The United States appeals, pursuant to 18 U.S.C. Sec. 3731 (1982), a district court order dismissing on double jeopardy grounds an indictment for violation of section 1 of the Sherman Act, 15 U.S.C. Sec. 1 (1982). We reverse.

On July 19, 1984, a grand jury in the Eastern District of Pennsylvania returned an indictment charging Sargent Electric Company (Sargent), Lord Electric Company (Lord), W.V. Pangborne and Co. (Pangborne) and Pangborne's Executive Vice-President, J.A. Bruce Pinney, with violating section 1 of the Sherman Act by conspiring to rig bids for electrical construction work at the Fairless Hills Works of United States Steel Corporation in Bucks County, Pennsylvania (Fairless Hills). The four defendants moved to dismiss the indictment, all asserting that it was barred by the double jeopardy clause of the United States Constitution.

The first two defendants, Sargent and Lord, contended that the offense charged in the July 19, 1984 indictment was the same as that for which they previously had been convicted in the Western District of Pennsylvania. That conviction resulted from an indictment returned by a grand jury in the Western District of Pennsylvania on July 1, 1983, which charged that Sargent, Lord, and other electrical contractors had, for approximately seven years from 1974 to 1981, conspired to rig bids for electrical construction work at the Western Pennsylvania Works of U.S. Steel (Western Works). See Joint Appendix at 1230. This court affirmed judgments fining Sargent and Lord one million dollars each, the statutory maximum for violation of section 1 of the Sherman Act. United States v. Fischbach & Moore, Inc., 750 F.2d 1183 (3d Cir.1984), cert. denied, --- U.S. ----, 105 S. Ct. 1397, 84 L. Ed. 2d 785 (1985).

With respect to the second two defendants, Pangborne and Pinney, they contended that the offense charged in the July 19, 1984 indictment was the same as that for which they had pleaded nolo contendere in the Eastern District of Pennsylvania. Those pleas followed a March 8, 1984 indictment that charged that Pangborne, Pinney, and other electrical contractors, "beginning in or about the late 1970's and continuing thereafter into about 1982," had conspired to rig bids for electrical construction work at a Gulf Oil Company refinery in Philadelphia, Pennsylvania. See Joint Appendix at 1130. The district court accepted the pleas from Pangborne and Pinney on June 5, 1984 and sentenced them on July 19, 1984.

On the basis of these contentions, the district court ruled on November 13, 1984 that the four defendants had established prima facie cases of nonfrivolous double jeopardy claims and scheduled an evidentiary hearing on those claims. The court ordered the United States to disclose to the defendants all grand jury testimony regarding alleged illegal bid-rigging on steel and refinery jobs involving the indicted and unindicted co-conspirators named in the July 19, 1984 indictment. It then held a lengthy hearing and subsequently filed a memorandum concluding that there was at all relevant times one general understanding among all the participating electrical contractors that they would rig bids when circumstances permitted regardless of whether the job was at Western Works, at Fairless Hills, at the Gulf Oil refinery, or at several other locations. United States v. Sargent Electric Co., Crim. Action No. 84-00313 (E.D.Pa. March 26, 1985). "Bid rigging at each of these facilities," the court concluded, "did not reflect separate conspiracies but simply separate manifestations of the same unlawful agreement to rig bids wherever and whenever possible." Id., slip op. at 30. The district court therefore dismissed the indictment. This appeal followed.

The United States does not dispute the district court's ruling that the four defendants made out nonfrivolous claims of double jeopardy. Consequently, there is no question that once the defendants advanced their claims, the government assumed the risk of nonpersuasion of demonstrating by a preponderance of the evidence that the double jeopardy clause did not bar the challenged indictment. United States v. Inmon, 568 F.2d 326, 331-32 (3d Cir.1977).

While the parties agree upon the applicability of Inmon to the proceedings in the district court, they disagree as to the scope of our review of that court's holding. The defendants urge that the district court's conclusion that "[b]id rigging did not reflect separate conspiracies but simply separate manifestations of the same unlawful agreement to rig bids wherever and whenever possible" is a factual determination that we cannot reverse on appeal without finding that the determination was clearly erroneous. The United States, on the other hand, contends that whether it met its burden of proving separate offenses is a question of law as to which this court exercises plenary review.

Although this court did not address in Inmon the scope of review issue that this case raises, we did do so in United States v. Felton, 753 F.2d 276 (3d Cir.1985). In Felton, in which we reviewed a district court's denial of a motion to dismiss an indictment on double jeopardy grounds, the United States, as appellee, urged a position similar to that which the defendants here espouse. We rejected that argument, observing,

We find it necessary to reiterate that on review the trial court's findings of narrative or historical facts are measured by the clearly erroneous test, but as to the legal component of its conclusion, this court has plenary review. Accordingly, what we must examine here is whether the district court correctly determined that as a matter of law the government met its burden of proving the separate conspiracies by a preponderance of the evidence. Such an inquiry requires us to examine the basic facts of record and the permissible inferences that may be drawn therefrom.

Id. at 278 (citation omitted). Thus Felton has classified as a legal issue, subject to plenary review, the question whether, from the narrative or historical facts the government established to the satisfaction of the district court, an inference of multiple conspiracies rather than a single conspiracy was permissible.

The procedural posture of this case is not identical with Felton, however, because in this case the district court granted the defendants' motion to dismiss. In Felton we held that the government failed to establish a prima facie case on the existence of conspiracy. Here, in order to reverse, we must hold not only that the United States made out a prima facie case on the existence of multiple conspiracies, but also that an inference of a single conspiracy was on this record clearly erroneous as a matter of fact or legally impermissible. Such a conclusion involves a much greater assertion of authority to set aside rulings of the court of first instance than was made in Felton. There may be cases where it would be possible to hold that as a matter of law no inference other than that of multiple conspiracies was permissible. In most cases, however, in which the government has made out a case satisfying the Felton threshold, competing inferences may be logically permissible, and in such cases the inference drawn by the court of first instance should stand.

Of course a district court's decision, even when couched in terms of inference drawing, may be predicated upon legal assumptions such as the elements of the offenses charged in successive indictments. As to such legal assumptions our reviewing function is obviously plenary. If, however, we should disagree with the district court's legal assumptions of the elements of the offenses charged--here Sherman Act conspiracies--such disagreement would not necessarily permit us to select among competing permissible inferences. At most such a disagreement would require a remand for reconsideration. See United States v. Young, 503 F.2d 1072, 1076 (3d Cir.1974).

