The New York Court of Appeals recently struck a blow to policyholders by relieving an insurer of its obligation to indemnify for that part of a long-term harm that occurred when applicable insurance did not exist in the marketplace. See KeySpan Gas East Corp. v. Munich Reinsurance America Inc. et al., No. APL-2016-00236. As is often the case when a loss occurs over multiple, successive policy periods, gaps in available insurance coverage may appear. Gaps in coverage may exist for a number of reasons, including insurer insolvency, lost policies, policy exclusions or occasionally a policyholder’s failure to purchase insurance. Those gaps present problems when liability for long-term harm is allocated pro rata among triggered insurance policies. In such cases, courts are usually faced with two basic options, depending on the law of the jurisdiction: A court may either shift liability for those gaps in coverage to other insurers of the risk, or it may put policyholders on the hook for the gaps. In Keyspan, New York’s highest court opted for the second option, requiring the insured to shoulder the pro rata share of all those periods in which applicable insurance did not exist in the marketplace.

Keyspan’s basic facts are all too familiar. The insured, Keyspan, is statutorily liable for cleaning up contamination that resulted from years of operations beginning in the late 1800s. Keyspan sought coverage and asserted that any pro rata allocation of cleanup costs should not include periods where insurance covering pollution did not exist in the market place (whether due to a pollution exclusion or the fact that general liability coverage was unavailable to public utilities prior to 1925, as Keyspan’s expert testified).

One of Keyspan’s insurers, a Chubb company (Century Indemnity), challenged Keyspan’s position, asserting that insurance policies it issued in the 1950s and 60s obligate it to pay only for damages attributable to occurrences during those policy periods. Chubb argued the insured should have to pay for the coverage gap due to the unavailability of insurance coverage. The New York Court of Appeals agreed, stating that, in situations where cleanup costs must be allocated pro rata, the “during the policy period” proviso in the policies was inconsistent with excluding from the allocation years where insurance was unavailable. The Court essentially reasoned that the “during the policy period” language controlled and relieved insurers from having to pay damages attributable to periods when the insurer’s policies were not in force.

There are a number of takeaways from Keyspan worth noting.

First, and perhaps most importantly, Keyspan does not address the separate and distinct duty to defend. The duty to defend is not mentioned once in the entire opinion (or in the decisions below). This should come as no surprise because the question before the Court of Appeals was who had to pay damages in the form of cleanup costs. Neither defense costs nor the duty to defend were at issue. In fact, the Court never once mentioned its longstanding, oft-cited seminal decision addressing the duty to defend in long-tail cases: Continental Casualty Co. v. Rapid-American Corp., 609 N.E.2d 506, 514 (N.Y. 1993). Accordingly, Keyspan has no impact on the New York Court of Appeals’ several decisions stating that once the duty to defend is triggered by at least one potentially covered claim, the insurer owes a duty to defend the entire claim, including non-covered allegations, without foisting part of the defense onto its insured.  See, e.g., id.; Fieldston Prop. Owners Ass’n, Inc. v. Hermitage Ins. Co., 945 N.E.2d 1013, 1018 (N.Y. 2011).

Second, the Court of Appeals took care to note that the “unavailability” issue was irrelevant in cases where pro rata allocation is inappropriate, citing to its 2016 decision in Matter of Viking Pump, Inc., 27 NY.3d 244, 255 (2016). Pursuant to Viking Pump, if an insurance policy contains a non-cumulation or prior insurance or similar clause, the “all sums” method applies to the indemnity obligation. In such instances, the policy with the non-cumulation or similar clause and the insurance tower above it must pay all damages up to policy limits, with no proration to the insured for uninsured periods.

Third, Keyspan is at odds with the federal Second Circuit Court of Appeals’ Stonewall decision and the decisions of courts in several other “pro rata” jurisdictions that have refused to shift to the policyholder gaps in coverage due to the unavailability of insurance.  Compare Keyspan, No. APL-2016-00236, with Stonewall Ins. Co. v. Asbestos Claims Mgmt. Corp., 73 F.3d 1178 (2d Cir. 1995); Owens-Illinois, Inc. v. United Ins. Co., 650 A.2d 974 (N.J. 1994). The Court defended its departure from other courts’ rulings by asserting that, in New York, the language of the policy is paramount, whereas other courts, it asserted, focused instead on the public policy of maximizing insurance coverage.

This justification illustrates the tension between touting the policy language as paramount while at the same time adopting the legal fiction that an indivisible long-term loss can be fairly allocated based on time on the risk. By way of illustration, isn’t it true that in most cases environmental damage continuing after 1986 is at least in part attributable to discharges from preceding decades? For example, a migrating contaminated groundwater plume may cause continuing damage, but that continuing damage is undoubtedly also due to the fact that the groundwater is contaminated in the first place from discharges long ago. How is it improper to ask an insurer that issued coverage while discharges were occurring to pay for a part of the damage that results from the continued presence of pollutants in periods where coverage is no longer available in the marketplace? What about the proviso that the insurance policy provides coverage for all sums the insured must pay as damages due to “continuous or repeated exposure to substantially the same general conditions”—i.e., the continued exposure of soil and groundwater to preexisting contamination?

For that matter, what if most of the harm at issue occurred during early operations when discharges were often unregulated, more frequent and more toxic? It may still be impossible to establish the precise quantity and quality of harm during every policy period, but generally speaking, why should insurers on the risk during those earlier years be allowed to pass off a disproportionate share of an “indivisible” liability to the insured for years when no coverage was available? Isn’t this especially odd when, often, the cause of the harm during that post-1986 period—continuing migration of groundwater contaminated by discharges from long ago—begins to attenuate in the later years? It is difficult to find justification for these incongruities in the policy language.

Fourth, one might expect the Keyspan ruling to raise a whole host of new problems in allocation cases where New York law applies. Perhaps some policyholders may be more likely to take on the fact burden of proving that a loss is divisible and can be attributed to certain policy periods on a principled basis—certainly not something that will promote judicial efficiency or economy. At the very least, policyholders will be sure to assert, rightly so, that the coverage block for purposes of allocation should end as soon as the potential liability for environmental harm became known, if not earlier. Indeed, how can an allocation fairly be accomplished with an open-ended coverage block? Will coverage end when a slurry wall is installed? When regulatory standards are met and there is no longer a legal obligation to remediate? When a remedial investigation/feasibility study is complete and estimated cleanup costs are more or less known? When “additional” or “new” damage is de minimis?

None of these latter end-dates makes sense, especially in the context of a pro rata allocation that rests on the legal fiction that an equal amount of harm resulted from every occurrence in every policy period from inception to the present. Needless to say, these issues may make pro rata allocation in New York more difficult, not less, and will likely require more guidance from New York courts. It’ll be interesting.