Gould Electronics Inc. v. United States

220 F.3d 169, 2000 WL 1053828

• Court: Court of Appeals for the Third Circuit
• Decided: July 31, 2000
• Docket: 99-1893
• Status: Unknown

Opinion

OPINION OF THE COURT

SCHWARTZ, Senior District Judge

I. INTRODUCTION

Plaintiffs/appellants Gould Electronics, Inc. (“Gould”) and American Premier Underwriters, Inc. (“APU”) were co-defendants in a toxic tort case captioned Cheryl Allen, et al. v. Marathon Battery Co., et al., No. 1074/90 (N.Y.Sup.Ct.) (the “Allen case”). The Allen case arose out of personal injuries and property damage allegedly caused by air and water pollution from a battery manufacturing plant in Cold Spring, New York. The plant was designed, constructed, owned, and operated by the United States Army, via its office in Philadelphia, Pennsylvania, and owned and operated by Gould and APU, and their predecessors, at various times. The Army was not a party to the Allen litigation. Gould and APU settled the Allen case for $4.5 million.

Gould and APU filed a Complaint against defendant/appellee United States of America (“United States”), in the Eastern District of Pennsylvania, seeking contribution and indemnity because of their entry .into the Allen case settlement, pursuant to the Federal Tort Claims Act (“FTCA”), 28 U.S.C. §§ 1346(b), 2671 et seq. The District Court granted the United States’ Motion to Dismiss for lack of subject matter jurisdiction, pursuant to Fed.R.Civ.P. 12(b)(1) (“Rule 12(b)(1)”), holding: (1) under Pennsylvania choice of law rules, New York contribution and indemnity law governs the FTCA jurisdictional inquiry; (2) the court lacks jurisdiction over the contribution claim because the United States would not be liable for contribution under N.Y. Gen.Oblig. Law § 15-108(c), which bars settling parties from bringing contribution claims against non-settling parties (hereinafter “§ 15-108(c)”); and (3) the court lacks jurisdiction over the indemnification claim because the United States would not be liable for indemnification under New York law, which bars indemnification when the plaintiff is at least partially at fault.

Gould and APU contend the District Court erred by: (1) misapplying the standards governing a motion to dismiss under Rule 12(b)(1); (2) holding New York, rather than Pennsylvania, contribution and indemnity law governs the jurisdictional inquiry, under Pennsylvania choice of law rules; (3) holding the United States would not be liable for contribution under § 15- 108(c) because the United States waived the protection of § 15 — 108(c); and (4) holding the United States would not be liable for indemnity under New York law.

We find the District Court properly applied the standards used for analyzing a Rule 12(b)(1) motion to dismiss, but erred in determining New York contribution and indemnity law controls the outcome. Rather we hold that Ohio law governs the jurisdictional inquiry and, under Ohio law, the United States would be liable for contribution, but not indemnity. 1 As such, the District Court has subject matter jurisdiction over Gould/APU’s FTCA claim for contribution, but not for indemnity. The District Court’s ruling is AFFIRMED IN PART and REVERSED IN PART and the case is REMANDED for proceedings consistent with this opinion. 2

II. FACTS

The plaintiffs/appellants are Gould, an Ohio corporation with its principal place of business in Ohio, and APU, a Pennsylvania corporation with its principal place of business in Ohio (hereinafter “Gould/APU”). Defendant/appellee is the United States.

In 1951, the Signal Corps of the United States Army (“Army”), through its office in Philadelphia, Pennsylvania, entered into a Letter Contract and Facilities Contract with Sonotone Corp. (“Sonotone”), a predecessor corporation to Gould. Pursuant to these contracts, the United States agreed to design and construct a battery manufacturing plant to be located on government property in Cold Spring, New York. Sonotone was to assist in the design of the plant and then operate the plant on a contract basis to produce batteries for the government. However, the United States retained ultimate supervision and control over the day-to-day operations of the plant.

Between 1951 and 1952, the Army designed the plant, including its industrial waste water disposal and air emissions systems. For waste water disposal, Sono-tone recommended a closed system to allow removal of hazardous material before releasing waste water from the plant. The Army rejected the closed design, opting for an opén system which did not remove hazardous material from the waste water. By January, 1953, the Army had caused the construction of the plant, including the waste water and air emission systems.

From 1953 to 1962, Sonotone acted as contractor-operator of the plant, which produced nickel-cadmium batteries. During this time period, according to the Complaint, the Army owned and retained ultimate supervision and control over the plant, including:

(a) Title to all real and personal property remained with the government;

(b) All equipment was to be installed by the government;

(c) Title to all materials, supplies, work-in-process and other property vested with the government;

(d) The Plant was to be used solely to fulfill government contracts for an initial five year period;

(e) The government was to reimburse Sonotone for all repairs, replacements and restorations “in excess of normal requirements for maintenance and in excess of fair wear and tear.” Such- reimbursed expendi tures were required to be pre-ap-proved by the government;

(f) The government was provided with access to the Plant at all times;

(g) The Plant and facilities were to be erected, made available, delivered and installed by the government;

(h) The government reserved to itself the right to dismantle, remove and ship the Plant and facilities when deemed in the best interest of the government to do so; and

(i) The government reserved to itself the right to terminate use of the facility when the government determined it was in its best interests to do so.

Complaint P 19. During this time, the plant discharged industrial waste water into Foundry Cove and the Hudson River and discharged contaminated dust and vapors into the air surrounding the plant.

In 1962, the Army sold the plant to Sonotone. From 1962 to 1969, Sonotone continued to operate the plant as a battery manufacturing plant and continued to discharge industrial waste water into Foundry Cove. 3 In 1969, Sonotone sold the plant to Business Funds, Inc., which through a series of corporate mergers over the next several years, became Marathon Battery Co. and then plaintiff APU. From 1969 to 1979, APU owned and operated the plant, continued to produce batteries, and continued to discharge industrial waste water into Foundry Cove. 4 In 1979, APU sold the plant to Merchandise Dynamics, Inc., which ceased manufacturing batteries at the plant.

In 1990, residents of Cold Spring, New York filed the Allen lawsuit against, inter alia, Gould/APU in the Supreme Court of New York. The Allen plaintiffs alleged negligence and strict liability claims arising out of injuries caused by air and water pollution released from the plant. The United States was not named as a defendant and could not be joined as a co-defendant because of its sovereign immunity in state court. Gould/APU attempted to remove the case to the United States District Court for the Southern District of New York in order to join the United States as a defendant, but the request was denied.

In 1991, the United States Environmental Protection Agency (“EPA”) listed the plant and surrounding area (the “Site”) on the National Priorities List for the Federal Superfund Program under the Comprehensive Environmental Response, Compensation, and Liability Act (“CERCLA”), 42 U.S.C. § 9601 et seq. In 1991, the EPA entered into a first consent decree (“First CERCLA Consent Decree”) with Gould/ APU and the Army providing for clean up of the Site. The First CERCLA Consent Decree contained a provision providing:

N. All Parties reserve all rights and legal obligations with respect to any toxic tort claims including, but not limited to, all claims asserted in Cheryl Allen, et al. v. Marathon Battery Co., et al., Index No., 1074/90 (N.Y.Sup.Ct.).

Joint Appendix (“App.”) 648a.

In 1993, the EPA entered hilo a second consent decree (“Second CERCLA Consent Decree”) with Gould/APU and the Army covering the Site. The Second CERCLA Consent Decree contained, inter alia, two provisions providing:

5. Settling Parties specifically reserve and do not hereby waive any defenses which they may have with respect to any asserted liability related to the Site. Settling Parties reserve all rights, defenses, and legal contentions with respect to any third party claims, including, but not limited to, all claims asserted in Cheryl Allen et al. v. Marathon Battery Co., et al.....

108. Settling Parties reserve, and this Consent Decree is without prejudice to, (i) claims in the nature of contribution among Settling Parties which may arise from toxic tort claims, including those related to the pending action in Cheryl Allen et al. v. Marathon Battery Co., et al .... (iv) actions against the United States based on negligent actions taken directly by the United States ... that are brought pursuant to any statute other than CERCLA and for which the waiver of sovereign immunity is found in a statute other than CERCLA.

App. 549a-550a, 628a-629a.

In 1997, Gould/APU settled the Allen litigation for $4.5 million. In 1998, Gould/ APU filed administrative claims with the Army seeking contribution and indemnity for the $4.5 million settlement. After these claims were finally denied, Gould/ APU filed this lawsuit against the United States.

III. STANDARD OF REVIEW

The standard of review is plenary where the District Court dismisses for lack of subject matter jurisdiction. See Dresser Industries, Inc. v. Underwriters at Lloyds of London, 106 F.3d 494, 496 (3d Cir.1997).

IV. DISCUSSION

A. Standards for a Rule 12(b)(1) Motion

Gould/APU contend the District Court erred in applying the standards for evaluating a Rule 12(b)(1) motion by: (1) considering evidence outside of the pleadings; (2) dismissing the case on an inadequate factual record; and (3) improperly treating the Rule 12(b)(1) motion as a Fed.R.Civ.P. 12(b)(6) (“Rule 12(b)(6)”) motion, for failure to state a claim, and ruling on the merits of the case. 5 We find the District Court did not err in any of these three respects.