Thus for the United States to prevail in this appeal we must (1) hold that the district court misunderstood the elements of a Sherman Act conspiracy or (2) hold that the court's findings on historical or narrative facts are clearly erroneous, or (3) hold that the only inference that may be drawn from those facts is that the bid-rigging conspiracy aimed at Fairless Hills was separate from those aimed at Western Works and Gulf's Philadelphia refinery.

Unlike statutes such as 18 U.S.C. Sec. 371 (1982) and 21 U.S.C. Sec. 846 (1982) that speak of conspiracy to commit crimes defined in other federal statutes, section 1 of the Sherman Act is self-contained. It makes illegal contracts, combinations, or conspiracies "in restraint of trade" and punishes as a felony every contract, combination, or conspiracy "hereby declared to be illegal." 15 U.S.C. Sec. 1 (1982). Thus unlike other federal conspiracy statutes section 1 of the Sherman Act does not refer us elsewhere for the substantive elements of the prohibited conduct. As in all conspiracy statutes, concert of action is a substantive element, but under governing caselaw more than mere concert of action is required. The illegal object of a Sherman Act conspiracy must be identified in terms of an intended or achieved effect upon commerce in a relevant market, a market not defined by statute.

When the concerted action is price-fixing or bid-rigging, the Sherman Act caselaw shortcuts the inquiry into market effect, by treating such concerted actions as per se violations. Those activities are deemed to be per se illegal, in the sense that the court will not consider evidence that their effects upon commerce in a relevant market might have had a pro-competitive purpose and effect. The bid-rigging activities charged in the several indictments before us fall within that per se category. But while the per se rule proscribes inquiry into competitive effects, it does not excuse identification of relevant markets. An agreement among persons who are not actual or potential competitors in a relevant market is for Sherman Act purposes brutum fulmen. An agreement "to rig bids wherever and whenever possible" is meaningless for Sherman Act purposes unless there are in the real world of the marketplace some "whens" and "wheres."

To some extent, of course, a horizontal agreement tends to define the relevant market, for it tends to show that the parties to it are at least potential competitors. If they were not, there would be no point to such an agreement. Thus its very existence supports an inference that it would have an effect in a relevant market. Where, as here, however, the disputed issue is the existence or scope of the alleged horizontal agreement that is to be inferred from circumstantial evidence, the first inquiry must be whether or not each firm alleged to have been a party to it was an actual or potential competitor in that market. Membership in a bid-rigging conspiracy aimed at Western Works, for example, would not have any actual or potential effect upon Gulf Oil Company's Philadelphia refinery unless there was actual or potential competition for electrical work at that refinery by the Western Works conspirators.1

The district court's analysis placed principal reliance upon evidence that all alleged conspirators had a common objective: the elimination of price competition whenever and wherever possible. But as we note above, that common purpose is not alone sufficient to establish a violation of section 1 of the Sherman Act. Thus the district court's focus upon facts suggesting such common purpose arguably ignored the essential market element of a Sherman Act offense and gave insufficient weight to the government's proofs respecting that element.

Because Sherman Act conspiracies involve a relevant market and that market may vary over time, the government's task in drafting indictments is somewhat more complex than in other conspiracy contexts. The difficulty is illustrated by the contrasting positions taken by Sargent and Lord in this case and in their appeal from the sentence imposed in the Western District of Pennsylvania. In that case the indictment charged a conspiracy to fix the prices at which electrical construction projects at Western Works were bid. Sargent argued on appeal that there was a fatal variance between the indictment and the proofs at trial in that they failed to establish that all bids at Western Works were rigged. Alternatively Sargent argued that the government had proved not one continuing conspiracy, but a series of independent conspiracies. 750 F.2d at 1189. Lord also argued that the government proved fourteen separate conspiracies, while the indictment charged one. See id. at 1190. This court held that the indictment alleged a single conspiracy to rig bids at Western Works and that proof of fourteen instances of such bid-rigging permitted the jury to conclude that the conspiracy encompassed all bids. Id. In this appeal, on the other hand, Sargent and Lord both allege that the trial of the prior indictment put them in jeopardy with respect to a conspiracy that was aimed at different steel mills of U.S. Steel and different corporations over extended time periods. Had the government known about bid-rigging at other locations when the July 19, 1984 indictment was drawn and had it drawn that indictment to charge an agreement to rig bids "wherever and whenever possible," we may be certain that the defendants would have urged prior to trial that the indictment should be stricken as prejudicially duplicitous. Had the July 19, 1984 indictment been drawn with a separate count addressed to the bid-rigging activities of the defendants at different locations, both their pretrial and posttrial claims would have been that the indictment was multiplicitous.

For purposes of the double jeopardy clause it is appropriate to consider how the government could have drawn the first indictments. None of the four defendants before us in this appeal was acquitted in a prior proceeding. Thus they cannot rely upon the collateral estoppel component of the double jeopardy defense announced in Green v. United States, 355 U.S. 184, 78 S. Ct. 221, 2 L. Ed. 2d 199 (1957), and applied to the states in Ashe v. Swenson, 397 U.S. 436, 90 S. Ct. 1189, 25 L. Ed. 2d 469 (1970). They rely on two different and complementary aspects of double jeopardy: protection against more than one prosecution for the same offense, In re Nielsen, 131 U.S. 176, 187, 9 S. Ct. 672, 675, 33 L. Ed. 118 (1889), and protection against multiple punishments for the same offense, Ex Parte Lange, 85 U.S. (18 Wall.) 163, 168-69, 21 L. Ed. 872 (1873). The standard of these aspects of double jeopardy is no different pretrial and posttrial.2 The defendants' position, if accepted, means that in their initial prosecution the same grand jury could not have charged them with separate Sherman Act offenses. Our inquiry now is no different, legally, than that which would have faced the district court had multiple counts been pleaded in the initial indictment against each defendant. We must determine whether there was a conspiracy aimed at a single relevant market or conspiracies aimed at several separate markets. To put the same question in slightly different words, we must determine whether charging several Sherman Act violations was multiplicitous.

Because the district court placed primary reliance upon the common objective of price-fixing, its discussion of the narrative or historical facts emphasized those facts tending to support the inference that there was such a common objective. In the only real discussion of the market aspect of the Sherman Act offense the court observed,

In the present case, there is the potential for competition. If the contractors were added to the bid list at a facility, they would be able to compete for jobs. The contractors were not physically or economically incapable of competing for contracts, they were only prevented from doing so by the regulations at each facility. There was always the possibility that a contractor would be added to the bid list and, thus, all the contractors were potential competitors.

