In the video below, Liability Insurance project Reporter Kyle Logue discusses Long Tail Claims, and the issues surrounding them. Included below the video is the Black Letter and Comment from the 2018 Annual Meeting draft.
§ 41. Allocation in Long-Tail Harm Claims Covered by Occurrence-Based Policies
(1) Except as stated in subsection (2), when indivisible harm occurs over multiple policy periods, the amount of any judgment entered in or settlement of any liability action arising out of that harm is subject to pro rata allocation under occurrence-based liability insurance policies as follows:
(a) For purposes of determining the share allocated to an occurrence-based liability insurance policy that is triggered by harm during the policy period, the amount of the judgment or settlement is allocated equally across years, beginning with the first year in which the harm occurred and ending with the last year in which the harm would trigger an occurrence-based liability insurance policy; and
(b) An insurer’s obligation to pay for that pro rata share is subject to the ordinary rules governing any deductible, self-insured retention, policy limit, or exhaustion terms in the policy.
(2) When an insurance policy contains a term that alters the default rule stated in subsection (1), that term will be given effect, except to the extent that the term cannot be harmonized with an allocation term in another policy that provides coverage for the claim.
(3) Defense obligations relating to multiple triggered policies are subject to the rules in § 20.
a. The special case of long-tail harm. Liability claims for long-tail harm present difficult issues of contract interpretation and application for occurrence-based commercial general-liability (“CGL”) insurance policies as well as for other similarly worded insurance policies. As discussed in Comment f of § 33, the term “long-tail harm” describes indivisible harm, whether bodily injury or property damage, that is attributable to continuous or repeated exposure over time to the same or similar substances or conditions or that has a long latency period.
b. Divisible harm. The rule in this Section addresses allocation in liability claims involving indivisible harm. For liability claims involving divisible harm, courts generally will attempt to allocate among the policy periods according to the actual injury or harm that occurred during the policy period even if the total harm occurred over a long period of time. For example, in some toxic-tort cases, courts have allocated harm among policy periods and thus among multiple triggered insurers based on the relative amount of harm that occurred, or the relative volume of the injuring substance that was released, in each period.
c.Theories of allocation for long-tail harms. Once the trigger question has been decided in a long-tail harm claim (see § 33, Comment f), the question arises how to allocate the amount of any settlements or judgments arising out of that claim among the triggered policies and, to the extent the insured does not have coverage for part of the triggered period, to the insured. Courts have developed two general approaches to this allocation question when applying occurrence-based liability insurance policies: the “all sums” approach and the “pro rata” approach. Under the all-sums approach, the insured may recover from any of the triggered policies for the full amount of that policy’s coverage limits. The insurance case law uses the term “all sums” to refer to this approach because one of the justifications commonly provided for this approach is the presence of the words “all sums” in the insuring agreement of the version of the standard commercial general-liability insurance policy at issue in the cases that have adopted this approach. The contrary, pro rata approach adopted as the default rule in this Section is sometimes referred to as the pro rata by years or “time on the risk” approach. Under this method of allocation, courts allocate the amounts paid to claimants in long-tail liability actions equally across all triggered years, beginning with the first year in which harm occurred and ending with the last year in which harm triggered an occurrence-based policy, including years for which the insured does not have liability insurance coverage.
The all-sums approach is also sometimes referred to as the joint-and-several liability approach, because it is analogous to joint and several liability in tort, with the obvious difference being that, under the all-sums approach no insurer can be held responsible for more than the applicable policy limit (unless there is a breach of the duty to make reasonable settlement decisions or a violation of the duty. of good faith and fair dealing). Under the most common all-sums approach—sometimes called the all-sums-with-stacking approach—an insured may seek recovery from one triggered insurer until the limits of that policy are exhausted, then seek recovery from another triggered insurer until the limits of that policy are exhausted, and so on, until either the claim is fully paid or the limits of all the triggered policies are exhausted. Under this approach, the insured becomes responsible for the costs of covered claims only after all triggered policies have exhausted their policy limits (ignoring deductibles and self-insured retentions). Under this approach, the risk of uninsured years is borne by the triggered insurers, subject to those insurers’ policy limits.
