiry mandated in Daubert, (iii) and why the testimony will not assist the trier of fact to understand the evidence or to determine a fact in issue. See Order on Status Conference, dated March 9, 1999 [D.E. # 972]. The responding party was then permitted to reply with its own memorandum and affidavits.
Upon the completion of the depositions, Exxon moved to exclude the testimony of Dr. Raymond Fishe [D.E. # 1020]. In support, it filed an accompanying affidavit of Dr. Joseph Kalt, dated June 6, 1999. Exxon also filed a motion to preclude expert testimony on the amount and basis for damages [D.E. # 1028], and to strike Paragraph 5 of the May 17, 1999 affidavit of Dr. Raymond Fishe [D.E. # 1042].
In response, the Plaintiffs filed their own affidavits. The first affidavit was of Dr. Raymond Fishe, dated May 17, 1999. The second affidavit was of Dr. William D. Nordhaus, dated May 17, 1999. Exxon moved to dismiss the Nordhaus affidavit, but the court previously denied its motion by order dated July 1, 1999. [D.E. # 1125].
Plaintiffs then responded to Exxon's motion to exclude Dr. Fishe's expert testimony with a memorandum of law and a new 50 page affidavit of Dr. Fishe. This affidavit contained a number of new charts and exhibits to further explain or justify Dr. Fishe's April 15th report. It also responded to various attacks made by Exxon during Dr. Fishe's deposition, and by Dr. Kalt in his affidavit of June 6, 1999. Exxon then moved to preclude the testimony by Plaintiffs' expert on analysis not contained in his final report [D.E. # 1091]. Finally, Plaintiffs filed their motion to exclude or limit the testimony of Dr. Kalt [D.E. # 1018].[4] For reasons stated below, the parties' motions are denied.
Federal Rule of Evidence 702,[5] as explained by the United States Supreme Court in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589, 113 S. Ct. 2786, 2794-95, 125 L. Ed. 2d 469 (1993), and in Kumho Tire Co. v. Carmichael, 526 U.S. 137, 119 S. Ct. 1167, 143 L. Ed. 2d 238 (1999), controls determinations regarding the admissibility of expert testimony. Expert testimony may be admitted into evidence if: (1) the expert is qualified to testify competently regarding the matters he intends to address; (2) the methodology by which the expert reaches his conclusions is sufficiently reliable as determined by the sort of inquiry mandated in Daubert; and (3) the testimony assists the trier of fact, through the application of scientific, technical, or specialized *1338 expertise, to understand the evidence or to determine a fact in issue. See Fed. R.Evid. 702; Daubert, 113 S.Ct. at 2794 (holding that "under the [Federal] Rules the trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable"). Here, the court concludes that: (1) both experts are qualified to testify, and (2) that their testimony, through their specialized knowledge, would assist the trier of fact to determine whether the Plaintiffs are entitled to damages.[6] The sole remaining issue for determination is the second inquiry; that is, whether the methodologies by which Dr. Fishe and Dr. Kalt reached their respective conclusions are sufficiently reliable for consideration by the jury. Each party bears the burden of proof to demonstrate reliability by the preponderance of the evidence. Daubert, 113 S.Ct. at 2796 n. 10. As discussed more fully below, the reliability assessment entails preliminary consideration as to whether each expert's reasoning and methodology are sufficiently valid so as to be properly applied to the facts in issue. Id. at 113 S. Ct. at 2797.
The Daubert Court noted that "[t]he inquiry envisioned by Rule 702 is, we emphasize, a flexible one." Daubert, 113 S.Ct. at 2797. Many factors will "bear on the inquiry [into whether an expert's reasoning or methodology is reliable], and we do not presume to set out a definitive checklist or test." Daubert, 113 S.Ct. at 2796. Nevertheless, the Court did identify certain factors that may be pertinent to such an inquiry. These factors include: "whether [the theory or technique at issue] can be (and has been) tested"; whether it "has been subjected to peer review and publication"; the "known or potential rate of error" of the technique, as well as the "existence and maintenance of standards controlling [its] operation"; and the degree to which the relevant scientific community accepts the theory or technique as reliable. Daubert, 113 S.Ct. at 2796-97. The district judge is assigned the task "... of ensuring that an expert's testimony both rests on a reliable foundation and is relevant to the task at hand." Daubert, 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469.
In Kumho, the Supreme Court held that Daubert's "gate-keeping" obligation, requiring the trial judge's inquiry into both the expert's relevance and reliability, applies not only to "scientific" testimony, but to all expert testimony. Kumho, 526 U.S. 137, ___, 119 S. Ct. 1167, 1174, 143 L. Ed. 2d 238; see also United States v. Paul, 175 F.3d 906, 910 (11th Cir.1999). The Supreme Court further noted that Federal Rules of Evidence 702 and 703 give all expert witnesses testimonial leeway unavailable to other witnesses on the presumption that the expert's opinion "will have a reliable basis in the knowledge and experience of his discipline." Kumho, 526 U.S. at ___, 119 S.Ct. at 1174 (citing Daubert, 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469). Moreover, the Court held that a trial judge may consider one or more of the specific Daubert factors when doing so will help determine that expert's reliability. Kumho, 526 U.S. at ___, 119 S.Ct. at 1175. But, as the Court stated in Daubert, the test of reliability is a "flexible" one, and Daubert's list of specific factors neither necessarily nor solely applies to all experts or in every case. Kumho, 526 U.S. at ___, 119 S.Ct. at 1175 (citing Daubert, 113 S.Ct. at 2786).[7] Alternatively, *1339 Kumho provides that "the law grants a district court the same broad latitude when it decides how to determine reliability as it enjoys in respect to its ultimate reliability determination." Kumho, 526 U.S. at ___, 119 S.Ct. at 1171 (citing General Electric Co. v. Joiner, 522 U.S. 136, 143, 118 S. Ct. 512, 139 L. Ed. 2d 508 (1997)).
Thus, the district court is required to exercise a special "gate-keeping" function to ensure that an opinion offered by an expert is reliable. Kumho, 526 U.S. at ___, 119 S.Ct. at 1176. Indeed, "where [expert] testimony's factual basis, data principles, methods, or their application are called sufficiently into question ... the trial judge must determine whether the testimony has a `reliable basis' in the knowledge and experience of [the relevant] discipline." Id. at 1175[8] (emphasis added). The "`gate-keeping' inquiry must be tied to the facts of a particular case." Id. Moreover, the testimony must be helpful or "fit" with the issues to be resolved in the case; that is, the district judge must also determine whether the expert's reasoning and methodology can be properly applied to the facts in issue. Daubert, 113 S.Ct. at 2796.
The focus is not on the conclusions generated by the expert's methodology, but on the reasonableness of the expert's use of such an approach, together with his or her particular method of analyzing the data obtained, to draw a conclusion regarding the specific matter to which the expert testimony is directly relevant. Kumho, 526 U.S. at ___, 119 S.Ct. at 1177. The "overarching" goal of Daubert's gate-keeping requirement "... is to make certain that an expert, whether basing testimony *1340 upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field." Kumho, 526 U.S. at ___, 119 S.Ct. at 1175. The expert's "self-proclaimed" accuracy is insufficient. "Nothing in either Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert." Kumho, 526 U.S. at ___, 119 S.Ct. at 1179 (citing General Electric Co. v. Joiner, 118 S.Ct. at 519). Instead, in terms of reliability, it is not the general acceptance of the methodology that is relevant. "[r]ather, it was the reasonableness of using such an approach, along with [the expert's] particular method of analyzing the data obtained, to draw a conclusion regarding the particular matter to which the expert testimony was directly relevant." Kumho, 526 U.S. at ___, 119 S.Ct. at 1177 (emphasis added).
