United States v. David M. Pharis Edward J. Habina William M. Dull Harry Gangloff, 298 F.3d 228

Lead Opinion

Argued May 31, 2001.

Before: SLOVITER, FUENTES and COWEN, Circuit Judges.

Reargued En Banc Feb. 13, 2002.

Before: BECKER, Chief Judge, SLOVITER, MANSMANN,^* SCIRICA, NYGAARD, ALITO, ROTH, MсKEE, RENDELL, BARRY, AMBRO, FUENTES and COWEN, Circuit Judges.

Filed July 26, 2002.

OPINION OF THE COURT

SLOVITER, Circuit Judge.

After the jury for this federal criminal case had been sworn, after two witnesses had testified, and after the testimony of a third witness had been stipulated, the Government filed a notice of appeal from an adverse ruling by the District Court. The issue before us is whether we can hear its appeal.

I.

FACTS AND PROCEDURAL BACKGROUND

On November 23, 1999, a federal grand jury returned a six-count indictment charging that from January 1988 until June 1995 David Pharis, Edward Habina, William Dull, and Harry Gangloff (Defendants) committed mail fraud, in violation of 18 U.S.C. § 1341 (2001), by inflating consulting bills that they submitted to insurance companies. Defendants’ business, S.T. Hudson International, Inc., and its affiliates specialized in providing services to insurance companies which had large influxes of claims following disasters. Pharis was the CEO and president of Hudson, Habina the vice president, Dull an associate, and Gangloff a computer consultant.

According to the indictment, Defendants began their fraudulent scheme in 1989 by manually changing consulting bills. Specifically, the indictment alleged that starting in 1989 and continuing until February 1994, Pharis, Habina, and Dull “manually changed, or instructed their employees to manually change, the ‘pre-bills’ that accurately reflected the consultants’ billings, by inflating the number of hours the consultant worked.” App. at 14. The indictment further alleged that in February 1994, Pharis directed Gangloff to develop a computerized billing program, known as the “gooser,” that automatically multiplied the hours each consultant worked by a factor of 1.15 and then added an additional half hour to the total hours billed. App. at 15. In payment of bills produced by this computer program, Hudson’s clients mailed to Hudson the six checks that form the basis for the Governmеnt’s charge of mail fraud.

The trial was scheduled to start on Monday, September 25, 2000. When the Government filed its trial memorandum on Wednesday, September 20, it included a motion in limine seeking to offer, under either Federal Rule of Evidence 402 or 404(b), evidence of an incident of uncharged misconduct by Dull (specifically, the wrongful retention of a client’s overpayment) that appears to be unrelated to the alleged inflation of any client’s bill. On Friday, September 22, Habina responded in opposition to the Government’s in limine motion, arguing that the proposed evidence was “utterly irrelevant.” App. at 47. That motion was among the matters addressed in the order entered by the District Court on September 26, 2000 and was denied. The Government does not press that matter on appeal.

Also on Friday, September 22, Pharis filed a Joint Trial Motion to Redact Indictment and Motion In Limine No. 1 (hereafter “Motion to Redact”), accompanied by a supporting memorandum of law from all Defendants. In that memorandum, Defendants argued that the Government exhibits revealed that there were really two separate schemes — the manual billing scheme which ended in February 1993 and the computerized billing scheme which began in February 1994. Defendants claimed that the schemes differed in methodology, scope, and participants, that the statute of limitations barred criminal liability for the manual billing scheme, that there was no federal jurisdiction alleged as to that manual billing because there were no related mailings alleged and that only the computerized billing scheme was actionable as federаl mail fraud. The motion, which specifically alleged that the Government was improperly charging two separate fraud schemes, requested that the District Court redact from the indictment all references to the earlier scheme and exclude all evidence relating to it or, in the alternative, that the court direct the Government to show the admissibility of such evidence under Rule 404(b) by demonstrating that the probative value of the pre-1994 evidence was substantially greater than its prejudicial effect. Defendants sought by their Motion to Redact to preclude the Government from presenting any evidence about the manual changes to bills that occurred from 1989 to February 1994, a period that covers all but roughly a year and a half of the time described in the indictment.

On Monday, September 25, the Government responded to the Motion to Redact, defending the indictment’s allegation of a single scheme. The jury was sworn in that afternoon.

On Tuesday, September 26,¹ after the jury had been given preliminary instructions but before opening statements in the trial, the District Court granted Defendants’ Motion to Redact. In its memorandum, the District Court, in addressing the Government’s motion regarding the evidence of Dull’s wrongful retention of a client’s overpayment, discussed the circumstances under which evidence of uncharged misconduct could be admitted pursuant to Rule .404(b). Interspersed with the District Court’s discussion of that evidence were comments more pertinent to the issue raised by the Defendants’ Motion to Redact. Thus, the court stated, “Here, the Government seeks to introduce evidence of acts allegedly occurring between 1989 and February 1994.” United States v. Pharis, No. 99-CR-743, slip op. at 3, 2000 WL 1469330 (E.D.Pa. filed Sept. 26, 2000) [hereinafter Sept. 26 order].

After additional comments on the wrongfully retained payment, the court then shifted to a discussion of the scope of the federal mail fraud statute, stating that the statute reaches “only those limited instances in which the use of the mails is part of the execution of the fraud.” Id. at 3 — 4. The court, noting that “ ‘[t]he mailings in this case are checks that customers sent to Hudson in payment for Hudson’s services’ ” and that “ ‘[t]he scheme was not complete until Hudson received the fruits of the scheme, that is, the checks,’ ” concluded that inasmuch as all six counts in the indictment relate to mailings in 1994 and 1995, those “acts relating to conduct other than that alleged in the indictment constitute other acts that do not demonstrate violations of the mail fraud statute, but show other acts of Defendants’ bad character.” Id. at 4 (quoting Government’s in limine motion) (first alternation in original). This led the District Court to hold that evidence of such other acts was inadmissible under Rule 404(b) and to conclude, after performing the balancing test required under Rule 403, that the probative value of such evidence was substantially outweighed by the danger of unfair prejudice. Id. at 5.

After the District Court granted Defendants’ motion, Government counsel, prior to delivering her opening statement, requested a sidebar conference for clarification of the court’s order. The court denied this request. Counsel then presented her opening statement and called a witness. During this testimony, the District Court reiterated that the Government would not be permitted to prove any acts of altering of bills prior to 1994. App. at 174-76. After the Government’s first witness, the court recessed until the following morning.

On September 27, the day after the District Court’s original ruling was announced but before the trial continued, the Government filed a motion for reconsideration. In that motion the Government made essentially the same arguments it has raised before us on appeal. The District Court declined the Government’s request for a continuance until the motion for reconsideration was ruled upon. The Government proceeded to call another witness and the parties agreed to a stipulation about the testimony of a third witness.

The following day, September 28, in rejecting the motion for reconsideration, the District Court again quoted Rule 404(b) at length, said the manually changed bills were a “collection of uncharged bad acts ... [that are] utterly unrelated to the offenses which are charged,” United States v. Pharis, No. 99-CR-743, slip op. at 3 (E.D.Pa. Sept. 28, 2000) [hereinafter Sept. 28 order] (emphasis added), and held those “prior bad acts” to be inadmissible. Id. at 4-5. And again, citing Rule 403, the District Court, using the language of Rule 404(b), repeated that, although the evidence could be used to show Defendants’ “motive, intent, plan, scheme or course of conduct,” the probative value of the evidence was substantially outweighed by the danger of unfair prejudice. Id. at 3-4. The Court concluded:

To the extent to which the Court’s prior Order intimated that it redacted the indictment, the Court now clarifies its prior order. First, the Court based its decision to bar admission of any evidence during the period 1989 to 1992 on the basis of Federal Rules of Evidence 403, rather than on any authority to redact the indictment....

[U]nder Federal Rule of Evidence 403, this Court is granted substantial discretion in balancing the probative value of evidence on the one hand and the unfair prejudice of evidence on the other.

Id. at 4.

After receiving the District Court’s denial of its motion for reconsideration, the Government filed a notice of appeal from both the District Court’s order filed September 26, 2000, the day after the jury was sworn, and its order dated September 28, 2000 denying the Government’s motion for reconsideration. The Government requested that the case be stayed pending the appeal. Defendants filed motions for judgment of acquittal. Defendants also asked the District Court to continue the trial with the current jury. The District Court denied these three requests and dismissed the jury.

II. ANALYSIS

A. The District Court’s September 26 Order

On appeal, Defendants do not defend the District Court’s order of September 26, its order denying the Government’s motion for reconsideration and the court’s justification for these orders. See Transcript of Oral Argument (Feb. 13, 2002) at 57 [hereinafter Tr.] (agreeing that the indictment charged a single scheme) and Tr. at 59-60 (stating that the District Court misunderstood both the Government’s motion in limine and Defendant’s Motion to Redact).

The District Court’s September 26 order, in addition to denying the Government’s motion to admit evidence of other misconduct of Defendants, also denied “the Government’s motion to admit evidence of uncharged misconduct” under Rule 404(b) and granted Defendant’s Motion to Redact. Sept. 26 order at 7-8. In so ruling, the District Court appears to have operated under a misunderstanding as to the nature of the Government’s charge and the applicable law.² The Court stated that because all six counts of the indictment rested on mailings in 1994 and 1995, the earlier manually changed bills could not be part of the mail-fraud scheme. On this point the District Court was plainly mistaken. The elements of the mail-fraud counts are: (1) a scheme to defraud; (2) use of the mails to further that scheme; and (3) fraudulent intent. See United States v. Sturm, 671 F.2d 749, 751 (3d Cir.1982). The indictment charged that Defendants perpetrated a single scheme of mail fraud-one that encompassed both the manually altered bills and the computerized bills. It is not of any legal significance that the mailings used to bring the scheme under the mail-fraud statute occurred at the end of this single scheme. A fraudulent scheme can, span many years with the mailings occurring only at the end of the period. See Schmuck v. United States, 489 U.S. 705, 711-14, 109 S.Ct. 1443, 103 L.Ed.2d 734 (1989); United States v. Sampson, 371 U.S. 75, 80 & n. 5, 83 S.Ct. 173, 9 L.Ed.2d 136 (1962); United States v. Ruuska, 883 F.2d 262, 264-65 (3d Cir.1989); United States v. Lebovitz, 669 F.2d 894, 896 (3d Cir.1982).

