A Viking on Choppy Waters

The Viking Pump court's application of “all sums” allocation begins with an explication of New York's trigger-of-coverage doctrine that is at odds with controlling authority in New York itself. After stating the rule correctly ' that “New York law ' generally holds that an occurrence based policy is triggered upon an 'injury-in-fact to a tort plaintiff” ' the court in Viking Pump in the same breath asserted that this means “where contract language indicates that an 'occurrence' is an injurious exposure to conditions, which results in personal injury,' then the injury-in-fact theory dictates that the plaintiffs' exposure to asbestos attributable to the insured during the policy period triggers the policy.” 2009 WL 3297559, at *22. The court went on to conclude that “the words 'during the policy period'” included as a limiting phrase within the definition of a covered bodily injury “simply require that the insured's liability for the claim in question be attributable to an occurrence during the policy period.” Id. at 29. By treating the concepts of “occurrence” and “injury” as interchangeable, the court failed to grasp that the coverage at issue was limited to injury “during the policy period” ' and therefore failed to allocate to each policy only that portion of the loss attributable to it, as required under Con Ed.

The presumption in Viking Pump that exposure to asbestos constitutes an injury-in-fact triggering coverage has no basis in New York law. On the contrary, the First Department recently observed that “actual [asbestos-related] injury generally develops over time depending on a range of circumstances and conditions, but does not occur upon exposure by inhalation.” Continental Casualty Co. v. Employers Insurance Co. of Wausau, 871 N.Y.S.2d 48, 62 (App. Div. 1st Dep't 2008). To be sure, a number of New York cases found that asbestos-related injury-in-fact did occur upon exposure to the policyholder's asbestos-containing products. However, these findings reflect the results of case-by-case adjudication based on a full factual record, not application of a presumption of injury upon exposure. See, e.g., Stonewall Ins. Co. v. Asbestos Claims Mgmt. Corp., No. 86 Civ. 9671 (JSM), 1998 U.S. Dist. LEXIS 10954, at *7 (S.D.N.Y. July 16, 1998) (discussing resolution of injury-in-fact question by trial verdicts). The New York cases could not be more clear that such a presumption is an alternative to injury-in-fact, and is viewed by New York courts to be at odds with the injury-in-fact doctrine. See American Home Prods. Corp. v. Liberty Mut. Ins. Co., 565 F. Supp. 1485 (D.N.Y. 1983), aff'd 748 F.2d 760 (2d Cir. 1984).

The only New York case cited by Viking Pump as authority for applying an exposure trigger of coverage was In re Midland Ins. Co., 709 N.Y.S.2d 24 (1st Dep't 2000). However, Midland interpreted atypical “occurrence trigger” policy language. While coverage under most general liability policies ' including those at issue in Viking Pump ' is triggered by an injury during the policy period, the policies at issue in Midland were triggered by proof that the causative occurrences happened during the policy period, regardless of the timing of the resulting injuries. See Id. at 31; Robert F. Cusumano and Steve Vaccaro, “The Final Pieces of the Trigger and Allocation Puzzle in New York,” Ins. Coverage L. Bull. (Mar. 2004) at 1, 5. Based on this misplaced reliance on Midland, the Viking Pump court went seriously astray when it concluded that “New York courts have generally found that a plaintiff who proves that she suffered compensable damage as a result of asbestos exposure is injured during all periods of material exposure and therefore that any policy is triggered if it was in existence when the exposure occurred.” 2009 WL 3297559, at *22.

One Possible Lesson

A further possible reason for the Viking Pump court's trigger presumption may have been the court's desire to resolve the allocation issue “at a high level” and in the abstract, without the factual record necessary to support an appropriate injury-in-fact determination under New York law. See Excess Insurers' Joint Brief (D.I. 2149, Apr. 29, 2009), at 21, n.11 (describing court's order that its allocation determination “would only include a high level determination of the proper allocation method (i.e., pro rata vs. all sums),” and not consider closely linked issues such as the policyholder's obligation for uninsured or underinsured periods). If so, one lesson of Viking Pump may be that an attempt to resolve the allocation method prior to determining trigger of coverage creates a serious risk that a trigger presumption, rather that the required injury-in-fact inquiry, will serve as the basis for the allocation.

