These facts show how rare and difficult it is rationally and

logically to combine all the several parts of

legislation.1

We review the district court's construction of a recent federal statute, The Product Liability Risk Retention Act of 1981.2 Most immediately at issue is the validity of an insurance marketing enterprise with outstanding insurance contracts valued at $70,000,000,000. Our decision also inevitably implicates the balance of state and federal power in the regulation of insurance sales.

The question presented is whether an insurance organization selling what Georgia law would characterize as a home owner's policy may, by virtue of federal law, insure homes sold to Georgia home owners without complying with otherwise applicable Georgia insurance regulations. The Product Liability Risk Retention Act of 1981 is the basis for this claim of exemption. The district court, accepting the appellee's,3 Home Warranty's, arguments, permanently enjoined the appellant, the Insurance Commissioner for the State of Georgia, from interfering with the appellee's activities in the state, except insofar as authorized by the federal act. The Commissioner appealed. Under the terms of the Risk Retention Act, the dispositive issues are these: (1) whether Home Warranty assumes or distributes a product liability risk; (2) whether that is its primary activity; and (3) whether it assumes or distributes this risk on behalf of its members.

We address an area of the law that is unsettled4 and rife with far-reaching political and economic choices. In adding judicial gloss to this field we are mindful that these choices are for the Congress, not this court. Yet we must also recognize that our interpretation of congressional action is writ large in practical effect. For that reason, we seek to ground our opinion upon the broadest range of information available to us. We begin, in Part I of the opinion, with the evolution of product liability. Part II examines the economic consequences of that liability, the climate in which the Congress enacted the Product Liability Risk Retention Act. Part III reviews the development and passage of the Act itself, its legislative history, text and amendments. Part IV examines the development and purpose of the Home Warranty Corporation, its subsidiaries, its product, and prior judicial conclusions drawn about the activities of this entity. In Part V we set out the arguments presented, our decision, and its rationale.

I. EVOLUTION OF PRODUCT LIABILITY LAW IN THE UNITED STATES

Product Liability, as a branch of tort law, is traditionally concerned with one's duty to third parties. It begins with the notion that, "by entering into a contract with A, the defendant may place himself in such a relation toward B that the law will impose upon him an obligation, sounding in tort and not in contract, to act in such a way that B will not be injured." W. Prosser, The Law of Torts 622 (4th Ed.1971). The crucial point is that this duty arises by operation of law and not by operation of the bargain of the parties themselves. In this way the courts obviated the earlier necessity of contract as the basis of duty, the celebrated "privity" requirement. See National Savings Bank v. Ward, 100 U.S. 195, 25 L. Ed. 621 (1879); Bohlen, Fifty Years of Torts, 50 Harv.L.Rev. 1225, 1232 (1937). Under this evolving theory, as it concerned a plaintiff, the fact of contractual relations between the defendant and A was incidental; liability was predicated upon the expectation which the defendant had, or should have had, that his affirmative conduct toward A would affect the interests of B, a plaintiff. Prosser, supra at 622. In the years intervening between the inception of this concept and the present the expansion of liability arising out of product sales has been " 'spectacular' in the extreme...." Id.

The erosion of the nineteenth century "privity" requirement began in the New York case of Thomas v. Winchester, 6 N.Y. 397 (1852), in which it was held that a seller who negligently prepares and sells an article inherently or imminently dangerous to human life may be held liable to one injured by the article on the basis of his negligence alone. Prosser, supra at 642. The existence or absence of liability under this vague standard was uncertain for a number of years.5 It remained for Judge Cardozo to gather these miscellaneous authorities into an articulable formulation in the case of MacPherson v. Buick Motor Co., 217 N.Y. 382, 111 N.E. 1050 (1916). Although purporting merely to expand the "inherently dangerous" exception to the general requirement of privity, the case as persuasively reasoned exploded the general rule itself. The dangerous nature of an instrumentality would or should lead a manufacturer to foresee potential harm caused by the negligent design or manufacture of that instrumentality. Out of this foreseeability arose a duty to design and manufacture the instrumentality with due care.

If the nature of a thing is such that it is reasonably certain to place life and limb in peril when negligently made, it is then a thing of danger. Its nature gives warning of the consequences to be expected. If to the element of danger there is added knowledge that the thing will be used by persons other than the purchaser and used without new tests, then, irrespective of contract, the manufacturer of this thing of danger is under a duty to make it carefully.

MacPherson, 217 N.Y. at 389, 111 N.E. at 1053. In the aftermath of MacPherson it became clear "that the duty is one imposed by the law because of the defendant's affirmative conduct, which he must know to be likely to affect the interests of another." Prosser, supra at 643.

Over time the MacPherson rule was extended to include harm to property, see e.g., United States Radiator Corp. v. Henderson, 68 F.2d 87 (10th Cir.1933); Todd Shipyards Corp. v. United States, 69 F. Supp. 609 (D.Me.1947), and to liability caused by goods which involved no recognizable risk of personal injury. See e.g., Cohan v. Associated Fur Farms, Inc., 261 Wis. 584, 53 N.W.2d 788 (1952); Dunn v. Purina Co., 38 Tenn.App. 229, 272 S.W.2d 479 (1954); Brown v. Bigelow, 325 Mass. 4, 88 N.E.2d 542 (1949). Broadened from its traditional protection of the purchaser, the MacPherson rule was also extended to cover employees, see e.g., Rosebrock v. General Electric Co., 236 N.Y. 227, 140 N.E. 571 (1923); Marsh Wood Products Co. v. Babcock & Wilcox, 207 Wis. 209, 240 N.W. 392 (1932) (dictum); Kalash v. Los Angeles Ladder Co., 1 Cal. 2d 229, 34 P.2d 481 (1934) (dictum); O'Donnell v. Geneva Metal Wheel Co., 183 F.2d 733 (6th Cir.1950), members of the purchaser's family, see e.g., White Sewing Machine Co. v. Feisel, 28 Ohio App. 152, 162 N.E. 633 (1927); Baker v. Sears, Roebuck & Co., 16 F. Supp. 925 (S.D.Cal.1936), subsequent purchasers, Beadles v. Servel, Inc., 344 Ill.App. 133, 100 N.E.2d 405 (1951); Quackenbush v. Ford Motor Co., 167 A.D. 433, 153 N.Y.S. 131 (1915); State ex rel. Woodzell v. Garzell Plastics Industries, Inc., 152 F. Supp. 483 (E.D.Mich.1957), other users of the defective chattel, Hoenig v. Central Stamping Co., 273 N.Y. 485, 6 N.E.2d 415 (1936); Lill v. Murphy Door Bed Co., 290 Ill.App. 328, 8 N.E.2d 714 (1937); Reed & Barton Corp. v. Maas, 73 F.2d 359 (1st Cir.1934); Coakley v. Prentiss-Wabers Stove Co., 182 Wis. 94, 195 N.W. 388 (1923), casual bystanders, see e.g., McLeod v. Linde Air Products Co., 318 Mo. 397, 1 S.W.2d 122 (1927); Statler v. George A. Ray Manufacturing Co., 195 N.Y. 478, 88 N.E. 1063 (1909) (dictum); Hopper v. Charles Cooper & Co., 104 N.J.L. 93, 139 A. 19 (Ct. of Error & Appeal 1927), and others located in the general vicinity of the chattel's probable use. Restatement, Torts Sec. 395 (1934); Flies v. Fox Brothers Buick Co., 196 Wis. 196, 218 N.W. 855 (1928); Gaidry Motors, Inc. v. Brannon, 268 S.W.2d 627 (Ky.1954); Carpini v. Pittsburg & Weirton Bus Co., 216 F.2d 404 (3d Cir.1954); Whitehead v. Republic Gear Co., 102 F.2d 84 (9th Cir.1939) (dictum); Ford Motor Co. v. Zahn, 265 F.2d 729 (8th Cir.1959). This rapid expansion of liability notwithstanding, the cases shared a common theme: the operative theory of liability was negligence--a duty arising in the manufacturer by virtue of the foreseeability of harm. While these developments were taking place another branch of products liability was developing.

