Claudio Calderone and Domenico Catalano appeal on double jeopardy grounds from an order of the United States District Court for the Southern District of New York, Robert J. Ward, Judge, denying their motions to dismiss Indictment 89 Cr. 786 (RJW), which charges them with conspiracy to distribute heroin, use of a telephone to facilitate narcotics crimes, and possession of heroin with intent to distribute it. In a previous prosecution, Calderone and Catalano were acquitted of membership in a larger conspiracy that involved, in addition to the heroin-selling activities alleged here, the distribution of large quantities of cocaine and marijuana. We hold that the "same conduct" test that is now to be applied to double jeopardy claims arising in the context of successive prosecutions, see Grady v. Corbin, --- U.S. ----, 110 S. Ct. 2084, 109 L. Ed. 2d 548 (1990), bars prosecution of all counts of the present indictment. We therefore reverse the order of the district court and remand with a direction to dismiss the indictment.

A. The Adamita Indictment.

On June 30, 1988, a grand jury sitting in the Southern District of New York returned an indictment against Calderone, Catalano, and 26 others. United States v. Adamita, 701 F. Supp. 85 (S.D.N.Y.1988). Count One of the Adamita indictment charged all 28 defendants with participating in a wide-ranging, multi-drug conspiracy between January 1, 1985, and June 30, 1988, that involved, among other things, the importation of kilograms of heroin from Europe into the United States, sometimes in exchange for kilograms of cocaine; the distribution of kilograms of heroin and cocaine and multi-tons of marijuana; the use of various businesses as fronts for illegal activities; and the use of weapons to protect the narcotics business.

The Adamita indictment listed 105 overt acts committed in furtherance of the alleged conspiracy, but named Calderone or Catalano in only eight of them. In addition to the conspiracy count, Catalano was charged in only one substantive count; Calderone was charged in no other counts. The remaining 26 defendants were charged in one or more of the numerous substantive counts of possessing and distributing narcotics and using firearms in connection with drug trafficking crimes.

Trial before district judge John E. Sprizzo and a jury commenced on October 31, 1988, against 17 of the original 28 defendants, including Calderone and Catalano. To prove Calderone's and Catalano's membership in the alleged conspiracy, the government produced testimony showing that an undercover agent posing as a heroin distributor was introduced by a co-conspirator to Calderone and Catalano, both of whom acknowledged that heroin was available, but neither of whom actually sold any heroin to the agent. The agent also testified that after he had purchased heroin from another source, Calderone was present and received some of the money when the agent returned to pay the outstanding balance. In addition, both Calderone and Catalano were observed in the company of some of the co-conspirators at various times during the alleged conspiracy.

At the close of the government's case, all remaining defendants moved for judgments of acquittal pursuant to Fed.R.Crim.P. 29. After extensive argument, Judge Sprizzo granted the motions of seven defendants, including Calderone and Catalano. Judge Sprizzo ruled that proof of Calderone's and Catalano's membership in the conspiracy was legally insufficient to go to the jury. During an extensive colloquy with counsel, Judge Sprizzo stated, as to Calderone:

There is no question in my mind, and I will say it for the record, that if the government had proceeded against Mr. Calderone with a more narrowly charged conspiracy, I would have sent it to the jury. But you chose * * * to charge Mr. Calderone in a massive conspiracy, which according to the government's opening in this case involved all of these things. The fact that he participated in one or two narcotics transactions and that he knows these other conspirators is legally insufficient for that purpose. [Tr. 7700-01].

* * * * * *

I am not saying your evidence would not have been sufficient to prove Mr. Calderone to be a member of a heroin conspiracy involving [two alleged co-conspirators], but you didn't charge a heroin conspiracy. You charged a * * * broadly based conspiracy that involved different types of drugs. There is no evidence in the record from which I can infer that as to Mr. Calderone. [Tr. 7702-03].

[t]here is no evidence to permit a rational jury to find that Mr. Catalano was involved in this massive, all-encompassing multidrug conspiracy, none whatsoever, under the case law as I know it and upon the case law relied upon by the government. [Tr. 7704].

* * * * * *

The case against Catalano is not as strong as it is against Calderone. What evidence is there in this record other than the fact that people have a meeting, * * * any knowledge that people were shipping cocaine out of the country, that they were bringing heroin back in exchange for cocaine, or that they were discussing multiton marijuana lots? You have at best an ordinary heroin dealer. * * * And the government says that any time they have a kilogram dealing in heroin that automatically the jury can properly infer that this person is charged with activities relating to other drugs--which is not only other drugs, * * * but the very peculiar aspect of this distribution of cocaine, which is to sell it abroad in exchange for heroin. Nonsense. The motion is granted. [Tr. 7707].

