493 F.3d 283
2d Cir.
2007
GUIDO CALABRESI, Circuit Judge:

Plaintiff-Appellant Industrial Risk Insurers (IRI) — the subrogee of non-party, amicus Silverstein Properties, Inc. (Silver-stein) — brought, in the United States District Court for the Southern District of New York, a claim of gross negligence against, inter alia, Defendants-Appellees Citigroup Inc. and Citigroup Global Market Holdings, Inc. (together, Citigroup). Upon reviewing the parties’ submissions, the district court dismissed IRI’s complaint pursuant to Rule 12(b)(6). See Indus. Risk Insurers v. Port Auth. of N.Y. & N.J., et al., 387 F.Supp.2d 299 (S.D.N.Y. 2005).

IRI now appeals from that judgment and, in addition, IRI has moved for partial vacatur of the district court’s opinion and order. For the reasons stated below, while we (1) affirm the judgment of the district court, we (2) remand IRI’s motion for partial vacatur to the district court, to allow that court to consider that motion in the first instance.

BACKGROUND

Citigroup was the largest tenant in 7 World Trade Center (7 WTC), a 47-story office tower that had stood adjacent to the Twin Towers of the World Trade Center. On September 11, 2001, as a result of the attacks on the nearby Twin Towers, 7 WTC caught fire. For the next seven hours, 7 WTC burned wildly, until the building ultimately collapsed to the ground. IRI had provided property insurance to Silverstein for 7 WTC, and in the wake of this damage, has paid in excess of $400 million in property loss to Silverstein.

I. The Industrial Risk Insurers Action

In the present case, IRI charged Citigroup with gross negligence. Citigroup, as 7 WTC’s largest tenant, assertedly chose to design, construct, and install a diesel-fuel-powered generator system in 7 WTC, which pumped fuel through the entire building at all times. IRI claims that this generator system was unreasonably dangerous, and that, while Citigroup’s alleged gross negligence did not cause the initial combustion in 7 WTC, it directly-led to an aggravation of fire damage and the ultimate collapse of the building. As noted above, the district court dismissed IRI’s complaint pursuant to Rule 12(b)(6).

On appeal, IRI argues (1) that the district court improperly concluded, as a matter of law, that Citigroup’s conduct was not gross negligence; (2) that the district court erred in holding that New York’s “subrogation waiver” doctrine precluded IRI from asserting a claim sounding in gross negligence; and (3) that the district court erred in applying the doctrine of assumption of risk to bar IRI’s gross negligence claim.

IRI submitted its appellant brief to this court on May 2, 2005. Before Citigroup had submitted its appellee brief, however, our court decided St. Paul Fire & Marine Ins. Co. v. Universal Builders Supply, 409 F.3d 73 (2d Cir.2005) (holding that a waiver of subrogation clause bars a claim of gross negligence under New York law). Citigroup then filed its appellee brief on June 1, 2005, in which it argued — persuasively — 'that St. Paul required us to affirm the district court’s “subrogation waiver” holding.

The district court’s decision with respect to the waiver of subrogation clause is independent of the district court’s other two bases for granting the motion to dismiss— 1.e., that Citigroup’s conduct, as a matter of law, was not gross negligence, and that the assumption of risk doctrine would bar IRI’s gross negligence claim even if that claim were otherwise meritorious. Accordingly, given St. Paul —which requires us to affirm the judgment of the district court—it is unnecessary, and hence would be improper, for us to consider the alternative grounds relied on by the district court.

IRI recognized as much, and, since it feared that the district court’s far-reaching — and by no means unassailable 1 — holding on assumption of risk might have collateral estoppel effects in other litigation arising out of September 11, IRI moved for partial vacatur of the district court’s decision on the assumption of risk issue. 2 Citigroup consented to IRI’s motion.

This court, by order dated May 24, 2006, ruled that IRI’s motion seeking partial vacatur was “more accurately an amendment to its appellate brief.... Thus, it should be heard, after full briefing by both parties, by the panel deciding the appeal.” Indus. Risk Insurers v. Port Auth. of N.Y. & N.J., et al., No. 05-0664-cv (2d Cir. May 24, 2006) (unpublished order). We treated IRI’s motion in this manner because IRI has not formally abandoned its appeal.

II. The Aegis Action

In a completely separate lawsuit in which both AMEC Construction Management, Inc. (AMEC) and Silverstein are parties, the assumption of risk issue was raised. See Aegis Ins. Servs., Inc. v. 7 World Trade Center Co., L.P. (In re Sept. 11 Property Damage and Business Loss Litigation), 481 F.Supp.2d 253 (S.D.N.Y. 2007) (“Aegi s”). In the Aegis action, Aegis Insurance Services (Aegis), the subrogee of Consolidated Edison Co., brought a gross negligence claim against Citigroup, on the theory that Citigroup was negligent in designing, constructing, installing, and maintaining its diesel-fuel-powered generator system in 7 WTC. Silverstein and Citigroup were both parties to the Aegis action, but AMEC initially was not. Silverstein, however, impleaded AMEC as a third-party defendant.

