Commentary and analysis on the
law of insurance coverage

When an Absolute Pollution Exclusion is Not So Absolute

Smoke Stacks

Author: Carl A. Salisbury

Beginning in the early 1980s, an explosion of litigation across the country occurred, in which corporate policyholders sought insurance coverage for billions of dollars in defense and indemnity for environmental liabilities.  The passage by Congress in 1980 of the Comprehensive Environmental Response, Compensation and Liability Act, or CERCLA (also known as “the Superfund law”) was the catalyst for an ensuing cottage industry in coverage lawsuits.  Corporate policyholders large and small received “Potentially Responsible Party” letters from the U.S. Environmental Protection Agency and from state departments of environmental protection, holding them strictly liable for the creation, transportation, and disposal of hazardous wastes at thousands of private and municipal landfills across the country.  A small handful of visionary lawyers (a couple of whom I had the good fortune to work with in the late 1980s and early 1990s) believed — correctly, as it happens — that the standard Comprehensive General Liability insurance policy that essentially every company in America had been purchasing since the 1940s should provide coverage for this new kind of statutory liability.

Very soon, the dockets of state and federal courts across the country began to fill up with complex monster lawsuits.  Each one of these suits might easily involve scores of insurance companies as defendants and many tens or even hundreds of millions of dollars in potential insurance coverage.  As these cases made their way slowly through the system, judges began providing interpretations of the key words and phrases of the standard CGL policy: “occurrence,” “accident,” “as damages,” “neither expected nor intended from the standpoint of the insured,” “damage during the policy period.”  Starting in 1970, the CGL policy included a standard, limited, pollution exclusion.  It excluded coverage for the discharge or release of pollutants unless the discharge or release was “sudden and accidental.”  That, of course, became another key phrase that courts across the country had to interpret to determine if the CGL policy covered a policyholder’s CERCLA liabilities.  Not surprisingly, perhaps, a patchwork of case law developed in which some courts interpreted the key coverage provisions favorably to policyholders and others interpreted them favorably to insurers.  Soon, the question whether a policyholder might or might not recover insurance proceeds turned on which party won the race to the courthouse, with policyholders running to pro-insured jurisdictions to file their coverage cases and insurance companies running to pro-carrier states to file pre-emptive suits.


“If an insured throws a knife and injures someone, the insurance carrier will now deny coverage on the ground that the harm arose from ‘air pollution.'”


Beginning as early as 1982, but occurring in earnest by about 1984, the insurance industry started to include what it called the “Absolute Pollution Exclusion” in the standard CGL policy.  This provision eliminated the “sudden and accidental” exception that had been written into the old exclusion.  The new exclusion precluded coverage for any liability arising from or relating to bodily injury or property damage from the release of “pollutants,” which the provision defined very broadly.  Soon enough, courts began interpreting the Absolute Pollution Exclusion and applying it to an ever-expanding class of claims involving harm caused by the use or release of essentially any chemical or substance.  The late Eugene Anderson (who became known as the “Dean of the Policyholder Bar” and was a partner of mine at the time) once quipped that if an insured throws a knife and injures someone, the insurance carrier will now deny coverage on the ground that the harm arose from “air pollution.”

By about the late 1980s, attorneys who practiced insurance law predicted that the Absolute Pollution Exclusion would bring an end to all environmental coverage litigation in just a few short years.  In fact, it took nearly another decade and a half for the pace of filing environmental coverage litigation to slow to a comparable crawl.  As time has passed, though, policyholders have had success in some states challenging insurers on claim denials under the Absolute Pollution Exclusion that, once upon a time, were routinely upheld by the courts.

In my home state of New Jersey, for instance, our Supreme Court was one of the first in the nation to narrow the ever-expanding scope of the Absolute Pollution Exclusion in a case called Nav-Its, Inc. v. Selective Ins. Co., 183 N.J. 110 (2005).  The court relied, among other things, on a number of representations the insurance industry made in testimony before the New Jersey Insurance Commissioner at the time the industry sought approval of the exclusion for use in CGL policies sold in the state.  Those representations strongly suggested that the exclusion would never be used to try to exclude coverage for pollution claims other than traditional industrial waste handling and disposal.  Accordingly, the court limited the application of the Absolute Exclusion to hazards that were historically considered “environmental catastrophes related to intentional industrial pollution.”  A number of federal courts in New Jersey thereafter rejected the application of the exclusion to fact patterns that didn’t fit the “traditional environmental pollution” template.