Viking Pump's Departure From New York Number-of-Occurrences Doctrine

As explained below, the Viking Pump court's “all sums” holding relies heavily on non-cumulation and prior insurance clauses, which provided for reduction of policy limits due to payments from other policies based upon the same occurrence or loss. The court emphasized the conflict between these clauses and pro rata allocation method, because the court believed that each claim against the policyholder reflects a separate and single occurrence. The court based its view upon an examination of neither the facts of the underlying claims nor the relevant New York law, but rather on the presumption that each claim constituted a separate occurrence. See 2009 WL 3297559, at *22 (“New York law can be seen as treating each asbestos plaintiff's exposure and injury as a single “occurrence” (that triggers all the multiple policies in place during the period of asbestos exposure)); Id. at *30 (“[I]n the case of an asbestos plaintiff with Multi-Period Exposure, the plaintiff's injury is treated as indivisible and resulting from one occurrence.”).

The Viking Pump court cited to Appalachian Insurance Co. v. General Electric Co., 863 N.E.2d 994 (N.Y. 2007), in support of its multiple occurrence presumption. However GE cautioned that “[e]ach mass tort scenario must be examined separately,” and explained that determination of the number of occurrences in the asbestos bodily injury context required an analysis of the temporal and spatial proximity between the causative events and the loss, and the existence of an unbroken causal continuum spanning from those events to the resulting loss. Id. at 999. If the Viking Pump court embarked upon such an inquiry, it did not recount it in its decision.

The court's number-of-occurrences presumption, like its trigger presumption, illustrates the potential dangers of attempting to adjudicate allocation issues prior to establishing a clear factual record regarding the exact nature of exposures and circumstances giving rise to the underlying claims. Had the court applied the analysis required by GE, it might well have found that each claimant's injuries arose from separate occurrences ' and, therefore, the non-cumulation and prior insurance clauses did not come into play. Or, it might have concluded that all of the claims against the policyholder constituted a single occurrence, in which case the allocation dispute would have been rendered largely moot by application of the non-cumulation and prior insurance clauses.

Viking Pump's Departure from New York Allocation Doctrine

Based on these questionable presumptions as to the trigger of coverage and number of occurrences, the Chancery Court turned to the issue of the proper allocation method. Before discussing the factors it believed rendered the Con Ed and other New York pro rata allocation authorities distinguishable, the court criticized both, accusing New York federal courts of showing a “surprising” disregard for Erie on the allocation issue, and suggesting that the Con Ed decision itself is too “abbreviated” and “terse” to guide subsequent allocation decisions. 2009 WL 3297559, at *25-*26. Throughout this preliminary discussion of New York allocation authorities, the court relied on its exposure trigger presumption to vitiate the significance of the limiting phrase “during the policy period” applied to concept of “injury” in the policies at issue: “If, as appears to be the case, a policy would be deemed triggered under New York law simply because there was material asbestos exposure during the policy period, the use of the 'during the policy period' language does not itself answer the key [allocation] question ' . ” Id. at *26. Despite the express reliance of the Con Ed court on the limiting language “during the policy period” in concluding that pro rata allocation would apply in the asbestos bodily injury claims before it, the Viking Pump court nonetheless concluded that under New York law, “[a]t least as applied to asbestos exposure cases, the words 'during the policy period' do not shed much light on what method of allocation is intended.” Id.

Though critical of Con Ed as “extremely abbreviated,” the Viking Pump court omitted mention of the key rationale for pro rata allocation, discussed at length in that case. In Con Ed, because there was no evidence of specific damage during any particular policy period, the damage at issue was found to have occurred continuously during the 50-year coverage period. The policyholder argued that this meant every policy in effect during that period was triggered and available to respond on an “all sums” basis. 746 N.Y.S.2d at 630. The court found this contradiction in the policyholder's position untenable, because “collecting all the indemnity from a particular policy,” as permitted under an “all sums” allocation method, “presupposes [the] ability to pin an accident to a particular policy period.” Id. In contrast, “[p]roration of liability among the insurers acknowledges the fact that there is uncertainty as to what actually transpired during the particular policy period.” Id. The court further found, “[m]ost fundamentally, [that] the policies provide indemnification for liability incurred ' during the policy period, not outside that policy period.” Id. (emphasis added). The failure of the Viking Pump court to address this core rationale of Con Ed is inexplicable.