In the early part of the twentieth century a national uproar over the sorry state of manufactured food began. A dismayed public learned that its daily bread (and other eatables) was in large part adulterated and not fit to eat.6 This public concern resulted in a national movement for the purification of food and drugs. The common law fell in line with the procession. The Supreme Court of Washington, in Mazetti v. Armour & Company, 75 Wash. 622, 135 P. 633 (1913), held a meat packer strictly liable to the consumer without privity. The basis of liability was an implied representation that the food was safe to eat. Prosser, Assault Upon the Citadel (Strict Liability to the Consumer ), 69 Yale L.J. 1099, 1106 (1960) [hereinafter cited as Assault Upon the Citadel ]. Other states followed suit.7 Finally, in 1927, the Mississippi Supreme Court decided the case of Coca Cola Bottling Work v. Lyons, 145 Miss. 876, 111 So. 305 (1927), which drew the analogy between the sale of goods and the sale of land. As a covenant might run with land upon its sale, so a "warranty" of a manufacturer ran with a product upon its ultimate sale to a consumer. However, where the real property covenant provided the land purchaser with a right to recover the value of good title to the land, the covenant implied in the sale or distribution of a product created a right to recover for injury resulting from the use of the product. Strict liability on a theory of warranty for food was soon thereafter established as "the law of the future." See W. Prosser, Assault Upon the Citadel at 1110.

Predictably, strict liability predicated upon warranty also gathered momentum in areas beyond the manufacture and distribution of foodstuffs. The rationale for this rapid expansion was the perceived deficiency of negligence as a theory of liability. Negligence allowed recovery only against the manufacturer and "[t]here are other sellers than the manufacturer of the product.... If the plaintiff is to recover at all, he must often look to the wholesaler, the jobber, and the retailer." W. Prosser, Assault Upon the Citadel at 1116-17. Negligence often could not be proved against these parties. Moreover, the means of recovery afforded under prior law, the old sales warranties of merchantable quality and fitness for the purpose,8 were unavailing because of their traditional requirement of privity. Absent privity, warranties of fitness would protect neither the purchaser nor those in his immediate environment, his "wife or child, his employee, his guest, his donee, or his subpurchaser." W. Prosser, Assault Upon the Citadel at 1118 (citations omitted). For this reason there was a movement in the law to expand warranty liability for tainted food to include warranty liability for other products without the necessity of privity or even negligence.

This movement reached its apogee in the famous case of Henningsen v. Bloomfield Motors, Inc., 32 N.J. 358, 161 A.2d 69 (1960). In that case the Supreme Court of New Jersey stated, after surveying other cases in which privity was lacking, including food and drug cases:

We see no rational doctrinal basis for differentiating between a fly in a bottle of beverage and a defective automobile. The unwholesome beverage may bring illness to one person, the defective car, with its great potentiality for harm to the driver, occupants, and others, demands even less adherence to the narrow barrier of privity.

Henningsen, 32 N.J. at 383, 161 A.2d at 83. So stating, the court upheld the judgment against the dealer of an automobile for damage and personal injury caused by that automobile to one not in privity with the dealer and unable to show that the damage and injuries occurred as a consequence of the dealer's negligence. Strict liability on a theory of warranty was established. See W. Prosser, The Fall of the Citadel (Strict Liability to the Consumer ), 50 Minn.L.Rev. 791, 792-93 (1966) [hereinafter cited as Fall of the Citadel ]. It is interesting in the context of our present inquiry to note that the rationale underlying this liability was distribution of risk--liability should be imposed upon the supplier of a product because he is in the position to recover the cost of that liability by an increase in price thus distributing the risk among all consumers of a product.9

Further development of product liability occurred quickly. Other cases soon adapted the warranty theory, borrowed from Henningsen which had, in turn, borrowed it from the earlier food and drug cases,10 to a wide variety of product liability claims. The fictive warranty concept proved unwieldy and was eventually discarded. As Prosser relates:

[W]arranty, a freak hybrid born of the illicit intercourse of tort and contract, had always been recognized as bearing to some extent the aspects of a tort. In time the idea of running with the goods was discarded, and the warranty was considered to be made directly to the consumer. Until 1962 warranty had held the field, and no court proceeded on any other basis, although a good many of them had realized that this was a new and different kind of "warranty," not arising out of or dependent upon any contract, but imposed by law, in tort, as a matter of policy.

....

[T]he suggestion was sufficiently obvious that all of the trouble lay with the one word "warranty" which had been from the outset only a rather transparent device to accomplish the desired result of strict liability. No one disputed that the "warranty" was a matter of strict liability. No one denied that where there was no privity, liability to the consumer could not sound in contract and must be a matter of tort. Why not, then talk of the strict liability in tort, a thing familiar enough in the law of animals, abnormally dangerous activities, nuisance, workmen's compensation, libel, misrepresentation, and respondeat superior, and discard the word "warranty" with all its contract implications?

W. Prosser, The Fall of the Citadel at 800-802. The Second Restatement of Torts adopted this view. See Restatement (Second) Torts Sec. 402(a) & comment m at 355-56 (1965). Case law to this effect was also soon established. See Greenman v. Yuba Power Products, Inc., 59 Cal. 2d 57, 377 P.2d 897, 27 Cal. Rptr. 697 (1963).11 Strict liability for defective products had come of age.

This historical retrospective illustrates how rapid and radical was the transformation of product sales law over the last century. It is the pace of this change rather than the new substance of the law which here concerns us. As the author of MacPherson wrote in another context, we look not to the law but to the path of its advance--"neither a straight line, nor a curve ... [but] ... a series of dots and dashes ... [p]rogress [coming] per saltum...." B. Cardozo, The Paradoxes of Legal Science, 26-27 (1928). In all this, one thing is clear. The ultimate risk of injury resulting from the product was transferred from the injured person (whether purchaser, user, or otherwise) to those parties who put the product in the stream of commerce. Product liability created and expanded an exposure to damages imposed by law upon the providers of products who were perceived as being able to distribute the risk of this exposure by price increases and, ultimately, by insurance providing the providers with protection against the exposure to liabilities--products liability insurance.12

II. THE ECONOMIC CONSEQUENCES OF PRODUCT LIABILITY--PRODUCT

A new economic theory was taking shape beside the new legal theories of product liability.13 Accompanying the case law's drift from its mooring of privity as the precondition of liability was an increased judicial willingness to abandon the rigors of laissez-faire for a new policy of risk distribution through pricing decisions made in the marketplace. See generally R. McKean, Products Liability: Trends and Implications, 38 U.Chi.L.Rev. 3 (1970) [hereinafter cited as Trends and Implications ]. By expanding the scope of liability to encompass manufacturers, wholesalers, and retailers, the courts were forcing those parties to modify their products in the interest of safety and to pass on the costs of these modifications to the ultimate consumers. In this way the "cost" of five less amputated fingers, for example, might be borne by all users of the product in question.

The wisdom of this new policy has been hotly debated, its advent greeted as often with hosannas as with bitter lament. Whatever its merit, however, product liability clearly seeks to accomplish two social aims. One is compensatory, the reallocation of the costs of injury in an industrial society. The other is normative, the reduction of injury through the enforcement of certain minimal standards of care in the manufacture and sale of products. In order to accomplish these aims settled expectations were altered, with the inevitable result of uncertainty in the marketplace. Predictably, this uncertainty had economic consequences. These consequences were particularly apparent in two areas. One was an increase in the size of awards given by courts and juries in particularly severe cases of injury or in cases involving multiple victims. Drug litigation, asbestosis cases and airline crashes are examples coming readily to mind. Another was an increased reliance upon product liability insurance as a means of leveling the costs of unpredictable awards, the usual function of insurance as risk distribution. We turn to particular examples.

In June, 1976, a Wall Street Journal article called attention to a problem already very much on the minds of American businesses. Citing an example of one manufacturer forced to close its doors and idle its eighty employees,14 the article went on to examine the pervasive problem of increased products liability litigation, increased claims and claim awards and, particularly, the increasing difficulty encountered by manufacturers in obtaining affordable product liability insurance. The Wall Street Journal, June 3, 1976, at 1, Col. 6, quoted in W. Keeton, D. Owen & J. Montgomery, Products Liability and Safety (1980). As the article noted, the options available to a company were often unpalatable. In many instances annual premiums for product liability coverage amounted to as much as six percent of gross sales. Instances in which insurance rates jumped twenty-five fold in a single year were not uncommon.15

In 1976 President Ford created the Federal Interagency Task Force on Product Liability. Working under the auspices of the Department of Commerce, the task force was directed to study the factors giving rise to the perceived product liability problem. In 1977 the task force concluded its investigation and issued a Final Report. That report began with the following observation.

In 1975, an apparent problem (some sources said 'crisis') arose in the field of product liability. A number of manufacturers and business periodicals alleged that product liability insurance had become unavailable or unaffordable. The consequences of this situation included the possibilities that business might terminate because they were unable to get coverage; that injured persons would be unable to enforce product liability judgments; and that manufacturers would be hesitant to produce some products that would be useful in our society. It was also alleged that the system of private insurance in the field of product liability was breaking down. Finally, it was alleged that relatively few injured persons benefited from the system.

