Maureen Mcgurl; Dewey Cannella; Richard E. Mcfeeley;robert B. Davidson; Joseph Rizzo; Harvey Whille; Michaelkinsora; Louis Marcucci, As Trustees of the Ufcw Local 1262and Employers Welfare Fund; Joseph Rizzo; Harvey Whille;michael Kinsora; Gilbert C. Vuolo; John Polding; Robertf. Ennis, As Trustees of the Ufcw Local 1262 and Employershealth and Welfare Fund, Appellants in No. 96-5330, v. Trucking Employees of North Jersey Welfare Fund, Inc.,* Appellant in No. 96-5348.nos. 96-5330, 96-5348, 124 F.3d 471 (3d Cir. 1997)

In this case raising a question of first impression in the federal courts, we are faced with an apparent conflict between "other insurance" provisions in two self-funded ERISA plans, each of which purports to provide at most secondary coverage to the same claimants.

To resolve the conflict, the district court crafted a federal common law order of benefits determination rule that would impose primary liability on the fund whose participants are the employers of the claimants, in this case the Appellants. Appellants argue that the district court erred by concluding that the two plans are not reconcilable on their terms, that the court should have adopted New Jersey state law as the appropriate rule of decision for resolving any apparent conflict between the plans, and that the federal common law rule settled upon is a poor choice.

BACKGROUND

The Parties and Their Plans

Appellants are the trustees of two self-funded welfare benefit plans of the United Food and Commercial Workers Local 1262 and Employers Welfare Fund and the U.F.C.W. Local 1262 and Employers Health and Welfare Fund (collectively "Local 1262 Funds"), which cover employees of several contributing employers in the supermarket industry. Appellee Teamsters Local 560 Trucking Employees of North Jersey Welfare Fund (the "TENJ Fund") is also a self-funded welfare benefit plan that covers employees of participating supermarkets. The respective plans contain "other insurance" clauses, more particularly referred to as "coordination of benefits" clauses in the group health insurance industry, that set forth circumstances under which the plans will assume primary coverage liability for a claimant who is also covered by another plan.

Group health care insurance plans have increasingly included coordination of benefits clauses because the enlarged number of two-employee families has increased the possibility that a claimant could be covered under more than one plan. By conditioning coverage on specified circumstances, the clauses seek to limit their costs and prevent a claimant from acquiring coverage from multiple plans in excess of the claimant's covered medical expenses. See Jack B. Helitzer, Coordination of Benefits: How and Why it Works, 4 Benefits L.J. 411, 412 (1991).

This dispute concerns the obligation of the plans to certain part-time supermarket employees who are covered under both plans. The Local 1262 Funds describe their coverage obligations to part-time employees under the heading, "Coordination of Benefits":

If a Part-time Member who is a Covered Member ... is also covered under one or more Other Plans, the Benefits payable under this Plan will be coordinated with Benefits payable under all Other Plans. When there is a basis for a claim under this Plan and the Other Plan, this Plan is a Reimbursement Plan which has its Benefits determined after those of such Other Plan.

As this is a Reimbursement Plan for Part-time Members who are Covered Members ... payments will be made after all other sources of coverage have been exhausted.

App. at 89.

The Local 1262 Funds' Summary Plan Description also states with respect to coverage for part-time employees:

[T]his Plan is always a reimbursement plan; if you are covered under another medical plan, this Plan will only take effect when the limits of your other Plan have been exceeded. This means that, you can receive benefits from this Plan (in the form of reimbursement payments) only after the other plan pays benefits to the full extent of the terms of that Plan.

App. at 143 (emphasis in original). Thus, the Local 1262 Funds attempt to defer any medical payments for their part-time employees until after the employee has exhausted all other possible sources of coverage, and the Funds refer to this proviso alternatively as a "reimbursement clause," an "excess clause," or an "always secondary clause."

The TENJ Fund, in an effort to avoid always being left with a claimant's bill, has a coordination of benefits provision that disclaims liability altogether for employees who are participants in a plan such as that of the Local 1262 Funds. The TENJ Fund plan provides:

In determining whether this plan is primary for a spouse [or dependent] the following will apply:

"The Plan covering the patient as an employee or in which the employee is a participant ... will be the primary plan. If the primary plan denies coverage because of the application of a Rule which is unique to that Plan and which is not a rule of this Fund, then this Fund will provide only that coverage which it would have provided if the primary plan had granted primary coverage.

This Fund does not afford coverage to a participant's dependent who herself/himself is a participant in a Reimbursement or similar plan that affords coverage only if there is no other health/welfare coverage."

App. at 217 (emphasis in original).

In an effort to further clarify the limits of its coverage, the TENJ Fund describes the following hypothetical:

Mr. ABC is a participant under our Welfare Fund. His spouse works for the XYZ company. Under normal coordination of benefits, Mrs. ABC's medical claims are submitted to her company first. After they pay the claim, in accordance with their Plan, she submits the claim with a copy of the explanation of benefits from her Plan to our Fund showing the amount paid. Our Fund then pays our portion of the claim under the coordination of benefits rule as the secondary payor and pays the difference up to the Fund's allowable amount. However, if Mrs. ABC's XYZ Plan rejects her claim because XYZ says its Plan is a Reimbursement Plan and will not pay claims if there is any other coverage, such as her being covered as a dependent under her husband's plan, then our Fund will not pay any portion of Mrs. ABC's claim.

Id. (emphasis added).

In May 1993, Susan Armstrong, a part-time employee of Shop-Rite Supermarkets, a contributing employer to the Local 1262 Funds, submitted medical expense claims to the Local 1262 Funds in an aggregate amount of $243,993. Because Armstrong's father was a participant in the TENJ Fund, she would ordinarily have also been eligible for secondary coverage as a dependent under the TENJ Fund plan. The Local 1262 Funds denied primary liability for Armstrong's claims on the ground that it was a reimbursement or excess plan only, and instead notified the TENJ Fund that it was primarily liable for paying Armstrong's expenses. In the following months, the Local 1262 Funds received similar claims from Karen Iler, Esther Owens, and Patricia Kelly, all of whom were also part-time employees of contributing employers to the Local 1262 Funds as well as dependents of participants of the TENJ Fund. The Local 1262 Funds similarly denied these claims and sent notification that the TENJ Fund bore primary responsibility.