Sargent Electric, Crim. Action No. 84-00313, slip op. at 28 (emphasis supplied). The court's reference to the "regulations at each facility" is to the undisputed evidence that at each facility where bid-rigging charged in any of the indictments is alleged to have occurred, only electrical contractors on a pre-approved bid list were permitted to examine construction documents and submit bids. The court found, moreover, that all the companies affected usually invited the same contractors to bid on each of their projects. Id. at 4. In the case of U.S. Steel, separate bid lists were maintained for Western Works and Fairless Hills. Meetings were held among representatives of the conspirators separately for each project. The district court found,

At the meetings, the parties decided among themselves who the low bidder would be for that particular project. The bid each contractor was to submit on the project was determined either during or after the meeting. The jobs were allocated among the participants so that each contractor on the bid list would receive a fair share of the projects let by the particular facility....

The members of the conspiracy kept records concerning how the jobs were allocated. Separate records were kept for each facility. There was no bookkeeping system for the conspiracy as a whole.

Id. at 7.

If all bidders were on the bid list at each facility, or even if each bidder, by some action within its own control or the control of the other alleged conspirators, could get on each bid list, the court's conclusion that there was a potential for competition at each facility would plainly be a permissible inference. The undisputed facts, however, are otherwise.

Four companies, Pangborne, H.B. Frazer Co., Ortlip Co., and The Foley Co., were charged in the indictment to which Pangborne and Pinney pleaded nolo contendere to having rigged bids at Gulf's Philadelphia refinery. These bidders were qualified by Gulf to submit bids for work at the facility. Of these, only Pangborne, Ortlip Co., and The Foley Co. were on the Fairless Hills qualified list. That list included, besides Pangborne, Ortlip Co., and The Foley Co., Sargent, which was not qualified by Gulf to bid for work at the Philadelphia refinery. Lord, Sargent, The Foley Co., and Fishbach and Moore (Pangborne's parent corporation) were on U.S. Steel's qualified list for Western Works. Pangborne, H.B. Frazer Co., and Ortlip Co. were not. Sargent and Lord were on the qualified list at Fairless Hills, but neither Lord nor Sargent nor Fishbach and Moore were qualified at Gulf. Thus the only alleged co-conspirator qualified to bid at all three facilities was The Foley Co., an indicted co-conspirator in the July 19, 1984 indictment and an unindicted but named co-conspirator in the other two indictments.

The district court did not find, and the record would not support a finding, that it was within the power of bidders, singly or collectively, to get on the list of qualified bidders at these separate facilities. That was entirely within the control of the management of those facilities. Instead the district court found that "[t]he contractors could not place themselves on the bid list, nor could they attempt to compete for projects at that facility unless they were on the bid list." Id. at 27.

Given the incongruity in membership of the approved bidder lists and the fact that those lists were controlled by the management of the separate facilities, it is not surprising that, as the court found, "Only the contractors on the bid list for that project would attend the [bid-rigging] meeting. If a contractor was not on the bid list for a particular project, the contractor would not even be informed that a meeting was to take place." Id. at 6. The court found, moreover, that "[o]n those occasions when a contractor would be dropped from the list for a period of time it would not participate in the scheme. If, later, it was put back on the list it would resume its conspiratorial activities." Id. at 12.

The court's observation that "there is the potential for competition, " id. at 28, if it is intended as a legal conclusion, is inconsistent with the underlying factual findings and the undisputed facts. There is no potential competition between a party not on an approved list of vendors and a party on such a list. If it is intended as a finding of fact, that finding is on this record clearly erroneous. The government's evidence, which is not disputed, is that the managements of the purchasers by their independent actions determined the competitors at each facility, and thus the relevant market for purposes of section 1 of the Sherman Act.

The district court bolstered its single violation analysis by discounting certain other evidence tending to show that there were separate agreements or understandings with respect to each facility. Conceding that separate meetings took place for bid-rigging on different facilities, the court noted that "[t]he differences in meeting places does not indicate separate facilities, it simply demonstrates the adaptability of the scheme to the circumstances." Id. at 30. Differences in the manner in which bids were prepared and submitted at different facilities were dismissed as "simply adaptations of the same scheme to the individual requirements of each victim." Id. at 29. While observing that "[t]he parties' relationships at one facility depended upon their relationship at the others," the district court discounted the undisputed fact that The Foley Co. did not rig bids at Western Works or at Lukens Steel while it participated in bid-rigging at Gulf's Philadelphia refinery and at Fairless Hills. The court disregarded, as well, the undisputed fact that while Lord rigged bids at Western Works and at Fairless Hills, it did not do so at Lukens Steel. Conceding that not all the participants were active at every facility and that not all the participants even knew each other, the district court discounted the significance of these facts by noting that "members of a single conspiracy do not have to know or have direct contact with each other." Id. at 21.

These explanations of undisputed facts, which on their face tend to suggest separate agreements or undertakings with respect to each facility, would be plausible except for the absence of any evidence tending to show that the conspirators had any control over the dimensions of the market defined by the purchasers' independent decisions to deal only with pre-qualified bidders. We accept as true the underlying facts. The district court's explanations for their insignificance, if those explanations are intended as findings of fact, are clearly erroneous. Probably they are not so intended, but are merely legal argument. As such they suffer from the fundamental error that permeates the district court memorandum; namely, focusing on the common motive of fixing prices while disregarding the other essential element of a section 1 Sherman Act case, relevant market.

The district court also relied upon the fact that the time periods during which bid-rigging occurred at the various facilities were "basically congruent." Id. at 20. This finding is supported by the record in a limited sense. For some period between 1971 and 1981 bid-rigging was going on at all three facilities. However, the undisputed evidence--in particular the immunized testimony of Paul Arbogust of Lord--suggests that the bid-rigging began at each facility at a different time and ended at each facility at a different time and for different reasons. Thus the fact that the times of the bid-rigging were "basically congruent," while of some evidentiary significance, cannot be relied upon as establishing that there was a single conspiracy. The district court discounted the different reasons for terminating bid-rigging at each facility as irrelevant, asserting that the events leading to termination "were events over which the conspirators had no control." Id. at 20. This very absence of control, however, is strong evidence of the absence of interdependence among the separate markets and thus undercuts the court's ultimate conclusion that the participants acted pursuant to a single overarching conspiracy.3V.