1. Considering Evidence Outside of the Pleadings

Gould/APU argue “[b]ecause there were no affidavits, depositions, or testimony from which the district court could adduce facts outside of the Complaint, it should have accepted as true the facts set forth in the Complaint and not looked beyond them.” Brief of Appellants at 7. A Rule 12(b)(1) motion may be treated as either a facial or factual challenge to the court’s subject matter jurisdiction. See Mortensen v. First Fed. Sav. and Loan Ass’n, 549 F.2d 884, 891 (3d Cir.1977). In reviewing a facial attack, the court must only consider the allegations of the complaint and documents referenced therein and attached thereto, in the light most favorable to the' plaintiff. See id.; PBGC v. White, 998 F.2d 1192, 1196 (3d Cir.1993). 6 In reviewing a factual attack, the court may consider evidence outside the pleadings. See Gotha v. United States, 115 F.3d 176, 178-79 (3d Cir.1997) (citing Mortensen, 549 F.2d at 891).

Although not explicitly stated in its opinion, the District Court treated the United States’ motion as a factual attack. In addition to the allegations in the Complaint and the documents referenced in the Complaint and attached thereto, the District Court considered the CERCLA Consent Decrees, which were not referenced in or attached to the Complaint. By treating the motion as a factual attack, the District Court properly considered evidence beyond the pleadings.

2. Adequacy of Factual Record

In a closely related argument Gould/APU contend the District Court erred by ruling on an inadequate factual record instead of allowing Gould/APU “to develop the record and introduce additional relevant facts supporting jurisdiction.” Brief of Appellants at 8. In International Ass’n of Machinists & Aerospace Workers v. Northwest Airlines, Inc., this Court outlined procedures for ensuring that a ruling on a Rule 12(b)(1) factual attack be based on an adequate factual record. See 673 F.2d 700, 711-12 (3d Cir.1982). If the defendant raises no challenge to the facts alleged in the pleadings, the court may rule on the motion by accepting the allegations as true. See id. at 711. If the defendant contests any allegations in the pleadings, by presenting evidence, the court must permit the plaintiff to respond with evidence supporting jurisdiction. See id. at 711-12. The court may then determine jurisdiction by weighing the evidence presented by the parties. See id. However, if there is a dispute of a material fact, the court must conduct a plenary trial on the contested facts prior to making a jurisdictional determination. See id. at 712.

As to all issues, except for waiver of § 15-108(c), the United States did not challenge the truthfulness of any of the allegations in the Complaint or the attached documents. See App. 436a (“it appears that, based on Gould/APU’s own allegations, their case is not within this Court’s jurisdiction”); id. at n. 1 (“This statement of facts ... is taken entirely from the Gould/APU Complaint”). As to all issues, except for waiver of § 15-108(c), the District Court based its decision on accepting as true the allegations in the Complaint and the documents attached thereto. Since no facts were disputed, except waiver of § 15 — 108(c), the Court properly accepted as true all allegations of the Complaint and documents attached thereto and ruled on an adequate factual record.

As to the waiver issue, Gould/APU asserted the Second CERCLA Consent Decree operated as a waiver, by the United States, of the bar against contribution claims under N.Y. Gen. Oblig. Law § 15-108(c). 7 Gould/APU attached to its answering brief filed in the District Court a copy of a portion of the Second CERCLA Consent Decree. In its reply brief, the United States attached a copy of both CERCLA Consent Decrees in their entirety. The District Court considered the CERCLA Consent Decrees in evaluating whether it had jurisdiction.

The District Court properly considered the CERCLA Consent Decrees and did not need to take any additional evidence on the waiver issue. The allegations in the Complaint and attached documents did not relate to this issue. The CERCLA Consent Decree did not put into issue any of the allegations of the Complaint, but merely supplemented them. In addition, there was no dispute of any material fact in the CERCLA Consent Decrees. Rather, both parties agreed on the content of the CERCLA Consent Decrees, leaving the District Court to merely interpret the meaning of the relevant provisions. Finally, Gould/APU did not proffer any additional evidence relevant to this issue in the District Court. Accordingly, the Court evaluated the waiver issue based on an adequate factual record, consisting of the allegations in the Complaint, the documents attached thereto, and the CERCLA Consent Decrees.

Finally, Gould/APU’s request for additional discovery was properly denied as unnecessary and unwarranted because the United States did not contest the allegations in the Complaint. Moreover, while Gould/APU have argued to this Court that they be allowed to supplement the factual record, they failed to do so before the District Court, thereby waving the issue at the appejlate level.

3. Treating a Rule 12(b)(1) Motion as a Rule 12(b)(6) Motion

Gould/APU contend the District Court improperly treated the Rule 12(b)(1) motion to dismiss for lack of subject matter jurisdiction as a Rule 12(b)(6) motion to dismiss for failure to state a claim by improperly considering the merits of the case instead of assessing the court’s jurisdiction. This Court has previously cautioned against treating a Rule 12(b)(1) motion as a Rule 12(b)(6) motion and reaching the merits of the claims. See Kehr Packages, Inc. v. Fidelcor, Inc., 926 F.2d 1406, 1409 (3d Cir.1991). This concern arises because the standard for surviving a Rule 12(b)(1) motion is lower than that for a Rule 12(b)(6) motion. See id.

In a Rule 12(b)(6) motion, the court evaluates the merits of the claims by accepting all allegations in the complaint as true, viewing them in the light most favorable to the plaintiffs, and determining whether they state a claim as a matter of law. See In re Burlington Coat Factory Securities Litigation, 114 F.3d 1410, 1420 (3d Cir.1997). The defendant bears the burden of showing no claim has been stated. See Kehr Packages, 926 F.2d at 1409. In contrast, in a factual attack under Rule 12(b)(1), the court may consider and weigh evidence outside the pleadings to determine if it has jurisdiction. See Mortensen, 549 F.2d at 891. The plaintiff has the burden of persuasion to convince the court it has jurisdiction. See Kehr Packages, 926 F.2d at 1409. A claim may be dismissed under Rule 12(b)(1) only if it “clearly appears to be immaterial and made solely for the purpose of obtaining jurisdiction” or is “wholly insubstantial and frivolous.” Id. (internal quotations omitted).

This claim under the FTCA is somewhat unique because the merits of the case are closely intertwined with the jurisdictional proof — both are determined by evaluating whether the United States would be liable “in accordance with the law of the place where the act or omission occurred.” 28 U.S.C. § 1346(b)(1). However, when the merits and jurisdiction are closely related, a court may determine subject matter jurisdiction without reaching the merits, so long as the court “demand[s] less in the way of jurisdictional proof than would be appropriate at a trial stage.” Mortensen, 549 F.2d at 891 (permitting evaluation of jurisdiction for claim under Sherman Act where merits and jurisdiction closely intertwined); see also Gotha, 115 F.3d at 178-79.

In Gotha, the defendant asserted the court lacked jurisdiction under the FTCA because the claim fell under the discretionary function exception to the FTCA. See Gotha, 115 F.3d at 178-79. This case involves a determination of jurisdiction under the FTCA based on whether the United States would be liable “in accordance with the law of the place where the act or omission occurred.” 28 U.S.C. § 1346(b)(1). Although a different provision of the FTCA is implicated, we see no principled reason to distinguish between a jurisdictional determination based on the discretionary function exception and one based on a conflict of law analysis. Indeed, the Second Circuit Court of Appeals has allowed jurisdictional determinations, apart from merits consideration, based upon § 1346(b)(1). See Makarova v. United States, 201 F.3d 110, 113 (2d Cir.2000). Adhering to our precedent, a court may determine subject matter jurisdiction over an FTCA claim, involving a conflict of laws analysis, so long as it demands less in the way of jurisdictional proof than it would for a ruling on the merits.

The District Court did just that. First, it rehearsed the standards for a Rule 12(b)(1) motion and performed its analysis under these standards. 8 Next, the District Court performed a conflict of laws analysis under the FTCA by weighing the evidence before it, accepting as true all allegations in the Complaint, the documentary evidence attached thereto, and the CERCLA Consent Decrees. Finally, the District Court did not demand an inordinate amount of jurisdictional proof. Having done precisely what is required, the District Court did not err by implicitly treating the Rule 12(b)(1) motion as a Rule 12(b)(6) motion.

