back to news

Causation in Australasian Insurance Law

Newsletter Article 08 March 2011


In insurance law, causation is the identification of a sufficient link between an event which the policy recognises, and the loss for which indemnity under that policy is sought. The sufficiency of the link determines the policy’s response.

The challenges are to measure the sufficiency and, if there is more then one sufficient link, to determine how the policy responds to those many links.

The recent floods in Queensland revive a classic situation in which disputes about causation in insurance law often arise. They therefore provide a backdrop to a review of where causation in Australian insurance law has reached.

Floods invariably involve an invasion of water from various sources such as rivers, blocked drains or run-off, all of which can cause loss in one form or another. Insurance policies taken out to respond in the event of a flood, typically respond to some of those losses but not others.  Much can and will depend upon which source or sources caused the loss.

Causation Generally

In Australia, at least since the decision of the High Court in March v Stramare (E & MH) Pty Ltd [1], the test of causation at common law has been well established.

In Medlin v State Government Insurance Commission [2], the High Court of Australia had reason to consider causation in the context of a negligence claim. Deane, Dawson, Toohey and Gaudron JJ said:

‘The ultimate question must, however, always be whether, notwithstanding the intervention of the subsequent decision, the defendant’s wrongful act or omission is, as between the plaintiff and the defendant and as a matter of common sense and experience, properly to be seen as having caused the relevant loss or damage.’

In a separate judgment McHugh J said:

However, the ultimate question is whether, as a matter of common sense, the financial loss that the plaintiff has suffered was caused by the plaintiff’s act in resigning his office rather than by the defendant’s negligence.’

The High Court has rehearsed this principle several times since Medlin [3], most recently in Tabet v Gett [4] where Keifel J said:

‘The common law requires proof, by the person seeking compensation, that the negligent act or omission caused the loss or injury constituting the damage.  All that is necessary is that, according to the course of common experience, the more probable inference appearing from the evidence is that a defendant’s negligence caused the injury or harm.  “More probable” means no more than that, upon a balance of probabilities, such an inference might reasonably be considered to have some grave degree of likelihood; it does not require certainty.

The “but for” test is regarded as having an important role in a resolution of the issue of causation, although more as a negative criteria than as a comprehensive test.  The resolution of the question of causation has been said to involve the common sense idea of one matter being the cause of another.  But it is also necessary to understand the purpose for making an enquiry about causation and that may require value judgments and policy choices.

Once causation is proved to the general standard, the common law treats what is shown to have occurred as certain.  The purpose of proof at law, unlike science or philosophy, is to apportion legal responsibility.  That requires the courts, by a judgment, to “reduce to legal certainty questions to which no other conclusive answer can be given.”  The result of this approach is that when loss or damage is proved to have been caused by a defendant’s act or omission, a plaintiff recovers the entire loss (the “all or nothing” rule).’ [5]

At common law, it is common sense and experience (flavoured by any necessary policy or value judgment) which determines the existence of any causal link.

The Insurance Policy and its Interpretation

Insurance law offers an additional consideration to this general principle because there theparties’ relationship is governed by a specific contract, the insurance policy. So how are the terms of the insurance policy and the common law principle of causation reconciled?

In McCann v Switzerland Insurance Australia Limited [6], Gleeson CJ said:

‘A policy of insurance, even one required by statute, is a commercial contract and should be given a businesslike interpretation. Interpreting a commercial document requires attention to the language used by the parties, the commercial circumstances which the document addresses, and the objects which it is intended to secure.’ [7]

The first step down the path of reconciliation is the language which the parties to the insurance policy have adopted. It is the policy’s words rather than any broad overarching legal principle [8] against which the sufficiency of the causal link can then be measured. The link continues to be measured by common sense and experience as explained earlier. The policy’s words indicate how strong or sufficient the link must be.

There is a gloss. What happens when the terms of the insurance policy are imposed by statute?

