The opinion of the Court was delivered by O'HERN, J.
This appeal involves two aspects of a dispute between a manufacturer of an asbestos product and its insurers concerning the coverage afforded to the manufacturer under its liability insurance policies. First is the "trigger of coverage" issue, a shorthand expression for identifying the events that must occur during a policy period to require coverage for losses sustained by the policyholder. Second is the "allocation issue," which involves the scope of coverage afforded under a triggered policy. One question,
The Appellate Division has ordered a hearing, 264 N.J.Super. 460, 522, 625 A.2d 1 (1993), on issues of fraud related to the issuance of the policies in that officers of the manufacturing company or the captive company, interlocked as they were, may have failed to disclose fully the underwriting risks for which the manufacturer sought coverage. The court below also believed that "further airing" was necessary to resolve whether the manufacturer expected or intended that its product would cause injury. Id. at 515, 625 A.2d 1. Those and other issues decided by the Appellate Division are not part of this appeal.
The facts of the case are set forth fully in the Appellate Division opinion. We recite only those facts necessary to our disposition. Because the issues of coverage arose on cross-motions for summary judgment, we may accept as true in this appeal of defendants all the evidence supporting plaintiff's position, as well as all legitimate inferences that may be deduced therefrom. Boyer v. Anchor Disposal, 135 N.J. 86, 88, 638 A.2d 135 (1994). For purposes of this appeal, then, we adopt generally the factual version of the case set forth in the briefs of the plaintiff-manufacturer, Owens-Illinois (O-I).
The most salient feature of this case is that it concerns a decades-old manufacturing activity. From 1948 to 1958, O-I
In the mid-1970s, defendant American Risk Management, Inc., offered to develop a new program of insurance for O-I involving a captive insurance company. The captive company would be a subsidiary of O-I. For a fee, American Risk Management would manage the subsidiary and arrange to reinsure the policies with companies in the United States and abroad. In 1975, Owens Insurance Limited (OIL) was established as the captive insurance company and began providing O-I with a fire and extended peril reinsurance program and loss prevention services. In 1976, American Risk Management proposed that OIL also provide O-I's casualty insurance, including products-liability coverage.
In June 1977, O-I invited quotes from various sources, including American Risk Management, to replace its Aetna insurance coverage, which would expire September 1, 1977. The bid package solicited a comprehensive general liability (CGL) policy covering general and products liability. After receiving various proposals,
Toward the end of 1977, O-I's in-house legal department became aware of a number of asbestos-related lawsuits involving the Kaylo product. Early in 1978, O-I gave notice of those claims to Aetna, its carrier from September 1, 1963, to September 1, 1977. Aetna took the position, with which O-I originally agreed, that the cases should be reported on a manifestation basis, that is, the policy in effect when the disease manifested itself should respond to the claim. Because most statutes of limitations were of at least several years in duration, O-I assumed that Aetna would be the responsible carrier. As a precaution, O-I also informed the Insurance Companies of the asbestos claims.
Aetna rejected the claims submitted to it because it insisted that the $250,000 SIR was a per-claim figure. It thus estimated that none of the claims could reasonably be thought to call for coverage under its policies. As claims continued to mount, O-I recognized that the claims were no longer the exclusive responsibility of Aetna because the manifestation dates were now presumed to be after September 1, 1977. Accordingly, in 1980 O-I gave formal notice to the Insurance Companies of the pendency of those asbestos claims. The Insurance Companies adopted the same position as had Aetna with respect to the $250,000 SIR. They
By January 1980, O-I had established a $3 million reserve for general legal expenses, substantially generated by its growing number of asbestos cases.
In late November 1984, O-I sought a declaratory judgment from the Chancery Division that the two lead carriers, United and OIL, must provide coverage for O-I's asbestos-related personal-injury and property-damage claims. The other defendants joined or were joined in the proceedings.
In an early phase of the proceedings, Judge Keefe, then sitting as a Chancery Division judge, concluded that an "injury in fact" triggering coverage under the insurance policies occurs on the inhalation of asbestos fibers and continues up to and including manifestation of an asbestos-related disease. Following discovery, Judge Conley reaffirmed Judge Keefe's ruling and held all insurers whose policies were triggered to be jointly and severally liable to O-I to the extent of the policy limits. In addition, the court held that the continuous trigger should apply to claims for property damage.
The Chancery Division also resolved other issues that we do not review. Among those issues was whether the $250,000 SIR was on a per-occurrence or a per-loss basis. Because the vast majority of individual claims do not exceed $250,000, the Insurance Companies' position would have effectively eliminated coverage for most of the claims made against O-I. However, the courts below, consistent with other jurisdictions, see Owens-Illinois v. Aetna, supra, 597 F. Supp. at 1525, held that the $250,000 SIR applied not to individual claims but to the aggregate of exposures due to a single condition during the policy period. The Appellate Division held that "the manufacture and sale of Kaylo must be regarded as the single occurrence triggering liability for asbestos-related injury
In addition, the Chancery Division held, and the Appellate Division agreed, that certain exclusions in the policies were either internally contradictory or ineffective. Those questions of policy construction pertaining specifically to the policies in this case were not of such public importance as to warrant review under Rule 2:12-4. However, the Appellate Division did agree with the Insurance Companies that disputed issues of fact existed that required further hearings on certain other coverage issues. 264 N.J. Super. at 521-22, 625 A.2d 1. The issues of expected or intended environmental injury should be more readily resolved in light of this Court's decision in Morton International, Inc. v. General Accident Insurance Co. of America, 134 N.J. 1, 629 A.2d 831 (1993), cert. denied, ___ U.S. ___, 114 S.Ct. 2764, 129 L.Ed.2d 878 (1994).