B. Conflict of Laws

The FTCA waives sovereign immunity and grants district courts jurisdiction over tort claims against the United States “under circumstances where the United States, if a private person, would be hable to the claimant in accordance with the law of the place where the act or omission occurred.” 28 U.S.C. § 1346(b)(1) (emphasis added). In order to determine whether it has jurisdiction, the court must evaluate whether the United States would be hable under the “whole law” of the state in which the act or omission occurred. See Richards v. United States, 369 U.S. 1, 11, 82 S.Ct. 585, 7 L.Ed.2d 492 (1962); Small v. United States, 333 F.2d 702, 704 (3d Cir.1964). When a case involves multiple alleged acts or omissions occurring in more than one state, the FTCA, as construed by Richards, requires the District Court to engage in a complex conflict of laws analysis to determine which state law governs the jurisdictional inquiry. The Complaint alleges acts or omissions occurring in New York and Pennsylvania. 9

The District Court determined that most of the acts or omissions occurred in New York, but it noted that even if some occurred in Pennsylvania, its choice of law rules would indicate that New York had the more significant interest in the outcome of the litigation and, therefore, New York contribution and indemnity law should govern the jurisdictional inquiry. See id. at 4-5, 82 S.Ct. 585. Gould/APU contend the District Court erred in determining New York, and not Pennsylvania, contribution and indemnification law governs the jurisdictional inquiry, under Pennsylvania choice of law. 10 Thus, in the District Court and as originally briefed before this Court, the litigants differed over application of Pennsylvania choice of law rules, rather than addressing which state’s choice of law governs. 11

Because Richards interpreted the “law of the place where the act or omission occurred” to mean the “whole law” of the state where the act or omission occurred, including that state’s choice of law rules, a two step choice of law analysis is required when multiple acts or omissions have occurred in more than one state. First, the court must select between the states’ respective choice of law rules. See Richards, 369 U.S. at 11, 82 S.Ct. 585; Ducey v. United States, 713 F.2d 504, 509 n. 2 (9th Cir.1983); Bowen v. United States, 570 F.2d 1311, 1318 (7th Cir.1978); James A. Shapiro, Choice of Law Under the Federal Tort Claims Act: Richards and Renvoi Revisited, 70 N.C.L.Rev. 641, 669-75 (1992). Second, the court must apply that state’s choice of law rules to determine which state’s substantive tort law applies. See Richards, 369 U.S. at 11, 82 S.Ct. 585; Tyminski v. United States, 481 F.2d 257, 265 (3d Cir.1973). Because of this bifurcated analysis, the state whose choice of law rules are selected in the first step may or may not be the same state whose substantive law is chosen in the second step.

Before proceeding to the conflict of laws analysis, it is prudent to ensure that there is not a “false conflict” in the underlying choice of law rules or the underlying contribution and indemnity law making it unnecessary to engage in this complex bifurcated analysis. See Williams v. Stone, 109 F.3d 890, 893 (3rd Cir.1997). In the first step, Pennsylvania and New York, the two possible jurisdictions where acts or omissions occurred, differ in their choice of law rules. Pennsylvania employs an interest analysis, 12 while New York uses so-called Neumeier rules which focus on the parties’ domiciles. 13 In the second step, four possible jurisdictions’ substantive law may apply: New York (where acts or omissions occurred), Pennsylvania (where acts or omissions occurred), Ohio (domicile of Gould and APU), and the District of Columbia (possible domicile of the United States). As to the contribution claim, there is a conflict between New York law on one hand and Pennsylvania, Ohio, and District of Columbia law on the other hand, which necessitates a conflict of laws analysis. Compare N.Y. Gen. Oblig. Law § 15-108(e) 14 (barring contribution claims by settling tortfeasors against non-settling joint tortfeasors) with 42 Pa.Cons.Stat. Ann. § 8324; 15 Swartz v. Sunderland, 403 Pa. 222, 169 A.2d 289, 291 (1961) (allowing contribution claims by settling tortfeasors against non-settling tortfeasors); Ohio Stat. § 2307.31; 16 MetroHealth Medical Center v. Hoffmann-LaRoche, Inc., 80 Ohio St.3d 212, 685 N.E.2d 529, 532 (1997) (same); District of Columbia v. Washington Hosp. Ctr., 722 A.2d 332, 342 & n. 12-14 (D.C.1998) (and cases cited therein) (although leaving question unresolved, would likely allow contribution claims by settling tortfeasors against non-settling tortfea-sors). However, there is no conflict between the indemnification law of New York, Pennsylvania, Ohio, and the District of Columbia. Each bars indemnification where the party seeking indemnification is partially at fault. See Rosado v. Proctor & Schwartz, Inc., 66 N.Y.2d 21, 494 N.Y.S.2d 851, 484 N.E.2d 1354, 1356-57 (1985); Builders Supply Co. v. McCabe, 366 Pa. 322, 77 A.2d 368, 371 (1951); Motorists Mut. Ins. Co. v. Huron Road Hospital, 73 Ohio St.3d 391, 653 N.E.2d 235, 238 (1995); Quadrangle Development Corp. v. Otis Elevator Co., 748 A.2d 432, 435 (D.C.2000). Accordingly, we conduct the conflict of laws analysis to determine only which state’s contribution law governs.

1. Step 1: Selection of Choice of Law Under FTCA Where Acts or Omissions Occurred in More Than One State

Because multiple acts or omissions are alleged to have occurred in New York and Pennsylvania, the Court must elect between Pennsylvania and New York choice of law rules. The FTCA, as interpreted by Richards, requires the Court to select the choice of law rules of the state where the “acts or omissions” occurred, not where the injury occurred. See 28 U.S.C. § 1346(b)(1) (United States liable for “injury or loss of property ... in accordance with the law of the place where the act or omission occurred”) (emphasis added); Richards, 369 U.S. at 11-13, 82 S.Ct. 585. Neither the text of the FTCA nor Richards provides any guidance on how to choose between conflicting choice of law rules when the alleged acts or omissions occur in more than one state. Moreover, the legislative history of the FTCA sheds no light on this problem. The selection of what choice of law rules to apply at the first step of the analysis has been largely undeveloped. However, where courts have spoken, they have adopted five approaches.

In the first approach, when the injury can be parsed by the acts or omissions in the different states, one court applied the choice of law rules on an act-by-act basis, applying the relevant state’s choice of law rules for each act or omission. See Kohn v. United States, 591 F.Supp. 568, 572 (E.D.N.Y.1984). In Kohn, the plaintiffs sued the United States Army for emotional distress caused by acts or omissions occurring in Kentucky and New. York, relating to the treatment of a family member’s corpse. See id. at 571. In Kentucky, plaintiffs alleged the Army performed an unauthorized autopsy, failed to return organs to the body for burial, embalmed the body, and cremated missing organs. See id. In New York, plaintiffs alleged the Army communicated inaccurate and misleading information about the circumstances of the death, and failed to provide an honor guard at the burial. See id. Because each act by the Army was a distinct tort that, absent the others, could have caused an emotional distress injury, the Court applied Kentucky choice of law to the acts in Kentucky and New York choice of law to the acts in New York. See id. at 572. The Kohn approach is unworkable in this case because the Allen plaintiffs’ injuries are indivisible and cannot be parsed based on the alleged acts by the United States.

A second and a third approach were outlined in Bowen, supra, which held the court should elect the choice of law rales of “the place of the last act or omission having a causal effect” or “the place of the act or omission having the most significant causal effect.” Bowen, 570 F.2d at 1318. 17 In Bowen, an airplane en route from Arkansas to Indiana crashed in Indiana because of icé on its wings. See id. at 1314. The plaintiff sued the Federal Aviation Administration for failure to advise the pilot of icy conditions at various points along the route in Arkansas, Illinois, and Indiana. See id. While expressing a preference for the latter “most significant causal effect approach,” the court declined to decide which approach to take because both the last act having causal effect and the act with the most significant causal effect occurred in Indiana, where the plane crashed. See id. at 1318 & n. 14. Accordingly, the court applied Indiana choice of law rales. See id. at 1318.

In this case, either of the two Bowen criteria result in the selection of New York choice of law rules. Under the first Bowen approach, the last act or omission which caused the Allen plaintiffs’ injuries was the operation of the plant resulting in discharge of hazardous waste into the water and air. It is undisputed that the operation and discharge occurred in New York. 18 It follows, under the first Bowen approach, the Court should apply New York choice of law rules.

Under the second Bowen approach, the acts or omissions having the most significant causal effect on the Allen plaintiffs’ injuries were either constructing the plant or operating the plant to discharge hazardous waste into the water and air. According to the undisputed allegations in the Complaint, the United States constructed the plant in New York. Also, the United States operated the plant causing the discharge of hazardous waste, in New York. See, supra, note 18 and accompanying text. Since both of the “most significant” acts occurred in New York, under this approach, it is appropriate to apply New York choice of law rules.

Under a fourth approach, the court selects the choice of law rules of the state in which “physical acts” could have prevented the injury. See Ducey, 713 F.2d at 509 n. 2. In Ducey, plaintiffs brought an FTCA claim alleging injuries arising from the United States’ failure to place warning signs and erect fences at a park in Nevada, stemming in part from decisions made at a government office in California. See id. The Court elected Nevada choice of law rules because the injury could have been prevented by physical acts in Nevada. See id. Likewise, the Allen plaintiffs’ injuries were caused by the discharge of hazardous waste at a plant in New York, which the United States owned, constructed, and controlled its operation, 19 stemming from decisions made at a government office in Pennsylvania. Under this approach, application of New York choice of law rules is appropriate because New York is where physical acts could have prevented the injuries.

