ALL SUMS: REALLOCATION WITH NON-SETTLED INSURERS AND APPLICATION OF SETTLEMENT CREDITS. By Martin C. Pentz 1

Transcription

1 ALL SUMS: REALLOCATION WITH NON-SETTLED INSURERS AND APPLICATION OF SETTLEMENT CREDITS By Martin C. Pentz 1 We have now seen nearly three decades of high-stakes litigation over insurance coverage for liabilities arising from third-party claims for long-term, delayed manifestation bodily injury or property damage. Some of the issues that once were among the most hotly debated now appear to be the subject of something approaching a consensus resolution. One of these is the so-called trigger-of-coverage issue under occurrence basis general liability policies. Although unanimity remains elusive, it seems fair to say that the great majority of states have determined that all policies in effect during an extended period of exposure to harmful conditions will be triggered if their coverage threshold is met. See, e.g., Armstrong World Indus., Inc. v. Aetna Cas. & Sur. Co., 45 Cal. App. 4th 1, 52 Cal. Rptr. 2d 690, (1996) (asbestos bodily injury) citing Montrose Chem. Corp. v. Admiral Ins. Co., 10 Ca1. 4th 645, 42 Cal. Rptr. 2d 324, 913 P.2d 878, (1995) (environmental property damage). Whether such a resolution of the issue is labeled as a continuous trigger or an injury-in-fact trigger, and whether or not policies in effect after exposure but before manifestation also are deemed to be triggered, the cases almost uniformly contemplate that multiple successive policy periods will be called into play. And where successive policies are triggered, inevitably the extent-of-coverage question arises: how much must be paid under each triggered policy? To date, there clearly is not a judicial consensus regarding the answer to the extent-ofcoverage question. A number of state courts-of-last-resort have spoken, however, and there are 1 Martin C. Pentz is a partner in the Boston law firm Foley Hoag LLP who represents policyholders in coverage disputes with insurance companies. The author wishes to thank Jeremy A.M. Evans and Creighton Page of Foley Hoag for their assistance in the preparation of this paper.

2 now respectable groups of jurisdictions endorsing one or the other of two distinct and quite contradictory schools of thought. The first with which this paper is concerned is typically referred to as the all sums school, and posits, by reference to some of the wording of standard liability policy insuring agreements, that, once the policy is triggered, the insurer must pay all sums the insured becomes legally obligated to pay as damages, subject only to policy limits of liability, and regardless of the length of time the insurer was on the risk relative to the duration of the injurious process. See, e.g., Plastics Eng g Co. v. Liberty Mutual Ins. Co., 315 Wisc. 2d 556, 759 N.W.2d 613, (2009); Goodyear Tire & Rubber Co. v. Aetna Cas. & Sur. Co., 95 Ohio St. 3d 512, 769 N.E.2d 835, 841 (Ohio 2002). The second school calls for some form of proration of liability among coverage periods and uninsured periods on the theory that any policy applies only to so much of the claimed damages as can be traced to injury or damage that took place during the particular policy period. See, e.g., Boston Gas Co. v. Century Indem. Co., 454 Mass. 337, 910 N.E.2d 290, (2009); Towns v. Northern Sec. Ins. Co., 184 Vt. 322, 964 A.2d 1150, 1167 (2008); Consolidated Edison Co. of N.Y., Inc. v. Allstate Ins. Co., 98 N.Y.2d 208, 774 N.E.2d 687, (2002). Among the asserted virtues of proration is its tendency to result in comprehensive resolution of the coverage dispute in a single action: because each insurer is assessed only its pro-rata share of the liability, the policyholder has an incentive to sue all insurers who may have a share, and insurers have no basis to sue other insurers for contribution because, under proration, there is no common liability and no insurer will have paid more than their fair share. On the other hand, whatever its merits, the all sums approach holds the potential for follow-on litigation (or at least a second stage of litigation) in which one or more insurers who have indemnified, or have been held liable to indemnify the insured, seek to impose liability for a - 2 -

3 share, or perhaps even the entire amount, upon other insurers. See EnergyNorth Natural Gas, Inc. v. Certain Underwriters at Lloyd s, 156 N.H. 333, 934 A.2d 517, 527 (2007). This makes sense because, if each insurer whose policy is triggered is liable for all sums the insured must pay, then all insurers with triggered policies have a common liability. And, if there are insurers with triggered policies who have been assessed no share of the liability, then the insurers whose policies were selected by the insured often will have been assessed more than their fair share, and contribution rights against other insurers should lie. While jurisdictions adopting the all sums approach tend to broadly suggest the availability of a right to contribution for insurers upon whom all sums liability is imposed, see, e.g., J.H. France Refractories Co. v. Allstate Ins. Co., 534 Pa. 29, 626 A.2d 502, 509 (Pa. 1993); Stonelight Tile, Inc. v. California Ins. Guar. Assn., 150 Cal. App. 4th 19, 58 Cal. Rptr. 3d 74, 88 (Cal. App. 6th Dist. 2007), the case law defining the contours of that right in the context of successive policies remains at an early stage of development. This is not to say, however, that courts addressing this question are writing on an entirely clean slate. Indeed, there is a significant existing body of case law arising from intra-insurer disputes involving asserted common liability to a common insured the difference being that the coverage periods of the policies in issue typically have been concurrent, rather than successive. The rules adopted in those cases likely will at least contribute to the analysis of contribution cases following an all sums payment by a selected insurer. In addition to describing the rules under which contribution claims between successive insurers will be decided, the courts also are grappling with a second key issue that arises in all sums, but not pro rata, jurisdictions: the effect of a policyholder s prior settlement with other insurers whose policies also are triggered by the events or conditions which underlie the all - 3 -