While not relevant to this appeal, the Adamita jury ultimately returned guilty verdicts against five of the remaining seven defendants, and not guilty verdicts against the other two.

B. The Calderone Indictment.

Because Judge Sprizzo issued judgments of acquittal on the ground of insufficient evidence before the Adamita case reached the jury, the government could not appeal the ruling. See 18 U.S.C. Sec. 3731; United States v. Martin Linen Supply Co., 430 U.S. 564, 97 S. Ct. 1349, 51 L. Ed. 2d 642 (1977). Instead, in what it characterizes as a decision "to modify the charging instrument so as to charge the separate conspiracy", Brief of United States 17 n. **, the government filed a new indictment against Calderone and Catalano. United States v. Calderone, 89 Cr. 786 (RJW). In a nutshell, the Calderone indictment alleges that the activities for which Calderone and Catalano were unsuccessfully prosecuted in Adamita are sufficient to implicate them in a smaller conspiracy involving the distribution of heroin in New York.

Specifically, Count One of the Calderone indictment charges both defendants with participating in a conspiracy between January 1, 1987, and March 31, 1988, to distribute and possess with intent to distribute kilogram quantities of heroin in the New York metropolitan area. The indictment lists 21 overt acts in furtherance of the conspiracy, and names either Calderone or Catalano in 19 of them. Counts Two through Twenty-Five charge the defendants with using a telephone to facilitate narcotics crimes, and Counts Twenty-Six through Twenty-Eight charge them with possession of heroin with intent to distribute it.

On December 8, 1989, both defendants filed pre-trial motions to dismiss the indictment on double jeopardy grounds. District Judge Robert J. Ward, to whom the Calderone case had been assigned, denied the motions from the bench on February 2, 1990, following oral argument. The crux of the court's ruling on Count One (the conspiracy count) was that the government had alleged a different conspiracy in Calderone than it had in Adamita; that is, it had alleged a limited, single-drug conspiracy in Calderone as opposed to the massive, three-drug conspiracy in Adamita. In making this ruling, the court emphasized Judge Sprizzo's comments in Adamita that the government had, at best, proven a heroin conspiracy against Calderone and Catalano, but had failed to prove the larger conspiracy. As Judge Ward explained,

Let me put it this way. I would have no trouble with your [double jeopardy] argument if Judge Sprizzo had permitted that indictment, that charge to go to the jury, [and] the jury had found your client not guilty. I would have told the government in no uncertain terms this indictment is not appropriate.

But unfortunately what I have here is a directed verdict by a judge who has based his determination on a decision that he made that the government didn't charge a heroin conspiracy.

Judge Ward concluded that "since [Judge Sprizzo] determined there was no charge of a heroin conspiracy [in Adamita ], the defendants were not placed in jeopardy and therefore the present prosecution alleging a heroin conspiracy does not constitute double jeopardy." Judge Ward also denied the motions insofar as they sought dismissal of Counts Two through Twenty-Eight of the indictment.

Calderone and Catalano then brought this interlocutory appeal pursuant to Abney v. United States, 431 U.S. 651, 662, 97 S. Ct. 2034, 2041, 52 L. Ed. 2d 651 (1977). After this panel heard oral argument, the Supreme Court announced a new rule for double jeopardy claims arising in the context of successive prosecutions. See Grady v. Corbin, --- U.S. ----, 110 S. Ct. 2084, 109 L. Ed. 2d 548 (1990). We directed the parties to submit letter briefs addressing the impact of Grady on the present appeal. After careful consideration, we conclude that the "same conduct" test announced in Grady bars prosecution of all counts of the Calderone indictment. Accordingly, we reverse the order of the district court and remand for dismissal of the indictment.

Grady directs courts to follow a two-step analysis in resolving double jeopardy claims arising in the context of successive prosecutions. First, the court must apply the Blockburger test and its traditional exceptions to determine whether the offense charged in the subsequent prosecution "requires proof of a fact which the other does not." Blockburger v. United States, 284 U.S. 299, 304, 52 S. Ct. 180, 182, 76 L. Ed. 306 (1932); see generally Grady, 110 S. Ct. at 2090-93 (discussing application of Blockburger and "collateral estoppel" and "component offense" exceptions). If the indictment survives this first stage of analysis--and we assume for the purpose of this opinion that the Calderone indictment does--then the court must apply Grady's "same conduct" test, which "bars a subsequent prosecution if, to establish an essential element of an offense charged in that prosecution, the government will prove conduct that constitutes an offense for which the defendant has already been prosecuted." 110 S. Ct. at 2087.