Viking Pump's Interpretation Of Non-Cumulation and Prior Insurance Clauses

The prime factor on which Viking Pump distinguished Con Ed on the issue of allocation was the presence of non-cumulation and prior insurance clauses in the policies at issue. The language of these clauses is unremarkable and is found in many general liability policies; it calls for a reduction in limits based upon payments for related losses under prior policies:

If the same occurrence gives rise to personal injury, property damage or advertising injury or damage which occurs partly before and partly within any annual period of this policy, the each occurrence limit and the applicable aggregate limit or limits of this policy shall be reduced by the amount of each payment made by [Liberty Mutual] with respect to such occurrence, either under a previous policy or policies of which this is a replacement, or under this policy with respect to previous annual periods thereof.

* * *

It is agreed that if any loss covered hereunder is also covered in whole or in part under any other excess Policy issued to the Insured prior to the inception date hereof, the limit of liability hereon stated in the Items 5 and 6 of the Declarations shall be reduced by any amounts due to the Insured on account of such loss under such prior insurance.

2009 WL 3297559, at *28-*29.

It is apparent that these clauses have nothing to do with the trigger and coverage issues underlying the selection of an allocation method. As the Viking Pump court explicitly recognized, the purpose of such clauses is to limit the total amount that a policyholder can collect for all losses due to a single occurrence to the single highest per-occurrence limit of all policies covering any part of such loss ' i.e., to prevent “stacking” of policy limits. 2009 WL 3297559, at *29. However, because the clauses referred to the possibility that a single occurrence might give rise to losses occurring across multiple policy periods, the court concluded that these provisions, and not the language limiting coverage to injuries “during the policy period,” should govern the question of which losses each policy should cover. “By their very presence,” the court reasoned, these clauses “suggest that the words 'during the policy period,' do not have the drastic effect the Excess Insurers contend for.” Id.

The flaw in the court's reasoning is that application of such non-cumulation and prior insurance provisions are entirely consistent with pro rata allocation. Nothing stops insurer and policyholder from agreeing both that: 1) a policy will cover only that injury which occurs during the policy period; and 2) such coverage is subject to the further limitation that payments made in respect of related injuries arising from the same occurrence under previous policies will be deducted from the per-occurrence limits applicable to such coverage.

The Viking Pump court acknowledged that the non-cumulation and prior insurance clauses could both be applied as written, but rejected such an approach because it would “unreasonably reduce coverage,” amount to a “windfall” and “'double credit' for insurers,” and achieve what the court felt were “bizarre and inequitable results.” 2009 WL 3297559, *30-*31 & n.169. In declining to give effect to both the clauses and the express policy requirement that covered injury must occur “during the policy period,” the Viking Pump court made no mention of New York law holding that “[r]elying on social policy to justify imputing an expectation of complete coverage to the insured is legally ' unsupportable.” American Home Prods. Corp. v. Liberty Mut. Ins. Co., 565 F. Supp. 1485, 1511 (S.D.N.Y. 1983), aff'd, 748 F.2d 760 (2d Cir. 1984).

One reason for Viking Pump's decision to uphold the non-cumulation and prior insurance clauses over the interpretation of the controlling “during the policy period” limitation of the Con Ed court was the decision of the Delaware Supreme Court in Hercules Inc. v. AIU Ins. Co., 784 A.2d 481 (Del. 2001), which relied on similar clauses to reach an all sums allocation result. The Viking Pump court's heavy reliance on the Hercules decision raises further questions as to whether the court was faithfully applying New York law.

The Viking Pump court sought to buttress its all-sums holding with evidence of the parties' conduct with respect to allocation, but in so doing the court relied on two questionable propositions. First, the court proposed that the alleged waiver of the policyholder's primary insurer should bind its excess insurers, because the excess insurers were participating in a single integrated insurance program which should not admit differences in position among insurers. The court did not discuss authority in New York and elsewhere suggesting that one insurer cannot bind another in this manner. See United States Fidelity & Guaranty Co. v. Treadwell Corp., 58 F. Supp. 2d 77, 93-99 (S.D.N.Y. 1999) (holding that non-settling insurers were not bound by an allocation agreed to between insured and settling insurers); Allmerica Fin. Corp. v. Certain Underwriters at Lloyd's, London, 871 N.E.2d 418, 429 (Mass. 2007) (holding that “absent an explicit contractual commitment to do so, an insurer is not bound by the settlement another insurer makes for the same claim, even if the language of the nonsettling policy follows the form of the settling policy”).