Allianz Australia Insurance Limited v GSF Australia Pty Limited [9] concerned a claim for indemnity by an employer whose employee was injured in the course of unloading of trailer loaded with aircraft containers. The trailer was defective. For the relevant policy to respond, the employer had to establish that the defect in the trailer caused the loss. In addressing the question of causation in this matter McHugh J said:

‘In the end, the outcome of this appeal turns on the construction of the words “caused …. by a defect in the vehicle”.  The language of the Act reflects the concept of causation at common law.  This suggests that the inquiry into the question of causation under the Act does not differ materially from the “common sense” test for causation at common law.  However, because the task before the Court is one of statutory construction, the question of causation must be determined in light of the subject, scope and objects of the Act.  The common law concept of causation is concerned with determining whether some breach of a legal norm was so significant that, as a matter of common sense, it should be regarded as a cause of damage. In the present case, however, common law conceptions of causation must be applied having regard to the terms or objects of the Act. Those terms and objects of the Act operate to modify the common law’s practical or common sense concept of causation. The inquiry into the question of causality is therefore not based simply on notions of ’common sense’.  In NRMA Insurance Ltd v NSW Grain Corporation, Clarke JA said that the Act compels a ‘common sense’ approach to the question of causality of the injury (as prevails in relation to the common law):

[The Act] propounds an inquiry on causation identical with that undertaken in determining whether the negligence of the person claiming indemnity ‘caused’ the damage and thus was liable for it. That test has now been firmly based on common sense… It would not be reasonable, in my opinion, to attribute to the legislature an intention that the expression ‘caused by’ in the statutory policy should enliven a test of causation different from the test by which the party claiming indemnity had been found liable. 

However, the purpose of the inquiry must be ascertained before the application of any notion of ‘common sense’. The purpose of the causal inquiry is critical because it conditions the result. Once the purpose of the inquiry is ascertained, the question of causality must be determined in light of the subject, scope and objects of the Act.  Both Mason P and Santow JA acknowledged the importance of considering the purpose of the causal inquiry because the purpose ‘conditions the outcome of any application of common sense to its answer’.’ [10]

In a separate judgment, Gummow, Hayne and Heydon JJ accepted the need to identify immediately the purpose of the legislation [11].

The gloss therefore is that if the policy’s terms are a creature of statute, the subject, scope and objects of the governing Act need to be identified at the outset and the terms (which indicate the strength or sufficiency of the causal link) read in that context.

The Sufficiency of the Link


Insurance policies stipulate the sufficiency of the link between the loss and the event in different ways. The different expressions have their own distinct meanings as McHugh J observed in Insurance Commission of Western Australia v Container Handlers Pty Ltd [12]:

‘(I)n the context of the Act the expression "a consequence of" emphasises the result or effect of the driving rather than the driving causing the result. This distinction is important in an insurance context where cause is frequently — perhaps usually — equated with "proximate" or "dominant" cause. Although "consequence" involves notions of causation, the term "consequence" — with its emphasis on effect — places less emphasis on the proximity of cause and effect than the term "cause" may do in various contexts.’

Proximate Cause

The preferred way is to use the word (or an extension of the word) “cause”. Here the required “cause” is not a cause at large. It is the proximate cause of the loss in the sense that it is the most efficient cause of the loss.

On 20 September 1995 Professor Martin Davies delivered a paper at a seminar of the Western Australia Branch of the Australian Insurance Law Association in Perth [13].  The article was entitled ‘Proximate Cause in Insurance Law’ and after reviewing the relevant authorities on this issue as at the date of the article, Professor Davies lamented that:

‘It is difficult to state any useful conclusion about the law relating to proximate cause.  There is not and never has been any need for its principles to be baroquely complex, but those principles are too firmly established to be changed without a paradigm shift that would cause massive reverberations throughout the insurance industry.  Because reform seems impracticable, one is left with the anodyne observation that a subject made difficult should at least be treated as if it were difficult.  To solve complex causal questions by the application of undiluted “common sense” is to underrate the possibilities of analytical thought.  Despite the doubt expressed by Windeyer J in [Commonwealth v Butler] [14], it is (or should be) possible to explain causation and consequence, rather than merely “feeling” one’s way to an answer. Insurance litigants, both insurer and insured, are entitled to demand the problems of proximate cause be approached with at least an attempt at intellectual rigour.’

Notwithstanding this problem in the application of legal principle, the Courts have continued to ‘feel their way to an answer’ at times sweeping away answers that other judges have reached.  Here are some examples.

State Government Insurance Commission v Sinfein Pty Limited [15] concerned injuries which one Miller suffered as a result of a fire in a paddock which he and others were harvesting. He had driven a wheat truck over a wheat stubble and, in so doing, ignited the stubble. A fire truck on hand to put out any fires proved to be ineffective. Upon attempting to refuel the fire truck, Miller spilt some fuel on the ground and himself. He suffered burns when that fuel ignited.  He successfully sued the owner of the wheat truck for losses suffered as a result his burns. The owner sought indemnity from the insurer of the wheat truck.

