chevron-down Created with Sketch Beta.


Determining the Number of Occurrences and Its Effect on Coverage

Tred R Eyerly, Rina Carmel, and Karin Scherner Aldama


  • The number of occurrences is a key issue because it can dramatically affect the dollars involved in the claim.
  • A full understanding of several factors is often critical to determining the number of occurrences arising from an incident.
  • Factors include the policy language, limits of liability, non-cumulation clauses, the facts of the underlying case. and the governing law.
Determining the Number of Occurrences and Its Effect on Coverage
PeopleImages via Getty Images

Once it is determined that there is bodily injury and/or property damage caused by an occurrence, it may become critical for both the carrier and the policyholder to determine whether there was one or multiple occurrences. That determination is not always easy or straightforward. Disagreements about how many occurrences there were can arise, e.g., where the facts present numerous causes; where the insured committed numerous tortious acts; where there are numerous claimants, events occurring over several consecutive policy periods, or multiple layers of coverage; or some combination of all of the above.

The number of occurrences is a key issue because it can dramatically affect the dollars involved in the claim. For example, pre-1980s policies and some modern-day policies do not contain aggregate limits, so the number of occurrences directly affects the number of limits available under the policy. Most modern policies contain both per occurrence and aggregate limits, so if the insured is sued for an amount in excess of policy limits, establishing that there is more than one occurrence may increase the amount available for indemnity if the aggregate limit is more than the per occurrence limit. By the same token, the existence of more than one occurrence also increases the number of deductibles or self-insured retentions (SIRs) the insured must pay before a defense or indemnity is owed. The occurrence analysis could favor the insurer in one case and the policyholder in another, depending on policy language, the facts of the claim, the amount of the limits and deductible/SIR as compared with the potential exposure in the underlying claims, as well as how the court analyzes the issue.

To determine the number of occurrences, it is necessary to carefully consider the policy language, including the definition of “occurrence,” limits of liability, non-cumulation clauses, and other pertinent language; the facts of the underlying case; and the governing law. As we shall see in our survey of cases below, a full understanding of these factors is often critical to determining the number of occurrences arising from an incident.

“Occurrence” is defined in the most recent Insurance Services Office (ISO) commercial general liability (CGL) main form as “an accident, including continuous or repeated exposure to substantially the same general harmful conditions.” Some non-standard forms have added a significant word, “related,” so that the term is defined as “including continuous, repeated, or related exposure to substantially the same general harmful conditions.”

Courts employ several tests to determine the number of occurrences. The categorization of tests varies, and commentators and courts do not necessarily agree on what the tests are or how they should be categorized. A useful way to conceptualize the issue is that the tests analyze either cause or effect. The catch is that there are often different ways of viewing cause or effect, which can significantly complicate the analysis. In most cases, the outcome is likely to depend on how the court views the facts.

The majority of jurisdictions use the “cause” test. Under this test, courts variously determine the number of occurrences by looking to the number of causes or causal factors, liability-triggering events, or “unfortunate events” leading to the injury or damage.

In applying the cause test in situations involving negligent omissions by the insured, some courts add as another element of the test a determination of whether the cause and result are closely linked in time and space. Under this time-and-space variant of the cause test, there is only one occurrence if such a close linkage exists; if it does not, there are multiple occurrences.

A minority of jurisdictions employ the “effect” analysis. Under this approach, courts examine the number of injuries, or effects, of the insured’s act to determine the number of occurrences. The focus is on the number of injuries resulting from the act, not the number of acts performed by the insured.

Below, we summarize recent cases in which the courts considered the number of occurrences.

Cases Employing the Cause Test

Fellowship of Christian Athletes v. Axis Insurance Co., 758 F.3d 982 (8th Cir. 2014) (Missouri law). This case illustrates the occurrence-related issues that can arise when an insured has a tower of insurance. Typically, a higher-tier excess insurer’s interest is to argue that a lower-tier policy or policies cover multiple occurrences, so that the higher-tier policy will attach later (if at all).

In this case, two boys drowned at a summer camp run by the insured, the Fellowship of Christian Athletes (FCA). The two boys could not swim, and their camp permission forms indicated that they were non-swimmers. After a pool party, the FCA staff realized the two boys were missing. They had drowned, and their bodies were found lying side by side at the bottom of the deep end of the pool. The death certificate for one boy listed the time of death as 10:42 p.m., while the other boy’s time of death was listed as 10:44 p.m.

The FCA was insured under a tower of insurance consisting of three policies. The primary policy had limits of $1 million per occurrence and $5 million in the aggregate. The first‑tier umbrella policy provided up to $10 million in coverage in excess of the primary policy. The second-tier umbrella policy provided $5 million in excess of the lower-tier policies.

The FCA filed suit, seeking a determination on whether the deaths were caused by one or two occurrences under the primary policy. The district court granted the primary insurer’s motion for summary judgment and held that the drownings were caused by one occurrence. Therefore, the primary insurer’s liability was limited to $1 million, and the first-tier excess insurer would be liable for damages over $1 million. The first-tier excess insurer appealed, arguing that there were two occurrences under the primary policy, such that the primary insurer’s indemnity obligations should be $2 million.

The Eighth Circuit affirmed. Under Missouri law, the “cause” approach applied to determine the number of occurrences. Under the cause approach, an insured’s single act was considered the accident from which all claims flowed. The court observed that some states other than Missouri had “modified the causation approach to include a time and space test” to be applied in cases of negligent omission, which found two occurrences if two injuries were not so closely linked in time and space as to be considered one event. The court noted, however, that “Missouri has not adopted the time and space test . . . and we decline to apply it in this case.”

The first-tier excess insurer argued that there were two separate occurrences because the underlying lawsuit alleged negligent supervision, and the boys were under the care of two different camp counselors. The court rejected this argument, ruling that the conduct of the insured—the FCA—and not that of the camp counselors was the proper focus of the occurrence analysis. The underlying litigation alleged that the FCA was negligent in allowing the boys to attend the pool party while knowing they could not swim, and in failing to properly train and supervise the camp counselors. Further, the boys arrived at the pool at the same time, swam in the pool during the same one-hour period, and were discovered at the bottom of the pool at the same time.

Accordingly, the FCA’s alleged negligent conduct constituted one occurrence because the underlying lawsuit alleged that the drowning was caused by exposure to the same general harmful conditions caused by the FCA’s single insured act.

