Win with COVER & Emperor Asset Management


A requirement for a causal link between policy terms and the loss?

Recently, the Short-term Ombudsman has been requiring that insurers, should they repudiate a claim on the basis of a breach of a policy term, demonstrate a causal link exists between the breach of the term and the loss.   In adopting this stance, the Ombudsman received considerable support from Prof JP Van Niekerk (Unisa) who published his views in an article in the South African Mercantile Law Journal (2010), entitled: “The Requirement of a Causal Link between the Insured’s Breach of a Term in the Insurance Contract and the Insured’s Loss: An ‘Attractive Feature’ of South African Insurance Law?”

The purpose of this article is to answer the question, ‘Is a causal link required as a point of the existing law?’  Or simply is the link required as a matter of law?

The problem of insurance terms

It is useful to have some insight as to what is perceived to be the problem with policy terms and insurance claims, in order to understand the Ombudsman’s requirement.    The two cases usually referred to in this context will demonstrate the perceived problem.

Jordan v New Zealand Insurance Company 1968 2 SA238 ECD

Concern has been expressed for some time by academics about the use of insurance terms in policies to repudiate claims, especially where terms are declared to be warranties.   Thus, in 1965, the late Professor Boberg (Wits) published an article ‘Insurance warranties are trumps’ in the South African Law Journal where he essentially argued that via the use of warranties insurers, on immaterial grounds, can avoid paying otherwise valid claims.   The public were being deluded into thinking they had cover, when, in fact, they had none.  The industry became aware of his views and took exception.   Shortly thereafter, the case of Jordan v New Zealand Insurance Company took place.  Jordan, when completing a proposal form for motor insurance, incorrectly stated his age; the difference was, however, inconsequential and, in any event, favoured the insurer.   Jordan was older than he claimed to be.   If age is used as a risk rating factor, the slightly increased age would have resulted in a lower premium had the insurer rated the risk using age as a risk factor.  It was agreed that Jordan’s error was in law immaterial to the insurer.   Nevertheless, the insurer repudiated the claim as Boberg had predicted insurers would do.  The insurer’s decision was confirmed by the court in Jordan v New Zealand Insurance Company 1968.  The insured was clearly in breach of the term in which he warranted the information to be correct.  The information was, in fact, incorrect.  All that matters, in law, is that the warranty was breached and breached it was.   Boberg’s statement about the conduct of insurers turned out to correct.   The matter was referred to Parliament and Parliament was not impressed.  The Insurance Act was amended with the insertion of s63(3) which henceforth precluded insurers from repudiating claims for breach of warranty where the incorrect misrepresentation was immaterial to the determination of the risk.  In 1968, Prof Boberg wrote a second article on trumping warranties.

So the perceived problem becomes clear.  Insurers repudiate valid claims for immaterial, mere technical reasons.

SA Eagle Insurance Co Ltd v Norman Welthagen Investments (Pty) Ltd 1994 2 SA 122 A

In this more recent case, the insurer repudiated a claim for the breach of a term in a multi-peril policy in terms of which the insured warranted to keep all vehicles stored in the open, locked and to keep the keys of the vehicles in a locked safe.    A vehicle was stolen and a claim submitted. It then was discovered that the keys were kept in a locked office in a cupboard and not in a locked safe as required and warranted by a term in the policy.   The keys played no part in the theft of the vehicle and were found safely stored in the locked office in the cupboard after the theft.  The vehicle would have been stolen even if the keys were kept in a locked safe.  No causal connection existed between the failure to lock the keys in the safe and the theft; nevertheless, the insurer repudiated the claim on the basis that the insured had breached the term in the policy.  To breach the term, it was not necessary to show a causal link.  The trial court found in favour of the insured and the matter was taken on appeal to the Appellate Division where an attempt was made to bring the matter within the purview of s63(3).  The attempt failed and trial court’s decision overturned in favour of the insurer.

So, once again, it was shown that insurers were repudiating valid claims for factually immaterial reasons.   Academics expressed the opinion that s63(3) (now re-promulgated as s53(1) in the Short-term Act and s59(1) in the Long-term Act) did not go far enough and further legislation was called for.