In Kumho, the district judge had excluded the expert's testimony because of doubts about reliability. Despite the expert's qualifications, the district judge determined that the expert's testimony "... fell outside the range where experts might reasonably differ, and where the jury must decide among the conflicting views of different experts, even though the evidence is `shaky.'" Kumho, 526 U.S. at ___, 119 S.Ct. at 1177 (citing Daubert, 509 U.S. at 596, 113 S. Ct. 2786) (emphasis added). It is also important that "the court, after looking for a defense of [the expert's] methodology as applied in these circumstances found no convincing defense. Rather, the district judge found (1) that none of the Daubert factors, including that of `general acceptance' in the relevant expert community, indicated the [the expert's] testimony was reliable ... (2) that his own analysis `revealed no countervailing factors operating in favor of admissibility which could outweigh those identified in Daubert ... and (3) the `parties identified no such factors in their briefs.'" Kumho, 526 U.S. at ___, 119 S.Ct. at 1178 (emphasis added).
In applying the Daubert/Kumho tests, a district judge should be conscious of two guiding, and sometimes competing, principles. Westberry v. Gislaved Gummi AB, 178 F.3d 257, 261 (4th Cir.1999). On one hand, the court should be mindful that Rule 702 was intended to liberalize the introduction of relevant expert evidence. Id.; see also Cavallo v. Star Enter., 100 F.3d 1150, 1158-59 (4th Cir.1996). In this regard, the court need not determine that the expert testimony a litigant seeks to offer into evidence is irrefutable or certainly correct. Id. As with all other admissible evidence, expert testimony is subject to being tested by "[v]igorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof." Id.citing Daubert, 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469. Moreover, "as circumstantial evidence, the [expert's] data and testimony need not prove the plaintiff's case by themselves; they must merely constitute one piece of the puzzle that the plaintiffs endeavor to assemble before the jury." City of Tuscaloosa, 158 F.3d at 565.
On the other hand, the court must recognize that due to the difficulty of evaluating their testimony, expert witnesses have the potential to "be both powerful and quite misleading." Daubert, 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469. Therefore, given the potential persuasiveness of expert testimony, proffered evidence that has a greater potential to mislead than to enlighten should be excluded. See United States v. Dorsey, 45 F.3d 809, 815-16 (4th Cir.1995).
While the above statement of the law appears clear enough, its application to the case of "dueling, but well-qualified experts", each claiming that the other's methodology and use of data is "unscientific," is often troublesome, particularly when the disciplines involved are well-established, the general methodologies are well-known, and where there is no apparent *1341 "smoking gun" that can be characterized as "junk."[9] Merely because two qualified experts reach directly opposite conclusions using similar, if not identical, data bases, or disagree over which data to use or the manner in which the data should be evaluated, does not necessarily mean that, under Daubert, one opinion is per se unreliable. Daubert does not empower the district judge to simply "pick" one expert over the other, because that expert is more credible or convincing, under the guise of exercising the gate-keeping function. To do so would improperly usurp the jury's function. Circumstances may well dictate that both opposing experts' methodologies are sufficiently reliable to undergo further jury scrutiny.[10] Merely because it may be hard for the jury to understand the competing methodologies does not mean that the right to jury trial should be denied and the matter decide de facto by the court.[11]
Lacking advice from its own court-appointed expert,[12] how should the district *1342 judge balance and apply these countervailing factors to the "dueling expert" situation in light of the Daubert/Kumho gate-keeping requirements? Otherwise stated, in a case such as this, where striking the expert's testimony would effectively end the case for the affected party, where is the balancing point after the court considers the relevant factors?
Daubert/Kumho offer no brightline answer on where to strike the balance. But several guidelines are evident. Once the district judge considers the relevant Daubert factors, the court may strike that expert's opinion, data and methodology when the methodology or data used makes it more likely than not that (1) the research method was not a reliable indicator of the results to which the expert would testify; (2) where the opinion falls outside the range where experts might reasonably differ [and the jury should decide among the competing points of view of the differing experts]; and (3) where there are no convincing, countervailing "more flexible" factors which operate in favor of admission. While this list is not meant to be exhaustive, it does provide a benchmark on how to implement and decide what has become a complex mini-trial process.[13]
The central damage issue in this case is whether, following the imposition of Exxon's 3% credit card recovery fee ("CCR fee") in August, 1982, Exxon reduced its wholesale prices for motor fuel by an amount, which, on average, across all of its markets, over course of the Discount for Cash Program, offset the CCR fee collected from its dealers? The time period for damages claimed by the Plaintiffs is from March 1983 through August 1994. During this time period, Plaintiffs claim that Exxon "took back" any reduction initially given, such that the entire CCR fee subsequently charged are its damages. In Plaintiffs view, the offset was entirely gone by March 1, 1983.
Without dispute, the initial DFC adjustment factor was 1.7 cpg throughout the United States.[14] After that, Exxon adjusted the DFC offset factor three times. First, in 1986, Exxon changed the adjustment factor to 1.0 cpg. Second, in 1991, Exxon again changed the adjustment factor to 1.5 cpg. Then, in 1992, Exxon adjusted it to 1.3 cpg.
Assuming these adjustments were actually given during the entire 12 years of DFC, the total wholesale price adjustments to Exxon dealers would have exceeded the CCR fees collected by Exxon by approximately 43 million dollars. Under these circumstances, Plaintiffs would not be entitled to damages. Thus, the ultimate question for the jury is whether Exxon initially gave, and then "took back" the adjustment in total as of March 1, 1983 through August, 1994?[15] In essence, this is an "all or nothing" case.[16] Either Exxon *1343 took back the DFC adjustments entirely or it did not.
To explain Exxon's motivation, Plaintiffs opine that Exxon "took back" the offset to increase its profits. These profits, in turn, were invested by Exxon as capital expenditures to modernize certain of its dealers [the "keepers"] at the expense of other, less profitable dealers [the "non-keepers"]. Exxon identified the "keepers" as those stations which could best compete with other major dealers in changing markets. Plaintiffs characterize the economic conditions surrounding Exxon's exercise of its pricing discretion during the period of August 1983 to August 1994 as "not normal" such as to provide Exxon with the incentive to fail to provide the offset in wholesale prices.
The ultimate "take back" issue cannot directly be determined from Exxon's business records. Exxon did not establish or use a "mechanistic" or "formulaic" approach to automatically make adjustments throughout the DFC program. At best, Exxon's business records[17] provided information from which its managers could make wholesale pricing decisions over its many markets. Therefore, whether Exxon "took back" its wholesale price adjustment or not is a matter, in part,[18] of expert opinion based on extrapolation from various data bases. As noted in City of Tuscaloosa, 158 F.3d at 565, it is one piece, although here, an important piece, of the *1344 puzzle that the Plaintiffs [and Exxon] endeavor to assemble before the jury.
In analyzing whether Exxon adjusted its wholesale prices, on average, over its markets, to offset the CCR fees collected from its dealers over the twelve years of the DFC program, both Dr. Fishe and Dr. Kalt start with Exxon's own explanation of what it did. This explanation is set out in what has been referred to as "Exxon's Answer to Interrogatory # 35." In this interrogatory, Exxon states that it utilized a pricing methodology that provided a wholesale price reduction to its dealers in an amount which, on average, offset the charge for credit to its dealers.
At the outset of its Discount for Cash marketing program in 1982, Exxon reduced its prices 1.7 cpg throughout the United States. Thereafter, in order to continue to provide the wholesale gasoline price reduction and remain a cash seller, Exxon, as a regular part of its wholesale pricing methodology, took the prices of its credit competitors and subtracted from them the amount that Exxon determined would offset the revenues it received from the 3% credit card fee. Initially, this amount was 1.7 cpg ... As discussed elsewhere, Exxon also adjusted the wholesale prices of its competitors for other marketing programs. Exxon then calculated an average net cash price of its key competitors in each market. Exxon used that average net cash price as a target at which to set Exxon's price.