CHS/Community Health Systems, Inc. v. Lexington Insurance Co., No. 3:11-cv-00449, 2013 U.S. Dist. LEXIS 175787 (M.D. Tenn. Dec. 9, 2013) (Tennessee law). The main issue in this case was how many self-insured retentions (SIRs) the insured had to satisfy—71 separate SIRs for each of 71 allegedly misconducted surgeries or a single SIR based on the insured’s alleged mismanagement of the hospitals in which the surgeries were performed by individuals other than the insured. Declining to apply the effects test usually applied under Tennessee law to determine the number of occurrences, the court basically applied a cause test and found that there had been a single occurrence. Consequently, the insured had to satisfy only a single SIR.

Since late 2005, the insured had provided administrative services for a New Mexico hospital association (referred to in the decision as the Hospital). As a result, the insured had to defend against 71 lawsuits alleging that surgeries conducted by the hospital had been misperformed.

The insured’s policy included errors and omissions (E&O) coverage. With respect to that coverage, the policy required the insurer to provide coverage subject to a $5 million SIR for each occurrence, with a per occurrence limit of $25 million. The policy also provided healthcare professional liability (HPL) coverage subject to a $6 million SIR per “medical incident” and a per occurrence (or per incident) limit of $35 million. The issues presented to the court were whether the claims in the underlying litigation represented separate occurrences for purposes of E&O coverage and separate medical incidents for purposes of HPL coverage—in other words, whether the insured was required to satisfy a separate SIR for each of the 71 claims against it.

In answering this question, the court declined to apply the effects test normally used under Tennessee law to determine the number of occurrences in a coverage dispute. The court found that

[u]nder these circumstances, . . . [Tennessee precedent is] inapplicable and [the Court] declines to adopt the “effects test” for two reasons. First, the issue before the Court is not the limits of [the insurer’s] liability, but the number of SIRs [the insured] must meet before . . . coverage is triggered. As such, the purpose behind interpreting ambiguous contracts in favor of the insured espoused in [Tennessee precedent]—to provide adequate compensation for victims—is not served by the effects test here. Second, courts look to the effects test where multiple reasonable interpretations of the language at issue exist, rendering the policy ambiguous. Here, . . . only one reasonable interpretation of the word “occurrence” both ensures the language of all parts of the policy is in harmony and honors the intentions of the parties.

The court concluded that, because the policy applied the SIR to occurrences instead of claims, “the underlying cause of the various tort claims against [the insured]—[the insured’s] alleged mismanagement of the Hospital—must be interpreted as one occurrence under the terms of the Policy.” For the same reasons, the court held that the effects test was inapplicable to its analysis of the HPL coverage. Specifically, the court focused on the fact that the only professional services the insured provided were management services, with respect to which it might have engaged in “negligent supervision”—thus constituting a single occurrence.

United Services Automobile Ass’n v. Neary, 307 P.3d 907 (Alaska 2013) (Alaska law). Employing the “unfortunate events” test, the Alaska Supreme Court concluded that there was one occurrence under a homeowner’s policy in a case involving multiple insureds and multiple claimants. Under the “unfortunate events” test, “it is the unforeseen event [unforeseen by the insured], not every act of [the insured’s] negligence preceding it, that constitutes the accident or occurrence for purposes of insurance coverage.”

In Neary, a single gunshot killed one and injured another when the insureds’ teenage son fatally shot one friend and seriously injured another with one of his father’s guns. The victims’ parents separately sued the insureds, their son, and the insureds’ insurer for negligence and negligent infliction of emotional distress. The insurer argued that there was a single occurrence because only one gunshot was fired. The insured parents argued that (1) there were multiple occurrences because several separate acts of negligence had preceded their son’s shooting, and (2) there were six occurrences, because there were six victims, namely the two teenagers and their respective parents. The trial court held that there was one occurrence but that each of the three insureds was entitled to a separate coverage limit of $300,000 per occurrence, based on the policy’s severability provisions, for a total covered amount of $900,000.

The Alaska Supreme Court held that nothing in the declarations page suggested that the policy limits of $300,000 would be increased depending on the number of insureds. The single per occurrence limit applied regardless of the number of insureds.

With respect to the number of occurrences, the court determined there was a single occurrence. The insureds argued that each act of negligence that enabled their son to shoot his friends should be counted as a separate occurrence: (1) his own negligent handling of the gun, and (2) each negligent act of his parents in failing to secure their firearms and supervise their son’s activities. The court noted, however, that it was the unforeseen event, not every act of negligence preceding it, that constituted the accident or occurrence. The unforeseen event was the shooting—that was, from the perspective of the insured parents, the accident that gave rise to coverage.

In responding to the insureds’ argument that there were six occurrences (that is, essentially for application of an effects test), the court rejected that test as a way to determine the number of occurrences, observing “that the test renders insurers’ liability both unpredictable and limitless, since any one event can cause many injuries to many people.”

Roman Catholic Diocese of Brooklyn v. National Union Fire Insurance Co. of Pittsburgh, 991 N.E. 2d 666 (N.Y. 2013) (New York law). The Court of Appeals of New York also used the unfortunate events test in deciding that allegations of several instances of sexual abuse by one priest against the same victim constituted multiple occurrences, requiring exhaustion of the SIR for each implicated policy.

The underlying complaint alleged that the priest sexually abused the minor claimant on several occasions from August 10, 1996, through May 2002. The alleged molestations took place in several different locations. The diocese settled with the claimant for $2 million and then demanded reimbursement from its insurer.

The diocese held primary policies over three consecutive one-year periods: August 31, 1995, to August 31, 1996; August 31, 1996, to August 31, 1997; and August 31, 1997, to August 31, 1998. Each policy had a $250,000 SIR, after which the policy’s liability limitation was $750,000. The policies defined “occurrence” as “an accident, including continuous or repeated exposure to substantially the same general harmful conditions.”

The diocese sought coverage under the 1996–1997 and 1997–1998 policies. After the insurer disclaimed coverage based on two sexual abuse exclusions, the diocese filed suit for a declaratory judgment. The insurer then moved for summary judgment, asserting that, to the extent there was coverage, it was subject to multiple SIRs. The trial court determined that the incidents of sexual abuse constituted a single occurrence and that the language of the policies required the exhaustion of the SIR for each implicated policy. The Appellate Division reversed, declaring that the alleged acts of sexual abuse constituted multiple occurrences.

