Defendant-appellant Darrell F. Pierro labors to convince us that the district court erred in refusing to grant him a separate trial, in refusing to declare a mistrial, and, following his conviction, in refusing to reduce his sentence beneath the suggested guideline range. We are not persuaded by appellant’s exhortations and, therefore, affirm.
I. BACKGROUND
At the times material hereto, appellant earned his livelihood as a vice-president of the Moore Group (MoGro), a California company. Yielding to temptation, he also joined a criminal cartel that, during the years 1989 and 1990, engaged in the theft and subsequent resale of computer components manufactured by and for Digital Equipment Corporation (DEC). This scheme functioned on three levels. The initial step involved the thefts — a step in which appellant at first did not participate. The second step involved the sale of the stolen equipment; with appellant’s connivance, his employer, MoGro, purchased much of the contraband. 1 The third step involved the purchasers’ disposal of the bootleg merchandise.
For its part, MoGro, under appellant’s aegis, handled this third phase in two ways. It returned some components to DEC, after altering their serial numbers, as part of an established exchange program, thus converting stolen, often unusable components into new, state-of-the-art equipment. It resold the rest of the components on credit terms to a Wisconsin firm, and then pledged the invoices as security for bank loans. MoGro used the loan proceeds, inter alia, to pay the thieves for the stolen merchandise.
From and after late 1989, appellant assumed an active role in the looting of DEC’s warehouse. On several occasions, he and fellow MoGro employees (including John McComas) flew from California to Massachusetts and assisted in the unlawful asportation of computer components. These purloined parts subsequently were shipped to MoGro’s California headquarters and disposed of by one of the two methods we have described.
In early 1990, appellant and several confederates were spotted inside DEC’s warehouse, fled, and were eventually apprehended. Subsequently, a federal grand jury returned a 158-count indictment against 16 persons. It charged appellant with conspiracy to participate in a racketeering enterprise, see 18 U.S.C. § 1962(d), participating in a racketeering enterprise, see id. § 1962(c), and money laundering, see id. § 1956(a)(1). The predicate acts upon which the RICO charges rested included both money laundering and interstate transportation of stolen property, see 18 U.S.C. § 2314.
In response to a clutch of severance motions, including one filed to appellant’s be- *615 hoof, the district court split the defendants into two groups for purposes of trial. The court’s order called for appellant and seven other alleged coeonspirators (including McComas, Kleinerman, and Ruslan Moore, MoGro’s president) to be tried together, but apart from the other eight defendants. On September 8,1992, trial commenced for most members of appellant’s group. 2 During the trial, the court denied appellant’s renewed severance motion and his motion for a mistrial. The jury found appellant guilty on all counts. The court sentenced him to serve 121 months in prison. This appeal followed.
II. THE ALLEGED TRIAL ERRORS
Appellant contends that the district court erred in denying his renewed motion for severance and his motion to declare a mistrial. We examine each of these contentions.
A. The Severance Motion.
We need not linger long over the question of severance. “As a rule, persons who are indicted together should be tried together.”
United States v. O’Bryant,
On appeal, Pierro does not challenge the district court’s pretrial order segmenting the defendants into two groups for purposes of trial. However, he complains bitterly about the district court’s refusal to grant his mid-trial motion for severance — a device by which he sought to put some distance between himself and Kleinerman. The motion rested on twin rationales: first, Kleinerman’s testimony about a litany of “bad acts” which had nothing to do with appellant; second, Klein-erman’s courtroom antics, which, appellant alleges, reflected adversely on all the defendants. We bifurcate this complaint, considering these grounds separately. ■
1.
Spillover.
The first aspect of appellant’s argument amounts to a claim of spillover prejudice. To prevail on such a claim, a defendant must prove prejudice so pervasive that a miscarriage of justice looms.
See United States v. Sabatino,
To be sure, Kleinerman testified about a bogus burglary he staged at his home and about telling another witness that she should have dissembled when appearing before the grand jury. This testimony, like other bits and pieces of evidence about which appellant complains, while unsavory, was not in any way antagonistic to appellant’s defense. And nothing implicated appellant in the peccadilloes; to the contrary, the evidence suggested he was on the other side of the continent both when Kleinerman faked the break-in and when Kleinerman attempted to suborn perjury. Since it is settled that properly joined defendants need not be severed merely because a joint trial will require that the jury receive testimony — even a large amount of testimony — irrelevant to one defendant,
see Boylan,
Although perhaps supererogatory in light of the foregoing, we also take note of
*616
the trial court’s exemplary handling of the situation. The court carefully controlled the presentation of the proof, making the jury keenly aware that certain evidence was limited to particular defendants, and that, in all events, the evidence had to be considered separately against each defendant. In the first instance, a reviewing court must presume that the jury heeded these prophylactic instructions.