4 sums liability of the selected insurer(s). Do the settled insurers remain exposed to contribution claims, despite their settlement with the policyholder? If not, is a selected insurer entitled to a credit for amounts the policyholder has received or should have received from the settled insurers? How should the amount of that credit be determined? This paper discusses the sources from which courts are deriving, or are likely to derive, the rules governing contribution claims by selected insurers. It also canvasses the approaches courts are taking to cases in which a selected insurer s potential contribution targets have already settled their potential liability to the policyholder. The paper concludes with a discussion of practical considerations from the policyholder s perspective. Throughout, the scenario examined is one in which the policyholder, having the benefit of the all sums rule, has selected one or more policy periods to which its liability has been assigned, and certain primary and, perhaps, excess insurers have indemnified, or been held liable to indemnify, the policyholder. Other insurers whose policies are triggered either were not selected, for whatever reason, or already have settled with the insured. What happens next? I. THE SECOND STAGE: REALLOCATION A. Selected Insurer Contribution Claims Against Non-Settled Insurers At least with respect to insurers who have not yet settled with the policyholder, the next step, as the J.H. France opinion indicates, involves a cross-claim, third-party complaint, or separate proceeding for secondary apportionment between insurers with triggered policies. 626 A.2d at 509. The jurisdictions that have directly addressed the issue suggest differing apportionment strategies, looking to policy other insurance provisions and applying principles of equitable contribution. See FMC Corp. v. Plaisted & Cos., 61 Cal. App. 4th 1132, 72 Cal. Rptr. 2d 467, 499 (1998) ( In the usual case [allocation among insurers] would be determined in proceedings, independent of the determination of each individual insurer's liability to its insured, - 4 -

5 to allocate a paid loss among all insurers on the risk on the basis of contractual other-insurance provisions or equitable considerations. ); Keene Corp. v. Insurance Co. of N. Am., 667 F.2d 1034, 1051 (D.C. Cir. 1981) ( When more than one policy applies to a loss, the other insurance provisions of each policy provide a scheme by which the insurers liability is to be apportioned, but also recognizing insurers rights of equitable contribution); Armstrong World Indus., 52 Cal. Rptr. 2d at (approving trial court view that the most equitable method of allocation is proration on the basis of policy limits, multiplied by years of coverage ). It appears likely that many courts will look first to the triggered policies terms specifically other insurance provisions for guidance on how to apportion liability among insurers. See FMC Corp., 72 Cal. Rptr. 2d at 499; Keene, 667 F.2d at 1050; but see Plastics Eng g Co., 759 N.W.2d at ( other insurance clauses applicable only when coverage is concurrent). The use of other insurance clauses as a basis for secondary apportionment will be discussed in further detail below. See infra, Section I.A.2. Other insurance clauses may be held inoperative, however, if they are deemed to be mutually repugnant. See Oregon Auto. Ins. Co. v. United States Fid. & Guar. Co., 195 F.2d 958, 959 (9th Cir. 1952) ( where both policies carry like other insurance provisions, we think [they] must be held mutually repugnant and hence be disregarded ); State Farm Fire & Cas. Co. v. LiMauro, 492 N.Y.S.2d 534, 482 N.E.2d 13, 17 (1985) (identical excess other insurance clauses mutually repugnant, canceling one another out). In addition, some courts reject other insurance clauses as a basis of apportionment on the ground that such clauses are contracts between insurer and insured, not among insurers. See e.g., Signal Cos., Inc. v. Harbor Ins. Co., 27 Ca1. 3d 359, 612 P.2d 889, 895 (1980). For these reasons and because principles of equitable contribution may underlie apportionment decisions based primarily on other insurance clauses, this paper first addresses equitable contribution as - 5 -

6 a basis for apportionment of liability among insurers. See Rebecca M. Bratspies, Splitting the Baby: Apportioning Environmental Losses Among Triggered Insurance Policies, 1999 BYU L. Rev. 1215, (1999). 1. Equitable Contribution Principles The availability of equitable contribution between co-insurers of the same risk has been widely recognized for more than a century. See, e.g., Barnes v. Hartford Fire Ins. Co., 9 F. 813 (C.C.D. Minn. 1882) (where several insurers take separate risks upon same property, and loss occurs, companies liable in the ratio their risks bear respectively to the total risk); Thurston v. Koch, 23 F. Cas (C.C.D. Pa. 1800) (in case of double insurance, insurer that has paid whole loss can compel others to contribute); Wiggin v. Suffolk Ins. Co., 35 Mass. 145 (1836) (in case of double insurance, insured may elect to sue either or both insurers, and they may have contribution between themselves). The doctrine is grounded in principles of unjust enrichment. See Nat'l Cas. Co. v. Vigilant Ins. Co., 466 F. Supp. 2d 533, 544 (S.D.N.Y. 2006); Maryland Cas. Co. v. W.R. Grace & Co., 218 F.3d 204, (2d Cir. 2000). Because a co-insurer is unjustly enriched where a shared obligation is discharged by another insurer, the doctrine of contribution operates to spread the liability equitably. Id. 2 2 The author s research suggests that recognition of contribution rights may well be universal with respect to the duty to indemnify. It should be noted, however, that a minority of jurisdictions refuse claims for contribution with regard to the duty to defend, finding that duty to be personal to each insurer and not subject to division. John Burns Constr. Co. v. Indiana Ins. Co., 189 Ill.2d 570, 727 N.E.2d 211, 918 (2000) (where insured invoked coverage of only one of two policies, the chosen insurer was foreclosed from seeking equitable contribution from the other for defense expenses); Argonaut Ins. Co. v. Maryland Cas. Co., 372 So.2d 960, 963 (Fla. Dist. Ct. App. 1979) (holding that because the duty of each insurer to defend its insured is personal and cannot inure to the benefit of another insurer... [c]ontribution is not allowed between insurers for expenses incurred in defense of a mutual insured ); Transamerica Ins. Group v. Empire Mut. Ins. Co., 31 Conn. Supp. 235, 327 A.2d 734, 735 (Conn. Com. Pl. 1974) (duty to defend is personal to each insurer so that obligation is several and an insurer is not entitled to divide the duty to defend nor to require contribution for defense from another insurer without a specific contractual agreement to that effect); Fidelity & Cas. Co. of N. Y. v. Ohio Cas. Ins. Co., 482 P.2d 924, 926 (Okla. 1971) (duty of insurer to defend lawsuits against insured is personal to each insurer, and carrier is not entitled to divide duty nor require contribution from another carrier absent specific contractual rights)