The district court's faulty legal analysis, which focuses exclusively upon the concert of action component of a section 1 violation, so permeates its decision that the order appealed from cannot stand. If the court's observation that there existed the potential for competition among all the conspirators at each facility and its conclusion that the bid-rigging schemes at all three facilities amounted to but one Sherman Act offense are regarded as findings of fact, they are inconsistent with the court's findings of narrative or historical fact, are inconsistent with the record, and are clearly erroneous. If they are regarded as conclusions of law, they are erroneous for the reasons set forth in Part III.

The question that remains is whether we should remand for further proceedings on the double jeopardy issue or simply reverse. It seems to us that the historical or narrative facts found by the district court require a holding that the United States established, as a matter of law, that the bid-rigging at the different facilities constituted multiple Sherman Act offenses. The district court found that at each facility there were separate bid lists; that the purchasers, not the bid-riggers, controlled access to those bid lists; that the membership of different lists were not identical; that some firms qualified to bid at more than one facility did not rig bids at all of them; that separate bid-rigging meetings were held with respect to each facility; that the allocation of business among the bid-riggers was made on a facility-by-facility basis; that not all the conspiring personnel knew each other; that the bid-rigging commenced at each facility at a different time; and that bid-rigging ended at each facility at different times and for different reasons. These findings, absent others pointing the other way, establish that there were separate offenses. No other findings, except the legally flawed conclusions about potential competition, point the other way. Given these historical facts, any ultimate finding of a single conspiracy would be clearly erroneous. Thus we will reverse with a direction that trial on the July 19, 1984 indictment may go forward.

STAPLETON, Circuit Judge, concurring.

Since I agree that the record will support only a finding of site-by-site conspiracies, I join the disposition of the court.

I begin with the uncontested facts that (1) there was an agreement among bidders at Western Works, including Sargent and Lord, to rig bids at that site, (2) there was an agreement among bidders at Fairless, including Sargent, Lord, and Pangborne, to rig bids at that site, and (3) there was an agreement among bidders at Gulf, including Pangborne, to fix bids at that site.

To reach the legal conclusion that it did, the trial court necessarily found as a fact that the defendants had not simply made site-by-site decisions to cooperate in rigging bids, but also that each had committed itself to rig bids at a universe of sites including Western Works, Fairless, and Gulf. Since there was no evidence, economic or otherwise, which would meaningfully distinguish any subset of sites that included Western Works, Fairless, and Gulf from the total universe of sites at which rigging was known to have occurred, the trial court concluded that each of the defendants and their competitors had committed itself to rig bids whenever and wherever the conspirators (whether or not presently known to it) might find themselves bidding against one another on a restricted list.

Recognizing that the crucial issue is whether there was ad hoc, site-by-site decision-making or a commitment by each conspirator to rig bids across a potentially limitless universe of sites, it is helpful in determining what inferences may rationally be drawn from the circumstantial evidence. Experience teaches that one seeking to maximize his risk/benefit ratio will generally prefer to make ad hoc decisions that can take into account the facts of each situation bearing on benefit and risk unless there is some substantial benefit attainable from a broader commitment which is unobtainable through ad hoc decision-making. Because this record is barren of any suggestion of a substantial benefit available from an "overarching" conspiracy that was not available through ad hoc, site-by-site decision-making and because all of the circumstantial evidence either strongly favors or is consistent with a finding of multiple conspiracies, I conclude that the district court's finding of an overarching conspiracy is clearly erroneous.

Neither the trial court nor the dissent suggests that a single site conspiracy to rig bids could not be successful without an overarching conspiracy.1 Nor do I understand either to say that an overarching conspiracy was shown to have been safer or to have had the potential for generating greater rewards than a series of site specific conspiracies. Each finds it significant, however, that the bidders who were able to trust one another at one site would be able to trust one another at a second site. While this is true, it is beside the point. No one would dispute that to a firm interested in bid rigging, knowledge as to the identity and trustworthiness of others willing to do the same is of great value. But on this record, such knowledge was as available and as valuable to a firm making site-by-site decisions on whether to rig bids as it would have been to a member of a grander conspiracy. In short, the "trust factor" provided no motivation for a prospective conspirator to forego the natural preference for site-by-site decision-making.

The trial court found or the undisputed evidence demonstrated (1) that there was a different group of bidders at each site; (2) that separate bid rigging meetings were held,2 and separate records maintained3, for each site; (3) that precautions taken against discovery varied4; (4) that the allocation of business at Western Works, Fairless, Gulf and virtually all other sites was made exclusively on a site-by-site basis despite the feasibility of trade-offs between sites as demonstrated by those which occurred between Western Works and Jones & Laughlin Steel5; (5) that the method for designating the low bidder varied6; (6) that some firms qualified to bid at more than one site did not rig bids at all of them7; (7) that not all of the conspiring personnel knew one another; (8) that several of the purported "core" conspirators did not know that bid rigging was going on elsewhere than at their particular location8; (9) that the bid rigging commenced at each site at a different time9; and (10) that the bid rigging ended at each facility at different times and for different reasons.10 These findings of historical fact and this undisputed testimony, in the absence of unambiguous evidence suggesting an overarching conspiracy, cumulatively compel a conclusion that those rigging bids were making ad hoc decisions on whether to rig at each site.

The few facts relied upon by the trial court in reaching a contrary conclusion are equally consistent with a finding of multiple conspiracies and, accordingly, do not provide unambiguous evidence of an overarching conspiracy. The district court specifically looked to Tri-City's withdrawal both from the Fairless and the Western Works schemes [193a, 179a, 494a-95a] as evidence that the schemes were so interdependent that "[i]f they could not work together and agree at one facility, they could not work together anywhere." However, this withdrawal is no more probative of a single than of several agreements. If an individual who is involved in more than one illegal conspiracy resolves to discontinue his illegitimate activities, his departure from both does not unify otherwise unrelated schemes.

Also, the district court believed that Frazer had agreed to rig bids at Lukens without being rewarded at Lukens, implying that Frazer did this so it would be rewarded elsewhere [193a]. The alternative explanation of Frazer's actions is consistent with the existence of multiple conspiracies, however, and seems considerably more likely. Frazer bid on a project at Lukens that it did not want to obtain so that it could remain on Lukens' bidding lists for future work [793a-94a]. A failure to bid on current jobs might have precluded Frazer from working at Lukens in the future [802a-03a]. In fact, it is uncontradicted that Frazer did just that on at least one job at Gulf that it had no desire to obtain [783a-84a].

