United States: The Myth of "Accident-Based" Policies

Folklore sometimes develops regarding a particular insurance issue, to the point where the actual policy wording is ignored or forgotten. One such area is the common discussion about the supposed difference between "accident based" and "occurrence based" coverage. According to this fairy tale, "Once upon a time, policies covered only events called 'accidents'...." The story goes on to reveal how subsequently the insurance industry decreed that coverage would expand to events called "occurrences," and there was much rejoicing among insureds. This story, though, is a fallacy which completely ignores both the terms of the pre-1966 policy wording, and the entire context of the changes made in the 1966 ISO policy. In short, it is simply incorrect to say that the earlier policies covered "accidents" and the later policies covered "occurrences." This story (if believed) deprives insureds of the broad coverage to which they are entitled under the pre-1966 wording. By treating the word "accident" as a noun, carriers argue the pre-1966 policy applies only to "boom-type events" and not gradual and progressive harm, a severe (and non-existent) restriction on coverage.

In standard form CGL policies issued before 1966, the insuring agreements read as follows (there were separate clauses for property damage and bodily injury; here we just quote the property damage clause):

The bolded phrase "caused by accident" is in contrast to that used in the insuring agreement that became standard in 1966:

When analyzing (and arguing) coverage, virtually everyone, from judges to carrier counsel to policyholder counsel have focused on the change of the word "accident" to the word "occurrence" and treat the change as involving a simple substitution of one noun for another. Certainly, the post 1966 policy uses " occurrence" as a noun, referring to "an occurrence." The pre-1966 wording, however, does not say "caused by an accident," it simply says "caused by accident." Incredibly, many policyholder counsel overlook this point and indeed even feel compelled to add the word "an" (in brackets) when referring to the clause. In fact, "accident" is not meant as a noun to describe a type of event. It serves as an adverb, which qualifies the manner in which damages were caused.

The entire clause "caused by accident" refers to the manner in which the property damage was caused, and indicates it must be caused accidentally. In other words, "not on purpose" or "not intentionally." So deliberate actions which result in unintended consequences, as well as gradual processes, fall within the insuring agreement, despite carrier arguments that seek to impose a requirement there be a sudden "boom" type event or an unintentional act. Clearly the drafters understood the function of the "caused by accident" wording – when that phrase was abandoned, they added a clause (in the definition of "occurrence") to replace it, requiring that damage be "neither intended nor expected."

Early cases interpreted the pre-1966 clause correctly – that is, "by accident" referred to "an unexpected, unforeseen, or undesigned happening or consequence from either a known or an unknown cause." See Geddes & Smith, Inc. v. St. Paul Mercury Indem. Co., 51 Cal. 2d 558, 563-64 (1959). See also Ritchie v. Anchor Cas. Co., 135 Cal. App. 2d 245, 254 (1955) ("the unexpected and unforeseeable result of a voluntary act fulfills the term of the policy . . . "); Meyer v. Pacific Employers Ins. Co., 233 Cal. App. 2d 321, 327 (1965) (damage to underground pipes caused by vibrations from well drillers not intended or expected, and therefore accidental).

For unknown reasons, (but perhaps influenced by a growing familiarity with the newer form requiring "an occurrence" or possibly the definition of "occurrence" as "an accident . . . .") somewhere along the line the pre-1966 forms started to be read as though they said "caused by an accident." No matter the reason why, it is unfortunate that courts' discussions of the change from "by accident" to "an occurrence" ignore this distinction. For example, the court in Lockheed Corp. v. Continental Ins. Co., 134 Cal. App. 4th 187 (2005) stated that "[i]ndeed, prior to 1966 the typical liability policy was written to cover only accidents...." Such broad brush statements should not, of course, discourage insureds from arguing for the proper interpretation of the phrase "caused by accident." Insurance cases are not decided by "general rules" but by wording of the actual policy issued to the insured.

The effect of this misreading is to change the entire meaning of the clause, supposedly requiring "an accident" to trigger coverage. Although the word "accident" is not defined in any of the policy forms, it has been used to argue for restricting coverage to sudden "boom" type events, or acts (rather than consequences) that were unintended. Under such a reading, injury or damage caused by unintended consequences of deliberate acts would not trigger the policy, nor would gradual events.

As time marches on, the real-world effect of this distortion diminishes simply because there are fewer claims involving these old policies. But insureds with long tail risks such as environmental exposures should review their pre-1966 policies very carefully, and make sure that the right coverage arguments are being made when making claims.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.