In interpreting the federal mail-fraud statute, the Supreme Court has long held that "[i]t is not iiecessary that the scheme contemplate the use of the mails as an essential element." Pereira v. United States, 347 U.S. 1, 8, 74 S.Ct. 358, 98 L.Ed. 435 (1954). All that is required is that the defendants knowingly participated in a scheme to defraud and caused a mailing to be used in furtherance of the scheme. See, e.g., Sturm, 671 F.2d at 751; United States v. Pearlstein, 576 F.2d 531, 534 (3d Cir.1978).³

The fact that in 1994 Defendants changed the method that they used to make increases to the bills they sent their clients does not establish that there were two different schemes. The District Court was in , error if it believed that the shift from increasing bills manually to increasing them by use of the "gooser" computer program gave rise to a different scheme. By itself, changing the method used to commit a fraud does not inaugurate a new fraudulent scheme.

Once the District Court concluded that the manually altered bills were not properly part of the charged scheme, it applied Rule 404(b), finding that the evidence of the manually changed bills constituted prior bad acts that should be excluded from evidence. Again, the District Court was mistaken. Once certain fraudulent conduct is properly included as part of the charged scheme, Rule 404(b) poses no obstacle because evidence that is part of the charged crime does not fall under Rule 404(b). See, e.g., United States v. Gibbs, 190 F.3d 188, 217-18 (3d Cir.1999); see also United States v. Bowie, 232 F.3d 923, 929 (D.C.Cir.2000); United States v. Miller, 116 F.3d 641, 682 (2d Cir.1997); United States v. Chin, 83 F.3d 83, 87-88 (4th Cir.1996).

The District Court, apparently as part of its Rule 404(b) analysis, then applied Rule 403’s balancing test.⁴ The District Court explained that whatever the manually changed bills showed about Defendants’ motive, intent, plan, scheme, etc., the probative value of the evidence was “substantially outweighed by the danger of unfair prejudice.” Sept. 26 order at 4.

Both parties agree that the District Court made mistakes. What should be done about these mistakes is, however, a question we can reach only if we have jurisdiction to hear this appeal.

B. Jurisdiction

1. Dismissal

The threshold issue before us is whether we have jurisdiction to hear the Government’s appeal — an appeal filed after the jury was sworn, after several Government witnesses testified, and after the District Court, on the Government’s motion, reconsidered and clarified its initial ruling.⁵

In its brief, the Government rests our jurisdiction over its interlocutory appeal on 18 U.S.C. § 3731. Br. of Appellant at 20. That statute, the Criminal Appeals Act, states, in relevant part:

In a criminal case an appeal by the United States shall lie to a court of appeals from a decision, judgment, or order of a district court dismissing an indictment ... as to any one оr more counts, except that no appeal shall lie where the double jeopardy clause of the United States Constitution prohibits further prosecution.

An appeal by the United States shall lie to a court of appeals from a decision or order of a district court suppressing or excluding evidence ... in a criminal proceeding, not made after the defendant has been put in jeopardy and before the verdict or finding on an indictment or information, if the United States attorney certifies to the district court that the appeal is not taken for purpose of delay and that the evidence is a substantial proof of a fact material in the proceeding.

The provisions of this section shall be liberally construed to effectuate its purpose.

18 U.S.C. § 3731.

The Government argues that the District Court’s order amounted to a dismissal of the indictment by preventing it from proving substantial, material allegations of the indictment. However, nothing in the District Court’s order purports to dismiss the indictment, all six counts alleged in the indictment constituting the mail fraud charges remained in the indictment as returned by the grand jury, and Defendants’ Motion to Redact, which the court granted, did not ask for dismissal of the indictment. Further, in its original objection to Defendants’ Motion to Redact, the Government did not object that granting the motion would amount to a dismissal. That contention emerged only in connection with the Government’s motion for reconsideration.

The Government’s argument that the District Court’s order “amounted to” a dismissal of the indictment focuses on the supposed effect of the order, namely that the ordеr precluded the Government from proving the allegations of the mail fraud contained in the indictment. The Government argues that the District Court “barred the government from proving the fraudulent scheme which the grand jury described.” Br. of Appellant at 22. The Government amplified this contention at oral argument before us when it stated that the District Court’s ruling “gutted” its case because the ruling prevented it from proving the scheme to defraud. Tr. at 12. In effect, the Government is claiming that a dismissal occurred simply because it lost an evidentiary ruling important to its case. Thus, the underlying basis for the Government’s appeal appears to be the considerable effect the District Court’s evidentiary ruling may have had on the Government’s case.

Defendants concede that “hypothetically ... an order having the form of excluding evidence ... so sweeping that it precluded any and all proof of a count of an indictment ... might be characterized by an appellate court as a dismissal under § 3731 rather than as an evidentiary ruling,” yet they contend that “that hypothetical is far from this case.” Br. of Appel-lees at 28-29. We agree. Even if the District Court’s ruling were as preclusive as portrayed by the Government, it did not preclude the Government from attempting to prove a scheme to defraud using post-February 1994 evidence relating only to the computerized billing method. After all, the indictment alleged mail fraud, all six mailings that constituted the acts of mail fraud were made after February 1994 and arose out of the computer generated bills, and those counts remained intact following the order the Government seeks to appeal.

It is acknowledged by the parties that Defendants shifted to computerized billing as a result of the influx of bills that needed processing in the wake of the January 1994 Northridge Earthquake in Los Angeles. The Government contends that before computerization, Defendants or their employees would (a) draft a pre-bill that accurately reflected the hours worked as listed on a consultant’s time sheet, (b) manually “mark-up” that pre-bill to indicate that the consultant had worked more hours than listed by the consultant, and then (c) send a final bill charging the client for the hours as increased. The Government explains that once Defendants devised a computerized billing system, with its automatic fraudulently inflated hours, there was no need for any communication among Defendants and their employees and hence no evidence comparable to that available regarding the manual overbilling. The Government candidly concedes that its concern is that the District Court’s order would have prevented it from proving the fraudulent scheme beyond a reasonable doubt because the evidence of the earlier manual overbilling was necessary to rebut Defendants’ defense that the computerized over-billing was merely an error. But that effect cannot be the basis for transforming an evidentiary ruling into a dismissal, any more than a midtrial order rejecting essential evidence proffered by the government in a drug trial can be considered a dismissal.

The second paragraph of § 3731 authorizes the government to take an interlocutory appeal from a decision or order excluding evidence but only under certain conditions. One such condition is that the order be one “not made after the defendant has been put in jeopardy.” 18 U.S.C. § 3731. The Government concedes that Defendants were put in jeopardy by September 26, 2000. It thus follows that the statutory provision fashioned to cover precisely this situation, the exclusion of evidence that the Government would have certified was a “substantial proof of a fact material in the proceeding,” id., was not available to the Government.

The Government could have protected itself from the situation in ‍‌‌‌​​​​​‌‌​​​​‌‌‌​​​​‌​‌​​​‌​‌‌‌​‌​‌​​‌‌​​‌​‌‌‌‌‍which it finds itself. Defendants’ Motion to Redact, which, inter alia, moved to exclude all evidence pertaining to Hudson’s billings before February 1994, was filed in the afternoon of Friday, September 22. The jury was not impaneled until Monday, September 25. Federal Rule of Criminal Procedure 12(e) requires that the District Court decide pre-trial motions before the trial begins unless the court has deferred decision for good cause; however, the rule precludes such deferral if a party’s right to appeal is adversely affected. Fed. R.Crim.P. 12(e).

The Government concedes that it did not ask the District Court to withhold swearing the jury until the motion to exclude evidence was decided nor did the Government call Rule 12(e) to the attention of the District Court. Government’s counsel has candidly conceded that in failing to ask the District Court not to swear the jury until after the pending motion was decided, the Government “made a mistake.” Tr. at 7. It was a mistake that is not without consequences. Because the Government failed to timely invoke the procedural rule and the statutory provision designed for the specific situation in which it found itself, the Government seeks to turn § 3731 upside down, inside out, and have this court become the first to call the exclusion of evidence a dismissal of the indictment, just so it can appeal.

We cannot allow the Government’s concern about the impact of an order in a particular case to lead us to overlook the distinction between the first and second paragraphs of § 3731. The distinction between an evidentiary ruling and a dismissal exists and has meaning. The District Court’s ruling was no more than “a decision or order ... suppressing or excluding evidence ... in a criminal proceeding” that occurred “after the defendant[s] ha[d] been put in jeopardy,” and thus cannot be appealed. 18 U.S.C. § 3731.⁶

The Government conceded at oral argument that it knew of no other case in which a court has interpreted an evidentia-ry ruling as a dismissal of an indictment. Tr. at 29. Yet it has maintained before us that the present case is not unique but rather a “repetition” of United States v. Maker, 751 F.2d 614 (3d Cir.1984), which it asserts should dictate this court’s analysis. Maker is of no help, and indeed is irrelevant on the principal issue before us because, unlike the present case, that case came to this court after the district court had dismissed the indictment. Id. at 615. Therefore, in Maker, we did not have occasion to address the circumstances in which an evidentiary ruling can amount to the functional equivalent of a dismissal of an indictment falling within the first paragraph of § 3731.

In Maker, two defendants were indicted on eighteen counts of mail fraud and charged with a single scheme to defraud three different insurance companies arising out of two car accidents. After jeopardy attached, the district court ordered separate trials with respect to each accident and also for each defendant. Id. at 615. Then, relying on United States v. Camiel, 689 F.2d 31 (3d Cir.1982), it dismissed the indictment because the allegations in the indictment did not support a single scheme to defraud but rather two separate schemes, one arising from each accident. Maker, 751 F.2d at 615. However, the court denied the defendants’ motion for a judgment of acquittal. Id. at 620. The government appealed.