As Justice Brennan's opinion for the Court acknowledged, application of the "same conduct" test to the facts of Grady was "straightforward." 110 S. Ct. at 2094. The defendant Corbin drove his car across the center line of a highway while intoxicated, killing the driver of an oncoming vehicle. Through an oversight in the local prosecutor's office, Corbin was initially charged and convicted only of the misdemeanors of driving while intoxicated and failing to keep right of the median. When the state later indicted Corbin for vehicular homicide, it stated in a bill of particulars that it would establish Corbin's guilt by proving his intoxication at the time of the accident, his failure to keep right of the median, and the fact that he was driving too fast for the weather conditions.

In Grady, then, the state admitted it would attempt to prove conduct for which Corbin had already been prosecuted (driving while intoxicated and crossing the center line) in order to establish the negligent or reckless acts necessary for a homicide conviction. In other words, the state would "prove conduct that constitutes an offense for which the defendant has already been prosecuted" in order to establish an essential element of the subsequent offense. The Court did emphasize, however, that the state could have prosecuted Corbin for homicide if it had relied only on conduct for which Corbin had not previously been prosecuted, for example, his driving too fast for road conditions. But because the state had indicated in its pleadings that it would rely on conduct for which Corbin had already in fact been prosecuted, double jeopardy barred the second prosecution. Id.

In the present case, the government first disputes whether Grady should be applied at all to successive conspiracy prosecutions, claiming that the decision "represents a narrow response to a relatively limited category of cases." In place of Grady, the government would have us apply the balancing approach set forth in United States v. Korfant, 771 F.2d 660 (2d Cir.1985) (per curiam). But the government's attempt to limit Grady to its facts mischaracterizes the intended scope of the Court's decision.

Although Grady involved the successive prosecution of separate crimes arising from a single event, nothing in the opinion suggests that the Court intended to limit the "same conduct" test to those particular circumstances. Instead, the Court's concern with the dangers of multiple trials, and its belief that the Blockburger test by itself was insufficient to guard against these dangers, see 110 S. Ct. at 2090-93, suggest that the Grady test was intended to guide double jeopardy analysis in all cases involving successive prosecutions. The "same conduct" test is described in inclusive, generally applicable terms throughout the opinion. See, e.g., id. 110 S.Ct. at 2093 ("the Double Jeopardy clause bars any subsequent prosecution" if Grady test is met) (emphasis added). Indeed, the opinion is structured so that it first declares a general rule, id. 110 S.Ct. at 2090-93, and then applies that rule to the facts of the case. See id. 110 S.Ct. at 2094.

For these reasons, we are persuaded that the "same conduct" test announced in Grady applies to all double jeopardy claims arising in the context of successive prosecutions. To the extent that Korfant and our other opinions in this area conflict with Grady, they are no longer good law.

Turning to the merits, the government contends that the conspiracy count of the Calderone indictment will not require proof of "conduct that constitutes an offense" for which the defendants have already been prosecuted. According to the government, the "conduct that constitutes the offense" in conspiracy cases is the alleged agreement itself rather than the specific conduct from which the agreement can be inferred. Therefore, the government argues, it will not establish an essential element of the conspiracy charged in Calderone by proving the "conduct" (i.e., the "agreement") prosecuted in Adamita. We reject this argument for several reasons.

To begin with, the "agreement" in nearly all conspiracy prosecutions is not proved directly, such as by testimony of conspirators that "we agreed to sell narcotics". Rather, the "agreement" is simply alleged, and to prove its existence the government presents evidence of the conspirators' conduct--what they said and did--and asks the jury to infer from this conduct that the criminal agreement has been established. The conduct in most conspiracy cases, as was the conduct in Adamita, is the focus of the case. Under Grady, this conduct may not be prosecuted a second time in order to establish an "agreement" that differs from the first crime only in that the indictment happens to describe it differently.

Moreover, the government's approach would nullify Grady by recasting it as an alternative formulation of Blockburger. In conspiracy cases, the agreement "is all but synonymous" with the offense itself, United States v. Broce, 488 U.S. 563, 109 S. Ct. 757, 763, 102 L. Ed. 2d 927 (1989); it may be proved by, but it is not synonymous with, the conduct that gives rise to the conspiracy. Hence, if we simply were to compare the "agreements" alleged in successive conspiracy prosecutions, we would be looking to the offenses charged, as Blockburger requires, rather than to the conduct that constitutes the offenses, as Grady demands.

The problems with the government's interpretation of Grady become even clearer if we recall "the ease with which prosecutors can draft indictments that allege what appear to be separate conspiracies". United States v. Abbamonte, 759 F.2d 1065, 1068 (2d Cir.1985). A double jeopardy test that depends entirely on the agreement charged in the indictment would permit a series of conspiracy prosecutions based on identical conduct, provided that the succeeding indictments allege smaller (or larger, or different) "agreements" than the ones previously charged. Not only would this subject defendants to the "embarrassment, expense and ordeal" of trials seriatim, Green v. United States, 355 U.S. 184, 187-88, 78 S. Ct. 221, 223-24, 2 L. Ed. 2d 199 (1957), it would also give the government "an opportunity to rehearse its presentation of proof, thus increasing the risk of an erroneous conviction". Grady, 110 S. Ct. at 2091-92. In short, we conclude that Grady's "same conduct" test bars prosecution of the conspiracy count of the Calderone indictment.