The economic realities of this litigation are stark. By 1991, when O-I filed its Appellate Division briefs, it had settled 43,000 bodily-injury lawsuits. More than 90,000 bodily-injury and sixty-three property-damage cases were pending in all states and some territories. With bodily-injury lawsuits accumulating at the rate of 1700 per month, O-I's unreimbursed costs of defending and settling those cases had by then exceeded $95 million. O-I had already spent close to $10 million in defense and settlement costs associated with the property-damage cases. The questions raised here concern which of the insurance policies issued from 1977 to 1985 provide indemnity to O-I and to what extent.
Our grant of certification, 135 N.J. 301, 639 A.2d 301 (1994), was limited to two issues: (1) the application of the "continuous trigger" theory, and (2) any consequent apportionment of liability.
The Policy Language
The source of any duty on the part of the Insurance Companies to defend actions or to pay any judgments is obviously in the
The OIL policy provides:
Both the United and OIL policies are subject to policy limits and certain exclusions. Although they are not boilerplate policies, they closely resemble a standard CGL policy.
The United policy contains the following definitions: An "occurrence" means "an accident, including continuous or repeated exposure to conditions, which results in Bodily Injury or Property Damage neither expected nor intended from the standpoin[t] of the Insured." "Bodily injury" means "bodily injury, sickness or disease sustained by any person which occurs during the policy period * * *." "Property damage" means "physical injury to or destruction of tangible property which occurs during the policy period, including the loss of use thereof at any time resulting therefrom * * *." The corresponding definitions in the OIL policy are similar.
The policies do not refer to a "trigger"; "the term `trigger' is merely a label for the event or events that under the terms of the insurance policy determines whether a policy must respond to a claim in a given set of circumstances." Robert D. Fram, End Game: Trigger of Coverage in the Third Decade of CGL Latent Injury Litigation, in 10th Annual Insurance, Excess, and Reinsurance Coverage Disputes 9 (PLI Litig. & Admin. Practice
Despite the relative familiarity of these concepts, the one hundred or so pertinent words in the coverage clause have spawned "a bewildering plethora of authority" interpreting their meaning. Gottlieb v. Newark Ins. Co., 238 N.J.Super. 531, 534 (App.Div. 1990).
To place the issues in context, we shall use an example to which most of us can relate. (The example is intended not to suggest that workers might in fact contract disease in the manner as stated but only to simplify our analysis.) Assume that a group of workers occupied an office building for nine years under the following circumstances. For the first three years, the building owners had no liability insurance, assuming any risk of loss. During each of the middle three years, the owners were insured under a CGL policy with the Trustworthy Insurance Company for $5,000,000 per occurrence. For the remaining three years of the period, the owners were again uninsured. Assume, too, that during the first three years the building occupants were exposed to asbestos fibers in the ceilings and insulation but that all asbestos products were removed at the end of the third year. During the first three years, no occupants of the building manifested any symptoms of disease. During the fourth, fifth, and sixth years, there was "exposure in residence," that is, some building occupants began to develop breathing problems, but no disease was diagnosable. In the final three years, some of the building's occupants were diagnosed with asbestos-related diseases.
Trustworthy's CGL policy is as broad as those in this case. The policy promised that the company would pay all sums that the insured should become legally obligated to pay as damages for bodily injury caused by an occurrence during the policy period. Each of the thirty claimants in our hypothetical might recover from the owners a large sum, let us say $500,000, for a total of $15,000,000 in damages. Was there an occurrence during the fourth, fifth, and sixth years that calls for full indemnity? The owners contend that the fibers once inhaled by the building occupants continued to cause injury to the occupants during those middle years, and that they paid premiums to cover their risk of liability for injury during those three years even if caused by earlier exposure. The owners seek indemnity to the extent of the full $5,000,000 for each year of coverage, for a total of $15,000,000. The first question is: Were the policies issued in the middle years "triggered"?
Trigger of Coverage
Not surprisingly, the answer to the trigger-of-coverage question varies by jurisdiction. The most frequently offered theories for the trigger of coverage are (1) the exposure theory, (2) the manifestation theory, and (3) the continuous-trigger theory. A concise summary of these approaches is contained in a law
The conceptual underpinning of the continuous-trigger theory, then, is that injury occurs during each phase of environmental contamination — exposure, exposure in residence (defined as further progression of injury even after exposure has ceased), and manifestation of disease.
At least two other less-frequently followed theories exist. One is the "injury-in-fact" (or "damages-in-fact") approach, which holds that coverage is triggered by a showing of actual injury or damage-producing event. See, e.g., American Home Prods. Corp. v. Liberty Mut. Ins. Co., 565 F.Supp. 1485 (S.D.N.Y. 1983), aff'd as modified, 748 F.2d 760 (2d Cir.1984). Under that theory, coverage is triggered by "a real but undiscovered injury, proved in retrospect to have existed at the relevant time * * * irrespective of the time the injury became manifest." Id. at 1497. "That is, after an injury has been diagnosed, it may be inferred, from the nature of the gestation period and from the stage of the illness, that the harm actually began sometime earlier." Armstrong World Indus., Inc. v. Aetna Casualty & Sur. Co., 25 Cal.App.4th 1316, 26 Cal.Rptr.2d 35, 52 (1993), review granted sub nom. In re Asbestos Ins. Coverage Cases, 27 Cal.Rptr.2d 488, 866 P.2d 1311 (1994).