Finally, the court found that the bid rigging at Western Works and Fairless was interconnected partly because of a meeting called to resolve a dispute after Sargent had underbid the designated low bidder on a Fairless job [192a-93a]. However, the testimony was undisputed that the meeting only related to Fairless, and no mention was made of non-Fairless jobs. The only connection to Western Works was that the meeting took place in Pittsburgh, not Philadelphia. However, this meeting took place at the Pittsburgh airport and not at the Duquesne Club, where the Western Works conspirators met. Only those who were involved with bid rigging at Fairless were present [546a-47a; 621-22a].11

In sum, while there are a few record facts which are consistent with the hypothesis of an overarching conspiracy, no one has provided a persuasive reason why the defendants would have been interested in joining one and the evidence as a whole cumulatively compels a conclusion that they did not. Accordingly, proceedings should go forward on the July 19, 1984 indictment.

ADAMS, Circuit Judge, dissenting.

Defendants' guilt is not at issue in this appeal. Indeed, it is conceded by defendants that for many years they engaged in the rigging of bids for electrical contracting projects at various steel plants and oil refineries throughout Pennsylvania, in violation of the antitrust laws. What is at issue here is whether the indictment of defendants for bid-rigging in the present case, following their convictions or nolo contendere pleas in previous prosecutions for bid-rigging, violates their constitutional right not to be twice placed in jeopardy for a single crime.

The Fifth Amendment of the Constitution provides that no person shall "be subject for the same offense to be twice put in jeopardy of life or limb." The bar against double jeopardy was taken from the English common law, where the concept developed as early as the thirteenth century. See, e.g., United States v. Wilson, 420 U.S. 332, 339-40 & n. 6, 95 S. Ct. 1013, 1019-20 & n. 6, 43 L. Ed. 2d 232 (1975); Bartkus v. Illinois, 359 U.S. 121, 151-55, 79 S. Ct. 676, 695-697, 3 L. Ed. 2d 684 (1959) (Black, J., dissenting); Kirk, "Jeopardy" During the Period of the Year Books, 82 U.Pa.L.Rev. 602 (1934). Among other things, the clause was designed to protect against a second prosecution for a single offense and multiple punishments for the same offense. See North Carolina v. Pearce, 395 U.S. 711, 717, 89 S. Ct. 2072, 2076, 23 L. Ed. 2d 656 (1969). The Supreme Court has repeatedly stressed that the guarantee against double jeopardy is "a fundamental ideal in our constitutional heritage." Benton v. Maryland, 395 U.S. 784, 794, 89 S. Ct. 2056, 2062, 23 L. Ed. 2d 707 (1969).

The opinion announcing the judgment of the Court concludes that the district court erred both in fact and in law in determining that defendants were potential competitors at the various facilities at which bid-rigging was practiced, and that their bid-rigging activities were conducted pursuant to a single over-arching conspiracy in violation of section one of the Sherman Act, 15 U.S.C. Sec. 1 (1982). It asserts that defendants and their alleged co-conspirators were not competitors at each of the locations, and holds as a matter of law that noncompetitors cannot conspire in restraint of trade. The opinion thus determines that the conspiracy charged in this prosecution and those charged in prior indictments are not the same, and that defendants' double jeopardy rights therefore are not abridged. Because I disagree with that legal analysis, and with the application of such an analysis to the facts present in the record, and because I differ from the conclusion of the concurring opinion that this Court is free to disturb the trial court's finding of a single conspiracy, I respectfully dissent.

Defendants Lord Electric Company (Lord) and Sargent Electric Company (Sargent) were indicted in July 1983 for rigging bids at the Western Pennsylvania Works of United States Steel Corporation. Also indicted were Fischbach & Moore, Inc., the parent corporation of W.V. Pangborne & Co., Inc. (Pangborne), a defendant in this action, and a number of other electrical contracting companies and their individual officers. On March 7, 1984, Lord, Sargent, and Fischbach & Moore were convicted of violating section one of the Sherman Act and each was fined $1,000,000, the maximum penalty allowable under the Act. See United States v. Fischbach & Moore, Inc., 750 F.2d 1183 (3d Cir.1984), cert. denied, --- U.S. ----, 105 S. Ct. 1397, 84 L. Ed. 2d 785 (1985).

Defendants Pangborne and J.A. Bruce Pinney, the Executive Vice President of Pangborne, were indicted on March 8, 1984 on charges of rigging bids for electrical contracting work at the Philadelphia refinery of Gulf Oil Company, in violation of section one of the Sherman Act. They pleaded nolo contendere to the charges and were sentenced on June 19, 1984.

The indictment underlying the present prosecution was returned by a grand jury in June 1984, on the same day that Pangborne and Pinney were sentenced for their bid-rigging activities at the Gulf Oil refinery. It charged Pangborne, Pinney, Lord and Sargent with conspiring to rig bids at the Fairless Hills plant of United States Steel Corporation, in violation of section one of the Sherman Act. Prior to trial, all four defendants moved to dismiss the indictment on the ground that the activities leading to their previous convictions or nolo pleas were part of the same conspiracy as the bid-rigging alleged in the Fairless Hills indictment. They thus contend that the second indictment for the same offense under the Sherman Act, based upon identical evidence of conspiracy, infringed their rights under the double jeopardy clause. After conducting an extensive hearing on the issue, the district court found as a fact that the bid-rigging conducted at the Western Pennsylvania Works, the Gulf Oil refinery, and the Fairless Hills plant, as well as that at several other Pennsylvania facilities,1 had taken place pursuant to a single, broad conspiracy in which Lord, Sargent, Pangborne and Pinney had been participants. It therefore granted defendants' motion to dismiss the indictment on the basis of double jeopardy.

As explained in Judge Gibbons' opinion, each of the companies that solicited bids from defendants and their co-conspirators maintained selected bidders' lists. When a project became available, only those electrical contracting firms currently named on the bidders' list were invited to submit bids. The same names did not always appear on a given company's list, although some names were included more often than others. Because of the policy of employing selected bidders' lists, competition for any given project was limited to those electrical contractors who were invited to submit bids.

Judge Gibbons' opinion asserts that the use of bidders' lists by the targets of the bid-rigging scheme precludes as a matter of law a finding that the bid-rigging conducted at Fairless Hills, the Western Pennsylvania facilities, and Gulf Oil was embraced by a single conspiracy in which all four defendants were participants. It states that, because the bidders' lists restricted the scope of competition for a project to those named in the lists, a conspiracy between contractors "qualified" to bid and those not invited to bid could not in any way restrain trade. Since section one of the Sherman Act prohibits only conspiracies in restraint of trade, the argument continues, it follows that qualified bidders and unqualified bidders cannot be parties to a single conspiracy to rig bids.