Affirming the Appellate Division, the Court of Appeals noted that nothing in the policies demonstrated an intent to aggregate the incidents of sexual abuse into a single occurrence. Applying the unfortunate events test, which it defined as “requiring consideration whether there is a close temporal and spatial relationship between the incidents giving rise to injury or loss, and whether the incidents can be viewed as part of the same continuum, without intervening agents or factors,” the court found that the incidents of sexual abuse constituted multiple occurrences. The incidents spanned a six-year period and happened at multiple locations, and therefore lacked the requisite temporal and spatial closeness to be considered a single incident. Further, the incidents were not part of a singular causal continuum.

Consequently, the diocese had to exhaust the SIR for each occurrence that happened within an implicated policy under which it sought coverage.

National Liability & Fire Insurance Co. v. Itzkowitz, 2015 U.S. App. LEXIS 16387 (2d Cir. Sept. 15, 2015) (New York law). Following New York law, the Second Circuit applied the unfortunate events test to find three occurrences resulting from a highway accident.

A dump box attached to a dump truck struck and damaged an overpass. The impact separated the dump box from the truck, and the dump box landed in the right lane of the highway. Some 30 seconds to 5 minutes later, a first vehicle struck the detached dump box. Then, at some point between a few seconds and 20 minutes later, a second vehicle struck the dump box.

The insurer for the dump truck owner, National Liability & Fire Insurance Company, argued there had been one accident, or at most two separate accidents, under the policy. The district court found, however, that there had been three occurrences, and National appealed.

In applying New York’s unfortunate events test, the court used a two-part inquiry, first identifying the operative incident giving rise to the liability, and second considering whether there was a close temporal and spatial relationship between the three incidents and whether they could be viewed as part of the same causal continuum, without intervening factors.

National argued that the policy language providing that “all ‘bodily injury’ and ‘property damage’ . . . resulting from continuous or repeated exposure to substantially the same conditions will be considered as resulting from one ‘accident’” indicated an intent to aggregate separate accidents into a single occurrence. But both the New York Court of Appeals and the Second Circuit had previously applied the unfortunate events test in considering the same or similar policy language, concluding that such language did not evidence an intent to aggregate incidents.

Therefore, applying the unfortunate events test, the court first determined that each of the three collisions was a separate operative incident. Further, although the three incidents occurred close in time (and even assuming the shortest possible difference in time between the three incidents), nothing suggested that the narrow time span between each incident played a role in causing any of the other incidents.

The spatial proximity of the events presented a closer question. The dump box struck the overpass at a location different from where the first vehicle struck the dump box. The second and third incidents, however, were spatially proximate. The two collisions with the dump box occurred at virtually identical spots and involved the same dump box. But this was not the end of the analysis. The unfortunate events test did not dictate that separate incidents were part of the same accident if they met any one of the three criteria—spatial proximity, temporal proximity, or occurrence in a causal continuum. Rather, the test reflected a “common sense” balancing of the three elements.

Accordingly, the court looked to whether there was an “unbroken” continuum between the events. The dump box striking the overpass was not responsible for the second and third incidents. For example, the first incident did not weaken the overpass’ structure in a way that caused further injury. Rather, the dump box was released and came to a stop on the highway. Then, after 30 seconds passed, the first vehicle struck the dump box. This was distinct from the dump box striking the overpass. Finally, the second vehicle struck the dump box, and this collision was unrelated to the preceding collision involving the first vehicle as it was the dump box, and not the first vehicle, that caused the collision with the second vehicle.

Therefore, applying the unfortunate events test, there were three separate accidents under the National policy.

Mitsui Sumitomo Insurance Co. v. Duke University Health System, 509 F. App’x 233 (4th Cir. 2013) (North Carolina law). The court held that, under North Carolina’s cause test for the number of occurrences, there had been a single occurrence where employees of Duke University Health System mistook discarded hydraulic elevator fluid for surgical detergents and lubricants, leading to at least 127 patients being operated on with surgical instruments allegedly contaminated by hydraulic fluid.

Duke had engaged Automatic Elevator Company to renovate two elevators at Duke’s parking deck. When Automatic Elevator completed the work, it left barrels full of used hydraulic fluid in its designated storage area at Duke’s hospital. Duke employees mistakenly thought the barrels contained surgical detergents and lubricants and “ultimately used the hydraulic fluid to wash hundreds of surgical instruments.” Approximately 127 patients alleged they came into contact with the contaminated instruments and sued Duke, which settled the claims for over $6 million and then sued Automatic Elevator.

Automatic Elevator had obtained two CGL policies for the relevant time period, one in effect while the company worked on the first elevator and one in effect while it worked on the second. Both policies included a $1 million limit for “any one occurrence” and a $3 million aggregate limit, and both defined occurrence as “an accident, including the continuous repeated exposure to substantially the same harmful condition.” Both also contained a “per elevator” endorsement that applied the aggregate limit separately to each elevator serviced by the company.

After Duke sued Automatic Elevator, the company’s insurer brought an action seeking a declaration that it had no further obligations to the company because the hydraulic fluid mistake was a single occurrence involving only one elevator, so that it was obligated to pay only $1 million under the company’s policies, which it had already done. The district court entered summary judgment for the insurer, and Duke appealed.

Duke argued that there were multiple occurrences for two reasons. First, Automatic Elevator’s leaving the hydraulic fluid on Duke’s premises was an intentional act and therefore could not be an “accident” under Automatic Elevator’s policy. Second, the court should look to the “most immediate cause” of the claimed injuries, namely “each surgery or each use of hydraulic fluid to wash surgical instruments,” to determine the number of occurrences.

The Fourth Circuit rejected both arguments. With respect to Duke’s first argument, the court concluded that the dictionary definition of “accident” as something unintended and unforeseen had no bearing on the number of occurrences, but was relevant only to determining whether there even was an occurrence, or determining an occurrence’s date. Because the insurer disputed neither that there had been an occurrence nor the trigger-of-coverage date, Duke’s definitional argument was misplaced, particularly since North Carolina had adopted a specific test for evaluating the number of occurrences.

