CALLAHAN v. A.E.V., INC

BECKER, Chief Judge:

Prior to 1985, the retail sale of beer in the Pittsburgh area was conducted exclusively by "mom and pop"-type beer distributorships, such as those operated by plaintiff Michael W. Callahan and his fifteen co-plaintiffs. In that year, defendant David Trone opened the first "Beer World" store, a supermarket-style beer distributorship ten times the size of the traditional stores. He opened four more such stores in the Pittsburgh area between 1986 and 1988, offering a larger selection and lower prices. This case involves antitrust and RICO claims arising out of the manner in which Trone operated these stores.

The Pennsylvania Liquor Code limits the ability of one entrepreneur to own or operate more than one beer distributorship. Trone apparently evaded these restrictions by placing the Beer World stores in the names of others, and, while acting as a "consultant," effectively running the stores himself. According to the plaintiffs, Trone deceived the Pennsylvania Liquor Control Board (LCB) as to the true state of affairs by filing of false statements and affidavits.

Trone negotiated purchases of beer from wholesalers for all of the Beer World stores collectively. By doing so, the stores were able to purchase at a wholesale price lower than they would have been able to obtain in individual purchases. Central to this case are Trone's negotiations with defendant Frank Fuhrer, the master distributor in the Pittsburgh area for Anheuser-Busch and Coors, in the course of which Trone allegedly forced Fuhrer to agree to give a quantity discount to the Beer World stores based on their purchases as a group, but not to give this discount to any other retailers. Trone is said to have been able to do this because the Beer World stores held a substantial portion (at least 25%) of the Pittsburgh beer market, and because he threatened to place Fuhrer's products poorly within the stores. The Beer World stores allegedly received this discount even though their orders in the aggregate did not always reach the 4500-case level Fuhrer set for the discount. According to the plaintiffs, this discount was not disclosed to anyone else; it was not included on Fuhrer's ordinary price list and was excluded from loading sheets posted at Fuhrer's distribution center. Not surprisingly, the Beer World stores' advantage in pricing, as well as other areas, cut sharply into the business of the smaller stores.

Several other master distributors were involved in similar arrangements with the defendants. Although they were apparently named in the original complaint, they have now settled with the plaintiffs and are no longer participating in this case.

This state of affairs has spawned this unusual antitrust and civil RICO case with state tort law claims appended, brought by the plaintiffs against Trone, the Beer World stores, and Fuhrer. The plaintiffs' antitrust theory is that Trone, his employees, and the separately incorporated stores have contracted, combined and conspired to restrain trade in beer in Allegheny County, by confronting wholesalers as a group and using their buying power and the threats described above to force the wholesalers to sell them beer at a price lower than that available to other retailers. The plaintiffs' RICO theory is that Trone and others, by submitting false statements and affidavits to the LCB, as well as lying to a grand jury to cover up these false statements, were able to maintain illegal consolidated control of the Beer World stores. The plaintiffs submit that, as a result of this control, Trone and the Beer World stores obtained the advantages that enabled them to sell beer at prices below that of the plaintiffs. Although in a free market, these different approaches to operating a beer distributorship might not seem to offer grounds for a federal antitrust or civil RICO suit, in the context of Pennsylvania's detailed malt and brewed beverages regulatory scheme, the plaintiffs have found grounds for a lawsuit.

The Beer World stores are separately incorporated and named in the complaint as A.E.V., Inc., Beer and Pop Warehouse, Inc., Jet Distributor, Inc., Q.F.A., Inc., and Red Sky, Inc., all of which operate under the Beer World name. Trone is named personally in the complaint, along with the consulting business he runs, Retail Services and Systems, Inc. Fuhrer includes both Frank B. Fuhrer, Jr. himself and his business, Frank B. Fuhrer Wholesale Co.

The District Court granted summary judgment for the defendants on all claims, including both the state tort law claims and the federal claims, and the plaintiffs have appealed. Strangely, antitrust liability issues are not presented in this appeal. The District Court, in deciding the defendants' motion for summary judgment, did not consider antitrust liability issues at all; rather, the District Court disposed of the antitrust and RICO claims on the ground that the plaintiffs had not produced sufficient evidence that they suffered actual losses that were in fact a result of the defendants' actions. Accordingly, and given the incomplete state of the record as presented to us by the parties, we do not intend to engage in an examination of the nature and scope of the plaintiffs' theory or proof of antitrust violations (and, consequently, we express no view as to their correctness). Instead, we will assume, for the purposes of this appeal, that the plaintiffs can offer sufficient proof that the defendants engaged in antitrust violations throughout the relevant time periods. We will accordingly concentrate on the issues — actual loss and causation in fact (termed "fact of damage") with respect to the antitrust claims, and proximate causation with respect to the RICO claim — that are fairly presented by this appeal.

In order to prove that the plaintiffs suffered losses and that the defendants' antitrust violations caused the injuries as a matter of fact, the plaintiffs offered (1) testimony that various customers no longer came to their stores and that the customers explained that this was because the Beer World stores offered cheaper prices, along with (2) the report of an expert who opined that the defendants' actions had caused harm to the plaintiffs. The defendants contend that this evidence is insufficient to meet the plaintiffs' burden of production. They first submit that the plaintiffs' anecdotal evidence is inadmissible hearsay on which the plaintiffs cannot rely. We disagree. The plaintiffs themselves can testify that the customers are in fact no longer shopping at their stores. Furthermore, although the reports of the customers' statements are hearsay, they are admissible as evidence of the customers' states of mind, i.e., their reasons for no longer shopping at the plaintiffs' stores. This combined evidence is sufficient to meet the plaintiffs' burden of producing enough evidence of loss and causation with respect to the plaintiffs' antitrust claims to overcome a motion for summary judgment.

Also on the antitrust issues, the defendants argue that the plaintiffs' proffered expert testimony is inadequate to prove fact of injury and causation because, inter alia, the expert failed to discuss numerous other possible causes of the plaintiffs' losses. Furthermore, the defendants challenge the expert's methodology for estimating the amount of damages. In spite of these flaws, we conclude that the expert's testimony is sufficient to meet the plaintiffs' burden of proof. At all events — taking into consideration both the customer evidence and the expert reports — we believe that the District Court erred in dismissing the plaintiffs' antitrust claims on the ground that there was inadequate proof of fact of injury and causation in fact.

With respect to the RICO claim, the defendants contend that the alleged causal connection between the defendants' fraud and the plaintiffs' losses is not sufficiently close to meet the requirement of proximate causation. The plaintiffs' RICO claim runs as follows: If Trone and others associated with the Beer World stores had not defrauded the Pennsylvania Liquor Control Board by submitting sworn statements that Trone did not own and control all of the stores, the Liquor Control Board would have put Trone out of business. Since he stayed in business, Trone was able to use his control of several stores to obtain volume discounts by buying for the stores in the aggregate. The plaintiffs were then harmed by the defendants' ability to sell at lower prices.

We think this case is similar to Steamfitters Local Union No. 420 Welfare Fund v. Philip Morris, Inc., 171 F.3d 912 (3d Cir. 1999), in which we recently held that the plaintiffs had failed to prove proximate causation. In Steamfitters, we recognized three factors the Supreme Court has identified for determining proximate causation in RICO cases: the directness of the injury, the difficulty of apportioning treble damages among potential plaintiffs, and the possibility of other plaintiffs vindicating the goals of RICO. Given that the plaintiffs are relatively remote third-party "victims" of the fraud and that the LCB itself, or the wholesalers, could take steps to counter the defendants' allegedly illegal actions, we think the plaintiffs' claim meets none of the factors. Accordingly, we believe that the District Court properly dismissed the plaintiffs' RICO claim, although not for the appropriate reason. For these reasons, we will affirm the judgment of the District Court to the extent it dismissed the plaintiffs' RICO claim, but reverse its judgment with respect to the antitrust claims.

I. Facts and Procedural History A. The Pennsylvania Beer Sales Regulation Scheme

Pennsylvania is a state in which temperance with respect to alcoholic beverages has always been an important policy, and statutory regulation of alcoholic beverage sales is extensive. The best known example, of course, is the"state store" system, under which liquor can only be sold in state-owned stores. With respect to malt and brewed beverages there is likewise a panoply of regulations. See, e.g., Pa. Stat. Ann. tit. 47, § 4-441(b) (West 1997) (prohibiting sales in units smaller than one case); § 4-447 (limiting sellers' ability to change prices); § 4-492(2) (prohibiting sales by licensees for consumption on the premises); § 4-492(4) (prohibiting sales on Sunday); § 4-493(2) (prohibiting credit sales of alcoholic beverages other than by credit card); § 4-493(3) (prohibiting exchange of alcoholic beverages for goods or services); § 4-493(8) (prohibiting the use of labels or advertisements containing the alcoholic content of brewed or malt beverages).

For present purposes, we are concerned with the regulation of beer sales. Under Pennsylvania law, beer sellers are divided into four classes for licensing purposes: manufacturers, master distributors, importing distributors and distributors. See Pa. Stat. Ann. tit. 47, § 4-431 (West 1997). The first category consists of breweries. An out-of-state brewer is required to designate a particular importing distributor as the master distributor for a particular geographic area within which only that master distributor is permitted to buy that brewer's beer directly from the brewer. See § 4-431(b). Thus, any beer sold in a particular area must at some point pass through the master distributor designated for that brand in that area. A master distributor can sell beer to importing distributors, (ordinary) distributors or the public. An importing distributor can also sell beer either to other importing distributors, (ordinary) distributors or the public. A distributor can only sell beer to the public. The Beer World stores all have importing distributor licenses, and can therefore sell to each other and to the public. Only some of the plaintiffs have such licenses.

Highly relevant here is the extent to which Pennsylvania law limits the ability of a participant — e.g., a partner, member or shareholder — in one beer distributor to participate in another. See § 4-438 ("No person shall possess more than one class of license. . . ."); § 4-443 (prohibiting interlocking ownership in various forms). In particular, the law restricts the ability of an individual to participate in companies that operate at the same level, although the parties debate the extent to which the law does so. See § 4-438(b) ("No person shall possess or be issued more than one distributor's or importing distributor's license."); § 4-436(e) (application for brewed or malt beverage license must state "[t]hat the applicant is not, or in case of a partnership or association, that the members are not, or in the case of a corporation, that the officers or directors are not, in any manner pecuniarily interested, either directly or indirectly, in the profits of any other class of business regulated under this article, excepts as hereinafter permitted"); § 4-436(f) (applicant must state "[t]hat applicant is the only person in any manner pecuniarily interested in the business so asked to be licensed. . . .").

B. Trone's Beer Business Arrangements

Trone's family had been in the beer business in Harrisburg and Pittsburgh for some time. While a business student at the University of Pennsylvania's Wharton School, Trone apparently came up with a plan for a new type of beer distributorship business. Prior to his plan, beer distributors were typically small, low-capitalization "mom-and-pop" stores of the kind operated by the plaintiffs. They usually had ordinary distributor licenses and operated relatively small stores, selling beer by having people come in and ask for a particular brand. Trone's idea was to create much larger stores, roughly ten times the square footage of the plaintiffs' stores, to be operated like a supermarket. The cases of beer would be set out on shelves so that shoppers could wander through the store picking out particular brands themselves. In addition, Trone planned to offer soda and snacks in addition to the beer. This business plan became the "Beer World" concept. We chronicle the history and management structure of the stores because it bears on the contention that Trone improperly controls all of the stores in violation of the Pennsylvania liquor control scheme, an important part of the plaintiffs' antitrust and RICO claims.

It might seem surprising that Trone was the first to come up with the beer supermarket concept. Indeed, one might think that it would have been around for decades. Perhaps he was simply the first to bring this idea to Pennsylvania. At all events, these ruminations have no bearing on the outcome of this case.

The first Beer World opened in the Pittsburgh area in 1985. Two more stores opened in Pittsburgh in 1986, followed by the last two in 1987 and 1988. The first store, incorporated as Jet Distributors, Inc., is apparently owned by Paul Piho, a childhood friend of Trone's. Piho initially worked full-time in Chicago after the store opened. For a short time, he moved to Pittsburgh and managed the store. Currently, he works at a Delaware branch of a chain of liquor stores apparently owned by Trone. The second store is apparently owned by Trone's wife, who for a time worked at the store, but presently spends less than five hours per week there. The third is apparently owned by Thomas Esper, a retired schoolteacher who apparently knows little about either the store or the liquor business. The fourth store is apparently owned by Trone's sister, who has been in school or working at other jobs for the relevant period. Before 1990 and since 1994, she has lived outside of Pennsylvania. The last store was apparently owned by Albert Vivio, the father of one of Trone's employees. He stated that he did not pay anything to own the store, but that Trone asked him to put his name on a license. He testified that he had "no duties at the store," pursuant to an "agreement with Mr. Trone."