Finally, the "double-trigger" theory holds that injury occurs at the time of exposure and the time of manifestation, but not necessarily during the intervening period. E.g., Zurich Ins. Co. v. Raymark Indus., Inc., 118 Ill.2d 23, 112 Ill.Dec. 684, 695, 514 N.E.2d 150, 161 (1987).
The court in Keene Corp. v. Insurance Co. of North America, 667 F.2d 1034, 1041, 1048-49 (D.C. Cir.1981), cert. denied, 455 U.S. 1007, 102 S.Ct. 1644, 71 L.Ed.2d 875 (1982), relied on the presumption
Certain things are well settled: As a general rule, the time of the occurrence of an accident within the meaning of an indemnity policy is not the time the wrongful act is committed but the time when the complaining party is actually damaged. Hartford Accident & Indem. Co. v. Aetna Life & Casualty Ins. Co., 98 N.J. 18, 483 A.2d 402 (1984). Hartford concerned the exposure of a child, Ann Marie Sherman, to the medication Atropisol, which resulted in the child's illness. Two companies had insured the drug's manufacturer at the different times of exposure to the drug and manifestation of illness. Each contended that the other was liable to indemnify the manufacturer. In adopting the opinion of Judge Skillman, then sitting in the trial court, we agreed that the occurrence triggering the indemnity policy was not the administration of the drug to Ann Marie but the time when her injuries manifested themselves.
Later in the opinion the court said:
The Insurance Companies maintain that Hartford is controlling and that "the time when the damage has been suffered" must be proven on a case-by-case basis in the tens of thousands of pending lawsuits to establish a duty to indemnify under the policies. (They do concede, however, that some generalized formula for progression might suffice.) Hartford, however, expressly declined to resolve the time of an occurrence in the case of progressive bodily disease. Judge Skillman explained:
Hartford, the company on the risk when Ann Marie's illness became manifest, had simply failed to offer any evidence that the medication administered to the child had caused her any damage before the Hartford coverage took effect.
The Insurance Companies insist that the record in this case is incomplete because it does not contain "medical testimony that the inhalation of asbestos causes immediate tissue damage." Ibid. However, Judge Keefe, the judge presiding over the early phase of this matter in the Chancery Division, had an extraordinary understanding of the nature of asbestos-induced disease. He was at that time presiding over the unified statewide administration of all asbestos trials, and had conducted many asbestos trials. He found
The "overwhelming weight of authority" elsewhere acknowledges the progressive nature of asbestos-induced disease, and affirms that "`bodily injury' occurs when asbestos is inhaled and retained in the lungs." Lloyd E. Mitchell, Inc. v. Maryland Casualty Co., 324 Md. 44, 595 A.2d 469, 478 (1991). Generalities about asbestos may be overblown. See Becker v. Baron Bros., 138 N.J. 145, 649 A.2d 613 (1994) (finding erroneous charge to jury that as matter of law any product containing asbestos, without regard to type of asbestos, was defective if it did not contain warning). Still, we are satisfied, like most American jurisdictions, that medical science confirms that some injury to body tissue occurs on the inhalation of asbestos fibers, and that once lodged, the fibers pose an increased likelihood of causing or contributing to disease. Lloyd E. Mitchell, Inc., supra, 595 A.2d at 478-81. See also Fischer v. Johns-Manville Corp., 103 N.J. 643, 660-61 n. 2, 512 A.2d 466 (1986) in which we recognized the progressive nature of asbestos-related disease.
The record in this case is less persuasive on the issue of asbestos-related property damage. The underlying complaints against O-I allege that Kaylo damages buildings from the moment it is installed and that the damage continues throughout the period that Kaylo remains in a property. The Appellate Division held that "an ongoing process of property damage triggers every policy during the destructive process." 264 N.J. Super. at 510, 625 A.2d 1.
Other courts have applied multi-year triggers in cases of delayed manifestation of property-damage claims. In Dayton Independent School District v. National Gypsum Co., 682 F.Supp. 1403,
Gottlieb, supra, 238 N.J. Super. at 537-38, 570 A.2d 443, held that, pending a hearing to determine various factual matters, homeowners could, under a continuous-trigger theory, recover from a pest control company's earlier insurer for chemical poisoning. Lac d'Amiante du Quebec, Ltee. v. American Home Assurance Co., 613 F.Supp. 1549 (D.N.J. 1985), held that under New Jersey law the progressive nature of asbestos-related property damage triggers every policy on the risk from installation to removal. The court took judicial notice of the fact that asbestos products cause continuous property damage from installation until removal:
Property-damage cases are analogous to the contraction of disease from exposure to toxic substances like asbestos. Like a person exposed to toxic elements, the environment does not necessarily display the harmful effects until long after the initial exposure. "Thus, while property damage is not, of course, an insidious disease, many of the same considerations apply." Ibid. The court in Armstrong World Industries, supra, also recognized that property damage caused by asbestos might not be complete at installation. The California Court of Appeals in that case
O-I informs us that a small percentage of its asbestos-related expenditures has gone to satisfy property-damage cases. At least in the context of this case (in that no one suggests that the process was anything but continuous), we hold that claims of asbestos-related property damage from installation through discovery or remediation (the injurious process) trigger the policies on the risk throughout that period. There is some question about when the injurious process ends. We do not reach that issue in this opinion. We shall concentrate in this opinion primarily on the issues of asbestos-related personal injury.
