Causation in Negligence Law — Unravelling the Difficulties
The aim of this article is to examine and order the underlying principles of causation and to describe the interrelationship of Civil Liability legislation on causation and the common law. The overarching goal is to provide a set of signposts for the profession to navigate to a resolution of the issue of causation in any given circumstance.
Causation in the law of negligence does not parallel causation in science or philosophy.[i] Tort law or Delict is about the allocation of fault or wrongdoing and consequently causes which science may determine produced a result may be ignored in judicial assessment of fault and blameworthiness
Even with an understanding of this caveat a lawyer seeking clarity of the law of causation in negligence, having waded through judicial analysis and the general provisions in Civil Liability legislation, is unlikely to distil a clear set of guiding principles. Not least because of the varied factual circumstances where multiple causes of harm are present. As Fleming notes in his seminal work on The Law of Torts ‘there is an increasing appreciation of the futility to elevate statements of law, made in relation to the particular facts of an individual case, to principles of general application’.[ii]
The application of overarching general principles, while providing assistance in the majority of causation issues, can and do lead in exceptional circumstances to absurd and unjust results.
A further cause of confusion is the multitude of labels found in case law when discussing causation. This is further complicated where those labels are attaching to the actions of plaintiffs, defendants, third parties, as well as simultaneous acts and consecutive acts. The labels referred to include: “causa sine qua non; causa causans; proximate cause; direct cause; operative cause; real and efficient cause; novus actus interveniens; ‘but for’ test.”[iii]
A more fundamental source of confusion arises with the intermingling of causation in fact with issues of causation in law, the latter encompassing remoteness of damage. It is unfortunate and apt to mislead that a question of whether a defendant could reasonably have foreseen the type of damage that occurred, is labelled under the heading of causation. The reasonable foresight test is applied to determine the scope of liability and to limit the defendant’s responsibility for damage. Put simply, not all damage caused by the defendant’s conduct is recoverable by the plaintiff, but only such damage as is reasonably foreseeable. Reasonable foreseeability is not a test of causation but of legal responsibility.
Who Bears the Burden of Proof
The question of who bears the onus of proof in establishing causation is put beyond doubt by Civil Liability legislation in Australia. “The plaintiff always bears the onus of proving on the balance of probabilities any fact relevant to the issue of causation”.[iv]
The “But For” test or “Causa Sine Qua Non”
This test is expressed by answering the question “but for” the negligence would the harm or damage have occurred. Or translated from the latin, causa sine qua non, is the negligence a cause without which the harm would not have happened. Or further, to use the language of Civil Liability legislation; “was the breach of duty a necessary condition of the occurrence of the harm”.[v]
At base, it is a test of factual causation only and while its shortcomings have been acknowledged judicially,[vi] it serves as a starting point[vii] for the analysis of the ultimate question of causation at law.
Nor can it be described as even a prima facie test or necessary precondition to a finding of legal causation since it is recognised and acknowledged in Civil Liability legislation that a court can find as a legal cause negligent conduct which was not a necessary condition of the occurrence of the harm.[viii] Underlying this legislative exception is recognition of the overarching duty of the court to address the tort issue of blameworthiness and fault in the assessment of legal causation. While a party’s negligence may not, in a factual analysis, have been necessary to cause the ultimate result, given that the same result would have occurred anyway, the degree of fault may justify a finding that the negligent conduct was a legal cause.
If the “but for” test is applied and a negative answer is given to the question “would the harm have occurred but for the negligence of that party”, this does not mean that, either automatically or even prima facie, the negligence will qualify as a legal cause. The only reliable conclusion from satisfying the “but for” test is that factual causation only is established.
Alternatively, if the “but for” test is applied and a positive answer is given to the question “would the same harm have occurred but for the negligence of that party” this again does not automatically disqualify the conduct from being held as a legal cause of the harm.[ix]
Finding legal causation where the “but for’ test satisfied
Negligent conduct may be the sole cause of damage and satisfy both the elements of factual causation (“but for” test) and legal causation (blameworthy conduct with no mitigating factors and the damage not too remote). The assessment of legal causation involves in addition to issues of culpability, the question of remoteness of damage. A tortfeasor will only be legally responsible where the damage caused was of a type that was reasonably foreseeable as flowing from the negligence.[x]
A trite example is where the driver of vehicle A ignores a Give Way sign at an intersection and thereby collides with vehicle B. There was no Give Way sign facing vehicle B as it entered the intersection. The elements satisfying both factual causation and legal causation are present. The damage would not have occurred to the plaintiff driver of vehicle B “but for” the defendant’s negligence, there being no other cause operating. The defendant driver’s conduct in failing to give way and ignoring the sign, is blameworthy with no mitigating circumstance. Nor is the damage too remote in that the defendant driver could reasonably foresee the type of damage that occurred, namely a collision with potential property damage and personal injury.