Since the Beer World stores opened, Trone has been employed as a "consultant" for all of them. The plaintiffs allege, however, that Trone's role in the stores is much greater. When the stores opened, he did much of the work in preparing the stores, choosing product line and layout, and selecting employees. He also set up purchasing and delivery systems. Since then, Trone has apparently controlled the day-to-day operations of the stores. He set the salaries for Beer World employees. Employees were routinely moved from store to store while remaining on the payroll of the store in which they began. Although each store maintains a separate bank account in the owner's name, Trone has a stamp of each owner's signature which he uses for checks. He also designed all the advertising for the stores, which included aggressive price advertising until July 1, 1987, when Pennsylvania banned it. And he determined purchasing and product placement within the stores. Finally, Trone purchased a single insurance policy and used one law firm for all of the stores.

Of particular relevance to the plaintiffs' claims are Trone's efforts in coordinating purchasing. Trone negotiated purchases of beer from wholesalers for all of the Beer World stores at once, obtaining an agreement that the Beer World stores could order together in order to obtain substantial volume discounts. The parties focus particularly on the negotiations between Trone and Fuhrer, who was the master distributor in the Pittsburgh area for Anheuser-Busch and Coors. All of Fuhrer's negotiations regarding the prices he would charge Beer World stores were conducted with Trone.

Even before the Beer World stores opened, beer wholesalers offered various quantity discounts, although they were relatively small. From September 1, 1987, until the end of 1989, pursuant to an agreement with Trone, Fuhrer implemented a $.25 per case discount for purchases of 4500 or more cases, a purchase amount substantially larger than that required for other, smaller volume discounts wholesalers offered. The Beer World stores were the only ones ever able to achieve this level of purchasing, which they did by ordering as a unit. Although each store would place separate orders that were delivered separately, they were placed in the name of Jet Distributors, one of the stores, in order to aggregate the order size to reach the 4500 case level. Each store's order was substantially less than this, usually in the range of 1000 cases. Although the plaintiffs attempted to take advantage of this discount, they were never able or permitted to do so.

Although the parties focus primarily on these quantity discounts, the plaintiffs allege that Trone was also able to obtain other benefits for the Beer World stores from wholesalers. For example, Fuhrer allegedly gave the Beer World stores a full-time employee, paid by Fuhrer, who stocked shelves at all of the stores. The plaintiffs further contend that Trone forced Fuhrer to sell him out-of-code beer, i.e., beer past its expiration/freshness date, at a discount. Apparently state law prohibits this and requires wholesalers to give retailers new beer in exchange for out-of-code beer. Trone allegedly got such beer at a discount and sold it while concealing the fact that it had expired from customers and inspectors sent by the beer brewers.

The plaintiffs criticize several aspects of these arrangements. First of all, they contend that Trone forced Fuhrer to agree not to give the discount to any other retailers. He allegedly could do so because, since the Beer World stores held a substantial portion (at least 25%) of the Pittsburgh beer market, Trone's threat to place Fuhrer's products in unfavorable locations within the stores carried force. Second, the plaintiffs point out that the Beer Worlds consistently received this discount even though their orders in the aggregate did not always reach the 4500-case level. In addition, many of the individual orders were fairly small:

29% were below 500 cases and 14% were below 200 cases, roughly the level at which the plaintiffs ordered. Finally, this discount was not disclosed to anyone else; it was not even included on Fuhrer's ordinary price lists.

In response to the defendants' actions, the plaintiffs instituted a state lawsuit against the defendants and convinced the Commonwealth to commence criminal proceedings. Neither of these actions achieved their desired results.

C. The Present Lawsuit

The plaintiffs filed the present lawsuit in March of 1992. Their primary claims include price fixing, engaging in a group boycott, and attempting and conspiring to monopolize the beer market in Pittsburgh, all in violation of the Sherman Act, 15 U.S.C. §§ 1 2, and civil RICO claims predicated on money laundering and mail fraud in connection with the license applications to the LCB, said to be a violation of 18 U.S.C. §§ 1341, 1956, 1962. They also brought various other claims that have been dismissed and not appealed or that we may dispose of summarily. Although the plaintiffs moved for class certification, this motion was denied, at which point some additional plaintiffs joined the suit.

These claims include price discrimination in violation of the Robinson-Patman Act, 15 U.S.C. § 13; common-law fraud; common-law conspiracy to defraud; and RICO violations predicated on mail fraud in the mailing of price lists by Fuhrer, 18 U.S.C. §§ 1341, 1962. The Robinson-Patman Act claim was dismissed early on, and the plaintiffs have not appealed from that dismissal. See Callahan v. A.E.V., Inc., Civ. A. No. 92-556, 1994 WL 682756 (W.D. Pa. Sept. 26, 1994). The plaintiffs' others claims are discussed in infra note 7.

The antitrust claims arise out of the joint operation of the Beer World stores. The plaintiffs contend that, by operating as a group, the Beer World stores were able to obtain an illegal competitive advantage. As evidence of such joint operation, they point to inter alia Trone's collective control of the stores, the aggregated orders through Jet Distributing, and coordinated advertising. The plaintiffs contend that this conduct violated the antitrust laws in several ways. First, the "quantity" discounts the Beer World stores were able to obtain are said to have constituted unfair price fixing, i.e., the price for other beer distributors was fixed at a level $.25 higher than that for the Beer World stores. Second, the discounts are claimed to have resulted in a group boycott, i.e., Beer World convinced the wholesalers to sell to the other distributors only on unfairly disadvantageous terms. Finally, the plaintiffs allege that all of the actions of Trone and the Beer World stores constituted an effort to monopolize the beer retail market in Allegheny County, which includes Pittsburgh. These efforts were aggravated by the fact that, pursuant to the Pennsylvania Liquor Code, the plaintiffs could only purchase beer through the single, designated master distributor for each brand for Allegheny County.

The RICO claim arises out of the various statements made during and concerning the Beer Worlds' efforts to obtain licenses from the LCB. First, various of the defendants and others allegedly lied about the true ownership of the Beer World stores in affidavits and other documents filed with the LCB via mailings in order to obtain and retain their licenses. Second, Trone and others allegedly lied before a grand jury investigating their operation when asked about the ownership of the Beer World stores. The plaintiffs contend that, as a result of this fraud, the Beer World stores were able to remain in business illegally under the control of Trone. Furthermore, Trone is said to have engaged in transactions involving the proceeds of this fraud, i.e., the income of the stores, by reinvesting the money in the stores, allegedly in violation of the money laundering statute. The plaintiffs contend that these various activities violated RICO.

D. The District Court's Rulings

Following extensive discovery, the parties each moved for summary judgment on various of the claims. The plaintiffs moved for summary judgment on their RICO claim relating to the Trone and Beer World defendants' statements to the LCB. The District Court denied the plaintiffs' motion because they did not "provide [any] substantive analysis of the meaning or application of § 1962 or its various subsections." Dist. Ct. Op. I, at 3.

Since we will affirm the District Court's judgment in favor of the defendants on the plaintiffs' RICO claim, we need not consider specifically whether it erred in denying the plaintiffs' motion for summary judgment.

The defendants moved for summary judgment on all of the plaintiffs' claims. The District Court, in a series of orders, granted the defendants' motions in part and denied them in part, and granted judgment in favor of the defendants on all of the plaintiffs' claims. First, the District Court dismissed part of the plaintiffs' RICO claim on statute of limitations grounds to the extent it was based on matters that occurred more than four years before the suit was filed. Second, the District Court dismissed all of the plaintiffs' remaining claims — the antitrust and RICO claims — because it concluded that the plaintiffs had not offered sufficient evidence of fact of damage, i.e., loss and causation in fact.

The District Court granted the defendants' motion for summary judgment on the RICO claim to the extent it was based on actions prior to March 1988, four years before the present suit was filed, because the plaintiffs should have been aware of the defendants' acts prior to that time. The plaintiffs contend that the District Court's conclusion erroneously rested on the fact that some of them filed a state lawsuit against Trone and the Beer World stores in 1986 alleging similar concerns, during which they could have obtained sufficient discovery to bring their present claims. They argue that their attorney misled them into believing they could not pursue their claim in that context, and that the statute of limitations should be tolled equitably.
We recently explained that attorney misconduct can give rise to equitable tolling only in unusual circumstances. See Seitzinger v. Reading Hosp. Med. Ctr., 165 F.3d 236, 240 (3d Cir. 1999). The plaintiffs contend that such unusual circumstances are present here, because their attorney allegedly was conflicted in that he also represented Fuhrer, and because, unlike Seitzinger, the lack of information on which to base a claim was at least arguably a result of the defendants' fraud. Furthermore, the plaintiffs note that, given the tremendous difficulties they faced in obtaining adequate discovery from the defendants in this case, the defendants cannot contend that the plaintiffs would have been able to obtain sufficient discovery in the previous state case. On the other hand, the defendants point out that, even if they fraudulently concealed certain facts, the plaintiffs were aware of those facts by the end of 1987. We need not decide this issue, because we will affirm the District Court's dismissal of the RICO claim in its entirety on other grounds.

As noted above, the plaintiffs brought additional RICO and common-law tort claims. In the same series of orders identified in the text, the District Court granted summary judgment on these claims in the defendants' favor. We will affirm those aspects of the judgment summarily.
The plaintiffs' common-law claims are that Fuhrer issued price lists that were fraudulent because they did not state the volume discount the Beer World stores received, and that Trone and Fuhrer conspired to misrepresent the prices through the same mechanism. Claims for common-law fraud and conspiracy are governed by a two-year statute of limitations. See 42 Pa. Cons. Stat. § 5524(7). The discount was discontinued at the end of 1989, and the plaintiffs were aware of the discount before then. The complaint was filed in March of 1992. Accordingly, the District Court concluded that more than two years had elapsed between the defendants' fraudulent acts and the filing of the complaint, and that the claim was therefore time-barred. Since the plaintiffs have not addressed this issue in the briefs (or, apparently, before the District Court), and the District Court's decision appears to be correct, we will affirm the District Court's judgment as to the common-law claims summarily.
The other RICO claim was based on Fuhrer's allegedly fraudulent mailing of price lists that did not include the $.25/case volume discount offered to the Beer Worlds. This discount was begun in September of 1987. Fuhrer did not mail a price list thereafter until March of 1988, and the plaintiffs were aware of the discount by October of that year. The District Court analyzed whether this constituted a "pattern of racketeering activity," 18 U.S.C. § 1962, in light of long-standing precedent. See, e.g., H.J., Inc. v. Northwestern Bell Tel. Co., 492 U.S. 229, 241 (1989). First, the Court concluded that this was not an open-ended pattern because, as Fuhrer discontinued the discount in 1989, the alleged fraud was unlikely to recur. Second, the Court found that fraud of six months' duration could not constitute a closed-ended pattern. See, e.g., Tabas v. Tabas, 47 F.3d 1280, 1293 (3d Cir. 1995) ("Since H.J., Inc., this court has faced the question of continued racketeering activities in several cases, each time finding that conduct lasting no more than twelve months did not meet the standard for closed-ended continuity." (citing cases)). Because the plaintiffs could prove no pattern of racketeering activity, the District Court concluded that they could not bring a successful RICO claim based on the price lists. Since the plaintiffs have not discussed this issue in their briefs and the District Court's reasoning is persuasive, we will affirm the District Court's judgment in favor of the defendants on this other RICO claim, also summarily.

Summary judgment "shall be rendered forthwith if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law." Fed. R. Civ. P. 56(c). Although the plaintiffs must prove loss, causation and specific damages, at the summary judgment stage, the court's main concern should be with determining loss and causation in general, rather than proof of specific amounts of damages:

At this procedural juncture, reviewing the district court's grant of summary judgment, we are not, as we would be upon reviewing a jury verdict, determining whether a plaintiff has brought sufficient evidence to justify the actual damages awarded. Rather, here, all we are concerned with is whether Rossi has established that the defendants' illegal conduct was a material cause of [his] injury.

