JACOBS, Circuit Judge:
Defendants-appellants John Barberio and Marat Balagula (collectively, the “defendants”), having previously been convicted for their roles in a five-month scheme to defraud the United States of federal gasoline excise taxes, now interpose the double jeopardy clause of the Fifth Amendment to bar the present prosecution for their roles in a broad-gauge tax evasion conspiracy employing similar techniques on a greater scale, over a period of years that wholly subsumes the period of the first conspiracy.
The United States District Court for the Eastern District of New York (Wexler, J.) conducted a thorough analysis under United States v. Korfant, 771 F.2d 660 (2d Cir.1985), and determined that the two indictments describe distinct conspiracies and therefore do not subject the defendants to double jeopardy. See United States v. Macchia, 845 F.Supp. 953 (E.D.N.Y.1994). We affirm.
BACKGROUND
The facts of this appeal are set forth in every useful detail in Judge Wexler’s opinion. [665]*665We recapitulate only the facts we deem dis-positive.
A. The Tarricone Indictment
In January 1992, a federal grand jury returned an indictment charging Balagula, Bar-berio and three co-defendants — Arthur Tar-ricone, Dominic A. Bombaee, and John Pa-bone — with one count of conspiring to evade the federal gasoline excise tax in violation of 18 U.S.C. §§ 371 and 3623, and two counts of attempted excise tax evasion in violation of 26 U.S.C. § 7201 and 18 U.S.C. .§§ 2 and 3623 (the “Tarricone Indictment”). .
The Tarricone Indictment described the alleged evasion of over $400,000 in federal gasoline excise taxes between November 1985 and April 1986. At that time, federal law imposed an excise tax of nine cents per gallon on the sale of gasoline, but exempted sales between gasoline distributors holding a Registration for Tax-Free Transactions issued by the Internal Revenue Service on its “Form 637”. In tax parlance, a holder of a Form 637 was a “licensed” company; a company without such a form was “unlicensed.” Transactions between licensed companies were “tax-free” because the seller was not required to pay the federal excise tax. The one-time payment of the excise tax on any gasoline shipment occurred when it passed from a licensed company to an unlicensed one. In that transaction, the licensed seller incurred the obligation to pay the tax. Since the licensed seller generally included the excise tax in the sales price, such sales were known as “tax included” or “tax paid” transactions. Subsequent sales of the same gasoline were exempt because the one-time excise on that gasoline had been paid.
The defendants named in the Tarricone Indictment evaded the federal excise tax with respect to the sale of some 41 million gallons of gasoline from Arthur Tarricone, Inc. (“AT”), a licensed supplier, to unlicensed purchasers. This was accomplished by false invoices that created a paper trail — or “daisy chain” — of fictitious transactions. Through these fictitious transactions, the gasoline appeared to be sold first to a licensed company, and then, in apparently tax-paid transactions, to unlicensed companies. However, .the licensed company that made the paper sale to the unlicensed purchasers was a shell company, consisting of little more than a Form 637 and a rented office. The company existed solely to record the receipt of excise taxes, incur the tax liability to the United States, and eventually collapse. The conspirators called this instrumentality a “burn” company.
The daisy chain outlined in the Tarricone Indictment was set in motion when AT bought a total of 21 barge loads of gasoline from licensed New Jersey wholesalers in tax-free transactions. AT then created invoices reflecting fictitious sales to a burn company that was called Conlo, Inc. (“Conlo”). John Pabone and John Quock had purchased Con-lo as a burn company on December 11, 1985, from individuals who had been using Conlo for the same purpose. Since both Conlo and AT were licensed companies, the sales of gasoline to Conlo were, on paper, tax free. In the second step of the daisy chain, Conlo sold the gasoline to Beck Equities, Inc. (“Beck”), an unlicensed shell company also controlled by Pabone and Quock. This transaction triggered Conlo’s obligation to pay the federal excise tax. The conspirators then prepared false invoices indicating that Beck had sold the gasoline to unlicensed purchasers and had paid the excise tax to Conlo. At the completion of the daisy chain, Conlo, which was responsible for the federal excise taxes, simply “burned” up — the company and its records disappeared.