Turning to that test, the court noted that North Carolina used the “proximate cause theory” of the cause test to assess the number of occurrences. Under that test, “courts consider an event to constitute one occurrence when there was but one proximate, uninterrupted, and continuing cause which resulted in all of the injuries and damage.” The court then held that, as Automatic Elevator was the insured, it was Automatic Elevator’s conduct on which the analysis had to focus, finding that “looking at the number of surgeries or instances of using hydraulic fluid to wash surgical instruments . . . would [improperly] turn the focus in this case” from Automatic Elevator’s actions to Duke’s. Because Automatic Elevator’s only relevant action had been leaving the hydraulic fluid on Duke’s premises when it finished its work, the court held that there was only a single occurrence, for which reason the “per elevator” endorsement also did not apply.

This case is noteworthy because, as pointed out in the vigorous dissent, the court gave short shrift to Duke’s definitional argument based on the policy language and essentially read the word “accident” out of the analysis of the number of occurrences. The dissent also pointed out that injury, which was also required for coverage, occurred only when the tainted surgical instruments were used—had the left-behind hydraulic fluid been disposed of, for example, no injury would ever have occurred. The dissent therefore would have found that there were 127 occurrences—one for each of the surgeries at issue. It would also have applied the “per elevator” endorsement (given that the hydraulic fluid had come from both of the elevators on which Automatic Elevator had worked), such that Automatic Elevator would have been entitled to $6 million in coverage.

Republic Underwriters Insurance Co. v. Moore, 493 F. App’x 907 (10th Cir. 2012) (Oklahoma law). Applying the cause test, the court concluded that there was one occurrence when food infecting several hundred people with E. coli was prepared and served at two different locations.

The insured restaurant prepared and served E. coli-contaminated food between August 15 and August 25, 2008. Some of the food was served at the restaurant, and some of it was served at a church gathering catered by the restaurant. Three hundred forty-one people were infected, and one person died.

The insurers argued that the entire contamination period constituted one continuing occurrence as defined by the policies. The event giving rise to all the alleged injuries was the restaurant’s preparation, handling, or storage of food that purportedly became contaminated with E. coli.

The restaurant relied on an investigation report that was inconclusive as to how the bacteria had contaminated the restaurant. The report suggested multiple likely contributing factors, including contamination by food handlers as well as cross-contamination from food-preparation equipment, counter surfaces, and storage areas.

The magistrate judge found two occurrences because two separate locations were used to prepare and serve the food.

The Tenth Circuit reversed, based on Oklahoma’s cause test. As long as the injuries stemmed from one proximate cause, there was a single occurrence. Here, all the injuries were proximately caused by the restaurant’s ongoing preparation of contaminated food. Therefore, there was only one occurrence.

Mid-Continent Casualty Co. v. Basdeo, 477 F. App’x 702 (11th Cir. 2012) (Florida law). Here, too, the court applied a cause test, but concluded that negligent roof tarping and faulty repair work to two roofs constituted three occurrences under a CGL policy.

Southgate Gardens Condominium owned buildings that were damaged by Hurricane Wilma in 2005. Southgate hired the insured, First State Development Corporation, to make repairs to the roofs.

On November 1, 2005, First State completed tarping to protect the buildings until more extensive repairs could be made. Thereafter, on November 11, 2005, First State contracted with Southgate to remove and replace the roofs of the Southgate buildings.

Eventually, it became apparent that the tarps placed by First State were inadequate, allowing water to enter the buildings and cause damage. Further, when it attached the tarps, First State created holes in the roofs of buildings, which caused additional damage. First State also left open the mansards (a type of sloped roof). Finally, the peeled-back condition of flat roofing caused by First State allowed rain to enter the buildings.

The insurer filed a declaratory judgment action against its insured First State, Southgate, and building resident Basdeo. The district court found there were three occurrences.

The Eleventh Circuit affirmed. Both parties agreed that the faulty tarping performed by First State constituted one occurrence. But they disagreed on whether the roof repair work performed by First State constituted a single occurrence for damage caused to the two roofs or two separate occurrences, one occurrence for each roof.

Florida used the “cause theory,” under which the act that caused the damage, and that was neither expected nor intended from the standpoint of the insured, constituted an occurrence. The insurer argued that the damage caused by work on the flat-top portion of the roofs was not a separate occurrence from the damage caused in connection with First State’s work on the sloped roofs. Work on both the sloped and the flat-top portions of the roofs was performed under the same contract, so that the true cause of the damage was First State’s breach of its contract with Southgate, and all damages that flowed from that breach were a single occurrence.

The court disagreed, reasoning that the insurer’s argument would redefine an occurrence to mean all damages caused by a breach of contract. The court also rejected the insurer’s second argument, that a proximate causal link between the work on the sloped and the flat portions of the roofs meant that there was a single occurrence, because it found this argument was unsupported under Florida law.

Instead, the court ruled that there were three occurrences: one in connection with the tarping and resulting damage to various units, and two in connection with the contracted repairs to unrelated portions of the roofs.

With respect to policy limits, the policy contained a $1 million per occurrence limit and, apparently, two aggregate limits, a General Aggregate Limit and a Designated Construction Project General Aggregate Limit. The resident argued that an occurrence could fall under each of the aggregate limits, such that a $4 million limit should apply. The court rejected this argument, stating that to apply a $2 million limit to one of the occurrences would improperly create additional coverage for that occurrence, contrary to the unambiguous per occurrence limit of $1 million per occurrence. Thus, the court held that $3 million, at $1 million per occurrence, was the appropriate coverage.

Dragas Management Corp. v. Hanover Insurance Co., 798 F. Supp. 2d 758 (E.D. Va. 2011) (Virginia law). This case involved damages resulting from the installation in two housing developments of defective Chinese drywall. Using Virginia’s cause test, the court concluded that each replacement of non-defective components in a particular residence, made necessary to replace defective Chinese drywall, constituted a separate occurrence under a CGL policy, for a total of 74 occurrences. (However, replacement of the defective Chinese drywall itself was not an occurrence and hence not covered.)

Dragas Management Corporation, a developer, hired the insured, Porter-Blaine Corporation, to supply and install drywall for the homes in two of Dragas’s developments. Porter-Blaine eventually installed defective Chinese drywall in 74 of the homes in the two developments. A high level of elemental sulfur contained in the Chinese drywall caused property damage in the homes, including corroding heating, ventilation, and air conditioning coils, damaging wiring, tarnishing or corroding metal objects, and causing a bad odor.

