359 F. 3d 312 - Government of the Virgin Islands v. Hodge

This is an appeal by four codefendants, Selvin Hodge, Ottice Bryan, Kirsten Greenaway, and Eladio Camacho, of an order of the District Court of the Virgin Islands, Appellate Division in an interlocutory appeal brought by the Government of the Virgin Islands from the Territorial Court of the Virgin Islands. Seeking reversal, pursuant to 4 V.I.Code § 39(a)(1), the Government sought review of the Territorial Court's pretrial order redacting the confessions that the Government planned to use against the defendants. The Appellate Division held that the Territorial Court had erred in redacting the confessions more stringently than required by the Supreme Court's holdings in Bruton v. United States, 391 U.S. 123, 88 S.Ct. 1620, 20 L.Ed.2d 476 (1968), Richardson v. Marsh, 481 U.S. 200, 107 S.Ct. 1702, 95 L.Ed.2d 176 (1987), and Gray v. Maryland, 523 U.S. 185, 118 S.Ct. 1151, 140 L.Ed.2d 294 (1998). Accordingly, the Appellate Division vacated the Territorial Court's order and remanded for that Court to reconsider the redactions in the first instance.

Having lost before the Appellate Division, the defendants seek review in this Court, arguing that the Appellate Division either (1) lacked jurisdiction over the Government's interlocutory appeal (and hence that the Territorial Court's order should stand for now), or (2) erred on the merits in vacating the Territorial Court's order. This being an interlocutory appeal from an order entered in an interlocutory appeal, the threshold issue is our own appellate jurisdiction. We conclude that we have appellate jurisdiction over the Appellate Division's determination of its own jurisdiction under the collateral order doctrine. We also conclude that we lack appellate jurisdiction to review the merits of the Appellate Division's ruling. Most notably, we decide that a certification by the Government that the Territorial Court's grant of a pretrial suppression motion deprives the Government of "substantial proof of the charge pending against the defendant" satisfies the requirements of 4 V.I.Code § 39(a)(1), without a separate substantiality determination by the court. Accordingly, in this case we hold that the Appellate Division had jurisdiction under 4 V.I.Code § 39(a)(1) to hear the Government's interlocutory appeal, dismiss the appeals in all other respects, and remand to the Territorial Court for further proceedings in accordance with the Appellate Division's opinion.

We have recently described the structure of the court system in the Virgin Islands in some detail, see Gov't of V.I. v. Rivera, 333 F.3d 143, 145-46 (3d Cir.2003), cert. denied, ___ U.S. ___, 124 S.Ct. 1171, 157 L.Ed.2d 1205, 72 U.S.L.W. 3373 (U.S. Jan. 26, 2004) (No. 03-736), and need not recount it here, though some details bear repeating. There are two trial courts: The Territorial Court is comparable to a state court of general jurisdiction, see 4 V.I.Code § 76, while the District Court of the Virgin Islands has "the jurisdiction of a District Court of the United States," 48 U.S.C. § 1612(a). Appeals from the District Court come to this Court under the familiar provisions of 28 U.S.C. §§ 1291-1292 and other special-purpose statutes. Until the Virgin Islands establishes a local intermediate appellate tribunal, appeals from the Territorial Court go to a three-judge panel known as the District Court of the Virgin Islands, Appellate Division (the "Appellate Division").1 See 48 U.S.C. § 1613a(a). Though established by federal law, the Appellate Division exercises "such appellate jurisdiction over the courts of the Virgin Islands established by local law [i.e., the Territorial Court] to the extent now or hereafter prescribed by local law." Id. In other words, the Virgin Islands Legislature decides (subject to some reservations in 48 U.S.C. § 1613a(a)) who can appeal to the Appellate Division, and when they can appeal. At issue in this case is a provision regarding interlocutory appeals by the Government of certain pretrial orders in criminal cases, 4 V.I.Code § 39(a)(1). Appeals from decisions of the Appellate Division may be taken to this Court under 48 U.S.C. § 1613a(c).

B. The Proceedings in the Territorial Court and Before the Appellate Division

In November 1999, Duvalier Basquin was lured to a lonely road in the Bolongo Bay area of St. Thomas. There, he was robbed and murdered. Following an investigation by the Virgin Islands Police, the Government of the Virgin Islands (the "Government") charged Selvin Hodge, Ottice Bryan, Kirsten Greenaway, and Eladio Camacho (collectively, the "defendants") with robbery, felony murder, and conspiracy to commit murder. During the investigation, Hodge and Camacho gave statements inculpating themselves and the other defendants in Basquin's murder. Greenaway gave a statement exculpating herself, but potentially inculpating the other defendants. Bryan gave no statement.