A fifth approach was taken by the District of Columbia Circuit Court of Appeals, which made a choice of choice of law based on where the “relevant” act or omission occurred. See Hitchcock v. United States, 665 F.2d 354, 359 (D.C.Cir.1981). In Hitchcock, the plaintiff was injured by a vaccine administered by a government nurse in Virginia, but where the protocol for its use was developed by government officials in the District of Columbia. See id. The Court applied District of Columbia choice of law because, although the vaccine was administered in Virginia, no relevant act occurred in Virginia. The nurse was the only staff member at the Virginia facility and was given no instructions about the protocol for administering the vaccine or the risks from administration. See id. According to Hitchcock, the only relevant negligent acts and omissions were the development of the protocol for administration performed by officials in the District of Columbia. See id. In contrast, in this case, the relevant acts of ownership, construction, and control over operation of the plant were performed by the United States Army in New York. Under this approach, it is appropriate to select the choice of law rules of New York.

In conclusion, there are five approaches to the election of choice of law. The Kohn approach is unworkable. However, the Court need not decide whether to adopt the second, third, fourth, or fifth approach because each leads to application of New York choice of law rules. Under these approaches, the last causal act, the most significant causal act, physical acts to prevent injury, and the relevant acts all occurred in New York. It follows, the District Court should have applied New York choice of law rules.

2. Step 2: Application of New York Choice of Law Rules

New York choice of law rules differ depending on whether the underlying substantive law is conduct regulating or loss allocating. See Padula v. Lilarn Properties Corp., 84 N.Y.2d 519, 521-22, 620 N.Y.S.2d 310, 644 N.E.2d 1001 (N.Y.1994). Conduct regulating law has “the prophylactic effect of governing conduct to prevent injuries from occurring.” Id. at 522, 620 N.Y.S.2d 310, 644 N.E.2d 1001. Loss allocating law “prohibit[s], assign[s], or limit[s] liability, after the tort occurs.” Id. Contribution law is loss allocating because it assigns or limits liability after the tort occurs. See id. (citing Cooney v. Osgood Mach., 81 N.Y.2d 66, 595 N.Y.S.2d 919, 612 N.E.2d 277, 280 (1993)); Ray v. Knights, 194 A.D.2d 131, 605 N.Y.S.2d 536, 538 (1993). 20

Under New York choice of law, there are three rules for choosing between loss allocating law, all keyed to the domicile of the litigants. See Neumeier v. Kuehner, 31 N.Y.2d 121, 335 N.Y.S.2d 64, 286 N.E.2d 454, 457-58 (1972); Cooney, 595 N.Y.S.2d 919, 612 N.E.2d at 281. First, when the plaintiff and defendant share a common domicile, the law of that state controls. Second, when the plaintiff and defendant are not domiciled in the same state (a “split domicile”), the states have conflicting law, and the tort occurs in one of the domiciles, the law of the state where the tort occurred controls. Third, in all other situations where there is a split domicile, the law of the state where the tort occurred controls, unless “it can be shown that displacing that normally applicable rule will advance the relevant substantive law purposes without impairing the smooth working of the multistate system or producing great uncertainty for litigants.” Neumeier, 335 N.Y.S.2d 64, 286 N.E.2d at 458. Originally developed for guest statutes, these three Neumeier rules have been extended to choice of law for all loss allocating law. See Cooney, 595 N.Y.S.2d 919, 612 N.E.2d at 281; Schultz v. Boy Scouts of Am., Inc., 65 N.Y.2d 189, 491 N.Y.S.2d 90, 480 N.E.2d 679, 684 (1985).

The New York choice of law analysis is performed in two steps. First, in order to choose the correct Neumeier rule, we determine the domicile of the parties. Second, based on the domiciles of the parties, we choose and apply the appropriate Neumeier rule.

a. Domicile of the Parties

For purposes of New York choice of law, a corporation is domiciled where it has its principal place of business. 21 See Schultz, 491 N.Y.S.2d 90, 480 N.E.2d at 682; Morgan Guar. Trust Co. of N.Y. v. Garrett Corp., 625 F.Supp. 752, 754, 760 & n. 11 (S.D.N.Y.1986); In re DES Cases, 789 F.Supp. 552, 590 (E.D.N.Y.1992). Gould, an Ohio corporation with its principal place of business in Ohio, is domiciled in Ohio. APU, a Pennsylvania corporation with its principal place of business in Ohio, is also domiciled in Ohio.

There are three possibilities for the domicile of the United States, for purposes of New York choice of law. First, the parties and the District Court assumed the United States is domiciled in all 50 states. See United States v. Whitcomb, 314 F.2d 415 (4th Cir.1963) (under Maryland law, United States domiciled in all states) (cit ing Helvering v. Stockholms Enskilda Bank, 293 U.S. 84, 91-93, 55 S.Ct. 50, 79 L.Ed. 211 (1934); Vaughan v. Northup, 40 U.S. (15 Pet.) 1, 6, 10 L.Ed. 639 (1841)), but see Ward Electronic Services, Inc. v. Property & Cas. Ins. Guar. Corp., 325 Md. 1, 599 A.2d 81, 84 (1991) (the United States is not domiciled in all 50 states under Maryland law). However, neither Helvering nor Vaughan supports the United States being domiciled in all 50 states. See Helvering, 293 U.S. at 91-93, 55 S.Ct. 50 (holding United States has residence, not domicile, in all states); Vaughan, 40 U.S. at 6 (“[t]he United States, in their sovereign capacity, have no particular place of domicile, but possess, in contemplation of law, an ubiquity throughout the Union ....”) (emphasis added). Finally, and most importantly, for present purposes, the United States being domiciled in all 50 states runs counter to New York’s rule that a person may have only one state of domicile. See In re Strobel’s Estate, 200 Misc. 483, 109 N.Y.S.2d 848, 850 (N.Y.Sur.Ct.1951); 49 N.Y.Jur.2d, Domicile & Residence § 1 (1985). Therefore, under New York choice of law, it is unlikely the United States would be considered domiciled in all 50 states.

The second possibility is the United States is domiciled nowhere. See Vaughan, 40 U.S. at 6 (“[t]he United States, in them sovereign capacity, have no particular place of domicile ....”) (emphasis added); O’Rourke v. Eastern Air Lines, 730 F.2d 842, 851 n. 12 (2d Cir.1984) (without citation, under New York choice of law, “[t]he United States, of course, is neither incorporated nor domiciled in any one state”); Foster v. U.S., 768 F.2d 1278, 1284 (11th Cir.1985) (without citation, under Florida choice of law, “the United States is obviously a domiciliary of neither [Florida nor Illinois]”); Clawans v. U.S., 75 F.Supp.2d 368, 373-74 (D.N.J.1999) (without citation, under New Jersey choice of law, the United States is not domiciled in New Jersey). However, this position is not as iron clad as it first appears. Each of the cited cases lack citation to authority. See id. Also, none of these cases is from a New York state court determining the domicile of the United States under New York choice of law. See id. Indeed, O’Rourke, Clawans, and Foster merely state the United States is not domiciled in the state under discussion without mention of where the United States is domiciled. See O’Rourke, 730 F.2d at 851 n. 12; Foster, 768 F.2d at 1284; Clawans, 75 F.Supp.2d at 373-74. Nonetheless, it is possible the United States is not domiciled in any state.

The third possibility is the United States is domiciled in the District of Columbia. See Fisher v. Fisher, 250 N.Y. 313, 165 N.E. 460, 462 (1918); Hitchcock v. United States, 665 F.2d 354, 360 (D.C.Cir.1981) (without citation, under District of Columbia choice of law, “[i]t is thus useful to analogize the Government in this case to a corporation of national scope, headquartered in Washington”). In Fisher, the highest court of New York applied New York choice of law and determined the United States is domiciled in the District of Columbia. See Fisher, 165 N.E. at 463. 22 The Fisher position runs counter to Vaughan and O’Rourke, discussed supra. Also, the Fisher decision is quite old and predates modern New York choice of law rules. Nonetheless, because Fisher is the only decision from a New York court applying New York choice of law to the question of domicile of the United States, it is possible a New York state court would consider the United States to be domiciled in the District of Columbia.

In conclusion, there is some authority supporting the mutually exclusive positions of the United States being domiciled in all 50 states, in the District of Columbia, or nowhere. We need not select among these positions because the result under New York choice of law rules will be the same.

b. Application of Neumeier Rules

(1) Assuming United States Domiciled in All 50 States

If the United States is domiciled in all 50 states, the first Neumeier rule applies. 23 See Neumeier, 335 N.Y.S.2d 64, 286 N.E.2d at 458 (when parties share common domicile, the law of that state controls). Assuming the United States is domiciled in all 50 states, the United States, Gould, and APU share the common domicile of Ohio and Ohio contribution law would govern.

