Download Confusion in Negligence Law: Proximate Cause, Legal Cause, and Substantial Factor and more Summaries Law in PDF only on Docsity!
FACTUAL CAUSATION
Jane Stapleton*
The doctrinal parameters of the tort of negligence are remarkably open textured which is why it has typically been in negligence cases that foundational formulations of factual causation have been made. This area of law has recently undergone an extensive restatement by the American Law Institute ('ALI') and been the subject of legislative attention in all Australian states. In the light of these developments this essay sketches some essential issues relevant to factual causation which apply not only to the tort of negligence but throughout the law.
I THE RESTATEMENT OF US TORT LAW
At the end of the 20 th^ century the ALI launched a project to compile a Third Restatement of the general principles of the US law of torts. Previous restatements had covered a field described as 'legal causation' so the question arose as to how to restate this field. The orthodox starting point for common law analysis is to read the cases and deduce there from the meaning courts ascribe to relevant terms such as 'duty' or 'breach'. This most basic of analytical techniques fails utterly in the area of causal terms. In the First and Second Restatement of Torts the term 'legal cause' signified an amalgam concept consisting of both an historical connection element (did the breach contribute to the injury?) and a truncation of legal responsibility element (should the party in breach be liable for this injurious consequence of the breach?). Yet in some case law 'legal cause' signified only the truncation of legal responsibility element. Exactly the same terminological disarray was present in relation to the term 'proximate cause': in some contexts it signified the amalgam of the two elements; in others, the most common usage, it referred only to the truncation issue. So it was that US case law contained both: statements that proximate cause was the second component of legal cause; 1 and statements that legal cause was the second component of proximate cause.^2
Compounding this confusion was the indiscriminate deployment by US courts of the term 'substantial factor'^3 (akin to the shifting use of 'material contribution' in the Commonwealth). 4 Sometimes 'substantial factor' was used in relation to the
- (^) Professor, Australian National University. (^1) See, eg, Winn v Posades , 913 A 2d 407, 411 (Conn, 2007). (^2) See, eg, Harrison v Binnion , 214 P 3d 631, 638 (Idaho, 2009). (^3) See, eg, Dan Dobbs, The Law of Torts (2000) 416: 'the substantial factor test is not so much a
test as an incantation'. (^4) See Jane Stapleton, 'Cause in Fact and the Scope of Liability for Consequences' (2003) 119
Law Quarterly Review 388, 394–5 (hereafter 'Cause in Fact').
468 Federal Law Review Volume 38
requirement that the legal complaint relate to an injury that is more than trivial. Elsewhere it was used: as a synonym for the 'but for' test; or as a fudge to mask the inadequacy of the but for test of historical connection in cases such as where there are two or more sufficient factors; 5 or to mask radical rules developed to permit a plaintiff to jump an otherwise unbridgeable evidentiary gap. 6
There was no value in the Institute attempting to 'restate' this causal usage. Clarity of exposition required that causal terminology be confined to a single idea, namely the objective idea of historical connection, conveniently captured by the term 'factual causation'. It was crucial that this question of fact (and the special rules of law relating to its proof) be recognised as completely distinct from the issue of where and why responsibility for the infinite chain of consequences of conduct should be truncated. This truncation issue, sometimes known outside the US as 'remoteness', rests entirely on the normative analysis of the facts. Accordingly it was recommended to the ALI:^7 that each of these two issues be given its own separate chapter in the new Restatement; that terms such as 'legal cause', 'proximate cause' and 'substantial factor' should be completely abandoned; and that henceforth, the truncation issue should be described in the completely non causal terms of 'the scope of liability for the consequences of breach', or 'scope of liability' for short.
It was a reflection of the deep dissatisfaction with the state of US doctrine in the area, and in particular with the description of the truncation issue in causal terms, that all these suggestions were smoothly adopted by the ALI, despite their radical nature both in terms of their departure from the terminology of case law and in terms of the structure of preceding restatements of torts. Accordingly in volume 1 of the Restatement Third, Torts: Liability for Physical and Emotional Harm (2009) Chapter 5 deals with the issue of 'Factual Cause' while a completely separate chapter, Chapter 6, deals with the distinct truncation issue of 'Scope of Liability'.
