Skip to Content
Liability Insurance

Variations on a Theme: When the Cause Theory Determines the Number of Occurrences

Steven Rawls | May 1, 2008

On This Page
Court building

To determine the number of occurrences under a general liability insurance policy, jurisdictions typically follow either the cause theory or the effect theory.

As the name implies, under the cause theory, to determine the number of occurrences, courts look to the cause or causes of the damage. See e.g., Donegal Mut. Ins. Co. v. Baumhammers, 938 A.2d 286 (Pa. 2007). Conversely, courts using the effect theory calculate "the number of occurrences by looking to the effect of the accident or ... how many individual claims or injuries resulted therefrom." Id. at 293.

Most jurisdictions follow the cause theory. A minority of jurisdictions are said to follow the effect theory; however, the cases addressing the effect theory merely reject it. Courts applying the cause theory focus on several different considerations including the policy language, different facts, and public policy considerations. Different views of these considerations result in inconsistency that affects uniformity and predictability of result.

Applying the Cause Theory

In Koikos v. Travelers Ins. Co., 849 So. 2d 263 (Fla. 2003), the Supreme Court of Florida addressed the certified question of whether the facts "constitute one occurrence, or multiple occurrences as that term is defined in the policy of liability insurance" issued to the insured, Koikos. Two shooting victims sued Koikos, the owner of a restaurant, for negligent security when, during a fight on the premises, they were shot, and three others were injured by one person firing "two separate-but nearly concurrent-rounds." Koikos at 265.

Koikos argued that "the force that caused the injuries was the gunshots and, therefore, each shot injuring a victim was a separate occurrence." Id. at 265. Conversely, Travelers argued that "the injuries resulted from Koikos's alleged negligence and that negligence constituted a single 'occurrence' under the terms of the policy" subject to the $500,000 per occurrence limit. Id.

The Travelers policy defined "occurrence" as "an accident, including continuous or repeated exposure to substantially the same general harmful conditions." Koikos at 266. When "accident" is undefined, it is ambiguous and, under Florida law, encompasses "not only 'accidental events,' but also injuries or damage neither expected nor intended from the standpoint of the insured." Koikos at 267 (citation omitted).

Travelers argued that the occurrence was Koikos's negligence and that the plaintiffs were continually exposed to Koikos's inability to keep the premises safe. The Supreme Court of Florida disagreed and explained that the continuous or repeated exposure language of the policy expands the definition of occurrence to address toxic torts. Favorably citing Lee v. Interstate Fire & Cas. Co., 86 F.3d 101 (7th Cir. 1996), the court stated that the "exposure" was to the bullets, not the negligent failure to provide security. See Koikos at 268.

The court explained that the insured's alleged negligence was not the occurrence. Instead, "it is the act that causes the damage which is neither expected nor intended from the standpoint of the insured that constitutes the 'occurrence.'" Koikos at 271.

Thus, when the insured is sued for negligent failure to provide security, "occurrence" is defined by the immediate injury-producing act and not by the underlying tortious omission. Here, the immediate causes of the injury were the intervening intruder's gunshots. The court rejected Travelers' argument that there should be one occurrence due to the close proximity in time and space of the shots fired and instead concluded that "using the number of shots fired as the basis for the number of occurrences is appropriate because each individual shooting is distinguishable in time and space." Koikos at 272.

The court further supported its holding by citing to the difficulty in determining the facts with respect to the intervals of the shots and that Travelers could have written language making multiple shootings one occurrence. See e.g., SR International Business Ins. Co. v. World Trade Ctr. Prop., LLC, 222 F. Supp. 2d 385, 398 (S.D.N.Y. 2002) (defining "occurrence" as "all losses or damages that are attributable directly or indirectly to one cause or to one series of similar causes ... will be treated as one occurrence irrespective of the period of time or area over which such losses occur").

In Donegal, the Supreme Court of Pennsylvania distinguished Koikos and applied the cause theory to find that substantially similar facts gave rise to one occurrence. Donegal issued a homeowners policy to the parents of the shooter, an adult resident of the household. The shooter left the house, and for 2 hours drove through four towns and killed 6 people. The parents were sued for negligence in failing to take the gun or alert police or mental health professionals about the shooter's dangerous propensities, which constituted an accident under Pennsylvania law.

Using the cause approach, the Supreme Court of Pennsylvania held that:

to determine the number of "occurrences" for which an insurance company is to provide coverage, the more appropriate application of the cause approach is to focus on the act of the insured that gave rise to their liability.

Donegal at 295.

The court favorably cites the rationale of the Supreme Court of Nevada's approach to the cause theory in Washoe Cty. v. Transcontinental Ins. Co., 878 P.2d 306 (Nev. 1994).