There were no sales between AT and Con-lo or between Conlo and Beck. In reality, AT sold the gasoline directly to the unlicensed purchasers invoiced by Beck, including Westchester Hudson Petroleum Corp. (‘WHPC”),' which employed Barberio, and Hamilton Oil Brokers, Inc. (“Hamilton”). Hamilton was officially owned by an unindict-ed coconspirator named John Byrne, but actually was controlled and operated by Bala-gula. After receiving the gasoline, Hamilton would turn around and sell it to other unlicensed companies, such as Energy Makers of America, Inc. (“EMA”), a company owned by Balagula, and Shore Line Oil Co., Inc. (“Sho-reco”), a company that employed Barberio as a consultant. The invoices reflecting the sales from Hamilton to these companies [666]*666falsely indicated that all excise taxes had been paid.
Count I of the Tamcone Indictment charged the defendants with conspiring to defraud the United States by evading and impeding the collection of federal gasoline excise taxes. The remaining two counts charged the defendants with specific instances of attempted tax evasion. After an eight day jury trial beginning on June 22, 1992, Balagula was convicted on all three counts and Barberio was convicted on the conspiracy count and on one substantive count.
B. The Macchia Indictment
In June 1993, a federal grand jury returned an indictment against Balagula, Bar-berio and six co-defendants — Joseph Macchia Sr.; his three sons Lawrence, George, and Joseph; Viktor Batuner; and Michael Var-zar. Count I of the indictment charged the eight defendants with conspiring to evade the federal gasoline excise tax in violation of 18 U.S.C. §§ 371 and 3623. The six additional counts charged them with attempted tax evasion in violation of 26 U.S.C. § 7201 and 18 U.S.C. §§ 2 and 3623 (the “Macchia Indictment”). Balagula is named in four of the substantive counts, while Barberio is charged only in the conspiracy count.
As detailed in the Macchia Indictment, which is summarized in the following paragraphs, the conspirators used a daisy chain that allowed the defendants to evade over $85 million in federal gasoline excise taxes from the beginning of 1983 through the middle of 1988. According to the “Manner and Means” portion of the Macchia Indictment, a wholesale gasoline distributor called New York Fuel Terminal (“NYFT”), controlled by Joseph A. Macchia and his sons, sold large quantities of gasoline to unlicensed companies without paying the required excise tax.
NYFT sold most of its gasoline directly to unlicensed purchasers through a series of book transfers — the gasoline remained in storage at NYFT’s M & Q Terminal in Brooklyn while title passed from NYFT to another company. The remainder of the gasoline was disposed of through barge sales conducted in New York and New Jersey or through book transfers conducted in New Jersey. The unlicensed companies that actually purchased the gasoline from NYFT typically received false invoices indicating that all excise taxes had been paid. According to the indictment, many of the unlicensed companies to which NYFT distributed its bootleg gasoline were controlled by Balagula; the rest were controlled by Batuner and Varzar.
The Macchia Indictment alleges that NYFT used fraudulent invoices and false book transfers to create the appearance that NYFT’s sales were simply tax-free sales to licensed companies. The false invoices indicated that NYFT had sold hundreds of millions of gallons of gasoline to roughly 18 licensed companies, including AT and Conlo. Additional invoices were then used to reflect fictitious sales from these eighteen licensed companies to third parties. None of the eighteen companies ever purchased the gasoline from NYFT; nor, for that matter, did they ever resell the “purchased” gasoline to third parties. The false book transfers created the appearance that the gasoline passed through the accounts of several licensed and unlicensed companies before reaching the actual unlicensed purchaser (while in reality the gasoline was transferred directly to the unlicensed purchaser). The Macchia Indictment identifies 55 of these false book transfers, 14 of which reflect “sales” from NYFT to either Conlo or AT. EMA and Hamilton, companies involved in the Tarricone daisy chain, were also the subjects of false book transfers.