Rossi v. Standard Roofing, Inc., 156 F.3d 452, 484 (3d Cir. 1998) (citation and quotations omitted); see also Stelwagon Mfg. Co. v. Tarmac Roofing Sys., Inc., 63 F.3d 1267, 1276 n. 19 (3d Cir. 1995) (declining to consider whether the plaintiff had offered sufficient proof of the amount of damages, since the plaintiffs' proof of loss in general was inadequate).

On appeal, our review of a District Court's grant of summary judgment is plenary. See In re Baby Food Antitrust Litig., 166 F.3d 112, 123 (3d Cir. 1999). "We evaluate the evidence using the same standard the District Court applied in reaching its decision." 166 F.3d at 123-24.

The District Court had subject matter jurisdiction pursuant to 28 U.S.C. § 1337 and 1367, as well as 15 U.S.C. § 15. We have appellate jurisdiction pursuant to 28 U.S.C. § 1291.

II. Antitrust Claims: Antitrust Liability

In the ordinary case, liability is the first question that must be decided. Accordingly, we would usually begin our analysis of this case with a discussion of whether the plaintiffs have produced sufficient evidence to prove that the defendants violated the Sherman Act. Although that would appear to be an obvious question in this case, for the reasons set forth below we are not presently in a position to evaluate the plaintiffs' theory of antitrust liability. We will, however, briefly summarize that theory and the defendants' arguments against it in order to provide a background for our discussion of fact of damage, and for the benefit of the District Court and the parties on remand.

The plaintiffs' antitrust claims begin with the premise that Trone coordinated the activities of all of the Beer World stores. In support of this contention, they note that Trone dictated most aspects of store policy, was in charge of hiring and managing employees, and had sole control of the stores' accounts. In addition, Trone coordinated the stores' interactions with other people, including wholesalers and customers. He negotiated a single set of wholesale prices for all of the Beer World stores. When one wholesaler would not agree to a discount, he organized a joint advertising campaign among the stores against the wholesaler. He also published joint advertising for the stores.

Furthermore, Trone and the stores allegedly conspired with wholesalers, Fuhrer in particular, so that the stores could obtain a competitive advantage over other retailers. Most prominently, the plaintiffs allege that Trone convinced Fuhrer to grant the stores a volume discount $.25/case lower than that available to any other retailer. This discount was concealed from other customers and wholesalers in several ways, and denied to the customers when they requested it. The Beer World stores' orders pursuant to the discount were placed jointly. Furthermore, the discount was always given even though the minimum order required for the discount was not always met by the Beer World stores in the aggregate. In addition, the plaintiffs contend that the evidence shows that Fuhrer granted the stores other advantages, including special delivery terms and assistance in placing beer in the stores.

The plaintiffs contend that the advantages the Beer World stores obtained caused losses to the plaintiffs. As a result of the advantages, the Beer World stores were able to undersell the plaintiffs. Accordingly, the plaintiffs contend, they lost customers to the Beer World stores. The plaintiffs submit that these harms were particularly aggravated because of the geographical limitations the Liquor Code places on distributors. The Code requires that, for each brand of beer sold in a particular area, a specific wholesaler be designated as the master distributor. A beer retailer within that geographic area, must buy that brand either from the master distributor, or from someone who bought it from the master distributor. Since the plaintiffs allege that the defendants were conspiring with the master distributors, they were at a particular competitive disadvantage.

Although, as noted above, the plaintiffs identify several antitrust liability theories, they focus on one in particular in their briefs. They argue that the aforementioned actions constitute a group boycott on the part of Trone, the Beer World stores, and Fuhrer. They contend that Trone convinced Fuhrer to agree to sell beer to the Beer World stores at a lower price than would be available to any other retailer. They rest their legal theory on, inter alia, Klor's, Inc. v. Broadway Hale Stores, Inc., 359 U.S. 207 (1959), and Rossi v. Standard Roofing, Inc., 156 F.3d 452 (3d Cir. 1998).

The defendants contend that the plaintiffs' theory of antitrust liability is untenable for several reasons. First, they argue that the plaintiffs' theory is simply a Robinson-Patman Act price-discrimination claim recast as a Sherman Act claim. They note too that the plaintiffs did bring a Robinson-Patman Act claim that was dismissed on jurisdictional grounds. The defendants also submit that price discrimination without much more cannot be a violation of the Sherman Act.

We agree that price discrimination simpliciter — even when it violates the Robinson-Patman Act — is usually not a Sherman Act violation. But we do not think this necessarily means that the plaintiffs are barred from bringing a price discrimination claim under the Sherman Act. The plaintiffs' claims are unlike an ordinary price discrimination case, in which a single supplier offers different prices to different purchasers in order to advance its own interests. They allege that Fuhrer was convinced to offer different prices in order to advance the defendants' — the plaintiffs' competitors — interests. We see no reason why price discrimination, under appropriate circumstances, could not be part of an agreement in restraint of trade or a monopolization attempt. See, e.g., Black Gold, Ltd. v. Rockwool Indus., Inc., 729 F.2d 676, 683-84 (10th Cir. 1984); Peelers Co. v. Wendt, 260 F. Supp. 193, 198 (W.D. Wash. 1966); McKeon Constr. v. McClatchy Newspapers, Civ. No. 51627, 1969 WL 226 (N.D. Cal. Nov. 24, 1969). So long as the price discrimination involves a conspiracy to restrain trade or create a monopoly in some market — along with a substantial effect on competition in the market, see J.F. Feeser, Inc. v. Serv-A-Portion, Inc., 909 F.2d 1524, 1541 (3d Cir. 1990) (quoting Zoslaw v. MCA Distributing Corp., 693 F.2d 870, 887 (9th Cir. 1982)); see also United States v. Arnold, Schwinn Co., 388 U.S. 365, 375 (1967) — it would violate the Sherman Act. The proper evidence in this case might support the conclusion that this constituted a conspiracy or agreement to restrain trade or create a monopoly, although we express no opinion as to whether the plaintiffs have produced such evidence.

The defendants also contend that the plaintiffs cannot prove that they engaged in a group boycott. Relying on Klor's and Rossi, they submit that a group boycott only exists where the defendants' actions result in the product's not being available to the plaintiffs at all, or only being available at highly unfavorable terms. Of course, when one thinks of a boycott, one ordinarily thinks of preventing access to something entirely. Moreover, the defendants contend that the putative quantity discount is modest. The plaintiffs respond, however, that the evidence here is sufficient to conclude that, as a result of the defendants' actions, beer was only available to them on highly unfavorable terms, i.e., $.25/case more than their competitors were paying.

Finally, the defendants contend that the antitrust violations were limited to a narrow array of conduct, specifically the $.25/case discount discussed above. Plaintiffs contest this point vigorously. They suggest that, solely with respect to Fuhrer, the evidence supports the conclusion that he engaged in other activities over a longer period of time, including delivery and product placement assistance, that gave the Beer World stores an advantage. Furthermore, the plaintiffs point to evidence that suggests that other wholesalers were giving the Beer World stores discounts and other benefits throughout a substantially broader time frame.

The District Court did not address these questions of antitrust liability because it thought it could dispose of the case on other grounds. In part, this may have been because the Court came to the case late, upon transfer of the case from the docket of another judge. In addition, it undoubtedly seemed to it to be a more straightforward way in which to dispose of the case. We imply no criticism of the District Court's approach. As discussed further below, however, liability is not an issue that ultimately can be avoided in this case. The defendants have suggested that it is an appropriate alternative grounds upon which we can rest our judgment, but we do not think so. Although the parties have set forth in their briefs their legal analyses of the liability questions, the record as presented to us is not sufficiently adequate for us to give the careful and thorough consideration these issues merit. Since the case must go back to the District Court, we think these issues would benefit from further elaboration there in the first instance.

We also note that the present plaintiffs' counsel came to the case late as well, after much of its present contours had been fixed.

On remand, in determining whether the plaintiffs can prove that the defendants violated the Sherman Act, the District Court can answer the questions discussed above. The Court will be able to determine under which of their variegated antitrust theories the plaintiffs may proceed. In addition, the Court can clarify the precise temporal scope and nature of the defendants' antitrust violations. Explication of this last issue in particular will provide a better framework for more precise analysis of the questions to which we turn next (and which will remain a matter in controversy on remand). At this juncture, because of the lack of clarity concerning the precise nature and scope of the plaintiffs' antitrust liability proofs, we will assume that the plaintiffs can prove that the defendants engaged in antitrust violations throughout the relevant period. Based on this assumption, we turn to the issue upon which the District Court rested its decision: whether the plaintiffs have offered sufficient proof of fact of damage.

III. Antitrust Claims: Fact of Damage

The primary issue actually before us on the antitrust claims is whether the plaintiffs have proffered sufficient evidence to raise a genuine issue of material fact as to whether the defendants' alleged antitrust violations caused harm to the plaintiffs. "[A] plaintiff must prove a causal connection between [the antitrust violation] and actual damage suffered." Stelwagon Mfg. Co. v. Tarmac Roofing Sys., Inc., 63 F.3d 1267, 1273 (3d Cir. 1995); see also Rossi v. Standard Roofing, Inc., 156 F.3d 452, 483 (3d Cir. 1998) ("To recover damages, an antitrust plaintiff must prove causation, described in our jurisprudence as `fact of damage or injury.'" (citations omitted)); II Phillip E. Areeda Herbert Hovenkamp, Antitrust Law P 360c2, at 195 ("The plaintiff must show actual injury that was `caused' by the violation."). Although we suspect our factual analysis of loss and causation would apply equally to both the plaintiffs' antitrust and RICO claims, we will focus in this section only on the former. We can put the RICO claim to the side because, although we are unsure that the District Court's reasons for dismissing it was correct, we think they should be dismissed for other reasons, i.e., lack of proximate causation.

In brief, the plaintiffs' theory of antitrust fact of damage is as follows: Trone, the Beer World stores, and Fuhrer engaged in various joint actions, including but not limited to granting the Beer World stores secret discounts on wholesale purchases, which resulted in the plaintiffs' losing business. In support of this theory of fact of damage, the plaintiffs offer two types of evidence: (1) testimony concerning customers who no longer shop at the plaintiffs' stores and their statements about their reasons for not doing so; and (2) expert opinion testimony concerning the cause of the plaintiffs' loss of income. We must decide whether the former type of evidence is admissible, and whether either is sufficient, individually or together, to establish actual injury and causation in fact.

Given our conclusion that the customer and expert evidence of causation is sufficient, we need not consider the plaintiffs' other arguments for reversing the District Court's conclusion that they had not adduced sufficient evidence of causation: (1) that a price differential permits an automatic presumption of causation of loss to those who pay the higher price, see Bogosian v. Gulf Oil Corp., 561 F.2d 434, 455 (3d Cir. 1977), and (2) that the defendants' own statements and "admissions" constitute proof of causation.

A. Customer Evidence

1. Summary of the Evidence: In opposition to the defendants' motion for summary judgment, the plaintiffs offered deposition testimony concerning their customers. It included testimony of various plaintiffs that certain customers ceased purchasing beer from them after the Beer World stores opened, and that the customers stated that they had done so because the Beer World stores had cheaper beer. The District Court concluded that this testimony was inadmissible hearsay and therefore could not meet the plaintiffs' burden of production to defeat the defendants' motion for summary judgment.

Five of the plaintiffs offered testimony concerning customers' behavior and statements. This testimony can be divided into two categories. First, several of the plaintiffs testified that, during the time at issue in this litigation, some people who had formerly been their customers stopped coming to their stores. Carl Altenhof testified that, "Retail customers that I had as steady customers, I don't have anymore when Beer World came in. . . ." App. at 750. Likewise, Douglas J. Berthold stated in his deposition that, although he could not document his losses, he had"lost forty percent of [his] business, probably most of them are one case purchase customers, some of them two case purchase [sic]." App. at 754. Finally, Kathleen Kapres said that she lost customers, purportedly to Beer World. App. at 819.

Second, several of the plaintiffs testified that various customers, some identified and some not, told them that they no longer shopped at the plaintiffs' stores because of the Beer World stores' operations. Berthold testified that one customer, David Begg, told him that he was going to shop at Beer World because "I like selection" and "money talks." App. at 755. Kapres also stated that she"had quite a few customers come in and say they wanted the same deal [lower prices] from me or they were just going to buy their beer from [Beer World], and I said I just can't give you that deal." App. at 819. In addition, Paul Kelly identified by name three customers of his who began to buy from Beer World, and discussed at length conversations with one of them in which the customer revealed that he was going to Beer World because of the prices. App. at 822-27.