The insurance policy held in respect of the wheat truck provided insurance ‘against any liability which may be incurred by (the) owner … in respect of the death of or bodily injury to any person directly caused by, or by the driving of, such motor vehicles’ (our emphasis).

There were a number of potential causes of the losses for which the employer was liable. They included the condition of the wheat truck that Mr Miller was instructed to drive which rendered his place of work unsafe, Mr Miller’s employer’s failure to instruct him as to the safe routes for him to take and defects in the fire truck used to douse the fire. The task before the Court was to determine from all of the events which common sense and experience said caused the losses, which was the most efficient cause.

The majority of the Full Court of the Western Australian Supreme Court held that, the proximate cause was not the driving of the wheat truck.  As that was the contrary to what the insurance policy required, indemnity was not available.  Indemnity under the insurance policy was therefore not available.

In HIH Casualty & General Insurance Limited v Waterwell Shipping Inc & Anor [16], the New South Wales Court of Appeal considered the possible causes of the sinking of a ship while at berth in Kenya including the failure to keep the ship watertight and evidence of corrosion of the mud (strainer) box. The relevant policy provided that the insurer was liable ‘for any loss proximately caused by a peril insured against, but, subject to the aforesaid, he is not liable for any loss which is not proximately caused by a peril insured against’.

While the various explanations for the sinking of the ship were put forward both at trial and before the Court of Appeal, the Court of Appeal accepted that applying common sense standards, that  the proximate cause for the sinking of the ship was the failure to seal a number of sea valves.

Lasermax Engineering Pty Limited v QBE Insurance [17] concerned a fire on a power pole approximately 55 metres from the appellant’s premises. It caused the upper arm on the burned pole to fail, the wiring on the upper arm came into contact with the wiring on the lower arm thus causing a high voltage/low voltage intermix. The intermix caused a power surge to the premises which caused damage to the laser.  The relevant policy provided that cover was available up to the limit specified in the schedule in respect of physical loss or damage to contents directly caused by fire.

The New South Wales Court of Appeal determined that it was the fire that was the proximate cause for the damage to the contents of the insured’s premises. The relevant policy therefore responded.

This approached has been subsequently entrenched in McCarthy v St Paul International Insurance Co Ltd  [18] (‘McCarthy’), Caine v Lumley General Insurance Limited [19], Zotti v Australian Associated Motor Insurers Limited [20] and El Hayek v Vasic [21].

Other Expressions of Causation

Typically insurance policies use the word ‘cause’ or variations of it.  Other expressions used include ‘arising out of’ and ‘brought about’. Where they are employed, the sufficiency of the link is weaker.

In Dickinson v Motor Vehicle Insurance Trust [22], the High Court of Australia commented upon the words ‘arising out of’ in the following terms:

‘The test posited by the word “arising out of” is wider than that posited by the words “caused by” and the former, although it involves some causal or consequential relationship between the use of the vehicle and the injuries, does not require the direct or proximate relationship which would be necessary to conclude that the injuries were caused by the use of the vehicle:  State Government Insurance Commission v Stephens Bros Pty Limited (1984) 154 CLR 552 at 555.’

The decision of the High Court of Australia in McCann v Switzerland Insurance Australia Limited concerned a claim by a firm of lawyers where there had been allegations of dishonesty. This was a case where the policy would not respond if the alleged dishonest conduct caused the loss. The expression used here was ‘brought about’. Gaudron J observed:

‘The different expressions used in the various subclauses of the exclusion clause convey different shades of meaning or different degrees of causal connection. This notwithstanding, each subclause must be read as a whole and read in its own particular context. One thing, however, is common to each of the subclauses:  as the Court of Appeal pointed out, the subclauses are concerned to specify a causal connection between particular acts or events and the liability of the insured, not with the loss sustained.’

More Than One Sufficient Link

The analysis becomes more complicated when more than one causal link is identified.

If, as this article has discussed, only one is sufficient for the purposes of the insurance policy, that link will determine the policy’s response. But what if there is more than one sufficient link and one link triggers the insuring clause while the other triggers an exclusion?