In simple terms, before DFC, Exxon was including in its wholesale price to its dealers a factor for credit card processing. As such, it was a "credit based" major. Other majors which separately charged a credit card processing fee to its dealers [i.e., a non-bundled CCR fee] were known as "cash based majors." In its pricing system, known as the "Wholesale Price Development System" ("WPDS"),[19] Exxon adjusted its key competitor's wholesale prices in WPDS up by 1.7 cpg. After DFC, Exxon stopped adjusting the prices of its cash competitors up and began adjusting the prices of its credit competitors down in WPDS by 1.7 cpg. According to Exxon, it then calculated the average net wholesale price of its key competitors in each market. It then used the average net price amount as a target in setting its prices. Exxon's primary pricing parameter stated that it would seek to set its dealer wholesale prices competitive with the average of the net prices of its key competitors. According to Exxon, "because [it] set its prices on a net basis after subtracting the amount of the DFC adjustment (i.e. 1.7 cpg) from credit competitors prices, Exxon ensured that it priced in the cash market and that its dealers received the wholesale price reduction." Exxon Interrogatory # 35, page 10. Furthermore, *1345 Exxon said that its pricing system was not "formulaic," meaning that its wholesale prices may be above or below the average net wholesale prices of its key competitors on any given day.
Dr. Fishe reached three ultimate opinions which are summarized as follows:
Immediately following the imposition of the 3% credit cost recovery fee in August, 1982, Exxon reduced its wholesale prices for motor fuel by an amount which, after several months, was sufficient to provide a full, on average, offset to Exxon's 3% CCR fee. It took several months to achieve the full offset because of competitive responses at wholesale and retail and because Exxon introduced the CCR fee in the context of a falling market (i.e., all fuel prices were failing). Sometime between January 1983 and March, 1983, Exxon raised its wholesale prices to take away the offset and never provided it again for the duration of the DFC program. During the period from March 1983 until August 1994, Exxon did not offset the 3% fee, on average, with wholesale price reductions. The effect was that Exxon, on average, overcharged its dealers on a cents-per-gallon basis for motor fuel by an amount determined by dividing the total revenues collected through the 3% fee by the total volume of motor fuel sold to the dealers.
Exxon's contention that it complied with the obligation to provide the offset through the DFC adjustment in its price monitor system ("WDPS") is not support by further analysis.
The economic conditions which surrounded Exxon's exercise of pricing discretion during the period of August 1983 until August 1994, were not normal and provided Exxon an incentive to fail to provide the offset in wholesale prices.[20]
As noted in City of Tuscaloosa, 158 F.3d at 566 n. 25, the proper inquiry regarding the reliability of the methodologies implemented by economic and statistical experts in this context is not whether other experts, faced with substantially similar facts, have repeatedly reached the same conclusions (because there will be few or no cases that have presented substantially similar facts). Instead, the proper inquiry is whether the techniques utilized by Dr. Fishe (and Dr. Kalt) are reliable in light of the factors (other than testibility) identified in Daubert and in light of other factors bearing on the reliability of the methodologies. Id. In this case, the two principal methodologies used to verify compliance are: (1) an analysis of the retail margins of Exxon dealers, and (2) a comparison of Exxon's wholesale prices to the net average wholesale prices of its key competitors.
Dr. Fishe's opinion that Exxon failed to provide the DFC offset is based on the decline in Exxon dealer margins versus the margins of Exxon's competitors shortly after DFC was introduced. Exxon itself, in its answer to Interrogatory # 35, recognized that Exxon's wholesale price reduction to its dealers can be "objectively" verified by margin analysis.[21] Dr. Kalt evidently concurred. In an early affidavit (which Dr. Kalt now suggests was offered in a different context), he stated that if the alleged overcharge was borne by the Plaintiffs, "... the existence ... would be reflected in reduced dealer profitability, resulting from depressed margins and/or lost volume." Affidavit of Dr. Joseph Kalt, dated November 25, 1994.[22] Moreover, Exxon documents offered at the Daubert hearing show that Exxon managers did their own margin analysis to measure the effects of the DFC program on Exxon, its dealers, and on its dealers' key competitors. In effect, Dr. Fishe's margin analysis both utilized, and expanded upon, Exxon's own analyses, with results on a nationwide basis that were not significantly dissimilar.[23] Thus, while the margin *1347 analysis used would not ordinarily be the subject of peer review, publication or general acceptance in the scientific field, its use in the context of this case is appropriate in terms of general methodology. Whether margins increased, decreased, or stayed the same versus the margins of other majors before and after introduction of DFC may be measured objectively with data maintained by Exxon during the ordinary course of business.
Of course, merely using an appropriate methodology does not ensure reliability in terms of its application. In fact, a considerable part of Exxon's challenge is to its application based on available data. Before discussing the approach and the challenge, however, it is important to note that both experts employed regression analysis as part of their methodology, although again, there is a dispute over whether the specific analysis used was appropriate to the data.
Generally, econometric and regression analyses are considered reliable disciplines. City of Tuscaloosa, 158 F.3d at 566 (expert's compilation of data from business records and analysis with simple arithmetic, algebra, and multiple regression analysis is well-established as a reliable methodology); Askew v. City of Rome, 127 F.3d 1355, 1365 n. 2 (11th Cir. 1997) (in voting rights case, district court admitted expert statistical testimony based on part on multiple regression analysis); Petruzzi's IGA Supermarkets v. Darling-Delaware Co., 998 F.2d 1224, 1238 (3d Cir.) (finding use of multiple regression analysis reliable under Rule 702); see also, Daniel L. Rubinfield, Econometrics in the Courtroom, 85 COL.L.REV. 1084, (1985). Regression and statistical analysis also have been admitted in antitrust cases to prove injury and to determine damages. State of Colorado v. Goodell Bros. Inc., 1987 WL 6771 (D.Colo. Feb.17, 1987) (admitting one and excluding one of expert's econometric models estimating damages).
The question at issue is whether available wholesale price comparisons establish that, on average, over the DFC period, across all grades of gasoline, and across all of the markets in which Exxon operates, Exxon's wholesale prices were below [or above] the average of its cash-basis competitors (i.e. prices charged by competitors that reflect the net prices paid by dealers when not utilizing credit card services). Each expert reaches directly opposite conclusions.
From Dr. Fishe's standpoint, Exxon's internal business records contained errors and omissions in the process by which Exxon attempted to adjust competitors' posted wholesale prices to net prices. In his view, documented errors include relative rent adjustments, credit cost adjustments, and the improper categorization of one of Exxon's significant competitors as one which had offset credit cost recovery fees when Exxon knew that it had not. According to Dr. Fishe, Exxon's records also established that Exxon recognized brand value as an essential factor in motor fuel pricing and that Exxon's brand value varied significantly over markets and across time but that Exxon did not include brand value as one of the adjustments in its WPDS. Finally, he contends that Exxon's wholesale price prior to DFC was consistently reported by Exxon's pricing system to be lower than the wholesale price of its other major competitors. He concludes that during the pre-DFC period, *1348 when Exxon attempted to maintain prices which appeared to be equal to its competitors, it lost volume and market shares. This failure to take the "initial condition" into account through an adjustment in WPDS, in his view, created a bias in the comparison of wholesale prices after DFC. Thus, his premise is that because WPDS did not report accurate adjusted net prices which reflected the absolute relationship between the wholesale prices of Exxon and its competitors (and because Exxon destroyed the before and immediately after key competitor wholesale price data), comparison of the net wholesale prices does not provide an objective basis to measure whether Exxon provided the DFC offset.
To determine whether Exxon actively managed the WPDS and set prices by targeting the average of the reported adjusted net wholesale prices of its key competitors so as to provide the offset over time, Dr. Fishe applied a "runs" test—a standard statistical technique—tohe delta between Exxon's wholesale price and the average reported adjusted net price of its key competitors.[24] According to Dr. Fishe, the small number of "runs" in the deltas (evidencing lengthy periods of time in which Exxon was above or below the average adjusted net wholesale price of its key competitors) indicated that Exxon was not managing WPDS wholesale prices so as to provide the wholesale price offset.