The Government sought to use these statements at trial. However, since the Government proposed to try the defendants jointly, and none of the defendants who offered statements would testify, the statements would have to be redacted—or even rewritten—to preserve the defendants' Sixth Amendment Confrontation Clause rights. See Bruton v. United States, 391 U.S. 123, 88 S.Ct. 1620, 20 L.Ed.2d 476 (1968), Richardson v. Marsh, 481 U.S. 200, 107 S.Ct. 1702, 95 L.Ed.2d 176 (1987), and Gray v. Maryland, 523 U.S. 185, 118 S.Ct. 1151, 140 L.Ed.2d 294 (1998).2 At a pretrial hearing before the Territorial Court, the Government offered proposed redactions of the statements, but after lengthy argument, the Territorial Court concluded that the Government's proposal did not satisfy Bruton and its progeny. Ruling from the bench, the Territorial Court described the further redactions that would be required to admit the confessions.3

Title 4, section 39(a)(1) of the Virgin Islands Code provides:

The United States or the Government of the Virgin Islands may appeal an order, entered before the trial of a person charged with a criminal offense under the laws of the Virgin Islands, which directs the return of seized property, suppresses evidence, or otherwise denies the prosecutor the use of evidence at trial, if the United States Attorney or the Attorney General conducting the prosecution for such violation certifies to the Judge who granted such motion that the appeal is not taken for purpose of delay and the evidence is a substantial proof of the charge pending against the defendant.

The Government, relying on 4 V.I.Code § 39(a)(1), noticed its appeal to the Appellate Division and on the same day provided the certification that the statute requires. On appeal, the Appellate Division opined that the Government's proposed redaction was insufficient to protect the defendants' Confrontation Clause rights, but concluded that the Territorial Court had directed more redaction than necessary. It offered some illustrations of how, on remand, the Territorial Court could solve the "Goldilocks problem" of crafting altered confessions that were not too lightly redacted, not too heavily redacted, but just right.

The defendants were disappointed in the outcome before the Appellate Division; they would have much preferred the redactions ordered by the Territorial Court.4 They appealed to this Court, arguing that either the Appellate Division did not have jurisdiction under 4 V.I.Code § 39(a)(1) to review the Territorial Court's order, or else that the Appellate Division erred on the merits in holding that the Territorial Court went further than required by Bruton and its progeny. Under the former disposition, we would simply reinstate the Territorial Court's order. Under the latter disposition, we would confront the merits of the defendants' Bruton argument.

At the threshold, we must examine whether we have appellate jurisdiction over one, both, or neither of the questions that the defendants present. See Gov't of V.I. v. Marsham, 293 F.3d 114, 116 (3d Cir.2002) (quoting Collinsgru v. Palmyra Bd. of Educ., 161 F.3d 225, 229 (3d Cir.1998) ("we have an independent obligation to examine our jurisdiction to hear this appeal.")). Three of the four defendants invoke this Court's jurisdiction under 28 U.S.C. § 1291. Although some of our cases are imprecise about the statutory source of our jurisdiction over the Appellate Division, we take this opportunity to clarify that, as a technical matter, it is 48 U.S.C. § 1613a(c), and not 28 U.S.C. § 1291, that confers jurisdiction on this Court over appeals from the Appellate Division. However, the distinction is only technical—our cases have uniformly held that 48 U.S.C. § 1613a(c) has the same requirements for appealability as 28 U.S.C. § 1291. See, e.g., Rivera, 333 F.3d at 147; Ortiz v. Dodge, 126 F.3d 545, 547 (3d Cir.1997).

Turning to the substance of our appellate jurisdiction, we consider whether we have jurisdiction over some or all of this case as a "final decision" of the Appellate Division within the meaning of 48 U.S.C. § 1613a(c). We conclude that we do not in the usual sense. We then consider whether we have appellate jurisdiction over some or all of this case under the collateral order doctrine. We conclude that we do have jurisdiction under the collateral order doctrine to review the Appellate Division's determination of its own jurisdiction.

We are the second appellate court to address this case. Nonetheless—to reiterate the point made above about the parallel construction of 48 U.S.C. § 1613a(c) and 28 U.S.C. § 1291—"with regard to the question of finality, we have treated appeals from the Appellate Division ... no differently than appeals taken from any other federal district court." Ortiz, 126 F.3d at 548 (citing as examples Gov't of V.I. v. Blake, 118 F.3d 972 (3d Cir.1997); In re A.M., 34 F.3d 153 (3d Cir.1994)).

The key question is whether the vacate-and-remand order of the Appellate Division was a final decision under 48 U.S.C. § 1613a(c). It was not a "final decision" in the most common sense of the term—for two reasons. First, it was a remand order, and we have explained in a similar context that remand orders are not final under § 1613a(c). In re Alison, 837 F.2d 619 (3d Cir.1988), considered our appellate jurisdiction over an order of the Appellate Division reversing the Territorial Court's grant of a Fed.R.Civ.P. 12(b)(6) motion to dismiss. Since the Appellate Division had reversed, it remanded the case to the Territorial Court for further proceedings. We concluded that such a remand was not a final decision under § 1613a(c). Remand orders are not generally appealable because they are not final decisions within the meaning of 28 U.S.C. § 1291 and 48 U.S.C. § 1613a(c). We recently reiterated that "[a] final decision `ends the litigation on the merits and leaves nothing ... to do but execute the judgment.'" Rivera, 333 F.3d at 150 (alteration in original) (quoting Catlin v. United States, 324 U.S. 229, 233, 65 S.Ct. 631, 89 L.Ed. 911 (1945)). The remand in Alison left more to do than mere execution of the judgment, and thus the remand order was not appealable.