Petersen & Ors v Union des Assurances [23] concerned premises which had been damaged by water following torrential rains. Some of the water had reached the premises via flooded drains and some water by way of an unimpeded overland route. There were therefore two concurrent causes of the damage. One was the water flowing from the drains. The policy covered that damage. The other was from an unimpeded overland route. The policy excluded cover for that damage. But they operated together to damage the insured’s premises.

The New South Wales Court of Appeal upheld the trial judge’s conclusion that as one cause was specifically excluded under policy, the entire claim was excluded. The New South Wales Court of Appeal subsequently endorsed this approach in Waterwell.

In McCarthy [24], the Full Federal Court of Australia looked closely at the role of insurance in circumstances where only part of the claim fell within the cover which the policy afforded while the balance did not (either because it was excluded or because the policy simply did not respond).  In those circumstances, the Full Federal Court said that it was a matter of close analysis of the relevant policy wording to determine how it responded in the face of multiple causes (whether they be proximate or otherwise). Allsop J (as His Honour then was) said:

‘In each such case the solution was seen as an application of the revealed contractual intentionof the parties. The scope of the insurance cover is identified by reading the policy as a whole (insuring clause and exclusion, in particular) and appreciating that loss caused in a particular way is excluded.  Given that the two causes are interdependent and that the loss would not have occurred without the operative effect of the excluded cause, the non-response of the policy can be comfortably and logically accepted as the intended result of the revealed agreement of the parties.

More difficulty may be encountered in circumstances where a policy excludes one cause, includes another and the loss is occasioned by the two causes operating concurrently, but independently, in the sense that each would have caused the loss without the other. At the outset, it may doubted that the solution in any given case is to be found in the application of any principle of insurance law, other than one which states that the rights of the parties to the policy are to be determined by reference to the terms of the contract as found. This was the principle applied by all three Lords Justices in Wayne Tank. Thus, it is always essential to pay close attention to the terms of any policy and the commercial context in which it was made, for it is out of these matters that the answer to the application of the policy to the facts will be revealed.’

If the concurrent causes were dependent upon each other but one is excluded, the exclusion of the entire claim is more likely than not to conform to the parties’ intent. If the causes are independent of each other, the outcome will vary from one policy to the next and depend upon a very close consideration of the policy’s terms.


In Australia, causation and related causal problems in the operation of an insurance policy will largely be determined by the terms of the insurance policy.

Beyond that it is a matter of common sense and experience. Exactly what that involves takes us back to the words of Justice Windeyer in Butler’s Case [25]:

‘Attempted explanations of causation and consequence can, I feel, be as unhelpful and unhappy as definitions of reasonable doubt’

Professor Davies pleaded in his article for a more concrete process. For fifteen years his pleas remain unanswered. Then there came a flood.

[1] (1991) 171 CLR 506
[2] (1995) 182 CLR 1 at 6 and 21
[3] For example Chappel v Hart (1998) 195 CLR 232 and Roads and Traffic Authority v Royal (2008) 245 ALR 653
[4] (2010) 240 CLR 537
[5] (2010) 240 CLR 537 at 578
[6] (2000) 203 CLR 579
[7] (2000) 203 CLR 579 at 589
[8] Contrast the statement in paragraph 21 of Volume 8.7 of the Laws of Australia which requires a relevant event to be ‘the dominant, effective or operative cause of the loss’.  That probably overstates the position and is insufficiently flexible to embrace the language that the parties to the contract may choose.
[9] (2005) 221 CLR 568
[10] (2005) 221 CLR 568 at 581 to 582
[11] (2005) 221 CLR 568 at 596
[12] (2004) 218 CLR 89 at
[13] A version of that paper was subsequently published in Volume 7 (1995) Insurance Law Journal. 
[14] (1958) 102 CLR 465 at 479
[15] (1996) 15 WAR 434
[16] (1998) 43 NSWLR 601
[17] (2005) NSWCA 66
[18] (2007) 157 FCR 402
[19] (2006) NSWSC 337
[20] (2009) NSWCA 323
[21] (2010) NSWSC 634
[22] (1987) 163 CLR 500 at 505
[23] (1995) 8 ANZ Ins Cas 61-244
[24] (2007) 157 FCR 402 at 434
[25] The Commonwealth v Butler (1958) 102 CLR 465 at 479

Mark Sheller, Partner

Newsletter Article 08 March 2011
back to news