No Finding of legal causation where “but for” test satisfied
Civil Liability legislation in Australia refers to the elements involved in a finding that a breach of duty (negligence) caused particular harm.[xi]
The first element is simply factual causation established by examining the “but for” question, or to use the language of the Civil Liability Acts, whether the negligence “was a necessary condition of the occurrence of the harm”.[xii] The second element referred to in the legislation is the question whether the “scope of liability of the person in breach should extend to the harm so caused”.[xiii]
While factual causation may be established with respect to the negligence of a party, the examination of “the scope of liability” involving issues of culpability and reasonable foreseeability of harm, may, on balance, not be sufficient for the court to conclude that the negligence of that party should be found as a legal cause of the harm.
The following example indicates circumstances where factual causation is present but a court ultimately concludes that the negligence of a party (defendant in the example) should not be found as a legal cause:-
An intoxicated pedestrian walks onto a carriageway from between parked cars, into the path of an oncoming vehicle and is hit, thereby suffering severe personal injuries. The driver of the vehicle, while not speeding, failed to take evasive action soon enough, as expected of a competent driver which action would have avoided the collision.
The “but for’ test is satisfied both with respect to the victim’s gross negligence and the negligence of the driver, who, if he had taken reasonable evasive action could have avoided the collision with the pedestrian. A court in these circumstances, moving beyond factual causation to an examination of culpability and reasonable foresight of likely risk, may justifiably conclude that the contributory negligence of the plaintiff victim is so gross and culpable as to carry 100% of the blame for his own injuries. Such a finding is recognised in the Civil Liability Acts whereby “in deciding the extent of a reduction in damages by reason of contributory negligence, a court may decide a reduction of 100% if the court considers it just and equitable to do so, with the result that the claim for damages is defeated”[xiv]
Such a finding in the above scenario involves a conclusion that the degree of blameworthiness of the driver should be ignored as an operative cause. This finding may be justified on the grounds that one should not look too closely, nor set expectations too high when examining the conduct of a person presented with an unexpected emergency situation created solely by the negligence of another.
A rejection by the court of the defendant driver’s negligence as an operative cause, is facilitated in the Civil Liability legislation wherein:-
“For the purpose of deciding the scope of liability, the court is to consider (among other relevant things) whether or not and why responsibility for the harm should be imposed on the party who was in breach of the duty”.[xv]
Novus Actus Interveniens
A party’s negligence may be deemed “spent” as a causative factor in the harm done, due to the intervention of an overriding and unforeseen circumstance. This novus actus interveniens (new intervening cause) may be such as the court will find the operative cause of the harm despite the earlier negligence. A court may find the chain of causation, from the earlier negligence to the harm done, is broken by the novus actus interveniens even where the harm would not have occurred without the first negligence.
The finding that the later intervening circumstance is a novus actus interveniens is not an arbitrary choice by the court, but is based on an assessment of whether the first wrongdoer should have foreseen as likely to happen, the new intervening circumstance.[xvi] If the first wrongdoer should have foreseen the later intervening circumstance as very likely to happen, there is no room for a finding of novus actus interveniens breaking the chain of causation from the first wrongdoer’s negligence. This is so, since the risk created by the first wrongdoer’s negligence encompasses the foreseeable likelihood of the very intervening circumstance which occurred.[xvii] The chain of causation from the wrongdoer’s negligence to the ultimate harm is not broken when the wrongdoer can foresee that the situation created by the original negligence is likely to be followed by an intervening circumstance that will result in the very kind of damage actually occurring.
Having foreseen the intervention as very likely to happen, it would also be necessary that the damage actually caused by the intervention was of a type which the first wrongdoer could reasonably foresee as resulting from the intervention[xviii] (eg personal injury and property damage from skidding on bald tyres on a wet road, the intervention of rain being very likely in the high rainfall locality).
A mere foreseeable possibility by the first wrongdoer of such an intervening circumstance (eg an earthquake) is not sufficient to negate a finding of a novus actus interveniens.[xix] It must be an intervention that was very likely to happen.[xx]
The following hypothetical scenario may be useful in exemplifying a novus actus interveniens.
A workman at a factory site negligently fails to replace a safety cover over a deep hole in the grounds of the factory. The factory perimeter is surrounded by a high fence with barbed wire at the top and the entrance gates are locked each evening. Two ten year old boys, break the lock and enter the grounds of the factory at night. Both boys are injured as a result of falling into the unguarded hole.
The first negligence of the workman satisfies the “but for” test of causation. A court, however, may find a novus actus interveniens with respect to the later circumstance of the young boys entering the site. A court would be justified in finding that the intervening circumstance of the boys’ entry at night, was not something which the workman should foresee as a likely happening. The risk created by the workman did not encompass, as a likely event, the further risk of the illegal entry by the boys. Consequently, the workman’s negligence, while a necessary condition for the occurrence of the harm, is not a legal cause in the face of the novus actus interveniens. Or to use the language of the Civil Liability Acts, “it is [not] appropriate for the scope of the liability of the person in breach [the workman] to extend to the harm so caused.”[xxi] The workman’s negligence, both in science and fact was a cause of the boys’ injuries but the law is deeming the workman’s negligence not to be a cause based on what he ought to have foreseen as likely to happen.