The cash that NYFT received from the actual unlicensed purchasers of the gasoline was falsely reported on the NYFT books as having been received from one of the eighteen licensed companies to which NYFT had purportedly sold the gasoline. The false entries on NYFT’s books were substantiated through the creation of false cash receipts. The trail of fictitious transactions created the appearance that insolvent burn companies had first sold the gasoline to unlicensed purchasers in tax-paid transactions. In this manner, the tax liability was diverted from NYFT to the burn companies.
The Macchia Indictment charges the Macchias with operating NYFT and the M & [667]*667Q storage facility. Barberio, meanwhile, allegedly worked for NYFT as the manager of its wholesale gasoline sales. As for Balagula, the indictment alleges that he controlled many of the unlicensed companies that purchased gasoline from NYFT.
After the return of the- Macchia Indictment, Barberio and Balagula moved to dismiss the conspiracy count (Count I), arguing that the conspiracy alleged in the Macchia Indictment (the “Macchia Conspiracy”) was, for double jeopardy purposes, the same conspiracy alleged in the Tarrieone Indictment (the “Tarrieone Conspiracy”). A hearing was held on January 21,1994, and by opinion and order dated March 9, 1994, Judge Wex-ler denied defendants’ motion:
[U]nder the totality of the circumstances, notwithstanding that there is some overlap in time, place and participants of the two conspiracies, and both indictments charged conspiracy to evade federal gasoline excise taxes, these similarities are far outweighed by the differences in operations, overt acts and objectives of the two conspiracies, and the relative independence of the two conspiracies. The similarities do not alter this Court’s determination that the two conspiracies appear to be distinct as a matter of law and fact.
Macchia, 845 F.Supp. at 959.
DISCUSSION
The double jeopardy clause of the Fifth Amendment provides: “[N]or shall any person be subject for the same offence to be twice put in jeopardy of life or limb.” U.S. Const. Amend. V. In determining whether two conspiracies amount to the “same of-fence” for that purpose, we have traditionally considered a variety of factors, see, e.g., United States v. DeFillipo, 590 F.2d 1228, 1234-35 (2d Cir.), cert. denied, 442 U.S. 920, 99 S.Ct. 2844, 61 L.Ed.2d 288 (1979); United States v. Papa, 538 F.2d 815, 821-22 (2d Cir.), cert. denied, 429 U.S. 961, 97 S.Ct. 387, 50 L.Ed.2d 329 (1976); United States v. Mallah, 503 F.2d 971, 985-86 (2d Cir.1974), cert. denied, 420 U.S. 995, 95 S.Ct. 1425, 43 L.Ed.2d 671 (1975), which were reviewed and catalogued in United States v. Korfant, 771 F.2d 660, 662 (2d Cir.1985):
(1) the criminal offenses charged in successive indictments; (2) the overlap of participants; (3) the overlap of time; (4) similarity of operation; (5) the existence of common overt acts; (6) the geographic scope of the alleged conspiracies or location where overt acts occurred; (7) common objectives; and (8) the degree of interdependence betweén alleged distinct conspiracies.
The viability of the Korfant analysis was called into doubt following the Supreme Court’s decision in Grady v. Corbin, 495 U.S. 508, 521, 110 S.Ct. 2084, 2093, 109 L.Ed.2d 548 (1990). According to Grady, “the Double Jeopardy Clause bars any subsequent prosecution in which the government, to establish an essential element of an offense charged in that prosecution, will prove conduct that constitutes an offense for which the defendant has already been prosecuted.” In the conspiracy context, we reasoned that the offense for which a defendant is prosecuted is not simply the conspiratorial agreement itself, but also the conduct from which that agreement can be inferred. United States v. Calderone, 917 F.2d 717, 721-22 (2d Cir.1990) (“Calderone I”). Therefore, under Grady, conduct used to prove one conspiratorial agreement could not be reintroduced in a subsequent prosecution to establish an agreement “that differs from the first crime only in that the indictment happens- to describe it differently.” Calderone I, 917 F.2d at 721. Since this result was not necessarily mandated under Korfant, we concluded that the Korfant factors “are inadequate to satisfy double jeopardy concerns after Grady.” United States v. Gambino, 920 F.2d 1108, 1112 (2d Cir.1990) (“Gambino I”); see also Calderone I, 917 F.2d at 721 (“To the extent that Korfant and our other opinions in this area conflict with Grady, they are no longer good law.”). We therefore abandoned the Korfant analysis in Gambino I and Calder-one I. The result was an effective prohibition on successive conspiracy prosecutions where there was substantial overlap of players and conduct. See Gambino I, 920 F.2d at 1112 (defendant’s prior acquittal on an eight-month conspiracy charge jeopardy-barred the government from prosecuting the [668]*668defendant for participation in a 14-year conspiracy); Calderone I, 917 F.2d at 721-22 (defendants’ acquittal on charges of extensive multi-drug conspiracy barred subsequent prosecution for narrower, single-drug conspiracy).