Accepting that inhalation of asbestos fibers causes some injury to tissue, does that injury trigger coverage under a CGL policy? Many courts have answered that question by finding ambiguity in the language of the policy and construing the policy in favor of the policyholders. For example, the District of Columbia Circuit, which applied the continuous-trigger theory in Keene, supra, noted the difficulty of interpreting the typical language of a CGL policy:
However, just as we did not find ambiguity in the language of the pollution-exclusion clause in Morton International, supra, 134 N.J. at 28-29, 629 A.2d 831, we do not find ambiguity in the language of the "occurrence" clause. The words are all familiar and easily understandable. "The plain meaning of the `occurrence' clause is no secret to the parties." American Home Prods., supra, 565 F. Supp. at 1497.
What is not so easily understandable is the point at which the law will say that injury requires indemnity. In that sense, the concept of injury, like the related concepts of duty and causation, is an instrument of policy. After all, the air we breathe and the water we drink contain trace elements of toxic substances. The law decides when an invasion of the body constitutes an injury entitling one to damages.
In addition, injury may mean different things in different contexts. For example, in Coughlin v. Owens-Illinois, Inc., 26 Cal.App.4th 1511, 27 Cal.Rptr.2d 214, 228 (Ct.App. 1993), review granted, 29 Cal.Rptr.2d 538, 871 P.2d 1134 (Cal. 1994), O-I contended that for purposes of determining the effective date of California's Proposition 51, modifying rules of common-law liability, an action for asbestos-related injury accrues when disease is diagnosed or discovered, a position different from that which it holds today. See also Owens-Illinois, Inc. v. Armstrong, 326 Md. 107, 604 A.2d 47, 53, cert. denied, ___ U.S. ___, 113 S.Ct. 204, 121 L.Ed.2d 145 (1992) (rejecting O-I's argument regarding applicability of newly-established cap on non-economic damages that action
Our own law has drawn the line on liability to indemnify for injury in different ways in different contexts. In Ayers v. Township of Jackson, 106 N.J. 557, 598-99, 525 A.2d 287 (1987), we concluded that absent diagnosed disease, the ingestion of toxic substances that enhanced the risk of progressive growth of cancer did not constitute a legal injury warranting compensation. In Mauro v. Raymark Industries, Inc., 116 N.J. 126, 561 A.2d 257 (1989), we held that even when the body suffered diagnosable injury due to asbestos exposure (thickening of chest walls and calcification of diaphragm), the injury was insufficient to allow recovery for the enhanced risk of contracting cancer.
What we did, however, in the face of doubtful scientific premises was to adapt our law to the uncertainties of medical causation. In Ayers, supra, we held that even when there is no bodily injury (or none that can be found), "the public health interest may justify judicial intervention even when the risk of disease is problematic." 106 N.J. at 605, 525 A.2d 287. We said that "mass-exposure toxic-tort cases involve public interests not present in conventional tort litigation." Id. at 609, 525 A.2d 287. We thus allowed tort recovery for medical-surveillance damages even without evidence of physical injury. Id. at 610-11, 525 A.2d 287. In Mauro we relaxed our statute of limitations and single-controversy doctrines to allow victims to assert later claims when the foreshadowed disease eventually occurs. In candor, we "acknowledge[d] that our resolution of [that] issue [was] imperfect." 116 N.J. at 143, 561 A.2d 257.
So too here, our resolution of the issues is necessarily imperfect. Our concepts of legal causation were developed in an age of Newtonian physics, not of molecular biology. Were it possible to know when a toxic substance clicks on a switch that alters irrevocably the composition of the body and before which no change has "occurred," we might be more confident that occurrence-causing damages had taken place during a particular policy
Mass-exposure toxic-tort cases have simply exceeded the capacity of conventional models of judicial response.
No such procedure has been forthcoming. Hence, courts must adapt common-law doctrines "to the peculiar characteristics of toxic-tort litigation." Ibid. We advert to those principles because we believe that common-law resolution of the trigger-of-coverage issue requires that we consider, at the same time, the issue of scope of coverage if a policy is triggered. "[T]he choice of trigger theory is related to the method a court will choose to allocate damages between insurers." Northern States Power Co. v. Fidelity and Casualty Co. of New York, 523 N.W.2d 657 (Minn. 1994).
Scope of Coverage
Assuming that every phase from exposure to manifestation of disease is a period of continuous bodily injury, does that trigger the sum of all policies in force during the years of exposure, exposure in residence, and manifestation? Recall that in our hypothetical the building owners were insured only during the middle three years of the workers' occupancy. Keene, supra, 667 F.2d 1034, is the leading case for the proposition "that any triggered policy must respond for the entirety of a claim, subject
The conceptual model employed by the Keene court is that of a pleated accordion surrounding the entire "occurrence" and representing the time span from exposure to manifestation. Its solution to the problem of indivisible injury was to collapse the injuries in the accordion into a single year. It wrote:
That theory is sometimes referred to as one of joint-and-several allocation. The Keene court explained that it did not mean that a single insurer will be saddled with full liability for any injury.
What the Keene decision would mean in the context of our office-building example is not perfectly clear to us. By way of illustration, the Keene opinion states that "only one policy's limits can apply to each injury." Id. at 1049. But what did the Keene court mean by "injury"? Did it mean each claim of injury, or each cause of injury? Some language in the opinion has led commentators to believe it meant the latter: "The key to the Keene court's decision was its conclusion that there was only a single occurrence. The court's concern was that the insurers' liability for a long-term exposure injury be the same as their obligations for other types of losses." William R. Hickman & Mary R. DeYoung, Allocation of Environmental Cleanup Liability Between Successive Insurers, 17 N.Ky.L.Rev. 291, 301 (1990).