7 Because contribution is an equitable doctrine, courts may consider a wide variety of factors in determining how to spread liability among co-insurers when other insurance clauses are insufficient. See Signal Cos., 612 P.2d at 895 (declining to adopt a bright-line rule for apportionment of liability among insurers). Among these considerations are the particular terms, exclusions and limits of the respective insurance policies in effect;... the nature of the given claim; the relation of the insured to the several insurers; and the relative amount of premiums paid. Centennial Ins. Co. v. United States Fire Ins., 88 Cal. App. 4th 105, 105 Cal. Rptr. 2d 559, 565 (2001) (mentioning also the time each co-insurer is on the risk, a factor applicable to apportionment between successive insurers, a topic considered in greater detail below). Although they may consider a wide array of equitable factors, courts also have developed several set methods for apportioning liability among concurrent insurers: based on policy limits; based on premiums paid; and in equal shares up to the maximum loss each insurer would incur in the absence of other insurers. See CNA Cas. of Cal. v. Seaboard Sur. Co., 176 Cal. App. 3d 598, 222 Cal. Rptr. 276, 289 (1986) (apportionment by policy limits); Insurance Co. of Tex. v. Employers Liab. Assur. Corp., 163 F. Supp. 143, 147, 151 (S.D. Cal. 1958) (apportionment based on premiums paid); Reliance Ins. Co. v. St. Paul Surplus Lines Ins., 753 F.2d 1288, 1292 (4th Cir. 1985) (apportionment in equal shares). Because courts apportioning liability among successive insurers are likely to draw on established methods of apportioning liability among concurrent insurers, each of these methods will be briefly described. First, the majority approach apportionment based on policy limits apportions liability among triggered policies by assigning liability in proportion to the ratio between a triggered policy s limits and the sum of all triggered policies limits. See Seaboard, 222 Cal. Rptr. at 289; - 7 -

8 see also cases cited in Ostrager & Newman, Handbook On Insurance Coverage Disputes 6.021a[11] (12th ed. 2003). So, for example, where an insured had concurrent policies with limits of $100,000 and $200,000, respectively, a court apportioning liability based on policy limits would allocate one third of the liability to the first policy (because its $100,000 limit represents one third of the $300,000 sum of the limits) and two thirds to the second (because its $200,000 limit represents two thirds of the $300,000 sum of the limits) until the policy limits were met. The rationale for this method of apportionment is that, in addition to receiving higher premiums, insurers with higher policy limits have greater potential liability and thus have greater interest in the litigation. See Federal Ins. Co. v. Cablevision Sys. Dev. Co., 662 F. Supp. 1537, (E.D.N.Y. 1987). As one commentator notes, however, the problem with this approach is that although simple to administer, it completely ignores the layers of coverage a policyholder may have purchased and assumes that all insurers retain the same risk of loss vis-avis a claim by the policyholder. Bratspies, 1999 BYU L. Rev. at 1254; see Reliance, 753 F.2d at In addition, it may lead to inequitable results because the cost of purchasing insurance does not increase proportionately with the policy limit, rather, the principal expense in purchasing insurance is the cost of minimum coverage, with additional amounts of coverage being relatively inexpensive in comparison. Reliance, 753 F.2d at Second, apportionment based on premiums paid seeks to avoid the possible inequitable results under the policy limits approach. See Insurance Co. of Tex., 163 F. Supp. at 147. As its name indicates, this method allocates liability among insurers in proportion to the policies relative premium amounts. See id. However, unless the policies provide identical coverage, many variables may effect [sic] the amount of premium charged thereby making the amount of - 8 -

