This article was first published in Law Talk on 1 November 2013
Insurance law has its own way of thinking about causation. Anyone instructed in respect of an insurance coverage dispute would be well-advised to gain awareness of this perspective.
Mention of an insurance company relying on an exclusion of liability to decline a claim sometimes meets a knowing nod and a sigh from members of the general public. Anyone who believes they had a valid insurance claim, in circumstances where in fact they do not, deserves sympathy. However, it pays to understand that this expectation may be based on a misconception about how insurance policies work, and the specific effect of the language used, particularly with reference to exclusion clauses. The “missing link” in people’s understanding is sometimes the requirements of causation, as it is reflected in the language of the policy. This is a needy concept that deserves more care and attention than it generally receives.
Insurance companies are in the business of earning premium and paying claims. How this fits together in a way that works commercially is a matter of underwriting. This should enable the company to pay all eligible claims and at the same time make a profit by investing the difference between all of the premium collected from customers less the amounts it must pay in claims. A properly conducted underwriting analysis should set the premium at a level which makes into account the probability of a particularly event occurring giving rise to an eligible claim, and also taking into account other factors. The relevant probabilities are a matter of actuarial analysis.
This analysis is reflected in the language used in an insurance policy. Some risks will be met, others will not. In respect of risks which are excluded, it is not necessarily the case that there needs be a direct causal link between that risk and the loss or damage the customer seeks to cover. That is a product of the language used in the agreement. This language will usually be selected deliberately to reflect that, from an actuarial perspective, it is difficult or perhaps impossible for an insurer to gauge the likelihood of the risk occurring. It is not generally reflective of a desire by an insurance company to shirk its responsibilities.
The lesson to be learnt is that wording of the policy is absolutely central to a properly conducted analysis. A view of the world that an insurer has deep pockets and a customer is being hard done by where a claim is excluded will not greatly assist the client. As set out below, in an insurance context, causation is dominantly a matter of contractual analysis. A judge will generally not be moved to depart from this analysis without good reason.
When does one thing can be said to have caused another is a question of almost universal application in civil litigation practice. At law school we are raised on a steady diet of common law tort, where the purpose of the inquiry is to identify who is – and who is not- at fault. This attribution of responsibility is the key area of interest. This sometimes leads to a subtle blending of causation and duty/ breach of duty concepts. Further, the characterisation of causation may sometimes be motivated by considerations of fairness and public policy.
On the other hand, an insurance coverage dispute by its nature is concerned with an insurance contract where the key terms are contained in an insurance policy. The Court will therefore ordinarily limit its inquiry to what the policy says about causation.
These different approaches may produce different outcomes. Consider the case of Miss Jay Jay  1 Lloyd’s Rep 32. If a yacht has been negligently designed or built, and then breaks up while at sea, in calm waters, the tort lawyer is interested in whether a breach of duty by either the designer or the manufacturer, or both, can be said to have caused the damage. In Miss Jay Jay the Court was required to address causation in the context with insurance against perils of the sea. Once the Court had located a proximate cause – the incursion of seawater (being the relevant peril), the inquiry ended.
To what extent the Court was required to keep searching depending on the provisions of the policy. So, where there are a range of potential causes “if one of these causes is insured against under the policy, and one of the others is expressly excluded from the policy, the insured will be entitled to recover” (an extract from Halsbury, quoted with approval in Miss Jay Jay at 36 and 40). It followed that “where there were no relevant exclusions or warranties in the policy the fact that there may have been another proximate cause did not call for specific mention since proof of a peril which was within the policy was enough to entitled the plaintiffs to judgment” (Miss Jay Jay, page 37). On this view of things, any negligence by the designer or manufacturer does not make any difference to the policy analysis.
So far so good. But in fact “causation” is not always a term that is expressly identified in a policy, rather, it is a concept implicit in the words used and the approach towards interpreting them.
In the past, occurrence based insurance policy would helpful listed out the “perils” to which the policy would respond if one of those perils caused accidental damage: fire, tsunami, a plague of locusts and so on. Such a policy may have also contained a list of exclusions. So, the inquiry included a consideration of whether something was “in” (ie there was a listed peril which was the proximate cause of the damage) or “out” (excluded). Nowadays, an occurrence based insurance policy is more likely to be formulated on an “all risks” basis for accidental loss or damage. That is to say something should be covered, other things equal, unless the loss or damage is caused by something that is excluded. The causation focus is therefore more on exclusions.
The causation focus also tends to be more on exclusions with regards to liability policies. Liability policies ordinarily respond to claims made during the policy period (and are subject to other terms and conditions outside the scope of this brief paper). Other things equal, either there is a claim to which the policy responds or there is not. Any causation issues are more likely to focus on whether the claim is excluded.
All of the above is most usefully demonstrated by example, with reference to the recent case of IAG v Jackson  NZCA 302. According to the judgment, a Christchurch couple, the Marchands, engaged an insurance broker, Mr Jackson, to arrange insurance cover for their home and contents, motor vehicles and a medical practice. Mr Jackson placed business interruption cover for the medical practice, but he arranged none of the other insurance. This was initially an oversight. On several occasions he assured the Marchands he had placed cover with NZI whereas he had not. The Marchands’ home was badly damaged in the 4 September 2010 Canterbury earthquake. Mr Jackson then attempted to arrange cover with NZI by submitting an application form which he dated 30 August 2010.
The Marchands sued Mr Jackson for their uninsured loss. Mr Jackson joined IAG, his professional indemnity insurer. IAG applied for defendant’s summary judgment relying, amongst other grounds, on an exclusion for dishonesty. The Marchands succeeded at trial against Mr Jackson. IAG did not participate. An Associate Judge declined IAG’s application for summary judgment, concluding that dishonesty would need to be decided on the facts at trial. IAG appealed against that ruling resulting in the present judgment.
The IAG policy contained an exclusion which provided that Mr Jackson was “…not insured for civil liability in connection with any dishonest, fraudulent, criminal or malicious acts or omissions by you…”.
The Court of Appeal was required to decide what “dishonest” meant in this setting and what the connection needed to be shown between the insured’s dishonest conduct and his civil liability.
On the evidence, the Court concluded that Mr Jackson had no case to answer to IAG’s allegation that he acted dishonestly.
The next question was whether Mr Jackson’s dishonest acts or omissions were “in connection with” his civil liability to the Marchands. The question arose because it was apparently common ground that Mr Jackson did not act dishonestly when he first incurred a liability to the Marchands by failing to act on their instructions to secure cover for their home, contents and vehicles. The relevant dishonesty occurred later.
The Court stated that IAG was required to establish a nexus or relationship between dishonest conduct and civil liability. It noted that the dishonest act “did not need to be the direct or proximate cause of the civil liability, and it need not precede the liability in time” (para 29). However: ‘“in connection with” does demand some causal or consequential relationship between the two things in this setting” (para 29).
The Court considered whether IAG has discharged that burden. It had been agreed that Mr Jackson had acted honestly when he first failed to place cover. He incurred a liability for breach of this retainer at that time. However, the trial judge held that had Mr Jackson not hidden the truth from the Marchands, they would have secured cover before the earthquake. That being so, the required nexus was established. The exclusion therefore applied. The Court therefore permitted the appeal (the procedure aspects warranting separate consideration).
POST SCRIPT (added 30 December 2013): nzinsurancelaw is aware that Mr Jackson has requested leave to appeal in the Supreme Court.