In Washoe Cty., the County was sued for negligently licensing a daycare at which an employee sexually abused the children over a 3-year period. The Supreme Court of Nevada reasoned that:

liability was premised on the insured's negligence in performing a duty, which permitted the intervening conduct of those who actively caused the victims harm ... [A]s long as the injuries stemmed from one proximate cause there is a single occurrence ... for purposes of liability.

Washoe Cty. at 308.

In Washoe Cty., each act of molestation arose from the same proximate cause, the County's alleged negligence. Applying that rationale to Donegal, the Pennsylvania Supreme Court held that "coverage is predicated on Parents' inaction, and the resulting injuries stem from that one cause," so that Parents' negligence constituted one occurrence. See Donegal at 295.

Supporting its application of the cause theory, the court explained that:

looking to the underlying negligence of the insured recognizes that the question of the extent of coverage rests upon the contractual obligation of the insurer to the insured. Since the policy was intended to insure Parents for their liabilities, the occurrence should be an event over which Parents had some control.

Donegal at 296.

To take "the covered risks out of the insured's hands [and to determine coverage] by the acts of the unfettered shooter, the insurer would have no basis for setting premiums ... and the insurance contract would be illusory." Id. (citing RLI Ins. Co. v. Simon's Rock Early College, 765 N.E.2d 247, 251 (Mass. 2002)).

Time and Space Considerations

Illinois includes an analysis of time and space when applying the cause theory to determine the number of occurrences. Previously, in Nicor, Inc. v. Associated Elec. & Gas Ins. Svcs. Ltd., 860 N.E.2d 280 (Ill. 2006), the Supreme Court of Illinois approved the following application of the cause theory:

Where each asserted loss is the result of a separate and intervening human act, whether negligent or intentional, or each act increased the insured's exposure to liability, Illinois law will deem each such loss to have arisen from a separate occurrence within the meaning of liability policies containing [per occurrence] language.

Nicor at 294.

In applying Nicor, the Illinois Court of Appeals noted that "the distinction between cause and effect isn't always clear." Addison Ins. Co. v. Fay, 875 N.E.2d 190, 194 (Ill. App. 2007). Fay presented "unfortunate and unique facts" where two young boys died when they got stuck in the clay bottom of an excavation pit owned by the insured during a rainstorm in 34-degree weather. They died of hypothermia due to entrapment in the sand. According to the facts, the first boy got stuck and then his friend entered the pit to try to help him. The boys' parents sued the insured for negligent site security.

In applying the cause theory with a time-and-space analysis, the court created a two-part test courts must consider to determine the number of occurrences:

  • (1) the negligent act or condition that caused the injury, and
  • (2) how the temporal and spatial nature of the incident may have affected any "separate or intervening acts" or "increased the insured's exposure to liability under Nicor.

Fay at 195.

Here, the boys' deaths resulted from one cause: the insured's failure to properly secure entry into the excavation pit. Applying the second prong of its cause theory test, the court explained that "the unfortunate events resulting in the boys' deaths are so closely linked in time and space as to be considered by a reasonable person as one 'occurrence.'" Id. The facts showed that:

the boys entered the property together ... became entrapped in the muddy soil within moments of each other ... deaths were both caused ... by the freezing temperature of the air and water ... their bodies were found lying only inches apart.

Fay at 195.

Wisconsin also considers time and space in applying the cause theory. The cause theory test is applied in Wisconsin as follows; "if a single, uninterrupted cause which results in a number of injuries or separate instances of property damage," it is considered one occurrence. Olsen v. Moore, 202 N.W.2d 236, 240 (Wis. 1972). If, however, that cause is interrupted or replaced by another cause the chain of causation is broken and more than one accident or occurrence has taken place." Id. Additionally, "a cause and result that are closely linked in time and space are likely to be viewed as one occurrence." Plastics Eng'g Co. v. Liberty Mut. Ins. Co., 514 F.3d 651, 655 (7th Cir. 2008) (citation omitted).

In Olsen, the question at issue was whether the insured's negligent driving resulted in two accidents or occurrences or only one. The court found one occurrence where "there was virtually no time or space interval between the two impacts," the whole incident occurred in less than a second, and the driver did not regain control of the vehicle prior to striking the second auto. Olsen at 241.

Effect upon Policy Deductibles

The number of occurrences question often arises when the policies at issue contain per occurrence deductibles or retained limits or limit of liability clauses providing through various language that the per occurrence limit is the most that will be paid for all damages arising out of a single occurrence. In Nicor, Inc., supra, the application of the Illinois cause theory resulted in 195 separate occurrences and no coverage for the insured.