Less than two years after Grady, the Supreme Court cautioned in United States v. Felix, — U.S. —, 112 S.Ct. 1377, 118 L.Ed.2d 25 (1992), that Grady is not to be read “literally” or “expansively.” Id. at --•, 112 S.Ct. at 1383-84. The Court also recognized that the line between the “same evidence” test (which was rejected by Grady) “and the ‘same conduct’ language of Grady is not easy to discern.” Id. at-, 112 S.Ct. at 1385. Accordingly, when the Supreme Court granted certiorari in Calderone I and Gambino I, see United States v. Calderone, — U.S.—, 112 S.Ct. 1657, 118 L.Ed.2d 381 (1992); United States v. Gambino, — U.S. —, 112 S.Ct. 1657, 118 L.Ed.2d 381 (1992), and remanded those cases for further consideration in light of Felix, we returned to the Korfant analysis in determining whether two conspiracies are the “same offense” for double jeopardy purposes. See United States v. Gambino, 968 F.2d 227, 232 (2d Cir.1992) (“Gambino II”) (“Grady’s ‘same conduct’ test did not supplant our Korfant multi-factor analysis for determining whether successive conspiracy prosecutions run afoul of double jeopardy”); United States v. Calderone, 982 F.2d 42, 45 (2d Cir.1992) (“Calderone II ”) (same). These decisions proved to be prescient: Grady was subsequently overturned by the Supreme Court in United States v. Dixon, — U.S. —, —, 113 S.Ct. 2849, 2864, 125 L.Ed.2d 556 (1993).
We therefore consider the several Korfant factors with the lively awareness that no dominant factor or single touchstone determines whether the Tarricone Conspiracy and the conspiracy alleged in Count I of the Macchia Indictment “appear in fact and in law the same.” See United States v. Reiter, 848 F.2d 336, 340 (2d Cir.1988). Defendants argue that the Tarricone Conspiracy was merely a subset of the larger Macchia Conspiracy. “Where the facts of the smaller conspiracy were substantially overlapping with those of the larger conspiracy, we have either held the conspiracies to be the same ... or sufficiently similar to require the Government to prove that they are differ-ent_” Calderone II, 982 F.2d at 47. “[Ojnce a defendant introduces sufficient evidence that the two conspiracies alleged were in fact one, the burden shifts to the government to rebut the inference of unity.” United States v. Abbamonte, 759 F.2d 1065, 1069 (2d Cir.1985) (quoting Papa, 533 F.2d at 821). In a thoughtful and lucid decision, Judge Wexler determined that while the defendants “arguably have made a sufficient showing under the Korfant factors to shift the burden to the government to prove distinct conspiracies,” the government successfully “rebutted any inference of unity.” Macchia, 845 F.Supp. at 957. We agree and therefore affirm the judgment of the district court.
At a certain level of generality, the Tarri-cone Conspiracy and the conspiracy alleged in the Macchia Indictment overlap with respect to a number of characteristics, including time frame, geographic locale, ■ participants, and criminal objective. At a more particular level, however, the dissimilarities of the two tax evasion schemes substantially outweigh any commonality. There exist sufficient distinctions between the schemes charged in the two indictments to support the conclusion that the defendants have not been twice been put in jeopardy for the same offense.