Final Report, Interagency Task Force on Product Liability I-1 (1977). The findings of the task force indicated that, to an uncertain degree, the concerns over product liability availability and the breakdown of the private product liability insurance system were overblown. While it was conceded that a problem existed, that problem was not, so the task force found, of crisis proportions. The problem which did exist was determined by the task force to be caused by three principal factors. While it was not able to rank these factors according to the significance of their contribution to the problem, it was able to articulate them.

Partly at fault were the ratemaking procedures of liability insurers. Product liability policy premiums remained static for a number of years after significant legal changes had affected the potential exposure of insureds. When significant losses were sustained by the insurers on the basis of product liability claims, the insurers, foreseeing a devastating new trend in adjudication of product liability claims, often engaged in overreactive escalations of premium prices--"panic pricing." Prices were often set without regard for actual claims experience data, data the industry stated it could not compile. A second factor, according to the task force, was found in manufacturing practices which were in fact unsafe. It was concluded by the task force that in many respects litigation in these areas was commercially reasonable and necessary to fulfill the normative and compensatory policies of product liability law in general. Finally, uncertainties in the judicial and statutory treatment of product liability claims contributed significantly to the problem. Variations in tort law from state to state and the lack of any predictable means of foreseeing potential product exposure contributed to the inevitable confusion clouding this field. That uncertainty was reflected in the commercial practices of those operating in the product liability insurance marketplace. See Birnbaum, Legislative Reform or Retreat? A Response to the Product Liability Crisis, 14 Forum 251 (1978).

With these findings, the task force concluded that an accurate means of gathering data on product liability claims nationwide was urgently needed. Only when this information was obtained could accurate forecasts and reasonable reactions to the product liability problem be made. U.S. Department of Commerce, Interagency Task Force on Product Liability, Final Report V-48 (1977). The stage was set for legislative action.

III. THE DEVELOPMENT, PASSAGE, AND AMENDMENT OF THE PRODUCT

In April, 1977, the Senate held hearings on Senate Bill 403, a bill "to regulate the flow of interstate commerce by establishing programs, standards, and procedures for determining responsibilities and liabilities arising out of product related injuries, and for other purposes." National Product Liability Insurance Act: Hearings on S. 403 Before the Subcommittee for Consumers of the Senate Committee on Commerce, Science, and Transportation, 95th Cong., 1st Sess. (1977) [hereinafter cited as 1977 Hearings ]. This legislation proposed to address national problems of product liability availability and affordability by establishing a national product liability insurance administration, a national product liability arbitration program, and a product liability insurance pool which would increase the availability of insurance coverage. Additionally, the Act sought, through federal financial assistance, to encourage states to participate in product liability arbitration programs in a way consistent with national standards for the creation and regulation of such programs. 1977 Hearings at 6-7.

In three days of hearings the committee received testimony much of which was duplicative of that already established by the Interagency Task Force. However, a new idea also emerged from the testimony. In addition to manifest concern over the unpredictable nature of product liability recoveries and escalating insurance premiums, there appeared the first rumblings of legislative concern over the federal nature of the problem--the interstate implications of product liability itself.16B. The 1979 Hearings

Senate Bill 403 was not enacted into law. However, the subject of national product liability reform was again taken up by the next Congress. In October and November of 1979, the House Subcommittee on Consumer Protection and Finance of the Committee on Interstate and Foreign Commerce held hearings on a series of bills addressing the product liability problem. In an opening statement, Representative Rinaldo, a member of the Subcommittee, made the following remarks:

In October of 1977, the interagency task force on product liability reported that most businesses were experiencing significant increases in their product liability premiums. The report stated that many industries with high risk product lines were having difficulty obtaining product liability insurance at affordable rates. The report also observed that some industries with product liability insurance were being forced to accept deductibles so high that it amounted to self-insurance. Finally, the report noted that other businesses were unable to obtain product liability insurance at any cost, and that an increasing number are going without insurance coverage.

Two years after that report, the subcommittee has heard testimony that all of these problems still exist....

....

Two of the primary causes of the product liability problem identified by the interagency task force are, first, questionable insurer ratemaking and reserving practices and second, two types of uncertainties in the tort law.

The legislation pending before the subcommittee attempts to address both of these causes. We are looking forward to hearing from the witnesses today as to the merits of the various bills before us. We shall be considering the Product Liability Risk Retention Act, one approach to the problem, as well as several tort reform bills.

Product Liability Risk Retention Act: Hearings on H.R. 5571, H.R. 5258, H.R. 1061, H.R. 2891, H.R. 4204, H.R. 1675, H.R. 1676, H.R. 2964, H.R. 5626 Before the Subcommittee on Consumer Protection and Finance of the House Committee on Interstate and Foreign Commerce, 96th Cong., 1st Sess. 1-2 (1979) [hereinafter cited as 1979 Hearings ].

For present purposes the pertinent provisions of these hearings pertain to the Product Liability Risk Retention Act of 1979, H.R. 5571 and H.R. 5258 (identical bills) [hereinafter cited collectively as H.R. 5571]. H.R. 5571 attempted to control that portion of the product liability problem caused by "subjective" rate making practices of product liability insurers.17 H.R. 5571 authorized the creation of interstate, industry-wide insurance groups insuring their members against product liability and completed operations claims. The approval and regulation of these groups, named "risk-retention groups," was delegated to the Secretary of the Department of Commerce. The Secretary was given broad powers to audit and regulate the sale of insurance by these groups. Annual reports were required to be filed with the Secretary in such form as the Secretary, by regulation, might prescribe. One provision of the bill provided that "[t]he annual report shall include a statement of financial condition and operations substantially similar to the annual statement approved by the National Association of Insurance Commissioners." H.R. 5571, 1979 Hearings at 16.

The approval of risk retention groups was governed by Title 1 of the bill. In summary, Title I authorized the Secretary of Commerce to approve the formation of the risk retention groups pursuant to regulations considering a number of general factors.18 The Act then went on to describe necessary limitations on the membership of a risk-retention group and upon the risks which that group might undertake to insure.19

H.R. 5571, a predecessor to the statute at issue in this case, thus contemplated the creation of a new insurance concept, the risk-retention group, with two lines of limitations upon its existence.20 A risk-retention group, by definition, was an organization providing insurance only to its members. That insurance was limited to a certain kind of risk, products liability and completed operations hazards. The creation and regulation of the group was regulated by the Secretary of Commerce. With these limitations in place, the Act then went on to preempt state law with respect to state regulation of this new creature, to the end of allowing an interstate insurance capability and to inject an element of competition to modify the heretofore subjective ratemaking policies of product liability insurers.21 In this way H.R. 5571 sought both to provide obtainable and affordable product liability coverage for market participants in high risk industries, and to force existing product liability insurers to tailor their premiums to the actual claims experience of their customers.

These purposes of H.R. 5571 required for their accomplishment a preemption of state law said to have previously frustrated interstate product liability insurance capabilities. But preemption was necessarily a radical remedy in a field so traditionally relegated to state law as insurance. See e.g., Product Liability Risk Retention Act of 1979: Hearings on S. 1789 and H.R. 6152 Before the Committee on Commerce, Science, and Transportation., 96th Cong., 2d Sess. 62 (1980) (colloquy between Sen. Stevenson and Victor Schwartz) [hereinafter cited as 1980 Hearings ].22 For this reason much attention in the 1979 Hearings was given to the preemption question, the manifest concern being that preemption be only so broad as necessary to accomplish the specific purposes of the Act.

Testimony was taken from the representatives of the Commerce Department the author of H.R. 5541. In an overview of the bill, Mr. C. L. Haslam, General Counsel of the Department of Commerce stated:

The Risk Retention Act will help product sellers address their product liability concerns in two ways. First, title I of the Act enables product sellers to form their own insurance cooperatives called Risk Retention Groups. Members of these groups will be able to pool all or a portion of their product liability exposure. These groups will be exempted from state insurance regulation. These regulations are directed at commercial insurers that deal with the public, but have had the practical effect of preventing produce sellers located in different states from forming self-insurance groups.

1979 Hearings at 185. Thus, from its inception the scope of preemption authorized by Congress to effect the creation of risk retention groups turned upon the limited field of customers that those groups could serve. Risk retention groups were member financed and member servicing organizations only; the state's interest in regulating insurers dealing with the public was to remain untouched by this legislation.23

An additional concern of the subcommittee considering enactment of H.R. 5571 was that in defining a risk retention group it was necessary to limit the nature of the risks those groups could insure. However, because of the interstate nature of the concept, it was necessary to include within this definition a range of risks encompassing those allowed as product liability claims by all states. By 1979 disparate concepts of product liability had evolved in the various states. For example, several states had extended product liability law to include claims against builders for economic loss sustained by purchasers of defective housing. See R. McKean, Trends and Implications at 18.24 In this way, the scope of preemption necessary to effectuate the purposes of the Act--a scope designed to be as narrow as possible--became identified with the definition of product liability put forward by the Congress. The parameters of this definition were to be considered in detail at a later time.