The most recent example of a New Jersey case to apply the exclusion narrowly came in an unpublished trial court decision in Westwood v. Great American E&S Ins. Co., et al., MID-L-2320-13 (2013). (Full disclosure: I represented the policyholder in that case.)  Westwood manufactures hoses, filters, and other peripherals for the oil and gas industry. Its products are sold and distributed in the U.S., Mexico, and Canada.  A Westwood filter was installed on a home heating oil tank in Canada.  It appears that the installer did not properly fit the gasket to the device, leaving a gap through which thousands of gallons of oil were released onto the ground during several refills of the tank.  The resulting pollution caused extensive environmental damage for which the homeowner sued Westwood, among others.  Westwood argued, and the trial court agreed, that the claim at issue was not an “environmental catastrophe related to industrial pollution” but was, instead, in the nature of a defective-product suit.  Therefore, the pollution exclusion did not apply.


Injury or damage must have a legal cause, even if there is no such thing as a sole cause.


In a case decided last month, the Washington Supreme Court employed a causation analysis to reject a coverage denial in what otherwise would have been a straightforward application of the Absolute Pollution Exclusion.  The case is Xia v. ProBuilders Specialty Ins. Co., No. 92436-8 (Apr. 27, 2017).  (Get a copy here.)  This is how it happened.

The plaintiff in the case moved into a new home in May 2006.  Soon thereafter, she became ill.  A service contractor discovered that the builder had incorrectly installed a vent on the hot water heater, a defect that released carbon monoxide into the home.  She sued the builder, which put its liability carrier, ProBuilders Specialty Insurance Company, on notice.  ProBuilders declined coverage on the basis of a broad Absolute Pollution Exclusion in its policy.  The plaintiff eventually settled with the builder for $2 million and took an assignment of the builder’s insurance claim against ProBuilders in exchange for an agreement that she would seek to recover the settlement only through the proceeds, if any, from the ProBuilders policy.  This is a pretty common way for insurance coverage claims to reach the courts.

The case made its way through the lower courts and ultimately got accepted for appeal to the Washington Supreme Court.  The court framed the questions on appeal in this way: “Does the specific pollution exclusion include the carbon monoxide in this case, and does the exclusion preclude coverage when the cause of the loss was a covered occurrence under a different provision.”  The answer to those questions was, according to the court, “Yes,” and “No,” respectively.

The court reviewed the development of earlier case law on the absolute exclusion and found that Washington courts had applied the exclusion “to preclude coverage in the case of traditional environmental harms or where the pollutant acted as a pollutant.”  In other words, the interpretation of the absolute exclusion in Washington had followed a course similar to that employed by courts in New Jersey (and a few others), and had reached a similar conclusion.  Under this analysis, the first question is whether the occurrence stems from traditional pollution and whether it involves a “pollutant acting as a pollutant.”  The court concluded that, in fact, the carbon monoxide release did constitute a “traditional environmental harm” in which the pollutant acted as a pollutant.  The court did not end the analysis there, however.  It also considered whether the otherwise excluded occurrence was the “efficient proximate cause” of the claimed loss.  Here, it is worthwhile to take a short diversion.

In a post on December 5, 2016, I discussed a decision by the Florida Supreme Court that explored causation in the context of insurance coverage.  Causation is a central concept in very nearly all of litigation.  Cause and effect, cause and result; these are superficially simple concepts in theory that are vexatious to apply in practice.  Most have heard of the “butterfly effect.”  It is a concept in chaos theory that posits that tiny changes in the initial condition of a system can cascade into larger and larger perturbances.  It inspired a famous and influential scientific lecture by Edward Lorenz entitled Does the Flap of a Butterfly’s Wings in Brazil Set Off a Tornado in Texas?  The point of the butterfly effect is that every event that happens in the world alters the future and is a “cause” of everything that occurs thereafter.  There is no such thing as a “sole cause” of anything.  But lawyers and judges and courts — and the parties who look to the courts to resolve their disputes — can’t live with such uncertainty.  Injury or damage must have a legal cause, even if there is no such thing as a sole cause.  In the law, the legal cause of injury or damage is the one we identify as the “proximate cause.”