The "same conduct" test likewise bars prosecution of the substantive telephone and heroin distribution counts. Although Calderone and Catalano were not charged with these specific crimes in Adamita, they were nonetheless prosecuted for the conduct that supports these offenses. And under Grady, they may not again be prosecuted for this conduct, even though the charges are brought under different criminal statutes.

For the reasons above, we hold that the "same conduct" rule announced in Grady bars prosecution of all counts of the Calderone indictment. Accordingly, the order of the district court is reversed, and the case is remanded for dismissal of the indictment.

I concur in Judge Pratt's opinion, upholding the double jeopardy claims of the appellants, but write these additional views to explain why I believe that the Supreme Court's new double jeopardy jurisprudence, announced in Grady v. Corbin, --- U.S. ----, 110 S. Ct. 2084, 109 L. Ed. 2d 548 (1990), requires this result, a conclusion I concede is not free from doubt.

Grady is arguably distinguishable from the pending case in three respects, none of which persuades me to deem its teachings inapplicable. First, Grady presented the issue of whether the Double Jeopardy Clause barred prosecution under one statute after prosecution under a different statute. In the pending case, we are primarily concerned with whether the prior prosecution under count one of the Adamita indictment bars prosecution under count one of the Calderone indictment. Both counts charge violations of the same statute, 21 U.S.C. Sec. 846 (1988). I can see no reason why this distinction should render Grady inapplicable, and the Government has not suggested any.

Second, Grady involved successive prosecutions of offenses arising out of a single occurrence, whereas we are concerned with offenses arising out of continuing conduct, i.e., numerous aspects of conduct relied on by the prosecution to show a large conspiracy in Adamita and proposed to be relied on to show a conspiracy of smaller scope in the pending case. The Government contends that it is "far from clear" whether Grady applies to this case. Some reason for doubt on this score is supplied by then Justice Rehnquist's opinion in Garrett v. United States, 471 U.S. 773, 105 S. Ct. 2407, 85 L. Ed. 2d 764 (1985). That opinion pointedly distinguishes single occurrence offenses from continuing conduct offenses, indicating, at a minimum, that different double jeopardy principles may be applicable. Id. at 787-90, 105 S.Ct. at 2415-17.

The force of this distinction is itself put in doubt by Justice O'Connor's concurring opinion in Garrett, which emphasized that the double jeopardy claim was being rejected, not because a continuing conduct offense was alleged, but because the defendant's conduct had in fact continued past the date of the first prosecution. Id. at 797-98, 105 S.Ct. at 2420-22. She pointedly noted that the Court was not deciding whether a double jeopardy claim would be rejected in the absence of such evidence, a circumstance she acknowledged would make the defendant's claim "more compelling." Id. at 799, 105 S.Ct. at 2422.

Neither the holding in Garrett, which rejected a double jeopardy claim where criminal conduct continued subsequent to a first prosecution, nor the holding in Grady, which upheld a double jeopardy claim where offenses arose out of a single occurrence, precisely informs us as to the Supreme Court's current thinking about the instant case, where offenses arise out of continuing conduct but no conduct of the defendants is alleged to have occurred after the first prosecution. In such circumstances, I agree with Judge Pratt that we should be guided by the broad statements in Grady, which is the more recent decision.

Third, it is arguable that Grady, even if applicable to some continuing conduct offenses, does not apply to what the Third Circuit has called "compound-complex crimes," evidently illustrated by RICO conspiracy offenses. See United States v. Pungitore, 910 F.2d 1084 (3d Cir.1990). The Third Circuit has accepted this distinction, relying on some of the language in Garrett, which involved a continuing criminal enterprise offense, 21 U.S.C. Sec. 848 (1988). I am not certain whether the Third Circuit would regard a violation of 21 U.S.C. Sec. 846 as a "compound-complex crime," and, if Grady applies to continuing conduct offenses, I doubt if the double jeopardy analysis should become less rigorous simply because Congress has defined more complicated crimes. Moreover, we have already applied Grady to uphold the double jeopardy claim of a defendant based on a prior prosecution for a RICO conspiracy. See United States v. Russo, 906 F.2d 77 (2d Cir.1990).

Having concluded that Grady is applicable to the pending case, I must then consider how it applies. Grady bars a second prosecution if the prosecutor seeks to establish an "element" of the second crime by proving "conduct that constitutes an offense for which the defendant has already been prosecuted." 110 S. Ct. at 2093. In applying Grady, we have to consider how precisely Justice Brennan intended to use the word "element" and what he meant by "conduct that constitutes an offense for which the defendant has already been prosecuted."