During the second session of the 96th Congress the Senate Committee on Commerce, Science, and Transportation took up consideration of the Product Liability Risk Retention Act of 1979 in hearings held in April and July of 1980. The concerns at issue in the House hearings were reiterated on the Senate side and testimony was largely duplicative. Concern again centered upon the necessity of creating a risk retention group to offset subjective ratemaking practices of the insurance industry. Preemption was designed to accommodate this concern, exempting risk retention groups from state regulation to the extent necessary to accomplish the interstate utility of this new concept. More particularly, in the Senate hearings it began to be clear that a two-pronged legislative effort was underway. Lobbyists, the Department of Commerce, and other interests were advocating, on the one hand, federal legislation reforming the standards of liability in American tort law. Another prong, however, represented by the Risk Retention Act, was aimed at making certain any decrease in claims experience caused by any reform in tort law was passed on to the consumers of product liability insurance. As stated in the Senate hearing, "[w]hat [was] needed [was] some assurance, a bellwether, so to speak, that savings wrought by tort law reform are passed along to individual product sellers." 1980 Hearings at 61 (testimony of Victor Schwartz).25

The 1980 Hearings demonstrated that a remaining barrier to the enactment of the Risk Retention Act was that the Act would create unnecessary and duplicative federal regulation of insurance organizations. It was argued in the 1980 Hearings that because of the limited nature of the groups to be policed by the federal regulatory body under the Department of Commerce, a very small organization26 would be required. The role of this objection in the ultimate formulation of the Product Liability Risk Retention Act of 1981 will be developed subsequently. The 1979 version of the Act was never enacted.

On February 25, 1981, Representative Florio introduced the subsequently enacted Product Liability Risk Retention Act of 1981. Given the number H.R. 2120, the Bill was referred to the House Committee on Energy and Commerce. H.R. 2120 was reported out of Committee with amendment, and committed to the Committee of the Whole on July 21, 1981. H.R. 2120 was passed by the House on July 26, 1981, and received in the Senate on July 30, 1981. On May 4, 1981, a parallel Senate bill, S. 1096, was introduced by Senator Kasten and others, and referred to the Senate Committee on Commerce, Science, and Technology. S. 1096 was reported out of committee on July 30, 1981. The Product Liability Risk Retention Act of 1981 was enacted on September 25, 1981. Product Liability Risk Retention Act, Pub.L. No. 97-45, 95 Stat. 949 (1981), codified at 15 U.S.C. Sec. 3901 et seq.27

H.R. 2120 differed from predecessor bills in its approach to the regulation of risk retention groups. Under prior proposals the Department of Commerce would be given the responsibility for approving and regulating risk retention groups. Objection to this regulatory approach was made by, principally, insurance organizations opposed to additional governmental regulation of their industry. Accordingly, as what was apparently a compromise measure, H.R. 2120 deleted provisions for federal regulation of risk retention groups. In their stead, H.R. 2120 substituted a scheme whereby a risk retention group approved by the insurance authority of any state could then function as a risk retention group on a national basis. The Act thus sought to incorporate the existing apparatus of state insurance regulation as a means of policing the new and interstate operation of risk retention groups. In theory, once a risk retention group was approved by any particular state, that state's minimum capitalization requirements and other requirements enacted under state law for the protection of state citizens would be binding upon the risk retention group on a national basis. Sufficient financial strength to support risks assumed in one state would suffice for underwriting risks nationwide. Functioning thus as an index of public accountability, this qualification in a single state was to be a license binding all states for the operation of risk retention groups nationwide.

On April 9, 1981, the Subcommittee on Commerce, Transportation and Tourism of the House Committee on Energy and Commerce held hearings on the Act. Product Liability Risk Retention Act: Hearings on H.R. 2120 Before the Subcommittee on Commerce, Transportation, and Tourism of the House Committee on Energy and Commerce, 97 Cong., 1st Sess. (1981) [hereinafter cited as the 1981 Hearings ]. This brief hearing largely incorporated the substance of testimony contained in the prior 1977, 1979, and 1980 hearings in both the House and the Senate. The purposes of the Act remained the same: to facilitate product liability group insurance and thereby to provide a competitive incentive for reducing subjective ratemaking policies of existing product liability insurers, all to the end of making product liability coverage available and affordable to manufacturers, wholesalers, and retailers of products.

By eliminating the provision for Commerce Department regulation of risk retention groups, the support of many members of the existing product liability insurance industry had been garnered. See e.g., 1981 Hearing at 17-18 (statement of John R. Cox, President, Insurance Company of North America, Inc.). Others, however, argued that state insurance authorities of non-chartering states were incapacitated by the Act from policing requirements necessary for the protection of their citizens.

As proposed, the Bill provided for an investigation of the financial status of a risk retention group if the state insurance commissioner should gain notice that the risk retention group was in a financially impaired status. It was the concern of many witnesses that such a limited power to review the financial condition of these entities was inadequate. Once it became possible to make the objective determination of financial impairment the risk retention group could be already too far gone. See e.g., 1981 Hearings at 47 (testimony of Lawrence Herman, Director of Congressional Relations, Independent Insurance Agents of America, Inc.). The countervailing concern was that, if the individual states were given the power to regulate the capitalization requirements of individual risk retention groups in more than a de minimis capacity, the essential interstate character of the Act would be frustrated. See 1981 Hearings at 47-48 (Comments of Representative Florio). Again the Committee returned to the same concept that guided it previously. Risk retention groups were appropriately exempted from state regulation of insurance companies because risk retention groups were not in the business of selling insurance to the public. Rather, they were group insurance organizations selling insurance to group members and as such the policies behind state regulation of the insurance business did not attach to these entities. Id.

Apart from testimony on the removal of regulatory authority from the federal government and its replacement in the insurance authority of any particular state, the discussions of H.R. 2120 replicated in large part the earlier discussions of predecessor bills. The "problem" defined and addressed by H.R. 2120 was the obtainability and affordability of product liability insurance by manufacturers, wholesalers, and retailers of products causing damage to persons or property. The Act was designed to preempt state regulation which would interfere with the establishment of the new concept of a risk retention group, which was, in turn, aimed at ridding the marketplace of subjective insurance ratemaking policies.

After passage by the House and Senate the Product Liability Risk Retention Act of 1981 was signed into law by the President on September 25, 1981. As enacted, the Act provided for preemption of state regulation of products liability insurers who qualified as risk retention groups under the terms of the Act.

Legislative history on the enactment of the Risk Retention Act generally summarized information developed at the hearings discussed previously in the Senate and the House. The Act addressed one of the principal causes of the product liability problem, the questionable insurance ratemaking practices. As stated in H.R. 97-190,

The Act will reduce the problem of the rising cost of product liability insurance by permitting product manufacturers to purchase insurance on a group basis at more favorable rates or to self-insure through insurance cooperatives called "risk retention groups."

The bill should reduce insurance costs for some businesses, particularly small firms, which have had good claims experience, but have not benefitted from that experience under the current insurance ratemaking system. The legislation should also ensure the prompt payment of legally valid claims made by persons injured by products. Further, the Act will promote greater competition among product liability insurers which may encourage private insurers to set rates to reflect experience as accurately as possible.

....

Federal action is necessary because experience has shown that individual state legislation cannot facilitate the formation of self-insurance groups. Individual states can only enact legislation affecting their respective insurance law requirements. The practical effect of these laws is to prevent product sellers located in several states from forming such groups.

....

Though the product liability insurance problem is a matter for federal attention, no continuing federal presence is created by the Risk Retention Act. Moreover, state law is preempted only to the extent necessary to carry out the purposes of the Act. Only state laws which prohibit or state laws of a non-chartering state which attempt to regulate, directly or indirectly, the formation and operation of approved risk retention groups or the group purchase of product liability insurance are preempted.

....

The bill also enables persons engaged in businesses to purchase product liability insurance, completed operations liability insurance, and comprehensive general liability insurance containing either or both of these coverages on a group basis. This portion of the Act includes comprehensive general liability insurance because product liability insurance is usually sold as an endorsement to such coverage.

....

... Since ownership interests in risk retention groups will not be sold to the public, making them comply with Federal and State securities laws is unnecessary.

H.R.Rep. No. 97-190, 97th Cong., 1st Sess. 4-7, reprinted in 1981 U.S.Code Cong. & Ad.News 1432, 1432-1436.

Most significantly, the Committee report also contained this elaboration upon the effect of the Risk Retention Act's definition of "product liability."