II THE COMMONWEALTH
What then about the state of case law on these issues in the Commonwealth? Remarkably, here there has been even more disarray for, whereas US courts had long accepted that there were two separate issues at stake, the historical connection issue and the truncation issue, Commonwealth courts have struggled to express this separation in a consistent and coherent manner. One major source of difficulty in Australian courts has been the frequent and lamentable recourse to the slogan of 'common sense' causation. 8
(^5) See, eg, Anderson v Minneapolis, St Paul & Sault Ste Marie Railway Co, 146 Minn 430 (Minn,
- a case involving the merging of two fires where the term was first coined by a US court. (^6) See Jane Stapleton, 'The Two Explosive Proof of Causation Doctrines Central to Asbestos
Claims' (2009) 74 Brooklyn Law Review 1011 (hereafter 'The Two Explosive Proof of Causation Doctrines'). (^7) See Jane Stapleton, 'Legal Cause: Cause in Fact and the Scope of Liability for
Consequences' (2001) 54 Vanderbilt Law Review 941. (^8) A typical example is Allianz Australia Insurance Ltd v GSF Australia Pty Ltd (2005) 221 CLR
568, 581 [41] (McHugh J). Appeal to the idea of 'common sense' causation is extremely rare in the US and is widely deprecated in the UK, see, eg, Stone & Rolls Ltd (in liq) v Moore Stephens (a firm) [2009] 3 WLR 455, 460 [5] (Lord Phillips of Worth Matravers).
470 Federal Law Review Volume 38
difficulties in their complex approach. 15 But it is worth looking at one of their elaborate and ultimately unhelpful 'principles' because it illustrates the sort of textual obfuscation that runs through key Australian decisions. In looking at cases which turned on the truncation issue they found that it was often the case that liability was denied when, after the defendant's breach of obligation, a third party or an abnormal event (such as lightning) had intervened. Australian courts often communicated this result by stating that the intervention 'broke the chain of causation' while, as we have seen, American courts typically did so by stating that the tort was not the 'legal cause' or was not the 'proximate cause' of the outcome. Since Hart and Honoré were committed to the project of mapping causal usage they concluded that:
courts have often applied, in their determinations of causal questions, a central concept in which great emphasis is laid on voluntary action or abnormal and coincidental events as negativing causal connection.^16 But the fact remained that there were myriad situations where the law imposed liability in the presence of just such an intervention. 17 While the fact of intervention of a third party or abnormal event certainly may be relevant to our enquiry about where liability should be truncated, it is not clear from Hart and Honoré when and why the intervention might be relevant: what they produced was merely a topology of causal usage, not a geology of the normative reasoning lying beneath that usage. This absence of normative rationale exposes a judge to the temptation of merely asserting a conclusion on the truncation issue without providing reasons, while reciting some version of Hart and Honoré's 'central concept' of causal connection — a temptation to which Justice McHugh fell victim on a number of occasions. 18 This is hardly an advance on what Hart and Honoré rightly called the 'obscure metaphor' 19 of the intervention 'breaking the chain of causation'.
The temptation to mere assertion was further exacerbated by Hart and Honoré's characterisation of their observed patterns of truncation in terms of causal connections. Just as the Americans found that giving the truncation issue causal names such as 'proximate cause' or 'legal cause' carried with it the risks of confusing jurors, jury free jurisdictions also face risks if the truncation issue is characterised as a 'causal' question. This is because for many of us the notion of causation has a factual ring. 20 In ordinary speech we tend to think of something either being a cause or not, and we often do not see our conclusions on the matter as requiring normative justification. In short, so long as the truncation issue is framed in causal language, some judges will be tempted to present their determinations relating to truncation without adequate normative
(^15) See Jane Stapleton, 'Choosing What We Mean by "Causation" in the Law' (2008) 73 Missouri
Law Review 433, 458–65 (hereafter 'Choosing'). (^16) Hart and Honoré , above n 14, 130–1. (^17) For example sometimes there is liability even though lightning has intervened and
sometimes there is no liability: we need to delve into the reasons why this is so, see Stapleton, 'Choosing', above n 15, 461–4. (^18) See, eg, Bennett v Minister of Community Welfare (1993) 176 CLR 408, 429–30 [13]; Nominal
Defendant v Gardikiotis (1996) 186 CLR 49, 55. On the general point see Stapleton, 'Choosing', above n 15, 463–4; Jane Stapleton, 'Occam's Razor Reveals an Orthodox Basis for Chester v Afshar ' (2006) 122 Law Quarterly Review 426, 431–6. (^19) Hart and Honoré , above n 14, xxxiii. (^20) Indeed, Hart and Honoré even characterised their truncation notions as not just causal but
'factual': Hart and Honoré, above n 14, lii, 91.