As noted previously, the defendants contend that the plaintiffs have presented evidence of antitrust violations at most during a fairly brief period of time, for which only some of the customer evidence is relevant. The District Court did not consider this issue and, as we have stated, neither will we. Instead, we assume that the plaintiffs can establish antitrust violations throughout the relevant period. On remand, the District Court will have to analyze the extent of the defendants' antitrust violations and then determine whether the plaintiffs' evidence of loss and causation remains sufficient in light of the more specific temporal scope. If it appears that the defendants did not engage in antitrust violations during some of the relevant period, the District Court is free to revisit the question whether the plaintiffs' proof of causation remains sufficient.

2. Admissibility: The District Court, relying on our decision in Stelwagon Manufacturing Co. v. Tarmac Roofing Systems, Inc., 63 F.3d 1267 (3d Cir. 1995), held this evidence inadmissible, finding that the plaintiffs' testimony concerning their customers' statements was inadmissible hearsay. It also noted that, although this litigation has been proceeding for some six years, the plaintiffs had not taken the simple step of obtaining affidavits from customers concerning their reasons for ceasing to purchase beer from the plaintiffs. We disagree with the District Court's reading of Stelwagon.

In Stelwagon, the plaintiff proffered the testimony of its employees concerning the statements of their customers.

The employees proposed to testify, based on "out-of-court conversations with Stelwagon customers . . . that the customers could and did purchase Tarmac MAPs from Standard at prices lower than Stelwagon's prices."

Stelwagon, 63 F.3d at 1274. The plaintiff argued that this testimony was admissible to prove fact of damage, i.e., both loss and causation, under Federal Rule of Evidence 803(3), which provides:

The following are not excluded by the hearsay rule, even though the declarant is available as a witness: . . . A statement of declarant's then existing state of mind, emotion, sensation, or physical condition (such as intent, plan, motive, design, mental feeling, pain, and bodily health), but not including a statement of memory or belief to prove the fact remembered or believed unless it relates to the execution, revocation, identification, or terms of declarant's will.

Fed. R. Evid. 803(3).

In Stelwagon, the plaintiff offered the customers' statements to prove, not only causation, i.e., the reason it lost business — for which purpose it would be admissible evidence of motive under Rule 803(3) — but also loss, i.e., the fact that it lost business to the defendants. We concluded that the customers' statements about why they purchased from Standard was inadmissible to prove that they actually did so. See 63 F.3d at 1274 ("Statements that are considered under the exception to the hearsay rule found at Fed. R. Evid. 803(3) . . . cannot be offered to prove the truth of the underlying facts asserted." (footnote omitted)). As we have explained, "`[s]tatements of a customer as to his reasons for not dealing with a supplier are admissible for this limited purpose,' i.e., the purpose of proving customer motive, but not as evidence of the facts recited as furnishing the motives." J.F. Feeser, Inc. v. Serv-A-Portion, Inc., 909 F.2d 1524, 1535 n. 11 (3d Cir. 1990) (quoting Herman Schwabe, Inc. v. United Shoe Mach. Corp., 297 F.2d 906, 914 (2d Cir. 1962)).

We think that the District Court's dismissal of the plaintiffs' evidence on the basis of Stelwagon was inappropriate. The purpose for which the customers' statements are offered in this case differs in substance from the purpose for which the court in Stelwagon found them inadmissible. In that case, the only evidence of actual loss, i.e., that customers stopped purchasing from the plaintiff, was the employees' reports that customers had said that they were no longer buying from the plaintiff because the plaintiff's competitors had lower prices. We concluded that this evidence could not be used to prove such loss. While the plaintiffs here have also offered similar testimony that their customers told them that they were purchasing beer from the Beer World stores and not the plaintiffs, they offer it only for "the [limited] purpose of proving customer motive," for which purpose we found such evidence admissible under Rule 803(3). Stelwagon, 63 F.3d at 1274.

The defendants also complain that, even if the testimony is otherwise admissible under Rule 803(3), it is not admissible, particularly for use at the summary judgment stage, because the declarants are unidentified or inadequately identified. First, the defendants contend that this means the evidence cannot meet the plaintiffs' burden to defeat a motion for summary judgment because it is not in an admissible form. See Fed. R. Civ. P. 56(e) ("Supporting and opposing affidavits [introduced at the summary judgment stage] . . . shall set forth such facts as would be admissible in evidence. . . ."); Petruzzi's IGA Supermarkets, Inc. v. Darling-Delaware Co., 998 F.2d 1224, 1234 n. 9 (3d Cir. 1993) (noting that evidence introduced to defeat a motion for summary judgment must be "capable of being admissible at trial"). In particular, the defendants argue that, since the declarants are unidentified, there is no way to ensure that they will be able or willing to testify at trial. We need not consider this issue, however, because the statements are admissible hearsay as discussed in the text.
Moreover, contrary to the defendants' apparent suggestion, we do not think that the fact that the declarants are not specifically identified is relevant for determining whether their statements fall within the Rule 803(3) hearsay exception. The defendants cite Philbin v. Trans Union Corp., 101 F.3d 957 (3d Cir. 1996), for the proposition that a hearsay statement by an unidentified or unknown person "is not `capable of being admissible at trial.'" Philbin, 101 F.3d at 961 n. 1 (quoting Petruzzi's, 998 F.2d at 1234 n. 9). Philbin is distinguishable, however. The plaintiff in that case relied on the statement of an unidentified official of the defendant as direct evidence of the defendant's allegedly unlawful motive in a Fair Credit Reporting Act suit. The declarant's identity was important to ensure that he or she was in fact an official of the defendant company whose statement would be admissible nonhearsay under Fed. R. Evid. 801(d). Identity was a critical element of admissibility.
The customers' statements in this case, however, are different. In a practical sense, their identities are not important. The relevance of their statements depends only on the fact that they were the plaintiffs' customers, not their particular identities. Furthermore, we do not think that the admissibility of their statements under the Rule 803(3) hearsay exception depends on their being identified. Knowing the specific identity of the declarant will not make the statements more trustworthy evidence of the declarants' descriptions of their states of mind, the primary concern in interpreting hearsay exceptions.
In United States v. Mitchell, 145 F.3d 572, 576-77 (3d Cir. 1998), we held that the identity of the declarant is a substantial, although not determinative, factor in determining whether a hearsay statement is admissible under the present sense impression or the excited utterance exceptions to the hearsay rule. The proposed evidence in that case was an anonymous note purportedly identifying a getaway car. We held that the note was not admissible as a present sense impression or excited utterance because there was no evidence that the unidentified declarant personally perceived the event or condition about which the statement is made. With respect to a state-of-mind statement, however, it is only important that the declarant be the person whose state of mind the statement concerns, which is true by definition.

In addition, however, the record contains other non-hearsay evidence of a type not before the court in Stelwagon, that the plaintiffs offer to prove the fact of loss, the issue for which the court in Stelwagon found the customers' statements inadmissible. Here, the plaintiffs themselves testified that they knew of customers who used to purchase beer from them, but no longer did. This is direct evidence of an actual loss of customers. Although in Stelwagon we held that customers' hearsay statements were not admissible to prove lost business, the plaintiffs' own testimony about the actual behavior of their customers is not hearsay. Rather, it is admissible evidence of lost business, although not of the reason therefore. Thus, in the present case, the plaintiffs' testimony that certain customers no longer purchased beer from them, coupled with their testimony concerning the customers' statements of their motive, which is admissible hearsay under Rule 803(3), are together evidence of the fact of damage.

3. Sufficiency of the Evidence to Prove Causation: The next question is whether this evidence is sufficient to defeat a motion for summary judgment. "[O]ur jurisprudence does not require the summary judgment opponent to match, item for item, each piece of evidence proffered by the movant, but rather he or she must only exceed the`mere scintilla' standard." Rossi, 156 F.3d at 466 (citations and some quotations omitted). We recently confronted the question of the sufficiency of this sort of evidence of causation in antitrust cases. In Rossi, the plaintiff offered, in opposition to the defendants' motion for summary judgment, the testimony of several potential customers that they would have purchased a certain product from him if he had not been deprived of it in violation of the antitrust laws. We concluded that this evidence was sufficient evidence of fact of damage to defeat a motion for summary judgment:

Rossi has proffered evidence from five specific customers that they would have purchased GAF product from Rossi if he had been able to sell it to them, and Rossi's inability to consummate those sales (leading to a loss of business and therefore injury) is a direct result of the alleged antitrust violation — the group boycott. In addition, Richard Droesch, Rossi's partner in the failed Rossi Florence venture, backed out of that venture at least in part based upon his understanding that the company would not be able to get the products it needed, particularly GAF product, to compete successfully in the market. For all these reasons, we believe that the record supports Rossi's allegations that he suffered antitrust injury, and that it was caused by the defendant's [sic] allegedly unlawful actions.

156 F.3d at 485. We think that Rossi supports the conclusion that the plaintiffs' testimony concerning their customers' actions and statements is sufficient to meet their burden to produce evidence of loss and causation.

Initially, we reject the defendants' attempt to distinguish Rossi on the ground that the customers there stated that they would have purchased product from Rossi but for circumstances that were the direct and intended result of the conspiracy. As noted previously, we have had to assume for the purposes of this appeal that the plaintiffs will be able to prove that the defendants violated the antitrust laws. The direct result of these violations would be the Beer World stores' ability to sell beer at a lower price than the plaintiffs, the precise circumstance the customers cited as a reason for their actions.

The defendants also submit that Rossi is distinguishable because in this case there was no admissible evidence that the customers purchased beer from the Beer World stores. Of course, the defendants are correct that the testimony at issue is not admissible to prove that the customers purchased beer from the defendants. See Stelwagon, 63 F.3d at 1274. But the plaintiffs do not need to prove that point; in order to establish antitrust liability and damages, all the plaintiffs must show is that they suffered an economic loss as a result of the defendants' antitrust violations; not that the defendants benefitted from that loss directly. See Rossi, 156 F.3d at 464-65 (plaintiff, in order to recover on an antitrust claim, must prove an antitrust violation and "that the plaintiffs were injured as a proximate result of that" violation (citation omitted)). As long as the plaintiffs can prove that they lost business, and that this loss was a result of the defendants' antitrust violations, they can bring a successful antitrust claim.

At all events, Rossi makes no mention of any evidence, or even any requirement, that the customers in that case purchased product from the defendants instead of the plaintiff. See Rossi, 156 F.3d at 485. For all we know, the customers who offered testimony in Rossi simply decided not to purchase GAF product at all, instead of buying it from the defendants. What the customers did instead of purchasing product from the plaintiff is irrelevant, so long as there is evidence that they did not purchase from the plaintiff because of the defendants' antitrust violations.

In addition to these points, we also find it significant that neither here nor in Rossi did the customer evidence purport to prove any specific amount of damages. Although the plaintiff in Rossi proffered the testimony of five customers, these customers gave no indication of exactly how much product they would have bought from him if they could. Yet we concluded that the evidence of causation was sufficient. This conclusion was driven by the principle that, on review of a grant of summary judgment, we should focus on whether there is sufficient evidence of fact of damage in general, not on the sufficiency of the evidence of a specific amount of damages. See Rossi, 156 F.3d at 484. Accordingly, just as in Rossi, the lack of specific evidence of the total amount of lost beer sales does not preclude our ultimate conclusion that the customer evidence, especially in conjunction with the expert evidence discussed next, is sufficient to meet the plaintiffs' burden of producing evidence of fact of damage.

B. Expert Evidence

The plaintiffs also offered expert opinion evidence in support of their contention that the defendants' alleged antitrust violations caused actual injuries to them. In particular, they offered the report and testimony of their primary expert, Garth Seidel, along with the report and testimony of their rebuttal expert, Brian Sullivan, to that effect. Neither the defendant nor the District Court raised a question about the admissibility of Seidel's or Sullivan's opinion, see Kumho Tire Co. v. Carmichael, 119 S.Ct. 1167 (1999); Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993). The District Court concluded, however, that this evidence was insufficient as a matter of law to permit a finding of causation. In particular, the Court noted that Seidel's opinion appeared to be based primarily on timing, and that he did not consider a number of possible alternative causes of the plaintiffs' losses. It therefore concluded that Seidel's report was deficient under the standards we have set forth in previous cases. We disagree.