The Maker defendants challenged this court’s jurisdiction under § 3731, not because there was no dismissal but on the ground the appeal was barred by the Double Jeopardy Clause. We concluded that the district court had made a legal determination, not a factual one, so that the Double Jeopardy Clause was not implicated. Id. at 624. On review of the merits of the government’s appeal, we concluded that the indictment charged only one scheme to defraud, and that the district court erred in dismissing the indictment. Id. at 626.

The only questions at issue in Maker, whether the court implicitly made a factual determination in dismissing the indictment and whether the indictment charged one or two schemes to defraud, are not contested in this case. Maker’s analysis of the parameters of the scheme to defraud might prove useful were we to reach that issue here, but because the Maker decision does not pertain to the question when a court should recharacterize an evidentiary ruling as a dismissal of an indictment, that case does not support the Government’s position.

The Government further relies on this court’s decision in Camiel, but that is of no help either. In Camiel, following a jury verdict of guilty, the district court found that the indictment improperly charged a single fraudulent scheme, and accоrdingly entered a directed verdict of acquittal. The government argued that the proper remedy for the defective indictment was retrial and severance of the three defendants’ trials, rather than acquittal. We rejected that view because a court may not correct an indictment, even if flawed. Camiel, 689 F.2d at 39-40. Accordingly, we determined that “a new trial with the possibility of severance [was] not the answer,” and thus affirmed the judgment of acquittal. Id. at 40.

As with Maker, Camiel lends support to the conclusion that the District Court in this case erred in attempting to compartmentalize the evidence. See, e.g., Maker, 751 F.2d at 625 (“Camiel ... teaches that in evaluating whether one scheme has been alleged or proved, a court should focus on the identity of the actors and the overall purpose of their conduct.”). But the Camiel court never questioned whether it lacked jurisdiction to hear the government’s appeal under § 3731. Thus, Cam-iel does not establish, or even intimate, that a district court order restricting the admission of evidence, even one that amounts to an amendment of an indictment, effectively constitutes a dismissal of that same indictment. The irrelevance of Maker and Camiel, the legal props on which the Government relies, demonstrates the weakness of the Government’s argument.

More pertinent to the case before us is the Supreme Court’s opinion in Sanabria v. United States, 437 U.S. 54, 98 S.Ct. 2170, 57 L.Ed.2d 43 (1978). The first sentence of the opinion characterizes “the issue presented [as] whether the United States may appeal in a criminal case from a midtrial ruling resulting in the exclusion of certain evidence and from a subsequently entered judgment of acquittal.” Id. at 56, 98 S.Ct. 2170. The indictment in San-abria charged several dеfendants with one count of conducting an illegal gambling business in violation of federal law, a business that involved two different types of conduct — numbers betting and horse betting, both allegedly in violation of Massachusetts law. Under the federal statute defendants were charged with violating, 18 U.S.C. § 1955 (1976), the gambling business must have been in violation of the law of the state where the business was located. At the close of the defendants’ case at trial, the court excluded evidence of numbers betting because the state statute referenced in the indictment addressed only horse betting and not numbers betting, but the court did not strike or amend any language in the indictment. The court then granted the motion for judgment of acquittal filed by Sanabria, one of the defendants, because there was no evidence linking him to the horse-betting activities. Id. at 59, 98 S.Ct. 2170.

Following the conviction of the other defendants, the government appealed from the district court’s order excluding evidence of the numbers-betting activities and entering a judgment of acquittal. The government recognized it could not appeal the ruling regarding Sanabria’s involvement with horse betting, as that aspect of the acquittal was based on the insufficiency of the evidence, but it sought a new trial on the aspect of the charge involving numbers betting. The court of appeals interpreted the district court’s evidentiary ruling to be a “dismissal” of the numbers-betting “charge,” thus finding it had jurisdiction under § 3731 to consider the government’s claim that the “dismissal” was in error. Id. at 61, 98 S.Ct. 2170. Reviewing the merits, the court of appeals concluded that the “dismissal” was in error and ordered a new trial. The Supreme Court granted cеrtiorari and reversed.

' The order granting certiorari was limited to the question whether “the Government’s appeal was authorized by statute and not barred by the Double Jeopardy Clause.” Id. at 62 n. 13, 98 S.Ct. 2170. However, the Court did not thereafter discuss the statute on which the court of appeals had relied for its jurisdiction. It explained that it had previously observed that “the jurisdictional statute authorizing Government appeals, 18 U.S.C. § 3731 (1976), was ‘intended to remove all statutory barriers’ to appeals from orders terminating prosecutions. United States v. Martin Linen Supply Co., 430 U.S. 564, 568, 97 S.Ct. 1349, 51 L.Ed.2d 642 (1977), quoting United States v. Wilson, 420 U.S. 332, 337, 95 S.Ct. 1013, 43 L.Ed.2d 232 (1975).” Id. at 63 n. 16, 98 S.Ct. 2170 (quotation omitted). It therefore turned immediately to the constitutional issue, which was whether the Double Jeopardy Clause barred the appeal.

However, in deciding the double jeopardy issue, i.e., whether the horse-betting and numbers-betting activities were part of the “same offense” so that the acquittal for insufficient evidence as to Sanabria’s participation in horse betting barred the new trial ordered for Sanabria on the numbers-betting conduct, the Supreme Court first considered the nature of the district court’s order with respect to the numbers-betting aspect of the charge. It did so to ascertain whether “the numbers theory was dismissed from the count before the judgment of acquittal was entered,” id. at 65, 98 S.Ct. 2170, and therefore whether Sanabria was acquitted of numbers betting, which would have entailed a resolution of factual issues precluding a second trial for the same offense.⁷ The government argued that the district court’s order was a “dismissal” of a discrete portion of the count in which the two types of gambling were included. The Supreme Court was “not persuaded that it is correct to characterize the trial court’s, action as a ‘dismissal’ of a discrete portion of the count.” Id. at 66, 98 S.Ct. 2170. It continued, “[w]hile form is not to be exalted over substance ..., neither is it appropriate entirely to ignore the form of order entered by the trial court.” Id.

The Court concluded that the district court’s order simply constituted “an evi-dentiary ruling based on its interpretation of the indictment,” and not a dismissal. Id. at 68, 98 S.Ct. 2170. As the Court noted,

[N]ot every erroneous interpretation of an indictment for purposes of deciding what evidence is admissible can be regarded as a ‘dismissal.’ Here the District Court' did not find that the count failed to charge a necessary element of the offense; rather, it found the indictment’s description of the offense too narrow to warrant the admission of certain evidence. To this extent, we believe the ruling below is properly to be characterized as an erroneous evidentiary ruling.

Id. (citation omitted).

The Supreme Court concluded that the evidentiary ruling on the numbers charge, albeit erroneous, “led to an acquittal for insufficient evidence,” id., and held that the Double Jeopardy Clause would be offended by a new trial relating to the numbers charge, as that charge was part of the “same offense” as the horse-betting charge on which the • defendant had been acquitted. Id. at 69, 98 S.Ct. 2170.

The Government tries to minimize the significance of Sanabria by limiting it to cases where the trial court entered a judgment of acquittal. However, Sanabria’s analysis of an evidentiary ruling based on an interpretation of an indictment as distinguished from a dismissal of that indictment is applicable here.

The district court in Sanabria “found the indictment’s description of the offense too narrow to warrant the admission of certain evidence.” Id. at 68, 98 S.Ct. 2170. Thé District Court in this case made a similar ruling. Accordingly, the District Court’s exclusionary ruling in this case did not constitute a dismissal of one or more counts of the indictment within the scope of § 3731, even construing that statute liberally, as the statute requires. Moreover, the Government does not argue that the District Court’s ruling dismissed “a portion of a count,” from which order an appeal to this court would lie under § 3731 when “the dismissed portion of the count constitutes an independent ground of criminal liability.” United States v. Serafini, 167 F.3d 8Í2, 814 (3d Cir.1999). That argument would have been meritless, as it is plain that the excluded evidence included no alleged mailings and thus could provide no independent ground of liability.

Because the District Court’s evidentiary ruling did not constitute a dismissal of the indictment, we reject the Government’s argument that this court has jurisdiction to hear its appeal under § 3731.⁸

We need not rest our disposition only on our interpretation of the meaning of “dismissal” in § 3731. Even if we were to limit our consideration to § 3731 and assume that the order of the District Court were a dismissal of the indictment (which we have held it was not), the first paragraph of § 3731 permits the Government to retry Defendants only if the Double Jeopardy Clause is not a bar to further prosecution. The Government’s opening brief conceded this in its section headed “There Is No Double Jeopardy Bar to a Retrial.” Br. of Appellant at 30.

Moreover, there is a plausible argument that the Supreme Court’s statement in Wilson cited in Sanabria, 437 U.S. at 63 n. 16, 98 S.Ct. 2170, that the jurisdictional statute was “intended to remove all statutory barriers to Government appeals and to allow appeals whenever the Constitution would permit,” Wilson, 420 U.S. at 337, 95 S.Ct. 1013, requires us to move directly to the Constitution in determining our jurisdiction over appeals from orders terminating prosecutions.⁹ There is legislative history to support the intent described in Wilson. See, e.g., id. at 338-39, 95 S.Ct. 1013. Under that theory, our interpretation of § 3731 would not be dispositive, notwithstanding the Government’s reliance upon it as ‍‌‌‌​​​​​‌‌​​​​‌‌‌​​​​‌​‌​​​‌​‌‌‌​‌​‌​​‌‌​​‌​‌‌‌‌‍a basis for our jurisdiction and we turn to the Double Jeopardy Clause for consideration of its effect on our jurisdiction.¹⁰

2. Double Jeopardy

The first relevant question is whether Defendants were placed in jeopardy. As discussed above, all parties agree that the jury had been sworn and the trial had been started. Defendants were, therefore, placed in jeopardy. This does not end our inquiry. As we observed in Maker,

[although the Supreme Court has explicitly held that “[t]he federal rule that jeopardy attaches when the jury is empaneled and sworn is an integral part of the constitutional guarantee against double jeopardy,” Crist v. Bretz, 437 U.S. 28, 38, 98 S.Ct. 2156, 57 L.Ed.2d 24 (1978), this proposition ... is the beginning rather than the end of our analysis.