2010 Factual Causation 471
justification. Accompanying this is the risk that, so long as both the factual issue of historical involvement and the normative issue of the truncation of liability are framed as 'causal' questions, a trial judge may not easily recognise whether statements in previous appellate cases concerning 'causation' relate to historical involvement or truncation. Judges would better communicate their reasons if the historical contribution question was kept absolutely separate from the truncation question and if causal terminology was confined to the former.
The argument for such clarification of terminology 21 was referred to with approval in the final report of the Review of the Law of Negligence, the 'Ipp Report' which recommended the clear separation of factual causation and the scope of liability. 22 Bizarrely, however, the Ipp Report chose to retain the umbrella term of 'causation' to signify the amalgam of both issues. So it was that, at the very time the American Law Institute was stripping the truncation issue of its misleading causal label throughout the law of torts, the Ipp Report was entrenching that barrier to clarity of legal analysis in Australia: for all Australian States plus the Australian Capital Territory 23 followed the Ipp Report recommendation and legislatively adopted the 'causation' umbrella term albeit only in the limited field covered by that Report, namely where the focus is on the 'fault of a person (the "tortfeasor")', 24 'negligence'^25 or a 'breach of duty'. 26 So long as the 'causation' term is used as such an umbrella concept, it renders incoherent the many judicial statements that 'causation is essentially a question of fact'. 27
Nothing would be lost and much would be gained if Australian courts quietly ignored the umbrella term both under these civil liability statutes where it does no substantive work and, elsewhere, resisted the temptation to refer to 'common sense causation' and proceeded directly to the analysis of the separate issues of factual cause and scope of liability. This substantial improvement in the clarity of exposition of judicial reasoning would, it is to be hoped, then work its way into legislative drafting which has been so bedevilled by cryptic terms aimed at capturing some amalgam of factual cause and scope of liability. 28
(^21) See Stapleton, 'Cause in Fact', above n 4. (^22) Commonwealth of Australia, Review of the Law of Negligence: Final Report (2002) 109
(footnote 6); 117 8 (Recommendation 29). (^23) Civil Liability Act 2003 (Qld) s 11; Civil Liability Act 2002 (NSW) s 5D; Wrongs Act 1958 (Vic)
s 51; Civil Law (Wrongs) Act 2002 (ACT) s 45; Civil Liability Act 2002 (Tas) s 13; Civil Liability Act 1936 (SA) s 34; Civil Liability Act 2002 (WA) s 5C. Neither the Northern Territory nor the Commonwealth has implemented any provision relating to 'causation'. (^24) Civil Liability Act 2002 (WA) s 5C(1). (^25) Civil Liability Act 2002 (NSW) s 5D(1); Wrongs Act 1958 (Vic) s 51(1); Civil Law (Wrongs) Act
2002 (ACT) s 45(1); Civil Liability Act 1936 (SA) s 34(1). (^26) Civil Liability Act 2003 (Qld) s 11(1); Civil Liability Act 2002 (Tas) s 13(1). (^27) Bennett v Minister of Community Welfare (1993) 176 CLR 408, 412 3 (Mason CJ, Deane and
Toohey JJ); Roads and Traffic Authority v Royal (2008) 245 ALR 653, 674 [81] (Kirby J). (^28) All liabilities, including those arising under statute, are limited. A statute may expressly
limit the type of consequence that comes within its scope: see eg, Allianz Australia Insurance Ltd v GSF Australia Pty Ltd (2005) 221 CLR 568. Or such limits may be generated implicitly by the clear purpose of the statute: eg, Gorris v Scott (1874) 9 LR Exch 125. More often these must be divined by the court from more general interpretations of the purpose of the rule in the light of the general law.