This Restatement follows the pro rata by years default rule for allocation in the case of long-tail harms, because that approach is the most consistent, simplest, and fairest solution to this problem. It is consistent because it provides the same result for every triggered year. It is simple because it requires very little information to determine the pro rata percentage to be applied, and it presents the fewest complications regarding exhaustion, deductibles, and settlement. It is fair because all triggered years, including the years in which the insured did not purchase insurance, share equally in the indivisible losses. In addition, this approach reflects the best effort to accommodate the language in insurance policies that links the coverage to harm that occurs during the policy period. Of all the alternative theories, this approach comes closest to allocating to each policy only those bodily injuries or property damage that occurred during each policy period, given the indivisibility of the harms at issue.
d. Pro rata versus all sums as a matter of interpretation. The split of authority regarding the allocation rule reflects the fact that the liability risks presented by the rise of mass toxic-tort suits and environmental-cleanup and property-damage causes of action were not adequately anticipated and addressed in the standard commercial general-liability (CGL) insurance policies sold before the nature and extent of those risks became apparent in the 1980s. A careful assessment of the relevant policy language in those earlier policies must acknowledge that the language is susceptible to both pro rata and all-sums interpretations. The earliest edition of the occurrence form of the CGL provides that the insurer will pay “all sums that the policyholder shall become legally obligated to pay as damages because of bodily injury or property damage to which the insurance applies caused by an occurrence.” The term “occurrence” is then defined to mean “an accident, including injurious exposure to conditions, which results, during the policy period, in bodily injury or property damage.” In later versions of the CGL, the language regarding bodily injury or property damage during the policy period was moved out of the definition of occurrence, first into the bodily-injury and property-damage definitions and then into the insuring agreement.
In all three of these versions of the CGL, it is possible to read the during-the-policy-period requirement as applying only to the trigger of coverage and not to allocation, thus leaving open the possibility that the policy covers all the damages awarded in a legal action as long as any part of the harm upon which the damages are based occurred during the policy period. According to this interpretation, for a standard CGL policy to be potentially available to cover a given legal action, there must at least be some bodily injury or property damage that occurs during the policy period. However, once it is determined that there is some bodily injury or property damage during the policy period, and thus that the CGL policy issued to cover that policy period has been triggered, the during-the-policy-period language has no other effect. On this view, the during-the-policy-period language does not provide any justification for limiting an insurer’s responsibility to harm that occurs during the policy period. At most, the during-the-policy-period language creates an ambiguity regarding the question of allocation, and such ambiguity should be construed against the drafter.
As the latter concession indicates, however, it is also possible to read the timing requirement as applying to both trigger of coverage and allocation, meaning that the policy covers only the damages that are attributable to the harm that occurred during the policy period. Indeed, many courts have found this to be the more reasonable interpretation of the relevant policy language and some courts have determined that it is the plain meaning of that language. Given the indivisible nature of the harm in such cases, the pro rata by years approach provides the best method under the circumstances for achieving the goal of limiting the insurer’s liability to the harms that occur during the policy period. Because it is impossible to determine how much harm occurred in each year, the best that can be done is to spread those costs evenly across all triggered years. (If it were possible to determine how much harm occurred during the policy period, the harm would be divisible and this Section would not apply.) By contrast, the all-sums approach allows an insurer to be held responsible for a large amount of losses that did not occur during the policy period that the insurer agreed to cover.
Most of the courts that have considered this issue have rejected the all-sums approach in favor of the pro rata approach. Not all these courts provide the same reasons for their choice, but their results are all consistent with the following reasoning: (a) pro rata by years is the default allocation rule for long-tail claims under occurrence policies with harm-based triggers,
(b) ambiguous or uncertain terms that can be read in two ways—as consistent with the default rule or to the contrary—are insufficient to alter that default rule, and (c) the “all-sums” wording in the pre-1986 policies is, at best for policyholders, ambiguous or uncertain in that regard and, thus, insufficient to displace the default rule. It is important to note that most of the cases reaching the contrary, all-sums result are also consistent with a pro rata default rule, with the crucial difference that the courts in these cases differ from the majority in treating the “all-sums” language in the earlier CGL policies as sufficient to alter that default rule.