I think the word "element" was chosen with some care and must be read technically to avoid letting the Grady test become a "same evidence test," a result that Grady explicitly disclaims. 110 S. Ct. at 2093 n. 12. Grady itself illustrates a precise use of the concept of "element." Grady was first prosecuted for driving while intoxicated and failing to drive on the right. The subsequent homicide and assault charges required reckless or negligent acts to be established as elements of the subsequent offenses. What Grady holds is that the State may not use Corbin's conduct in driving while intoxicated or driving across the center line to establish these required elements of the subsequent offenses.1 Grady does not bar the State from introducing in the second trial evidence that Corbin was intoxicated or drove across the center line, but these acts (or that conduct) cannot be the elements of reckless or negligent action underlying the homicide and assault charges.

Justice Scalia's dissent assails the idea that the "element" component of the Grady test is a real limitation. As he points out, "All evidence pertaining to guilt seeks 'to establish an essential element of [the] offense,' and should be excluded if it does not have that tendency." 110 S. Ct. at 2103. He may well be right, but he is in dissent. I think we are obliged to apply Grady in a way that gives the "element" component significance. That means barring the second prosecution only when the conduct previously prosecuted is to be used to "establish" the element of the second crime, which I think must mean "constitute the entirety of" the element. If Grady is read more broadly, that is, if the second prosecution is barred whenever the previously prosecuted conduct is to be used only as evidence of an element of the second offense, then we would almost be applying a "same evidence" test.2 Instead, I think it more likely that the Supreme Court expected Grady to apply only when the conduct prosecuted at the first trial is or may constitute the entirety of an element of the offense at the second trial.

Dowling v. United States, --- U.S. ----, 110 S. Ct. 668, 107 L. Ed. 2d 708 (1990), may be used to illustrate the point. Dowling was first prosecuted for robbery A. After an acquittal, he was prosecuted for robbery B. Evidence of robbery A, indeed, evidence of his conduct in committing robbery A, was offered to prove his identity as the perpetrator of robbery B. The Court ruled that this did not violate the Jeopardy Clause. The Grady dissenters believe that Grady overrules Dowling, 110 S. Ct. at 2095-96 (O'Connor, J., dissenting); id. at 2102 (Scalia, J., with whom Rehnquist, C.J., and Kennedy, J., join, dissenting), but, again, they are dissenters. The Grady majority opinion does not claim to have overruled Dowling and, in the face of the challenge raised in the dissents, should be read, if at all possible, not to overrule Dowling. This can be done by recognizing that in Dowling, the defendant's conduct in committing robbery A was introduced only as evidence that the defendant committed robbery B, not as the entirety of an element of robbery B.

Consideration must also be given to Justice Brennan's phrase "conduct that constitutes an offense for which the defendant has already been prosecuted." In Grady, that phrase had clear meaning because the prior prosecution was precisely for the two aspects of conduct sought to be used to establish the elements of the subsequent offenses. Corbin was first prosecuted for the offenses of driving while intoxicated and driving across the center line. What is not clear is whether Grady applies only when the conduct, which is sought to be used to establish the entirety of an element in the second prosecution, constitutes the entirety of a previously prosecuted offense, or also when such conduct is only an element of the prior offense or only evidence of an element of the prior offense.

I think it is clear that Grady does not go so far as to preclude proof of conduct that was only evidence of an element of the prior offense. For example, if Corbin had previously been prosecuted not for driving while intoxicated but for robbing a bank, and if in the bank robbery trial the prosecutor had shown that Grady got up the nerve to rob the bank by drinking heavily and thereafter drove to the bank in an intoxicated state, that evidence would not be barred in the negligent homicide trial. The drunken driving would not have been the offense for which he had been prosecuted. A contrary interpretation would resurrect the "same evidence" test.

It is less clear, however, whether the conduct must be the entirety of the offense previously prosecuted or may be only the entirety of an element of the prior offense. On this issue, the Government has conceded that the latter view is correct, at least where the element constitutes a distinct offense. In United States v. Russo, supra, the Government conceded that a prosecution for obstruction of justice was barred by a prior prosecution for a RICO conspiracy in which the same obstruction of justice offense had been alleged as a predicate act. Russo had not previously been prosecuted for the obstruction of justice, but that offense was an element of the prior RICO offense. In fact, the matter is slightly more complicated. The obstruction was not the entirety of an element of the previously prosecuted RICO offense; the pertinent element was a pattern of racketeering activity, and the obstruction was one of the predicate acts forming the requisite pattern. But this wrinkle shows only that the conduct must be either the entirety of an element of a previously prosecuted offense or at least a distinct component of such an offense, such as a RICO predicate act. The Government was simply acknowledging in Russo that for purposes of Grady a RICO predicate act should be treated as if it is an element of the offense. However, the concession in Russo does not necessarily mean that conduct previously charged as a RICO predicate act, or, as in Adamita, as an overt act, that is not charged as a distinct crime may not be used to prove an element in a subsequent prosecution. Where overt acts not charged as distinct offenses have been alleged in a prior conspiracy indictment, the question remains whether it may be said that a defendant has been "prosecuted" for such acts? I think the answer is yes.