2010 Factual Causation 473
It may well be that the reason why conduct is judged to be a breach provides a sound basis on which to determine whether a particular consequence of that breach should be judged to be within or outside the appropriate scope of liability for consequences. For example, if, as in Chappel v Hart , 32 the central ground on which the defendant's conduct was held to be a breach of the duty of care was that a reasonable person would have warned about the risk of a certain consequence, this provides a straightforward ground for including such a consequence within the scope of liability even if the risk had been of extremely low probability. Conversely, the mischief to which a statutory liability is expressly targeted provides a sound reason for judging some other consequence to fall outside the appropriate scope of that liability. 33
But typically, outside the warning and some statutory contexts, the bearing that the nature of the breach has on the truncation (ie scope) issue is much more complex and contestable, and these qualities are masked when the judgment on the truncation issue is communicated in terms of a merely asserted 'scope of the duty'. 34 By 2005 Lord Hoffmann had recognised the inappropriate bootstraps quality of his 'scope of the duty' approach and explicitly disapproved it, recognising that by the term 'scope of the duty' he had intended to refer to what is required to avoid breach, which is always and simply to take reasonable care in the circumstances. This question 'has nothing to do with the extent of the consequences for which the valuer is liable'. 35
Regrettably Australian courts often refer to the 'scope of the duty'. It is preferable for such usage to cease: where courts want to refer to the issue of breach, namely what it was that reasonable care required of the defendant in the circumstances, safer more transparent terms are available such as the 'content of the duty of care' or the 'content of the obligation'.
IV FACTUAL CAUSE
Necessary factors
Courts throughout the world are agreed that the relation of necessity (ie 'but for') between the breach and outcome is one that the law should designate as causal. If then, according to the fact finder's evaluation of the evidence, the plaintiff has established to the requisite standard of proof that the outcome would not have occurred if the breach had not occurred; the plaintiff has shown that the breach is a factual cause of the outcome. Thus in Chappel v Hart it was because, in the evaluation of the fact finder, it was clear that Mrs Hart would not have had the operation where and when she did had Dr Chappel given her the warning he should have given, that his breach was held
(^32) (1998) 195 CLR 232. See generally Stapleton, 'Occam's Razor Reveals an Orthodox Basis for
Chester v Afshar ', above n 18, 447–8. But even if it is completely uncontroversial that the outcome was 'the very thing' the risk of which had been a reason why the defendant's conduct was judged to be a breach, this does not relieve the plaintiff from proving factual cause: see Adeels Palace Pty Ltd v Moubarak (2009) 239 CLR 420, 442 [51]. (^33) Gorris v Scott (1874) 9 LR Exch 125. (^34) Jane Stapleton, 'Negligent Valuers and Falls in the Property Market' (1997) 113 Law
Quarterly Review 1, 1–7; Stapleton, 'Cause in Fact', above n 4, 390–1. (^35) Leonard Hoffmann, 'Causation' (2005) 121 Law Quarterly Review 592, 596.
474 Federal Law Review Volume 38
a factual cause of her perforated oesophagus. 36 Similarly, in March v E & M H Stramare Pty Ltd the breach was clearly a factual cause: had the defendant not carelessly parked his truck in the mid line of the street, the intoxicated plaintiff would not have hit it. 37 The controversial issue in both Chappel and March concerned the normative truncation question: should this outcome of the breach be regarded as within the appropriate scope of the defendant's liability? This is not a question of fact or 'common sense' but of normative judgment on which reasonable minds might differ.