Dr. Fishe also concluded that the absence of key competitor data for the critical periods pre-dating and post-dating the introduction of DFC called into question the use of WPDS to establish compliance. The earliest WPDS electronic data for this time period was missing, as was the hard copy, until January of 1983. In Dr. Fishe's view, the absence of such data during this period precludes any scientific analysis by Exxon of whether it provided the wholesale price offset in accordance with the WPDS pricing methodology as presented in Interrogatory # 35. Accordingly, by using Exxon's internal business records as the source data, Dr. Fishe purported to "correct" the so-called WPDS errors on a cents-per-gallon basis by accounting for brand value studies, reclassifying Texaco as a credit price major, correcting for Shells' rents, and by making adjustments for competitors' proprietary credit card costs. He calculated the extent of the claimed errors and omissions at a positive 1.37 cents per gallon. He applied regression analysis to control for errors in the WPDS wholesale price delta series. He concluded, based on these statistical tests, that the biases in the system made Exxon's prices appear to be lower than they actually were relative to the prices of competitors.
The economic and statistical tests applied by Dr. Fishe have standard acceptance in the relevant scientific community. He applied a standard regression model to control for anomalies in his corrected WPDS time series analysis. His "before and after" test and his "initial conditions" analysis are accepted economic analyses used to evaluate causes and effects associated with the introduction of new or extraneous market events or conditions. See In re Industrial Silicon Antitrust Litigation, 1998 WL 1031507 (W.D.Pa.1998) (in price-fixing suit, before and after models of pricing relationships satisfies Daubert requirements as analysis is generally accepted in field of economics). Dr. William Nordhaus, Plaintiffs' rebuttal witness, supports Dr. Fishe's approach and conclusions.
Dr. Fishe's third area of inquiry was whether there were economic circumstances that existed during 1982 through 1983 and thereafter that would, in a competitive market, cause Exxon to "take *1349 back" the DFC offset. In other words, from an economic standpoint, why would Exxon cause harm to its dealers? The Daubert factors do not readily apply to Dr. Fishe's opinions concerning Exxon's purported business plan or purpose to eliminate "non-keeper" dealers.[25] Nonetheless, Dr. Fishe's economic opinions are well-grounded on Exxon's own documents and actions.[26] Reliability is sufficiently established on that basis. Dr. Fishe has made permissible inferences from available documentation to support his ultimate conclusion.
According to Dr. Fishe, in the 1970's, all major oil companies were reducing their dealer network due to a drop in demand in the United States market for gasoline. During that period, the number of Exxon store closings was below that of Exxon's major competitors. During the 1980's, however, Exxon changed its business strategy and implemented a higher rate of divestment. Its percentage of divestment went from the lowest to the highest among its major competitors starting at the time of DFC and thereafter throughout the 1980's. Obviously, Exxon's divestment rate may have been caused by numerous factors, exclusive of DFC, but it does constitute one piece of the puzzle of circumstantial evidence that Plaintiffs propose to offer to the jury.
According to Dr. Fishe, by imposing an additional cost on its dealers — the credit cost recovery fee — Exxon could observe which outlets could survive economically with lower margins. Those that could survive with lower margins offered greater potential return to Exxon in the form of higher wholesale margins. Those dealers then became more viable candidates for the investments that Exxon had to make to modernize and implement its "pacesetter model."[27] The essence of his opinion is that the failure to offset the credit cost recovery fee was an economically rational decision given Exxon's goals at the time but that these circumstances, collectively, were not "normal" from an economic standpoint.
Dr. Fishe controlled for factors extraneous to the issue of DFC compliance by comparing Exxon dealer margins to the margins of other dealers in the industry for the same period.[28] Dr. Fishe also controlled *1350 for the potential bias which would be introduced by aggregating full-service and self-service margins into composite pool margins for the purposes of the relevant analysis.[29] Dr. Fishe than applied standard statistical regression models which confirmed with a high degree of statistical confidence that the change in the level of Exxon dealer margins, on both an absolute and relative basis, were statistically significant, thereby excluding other causes for the decline.[30]
Although disputed by Exxon,[31] the methodology employed by Dr. Fishe, including the use of controls for extraneous causal factors, together with the application of regression analysis to test the validity of the relevant statistical hypothesis, were appropriate both in their theory and in their application and were of the type routinely utilized by economists and statisticians in conducting such inquires. Moreover, the results drawn from the margin tests were consistent with the various data sets and analyses including Exxon's own internal margin analyses. While Exxon, by cross-examination, and through Dr. Kalt, challenged Dr. Fishe's selection of data, choice of controls, and the like as being fatally flawed, the court respectfully disagrees and concludes that much of Exxon's attack goes more to weight rather than to reliability.
Dr. Fishe has applied appropriate control mechanisms and standard statistical techniques as appropriate to his various analyses. The results of the "runs test" across all grades of Exxon motor fuel were *1351 consistent and statistically significant at a 99 % level of confidence. In addition, Dr. Fishe applied regression analysis to his corrected WPDS data series in concluding that wholesale prices are biased in Exxon's favor by 1.37 cents per gallon. The regression analysis was performed at a 99 % level of confidence. In response to Exxon's criticism that the data Dr. Fishe used as Exxon's initial condition versus its competitors was not representative, Dr. Fishe provided a matrix of all possible values for the pre-DFC period. The matrix, according to Dr. Fishe, establishes that 98.4 percent of all possible values show Exxon's pre-DFC wholesale price was below other majors, and that the average of all these values is -.9 cents, thereby, in his view, confirming the -.94 cent value from January to August 1982 which Dr. Fishe used as his "initial condition" benchmark.
During the course of DFC, Exxon managers repriced motor fuel over 106,000 times. Yet, conspicuously absent from the Daubert hearing is any coherent paper trail showing how, and the manner in which, these pricing adjustments were implemented throughout the DFC period so as to document compliance with Exxon's DFC commitments.[32] While the absence of such records is certainly not determinative, considerable weight must be given to those Exxon business records which are available and are relevant to the questions posed. Dr. Fishe, in fact, considered such records, as produced in discovery, as part of his analysis.[33]
The Exxon documents offered by Plaintiffs during the Daubert hearing measure the financial effects of the Discount for Cash Program to Exxon on both a short term basis and a long term basis. One possible interpretation of these documents is that Exxon had a business plan to eliminate the DFC adjustment[34] by 1985 as part of an effort to generate a total net financial effect of $85 million dollars at 75% dealer participation. Whether this was Exxon's intent must await further testimony at trial. But, it is fair game for an expert to consider such documents in conjunction with the application of his own methodologies, particularly where the opposing expert is contending that his opinions are unreliable. At a minimum, these *1352 documents suggest that Dr. Fishe's opinions are supported by more than unsupported speculation or subjective belief. Daubert, 113 S.Ct at 2795.
Dr. Kalt opines that economic evidence does not support Plaintiffs' theory that Exxon failed to effectuate cash base prices for its gasoline by removing from its wholesale gasoline price an amount sufficient to offset the cost of credit card processing. To the contrary, he asserts that wholesale price comparisons show that on average over the DFC period, across all grades of gasoline, and across all of the markets in which Exxon operated, its wholesale prices were below the average of its key cash-basis competitors. If Exxon took back the DFC offset, in his view, Exxon's prices would be at the level of its credit-basis competitors and, in fact, Exxon was .14 cents below the average cash basis price of its competitors and 1.52 cents per gallon below the average of its competitor's credit-basis prices. Moreover, he concludes that the comparison of Exxon's dealers' margins to the margins of Exxon's competitors' dealers show that, on average over the DFC period, across all grades of gasoline, and across all of the markets in which Exxon operated, Exxon dealers enjoyed generally higher gross margins than their competitors.
Dr. Kalt further opines that during the time of the purported take back of the DFC reduction in price, Exxon's wholesale price actually fell in absolute terms and was essentially unchanged relative to its competitors prices. Finally, contrary to Dr. Fishe's conclusions, Dr. Kalt opines that basic economic principles indicate that it would not have been in Exxon's rational self-interest to fail to implement truly cash-basis pricing for its gasoline, or to take back such pricing after some period of initial implementation.