Thus, I read Grady to bar a second prosecution whenever the defendant is at risk that the entirety of an element of an offense in a pending prosecution might be established by conduct constituting the entirety of a previously prosecuted offense (as in Grady ), or the entirety of an element of such an offense, or the entirety of a distinct component of such an offense (as in Russo). The further question is how this reading applies to the pending case.

The "elements" to be proven for the offense charged in count one are the agreement charged in the Calderone indictment and the defendants' willful joining of the conspiracy. Clearly the Government may not prove the element of willful joining by proving any of the acts identified as overt acts in the Adamita indictment. Even though 21 U.S.C. Sec. 846 does not require proof of an overt act, once the Government alleges overt acts, as it did in Adamita, the defendants were being prosecuted for those acts, and, under Grady, those acts may not be used to establish an element of the Calderone offense. Though the overt acts that the prosecution sought to prove these appellants committed in the course of the Adamita conspiracy--meetings and telephone calls--were not crimes, it was the Government's contention that the appellants committed these acts in furtherance of the Adamita conspiracy and that these acts showed their participation in that conspiracy. The Government did not think these were innocent acts; it sought to persuade the jury in the prior case that these acts established appellants' guilt of the Adamita conspiracy, when assessed in light of the other evidence in the case. Thus, in a very real sense, the appellants were prosecuted for these acts.

The harder question is whether the Adamita prosecution leaves the Government with any way of proving that the defendants joined the Calderone conspiracy, apart from conduct for which they were prosecuted in Adamita. It might be possible for the Government to prove that the defendants joined the Calderone conspiracy on the basis of conduct wholly distinct from the conduct relied on in the Adamita case to show that they joined the Adamita conspiracy. But the Government has made no claim that it will rely on such distinct conduct. Indeed, the lack of distinction is apparent from the two indictments; some of the overt acts charged to the defendants are common to both indictments:

Adamita indictment:  Calderone indictment:
overt act 32         overt act 6
overt act 33         overt act 7
overt act 34         overt act 8
overt act 63         probably the same conduct as overt act 10

Even if the Government has conduct independent of the Adamita prosecution to establish the element of joining the Calderone conspiracy,Grady cannot be overcome. The conduct for which the defendants were prosecuted in Adamita, i.e., their conduct in joining what was alleged to be a large international drug conspiracy, would have sufficed to show their joining of the allegedly smaller Calderone conspiracy. In Grady's terms, the conduct for which they were prosecuted in Adamita would have established the entirety of an element of the Calderone offense.

A similar analysis applies to the element of the conspiracy itself. Even if it can be said that the two conspiracies are "different" in the sense that United States v. Korfant, 771 F.2d 660 (2d Cir.1985), viewed conspiracies as different, the conduct alleged to show the existence of the big conspiracy in Adamita, conduct for which these defendants were prosecuted, would suffice to establish the element of the smaller conspiracy in Calderone. Evidence that defendants have agreed to sell heroin and other drugs over a long period of time will surely establish the element of an agreement to sell only heroin during an interval within that longer period.

If Grady bars count one, it also bars counts two through twenty-five (the "telephone" counts, 21 U.S.C. Sec. 843(b)) (1988)), because each of these counts alleges as an element the facilitation of the conspiracy charged in count one. If the defendants have already been prosecuted for conduct that establishes this element, then they cannot be prosecuted for telephone counts that include this element. And the Government has acknowledged that proof of the events alleged in counts two through twenty-eight was introduced in the Adamita trial to prove the conspiracy, thereby precluding prosecution of the possession with intent to distribute offenses charged in counts twenty-six through twenty-eight, 21 U.S.C. Secs. 812, 841(a)(1), 841(b)(1)(A) (1988).

Application of Grady to the pending case leaves the Government with precisely the same option with respect to conspiracy cases that Grady left New York with respect to single occurrence cases--package in one indictment all of the offenses to be charged. Nothing prevented the Government from including in the Adamita indictment a count charging the appellants with the smaller conspiracy now sought to be charged. Had that occurred, Judge Sprizzo, upon dismissing the broader conspiracy offense for lack of evidence, would surely have submitted the narrower conspiracy offense to the jury. Whether both conspiracy offenses could have gone to the jury if supported by evidence, whether the jury in that event could have convicted for both offenses or would have been instructed to consider the offenses in sequence and to convict only on the "greater" offense if persuaded of guilt, or whether the Government must elect among conspiracy offenses prior to jury consideration are matters that need not be resolved at this time. But the Government's clear opportunity to have charged both conspiracy offenses in one prosecution supports application of Grady to the circumstances of this case.