e. The expectations argument in favor of pro rata. Proponents of the pro rata rule contend that to hold an insurer that issued a policy to cover one year responsible for harms that are statistically certain to have occurred in other years not only runs counter to the language of the policy but also conflicts with commonsense expectations regarding the difference between buying and not buying insurance. A policyholder who does not buy insurance for liability attributable to harm that occurs during a given period should bear greater financial responsibility for harm that in fact occurs during that period than a policyholder who does buy insurance for that liability. This argument can be seen in a simple hypothetical example. Insured A purchases a CGL policy with $1 million coverage limits in each of years one through five and does not purchase a CGL policy in years six through 10. Insured B purchases a CGL policy with $1 million coverage limits in each of years one through 10. Both Insured A and Insured B experience a liability claim totaling $5 million that results from continuous exposure to a long-tail harm over years one through 10. Under the most common all-sums approach, which includes stacking, both Insured A and Insured B would have, in effect, the same amount of coverage for the $5 million claim. Under the pro rata by years approach, however, the amount of coverage would be different: Insured A would have a total of $2.5 million of coverage, which results from $5 million of damages allocated over 10 years of exposure ($500,000 per year) times five years of coverage. Insured B would have a total of $5 million of coverage for the $5 million claim. The pro rata by years result makes the amount of total insurance coverage provided to the insureds over a given period of time a function of the number of years in which coverage was purchased.
f. The extrinsic evidence in favor of the all-sums approach. All-sums proponents contend that the available extrinsic evidence supports their approach. Specifically, based on records from the drafting history of the earlier standard CGL forms, as well as statements made by industry representatives who were involved in the drafting process, they contend that (a) the insurance industry itself interpreted the language in the pre-1986 CGL forms consistently with the all-sums-with-stacking approach and (b) the industry considered several explicit allocation terms that were consistent with the pro rata approach and ultimately rejected them. On that basis, they contend that it is reasonable to interpret the drafting history as supporting the conclusion that the insurance industry acknowledged and accepted, or at least acquiesced in, the all-sums interpretation of the pre-1986 standard-form CGL insuring agreement.
Careful analysis of these materials, however, reveals that many of the records and statements referenced by the all-sums proponents primarily if not exclusively support the concept of stacking (i.e., the proposition that multiple per-occurrence policy limits are available in the event of harm that takes place over multiple years), which is consistent with a pro rata approach as well as an all-sums approach. Although some of the records and statements are inconsistent with the pro rata approach, these records and statements primarily serve to demonstrate that it is possible to interpret the policy language in favor of the all-sums approach. They cannot change the fact that the policy language is also susceptible to the pro rata interpretation. All things considered, the records and statements by the drafters simply corroborate the point made in Comment d: the rules for allocating liability risks presented by the rise of mass toxic-tort suits and environmental-cleanup and property-damage causes of action were not adequately anticipated and addressed in the standard general-liability insurance policies before the nature and extent of those risks became apparent in the 1980s. This Restatement follows the pro rata rule for the earlier CGL policies for the same reasons of consistency, simplicity, and fairness that lead to the conclusion that the default rule is pro rata, and because a majority of the courts that have considered the question have concluded that the pro rata rule represents the more reasonable interpretation of the language in those policies.
1. A series of asbestos-related lawsuits is brought against the insured involving $40 million of total liability costs. The bodily injuries that give rise to the suits occurred continuously over a period of 10 years. Although it can be determined that some bodily injury occurred in each of the 10 years, it cannot be determined precisely how much of the $40 million of harm occurred in each of the 10 years. During this 10-year period, the insured was covered under an array of CGL policies issued by three different insurers, as follows: Insurer A issued policies covering years 1-4 with annual policy limits of $500,000; Insurer B issued policies covering years 5-8 with annual policy limits of $5 million; and Insurer C issued policies covering years 9 and 10 with annual policy limits of $20 million.