On March 9, 2007, Judge Hellerstein dismissed Silverstein’s third-party claims for two, independent, reasons: (1) that AMEC owed no duty to Silverstein because there was no contractual privity or its functional equivalent between the parties, id. at 260-61, and (2) that, as Judge Hellerstein had previously held in the Industrial Risk Insurers action, Silverstein assumed the risks associated with the Citigroup Fuel System, id. at 261-62.

III. The Motions to Intervene in the Industrial Risk Insurers Action

Shortly before Judge Hellerstein issued his ruling dismissing Silverstein’s claims against AMEC in the Aegis case, AMEC learned of IRI’s motion for partial vacatur in the Industrial Risk Insurers case. AMEC sought to intervene, because,

if vacatur was granted, it would cast a cloud of disapproval on Judge Heller-stein’s assumption of risk holding, perhaps influencing his consideration of AMEC’s then pending motion to dismiss in the Aegis Action. Additionally, AMEC was concerned that, because IRI and Citigroup no longer had any real stake in the assumption of risk issue on appeal, the arguments in favor of Judge Hellerstein’s ruling on that question might not be adequately presented to this Court by the parties to the appeal in the IRI Action....

Brief for AMEC Construction Management, Inc. as Amicus Curiae at 8, Indus. Risk Insurers v. Port Auth. of N.Y. & N.J., et al., No. 05-0664-cv (2d Cir. June 1, 2007). Soon thereafter, Silverstein also moved to intervene.

After AMEC and Silverstein had moved to intervene in the Industrial Risk Insurers case, Judge Hellerstein issued his ruling in the Aegis action. Because Judge Hellerstein offered two independent grounds for dismissing Silverstein’s third-party claims against AMEC — both lack of contractual duty, and assumption of risk— AMEC now concedes that “any influence that vacatur might have had prospectively on Judge Hellerstein’s decision is no longer a concern for AMEC.” AMEC Amicus Brief at 9.

IV.This Court’s May J, 2007 Order

On May 4, 2007, this court issued an order denying AMEC’s and Silverstein’s motions to intervene in the Industrial Risk Insurers case, but inviting them to file amicus briefs and participate in oral argument. We also noted that since we could affirm the district court’s judgment in the Industrial Risk Insurers case on its St. Paul holding alone, there was no need for us to review Judge Hellerstein’s assumption of risk holding in this appeal. Indus. Risk Insurers v. Port Auth. of N.Y. & N.J., et al., No. 05-0664-cv (2d Cir. May 4, 2007) (unpublished order). Accordingly, the only issue still in active dispute today is whether to grant IRI’s motion for partial vacatur.

DISCUSSION

Having carefully considered the parties’ arguments, we now affirm the judgment of the district court, doing so solely on the basis of its “subrogation waiver” holding, and turn to whether IRI’s motion for partial vacatur should be granted. We have recognized that “a party ‘who seeks review of the merits of an adverse ruling, but is frustrated by the vagaries of circumstance, ought not in fairness be forced to acquiesce in the judgment.’ ” Major League Baseball Props., Inc. v. Pacific Trading Cards, Inc., 150 F.3d 149, 151 (2d Cir.1998) (quoting U.S. Bancorp Mortgage Co. v. Bonner Mall P’ship, 513 U.S. 18, 25, 115 S.Ct. 386, 130 L.Ed.2d 233 (1994)). At the same time, we have acknowledged that “vacatur should be regarded as an extraordinary remedy because judicial precedents have social value.” Id.

AMEC argues, in its amicus brief to us, that IRI should not be entitled to partial vacatur of the district court’s opinion because, “[ajlthough it is clear, in light of this Court’s ruling in St. Paul, that IRI’s appeal stood virtually no chance of success, nevertheless the St. Paul decision in and of itself did not moot IRI’s appeal, as that decision did not require IRI to abandon its own appeal.” AMEC Amicus Brief at 11 (footnote omitted). This argument misconstrues IRI’s position. IRI’s motion for partial vacatur was effectively an amendment to its appellate brief; it was not an abandonment of the appeal. Indus. Risk Insurers v. Port Auth. of N.Y. & N.J., et al., No. 05-0664-cv (2d Cir. May 24, 2006) (unpublished order). Accordingly, this case squarely raises the question of whether a motion for partial vacatur should be granted where, as here, the moving party is unable to obtain review on the merits of one aspect of a district court’s holding because the holding is affirmed on other, independent, grounds.