We acknowledge that this determination does not foreclose the possibility that the conduct alleged in the Tarricone Indictment could have been properly pleaded within the parameters of the overarching conspiracy alleged in the Macchia Indictment. Korfant, however, is not a prescription for drafting indictments; nor is it akin to the “same conduct” test announced in Grady. As a practical matter, the Korfant test draws the Constitutional outer limits of prosecutorial discretion in framing indictments for successive conspiracy prosecutions.
With this principle in mind, we turn our attention to the Korfant analysis.
[669]*6691. Criminal Offenses Charged
The Tarricone Indictment and the Macehia Indictment both charge conspiracy to evade federal gasoline excise taxes in violation of 18 U.S.C. §§ 371. Similarity at this general level, however, is of limited import. See United States v. Chiattello, 804 F.2d 415, 419 (7th Cir.1986) (the fact that the same statutory offense is charged in both conspiracy indictments “is far from sufficient to establish the existence of one conspiracy”).
2. Overlap of Participants
Many of the Tarricone participants and companies also played roles in the Macehia Conspiracy, but they played different roles in each conspiracy. After acquiring Conlo on December 11, 1985, Quock and Pabone proceeded to organize the entire Tarricone Conspiracy. At the Tarricone trial, Quock accurately identified himself as the “master mind” of the Tarricone Conspiracy. In the Macehia scheme, however, these two individuals simply operated companies that were party to some of NYFT’s false book transfers. Meanwhile, the organizers of the Macehia Conspiracy, the Macchias, were not mentioned in the Tarricone case. Barberio also functioned in different capacities in the two conspiracies. In the Tarricone Conspiracy, Barberio served as a consultant to Shore-co and as an employee of WHPC, two of the unlicensed companies that purchased bootleg gasoline from AT. In the Macehia conspiracy, Barberio worked as the manager of wholesale gasoline sales for NYFT.
Balagula, the other defendant common to both the Tarricone and Macehia Indictments, provided both AT and NYFT with outlets to end-users. Nevertheless, the existence of a single conduit does not establish proof of a single conspiracy. See Reiter, 848 F.2d at 341 (the fact that defendant served as an intermediary to multiple distribution networks “does not establish the existence of a single conspiracy”); see also Papa, 533 F.2d at 822 (double jeopardy did not bar successive prosecutions where defendant was the director of two unrelated chains of distribution). According to testimony presented at trial, Balagula controlled 100% of the bootleg gasoline market in and around New York City. From the beginning of 1983, he provided an assortment of unlicensed purchasers for the NYFT scheme. Some years later, Quock and Pabone, who had acquired Conlo to use as a burn company, approached Bala-gula seeking (i) access to purchasers, and (ii) a supply of tax-free gasoline to transfer through Conlo. Balagula then arranged for AT to become a source for Conlo. If, as defendants contend, the Tarricone and Mace-hia conspiracies were one and the same, Quock and Pabone would have had no need to seek access to Balagula’s chain of distribution. That they did so demonstrates that Balagula served as a conduit for two distinct conspiracies: one using AT as a supplier and Conlo as the burn company; and the other using NYFT as a supplier and a score of burn companies.
There is a substantial overlap of the gasoline trading companies involved in the Tarri-cone and Macehia schemes. Again, however, many of the players are performing different roles. AT, for example, was initially used as a burn company in the Macehia daisy chain to conceal NYFT’s tax obligations. After Quock and Pabone acquired Conlo in December 1985, AT was no longer used in the Macehia scheme. Instead, AT became the supplier of bootleg gasoline. Similarly, after becoming the burn company for AT, Conlo ceased being a party to NYFT’s false book transfers and fraudulent invoices.