If the Keene court's concern was that the insurance companies' liability for long-term exposure to injury be the same as their obligations for other types of losses, only one policy's limits would seem to apply to all claims arising out of the same occurrence, as would have been the case had an asbestos-laden steam pipe exploded in the fourth year of our illustration, causing immediate injury to thirty of the building occupants. We surmise, however, that the Keene court's holding is that one policy's limits apply to each claim of injury, and that in our office-building hypothetical the building owner might have assigned ten of the claims to year four, ten to year five, and ten to year six. The cross-indemnity among the policies would neutralize the effect of contribution on policy limits. In Air Products and Chemicals, Inc. v. Hartford
One anomaly in the Keene court's analysis is that a single claim for the cost of cure of a long-term release of contaminants that polluted a city water supply would be limited to one policy's limits, whereas if 300 residential wells were affected, the limits of multiple policies would be available, though the occurrence (cause) was the same. Another anomaly is that although the opinion's premise is that all damages can be claimed in any one of the years, it nonetheless calls for contribution from other policies. By definition, if all damages occurred in one of the years (in the sense of that year's injury establishing the damages), none of the other policies would be triggered.
Several courts selecting multi-year triggers of coverage have held that "triggered policies must respond to a claim on a prorated basis and that a policyholder is responsible for a portion of
Forty-Eight Insulations concluded that a reasonable means of allocating costs among the triggered policies was available based on the number of years of exposure. 633 F.2d at 1225. (The policyholder disputed only the allocation of defense costs in the Court of Appeals in Forty-Eight Insulations.) In Uniroyal, supra, 707 F.Supp. 1368, Judge Weinstein applied a different formula because evidence was available to differentiate between the various periods of coverage. He applied a pro-rata method under which the loss would be allocated to each policy according to the portion of injuries triggering that policy. Id. at 1393. He resolved that portion by the quantity of the substance released during the policy periods. Id. at 1393-94. He appeared to reject the Keene theory of joint-and-several liability
In Diamond Shamrock Chemicals Co. v. Aetna Casualty & Surety Co., 258 N.J.Super. 167, 222-23, 609 A.2d 440 (App.Div. 1992), certif. denied, 134 N.J. 481, 634 A.2d 528 (1993), the Appellate Division, applying New York law, let stand a similar allocation among policies covering liabilities for dispersal of Agent Orange.
Does the Language of the Policies Resolve the Allocation Issue?
Each side relies on the same language in the policies. O-I emphasizes that the language is unambiguous, the meaning clear. It contends that once a policy is triggered, an insurance company is liable, under the United policy, for "all sums which the insured shall become legally obligated to pay * * * as damages," and under the OIL policy, for "the sum actually paid * * * in the settlement or satisfaction of any claim or suit." The policies do not state that the insurer shall pay only some of the damages sustained, or that the insurers must pay only a portion of the ultimate net loss sustained during the policy period. Instead, the policies expressly require that when personal injury or property damage occurs during a policy period, the insurers must pay all sums resulting from that injury or property damage.
The Insurance Companies emphasize with equal certainty that the language of the policies is clear. They insist that the Appellate Division's opinion ignores the plain language of the policies and applicable law. They contend that in its citation of the policy language, the Appellate Division omitted the language limiting the insurers' liability to injury "which occurs during the policy period."
Both arguments are flawed. As to the Insurance Companies' argument that all injury (or damages) must occur in the policy period or that indemnity is awarded for only the part of the injury that occurs during the policy period, consider the simple case of an automobile accident in 1994 with a definite prognosis that an injured occupant's spine will deteriorate in 1996 resulting eventually in paralysis. The policy in effect during 1994 must indemnify for all damages attributable to the 1994 accident even though the full extent of the damages or the injury will not take place until a future date. Conversely, to convert the "all sums" or "ultimate net loss" language into the answer to apportionment when injury
The problem is how to apply the abstract concepts of law and the related provisions of the insurance contract to the realities of environmental disease. The legal concepts of injury, defect, negligence, and damages are well-suited to the prototype accident of an exploding steam boiler. A defective weld in a boiler may have been present years before but the occurrence (an explosion) and the attendant injuries are easily identified as falling within a particular policy period. Even though "all sums" due from the accident might not be known with certainty at the time of the explosion, by the time of trial a claimant would be able to establish, within a reasonable degree of medical probability, what damages would flow from the injury.
That is not so in the case of gradual release of contaminants. Even in cases such as Ayers, supra, 106 N.J. 557, 525 A.2d 287, in which bodies were admittedly exposed to damaging contaminants, "all sums" due because of the injury simply cannot be determined in each of the years of exposure or exposure in residence, and perhaps not even when there has been a manifestation.
In a refreshing display of candor, Judge Barry acknowledged in Lac d'Amiante du Quebec, supra, 613 F. Supp. at 1551, that the differing interpretations of which occurrence constitutes the injury contemplated by the language of insurance policies is "due, in part, to the desire of some courts to maximize coverage even when to do so requires judicial sleight of hand." It was precisely that "sleight of hand" or "leap of logic" that all the damages can be thought to
The Reasons for Our Decision
No great difference in principle divides Keene and Forty-Eight Insulations. Using either method, allocation will exist among the insurance companies on the risk. And using either method, the use of a multi-year trigger will not end the litigation. The real difference between Keene and Forty-Eight Insulations is in their treatment of periods of self-insurance. Each court was equally certain that the policy language dictated the result. Forty-Eight Insulations said that "it requires only a straightforward interpretation of the policy language for us to adopt the exposure theory." 633 F.2d at 1222. Once the exposure theory was adopted, proration followed, in its view, because "[a]n insurer contracts to pay the entire cost of defending a claim which has arisen within the policy period. The insurer has not contracted to pay defense costs for occurrences which took place outside the policy period." Id. at 1224-25. Keene said: "As we interpret the policies, they cover Keene's entire liability once they are triggered. That interpretation
We are unable to find the answer to allocation in the language of the policies. The occurrence clauses undoubtedly contemplated indemnity for provable damages incurred by the policyholder because of injury that occurred during the policy period. The continuous-trigger theory coupled with joint-and-several liability is premised on a tenuous foundation: that at every point in the progression the provable damages due to injury in any one of the years from exposure to manifestation will be substantially the same (the collapsed accordion). As we have seen, our law has been developing in a different manner.