The question is a difficult one. IRI powerfully contends that if its motion is not granted, it may be collaterally es-topped from arguing the merits of the assumption of risk issues in other cases. Indeed, that is precisely what happened to Silverstein, at the district court level, in Aegis. The district court there applied collateral estoppel to. bar Silverstein’s claims against Citigroup, finding that “the identical issue was necessarily decided in the prior action [Indus. Risk Insurers, 387 F.Supp.2d. at 299] and is decisive in the present action.” Aegis, 481 F.Supp.2d at 261-62 (citation omitted).

But, of course, Aegis notwithstanding, it is more than possible that collateral estoppel would not ultimately be applied, given that we have found it inappropriate to review the district court’s original assumption of , risk holding. Since we are affirming the district court’s decision solely on the “subrogation waiver” holding and refusing to consider its assumption of risk ruling, it would seem that IRI has not had the opportunity to contest fully the merits of the assumption of risk issue. And such an opportunity is a prerequisite for the application of collateral estoppel under both federal and New York state law. Compare Boguslavsky v. Kaplan, 159 F.3d 715, 720 (2d Cir.1998) (“Under federal law, a party is collaterally estopped from relitigating an issue if a four-part test is met: (1) the identical issue was raised in a previous proceeding; (2) the issue was actually litigated and decided in the previous proceeding; (3) the party had a full and fair opportunity to litigate the issue; and (4) the resolution of the issue was necessary to support a valid and final judgment on the merits.” (internal quotation marks and footnote omitted and emphasis added)) with Curry v. City of Syracuse, 316 F.3d 324, 331 (2d Cir.2003) (stating that “under New York law, collateral estoppel prevents a party from relitigating an issue.... (1) identical to an issue already decided (2) in a previous proceeding in which that party had a full and fair opportunity to litigate,” and where (3) “the issue that was raised previously [is] decisive of the present action” (internal quotation marks omitted and emphasis added)).

The collateral estoppel issue is, in this case, complicated by the fact that some doubt exists as to whether New York state law or federal law controls. In the Aegis action, for example, Judge Hellerstein appeared to apply New York law to the collateral estoppel inquiry, Aegis, 481 F.Supp.2d at 261-62, perhaps following “an old line of cases that has never been overruled, [in which] the Supreme Court has held that the states may apply their own law to determine the preclusive effect of a federal diversity judgment,” Gelb v. Royal Globe Ins. Co., 798 F.2d 38, 42 n. 3 (2d Cir.1986); see also United States v. Frank, 494 F.2d 145, 159 (2d Cir.1974) (rejecting, in dicta, the government’s contention that federal law would determine the collateral estoppel effect of a federal criminal conviction in a subsequent diversity case). On the other hand, more recent case law from our circuit suggests that federal law should have controlled, See Marvel Characters, Inc. v. Simon, 310 F.3d 280, 286 (2d Cir.2002) (“We apply federal law in determining the preclusive effect of a federal judgment and New York law in determining the preclusive effect of a New York State court judgment.” (internal citations omitted)); Gelb, 798 F.2d at 42 n. 3 (holding that federal law governed the collateral estoppel effect of a federal criminal conviction in a subsequent diversity action and suggesting that, in light of the Federal Rules of Civil Procedure and Supreme Court cases requiring federal courts to give state judgments collateral estoppel effects, “[i]t would seem that ... federal law should determine the preclusive effect-of a federal judgment without regard to the basis of jurisdiction”). See also the much earlier case of Kern v. Hettinger, 303 F.2d 333, 340 (2d Cir.1962) (applying federal law, not state law, to determine the character and extraterritorial effect of a federal district court judgment because “[i]t would be destructive of the basic principles of the Federal Rules of Civil Procedure to say that the effect of a judgment of a federal court was governed by the law of the state where the court sits simply because the source of federal jurisdiction is diversity”).

In the end, however, we need not decide whether New York or federal law controls the question. This is so despite the fact that — -although they often ultimately reach the same result — these two bodies of law do appear to diverge in some particulars, see Curry, 316 F.3d at 331 n. 4 (noting that federal law and New York state law differ on the issue of collateral estoppel), and despite the fact that we cannot be sure that they are the same with regard to the situation before us. In Gelb, we held that under federal law, “if an appeal is taken and the appellate court affirms on one ground and disregards the other, there is no collateral estoppel as to the unreviewed ground,” and that “inability to obtain appellate review, or the lack of such review once an appeal is taken, does prevent preclusion.” 798 F.2d at 44-45. But, we have found no equivalent New York holding. 3

Nevertheless, we believe it clear that applying collateral estoppel in-later cases to a holding such as Judge Hellerstein’s assumption of risk decision would be highly problematic, under either New York or federal law. But the matter is not precisely before us, and hence it has not been fully argued to us. On the other hand, the merits of the motion for vacatur of the assumption of risk holding, which have been argued to us, do depend in part on the existence vel non of collateral estoppel. We hesitate to vacate the district court’s learned consideration of assumption of risk (whatever our own doubts about it may be) as it is entitled to have persuasive effect on district courts in later cases. We cannot, however, countenance its remaining in place, if it were to estop collaterally parties who have had no chance to contest it on appeal.