Paragraph (3) defines the term "product liability" as liability for damages because of personal injury, death, emotional harm, consequential economic damage or property damage (including damages resulting from the loss of use of the property). Liability for these damages arises out of the manufacture, design, importation, distribution, packaging, labeling, lease, or sale of a product.

....

The committee recognizes that the definition of "product liability" may be broader than the existing product liability law of many jurisdictions in several respects. In that regard, it should be noted that the definition of "product liability" as well as that of "completed operations liability" serves to define the coverage that may be provided by a risk retention group or made available to a purchasing group. In essence, the definition delineates the boundaries of the preemption of state law for these coverages under Section 3 of the Act and exemption from state law under Section 4 of the Act. Subsection (b) explicitly provides that neither state tort law nor the law governing the interpretation of insurance contracts is effected [sic] by the Act.

An example of this broader scope is the "Restatement (2d) of Torts" which provides that strict liability shall include damage to property as well as to persons. This rule has been adopted in a majority of jurisdictions and is included in the Act's definition of "product liability." In contrast, most courts do not allow recovery for consequential economic losses in tort cases, thus distinguishing such cases from warranty cases under the Uniform Commercial Code. As noted above, the Act includes damages for loss of use of property--an example of consequential economic losses--in the term "product liability."

....

The rationale for the inclusion of damages for loss of use of property ... is to allow product sellers to protect themselves through risk retention groups against these types of damages in jurisdictions where recoveries for such damages are permitted or where the applicable law may change in the future. The definition is permissive and concerns permissible coverage.

The same rationale motivated the inclusion of damages for ... loss of use of property in the definition of the term "product liability" which appears in the recently enacted extension of the loss carryback provisions of Section 172 of the Internal Revenue Code.

H.R. 97-190 at 9-11, 1981 U.S.Code Cong. & Ad.News at 1437-38.

Finally, legislative history also analyzes the concept put forward in the Risk Retention Act of State power and State regulatory authorities as the approving agency necessary for the creation of a risk retention group. The concept is contained in the definition of "risk retention group" under the Act.

Paragraph (4) defines "risk retention group" to mean a corporation or limited liability association taxed as a corporation or insurance company, formed under the laws of any State, Burmuda, or the Cayman Islands, which meets the five requirements set forth in subparagraphs (A) through (E) [the factors limiting the constituency of the risk retention group and the risks against which it can provide coverage].

Subparagraph (A) requires the group's primary activity to consist of assuming and spreading product liability and completed operations liability coverage among its members and not the business of investing, reinvesting, or trading in securities for purposes other than providing reasonable reserves to meet liability claims.

In this connection, it is the committee's intent that "members" include the equity owners of, or contributors to, the risk retention group, as well as any entities affiliated with or related to such owners or contributors. Membership in a risk retention group should be limited to active participants in a risk retention program. Active participants include persons whose own product liability or completed operations liability is currently assumed, in whole or in part, by the risk retention group. Thus, the groups will be similar to cooperatives.

Subparagraph (B) requires that the group be organized for the purposes recited in subparagraph (A), which will presumably be reflected in its certificate of incorporation (or other organizational documents in the case of limited liability associations).

Subparagraph (C) requires the group to be chartered under the laws of any state or, before the date of January 1, 1985, under the laws of Bermuda or the Cayman Islands. It is necessary for those groups chartering in the two offshore jurisdictions to meet the minimum capitalization requirements of at least one state. The place of incorporation or formation of the risk retention group does not have to be the place of chartering (e.g.), a Delaware corporation can be chartered as an insurance company in New York if the latter's insurance laws permit). However, for the purposes of this Act, the jurisdiction in which the risk retention group is chartered as an insurer is intended to govern its status in other jurisdictions (including the place of incorporation or formation, if that state is not the jurisdiction from which it received its charter as an insurer). Thus, it is intended that the chartering jurisdiction can apply the full panoply of its insurance laws regarding the organization and operation of the company (e.g., minimum capitalization, reporting, audits, etc.) to the group.

H.R. 97-190 at 10-11, 1981 U.S.Code Cong. & Ad.News at 1438-39.

In 1983 the following amendment was proposed "to clarify the applicability of" the Risk Retention Act. The text of the proposed amendment was as follows:

(2)(b) Nothing in this Act shall be construed to affect either the tort law or the law governing the interpretation of insurance contracts of any State, and the definitions of product liability and product liability insurance under any State law shall not be applied for the purposes of this Act, including recognition or qualification of risk retention groups or purchasing groups.

S.Rep. No. 172, 98 Cong., 1st Sess. 5 (1983). The proposed amendment was reported favorably out of committee on June 10, 1983. It was passed by the Senate on November 17, 1983, and by the House on November 18, 1983. See Risk Retention Act (1983 Amendment), Pub.L. 98-193, 97 Stat. 1344 (1983). No House Report was submitted with the proposed amendment and the only legislative history available on its passage is that contained in the Senate Report No. 172 and in the floor debates.

According to Senate Report No. 172, the Congress was concerned with efforts, by the National Association of Insurance Commissioners in particular, to gain passage of a model act which, in the Congressional view, conflicted with the provisions of the Risk Retention Act of 1981. The amendment was passed to clarify the definition of "product liability" contained within the Act, and in this way to specify the scope of preemption required by the federal law.

Under the original Act, the availability of risk retention group status depended upon whether an organization could fit within the two primary limitations of the Act. The limitation at issue in the 1983 amendment to the Act was the limited nature of the risk which a risk retention group could insure: product liability. Thus the definition of product liability defined the scope of preemption.28 The 1983 Amendment elaborated upon the congressional definition of this term, and was said to result in including as "product liability" the liability of a homebuilder for economic loss to the purchaser caused by a defect in the home.29

Senate Bill No. 1046 was passed by the Senate on September 27, 1983. Senator Robert Kasten made the following remarks at its offering for consideration:

S. 1046 makes clear, as was originally intended, that risk retention and purchasing groups may insure any coverage which constitutes a "product liability" loss, as that term is defined in the Risk Retention Act. This includes insurance by a risk retention group for damage to a product itself (such as builder warranties), as well as injury to persons and other property.

The bill will insure that the definitions in that act are controlling with respect to the scope of the insurance coverage that risk retention groups may provide. It does not expand the Risk Retention Act; it merely clarifies the existing scope of the law. The bill does not, in any way, affect the liability of manufacturers for loss occurring as a result of a product defect.

....

... [T]he Risk Retention Act permits self insurance cooperatives called "risk retention groups" to insure the product liability risks of its member. The Senate report stated unambiguously that the definition of "product liability" in the act served to define the coverage that may be provided by a risk retention group and to "delineate[s] [sic] the boundaries" of the exemption from State law.

Notwithstanding this clear statement of congressional intent, some disputes have arisen over whether Federal or State law definitions of "product liability" control the scope of the coverage that may be provided by a risk retention group. This has resulted in litigation in two Federal courts and a number of inquiries by State regulators.

S. 1046 is intended to remove any legal doubt that the product liability risks specified in the Risk Retention Act define the scope of coverage that risk retention groups may insure. I believe that this clarification will help avoid needless litigation regarding the scope of authorized coverage and will eliminate any existing uncertainty that could discourage groups from utilizing the authority granted under the act. This amendment also clarifies that nothing in the act or its structure authorizes any State, including a licensing State, to restrict or expand the coverages a group may provide. It is intended that a group will qualify, regardless of the features of its program, if its primary activity consists of assuming and spreading any portion of the legal liability of its members for their product liability risk exposure.

Mr. President, I appreciate the concern of some State insurance regulators that the Risk Retention Act not be used as a vehicle to avoid legitimate State regulation of automobile, health, life, and other lines of insurance. At the same time, however, insurance regulators should not construe this law--intended to simplify the regulation of the self-insurance of product liability risks--so narrowly that its purposes and objectives are defeated.

The amendment contained in S. 1046 will end debate over the scope of insurance coverage that risk retention groups and purchasing groups are authorized to offer to their members. It will provide guidance to those forming such groups, as well as to insurance regulators.

129 Cong.Rec.S. 12972 (daily ed. Sept. 27, 1983) (statement of Senator Kasten).

Following Senate passage of S. 1046, a similar colloquy was had in the House of Representatives on November 18, 1983. House debate verified that the amendment to the Risk Retention Act was designed solely to make clear that the scope of preemption authorized under the Act was limited to that necessary to effectuate the interstate formation of risk retention groups which concept was, in turn, defined as a group whose primary activity consisted of the assumption and spreading of the product liability and completed operations risk exposure of its group members.

Mr. LENT. Mr. Speaker, I have just one more question.