When Porter-Blaine refused to remove and replace the drywall, Dragas remediated the problem by having the homeowners temporarily move, tearing out the drywall, replacing it, and repairing the other property damage caused by the drywall, all at its own cost. Dragas then invoked the arbitration clause in its contract with Porter-Blaine and was awarded $4.9 million in damages. The award was converted into a judgment.

Unable to collect the judgment from Porter-Blaine, Dragas sued Porter-Blaine’s insurers to enforce the arbitration award. Porter-Blaine had primary CGL coverage with limits of $1 million per occurrence and $2 million in the aggregate. It also had an umbrella excess liability policy with limits of $10 million per occurrence and in the aggregate.

Dragas moved for partial summary judgment on the primary CGL policy. The court concluded, under principles of coverage for construction defect claims, that replacement for defects was not an occurrence, but any need to replace non-defective components of the affected residences was an occurrence or part of an occurrence.

Next, Dragas argued that each installation of the defective drywall constituted a separate occurrence because the introduction of the drywall was the cause of the resulting damage. The primary insurer argued that the damage had a single cause, i.e., the single purchase of the defective Chinese drywall by Porter-Blaine. While agreeing that the purchase was a “but-for cause of the damage,” the court found that focusing on the purchase improperly failed to consider the act that set in motion the chain of events that caused injury. Instead, it was the installation of the defective drywall in each of the affected residences that caused injuries and gave rise to the insured’s liability, thereby constituting a separate occurrence. Put another way, the act of installing the drywall in each home was what set in motion the chain of events culminating in the damage to that home. Thus, there were 74 occurrences, one for each affected home.

LSG Technologies, Inc. v. U.S. Fire Insurance Co., No. 2:07–CV–399–DF, 2010 U.S. Dist. LEXIS 140879 (E.D. Tex. Sept. 2, 2010) (Texas law). LSG Technologies illustrates the complex issues that arise in long-tail claims arising over multiple policy periods, where the issue of the number of occurrences can intersect with a variety of other issues—i.e., what constitutes an occurrence, what is the trigger of coverage, vertical versus horizontal exhaustion, and what constitutes proof of exhaustion.

In LSG Technologies, the insureds were defendants in numerous underlying actions brought by numerous claimants, all alleging injury from exposure to asbestos. The insureds sought coverage from their primary insurer, which had issued three years of consecutive policies, and from their excess insurer. The court held, under the Texas version of the cause test, that there was one occurrence for each set of claimants who were exposed to asbestos-containing gaskets at the same time and location.

Because Texas had not adopted a particular trigger theory for asbestos claims, the court began its analysis by reviewing the three tests for what constitutes a trigger in the asbestos context: (1) injury in fact to an individual claimant, (2) the manufacture and sale of asbestos-containing products, and (3) injury in fact to several individuals through exposure at the same time and place. The court then noted that in other contexts, Texas applies a cause test, “focusing on the events that cause the injuries and give rise to the insured’s liability, rather than on the number of injuries.”

The court next noted that there was no dispute that the cause of the claimants’ injuries was exposure to asbestos-containing gaskets. Based on the primary and excess policies’ definitions of “occurrence”—“continuous or repeated exposure to conditions” and “exposure to substantially the same general conditions,” respectively—the LSG Technologies court made an Erie guess that Texas would adopt the third approach; namely, that there was a single occurrence for each group of claimants who were exposed to the asbestos-containing gaskets at the same time and location. However, a factual dispute remained as to when each claimant was exposed to the products at issue, i.e., how many occurrences there actually were, precluding summary judgment.

Plastics Engineering Co. v. Liberty Mutual Insurance Co., 759 N.W.2d 613 (Wis. 2009) (Wisconsin law). Plastics Engineering addressed an issue very similar to that in LSG Technologies but reached a different conclusion, even though it also applied the cause test (emphasizing the inherent uncertainties of the number-of-occurrences analysis). In certified questions from the Seventh Circuit, the Wisconsin Supreme Court was asked the following: When the insured sells its asbestos-containing product without a warning, causing bodily injury to vast numbers of victims, is there only one occurrence? Or is each individual’s exposure to the asbestos and resulting injury an occurrence, creating multiple occurrences? The Wisconsin Supreme Court applied the state’s cause test, determining that each individual’s exposure was a separate occurrence.

The insured manufactured and sold asbestos-containing products from 1950 to 1983. The insured was named in a number of suits because of bodily injury or wrongful death allegedly caused by exposure to its asbestos-containing products. The insured held several primary and excess policies issued during the years of alleged exposure and injury.

At trial, the federal district court concluded that each person’s injury resulting from exposure to asbestos constituted a separate occurrence. On appeal, the Seventh Circuit determined that the issues presented important unresolved questions under Wisconsin law, and the court certified the questions to the Wisconsin Supreme Court.

The insurer relied on the policy’s limits of liability provision, which provided that “[a]ll bodily injury and property damage arising out of continuous or repeated exposure to substantially the same general conditions shall be considered as arising out of one occurrence,” to argue that the number of occurrences should be limited to a single one. The Wisconsin Supreme Court held this provision merely limited an individual claimant’s repeated and continuous exposure to asbestos-containing products to one occurrence (instead of multiple occurrences).

The insurer also argued that the insured’s manufacturing and sale of the asbestos-containing products constituted one occurrence regardless of the number of individuals injured. The insured, on the other hand, argued that each individual’s exposure to asbestos resulting in injury constituted a single occurrence.

The Wisconsin Supreme Court agreed with the insured. The policy defined “occurrence” as “an accident, including continuous or repeated exposure to conditions, which results in bodily injury. . . .” The most reasonable understanding of this provision was to find that each individual’s exposure to asbestos constituted a “repeated exposure to conditions” resulting in bodily injury. Accordingly, each individual’s repeated exposure constituted a separate occurrence. Under the cause theory, adopted in Wisconsin, where a single, uninterrupted cause results in all of the injuries and damage, there is one accident or occurrence. Therefore, each claimant’s repeated exposure was one occurrence, creating multiple occurrences.

Although the court stated that it was applying a cause test, its determination that each individual’s exposure to asbestos constituted a separate occurrence looks more like an effects analysis. Most cause tests would look to what the insured did (i.e., making decisions, manufacturing products, distributing products to numerous customers or employers in the supply chain). In contrast, an effects analysis usually looks to the impact on claimants. Here, the court concluded that the cause was the impact on, or exposure of, the claimants. This case thus serves as a reminder that even though a state’s approach might refer to one test, how the court applies that test in a particular case might result in an outcome that is different than expected.