Plaintiffs challenge four aspects of Dr. Kalt's opinions: (1) his "margin count" study for the period 1988-1994, both for all Exxon dealers and for the named Plaintiffs[35]; (2) his creation of wholesale pricing data which he refers to as the Lunberg adjusted data; (3) his "difference" analysis in which he reaches conclusions based upon a comparison of Exxon's wholesale prices as compared to the wholesale prices of Exxon's key competitors as reported in Exxon's WPDS system, and (4) his opinions based upon an aggregation of Exxon dealer margins on both a relative and absolute basis using "common weights" of full service and self service data into a composite pool. The court respectfully disagrees and concludes that Dr. Kalt's *1353 use of data and methodologies is sufficiently reliable for Daubert purposes.
In reviewing the Daubert factors as applied to Dr. Kalt's testimony, the court finds similar use of the data, methodologies and statistical techniques as those applied by Dr. Fishe. Obviously, each expert tested his hypotheses in different ways, but that does not mean that each approach was unreliable. What can be said about Defendant's attack of Dr. Fishe applies equally to Plaintiffs' of Dr. Kalt. It goes way beyond reliability and into weight. In the court's view, from a reliability standpoint, Dr. Kalt used reliable and scientifically supported data and implemented proper statistical controls to account for unrelated factors in reaching his conclusion. As such, he, like Dr. Fishe, may testify about his conclusions based on his various analyses. Ultimately, the jury will determine who is right or wrong by the preponderance of the credible evidence in the case.
Dr. Kalt's pricing analysis is reasonably based on publicly available posted prices from Lundberg Inc. and from Exxon's WPDS internal pricing system. While Dr. Fishe challenges aspects of the WPDS data, the court concludes it is sufficiently reliable for Daubert purposes.
Dr. Kalt's second and third regression tests examined the "but for" effect of changes in the adjustments used by Exxon in its wholesale pricing system on the actual prices that Exxon set over the DFC program and at the time of the DFC adjustment-line-item introduced in 1991. The methodology employed in reaching his conclusion is sufficiently reasonable. While Plaintiffs claim that Dr. Kalt's comparison of net wholesale prices during the period of the DFC program is unreliable because it fails to take account of the relative level of wholesale prices before the DFC program, the court concludes that this is a matter of weight, and that its omission, while certainly probative, does not make his analyses, methods, or conclusions sufficiently unreasonable as to be unreliable. Dr. Kalt adequately responds to the challenge raised by utilizing other tests to control for the prospect that wholesale price comparisons might be biased. The jury ultimately will consider the weight to be given to Dr. Kalt's failure to use an initial conditions benchmark.
Dr. Kalt also did two types of analyses regarding dealer margins. First, he tested Plaintiffs' contention that Exxon dealer margins were squeezed by Exxon's purported take back of the DFC offset by the effects of low brand values. Second, he utilized dealer margins relative to competitor margins. Plaintiffs challenge Dr. Kalt's methodology on the grounds that they were not based on proper factual assumptions. The court concludes, however, that the assumptions made were sufficiently reliable for Daubert purposes.
After considering the relevant Daubert and other factors, I conclude, based on the preponderance of the evidence, that the expert testimony of both Dr. Fishe and Dr. Kalt is reliable and can properly be applied to the issues in this case. Each expert's approach to test his respective hypotheses was reasonable; that is, each expert's use of his particular approach, along with his particular method of analyzing the data to draw a conclusion, was reasonable and employed a high level of intellectual rigor that well characterized the discipline of econometrics. None of the methodologies employed, or choice of data utilized, was based solely on guesswork, speculation or conjecture. Dr. Fishe's reliability is further established when his opinions are reviewed in context with Exxon's own documents, and in the case of pre-or-early-DFC, the conspicuous absence of important Exxon documents.
While the cross-examination, affidavits, and counter-testimony were significantly probing, and did in fact raise legitimate questions as to the weight to be given to each expert's methodology and data used, *1354 it is not a district judge's function at a Daubert hearing to determine that the expert's testimony was irrefutable or certainly correct.[36] It is sufficient that each expert's reasoning and methodology had a reliable foundation in the knowledge and experience of his discipline, regardless of claimed errors of interpretation. It is further significant that each expert has more than adequately responded to the opposing counter-attacks with convincing, countervailing factors which, on balance, operate in favor of admissibility. In sum, the opinions of both experts are well within the range where experts may honestly differ, and where the jury must decide among their competing points of view.
WHEREFORE, its is ORDERED:
1. Exxon's Motion to Preclude Expert Testimony on the Amount and Basis for Damages [D.E. # 1028] is DENIED.
2. Exxon's Motion to Strike Paragraph 5 of the May 17, 1999 Affidavit of Dr. Raymond Fishe [D.E. 1042] is DENIED.
3. Exxon's Motion to Exclude Expert Testimony by Dr. Raymond Fishe [D.E. 1020] is DENIED.
4. Exxon's Motion to Preclude the Testimony by Plaintiffs' Expert on Analysis not contained in his Final Report [D.E. # 1091] is DENIED.
5. Plaintiffs' Motion to Exclude or Limit Testimony of Exxon's Expert Witness, Dr. Joseph P. Kalt [D.E. # 1018] is DENIED.
6. The Plaintiffs and Exxon are precluded from citing, quoting or referring to any matter in this order to the jury during the trial of the case. The jury shall not be advised, directly or indirectly, that the court has held a Daubert hearing, or has determined that either expert's testimony was found to be reliable. The court shall impose significant monetary sanctions against any attorney, and the respective party, in the event of a violation of this order. If either party seeks to impeach the opposing expert by statements made during the Daubert hearing, such party shall reference the source of the statement as having been made "at prior proceedings."
[1] The court has entered a number of orders on recent pending motions and on pretrial procedures. The more substantive orders include those denying Exxon's Renewed Motion for Summary Judgment [D.E. # 1129]; denying Exxon's Corporation Motion in Limine to Exclude Extrinsic Evidence to Establish the Alleged Obligation under the Written Contracts [D.E. # 1119]; and denying Plaintiffs' Motion for Leave to Assert Claim for Punitive Damages [D.E. # 1156].
[2] See Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993) and Kumho Tire Co. v. Carmichael, 526 U.S. 137, 119 S. Ct. 1167, 143 L. Ed. 2d 238 (1999) which established "gate-keeping" procedures under Fed.R.Evid. 702.
[3] This case was filed in 1991 and assigned to the Judge Kehoe. He had set the case for trial in January, 1999. After Judge Kehoe's death, the undersigned, following consultation with the parties, reset the trial date for August 2, 1999. By agreement, the Daubert hearing was set for July, 1999.
[4] The Plaintiffs also filed additional motions which are addressed by separate orders. These include Plaintiffs' motion to preclude the testimony of Avrom Landesman [D.E. # 1018]; Plaintiffs' motion for leave to assert claim for punitive damages against Exxon [D.E. # 1016], and Plaintiffs' motion to allow plaintiffs' expert, Dr. Raymond Fishe, to publish academic article regarding oil industry pricing [D.E. # 1013].
[5] Federal Rule of Evidence 702, Testimony by Experts, reads: "If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise."
[6] Neither party challenges these conclusions.