The lone case cited that arguably adopts the position espoused by Judge Gibbons, which is of course crucial to his conclusion, is United States v. Ashland-Warren, Inc., 537 F. Supp. 433, 443, 445 (M.D.Tenn.1982). There, Ashland-Warren argued that its previous convictions for bid-rigging in the highway paving industry in Virginia barred its prosecution for similar bid-rigging in Tennessee. The district court rejected the argument, stating that, as Virginia paving contractors did not compete with Tennessee contractors for paving jobs, they could not all be participants in a single bid-rigging conspiracy.

Ashland-Warren is distinguishable from this case on a number of grounds. First, in Ashland-Warren, the only contractor that had participated in both the Virginia and the Tennessee bid-rigging schemes was Ashland-Warren. Id. at 440. With this minimal degree of overlap, there could have been no showing of interdependence between the two schemes, such as was made here. Second, the highway paving industry is significantly different from the electrical contracting industry in that the geographical area within which a paving contractor may provide services is limited by the nature of its business. As the Ashland-Warren court pointed out, since hot asphalt cannot be hauled more than forty miles before it hardens, the location of a paving contractor's plant imposes geographical restrictions upon the area within which it can compete. Id. at 438, 443. In light of these conditions, Ashland-Warren concluded that Virginia paving contractors and Tennessee contractors did not compete for the same jobs. Id. The record in this case reveals no such geographical constraints upon defendants' ability to compete.

Because of its unique facts, Ashland-Warren's statement that only competitors can conspire to restrain trade is subject to varying interpretations. However, it is significant that at least two courts have expressly declined to follow Ashland-Warren. See United States v. Waldbaum, Inc., 612 F. Supp. 1307, 1313 (D.Conn.), aff'd sub nom. United States v. Korfant, 771 F.2d 660 (2d Cir.1985); United States v. Beachner Construction Co., 555 F. Supp. 1273, 1282 (D.Kan.1983), aff'd, 729 F.2d 1278 (10th Cir.1984).

Even assuming that Judge Gibbons' view that noncompetitors can never conspire to restrain trade has some validity in theory, it seems unlikely that a situation would actually arise in which two conspirators that are truly not in competition would agree to engage in an anticompetitive practice when that agreement could in no way benefit either conspirator.2 Any such agreement to restrain trade would be ineffectual absent the participation of other conspirators who were in competition with both of the noncompeting parties. Only the involvement of such additional conspirators would create the potential for successful anticompetitive practices.

To illustrate, suppose contractors A and B conspire to rig bids with C at one facility, with D at a second facility, and with E at a third.3 Further suppose that C, D, and E would each submit bids at all three locations if they could, but are presently limited to bidding at a single location by the facilities' policy of using selected bidders' lists. Under the majority's analysis, each of these instances of bid-rigging is necessarily the object of a separate conspiracy to restrain trade, as C, D, and E do not compete with one another for contracts. This is so even if the three facilities are situated in the same city. Thus, under the majority's view, contractors A and B could be indicted three times, once for their activities at each of the three plants.

Yet, on the hypothesized facts, it is clear that the continuing cooperation of A and B is essential to successful bid-rigging at all three locations. If a dispute arises between A and B at the first location, which results in a rupture in their conspiratorial relationship, that rupture will cause them to cease rigging bids at the second and third locations as well. Thus, the effects of any disagreement or lack of trust between A and B will be felt by their three co-conspirators, regardless of whether the dispute arises out of bid-rigging at only one facility. Because this is so, although C, D, and E do not rig bids at all three plants, each has an interest in and will encourage bid-rigging at each location. This shared purpose that bid-rigging shall occur at all three plants is even stronger if C, D, and E have reason to believe that changes in the selected bidders' lists may result in their becoming eligible to bid at each facility. In my view, if the bid-rigging at the various plants is the purpose and result of a single agreement among the five contractors, the existence of selected bidders' lists does not preclude the conclusion that there is a single conspiracy to restrain trade both under section one of the Sherman Act and under the double jeopardy clause.

The hypothetical fact pattern set forth above fairly describes the situation presented by the record in the case at hand. While it may be true that neither Lord nor Sargent competed for contracts at the Gulf Oil refinery, both of them submitted bids at the Western Pennsylvania Works and at Fairless Hills, and thus competed with a number of other bidders at the latter two locations. Likewise, although Pangborne did not compete for bids at the Western Pennsylvania plants (its parent corporation, Fischbach & Moore, Inc., did, however), it competed at Gulf Oil and at Fairless Hills. In short, there was considerable overlap among the bidders competing for projects at the three locations specified in the indictments here, as well as among the bidders competing for jobs at all of the facilities that the district court found were within the scope of the bid-rigging conspiracy. As a consequence, Lord and Sargent had an interest in the continued successful rigging of bids and in the cooperation of competitors at Gulf Oil despite their lack of participation at that location. This was so because if the conspirators at Gulf Oil were unable to rely upon one another to rig bids at that location, they would inevitably withdraw from the bid-rigging at other locations where Lord and Sargent did compete.

Indeed, the district court found that two incidents occurred which showed that suspicions regarding the reliability of one conspirator at one location would affect the viability of the conspiracy at other locations. In one instance, Sargent inadvertently underbid the designated low bidder for a job at Fairless Hills. As a result, a meeting was immediately held to resolve the matter and to determine whether Sargent could still be "counted in." The district court found that "[t]he parties recognized that if Sargent could no longer be trusted, agreements could not be reached on any contract Sargent was a party to, both at the Fairless works and at the Western Pa. works." App. at 192-93 (emphasis added). The district court thus found that, before bid-rigging could continue at any location at which Sargent was a bidder, Sargent's co-conspirators had to be reassured of its reliability. If bid-rigging had ceased at both Fairless Hills and the Western Pennsylvania plants because of Sargent's lapse, even contractors not qualified to bid at Fairless Hills, where the incident occurred, would have been affected adversely.

The second indication of reciprocity that the district court found to have taken place involved an occurrence when Tri-City Electric Company, a contractor who bid at Fairless Hills and the Western Pennsylvania Works, was denied the position of designated low bidder for a project at one location, and in response, withdrew from the bid-rigging at both locations. Again, the relations of the conspirators at one plant affected those at the other plants, regardless of whether they were qualified to bid at the first plant. On these facts, the district court could reasonably have found, as it did, that the bid-rigging practiced at the various locations was interdependent, and that the success of each instance of bid-rigging depended to some extent upon the success of the bid-rigging elsewhere. Thus, although all the conspirators did not directly compete with one another for each job at each location, the district court reasonably determined that they all did share a common, continuing objective to rig bids, and that each was benefited by the continued successful rigging of bids, at all locations.