There can be more than one necessary factor, as illustrated by the facts of March itself where both the plaintiff's intoxicated driving and the defendant's careless parking of the truck on the mid line of the road were but for factors of the collision. 38
Multiple sufficient factors
Though the but for test of factual causation is notoriously inadequate, the Ipp Report made no recommendation on the issue of which relations besides this one of necessity should be designated as causal by the law. The grounds given for this omission were unconvincing. First it was said that the problem lay in cases of over determination by multiple sufficient factors. An example of this is what I call the 'double hit hunters' case' where two hunters carelessly shoot and a mountain walker is hit by both bullets each of which would have been sufficient to kill instantly. In fact the problem for the law is, as we will see, much more extensive than these special cases. The second argument for inaction, namely that 'the law has devised rules for resolving such cases in ways that are generally considered satisfactory and fair', 39 is wishful thinking: judicial treatment of these cases is confused, a state of affairs that is understandable given the relative neglect of causal analysis by legal scholars.
Though a regime of tort law might confine its interest to necessary (ie but for) factors, clearly none does this. In other words, the law is interested in the possibility of imposing liability on non necessary factors and in order to do so must recognise them as qualifying as 'causal'. The most often cited examples of non necessary relations that the law chooses to designate as causal are multiple sufficient factors that are independent of one another. In such cases each tortious factor is designated as causal even though not necessary: 40 so, in the example in the previous paragraph, each hunter is recognised as a cause. The equivalent recognition would no doubt be afforded where the multiple non necessary factor is an omission (which would be the case if, for example, the hunters were children and the defendants were the parents who had independently and culpably failed to control their respective children). Moreover, there are plausible reasons why the same approach should be taken to the relation to the outcome of multiple sufficient omissions which are dependent — that is where, to avoid the outcome one party would have had to act on a state of affairs set up by another but where the first omitted to act and the other omitted to set up: for example
(^36) '[T]he task of Dr Chappel was to demonstrate some good reason for denying to Mrs Hart
recovery in respect of injuries which she would not have sustained at his hands but for his failure adequately to advise her': Chappel v Hart (1998) 195 CLR 232, 257 [69] (Gummow J). (^37) (1991) 171 CLR 506. (^38) An illustration of multiple but for factors that are omissions was given in Bennett v Minister
of Community Welfare (1993) 176 CLR 408, 429 (McHugh J). (^39) Commonwealth of Australia, above n 22, 109. (^40) March v E & M H Stramare Pty Ltd (1991) 171 CLR 506, 516 (Mason CJ); Chappel v Hart (1998)
195 CLR 232, 282–3 [116].
476 Federal Law Review Volume 38
Three factories each independently and in breach of duty discharge oil into a bay. By a regulatory standard, fishing in the bay is forbidden if the concentration of oil is greater than a particular level. By the time the pollution is detected the concentration level far exceeds this regulatory standard. When the ban is triggered this causes grave economic injury to local commercial fishermen who are unable to quantify the contribution each factory's discharge made. Xavier, Yadra and Zach independently but fraudulently advise Penny that the published financial statements of a company, Tenron, are honest and strong. This advice is sufficient to persuade Penny to invest in Tenron, whose share price almost immediately collapses when it is revealed that the financial statements of the company dishonestly misrepresented the health of its finances. What these four examples show is that a non necessary non sufficient factor may contribute to the existence of a phenomenon. It does so by forming part of an undifferentiated whole that operates to bring about the existence of the phenomenon: a single vote formed part of the unanimous resolution; Able's push formed part of the total physical force that sent the car off the edge of the mountain; the discharge by one factory formed part of the total concentration that exceeded the regulatory standard; and Xavier's advice was part of the total information that persuaded Penny. The examples also illustrate how the law is interested in the possibility of imposing liability on such a non necessary non sufficient factor. To do so the law needs to designate a notion of factual causation that is wide enough to accommodate these contributions. This is why the 'but for' test of factual cause is under inclusive and why courts grasp at vague undefined labels such as 'substantial factor' and 'material contribution' in their attempt to recognise a non necessary non sufficient factor as a 'cause'.
It is only the seductive simplicity of the 'but for' test that distracts courts from enunciating an appropriately wide statement of the relation of 'factual cause': namely, that a factor is a factual cause if it contributes in any way to the existence of the phenomenon in issue. Courts should no longer allow the fact that in most cases the contribution in issue will have been one of necessity (but for), to mislead them into regarding necessity as the fundamental form of causal relation recognised by the law: courts should grasp that it is the relation of contribution.