[7] In City of Tuscaloosa v. Harcros Chemicals, Inc., 158 F.3d 548, 566 n. 25 (11th Cir.1998), rehearing denied en banc, 172 F.3d 884, petition for certiorari filed May 10, 1999, 67 USLW 3717 (1999), the Eleventh Circuit, in a pre-Kumho case, discussed, in a footnote, the proper inquiry regarding the reliability of the methodologies implemented by economic and statistical experts in the Daubert context. The entire footnote is set forth because of its importance to this court's analysis:
Twenty-nine states, along with the territories of Guam and Puerto Rico, have briefed this appeal as amici curiae in support of the appellant municipalities. The amici helpfully point out that, although "an important aspect of assessing scientific validity (and therefore evidentiary reliability) is the ability of other scientists to test or retest a proponent's theory," not every scientific or technical methodology applied by expert witnesses is susceptible to such an analysis. In re Paoli R.R. Yard PCB Litig., 35 F.3d 717, 758 (3d Cir.1994), cert. denied sub nom. General Elec. Co. v. Ingram, 513 U.S. 1190, 115 S. Ct. 1253, 131 L. Ed. 2d 134 (1995). Economic or statistical analysis of markets alleged to be collusive, for instance, cannot readily be repeatedly tested, because each such case is widely different from other such cases and because such cases cannot be made the subject of repeated experiments. The proper inquiry regarding the reliability of the methodologies implemented by economic and statistical experts in this context is not whether other experts, faced with substantially similar facts, have repeatedly reached the same conclusions (because there will be few or no cases that have presented substantially similar facts). Instead, the proper inquiry is whether the techniques utilized by the experts are reliable in light of the factors (other than testability) identified in Daubert and in light of other factors bearing on the reliability of the methodologies. Id. note 16; see also, e.g., Ohio ex rel. Montgomery v. Louis Trauth Dairy, Inc., 925 F. Supp. 1247, 1252-53 (S.D.Ohio 1996) (noting that Daubert factors should be applied flexibly in determining admissibility of economic and statistical expert testimony); In re Aluminum Phosphide Antitrust Litig., 893 F. Supp. 1497, 1506 (D.Kan.1995) ("To the extent that Daubert factors are relevant to its determination [whether an economist's methodologies are reliable], the Court considers them along with any other relevant factors. ... The inquiry is a flexible one, with the overarching subject being the evidentiary relevance and reliability of the principles that underlie a proposed submission."). One factor left unenumerated by the Daubert Court is whether "the methods used by the expert to derive his opinion satisfy the standards for scientific methodology that his profession would require of his out-of-court research." People Who Care v. Rockford Bd. of Educ., Sch. Dist. No. 205, 111 F.3d 528, 537 (7th Cir.1997). Taking this factor into account along with the factors identified in Daubert, we readily conclude that McClave's methodologies are quite reliable, with the small exception explained infra.
[8] Proposed amendments to Rule 702 are currently under consideration. Should these amendments take effect, trial judges conducting the gate-keeping function will be required to undertake a threepart reliability inquiry in connection with all expert testimony. Gate-keeping judges must make certain that all expert testimony admitted at trial "is sufficiently based upon reliable facts or data" and "is the product of reliable principles and methods," and that "the witness has applied the principles and methods reliably to the facts of the case." Proposed Fed.R.Evid. 702, quoted in 67 U.S.L.W. 2588 (April 6, 1999), and referred to in United States v. Charley, 176 F.3d 1265, 1277 n. 15 (10th Cir.1999).
[9] This is not a case of some "late arrival" expert who testifies without significant knowledge about the facts of the case based on some new, unproven, methodology. Here, both Dr. Fishe and Dr. Kalt are well-qualified in econometrics; each has substantial knowledge about the facts of the case based on extensive research and involvement over several years and, while the application of economic/statistical methodologies to the data are at issue, there is no "new" or "untested" methodology involved.
[10] This case does not lack in analysis. As described above, by April 15, 1999, each expert filed extensive reports with accompanying exhibits. Initially, Dr. Fishe submitted an 80 page report explaining his conclusions, data and methodologies. Dr. Kalt was equally expansive with his 40 page report with detailed charts and exhibits. Dr. Fishe responded to Dr. Kalt with his affidavit of May 17, 1999. Dr. Fishe's methodologies was supported a 12 page affidavit from Dr. William D. Nordhaus, a professor of economics from Yale University. He also concluded that Dr. Kalt's statistical tests were unreliable. In turn, Dr. Kalt filed his detailed 34 page affidavit, with numerous exhibits, challenging Dr. Fishe's conclusions and methodologies. In response, Dr. Fishe filed a supplemental 50 page affidavit responding to Dr. Kalt. Finally, Dr. Nordhaus filed a second affidavit, dated June 11, 1999, critical of the methods in Dr. Kalt's affidavit of June 5th. All of these matters were again addressed during the court's six-day Daubert hearing.
[11] Plaintiffs urge the court to consider decisions of the United States Supreme Court and the Eleventh Circuit that have repeatedly observed that wrongdoers may not argue that the plaintiff's damages methodology is speculative or uncertain, as long as it has a valid evidentiary basis. See Bigelow v. RKO Radio Pictures, Inc., 327 U.S. 251, 66 S. Ct. 574, 90 L. Ed. 652 (1946); Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 51 S. Ct. 248, 75 L. Ed. 544 (1931); Eastman Kodak Co. of New York v. Southern Photo Materials, 273 U.S. 359, 47 S. Ct. 400, 405, 71 L. Ed. 684 (1927); United States v. Killough, 848 F.2d 1523, 1531 (11th Cir.1988); Ramada Inns, Inc. v. Gadsden Motel Company, 804 F.2d 1562, 1565 (11th Cir.1986).
On the other hand, Exxon urges the court to consider those cases striking expert testimony when it is wholly deficient and fails to provide a just and reasonable basis for an award of damages. General Electric Company v. Joiner, 522 U.S. 136, 146, 118 S. Ct. 512, 139 L. Ed. 2d 508 (1997) (nothing in the "Federal Rules of Evidence requires a district court to admit opinion evidence [where] ... there is simply too great an analytical gap between the data and the opinion proffered"); Navarro v. Fuji Heavy Indus., 117 F.3d 1027, 1031 (7th Cir.), cert. denied, ___ U.S. ___, 118 S. Ct. 600, 139 L. Ed. 2d 488 (1997) ("a conclusion without any support is not one based on expert knowledge and entitled to the dignity of evidence"); Joy v. Bell Helicopter Textron, 999 F.2d 549, 564 (D.C.Cir.1993) (expert testimony that was unsupported by record and based solely on guesswork, speculation and conjecture inadmissible); Viterbo v. Dow Chemical Co., 826 F.2d 420, 423 (5th Cir. 1987) (excluding expert opinion based on incomplete data); Handi Caddy, Inc. v. American Home Products Corp., 557 F.2d 136, 141 (8th Cir.1977) (vacating jury award where expert testimony relied on assumptions contradicted by the record).
[12] In complicated cases such as this, a timely appointment of a court appointed expert to comment on the attacks and counter-attacks could help assist the district judge in understanding and analyzing the crucial differences between the experts, and the reasonableness of each expert's methodology, data and approach given the Daubert factors. See Fed. R.Evid. 706.
[13] In their final arguments, and during the course of the six day Daubert hearing, each party raised innumerable challenges to validity of each expert's use of data, assumption and methodology. To listen to the parties, both experts, who have spent years preparing their testimony, did nothing reliable, credible or worthy of further review by the fact finder. While the court complements counsels' advocacy, it is self-evident that each party perceived Daubert as an opportunity to vitiate the other's case or defense. While this may be a function of Daubert, it is not, nor should it be, its intended purpose.
[14] The methodology by which Exxon determined the initial 1.7 cpg is not disputed.
[15] Plaintiffs do not claim, nor will there be any expert testimony to support, that Exxon incorrectly applied the adjustment factor on a year by year basis, such that the Plaintiffs' damages would equal the difference between the factor applied verses the factor that should have been applied. Plaintiffs' position is that no adjustment occurred after March 1, 1983, and that there were no "give backs" by virtue of later adjustments.
[16] U.C.C. Section 1-106 specifies the measure of damages in a breach of contract action under the Code. It states: "The remedies provided by this code shall be liberally administered to the end that the aggrieved party may be put in as good a position as if the other party had fully performed but neither consequential or special nor penal damages may be had except as specifically provided in this code or by other rule of law." While Exxon argues that the use of on average nationwide data may mask the fact that some dealers may have post-DFC margins that are greater than pre-DFC margins, the court has previously held that Exxon's alleged good faith obligation is owed to each dealer, and to the class as a whole. Exxon's breach, if any, can be determined only by an "on average, over time, across all markets" analysis of Exxon relationship with all of its dealers. "Following such common proof of damages suffered by the class or by subclasses, the total recovery will then be allocated to individual members, a process that may be accomplished by various means, such as ... use of average damages formulae on a per capita basis .... Once the defendant's total damages liability has been determined, then the allocation of that aggregate sum among class members is an internal class action accounting question that does not directly concern the defendant...." Newberg on Class Actions at 4.26.