The gentleman mentioned that the "primary activity" of the risk retention group must be the assuming and spreading of the product liability or completed operations risk exposure of its group members. Do I understand the gentleman correctly to mean that if a risk retention group engages in a subsidiary activity such as the sale of other lines of insurance, such as life, health, automobile liability, or workers' compensation insurance, or any other line of insurance not related to product liability or completed operations insurance, such activity would be subject to State regulation?

Mr. FLORIO: The gentleman's interpretation is exactly correct.30

129 Cong.Rec.H. 10431 (daily ed. Nov. 18, 1983) (statement of Rep. Lent).F. Summary of the Act and its Legislative History

We have reviewed these materials to determine the nature of the congressional mandate contained in the Risk Retention Act of 1981. We find it manifestly clear that Congress created an interstate capacity for groups to self-insure against risks. Qualification of a particular entity for such interstate operation depended both upon the nature of the risks insured and the market for that insurance. The entire question before us, therefore turns upon whether the "primary activity" of the entity in question consists of the assuming and spreading of the product liability risk of its group members. That risk may include liability for economic harm; it may include damages for the loss of use of property; it may extend to builders' liability for defects in the manufacture of homes. We turn to the facts of the case before us.

IV. HOME WARRANTY CORPORATION AND THE HOW WARRANTY

In 1973, the National Association of Home Builders ("NAHB") formed the Homeowners Warranty Corporation ("HOW") under the laws of the District of Columbia to administer the Homeowner's Warranty Program (hereinafter the "HOW program" or the "Program"). The HOW program involves a 10-year new home protection plan consisting of (1) a one-year builder warranty against defects in material and workmanship; (2) a two-year builder warranty assuring that the plumbing, electrical, heating and cooling systems will perform according to approved standards and insuring against "major structural defects" in the new home as that term is defined in the insurance document; (3) an insurance policy covering the builder's performance under his warranty; and (4) insurance against "major structural defects" not covered by the express builder warranty, arising in the third through tenth years after purchase of the new home.

In marketing this program, a participating homebuilder makes the described warranties to the consumer/purchaser. HOW issues to the home purchaser a "Certificate of Insurance" purporting to underwrite the builder's express warranties during the first two years after sale, and insuring the home against these described structural defects. Under the "Certificate of Insurance," all claims are to be made by the consumer/purchaser directly against HOW; all benefits, if any, are to be paid by HOW directly to the home purchaser. All deductibles are to be paid by the consumer/purchaser. The home purchaser must also assign to HOW his rights under other insurance policies held by the purchaser. The HOW program is promoted and marketed to home purchasers as an inducement to purchase a home built by a participating builder, and the purchaser must sign the application in order for the coverage to become effective. The cost of the program, initially borne by the participating builder, is passed on to the purchaser through inclusion in the purchase price of the home.

Since its inception, the HOW program has grown rapidly. At the present time policies cover more than one million homes with exposure in excess of seventy billion dollars.

HOW originally administered the program by, among other things, screening builders, registering them in the program, developing standards for home construction, establishing a dispute settlement process for disputes arising under the two-year builder warranty, inspecting the homes built by registered builders to monitor compliance with HOW's approved standards, arranging for insurance to underwrite the builder's warranty and the major structural defect coverage in the third through tenth years of the program, and collecting insurance premiums which, in turn were forwarded to the insurer.

HOW remained a subsidiary of NAHB until May, 1981. At that time HOW incorporated a non-stock membership corporation in Delaware known as the Home Warranty Corporation ("HWC"). NAHB transferred its stock in HOW to HWC, thus making HOW a wholly-owned subsidiary of HWC. The membership of HWC consists entirely of registered participants in the HOW program, with each member being required to make a capital contribution to HWC for each home enrolled by the builder in the HOW program. From 1978 until March 1, 1982, INA Underwriters Insurance Co. ("INAU"), a subsidiary of the Insurance Company of North America ("INA"), served as HOW's underwriter for the builder's insured warranty in the program's major structural defect coverage. As a result of increasing claims losses and other costs, however, INAU informed HOW in 1980 that its loss experience on the HOW program was simply too great for it to continue underwriting the program without HOW's participation in the risk. Thus, in February, 1981, HOW restructured the program to substitute itself in INAU's place for the purpose of insuring HOW builder's liability under their two-year builder warranties.

In late 1981, INAU informed HOW that it would no longer continue as the HOW program underwriter. HOW's efforts to find another insurer capable of assuming the insurance obligations of the HOW program were unsuccessful.

After passage of the Risk Retention Act of 1981, HWC formed the HOW Insurance Company ("Company"). The Company thereafter qualified to operate as an insurance company in the State of Delaware, meeting all of the statutory and regulatory requirements for approval by one state as set out in the Risk Retention Act. The Company was incorporated and chartered in Delaware as an insurance company authorized to transact the business of insurance on July 27, 1981.31

At the time the Delaware insurance commissioner approved the company's proposed insurance coverages and rates, he announced reservations on the issue of whether the company was a risk retention group. The Delaware commissioner indicated that he would at that time seek advice from the National Association of Insurance Commissioners on the propriety of accepting the company's filing under the Product Liability Risk Retention Act. Thereafter, a public hearing was scheduled to discuss the issues of qualification under the Act. The hearing was postponed and has never been rescheduled. On March 4, 1982, however, the Delaware commissioner did expressly conclude that it was inappropriate to deal with the company's filing under the Risk Retention Act. He therefore forbade use of that filing by HOW as a basis for nationwide qualification as a risk retention group and for offering the program in other states under provisions of the Act. On August 4, 1982, the Delaware commissioner affirmed his earlier determination concerning HOW's insurance program. On March 29, 1983, the Delaware commissioner issued an Order to Show Cause which alleged that the company violated the Delaware law by misrepresenting the company's insurance through the use of Delaware's approval of its filing with that state as a basis for interstate operation as a risk retention group under the Risk Retention Act.32

In response, HOW filed a complaint in the United States District Court for the District of Delaware on April 22, 1983, seeking preliminary and permanent injunctive relief as well as an order declaring the HOW Insurance Company a risk retention group under the Risk Retention Act of 1981. Following a hearing on HOW's motion for a temporary restraining order, the Delaware Insurance Commissioner agreed to cancel the show cause hearings and to submit the question of the company's status as a risk retention group to the district court for resolution on cross-motions for summary judgment.

In two separate opinions, issued after motions for summary judgment were filed, the Delaware District Court held that the company was a risk retention group exempt from all state regulations except those permitted by the Risk Retention Act. Home Warranty Corporation v. Elliott, 572 F. Supp. 1059 (D.Del.1983) (Elliott I ); Home Warranty Corporation v. Elliott, 585 F. Supp. 443 (D.Del.1984) (Elliott II ).33

The district court, surveying the history of the enactment of the Risk Retention Act together with the activities undertaken by the company and its insurance program, concluded that the coverage provided by the first two years of the HOW program did not constitute product liability insurance under the Act because HOW reserved the right to seek reimbursement from the builder for amounts paid for the benefit of the homeowner. Elliott II, 572 F. Supp. at 1067-68. An issue therefore remained as to whether the major structural defect coverage offered during the third through tenth years of the HOW program was product liability insurance under the Act in such a way as to qualify HOW as a risk retention group. In Elliott II, the district court concluded that the HOW Insurance Company's primary activity was issuing product liability insurance and that the organization was therefore qualified to conduct business as a risk retention group under the Act. Elliott II, 585 F. Supp. 446-47.

The Delaware commissioner did not appeal these rulings. Home Warranty thereafter again began representing itself as a risk retention group and selling its policies on a nationwide basis. The Georgia Insurance Commissioner issued a cease and desist order against HOW's activities in Georgia on April 19, 1983. Thereafter HOW Insurance Company and two of its corporate parents sought a determination in federal district court that their activities within the state of Georgia were exempted by the Risk Retention Act from the Commissioner's normal regulatory power over insurance companies operating in the state. They also sought an injunction preventing enforcement of the cease and desist order.

On May 13, 1983, the Georgia district court entered a preliminary injunction enjoining the enforcement of the Commissioner's order and enjoining the Commissioner from taking any regulatory action against the appellees except that provided for in 15 U.S.C. Sec. 3902(a)(1)(A)-(G). The Commissioner appealed the entry of the preliminary injunction to this court, but that appeal was dismissed as moot, after briefing and oral argument, when the district court entered a permanent injunction. Home Warranty Corporation, et al. v. Caldwell, No. 83-8438, dismissed August 27, 1984.

On July 23, 1984, the district court entered its permanent injunction enjoining the Commissioner from enforcing the Cease and Desist Order and from "otherwise interfering or attempting to interfere or restrict [HOW] ... from underwriting the HOW program and issuing HOW product liability insurance in the State of Georgia." In issuing this injunction the court concluded that HOW's insurance program fell within the confines of the Risk Retention Act. The basis for this conclusion was the opinion of the United States District Court for the District of Delaware in its decision in the two cases previously mentioned, Elliott I and Elliott II.