In response, the TENJ Fund likewise denied primary coverage liability for the four part-time employees' claims. It took the position that, because the Local 1262 Funds provided only a reimbursement plan for the part-time employees, it was relieved from any liability by the express terms of the TENJ Fund. In a letter dated June 4, 1993, the TENJ Fund informed the Local 1262 Funds that "Since TENJ does not cover Susan Armstrong, your fund provides sole coverage." App. at 154.

In order to avoid undue hardship to the claimants throughout the period of time in which the two funds debated their respective liabilities, the Local 1262 Funds paid the claimants' benefits, without prejudice to their right to proceed against and seek reimbursement from the TENJ Fund. The TENJ Fund agreed to pay secondarily for the time being, but also "without prejudice to the rights of either party." App. at 413.

District Court Proceedings

On October 26, 1994, the Local 1262 Funds filed an action in the district court in New Jersey seeking a declaration that the TENJ Fund was primarily liable on the contested claims and an order directing the TENJ Fund to reimburse them for money paid to claimants in their assumed role as the primary provider. The Local 1262 Funds argued that the provision of the TENJ Fund plan which disclaims liability entirely if a beneficiary is covered by an alternate reimbursement plan was an invalid "escape clause." They then contended that once the escape clause is read out of the TENJ Fund plan, the remaining terms of both plans assign primary liability to the TENJ Fund plan. In response, the TENJ Fund argued that its plan did not contain an escape clause and that, regardless, the Local 1262 Funds plan was primarily responsible for the claims at issue according to its own coordination of benefits provision because the claimants are employees of participants of that plan.

In a thoughtful opinion, the district court granted the TENJ Fund's motion for summary judgment. See McGurl v. Teamsters Local 560 Trucking Employees of New Jersey Welfare Fund, 925 F. Supp. 280 (D.N.J. 1996). The court agreed that the provision of the TENJ Fund purporting to deny any liability if a beneficiary is separately covered by a reimbursement plan is an escape clause and thus unenforceable. Id. at 286. The court then concluded that the TENJ Fund's remaining coordination of benefits provision and the excess clause in the Local 1262 Funds plan were "mutually repugnant" because both attempted to deny primary coverage to these claimants, and would provide secondary coverage only if the other accepted primary liability. Id. at 289. In rejecting the Local 1262 Funds' suggestion that the remainder of the plans were still reconcilable in favor of the Local 1262 Funds, the court declined to apply the decision in Starks v. Hospital Serv. Plan of N.J., Inc., 182 N.J.Super. 342, 350, 440 A.2d 1353 (1981), aff'd, 91 N.J. 433, 453 A.2d 159 (1982), which held that an excess clause was secondary to an ordinary other insurance provision in an insurance contract. See McGurl, 925 F. Supp. at 288.

The district court in this case found the two plans at issue to be irreconcilable and chose to create a uniform federal common law rule in order to resolve the issue of how to prioritize the payment of benefits between two self-funded ERISA plans which have mutually repugnant coordination of benefits provisions. Id. at 293. The court adopted an "employer first" rule, recommended by the Model Regulations of the National Association of Insurance Commissioners ("NAIC"), which would impose primary liability for coverage on the plan which covers claimants as employees rather than as dependents. Id. The court also rejected the Local 1262 Funds' suggestion that the better federal common law rule would be to apportion liability on a pro-rata basis, reasoning that such a rule would provide an undesirable incentive for ERISA-regulated plans to include excess provisions. Id. at 292.

The Local 1262 Funds appeal from the district court's order, and the TENJ Fund, although successful, cross-appeals to preserve its argument that the district court erred in determining that its plan contains an unenforceable escape clause.

JURISDICTION AND STANDARD OF REVIEW

Both the Local 1262 Funds and the TENJ Fund are self-funded employee benefit plans, meaning that they do not purchase insurance policies in order to satisfy their obligations to pay for medical and disability benefits of their participants and they are, therefore, covered by the federal Employee Retirement Income Security Act of 1974 ("ERISA"), 29 U.S.C. §§ 1001-1461, § 1002(1). ERISA provides comprehensive regulation of employee benefit plans, §§ 1021-1031, §§ 1101-1114, and broadly preempts state laws that "relate to" such plans, § 1144(a). However, ERISA does not regulate the substantive terms of plans, see Shaw v. Delta Air Lines, Inc., 463 U.S. 85, 91, 103 S. Ct. 2890, 2896-97, 77 L. Ed. 2d 490 (1983), and makes no express mention about how to resolve conflicts between coordination of benefits clauses.

The district court had jurisdiction pursuant to 28 U.S.C. § 1331 because the case involved a dispute over the disbursement of payments under ERISA. See Northeast Dep't ILGWU Health and Welfare Fund v. Teamsters Local Union No. 229 Welfare Fund, 764 F.2d 147, 159, 166 (3d Cir. 1985). We have jurisdiction pursuant to 28 U.S.C. § 1291 and exercise plenary review over the district court's grant of summary judgment. United States v. Capital Blue Cross, 992 F.2d 1270, 1271-72 (3d Cir. 1993).

THE TENJ FUND'S ESCAPE CLAUSE

We begin with approval of the district court's construction of the coordination of benefits provision in the TENJ Fund plan. Under that provision, " [t]he Plan covering the patient as an employee ... will be the primary plan," but if that dependent is an employee of a plan that attempts to provide only reimbursement coverage, the TENJ Fund will not provide any coverage to the dependant at all. App. at 217 ("This Fund does not afford coverage to a participant's dependent who herself/himself is a participant in a Reimbursement or similar plan that affords coverage only if there is no other health/welfare coverage.")