(2) Assuming United States Domiciled in the District of Columbia or Nowhere

If the United States is domiciled in the District of Columbia or nowhere, there is a “split-domicile” between Gould/ APU and the United States, requiring application of the second or third Neumeier rules. Under the second rule, when the plaintiff and defendant are not domiciled in the same state, the states have conflicting law, and the tort occurs in one of the domiciles, the law of the state where the tort occurred controls. See Neumeier, 335 N.Y.S.2d 64, 286 N.E.2d at 458. The locus of the tort is the state where “the last event necessary to make the actor liable” occurred. Schultz, 491 N.Y.S.2d 90, 480 N.E.2d at 682-83. In Schultz, plaintiffs’ children, New Jersey residents, were sexually abused by defendants at a camp in New York. See id. at 681. The Court held the locus of the tort was New York, the place where the injuries occurred. See id. at 683. In this case, the last event which caused injuries to the Allen plaintiffs was the release of hazardous waste from the plant in New York, making New York the locus of the tort, for purposes of the Neumeier rules. Regardless of whether the United States is domiciled in the District of Columbia or nowhere, the locus of the tort was not one of the parties’ domiciles. Therefore, the second Neumeier rule does not apply.

The third Neumeier rule provides that in all other situations with a split domicile, the law of the locus of the tort governs, unless “it can be shown that displacing that normally applicable rule will advance the relevant substantive law purposes without impairing the smooth working of the multistate system or producing great uncertainty for litigants.” Neumeier, 335 N.Y.S.2d 64, 286 N.E.2d at 458. Regardless of whether the United States is domiciled nowhere or in the District of Columbia, New York contribution law governs unless there is an exception allowing displacement of New York law (“locus law”) with Ohio law, the law of the plaintiffs domicile (“domiciliary law”). It is appropriate to displace the locus law with domiciliary law when the plaintiffs domicile has a strong interest in protecting the rights of its domiciliary and displacement will not interfere with the interests of the locus state. See Schultz, 491 N.Y.S.2d 90, 480 N.E.2d at 687.

The purpose of this exception is to achieve New York’s strong preference for using domiciliary law as the loss allocating law. See Comer v. Titan Tool Inc., 875 F.Supp. 255, 259 (S.D.N.Y.1995); Hamilton v. Accu-Tek, 47 F.Supp.2d 330, 337 (E.D.N.Y.1999). In Comer, the Court explained:

When the law in conflict is loss allocating, the law of the state where at least one of the parties is domiciled generally applies. See Padula v. Lilam Properties Corp., 84 N.Y.2d 519, 620 N.Y.S.2d 310, 311-12, 644 N.E.2d 1001, 1002-03 (1994); Schultz, 491 N.Y.S.2d at 94, 480 N.E.2d at 683; Neumeier v. Kuehner, 31 N.Y.2d 121, 335 N.Y.S.2d 64, 286 N.E.2d 454 (1972)....

A state has a strong interest in enforcing its own loss-allocating rules, see Schultz, 491 N.Y.S.2d at 96, 480 N.E.2d at 685, especially in cases where such rules would serve to protect one of its own domieiliaries who was injured outside of the state. See id., Neumeier, 31 N.Y.2d 121, 335 N.Y.S.2d 64, 67, 286 N.E.2d 454, 455-56. In cases in which none of the parties share domicile (so-called “split-domicile” cases) the law of the domicile of at least one of the parties ought to apply. Neumeier, 31 N.Y.2d 121, 335 N.Y.S.2d 64, 67, 286 N.E.2d 454, 455-56....

Comer, 875 F.Supp. at 259 (emphasis added).

Courts have been more willing to displace locus law with domiciliary law when one or more of the following factors are present: (1) the plaintiffs and defendant’s domiciliary law is similar, see Tkaczevski v. Ryder Truck Rental Inc., 22 F.Supp.2d 169, 173 (S.D.N.Y.1998); (2) the purpose of the domiciliary law is superior to and does not interfere with the purpose of the locus law, see Schultz, 491 N.Y.S.2d 90, 480 N.E.2d at 687; (3) the locus of the tort is not one of the plaintiffs or defendant’s domiciles, see, e.g., id.; (4) plaintiffs and defendant’s contacts with the locus state were a matter of fortuity, happenstance, or randomness, and not voluntary action, see, e.g., Hamilton v. Accu-Tek, 47 F.Supp.2d 330, 341 (E.D.N.Y.1999); (5) displacement will not encourage forum shopping, see, e.g., Aboud v. Budget Rent A Car Corp., 29 F.Supp.2d 178, 182 (S.D.N.Y.1998); and (6) displacement will not create the appearance of favoring local litigants, see id. An evaluation of these factors, in light of the purpose of the exception, dictates that New York contribution law should be displaced with Ohio contribution law.

(a) Similarity of Domiciles’ Laws

First, courts are more likely to displace the locus law with domiciliary law when the domiciliary law of the plaintiff and defendant is similar. See Tkaczevski, 22 F.Supp.2d at 173. Three reasons have been articulated in support of this principle: (i) New York choice of law prefers domiciliary law to apply in loss allocating situations; (ii) neither party can complain about upset expectations when its own domiciliary law, or the equivalent, is applied; and (iii) the locus of the tort is only to be used as a tie breaker. See id. Further, there should be no difference in result where the plaintiff and defendant have the same domicile (where the first Neumeier rule governs and the law of the common domicile applies), and the plaintiff and defendant have a split domicile but with the same domiciliary law (where the third Neumeier rule governs).

Ohio contribution law permits contribution claims by settling tortfeasors against non-settling joint tortfeasors. See Ohio Stat. § 2307.31; MetroHealth, 80 Ohio St.3d at 214, 685 N.E.2d at 532. If the United States is considered to be domiciled nowhere, then it has no domiciliary law to compare to Ohio contribution law. However, since Gould/APU’s domiciliary law and the United States lack of domiciliary law are not in conflict, this factor weighs towards displacing New York law with Ohio law.

If the United States is considered to be domiciled in the District of Columbia, the District of Columbia will most likely permit contribution claims by settling tortfea-sors against non-settling joint tortfeasors. See District of Columbia v. Washington Hosp. Ctr., 722 A.2d 332, 342 & n. 12-14 (D.C.1998) (en banc) (and cases cited therein). 24 To the extent each party ex pected application of its own contribution law, it will not upset expectations to now apply Ohio law in place of New York law. Since the District of Columbia’s contribution law is similar to that of Ohio, this factor weighs towards displacing New York contribution law with Ohio contribution law.

(b) Purposes of Domiciles’ Laws

Courts are more likely to displace locus law with domiciliary law when the purpose of the domiciliary law is superior to and does not interfere with the purpose of the locus law. See Schultz, 491 N.Y.S.2d 90, 480 N.E.2d at 687; McCann v. Somoza, 933 F.Supp. 362, 366 (S.D.N.Y.1996). In Schultz, the plaintiff was domiciled in New Jersey, the defendant was domiciled in Ohio, and the tort occurred in New York. See Schultz, 65 N.Y.2d at 201, 491 N.Y.S.2d 90, 480 N.E.2d 679. The court displaced the charitable immunity law (loss allocating law) of New York with the charitable immunity law of New Jersey because New Jersey had a strong interest in protecting the charitable immunity of its domiciliary and New York had no significant interest in applying its charitable immunity law to non-domiciliaries. See id. Since New Jersey had a superior interest to New York which did not conflict with the purpose of the New York law, the court displaced New York law with New Jersey law. See id. 25

Similarly, in this case, the purpose of Ohio contribution law is superior to and does not interfere with the purpose of New York contribution law. Ohio contribution law allows settling tortfeasors to claim contribution from non-settling joint tort-feasors. See Ohio Stat. § 2307.31; MetroHealth, 685 N.E.2d at 532. The purpose of Ohio’s contribution law is “to make contribution readily available between joint tortfeasors” without exposing any joint tortfeasor to potential double liability. MetroHealth, 685 N.E.2d at 532.

The purpose of New York Gen. Oblig. Law § 15-108 is “encouraging settlements, fully compensating injured victims and equitably allocating liability among tortfea-sors.” Didner v. Keene, 188 A.D.2d 15, 593 N.Y.S.2d 238, 243 (1993). To achieve this purpose, § 15-108 contains three subsections which: (a) reduce plaintiffs recovery against a non-settling tortfeasor by the amount obtained in a settlement with a joint tortfeasor; (b) bar a non-settling joint tortfeasor from obtaining contribution from a settling joint tortfeasor; and (c) bar a settling joint tortfeasor from obtaining contribution from a non-settling joint tortfeasor. See § 15-108(a)-(e); Makeun v. State, 98 A.D.2d 583, 471 N.Y.S.2d 293, 296 (1984). The purpose of subsections (a) and (b) is “to encourage settlements by altering or eliminating certain rules of prior law which had an inhibiting effect on the settlement process.” Makeun, 471 N.Y.S.2d at 296 (quoting Rock v. Reed-Prentice Div. of Package of Mach. Co., 39 N.Y.2d 34, 382 N.Y.S.2d 720, 346 N.E.2d 520, 523 (1976)); N.Y. State Legis. Annual, 15 (1974). On the other hand, the purpose of subsection (c) is twofold. First, it acts as a “quid pro quo” preventing a settling tortfeasor from obtaining contribution when he is insulated from contribution. See id. (citing McDermott v. City of New York, 50 N.Y.2d 211, 428 N.Y.S.2d 643, 406 N.E.2d 460, 464 (1980); Rock, 346 N.E.2d at 524). Second, to the extent a settling tortfeasor pays more than his share, this section “may be justified on the ground that he is a volunteer as to the excess paid by him.” Id. (quoting Codling v. Paglia, 38 A.D.2d 154, 327 N.Y.S.2d 978, 986 (1978)).