Getting back to the Xia decision, then.  The Washington Supreme Court has adopted a theory of causation that further limits the legal cause of an event to the “efficient proximate cause.”  The court described it as follows: “Under Washington law, the rule of efficient proximate cause provides coverage ‘where a covered peril sets in motion a causal chain[,] the last link of which is an uncovered peril.'”  In other words, if the first cause in a chain of events is a covered peril, it doesn’t matter that other concurrent causes might be excluded perils.  The court will look only to the first cause to determine if coverage applies.

Now, the insurance industry has been writing exclusions for many years that contain language to circumvent the concepts of concurrent causation (one or more causes combining to result in an injury) and efficient proximate causation from coverage decisions.  The standard Absolute Pollution Exclusion contains such language.  In fact, the exclusion in Xia contained such language: “The Exclusion applies regardless of the cause of the pollution and whether any other cause of said property damage, bodily injury, or personal injury acted jointly, concurrently, or in any sequence with said pollutants or pollution.”

A number of states, including the State of Washington, have rejected anti-concurrent-causation provisions such as the one that appeared in the exclusion in the Xia case.  For this reason, the Washington Supreme Court voided the anti-concurrent-causation clause of the Absolute Pollution Exclusion:  “The exclusion cannot eviscerate a covered occurrence merely because an uncovered peril appeared later in the causal chain.  The efficient proximate cause rule exists to avoid just such a result, ensuring that an insurance policy offered for indemnity of a covered peril will provide coverage when a loss is proximately caused by that covered peril.”

From there, it was a relatively short leap to the conclusion that the event in this case should be covered.  “[I]t is clear that a polluting occurrence happened,” said the court, “when the hot water heater spewed forth toxic levels of carbon monoxide into Xia’s home.  However, by applying the efficient proximate cause rule, it becomes equally clear that the ProBuilders policy provided coverage for the loss.  The polluting occurrence here happened only after an initial covered occurrence, which was the negligent installation of a hot water heater that typically does not pollute when used as intended.”


Courts are increasingly willing to apply coverage rules that significantly restrict the application of the Absolute Pollution Exclusion.


There are problems for both policyholders and insurers inherent in the application of an “efficient proximate cause” rule to insurance coverage.  By definition, the rule focuses solely on the most remote cause in a chain of occurrences and ignores all of the more immediate causes, a choice that seems arbitrary.  In states that do not preclude the use of anti-concurrent-causation clauses, the efficient proximate cause rule may operate to preclude coverage unfairly to a policyholder when there are concurrent causes of a loss, at least one of which is covered, and the carrier chose not to employ an anti-concurrent-causation exclusion for that loss.  An efficient proximate cause can just as easily be excluded as it can be covered.  In a state such as Washington, which has outlawed anti-concurrent-causation clauses as a matter of public policy (California, North Dakota, and West Virginia have also done so), policyholders may find that they have no coverage when a long chain of causes fall within the coverage of a policy except for the very first one.  It just happens that the insurer’s ox was the one that got gored in the Xia case.

The broader take-away from the Xia case, though, is that courts are increasingly willing to apply coverage rules that significantly restrict the application of the Absolute Pollution Exclusion.  When it was first introduced, that exclusion appeared to mark the end of coverage for any and all losses that could remotely be viewed to involve a release of essentially any substance whatsoever into the environment.  Indeed, for a period of years after its introduction, insurance carriers were getting away with applying that exclusion in highly creative ways to deny claims that could never have been in the contemplation of its drafters, that was contrary to what those drafters routinely represented to insurance regulators at the time of adoption, and that was thwarting policyholders’ reasonable expectations of coverage.  It now appears that the pendulum might be in full swing in the other direction.

If you’d like us to notify you whenever there is a new post on The Insurance Recovery Journal, subscribing will make that happen. There’s no charge and we won’t use your email address for any purpose other than to send the notifications.

[simpleSubscribeForm]