See United States v. Olano,
- U.S. -, -,
2. Kleinerman’s Behavior. Appellant also assigns error to the district court’s denial of a severance based on what he calls Kleinerman’s “eheerleading” during the trial testimony of various witnesses. The conduct that appellant attributes to Kleinerman— mostly gestures and grimaces — obviously occurred; indeed, the district judge found Kleinerman in contempt for his courtroom antics.
If, during the course of a multi-defendant criminal trial, a defendant misbehaves in the jury’s presence, the misbehavior usually will not compel a separate trial for his codefendants.
See, e.g., United States v. Rocha,
Applying this standard, we do not think that Kleinerman’s pantomime, regrettable though it was, required the
nisi prius
court to grant a severance. In our view, this situation parallels — but is much less noxious than — other situations in which courts have concluded that well-chosen, well-timed curative instructions will satisfactorily ameliorate the adverse effects of a defendant’s inappropriate behavior on his codefendants.
3
See, e.g., Rocha,
3.
Recapitulation.
Trial courts are afforded considerable leeway in determining severance questions.
See O’Bryant,
B. The Motion for a Mistrial.
After 18 days of trial, the government negotiated a plea agreement with McComas and, after McComas pled guilty, called him as a prosecution witness. Appellant immediately moved in limine to bar McComas from testifying. The district court denied that motion. Appellant subsequently moved for a mistrial, arguing that McComas’s abrupt change of plea, followed by potentially in *617 criminating testimony, would be unfairly prejudicial. The district court refused to grant the requested relief. Appellant assigns error.
Motions to declare mistrials are directed primarily to the district court’s discretion,
see United States v. Sepulveda,
Of course, this does not mean that appellate courts should reflexively rubber-stamp the trier’s refusal to declare a mistrial. But because mistrials are strong medicine, disruptive not only to the parties but also to the judiciary’s efforts to manage crowded dockets, it is only rarely — -and in extremely compelling circumstances- — that an appellate panel, informed by a cold record, will venture to reverse a trial judge’s on-the-spot decision that the interests of justice do not require aborting an ongoing trial.
See Real v. Hogan,
As a general matter, a mistrial is not automatically required when a codefendant changes his plea in mid-trial.
See, e.g., United States v. Del Carmen Ramirez,
Appellant asserts that this case evades the usual rule because special circumstances exist. He points to the fact that the three “MoGro defendants” — appellant, McComas, and Moore — had been pursuing a common defense, and that McComas’s change of plea and his ensuing testimony knocked the pins out from under this defense. 6 He adds, moreover, that the deleterious side effects of Kleinerman’s misbehavior, see supra Part 11(A)(2), furnished a further basis for a mistrial.
We do not believe that this jeremiad derives sufficient support from the record. The trial judge found no cognizable prejudice, and close perserutation of the transcript *618 fails to shake this finding. At any rate, given the highly deferential standard of review, we are not prepared to second-guess the finding based on appellant’s self-interested speculation. 7
III. THE ALLEGED SENTENCING ERROR
To place appellant’s final assignment of error into proper perspective, we divide this portion of our analysis into three sections. First, we rehearse the sentencing calculus. 8 Second, we discuss the question of appellate jurisdiction. Third, we consider the substance of the claimed error.
A. The Sentence.
The district court classified the counts of convictions as comprising two groups, see U.S.S.G. § 2E1.1, comment, (n.l), one for interstate transportation of stolen property, see 18 U.S.C. § 2314, and one for money laundering, see 18 U.S.C. § 1956(a)(1). 9 The judge determined that, under U.S.S.G. § 2B1.2(a), the first group had a base offense level (BOL) of 4. The court then added 15 levels for bringing about a loss in excess of $2,500,000 (but less than $5,000,000), see U.S.S.G. § 2Bl.l(b)(l)(P); 4 levels for engaging regularly in the business of buying and selling stolen property, see U.S.S.G. § 2B1.2(b)(4)(A); and 3 levels for performing a managerial role in the offense, see U.S.S.G. § 3Bl.l(b). These calculations yielded an adjusted offense level of 26.