This latter provision is indeed an escape clause, which, as we have previously explained, is one which "provides for an outright exception to coverage if the insured is covered by another insurance policy." Northeast, 764 F.2d at 160. In Northeast, we recognized that both a well-developed body of state common law of insurance and the policies underlying ERISA are hostile to the inclusion of escape clauses in benefits plans because they undermine the reasonable expectations of a beneficiary who may have his or her coverage shifted to another insurer with less favorable terms. Id. at 162-63. Such clauses are particularly harsh because,

[A] plan with an escape clause does not provide participants who receive less in benefits from the other plan with the opportunity to return to the first plan for the difference. As a result, a participant of a plan with an escape clause, who thinks that he is covered by that plan and who expects to recover medical expenses in accordance with the terms of that plan, automatically loses this coverage in the presence of another insurance plan, even if the benefits he is entitled to receive under the other plan are much less favorable than those of his own.

Id. at 163. We concluded that a decision by a plan's fiduciary to include an escape clause is "arbitrary and capricious" and thus unenforceable under ERISA's regulatory scheme. Id.

As written, under the TENJ Fund plan, a dependent-beneficiary of the TENJ Fund, who is also an employee beneficiary of the Local 1262 Funds, and who would anticipate receiving secondary benefits from the TENJ Fund, will get nothing because the Local 1262 Funds are excess only. According to the analysis in Northeast, therefore, this provision in the TENJ Fund plan is an unenforceable escape clause.

The TENJ Fund nevertheless argues that "in practice" it has not followed the categorical exclusions of the purported escape clause, but has treated it as a secondary liability provision once a competing plan abandons its reimbursement provision and assumes primary responsibility. That argument is unpersuasive. First, it is clear from the record that the TENJ Fund has expressed its right to categorically deny any payment obligations based on its escape clause, as it did in its June 4, 1993 letter to the Local 1262 Funds' manager. App. at 154. That the TENJ Fund did, in fact, pay secondarily in this case was merely a litigation convenience undertaken expressly "without prejudice to the rights of either party," app. at 413; it was not done as a modification or amendment of the language of its plan.

Second, interpretation of self-funded plans cannot depend on the unilateral understanding or ad hoc application by the plan, lest the comprehensibility, predictability, and assurance that ERISA intends to provide be lost. See, e.g., Coleman v. Nationwide Life Ins. Co., 969 F.2d 54, 56 (4th Cir. 1992), cert. denied, 506 U.S. 1081, 113 S. Ct. 1051, 122 L. Ed. 2d 359 (1993). Finally, when we were presented with a similar categorical/as-applied distinction in Northeast, we expressly chose a remedy of "total invalidation of escape clauses" and "put the onus on trustees of plans with escape clauses to rewrite the plans." 764 F.2d at 164 n. 17. Thus, only by express revision can the TENJ Fund transform the escape clause into a secondary liability clause, should it so choose.

CONFLICT BETWEEN THE TERMS OF THE PLANS

Once the district court found the TENJ Fund plan's escape clause unenforceable, it proceeded to examine whether, if the escape clause were read out of the TENJ Fund plan, the two plans were reconcilable or, in other words, if the plans themselves could still provide a coherent order of benefits scheme. The TENJ Fund plan, as redacted, would read: "The Plan covering the patient as an employee or in which the employee is a participant ... will be the primary plan," and thus would deny primary coverage if a claimant is an employee of a participant in another plan. The Local 1262 Funds plan denies primary coverage to their participants' part-time employees when those employees are in any way covered by another plan. The district court held the plans to be "mutually repugnant" because " [b]oth deny primary coverage and are willing to provide benefits in a secondary capacity only after the other accepts the responsibility of primary coverage." McGurl, 925 F. Supp. at 287.

The Local 1262 Funds urged the district court to apply the analysis of Starks, 182 N.J.Super. 342, 440 A.2d 1353 (1981). In Starks, individual claimants were employee-beneficiaries of the Amalgamated Welfare Fund and were also covered by Blue Cross/Blue Shield, an insurer, as dependents of beneficiaries. The Amalgamated Fund plan was similar to that of the Local 1262 Funds in that it would only provide reimbursement or "always excess" benefits after a claimant exhausted coverage from another plan. The Blue Cross/Blue Shield plan contained a coordination of benefits provision under which it would be secondary to a plan that covered one of its member's dependents as a direct beneficiary. Id. The Starks court held that Blue Cross/Blue Shield was primarily liable, reasoning that the Blue Cross/ Blue Shield plan contemplated being primarily liable in some instances, i.e., where the claimant is an employee, but the Amalgamated Fund plan never contemplated being primarily liable vis-a-vis another plan. Therefore, the two plans could be ranked hierarchically. Id. at 350, 440 A.2d 1353.

The Starks court concluded that the Amalgamated Fund's trustees contemplated a "tertiary" role when competing against any other secondary coverage provider, and stated that " [w]here the two coverages are not, however, primary and secondary but rather secondary and tertiary, there being no primary coverage in the usual sense, the only rational result is to require the secondary coverage to pay first and the tertiary to pay second." Id. at 353-54, 440 A.2d 1353. Despite the rather complex reasoning, the essence of Starks' ultimate holding was that the plans at issue "d [id] not support the predicate of mutual repugnancy." Id. at 353, 440 A.2d 1353.

The district court declined to follow Starks based on its authority derived from ERISA's broadly worded preemption provision to categorically ignore state rules of decision that relate to the regulation of self-funded plans. See PM Group Life Ins. Co. v. Western Growers Assurance Trust, 953 F.2d 543, 546 (9th Cir. 1992). In so ruling, the court also expressed its disagreement with the logic applied by the Starks court.

The Local 1262 Funds argue that we should construe the two plans as the Starks court did, and interpret their plan as secondary to the TENJ Fund plan's coordination of benefits provision on the ground that the TENJ Fund plan (like the Blue Cross/Blue Shield plan) recognizes certain situations in which it could be primarily liable, whereas the Local 1262 Funds plan is always secondary. We do not agree.