Some drafting history suggests that more than one year's policy would be applicable in the case of progressive environmental disease. The O-I briefs refer us to industry-group acknowledgments in 1977 that coverage existed for each carrier throughout the period of time an asbestosis condition developed, that is, from the first exposure through the discovery and diagnosis. A majority of that same group also contended that each carrier on risk during any part of the period could be fully responsible for the cost of defense and loss. (The Insurance Companies' response is that those views were predictions of what courts might do, rather than what courts should do.)
On other occasions insurance industry officials acknowledged that multiple policies of insurance would be triggered by a gradual release of contaminants causing progressive injury or damage. See Eugene R. Anderson, et al., Liability Insurance Coverage for Pollution Claims, 59 Miss.L.J. 699, 729-30 (1989) (quoting (1) Gilbert L. Bean, a drafter of the CGL policy: "[I]f the injury or damage from waste disposal should continue after the waste disposal ceased, as it usually does, it could produce losses on each side of a renewal date, and in fact over a period of years, with a separate policy applying each year."; (2) a company claims manual: "When the injury is gradual, resulting from continuous or repeated exposures, and occurs over a period of time, coverage
Still, to say that coverage is afforded under more than one policy seems different from saying that a policyholder may stack the limits of all the policies in effect during a period of repeated or continuous exposure. The Keene court declined to make such an interpretation, at least in the context of a single claim as for governmental cleanup costs at a landfill site.
Other drafting history reflects the considerations that went into the writing of the occurrence clause. If the drafters had chosen the manifestation-of-disease single trigger, insurance companies would have been encouraged to "dump" risks when the first few cases of occupational disease appeared. American Home Prods., supra, 565 F. Supp. at 1501. Perhaps the drafters sensed, too, a corresponding inequity in absolving from responsibility carriers on the risk during early years of exposure. If absolved during periods of early exposure, those carriers would have less incentive to monitor the risks insured. See Developments in the Law, supra, at 1577-81. Although the drafters seemed to acknowledge that some continuum existed between exposure and manifestation of disease when coverage was provided, they did not (or could not) decide how to apportion the responsibility. Almost wistfully, one of the drafters acknowledged that at one time in the process, they sought a "simple approach that would allow some latitude in each case for the courts to make an equitable decision on the facts." Memorandum from George Katz, et al. to National Bureau of Casualty Underwriters, Joint Forms Committee 2 (Apr. 17, 1961), quoted in Eugene R. Anderson, History of Disputed Provisions of the 1966 Standard Form Comprehensive General Liability Insurance Policy, Drafting History, Sales History and Historical Review of Commentators, in Insurance, Excess, and Reinsurance Coverage Disputes 1989, at 203 (PLI Litig. & Admin. Practice
Except for periods of self-insurance involved, the Keene court sought to reach an equitable decision on the facts apportioning the damages among the successive carriers. Although it collapsed the injury/occurrence into a single year, the Keene court allowed contribution from the policies in force in the other years.
However, that court's invocation of the "other insurance clauses" in the policies is strained. Historically, "other insurance" clauses were designed to prevent multiple recoveries when more than one policy provided coverage for a given loss. Kahn, supra, at 591-92. An example of a typical multiple-coverage case is the situation in which a loss is incurred by an insured driver while driving an automobile of an insured owner with the owner's permission. 3 Rowland H. Long, The Law of Liability Insurance § 22.01 (1992). In such a case both policies clearly cover the entire loss.
Generally speaking, pro-rata provisions are intended to apply only "when the coverage is concurrent." St. Paul Fire & Marine Ins. Co. v. Vigilant Ins. Co., 919 F.2d 235, 241 (4th Cir.1990). If the policies do not overlap, such clauses are not generally applicable. Ibid. But see Glacier Gen. Assurance Co. v. Continental Casualty Co., 605 F.Supp. 126, 130 n. 6 (D.D.C. 1985) (explaining Keene as having "rendered much of traditional insurance law inapplicable" and adopting its own analysis for proration of loss when two consecutive malpractice policies are triggered, one by "medical incident" and the other by "bodily injury" during the policy period). The Glacier court declined to give effect to an "other insurance" provision that would have withdrawn coverage
The other usual principles of interpretation of contracts of insurance do not provide much guidance. As the Appellate Division noted, "O-I was a sophisticated insured and cannot seek refuge in the doctrine of strict construction by pretending it is the corporate equivalent of the unschooled, average consumer." 264 N.J. Super. at 489, 625 A.2d 1. Assessing the objectively reasonable expectations of a policyholder in this context of long-tail injuries is also very difficult. Sparks v. St. Paul Ins. Co., 100 N.J. 325, 330, 495 A.2d 406 (1985). At least in the case of property damages due to environmental contamination, the retroactive imposition of absolute liability under laws like CERCLA, 42 U.S.C. §§ 9601-9675, was surely unknown, if not unknowable.