1. Seidel's Report: Seidel concluded, based on the facts provided to him, that "the plaintiffs experienced significant drops in gross profits in the period subsequent to the start of Beer World operations and which were caused by the Beer World Stores' unique advantage." App. at 830. He began by collecting data on gross profits of the sixteen plaintiffs from 1980 to 1995, although such information was not available from every plaintiff for every year, or even for many of the years. He then made two calculations. First, using the admittedly incomplete data he had, he calculated the plaintiffs' average annual gross profits for the periods from 1980-84 and 1985-95. He then estimated that the plaintiffs' damages were the difference between these two numbers, multiplied by eleven years and sixteen plaintiffs, or approximately $6.6 million. Second, he made a similar calculation, but using only data from the six plaintiffs for whom data was available for most of the years. This method gave a damages estimate of $2 million.

Next, he concluded that these lost profits "were caused by the Beer World Stores' unique advantage." App. at 830. He based this opinion initially on his conclusion that the Beer Worlds' ability to purchase beer at a lower wholesale cost "must have had a significant impact on the market." App. at 831. He also stated that the Beer World stores' aggressive price advertising would have magnified the effect of the special discount. In addition, he examined Fuhrer's profits between 1989 and 1994, and observed that they increased substantially during this period. Based on this, Seidel concluded that the malt-beverage market experienced no downturn during this time. Third, he noted that, beginning in 1991 — after the grand jury investigation of Trone began — the Beer World stores' volume of business declined each year until 1995, while at the same time the plaintiffs' gross profits increased. Finally, he noted that two other stores had opened using a "supermarket" approach similar to the Beer Worlds'. One of them opened shortly after the Beer World stores, but failed within a matter of months in spite of aggressive promotions. The other was open from the mid-1970's to the mid-1990's, but did not appear to have had any effect on the plaintiffs. App. at 829-32.

The District Court found Seidel's report inadequate to meet the plaintiffs' burden of production because it failed to consider other market forces that could have explained the plaintiffs' losses. The Court began with the proposition that "any analysis of antitrust, RICO or similar damage that fails to exclude or take account of any adverse effects caused by other factors, including lawful competition on the part of the defendants, is fatally flawed." Dist. Ct. Op. IV, at 6. It observed that Seidel's report did not include a comparison of costs and business practices, price advertising, the availability of pool-buying and other discounts, store size, purchasing capacity, or proximity to a Beer World store, any one of which might have provided an alternative explanation for the plaintiffs' losses. Furthermore, it noted that Seidel had specifically failed to consider whether any other differences between the plaintiffs and the Beer World stores accounted for the plaintiffs' loss of business to the latter. Given these omissions, the District Court concluded that Seidel's report provided insufficient evidence of causation.

The plaintiffs contend that we must reverse the judgment because the District Court relied, in its legal analysis, on the district court's opinion in Rossi rejecting Rossi's proffered expert evidence, which opinion we later reversed on these exact grounds, although not until well after the District Court in the present case had issued its opinions. See Rossi v. Standard Roofing, Inc., 958 F. Supp. 976 (D.N.J. 1997), revd., 156 F.3d 452 (3d Cir. 1998). As this question is before us on an appeal from a grant of summary judgment and our review is plenary, we will start from the premise that it is the defendants's burden to show that Seidel's report is inadequate to create a genuine issue of material fact as to causation. We begin with a review of our case law in this area, and then apply that law to the evidence before us.

2. Precedent: Stelwagon and Rossi: We have twice recently considered the sufficiency of expert evidence offered as proof of causation in antitrust cases. See Rossi, 156 F.3d at 485-87; Stelwagon, 63 F.3d at 1275-76. In addition to the customer evidence discussed above, the plaintiff in Stelwagon offered expert opinion evidence to prove causation. In brief, "based on the assumption that but for Tarmac's price discrimination, Stelwagon's sales of MAPs would have tracked its sales of [other] products [not subject to anticompetitive practices], Dr. Perry concluded that Stelwagon lost $257,000 in profits as a result of Tarmac's illegal pricing policy." Stelwagon, 63 F.3d at 1275.

More specifically, the expert's report based its analysis on two premises. First, it assumed that Stelwagon's sales of the product at issue during the year in which it did not allegedly suffer antitrust harm were representative of what the sales would have been in the absence of such harm. Second, the expert assumed that its sales of this product would follow a pattern similar to that of other products Stelwagon sold. Finally, the expert posited that Stelwagon would have been able to charge the same retail markup on the product at issue as it did on other products. Based on these assumptions, the expert calculated Stelwagon's lost sales and profits. See Stelwagon, 63 F.3d at 1275.

We concluded that the expert's testimony, although admissible evidence, was insufficient by itself to prove that the antitrust violations had in fact caused Stelwagon's losses:

Significantly, Dr. Perry's analysis failed to sufficiently link any decline in Stelwagon's MAPs sales to price discrimination. The sales may have been lost for reasons apart from the price discrimination — reasons that Dr. Perry's analysis apparently did not take into account. For example, the evidence showed that Stelwagon had higher overhead costs than his competitors. In addition, there was undisputed evidence that Stelwagon experienced other business complications during the relevant time period. In 1988, for example, Stelwagon terminated a vice-president, two territorial managers and three key employees for their part in an embezzlement scheme.

Stelwagon, 63 F.3d at 1275. Given that Stelwagon had not offered any other evidence of loss — as discussed previously, its employees' anecdotal testimony concerning lost customers was not admissible to prove that it actually lost customers, see Stelwagon, 63 F.3d at 1274-75 — we concluded that he could not meet his burden of proof, Stelwagon, 63 F.3d at 1275-76.

In Rossi, by contrast, we considered an expert opinion and found it sufficient to prove loss and causation. The expert in Rossi rested his calculation of damages on two assumptions:

First, he estimated that Rossi[`s businesses] would have achieved the same pattern of sales revenues (and revenue growth) beginning in 1989 and extending to 2008 that ABC's Morristown sales branch actually achieved from 1990-93, operating out of the same location, with Rossi as branch manager. . . . The second major assumption in the Rockhill Report is that Rossi would have been able to manage [his proposed businesses] in the manner that he had run Standard's Morristown branch from 1984-87. Rockhill used Standard's Morristown branch financial statements to develop 14-year averages for [costs] and applied them to the sales estimate.

Rossi, 156 F.3d at 486 (footnote omitted). Based on these assumptions, the expert estimated Rossi's losses as a result of the defendants' antitrust violations.

We determined that this expert evidence was sufficient proof of causation to defeat a motion for summary judgment. We began our analysis with the recognition that the expert's opinion was a "but for" damage model — one that "aggregates the defendant's alleged violations and creates a hypothetical calculation projecting the plaintiff's profits and losses `but for' the defendant's antitrust violations" — which several courts have rejected. Rossi, 156 F.3d at 485 (citing Southern Pac. Com. Co. v. American Tel. Tel. Co., 556 F. Supp. 825 (D.D.C. 1982), affd., 740 F.2d 980 (D.C. Cir. 1984)); Van Dyk Research Corp. v. Xerox Corp., 478 F. Supp. 1268 (D.N.J. 1979), affd., 631 F.2d 251 (3d Cir. 1980)). We identified two key problems with the use of "but for" damage models:

First, they do not attempt to measure the particularized effects of any specific alleged illegal activities, but rather rely on an aggregation of injury from all factors. Second, their hypothetical "but for" calculations usually rely upon unrealistic ex ante assumptions about the business environment, such as assumptions of perfect knowledge of future demand, future prices, and future costs that tend to overstate the plaintiff's damages claim. Thus, using a "but for" damage model arguably makes it impossible for the trier of fact to determine what, if any, injury derived from the defendant's antitrust violations as opposed to other factors, and courts sometimes reject such models as the basis of either causation or the amount of injury.

Rossi, 156 F.3d at 486 (citations omitted).

We concluded that, although the Rockhill Report rested on a "but for" damage model, this did not mean it was inadequate proof of causation, because it did not have the usual problems of "but for" damage models. We noted that, since the Report was based on the actual performance of other businesses — the business Rossi managed instead of running his own and the business he formerly managed — it did not involve any "unrealistic ex ante assumptions about the business environment." We concluded that, "This kind of estimate, while perhaps not one upon which we would base our own personal investment decisions, nevertheless is sufficient to establish causation. . . ." Rossi, 156 F.3d at 485.

We also rejected the defendants' argument, upon which the district court had rested its decision, see Rossi, 958 F. Supp. at 991, that the Rockhill Report was inadequate because it failed to consider possible alternative causes of Rossi's losses. In particular, the defendants contended that Rossi's businesses failed because: "(1) they were start-up operations, (2) they were founded during one of the worst recessions ever to hit the New Jersey housing market, (3) Rossi, as a manager, failed to control his costs, and/or (4) Rossi worked on other ventures to the detriment and ultimate failure of both companies." Rossi, 156 F.3d at 486. Although we recognized that these explanations might ultimately prove to be correct, we found that they were issues of fact best left to the jury, not reasons for concluding that the Rockhill Report was insufficient evidence of causation as a matter of law.

3. Application to Seidel's Report: We believe that our jurisprudence supports the conclusion that the plaintiffs, by offering Seidel's report, have produced sufficient evidence of causation to defeat a motion for summary judgment. Here, as in Rossi, Seidel's report rests on assumptions that are based on past performance, not guesses as to the future. His opinion was based on the assumption that the plaintiffs' performance in the years before Beer World entered the Pittsburgh market provided an appropriate benchmark for their performance thereafter. This assumption does not rest on any "assumptions of perfect knowledge of future demand, future prices, and future costs" of the sort we condemned in Rossi. At most, it requires some consideration of whether general economic conditions were substantially similar before and after the Beer World stores opened. Seidel's observation that Fuhrer's sales increased substantially during the period after the Beer World stores opened strongly suggests that economic conditions were at least as good during this period. See App. at 831.

The defendants, in response, identify several problems that they believe render Seidel's report inadequate. In general, the defendants criticize Seidel's report for failing to take into account potential alternative causes for the plaintiffs' losses not attributable to the defendants' actions. Seidel stated that, "In my opinion, it does not appear that the losses in plaintiffs' profits . . . were caused by market factors other than plaintiffs' competition with the Beer World stores in the face of the availability to the Beer Worlds of unique discounts and special services, such as free delivery." App. at 831. The defendants contend that, just as we found in Stelwagon that the expert's report was inadequate because it failed to consider alternative causes, so must we find Seidel's report inadequate because he failed to consider certain specific factors that might have affected the plaintiffs' business success, such as general economic conditions, changes in their operations during the relevant time period, or changes in costs.

The defendants also contend that, even if Seidel has shown that the defendants' acts caused the damages, he has not shown that the defendants' illegal acts caused damages. The defendants are correct that Seidel failed to account for many variations in what the defendants did during the periods that Seidel aggregated for analysis. We cannot evaluate this contention without a clearer picture of the scope of the defendants' antitrust violations, however. In the absence of such a clarification, we will leave it to the District Court to consider this objection in the first instance after examining the plaintiffs' antitrust liability theories in greater depth. Only after such a closer examination can this criticism of Seidel's report be given adequate consideration.

Initially, we note that, as discussed above, Seidel does discuss some of these factors the defendants suggest he should have — including general economic conditions — albeit not to the degree the defendants might prefer. In addition, he specifically noted a correlation between declining profits for the Beer World stores and increasing profits for the plaintiffs after the criminal indictment came down in 1991. Furthermore, we think that the factors Seidel failed to explain are more like those at issue in Rossi, in which we found the export's report acceptable in spite of certain gaps, than the factors in Stelwagon. In the latter case, the expert failed to discuss certain factors — higher overhead costs and embezzlement by the plaintiff's employees — about which the defendants introduced specific evidence. In Rossi, by contrast, the defendants argued that the Rockhill Report was inadequate because of factors the effects of which were pure speculation on the defendants' part. Similarly, the defendants here propose numerous factors extrinsic to the defendants that might explain the plaintiffs' losses. But they have not directed us to any point in the record that suggest that these concerns were actually relevant in this case. Accordingly, we will leave these questions to be resolved during further proceedings in the District Court. See Rossi, 156 F.3d at 487 ("[Although o]ne or more of these reasons . . . might explain Rossi's failure and could conceivably result in a verdict for the defendants at trial . . . they all involve factual disputes that need to be resolved by the trier of fact, not by this court on a motion for summary judgment.").