Exxon's stated position does not seem to be inconsistent. The use of "average nationwide" data for damage purposes is supported by Exxon's own answers to interrogatories. As set forth in Exxon's Interrogatory Response 67 (7/29/96), "Exxon did not attempt to reduce each dealer's individual wholesale price by an amount which offset the amount of the credit card charges actually paid [sic] [by] each dealer." The Discount for Cash program did not contemplate any such reduction, and to do so would have made the program administratively impossible since it would require different reductions for each dealer based on that dealer's own individual retail prices and the percentage of gasoline the dealer sold on credit. Exxon dealers were informed, and should have understood, that the reduction would only be "on average" and that therefore the amount of their individual reduction in the wholesale prices would not necessarily offset the amount of the credit card processing charge each dealer individually paid. In some cases, the reduction in price would have exceeded the amount of an individual dealer's credit card charges. In other cases, the amount of the credit card charges would have exceeded the amount of an individual dealer's price reduction. (Emphasis added).
[17] In its answer to Interrogatory # 35, Exxon states that "the fact that Exxon provided the wholesale price reduction can be objectively verified. The price reduction is clearly shown in Exxon's Wholesale Price Development System documents and data." While there is a "DFC" adjustment shown on what is known as "2s" and "5s" reports which were prepared each day for each market throughout the United States, both experts agree that this information, alone, does not answer the ultimate question presented. It merely provided Exxon managers information upon which to make adjustments to its wholesale prices and does not prove that the appropriate adjustments were actually made throughout the DFC program.
[18] The Plaintiffs have submitted various Exxon documents to substantiate their claims. Dr. Fishe's opinion rests, in part, on these documents.
[19] In an effort to create comparisons between its wholesale prices and the wholesale prices of its key competitors, Exxon, in early 1982, created what is called the Wholesale Price Data System. The objective of WPDS was to adjust key competitors' posted prices for differences known to Exxon so as to create comparable "net" or "adjusted" prices. The use of WPDS data has been a subject of significant debate between the two experts. For instance, Dr. Fishe argues that the data contains certain bias, while Dr. Kalt argues that its does not. Exxon challenges Dr. Fishe for using the WPDS data as part of his analysis notwithstanding the claimed bias. The court concludes that the use of the data by each expert, as part of their methodology, is sufficiently reliable. The weight to be given to the data, and its use by each expert, shall be left for the jury.
The use of other data also was at issue between the parties. For the period January 1982 until sometime in 1988, Exxon maintained data reflecting both its and its competitor's retail prices in what it called its Retail Price Data System ("RPDS"). RPDS also was used by Exxon to record its dealer margins (the difference between Exxon's wholesale prices and its dealers' retail prices) and competitor dealers' margins (the difference between its competitors' "net" wholesale prices, as computed by WPDS, and the competitor dealers' retail prices). Exxon used RPDS data to periodically provide management reports reflecting the "delta" or difference between Exxon dealer margins and the dealer margins of its competitors.
[20] In its Daubert memorandum, Exxon challenges the following opinions of Dr. Fishe: (1) Exxon did not reduce wholesale prices to dealers on average over time across all markets sufficient to provide an offset to the three percent credit cost recovery fee; (2) Exxon's DFC marketing program was used as a secret method by which to identify, segregate and eliminate inefficient dealers, creating an "abnormal" business environment and an incentive to overcharge dealers, and (3) Plaintiffs damages equal the total amount of credit card cost recovery fees collected by Exxon during the DFC marketing program.
While Exxon does not challenge Dr. Fishe's qualification as an expert, it does raise innumerable challenges to the reliability of his analysis. Exxon claims that he used improper methods of analysis that failed to control for factors other than those Dr. Fishe sought to investigate; that Dr. Fishe's analysis assumes facts that are not supported or are contradicted by the record; and that Dr. Fishe used unreliable or unrepresentative data. Exxon also challenges Dr. Fishe because he formulated his opinions before conducting any meaningful work, and because he repeatedly changed the analyses on which he then relied to support those conclusions.
The court respectfully disagrees and offers an initial observation. Dr. Fishe's opinions [as well as those of Dr. Kalt's] have evolved over the case as discovery has proceeded. Some of the discovery provided, particularly of the Exxon's electronic data base, was not in any clearly understandable form. Dr. Fishe has had to spend considerable time recasting the data into useable form for proper statistical analysis. The fact that his analysis evolved does not necessarily equate to unreliable. Importantly, the court has had an opportunity to evaluate Dr. Fishe's explanations during his cross-examination.. Contrary to Exxon's position, the court does not view Dr. Fishe as the "hired gun" called upon to speculate and hypothesize and "plug holes" in Plaintiffs case. Of course, Plaintiffs have raised such a charge against Dr. Kalt who has received approximately one million dollars from Exxon for his testimony in this case and who appears to testify regularly for Exxon in other cases, as well as for other major oil companies. Merely because Dr. Kalt was paid well for his time, however, does not necessarily mean that his work effort was unreliable.
In this order, the court has endeavored to set forth the essential reasons for its conclusions rather than to address in detail each one of Exxon's [and the Plaintiffs'] innumerable challenges. This does not mean, however, that each such challenge was not considered and evaluated as part of this court's Daubert reliability analysis. The court has been fully briefed on each such challenge, has heard comprehensive final arguments, and has considered proposed orders from the parties. The court also has read the extensive expert reports, reply affidavits, and the legal memoranda of the parties. Time constraint, however, [given the start of this nine year old case on August 2, 1999] precludes the crafting of a more detailed order which addresses each specific challenge.
[21] Exxon stated, "Margin data further demonstrate that Exxon provided the wholesale price reduction."
[22] Dr. Kalt also stated in his affidavit: "I believe that if an overcharge or duplicative charge did occur as plaintiffs allege, properly applied economic analysis would distinguish the effect of the alleged overcharge or duplicative charge from other factors that might affect retailer profitability and margins. If dealer profitability and margins are found not to have been depressed by the alleged overcharge or duplicative charge, however characterized, that would, in my opinion, be substantial evidence that the behavior the plaintiffs charge did not occur." Id. at Paragraph 6. In his initial expert report, filed January 22, 1997, pages 13 through 14, Dr. Kalt reached a similar conclusion. He agreed that if the overcharge took place, it would be borne by the Exxon dealers and not their ultimate customers. He stated: "... if Exxon retailers were subjected to the alleged overcharge or duplicative charge, market-place forces would preclude the formulaic pass through of such an overcharge or duplicative charge and would cause Exxon's dealers margins to be reduced and the resulting economic burden of the alleged overcharges of duplicative charges to be borne all or in substantial part by Exxon retailers." Id.
[23] Exxon's own margin analysis for the Gulf/East Coast region is instructive. While it represents only a margin sample between pre DFC and post DFC through the second half of 1984, it does document that the margins of other dealers either went up or stayed the same (relative to the margins of Exxon dealers) after Exxon adopted its DFC Program. It also demonstrates that after an initial reduction, Exxon's wholesale margin returned to pre-DFC levels, and that Exxon received, in addition, the income from the 3% CCR fee (called the "merchant's fee" on the chart). Finally, the chart shows that, while Exxon's wholesale margin together with the CCR fee increased relative to its competitor's wholesale margins, the margins of Exxon dealers were depressed following the imposition of the CCR fee. Thereafter, Exxon updated this chart with more current information which, according to Dr. Fishe, further establishes the existence of the overcharge; that is, the overcharge is reflected in Exxon dealer's "reduced" profitability resulting from depressed margins.