And the doctrine of contra preferentem, construing any ambiguity against the insurer as drafter, Uniroyal, supra, 707 F. Supp. at 1376, can produce uneven results. For example, in Eagle-Picher Industries v. Liberty Mutual Insurance Co., 523 F.Supp. 110 (D.Mass. 1981), aff'd as modified, 682 F.2d 12 (1st Cir.1982), cert. denied, 460 U.S. 1028, 103 S.Ct. 1280, 75 L.Ed.2d 500 (1983), the insured preferred a manifestation theory. (On appeal, the insured in Eagle-Picher advocated the continuous-trigger theory. 682 F.2d at 16.) In Forty-Eight Insulations, supra, 633 F.2d 1212, the exposure theory afforded the insured the most coverage. To have shifting rules of interpretation that depend on the configuration of insurance coverage is unacceptable to us.
The language of the policies does not itself yield either result and the usual rules of interpretation are less helpful in this context. See Note, Adjudicating Asbestos Insurance Liability: Alternatives to Contract Analysis, 97 Harv.L.Rev. 739, 743 (1984), observing that traditional techniques of contract interpretation "cannot produce a coherent result." Therefore, the public interest factors set forth in Ayers, supra, 106 N.J. at 608-10, 525 A.2d 287,
One important argument generally advanced for imposing strict liability is that the manufacturers and distributors of defective products can best allocate the cost of injuries resulting from those products. The premise is that the price of the product should reflect all its costs, including the cost of injuries caused by the product. Those manufacturers and distributors can incorporate the cost in the price of the product. The cost of the product will thus be borne by all those who profit from it, including manufacturers and distributors who profit from its sale, and buyers who profit from its use. The policy considerations underlying those principles include the relative bargaining power of the parties and the allocation of the loss to the better risk-bearer in a modern marketing system. Spring Motors Distribs., Inc. v. Ford Motor Co., 98 N.J. 555, 576, 489 A.2d 660 (1985).
The theory of insurance is that of transferring risks. Insurance companies accept risks from manufacturers and either retain the risks or spread the risks through reinsurance. John A. Appleman & Jean Appleman, 13A Insurance Law and Practice § 7681 (1976). Because insurance companies can spread costs throughout an industry and thus achieve cost efficiency, the law should, at a minimum, not provide disincentives to parties to acquire insurance
Almost all such insurance controversies are retrospective, and to reflect now on what might have been done if the parties had contemplated today's problem is almost fatuous. Our job, however, is not just to solve today's problems but to create incentives that will tend to minimize their recurrence. "[T]o send the correct signals to the economic system, a judge must appreciate the consequences of legal decisions on future behavior." Hallett & Berney, supra, at [*]15. Future actors would know that if they do not transfer to insurance companies the risk of their activities that cause continuous and progressive injury, they may bear that untransferred risk.
To return to our hypothetical of the building occupants, the Keene rule of law reduces the incentive of the property owners to insure against future risks. Recall the circumstances in the final three years: The exposure had taken place and some symptoms of disease had occurred, but grave cases of injury had not yet arisen. Assuming the availability of insurance, a principle of law that would act as a disincentive to the building owners in the hypothetical might serve in the long run to reduce the available assets to manage the risk. O-I's counsel counters that these are not correct assumptions about the way in which the "real world" responds. We cannot be sure that the policy will be effective. We believe, however, that the policy goal is sound.
Finally, principles of simple justice cannot be entirely discounted. To rebut effectively the question posed in Forty-Eight Insulations is difficult:
O-I emphasizes that its policies were in effect for eight years, but the principle that it advocates would apply if the policies had been in force for one day.
The Remedy That We Propose
The final question that we must address, however, is whether our proposed solution will be an efficient response to the problem of insurance coverage for long-term environmental damage. The court in Forty-Eight Insulations, aptly described the challenge:
One thing is certain: The present system is inefficient. "`[T]he largest transaction cost today is money being spent by insurance companies and industry making claims. [The cost is estimated] at about $500 million annually. These are the litigation costs between insured and insurers.'" Hallett & Berney, supra, at [*]21-22 (quoting testimony of Doctor Joel Hirschhorn before a Congressional Subcommittee). "The legal costs of environmental coverage litigation today may run as high as 70 percent of total cleanup costs." Eugene R. Anderson & Giovanni Rodriguez, Settling Environmental Coverage Disputes: What You Know About Your Enemy Cannot Hurt You in Environmental and Toxic Tort Claims: Insurance Coverage in 1991 and Beyond 383 (PLI Comm. Law & Practice Course Handbook Series No. 579, 1991), available in WESTLAW, PLI-COMM Database, [*]3. Concededly, legal disputes have necessarily preceded the cleanup work under new and complex laws like CERCLA, and once the rules are settled, litigation costs may decline. Hallett & Berney, supra, at [*]22. Still, the present rules do not seem to be working.