The judge performed a similar set of computations for the second group of convictions. He determined that this group revolved around money laundering and, therefore, used U.S.S.G. § 2Sl.l(a)(l) to fix the BOL at 23. 10 The judge then added 7 levels because the value of the laundered funds exceeded $3,500,000, see U.S.S.G. § 2Sl.l(b)(2)(H), bringing the adjusted offense level to 30.
Since the second group produced a substantially higher adjusted offense level than the first group, the district court, following the praxis specified by the Sentencing Commission, see U.S.S.G. § 2El.l(a), set the first series of computations to one side and, instead, added 2 levels to the second group’s adjusted offense level, pursuant to section 3D1.4, thus bringing the final offense level to 32. This produced a guideline sentencing range (GSR) of 121-151 months for a first-time offender. The court imposed an incar-cerative sentence at the bottom of the range.
On appeal, Pierro concedes that these calculations are supportable, but claims that the lower court erred in not venturing a downward departure premised on “mitigating circumstances.” See 18 U.S.C. § 3553(b) (providing, inter alia, for departures if the court ascertains that there exists a “mitigating circumstance of a kind, or to a degree, not adequately taken into consideration by the Sentencing Commission in formulating the guidelines that should result in a sentence different from that described [in the GSR]”); U.S.S.G. § 5K2.0 (implementing statute). The government demurs.
B. Appellate Jurisdiction.
As a matter of first principles, an appellate court is duty bound to confirm the existence of its own jurisdiction.
See Juidice
*619
v. Vail,
It is by now axiomatic that a criminal defendant cannot ground an appeal on a sentencing court’s discretionary decision not to depart below the guideline sentencing range.
See, e.g., United States v. Tardiff,
Counsel often confuse the exception and the rule. If the judge sets differential factfinding and evaluative judgments to one side, and says, in effect, “this circumstance of which you speak, even if it exists, does not constitute a legally sufficient basis for departure,” then the correctness of that quintessentially legal determination may be tested on appeal. But if the judge says, in effect, either that “this circumstance of which you speak has not been shown to exist in this case,” or, alternatively, that “while this circumstance of which you speak might exist and might constitute a legally cognizable basis for a departure in a theoretical sense, it does not render this particular case sufficiently unusual to warrant departing,” then, in either such event, no appeal lies.
We think that this case fits within the exception rather than the rule, and, hence, that we have jurisdiction to consider the assigned error. At the disposition hearing, appellant identified three possible grounds for departure, namely, (1) that his case was, in essence, a sheep in wolves’ clothing — a garden-variety theft-of-property case, treated by the guidelines as a money laundering case, and, therefore, well outside the heartland of the money laundering guideline; (2) that his GSR was skewed by “double counting”; and (3) that, if sentenced within the GSR, his punishment would be vastly disproportionate to his codefendants’ sentences. The district court rejected all three bases for departure. Read objectively, the district court seems to have said, in effect, that even if appellant could prove the subsidiary facts upon which his arguments rested— that his conviction grew out of a scheme to steal property rather than a scheme to launder money, that double counting influenced the composition of the GSR, and that his coconspirators received sentences milder than the GSR in his case prophesied — none of the cited circumstances would constitute a legally cognizable reason for imposing a sentence below the GSR. If the district court erred in this determination, the error was a purely legal one. Thus, appellate jurisdiction attaches.
C. The Merits.
We address separately appellant’s claim that his conduct fell outside the heartland of the money laundering statute, thereby justifying a downward departure. We then proceed to examine appellant’s remaining sentence-related claims.
1.
The Essence of the Offense.
In analyzing appellant’s “heartland” claim, we first step back to review the anatomy of “mitigating circumstance” departures. The method of the sentencing guidelines demands that, in the ordinary case, the judge apply the guidelines, make such interim adjustments as the facts suggest, compute a sentencing range, and then impose a sentence within that range.