First, it is not helpful to speak in terms of secondary and tertiary, and indeed it is somewhat misleading. The Starks court ranked the plans in this manner to emphasize that neither of the plans before it accepted primary liability and it was thus forced to rank the payment obligations of a plan that it construed as providing some coverage, albeit secondary, and a plan that was always excess, which it denominated as "tertiary". However denominated, the task required in this case is to determine which Fund's plan is primary. It would be arbitrary to adopt the Local 1262 Funds' suggestion that because the TENJ Fund concedes primary liability in the situation of another plan's beneficiary but who is an employee of its participant, it must always be primary vis-a-vis Local 1262 Funds' always excess clause. In fact, the TENJ Fund plan flatly denies primary coverage when presented with a claim of a dependant-beneficiary such as Susan Armstrong's. The mere fact that in other circumstances the TENJ Fund would be primary does not obviate the inescapable fact that it is not primary in the circumstances here. However the Starks court chose to interpret the language of the Blue Cross/Blue Shield plan before it, we cannot fairly ignore the certain, evident conflict from the faces of the two plans.

Second, there would be substantial and adverse fiscal consequences were a court to impose primary coverage on a plan, such as that of the TENJ Fund, which intended to provide the nominal, secondary coverage for this group of claimants merely because the plan provides primary coverage for certain other claimants. As the district court recognized, "a court cannot deem one plan primary without shifting unanticipated costs to that plan and frustrating the intent of its trustees." McGurl, 925 F. Supp. at 289.

The Local 1262 Funds argue that the district court erred in failing to follow what they claim is the majority view that entitles a plan that describes itself as a pure excess plan to pay secondarily because its coverage is not implicated until another policy's limits have been exhausted. See Insurance Co. of N. America v. Continental Cas. Co., 575 F.2d 1070, 1071 (3d Cir. 1978) (" [s]ince [an excess clause] does not provide that supplemental protection until the other policy has been exhausted, it is 'excess' to the other coverage"); Institute for Shipboard Educ. v. Cigna Worldwide Ins. Co., 22 F.3d 414, 419 (2d Cir. 1994) (excess plan " 'kicks in' to provide additional coverage once the policy limits of other available insurance are exhausted"). They argue based on these cases that their excess clause exempts their plan from any coordination of benefits with other plans, so that their coverage will "kick in" only after any other plan's coverage is depleted.

As the TENJ Fund points out, the concept of "excess insurance" typically applies to casualty insurance policies which cover a single party for a single risk. See, e.g., Couch on Insurance 2d §§ 62.48-49. The cases cited by the Local 1262 Funds fall within that category. These "pure" excess policies, which are also commonly referred to as "umbrella" policies, are contingent on the existence of another, primary policy, and are intended to provide a separate, additional layer of coverage, never primary coverage. An insured will purchase this separate layer typically at a discounted price because it "will pick up where primary coverages end in order to provide extended protection." Occidental Fire and Cas. Co. of North Carolina v. Brocious, 772 F.2d 47, 53 (3d Cir. 1985). Since such layered policies are "not an attempt by a primary insurer to limit a portion of its risk by labelling it 'excess' nor a device to escape responsibility, they are regarded as a 'true excess over and above any type of primary coverage, excess provisions arising in regular policies in any manner, or escape clauses.' " Id. (quoting 8A J. Appleman, Insurance Law and Practice § 4909.85 at 453-54 (1981)). Rates for excess insurance are set "after giving due consideration to known existing and underlying basic or primary policies". 46 C.J.S. Insurance Law § 1138. Such policies are categorically separate and do not attempt to coordinate with other policies.

The TENJ Fund argues that pure excess coverage as applied in casualty insurance cannot apply to group health plans covering numerous persons where duplicate, overlapping coverage is often likely.¹  Coordination of benefits rules have evolved to cover these circumstances and are routinely followed.

We need not resolve the parties' disagreement as to whether it is theoretically possible or desirable to have pure excess coverage in the group health care context. The relevant portions of both plans' terms are in fact coordination of benefits clauses because both represent a method for determining how and when two plans may be responsible for covering a common beneficiary. In this case, where the claimants are dependent-beneficiaries of the TENJ Fund plan and part-time employee-participants in the Local 1262 Funds plan, each of the plans views itself as "excess" or "secondary" and each looks to the other as primary. Thus, both plans attempt to coordinate benefits with potentially competing plans.

The very terms of the Local 1262 Funds plan manifest an intent to coordinate benefits with other competing welfare plans. Under the heading "Coordination of Benefits," the Local 1262 Funds plan states that: "If a Part-time Member who is a Covered Member ... is also covered under one or more Other Plans, the Benefits payable under this Plan will be coordinated with Benefits payable under all Other Plans." App. at 89 (emphasis added).

Moreover, unlike the prototypical pure excess or umbrella policy, the Local 1262 Funds plan itself contemplates assuming primary liability in instances where there is no other coverage available, and thus no other plan with which to coordinate benefits. The plan's Summary Plan Description states:

This Plan has a coordination of benefits provision for both Full-time and Part-time members. In most instances, this means that if your covered dependents are covered primarily under another medical plan, they can also receive benefits ... from this Plan, up to the amount this Plan would have paid as your primary plan, but only after they receive reimbursement from the other Plan.... The benefits you receive from this Plan cannot exceed the amount this Plan would have paid if it was your primary plan.

App. at 142-43 (emphasis added).

Thus, we agree with the TENJ Fund that the Local 1262 Funds plan's "so called 'always excess' provision is no more than a subtle attempt to impose coordination of benefits using a biased order of benefits determination rule." Brief of Appellee at 27. In sum, the disputed reimbursement provision in the Local 1262 Funds plan is essentially a coordination of benefits provision, and, therefore, does not have a categorically secondary status to every plan with which it comes in conflict. In this case, the coordination of benefits provisions are "mutually repugnant," forcing the court to look outside the plans to resolve the apparent conflict.