751 F.2d at 620 n. 22.

In Sanabria, the Court summarized the “limited circumstances” when a second trial on the same offense is constitutionally permissible. 437 U.S. at 63, 98 S.Ct. 2170. It stated,

A new trial is permitted, e.g., where the defendant successfully appeals his conviction, United States v. Ball, 163 U.S. 662, 672, 16 S.Ct. 1192, 41 L.Ed. 300 (1896); where a mistrial is declared for a “manifest necessity,” Wade v. Hunter, 336 U.S. 684, 69 S.Ct. 834, 93 L.Ed. 974 (1949); where the defendant requests a mistrial in the absence of prosecutorial or judicial overreaching, United States v. Dinitz, 424 U.S. 600, 96 S.Ct. 1075, 47 L.Ed.2d 267 (1976); or where an indictment is dismissed at the defendant’s request in circumstances functionally equivalent to a mistrial, Lee v. United States, 432 U.S. 23, 29-30, 97 S.Ct. 2141, 53 L.Ed.2d 80 (1977). See also Jeffers v. United States, 432 U.S. 137, 97 S.Ct. 2207, 53 L.Ed.2d 168 (1977).

Id. at 63 n. 15, 98 S.Ct. 2170.

The situations listed in the passage quoted above from Sanabria can be characterized as situations in which the defendant consented to or requested the termination. This is not such a situation, as illustrated by this court’s precedents.

In Love v. Morton, 112 F.3d 131, 136-37 (3d Cir.1997), towards the close of the first day of Love’s criminal trial, the trial judge learned that his mother-in-law had died unexpectedly. Without the explicit consent of the parties, the judge declared a mistrial and left the courthouse. Id. at 134-35. The next day, a new trial began with a new jury and a new judge. Love’s request for dismissal on double jeopardy grounds was overruled by the new judge, and he was thereafter convicted. In considering Love’s habeas petition, we held that “[t]o demonstrate manifest necessity, the state must show that under the circumstances the trial judge ‘had no alternative to the declaration of a mistrial.’ The trial judge must consider and exhaust all other possibilities.” Id. at 137 (quoting United States v. McKoy, 591 F.2d 218, 222 (3d Cir.1979)). Applying this principle, we held that there was no manifest necessity for declaring a mistrial because there were other possibilities available. For example, the trial could have been continued with a replacement judge presiding after the new judge reviewed the trial transcript. Further, we held that under the circumstances, Love’s consent to the termination could not be inferred from his failure to object to it. Id. at 138-39. We therefore concluded that the Double Jeopardy Clause prohibited Love’s second trial. See also McKoy, 591 F.2d at 222-23 (finding no manifest necessity for declaring mistrial and, hence, holding further prosecution was prohibited by double jeopardy).

In the present case, there were also clear alternatives to termination. Before the District Court dismissed the jury, the Government suggested that the trial be stayed briefly without discharging the jury during the pendency of an expedited appeal. Defendants suggested that the trial could continue unless and until this court granted a stay. Alternatively, as we noted previously, had the Government raised and the District Court follоwed Federal Rule of Criminal Procedure 12(e) by ruling on the pending motions before the trial began, an appeal would have been feasible without any attachment of jeopardy. In the present case, as in Love, there was no manifest necessity to terminate the trial.

In light of the lack of manifest necessity, we next consider whether Defendants in effect consented to the termination of the trial. The Government argues that the termination was prompted by Defendants’ motion, and thus, there is no double jeopardy bar to a remand for a new trial on the theory that Defendants consented to the termination.

We observed in Love that “[mjistrials declared with the defendants’ consent do not bar a later prosecution,” and cited United States v. Dinitz, 424 U.S. 600, 607, 96 S.Ct. 1075, 47 L.Ed.2d 267 (1976). Love, 112 F.3d at 133. In Dinitz, the trial judge dismissed Dinitz’s counsel during his opening statement for repeated misconduct. The trial judge asked co-counsel, who had not discussed the case with any of the witnesses, to proceed with trial. Dinitz expressed desire that his original counsel be permitted to defend him. The trial judge offered Dinitz the choice of (a) continuing the trial with co-counsel, (b) recessing the trial during an attempt at interlocutory review of the dismissal of original counsel, or (c) declaring a mistrial to allow Dinitz to obtain new counsel. Dinitz chose a mistrial. Before his subsequent trial, Dinitz argued that the Double Jeopardy Clause barred his further prosecution. The Supreme Court held that when a defendant requests a mistrial, even in response to prosecutorial or judicial error, double jeopardy does not bar retrial, Din-itz, 424 U.S. at 606-612, 96 S.Ct. 1075, unless the error that prompted it was “ ‘bad-faith conduct by judge or prosecutor.’ ” Id. at 611, 96 S.Ct. 1075 (quoting United States v. Jorn, 400 U.S. 470, 485, 91 S.Ct. 547, 27 L.Ed.2d 543 (1971) (plurality opinion)).

In Love, we held that defendant’s failure to object to the mistrial did not constitute express consent or implied consent because he did not have a meaningful opportunity to object to the mistrial. Because there was no consent to the termination, the Double Jeopardy Clause barred a new trial. Love, 112 F.3d at 138-39.

In Lee v. United States, 432 U.S. 23, 97 S.Ct. 2141, 53 L.Ed.2d 80 (1977), defense counsel, immediately before the start of a trial for theft, moved to dismiss the information for failure to include allegations of knowledge and intent. The trial court denied the motion but said that it would be reconsidered. At the end of the trial, the court granted the defendant’s motion to dismiss for failure to charge knowledge and intent, even though the trial judge was convinced that the defendant was guilty. The prosecution filed a corrected information and, after the second trial, the defendant was convicted. The Supreme Court held that the Double Jeopardy Clause did not bar the second prosecution because the first prosecution was terminated at the defendant’s request. See also United States v. Scott, 437 U.S. 82, 98-99, 98 S.Ct. 2187, 57 L.Ed.2d 65 (1978) (holding that “defendant, by deliberately choosing to seek termination of the proceedings against him on a basis unrelated to factual guilt or innocence ... suffers no injury cognizable under the Double Jeopardy Clause if the Government is permitted to appeal from such a ruling of the trial court in favor of the defendant”).

In United States v. Kennings, 861 F.2d 381 (3d Cir.1988), we held that the fact that the district court rather than the defendant first suggested dismissal does not bar appeal so long as the defendant agreed to dismissal. Id. at 383-86. We found that the defendant’s consent and active support of dismissal undercut the argument that the defendant should only be tried by the first jury empanelled to hear his case. Id. at 385.

Unlike the situations in Lee, Scott, and Kennings, Dеfendants in the present case did not move for a mistrial, for a termination of the trial, for the discharge of jury, or for the dismissal of the indictment or any count of it. Moreover, they objected to the District Court’s decision to terminate the trial and asked the court to continue with the trial.¹¹ Although Defendants, through their Motion to Redact, did set in motion the chain of events that led to termination of their trial, that alone is not enough to infer their consent to the resulting termination of the trial.

In Kennings, we contrasted the situation before us with that the court faced in United States v. Dahlstrum, 655 F.2d 971 (9th Cir.1981)-. In Kennings, the defendant “encouraged the dismissal solely on the grounds that the statute was inapplicable” and “argued that the appropriate action by the court was immediate, midtrial dismissal.” Kennings, 861 F.2d at 385 n. 6. We noted that in Dahlstrum, on the other hand, “the [trial] court dismissed the case sua sponte” and “the defendant passively acquiesced .... in response to the judge’s statements that he ‘did not believe the government had proved its case,’ [and t]hus, acquiescence could not have been read to mean that retrial would be acceptable to the defendant.” Id. (quoting Dahlstrum, 655 F.2d at 975). We found that these differences explained why the Double Jeopardy Clause did not bar the appeal in Kennings but it did bar the appeal in Dahlstrum. Id.

Our analysis in Kennings demonstrates that a defendant who merely sets in motion a series of events that leads to the termination of his or her trial is protected from being retried by the Double Jeopardy Clause. Defendants in the present case did no more than set in motion events that led, over their objections, to the dismissal of their trial. The real cause of the trial being interrupted was the Government’s desire to appeal the exclusion of some of its evidence. Because Defendants did not consent to the termination, even if the District Court’s decision to terminate the trial did constitute a dismissal, the Double Jeopardy Clause bars this court from having jurisdiction to hear the Government’s appeal.

3. The Dissent

The dissent’s interesting discourse on the important principle of separation of powers overlooks that the holding of the majority in this case is that the court lacks jurisdiction to hear the Government’s appeal. In each of the cases cited by the dissent on pages 246 through 248, the court’s appellate jurisdiction stood on firm statutory ground. Thus, in United States v. Giannattasio, 979 F.2d 98, 100 (7th Cir.1992), the court reviewed the dismissal of an-indictment; in United States v. Zabawa, 39 F.3d 279, 283 (10th Cir.1994), and United States v. Nakashian, 820 F.2d 549, 550 (2d Cir.1987), each court deemed the district court’s requirement that the government select a limited number of counts on which it would proceed to be a dismissal of one or more counts, and thus expressly covered by 18 U.S.C. § 3731; in United States v. Abdelhaq, 246 F.3d 990 (7th Cir. 2001), the defendant was appealing his conviction, and thus there was no question of jurisdiction under § 3731; in United States v. Marubeni America Corp., 611 F.2d 763, 764-65 (9th Cir.1980), the court (following a similar decision by the Second Circuit) held that dismissal of a substantial part of a count of an indictment fell within § 3731; and in United States v. Woolard, 981 F.2d 756, 757 (5th Cir.1993), the court held it had jurisdiction over the order of the district court striking the death penalty as a possible sentence because it effectively “removed a discrete basis of criminal liability.” The District Court’s order in this case is not comparable to any of those cases. The court neither struck nor dismissed a count or a portion of a count. Because the references to the manual billing were included in the indictment only as part of the history of Defendants’ scheme, the Government never included them in any count, so the court’s order did not remove any discrete basis for criminal liability.