See
18 U.S.C. § 3553(a), (b);
see also United States v. Rivera,
When a sentencing court considers a “mitigating circumstance” departure, the relevant circumstance must be of a kind cognizable under the guidelines,
see Rivera,
We accept appellant’s two subsidiary premises. First, his involvement in money laundering arose out of his use of proceeds from the sale of stolen property as security for bank loans. See 18 U.S.C. § 1956(a)(1)(A) (criminalizing the conduct of a financial transaction with knowledge that the property involved in the transaction represents the proceeds of specified forms of illegal activity). Second, the Sentencing Commission has chosen to punish money laundering with particular severity, and the introduction of the money laundering guideline into the sentencing calculus therefore resulted in a markedly higher GSR and a longer prison term for appellant, see supra Part 111(A).
Be that as it may, we cannot accept the conclusion that appellant draws from these two premises. The money laundering statute does not exempt from its reach those persons who launder money merely in the furtherance of underlying criminal activities. Nor does the statute, in terms, suggest that such persons’ actions perforce fall outside the statute’s proper scope. On the contrary, the crime colloquially known as money laundering is committed whenever a person, “knowing that the property involved in a financial transaction represents the proceeds of some form of unlawful activity” nevertheless “conducts or attempts to conduct ... a financial transaction which in fact involves [such] proceeds.” 18 U.S.C. § 1956(a)(1).
In our view, Congress meant this statute to address, among other things, conduct undertaken subsequent to, although in connection with, an underlying crime, rather than merely affording an alternative means of punishing the underlying crime itself.
See United States v. Johnson,
There is little question that the appellant’s conduct fits snugly within this framework. Appellant argues that “although the facts which formed the predicate for the convictions were within the strict linguistic parameters of the sentencing guidelines for money laundering, the defendant’s actual conduct did not fall within [what the Sentencing Commission intended to punish as] money laundering .•... ” This is virtually a replica of the argument we rejected in
LeBlanc,
*621
2. Remaining Bases for Departure. Appellant suggests two additional ways in which the trial court appropriately could have departed downward. Neither suggestion has any merit. 12
a.
First, appellant contends that a downward departure could have been predicated on the fact that “double counting” boosted his GSR to heights not contemplated by the Sentencing Commission. In this regard, appellant asserts that the same money was factored into the sentencing court’s computations twice — once in calculating the offense level for money laundering and once in calculating the offense level for interstate transportation of stolen property. We cannot accept this assertion.
It is not at all clear that any double counting took place. As discussed above, where an underlying crime occurs antecedent to money laundering, the offenses are considered separate and distinct for sentencing purposes. This distinctiveness requires that a separate computation be made for each group of offenses.
See United States v. Lombardi,
By the same token, each crime has its own measure of loss; the value of the property stolen from DEC and the dollar amount of ill-gotten sale proceeds used by MoGro to secure financial support may turn out to be the same, but they are arrived at differently. The mere existence of some indeterminate degree of overlap between these figures does not constitute double counting.
See, e.g., United States v. Lilly,
To say more would be to trespass on the reader’s indulgence. Even if the situation here could be described in some useful way as comprising double counting, the phenomenon is not sufficiently “special” or “unusual” *622 to warrant a downward departure. After all, in the sentencing context double counting is not rare — and the practice is often perfectly proper. See id. at 19.
b.
The final circumstance on which appellant relies in support of a downward departure is disproportionality — the comparative severity of his sentence as contrasted with the sentences to be served by other coconspirators.
13
The district court believed that it lacked authority to depart on this basis. We concur.
See, e.g., United States v. Wogan,
3.
Need for Remand.
The district court sentenced appellant in February 1993. Approximately five months later, this court decided
Rivera,
Building a body of precedent is an evolutionary process. If the mere fact that a new opinion sheds light on an area of the law automatically required appellate courts to remand for reconsideration all cases pending on direct appeal that dealt with the same area of the law, the system would become a shambles. Remand is required only when there is a realistic possibility that the new precedent, properly applied, will alter or otherwise materially affect the result reached in the trial court.
14
See, e.g., Gifford,
When a newly minted precedent clarifies a corner of the law while a case involving the same (or a closely related) point is pending on direct appeal, the threshold question is almost always whether the trial court’s analysis would have differed in some material respect if it had had the benefit of the clarification. Here, that question demands a negative answer: Judge Woodlock fully anticipated our opinion in
Rivera,
carefully sifted the record to determine whether any unusual circumstances existed that might warrant a downward departure, and, discovering none, correctly abjured the desired departure. Hence, a remand would serve no useful purpose as the analytic approach would be essentially unchanged and would, therefore, produce the same conclusions.