Addison Insurance Co. v. Fay, 905 N.E.2d 747 (Ill. 2009) (Illinois law). This sad case analyzed the number of occurrences under the time-and-space variant of the cause test usually applied in Illinois and concluded that the deaths of two teenage boys constituted two occurrences.

The two boys lived in rural Illinois. One afternoon, they left home to go fishing. When they did not return, a search party was assembled. Several days later, the boys’ bodies were found in an excavation pit on the insured’s property. The boys had become trapped in wet clay and sand at the edge of a pool of water in the excavation pit, essentially a quicksand condition. The boys were found facing different directions, although their bodies were in close proximity and touching each other.

The investigators concluded that one boy tried to jump across the water but became trapped. The second boy apparently attempted to help his friend but also became trapped. The investigators could not determine the time of death or how much time had elapsed between the two entrapments.

When the boys’ families sued the insured property owner, the insurer agreed to settle the claims for policy limits. The policy contained a “general aggregate” limit of $2 million and an “each occurrence” limit of $1 million. The insurer filed a declaratory relief action to determine whether the boys’ deaths constituted one or two occurrences.

The trial court found the deaths were the result of two occurrences because the causes of death and the circumstances immediately prior to the deaths were different. The appellate court reversed, determining the boys’ deaths were so closely linked in time and space as to be considered one occurrence.

The Illinois Supreme Court reversed the appellate court. First addressing the burden of proof, the court determined that the boys’ families had demonstrated the existence of coverage and that the insurer’s assertion that there was one occurrence essentially acted as a limitation on the amount of coverage, such that the insurer bore the burden of proof on the issue of the number of occurrences. Next, the Supreme Court agreed with the appellate court’s adoption of the “time-and-space” addition to the cause test. Where negligence resulted from an ongoing omission rather than separate affirmative acts, a time-and-space test effectively limited what would otherwise potentially be a limitless bundling of injuries into a single occurrence under a straightforward application of the cause test.

But the court disagreed that the facts at issue conclusively demonstrated that the boys’ injuries constituted only a single occurrence. From the evidence, the court could infer that the boys had not been trapped simultaneously; rather, one boy had been trapped first and the other had become trapped while trying to free his friend. Otherwise, there was little evidence to support the insurer’s assertion that the injuries were the result of a single occurrence. The investigators could not determine how closely in time the boys had become trapped. Because the insurer could not meet its burden of proof, the court found that the boys’ injuries constituted two occurrences.

It is important to note that, even in Illinois, the time-and-space variant of the cause test applies only in situations where the injury results from the insured’s ongoing negligent omission. Where injuries are the result of a discrete act or event, Illinois applies a straightforward cause test, without any time-and-space analysis. As discussed above, Missouri has specifically rejected the time-and-space variant of the cause test, even where the case involves an insured’s ongoing negligent omission.

Cases Employing the Effect Test

American Modern Select Insurance Co. v. Humphrey, No. 3:11-CV-129, 2012 U.S. Dist. LEXIS 20800 (E.D. Tenn. Feb. 17, 2012) (Tennessee law). Under Tennessee’s effects test, the court held that an attack by seven dogs that inflicted 147 wounds on the claimant constituted a single occurrence because the policy language focused on the injury to the claimant, and not the number of wounds she sustained.

The claimant, while jogging, was attacked by seven dogs owned by the insureds. During the 20-minute attack, each of the dogs bit the claimant, who sustained a total of 147 wounds. The insureds’ homeowner’s policy provided:

However, we will pay no more than $10,000 for any claim made or suit brought against an insured for bodily injury . . . caused by any animal owned by, or in the care, custody or control of, any insured. This limit is the maximum we will pay for any one occurrence.

The policy defined “occurrence” as “an accident, including continuous or repeated exposure to substantially the same general harmful conditions, that results, during the policy period, in: a. Bodily injury[.]”

In the underlying action, the claimant was awarded $100,000. She agreed with the insureds that the judgment would be effective only to the extent of any insurance recovery. In the subsequent declaratory relief action, the insurer asserted that there was a single occurrence, while the claimant argued that there were 147 occurrences (one for each wound) or, in the alternative, seven occurrences (one for each dog).

The court explained that the effects test “states, generally, that limits in a liability policy which limit the insurer’s liability to a specified amount ‘per occurrence’ or ‘per accident’ refers [sic] to the effect of the occurrence or accident, thus making the entire policy limits available to each injured or damaged party.” To treat each of the seven dogs as a separate occurrence would, according to the court, improperly apply a cause test, in that each dog was a cause of the dog attack—not an effect of it. The court further ruled that the 20-minute attack constituted a single occurrence because the policy’s per occurrence provision focused on injury to the claimant, not the number of wounds she sustained. Therefore, there could not be 147 occurrences; rather, there was only a single occurrence.

Cases Based on Policy Language


Nesmith v. Allstate Insurance Co., 24 N.Y.3d 520, 25 N.E.3d 924, 2 N.Y.S.3d 11 (2014) (New York law). The court held that injury suffered by children of different families living at different times in the same apartment was limited to one occurrence under the non-cumulation clause of the landlord’s liability policy.

The liability policy provided a $500,000 limit for “each occurrence.” The policy contained the following non-cumulation clause:


Regardless of the number of insured persons, injured persons, claims, claimants or policies involved, our total liability . . . for damages resulting from one accidental loss will not exceed the limit shown on the declarations page. All bodily injury . . . resulting from one accidental loss or from continuous or repeated exposure to the same general conditions is considered the result of one accidental loss.


The Young family lived in the apartment from November 1992 to September 1993. In July 1993, the Department of Health notified the landlord that one of the children had an elevated blood lead level and that several areas in the apartment were in violation of state regulations. Repairs were made, and the department advised the landlord in August 1993 that the violations had been corrected.

After the Young family moved out, the Patterson family moved into the apartment. Again, one child was found to have an elevated blood lead level. The Department of Health sent another letter stating that violations had been found and needed to be corrected.