The dissent concludes its discussion of this portion of its opinion with the sentence that “the law governing the separation of powers between the executive and judicial branches is sufficient to demonstrate that the District Court’s actions in this case must have resulted in a dismissal.” Dis. op. at 250 (emphasis added). The dissent’s non sequitur is not an effective substitute for an order or action by the District Court that does indeed fall within 18 U.S.C. § 3731. Had its separation of powers argument been of any force, the Government, which has vigorously pursued this appeal, would be likely to have included it. But it didn’t.

When the dissent turns to the issue of the order the District Court did enter, it appears to misread it, as it fails to acknowledge that, whatever the effect of the order, it did nothing more than exclude evidence of the manual scheme (which, as we previously acknowledged, was erroneous). Because no count in the indictment charged the manual billing as an offense, the District Court’s exclusion of evidence of the manual billing cannot be viewed as a dismissal of any count or a part thereof. It follows that the points made by the dissent are not persuasive.

III.

CONCLUSION

Despite the various mistakes that the District Court made, this court lacks jurisdiction to hear the Government’s appeal under § 3731 both because the District Court did not dismiss the indictment and because a retrial is barred by the Double Jeopardy Clause. We will therefore dismiss the Government’s appeal.

1 . The court’s order and memorandum were dated September 25 but were issued and filed September 26.

2 . We note that the misunderstanding was likely initiated by Defendants' mischaracteri-zation of the indictment in their Motion to Redact.

3 . Defendants' Motion to Redact noted that any loss to Hudson's clients that resulted from the manual billing scheme occurred before the statute of limitations. That was not an issue because mailings that fall outside the statute of limitations can be considered as evidence to prove later fraud that was within the statute of limitations. See, e.g., United States v. Ashdown, 509 F.2d 793, 797-98 (5th Cir.1975). Moreover, a mailing that is within the statute of limitations can impose criminal liability for conduct that was part of the same scheme but that was initiated outside the statute of limitations. See, e.g., United States v. Crossley, 224 F.3d 847, 857-59 (6th Cir.2000). Since a mailing is an element of mail fraud, "the ,offense of mail fraud is completed and the statute of limitations begins to run on the date on which the defendant, depending on the specific use of the mails charged in the indictment, `places,' `deposits,' `causes to be deposited,' `takes,' or `receives' mail, or `knowingly causes' mail `to be delivered' as part of the execution of a scheme to defraud." Id. at 859; see also United States v. Bach, 172 F.3d 520, 521-22 (7th Cir.1999); United States v. Pemberton, 121 F.3d 1157, 1163-64 (8th Cir.1997); United States v. Eisen, 974 F.2d 246, 263-64 (2d Cir.1992).

4 . The court’s analysis in this respect is consistent with our holding that Rule 403’s balancing test is one prong of a four-part test for admitting evidence under Rule 404(b). See, e.g., Becker v. ARCO Chem. Co., 207 F.3d 176, 189 (3d Cir.2000); United States v. Himelwright, 42 F.3d 777, 781 (3d Cir.1994).

5 . The District Court's initial order precluded the Government's use of evidence through February 1994. The District Court's subsequent order, in response to the Government's motion for reconsideration, modified its initial ruling and precluded evidence up until the end of 1992. Sept. 28 order at 4. The District Court did not explain its modification. The dissent, using its telepathic powers to read between the lines, ascribes the District Court’s barring of evidencе "during the period 1989 to 1992” to "merely a scrivener's error.” Dis. op. at 253. Although it does not bear on our disposition, we are not convinced by the dissent's attempt at mind reading.

6 . Further, we are not convinced that the ruling would have the draconian effect that the Government posits. If the purpose of the evidence was to show the intent to defraud over a lengthy period of time, a skilled Government prosecutor may very well have been able to question witnesses about that issue and elicit that background ‍‌‌‌​​​​​‌‌​​​​‌‌‌​​​​‌​‌​​​‌​‌‌‌​‌​‌​​‌‌​​‌​‌‌‌‌‍without directly violating the court's evidentiary ruling. As the trial unfolded, the District Court may have modified what may have appeared at first to be its rigid prohibition. Moreover, the Government did not attempt to elicit testimony concerning what may have been an additional year period given by the District Court, see supra note 5, about which to question the witnesses.

7 . The Double Jeopardy Clause of the Fifth Amendment provides that no person shall "be subject for the same offense to be twice put in jeopardy of life or limb.” U.S. Const, amend. V.

8 . Because we believe that the District Court’s ruling did not amount to a dismissal at all, we need not consider further whether the District Court's order amounted to a dismissal of the indictment as it pertained to defendant Harry Gangloff in particular, inasmuch as he appears to have been implicated only in the inflated computer billing scheme.

9 . The Court in Sanabria recognized some undefined limitation other than the obvious Double Jeopardy Clause, as it stated in another context, without explanation, "the Government is not authorized to appeal from all adverse rulings in criminal cases." 437 U.S. at 67 n. 21, 98 S.Ct. 2170.

10 . In Sanabria, the Supreme Court, after concluding that further prosecution was barred because it viewed the district court’s order to have been a judgment of acquittal, nevertheless proceeded to discuss whether, even if the government were correct that the order was merely a dismissal, further prosecution of Sanabria was barred by the Double Jeopardy Clause. 437 U.S. at 69, 98 S.Ct. 2170. Although there was no acquittal in this case, all the parties acknowledge the relevance of the Double Jeopardy Clause to our jurisdiction.

11 . Defendants did, however, ask for an acquittal, but their motion for acquittal was not granted. If Defendants had been acquitted, then the Double Jeopardy Clause would bar any further prosecution. Defendants do not claim that they were acquitted and do not base their argument that double jeopardy bars their further prosecution on this ground.

---

Dissenting Opinion

COWEN, Circuit Judge,

dissenting.

Under our system of government, judges do not draft indictments. The power to decide whether charges will be brought, which violations to allege, and what the scope of those charges will be belongs to the executive branch and grand jury, not to the judiciary. United States v. Armstrong, 517 U.S. 456, 464, 116 S.Ct. 1480, 134 L.Ed.2d 687 (1996); Wayte v. United States, 470 U.S. 598, 607-608, 105 S.Ct. 1524, 84 L.Ed.2d 547 (1985); Heckler v. Chaney, 470 U.S. 821, 832, 105 S.Ct. 1649, 84 L.Ed.2d 714 (1985); United States v. Goodwin, 457 U.S. 368, 382, 102 S.Ct. 2485, 73 L.Ed.2d 74 (1982); Bordenkircher v. Hayes, 434 U.S, 357, 364, 98 S.Ct. 663, 54 L.Ed.2d 604 (1978); Oyler v. Boles, 368 U.S. 448, 456, 82 S.Ct. 501, 7 L.Ed.2d 446 (1962); United States v. Chemical Foundation, Inc., 272 U.S. 1, 14-15, 47 S.Ct. 1, 71 L.Ed. 131 (1926). See also, United States v. Williams, 504 U.S. 36, 47-50, 112 S.Ct. 1735, 118 L.Ed.2d 352 (1992). This separation of powers is fundamental to our Constitution’s design. “Time and again we have reaffirmed the importance in our constitutional scheme of the separation of governmental powers into the three coordinate branches.” Morrison v. Olson, 487 U.S. 654, 693, 108 S.Ct. 2597, 101 L.Ed,2d 569 (1988) (citations omitted). Just as the legislative branch cannot alter the outcome of a case that the judiciary has decided, Plant v. Spendthrift Farm, Inc., 514 U.S. 211, 115 S.Ct. 1447, 131 L.Ed.2d 328 (1995), so too the judicial branch cannot assume the role of writing indictments that initiate cases.

Undermining the important principle of separated pоwers, the majority today grants trial courts substantial powers to reshape indictments brought by the executive and approved by a grand jury. The majority concludes that the District Court in this case was permitted — without dismissing the indictment — to cut in half the fraudulent scheme alleged in each of the six counts of mail fraud brought in the indictment, and to do so based on what the majority admits are pure errors of law. This erroneous alteration of the indictment changed allegations addressing an essential element of the offense of mail fraud, the fraudulent scheme, see United, States v. Lane, 474 U.S. 438, 451, 106 S.Ct. 725, 88 L.Ed.2d 814 (1986) (quoting Kann v. United States, 323 U.S. 88, 94, 65 S.Ct. 148, 89 L.Ed. 88 (1944)); United States v. Sturm, 671 F.2d 749, 751 (3d Cir.1982), and the evidence encompassed by the excluded half of the fraudulent scheme is crucial, as the government very plausibly insists, to proving any claim of fraud.

The upshot of the majority’s rule is that district courts in this circuit now possess unchanneled discretion to rewrite indictments without having the effect of dismissing them, even though the changes to the offense elements are substantial enough to prevent the government from proving the defendants’ crimes. At most the only limitation that can be extracted from the majority’s opinion is that perhaps the number of counts and type of offense must be maintained, together with some unspecified facts about some unspecified elements of the offense. See Majority Op. at 235-36. Thus, it would seem that a district court could, without dismissing the indictment, force the government to proceed on a lesser included offense that was never expressly charged in the indictment, even though the greater offense was supported by the law and evidence. Or a district court could alter the type of fraudulent scheme alleged by changing, for examplе, a scheme alleging fraudulent over billing to one charging fraudulent billing for services that were never performed, even though the latter was unsupported by the evidence. Or a district court could decide to reduce by a third the total losses alleged from a fraudulent scheme. If the majority’s rule is limited only to changes to certain categories of offense elements, or even more specifically to just the duration of a fraudulent scheme in mail-fraud prosecutions, the majority certainly does not say so, nor does its opinion offer any explanation of how such an anomalous special limitation could be justified.

