So this dump truck can’t make it through an overpass on I-90; the crash knocks the dump box off the truck and into the road. Five minutes later (or 30 seconds, if you believe some people), along comes Mr. Itzkowitz, with his wife and five children, and plows right into the dump box. Then, after another few seconds (or 20 minutes, according to somepeople), ka-pow: Mr. Compton’s car, with another six passengers, does the same thing. What are you gonna do? In National Liability & Fire Ins. Co. v. Itzkowitz, No. 14-3651 (2d Cir. Sept. 15, 2015), the Second Circuit said these facts call for the “unfortunate event test.” It says the test shows there were three different accidents, because the crashes were not part of the “same causal continuum”—even though the owner of the dump truck was liable for “causing” all of them. Which meant that the dump truck’s insurer was on the hook for the $1 million per-accident limit three times, rather than once.

Digging Up the Pertinent Provisions

The owner of the dump truck was insured under a policy that contained this provision:

Regardless of the number of covered “autos”, “insureds”, premiums paid, claims made or vehicles involved in the “accident”, the most we will pay for the total of all damages … resulting from any one “accident” is the Limit of Insurance for Liability Coverage shown in the Declarations. 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.

The policy further defined “accident” to include “continuous or repeated exposure to the same conditions resulting in ‘bodily injury’ or ‘property damage.’”

In Itzkowitz, the insurer argued that all three collisions constituted “one accident,” and so the insurer was liable only once for the $1 million per-accident limit.

Examining the insurer’s claim, the court found that the policy language did not clearly indicate “an intent to aggregate separate incidents into a single occurrence,” noting that the New York Court of Appeals, as well as the Second Circuit, had come to the same conclusion in previous cases involving similar language.  Under New York law, the court held, the absence of such language requires a court to apply the “unfortunate event test” to “determine how many occurrences are subject to coverage.

That test involves a two-part inquiry:

First, we identify the ‘operative incident … giving rise to liability in this factual context.’ Second, after identifying the operative incident or incidents, we consider ‘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 causal continuum, without intervening agents or factors.’

The court acknowledged that assumption only in passing, by stating:

The three incidents here share a common origin: the initial negligence that caused the dump truck’s collision with the overpass.

In reality, the court assumed that all three collisions shared both a common “origin” and a common proximate cause: if the driver’s “initial negligence” had not been the proximate cause of the Itzkowitz and Compton collisions, then the dump truck’s owner would not be liable for them. Yet the court never noted that assumption in the ensuing inquiry into “causal continuum.”

Keep on Trucking

To identify the operative incident giving rise to liability, the court consulted Appalachian Ins. Co. v. Gen. Elec. Co., 863 N.E.2d 994 (N.Y. 2007). In that case, New York’s high court held that each individual’s exposure to a common source of asbestos could be treated as a different “operative incident.” It also warned:

Common causation is pertinent once the incident—the fulcrum of our analysis—is identified, but the cause should not be conflated with the incident.

Heeding that warning, the Second Circuit concluded—without further explanation—that each of the three collisions on I-90 was a distinct operative incident.

Next, the court considered the spatial and temporal proximity of the three incidents, and whether they could be considered part of a causal continuum. The court declined to draw a line in the asphalt defining “any particular number of seconds or minutes that must elapse before two incidents are distinct accidents.” Instead, the court inquired “whether the relative timing of the various incidents played a role in causing any of the incidents.” In this case, the court found that timing was irrelevant, because the dump box remained a hazard for as long as it sat in the road:

No evidence in the record suggests that the short timespan between the dump box’s collision with the overpass and the Itzkowitz vehicle’s collision with the dump box played any role in the Itzkowitz vehicle’s collision with the dump box.

The court considered spatial proximity was a “closer question.” The two collisions with the dump box were “spatially proximate,” but not so the collision with the overpass, which occurred “farther down the road.” This limited proximity was deemed insufficient; the court applied a “’common sense’ balancing” of all three of the unfortunate event test’s criteria, and proximity was found wanting.

Lastly, the court considered whether the collisions were part of a causal continuum—an inquiry that ostensibly focused on the concept of intervening factors, but which actually appears to have snuck temporal proximity in through the back door. As the court explained, the fact that all three events had a “common origin”—the negligence of the dump truck driver—was not dispositive.

[C]ommon causation, while relevant to our inquiry, is insufficient to aggregate incidents into one accident. Instead, we look to whether there was an ‘unbroken’ continuum between the events. To be part of the same accident, the operative incidents must be part of the same causal chain. Once an incident occurs and that incident does not then cause further injury, the causal chain is broken.

In this case, the court ruled, the collision with the overpass “did not immediately cause further damage,” even though it caused the dump box to land in the road. On that basis, the court distinguished cases involving “chain-reaction” collisions, in which a vehicle ricochets off one car into another. Instead, the court decided Mr. Itzkowitz’s collision with the dump box started a “second causal chain,” and that the ensuing collision started a third one.

We would be facing a different set of facts if the third incident … occurred because of the Itzkowitz collision; if, for example, the Itzkowitz vehicle had ricocheted off the dump box before hitting the [other] vehicle. There might then have been an unbroken chain between the second and third collisions. But that is not what the record indicates. The second and third incidents were therefore not part of the same unbroken continuum.

It seems fair to ask, however, whether a ricochet is really the only way in which two or three collisions might be causally related. The collision with the overpass was certainly a but-for cause of the later accidents, and it was at least arguable that the judgment of the drivers in those accidents was affected by the relative improbability of encountering a dump box in the middle of an interstate highway. The Second Circuit’s opinion did not take up those questions.

Nor did the court explain how the dump truck’s owner could be liable for accidents that were not part of the “same causal chain” as the driver’s negligence. If the driver’s negligence really “[did] not then cause further injury,” why was the driver’s insurer still on the hook?

Thinking Outside the Dump Box

The court concluded that three separate accidents had occurred for purposes of coverage. Thus, at least in those courts bound by Second Circuit precedent, insurers should carefully consider the issue of aggregation in multiple-claimant incidents, as courts may not be willing to dump all the liabilities into a single per-accident limit.