Proposal to ‘solve’ the perceived problem

Against this background, how are claims to be dealt with involving a breach of a contract term?  It seems attractive to hold that a causal link should exist between the breach of the term and the loss.  If this ‘solution’ is applied to the second case, the question becomes, “Did the failure to keep the keys in the locked safe cause or contribute to the theft of the vehicle?”  Clearly, the answer is no, in which event, if a causal link is required, the insurer cannot repudiate the claim because of the breach of the term.  However, if applied to the first case, the Jordan case, the matter is not at all clear.  It is difficult to see how anyone can prove one way or another that age caused any specific insured event.   The causal link ‘solution’ does not seem to assist in any way in the first case.  In the first case, a more appropriate test would be was the insurer prejudiced by the mis-representation.

That the proposed link test does not work in both cases is important because it suggests that the proposed test is not of universal application and thus as a test cannot be applied to all claims.  It is not clear to which claims the test can in fact be applied.

Van Niekerk’s article

With this background Van Niekerk’s article and the Ombudsman’s reliance on it is now examined.

I have repeatedly pointed out that, in recent years, a characteristic of our modern judiciary is the almost complete absence of law and the substitution for the law of what I have called the Rule of Jolly-Good-Ideas.  Judges (and academics) no longer make decisions in terms of the existing law, but in terms of their own moral lights.   This as I have previously pointed out is a return to Greek system of Themis which existed before the Rule of Law.

So, when reading Van Niekerk’s article, I suggest the question must be kept in mind is: “Is he expressing a view on the existing law or expressing his view as to what he thinks or would like the law to be?”     Now even the most cursory reading of his article leads to the conclusion that the idea that a causal link must exist between the breach of the policy term and the loss is his idea of what the law should be. It is not what the law currently, in fact, is.   His article makes it clear the current law is, in fact, is exactly the opposite to what he is suggesting.   The current law is as indicated by the above two cases.  A breach of a warranted term in the policy is clear grounds for repudiation.  No link is required.  Van Niekerk argues for an alternative, the link approach.  He concludes to get to his position, “Although the causal requirement should ideally be taken up in legislative reform … it can be argued that this is not necessary.   All that is required to put the matter beyond doubt is an appropriate … case … and a suitable pronouncement by that court on the issue.”  In other words, judges should endorse his idea and convert his idea into legal practice.  It is interesting to note the diminished role of Parliament.  It has been completely abrogated and replaced with judges.  Its role as the sovereign law-making institution which existed in 1968 when dealing with the Jordan matter has been replaced by judges.  So, to change the law, it is not necessary to take it back to parliament, merely take it to a judge.  Themis will guide the judge to endorse Van Niekerk’s idea.

It thus can be accepted that, at present, as a matter of law, there is no general requirement that a link exist between the breach of a policy term and the loss.  As indicted when the mere two cases discussed above are considered, that the link theory is, in any event, not universally workable in practice.  It does not meet the requirement of a being a law of general application.

Short-term Ombudsman’s requirement

Having come to the conclusion that Van Niekerk’s views do not, or, claim to, reflect the current legal position, where does that leave the industry with the Ombudsman’s requirement that a link must exist between the breach of the policy term and the loss?   Although one can be very sympathetic to what the Ombudsman is attempting to achieve, the conclusion is inevitable; His requirement is not in terms of the existing law.   It is thus concluded that as matters now stand the Ombudsman’s requirement cannot be sustained and this places individual insurers in a very uncomfortable position.    It can also be shown that a requirement causal link exist cannot be sustained in terms of equity.

This conclusion does not solve the perceived problem of repudiating claims for a breach of a policy term as highlighted by the above two cases.   It is suggested that a workshop be convened with claims managers of the insurance companies to discuss this issue and to decide what action, if any, is needed to address the perceived problem.

Related posts

Hollard calls on all social partners to collaborate more to reduce youth unemployment


Here’s how the new drunk driving law will affect your car insurance


Tokyo Sexwale: Another scammed Farmer?


Consumers call for a more digitised InsurTech industry