“Why The ‘Unavailability Exception’ Makes Sense”
Law360 03/29/17 By Jeffrey O. Davis
The issue of allocation in long-tail claims has vexed courts for about as long as insurance coverage has been sought for such claims. The fundamental debate is this: when a policyholder under a standard commercial general liability policy is sued on a claim based on "continuous" injury, that is, an ongoing injury that occurs both within and outside the time period in which the policy was in effect, should the policyholder be fully defended and indemnified by the insurer or does the insurer owe only a "pro rated" portion of defense and indemnity costs. And, if it is the latter, should the rule be different if it can be shown that insurance for the type of claim at issue was not "available" for the uninsured period that would otherwise be allocated to the policyholder.
Prompted by a couple of recent decisions, much proverbial ink has recently been spilled in this publication over the so-called "unavailability exception." Most recently, Aidan McCormack and Brian Seibert weighed in with an article that rejects the unavailability exception. (Why the Unavailability Rule should be Rejected, Law360, March 15, 2017). The primary basis for this rejection lies in the authors' repeatedly stated premise that pro rata allocation is rooted in the language of the policies. With that assertion in tow, they conclude that an exception based on unavailability "calls for a re-drawing of the contracts in direct contradiction of the parties' agreement and principles of equity."
With all due respect to McCormack and Seibert, their fundamental premise fails. Whatever the merits or shortcomings of pro rata allocation as a matter of fairness and public policy, it is definitely not consistent with the language of the policy. In fact, the exact opposite is true.
To support their "driven by contract language" argument, McCormack and Seibert recite a snippet of the policy stating that it "applies" to injury that occurs "during the policy period". This snippet, while dictating when a policy is triggered, has nothing to do with the scope of coverage, which is controlled by other language. Because this is the entire thrust of their argument a closer look is in order.
Here is what the policy actually says (in various permutations though it is always words to this effect): "We [the insurer] will pay all sums that the insured becomes legally obligated to pay as damages because of bodily injury or property damage to which this insurance applies." The policy goes on to say that the policy "applies" to "bodily injury or property damage that occurs during the policy period." So the language on which McCormack and Seibert rely addresses when the policy "applies" — or to use the term of art that has universally arisen among insurance lawyers, when the policy is triggered.
The real issue is, once a policy is triggered, how much coverage and defense is afforded under that policy? The appropriate starting point is the language of the policy, which McCormack and Seibert, somewhat ironically given their thesis, ignore. Preliminarily, the scope of coverage analysis can be broken down between the two principal duties the insurer has: the duty to defend and the duty to indemnify.
With respect to defense, insurers seeking a pro rata result are essentially arguing under a policy that doesn't exist. The policy they want to apply reads as follows:
We will have the right and duty to defend that portion of any suit seeking damages to which this insurance applies.
The policy that actually exists reads:
We will have the right and duty to defend any suit seeking damages to which this insurance applies.
Words matter — or at least they're supposed to — and obviously there is a difference between an obligation to defend "any suit" and an obligation to defend some "portion" of a suit. "Any suit" can only mean the whole suit, a conclusion that is only bolstered by the near universally accepted principle that if any part of a complaint triggers the duty to defend the insurer owes a duty to defend the case in its entirety.
The argument for pro rata indemnity employs logic that is every bit as tortured. An insurer seeking a pro rata result is also attempting to limit its liability under a nonexistent policy, one that reads:
We will pay that portion of those sums which the insured shall become legally required to pay as damages which is allocable to injury that occurs during the policy period.
The policy that actually exists reads:
We will pay on behalf of the insured all sums which the insured shall become legally obligated to pay as damages because of bodily injury or property damage to which this policy applies [i.e., to injury that occurs during the policy period] ...
This language, too, leaves no doubt about an insurer's obligation to pay all of the insured's legal damages when a policy is triggered by injury that occurs, even if only in part, during the policy period. Under joint and several liability principles, the insured is "legally required" to pay all damages assessed for bodily injury or property damage even if only some of it occurred during a period for which the insured had responsibility, a point succinctly made long ago in ACandS Inc. v. Aetna Cas. and Surety Co., 764 F.2d 968, 974 (3d Cir 1985):
The policies require the insurers to pay all sums which ACandS becomes "legally obligated to pay" because of bodily injury during the policy period. It is uncontested that under principles of tort law ACandS may be held fully liable for a personal injury plaintiff's damages caused in part by ACandS' asbestos during a particular period, even though plaintiff's damages may also have been caused, in part, at other times. It follows that if a plaintiff's damages are caused in part during an insured period, it is irrelevant to ACandS' legal obligations and, therefore, to the insurer's liability that they were also caused, in part, during another period.