3.Overlap of Time
The time frame of the Tarricone Indictment (November 1985 through April 1986) is wholly contained within that of the Macehia Indictment (early 1983 through the middle of 1988). It would give us pause if the reverse were true: “The Government cannot be permitted to retry defendants on smaller and smaller conspiracies, wholly contained within the scope of a large conspiracy, until it finds one small enough to be proved to the satisfaction of a jury.” Calderone II, 982 F.2d at 48. However, where the smaller conspiracy is charged first, there is not the same opportunity for prosecutorial abuse, and the overlap of time is therefore a less important consideration. Compare Gambino II, 968 F.2d at 232 (conspiracies held to be [670]*670distinct where time fráme of the first conspiracy was subsumed in that of the second), and United States v. Rivera, 844 F.2d 916, 923-25 (2d Cir.1988) (same) with Mallah, 503 F.2d at 987 (in similar circumstances, conspiracies held to be the same). Accordingly, the fact that the time frame of the first conspiracy is a subset of that of the second does not preclude a finding of distinct conspiracies.
4. Similarity of Operation
Both the Tarricone and Macchia schemes used a daisy chain involving one or more burn companies to evade federal excise taxes. However, as the district court recognized, similarity between modes of operation at this level of generality does not signify a single conspiracy. See, e.g., United States v. Futch, 637 F.2d 386, 390 (5th Cir.1981) (the use of shrimpboats and trucks to import and off-load marijuana is a common means of implementation that does not integrate two separate conspiracies).
The operations of the Tarricone and Macchia conspiracies were in fact quite different. In the Tarricone Conspiracy, AT first purchased barge loads of gasoline from New Jersey wholesalers in tax-free transactions. The gasoline was then sold directly to unlicensed purchasers and delivered to various fuel terminals. To disguise these taxable transactions, AT created false invoices reflecting sales from AT to a single burn company, Conlo. Additional false invoices were then used to indicate tax-paid sales from Conlo to Beck, an unlicensed company.
The Macchia Conspiracy employed a more complex modus operandi. NYFT served as the wholesaler, storing the gasoline at its M & Q Terminal in Brooklyn. The gasoline was sold to unlicensed companies primarily through book transfers: title to the gasoline held within the M & Q Terminal passed from one company to another. To avoid incurring tax liabilities, NYFT issued false invoices reflecting sales to 18 different burn companies. Moreover, NYFT used false book transfers and false cash receipts to further disguise the actual whereabouts of the gasoline. The Tarricone and Macchia schemes therefore used different suppliers, different means of distributing the gasoline to unlicensed purchasers, and different fraudulent techniques-for concealing the taxable transactions. Accordingly, there is ample support for the district court’s determination that “both conspiracies had a full complement of players, and the actual and contrived transactions to effect each of the schemes do not indicate that one conspiracy was part of the other.” Macchia, 845 F.Supp. at 959.
5. Existence of Common Overt Acts
The Tarricone Indictment and the Macchia Indictment, contain no common overt acts. As the district court noted, “[t]he overt acts charged in the Tarricone Conspiracy dealt with the supply of gasoline from a different source than in the Macchia Conspiracy, with the evasion of the tax liability of a different company than in the Macchia Conspiracy, and concealed by different “burn” companies (Conlo and Beck as then owned by Pabone and Quock) than used in the Macchia Conspiracy.” Macchia, 845 F.Supp. at 959. Defendants attribute the absence of common overt acts to the fact that the government simply carved out the Tarricone Conspiracy from the larger Macchia Conspiracy.
We have previously “recogniz[ed] the ease with which prosecutors can draft indictments that allege what appear to be separate conspiracies but may actually be parts of an overall conspiracy.” Abbamonte, 759 F.2d at 1068. It is precisely because of this opportunity for procedural abuse that “we have not resolved double jeopardy claims involving conspiracy charges solely by the ‘same evidence’ test, which inquires only whether evidence required to support one charge would have sufficed to support the other.” Id.; see also Mallah, 503 F.2d at 985 (“a test measuring only overt acts provides no protection against carving one larger conspiracy into smaller separate agreements”).1 [671]*671Instead, we apply the multi-factor Korfant test, of which the existence (or non-existence) of common overt acts is but one factor.
6. The Geographic Scope of the Alleged Conspiracies
The Tarrieone and Macchia schemes were played out in the New York/New Jersey Metropolitan area. While geographic overlap might be probative if there were “other factors in this ease that point toward a single conspiracy,” Abbamonte, 759 F.2d at 1070, the other factors here show real distinctions in instrumentalities, in techniques, and in the roles played by the conspirators. Moreover, the New York/New Jersey Metropolitan area is no mere locality; in commercial terms, it is a world. As the district court noted, it is certainly large enough to harbor two simultaneous conspiracies dealing in bootleg gasoline. See Mallah, 508 F.2d at 983 (“New York City is large enough to harbor two simultaneous narcotics conspiracies”).