B. The Delaware Litigation-Elliott I and Elliott II

In Elliott I, the Delaware district court analyzed the Risk Retention Act noting, particularly, that preemption of state law was dependent upon qualification under the terms of the Act and that this qualification, in turn, required any putative risk retention group both to limit the risk it insured and to limit the market to which that insurance might be sold. Elliott I, 572 F. Supp. at 1061-62. Applying these limitations to the HOW program, the district court noted two components of the insurance in question. The program contained both a home warranty and a risk retention policy. As the district court stated:

A home builder participating in the program issues to the initial purchaser of a home a two-year 'Home Warranty.' For the first year following the sale, the builder warrants to the homeowner that the house will be free of defects due to non-compliance with the 'Approved Standards' attached to the warranty. During the second year of occupancy, the warranty continues to cover defects in the plumbing, electrical, heating, and cooling systems due to non-compliance with the 'Approved Standards.' During this initial two-year period, the builder also warrants that the house will be free from 'Major Structural Defects'....

....

Under the HOW program, the builder purchases a 'Risk Retention Insurance Policy' from HOW Insurance Company. In return for premium payments which vary depending on the length of time the builder has been in the Program and its claims record, the insurer agrees, in the language of the policy, to insure against loss resulting from: (1) Builder Default under the Home Warranty, and (2) 'Major Structural Defects' of the home which first occur after expiration of the Home Warranty and before the termination of this Policy. First, if a 'Builder Default' occurs under the two-year Home Warranty, the insurer will either 'repair, replace, or pay to the Purchaser on behalf of the Builder the reasonable cost of such repair or replacement.' In addition, the policy protects the builder against liability for Major Structural Defects in the home that occur after two years but within ten years of the initial sale. If a Major Structural Defect occurs during this period, HOW Insurance will repair or replace the defect or will pay the homeowner the reasonable costs of the repair, whichever it chooses. The policy contains deductible and excess insurance clauses, and runs for ten years, regardless of any transfer of the ownership of the home. Claims under the policy are to be pursued by the homeowner with HOW Insurance, not the builder. Finally, if HOW Insurance incurs expense when a builder defaults on the Home Warranty, the insurer retains a right to be reimbursed by the builder.

Elliott I, 572 F. Supp. at 1063 (footnotes and citations omitted).

Applying its analysis of the Act to these provisions, the district court held: (a) that a builder of new homes is a "manufacturer" and a new home a "product" within the meaning of the Act; (b) that the major structural defect coverage provided during the third through tenth years of the HOW program is coverage which assumes or spreads the product liability risk exposure of HOW's builder members and thus that HOW's policy covers an appropriately limited risk under the terms of the Act; and (c) that the coverage provided during the first two years of the HOW program does not constitute an assumption and spreading of the product liability risk exposure of the builder members because HOW reserves the right to seek reimbursement from the builder for the amounts paid for the benefit of the homeowner. Elliott I, 572 F. Supp. at 1065-67.

These conclusions drawn, the remaining issue was whether HOW's "primary activity" was the product liability insuring activity of providing major structural defect coverage during the third through tenth years of the program. Since the parties did not address that issue in the pleadings, the court denied both motions for summary judgment, giving the parties an opportunity to brief that issue in a subsequent motion.

In Elliott II, the Delaware district court concluded that the HOW Insurance Company's primary activity was issuing product liability insurance. The court relied in part on evidence which showed that, although none of HOW's policies had run through a full ten-year policy period, only 27 percent of HOW's losses were for claims made during the first two years of the ten-year program, and its finding that the two-year builder warranty was a "necessary and integral part" of the ten-year HOW program "which is designed overall to assume and spread the product liability risks of HOW's members." 585 F. Supp. at 445-56.34 The district court concluded that the incidental, integral, and necessary two-year builder warranty coverage contributed to the efficient operation of the risk distribution effected by HOW's participation as a risk retention group. The district court therefore found that this incidental activity, because it contributed to the effectiveness of the overall distribution of risk, should not invalidate the organization's ability to operate on an interstate basis as a risk retention group. Elliott II, 585 F. Supp. at 447.

We turn now to the arguments of the parties and to our resolution of the issues raised.

V. THE CASE BEFORE US--DISCUSSION AND DECISION

The Georgia Commissioner claims that the district court erred in enjoining his regulation of HOW and its program. He argues HOW is not a risk retention group under the Risk Retention Act because the insurance it markets is not product liability insurance but property insurance marketed as such to home purchasers as an inducement to purchase from a participating builder. He points out that HOW's program breaks down into two component parts. The first is a series of one and two-year warranty programs which force builder compliance with standards promulgated by HOW. Because HOW may obtain reimbursement from the builder for any claims paid to a home purchaser as a result of a builder default under these warranties, there is no distribution of risks under this aspect of the program. Consequently, the Georgia Commissioner argues, HOW's interpretation of its activities with respect to these warranties as "insurance" is incorrect. Moreover, in Georgia, these builder default claims for which HOW may obtain reimbursement from the builder constituted a substantial 34 percent of the claims made numerically and 44 percent of the dollar amount of claims paid homeowners by HOW.

HOW responds that these warranties running during the first two years of home purchase are an integral and nonseverable part of the product liability coverage primarily provided under the remaining warranty running from the third through tenth years of purchase. It is HOW's position that these initial warranties, whether characterized as insurance, or deductibles, guarantees, risks-sharing, or claims-handling aspects of the policy, are necessary to its overall operation as a risk retention group. As such, these incidental activities should not disqualify the operation of the entire program.

The Georgia Commissioner also argues that even the remaining portion of the HOW warranty, that running from the third through tenth years of purchase, is not product liability insurance within the meaning of the federal act and, as such, the purveyor of this policy is not a risk retention group entitled to the exemption from state regulation which the preemption provisions of the Act provide. Under this portion of the HOW program, the Georgia Commission urges, the insured entity is the homeowner, not the builder. The interest insured is the homeowner's economic interest in the home. The risk insured against is damage to the home, not the legal liability of the builder for product failure. Moreover, the manner in which the HOW program is marketed, the Commissioner argues, also illustrates its character as a homeowner's policy, not insurance against product liability. The home purchaser signs an application for the insurance program. He or she ultimately pays for it through the purchase price of the house. The program is promoted and marketed to consumer home purchasers, who are given the Certificate of Insurance which reads like a homeowner's insurance policy. The homeowner makes all claims directly against HOW. The homeowner pays the deductible. It is therefore the Commissioner's position that the substance of the insurance contract at issue, in contrast to its form, reveals that HOW is merely engaged in the interstate marketing of home insurance under the guise of product liability, and that the sale of this insurance in Georgia should be subject to regulation for the benefit of the insured public in that state.

To these arguments HOW responds that the substance of the insurance contract at issue is risk distribution of the product liability risk of member builders. The HOW policy defines "product liability" as:

(1) That liability of the Builder which arises solely from non-performance by the Builder under the Home Warranty and which liability extends only to the repair or replacement of defects in material or workmanship used or employed by the Builder in the construction of the Home; and

(2) that liability of the Builder for property damage to the Home itself which damage constitutes a Major Structural Defect and which defect first occurs after the expiration of the Home Warranty and before the termination of this Policy.

The policy goes on to define "major structural defect" as:Actual physical damage to the following designated load-bearing portions of the Home caused by failure of such load-bearing portions which affects their load-bearing functions to the extent that the Home becomes unsafe, unsanitary or otherwise unlivable:

1. Foundation system and footings;

2. Beams;

3. Girders;

4. Lintels;

5. Columns;

6. Walls and partitions;

8. Roof framing systems.

Record on Appeal at 515. Under this definition, HOW argues, the insured risk is the liability of the manufacturer for economic injury or property damage resulting from a defect in the manufacturer's product.35

The aspects of the policy to which the Georgia Commissioner objects, Home Warranty argues, are incidental and should not be allowed to affect the conclusion that this insurance is product liability risk insurance as set out in the Act. The signature of the home buyer on the policy is obtained to confirm the accuracy of information provided by the builder. That insurance costs are passed on to the purchaser of the home is not remarkable; all costs of a manufacturer, including insurance costs, are ultimately passed on to the consumer of the product.36 That the HOW policy and the building standards set out in it are marketed to a potential consumer as an inducement for purchasing a home built by a member builder, HOW argues, does not change the character of the insurance offered under its policy.

