On May 23, 1986, an eighteen-count indictment was filed in the Western District of Pennsylvania. The indictment charged George Thornton and others known and unknown with conspiracy to possess with intent to distribute more than 1,000 pounds of marihuana in violation of 21 U.S.C. §§ 841(a)(1) and 846. According to the indictment and the allegations at trial, this conspiracy ran from March 1, 1981, to approximately May 14, 1981. This indictment also charged Thornton with three other counts of various related substantive offenses. A jury tried and acquitted Thornton of all charges.

Then, a little over one year later, on September 29, 1987, a two-count indictment was filed in the Southern District of Illinois charging Thornton with conspiracy to knowingly and intentionally distribute more than 1,000 pounds of marihuana in violation of 21 U.S.C. §§ 841(a)(1) and 846. This indictment alleged that the conspiracy spanned from March 1980 through June 1987.

Thornton moved to dismiss the conspiracy charge under the Illinois indictment on double jeopardy grounds, arguing that the prosecution under the Illinois indictment puts him twice in jeopardy for the conspiracy alleged in the Pennsylvania indictment. The district court denied this motion. We affirm.

The double jeopardy clause of the Fifth Amendment to the United States Constitution provides: "[N]or shall any person be subject for the same offence to be twice put in jeopardy of life or limb." This means that the United States cannot twice prosecute the same person for the same offense. Deciphering what constitutes prosecution for the same offense for purposes of double jeopardy is not an easy task. And, the Supreme Court and this court have recognized that this task becomes even more difficult when we move from single layered crimes such as bank robberies to prosecution for multilayered crimes such as conspiracies which expand over time and place. See United States v. Felix, --- U.S. ----, 112 S. Ct. 1377, 1385, 118 L. Ed. 2d 25 (1992); United States v. O'Connor, 953 F.2d 338 (7th Cir.), cert. denied, --- U.S. ----, 112 S. Ct. 1979, 118 L. Ed. 2d 578 (1992). The reason for the added complexity is that it is difficult to apply double jeopardy's notions of finality to crimes which have no easily discernable boundaries with regard to time, place, persons, and objectives.

Both the Supreme Court and this court have recently faced the task of applying double jeopardy's notions of finality to continuing, multilayered crimes. The Supreme Court in Felix and this court in O'Connor addressed whether it violated double jeopardy to prosecute for underlying substantive crimes and then to later prosecute the defendant for multilayered crimes which somehow involve these underlying crimes. More specifically, the Supreme Court in Felix addressed whether it violated double jeopardy to first prosecute Felix for attempting to manufacture methamphetamine in Missouri and then later prosecute Felix for conspiracy to manufacture, possess, and distribute methamphetamine where some of the alleged overt acts in the conspiracy case were based on the attempt charge in Missouri. The Court held it did not. Felix, --- U.S. at ----, 112 S. Ct. at 1385. Likewise, in O'Connor, a decision which preceded Felix, this court addressed whether it violated double jeopardy to first prosecute for substantive offenses and then later rely on these substantive offenses as predicate acts in a prosecution for engaging in a pattern of racketeering activity under the Racketeer Influenced and Corrupt Organizations Act, 18 U.S.C. § 1962(c). We held that it did not. O'Connor, 953 F.2d at 344-45.

Both Felix and O'Connor indicate that the multilayered, continuing, criminal conduct at issue in those cases was distinct from the underlying substantive offenses, meaning that the government could prosecute persons in two separate proceedings for the underlying substantive conduct and the multilayered, continuing criminal conduct without violating the United States Constitution's protections against double jeopardy. See Felix, --- U.S. ----, 112 S. Ct. at 1384-85; O'Connor, 953 F.2d at 344-45. This conclusion, however, says nothing about when it constitutes double jeopardy to prosecute the same person for conspiracy under the same statute for what may be one or two separate agreements. And, for a discussion more directly applicable to this decision we turn to United States v. Castro, 629 F.2d 456 (7th Cir.1980). In Castro we held that the double jeopardy clause prohibits the government from arbitrarily subdividing one conspiracy into several and then prosecuting a person multiple times for what essentially constitutes one conspiracy. Castro, 629 F.2d at 461. The rationale underlying this proposition is simple: the double jeopardy clause prohibits multiple prosecutions for the same offense, and because the agreement is the sine qua non of conspiracy, if the government twice prosecutes an individual under the same statute for what essentially constitutes one agreement, this must constitute prosecution for the same offense in violation of double jeopardy. See United States v. Chiattello, 804 F.2d 415, 418 (7th Cir.1986). Indeed, the Felix decision, which emphasizes that the essence of conspiracy is the agreement, supports this principle. See Felix, --- U.S. at ----, 112 S. Ct. at 1384.