In describing its standard, the majority says, quoting the appellees’ brief, that “hypothetically ... an order having the form of excluding evidence ... so sweeping that it precluded any and all proof of a count of an indictment ... might be characterized by an appellate court as a dismissal under § 3731....” Id. at 235. But the majority concludes that this standard was not met in this case because, “[ajfter all, the indictment alleged mail fraud, all six mailings that constituted the acts of mail fraud were made after February 1994 and arose out of the computer generated bills, and those counts remained intact following the order the Government seeks to appeal.” Id. at 235. In other words, it is of no consequence that the District Court cut in half the scope of one element of the offense, the fraudulent scheme, and did so in a manner that made proof of the remaining, narrower criminal offense considerably more difficult if not impossible. Conspicuously, the majority does nоt cite any legal authority supporting its sweeping standard for altering indictments without dismissing them.

And there is legal authority that concludes that the separation of powers prevents district courts from reaching so far into the executive branch’s role of selecting the type and scope of criminal charges. Cases like Armstrong, Wayte, Heckler, Goodwin, Bordenkircher, Oyler, and Chemical Foundation, Inc. establish that the task of selecting charges belongs to the executive. Similarly, several courts have concluded that it transgresses the separation of powers for a district court to force the government to proceed on only parts of an indictment. See, e.g., United States v. Zabawa, 39 F.3d 279 (10th Cir.1994); United States v. Giannattasio, 979 F.2d 98 (7th Cir.1992). These cases embody the fundamental principle that unelected federal judges are not vested under the Constitution with the responsibility to exercise prosecutorial discretion and “take Care that the Laws be faithfully executed.” Art. II, Sec. 3. A more democratically responsive branch must be left to make the difficult choices about whom to charge, which crimes to pursue, and how numerous and serious the charges will be. District courts of course retain the power to dismiss legally defective indictments, but they cannot engage in a freewheeling process of rewriting indictments and offering defendants lighter or different charges.

In Zabawa after the district court required the government to proceed to trial on a limited portion of the indictment, the government objected and sought to appeal because it believed that reducing the counts prevented the government from “presenting the necessary evidence to convict all defendants.” Zabaiva, 39 F.3d at 284. Although the district court did not dismiss those counts that the government was not permitted to pursue at trial, the Tenth Circuit concluded that “the district court’s order — -forcing a choice of counts without a formal dismissal of the other counts — is not significantly different from ordering a formal dismissal without prejudice.” Id. at 283. The Tenth Circuit continued that the district court’s attempt to limit the counts presented would violate the separation of powers: “Because the district court’s ruling forces the government to abandon, at least temporarily, the prosecution of separate crimes it has charged against defendants who are scheduled to be tried, we believe the ruling goes beyond those subject to the court’s discretionary control and impinges upon the separation of powers. Prosecutorial discretion is a function of the executive branch, not the judiciary.” Id. at 284.

In Giannattasio the Seventh Circuit similarly held that a district court violated the separation of powers when it forced the government to select five of fifteen counts to present at trial. Judge Posner’s opinion explained:

A judge in our system does not have the authority to tell prosecutors which crimes to prosecute or when to prosecute them. Prosecutorial discretion resides in the executive, not in the judicial branch, and that discretion, though subject of course to judicial review to protect constitutional rights, is not reviewable for a simple abuse of discretion. This principle is most often involved when the issue is whom to prosecute ... but it has equal force when the issue is which crimes of a given criminal to prosecute. If Dr. Giannattasio committed fifteen Medicаre frauds, a judge cannot tell the Justice Department to prosecute him for only five of the frauds, or to prosecute him for five now and the rest later, if necessary.... No rule authorizes the judge to sever offenses in an indictment because he believes that a trial of all the counts charged would clog his docket without yielding any offsetting benefit in the form of a greater likelihood of conviction or a more severe punishment.

Giannattasio, 979 F.2d at 100 (citations omitted). Cf. United States v. Abdelhaq, 246 F.3d 990, 992 (7th Cir.2001) (“The judge felt that the government didn’t need the extra counts. He may have been right. There would be no sentencing increment from conviction of the other charges if the government succeeded in convicting the defendant of the main charge.... And indeed, after she was convicted on that charge and sentenced to 21 years in prison, the government dismissed the counts that had been severed. But the decision on how many counts are needed to present an effective case is a managerial decision committed to the discretion of the prosecution.”); United States v. Woolard, 981 F.2d 756, 757 (5th Cir.1993) (finding jurisdiction for “orders that did not dismiss an entire count but altered it in a significant way from the grand jury’s charge.”); United States v. Nakashian, 820 F.2d 549, 550 (2d Cir.1987) (“an order compelling an election between counts is a ‘dismissal’ for Section 3731 purposes.”); United States v. Marubeni America Corp., 611 F.2d 763, 764-65 (9th Cir.1980) (finding jurisdiction for a government appeal of an order striking forfeiture allegations from RICO indictment).

Two other lines of authority confirm that the District Court should not be permitted to alter the scope of the fraudulent scheme alleged in the indictment. The first line of authority interprets the scope of Fed. R.Crim.P. 7(d), which allows a defendant to request that a court remove surplusage from the charging indictment. Under the rule, the “scope of a district court’s discretion to strike material from an indictment is narrow.” United States v. Oakar, 111 F.3d 146, 157 (D.C.Cir.1997) (citing United States v. Jordan, 626 F.2d 928, 931 n. 1 (D.C.Cir.1980)); “Words descriptive of what is legally essential to the charge in the indictment cannot be stricken as sur-plusage.” Wright, Federal Practice and Procedure § 127, at 635. See also United States v. Collins, 920 F.2d 619, 631 (10th Cir.1990); United States v. Figueroa, 900 F.2d 1211, 1218 (8th Cir.1990); United States v. Behenna, 552 F.2d 573, 576 (4th Cir.1977); United States v. Root, 366 F.2d 377, 381 (9th Cir.1966). “Material that can fairly be described as ‘surplus’ may only be stricken if it is irrelevant and prejudicial.” Oakar, 111 F.3d at 157. See also United States v. Rezaq, 134 F.3d 1121, 1134 (D.C.Cir.1998); United States v. Huppert, 917 F.2d 507, 511 (11th Cir.1990); United States v. Scarpa, 913 F.2d 993, 1013 (2d Cir.1990).

The second line of authority is based not on the separation of powers or on the scope of Fed.R.Crim.P. 7(d), but on the Fifth Amendment’s Grand Jury Clause limiting amendments to indictments. See Ex parte Bain, 121 U.S. 1, 7 S.Ct. 781, 30 L.Ed. 849 (1887). While Bain has been rejected and limited in a number of important respects, the Supreme Court recently noted that by limiting Bain, the Court “need not retreat” from the “settled proposition” in Bain that “an indictment may not be amended except by resubmission to the grand jury, unless the change is merely a matter of form.” United States v. Cotton, - U.S. -, -, 122 S.Ct. 1781, 1785, 152 L.Ed.2d 860 (2002) (quoting Russell v. United States, 369 U.S. 749, 770, 82 S.Ct. 1038, 8 L.Ed.2d 240 (1962)). In Bain the Supreme Court stated:

If it lies within the province of a court to change the charging part of an indictment to suit its own notions of what it ought to have been, or what the grand jury would probably have made it if their attention had been called to suggested changes, the great importance which the common law attaches to an indictment by a grand jury, as a prerequisite to a prisoner’s trial for a crime, and without which the Constitution says “no person shall be held to answer,” may be frittered away until its value is almost destroyed.

121 U.S. at 10, 7 S.Ct. 781.

The one limitation of Bain relevant here is the rule that a prosecutor may choose to withdraw a portion of the indictment that the evidence does not support and proceed on a lesser included offense without resubmitting the indictment to the grand jury. United States v. Miller, 471 U.S. 130, 105 S.Ct. 1811, 85 L.Ed.2d 99 (1985). The defendant in Miller had been charged with defrauding his insurer both by consenting to a burglary of Ms place of business and by lying to the insurer abоut the value of his losses. After the grand jury returned the indictment, the government moved to strike the part of the indictment alleging that the defendant had prior knowledge of the burglary, but the government correctly maintained that the remaining allegations of lying about the value of the loss still supported the indictment’s allegation of mail fraud. In finding no violation of Bain, the Supreme Court explained, “A part of the indictment unnecessary to and independent of the allegations of the offense proved may normally be treated as a ‘useless averment’ that ‘may be ignored.’ ” Miller, 471 U.S. at 136, 105 S.Ct. 1811 (quoting Ford v. United States, 273 U.S. 593, 602, 47 S.Ct. 531, 71 L.Ed. 793 (1927)).

In our case, however, the government has not sought to abandon any allegations or charges and pursue a separately alleged or lesser included offense, and the government has steadfastly maintained, quite plausibly, that the allegations excluded by the District Court are essential for proving the defendants’ guilt of any charge of mail fraud. The allegations can hardly be described as “unnecessary” and “independent” to the offense that the government wished to pursue, or even to the truncated offense that the District Court wanted the government to pursue. If there is any meamng to the screening function performed by a grand jury, it must be implicated when a district court amends an indictment in a fashion that excludes evidence that the government itself believes makes its ease difficult if not impossible to prove, and thus very likely did play an important role in the grand jury’s decision to indict.

One potential caveat presented by applying Bain’s prohibition on amending indictments is that the rule might be regarded as only a defendant’s right. The government may be the wrong party to complain about amendments that undermine the grand jury’s role in screemng charges. Cf. Rakas v. Illinois, 439 U.S. 128, 134, 99 S.Ct. 421, 58 L.Ed.2d 387 (1978) (limiting those who have standing to assert Fourth Amendment violations). See also Minnesota v. Carter, 525 U.S. 83, 119 S.Ct. 469, 142 L.Ed.2d 373 (1998); Rawlings v. Kentucky, 448 U.S. 98, 105, 100 S.Ct. 2556, 65 L.Ed.2d 633 (1980). And Bain’s rule can be waived. United States v. Cotton, 122 S.Ct. at 1786. Of course, the defendants have not waived their right to a grand jury indictment, and the government has steadily asserted its objections to the District Court’s rulings.