ESTABLISHMENT OF ORDER OF BENEFITS DETERMINATION RULE

Federal Common Law

Because the plain terms of the individual plans would not resolve the conflict, the district court exercised its authority to devise federal common law, and settled on the "employer first" rule suggested by the National Association of Insurance Carriers ("NAIC"). See McGurl, 925 F. Supp. at 293. Before examining the merits of the district court's selection, we consider the Local 1262 Funds' objections to the courts' federal common law-making authority to impose a rule of decision independent of state law.

Federal common law refers to the development of legally binding federal rules articulated by a federal court which cannot be easily found on the face of a constitutional or statutory provision. See Larry Kramer, The Lawmaking Power of the Federal Courts, 12 Pace L.Rev. 263, 267 (1992); see also Thomas W. Merrill, The Common Law Powers of Federal Courts, 52 U. Chi. L. Rev. 1, 5 (1985) (" 'Federal common law' ... means any federal rule of decision that is not mandated on the face of some authoritative federal text--whether or not that rule can be described as the product of 'interpretation' in either a conventional or unconventional sense."). Notwithstanding the decision in Erie Railroad Co. v. Tompkins, 304 U.S. 64, 58 S. Ct. 817, 82 L. Ed. 1188 (1938), which curtailed development of general federal common law, the power of federal courts to craft federal rules of decision is established in cases in which a federal common law rule is "necessary to protect uniquely federal interests," Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 426, 84 S. Ct. 923, 939, 11 L. Ed. 2d 804 (1964), such as federal proprietary interests, federal interests in international law and to resolve conflicts among the states, or where "Congress has given the courts the power to develop substantive law," Texas Indus., Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 640, 101 S. Ct. 2061, 2067, 68 L. Ed. 2d 500 (1981).

The Court has recognized that while at times state law would be appropriate, " [t]he desirability of a uniform rule is plain" where "identical transactions subject to the vagaries of the laws of the several states" would lead to great diversity in results. Clearfield Trust Co. v. United States, 318 U.S. 363, 367, 63 S. Ct. 573, 575, 87 L. Ed. 838 (1943). This would be true not only when the issue involves the rights and duties of the United States, as it did in Clearfield Trust, but also when a federal statute encompasses a broad mandate that requires uniform rules to effectuate the congressional purpose. See, e.g., Textile Workers Union of America v. Lincoln Mills of Alabama, 353 U.S. 448, 456-57, 77 S. Ct. 912, 917-19, 1 L. Ed. 2d 972 (1957) (upholding federal jurisdiction for labor-management disputes because of congressional authorization to develop federal common law pursuant to the LMRA); National Soc'y of Prof'l Eng'rs v. United States, 435 U.S. 679, 688, 98 S. Ct. 1355, 1363-64, 55 L. Ed. 2d 637 (1978) (in enacting Sherman Antitrust Act, Congress made "perfectly clear that it expected the courts to give shape to the statute's broad mandate by drawing on common-law tradition").

Justice Jackson, in his famous concurrence in D'Oench, Duhme & Co., Inc. v. FDIC, 315 U.S. 447, 470, 62 S. Ct. 676, 685, 86 L. Ed. 956 (1942), noted that the need to make common law stems from the inability of legislators to anticipate every possible contingency and the impracticability of judges returning all unanswered questions to the legislature. He stated, "Were we bereft of the common law, our federal system would be impotent. This follows from the recognized futility of attempting all-complete statutory codes, and is apparent from the terms of the Constitution itself." Id. Justice Jackson explained further that, "Federal common law implements the federal Constitution and statutes, and is conditioned by them. Within these limits, federal courts are free to apply the traditional common-law technique of decision and to draw upon all the sources of the common law." Id. at 472, 62 S. Ct. at 686 (citing Board of Comm'rs v. United States, 308 U.S. 343, 350, 60 S. Ct. 285, 288, 84 L. Ed. 313 (1939)).

Relevant to the determination whether to adopt a federal rule in this case is the scope of the ERISA preemption provision which states that the provisions of ERISA "shall supersede any and all State laws insofar as they may now or hereafter relate to any employee benefit plan."²  29 U.S.C. § 1144(a). That provision, "conspicuous for its breadth," FMC Corp. v. Holliday, 498 U.S. 52, 58, 111 S. Ct. 403, 407, 112 L. Ed. 2d 356 (1990), was drafted expansively in order to establish pension and welfare benefits "as exclusively a federal concern," Alessi v. Raybestos-Manhattan, Inc., 451 U.S. 504, 523, 101 S. Ct. 1895, 1906, 68 L. Ed. 2d 402 (1981), and to relieve plans of the burden of adhering to diverse state regulations.

Thus, by preempting any law that even relates to ERISA plans Congress anticipated the development of a "federal common law of rights and obligations under ERISA-regulated plans." Pilot Life Ins. Co. v. Dedeaux, 481 U.S. 41, 56, 107 S. Ct. 1549, 1558, 95 L. Ed. 2d 39 (1987). As one court aptly stated, " [w]here state law is preempted and no specific federal provision governs, a court is forced to make law or leave a void where neither state nor federal law applies. In such a situation it is a reasonable inference that Congress intended some law, and therefore federal law, to apply." Wayne Chemical, Inc. v. Columbus Agency Serv. Corp., 426 F. Supp. 316, 322 (N.D. Ill.) (internal quotations omitted), aff'd on other grounds, 567 F.2d 692 (7th Cir. 1977); see also Fox Valley & Vicinity Constr. Workers Pension Fund v. Brown, 897 F.2d 275, 281 (7th Cir.) (en banc) ("When ERISA is silent on an issue, a federal court must fashion federal common law to govern ERISA suits."), cert. denied, 498 U.S. 820, 111 S. Ct. 67, 112 L. Ed. 2d 41 (1990).

Therefore, although a federal court has the discretion to adopt state law as part of a federal rule of decision in order to resolve ERISA-related disputes, see Clearfield Trust, 318 U.S. at 367, 63 S. Ct. at 575, a federal court certainly has the power pursuant to ERISA to reject any state rules, particularly a non-legislative rule such as that promulgated in Starks, which do not complement ERISA's policy goals. As we stated in Northeast, "judge-made rules regarding interpretation of insurance contracts are not the kind of state insurance regulations that the Congress intended to preserve." 764 F.2d at 158 n. 8.