9 premium received an inaccurate reflection of the insurer's ultimate obligation. Reliance, 753 F.2d at Thus, apportionment based on premiums paid also may lead to inequitable results. Third, the equal shares or maximum loss method allocates liability equally among insurers until the lowest policy limit is reached, then equally among the remaining policies until the limits of the next-lowest limit is met, and so on until all limits are exhausted or liability is discharged. See Mission Ins. Co. v. Allendale Ins. Co., 95 Wash. 2d 464, 626 P.2d 505, 507 (1981). 3 In addition to the apportionment methods developed in the concurrent insurance context, courts allocating liability among insurers following an all sums award in long-tail coverage litigation may also look to proration precedents for guidance, albeit they presumably will confine their analysis to periods and layers for which insurance was purchased, and to policies whose material terms can be proved. In other words, whereas courts in pro rata jurisdictions apportion liability based on the ratio between an insurer s coverage period and the duration of the triggering occurrence, courts in all sums jurisdictions likely will consider the ratio between an insurer s coverage period and the total triggered time for which insurance is available. See Bratspies, 1999 BYU L. Rev. at Armstrong World Industries provides an example of second-stage apportionment in an all sums jurisdiction that was based, in part, on time on the risk and, in part, on principles developed in the concurrent insurance context. See 52 Cal. Rptr. 2d at In that case, which involved asbestos bodily injury claims, the trial court employed a method of allocation 3 Many courts refer to the equal shares and maximum loss methods as distinct apportionment methods. However, as described by the case most frequently cited as an example of equal shares allocation, [t]he equal share method apportions the loss equally between the two insurers until the coverage provided by one is exhausted, with any remaining portion of the loss being paid by the other insurer until its policy limit is reached. Reliance, 753 F.2d at This description indicates that the equal shares and maximum loss methods are essentially the same

10 based on policy limits multiplied by years of coverage. Id. at 707. The trial court described its method thus: [W]hen a particular claim triggers more than one policy, each insurer s share of liability shall be determined by the proportion that each policy's applicable per occurrence limits multiplied by years the policy was in effect bears to the sum total of the applicable per occurrence limits of all triggered policies multiplied by the years each policy was in effect. When a policy does not contain a per occurrence limit, the per person limit shall be used in this calculation. Id. at 708 (quoting from trial court opinion). Although this apportionment method was not challenged on appeal, the appellate court nevertheless commented on it, describing it as nontraditional but sound. Id. The appellate court was quick to point out, however, that different approaches may be appropriate in other cases in light of varying equitable considerations that may arise. Id. The Armstrong World Industries apportionment method is, of course, subject to the same criticism frequently leveled at the policy limits method applied in the concurrent coverage context: arguably, it gives undue weight to limits and insufficient significance to the insurer s coverage threshold. Other courts likely will adopt different approaches. Some may endorse the concepts that underlie the equal shares or maximum loss method. Some may even borrow from the analysis of proration cases that purport to account for changes over time with regard to the perceived magnitude of the risk transferred to insurers. See Owens-Illinois, 650 A.2d at ; Carter-Wallace, Inc. v. Admiral Ins. Co., 154 N.J. 312, 712 A.2d 1116, (1998). Given the complexity and variety of long-tail allocation scenarios, it seems unlikely that any one approach will fit all cases as courts in all sums jurisdictions endeavor to do equity in response to second-stage contribution claims

11 2. Other Insurance Clauses As discussed briefly above, another way in which courts have addressed the apportionment of liability among insurers is to examine and attempt to reconcile other insurance clauses in the policies. There are three main classifications of other insurance clauses: pro rata, excess, and escape. Pro rata clauses limit an insurer's liability to a proportion of the total loss; excess clauses provide that a policy is excess after other policies; and escape clauses allow the insurer to avoid all liability. See Am. Cas. Co. of Reading, Pa. v. Health Care Indem., Inc., 613 F. Supp. 2d 1310, 1318 (M.D. Fla. 2009); Planet Ins. Co. v. Ertz, 920 S.W.2d 591, 593 (Mo. Ct. App. 1996). An example of a pro rata clause provides that: If the insured has other insurance against liability or loss covered by this policy, the company shall not be liable for a greater proportion of such liability or loss than the applicable limit of liability bears to the total applicable limit of liability of all collectible insurance against such liability or loss. A typical excess clause may state that: If other valid and collectible insurance with any other Insurer is available to the Assured covering a loss also covered by this insurance other than insurance that is specifically stated to be excess of this insurance, this insurance shall be in excess of and shall not contribute with other insurance. A standard escape clause provides that: If any other Assured included in this insurance is covered by valid and collectible insurance against a claim also covered by this Policy, he shall not be entitled to protection under this Policy. In cases where two policies contain the same type of other insurance clause (e.g. pro rata v. pro rata), courts will apportion liability such that each pays its pro rata share. See Am

12 Cas. Co. of Reading, Pa., 613 F. Supp. 2d at 1319; State Auto Mut. Ins. Co. v. State Farm Mut. Auto Ins. Co., 456 F.2d 238, 239 (6th Cir. 1972). In cases involving conflicting other insurance provisions, courts have applied divergent allocation methods. A minority of jurisdictions hold that all other insurance clauses are mutually repugnant and consequently prorate liability. See Hoffmaster v. Harleysville Ins. Co., 441 Pa. Super. 490, 657 A.2d 1274, 1277 (Pa. Super. Ct. 1995); State Farm Mut. Auto. Ins. Co. v. United States Fid. & Guar. Co., 490 F.2d 407, 411 (4th Cir. 1974). The majority rule, however, dictates that the contract language be examined in order to give effect to the parties intentions. See Citizens Ins. Co. v. Ganschow, 859 N.E.2d 786, 795 (Ind. Ct. App. 2007) (if the policies respective provisions are capable of being harmonized, the court will give effect to the parties intent). In allocating liability between policies that contain a pro rata and an excess clause, courts generally hold that the pro rata policy provides primary coverage. See W9/PHC Real Estate LP v. Farm Family Cas. Ins. Co., 407 N.J. Super. 177, 970 A.2d 382, 398 (N.J. App. Div. 2009); American Auto Ins. Co. v. Transportation Ins. Co., 288 F. App'x 219, 225 (6th Cir. 2008); Orsi v. Aetna Ins. Co., 41 Wash. App. 233, 703 P.2d 1053, 1058 (1985). The reasoning behind this view is that the excess policy does not qualify as other valid and collectible insurance in order to trigger the pro rata clause. Jones v. Medox, Inc. 430 A. 2d 488, 491 (D.C. 1981). Similarly, in reconciling a pro rata and escape clause, courts will find that the pro rata policy provides primary coverage. The policy with the pro rata clause is designated as constituting valid and collectible insurance in order to trigger the escape clause. See Aetna Cas. & Sur. Co. v. Continental Cas. Co., 413 Mass. 730, 604 N.E.2d 30, 32 (1992)