A second, independent reason leads us to conclude that the Appellate Division's order was not a final decision: The first appeal (i.e., the appeal to the Appellate Division) was interlocutory, but, as we explain in Part III below, was nonetheless proper. The subsequent appeal to this Court asks us, in effect, to (re)consider an interlocutory order of a trial court. But, in view of the finality policy of 28 U.S.C. § 1291 and 48 U.S.C. § 1613a(c), this is something which we do not generally engage in (absent specific statutory authorization).5 Such statutory authorization comes from Congress. See U.S. Const. Art III. § 1 ("The judicial Power of the United States, shall be vested in ... such inferior Courts as the Congress may from time to time ordain and establish."). If we were to blithely take jurisdiction over appeals of decisions that the Appellate Division rendered on interlocutory appeal, we would in practice be allowing our jurisdiction to expand based on the Virgin Islands Legislature's exercise of its authority, under 48 U.S.C. § 1613a(a), to determine the appellate jurisdiction of the Appellate Division. Of course, the scheme in § 1613a means that, for a Territorial Court case to appear on our docket on appeal, it is necessary that the Virgin Islands Legislature confer intermediate appellate jurisdiction on the Appellate Division; but it does not follow that such a jurisdictional statute is sufficient to confer jurisdiction, in turn, on this Court. Hence we decline to conclude that in enacting § 1613a Congress intended to cede to the Virgin Islands Legislature such control over this Court's jurisdiction.

Thus we hold that the Appellate Division's decision is not a "final decision" in the most common sense under 48 U.S.C. § 1613a(c), and therefore this Court does not have appellate jurisdiction in the normal sense. We next consider whether this Court has jurisdiction under the collateral order doctrine.

This Court's recent definitive treatment of the collateral order doctrine is In re Ford Motor Co., 110 F.3d 954 (3d Cir.1997). There we explained:

[T]he collateral order doctrine, first enunciated by the Supreme Court in Cohen v. Beneficial Indus. Loan Corp., 337 U.S. 541, 69 S.Ct. 1221, 93 L.Ed. 1528 (1949), provides a narrow exception to the general rule permitting appellate review only of final orders. An appeal of a nonfinal order will lie if (1) the order from which the appellant appeals conclusively determines the disputed question; (2) the order resolves an important issue that is completely separate from the merits of the dispute; and (3) the order is effectively unreviewable on appeal from a final judgment. See Rhone-Poulenc Rorer Inc. v. Home Indem. Co., 32 F.3d 851, 860 (3d Cir.1994).

Id. at 958. As the Cohen Court explained, 28 U.S.C. § 1291 has been given a "practical rather than a technical construction." 337 U.S. at 546, 69 S.Ct. 1221. To this end, as a doctrinal matter, orders that meet the three prongs described above are deemed to be "final decisions" within the meaning of the statute.

Ford Motor Co. paid special attention to the question of what makes an issue "important" under the second prong. We described the task as one of "compar[ing] the apple of the desire to avoid piecemeal litigation to the orange of, for example, federalism." Ford Motor Co., 110 F.3d at 960. In cases where the Supreme Court has blessed interlocutory appeals, we observed, it was because "the imperative of preventing impairment of some institutionally significant status or relationship" made "the danger of denying justice by reason of delay in appellate adjudication outweigh[ ] the inefficiencies flowing from interlocutory appeal." Id.

We will apply the doctrine separately to both of the questions that the defendants urge us to consider: (1) the merits of the Appellate Division's decision, and (2) the Appellate Division's determination of its own jurisdiction.

1. Appellate Jurisdiction to Review The Merits of the Appellate Division's Decision

As to the first prong of the collateral order doctrine, the Appellate Division's order did not conclusively resolve much of anything. To be sure, it established some guideposts for "too much" and "too little" redaction, but at bottom, it remanded the issue to the Territorial Court to settle on the exact redaction to use.

On the second prong, the redaction question is clearly separable from the merits, and this favors appealability. The question about the redactions goes to how much identifying information can be contained in a nontestifying codefendant's statement and still preserve the other defendants' Confrontation Clause rights. This is an exercise in applied constitutional law, as it were, and it does not implicate the merits of whether some or all of the defendants did or did not participate in the robbery-murder of the victim. As for the importance of the question, there are mixed signals. On the one hand, the Confrontation Clause articulates a fundamental constitutional right, and one might assume that such rights cry out most strongly for vindication on interlocutory appeal. Cf., e.g., P.R. Aqueduct & Sewer Auth. v. Metcalf & Eddy, Inc., 506 U.S. 139, 145, 113 S.Ct. 684, 121 L.Ed.2d 605 (1993) (holding that determination of sovereign immunity was a proper subject for interlocutory appeal because it "involves a claim to a fundamental constitutional protection"). On the other hand, Confrontation Clause rights are vindicated through evidentiary rulings, and a prime target of the policy against interlocutory appeals is the avoidance of piecemeal review of the many evidentiary rulings in a typical case. Thus we find this factor inconclusive.