The government, however, seems to be arguing for a deviation from the principles set forth in Castro. The government has taken great pains to emphasize that the conspiracy alleged in the Pennsylvania indictment lasted only a few months, involved many fewer people, and was therefore much smaller in scope than the conspiracy alleged in the Illinois indictment, which involved some forty plus coconspirators, trafficking to numerous states, and encompassed a seven-to-nine-year time frame. Moreover, the government emphasized in the hearing before the district judge that the agent involved in the Pennsylvania indictment knew nothing about the activities alleged in the Illinois indictment. It appears that in making such arguments the government is implying that even assuming that the Pennsylvania indictment charged the same conspiracy as the Illinois indictment, there is no double jeopardy problem because the first-charged conspiracy was only a small subset of the later-charged conspiracy and because the government did not know that this was one conspiracy. We must remember, however, the double jeopardy clause imposes limits on a defendant's criminal exposure. In order to stay true to these finality requirements, the government cannot reprosecute a defendant for the same offense whenever it obtains broader evidence of criminal culpability.

We now move to the question at the heart of this case: whether there has been a sufficient showing that both prosecutions alleged the same conspiracy. The first issue we confront in this context is who bears the burden of proof. We indicated in Castro that this burden belongs to the defendant. Castro, 629 F.2d at 461. We further explained in United States v. Dempsey, 806 F.2d 766, 768 (7th Cir.1986), cert. denied, 481 U.S. 1014, 107 S. Ct. 1889, 95 L. Ed. 2d 497 (1987), that the defendant bears the burden of making a prima facie showing that the two indictments cover the same offense, and thereafter the burden shifts to the government to demonstrate that it has not twice prosecuted the defendant for the same conspiracy. Id.

In Castro we held that in order to determine if the government's prosecution of two conspiracies violates double jeopardy, courts should look at the indictments and other evidence to determine whether the two alleged conspiracies contain an overlap in time, alleged coconspirators, overt acts and/or places. Castro, 629 F.2d at 461. And, in other decisions this court has also looked to such factors as whether the conspiracies mutually support each other, whether the modus operandi are similar, and the specific nature of the objectives embodied in each conspiracy. See United States v. Cerro, 775 F.2d 908, 914 (7th Cir.1985) (indicated that mutual support is relevant to the single versus double conspiracy analysis); Chiattello, 804 F.2d at 419 (focused on the specific objectives of the criminal agreements and the modus operandi in achieving these objectives in concluding that the subsequent prosecution constituted a prosecution for a separate conspiracy). Thornton argues that he has demonstrated an overlap in conspirators, place, and time. Thornton, therefore, argues that he has made a prima facia showing that the prosecution under the Illinois indictment constitutes double jeopardy.

In order to understand Thornton's argument we first look to the scope of the conspiracy alleged in the Illinois indictment. The Illinois indictment alleges a broad conspiracy to knowingly and intentionally distribute more than 1,000 pounds of marihuana in violation of 21 U.S.C. §§ 841(a)(1) and 846. The indictment alleges that over forty persons participated in this conspiracy (the "Lanier conspiracy," "Lanier network," or "Lanier operation") over a seven-year period beginning around March 1980 and continuing until about June 1987. Moreover, the indictment alleges that the conspirators' activities spanned numerous states including Illinois, Kentucky, Missouri, Pennsylvania, Michigan, Indiana and elsewhere.