But more important, the grand jury through its investigatory powers serves a public role beyond protecting the defendant charged, and hence the rule against amending indictments may serve more than the defendant’s interests. “The investigation of crime by the grand jury implements a fundamental governmental role of securing the safety of the person and property of the citizen.” United States v. Calandra, 414 U.S. 338, 344, 94 S.Ct. 613, 38 L.Ed.2d 561 (1974) (quoting Branzburg v. Hayes, 408 U.S. 665, 700, 92 S.Ct. 2646, 33 L.Ed.2d 626 (1972)) (internal quotations and brackets omitted). The Supreme Court in Williams explained:

Rooted in long centuries of Anglo-American history, the grand jury is mentioned in the Bill of Rights, but not in the body of the Constitution. It has not been textually assigned, therefore, to any of the branches described in the first three Articles. It is a constitutional fixture in its own right. In fact the whole theory of its function is that it belongs to no branch of the institutional Government, serving as a kind of buffer or referee between the Government and the people.

504 U.S. at 47, 112 S.Ct. 1735 (citations and quotation marks omitted). While refusing in Williams to expand judicial oversight of the grand jury’s evidence-taking process, the Supreme Court added: “The grand jury’s functional independence from the Judicial Branch is evident both in the scope of its power to investigate criminal wrongdoing and in the manner ‍‌‌‌​​​​​‌‌​​​​‌‌‌​​​​‌​‌​​​‌​‌‌‌​‌​‌​​‌‌​​‌​‌‌‌‌‍in which that power is exercised. ‘Unlike a court, whose jurisdiction is predicated upon a specific case or controversy, the grand jury can investigate merely on suspicion that the law is being violated, or even because it wants assurance that it is not.’ ” Williams, 504 U.S. at 48, 112 S.Ct. 1735 (iquoting United States v. R. Enterprises, Inc., 498 U.S. 292, 297, 111 S.Ct. 722, 112 L.Ed.2d 795 (1991) and United States v. Morton Salt Co., 338 U.S. 632, 642-643, 70 S.Ct. 357, 94 L.Ed. 401 (1950)) (brackets omitted). See also In Re Grand Jury, 286 F.3d 153 (3d Cir.2002); In re Grand Jury, 223 F.3d 213 (3d Cir.2000). These considerations may militate against allowing only the defendant to invoke the rule against amending indictments. Chief Justice Marshall’s famous declaration that in a grand jury proceeding, “The public has the right to every man’s evidence,” United States v. Burr, 25 Fed.Cas. 38 (No. 14,692e) (CCVa.1807), emphasizes that body’s purpose of serving the broader public’s interests, and not just the defendant’s.

In the end, however, I believe that the law governing the separations of powers between the executive and judicial branches is sufficient to demonstrate that the District Court’s actions in this case must have resulted in a dismissal. For that reason, I need not reach whether the government can assert a violation of Bain’s rule against amending indictments.

The majority is obliged, however, to address this issue since the government has invoked the rule against amending indictments. But not only has the majority failed to address amendments adequately, it also insists that it may ignore the implications that its decision has for the separation of powers because the government never discussed, and hence waived, the issue.

I believe that by arguing against the amendment to the indictment, the government has not slept on its rights and has invoked rules designed to protect the separation of powers. But regardless, the threat posed to the separation of powers is substantial and cannot be waived any more than jurisdiction can be. As the Supreme Court has said on numerous occasions, the separation of powers prevents “encroachment or aggrandizement of one branch at the expense of the other. To the extent this structural principle is implicated in a givеn case, the parties cannot by consent cure the constitutional difficulty for the same reason that the parties by consent cannot confer on federal courts subject-matter jurisdiction beyond the limitations imposed by Article III, Section 2.” Commodity Futures Trading Comm. v. Schor, 478 U.S. 833, 850-51, 106 S.Ct. 3245, 92 L.Ed.2d 675 (1986) (citations omitted). See also Freytag v. Comm. of Internal Revenue, 501 U.S. 868, 879, 111 S.Ct. 2631, 115 L.Ed.2d 764 (1991). (“[T]he disruption to sound appellate process entailed by entertaining objections not raised below does not always overcome what Justice Harlan called ‘the strong interest of the federal judiciary in maintaining the constitutional plan of separation of powers.’ ”) The Court’s conclusion in Freytag about why a challenge under the Appointments Clause could not be waived applies equally here. “The structural interests protected by the Appointments Clause are not those of any one branch of Government but of the entire Republic.” Id. at 880, 111 S.Ct. 2631. Cf. Steel Co. v. Citizens for a Better Environment, 523 U.S. 83, 94, 118 S.Ct. 1003, 140 L.Ed.2d 210 (1998) (The doctrine of “hypothetical jurisdiction” that was once invoked by lower courts improperly “carries the courts beyond the bounds of authorized judicial action and thus offends fundamental principles of separation of powers.”).

The majority’s best effort to avoid the subject of impermissible amendments to indictments is to rely on the defendant’s argument that the District Court merely made an evidentiary ruling. Like the District Court in its second opinion responding to the government’s motion for reconsideration, the majority attempts to characterize the District Court’s decision as ■ just an evidentiary ruling that happened to be especially damaging to the government’s case. “In effect, the Government is claiming that a dismissal occurred simply because it lost an evidentia-ry ruling imрortant to its case.” Majority Op. at 235.

As a threshold matter, whether a district court’s ruling resulted in a dismissal turns on what the ruling accomplished and not whether the district court called it an evidentiary ruling rather than a dismissal. In evaluating when there was a dismissal and when the government’s appeal is barred by the Double Jeopardy Clause, the Supreme Court has repeatedly said, “The trial judge’s characterization of his own action cannot control.... ” Smalis v. Pennsylvania, 476 U.S. 140, 144 n. 5, 106 S.Ct. 1745, 90 L.Ed.2d 116 (1986) (quoting United States v. Scott, 437 U.S. 82, 96, 98 S.Ct. 2187, 57 L.Ed.2d 65 (1978) (internal brackets and quotation marks omitted)); Martin Linen Supply Co., 430 U.S. 564, 567 n. 4, 97 S.Ct. 1349, 51 L.Ed.2d 642 (1977); United States v. Maker, 751 F.2d 614, 621 n. 26 (3d Cir.1984); United States v. Ember, 726 F.2d 522, 524 (9th Cir.1984). The principle is analogous to the Supreme Court’s extension of Bain that even if a trial court does not expressly allow an amendment to an indictment, the court can constructively amend the indictment by allowing the government to rest on proof of an offense crucially different than the one alleged in the indictment. See Stirone v. United States, 361 U.S. 212, 217, 80 S.Ct. 270, 4 L.Ed.2d 252 (1960). In flatly declaring that there was no dismissal, and hence no jurisdiction under § 3731, the majority never acknowledges or discusses these cases and begs a central question in this case.

The fact that a district court’s label does not control is not of course an open-ended license for the government to claim that there was a dismissal of an indictment every' time the government loses an evi-dentiary ruling significant to its case. The Supreme Court’s rule was not intended to give the government in practical effect a broader right of interlocutory appeal than even ordinary civil litigants have. If the majority is motivated by a concern that the government not secure such a broad fight of appeal, then I am in agreement with their sentiment — but not with their estimation of what occurred in this case. To call the District Court’s ruling a merely evidentiary one is to misunderstand what transpired at trial and to jeopardize a constitutional value — the separation of judicial and executive functions — carrying as great importance as the concern with limiting government appeals.

To understand why this case does not simply involve a damaging evidentiary ruling, it is important to review both the District Court’s initial ruling, which granted the defendants’ motion to redact the indictment, as well as the Court’s second opinion, which denied the motion for reconsideration but purported to recast the ruling as an evidentiary one. In both opinions, the District Court made it abundantly clear that it believed that there were two schemes alleged in the indictment and that the earlier scheme should not properly be considered part of the indictment. The District Court was not limiting a way of proving allegations in the indictment, such as by excluding hearsay evidence of the manually altered bills. Rather, the Court was saying that the scheme alleged in the indictment was contrary to the law and could not be validly proved by any means.

When the District Court was discussing what evidence it was excluding, the Court said in its opinion after the motion for reconsideration:

The evidence that the Government pro- , poses is utterly unrelated to the offenses which are charged. Those prior bad acts have nothing to do with the computer programming. of bills, they have nothing to do with the California earthquake program, they are not temporally continuous; and they involve different insurance companies. The six counts involve automatic computer adjustment of bills. The legitimacy of that process is the issue in the case.

Supp.App. at 11 (emphasis added).

The government also points out that at trial the district judge said: “All six counts relate to acts the Government alleges were violated in 1994 and 1995. The Court concludes that acts relating to conduct other than that alleged in the indictment in those years constitute other acts that" do not demonstrate violations of the mail fraud statute.” App. at 176. Like the statement quoted from the District Court’s second opinion, this statement at trial also strongly supports the government’s contention that the District Court sought to preclude all evidence about the manually altered bills.

Moreover, when the government asked its first witness at trial about her observation that the company began having financial problems in 1989, the defense objected that the part of the indictment relevant to that period had been redacted. App. at 173-74. And the District Court sustained the objection, stating, “Anything she has to tell us about 1988 and 1989 is not relevant.” App. at 175. A fair reading of the record makes plain that what the District Court did was exclude categorically all evidence about the defendants’ conduct prior to the period when the gooser program was created and used:

Seizing on a clever but mistaken argument by the defendants, the majority suggests that despite all the passages quoted above, the District Court did not flatly prohibit all evidence about the earlier scheme of manually altered bills. See Majority Op. at 235 n. 5 and 237 n. 6. The defendants’ argument, endorsed by the majority, is that once the District Court switched so that its decision purported to rest on evidentiary grounds, the District Court struck a compromise, allowed one year’s worth of evidence about the manually altered bills, and for the other, earlier years still did not foreclоse all discussion of the defendants’ business. All of these points, the defendants maintain, indicate that the District Court did not prohibit the government from going forward with its prosecution of the single scheme covering the period described in the indictment, and hence the District Court’s ruling did not effectively dismiss the indictment.