Under the pro rata by years allocation method, 10 percent of the total $40 million liability cost ($4 million) is allocated to each of the 10 years, as if that portion of the harm occurred in that year. Thus, in the absence of any deductibles or self-insured retentions in any of the policies, the insured is entitled to a total of $26 million of liability coverage allocated as follows: $2 million from Insurer A (4 x $500,000 annual policy limit), $16 million from Insurer B (4 x $4 million annual allocation), and $8 million from Insurer C (2 x $4 million annual allocation). The insured is responsible for the remaining $14 million in liability costs.
2. Same facts as Illustration 1, except that in years 5-8 Insurer B issued policies with annual limits of $10 million rather than $5 million, and the insured chose not to purchase liability insurance policies for years 9 and 10. Under the pro rata by years approach, again $4 million of the $40 million total costs is allocated to each of the 10 years. As a result, Insurers A and B are responsible for $2 million and $16 million, respectively. The insured is responsible for the remaining $22 million of losses.
g. Pro rata by limits. A few courts have adopted a pro rata by limits rule, which is a common formula used for contribution among insurers in the context of concurrently overlapping policies with no or conflicting other-insurance clauses. See § 42, Comment b. The pro rata by limits rule differs from the ordinary pro rata by years rule in two respects. First, the pro rata by limits approach uses policy limits in the calculation of the amount allocable to each of the relevant years, so that more of the indemnity obligation is allocated to policies with higher limits. Second, the pro rata by limits approach allocates long-tail losses to uninsured years only to the extent that the policyholder intentionally opted not to purchase coverage that was available—and then only to the extent of that available coverage. This second aspect of the pro rata by limits approach is sometimes called the “unavailability rule,” which is addressed separately in Comment h.
Proponents of the pro rata by limits approach contend that it is more consistent with the pricing of those policies and thus with the parties’ expectations. On this view, the higher the limit of a policy, the larger was the premium paid for the coverage in that year (all other things being equal), and hence the greater the amount of the long-tail harm that should be allocated to that year. Further, it is argued that the pro rata by limits formula has the beneficial effect of encouraging the purchase of relatively high policy limits, because the higher the limits of coverage purchased for any year, the larger the fraction of total losses will be allocated to that year. In addition, this approach provides some relief to an insured—and to the claimants of the insured—who was unable to purchase coverage for reasons beyond its control. Finally, the courts adopting this approach have concluded that strong public-policy considerations justify rewriting the insurance policy in the special circumstances of unanticipated long-tail harm.
The pro rata by limits approach is subject to several critiques. First, the pro rata by limits rule puts a disproportionate burden on the higher-limit insurer by increasing the share of the loss that insurer must pay if the policyholder purchased less insurance in another year. There are fairness concerns about making insurers that issued policies in one year more (or less) responsible because the policyholder purchased lower (or higher) policy limits in another year; and there are fairness concerns about making insurers that, by assumption, expressly chose not to provide coverage in the years in which liability insurance was unavailable, responsible for those harms under other policies that they sold that covered different policy periods. Second, there is no textual basis for either aspect of the pro rata by limits rule in the standard-form commercial general-liability insurance policies to which it is applied. Third, the treatment of uninsured years in the pro rata by limits rule can be complicated and information intensive. Finally, it is by no means certain that the pro rata by limits approach in fact does more to encourage the purchase of insurance than any other allocation method. Any time an insured purchases greater policy limits, whatever the allocation method, there is more coverage potentially available to pay claims.
h. The unavailability rule. Some courts that have applied the pro rata by years approach to long-tail indivisible claims have adopted the unavailability rule aspect of the pro rata by limits approach. These courts allocate to the policyholder a share for an uninsured year only if liability insurance coverage was available in that year for the risks in question, based on public-policy considerations similar to those relied upon by the courts adopting the pro rata by limits approach to allocation. In these jurisdictions, the burden of proof is generally on the policyholder to show that insurance was unavailable during the relevant policy period. Among the jurisdictions that follow some version of the pro rata approach, the courts that have considered the issue are evenly split on the application of the unavailability rule. When the number of jurisdictions adopting the all-sums approach to allocation is combined with the number of jurisdictions adopting the unavailability rule, however, it can be said that most courts that have considered the issue do not allocate to insureds the costs of damage that occurs during years in which insurance was unavailable. Courts that reject the unavailability rule do so on the grounds that it is inconsistent with the language of the insurance policy. Other objections include: the treatment of uninsured years can be complicated and information intensive and the rule imposes responsibility on insurers for harm that occurred during periods for which the insurers expressly chose not to provide coverage.