Furthermore, the district court found that, despite the target companies' policy of maintaining selected bidders' lists, all conspirators were potential competitors. It reasoned that the contractors "were all available and capable of being added to the bid list at each facility," and that if "added to the list, they would then [have been] eligible to bid for jobs." App. at 194. Judge Gibbons rejects this finding, stating that there was no potential for competition where an unqualified bidder could not "by some action within its own control or the control of the other alleged conspirators" have its name placed on a selected bidders' list. Ante at 1129.

The linchpin of Judge Gibbons' reasoning is that because inclusion on the bidders' lists was controlled by the target corporations rather than by the conspirators, a finding that the conspirators were potential competitors is foreclosed as a matter of law or is clearly erroneous. There was evidence that the make-up of the lists fluctuated over the period of the conspiracy. These variations in the bidders' lists raised the possibility that contractors who had been excluded from the bidding at a given location might be added to the list, and thereby become qualified to bid. Under these circumstances, a conspirator could be said to be a potential competitor at all locations at which it was physically able to provide its services. Cf. Ashland-Warren, 537 F. Supp. at 443. And, perhaps more importantly, each conspirator would have an additional purpose in encouraging continued bid-rigging at all locations because of the possibility that a change in the bidders' list of one of the target companies might make the contractor eligible to bid, and thus to rig bids, at that target's facility.

Because I disagree with Judge Gibbons' conclusion that the district court's finding of a single conspiracy was reached under an erroneous application of principles of antitrust law, it remains to address the question whether the district court improperly concluded that the evidence established a violation of defendants' double jeopardy rights. As Judge Gibbons concedes, since defendants made a nonfrivolous showing that their double jeopardy rights were infringed by the indictment at issue, the government bore the burden of proving by a preponderance of the evidence that the indictment was not based upon the same offense for which defendants had already been indicted and sentenced. See United States v. Felton, 753 F.2d 276, 278 (3d Cir.1985); United States v. Inmon, 568 F.2d 326, 332 (3d Cir.1977) (Inmon I). It is also conceded that we may reject the district court's findings of historical fact only if clearly erroneous, and that we must affirm the district court's finding of a single conspiracy unless the "inference of a single conspiracy was on this record clearly erroneous as a matter of fact or legally impermissible." Ante at 1126; see Felton, 753 F.2d at 278; Inmon I, 568 F.2d at 332.

This Court has adopted the "same evidence" test for assessing double jeopardy claims. See United States v. Young, 503 F.2d 1072, 1076 (3d Cir.1974). Under this test, a second indictment for an offense violates the double jeopardy clause "only when the evidence required to support a conviction upon [the offense] would have been sufficient to warrant a conviction upon" a prior indictment. Id. at 1075 (quoting United States v. Pacelli, 470 F.2d 67, 72 (2d Cir.1972)). However, Young cautioned that application of the same evidence test "must be tempered ... with the consideration that a single conspiracy may not be subdivided arbitrarily for the purposes of prosecution." Id. As we explained there, "[d]ifferent alleged overt acts are not necessarily inconsistent with an improper division of a single conspiracy into multiple crimes. It is the agreement which constitutes the crime, not the overt acts." Id. at 1076. See also Felton, 753 F.2d at 278.

Other courts, noting the difficulties of applying the same evidence test in assessing claims that a second indictment on a charge of conspiracy violates the double jeopardy clause, have adopted a "totality of the circumstances" standard. See, e.g., Korfant, 771 F.2d at 662; United States v. Thomas, 759 F.2d 659, 661-62 (8th Cir.1985); United States v. Sinito, 723 F.2d 1250, 1256 (6th Cir.1983); cert. denied, --- U.S. ----, 105 S. Ct. 86, 83 L. Ed. 2d 33 (1984); United States v. Castro, 629 F.2d 456, 461 (7th Cir.1980); United States v. Marable, 578 F.2d 151, 154 (5th Cir.1978). Under the totality of the circumstances approach, a court will consider the degree of overlap between the two charged conspiracies in the following factors: (1) the criminal offenses charged in the successive indictments; (2) the participants; (3) the time periods involved; (4) similarity of operation; (5) the overt acts alleged; (6) the geographic scope of the alleged conspiracies or locations where overt acts occurred; and (7) the objectives of the alleged conspiracies. Additionally, it will look to whether any other signs of interdependence between the two charged conspiracies exist. See Korfant, 771 F.2d at 662; Thomas, 759 F.2d at 661-62. The ultimate question under the totality of the circumstances approach, as under the same evidence test, is "whether there [was] more than one agreement." Sinito, 723 F.2d at 1256. See also Beachner Construction Co., 729 F.2d at 1281 (bid-rigging case); Braverman v. United States, 317 U.S. 49, 53, 63 S. Ct. 99, 101, 87 L. Ed. 23 (1942) ("The gist of the crime of conspiracy is the agreement."); Young, 503 F.2d at 1076.

Although this Court has thus far declined expressly to adopt the totality of the circumstances approach, Felton, 753 F.2d at 281, it has followed that approach in practice, see id. at 278-81; United States v. Inmon, 594 F.2d 352, 353-54 (3d Cir.), cert. denied, 444 U.S. 859, 100 S. Ct. 121, 62 L. Ed. 2d 79 (1979) (Inmon II). And it would appear that consideration of all relevant facts is necessary to a determination whether two charged offenses are "in law and in fact the same offense." Young, 503 F.2d at 1075 (citing United States v. Ewell, 383 U.S. 116, 124-25, 86 S. Ct. 773, 778, 15 L. Ed. 2d 627 (1966)). I therefore believe that the district court properly considered all relevant circumstances in determining that the government had failed to carry its burden of establishing by a preponderance of the evidence that the Fairless Hills bid-rigging and that conducted at Gulf Oil and in Western Pennsylvania were the subjects of separate conspiracies rather than one over-arching conspiracy.

As the district court observed, the time spans of the three bid-rigging schemes charged overlapped "for a substantial part of the period between 1971 and 1981," and "[t]he arrangements at each facility ended at approximately the same time." App. at 187. It also found that, to the extent that the arrangements at each facility ended at different times, they did so because work was drying up at the different locations at different times, an event over which the conspirators had no control. Thus, any variance in the termination dates of the bid-rigging at the different plants did not indicate a lack of interdependence among the arrangements.