The third prong strongly disfavors appealability. Practice alone—in Bruton and Gray themselves—suggests that interlocutory appeal is unwarranted because the constitutional defect in Bruton's and Gray's trials were, in fact, remedied by vacating their convictions and remanding for a new trial.6

In sum, the prongs range from inconclusive to strongly disfavoring appealability. As the collateral order doctrine is a "narrow" exception and the Supreme Court has "described the conditions for collateral order appeal as stringent," Digital Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 868, 114 S.Ct. 1992, 128 L.Ed.2d 842 (1994), failure to meet one prong makes the doctrine inapplicable no matter how compelling the other prongs may be (and here, not even one prong is in favor of appealability). Thus we conclude that this Court does not have appellate jurisdiction to hear an appeal of the merits of the Appellate Division's order.

2. Appellate Jurisdiction to Review the Appellate Division's Determination of Its Own Jurisdiction

Turning to the question of the reviewability of the Appellate Division's determination of its own jurisdiction, it is clear that we may at least review this limited question. This Court's indistinguishable precedent in Government of the Virgin Islands v. Blake, 118 F.3d 972 (3d Cir.1997), compels this conclusion. In that case, the Government had taken an interlocutory appeal from the Territorial Court to the Appellate Division under 4 V.I.Code § 39(d), a provision which allows an interlocutory appeal during trial under some circumstances. (In Blake, the Territorial Court had suppressed—during motions decided after the jury had been selected and sworn—a variety of testimony and other evidence the Government sought to present.) The Appellate Division in Blake decided that it did not have jurisdiction to hear the Government's appeal. On appeal we held that although we had no jurisdiction to reach the merits, we did have jurisdiction under the collateral order doctrine to review the Appellate Division's jurisdiction over the appeal. Blake, 118 F.3d at 975-76. We of course adhere to Blake in this case, see Third Circuit IOP 9.1, but we do add a few words of analysis since the discussion in Blake was quite summary.

The first prong of the collateral order doctrine is clearly satisfied here because the Appellate Division did finally determine its own jurisdiction over this sort of interlocutory appeal. The third prong is also clearly satisfied because such a determination cannot be effectively reviewed on appeal from a final judgment because, by hypothesis, the Appellate Division's jurisdiction to hear interlocutory (i.e., not final) appeals would not be implicated in that posture.

The second prong is more complex, but it too favors our jurisdiction. Part of it is clear: The issue of the Appellate Division's jurisdiction is separate from the merits. Whether the question is important enough requires some discussion. On the one hand, issues involving the scope of federal jurisdiction are good candidates for the collateral order doctrine. See, e.g., Quackenbush v. Allstate Ins. Co., 517 U.S. 706, 116 S.Ct. 1712, 135 L.Ed.2d 1 (1996) (holding that an abstention-based remand to state court was immediately appealable under collateral order doctrine). On the other hand, a vague reference to the "scope of federal jurisdiction" may denominate the category too broadly, for the cases involving the collateral order doctrine and the scope of federal jurisdiction are by and large abstention cases, see id. at 712-15, 116 S.Ct. 1712 (canvassing cases), which "put the litigants `effectively out of court,'" id. at 713, 116 S.Ct. 1712 (quoting Moses H. Cone Mem'l Hosp. v. Mercury Contr. Corp., 460 U.S. 1, 11 n. 11, 103 S.Ct. 927, 74 L.Ed.2d 765 (1983) (quoting Idlewild Bon Voyage Liquor Corp. v. Epstein, 370 U.S. 713, 715 n. 2, 82 S.Ct. 1294, 8 L.Ed.2d 794 (1962))), and some courts have explicitly held that "non-immunity based motions to dismiss for want of subject matter jurisdiction are not ordinarily entitled to interlocutory review." Merritt v. Shuttle, Inc., 187 F.3d 263, 268 (2d Cir.1999) (citing Catlin, 324 U.S. at 236, 65 S.Ct. 631).

The dispositive differences in this case are twofold. First, we are considering the ability to appeal an interlocutory determination of appellate jurisdiction, not original jurisdiction, making cases like Merritt distinguishable. Second, the order at issue here is not so much effectively unreviewable as it is procedurally unreviewable if we do not take jurisdiction now. "Effective" unreviewability arises because a party's putative rights will be irreparably harmed. For example, a party may have to forego an injunction guarding against irreparable harm because the security bond that is the price of the injunction may have been made too costly by the lower court; or a party wrongly determined to lack qualified immunity may be subjected to a trial. In such situations, although the aggrieved party cannot be made whole after the fact, the legal question will, as a matter of procedure, still be preserved for the appellate court's review at a later time. In contrast, only in the most convoluted and improbable of hypotheticals will the jurisdictional issue presented here ever make its way to this Court on appeal from a final decision.7 As a procedural matter, now is this Court's only opportunity to pass on the issue.