i. Exhaustion, deductibles, SIRs, and settlement. In addition to the issues of trigger and allocation, long-tail-harm claims raise related issues such as the application of deductibles, self-insured retentions (SIRs), exhaustion, and the effect of settlements. The pro rata by years rule addresses each of these issues in a more straightforward and easier to administer manner than the all-sums rule. Long-tail harms are allocated to each policy period as if the pro rata portion of the loss occurred in each triggered year, with the application of deductible, SIR, policy-limit, and exhaustion provisions following the usual course without the need for any special rules for the long-tail-harm situation. As soon as one policy in a given year is exhausted, the next-level policy takes over for the remaining portion of the liability allocated to that year, and so on, until that tower of insurance is exhausted, at which point the insured is financially responsible for losses allocated to that policy period. This is a form of “horizontal exhaustion,” which is the majority approach to exhaustion in pro rata by years jurisdictions. Under the all-sums approach, by contrast, insureds exhaust the coverage available in one year using a “vertical-exhaustion” approach before accessing the coverage available in another year, once again requiring all of the insurers that have not yet exhausted to track the payments. Moreover, vertical exhaustion under the all-sums approach puts some excess insurers in the position of paying long before primary insurers, which is inconsistent with the pricing of excess and primary coverage. One of the benefits of the pro rata allocation approach is that it avoids these problems as well.
Under the pro rata rule, deductibles and SIRs are also easy to administer. Because a pro rata share of the liability is allocated to each policy year, the deductibles and SIRs for each policy period are applied to that pro rata amount, just as they would apply to any other liability that occurred during the period. Moreover, because the amount allocated to each policy period is identical, no insurer needs to keep track of how much coverage remains under any other insurance policy. Further, under the pro rata rule, there need be no subsequent contribution action, as the allocation of responsibility among multiple triggered insurers is determined by the pro rata by years rule. (If an insurer pays more than its share under the pro rata by years rule, contribution is nevertheless ordinarily available under § 42.)
The treatment of settlements under the pro rata approach is also simple. Because each year is allocated a pro rata portion of the overall liability, settlements by insurers during one policy period have no effect on the liability of insurers in other policy periods. By contrast, under the all-sums approach, a complex and difficult decision must be made regarding how much to credit one insurer’s settlement payment against the other insurers’ overlapping liability taking into account exhaustion. See § 43 (regarding the effect of partial settlements). Moreover, because the all-sums approach requires contribution actions to allocate among insurers, it is susceptible to agreements between the insured and one or more insurers that work to the disadvantage of other insurers. See § 42 (regarding contribution).
3. A series of asbestos-related lawsuits is brought against Insured involving $100 million of total liability costs. The bodily injuries that give rise to the suits occurred continuously over a period of 10 years. Although it can be determined that some bodily injury occurred in each of the 10 years, it cannot be determined precisely how much of the harm occurred in each of the 10 years. During this 10-year period, the insured was covered under several different towers of CGL policies, as follows: For years 1-4, Insurer A issued primary policies with annual policy limits of $500,000, and Insurer B issued excess policies that attached at $500,000 with annual policy limits of $5 million; for years 5-6, Insured did not purchase liability insurance; for years 7-10, Insurer C issued primary policies with annual policy limits of $5 million, and Insurer D issued excess policies attaching at $5 million with annual limits of $30 million. There were no deductibles or retentions.
Under the pro rata by years allocation method, 10 percent of the total $100 million liability cost ($10 million) is allocated to each of the 10 years, as if that portion of the harm occurred in that year. For each of the policy years 1 through 4, all $500,000 of primary coverage is exhausted, as well as all $5 million of excess coverage, with the remaining $4.5 million of annual asbestos liability being borne by Insured. Insured bears all of the asbestos losses for years 5 and 6. For each of the policy years 7 through 10, all $5 million of the primary coverage is exhausted, as well as $5 million of the $30 million excess coverage. Insured bears no losses for those years.