The Local 1262 Funds argue that a uniform coordination of benefits rule sacrifices an important pursuit of intra state uniformity for an overstated goal of inter state uniformity. Specifically, they contend that because state law governs coordination of benefits disputes for non-ERISA regulated plans, the plans will face the risk of different outcomes depending upon whether the competing plan is regulated by ERISA or not. By way of example they cite the unpublished opinion in Zalkin v. Teamsters Local 469 Welfare Fund, No. 92-477 (D.N.J. 1993), which held that under New Jersey order of benefits determination rules the Local 1262 Funds plan, which was an excess plan, would not be primarily liable for an employee-participant of its plan vis-a-vis a non-ERISA regulated plan.

ERISA's statutory mandate is to impose uniformity and predictability for the administration of self-insured plans so that beneficiaries can be guaranteed their expected benefits and so that administrators are not subject to " 'conflicting or inconsistent State and local regulation of employee benefit plans.' " Shaw v. Delta Air Lines, Inc., 463 U.S. 85, 99, 103 S. Ct. 2890, 2901, 77 L. Ed. 2d 490 (1983) (quoting remarks of Senator Williams, 120 Cong. Rec. 29933 (1974)); see also FMC Corp., 498 U.S. at 60, 111 S. Ct. at 408; Pilot Life, 481 U.S. at 56, 107 S. Ct. at 1557-58.

It is not difficult to foresee the complications and "considerable inefficiencies" that would arise from having a "patchwork scheme" of differing state coordination of benefits rules. See Fort Halifax Packing Co., Inc. v. Coyne, 482 U.S. 1, 11, 107 S. Ct. 2211, 2217, 96 L. Ed. 2d 1 (1987); Keystone Chapter, Associated Builders and Contractors, Inc. v. Foley, 37 F.3d 945, 954 (3d Cir. 1994), cert. denied, 514 U.S. 1032, 115 S. Ct. 1393, 131 L. Ed. 2d 244 (1995). See generally Helitzer, Coordination of Benefits at 411-15. For example, an ERISA-regulated plan may cover thousands of participants, some of them residing in jurisdictions that follow NAIC order of benefits determination rules and some residing in jurisdictions that do not. That plan's obligation to assume primary liability for a claimant would depend on the fortuity of a claimant's place of residence and application of that state's coordination of benefits law, a consequence clearly disfavored by ERISA. See Alessi, 451 U.S. at 523-26, 101 S. Ct. at 1906-08 (ERISA preempts state statute that would force employer to adopt different payment formulae for employees inside and outside state). As stated by the Ninth Circuit in PM Group, such indeterminacy and conflict "would almost certainly lead to litigation, thereby burdening the insured employees, the providers of covered services, and the plans themselves as well as the federal courts. Adoption of a uniform federal rule avoids such confusion and expense, and thus best serves the purposes of ERISA." 953 F.2d at 547; see also FMC Corp., 498 U.S. at 60, 111 S. Ct. at 408 ("To require plan providers to design their programs in an environment of differing state regulations would complicate the administration of nationwide plans, producing inefficiencies that employers might offset with decreased benefits").

Contrary to the Local 1262 Funds' suggestion, there are very few instances in which the federal common law is concerned with promoting uniformity within a state. This is not a situation analogous to those where the Court has approved adoption of a state's law when the federal statute incorporates a matter which is one primarily of state concern. See De Sylva v. Ballentine, 351 U.S. 570, 580, 76 S. Ct. 974, 979-80, 100 L. Ed. 1415 (1956) (instructing federal courts to defer to state law for meaning of terms like "children" and "widower" in the relevant portions of the federal Copyright Act rather than formulate a federal law of domestic relations); Reconstruction Fin. Corp. v. Beaver County, 328 U.S. 204, 209-10, 66 S. Ct. 992, 995, 90 L. Ed. 1172 (1946) (definition of "real property" which Congress authorized to be taxed should be defined by settled state rules, because Congress obviously contemplated various results among the states).

There is no evidence that state commercial or other domestic interests would be upset by imposition of uniform federal order of benefits determination rules. In fact, New Jersey law regulating "other insurance" provisions is similar to the NAIC Model Regulation in all major respects except that it does not provide the complying plan with the right to sue the noncomplying plan for subrogation. See N.J. Admin. Code tit. 11, § 4-28.9. Nevertheless, because of preemption the possibility that the Local 1262 Funds plan will face a different rule, and thus disuniformity within the state, when it conflicts with a non-ERISA insured plan, is of little concern under ERISA. Cf. Keystone, 37 F.3d at 959 n. 19 ("While state regulations may affect the cost of doing business in a state, they may not, consistent with ERISA, place administrative burdens and costs on ERISA plans that make it impractical for an employer to provide a nationwide plan.").

We thus conclude that Congress envisioned establishment by the federal courts of a uniform set of federal rules rather than subjecting to diverse state laws ERISA-regulated plans involving competing benefits clauses. See PM Group, 953 F.2d at 547; Northeast, 764 F.2d at 158.

Selection of the "Employer First" Rule

As we have stated, the district court exercised its common law-making authority to select the "employer first" rule advocated by NAIC as the method for determining which competing ERISA plan should pay the claimed benefits. In 1970, in order to deal with the increasing problem of duplicate coverage, NAIC, an independent group of state insurance regulatory commissioners, promulgated a set of rules under the heading of Group Coordination of Benefits Model Regulation ("Model Regulation"), based in large part on rules that had been established and followed by the group insurance industry in the previous decade. See Helitzer, Coordination of Benefits, at 413-14. The Model Regulation contains a recommended order of benefits scheme covering potential conflicts among health benefit plans or policies. NAIC recommendations do not have the force of law, but many states have incorporated part or all of particular recommendations into their insurance statutes.