At the same time, New York has “no interest in applying its [contribution] laws for the benefit of nonresidents and to the detriment of its residents.” Brewster v. Baltimore & Ohio Railroad Co., 185 A.D.2d 653, 585 N.Y.S.2d 647, 648 (1992). In Brewster, the court refused to give contribution claim defendants the benefit of New York contribution law when they were both domiciled in other states. See id. at 648-49. On the present facts, New York has no interest in extending the protection of its contribution law to the United States, assuming it is domiciled in the District of Columbia or nowhere. By the same token, New York has no interest in inhibiting the contribution rights of Gould/ APU, domiciliaries of Ohio. Finally, the Allen plaintiffs, the only New York domi-ciliaries related .to this case, have been compensated by virtue of the underlying tort settlement and are not parties to this contribution claim.

New York’s interest in achieving settlement of the Allen claims has been satisfied. See Lang Tendons, Inc. v. Great Southwest Marketing, Inc., 1994 WL 159014, at *4 (E.D.Pa.1994). 26 While subsections (a) and (b) achieve the purpose of encouraging settlement, subsection (c) does not. See Makeun, 471 N.Y.S.2d at 296. In fact, commentators have argued subsection (c) acts as “a barrier to settlements, rather than an aid.” Id. at 297 (citing Green, General Obligations Law, Section 15-108: An Unsettling Law, New York State Bar Journal, 28 (Oct.1983)). Accordingly, any interest New York has in promoting settlement will not be advanced by application of § 15-108(c) to bar Gould/ APU’s contribution claim against the Unit ed States. 27

In contrast, Ohio has a strong interest in applying its contribution law to protect its domiciliary’s rights to obtain contribution. Cf. Sullivan v. J.V. McNicholas Transfer Co., 224 A.D.2d 966, 638 N.Y.S.2d 260, 262 (1996) (New York has no interest in applying New York loss allocation law to Ohio domiciliary). Because Ohio’s interest in protecting the contribution rights of Gould/ APU is superior to and does not interfere with the purpose of New York’s contribution law, this factor weighs in favor of displacing New York law with Ohio law.

(c) Locus State Not Domicile of Any Party

Third, courts are more likely to displace the locus law when the locus state is not one of the plaintiffs or defendant’s domiciles. Compare Schultz, 491 N.Y.S.2d 90, 480 N.E.2d at 682-83 (locus law displaced when locus not a domicile of any party); Sheldon v. PHH Corp., 135 F.3d 848, 853 (2d Cir.1998) (same) with LaForge v. Normandin, 158 A.D.2d 990, 551 N.Y.S.2d 142, 142 (1990) (locus law not displaced when locus is one party’s domicile); Hamilton, 47 F.Supp.2d at 337 (E.D.N.Y.1999) (same). If the United States is domiciled in the District of Columbia or nowhere and Gould/APU are domiciled in Ohio, neither party is domiciled in the locus of New York. This factor favors displacement of New York law with Ohio law.

(d) Fortuity of Contacts with Locus State

Fourth, courts are more likely to displace locus law with domiciliary law when the parties have minimal contact with the locus jurisdiction and the tort occurring there was due to fortuity, happenstance, or randomness. Compare Pescatore v. Pan Am. World Airways, Inc., 97 F.3d 1, 13 (2d Cir.1996) (Scottish law displaced by plaintiffs domiciliary law because plane crashing in Scotland was random); Sheldon, 135 F.3d at 853 (Michigan law displaced by domiciliary law because car accident occurring in Michigan was fortuity); Stevens v. Shields, 131 Misc.2d 145, 499 N.Y.S.2d 351, 352 (N.Y.App.1986) with Simons v. Marriott Corp., 1993 WL 410457, *7 (S.D.N.Y.1993) (Texas law not displaced by domiciliary law because defendant voluntarily built hotel in Texas); Hamilton, 47 F.Supp.2d at 341 (California law not displaced by domiciliary law because plaintiff voluntarily came to California on business trip). There was nothing random or fortuitous about Gould/APU and the United States voluntarily building and operating a battery manufacturing plant in the locus state of New York. Therefore, the fortuity factor counsels against displacement of New York law with Ohio law.

(e) Forum Shopping

Fifth, courts are hesitant to displace locus law with domiciliary law if it encourages forum shopping. See, e.g., Aboud, 29 F.Supp.2d at 182. This concern is not present in this case because the FTCA does not allow forum shopping. Rather, under the FTCA, the court must apply the “whole law” of the state where the act or omission occurred, regardless of where the suit is filed. See Richards, 369 U.S. at 11, 82 S.Ct. 585. Because there can be no forum shopping for substantive law under the FTCA, this factor weighs in favor of displacement of New York law with Ohio law.

(f) Favoring Local Litigants

Sixth, courts are hesitant to displace locus law with domiciliary law if it causes the appearance of favoring local litigants. See, e.g., Aboud, 29 F.Supp.2d at 182. This concern is not present here because Gould and APU are Ohio domicili-aries and not local litigants in Pennsylvania federal court. Accordingly, this factor does not counsel against displacement of New York law with Ohio law.

In sum, five of the six factors favor displacement of New York law with Ohio law: the domiciles of Gould, APU, and the United States have similar, or at least non-conflicting, contribution law; Ohio’s strong interest in protecting its domiciliaries’ rights to contribution is superior to and does not interfere with New York’s interest in achieving settlement, which has al: ready been satisfied; the locus of the tort is not in the domicile of any party; and displacing New York law will not encourage forum shopping or give the appearance of favoring local litigants. Weighed against these five considerations is the fortuity factor, which, standing alone, is simply not enough to prevent displacement of New York locus law with Ohio domiciliary law. This result will advance the interests of Ohio without upsetting the smooth working of a multi-state system or causing uncertainty for litigants. It follows, if the United States is domiciled in the District of Columbia or nowhere, the exception to the third Neumeier rule applies and Ohio contribution law governs the FTCA jurisdictional inquiry.

* * * * * *

In conclusion, regardless of whether the United States is domiciled in all 50 .states, thq District of Columbia or nowhere, New York choice of law rules dictate that Ohio contribution law governs the FTCA jurisdictional inquiry. Ohio contribution law allows settling tortfeasors to claim contribution from non-settling joint tortfeasors, so long as the settlement extinguishes liability of the non-settling party to the, underlying plaintiff. See Ohio Stat. § 2307.31; MetroHealth, 685 N.E.2d at 532. In the Release presented to the District Court the Allen plaintiffs released “all other persons including but not limited to the federal government, the U.S. Army and/or other federal agencies or entities from any and all ... claims asserted or which could have been asserted, in the action captioned Cheryl Allen, et al. v. Marathon Battery Co., et al .,....” App. 705a. Under Ohio law, the United States will be liable to Gould/APU for contribution if the facts at trial call for that result. Accordingly, the District Court’s ruling it lacked subject matter jurisdiction over the contribution claim must be reversed.

C. Indemnification Under New York, Pennsylvania, Ohio, and District of Columbia Law

Indemnification occurs when one person is held solely liable for the acts of another person. See Rogers v. Dorchester Assocs., 32 N.Y.2d 553, 347 N.Y.S.2d 22, 300 N.E.2d 403, 410 (1973); Builders Supply Co. v. McCabe, 366 Pa. 322, 77 A.2d 368, 371 (1951); Motorists Mut. Ins. Co. v. Huron Road Hospital, 73 Ohio St.3d 391, 653 N.E.2d 235, 238 (1995); Quadrangle Development Corp. v. Otis Elevator Co., 748 A.2d 432, 435 (D.C.2000). 28 One may obtain indemnification only if he is not responsible in any degree for the harm caused. See Rosado, 494 N.Y.S.2d 851, 484 N.E.2d at 1356-57 (N.Y.); Builders Supply Co., 77 A.2d at 371 (Pa.); Motorists Mut. Ins. Co., 653 N.E.2d at 238 (Ohio); Quadrangle Development Corp., 748 A.2d at 435 (D.C.). If the party seeking indemnification is partially at fault, it cannot receive indemnification from another. See id.

The District Judge correctly decided Gould/APU would not be entitled to indemnity from the United States because they were at least partially at fault for the injuries to the Allen plaintiffs. According to the Complaint, Gould’s predecessors operated the plant and discharged hazardous waste under government contract from 1953 to 1962. From 1962 to 1969, Gould and its predecessors owned and operated the plant and continued to discharge hazardous waste. From 1969 to 1979, APU and its predecessors owned and operated the plant and continued to discharge hazardous waste. The Complaint alleges Gould/APU were at least partially negligent and at fault for the Allen plaintiffs’ injuries. Accordingly, Gould/APU may not seek indemnification from the United States.