Finally, the defendants make a number of arguments to the effect that Seidel's method of calculating the plaintiffs' losses is unsupported and inappropriate. Seidel's calculations were based on the average or aggregate gross profits of the plaintiffs. The defendants contend that this use of averages was inappropriate, as it ignored potential differences among the plaintiffs. For instance, it ignores the problem that data was not available for all of the plaintiffs for all of the relevant time period. Furthermore, there is no way to determine based on this calculation how the damages are to be allocated among the plaintiffs. Finally, it masks the fact that some of the plaintiffs in fact had higher gross profits during the relevant period as compared with earlier. The defendants' observations are, of course, correct. They ignore a vital distinction, however: proof of fact of damage and proof of the actual amount of damages are two distinct steps. Cf. Stelwagon, 63 F.3d at 1276 n. 19 ("Because of our conclusion on the issue of Stelwagon's entitlement to damages under the Clayton Act [i.e., he failed to present sufficient evidence to prove causation], we do not reach Tarmac's argument that the amount of damages is unsupported by the evidence.").

As we stated in Rossi, at the summary judgment stage "we are not, as we would be upon reviewing a jury verdict, determining whether a plaintiff has brought forth sufficient evidence to justify the actual damages awarded." Rossi, 156 F.3d at 484. Rather, before us is only the question whether the defendants' unlawful actions caused the plaintiffs' losses. See 156 F.3d at 484. Although in Rossi we did specifically note that the Rockhill Report would support a damages judgment in the amount the expert estimated, we did so only for the future guidance of the district court, and not for any purposes related to deciding motions for summary judgment. See Rossi, 156 F.3d at 486 n. 22 ("For the guidance of the district court on remand, we note that the Rockhill Report satisfies the relaxed Bigelow standard of proof for estimating the amount of damages. . . ." (emphasis added)).

In sum, we believe that, although the question is close, Seidel's report, like the Rockhill Report in Rossi, is sufficient to create a genuine issue of material fact concerning fact of damage. This is in contrast to the expert evidence offered in Stelwagon, in which the expert's opinion involved more speculation and failed to explain certain factors concerning which the defendants had presented specific evidence at trial.

4. Sullivan's Report: We find additional evidence of loss and causation, contributing to our ultimate conclusion that the plaintiffs have adduced sufficient evidence of fact of damage, in the report of the plaintiffs' rebuttal expert, Brian Sullivan. Sullivan examined the defendants' expert's report and rebutted it in part. Although his report focused primarily on antitrust liability issues, Sullivan observed that, between 1985 and 1993, beer distributors in Allegheny County failed at a rate nearly twice that in Pennsylvania as a whole. The plaintiffs argue that, since the record suggests no other distinction between Allegheny County and the remainder of the Commonwealth than the presence of Beer World stores, the logical conclusion is that these failures were caused by the Beer World stores.

The defendants contend that we cannot consider Sullivan's report for several reasons. First, they submit that we cannot do so because he functioned only as the plaintiffs' rebuttal expert to respond to the defendants' expert, and that his report and testimony therefore cannot be introduced to support the plaintiffs' substantive case. The District Court refused to consider Sullivan's report on precisely these grounds. We think that that refusal was inappropriate. See Bowers v. Northern Telecom, Inc., 905 F. Supp. 1004, 1008 (N.D. Fla. 1995) (holding that labeling of a witness as a rebuttal expert did not preclude consideration of his testimony to defeat a motion for summary judgment). The Federal Rules of Civil Procedure, and Rule 26(a)(2) governing the disclosure and discovery of expert witnesses in particular, make no distinction between the permissible uses of "regular" experts and"rebuttal" experts. Furthermore, we see no reason to prevent the plaintiffs from using Sullivan in their case-in-chief at trial. Accordingly, it is appropriate to consider Sullivan's report as evidence in opposition to the defendants' motion for summary judgment.

Second, the defendants submit that the distinction upon which the plaintiffs base their reasoning is flawed, in that Sullivan's own report reveals that there were Beer World stores in other parts of Pennsylvania. Although there may have been other beer supermarkets in Pennsylvania, as far as we can determine from the record the only other Beer World store was one in Harrisburg. We do not think the existence of this one store can be sufficient to render Sullivan's opinion a nullity as a matter of law.

Third and last, the defendants argue that Sullivan's report is irrelevant, since it focuses on predatory pricing and the Beer World operations in general, rather than the specific discriminatory discount. Once again, we note that the record before us is not sufficiently developed for us to address this contention. We will assume for present purposes that the Beer World stores committed antitrust violations. Accordingly, we leave the defendants' contention to the District Court to consider in the first instance. We conclude that Sullivan's report provides some additional evidence of causation that, together with Seidel's report, meets the plaintiffs' burden of production on the issue of actual loss and causation in fact.

C. Is the Evidence in the Aggregate Sufficient to Prove Causation in Fact?

At all events, we are satisfied that Seidel's report, as well as Sullivan's, in conjunction with the customer evidence discussed above, constitutes sufficient evidence of causation. In Stelwagon, we noted that there was no admissible evidence that the plaintiffs had suffered injuries attributable to the defendants' price discrimination. See Stelwagon, 63 F.3d at 1275. We therefore held that the expert's report was not sufficient evidence of causation and loss. Here, by contrast, there is direct evidence — i.e., the plaintiffs' testimony about their customers' behavior — that identifies customers whom the plaintiffs lost as a result of the defendants' actions. See supra section III.A. Furthermore, there is evidence — the customers' hearsay statements, which are admissible under Rule 803(3), in addition to the expert reports — of the reasons for such loss. Thus, the plaintiffs have "adduced evidence of specific lost transactions showing causation or fact of injury, which is bolstered by an expert damage report that is not overly speculative as a matter of law." Rossi, 156 F.3d at 487. We conclude that all of this evidence taken together defeats the defendants' motion for summary judgment on the ground that the plaintiffs have not adduced sufficient evidence of fact of damage on their antitrust claims. Accordingly, we will reverse the District Court's grant of summary judgment on the antitrust claims, and remand for further proceedings.

The defendants offer as an alternative ground for affirming the dismissal of the plaintiffs' antitrust claim that the plaintiffs have failed to show that they suffered an "antitrust injury." Since we have declined to consider whether the plaintiffs have offered sufficient evidence of antitrust violations at the present time, we will also refrain from considering the defendants' contention that the plaintiffs cannot prove an antitrust injury. See supra Part II.

IV. RICO: Proximate Causation A. Basic Principles

In addition to establishing that the defendants' unlawful actions in fact caused the plaintiffs' losses, the plaintiffs must also establish proximate causation, i.e., that this causal connection is not too remote. Although this requirement applies to both antitrust and RICO claims, in this section we focus on the latter, because the plaintiffs' RICO claim founders on these grounds. A causal connection simpliciter between the defendants' actions and the plaintiffs' injuries is insufficient to give rise to a RICO claim; the plaintiff must show that that connection is proximate, i.e., not too remote. See Holmes v. Securities Investor Protection Corp., 503 U.S. 258, 268 (1992); Steamfitters Local Union No. 420 Welfare Fund v. Philip Morris, Inc., 171 F.3d 912, 932 (3d Cir. 1999).

The defendants contend that, with respect to the plaintiffs' RICO claim, the causal connection between the defendants' racketeering activities — defrauding the LCB — is too remote as a matter of law from the plaintiffs' losses — lost business. We agree. The LCB and the Commonwealth more generally were the direct victims of the defendants' actions; the plaintiffs' losses are at most derivative of any injuries to the LCB's regulatory mission. The plaintiffs are simply to remote to be able to bring a claim based on the defendants' actions.

In Holmes, the Court identified three key factors in determining whether a RICO claim is based on an injury too remote from the alleged racketeering activity:

First, the less direct an injury is, the more difficult it becomes to ascertain the amount of a plaintiff's damages attributable to the violation, as distinct from other, independent, factors. Second, quite apart from problems of proving factual causation, recognizing claims of the indirectly injured would force courts to adopt complicated rules apportioning damages among plaintiffs removed at different levels of injury from the violative acts, to obviate the risk of multiple recoveries. And, finally, the need to grapple with these problems is simply unjustified by the general interest in deterring injurious conduct, since directly injured victims can generally be counted on to vindicate the law as private attorneys general, without any of the problems attendant upon suits by plaintiffs injured more remotely.

Holmes, 503 U.S. at 269-70 (citing, inter alia, Associated General Contractors, Inc. v. California St. Council of Carpenters, 459 U.S. 519, 540-42 (1983)); see also Steamfitters, 171 F.3d at 932 (citing Holmes).

Although the Court in Holmes adopted these three factors for RICO cases from Associated General Contractors, we recognized in Steamfitters that the Court in the latter case had outlined six factors relevant to antitrust proximate causation analysis. See Steamfitters, 171 F.3d at 924. These factors included:

(1) the causal connection between defendant's wrongdoing and plaintiff's harm; (2) the specific intent of defendant to harm plaintiff; (3) the nature of plaintiff's alleged injury (and whether it relates to the purpose of antitrust laws, i.e., ensuring competition within economic markets); (4) "the directness or indirectness of the asserted injury"; (5) whether the "damages claim is . . . highly speculative"; and (6) "keeping the scope of complex antitrust trials within judicially manageable limits," i.e., "avoiding either the risk of duplicate recoveries on the one hand, or the danger of complex apportionment of damages on the other."

Steamfitters, 171 F.3d at 924 (quoting Associated General Contractors, 459 U.S. at 537-38, 540, 542-44). In our discussion of proximate causation for the RICO claims in Steamfitters, in addition to analyzing the Holmes factors, we incorporated by reference our discussion of the Associated General Contractors factors from the antitrust analysis. We did not express an opinion as to whether the Holmes factors replace the Associated General Contractors factors for RICO claims or merely supplement them.
At all events, to the extent the Associated General Contractors factors are relevant only to antitrust analysis, they are irrelevant in the present case. In addition, to the extent that any of the issues raised in these factors are not included in Holmes, they only weigh against a finding of proximate causation. For example, the second factor, specific intent to injure, is arguably not included in Holmes. To the extent it is not, however, we think it would be difficult on the present facts to conclude that the defendants specifically intended to harm the plaintiffs. Although harm to the plaintiffs may have been a probable ultimate consequence of the defendants' actions, we do not think they specifically intended to cause such harm. Accordingly, considering this factor in addition to the Holmes analysis would only provide an additional reason to conclude that proximate causation is lacking.

Both Holmes and Steamfitters clarified how the three factors set forth above would apply in particular cases.

In a recent case, we concluded that RICO proximate causation existed without specifically analyzing the Holmes factors. See Brokerage Concepts, Inc. v. U.S. Healthcare, Inc., 140 F.3d 494 (3d Cir. 1998). That case is distinguishable, however. The plaintiff BCI had formerly been the administrator of a pharmacy's self-funded employee health-insurance plan. When the pharmacy opened a new branch that it wanted to be a part of U.S. Healthcare's network, U.S. Healthcare essentially forced the pharmacy to use a U.S. Healthcare affiliate as its health-plan administrator, rather than the plaintiff. We concluded that, in such circumstances, proximate causation could exist:

The injury proved by BCI, the loss of its TPA contract with Gary's [the pharmacy], is not derivative of any losses suffered by Gary's. Unlike the injuries suffered by the non-purchasing customers in Holmes, BCI's injury was not contingent upon any injury to Gary's, nor is it more appropriately attributable to an intervening cause that was not a predicate act under RICO. Here, BCI's[administrator] relationship with Gary's was the direct target of the alleged scheme — indeed, interference with that relationship may well be deemed the linchpin of the scheme's success.

Brokerage Concepts, 140 F.3d at 521. The plaintiffs contend, similarly, that interference with their relationship with their customers, i.e., attracting the plaintiffs' customers to shop at Beer World stores, the precise harm the plaintiffs suffered, is the "linchpin of [Trone's] scheme's success." Although it may be true that interference in the relationship between the plaintiffs and their customers was the linchpin of the success of Trone's scheme, we think Brokerage Concepts is distinguishable.
The relationship between the alleged racketeering activities and the injuries to the plaintiffs are more distant than they were in Brokerage Concepts. In the latter case, the pharmacy, the party with whom BCI had a relationship with which U.S. Healthcare interfered, was the direct target of U.S. Healthcare's alleged racketeering activities, which included extortion and commercial bribery. See Brokerage Concepts, 140 F.3d at 521. Here, although the ultimate goal of Trone and the Beer World stores was presumably to woo customers away from the plaintiffs, the direct target of its alleged fraudulent scheme was the LCB, not customers. Unlike Brokerage Concepts, this case involves two third parties, one that was the target of the defendants' racketeering and another that had a relationship with the plaintiffs with which the defendants interfered.

The factual circumstances of these two cases, which inform our decision, are briefly summarized in the margin.