Jack Helitzer, who was the former chairman of the Industry Advisory Committee to the NAIC Task Force on Coordination of Benefits, attributes the widespread acceptance of the recommended regulation to the participants' need for absolute uniformity in this area. "The validity of the [coordination of benefits] rules is established not by law or regulation, but rather by the fact that there will be chaos without uniform rules to determine the order of benefit payment." Id. at 412.

The sequential system for determining the order of payment by benefit plans in the comprehensive Model Regulation provides, as relevant here, that the "benefits of the plan which covers the person as an employee, member or subscriber (that is, other than as a dependent) are determined before those of the plan which covers the person as a dependent." Model Regulation § 5B(1) (quoted in Helitzer, Coordination of Benefits, at 414). As Helitzer explains, " [t]he plan covering the person as an employee pays benefits first. The plan covering the same person as a dependent pays benefits second." Helitzer, Coordination of Benefits, at 415.

The Model Regulation does not recognize excess or always secondary plans or incorporate them into the order of benefits scheme because such clauses "will doom at least some of their employees to a double-secondary situation, in which the individual has double coverage and neither plan has the obligation to pay anything substantial. Responsibility for the resulting problem lies with the employer or plan that adopts a unique order of benefits determination rule, and not with the one who follows accepted practices." Id. at 421-22.

If the excess plan refuses to pay for primary coverage, when it would be obligated to pay as primary under the Model Regulation, the plan that would be secondary is instructed to advance to the claimant the amount it would have paid as primary and execute a right of subrogation against the noncomplying plan. Model Regulation § 7(B).

The NAIC approach to conflicts involving always excess coordination of benefits provisions has garnered widespread acceptance among the states. Twenty-four states have adopted the NAIC order of benefits determination rules in full, providing a right to subrogation against noncomplying plans. Jack B. Helitzer, State Developments in Employee Benefits: State Adoption of Coordination of Benefits Rules,4 Benefits L.J. 435, 442-43 (1991). Fifteen states, including New Jersey, have adopted the NAIC order of benefits scheme but, unlike the other twenty-four states, they do not incorporate the subrogation rule as a vehicle whereby a complying plan can compel payment from a noncomplying plan.³  Id. at 443.

In deciding to adopt the NAIC recommended "employer first" rule, the district court explained that the NAIC rule would provide a uniform coordination of benefits scheme and thereby would best further the statutory objectives of ERISA. McGurl, 925 F. Supp. at 293. The Local 1262 Funds urge us to overturn that decision in favor of a pro-rata rule pursuant to which the two plans would divide coverage on an equal basis. They contend the pro-rata rule is more equitable and is the rule applied in the majority of federal and state courts. They cite our opinion in Northeast, where we noted that other courts faced with incompatible other insurance clauses have chosen the pro-rata formulation. 764 F.2d at 161 n. 13. In fact, the majority of states have not adopted the pro-rata rule for health benefit plans, such as those covered by ERISA. And in Northeast we did not consider the merits of the pro-rata rule and thus the statement on which the Local 1262 Funds rely was merely dictum which is not binding upon our consideration here.

The two courts of appeals that have considered whether to apply the pro-rata rule, albeit under somewhat different circumstances than those presented here, have divided on its merits. In Winstead v. Indiana Ins. Co., 855 F.2d 430, 432 (7th Cir. 1988), cert. denied, 488 U.S. 1030, 109 S. Ct. 839, 102 L. Ed. 2d 971 (1989), the Seventh Circuit considered a conflict between the coordination of benefits clauses in a claimant's ERISA-regulated health and welfare fund and an applicable no-fault automobile insurance policy regulated by Michigan law. After finding the plan and the policy to be irreconcilable, the court affirmed the district court's decision to hold both insurers liable on a pro-rata basis, citing our dictum in Northeast. Id. at 434.

More recently, in Auto Owners Ins. Co. v. Thorn Apple Valley, Inc., 31 F.3d 371, 375 (6th Cir. 1994), cert. denied, 513 U.S. 1184, 115 S. Ct. 1177, 130 L. Ed. 2d 1129 (1995), the Sixth Circuit rejected the application of a pro-rata liability rule between an ERISA plan and an insurance policy nearly identical to those considered in Winstead. The court concluded that in light of ERISA's broad preemption provision, the ERISA-regulated plan's diversion of liability should be given priority over the state plan's attempt to do the same. Id. at 374. The court reasoned that although the pro-rata rule may be equitable in the context of two non-regulated, private plans, the equal apportionment formula would not "comply with a primary goal of ERISA, which is to safeguard the financial integrity of qualified plans by shielding them from unanticipated claims." Id. at 375.

Of course, neither case is apposite here because those courts were not presented with conflicts in "other insurance" clauses in which both plans are regulated by ERISA. However, in rejecting the pro-rata rule, the court in Auto Owners gave dispositive weight to the policy considerations underlying ERISA, a principal consideration in the district court's selection here.

In PM Group, 953 F.2d at 547-48, the Ninth Circuit, faced with incompatible "other insurance" provisions in two self-funded ERISA plans, relied in part on NAIC regulations to create a uniform federal common law solution. In that case, a husband and wife were each covered primarily by their respective plans for hospital expenses related to the premature birth of their daughter. The father's plan provided that, in all cases, the father's plan would be the primary insurer (the "gender rule") while the mother's plan provided that the plan covering the parent whose birthday fell earlier in the year--in this case, the mother's--was the primary plan (the "birthday rule"). Id. at 548.

The court recognized that ERISA was silent on the issue of conflicting "other insurance" provisions and decided that it must craft a common law rule that would take account of ERISA's stated goal of uniformity. Id. at 547 ("uniformity enables employers 'to predict the legality of proposed actions without the necessity of reference to varying state laws' " (quoting Pilot Life, 481 U.S. at 56, 107 S. Ct. at 1557)). The court then looked for guidance to the NAIC Model Regulation, which provided an order of benefits rule for such a scenario, and adopted the "birthday rule" for resolving all such conflicts in ERISA-regulated plans.⁴  Id.