Deductibles paid by the homeowner do not affect the character of the insurance offered because the home builder remains fully liable for the payment of those deductibles and, even though the policy requires the homeowner to pay them, the homeowner may recover this amount from the builder by suit. Finally, that the homeowner receives a certificate evidencing the fact that his home is covered by the HOW program and the fact that the claims are generally processed directly by HOW rather than through the builder does not change the character of the risk insured because, even though it may be the homeowner which ultimately benefits from the policy, the homeowner may not purchase that policy himself. Moreover, as the district court noted in Elliott I, all liability insurance benefits third parties. See Elliott I, 572 F. Supp. at 1067.

As a concluding argument, Home Warranty relies upon its assertion that its primary activity consists of assuming and spreading a portion of the legal liability of its members for their product liability risk exposure. As such, it argues, the program comports with the qualifications for risk retention status set out in the 1983 amendment to the Risk Retention Act and qualifies as a risk retention group under that Act.

In establishing the necessary foundation for our decision, we have reviewed the development of product liability itself, the economic consequences of that development, and, finally, the congressional response to those consequences, the Risk Retention Act. We may now reach the ultimately simple question of whether the insurance relationship created by the HOW warranty falls within the qualifications set out in the Act for a risk retention group. We note that it is the congressional definition, rather than the historical concept of what products liability insurance may be, that controls our determination as to whether HOW's policy is product liability insurance. It is possible that proper application of this definition may lead to a result not contemplated by Congress nor consistent with its purposes when it passed the Act. We are mindful, however, of our duty to apply the Act as it is written and thus we defer any subsequent remedial measures to the legislative process.

We observe at the outset that the HOW policy, if it is insurance at all, is certainly not product liability insurance as generally understood prior to the Risk Retention Act. Historically--and, generally today--product liability insurance seeks to distribute a product manufacturer's or seller's risk that he will be held liable for damages that his product will, through defect, cause to another. Be that harm physical, economic, or even emotional, the definitive aspect of product liability coverage of such a risk of harm is that the one making or selling the product may be held liable as a matter of tort law, not contract.

In contrast, the HOW policy creates contractually the liability the company will insure, a liability to which under tort law the builder might well not be subject. Yet HOW is correct in asserting that the Risk Retention Act sanctions conduct very similar to that in which it presently engages. We note particularly the provisions of the Act which serve as the instruments for this result; (1) the expansive congressional definition of products liability to include damage to or loss of the property itself regardless of any jurisdiction's definition of products liability; and (2) permitting a risk retention group to be created even though it underwrites only "any portion" of the congressionally defined products liability of its group members. The effect of these two provisions is that a risk retention group may issue insurance coverage against a risk not recognized as a products liability risk by the vast majority of American jurisdictions without underwriting those risks that are recognized as products liability risks by all American jurisdictions.37

Indeed, that result has materialized in the present case. It is a traditional concept of products liability law that the products liability risk does not include the loss of or damage to the product itself. In contrast, all jurisdictions hold that claims for death, injury or damage caused by the product are included in the definition of products liability. The HOW Insurance Company has discovered in the Risk Retention Act's definition of product liability that damage to or loss of the product itself, in this instance a house, is within the purview of the Act because a distinct minority of jurisdictions so define such a loss and the Congress includes it. Being authorized to underwrite only "any part" of the defined risk, HOW Insurance Company undertakes to provide insurance applicable to only this risk, damage to or loss of the product, even though this risk is not generally recognized as a products liability risk. In turn, the company proposes to underwrite no risk of the builder arising from claims of death, injury or damage caused by the property, claims universally included in the definition of product liability exposure.

As discussed in Part III, supra, the premise for the congressional action which produced the Risk Retention Act was the mounting exposure of manufacturers to expanding liabilities for death, injury, or damage caused by their product and what was perceived as unreasonably high insurance costs for protection against those risks. There is no indication in the extensive legislative history that claims asserted, contract or tort, for damage to the product itself threatened manufacturers, distributors or sellers. Yet HOW, whose insurance program addresses only a carefully limited part of damage to the product (the house), cloaks itself within the Risk Retention Act to evade regulation by all states save one. Our task is, however, to apply the laws written by the legislative branch. We conclude that, if properly orchestrated, issuing insurance against only risk of damage to the product itself, even in states which do not recognize such as a products liability claim, is sanctioned by the Risk Retention Act.38

Our interpretation of the Act concludes that it is Congress' decision to cloak one state with the power to license an insurer which will issue its policies nationwide. However, no federal regulatory requirements are placed upon the insurance commissioner (or the state agency).39 In oral argument it was stated that Delaware requires a capitalization of only $750,000 for a company to underwrite the risks borne by the HOW Insurance Company. It is Congress' judgment that this requirement, adequate in this case for Delaware, is adequate for operation nationwide.40 We defer to the wisdom of the legislative process.

We thus conclude that the subject matter of this insurance does fall within the scope of the Risk Retention Act. It remains to be ascertained whether or not the nature of HOW's undertakings vis-a-vis this subject matter is that which is contemplated by the Act. Central to this determination is whether the primary activity of the risk retention group is the assumption and spreading of the products liability exposure of its members, the builders. In making this "primary activity" determination, we look initially to the one and two year builder default protection. This court agrees with the decision of the district court in HOW v. Elliott, 572 F. Supp. 1059, 1065-67 (D.Del.1983) that the coverage provided during the first two years of the HOW program does not constitute assumption and spreading of the product liability risk exposure of its member builders as HOW reserves the right during the term to seek reimbursement for amounts paid for the benefit of the homeowner. Such coverage more closely resembles surety insurance than liability insurance. The trial court must exclude this segment of HOW's activity in calculating the portion of the program which is the insuring of product liability risks of its members.

The remainder of the HOW coverage, years three through ten of the program, must therefore sufficiently constitute the insuring of products liability risks such that the entire program satisfies the primary activity requirement. Among the factors the trial court should consider relevant in this calculation are; (1) what portion of the items listed as reimbursable in the policy are product liability risks of the builder as defined by Congress in the Risk Retention Act and what portion are risks created by and in the policy; and (2) what portion of the protection afforded by the program is merely an alternative source from which the homeowner can satisfy certain claims respecting damage to his home. The proper standard for this analysis is a national one.

This "primary activity" analysis, however, should not mask an even more fundamental question. Does HOW assume any of the liability of its members? The record is far from clear. Classically, liability insurance insures the insured against loss, damage and expense resulting to the insured on account of claims against it and imposed upon it by law. The policy in this record includes the builder among a group collectively defined as the policyholder. However, we can find no undertaking of HOW in the policy running to the builder. The contract appears to give the builder no right against HOW, the undertakings of the latter running directly to the homeowner, presumably the adverse party to the builder in the event of a claim.

We also find no undertaking in the policy to protect the member builder. The contract does prudently require that, if a homeowner's claim be resolved, the homeowner give a complete release as consideration for payment or replacement. However, the release is required to be in favor of HOW,41 not the builder, who insofar as the policy is concerned, remains exposed to the claim. The owner is nowhere required to give credit to the builder for any amount (perhaps less than the entire claim) paid by HOW. On the other hand, HOW Insurance Company's president presents an affidavit that payments are made "on behalf of" the builder member and indicates that, under some circumstances, it will defend the builder against claims by the homeowner, but it appears that the express obligation to defend applies only to claims under the first and second year builder warranties, not to the last eight years of the HOW program.

Any right to recovery in the homeowner must imply a liability on the part of another. The source of the right must be explored. If that right arises from products liability imposed upon a member builder such would be significant. If it is merely a right created by the contract, not otherwise existing as a product liability of the builder, the insurance entity could not be found to be assuming or spreading the products liability of its members. The Risk Retention Act does not undertake to expand products liability; it seeks to contain the mounting expense of such liability. The policy undertakes to pay for certain losses suffered by the homeowner without regard to the builder's legal liability therefor. To determine the homeowner's right of recovery, one would look to the policy, not to any jurisdiction's products liability law. To determine its obligation, HOW would look to the policy, not to any liability imposed on the builder by law. There may be an overlapping of the rights of a homeowner provided by law and rights provided by the policy. If they overlap, incidentally, to what extent does HOW assume and spread products liability exposure (imposed by law) of its members? To qualify as a risk retention group, this must be its primary purpose.

There are tensions between the affidavit of HOW Insurance Company's president and the policy attached to the motion for summary judgment. The motion does not provide undisputed facts sufficient to show: (1) whether products liability risks of the member builders are retained and spread; (2) whether retention and spread of products liability risks of the members is the primary purpose of the group.

Upon remand, the district court should determine these issues, mindful that loss of or destruction of the product is included in the congressional definition of products liability. The district court may be able to make the required determination on undisputed facts should they be developed in further proceedings or by trial.

We vacate the summary judgment and remand to the district court. The district court shall determine whether or not the preliminary injunction should continue during the proceedings contemplated.

VACATED and REMANDED.