Gould/APU argue, under New York law, the doctrine of “partial indemnification” permits indemnification between joint tort-feasors where one was primarily responsible for the injuries. See Dole v. Dow Chemical Co., 30 N.Y.2d 143, 331 N.Y.S.2d 382, 282 N.E.2d 288, 291 (1972). However, Gould/APU concede Dole has been superseded and clarified by N.Y. Gen. Oblig, Law § 15-108, which allows a party who is partially at fault to seek contribution, but not indemnification. See In re Brooklyn Navy Yard Asbestos Litigation, 971 F.2d 831, 842 (2d Cir.1992). Moreover, the Supreme Court of Pennsylvania, the Supreme Court of Ohio, and the District of Columbia Court of Appeals have squarely rejected similar partial indemnification theories, instead allowing contribution from a primarily, but not solely, liable joint tortfeasor. See Builders Supply Co., 77 A.2d at 371 (Pa.1951); Motorists Mut. Ins. Co., 653 N.E.2d at 238; District of Columbia v. Washington Hosp. Center, 722 A.2d 332, 340-41 (D.C.1998). It follows, there is no doctrine of partial indemnification that would permit Gould/APU to pursue an indemnity claim against the United States.

Finally, Gould/APU assert that they are entitled to indemnity for the Allen claim for abnormally dangerous activity, a strict liability tort for which there is no fault. See Doundoulakis v. Town of Hempstead, 42 N.Y.2d 440, 398 N.Y.S.2d 401, 368 N.E.2d 24, 28 (1977) (strict liability for abnormally dangerous activity); Diffenderfer v. Staner, 722 A.2d 1103, 1106 (Pa.Super.Ct.1999) (same); Brown v. County Commissioners, 87 Ohio App.3d 704, 622 N.E.2d 1153, 1159 (1993) (same); Beard v. Goodyear Tire & Rubber Co., 587 A.2d 195, 201 (D.C.1991) (same). However, if two parties are jointly hable for a strict liability tort, the appropriate remedy is contribution, not indemnification, between the parties. See Doundoulakis, 398 N.Y.S.2d 401, 368 N.E.2d at 29 (N.Y.); Baker v. AC & S, Inc., 729 A.2d 1140, 1149 (Pa.Super.Ct.1999) (citing Walton v. Avco Corp., 530 Pa. 568, 610 A.2d 454, 462 (1992)); Hamilton v. RB & W Corp., 1998 WL 32777, at *2-*3 (Ohio.App.1998) (citing Bowling v. Heil Co., 31 Ohio St.3d 277, 511 N.E.2d 373, 380-81 (1987)); Washington Hosp. Center, 722 A.2d at 340-41 (D.C.). If Gould/APU could prove the United States was jointly liable, the appropriate remedy would be contribution, not indemnification.

Under New York, Pennsylvania, Ohio, or District of Columbia law, the United States would not be liable to Gould/APU for indemnification. Accordingly, the District Court’s ruling it lacked subject matter jurisdiction over the indemnity claim will be affirmed.

V. CONCLUSION

For the foregoing reasons, we conclude the District Court has jurisdiction over the contribution claim but lacks jurisdiction over the indemnification claim. The District Court’s ruling is AFFIRMED as to indemnification and REVERSED as to contribution and REMANDED for proceedings consistent with this opinion.

1 . Because of our disposition it is unnecessary to reach the third and fourth points of error raised by Gould/APU.

2 . Gould/APU asserted jurisdiction of the District Court pursuant to 28 U.S.C. § 1346(b). This Court has jurisdiction over an appeal from a dismissal for lack of jurisdiction pursuant to 28 U.S.C. § 1291. In this case, we use the word "jurisdiction” not in the usual subject matter context, but rather to refer to the extent of the sovereign immunity waiver in the Federal Tort Claims Act. As the Supreme Court explained in Smith v. United States, 507 U.S. 197, 201, 113 S.Ct. 1178, 122 L.Ed.2d 548 (1993), "by its terms [28 U.S.C. § 1346(b)] is more than a choice-of-law provision: It delineates the scope of the United States’ waiver of sovereign immunity.”

3 . In 1967, Sonotone became a wholly owned subsidiary of Clevite Corporation and in 1969, Clevite merged with Gould. For simplicity, we use the name Sonotone to refer to the owner-operator of the plant from 1962 to 1969.

4 . Between 1969 and 1979, Business Funds, Inc. became Marathon Battery Co. which eventually became APU. For simplicity, the name APU is used to refer to the relevant owner/operator of the plant from 1969 to 1979.

5 . Gould/APU also challenge the application of standards on a Rule 12(b)(1) motion by arguing the Complaint alleges a claim for contribution and indemnity under the law of both New York and Pennsylvania. This argument relates to the substance of the District Court's conflict of laws analysis and is addressed in Section IV.B, infra.

6 . PBGC held, in evaluating a Rule 12(b)(6) motion, the court may consider documents attached to the complaint. See Gotha, 115 F.3d at 178-79. We can think of no principled reason why a court,, in resolving a Rule 12(b)(6) facial attack should not also consider documents attached-to the complaint.

7 . Section 15-108(c) provides: "A tortfeasor who has obtained his own release from liability shall not be entitled to contribution from any other person.” Id.

8 . In contrast, in Growth Horizons, Inc. v. Delaware County, Pa., this Court reversed the District Court’s purported dismissal under Rule 12(b)(1) as being, in essence, a 12(b)(6) dismissal, in part, because the District Court failed to mention or apply the standards for a Rule 12(b)(1) motion. See 983 F.2d 1277, 1281 (3d Cir.1993).

9 . The Complaint alleges the United States committed the following acts or omissions in New York: ownership of the property where the plant was located; construction of the plant; ownership of the plant; and installation of equipment in the plant. In addition, whether or not performed by the United States, the following acts or omissions allegedly occurred in New York: operation of the plant; discharge of hazardous waste into the water and air; and personal injuries and property damage to the Allen plaintiffs. Finally, the underlying Allen case was filed in New York and settled in New York.

The Complaint alleges the United States committed the following acts or omissions in Pennsylvania: entering into Letter Contract and Facilities Contract with Sonotone to allow Sonotone to operate the plant; entering into lease agreements with Sonotone; and selling the plant to Sonotone.

10 . Gould/APU also argue the District Court erred in conducting any choice of law analysis. Gould/APU contend, because contribution and indemnification law is procedural, not substantive, the District Court was bound to choose the contribution and indemnification law for the state in which it sits, Pennsylvania. We disagree. While there is a procedure/substance distinction for a conflict of laws analysis when, as here, the District Court determines jurisdiction pursuant to the FTCA. The FTCA requires application of the relevant state’s “whole law,” without making a distinction between procedure and substance. See 28 U.S.C. § 1346(b)(1); Richards, 369 U.S. at 11, 82 S.Ct. 585.

11 . Prior to oral argument, the litigants were requested to file supplemental briefs addressing, inter alia, the following question:

Under the Federal Tort Claims Act, how should a court choose between conflicting choice of law provisions when the acts or omissions occurred in more than one state? See Richards v. United States, 369 U.S. 1, 11, 82 S.Ct. 585, 7 L.Ed.2d 492 (1962).

12 . See Laconis v. Burlington County Bridge Comm’n, 400 Pa.Super. 483, 583 A.2d 1218, 1222 (1990). In an interest analysis, the court determines which jurisdiction is most intimately concerned with the outcome of the action, considering four factors:

(1) the place where the injury occurred; (2) the place where the conduct causing the injury occurred; (3) the domicile, residence, nationality, place of incorporation, and place of business of the parties; and (4) the place where the relationship between the parties is centered.

Id. at 1222-23.

13 . See Padula v. Lilarn Properties, Corp., 84 N.Y.2d 519, 521-22, 620 N.Y.S.2d 310, 644 N.E.2d 1001 (N.Y.1994); Cooney v. Osgood Mach., 81 N.Y.2d 66, 595 N.Y.S.2d 919, 612 N.E.2d 277, 280 (1993); Neumeier v. Kuehner, 31 N.Y.2d 121, 335 N.Y.S.2d 64, 286 N.E.2d 454, 457-58 (1972); see also, infra, Section IV.B.2.

14 . See, supra, note 7.

15 . .42 Pa.Cons.Stat.Ann. § 8324 provides:

(a) General rule. — The right of contribution exists among joint tort-feasors.

(b) Payment required. — A joint tort-feasor is not entitled to a money judgment for contribution until he has by payment discharged the common liability or has paid more than his pro rata share thereof.

(c) Effect of settlement. — A joint tort-feasor who enters into a settlement with the injured person is not entitled to recover con tribution from another joint tort-feasor whose liability to the injured person is not extinguished by the settlement.

16 . Ohio Stat. § 2307.31 provides, in pertinent part:

(A) Except as otherwise provided in this section or section 2307.32 of the Revised Code, if two or more persons are jointly and severally liable in tort for the same injury or loss to person or property or for the same wrongful death, there is a right of contribution among them even though judgment has not been recovered against all or any of them. The right of contribution exists only in favor of a tortfeasor who has paid more than his proportionate share of the common liability, and his total recovery is limited to the amount paid by, him in excess of his proportionate share. No tort-feasor is compelled to make contribution beyond that tortfeasor's own proportionate share of the common liability. There is no right of contribution in favor of any tortfea-sor who intentionally has caused or intentionally has contributed to the injury or loss to person -or property or the wrongful death.