13 In contrast, where the conflict arises from an excess clause and an escape clause, the policy with the escape clause is found to provide primary coverage. The rationale is that the excess policy does not constitute valid and collectible coverage within the meaning of the escape clause. Protective Nat'l Ins. Co. of Omaha v. Bell, 361 So.2d 1058, 1060 (Ala. 1978). However, if the escape clause specifically states that the policy is exempt from coverage by the existence of the other insurance, including primary, excess or contingent coverage, this super escape clause is given effect. See Genie Indus. v. Fed. Ins. Co., 316 Fed. Appx. 540, 541 (9th Cir. 2008); New Hampshire Indem. Co., Inc. v. Budget Rent-A-Car Sys., Inc., 148 Wash. 2d 929, 64 P.3d 1239, 1242 (2003). The reasoning derives from giving priority to the parties clear intent. With the exception of California, the all sums jurisdictions have yet to produce published opinions addressing the application of other insurance clauses in contribution actions involving successive coverage. California appears to have adopted the minority rule regarding the mutual repugnancy of other insurance clauses. In Travelers Cas. and Sur. Co. v. Century Sur. Co., one of the insured s liability insurers brought a contribution claim against the other insurer for defense costs incurred in the underlying property damage action. 118 Cal. App. 4th 1156, 13 Cal. Rptr. 3d 526, 527 (2004). The plaintiff s policy contained a pro rata clause, whereas the defendant insurer s policy contained an excess other insurance provision. The court noted that the general rule regarding inconsistent other insurance clauses in policies sharing the same risk was to prorate according to policy limits. Id. at 529, quoting Fireman s Fund Ins. Co. v. Maryland Cas. Co., 65 Cal. App. 4th 1279, 77 Cal. Rptr. 2d 296, 312 (Cal. Ct. App. 1998). The court further cited to a prior decision (Century Sur. Co. v. United Pac. Ins. Co., 109 Cal. App. 4th 1246, 135 Cal. Rptr. 2d 879, 887 (2003)) involving a conflict between a pro rata and excess clause for the proposition that the two clauses are mutually repugnant. Id. In

14 affirming pro rata allocation, the court stated that giving effect to defendant's other insurance provision... would result in imposing on plaintiff the burden of shouldering that portion of a continuous loss attributable to the time when defendant was the only liability insurer covering [insured]. Id. at 530. In all sums jurisdictions that have yet to develop rules for allocation of loss among successive insurers, it would be wise for counsel to be cognizant of the potential role of other insurance clauses and the case law developed thereunder. That said, it seems reasonable to predict that other insurance clauses will have a subordinate position, at best, in the reallocation process. Cases embracing pro rata (as opposed to all sums ) allocation principles repeatedly have held that other insurance clauses have no role to play in allocation because such clauses are intended to apply only in the case of concurrent (rather than successive) insurance. See Owens-Illinois, 650 A.2d at 991; Plastics Eng g Co., 759 N.W.2d at ; Boston Gas Co., 910 N.E.2d at 309 n.36. If that proposition is correct, there is no reason why it should not also be correct in all sums jurisdictions. Moreover, given that contribution is an equitable doctrine, if a court does not deem the result dictated by other insurance clauses to be equitable and the results in other insurance contests do often seem arbitrary then presumably courts will feel free to look elsewhere. B. Reallocation with Settled Insurers: Settlement Credits Courts adopting all sums allocation generally recognize a selected insurer s right to seek contribution from other insurers with triggered policies. See cases cited in Section I.A., supra. It is another question entirely whether a selected insurer will be entitled to contribution from insurers that have already settled their liability to the policyholder. Some courts do not permit non-settling insurers to sue settling insurers for contribution. See, e.g., Eli Lilly and Co. v. Aetna Cas. & Sur. Co., No. 49D CP (Ind. Super. Ct. July 15, 2002) (an insurer