The Viking Pump court also reached the remarkable conclusion that the primary insurer's allocation of payments pro rata across multiple policy periods did not indicate a practice of pro rata allocation, but rather one of “all sums” allocation, because of the opportunity for insurers in an “all sums” regime to bring contribution actions to recover overpayments from other insurers and, ultimately, to allocate losses to particular policy periods. 2009 WL 3297559, *33 n.184. The court's analysis failed to consider that even an insurer recovering excess payments in contribution would have no reason to allocate its own share among the multiple policy periods, unless it viewed that share as allocable pro-rata.

In brief, the Viking Pump decision appears to pump novel interpretations into virtually every major doctrinal area involved in determining coverage for continuing injuries under New York law. How these novel interpretations will be received on the shores of New York remains to be seen.

When the New York Court of Appeals issued its decision in C onsolidated Edison Company v. Allstate Insurance Co. 98 N.Y.2d 208 (2002), the issue of allocating liability for continuing losses among multiple insurers consecutively liable for the loss appeared to be all but settled. Con Ed held that under policies providing coverage for “all sums” attributable to damage occurring “during the policy period,” where there was no ability to pinpoint exactly when the loss occurred, the most equitable means of apportioning the liability for the losses is in direct proportion to each insurer's time on the risk, absent policy language to the contrary. Although the Con Ed court cautioned that it had not delivered “the last word on pro-ration,” in practice New York courts have presumed that pro rata allocation will apply in continuing injury cases involving standard CGL policy language absent unusual circumstances. See, e.g., Serio v. Public Service Mut. Ins. Co. , 759 N.Y.S.2d 110 (App. Div. 2nd Dep't 2003).

This approach to allocation has been called into question, however, by the recent decision of the Delaware Court of Chancery in Viking Pump, Inc. v. Century Indem. Co., 2009 WL 3297559 (Oct. 14, 2009). Purportedly applying New York law, the Viking Pump court concluded that non-cumulation and prior insurance clauses in the policies at issue effectively trumped an express limitation of coverage to injuries taking place “during the policy period.” The court reached this conclusion by applying presumptions, unprecedented in New York law, that exposure to asbestos rather than injury triggers general liability coverage and that each claimant's exposure to asbestos invariably constitutes a separate occurrence. In charting this novel route to its all-sums result, Viking Pump not only distorts New York's allocation rules in favor of contrary Delaware law, but also muddies other deep waters of New York coverage jurisprudence including the law governing trigger and number-of-occurrences.

Viking Pump's Departure from New York Trigger Doctrine

The Viking Pump court's application of “all sums” allocation begins with an explication of New York's trigger-of-coverage doctrine that is at odds with controlling authority in New York itself. After stating the rule correctly ' that “New York law ' generally holds that an occurrence based policy is triggered upon an 'injury-in-fact to a tort plaintiff” ' the court in Viking Pump in the same breath asserted that this means “where contract language indicates that an 'occurrence' is an injurious exposure to conditions, which results in personal injury,' then the injury-in-fact theory dictates that the plaintiffs' exposure to asbestos attributable to the insured during the policy period triggers the policy.” 2009 WL 3297559, at *22. The court went on to conclude that “the words 'during the policy period'” included as a limiting phrase within the definition of a covered bodily injury “simply require that the insured's liability for the claim in question be attributable to an occurrence during the policy period.” Id. at 29. By treating the concepts of “occurrence” and “injury” as interchangeable, the court failed to grasp that the coverage at issue was limited to injury “during the policy period” ' and therefore failed to allocate to each policy only that portion of the loss attributable to it, as required under Con Ed.