Now, unlike many double jeopardy cases where the double jeopardy motion is entertained before the trial of any of the coconspirators, in this case Michael J. Canino, James G. Markum, John G. Flynn, and David L. Malkin had already been tried and convicted under the same indictment as was rendered against Thornton. See United States v. Canino, 949 F.2d 928 (7th Cir.1991), cert. denied, --- U.S. ----, 112 S. Ct. 1701, 118 L. Ed. 2d 410, --- U.S. ----, 112 S. Ct. 1940, 118 L. Ed. 2d 546, --- U.S. ----, 112 S. Ct. 1954, 118 L. Ed. 2d 558 (1992) ("Canino trial"). And, Randy Lanier, Benjamin Kramer, Albert Fischer, and Kay Bell were tried in yet another trial under a second indictment for this same conspiracy. See United States v. Kramer, 955 F.2d 479 (7th Cir.1992) ("Lanier trial"). At these trials government witnesses indicated that coconspirators had imported and distributed marihuana from 1978 through 1987. That testimony, along with other evidence, indicated that the parties were involved in a large drug distribution network.

The government paints the Lanier conspiracy as one spearheaded by Randy Lanier, Benjamin Kramer, George Brock, and Eugene Fischer. According to the government, these individuals worked with other coconspirators to arrange for the distribution of large amounts of marihuana throughout the United States. Much of the marihuana distributed under the Lanier conspiracy was imported on barges and boats into various United States's ports from Colombia, South America. The testimony at the Canino and Lanier trials indicated that several of these barge and boat loads carried in excess of 120,000 pounds of marihuana. The record indicates that Kramer, Brock, Fischer, and Lanier would arrange for the importation of this marihuana from Colombia, and then these key players would arrange for the marihuana's distribution throughout a large, multi-state drug network. According to the government and the evidence at the Lanier trial, one of the major customers of Lanier included Ronald Ball. Ball then sold the marihuana to various persons, including defendant Thornton, Conrad Ignold from New Orleans, Louisiana, and Michael Canino from Philadelphia, Pennsylvania. And, these persons would then distribute the marihuana to others.

Thornton does not disagree that the parties took on the roles described by the government. Rather, Thornton argues that these roles were not as fixed as the government paints them, and Thornton points to different persons as the key players in the conspiracy. Thornton particularly emphasizes that the testimony indicates that the various players, including himself, moved in and out of different roles during the conspiracy, and the record, to some extent, supports this assessment. For example, the record indicates that Canino, who started out as a purchaser, eventually became more deeply enmeshed in other conspiratorial activities. That is, Canino eventually made his Allentown, Pennsylvania stash house available for storage, he supplied workers to help with the unloading of the shipments of marihuana from Colombia, and assisted with the transport of these shipments of marihuana to his stash house. Moreover, Thornton's role sometimes shifted from buyer to seller. That is, Thornton at times sold marihuana to Ball, and Ball would then sometimes distribute such marihuana to Lanier. Furthermore, the record indicates that Thornton sometimes sold directly to Lanier. Nonetheless, the government emphasizes that the bulk of the marihuana involved in the Lanier conspiracy--over 600,000 pounds--came from Colombia through Lanier, Kramer, Brock, and Fischer.

Now we turn to the Pennsylvania indictment and trial. The Pennsylvania indictment alleged that Thornton conspired with Thomas Shea, Jay Smith, Cesar Fernandez, Steven Petrone, Gregory Cooper, Charles McKinney, Richard Sellars, Alexander Sorce, Rick White, and Daniel Petrone, and others to possess with the intent to distribute in excess of 1,000 pounds of marihuana in violation of 21 U.S.C. §§ 841(a)(1) and 846. This indictment alleged that this conspiracy to distribute lasted from March 1, 1981, to around May 15, 1981, and that it occurred in the Western District of Pennsylvania and elsewhere.