Under the circumstances, we deem it best to allow the district court to consider, in the first instance, whether the motion to vacate should be granted. As the Supreme Court has made clear,

[E]ven in the absence of, or before considering the existence of, extraordinary circumstances [that might warrant vaca-tur of a district court’s opinion], a court of appeals presented with a request for vacatur of a district-court judgment may remand the case with instructions that the district court consider the request, which it may do pursuant to Federal Rule of Civil Procedure 60(b).

U.S. Bancorp Mortgage Co. v. Bonner Mall P’ship, 513 U.S. 18, 29, 115 S.Ct. 386, 130 L.Ed.2d 233 (1994).

Accordingly, we remand and instruct the district court consider IRI’s motion for partial vacatur. In doing so, we underscore that, if the district court, on remand, construes its ruling on assumption of risk in such a way that IRI cannot in other cases be collaterally estopped by it, then the need to grant the motion to vacate would be significantly reduced. Similarly, the reverse is true: if the assumption of risk ruling is such that collateral estoppel might well .apply, that fact would likely suffice as an extraordinary reason justifying the granting of a vacatur motion. But we leave these issues for the district court.

CONCLUSION

The judgment of the district court is Affirmed, though solely with respect to its “subrogation waiver” holding. IRI’s motion for partial vacatur is Remanded to the district court, for that court’s consideration in the first instance in light of the above discussion.

Notes

1

. The district court was interpreting New York state law. We think that New York law with respect to assumption of risk in situations like those in the instant case is anything but certain. As a result we believe that, in the proper case, guidance on this issue should be sought from the New York Court of Appeals through certification. We recognize, of course, that under New York law certification was not an option for the district court. See O’Mara v. Town of Wappinger, 485 F.3d 693, 698 & n. 7 (2d Cir.2007) (noting that while under Second Circuit rules and New York law, our court may certify a question to the New York Court of Appeals, district courts do not have the authority to do so); Nicholson v. Scoppetta, 344 F.3d 154, 168 (2d Cir.2003) ("On appeal, we now have available to us an additional option not open to the District Court: We may certify questions of New York state law to the New York Court of Appeals.”).

2

. IRI did not seek vacatur of the district court's separate conclusion that Citigroup’s conduct, as a matter of law, was not gross negligence. Presumably this was because that holding would have no collateral consequences for IRI in other litigation.

3

. There is authority in New York that collateral estoppel may apply to a lower court decision pending appeal. See, e.g., DiSorbo v. Hoy, 343 F.3d 172, 183 (2d Cir.2003) (“Under New York law, the mere pendency of an appeal does not prevent the use of the challenged judgment as the basis of collaterally estopping a party to that judgment in a see-ond proceeding.”) (internal quotation marks and citation omitted). While there are many New York cases suggesting caution in applying preclusion in such situations, see, e.g., Duverney v. State, 96 Misc.2d 898, 410 N.Y.S.2d 237, 246 (N.Y.Ct.Cl.1978) (noting that under New York’s collateral estoppel law, "the full and fairness opportunity doctrine ... mandates consideration of the existence of an appeal”); In re Estate of Alexis, 14 Misc.3d 379, 823 N.Y.S.2d 886, 889-90 (N.Y.Sur.Ct. 2006) ("In determining whether collateral estoppel should apply, it has been recognized that consideration should be given to the existence of an appeal.”); In re Estate of Brown, 132 Misc.2d 171, 503 N.Y.S.2d 532, 533 (N.Y.Sur.Ct. 1986) (holding proceeds of a life insurance policy in escrow, rather than applying collateral estoppel, pending finalization of a judgment against the named beneficiary for murdering the insured), and while we have said that under New York law “collateral estoppel is an equitable doctrine — not a matter of absolute right,” PenneCom B.V. v. Merrill Lynch & Co., 372 F.3d 488, 493 (2d Cir. 2004), the law as to the instant situation is uncertain. We have found only one New York case directly on point, Tydings v. Greenfield, Stein & Senior, LLP, No. 110963/06, 14 Misc.3d 1233, 836 N.Y.S.2d 495, slip op. at 2 (N.Y.Sup.Ct. Jan. 12, 2007) (precluding the plaintiff from raising an issue that had been litigated in Surrogate Court and appealed, but that had not been addressed in the First Department’s affirmance of the Surrogate Court's decision). This opinion, unpublished and by a court of first instance, obviously does not resolve the question of New York law on the matter — quite apart from’the fact that it appears to confuse the legitimacy of a judgment and its effect under New York law prior to an appeal with the collateral estoppel effect of a judgment after an appeal.