The presumption in Viking Pump that exposure to asbestos constitutes an injury-in-fact triggering coverage has no basis in New York law. On the contrary, the First Department recently observed that “actual [asbestos-related] injury generally develops over time depending on a range of circumstances and conditions, but does not occur upon exposure by inhalation.” Continental Casualty Co. v. Employers Insurance Co. of Wausau , 871 N.Y.S.2d 48, 62 (App. Div. 1st Dep't 2008). To be sure, a number of New York cases found that asbestos-related injury-in-fact did occur upon exposure to the policyholder's asbestos-containing products. However, these findings reflect the results of case-by-case adjudication based on a full factual record, not application of a presumption of injury upon exposure. See, e.g., Stonewall Ins. Co. v. Asbestos Claims Mgmt. Corp., No. 86 Civ. 9671 (JSM), 1998 U.S. Dist. LEXIS 10954, at *7 (S.D.N.Y. July 16, 1998) (discussing resolution of injury-in-fact question by trial verdicts). The New York cases could not be more clear that such a presumption is an alternative to injury-in-fact, and is viewed by New York courts to be at odds with the injury-in-fact doctrine. See American Home Prods. Corp. v. Liberty Mut. Ins. Co. , 565 F. Supp. 1485 (D.N.Y. 1983), aff'd 748 F.2d 760 (2d Cir. 1984).

The only New York case cited by Viking Pump as authority for applying an exposure trigger of coverage was In re Midland Ins. Co., 709 N.Y.S.2d 24 (1st Dep't 2000). However, Midland interpreted atypical “occurrence trigger” policy language. While coverage under most general liability policies ' including those at issue in Viking Pump ' is triggered by an injury during the policy period, the policies at issue in Midland were triggered by proof that the causative occurrences happened during the policy period, regardless of the timing of the resulting injuries. See Id. at 31; Robert F. Cusumano and Steve Vaccaro, “The Final Pieces of the Trigger and Allocation Puzzle in New York,” Ins. Coverage L. Bull. (Mar. 2004) at 1, 5. Based on this misplaced reliance on Midland, the Viking Pump court went seriously astray when it concluded that “New York courts have generally found that a plaintiff who proves that she suffered compensable damage as a result of asbestos exposure is injured during all periods of material exposure and therefore that any policy is triggered if it was in existence when the exposure occurred.” 2009 WL 3297559, at *22.

One Possible Lesson

A further possible reason for the Viking Pump court's trigger presumption may have been the court's desire to resolve the allocation issue “at a high level” and in the abstract, without the factual record necessary to support an appropriate injury-in-fact determination under New York law. See Excess Insurers' Joint Brief (D.I. 2149, Apr. 29, 2009), at 21, n.11 (describing court's order that its allocation determination “would only include a high level determination of the proper allocation method (i.e., pro rata vs. all sums),” and not consider closely linked issues such as the policyholder's obligation for uninsured or underinsured periods). If so, one lesson of Viking Pump may be that an attempt to resolve the allocation method prior to determining trigger of coverage creates a serious risk that a trigger presumption, rather that the required injury-in-fact inquiry, will serve as the basis for the allocation.

Viking Pump's Departure From New York Number-of-Occurrences Doctrine

As explained below, the Viking Pump court's “all sums” holding relies heavily on non-cumulation and prior insurance clauses, which provided for reduction of policy limits due to payments from other policies based upon the same occurrence or loss. The court emphasized the conflict between these clauses and pro rata allocation method, because the court believed that each claim against the policyholder reflects a separate and single occurrence. The court based its view upon an examination of neither the facts of the underlying claims nor the relevant New York law, but rather on the presumption that each claim constituted a separate occurrence. See 2009 WL 3297559, at *22 (“New York law can be seen as treating each asbestos plaintiff's exposure and injury as a single “occurrence” (that triggers all the multiple policies in place during the period of asbestos exposure)); Id. at *30 (“[I]n the case of an asbestos plaintiff with Multi-Period Exposure, the plaintiff's injury is treated as indivisible and resulting from one occurrence.”).

The Viking Pump court cited to Appalachian Insurance Co. v. General Electric Co. , 863 N.E.2d 994 (N.Y. 2007), in support of its multiple occurrence presumption. However GE cautioned that “[e]ach mass tort scenario must be examined separately,” and explained that determination of the number of occurrences in the asbestos bodily injury context required an analysis of the temporal and spatial proximity between the causative events and the loss, and the existence of an unbroken causal continuum spanning from those events to the resulting loss. Id. at 999. If the Viking Pump court embarked upon such an inquiry, it did not recount it in its decision.