In other words, an insured does not get to pro-rate its legal liability for damages if it is responsible for only a portion of the time in which on-going injury occurred. An insurer who has agreed to pay "all sums" for which the insured is legally responsible should have no greater right to pro-rate its defense or indemnity obligation. The language of the insurance policy precisely tracks the same joint and several liability concept to which insureds are subject.
In fact, courts adopting pro rata allocation have generally not done so on the basis of policy language. Prudential Lines Inc. v. Am. Steamship Owens Mut. Prot. & Indem., 158 F.3d 65, 85 (2d Cir. 1998) ("The courts that have endorsed allocation when the losses paid have generally been motivated by considerations of equity and policy, rather than contract wording"). Instead, they more typically cite any number of public policy reasons including alleged "fairness", ease of administration and spreading of losses. These policy reasons seem almost oblivious to the overriding public policy that contracts should be enforced as written, particularly as against the drafter.
From a policyholder's standpoint, it is nothing short of outrageous that insurers have been able to convince some courts — not a "majority" as McCormick and Seibel contend but exactly half of those considering the issue, by my research — to rewrite policies drafted by insurers by adopting pro rata, in direct contravention of contra proferentem. That is really another article.
Getting back to the question on the table, should the pro rata period include periods where coverage is not available for the risk at issue? If you want to adopt the majority view, answer this question in the negative. Indeed, just this month, in a 250 page opinion adopting the unavailability exception, the Connecticut Appellate Court noted that a majority of courts adopting pro rata allocation had also adopted the unavailability exception. The court also observed that, when considered together with those jurisdictions applying joint and several liability, the vast majority of jurisdictions throughout the country do not hold a policyholder responsible for losses that occur during periods when insurance is not available. R.T. Vanderbilt Co. Inc. v. Hartford Accident & Indem. Co., 171 Conn. App. 61 (March 7, 2017).
This result is hardly surprising. To justify a pro rata result, perhaps no policy reason has been given more credence than the notion that insureds seeking full defense and indemnity are seeking coverage for periods in which they "chose" not to have insurance and to award full defense and indemnity thus provides them a "windfall". It’s a flawed argument because, as shown above, it elevates vague notions of equity over the contract-based verity that insureds paid premiums in exchange for insurers' agreements to fully defend and indemnify claims stemming from injuries that trigger the policy. Moreover, the insured who chooses not to purchase insurance for certain time periods has less coverage in terms of limits, and fewer periods in which insurance might be triggered — so there, in fact, is a consequence, and potentially a severe one, for an insured who for some periods chose not to be insured.
But in any event, the "windfall of their own making" argument leads directly to the rationale underlying the unavailability exception: those courts inclined to elevate a public policy-based approach over insurance policy language nonetheless recognize that a judicially created pro rata scheme revolving around the insured's choice not to purchase insurance does not hold together where there is no choice. Security Insurance Co. v. Lumbermens Mut. Cas. Co., 1999 WL 545745, *8 (July 12, 1999) ("The element of choice, and in turn, the conscious decision of an insured to assume the risks of being self-insured is lost if the insured cannot realistically acquire the particular coverage desired.”).
Seen in this light, the unavailability exception can be viewed as a logical limitation on courts' judicially-created, equitably driven pro rata scheme of allowing only a portion of coverage where injury includes periods where there is no insurance. One way of looking at it is this: but for the insured's decision not to purchase available insurance, a triggered carrier might have had contribution rights against that insurance, which would have had the effect of limiting the targeted carrier's liability to a pro rata obligation — even under the proper interpretation of the policies described above. An insured's decision not to purchase insurance puts that insured in the position of stepping into the shoes of another carrier that could have shouldered some of the risk through contribution. That rationale (which again from a policyholder's view is still divorced from policy language that should put the entire loss on the triggered insurer) makes sense only for periods where insurance was available for the risk.
In short, if a court insists on deviating from the “all sums” outcome that should be required by the policy language, it should assess pro rata allocation on the basis of equitable concerns. See, e.g., R.T. Vanderbilt, 171 Conn. App. 61 (“Ultimately, we believe that the allocation system that we have adopted … — pro rata time-on-the-risk, employing a continuous trigger and an unavailability rule — distributes the burdens equitably among all parties involved and maximizes the resources available to respond to claims while minimizing administrative hassles and transaction costs.”). In that respect, the court in Keyspan Gas E. Corp. v. Munich Reins. Am. Inc., 143 A.D.3d 86 (1st Dep’t 2016), got it wrong not once, but twice. The court first rewrote the policies to provide for pro rata, then refused to consider equity in rejecting the availability rule. While what is equitable may be in the eye of the beholder, a reasonable conclusion — and one that R.T. Vanderbilt and most pro rata courts have adopted — is that any period where the insured is going to be forced to take on a share of an insured loss should at least be limited to time periods when such insurance was available.