Finding legal causation where “but for” test not satisfied
It is unlikely in the absence of factual causation (“but for” test not satisfied) that a court will find legal causation established. The following example will suffice, where, for instance, a party moors by anchor a boat, dangerously close to a rocky shoreline and an earthquake intervenes causing a tidal surge which pushes the boat onto the rocks. This novus actus interveniens would have pushed the boat onto the rocks even if the party had moored the boat at a safe distance.
However, exceptionally, as recognised in judicial analysis[xxii] and Civil Liability legislation,[xxiii] there are circumstances where negligence which does not satisfy the “but for” test as a necessary condition for the occurrence of the harm, will be held as a legal cause of the harm:-
“In deciding in an exceptional case, in accordance with established principles, whether a breach of duty — being a breach of duty that is established but which can not be established [as a necessary condition of the occurrence of the harm]- should be accepted as satisfying subsection (1)(a), the court is to consider (among other relevant things) whether or not and why responsibility for the harm should be imposed on the party in breach”[xxiv]
An exceptional situation is the “well-known difficulty in cases where there are two or more acts or events which would each be sufficient to bring about the plaintiff’s injury. The application of the “but for” test ‘gives the result, contrary to common sense, that neither is a cause’”.[xxv]
Such an example may involve two drivers A and B simultaneously racing each other along a carriageway who collide, at an intersection, with a third vehicle driven by C. C’s vehicle entered the intersection on a green light. A and B were unable to stop at the red light due to their excessive speed. Evidence indicates that the same collision and harm would have happened solely from A’s vehicle or solely from B’s vehicle. If the “but for” test in these circumstances was the sole criterion for determining legal causation, an absurd result would follow whereby neither A nor B’s negligence was a cause of C’s damage, since the same harm would have happened in the absence of A’s negligence, and similarly in the absence of B’s negligence. Of course the outcome in this scenario is that blame would be apportioned between A and B on the basis that their respective negligence was each a legal cause of the harm. On the question of remoteness of damage, the collision and subsequent damage was the very type of damage that could reasonably be foreseen from the speeding.
The “but for” test provides a focal point for the analysis of the ultimate issue of whether a party’s negligence is a legal cause of the harm. But satisfying the “but for” test, per se, is neither a necessary precondition nor a sole determinant of a legal cause. The weighing of factors of culpability, foreseeable risk, novus actus interveniens, or remoteness of damage may ultimately lead to a finding that particular negligence is not a legal cause despite that negligence being a necessary condition of the occurrence of the harm. The corollary is that in exceptional circumstances, negligence that was not a necessary condition of the occurrence of the harm may be found as a legal cause.
The marshalling and ordering of factual circumstances into the matrix of the legal elements of causation presents a significant challenge for legal advisers. It is hoped that the above writings may assist in this task.
[i] March v E & MH Stramare Pty Ltd  HCA 12 at para 5 per Mason CJ
[ii] J Fleming The Law of Torts ( 3rd Ed, Law Book Co, Sydney, 1965) p 231
[iii] March v E & MH Stramare Pty Ltd  HCA 12 at para 15 per McHugh J for a similar list
[iv] Civil Liability Act 2003 (Qld) s12; Civil Liability Act 2002 (NSW) s5E; Civil Liability Act 1936 (SA) s35; Civil Liability Act 2002 (Tas) s14; Wrongs Act 1958 (Vic) s52; Civil Liability Act 2002 (WA) s5D
[v] Civil Liability Act 2003 (Qld) s11(1)(a) and its other States’ equivalents
[vi] March v E & MH Stramare Pty Ltd  HCA 12; Amaca Pty Ltd v Ellis  HCA 5
[viii] Civil Liability Act 2003 (Qld) s11(2) and its other States’ equivalents
[x] Overseas Tankship (UK) Ltd v Morts Dock and Engineering Co Ltd  AC 388; Overseas Tankship (UK) Ltd v The Miller Steamship Co Pty  1 AC 617; March v E & MH Stramare Pty Ltd  HCA 12 at paras. 6,7 and 8 per Mason CJ
[xi] Civil Liability Act 2003 (Qld) s11(1)(a) and (b) and its other States’ equivalents
[xii] Civil Liability Act 2003 (Qld) s11(1)(a) and its other States’ equivalents
[xiii] Civil Liability Act 2003 (Qld) s11(1)(b) and its other States’ equivalents
[xiv] Civil Liability Act 2003 (Qld) s24 and its other States’ equivalents
[xv] Civil Liability Act 2003 (Qld) s11(4) and its other States’ equivalents
[xvi] March v E & MH Stramare Pty Ltd  HCA 12 at paras. 23-27 per Mason CJ; Thorpe Nominees Pty Ltd v Henderson  2 Qld R 216
[xxi] Civil Liability Act 2003 (Qld) s11(1)(b) and its other States’ equivalents;( the parentheses are mine)
[xxii]March v E & MH Stramare Pty Ltd  HCA 12 at para 22 per Mason CJ
[xxiii] Civil Liability Act 2003 (Qld) s11(2) and its other States’ equivalents
[xxv] March v E & MH Stramare Pty Ltd  HCA 12 at para 22 per Mason CJ