During the Pennsylvania trial, the government alleged that the codefendants were parties to an agreement to distribute marihuana flown from Colombia, South America to the Pittsburgh, Pennsylvania area. In that case the government tried to demonstrate that Shea was involved with codefendants Fernandez and Smith, among others, in plans to smuggle several plane loads of marihuana into the Pittsburgh area. The government alleged that the first of these plane loads of marihuana left Florida to Colombia and went back to the Bahamas where it was seized before arriving to its intended destination in Johnstown, Pennsylvania. And, according to the government's evidence at the Pennsylvania trial, the second plane left Florida around April of 1981, went to Colombia, and returned to Johnstown, Pennsylvania. According to the government's allegations at trial, the marihuana was unloaded in Johnstown, Pennsylvania, taken to stash houses, and then sold to several individuals, including Thornton, Rick White, and Leo Sorce. During the Pennsylvania trial the government tried to prove that Thornton had knowingly entered an agreement with the individuals named in the indictment to distribute this Johnstown plane load of marihuana.

As the government has pointed out, other than Thornton, none of the players named in the Pennsylvania indictment or mentioned at the Pennsylvania trial were mentioned in the Illinois indictment or in the 14,000 pages of trial transcripts from the Lanier or Canino trials. The government argues that without such overlap, Thornton has failed to make a prima facia showing that the conspiracy charged under the Pennsylvania indictment (the "Shea conspiracy," "Shea network," or "Shea operation") was the same as that charged in the Illinois indictment (the "Lanier conspiracy," "Lanier network," or "Lanier operation").

Thornton, however, argues that because he and Ronald Ball (key players in the Lanier conspiracy) arranged for the sale of some of the marihuana from the Johnstown plane load to be sold to Canino (another key player in the Lanier conspiracy), transported by the Scimecas (two "drivers" and named codefendants in the Lanier conspiracy), stored at Canino's Allentown stash house (a stash house used in the Lanier conspiracy), within a time period encompassed within the Lanier conspiracy, he has demonstrated an overlap in time, geography, and coconspirators. And, according to Thornton, he has, therefore, made a prima facia double jeopardy showing. In short, Thornton argues that because some of this marihuana that he purchased from the Shea operation can be traced to the Lanier operation, he has made a prima facia double jeopardy showing.

Thornton has relied on the stipulated testimony of Ball in an attempt to make such a showing. This stipulated testimony reads as follows:

In April of 1981, Thomas Shea, while in Pittsburgh, Pennsylvania, telephoned George Thornton who was in Fort Lauderdale, Florida. Shea told Thornton to come to Pittsburgh to receive a shipment of marijuana. Thornton called Ronald Ball and told Ball of the conversation with Shea. Ball and Thornton agreed to fly to Pittsburgh to pick up the shipment. Ball and Thornton arrived in Pittsburgh and were met at the airport by a person known as "Boo." Boo drove them to the Greentree Marriott Hotel where Ball and Thornton stayed that night. The next day, Boo picked them up in a blue pickup truck and brought them to Monzo's Howard Johnson's Motor Inn in Monroeville, Pennsylvania. Ball and Thornton took Boo's blue truck to the stash house near Johnstown, Pennsylvania where they met Shea and his workers Boynton and Cooper. Ball attempted to negotiate a price of $300-350 per pound including a $25 per pound fee for him. Fifteen hundred pounds were loaded onto Boo's truck by Thornton and Ball. Ball contacted Michael Scimeca and Diane Scimeca and had them drive a brown Impala automobile to Monroeville to pick up 375 pounds of marijuana for delivery to Canino at his stash house in Allentown. Canino rejected the marijuana and had the Scimecas return it to Ball and Thornton.