The reason the defendants and the majority maintain that the District Court intended to permit evidence about one year’s worth of the manually altered bills, notwithstanding the Court’s emphatic language quoted above, is that in several places the Court’s opinion in response to the government’s motion for reconsideration said it was barring evidence “during the period 1989 to 1992.” SuppApp. at 12 (emphasis added). The references to “1992” appear to give the government one year of evidence about the manual bills, because, as the majority opinion notes,-the gooser program was not initiated until February 1994. By comparison, before the government’s motion for reconsideration, the Court’s first order referred to the period where the evidence would be excluded as covering 1989 to February 1994, which did correctly capture the end of the period when the indictment alleges that bills were manually altered.

There are a number of reasons fop believing that the District Court’s several references to 1992 were merely a scrivener’s error. First, the use of 1992 is incom sistent with the Court’s explanation of why it was ruling the way it did- — namely that it believed the manually altered bills could not properly be included in the fraudulent scheme. Given the passages quoted earlier, one of which comes directly from the Court’s second opinion, denying the government’s motion for reconsideration, the defendants’ and majority’s theory that the District Court meant to write “1992” leaves inexplicable the fact that the Court never discussed why it was purportedly having a change of heart midway through its second opinion and allowing roúghly one year of evidence about the manually altered bills. Why allow any of-the evidence, given the Court’s views? And why permit the roughly one year of evidence that the majority and defendants say the Court allowed? There was nothing particularly unique about that time span. Given the arbitrariness of that proposed line, it is reasonable to expect that the Court would give some explanation for this change from the earlier opinion and from other language in its second opinion. This is especially so given that the District Court, which made a point of stressing the substantial deference given to its rulings under Rule 408, was undoubtedly aware that we have held that when a district court fails to give its reasons for its exercise of discretion under Rule 403, and the court’s reasons are not apparent from the record, we are unable to conduct any meaningful review and do not defer to the district court’s determination. Becker v. ARCO Chemical Co., 207 F.3d 176, 181 (3d Cir.2000); United States v. Sriyuth, 98 F.3d 739, 749 n. 9 (3d Cir.1996); United States v. Himelwright, 42 F.3d 777, 781 (3d Cir. 1994).

Another reason to believe that the reference to 1992 was in error is that in a number of places the Court’s opinion copied language from defendant Habina’s brief, and, as it happens, that brief mistakenly referred to the period of manually altered bills as covering 1989 to 1992. Given these reasons for doubting that the District Court intended to write “1992” rather than “1994”, reasons that the majоrity never addresses, the majority’s reb-anee on the 1992 date in defending the District Court’s ruling as not excluding all evidénce about the manually changed bills is difficult to understand. Under the circumstances, I would think' that the most the majority could claim, if it believes anything of consequence rides on the 1992 date, is that a remand to the District Court for clarification is in order. Certainly the evidence calls for more than a flip reference to needing “telepathic powers” to interpret the District Court’s actions. See Majority Op. fn. 5 at’235. The majority’s failure to engage the evidence apparently flows from its refusal to acknowledge cases like Smalis, Scott, Martin Linen Supply, . Maker, and Ember, which holds that the name a District Court gives to its ruling does not control.

But in any event, I believe that whether the District Court intended to write “1992” or “1994” is irrelevant. The fact remains that the District Court erroneously believed that the fraudulent scheme alleged in the indictment was improper and as a result erroneously precluded the government from presenting any evidence about a very crucial period of the fraudulent scheme alleged in the indictment. It really doesn’t matter whether the ban on manually altered bills extends to 1992 or 1994, for under either scenario the District Court improperly changed the indictment on an offense element. Indeed, if the District Court did intend to include one year of the manually altered bills as part of the fraudulent scheme, that may only make the Court’s action worse because it would be even clearer that the Court was exercising the type of purely prosecutorial discretion in framing charges that should never rest with the judiciary.

Once it is estаblished that the District Court’s action must be deemed a dismissal of the indictment, the question then becomes whether the government is entitled to an appeal even though a jury had been sworn and testimony taken from witnesses. Although the majority’s discussion confuses the analysis of this point by discussing appeals from acquittals and mistrials, neither of which is implicated in this case, the analysis of an appeal from a dismissal is straightforward. Existing Supreme Court precedent holds that if a defense motion prompts a district court to grant a dismissal before any determination of the defendant’s guilt, as is true in this case, then the government is entitled to an appeal — even though the jury has already been empaneled and sworn. See United States v. Scott, 437 U.S. 82, 95-101, 98 S.Ct. 2187, 57 L.Ed.2d 65 (1978); Lee v. United States, 432 U.S. 23, 27-28, 33, 97 S.Ct. 2141, 53 L.Ed.2d 80 (1977). Just as a defendant whose conviction is reversed on appeal due to legal error may be subjected to retrial, so too there is no double-jeopardy bar “where the defendant, instead of obtaining reversal of his conviction on appeal, obtains the termination of the proceedings against him in the trial court without any finding by a court or jury as to his guilt or innocence.” Scott, 437 U.S. at 100, 98 S.Ct. 2187.

The majority attempts to distinguish Scott and Lee on two grounds. First, the majority says that the appeal is in fact controlled by Sanabria v. United States, 437 U.S. 54, 98 S.Ct. 2170, 57 L.Ed.2d 43 (1978), which dealt with an appeal where the trial court had entered an acquittal. Second, the majority attempts to limit Scott and Lee by grafting onto them a requirement that the defendants consent to the termination of their trial, a standard that the majority insists was not met here. This consent requirement, according to the majority, would in fact bar the appeal in this case even if the District Court’s ruling did constitute a dismissal. Both attempts to distinguish Scott and Lee are flawed.

Sanabria was handed down on the same day as Scott and amplifies on a crucial condition in Scott. As noted earlier, Scott stated that if the gоvernment appealed a dismissal after jeopardy attached, the appeal would only be allowed if the judge or jury had not made a factual determination about the defendant’s guilt or innocence. This caveat follows the longstanding rule that an acquittal bars appellate review and any further prosecution of that count. See e.g., Martin Linen Supply Co., 430 U.S. at 571, 97 S.Ct. 1349 (citing United States v. Ball, 163 U.S. 662, 671," 16 S.Ct. 1192, 41 L.Ed. 300 (1896)). The Court’s decision in Sanabria had to interpret how that rule would apply when the judge made an erroneous evidentiary ruling excluding allegations that could have independently supported criminal liability, but then entered an order of acquittal for insufficient evidence. In analyzing this scenario, the Supreme Court decided that it would not breach its firm rule against reviewing the correctness of acquittals and hence would not look behind them to see if they rested on an error of law. When the majority attempts to apply Sanabria to this case, what they overlook is that the central element of Sanabria’s analysis, an acquittal, is not present here. Indeed, the defendants in our case filed a motion for acquits tal, and the District Court denied it. San-abria is therefore not controlling, and Lee, the case handed down on the same day as Sanabria, is.

The majority’s second attempt to limit Scott and Lee is its “consent” requirement. According to the majority, even if there was a dismissal, and even if the defendants filed a motion that would necessarily require a dismissal, this case still does not fall within Scott and Lee because the defendants did not expressly ask for a dismissal of the indictment and termination of their prosecution before the empaneled jury. One obvious weakness with their theory is that if a district court’s labeling of its action does not control for the purposes of analyzing appeals under § 3731 and the Double Jeopardy Clause, see, e.g., Smalis v. Pennsylvania, 476 U.S. at 144 n. 5, 106 S.Ct. 1745; Scott, 437 U.S. at 96, 98 S.Ct. 2187; Martin Linen Supply Co., 430 U.S. at 567 n. 4, 97 S.Ct. 1349; Maker, 751 F.2d at 621 n. 26, then it is very difficult to understand why dеfendants’ labeling of their actions should — particularly given the defendants have strong motives for manipulating the names they give to their motions. The majority’s position practically invites defendants to wait until the eve of trial and file motions to redact the indictment when what they wish to achieve is a dismissal or substantial reworking of the indictment to make the government’s case impossible to prove.

It is also telling that the only authority the majority has for its hard and fast rule of what constitutes “consent” is a case where the defendant never filed any motion and simply expressed agreement with the district court’s dismissal. See United States v. Kennings, 861 F.2d 381, 385 (3d Cir.1988). If the majority intends to convert the focus on subjective intent in Ken-nings from a sufficient condition to a necessary ‍‌‌‌​​​​​‌‌​​​​‌‌‌​​​​‌​‌​​​‌​‌‌‌​‌​‌​​‌‌​​‌​‌‌‌‌‍one, then the majority has indeed limited Scott and Lee to their facts, for what defendant will ever agree to a government appeal instead of going scot-free?

One last point bears mention. The majority complains that the government should not be heard to complain now because it did not invoke Fed.R.Crim.P. 12(e) after the defendants filed their motion to redact on the eve of trial. In Lee, however, the Supreme Court rejected precisely this argument as a grounds for denying the government an appeal when there was a dismissal of the indictment. As the Court noted, the defendant “had only himself to blame” for the fact that the dismissal occurred after the court began to hear evidence, for by “the last minute timing of his motion to dismiss, he virtually assured the attachment of jeopardy.” Lee, 432 U.S. at 28, 97 S.Ct. 2141.

The majority’s position goes too far in allowing the judiciary to rewrite indictments and misinterprets controlling Supreme Court precedent. I dissent. Judge Fuentes joins in this dissent.

Case Details

Case Name: United States v. David M. Pharis Edward J. Habina William M. Dull Harry Gangloff

Court Name: Court of Appeals for the Third Circuit

Date Published: Sep 30, 2002

Citations: 298 F.3d 228; 00-2855

Docket Number: 00-2855

Court Abbreviation: 3rd Cir.