15 stands in the shoes of the insured in an equitable contribution action; to the extent that the insured has released its claims against settling insurers, a non-settling insurer has no claim against such settlers). If claims against settled insurers are permitted, so the reasoning goes, insurers will have no incentive to settle, knowing they could nevertheless have to litigate later with non-settling insurers whose policies are triggered by the same occurrences. Not surprisingly, courts also recognize that an insured should not receive from its insurers more than the total amount of its loss. Thus, the question arises: what treatment should the insured s settlement recoveries receive in calculating the liability of the non-settling selected insurers? This issue often is referred to as the settlement credits issue, as courts generally resolve it by applying settlement recoveries as a credit against the total liability in assessing indemnity amounts to non-settling insurers. In applying settlement credits, courts have tried to achieve the dual goals of promoting settlements by protecting settling insurers, while at the same time preventing the insured from obtaining a double recovery at the expense of non-settling insurers. The emerging body of case law addressing settlement credits reflects differing views as to how they should be calculated. Two principal methods have emerged: (1) so-called pro tanto credits, based on the actual amount of the insured s settlement recoveries; or (2) pro rata credits based on what the settled insurers purportedly should have paid, assuming the settled insurers liability, with the difference between those amounts and the actual recoveries borne by the policyholder. This Section will discuss the cases that have determined and applied settlement credits in all sums jurisdictions. 1. Credit Based on Actual Recoveries the Pro Tanto Approach Several courts applying the all sums method of allocation have ruled that non-settling insurers are entitled only to a pro tanto credit for the policyholder s settlement recoveries. See, e.g., Insurance Co. of N. Am. v. Kayser-Roth Corp., 770 A.2d 403, 413 (R.I. 2001)); Eli Lilly and

16 Co. v. Aetna Cas. and Sur. Co., No. 49D CP (Ind. Sup. Ct. July 15, 2002); Weyerhaeuser Co. v. Commercial Union Ins. Co., 15 P.3d 115, (Wash. 2000). Under the pro tanto approach, an insured s potential recovery against non-settlers is reduced dollar-fordollar by the amount of prior settlements pertaining to the same claim. Thus, if the insured s loss is $1 million, but it has already recovered $500,000 from other settling insurers for that loss, then it may recover only $500,000 in total from the non-settling insurers for such loss. Courts applying the pro tanto approach have held that such an approach is sufficient to remove the possibility that a policyholder will receive the windfall benefit of a double recovery. Id. The rationale for the pro tanto approach was explained by the Indiana Superior Court in Eli Lilly. The court first explained that where state jurisprudence provides that insurers are liable for all sums for which the insured becomes obligated, allowing a non-settling insurer to recover in a contribution action against settling insurers a pro rata reallocation of the insured s losses would produce a result with resounding consequences and one which has been firmly rejected [by the state courts]. Eli Lilly, No. 49D CP , at *3. The court noted further that allowing non-settlers to seek pro rata contribution from settling insurers would effectively eliminate early or piecemeal settlements. Id. at *4. Also, where the insured agrees to indemnify the settling insurer, granting a contribution right to the non-settling insurer would compel the insured to self-insure a loss for which it had paid insurance. Id. The Court concluded that application of the pro tanto settlement credit approach protected the non-settling insurer from paying more than the insured s damages and prevented a double recovery by the insured. Id; see also RSR Corp. v. Int'l Ins. Co., 2009 U.S. Dist. LEXIS at *45-46 (N.D. Tex. Mar. 23, 2009) (holding that non-settled insurers held liable for all sums are entitled to a pro tanto credit for any settlement payment already received by the insured)

17 Critics of the pro tanto approach argue that it is unfair to allow the insured unilaterally to apportion liability among insurers by settling with some insurers and holding the non-settled insurers liable for the entire remainder of the insured s loss. See GenCorp, Inc. v. AIU Ins. Co., 297 F. Supp. 2d 995, 1007 (N.D. Ohio 2003) (holding that settlement credits for less than the full amount of settled policy limits saddle the non-settling insurers with more than their contractedfor share of liability without any recourse to reduce their burden). In addition, courts employing the pro tanto approach have encountered difficulty in determining how much of a given settlement should be allocated to the particular claim for which the non-settled insurers are being held liable. In many instances, the insured s settlement with its other insurers resolves more than one dispute, and it is difficult, if not impossible, to discern what portion of the settlement was intended to compensate the policyholder for the particular loss on account of which the non-settled insurers have been adjudged liable. See Weyerhauser, 15 P.3d at 126. When the policyholder s prior settlements do not specifically allocate settlement amounts to particular claims, there may not be any easy way to determine the proper amount of the settlement credit to which non-settled insurers are entitled. Some courts have dealt with this complication by placing the burden of proving the proper allocation of the settlement on the non-settling insurers. See, e.g., Puget Sound Energy, Inc. v. ALBA Gen. Ins. Co., 149 Wn. 2d 135, 68 P.3d 1061, 1064 (Wash. 2003) ( if a non-settling insurer seeks to offset its responsibility for a claim using proceeds from such a settlement, it has the burden of establishing what part of the settlement was attributable to the claim it seeks to offset. ); Goodrich Corp. v. Commercial Union Ins. Co., 2008 Ohio App. LEXIS 2716, *25 (Ohio Ct. App. June 30, 2008) (refusing to provide settlement credits to non-settling insurers