For these reasons, I agree that the second indictment must be dismissed.

I respectfully disagree with my colleagues' conclusion that the double jeopardy rule announced in Grady v. Corbin, --- U.S. ----, 110 S. Ct. 2084, 109 L. Ed. 2d 548 (1990), operates as a bar to the prosecution of the indictment in this case.

The Grady rule forecloses a later prosecution only "if, to establish an essential element of an offense charged in that prosecution, the government will prove conduct that constitutes an offense for which the defendant already has been prosecuted." Id. 110 S.Ct. at 2093. Accordingly, we are not constrained, at this early stage in the proceedings, to examine the government's case with a view toward determining whether the government simply will prove conduct or adduce evidence that supported the earlier prosecution. Indeed, Grady requires the rejection of a "same evidence" or "actual evidence" test: "[T]he presentation of specific evidence in one trial does not forever prevent the government from introducing the same evidence in a subsequent proceeding." Id. 110 S.Ct. at 2093 (citations omitted).

What we are required to do here is to ascertain whether the government, in order to establish an essential element of an offense charged in the indictment at bar, intends to prove conduct for which Calderone and Catalano already have been prosecuted. While the government in this case very well may introduce evidence of conduct presented in the previous prosecution, and while some of that conduct may even constitute one or more offenses, it appears to me that the government has no intention of presenting evidence of conduct constituting any offense "for which the defendant[s] [have] already been prosecuted." Id. 110 S.Ct. at 2093 (emphasis added). Simply stated, the various items of conduct to be proved by the government, whether constituting offenses or not, have not been the subject of previous prosecution.

The Adamita indictment, the basis for the prior prosecution, was dismissed as to Calderone and Catalano and five other defendants at the close of the government's case. In that indictment, twenty-eight defendants originally were charged with a wide-ranging narcotics conspiracy played out over a period of three and one-half years and implicating forty-nine unindicted co-conspirators. The conspiracy, which was international in scope, involved the importation of heroin, the distribution of heroin, cocaine and marijuana, the exchange of heroin for cocaine, and various narcotics-related crimes. Listed in the indictment were one hundred and five "overt acts," with Calderone and Catalano named in but eight of those listed. As in the case at bar, the conspiracy was charged under the provisions of 21 U.S.C. Sec. 846. In granting the motions of Calderone and Catalano for dismissal under the Adamita indictment, Judge Sprizzo opined that the evidence simply was insufficient to link the movants to the broadly-based conspiracy charged, although a charge of a more narrowly-based heroin conspiracy might be viable.

The conspiracy charged in the indictment at bar is the more narrowly-based heroin conspiracy suggested by Judge Sprizzo. The conspiracy alleged covers the distribution of heroin in the New York metropolitan area and spans a period of just over one year. Only Calderone and Catalano are named as conspirators. Twenty-one "overt acts" are listed in the indictment, with these defendants named in all but two. There are but three such acts carried over from the Adamita indictment: a telephone conversation between Calderone and one Alfred Spavento on June 24, 1987; a meeting of Calderone and Spavento on the same day; and a meeting of Catalano and Spavento, also held on June 24. It would seem that elimination of evidence of the conduct of the defendants on June 24, 1987 could be accomplished without any harm to the government's case. It is not necessary to do so, however, because these items of conduct were not the subject of previous prosecution and they do not constitute offenses in any event. They are merely items of evidence that the government has chosen to set forth in both indictments under the heading, "Overt Acts."

"In contrast to the general conspiracy statute, 18 U.S.C. Sec. 371, which requires the performance of an overt act, no overt act need be alleged or proven as a necessary element of a conspiracy under 21 U.S.C. Sec. 846." United States v. Delvecchio, 816 F.2d 859, 864 (2d Cir.1987). See also United States v. Bermudez, 526 F.2d 89, 94 (2d Cir.1975), cert. denied, 425 U.S. 970, 96 S. Ct. 2166, 48 L. Ed. 2d 793 (1976). The agreement between the parties is said to be the "gist" of a narcotics conspiracy. United States v. Nusraty, 867 F.2d 759, 763 (2d Cir.1989). Of course, the requisite agreement may be established by circumstantial evidence. United States v. Cepeda, 768 F.2d 1515, 1517 (2d Cir.1985). Reasonable inferences may be drawn from the circumstantial evidence. United States v. Martino, 664 F.2d 860, 876 (2d Cir.1981). All of this is to say that the allegation and proof of "overt acts" is superfluous where defendants are charged under 21 U.S.C. Sec. 846 with a conspiracy to commit drug offenses. See United States v. Quinones, 906 F.2d 924, 926 (2d Cir.1990). If an "element" of an offense is a constituent part or ingredient of the offense that must be proved to sustain a conviction, see Black's Law Dictionary 467 (5th ed. 1979), the overt acts set out in the Adamita indictment and in the indictment at bar, in support of the conspiracies charged, do not fall within the definition. Unless one of the overlapping "overt acts" is not only an element but an "essential element" of the conspiracy charged in the Calderone indictment, the Grady double jeopardy test is not met.