This reasoning also explains why our holding here would not apply to the issue in Merritt, i.e., why a district court's determination of its subject matter jurisdiction is not generally reviewable under the collateral order doctrine.8 Questions of original jurisdiction are always automatically before this Court on appellate review. See, e.g., Wujick v. Dale & Dale, Inc., 43 F.3d 790, 792 (3d Cir.1994) ("`[E]very federal appellate court has a special obligation to satisfy itself not only of its own jurisdiction, but also that of lower courts in a cause under review.'" (alteration in original) (quoting Spring Garden Assoc., L.P. v. Resolution Trust Corp., 26 F.3d 412, 415 (3d Cir.1994) (quoting Employers Ins. of Wausau v. Crown Cork & Seal Co., 905 F.2d 42 (3d Cir.1990)))). In other words, there is no procedural posture where a question of original jurisdiction will escape this Court's review in an appeal from a (non-interlocutory) final decision. In contrast—as this case itself illustrates—there are procedural postures which render permanently unreviewable the judgment of a hierarchically inferior appellate court,9 and thereby prevent the automatic review of jurisdiction described in Wujick. Because review of a question of appellate jurisdiction is a now-or-never proposition, interlocutory review of a jurisdictional question is warranted here where it is not warranted in the case of a district court's determination of its own original jurisdiction.

In brief, coupled with the institutional importance of the question, the absolute unreviewability of the Appellate Division's jurisdiction in this case makes the question an important one. Thus this prong too favors appealability. Because all three prongs are satisfied, the collateral order doctrine affords us a basis for reviewing the Appellate Division's determination regarding its jurisdiction under 4 V.I.Code § 39(a)(1).

In reaching this conclusion, we have considered the dissent's contention that our "decision effectively grants an appeal as of right to question an appellate court's jurisdiction whenever it makes an interlocutory ruling," and that this result is the very sort of inefficiency that the collateral order doctrine should not countenance. We are underwhelmed by the dissent's in terrorem argument. First, it is a dubious empirical proposition that the holding here will increase the quantity of this sort of appeal. As the citations in the opinion in this case suggest, this Court has seen but a handful of cases like this in the past decade. Second, the fact that we here take the opportunity to give some guidance (both to litigants and to the Appellate Division) should decrease, not increase, the number of appeals taken in good faith.

Third, the dissent claims that "[t]he majority's decision effectively grants an appeal as of right." But it is the Congress, not this Court, that has granted litigants an appeal as of right from the Appellate Division. Even if we did dismiss this appeal in its entirety for lack of jurisdiction, as the dissent would, little efficiency would be gained as a practical matter: In a subsequent case, a litigant could still file a notice of appeal (as a matter of statutory right), and he could still advance a good faith argument in favor of review under the collateral order doctrine. A motions panel would likely refer the jurisdictional question to the merits panel, and the merits panel would address the question (hopefully in less extended fashion than we have had to here). In other words, the decision here makes it neither easier nor harder for a party who is dissatisfied with the ruling of the Appellate Division to drag out the process by appealing to this Court.10

At bottom, it seems to us that the dissent's problem is the presence of a system of two-tier appellate review as of right. In fact, the dissent states explicitly that "[t]his type of review is wisely not found elsewhere in the federal system, and should not exist here." Dissenting Op. post at 328. While we might agree with the dissent if we were drafting 48 U.S.C. § 1613a, that simply is not our task. Congress has provided that we have appellate jurisdiction (until such time as the conditions for certiorari jurisdiction are met, see 48 U.S.C. § 1613), and accordingly, we will turn our attention to the substance of the appeal.

We exercise plenary review in determining whether a court hierarchically below us had subject matter jurisdiction. Wujick v. Dale & Dale, Inc., 43 F.3d 790, 792 (3d Cir.1994) (citing Nat'l Union Fire Ins. Co. v. City Savings, F.S.B., 28 F.3d 376, 383 (3d Cir.1994)). As previously noted, we look to local law to determine the jurisdiction of the Appellate Division. See 48 U.S.C. § 1613a(a); Gov't of V.I. v. Warner, 48 F.3d 688 (3d Cir.1995). The asserted basis for the Appellate Division's subject matter jurisdiction is 4 V.I.Code § 39(a)(1), which provides:

The United States or the Government of the Virgin Islands may appeal an order, entered before the trial of a person charged with a criminal offense under the laws of the Virgin Islands, which directs the return of seized property, suppresses evidence, or otherwise denies the prosecutor the use of evidence at trial, if the United States Attorney or the Attorney General conducting the prosecution for such violation certifies to the Judge who granted such motion that the appeal is not taken for purpose of delay and the evidence is a substantial proof of the charge pending against the defendant.