In Holmes, the plaintiffs (actually the plaintiffs' subrogors, although that was not relevant to the result) were customers of broker-dealers that had failed as a result of the defendants' conspiracy to manipulate certain stocks. As a result of the broker-dealers' failure, their customers suffered losses. The particular plaintiffs in Holmes were customers of the broker-dealers who never purchased the particular stocks that the defendants had manipulated. Thus, the plaintiffs' losses were not an immediate result of the defendants' manipulations, but rather were a derivative effect of the collapse of the broker-dealers. The Court found this connection insufficient to establish proximate causation for RICO claims. See Holmes, 503 U.S. at 271.
Steamfitters involved claims of union health and welfare funds against tobacco companies. The funds alleged that the tobacco companies had defrauded the funds by misleading them into believing that tobacco products were safe and could not be made safer. As a result of this fraud, the funds did not take steps to reduce their costs by, for example, attempting to reduce smoking among their participants or undertaking legal efforts to shift the costs of smoking back to the companies. The funds were harmed because their participants continued to smoke and accumulate medical bills that the funds were obligated to pay. We concluded that this causal chain was too attenuated to satisfy the requirements of proximate causation. See 171 F.3d at 932-34.

B. Anatomy of the Plaintiffs' RICO Claim

The plaintiffs' RICO claim alleges that the defendants engaged in racketeering activities by fraudulently obtaining and retaining licenses to operate beer distributorships. Specifically, they contend that Trone and others made false and fraudulent statements to the Pennsylvania LCB in order to obtain or retain various liquor licenses. The plaintiffs themselves best summarize their contention that the defendants' actions proximately caused their injuries:

[C]ausation in plaintiffs' RICO case . . . is a simple claim: we say that absent the fraud, the Trone defendants would not have been able to assemble or operate the chain of stores, and that only by assembling the chain — by "aggregat[ing]" their purchases, as Mr. Fuhrer put it — were the Beer Worlds able to secure the discriminatory discount. The Trone defendants' brief (at 40) asks rhetorically how David Trone's fraudulent statements to the LCB caused the discount, but the above two sentences show exactly how: absent the fraud, no chain; absent the chain, no discrimination. It's that simple.

Appellant's Reply Brf. at 13 (emphasis added; alterations in original). Although this is a clever and well-phrased summary, we disagree with its conclusion because the claim does not satisfy the specific factors the Court in Holmes identified as indicative of proximate causation.

The plaintiffs argue that, since the defendants' antitrust violations were a proximate cause of the plaintiffs' losses, a point the defendants do not — unlike the issue of causation in fact — presently contest, the RICO violations must also be a proximate cause of their losses. They base this contention on our recognition in Steamfitters that proximate causation principles for antitrust and RICO claims are closely related. See Steamfitters, 171 F.3d at 921 ("[T]he standing requirements for RICO and antitrust claims are similar, and . . . the standing analysis under these federal laws is drawn from common-law principles of proximate cause and remoteness of injury. . . .").
The plaintiffs misread Steamfitters. Admittedly, we said in that case that "much (if not all) of what we have said above in our discussion of antitrust standing applies to the Funds' RICO claims." See 171 F.3d at 932. But that was simply a recognition that the factual underpinnings of the causation chains in the funds' antitrust and RICO claims was so similar. The plaintiffs' theory of antitrust proximate causation in this case, however, is factually distinct from their RICO theory. Causation in their antitrust claim rests on the simple notion that the defendants contracted, combined and conspired to force the wholesalers to offer them beer at a lower price, which gave them a competitive advantage over the plaintiffs. Their RICO claim, however, rests on the more complicated theory of causation discussed in the text. It includes the additional step that Trone was able to operate the Beer World stores as a group because of his fraud on the LCB, which enabled him to obtain discounts which hurt the plaintiffs. This additional step distinguishes the plaintiffs' RICO and antitrust claims, and bars the inference of proximate causation they suggest, for reasons amplified in the text, infra.

Reflection on these three factors reveals that the direct impact of the fraud is primarily on the LCB, not the plaintiffs.

C. Analysis

1. Directness of the Injury: The first f actor, and the one on which we focused primarily in Steamfitters, is the directness of the relationship between the defendants' actions and the plaintiffs' injuries. See Holmes, 503 U.S. at 269. This is significant because "the less direct an injury is, the more difficult it becomes to ascertain the amount of a plaintiff's damages attributable to the violation as distinct from other, independent, factors." Holmes, 503 U.S. at 269. The more difficult it is to distinguish between the effects of the defendants' legitimate activities and their alleged racketeering actions on the plaintiffs, the more likely we are to conclude that proximate causation is lacking.

In Holmes, the Court found this factor indicated a lack of proximate causation. "If the nonpurchasing customers were allowed to sue, the district court would first need to determine the extent to which their inability to collect from the broker-dealers was the result of the alleged conspiracy to manipulate, as opposed to, say, the broker-dealers' poor business practices or their failures to anticipate developments in the financial markets." Holmes, 503 U.S. at 272-73. In Steamfitters, we reasoned that

if the Funds are allowed to sue, the court would need to determine the extent to which their increased costs for smoking-related illnesses resulted from the tobacco companies' conspiracy to suppress health and safety information, as opposed to smokers' other health problems, smokers' independent (i.e., separate from the fraud and the conspiracy) decisions to smoke, smokers' ignoring health and safety warnings, etc.

Steamfitters, 171 F.3d at 933 (footnote omitted).

We believe that this case presents similar difficulties in ascertaining the proportion of the plaintiffs' losses that can be attributed to the defendants' alleged racketeering activity. We think it would be difficult to trace the chain from the fraud on the LCB to particular actions of the defendants, and then to particular portions of the plaintiffs' losses, because the fraud only directly affects the LCB. In order to determine how the fraud affected the plaintiffs, we would need to analyze the extent to which the defendants were permitted to act as they did as a result of the fraud as opposed to normal operating procedures. More specifically, focusing solely on the issue of volume discounts, even if we could say that the plaintiffs' losses were entirely attributable to the defendants' ability to obtain such discounts, we would be hard-pressed to say that those discounts were entirely attributable to Trone's fraud on a third party, the LCB. Rather, it is likely that the defendants' ability to obtain these discounts was attributable, in at least as substantial a part, to the size of the individual Beer World stores, Trone's negotiating ability, the operating methodology of the stores, or other legitimate actions. Accordingly, we conclude that, "As in Holmes [and Steamfitters], this causation chain is much too speculative and attenuated to support a RICO claim." Steamfitters, 171 F.3d at 933.

2. Apportionment of Damages: Holmes also directs that we inquire into the difficulty of apportioning damages among potential plaintiffs in determining whether proximate causation is present. "[R]ecognizing claims of the indirectly injured would force courts to adopt complicated rules apportioning damages among plaintiffs removed at different levels of injury from the violative acts, to obviate the risk of multiple recoveries." Holmes, 503 U.S. at 269. As a result, where granting plaintiffs relief would require us to apportion that relief among numerous plaintiffs of different standing, we are inclined to find an absence of proximate causation for those less directly involved.

Again, this factor as applied to the facts in Holmes suggested that proximate causation was missing. The Court noted that the broker-dealers had suffered at least as much at the hands of the defendants as their customers did, and thus any determination of liability to the customers would necessitate an inquiry into the defendants' liability to the broker-dealers. The Court concluded that the possibility of treble damages in favor of both groups of plaintiffs militated in favor of finding no proximate causation for the former. See Holmes, 503 U.S. at 273 ("[T]he district court would . . . have to find some way to apportion the possible respective recoveries by the broker-dealers and the customers, who would otherwise each be entitled to recover the full treble damages."). Likewise, in Steamfitters, we noted the potential difficulty in allocating recovery between the funds and their participants:

As we noted in our discussion of the Funds' antitrust claims, more directly injured parties, i.e., smokers, would be unlikely to bring federal claims against tobacco companies for the same damages claimed by the Funds. Yet, as we also noted above, Fund participants who have not been fully reimbursed for their out-of-pocket costs that are traceable to defendants' alleged fraud and conspiracy might bring RICO or antitrust claims. Therefore, as in Holmes, a court adjudicating the Funds' RICO claims would need to consider the appropriate apportionment of damages between smokers and others such as the Funds who suffered economic losses as a result of the tobacco companies' alleged fraudulent acts.

Steamfitters, 171 F.3d at 933.

We believe that these cases support the conclusion that the defendants' fraud on the LCB did not proximately cause the plaintiffs' injuries. In particular, we note that the master distributors from whom the defendants purchased their beer at an artificially lowered price — at least according to the plaintiffs' theory of the case — suffered an injury identical to the plaintiffs'. The wholesalers presumably were paying the same price to brewers for beer, regardless of the price at which they sold it to distributors. Any discounts they gave to the Beer World stores as a result of racketeering violations came out of their own pockets. Determining how to apportion damages between the wholesalers and the plaintiffs in this case would require exactly the same sort of apportionment determination condemned in Holmes and Steamfitters. This difficulty in apportioning damages among the potential plaintiffs suggests that proximate causation is not present in this case.

The plaintiffs in their supplemental memorandum discussing Steamfitters contend that we cannot consider the wholesalers in our Holmes calculus because "they have decided to make their separate peace with Beer World, no doubt for the same reasons they decided to acquiesce in this scheme in the first place." Appellant's Post-Arg. Memo. at 13. While it may be true that the wholesalers in this case have not attempted to recover from the defendants, we do not think this is relevant to our Holmes analysis. We do not think the question whether the defendants' fraud proximately caused the plaintiffs' injuries can turn on whether some other potential claimants have filed suit. Although the broker-dealers identified in Holmes as having a potential claim did in fact sue the defendants in that case, see Holmes, 503 U.S. at 273 n. 21, we recognized in Steamfitters that the fact that smokers themselves could bring claims against the tobacco companies was relevant to determining proximate causation with respect to the funds, even though the smokers were in fact unlikely to bring claims on their own, see Steamfitters, 171 F.3d at 933.

3. Vindication of Claims by Others: The final factor that the Court in Holmes recognized as significant for proximate causation analysis was whether the plaintiff's claim could be vindicated by another, more directly injured plaintiff. More specifically, the Court recognized that the searching inquiry into causation and apportionment of damages among plaintiffs discussed above is unjustified where the central focus of RICO in deterring unlawful conduct can be vindicated by other means. See Holmes, 503 U.S. at 269-70 ("[T]he need to grapple with these problems is simply unjustified by the general interest in deterring injurious conduct, since directly injured victims can generally be counted on to vindicate the law as private attorneys general, without any of the problems attendant upon suits by plaintiffs injured more remotely."). Where a more directly affected party is available to vindicate the public interest in enforcing the law, we have less need to stretch the limits of proximate causation in RICO cases.

In Holmes, the Court concluded that, since the broker-dealers were available to vindicate the public interest in deterring racketeering, it was unnecessary to extend proximate causation analysis to include the customers. "[T]he law would be shouldering these difficulties [of making fine distinctions among causes of the plaintiff's injuries and apportioning recovery among potential plaintiffs] despite the fact that those directly injured, the broker-dealers, could be counted on to bring suit for the law's vindication." Holmes, 503 U.S. at 273. In Steamfitters, however, we found this factor to be less helpful. We noted initially that, although the funds' participants might be able to pursue RICO claims against the tobacco companies, granting due deference to the funds' allegations we could not conclude that their suits would provide the same deterrence as the funds. We were, however, ultimately "unconvinced that this distinction [from Holmes was] sufficient to overcome the concerns about apportioning damages and, most fundamentally, the remoteness of the Funds' alleged RICO injuries from any wrongdoing on the part of the tobacco companies." Steamfitters, 171 F.3d at 933-34 (citation omitted).

The plaintiffs contend that this factor dictates a finding that proximate causation is present in this case because there is no other party that was more directly injured or that will otherwise be able to vindicate the public interest in deterring racketeering activity of the sort in which the defendants have engaged. Preliminarily, as noted above, the master distributors were injured by the defendants' activities, and accordingly they could presumably serve at least as well to vindicate the public interest in deterring violations of the law. More significantly, the LCB — the direct victim of the defendants' alleged fraud — is an additional possible alternative agent for vindicating the public interest.