4. A series of asbestos-related lawsuits is brought against the insured involving $40 million of total liability costs. The bodily injuries that give rise to the suits occurred continuously over a period of 10 years. Although it can be determined that some bodily injury occurred in each of the 10 years, it cannot be determined precisely how much of the harm occurred in each of the 10 years. During this 10-year period, the insured was covered under an array of CGL policies issued by three different insurers, as follows: Insurer A issued policies covering years 1-4 with annual policy limits of $500,000 and annual deductibles of $25,000, written as a standard deductible; Insurer B issued policies covering years 5-8 with annual policy limits of $5 million and annual deductibles of $100,000, written as a standard deductible; and Insurer C issued policies covering years 9 and 10 with annual policy limits of $20 million and annual self-insured retentions of $1 million.
Under the pro rata by years allocation method, 10 percent of the total $40 million liability cost ($4 million) is allocated to each of the 10 years, as if that portion of the harm occurred in that year. Thus, Insured is entitled to a total of $23.5 million, allocated as follows: $1.9 million from Insurer A for years 1-4 [(4 x $500,000 annual policy limit) – (4 x $25,000 annual deductible]); $15.6 million from Insurer B [4 x $4 million annual allocation) – (4 x $100,000 annual deductible)]; and $6 million from Insurer C [(2 x $4 million annual allocation) – (2 x $1 million retention)]. The insured is responsible for the remaining $16.5 million ($40 million – $23.5 million) in liability costs.
j. Other-insurance clauses. While some pro rata proponents have suggested, and a few courts have agreed, that the other-insurance clauses found in most CGL policies should be understood as a sort of allocation provision for the long-tail-harm situation, the majority of courts that have addressed the question conclude that such other-insurance clauses address a different situation: namely, the situation in which multiple insurance policies issued during the same policy period cover the same insured concurrently for a given loss. See § 40, Comment c.
k. Opting out of the default rule. The default rule of pro rata allocation can be altered by contractual terms that provide an alternative method of allocation or priority. For example, if an insurance policy contains a term that clearly specifies the all-sums approach to allocation (perhaps by eliminating the “during the policy period” language), such a term will be enforced. However, if such allocation terms conflict with each other, courts should apply the pro rata method of allocation as a matter of public policy. For example, if multiple policies contain allocation terms that purport to apply to the long-tail-claim situation and that amount to escape clauses, such terms should not be enforced. Rather, the pro rata default rule should apply in such situations. This result is analogous to how courts have interpreted other-insurance clauses. See § 40, Comments d and e.
l. Independent and concurrent allocation of defense costs. Defense costs for long-tail claims are subject to the ordinary default rule for multiple triggered policies, pursuant to which there is an independent and concurrent obligation to pay the defense costs under each triggered policy. See Comment e to § 20. This rule is a corollary to the rule that the duty to defend includes the obligation to defend the insured from all of the causes of action and remedies sought in a potentially covered legal action, including those not covered by the liability insurance policy. See § 15. Notably, the textual argument in favor of the pro rata allocation of the costs of settlement and judgments, which relies on the “during the policy period” language in standard liability insurance policies, does not apply with the same force to the costs of defense, because standard liability insurance policies typically obligate the insurer to defend the insured against a suit that seeks covered damages, without regard to whether the suit also seeks other damages. In this regard, there is nothing unusual about multiple insurers, which are on the risk for different portions of the period of damage at issue in a covered legal action, each having an independent and concurrent obligation to defend that action. An insurer that pays more than its pro rata share of the defense costs may seek contribution from other insurers. See § 20. Moreover, adopting a pro-rata approach to defense costs would mean that an insured would not be able to receive a fully funded defense unless and until it succeeded in persuading all of its insurers to agree upon an allocation formula. It should be noted, however, that some jurisdictions that have adopted pro rata allocation for the costs of judgments and settlements have also done so for defense costs.