We will not attempt a universal resolution of all issues of coverage for gradual release of pollutants or toxins. At least in the context of asbestos-related personal injury and property damage, the rules that we adopt will attempt to relate the theory of a continuous trigger causing indivisible injury to the degree of risk
To explain the concept, we return to our example of the office-building workers. If we were to accept the constant levels of the policy limits as evidence of constant risks assumed over the nine-year span from exposure to manifestation (in the case of disease manifested in the ninth year), the carriers on the risk in years four, five, and six would each pay one-ninth of the loss, or collectively thirty-three percent. If the facts of coverage had been otherwise — let us say policies had been in effect for years one through three in the amount of two million per year and in years four through six at three million per year — we might assess the
We realize that many complexities encumber the solution that we suggest involving, as it does, proration by time and degree of risk assumed — for example, determining how primary and excess coverage is to be taken into account or the order in which policies are triggered. See Hickman & DeYoung, supra, at 310-11; Roger Westendorf & Ronald R. Robinson, Insurance Coverage for Environmental Claims Under the Comprehensive General Liability Policy, in Pollution Liability: Managing the Challenges of Coverage and Defense in 1991, at 57 (ALI-ABA Video Law Review Study, Q205, 1991), available in WESTLAW, ALI-ABA Database, [*]54-57 (each discussing various theories of horizontal and vertical stacking and relationship to excess coverage issues). The parties did not focus on those issues. Still, we do not believe that the issues are unmanageable. Constructing the model for analysis of the self-insurance portion of the risk assumed by O-I is difficult but not impossible. We recognize the difficulties of apportioning costs with any scientific certainty. However, the legal system "frequently resolves issues involving considerable uncertainty." SL Indus., Inc. v. American Motorists Ins. Co., 128 N.J. 188, 216, 607 A.2d 1266 (1992).
This case is undoubtedly more difficult to manage than most because of the great number of claims involved. On the other hand, the record is reasonably well-developed on the measure of risk assumed or transferred, at least since 1963. To extrapolate back from 1963 to establish a rough measure of the risk assumed
In addition, we are informed that Aetna has paid its policy limits for the years 1963 to 1977, so that Aetna's policy proceeds will not be called on for future contribution. If that be so, rather than go back to revisit Aetna's contributions, we shall start forward and treat pending matters from the perspective of one having a long view of the entire occurrence (the extended accordion). Because of the large number of claims of asbestos exposure, the net effect of proration may simply be to cover more cases with partial indemnity in each case, rather than to cover fewer cases with full indemnity.
On remand, coincident with resolving the other coverage issues, the court shall appoint a master, one skilled in the economics of insurance, to create a model for allocating the claims. Above all, the master should develop a workable system for efficient assignment and administration of the claims. Because the defendants refused to involve themselves in the defense of the claims as presented, they should be bound by the facts set forth in the plaintiff's own records with respect to the dates of exposure and with respect to the amounts of settlements and defense costs. See Kahn, supra, at 612 (describing the remand procedure in Keene). Those losses for indemnity and defense costs should be allocated promptly among the companies in accordance with the mathematical model developed, subject to policy limits and exclusions. We stress that there can be no relitigation of those settled claims. Exact dates of exposure may not now be available. Available data should enable the master to grasp the generality of the underlying claims and the exposures involved. (Even under the Keene formula exposure dates were necessary to determine contribution.)
These cases, like the underlying environmental cases to which they are related, demand special attention, such as the use of
In future cases, insurers aware of their responsibility under the continuing-trigger theory might minimize their costs by assuming responsibility for or involving themselves in the defense of the actions with the ultimate allocation of costs to be determined in accordance with the same general formulas. See R. 4:42-9(a)(6) (allowing counsel fees in actions on indemnity policies). If, after experience, we are convinced that our solution is inefficient or unrealistic, we will not hesitate to revisit the issue. "We do not expect that this case will be the `last word' in this area. Environmental liability insurance law, like any other area of law, will have to develop over time and trial courts must be flexible in responding to new fact situations." Northern States Power Co., supra, 523 N.W.2d at 665.
To recapitulate, we hold that when progressive indivisible injury or damage results from exposure to injurious conditions for which civil liability may be imposed, courts may reasonably treat the progressive injury or damage as an occurrence within each of the years of a CGL policy. That is the continuous-trigger theory
Although the use of a continuous trigger for property damage attributable to long-term embedding of contaminants is more problematic (for example, the "property damage" may be attributable only to third-party intervention, as in the form of a government order to rip out material previously thought not to be defective), the latent nature of such property damage, at least in the case of asbestos products, is sufficiently analogous to that in personal injury to warrant use of a continuous trigger under the terms we have outlined. We need not resolve in these cases exactly when the continuum ends in the contexts of bodily injury and property damage.
Because multiple policies of insurance are triggered under the continuous-trigger theory, it becomes necessary to determine the extent to which each triggered policy shall provide indemnity. "Other insurance" clauses in standard CGL policies were not intended to resolve that question. A fair method of allocation appears to be one that is related to both the time on the risk and the degree of risk assumed. When periods of no insurance reflect a decision by an actor to assume or retain a risk, as opposed to periods when coverage for a risk is not available, to expect the risk-bearer to share in the allocation is reasonable. Estimating the degree of risk assumed is difficult but not impossible. Insurers whose policies are triggered by an injury during a policy period must respond to any claims presented to them and, if they deny full coverage, must initiate proceedings to determine the portion allocable for defense and indemnity costs. For failure to provide coverage, a policyholder may recover costs incurred under the provisions of Rule 4:42-9(a)(6). Policyholders must cooperate in furnishing information concerning coverage. Courts must take an active role in the management and resolution of such coverage controversies. A trial court may repose a large measure of discretion in a special master to aid the court in developing a formula for allocation of the costs of defense and indemnity. R.
The judgment of the Appellate Division is reversed insofar as it allocated none of the costs of indemnity or defense to periods of no insurance and otherwise directed contribution under the "other insurance" clauses of the policies. The matter is remanded to the Chancery Division for further proceedings in accordance with this opinion.
For reversal and remandment — Justices CLIFFORD, HANDLER, POLLOCK, O'HERN, GARIBALDI and STEIN — 6.
[264 N.J. Super. at 505, 625 A.2d 1.]