There do not appear to be any cases—either from this Court or from the Appellate Division—considering this provision in any depth.11 We are, however, fortunate to be guided in this novel exercise by interpretations of 18 U.S.C. § 3731, the analogous provision for interlocutory appeals by the government of suppression motions decided in federal district courts.12

The defendants' challenges to the Appellate Division's jurisdiction are fourfold: First, they argue that the Territorial Court's redactions were not "an order ... suppress[ing] evidence." Second, they submit that the Government's certification was inadequate because the Territorial Court was not afforded an opportunity to review it before the Government noticed its appeal to the Appellate Division. Third, they contend that the statements in question, even unretracted, are not "substantial proof of the charge pending against [them]." Fourth, they assert that the statute denies them equal protection of the laws, by permitting the government to appeal adverse orders but not affording a similar opportunity to defendants facing adverse orders. We will treat each of these challenges in turn.

A. "An Order Suppressing Evidence"

The defendants contend that the Territorial Court's redaction order was not "an order suppress[ing] evidence" within the meaning of 4 V.I.Code § 39(a)(1). But it would be a terribly crabbed reading of the statute to hold that admitting a statement subject to redactions does not amount to a suppression of evidence. The statute finishes its list of appealable orders with a catch-all category—those orders which "otherwise den[y] the prosecutor the use of the evidence at trial." Id. Even if the redaction order is not an order suppressing evidence, it surely denies the Government the use of the full confessions at trial. Thus we reject the defendants' contention that the Territorial Court's order was not a proper subject for appeal.

B. Territorial Court Review of the Certification

The defendants next argue that the Territorial Court should have had an opportunity to review the certification provided by the Government. While as a policy matter, one might conclude that the Territorial Court should be given an opportunity to review the certification, there is at present simply no basis in the statute for such a requirement. We decline to read one in.

Moreover, this Court has decided in interpreting 18 U.S.C. § 3731 that that statute requires nothing more than the delivery of the certification to the district court in question. See In re Grand Jury Investigation, 599 F.2d 1224, 1226 (3d Cir.1979) (adopting United States v. Comiskey, 460 F.2d 1293, 1297-98 (7th Cir.1972) (holding that no evidence was required to support a certification under 18 U.S.C. § 3731)). We hold, by analogy, that 4 V.I.Code § 39(a)(1) is also completely satisfied by simple delivery of the certification to the Territorial Court. Thus the Government's certification in this case was adequate.

C. "Substantial Proof of the Charge Pending"

The Courts of Appeals have generally held that the "substantial proof" requirement of 18 U.S.C. § 3731 is a requirement of the certification, not a requirement of actual fact. See In re Grand Jury Investigation, 599 F.2d at 1226.13 The certification under 18 U.S.C. § 3731 must state both (1) that the appeal is not taken for the purpose of delay; and (2) that the evidence is a substantial proof of a material fact in the proceeding. The United States Attorney's word is enough; the reviewing court does not consider the truth of the certification.

If 4 V.I.Code § 39(a)(1) were worded identically to 18 U.S.C. § 3731, on which In re Grand Jury Investigation is binding precedent, this would certainly be the end of it. But the Virgin Islands statute is not clearly drafted: It is missing an additional "that" immediately before "the evidence is a substantial proof," which would make it grammatically unambiguous. As it now stands, it could be read to require essentially the same two certification elements that 18 U.S.C. § 3731 does; or it could be read (as the defendants urge) to require a certification that the appeal is not taken for delay, and also require that, in actual fact, the evidence be "substantial proof."

We decline to read it as the defendants would have it, for four reasons. First, the Virgin Islands statute is plainly modeled on the federal statute, and we conclude it is best to follow the federal statute as a guide. Second, if the Virgin Islands Legislature actually wanted the statute to operate as the defendants would have it, there were far less cryptic ways of communicating that intent. For example, the Virgin Islands Legislature could have put the "substantial proof" requirement before the certification clause. Third, the statute is silent on what court would evaluate the substantiality of the proof, or on how that court would go about the evaluation, suggesting that no such review is to be undertaken. Fourth, review of the substantiality of the proof necessarily entails a look at the other evidence that the Government has available to it, an inquiry that could both take considerable time and prejudice the Government's case. Both of these seem at odds with the expedited interlocutory appellate review contemplated by the statute.14 Thus, we conclude that the Appellate Division did not need to evaluate the substantiality of the proof in order to accept the certification.

The defendants object that the statute affords the Government an appeal right that it does not confer on defendants, and that this disparity amounts to a violation of the equal protection component of the Fifth Amendment's Due Process Clause. But this argument is foreclosed by United States v. Heinze, 218 U.S. 523, 545-46, 31 S.Ct. 98, 54 L.Ed. 1139 (1910), which held that the Act of March 2, 1907, ch. 2564 (which was later codified at 18 U.S.C. § 682, which in turn was the forerunner of 18 U.S.C. § 3731) did not violate equal protection or due process principles.

Moreover, even under more recent equal protection jurisprudence, Heinze reached the right result: The Supreme Court has not announced that the status of "criminal defendant" is a suspect classification, nor has it held the right to appeal in a criminal case to be a fundamental right. Thus we subject this legislation to rational basis review. See, e.g., Ramsgate Court Townhome Ass'n v. W. Chester Borough, 313 F.3d 157, 160 (3d Cir.2002) (reiterating that equal protection challenges to legislation not based on a suspect classification or implicating a fundamental right require only rational basis review of the legislation).