18 who had not satisfied their burden of proof that, without a set-off, the policyholder would receive a double recovery); Kayser-Roth Corp., 770 A.2d at 414; Weyerhauser, 15 P.3d at 126. In Weyerhaeuser, the court noted that the insurance settlement in the case before it was not simply a payment to offset losses the insured incurred in connection with particular sites. Rather, the settlement was in exchange for a release of liability for all past, present and future environmental claims. 15 P.3d at 126. As a result, the settling insurers received far more than a simple release of liability at specific sites.... [T]hese companies also purchased certainty by avoiding the risks of an adverse trial outcome not to mention foregoing the expenses associated with a lengthy trial and appeal. Id. The court thus held that the burden was on the non-settling insurers to prove that the policyholder would receive a double recovery before any settlement credit would be allowed: [w]ere we to hold that the insured bears the burden of proving it has not received a double recovery, such a rule would encourage litigation and reward the nonsettling insurer for refusing to settle. Id. 2. Credit Based on the Settled Policy Limits the Pro-Rata Approach A second approach provides for a pro rata settlement credit, determined not by reference to what the policyholder actually received from the settled insurers, but by reference to what the settled policies would have paid had there been no settlement and the settled insurer was actually held liable. Often, this approach will allow non-settling insurers to enjoy the benefit of a settlement credit in the full amount of the settled policies limits. See Chemical Leaman Tank Lines v. Aetna Cas. & Sur. Co., 177 F.3d 210, 227 (3d Cir. 1999) ( an excess carrier is entitled to a credit, not based on the primary carrier's settlement, but based on the amount allocable to the primary under its policies. ) (emphasis in original). The pro rata approach to settlement credits has been adopted in Ohio (GenCorp, 297 F. Supp. 2d at ), New Jersey (Chemical

19 Leaman, 177 F.3d at ) and Pennsylvania (Koppers Co., Inc. v. Aetna Cas. & Sur. Co., 98 F.3d 1440, (3d Cir. 1996)). The leading case applying settlement credits on a pro rata basis is the Third Circuit s decision in Koppers. 98 F.3d at In Koppers, the defendant excess liability insurers were held jointly and severally liable for Koppers liability arising out of various long-term pollution sites. Prior to judgment against the defendant excess insurers, Koppers reached settlements with its primary insurers and certain of its excess insurers. The district court had held that the excess insurers were liable for the full amount of the claim ($70 million) and were not entitled to any credit or reduction to account for Koppers previous settlements. The Third Circuit reversed, finding that Pennsylvania law would reduce the amount of the judgment against the excess insurers by the settling insurers apportioned shares of liability. The court noted that, on remand, the district court would have to determine each settled insurer s apportioned share of liability, in part, by reference to other insurance clauses in the policies. The court observed that these clauses may require that, the apportioned share of a settled primary policy covering the same indivisible loss is its full policy limits. Thus, the district court may be required to deduct from the total loss the combined limits of all settled primary policies. Id. at The court held that a policyholder s settlement with its primary insurer functionally exhausts the primary policy limits and triggers the excess policy. Id. at However, the policyholder loses any right to recover the difference between the settlement amount and the primary policy s limits by virtue of its settlement below the policy limits; an excess insurer has no duty to drop down to cover any gap that may exist between the settlement amount and the underlying policy limits. Id. Thus, the excess insurers could be held liable only for amounts in excess of the underlying primary policy limits. The court stated that this rule was in accord with

20 Pennsylvania public policy, which favors placing the risk of a low settlement on the insured as means of guarding against the risk of collusive settlements. Id. at 1453 n 15. In GenCorp, the United States District Court for the Northern District of Ohio held that selected excess insurers were entitled to set-off against their liability the full policy limits of any settled underlying insurer. 297 F. Supp. 2d at The excess insurers, who were alleged to be jointly and severally liable for all of GenCorp s losses, filed third-party complaints against each of GenCorp s settled primary insurers, arguing that, to the extent the excess insurers could be found liable to GenCorp, they were entitled to contribution or indemnification from the primary insurers. Although the court dismissed the third-party complaints on the basis that GenCorp had settled with its primary insurers, thereby extinguishing their liability to GenCorp, the court also found that the settlement with the primary insurers exhausted GenCorp s primary insurance for purposes of making claims under the excess policies. Id. Subsequent to the dismissal of the third-party complaints, the excess insurers filed a motion for summary judgment requesting credit for the settlements that GenCorp had reached with its primary insurers in the full amount of the policy limits available under each of the settled policies. The court granted the motion, and although its opinion stated that it was reserving the question of whether Ohio law permits settlement credits, the practical effect of the decision was to give the non-settling excess insurers credit for the full policy limits of the settling underlying insurers. The court found that GenCorp s liability would not exceed the limits of its primary insurance coverage and, because GenCorp had chosen to settle with each of its primary insurers, thereby exhausting its primary coverage, it could not now recover from its excess insurers the difference between the settlement amounts received and the limits of the underlying policies. Id

21 at Otherwise, the court noted, the excess insurers would be subjected to more than their contracted-for share of GenCorp s liability. Id. at Critics of the pro rata approach argue that it undermines the all sums rule in at least two ways. First, the all sums rule is supposed to relieve the insured of the burden of litigating the division of responsibility among insurers; however, the pro rata approach to settlement credits would require the insured, in its suit against the selected insurer, to apportion liability among the settled policies. Second, whereas all sums allocation is intended to ensure that the policyholder is fully covered for its loss, so long as there is triggered insurance available to pay, the pro rata approach to settlement credits creates the possibility that an insured will not recover the full amount of its loss. As the Indiana Superior Court reasoned in Eli Lilly, to require an insured to shoulder the responsibility for the gap between the settlement amount and the full allocated share of liability for the settled policies usually would leave the insured undercompensated, despite the availability of triggered policies from the non-settling insurers. Eli Lilly, No. 49D CP , at *3; accord Cascade Corp. v. American Home Assur. Co., 206 Ore. App. 1, 135 P.3d 450, (Or. Ct. App. 2006) (reversing the trial court s application of the pro rata method based on settled insurers policy limits because it would have left the insured with a loss for which there is no coverage). Finally, critics of the pro rata approach argue that it provides a powerful disincentive for the insured to settle because it requires an insured to bear the uncertainty of how a court might apportion liability among insurers in the future. Similarly, the prospect that an insurer can reduce its exposure by waiting for the insured to settle with other triggered policies encourages insurers to litigate, rather than settle, their disputes with the insured