As the plaintiffs point out, the LCB would not be able to bring a private civil RICO action, since it is not a "person injured in his business or property by reason of" the defendants' alleged racketeering violations. 18 U.S.C. § 1964(c). In spite of the fact that the LCB cannot bring a private civil RICO action, we think that the LCB and the Commonwealth of Pennsylvania more generally are in a position to vindicate the public interest in the sense set forth in Holmes. If the facts justified it, the Commonwealth could bring a criminal charge against Trone and the other defendants under the state "little RICO" corrupt organizations statute, which is virtually identical to the federal racketeering statute. See 18 Pa. Cons. Stat. § 911(b). Section 911 includes in particular perjury, false swearing in official matters, and tampering with official records as predicate activities which can lead to racketeering liability. See § 911(h)(1)(i) (" `Racketeering activity' means any act which is indictable under any of the following provisions of this title: . . . Chapter 49 (relating to falsification and intimidation)."); see also 18 Pa. Cons. Stat. § 4902(a) (defining perjury); § 4903 (defining false swearing in official matters); § 4911 (defining tampering with public records or information).

In fact, the Commonwealth indicted Trone on state racketeering charges predicated on tampering with public records and perjury before the LCB. See App. at 99. These charges arose out of the same activities that the plaintiffs identify as the racketeering acts upon which their RICO claim is predicated. Although the indictment was dismissed, this does not affect our ultimate conclusion that the Commonwealth could vindicate the public interest. The racketeering indictment charged Trone only with racketeering predicates in which he participated as a principal, and was dismissed because all but one of these was found wanting. But the indictment included perjury charges against others involved in the Beer World operations, which, like the remaining charge against Trone, were eventually nolle prossed. See App. at 98, 100. The indictment could have charged these other acts of perjury as predicates to the racketeering charge against Trone, which would have created the pattern of racketeering activity necessary to support a "little RICO" charge.

See App. at 43. The court dismissed the tampering charges on the ground that they should have been brought under a more specific statute, the Liquor Code, see Pa. Stat. Ann. tit. 47, § 4-436(j) (West 1997), which makes false statements on liquor license applications a misdemeanor. See App. at 25-31. The alleged tampering therefore also could not serve as part of the pattern of racketeering activity necessary to support the racketeering charge, since Liquor Code violations are not specified as racketeering activities in section 911. See App. at 39 n. 6; 18 Pa. Cons. Stat. § 911(h). Since the only remaining racketeering activity was one alleged instance of perjury on the part of Trone, the court concluded that there was no "pattern of racketeering activity" as required to support a racketeering charge. See App. at 40.

Although the Commonwealth cannot now bring civil RICO claims against the defendants here, given the possibilities set forth above, we do not think this brings it outside the scope of the third Holmes factor. The Court's primary concern in Holmes was to ensure that some plaintiff be available to vindicate the law's "general interest in deterring injurious conduct." Holmes, 503 U.S. at 269. A civil RICO action is not specifically required to vindicate this general deterrence interest. See Laborers Local 17 Health Benefit Fund v. Philip Morris, Inc., 172 F.3d 223, 235 (2d Cir. 1999) (concluding that the possibility of independent tort claims by smokers, or subrogated claims based thereon by union health funds, would be sufficient to satisfy the requirements of the third Holmes factor). Although not providing for treble damages, we believe that the prospect of state criminal racketeering charges would provide an adequate deterrent to lawless conduct of the type alleged here to satisfy the concerns embodied in Holmes.

Judge Wellford in his dissent contends that we cannot consider the wholesalers or the Commonwealth as potential alternative agents for the vindication of the public interest, because in this case none of them brought suit against Trone and the Beer World stores. We think this circumstance is irrelevant to determining whether the plaintiffs' injuries are too remote from the defendants' actions to be a proximate cause for the RICO claim. The post-injury actions of intervening parties cannot make the plaintiffs' losses more or less of a direct result of the defendants' actions. The only question is whether these intervening parties are ones that possibly could take steps to deter illegal activity as contemplated in Holmes. See Holmes, 503 U.S. at 269-70 ("[D]irectly injured victims can generally be counted on to vindicate the law as private attorneys general. . . ." (emphasis added)); 503 U.S. at 273 ("[T]hose directly injured . . . could be counted on to bring suit for the law's vindication. . . .").

4. Summary: At all events, even to the extent that we have questions about whether the possibility of the Commonwealth bringing criminal racketeering charges against Trone and the other defendants falls within the scope of the third Holmes factor, such questions cannot alter our ultimate conclusion, based on the Holmes factors as a whole, that proximate causation is lacking here. To paraphrase Steamfitters, "we are unconvinced that [the potential lack of alternative plaintiffs] is sufficient to overcome the concerns about apportioning damages and, most fundamentally, the remoteness of [the plaintiffs'] alleged RICO injuries from any wrongdoing on the part of [Trone]." Steamfitters, 171 F.3d at 933-34. Considered as a whole, the Holmes factors dictate the conclusion that the plaintiffs cannot demonstrate a proximate causal connection between their injuries and the defendants' alleged racketeering activities.

D. Policy Issues: Were the Plaintiffs the Intended Beneficiaries of the Liquor Code?

The plaintiffs also contend that proximate causation is present in a civil RICO case where the alleged racketeering conduct effects violations of a regulatory regime designed to protect the plaintiffs. See, e.g., Rodriguez v. McKinney, 878 F. Supp. 744, 747-49 (E.D. Pa. 1995); Trautz v. Weisman, 819 F. Supp. 282, 287 (S.D.N.Y. 1993); see also In re Orthopedic Bone Screw Prods. Liab. Litig., 159 F.3d 817, 826-27 (3d Cir. 1998) (recognizing a similar principle in the context of state common-law fraud claims). Although this may be a valid principle, we find it inapposite in the present case, as the condition of its application is not present here.

The purpose of the Pennsylvania Liquor Code is to promote temperance, not to protect small-business owners or ensure competition among beer retailers:

The provisions of [the Liquor Code] are intended to create a system for distribution that shall include the fixing of prices for liquor and alcohol and controls placed on prices for malt and brewed beverages, and each of which shall be construed as integral to the preservation of the system, without which system the Commonwealth's control of the sale of liquor and alcohol and malt and brewed beverages and the Commonwealth's promotion of its policy of temperance and responsible conduct with respect to alcoholic beverages would not be possible.

Pa. Stat. Ann. tit. 47, § 1-104(d) (West 1997) (emphasis added); see also § 1-104(a) ("This act shall be deemed an exercise of the police power of the Commonwealth for the protection of the public welfare, health, peace and morals of the people of the Commonwealth and to prohibit forever the open saloon. . . ."); Altshuler v. Pennsylvania Liquor Control Bd., ___ A.2d ___, No. 2126 C.D. 1998, 1999 WL 298228, at *3 (Pa. Commw. Ct. May 13, 1999) ("The purpose of the Liquor Code is not to promote the sale of liquor, rather it is to regulate and restrain the sale of liquor."). At least one court has recognized that the Liquor Code was, in fact, not at all intended to protect the economic interests of liquor retailers. See Lancaster County Tavern Assn. v. Pennsylvania Liquor Control Bd., 14 Pa. D. C.3d 381 (Lancaster Cty. C.P. Ct. 1980).

The plaintiffs submit that even if the purpose of the Liquor Code is not to protect retailers like themselves, the effect of the Code, and one of the goals of the LCB in enforcing it, is to protect retailers and competition. But although the LCB's efforts to enforce the Code may have resulted largely in a predominance of beer retailers similar to the plaintiffs, that does not render large-scale stores like the Beer World stores automatically illegal. Accordingly, we do not think that the principle of Rodriguez, were we to adopt it, would compel a finding of proximate causation. We will therefore affirm the District Court's grant of summary judgment on the plaintiffs' RICO claim.

For the foregoing reasons, the judgment of the District Court will be reversed to the extent that it granted summary judgment to the defendants on the plaintiffs' antitrust claims, but affirmed in all other respects, and the case will be remanded to the District Court for further proceedings in accordance with this opinion.

[116] WELLFORD, Senior Circuit Judge, concurring in part and dissenting in part:

I concur in Chief Judge Becker's excellent analysis of the antitrust claims of plaintiffs against the defendants. I therefore share in the conclusion that the district court was in error in granting summary judgment to defendants on the antitrust claims before the court.

My disagreement is in respect to the treatment of the RICO claims. I dissent in that respect with some trepidation, realizing that Chief Judge Becker has recently authored several RICO decisions of this court flowing from the Supreme Court decision in Holmes v. Securities Investor Protection Corp., 503 U.S. 258 (1992). I begin, then, with an analysis of Holmes in respect to the RICO issues in this case.

First, however, I construe plaintiffs' RICO claim to be as follows: as an intended consequence of defendants' alleged predicate fraudulent actions and activities in attaining a special status as a Pennsylvania beer retailer/distributor and obtaining through that fraud from the state a special license and status (contrary to any legal entitlement), plaintiffs were economically damaged. The relevant cases discuss in the RICO context whether plaintiffs have standing to bring the claim against a defendant, and whether plaintiffs can establish commercial damages as a proximate cause of defendant's illegal predicate acts.

Holmes involved the issue of standing and of proximate cause under RICO by a party asserting securities fraud. Plaintiff, Securities Investor Protection Corp. ("SIPC"), was neither a buyer nor a seller of alleged manipulated stocks orchestrated by defendants. SIPC sued seventy-five defendant broker/dealers whose alleged illegal predicate acts brought about the collapse of several brokerage concerns which were members of SIPC, causing it to pay millions in damages to the failed member brokerage houses. Holmes acknowledged that § 1964(c) of RICO was "modeled on the civil-action provision of the federal antitrust laws." Id. at 267. We agree that plaintiffs in this case have set out an antitrust claim that survives summary judgment treatment. Holmes interpreted proximate cause in its RICO analysis:

At bottom, the notion of proximate cause reflects "ideas of what justice demands, or of what is administratively possible and convenient." W. Keeton, D. Dobbs, R. Keeton, D. Owen, Prosser and Keeton on Law of Torts § 41, p. 264 (5th ed. 1984). . . . [One requirement is] some direct relation between the injury asserted and the injurious conduct alleged.

Id. at 268.

Was the plaintiff in Holmes simply complaining about "harm flowing merely from the misfortunates visited upon a third person by the defendant's acts . . ."? Id. Holmes held that SIPC was complaining about an indirect injury, but it is important to consider why it reached that result. First, Holmes noted in footnote 19 that SIPC was not claiming to sue under a claimed right of any customer who actually purchased the manipulated securities. Id. at 272 n. 19. Second, it is important to note that in Holmes, the broker/dealers, directly defrauded, who went into bankruptcy "have in fact sued" the same defendants. Id. at 273. Those third parties might then vindicate the public interest in recouping the economic damages caused by the fraudulent defendants, and in punishing them by treble damages.

Because of the potential of multiple claims against defendants seeking damages as a direct result of the same illegal predicate acts and the necessity of difficult and complex apportionment, Holmes decided in favor of defendants that SIPC's damages claims did not meet the proximate cause test. Our case is a very different one factually from Holmes. Plaintiffs here assert actions arising from defendants' illegally attained status based on asserted fraud perpetrated on the state of Pennsylvania. This does not, in my view, vindicate the rights of private parties, such as plaintiffs, arising out of that fraud. Unlike defrauded third party customers who had also sued defendants for the RICO actions in Holmes, neither Fuhrer, nor any other master distributor, sought any such damages against the Trone defendants for the alleged illegal predicate activity.

As Chief Judge Becker indicates, Pennsylvania may only seek criminal penalties and withdrawal of defendants' special license, not damages.

Indeed, Fuhrer denied that any such illegal activity took place, and is an alleged co-conspirator in the antitrust activity.

In sum, I cannot construe Holmes as helpful to defendants in this case. Steamfitters Local Union Fund No. 420 v. Philip Morris, Inc., 171 F.3d 912 (3d Cir. 1999), I think, is distinguishable. In Steamfitters, customers or purchasers of the tobacco products had brought suit, or might be expected to bring suit, to vindicate plaintiff's clearly indirect claim. These customers or purchasers had varying degrees of proximate contributory or comparative negligence or knowledge about the danger of the tobacco product used or sold to them. Respectfully, I do not believe plaintiffs' claims in the instant case to be as attenuated as in Steamfitters. It is closer to the standing and proximate causal relationship of plaintiff in Brokerage Concepts, Inc. v. U.S. Healthcare, Inc., 140 F.3d 494 (3d Cir. 1998), in my view.

I do, therefore, respectfully dissent on the RICO element of this difficult case. I would hold that we should reverse and remand on both the antitrust and RICO claims.

A True Copy: Teste:

Clerk of the United States Court of Appeals for the Third Circuit