See, e.g., United States v. Smith,
We hasten to add that even when a district court has not fully anticipated an emergent clarification, a remand will not necessarily follow. For example, when the court’s subsidiary findings of fact are reasonably explicit, unaffected by its legal error, and subject to reuse, a remand would be an empty exercise.
See Societe des Produits Nestle v. Casa Helvetia, Inc.,
*623 This principle offers an alternative basis for denying appellant’s request for a remand. Even if, without the benefit of Rivera, the district court’s grasp of departure jurisprudence proved faulty — and we do not believe that to have been the case — a remand would not be exigible. For all the skillful lawyering that has been mustered on appellant’s behalf, he has been utterly unable to isolate any “special” or “unusual” feature of this case which, under Rivera, could support a downward departure.
IY. CONCLUSION
We need go no further. For aught that appears, appellant was fairly tried, justly convicted, and appropriately sentenced. He has been unable to advance any persuasive reason either for revising the outcome or for prolonging the proceedings. The judgment below must, therefore, be
Affirmed.
Notes
. The thieves stole computer components from DEC's warehouse in Massachusetts both during the week and on weekends. The components purchased by MoGro comprised, for the most part, the bounty from the weekend heists. The remaining contraband was sold mainly to a code-fendant, Fred Kleinerman.
. One of the eight defendants originally included in this cadre pled guilty prior to trial.
. Kleirierman’s misbehavior did not plummet to the depths reached in
Rocha,
. There is, moreover, an added fillip in this case that further erodes appellant’s position: his assumption that the jury must have been aware of Kleinerman's antics is sheer speculation. Klein-erman's gestures were nonverbal, and the judge did not comment on them in the jury’s presence.
. We recognize that when a codefendant switches sides and becomes a government witness, his testimony will likely help the government and harm the remaining defendants. Indeed, this is often a strong incentive to the government in the plea-bargaining process. But this sort of prejudice does not necessitate a mistrial. It is only
unfair
prejudice against which courts must guard.
See, e.g., United States v. Rodriguez-Estrada,
. Although three different lawyers represented the MoGro defendants, appellant asseverates that the lawyers coordinated their opening statements and their cross-examination of prosecution witnesses, and that they adopted a “unified theme” in regard to challenging the government’s proof and confronting its witnesses. After McComas changed his plea, appellant alone was caught in the toils; the third MoGro defendant, Moore, found sanctuary when the district court granted his motion for judgment of acquittal, Fed. R.Crim.P. 29(a).
. There is another reason why we are particularly reluctant to override the trier’s judgment call in this case. If any unfair prejudice sprouted— and we stress that we have found no sign of any — we believe it would have been cured by the trial judge’s painstaking instructions. It strikes us as significant in this regard that appellant has criticized neither the content nor timeliness of these instructions.
. The district court sentenced appellant in Feb-ruaiy of 1993, using the November 1992 edition of the sentencing guidelines.
See United States v. Harotunian,
. Since "racketeering comes in many shapes and sizes, and covers a wide range of activities,”
United States v. Winter,
. For purposes of computing the sentencing range, it is unnecessary to distinguish between the substantive money laundering offenses and money laundering as a RICO predicate.
. In
LeBlanc,
following convictions for,
inter alia,
illegal gambling activities and money laundering, the court computed a GSR for each defendant. There, as here, the range was pushed upward by the presence of money laundering. To correct this perceived bias, the district court departed downward on the theory that the defendants’ crime, at bottom, constituted bookmaking, and should be sentenced as such; the money laundering, the court thought, was merely incidental to the illegal gambling, and, therefore, the offenses of conviction, although technically including money laundering, fell outside the heartland for that crime.
See LeBlanc,
. Appellant gives these points only cursory treatment in his appellate brief. We could, therefore, simply dismiss them on that ground.
See Ryan v. Royal Ins. Co.,
. As we have previously indicated, this lack of symmetry stems primarily from appellant's involvement in money laundering — a circumstance that boosted his GSR well above what it would have been under the guideline covering interstate transportation of stolen property. None of the other defendants carried similar baggage into the sentencing arena.
. For purposes of this discussion, we assume that the new precedent constitutes a clarification, entitled to retroactive effect vis-a-vis cases still pending on direct appeal. Broader retroactivity concerns are beyond the proper scope of this opinion.