The district court reasoned that adoption of a pro-rata rule would have the effect of encouraging welfare plans to adopt excess clauses in order to avoid the disadvantage, vis-a-vis a plan with excess or always secondary reimbursement provisions, of having to assume primary liability if the claimant is an employee of a plan participant, and 50% liability if the claimant were not an employee. McGurl, 925 F. Supp. at 293. By contrast, the excess or always secondary plans would never have to assume more than 50% liability. An excess or always secondary plan will invariably save money by reducing the plan sponsor's cost of providing health care coverage to their employees.

The district court concluded that this incentive would produce a "race to the bottom" in the context of reimbursement provisions. Id.; see also Helitzer, Coordination of Benefits, at 421 ("If any plan can be free to set its own rules to determine the order of benefits, every other similarly situated plan should also be free to do the same. When other plans are affected by such a cost shift, they would have to be encouraged to adopt similar, always-secondary approaches causing large scale chaos"). In a regime where all plans have always excess provisions but no governing uniform coordination of benefits rule, resolution of a particular conflict between two such plans would depend on an ad hoc judicial determination. This would jeopardize the predictability and certainty for plan sponsors and beneficiaries that was central to ERISA's enactment. See Auto Owners, 31 F.3d at 375; Northeast, 764 F.2d at 163.

As we acknowledged in Northeast,

even a qualified endorsement of escape clauses might encourage benefit plans with excess or coordination of benefits clauses to replace such clauses with those of the escape variety in order to "fight fire with fire." A war between plans would cause uncertainty in the industry and could potentially catch participants and beneficiaries in the crossfire.

764 F.2d at 164 n. 17. It is also of some interest that health care insurance contracts subject to New Jersey regulation are not permitted to include always secondary provisions. N.J. Admin. Code tit. 11, § 4-28.5(b).

We are concerned that adoption of the pro-rata rule which the Local 1262 Funds propose would present some serious difficulties when two self-insured ERISA plans cover a family member as an employee-participant and dependent respectively. In the first place, it is unclear how the rule would operate in practice. Although a pro-rata rule may technically encompass proportional payment rather than the 50-50 payment the Local 1262 Funds suggested here, the Funds were unable to explain precisely how proportional payment would be fixed. Benefit plans are unlike casualty insurance, which is the field in which pro-rata payments primarily operate. There has been no satisfactory explanation of its feasibility in the medical benefits field, where different plans have different deductibles and coverages. Its operation under managed care programs is also uncertain. Counsel conceded at oral argument that calculation of the final benefit allocation pursuant to a pro-rata formula, which counsel presumed would be based on each plans' proportional primary liability coverage, is considerably more complicated than under the more traditional primary versus secondary scale. It may be that it was these difficulties that led the vast majority of states to adopt the NAIC recommended "employer-first" rule.

The district court also noted that the "employer-first" rule has been "incorporated into most self-insured employee benefit plans," McGurl, 925 F. Supp. at 292, and, in fact, the brief of the amici welfare funds confirms that their plans include such a provision. See infra note 5. Significantly, the Local 1262 Funds and the TENJ Fund themselves have adopted the Model Regulation "employer first" rule to govern employee/dependent conflicts with regard to coverage for their full-time employees, and thus the Local 1262 Funds only resist its applicability to their part-time employees.

Moreover, the "employer first" rule validates the natural disposition of an employee to look to his or her own employer for health care benefits as a reward for his or her own labor. Most important, the rule also allows employers to predict with more accuracy the extent of their own potential liability because it is easier to calculate the number of a plan's own employee-participants for which it is responsible than the uncertain but likely greater number of those employees' dependents who will look to the plan for their primary coverage.

The final objection by the Local 1262 Funds to the imposition of an "employer first" order of benefits determination rule is that the prospect of having to provide primary coverage to part-time plan participants might force them to discontinue providing welfare benefits altogether. They argue that this would undermine ERISA's goal of not deterring the creation of employee benefit plans. They cite Hozier v. Midwest Fasteners, Inc., 908 F.2d 1155, 1160 (3d Cir. 1990), where we stated that " [h]aving made a fundamental decision not to require employers to provide any benefit plans, Congress was forced to balance its desire to regulate extant plans more extensively against the danger that increased regulation would deter employers from creating such plans."

It is true, as we explained in Nazay v. Miller, 949 F.2d 1323, 1329 (3d Cir. 1991), that " [i]n enacting ERISA, Congress did not impose a duty on employers to provide health care and other benefits to their employees. Rather, the clear emphasis of the statute is to ensure the proper execution of the plans once established." See also Hlinka v. Bethlehem Steel Corp., 863 F.2d 279, 283 (3d Cir. 1988). But it is merely speculation that if the Local 1262 Funds are obliged to provide primary coverage for part-time employees they will be unable to afford any coverage whatsoever. Nor is there a sound reason for giving the financial interests of the Local 1262 Funds priority over those of the TENJ Fund, which naturally faces similar concerns about managing the escalating health care costs for part-time employees.⁵  Moreover, these plans have been established by collective bargaining, and it is in that process that inclusion, vel non, will be decided.

Thus, in weighing the interests served by ERISA against the negative effects generated by a rule that favors "always secondary" plans and thereby induces all plans to structure their benefits similarly, to the ultimate detriment of participants, we conclude that the balance is heavily in favor of the "employer first" rule. Over the long-run, a uniform "employer first" rule is actually more equitable since, assuming a generally even distribution of employees and dependents among various plans, plans such as those at issue here will tend to be primary half of the time and secondary half of the time. The "employer first" rule advances the goals of preserving "the financial integrity of qualified plans by shielding them from unanticipated claims," Auto Owners, 31 F.3d at 375, and preventing participants from being "deprived of compensation that they reasonably anticipate under the plan's purported coverage," Northeast, 764 F.2d at 163.VI.

CONCLUSION

We thus conclude that in those instances where the plans have competing provisions with respect to persons covered by both plans, the "employer first" rule provides the most appropriate basis for apportioning liability under federal common law for self-insured benefit plans regulated by ERISA. We will affirm the district court's grant of summary judgment.