22 II. PRACTICAL CONSIDERATIONS: THE POLICYHOLDER'S PERSPECTIVE From the policyholder s perspective, perhaps the greatest virtue of the all sums rule is that it avoids allocation of a portion of the underlying liability to the policyholder itself, whereas jurisdictions embracing proration concepts will assign a share of liability to the policyholder for, inter alia, periods during the injurious process for which no insurance can be found or when no coverage was purchased because it was unavailable or prohibitively expensive. See, e.g., Domtar, Inc. v. Niagara Fire Ins. Co., 563 N.W.2d 724, (Minn. 1997) (prorating liability and assigning to policyholder share of environmental cleanup costs allocable to 23-year period for which no policies could be located and 26-year period from advent of pollution exclusion to commencement of cleanup). Under all sums, ostensibly the policyholder is taken out of the equation, at least to the extent that the liability exceeds any selfinsured retention ( SIR ) in selected years and is within available limits of liability. This does not mean, however, that the fallout from an all sums ruling will never have a negative impact on a policyholder. Perhaps the most practically significant such impact would be costs and potential liability in connection with a contribution claim by the selected insurers against a settled insurer that the policyholder has agreed to defend and indemnify against claims by third parties. There may be public policy grounds on which such a claim can be rebuffed (see discussion, supra, at Section I.B.), but if the contribution case were to yield a finding that the selected insurer paid, or was held liable to pay, more than its fair share, and that the settled insurer paid less than its fair share, then the policyholder could be liable to the selected insurer for contribution, as indemnitor of the settled insurer. Because settlements tend to be struck at a time when insurer liability has yet to be established, the prospect is real that the settled insurer s compromise payment will be deemed to be less than its fair share. And, in any event, the costs of defending the contribution claim will be significant

23 Claims invoking indemnity agreements in favor of settled insurers are not the only potential impacts on the policyholder from a selected insurer s contribution action. The selected insurer s claims will not necessarily reflect the same considerations that dictated the policyholder s selection. For example, the policyholder may select policy periods so as to avoid periods governed by so-called retrospective rating plans, under which the insurer s defense or claim payments can result in retrospective premium assessments to the policyholder many years after the termination of the policy period. There is nothing to prevent a selected insurer from asserting its contribution claim against another insurer whose policies are governed by such a plan. If the claim is successful, then the contribution payments to the selected insurer may entitle the contribution defendant insurer to assess the policyholder for retrospective premiums. The selected insurer also could choose to seek contribution from an insurer that wrote a so-called fronting policy for the policyholder. Under a fronting arrangement, an insurer writes a policy typically to enable the policyholder to satisfy contractual financial assurance requirements and the policyholder agrees to indemnify the insurer for any liability under the policy. An all sums selected insurer presumably could assert a claim for contribution against such an insurer, and the policyholder would then have to make good on its promise to indemnify the fronting insurer. Somewhat less direct but no less real results might occur where the policyholder has formed a captive insurer whose policies would apply. Potential impacts on association captives insurers owned by a number of companies in the same industry also must be anticipated. 4 4 Some insurers may even take the position that they should be permitted to seek contribution from the policyholder for periods during which the policyholder has high self-insured retentions and maintains reserves as a substitute for lower attachment-point insurance. Policyholders undoubtedly will resist such claims vigorously, asserting that there is no common liability, and therefore no contribution right, because in such situations the insured has not actually undertaken a contractual obligation to indemnify itself. See Keene Corp v. Insurance Co

24 Another potential impact of a second-stage insurer contribution action is the exhaustion of applicable aggregate limits of liability particularly of primary policies. If we suppose that the selected insurers in the policyholder s coverage case include excess carriers in the chosen policy period, such that the initial imposition of insurer liability forms a so-called vertical spike in the coverage profile, then the excess carriers presumably will seek to redistribute the liability into lower layers in other years, potentially exhausting the aggregate limits of primary policies. At least where all primary coverage is thereby exhausted, this can have significant consequences for the policyholder. Whereas primary policies typically impose on the insurer an executory defense duty whose cost is excess of limits i.e., does not draw down limits of liability excess policies serving as the policyholder s next resort often will include only defense expenses within covered ultimate net loss, such that the policyholder will need to supervise and initially pay for the defense itself, with any subsequent reimbursement coming at the expense of eroding limits. If ultimate exhaustion of excess coverage is likely, as often is the case in the asbestos liability context, then the cessation of the primary carrier s excess of limits defense activity will hasten the arrival of exhaustion and effectively reduce the value of available coverage. The all sums approach to the extent-of-coverage issue thus has both short term impacts satisfaction of the policyholder's coverage claim and perhaps disproportionate imposition of liability upon selected insurers and longer term impacts that may result from a second-stage selected insurer s contribution action. In deciding how to structure a coverage case, and which coverage periods to select for indemnification, it will be important to anticipate both the short of N. Am., 667 F.2d 1034, (D.C. Cir. 1981) (declining to postulate self-insurance policy that is triggered for periods in which no other policy was purchased )