The court's number-of-occurrences presumption, like its trigger presumption, illustrates the potential dangers of attempting to adjudicate allocation issues prior to establishing a clear factual record regarding the exact nature of exposures and circumstances giving rise to the underlying claims. Had the court applied the analysis required by GE, it might well have found that each claimant's injuries arose from separate occurrences ' and, therefore, the non-cumulation and prior insurance clauses did not come into play. Or, it might have concluded that all of the claims against the policyholder constituted a single occurrence, in which case the allocation dispute would have been rendered largely moot by application of the non-cumulation and prior insurance clauses.

Viking Pump's Departure from New York Allocation Doctrine

Based on these questionable presumptions as to the trigger of coverage and number of occurrences, the Chancery Court turned to the issue of the proper allocation method. Before discussing the factors it believed rendered the Con Ed and other New York pro rata allocation authorities distinguishable, the court criticized both, accusing New York federal courts of showing a “surprising” disregard for Erie on the allocation issue, and suggesting that the Con Ed decision itself is too “abbreviated” and “terse” to guide subsequent allocation decisions. 2009 WL 3297559, at *25-*26. Throughout this preliminary discussion of New York allocation authorities, the court relied on its exposure trigger presumption to vitiate the significance of the limiting phrase “during the policy period” applied to concept of “injury” in the policies at issue: “If, as appears to be the case, a policy would be deemed triggered under New York law simply because there was material asbestos exposure during the policy period, the use of the 'during the policy period' language does not itself answer the key [allocation] question ' . ” Id. at *26. Despite the express reliance of the Con Ed court on the limiting language “during the policy period” in concluding that pro rata allocation would apply in the asbestos bodily injury claims before it, the Viking Pump court nonetheless concluded that under New York law, “[a]t least as applied to asbestos exposure cases, the words 'during the policy period' do not shed much light on what method of allocation is intended.” Id.

Though critical of Con Ed as “extremely abbreviated,” the Viking Pump court omitted mention of the key rationale for pro rata allocation, discussed at length in that case. In Con Ed, because there was no evidence of specific damage during any particular policy period, the damage at issue was found to have occurred continuously during the 50-year coverage period. The policyholder argued that this meant every policy in effect during that period was triggered and available to respond on an “all sums” basis. 746 N.Y.S.2d at 630. The court found this contradiction in the policyholder's position untenable, because “collecting all the indemnity from a particular policy,” as permitted under an “all sums” allocation method, “presupposes [the] ability to pin an accident to a particular policy period.” Id. In contrast, “[p]roration of liability among the insurers acknowledges the fact that there is uncertainty as to what actually transpired during the particular policy period.” Id. The court further found, “[m]ost fundamentally, [that] the policies provide indemnification for liability incurred ' during the policy period, not outside that policy period.” Id. (emphasis added). The failure of the Viking Pump court to address this core rationale of Con Ed is inexplicable.

Viking Pump's Interpretation Of Non-Cumulation and Prior Insurance Clauses

The prime factor on which Viking Pump distinguished Con Ed on the issue of allocation was the presence of non-cumulation and prior insurance clauses in the policies at issue. The language of these clauses is unremarkable and is found in many general liability policies; it calls for a reduction in limits based upon payments for related losses under prior policies:

If the same occurrence gives rise to personal injury, property damage or advertising injury or damage which occurs partly before and partly within any annual period of this policy, the each occurrence limit and the applicable aggregate limit or limits of this policy shall be reduced by the amount of each payment made by [Liberty Mutual] with respect to such occurrence, either under a previous policy or policies of which this is a replacement, or under this policy with respect to previous annual periods thereof.

* * *

It is agreed that if any loss covered hereunder is also covered in whole or in part under any other excess Policy issued to the Insured prior to the inception date hereof, the limit of liability hereon stated in the Items 5 and 6 of the Declarations shall be reduced by any amounts due to the Insured on account of such loss under such prior insurance.