In 2004, the two families brought two separate actions against the landlord for personal injury caused by exposure to lead paint. The first family settled in 2006 for $350,000, which the insurer paid. In 2008, the second family settled its claim pursuant to a stipulation that reserved the issue of the applicable policy limit. The insurer paid the $150,000 that it claimed was the remaining coverage under the landlord’s policy. The second family sought a declaratory judgment, asserting that a separate $500,000 limit applied to each family’s claim.

In a prior case, Hiraldo v. Allstate Insurance Co., the New York Court of Appeals had interpreted a non-cumulation clause found in successively issued liability policies. The Hiraldo court held that a person suing for exposure to lead paint during the terms of all the policies could recover no more than one policy limit. Therefore, under the plain terms of the Hiraldo policy’s non-cumulation clause, the number of claims and claimants could not make an additional limit available.

Following the rationale of Hiraldo, the New York high court in Nesmith rejected the second family’s argument that the injuries to the two families were separate losses, because they did not result “from continuous or repeated exposure to the same general conditions.” Both families were exposed to the same hazard, lead paint, in the same apartment. The landlord’s remedial efforts were not wholly successful, and the same general conditions—the presence of lead paint that endangered children’s health—continued to exist. Because both families were injured by exposure to the same general conditions, their injuries were part of a single “accidental loss,” and only one policy limit was available to both families.

A dissent strongly disagreed with the majority’s conclusion and would have found that “[f]airly read, [the non-cumulation clause] provides that the policy limit—$500,000 limit for ‘each occurrence’—applies to limit the liability for lead exposure of children in one family over the course of that family’s tenancy,” consistent with Hiraldo. The dissenting judge considered the majority’s interpretation of the clause “inconsistent with the reasonable expectations of the insured” because it meant that “when the insured renewed his policy and paid his premium, he procured less protection with respect to lead paint claims” than he would have expected.


Kosnoski v. Rogers, No. 13-0494, 2014 W. Va. LEXIS 151 (W. Va. Feb. 18, 2014) (West Virginia law). Injuries from carbon monoxide poisoning suffered by two families living in the insured’s apartment complex arose from a single occurrence. In deciding this case, the West Virginia high court concluded that, although a precedential case had suggested that West Virginia would apply an effects test, the facts and the language of the policy at issue in this case required application of a cause test.

The two claimant families lived in two apartments in the same insured complex. One night, members of both families suffered serious injuries from carbon monoxide poisoning and one person died. A gas boiler furnace in the basement of the apartment complex created the carbon monoxide.

The families sued the property owner and his insurer, seeking a declaration that there were two occurrences, one for each family. The families argued they were exposed to carbon monoxide at different times during the night. Further, they were exposed to varying levels of carbon monoxide.

The policy at issue had limits of $1 million per occurrence and $2 million in the general aggregate. It defined “occurrence” as “an accident, including continuous or repeated exposure to substantially the same general harmful conditions.” Here, there was a leak of carbon monoxide from a single source, the gas boiler furnace. The injuries were from continuous or repeated exposure to substantially the same general harmful conditions. Therefore, there was a single occurrence.

The court held that the case of Shamblin v. Nationwide Mutual Insurance Co. provided controlling precedent. The claimants argued that under Shamblin, an effects test applies: “The term ‘occurrence’ . . . refers unmistakably to the resulting event for which the insured becomes liable and not to some antecedent cause(s) of the injury.” But the court disagreed. Focusing the on the policy language and the facts at issue, it found that, because the gas had come from a single source, the claimants’ injuries “were from continuous or repeated exposure to substantially the same general harmful conditions,” so that there was only a single occurrence. In thus focusing on the cause of claimants’ injuries, the court rejected the claimants’ argument that the relevant “events” were their various injuries, and it refused to consider Reynolds v. Transcontinental Insurance Co., discussed below, on which the claimants relied.

Reynolds v. Transcontinental Insurance Co., Civ. A. Nos. 94-1367, 94-1432, 1995 U.S. Dist. LEXIS 589 (E.D. La. Jan. 17, 1995) (Louisiana law). In a case involving similar facts, the court in Reynolds came to the opposite conclusion from that reached in Kosnoski, ruling that there were five separate occurrences, one for each claimant.

In Reynolds, the insured manager of a duplex apartment hired a handyman to install a heater. The handyman improperly installed and vented the heater. The plumber (insured under a different policy) allowed the manager to obtain permits to feed gas to the heater, even though the heater had not been properly installed or inspected. Thereafter, the duplex tenant had four houseguests visiting, and they turned on the heater because it was cold. Due to the improper installation and venting, lethal levels of carbon monoxide accumulated in the unit. The tenant and two of his guests died from carbon monoxide poisoning, and the two other guests sustained injuries.

The court reviewed the precise events as to how the gas lines came to be fitted and the heater improperly installed and vented, noting that these could each be a separate occurrence. The court further noted that, based on the testimony of the survivors, each claimant sustained discrete injuries over a 24-hour period, and “each person in the apartment was individually exposed to varying degrees of the lethal gas.”

It is interesting that the court did not discuss whether it was applying a cause test or an effects test. Although it appears from the language used by the court that it applied a cause test, the outcome—one occurrence with respect to each of the victims—suggests that the court was, in fact, applying the effects test.

Lexington Insurance Co. v. St. Bernard Parish Government, No. 11-1865, 2013 U.S. Dist. LEXIS 819 (E.D. La. Jan. 3, 2013), aff’d in part and modified, 548 F. App’x 176 (5th Cir. 2013) (Louisiana law). In a case involving interpretation of excess policies, the courts looked to policy language to determine coverage, and essentially stated that they were applying neither a cause nor an effects test. The court concluded that under the policy language, one series of “related occurrences” took place, such that the court essentially batched all of the claims together, with the result that the insured had to satisfy only one retained limit.

After Hurricane Katrina, the insured, St. Bernard Parish, enacted an ordinance to demolish damaged homes. St. Bernard was sued by 73 homeowners, alleging inverse condemnation as to 57 affected properties.

St. Bernard had excess public entity liability insurance policies spanning three consecutive years, and each policy provided coverage for “property damage” and “personal and advertising injury” caused by an “occurrence.” The policies included a $10 million per occurrence and aggregate limit, but coverage was subject to a $250,000 retained limit. The policies provided that the retained limit is “[a]ny one occurrence . . . or series of continuous, repeated, or related occurrences . . . .” The policies further provided that the retained limit “applies separately to each and every occurrence . . . or series of continuous, repeated, or related occurrences.” The policies defined “occurrence,” in part, to provide that “[a]ll damages that arise from the same related, or repeated injurious material or act will be deemed to arise out of one occurrence, regardless of the frequency or repetition thereof, . . . and the number of claimants.”