Even if the overlapping "overt acts" were considered to be essential elements of each of the conspiracies, the Grady test would not be met. This is so because the conduct represented by those acts must constitute an offense previously prosecuted, in order to fit within the Grady rubric. Surely we are not prepared to say that two meetings and a telephone call constitute an offense or offenses, let alone offenses for which Calderone and Catalano were prosecuted. A meeting or a telephone call may be conduct, and may even be conduct in furtherance of a conspiracy. But a meeting or a telephone call is not conduct constituting an offense, such as driving while intoxicated or crossing the center line of a highway, as in Grady v. Corbin. Those offenses, entire in and of themselves, established essential elements--necessary constituent parts--of the homicide charges later faced by Corbin, who already had been prosecuted for the traffic violations. Here, the previous prosecution involved the greater conspiracy, an agreement among a large number of people to deal in a variety of drugs over a wide geographical area during a long period of time. The present conspiracy prosecution involves two people agreeing to deal in heroin in a limited area during a much shorter time period. Confronting us, then, are two entirely different conspiracies, and, while evidence that tends to prove the existence of one may also tend to prove the existence of the other, Grady affords no double jeopardy protection under such circumstances.

What is applicable here is the test established by this court in United States v. Korfant, 771 F.2d 660 (2d Cir.1985). The Korfant test is designed to address the merits of double jeopardy claims arising from successive conspiracy prosecutions. It requires a "totality of the circumstances" analysis, illuminated by eight factors identified "as relevant to the task of individuating conspiracies." Id. at 662. These factors are: "(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 between alleged distinct conspiracies." Id. (citations omitted).

Applying these factors, it seems clear that the indictment before us survives the Korfant test. The crimes charged in the successive indictments are distinct, involving entirely different agreements for the distribution of narcotics. The overlap of participants and time is small--only two participants in the second indictment overlap with 77 (28 defendants and 49 unindicted co-conspirators) in the first and there is a one year out of three and one-half year overlap in time. The operations were dissimilar; one involved a variety of drugs and one was limited to heroin. As has been demonstrated, only three "overt acts" overlapped. The geographic scope of the activities described in the Adamita indictment was international, while the activities described in the Calderone indictment were local to the New York metropolitan area. Other than gaining money from illegal drug transactions, there were no common objectives, and the degree of interdependence between the two conspiracies is nil. A totality of the circumstances analysis dictates rejection of the double jeopardy claim.

If Grady does not bar prosecution of the conspiracy charged in the indictment before us (Count One), it surely does not bar prosecution of the charges of using a communication facility to facilitate heroin distribution (Counts Two through Twenty-Five) and possession of heroin with intent to distribute (Counts Twenty-Six through Twenty-Eight). Calderone and Catalano never before have been prosecuted for the conduct alleged in the "telephone" and possession counts, and no conduct for which they have been prosecuted constitutes an "element" of the offenses charged in those counts. The most that can be said in this regard is that some of the same evidence used in the prior prosecution may be used in the present one to prove the "telephone" and possession counts. Such use of evidence, it must be emphasized, does not constitute double jeopardy. See Dowling v. United States, --- U.S. ----, 110 S. Ct. 668, 107 L. Ed. 2d 708 (1990).

United States v. Russo, 906 F.2d 77 (2d Cir.1990) (per curiam), a post-Grady opinion, is not inconsistent with any of the foregoing. In that case, "[o]ne of the predicate acts alleged in the RICO prosecution was the identical obstruction of justice for which [the defendant] was later tried and convicted in the present case." Id. at 78. The government conceded that the later prosecution "was inconsistent with the 'conduct' test announced in Grady." Id. Russo is a case much different from the one at bar. There, the obstruction of justice conduct did not merely establish an essential element of the offense charged in the subsequent prosecution; it was the entire crime charged in the later indictment. Moreover, the conduct itself, obstruction of justice, was prosecuted separately and specifically as a discrete offense necessary to establish a predicate act.

Because I believe that the reading of Grady by my colleagues is far too expansive, I am compelled to dissent.