The Virgin Islands Legislature certainly had a rational basis for distinguishing between the Government and defendants on the question of appeal rights. To identify only one such basis, the Virgin Islands Legislature could rationally conclude that the efficient administration of criminal justice demanded that the Government have expansive pre-trial appeal rights, and the defendant have expansive post-trial appeal rights. Thus we find no merit in the defendants' equal protection challenge.

In sum, we conclude that the Government followed the procedures set out in 4 V.I.Code § 39(a)(1), and that the Appellate Division had jurisdiction over the appeal.

We have concluded that the Appellate Division had jurisdiction over the appeal, and we will affirm the Appellate Division's holding regarding its own jurisdiction. Having determined that we lack jurisdiction over the other questions in this appeal, we will dismiss the appeal with respect to those questions. The case will be remanded to the Territorial Court to revise the redacted statements in conformity with the Appellate Division's opinion. Because we do not reach the merits of the defendants' Bruton claims, this opinion is without any preclusive effect to the defendants asserting on direct appeal, should they be convicted, that the redacted confessions used at trial were insufficiently altered to secure their Sixth Amendment Confrontation Clause rights.

NYGAARD, J. dissenting.

I respectfully dissent and would dismiss this appeal for lack of jurisdiction. I agree with the majority that we do not have jurisdiction to consider the merits of the decision of the District Court of the Virgin Islands, Appellate Division. I disagree, however, that we may consider whether the Appellate Division had jurisdiction to hear the interlocutory appeal. The majority asserts that its resolution of this issue is "clear" and compelled by the "indistinguishable precedent" of Government of Virgin Islands v. Blake, 118 F.3d 972 (3d Cir.1997). Maj. Op. at 320. I agree the issue is clear, but conclude that the precedent of Blake is first, quite distinguishable, and moreover, not germane to our decision.

The majority insists that we should treat appeals from the Appellate Division "no differently than appeals taken from any other federal district court." Maj. Op. at 318. While I agree in general, I do not believe this means we can simply be blind to the differences. The majority's decision effectively grants an appeal as of right to question an appellate court's jurisdiction whenever it makes an interlocutory ruling. The procedural equivalent is not our routine review of a decision by a typical district court, but review by some hypothetical higher court of our jurisdiction every time we take an interlocutory appeal.

This type of review is wisely not found elsewhere in the federal system, and should not exist here. Under the majority's holding, we pile an extra layer of interlocutory appellate review solely onto cases that stem from the Virgin Islands, without explaining why this extra layer is necessary — and why the delay it engenders is justified. I find the extra layer unnecessary and the delay unjustified.

The majority correctly concludes that the issue of the Appellate Division's jurisdiction is not reviewable as an ordinary final decision. It also uses the correct test for determining whether, nevertheless, we may pretend it is a final decision and review it under the collateral order doctrine. It is the manner in which the majority applies this test where they and I part company.

At the outset, I think it important to emphasize what the majority only notes in passing—that the collateral order doctrine is meant to provide a "narrow exception" to the general rule that permits appellate review only of truly final orders. See Digital Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 868, 114 S.Ct. 1992, 128 L.Ed.2d 842 (1994). This exception is to be made only when required to preserve "a healthy legal system," and should "never be allowed to swallow the general rule." Id. at 867-68, 114 S.Ct. 1992 (internal citation and quotations omitted). Thus, the three prongs of the collateral order doctrine are "stringent," and each of them must be met in order for a decision to be reviewable. Id. at 868, 114 S.Ct. 1992. With this in mind, it seems clear that in this case the requirements of the second and third prongs have not been met, and we therefore may not review the Appellate Division's decision as to its own jurisdiction.

In addition to being separate from the merits of the case, as this jurisdictional question admittedly is, the second prong also requires that the issue be sufficiently "important." Discussing the meaning of "important," the United States Supreme Court has explained that it involves an examination of the "value of the interests that would be lost through rigorous application of a final judgment requirement." Digital Equip. Corp., 511 U.S. at 878-79, 114 S.Ct. 1992. Similarly, we have required a showing of the "impairment of some institutionally significant status or relationship" presenting the "danger of denying justice." In re Ford Motor Co., 110 F.3d 954, 960 (3d Cir.1997). To qualify as "important," these interests must also outweigh the "inefficiencies flowing from interlocutory appeal." Id.

It is difficult for me to see how we can conclude that an extra layer of appellate review, of a kind that does not exist anywhere else in the federal system, qualifies as "important" under this standard. When this Court takes jurisdiction over an interlocutory appeal from a typical district court's proceedings, our decision to do so is not subject to automatic review, and yet we do not find that this deprivation presents the "danger of denying justice." It is incongruous, and more than a bit patronizing, to conclude that the interests of justice require that we take an interlocutory appeal to second-guess the jurisdictional conclusions of the Appellate Division when it is sitting in the same posture.