(B) A tortfeasor who enters into a settlement with a claimant is not entitled [sic] recover contribution from another tortfea-sor whose liability for the injury or loss to person or property or the wrongful death is not extinguished by the settlement, or in respect to any amount paid in a settlement which is in excess of what is reasonable.

17 . In Andrulonis v. United States, the Court utilized that latter Bowen approach. See 724 F.Supp. 1421, 1471 (N.D.N.Y.1989) (citing Bowen, 570 F.2d at 1318).

18 . Although the United States contracted with Sonotone for operation of the plant, it is still deemed to have been operating the plant by virtue of the control it retained over the operations. Cf. FMC Corp. v. United States Dep’t of Commerce, 29 F.3d 833, 843 (3d Cir.1994) (en banc). In FMC, a CERCLA case, this Court determined the United States was an operator of a plant even if it contracted out the operation, so long as it retained substantial control over the facility and has active involvement in the activities there. See id. In FMC, the United States was held to be an operator of a plant because it:

determined what product the facility would manufacture, controlled the supply and price of the facility's raw materials, in part by building or causing plants to be built near the facility for their production, supplied equipment for use in the manufacturing process, acted to ensure that the facility retained an adequate labor force, participated in the management and supervision of the labor force, had the authority to remove workers who were incompetent or guilty of misconduct, controlled the price of the facility’s product, and controlled who could purchase the product.

Id. Similarly, the United States was an operator of the Cold Spring battery plant because it owned all real and personal property, installed all equipment, owned all materials and supplies, controlled what the plant would produce, had access to the plant at all times, reserved the right to dismantle or repair the plant, reserved the right to shut down the plant, and had the ability to exercise day-today control over the operations of the plant, and, through contracts, set the price terms for products produced by the plant.

19 . As discussed in note 18, supra, the United States is deemed to have operated the plant in New York by virtue of the control it exercised over the plant’s operations.

20 . The United Slates asserts that loss allocating law relates only to the parties’ expectations before torts are committed. It argues the § 15-108(c) bar against contribution claims is not loss allocating because it does not relate to prior expectations of tort liability. Rather, the United States contends § 15-108(c) is "a mechanism to influence and control the conduct of litigation in its jurisdiction.” United States Letter Brief (June 9, 2000) at 4.

This argument fails. In New York, like most jurisdictions, contribution claims were generally not recognized at common law. See Gleason v. Holman Contract Warehousing, Inc., 170 Misc.2d 668, 649 N.Y.S.2d 647, 649-50 (N.Y.Sup.Ct.1996). Rather, New York, like most states, enacted a statute providing for contribution among joint tortfeasors. Sec-lion 15-108 is a portion of New York’s contribution statute. See § 15-108. Since New York courts have held contribution law to be loss allocating, see Padula, 84 N.Y.2d at 522, 620 N.Y.S.2d 310, 644 N.E.2d 1001, section 15-108(c) must be loss allocating.

21 . In general, for purposes of other areas of New York law, the domicile of a corporation is the state in which it is incorporated. See Sease v. Central Greyhound Lines, Inc., 306 N.Y. 284, 117 N.E.2d 899 (1954); Comer v. Titan Tool, Inc., 875 F.Supp. 255, 259 (S.D.N.Y.1995). A person, including a corporation, can have a different domicile for purposes of applying different laws. See Gladwin v. Power, 21 A.D.2d 665, 249 N.Y.S.2d 980, 982 (1964).

22 . The Court needed to determine the domicile of the United States as an owner of a ship. See id. at 462. The Court first determined the United States is not domiciled in New York. See id. Next, the Court found the agency that owned the ship was domiciled in the District of Columbia, making the United States domiciled there. See id. at 463.

23 . The United States argues, if the United States is domiciled in all 50 states, the second Neumeier rule applies because Gould/APU are domiciled in Ohio and the United States is domiciled in New York, creating a split-domicile situation. However, the first Neumeier rule applies whenever the parties "share a common domicile.” Cooney, 595 N.Y.S.2d 919, 612 N.E.2d at 281.

24 . The recent decision in Washington Hosp. Ctr. explicitly left this issue unresolved. See 722 A.2d at 342. The case law cited in Washington Hosp. Ctr. reveals the District of Columbia uniformly allows such a claim. See Taylor v. Tellez, 610 A.2d 252, 254 (D.C.1992) (settling defendant may bring contribution claim against non-settling joint tortfeasor); Early Settlers v. Schweid, 221 A.2d 920, 922 (D.C.1966) (same). The purported contrary authority cited by the Washington Hosp. Ctr. court involves different types of contribution claims. In Hall v. George A. Fuller Co., 621 A.2d 848 (D.C.App.1993), the court held a settling tortfeasor may not obtain contribution from another settling tortfeasor because the court could not determine if they were jointly liable. See 722 A.2d 332, 342-43 (D.C.1998). In Rose v. Associated Anesthesiologists, the court held a settling tortfeasor may not obtain contribution from non-settling tort-feasor who is found solely liable. See 501 F.2d 806, 808-10 (D.C.Cir.1974). In contrast Gould/APU seek contribution against a non-settling joint tortfeasor, whose joint liability can be determined and who has not been adjudged solely liable.

25 . In contrast, in McCann, the plaintiff was domiciled in New Jersey, the defendant was domiciled in New York, and the tort occurred in Connecticut. See McCann, 933 F.Supp. at 365. The court refused to displace Connecticut law with New Jersey or New York law because the three states had conflicting loss allocating law with irreconcilable purposes. See id. at 367. The court sought to promote certainty by utilizing the locus of the tort as a “tie-breaker’' because the laws of the domiciles conflicted. See id.

26 . The decision in Lang Tendons is illuminating. In Lang Tendons, the plaintiff, a Pennsylvania domiciliary, settled a tort claim in New York federal court arising from a tort that occurred in Korea. The plaintiff then brought a contribution claim against the defendant, a Texas domiciliary. The defendant was a non-party to the prior litigation because the New York federal court lacked personal jurisdiction over it. The defendant sought application of the New York contribution law, specifically the § 15-108(c) bar against contribution claims against non-settling joint tort-feasors. The plaintiff sought application of Pennsylvania contribution law. See id. at *3.

The court conducted a choice of law analysis, applying Pennsylvania's interest analysis, and concluded Pennsylvania had the greater interest in the outcome of the litigation because: (i) Pennsylvania had a strong interest in vindicating the interests of its domiciliary; (ii) New York's interest had been satisfied by the settlement and had no interest in third party contribution claims outside of New York; (iii) the locus of the tort was not in New York; and (iv) there was a strong interest in not having the defendant avoid all liability simply because it could not have' been joined as a co-defendant in the underlying tort suit.

In the present case, Ohio, the domicile of Gould/APU, has the stronger interest in the outcome of this litigation. First, Ohio has a strong interest in protecting the rights of its domiciliaries to receive contribution from joint tortfeasors. See MetroHealth, 685 N.E.2d at 532. Second, New York’s interest in encouraging settlement has been vindicated and New York therefore no longer has an interest in whether contribution claims are pursued in other jurisdictions. See Makeun, 471 N.Y.S.2d at 297. Finally, the United States could not have been joined as a defendant in New York state court because of sovereign immunity. Applying New York contribution law would allow the United States to escape all liability because the Allen plaintiffs sued in New York state court, where the United States is shielded from liability.

28 . As discussed, supra, there is a false conflict between New York, Pennsylvania, Ohio, and District of Columbia indemnification law, making it unnecessary to perform a conflict of laws analysis for the indemnification claim.

27 . The United States correctly indicates that § 15-108(c) has been applied to bar contribution claims by a settling tortfeasor against New York state, even if New York had immunity in the underlying suit and could not be impleaded as a third party defendant. See Makeun, 471 N.Y.S.2d at 298; Maryland Cas. Co. v. State, 66 A.D.2d 953, 411 N.Y.S.2d 706, 707 (1978). However, neither Makeun nor Maryland Cas. Co. precludes displacement of New York contribution law with Ohio contribution law. Both cases involved contribution claims by New York domiciliaries against New York State. See id. In both cases, the New York domiciliary plaintiffs argued the § 15-108(c) bar against contribution did not apply because New York State could not have been joined as a defendant in the underlying action. See id. The New York court held § 15-108(c) does not contain an exception for contribution claims against parties over whom the court lacked jurisdiction in the underlying action. See id. In contrast, this case involves a contribution claim by two non-New York domiciliaries (Gould/APU) against a defendant (United States) which could not have been joined in the underlying Allen lawsuit. The issue is not whether § 15— 108(c) provides an exception for a defendant who could not be joined, but whether application of Gould/APU’s Ohio contribution law will be superior to and will not interfere with application of the locus contribution law of New York.