[24] In scientific terms, a "runs" test is used to validate the operation of a process. In seeking to exclude Dr. Fishe's runs test, Exxon does not argue that this is not a scientifically valid analytical technique, but rather, that if it was consistently higher or consistently lower than the average it does mean very much as long as it was "close." The court finds this argument goes more to weight than reliability.
[25] Dr. Fishe's conclusions in this area are not subject to peer review, a known or potential error rate, or the existence and maintenance of controlling standards. Rather, Dr. Fishe makes inferences from Exxon's own proprietary documents which remain under seal and which are not subject to peer review. Moreover, Dr. Fishe's opinions are not readily subject to standard economic methodologies, such as regression analysis. For this reason, the court has considered other factors in determining that these opinions are sufficiently reliable to be offered to the jury. The significant other factor are Exxon's own documents as well as Dr. Fishe's charts documenting trends that allow for permissible inferences.
[26] In Dr. Fishe's April 15, 1999 report, he discusses Exxon documents that support the concept that Exxon was willing to impose a cost on its stores in order to identify its better performing dealers. Instead of DFC, Exxon was implementing a rebate program for its retail auto stores. The program was funded by increasing Exxon's wholesale prices, so that only some participating dealers received the rebate, while others had to face a higher cost of gasoline.
[27] Dr. Fishe reports that Exxon's documents show that it used a "pacesetter outlet" as a model of the future. The model was shown to dealers and analyzed for internal planning purposes. Exxon's documents discuss the pacesetter goal and the difference between a pacesetter store and an Exxon outlet.
[28] Because a variety of factors could affect dealer margins, Dr. Fishe sought to control those "other factors." He did so with this "before and after" comparison of Exxon dealer margins over time to the "before and after" margins of competitive dealers over time. One other factors was for the effect of inflation (nominal dollars versus inflation adjusted dollars) which he controlled by adjusting data to account for the changes in the value of money over time. Dr. Fishe also examined data on "absolute margins" over time to compare to the data reflecting changes in margins relative to competitors' margins. With respect to these data he performed a series of statistical tests using "regression analysis."
[29] Another key issue in dispute concerns the aggregation of full service and self service data. Exxon's aggregation shows no significant change in dealers' margins. Dr. Fishe examined Exxon's aggregation methodology and concluded that no adjustment had been made by Exxon for differences between full service and self service ratios verses competitors. In Dr. Fishe's view, because Exxon's ratio of full service to self service sales was higher than the group to which margins were being compared [i.e. a larger percentage of Exxon's dealers' sales occurred at full station verses split island or self-service stations], it created a false impression that Exxon dealers' margins had remained flat in relationship to competitors. In his view, they dramatically declined in early 1983 [the period of the "take back"]. To address this issue, he prepared a statistical analysis in which the separate full service and self service margin deltas were aggregated. He observed a "statistically significant break" occurring prior to March of 1983, which appeared whether the series began in September, 1981, or in February, 1982, and continued throughout DFC. Based on this analysis, he concludes that there was no "give back" of the "take back." According to Dr. Fishe, a recovery of the magnitude of the "take back" would have been observable in both the relative and absolute dealer margin data, and no such recovery occurred. Dr. Fishe is critical of Dr. Kalt whose statistical analysis of Exxon's pricing data begins about the time of the claimed "take back."
Plaintiffs support Dr. Fishe's position on margins with Exxon's internal management reports. For example, Plaintiffs rely on an Exxon management report in 1994 which observed that while its wholesale margins had dramatically increased in the 12 year DFC period, for the 10 year period, 1984 to 1994, its dealers' margins were "Flat; profits squeezed by cost pressures." Plaintiffs argue that Exxon's own internal records demonstrate that it focused on self service sales instead of "composite" margins which include the elements of full service.
[30] Dr. Nordhaus concurred with Dr. Fishe that Dr. Kalt's aggregation of full and self serve margins where Exxon's full serve ratio is higher than its competitors would not constitute proper economic analysis.
[31] Exxon argues that Dr. Fishe's opinions and testimony fail to satisfy the standards of relevancy and reliability required by Rule 702. Regarding the offset and damages, Exxon claims that Dr. Fishe's methodology is suspect because he failed to control for unrelated factors; relied on assumptions contrary to facts or where the facts were internally inconsistent; used arbitrary and unscientific data sets and time periods; failed to conduct independent research and reach independent conclusions, and in the case of his opinion regarding the business environment during DFC, lacked relevant expertise. The court respectfully disagrees.
[32] There were no internal Exxon documents offered at the Daubert hearing that tracked Exxon's own methodology as set forth in its answer to Interrogatory # 35; that is, documenting how Exxon actually used and then implemented its key competitor's net average wholesale price as a target for its own wholesale price, on a regular basis. Dr. Fishe raises a fair question in his April 15, 1999 report. "How could Exxon be sure that it was in compliance over the twelve years of the DFC program if it did not make the average DTT delta calculation on a regular basis?" Of course, as noted above, even if Exxon mismanaged its own mechanism for the offset, the results still favor the dealers assuming the wholesale price adjustments were not taken back.
[33] Another gap in Exxon's data pertains to the comparison of Exxon's prices to the prices of its key competitors for the period before and after introduction of discount for cash. Exxon's own electronic data base covered the period of January 28, 1983 through August 22, 1994. Discount for Cash started in August of 1982. Exxon did produce management reports for the period immediately before and after commencement of DFC which Dr. Fishe used in analyzing Exxon dealer margins on both a relative and absolute basis.
[34] The 1982 Exxon management documents included a "DTW Reduction" in measuring the program's financial effects. The comparable 1985 documents did not include the DTW reduction. Dr. Kalt did not address these documents in his testimony. Obviously, the importance of these documents must be left to trial to be placed in context with the testimony of Exxon officials. Yet, for Daubert purposes, their existence adds a factor in support of the "reliability" of Dr. Fishe's expert testimony. For instance, one Exxon marketing department document, dated November 2, 1990, in discussing original DFC assumptions, refers to the "Elimination of the short-term requirement to reduce the DTT.." Another document, in discussing the long term effects of the DFC, allocates reductions in DTT to volume protection and dealer participation, and not to DFC wholesale price reduction.
[35] In a prior motion, the Plaintiffs have sought to exclude various "individual" dealer margins, profits, and prices. Plaintiffs have argued that this evidence is irrelevant to the elements of a contractual duty and a breach of that duty relative to the class of dealers as a whole. Plaintiffs also contended that the evidence is prejudicial in that it may tend to mislead the jury to determine that, since certain of the dealers did not realize a substantial loss, if any, the class of dealers was not damaged by Exxon's breach of its class-wide obligations. At issue here is Dr. Kalt's margin analysis of the named Plaintiffs. By prior order, the court reserved ruling on the matter pending the completion of the Daubert hearing. After now hearing the matter in context, the court concludes that Dr. Kalt's margin analysis of the named Plaintiffs is relevant and, on balance, shall not be excluded under Fed.R.Evid. 403. The jury may hear that certain dealers benefitted from DFC and others did not "on average." Using the named Plaintiffs to make this point is "fair game." Plaintiffs, in turn, may request a limiting instruction when this evidence is offered, and may offer rebuttal evidence that other class dealers did not so benefit. In terms of this testimony, the court shall hold Exxon to Dr. Kalt's analysis of the named Plaintiffs. No additional witnesses may be called since such testimony would be cumulative. See Fed. R.Evid. 611(a) (the court shall exercise reasonable control over the mode and order of interrogating witnesses and presenting evidence to avoid needless consumption of time). Under the same rule, the court shall limit the number of rebuttal witnesses, or rebuttal expert analysis of the margins of class members, to no more than 12 class members (i.e. the same number as the named Plaintiffs).
[36] Each expert has sufficiently controlled for variables to establish his position on the existence, or lack of existence, of causal connections; each sufficiently has based his respective hypotheses on factual assumptions that are supported by Exxon documents, or other reliable data, although the interpretation and application of the data is in dispute, and each has presented to the court a sufficient explanation, founded on a scientific basis, consistent with the discipline involved, for the selection of the data and the time periods used.