We agree that this stipulated testimony clearly indicates that Thornton and Ball's arrangements to distribute 375 pounds of marihuana to Canino were activities taken in furtherance of the Lanier conspiracy, and were, therefore, acts taken as part of that agreement. And, by following the path of this sale of 375 pounds of marihuana it might appear that Thornton has met his burden under Castro and its progeny because of the overlap in time, place, and persons. Nonetheless, we must remember that the gist of the double jeopardy inquiry under Castro is whether the government has alleged the same conspiracy in both indictments. In making such a determination we cannot ignore case law outside the double jeopardy context that defines the scope of a conspiracy. One seminal case in this regard is United States v. Townsend, 924 F.2d 1385 (7th Cir.1991). In that case we indicated that it pushes the bounds of logic to assume that the farmers in South America, the smugglers, the importers, several levels of retailers, and the numerous dealers on the street are all parties to one single agreement to distribute drugs simply because they are all members of a single distribution chain and because each level in this chain has some general notion that the persons with whom they are involved deal with others in the sale or purchase of drugs. Townsend, 924 F.2d at 1391. And, we therefore concluded that in order to prove that persons at different levels of a drug distribution chain are parties to a single agreement to distribute drugs, the government must produce evidence beyond the fact that the individuals at each level had some general knowledge that the individuals with whom they dealt were somehow involved with others in the illegal distribution of drugs. Id. at 1392-93. In reaching such a conclusion, we explained that a conspiracy is essentially a business enterprise with illegal purposes that like a legal business enterprise is characterized by cooperative relationships between its members. Id. at 1394-95. And, we further explained that, just like other commodities, there is a limited supply of drugs, meaning that if the government could pool all levels of a distribution chain into one conspiracy without providing some evidence that the members intended to join in a cooperative association with a common goal, it would be just as likely, if not more likely, that the persons pooled together would be competitors rather than persons working in concert to achieve a common objective. Id.

In most cases it is the government that comes before the courts arguing that numerous individuals are members of a single conspiracy. Id. at 1388-89. And in those instances the government has the burden of proving to a jury beyond a reasonable doubt that what they have alleged constitutes a single conspiracy. Id. In this case, however, the tables are turned, and today it is the defendant who bears the burden of making a prima facia showing that the government has alleged the same conspiracy in both the Pennsylvania and Illinois indictments. In an attempt to meet this burden, Thornton has offered evidence that he and Ball arranged for 375 pounds of the marihuana from the Johnstown plane load to be used to further the later-charged Lanier conspiracy--Thornton apparently bought 1,500 pounds of marihuana from this plane load. The problem is that under Townsend tracing the distribution of the contraband is not enough. In other words, although Thornton eventually used some of the marihuana purchased from the Shea operation to benefit the Lanier operation, it does follow that Thornton's agreement to buy and distribute from the Shea operation was part of the same agreement that he had with the Lanier operation. Indeed, under Townsend, if Thornton wants to demonstrate that the Shea operation was but part of the same conspiracy as the Lanier operation, Thornton must make some showing that the Shea agreement's objectives went far beyond an agreement to distribute one or two plane loads of marihuana. Rather, Thornton must demonstrate that the parties to the Shea conspiracy either explicitly or implicitly intended to join the Lanier conspiracy--a conspiracy which involved a multi-state, widescale distribution of 600,000 plus pounds of marihuana. It is hard to envision how Shea's single call to Thornton and eventual sale of 1,500 pounds of marihuana to Thornton indicated that Shea or his associates intended to join the widescale Lanier distribution network. Certainly, this evidence alone is not sufficient to make a prima facia showing of a single conspiracy under Townsend, and Thornton has provided nothing more. That is, Thornton has failed to allege facts or produce evidence indicating that the coconspirators in the Shea operation had any real interest or stake in the success of the Lanier conspiracy. Yes, Thornton has indicated that he sometimes acted as a supplier in his dealings with the Lanier network, yet Thornton has failed to allege facts or submit evidence indicating that any of the individuals in the Shea network acted as his regular or even frequent suppliers for these sales. Indeed, there is no indication in the record that this sale of 1,500 pounds of marihuana from the Johnstown plane load was anything but a onetime transaction between Thornton and the Shea network. Moreover, Thornton has failed to allege facts or provide any evidence indicating that there was a cooperative relationship between the Shea network and the Lanier network. And, under Townsend the mere fact that Thornton and Ball were associated with both the Lanier and Shea networks does not alone turn these otherwise separate operations into one large conspiracy.

For the above reasons, we conclude that Thornton has failed to make a prima facia showing that his prosecution under the Illinois indictment violated the Fifth Amendment's double jeopardy clause. We, therefore,

AFFIRM.