One of the thorniest issues to deal with in insurance coverage matters is that of “concurrent causation”. The problem was illustrated again on the national stage last year when Albertans experienced massive overland flooding that also resulted in sewer back up. Typically, homeowner’s policies in Canada do not cover damage from flooding but many will include an endorsement for sewer back up. The events of 2013 presented insurers and insureds with the problem of how to handle losses directly caused by a sewer back up that is, itself, caused by an excluded flood.

Serial concurrent causes vs. Independent Concurrent Causes

A starting point is to distinguish between two different types of concurrent causes: Serial and independent.

Serial causes are those where each is a consequence of the one preceding it; where, in other words, there is a causal connection not only between each cause and the loss, but also among the various causes. This is usually described as a “chain of causation”.{1}

Independent causes, on the other hand, are unrelated. The fact that they operate simultaneously to produce the loss is a mere coincidence of time and situation. There is no causal connection between independent concurrent causes; the causal connection exists only between each of those causes and the loss.

Independent Concurrent Causes

The leading case in Canada on the issue of independent concurrent causes and exclusion clauses is the Supreme Court of Canada case of Derksen v. 539938 Ontario Ltd.{2}

In Derksen, the insured had two policies: an automobile liability policy which was restricted to non-pecuniary damages, and a commercial general liability policy (the “CGL”) which contained a typical exclusion for bodily injury or property damage arising out of the use or operation of a vehicle.

While cleaning up a work site, the insured negligently stored a steel plate on the truck. While driving the truck, the steel plate went through the window of an oncoming school bus, killing one child and injuring others.

In this case, the independent concurrent causes are: 1) the negligent clean up of the work site, and 2) the negligent use and operation of the truck.

Both the Ontario Supreme Court and the Court of Appeal agreed that under the CGL policy, coverage should apply as the injuries arose from both auto-related and non-auto-related causes. As the non-auto-related cause is covered by the CGL policy, the insurer was obligated to cover the insured for damages arising from that loss.

In the Supreme Court of Canada the court reaffirmed that the issue of exclusion is a matter of contractual interpretation. A presumption that coverage is excluded is inconsistent with the well-established principle that exclusion clauses in insurance policies are to be interpreted narrowly and generally in favour of the insured in case of ambiguity in the wording. The court noted that exculpatory language was available to the insurers. Examples included “caused directly or indirectly”, or “caused by, resulting from, contributed to or aggravated by”. Such language would exclude coverage for a loss with several independent concurrent causes.{3}

As there was no exculpatory language in the CGL’s auto exclusion, the Court affirmed the lower court decisions that the CGL had to respond to the pecuniary loss and that portion of the loss attributable to non-auto negligence.

Policies need to have clear exculpatory language in order to deny coverage in a situation involving multiple independent concurrent causes, one of which is excluded under the policy.

Serial Concurrent Causes

The second type of concurrent cause scenario is when a chain of events causes a loss or damage.

In a situation involving serial causes, there is in reality only a single cause involving one chain of causation. The causes are not genuinely concurrent as in Derksen which involved two chains of causation. In the case of serial causes, there is the single dominant cause and mere derivative causes in a single chain of causation. And if that single dominant cause is excluded from coverage, then there should be no coverage for the loss even in the presence of covered, derivative causes.

In the case of the 2013 Alberta floods, there is a strong argument that because the sewer back ups were caused by overland flooding, and the latter is clearly excluded, then there is no cause of loss independent to the excluded cause, Derksen does not apply, and there is no coverage for the loss.

Of course, there is no pronouncement equivalent to Derksen from the Supreme Court of Canada on the issue of serial concurrent causation. But the matter has been dealt with by lower courts.

In Pavlovic v. Economical Mutual Insurance Co.{4}, the water service line ruptured at the insured’s residence. After an indeterminable period of time, the insured’s brick and concrete work at the front of the residence and driveway showed signs of water damage. The policy excluded coverage for seepage and leakage of water below the surface of the ground.

The trial judge held that there was no coverage. The Court of Appeal overturned the decision by ruling that many events conspired to cause the damage, one of which was leakage of underground water, but the chain of events was set in motion by the rupture of the water service line. Therefore, it would be unfair to say that the loss was “caused by” the leakage of water below the ground. In order to succeed, the Court of Appeal stated, the insurer must prove that the events causing the loss or damage are clearly within the language of the exclusion.

Implicit in the Court of Appeal decision is the finding that the dominant or true cause of the damage must be excluded from the policy in order for an insurer to deny coverage in a serial concurrent cause incident.

In another BC Court of Appeal, Catalano v. Canadian Northern Shield Insurance Co.{5}, the local government diverted a flooding creek down the insured’s street causing damage. The insured’s policy excluded coverage for damage caused by floods, natural or man-made. Both the Supreme Court and Court of Appeal agreed that the cause of the insured’s loss was not the diversion, but the flood that necessitated the diversion.

The Court found reinforcement in the language of the exclusion that included the words “directly or indirectly”, noting at paragraph 16, “Those words typically are intended to broaden the ambit of causation between the peril and the loss and avoid a rigid and narrow proximate cause analysis.”

Serial but not Concurrent Causes

There may be situations where a court may interpret a loss caused by a series of events as a loss that was caused by only one event, thereby dismissing any concurrent cause claim. In Algonquin Power (Long Sault) Partnership v Chubb Insurance Co. of Canada {6}, a dam had failed. There were concurrent causes – a faulty design, which was excluded from coverage, and other causes including a sink-hole collapse. The court held at paragraph 205:

This is not a case where there was more than one cause for the dam’s failure. The only cause was faulty design. What we have is a chain of events leading to the compromise of the dam. Faulty design caused hydraulic piping. This led to voids that caused surface sinkholes, cracks and slumps that compromised the dam. The Partnership simply has not established any other concurrent cause for the event.

The result in Algonquin appears to be the same as when a court attempts to find the underlying cause of the loss. However, the court in Algonquin stated that the series of events and resulting loss was contributed solely by the underlying cause and no additional factors contributed to them.

Apportionment

In the above cases, coverage for the insured’s loss was considered as an all or nothing approach. Some academic commentators believe that in situations where there are concurrent causes, the apportionment approach should be adopted. Basically, the court would be invited to divide the damage proportionally between covered and non-covered losses. While no jurisdiction has adopted this approach yet, it may be considered by a court in the future.

Conclusion

Much of the controversy surrounding concurrent causation ought to be avoided with clear, exculpatory language. But as environmental claims mount from catastrophic, extreme weather events, the efforts of insurance companies to predict and control their exposure in the event of large claims will come up against immense social and political pressure to pay out notwithstanding whatever language is contained in the policy. That was certainly the experience in Alberta in 2013 as angry homeowners sought, and in many cases obtained, coverage for sewer back ups notwithstanding exculpatory language in their policies.

Ultimately, insurance companies may respond with higher premiums with the unfortunate and unintended consequence of more homeowners running the risk of remaining uninsured.