In fact, the conspiracies operated within different segments of the broad geographic region. See Papa, 533 F.2d at 819-20 (no geographic overlap where one narcotics conspiracy occurred in Long Island, Long Island City, and several locations in the Bronx and the other conspiracy occurred in Astoria, Queens, and different locations in the Bronx). The Tarrieone Conspiracy began when AT purchased gasoline from wholesalers in New Jersey and transported it to fuel terminals in New Jersey, Brooklyn, Westchester, and Long Island. None of the gasoline was stored at the Macchia’s M & Q Terminal in Brooklyn, the exclusive warehouse for NYFT’s gasoline.
7. Common Objectives
Defendants characterize the Tarrieone and Macchia schemes as having a common objective “to sell and distribute bootleg gasoline and pocket money that would otherwise be paid to the United States Treasury.” Under this view, however, every conspiracy to distribute heroin (or cocaine, or anything else) would share a common objective for double jeopardy purposes.
“The gist of the crime of conspiracy ... is the agreement ... to commit one or more unlawful acts.” Braverman v. United States, 317 U.S. 49, 53, 63 S.Ct. 99, 101, 87 L.Ed. 23 (1942). “[MJultiple agreements to commit separate crimes constitute multiple conspiracies.” United States v. Broce, 488 U.S. 563, 571, 109 S.Ct. 757, 763, 102 L.Ed.2d 927 (1989). Here, the record easily supports multiple agreements. In the Tarrieone Conspiracy, the defendants caused AT to sell gasoline without paying the excise taxes due. In contrast, the conspiracy alleged in Count I of the Macchia Indictment focuses exclusively upon the tax evasion efforts of NYFT. If, as defendants contend, the Tarrieone Conspiracy was simply a narrow segment of the broader Macchia Conspiracy, there would have been no reason for the Tarrieone conspirators to establish AT as a supplier of bootleg gasoline; the Tarrieone conspirators could have simply relied upon the gasoline supplied by NYFT.
8.Degree of Interdependence
This factor requires us to consider the extent to which the success or failure of one conspiracy is independent of a corresponding. success or failure by the other. Korfant, 771 F.2d at 663. The district court determined that “the Tarrieone Conspiracy did not depend on the operation of the ongoing, substantially larger Macchia Conspiracy and was not merely a part of the Macchia Conspiracy.” Macchia, 845 F.Supp. at 957. We agree.
It is true that both conspiracies used Bala-gula’s network to distribute bootleg gasoline to actual purchasers. Balagula’s unlicensed companies, however, received only three of the 21 barge loads that AT purchased. The bulk of AT’s gasoline went to Barberio’s employers, Westchester and Shoreco. In contrast, the Macchia Indictment alleges that Balagula served as NYFT’s primary conduit to unlicensed purchasers.
[672]*672Defendants also emphasize the significant roles that AT and Conlo played in both tax evasion schemes. These companies, however, were never used simultaneously in the Macchia and Tarricone conspiracies. Before Conlo was acquired by Quock and Pabone, it was one of the 18 licensed companies appearing in NYFT’s fraudulent invoices and false book transfers. After it was acquired by Quock and Pabone, Conlo no longer figured in NYFT transactions, and it functioned exclusively as the burn company in the Tarri-cone scheme. Similarly, after AT assumed its role as the supplier in the Tarricone Conspiracy, it was never again used as a burn company by the Maechias. Since neither Conlo nor AT ever functioned in a dual capacity, the Macchia scheme and the Tarri-cone scheme could, at any given time, function independently of one another.
CONCLUSION
For purposes of double jeopardy, we conclude that the Tarricone Conspiracy is separate from the conspiracy alleged in Count I of the Macchia Indictment. The judgment of the district court is therefore affirmed.