2009 WL 3297559, at *28-*29.

It is apparent that these clauses have nothing to do with the trigger and coverage issues underlying the selection of an allocation method. As the Viking Pump court explicitly recognized, the purpose of such clauses is to limit the total amount that a policyholder can collect for all losses due to a single occurrence to the single highest per-occurrence limit of all policies covering any part of such loss ' i.e., to prevent “stacking” of policy limits. 2009 WL 3297559, at *29. However, because the clauses referred to the possibility that a single occurrence might give rise to losses occurring across multiple policy periods, the court concluded that these provisions, and not the language limiting coverage to injuries “during the policy period,” should govern the question of which losses each policy should cover. “By their very presence,” the court reasoned, these clauses “suggest that the words 'during the policy period,' do not have the drastic effect the Excess Insurers contend for.” Id.

The flaw in the court's reasoning is that application of such non-cumulation and prior insurance provisions are entirely consistent with pro rata allocation. Nothing stops insurer and policyholder from agreeing both that: 1) a policy will cover only that injury which occurs during the policy period; and 2) such coverage is subject to the further limitation that payments made in respect of related injuries arising from the same occurrence under previous policies will be deducted from the per-occurrence limits applicable to such coverage.

The Viking Pump court acknowledged that the non-cumulation and prior insurance clauses could both be applied as written, but rejected such an approach because it would “unreasonably reduce coverage,” amount to a “windfall” and “'double credit' for insurers,” and achieve what the court felt were “bizarre and inequitable results.” 2009 WL 3297559, *30-*31 & n.169. In declining to give effect to both the clauses and the express policy requirement that covered injury must occur “during the policy period,” the Viking Pump court made no mention of New York law holding that “[r]elying on social policy to justify imputing an expectation of complete coverage to the insured is legally ' unsupportable.” American Home Prods. Corp. v. Liberty Mut. Ins. Co. , 565 F. Supp. 1485, 1511 (S.D.N.Y. 1983), aff'd, 748 F.2d 760 (2d Cir. 1984).

One reason for Viking Pump's decision to uphold the non-cumulation and prior insurance clauses over the interpretation of the controlling “during the policy period” limitation of the Con Ed court was the decision of the Delaware Supreme Court in Hercules Inc. v. AIU Ins. Co. , 784 A.2d 481 (Del. 2001), which relied on similar clauses to reach an all sums allocation result. The Viking Pump court's heavy reliance on the Hercules decision raises further questions as to whether the court was faithfully applying New York law.

The Viking Pump court sought to buttress its all-sums holding with evidence of the parties' conduct with respect to allocation, but in so doing the court relied on two questionable propositions. First, the court proposed that the alleged waiver of the policyholder's primary insurer should bind its excess insurers, because the excess insurers were participating in a single integrated insurance program which should not admit differences in position among insurers. The court did not discuss authority in New York and elsewhere suggesting that one insurer cannot bind another in this manner. See United States Fidelity & Guaranty Co. v. Treadwell Corp. , 58 F. Supp. 2d 77, 93-99 (S.D.N.Y. 1999) (holding that non-settling insurers were not bound by an allocation agreed to between insured and settling insurers); Allmerica Fin. Corp. v. Certain Underwriters at Lloyd's, London , 871 N.E.2d 418, 429 (Mass. 2007) (holding that “absent an explicit contractual commitment to do so, an insurer is not bound by the settlement another insurer makes for the same claim, even if the language of the nonsettling policy follows the form of the settling policy”).

The Viking Pump court also reached the remarkable conclusion that the primary insurer's allocation of payments pro rata across multiple policy periods did not indicate a practice of pro rata allocation, but rather one of “all sums” allocation, because of the opportunity for insurers in an “all sums” regime to bring contribution actions to recover overpayments from other insurers and, ultimately, to allocate losses to particular policy periods. 2009 WL 3297559, *33 n.184. The court's analysis failed to consider that even an insurer recovering excess payments in contribution would have no reason to allocate its own share among the multiple policy periods, unless it viewed that share as allocable pro-rata.

In brief, the Viking Pump decision appears to pump novel interpretations into virtually every major doctrinal area involved in determining coverage for continuing injuries under New York law. How these novel interpretations will be received on the shores of New York remains to be seen.