At issue in the declaratory judgment action was whether the policies’ $250,000 retained limit applied separately to each alleged demolition or property damage asserted in the underlying actions. If so, no coverage would exist because the value of each property was less than $250,000.

The district court found there was no “property damage” caused by an “occurrence” because the property damage was expected and intended by St. Bernard. The court, however, also looked to the “personal and advertising injury” coverage, which applied to “a wrongful eviction from, wrongful entry into, or invasion of the right of private occupancy of a room, dwelling or premises that a person occupies by or on behalf of its owner, landlord or lessor.” The court found that coverage existed under this provision.

With respect to the number of occurrences, the district court reasoned that even if each condemnation was a separate occurrence, under the policies’ retained limit provision, a series of continuous, repeated, or related occurrences could have the same retained limit:

The contractual language “regardless of the number of claimants,” . . . expressly provides for grouping the claims of individuals suffering damage from related acts into one occurrence. Moreover, the [inclusion of the] word “related” . . . here impacts the Court’s determination of what constitutes an occurrence.

Each condemnation was not a separate “occurrence.” But even if it were, a series of “related occurrences” could have the same retained limit. In other words, the court essentially batched all of the plaintiffs and properties together into one related occurrence, and the insured had to pay only one retained limit. And the court relied on the above-quoted policy language to distinguish otherwise binding precedent mandating the application of an effects test under Louisiana law—it basically said that the policy language, and particularly the inclusion of “regardless of the number of claimants,” made the precedent inapplicable.

The court further found that, although the resulting demolitions occurred at different times over the course of more than one year, they had their genesis in a single action by St. Bernard, i.e., the passing of the ordinance. Therefore, the injuries produced by St. Bernard’s condemnation of properties at its meeting were related under the terms of the policies.

On appeal, the Fifth Circuit noted that the policies’ retained limit applied “separately to each and every occurrence . . . or series of continuous, repeated, or related occurrences.” The Fifth Circuit agreed that the acts of demolition alleged in the underlying actions were related because they all resulted from St. Bernard’s ordinance condemning the properties that remained in disrepair after Hurricane Katrina. The fact that the properties in the underlying action were demolished at different times, in varying degrees, and at different locations did not mean that the acts were not related; rather, they constituted one occurrence or a series of related occurrences. Therefore, the $250,000 retained limit applied once to the alleged acts. The Fifth Circuit concluded by noting that the issue of indemnity was non-justiciable because the pleadings before the district court involved only allegations in the underlying action, and the insured’s liability had not yet been resolved.

Westfield Insurance Co. v. Continental Insurance Co., 2015 U.S. Dist. LEXIS 45437 (N.D. Ohio Apr. 7, 2015) (Ohio law). A similar result was reached in another asbestos exposure case, after careful examination of the policy language.

Mahoning Valley Supply Company (MVS) was sued by numerous claimants who alleged that they had been injured by asbestos-containing products manufactured by third parties but distributed by MVS. The claimants alleged exposure to asbestos fibers at a variety of job sites, on numerous dates, and under a variety of conditions. Two insurers shared defense and indemnity costs.

In 2013, Continental Insurance Company informed MVS that the three policies it had issued to MVS were nearly exhausted, based on its conclusion that the asbestos claims asserted against MVS arose out of a single occurrence. The policies defined “occurrence” as “an accident, including continuous or repeated exposure to conditions, which results in bodily injury or property damage neither expected nor intended from the standpoint of the insured.”

MVS brought suit, and cross motions for summary judgment sought to determine the number of “occurrences” caused by the claims against MVS. MVS argued that each claimant had been exposed to harmful asbestos fibers contained in the products distributed by MVS at different times and under varying circumstances. Therefore, the $500,000 per occurrence limit in each of the three policies was not approaching exhaustion because the injury to each claimant constituted a separate occurrence.

Continental cited cases in support of its single-occurrence argument in which “occurrence” was alternately defined as an “accident” or “event” or “happening.” But Continental’s policy defined “occurrence” solely as an “accident.” Continental’s attempt to limit “occurrence” to the MVS decision to distribute asbestos-containing products contradicted the plain language of the policies and would impermissibly ignore the term “accident” in the definition of “occurrence”—the MVS decision to distribute the products was an intentional act and not an accident; that is, unintended, the result of chance, unplanned, or unforeseen. Further, Continental’s interpretation of the policy would imply terms (such as “event” or “happening”) in the definition of “occurrence” that were not there.

Continental further argued that the cause test should be applied whereby the number of occurrences was determined by reference to the cause or causes of the damage or injury, rather than by the number of individual claims. Continental alleged that the sole cause of MVS’s asbestos liability was the distribution of asbestos-containing products. Although the court (and MVS) agreed that the “cause test” applied, the court rejected Continental’s application of that test as illogical because there were multiple asbestos-containing products and multiple distributions of the different products to numerous sites and to a multitude of customers in varying states over a period of decades. Under those circumstances, “claimants’ injuries were proximately caused by exposure to asbestos fibers under unique conditions and circumstances,” and there were multiple occurrences under the cause test.

Therefore, the “per occurrence” limitations of the policies had not been exhausted and partial summary judgment was granted in favor of MVS.


As noted above, determining the number of occurrences may be a crucial issue that can be complicated by a lack of uniformity in the various methods employed by the courts for making this determination. The advocate must fully understand the test, if any, applied in the relevant jurisdiction, the policy language, and the factual background presented, in order to argue the number of occurrences. The monetary consequences of a finding of one rather than multiple occurrences can be dramatic. Consequently, both insurers and policyholders facing this issue should thoroughly analyze all relevant factors as well as, if state law allows, any public policy arguments.

Keywords: litigation, insurance, coverage, occurrence, cause test, effect test, time-and-space test, unfortunate event test, comprehensive general liability policy, commercial general liability policy, CGL policy

Tred R. Eyerly is Of Counsel at Damon Key Leong Kupchak Hastert in Honolulu, Hawaii.

Rina Carmel is a partner at Zelle McDonough & Cohen LLP in Los Angeles, California.

Karin S. Aldama is a partner at Perkins Coie LLP in Phoenix, Arizona.