Nicor supplied natural gas to residential homes and also provided regulatory meters with relief valves which, as a safety feature, open if the pressure of the gas being supplied is excessive. Until 1961, the relief valves contained mercury. After 1961, Nicor began voluntarily changing the meter to relief valves that were spring loaded. Of the approximately 300,000 homes where the meter was replaced, 1,070, or one-half of 1 percent of the homes, were found to contain impermissibly high levels of mercury due primarily to mercury spilling during the replacement process. Of that number, 195 were subject to the policies at issue.

Each occurrence was subject to a self-insured retention of at least $100,000. Nicor spent approximately $90 million in investigating, identifying, and remediating the mercury contamination and then sought indemnification from its insurers. All insurers settled except the excess insurers in whose policies the 195 homes fall. There was no dispute that the costs for each of the 195 homes was less than the self-insured retention (SIR) amount. The insurers argued that Nicor's liability arose from 195 separate occurrences. The circuit court found one occurrence: Nicor's systematic failure to consistently remove the mercury regulators. See Nicor at 285. The appellate court reversed finding that the contamination was the product of separate and independent acts occurring in an isolated number of cases.

The Illinois Supreme Court affirmed finding that Nicor's liability did not arise from any inherent defect in the mercury regulator or manufacturing or installation process. Nor did Nicor's liability "derive from any systemwide policy or procedure regarding the methodology employed for removing the regulators." Nicor at 295. Instead, liability was only incurred when the mercury happened to spill. The facts showed that a spill was a rare event, the spills had no common cause, did not share a temporal or geographic pattern, and occurred at different times over a 17-year period. The court found that the facts did not support the theory that the 195 spills resulted from a common cause and instead found 195 separate occurrences, the costs for each falling within the per occurrence deductible, and resulting in no coverage for Nicor.

Nicor argued that this interpretation of the policies denied it "the benefit of its bargain" but the court rejected that argument. The court found Nicor to be a "major corporation with substantial experience and formidable bargaining power" who "voluntarily agreed to assume the risk of absorbing individual claims of the magnitude involved here." Nicor at 297. Nicor elected to absorb part of the risk in exchange for reduced premiums and:

no principle of public policy authorizes courts to relieve such entities of the consequences of business calculations merely because those calculations ultimately prove to be erroneous.


Other jurisdictions may not face the same decisional limitations when presented with such "unfortunate and unique" facts or may apply the cause theory in a way that would have resulted in multiple occurrences. See e.g., Koikos (if the immediate injury producing event was each individual boy getting stuck in the pit, not the insured's underlying tortious omission of negligent security). Other jurisdictions may find seemingly unambiguous language ambiguous as a way to maximize the coverage available so that the application of the cause theory is affected by public policy considerations.


The differences in outcomes when using the same underlying theory seem to stem from using different considerations in applying the cause theory. Florida focuses on the "immediate injury producing event," while Pennsylvania looks to the act giving rise to the insured's liability. Illinois and Wisconsin include a consideration of the time and space between acts. Wisconsin, Alabama, and to a lesser extent Illinois, each look at whether there was one continuing proximate cause of all the injuries/damage.

Koikos, Nicor, and Home Indem. Co. v. City of Mobile, 749 F.2d 659 (11th Cir. 1984) (finding that each discrete act or omission by the City in causing the drainage system to flood was the occurrence which resulted in a separate occurrence for each overflowing drain but a $100,000 per occurrence limit for all damages resulting from the failure of each drain) provide examples of situations where applying the cause theory yields results that would have been the same had the court applied the effects theory and looked to the number of victims or claimants.

Two divergent opinions with respect to multiple shooting victims are described above in Koikos and Donegal. Olsen is a case finding one occurrence resulting from auto collisions involving one tortfeasor and Illinois Nat'l Ins. Co. v. Szczepkowicz, 542 N.E.2d 90 (Ill. 1989), finds multiple occurrences resulting from auto collisions involving one tortfeasor. Like Washoe Cty. and Lee, the sexual abuse cases also result in different results when applying the cause theory to determine the number of occurrences. Accord Interstate Fire & Cas. Co. v. Archdiocese of Portland in Oregon, 35 F.3d 1325 (9th Cir. 1994) (focusing on whether there was one proximate, uninterrupted and continuing cause which resulted in all of the injuries and damage and finding one occurrence per policy period under the church's policy for the damages suffered by a boy molested over 4 years by a priest).

This small sampling of cases highlights the difficulties that sometimes plague this issue; seemingly similar facts considered by different courts may result in a different number of occurrences, despite each court's reliance on the cause theory.

Contributing author Rebecca C. Appelbaum is a senior associate practicing in the area of third-party coverage at Butler Pappas Weihmuller Katz Craig, LLP.

Opinions expressed in Expert Commentary articles are those of the author and are not necessarily held by the author's employer or IRMI. Expert Commentary articles and other IRMI Online content do not purport to provide legal, accounting, or other professional advice or opinion. If such advice is needed, consult with your attorney, accountant, or other qualified adviser.