The district court further noted that the participants at the different facilities overlapped substantially. The overlap is even more extensive when one considers Fischbach & Moore and Pangborne, its subsidiary, as one participant.4 The court dismissed the government's argument that not all participants were active in every area of the conspiracy and that not all participants knew each other by correctly pointing out that members of a conspiracy need not know or have direct contact with each other. App. at 188; Blumenthal v. United States, 332 U.S. 539, 550, 68 S. Ct. 248, 253, 92 L. Ed. 154 (1947); United States v. Consolidated Packaging Corp., 575 F.2d 117, 127 (7th Cir.1978). It stated that:

In accordance with the operating agreement in place here, there was no need to inform every member about every action taken in furtherance of the conspiracy. The only time it was necessary for a contractor to know of bid rigging activities on a particular project was when it was on the bid list for that project.

A single conspiracy may be made up of a small group of core conspirators who interact with different sub groups at different times to achieve an overall plan. United States v. Boyd, 595 F.2d 120 (3d Cir.1978). This is what occurred here. The core group, made up of Sargent Electric, Lord Electric, Foley, Ortlip and Fischback [sic] and Moore with its subsidiary Pangborne, conspired together and with different sub groups to eliminate competitive bidding in each location at which members of the conspiracy bid. The agreement was not limited to the Fairless works or the Western Pa. works or the Gulf refinery, it extended to every facility where bidders on a project were members of the conspiracy. Whenever a new bidder was added at a facility, the contractor simply plugged into that aspect of the conspiracy.

App. at 189.

Judge Gibbons' opinion states that it is "undisputed ... that the Foley Co. did not rig bids at Western Works or at Lukens Steel while it participated in bid-rigging at Gulf's Philadelphia refinery and at Fairless Hills," and that "while Lord rigged bids at Western Works and at Fairless Hills, it did not do so at Lukens Steel." Ante at 1130. The first of these assertions is clearly disputed: as the government concedes, and the record indicates, the Foley Company did rig bids at the Western Pennsylvania facilities. See Brief for Appellant at 6 n. 11. Furthermore, although it is true that Foley did not rig bids on the intermittent occasions when it was invited to submit bids at Lukens Steel, the district court explained that there was no evidence that the bid-rigging had spread to the Lukens Steel plant at the time Foley bid competitively. App. at 191.

As for Lord's nonparticipation at Lukens Steel, Lord only bid at that facility once or twice in the early 1970's, both times through a general contractor, and there is no evidence as to the identities of the other bidders on those occasions. App. at 925, 1003. Again, in the face of Lord's admitted participation in the bid-rigging at Fairless Hills and the Western Pennsylvania facilities, where it was regularly included on the bidders' lists, its failure to rig one or two bids, submitted through a general contractor, at Lukens Steel is not sufficient to satisfy the government's burden of proof.

Judge Gibbons also stresses that separate records were kept of the allocation of jobs at each facility, and that the meetings at which the designated low bidders were chosen were held at a different location for each facility. But as the district court explained, these were merely adaptations of the common practice which were necessitated by the individual characteristics of the different target companies. Only Gulf Oil consistently used the same list of selected bidders, so only there could a regular rotation be used to allocate jobs. At the other facilities, the conspirators tried to apportion the work fairly "based upon the value of the contract and who had been awarded the previous contracts, with a variation in the order of rotation available 'if someone needed the work.' " App. at 174. And surely the fact that the meetings to allocate jobs were held at locations convenient to the area participants is of little help to the government in establishing the existence of multiple conspiracies. In short, on this record the district court reasonably could have concluded that there was a single bid-rigging conspiracy that encompassed the activities at Fairless Hills, which underlie the indictment at issue, and the activities at the Western Pennsylvania facilities and Gulf Oil, for which defendants have already been sentenced.

Judge Stapleton in his concurring opinion concludes that "the district court's finding of an overarching conspiracy is clearly erroneous." Ante at 1132. He argues that the record "either strongly favors or is consistent with a finding of multiple conspiracies," id., and that the facts relied upon by the trial court "do not provide unambiguous evidence of an overarching conspiracy." Id. at 1133. I do not disagree that the record might support an inference that multiple conspiracies existed. That, however, is beside the point. The government bore the burden of proving the existence of multiple conspiracies, and thus the absence of a double jeopardy violation, by a preponderance of the evidence. See Felton, 753 F.2d at 278; Inmon I, 568 F.2d at 332. Where the record might reasonably support either the inference of multiple conspiracies or the inference of a single conspiracy, and the government had the burden of proving individual and separate conspiracies, it cannot be said that the district court was clearly erroneous in concluding that the burden was not sustained. The district court, sitting as the finder of fact, concluded after a full hearing at which witnesses testified regarding the nature of the conspiracy, that defendants' activities were carried on pursuant to a single all-encompassing conspiracy. As the Supreme Court has recently emphasized,

If the district court's account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently. Where there are two permissible views of the evidence, the factfinder's choice between them cannot be clearly erroneous.

Anderson v. City of Bessemer City, --- U.S. ----, 105 S. Ct. 1504, 1512, 84 L. Ed. 2d 518 (1985).

The district court's view of the evidence clearly is a plausible one. Furthermore, it is especially inappropriate to disturb the district court's finding where, as here, that finding is consonant with protection of defendants' constitutional rights. I would therefore hold that the district court did not err in dismissing the indictment on the ground that it infringed defendants, double jeopardy rights.

It is essential to emphasize that this is a case of constitutional dimensions. The policies underlying the double jeopardy clause are fundamental to our system of justice. As the Supreme Court has declared, "[W]hen a defendant has been once convicted and punished for a particular crime, principles of fairness and finality require that he not be subjected to the possibility of further punishment by being again tried or sentenced for the same offense." Wilson, 420 U.S. at 343, 95 S.Ct. at 1021. In the present case, each of the defendants has already been indicted and sentenced for participating in virtually identical criminal activities to those charged in the indictment that initiated this prosecution. I would hold that principles of fairness and finality militate against allowing defendants to be placed in jeopardy a second time.

Because I believe that the antitrust principles applied in the opinion announcing the judgment of the Court are not applicable in this case, and disagree as to the evaluation of the evidence and the weight to be assigned to certain portions of the evidence, I would affirm the district court order dismissing the indictment.