In distinguishing holdings that find issues of subject-matter jurisdiction ineligible for interlocutory review, the majority points out that this case is different because it presents a question of appellate, not original, jurisdiction.15 The majority does not, however, indicate why this distinction weighs in favor of review. Nor does the majority explain how the interests of justice implicated by an interlocutory jurisdictional issue are so weighty that they overcome the "inefficiencies flowing from interlocutory appeal." Notably, the majority actually has very little to say about the "importance" of reviewing the jurisdictional issue, blurring its discussion of this condition with its insistence that the order is "procedurally unreviewable," a factor that is properly evaluated under the third prong. Maj. Op. at 321.

In my view, this third requirement has also not been fulfilled. I fail to see why the question of interlocutory appellate jurisdiction as presented in this case is "effectively unreviewable on appeal from a final judgment." Ford Motor Co., 110 F.3d at 958. As the Supreme Court notes, most interlocutory orders are "only imperfectly reparable by appellate reversal," and if this prong were to be interpreted too broadly, it would render it meaningless. Digital Equip. Corp., 511 U.S. at 872, 114 S.Ct. 1992. "A fully litigated case can no more be untried than the law's proverbial bell can be unrung, and almost every pretrial or trial order might be called `effectively unreviewable' in the sense that relief from error can never extend to rewriting history." Id.

It is indeed likely that the issue of the Appellate Division's interlocutory appellate jurisdiction will be made irrelevant by further proceedings. If the Appellants reach a plea agreement or are acquitted, for example, then the question would be moot. In the larger context of a murder trial, the issue may be rendered immaterial. But these possibilities only indicate that our consideration of the issue at this point may well be a waste of time and resources.

If the Appellants are convicted, however, and the decision of the Appellate Division is material to the outcome of the trial, I see no reason why we could not then consider whether the Appellate Division had jurisdiction to render its decision. In fact, it seems that the issue could be raised rather straightforwardly upon appeal, without requiring, as the majority suggests, "the most convoluted and improbable of hypotheticals." Maj. Op. at 322. True, if we were to reverse, at that point we could not rewrite history and pretend that the Appellate Division had never rendered its interlocutory ruling, but as the Supreme Court has emphasized, effective reviewability of a decision does not require that we be able to unring the "law's proverbial bell." Id.

It is in regard to reviewability that this case is most easily distinguishable from Blake. 118 F.3d 972. In Blake, the Appellate Division had found that it did not have jurisdiction to consider the government's interlocutory appeal, and it was the government, not the defendants, who sought to appeal that ruling to us. Id. at 974. This denial of jurisdiction could not have been effectively reviewed after final judgment, especially because it is unlikely the government would have been able to appeal at all had it lost and the trial resulted in an acquittal. In contrast, here Appellants would have an automatic right to appeal upon conviction.

These distinctions also make our finding that the interlocutory issue was sufficiently "important" more palatable in the Blake case, since the denial of jurisdiction there meant the challenged order would not be reviewed by any court, and a miscarriage of justice was therefore more likely to result. In contrast, in a case such as this in which the Appellate Division took jurisdiction, the underlying issue has already been reviewed and decided by an appellate court.

This case implicates all of the interests that justify the existence of the final judgment rule, and illustrates why exceptions to this rule should be few. As the Supreme Court explained:

An interlocutory appeal can make it more difficult for trial judges to do their basic job — supervising trial proceedings. It can threaten those proceedings with delay, adding costs and diminishing coherence. It also risks additional, and unnecessary, appellate court work ... when it brings them appeals that, had the trial simply proceeded, would have turned out to be unnecessary.

Johnson v. Jones, 515 U.S. 304, 309, 115 S.Ct. 2151, 132 L.Ed.2d 238 (1995).

Fortunately, the scope of the majority's decision is limited to the unique appellate position of the courts of the Virgin Islands, but even within this narrow arena today's holding seems certain to encourage more unnecessary delay while this Court wades through more unnecessary appeals. With this holding, every ruling by the Appellate Division becomes appealable to this Court, since every such ruling must necessarily contain at least an implicit finding of jurisdiction.

To see the harm in this decision, we need look no farther than the case now before us. The trial of Appellants for a brutal 1999 murder was set to begin in September 2001, when the case was brought to a grinding halt by the government's appeal on the eve of trial. Memories of witnesses have surely faded and evidence gone stale as the case has wended its way through two appellate courts, producing what will be at least a three-year delay in trial. There is no doubt that when this trial is eventually held, it will be less coherent than it would have been three or more years earlier, and less likely to achieve a just result. It is unfortunate that there has been such a delay in this case; we should not encourage its recurrence.

The government's original interlocutory appeal was brought under a statute passed by the Virgin Islands legislature, which made a measured decision that the interests of justice warrant the delay caused by allowing interlocutory appeals in a few specified instances. Now, in making any such decision, the local legislature must weigh the interests of justice against far more delay, since this Court must now intrude on every interlocutory appeal. From now on, every interlocutory appeal allowed by statute will come stapled to a right to appeal to us the Appellate Division's decision to take the interlocutory appeal. This extra layer is not only unnecessary, but also conveys an unjustified lack of confidence in the decisions of the Appellate Division, and creates a further delay that can only serve to jeopardize the ultimate attainment of justice at trial.