An appreciation that contribution is the fundamental form of causal relation is especially important in the context of decision making where, as Justice McHugh emphasised in Henville v Walker : 44
… the long standing recognition of the possibility that two or more causes may jointly influence a person to undertake a course of conduct … a representation need not be the sole inducement in sustaining the loss. If 'it plays some part even if only a minor part', in contributing to the course of action taken ... a causal connection will exist. 45 Two associated points need to be mentioned here. First, notice that the four examples given so far involve an indivisible outcome and the discussion has been about a non necessary non sufficient contribution to the existence of that indivisible
(^44) (2001) 206 CLR 459. (^45) Ibid 493 [107]. See also Medlin v State Government Insurance Commission (1995) 182 CLR 1
where the plaintiff's injuries from the tort were a 'contributing cause of his decision to retire' (at 8).
2010 Factual Causation 477
outcome: if the breach did make such a contribution and if all other elements of the cause of action are established, the defendants are jointly and severally liable for the entire indivisible outcome. This should be sharply distinguished from cases where the issue is whether a factor is a factual cause of only part of a divisible outcome, say pneumoconiosis or deafness: here if all elements of the cause of action are established, the defendant is only liable for that part of the divisible outcome in relation to which the breach was shown to be a factual cause (unless a special proof rule is available: see below).
Secondly, though it is for normative reasons that the law requires its notion of 'cause' to be wider than the relation of necessity (ie but for factors) and to include any relationship of contribution to the existence of the phenomenon, whether this relationship is present in any individual case is a matter of objective fact not normative choice. Two amplifications of this latter point should, however, be noted when we move to the quite separate issue of how this factual causation relationship of contribution might be proved : the evaluative nature of assessing evidence; and the possibility of special proof rules.
The relevant evidence on causal contribution is often disputed even in the most banal of orthodox cases so that the decision maker must make an evaluative decision on whether this factual relation of causal contribution is to be treated as having been established to the requisite standard of proof, but the evaluative nature of this process does not alter the factual nature of the underlying relation in dispute. 46
The other useful amplification here is that, as we will see below, in a very few distinct areas the law constructs special localised proof rules: while these are prompted by particular normative concerns, they also do not alter the factual nature of the underlying relation in issue.
The Ipp Report
We are now in a position to consider the following Ipp Report recommendations with respect to factual causation:
(c) The basic test of 'factual causation' (the 'but for' test) is whether the negligence was a necessary condition of the harm. (d) In appropriate cases, proof that the negligence materially contributed to the harm or the risk of the harm may be treated as sufficient to establish factual causation even though the but for test is not satisfied. (e) Although it is relevant to proof of factual causation, the issue of whether the case is an appropriate one for the purposes of (d) is normative. (f) For the purposes of deciding whether the case is an appropriate one (as required in (d)), amongst the factors that it is relevant to consider are: (i) whether (and why) responsibility for the harm should be imposed on the negligent party, and (ii) whether (and why) the harm should be left to lie where it fell. 47
(^46) Contrast Roads and Traffic Authority v Royal (2008) 245 ALR 653, 675 [84] (Kirby J) that 'the
determination of causation in fact is not one that can be made without recourse to broader considerations'. (^47) Commonwealth of Australia, above n 22, 118 (Recommendation 29).
2010 Factual Causation 479
might call 'generic capacity' 53 ) and that it did so in this individual case (which we might call 'individual agency') before the critical factual cause question can be asked about whether the failure to warn contributed to the outcome.
Another complexity arises here. It is often easy to grasp the notion of how a contribution that is neither necessary nor sufficient may contribute to the existence of a phenomenon by forming part of an undifferentiated whole that operates to bring about the existence of the phenomenon: for example, how a single vote formed part of the unanimous resolution and how Able's push formed part of the total physical force that sent the car off the edge of the mountain (see above). But it may sometimes be helpful for courts or commentators to be able to refer to an algorithm that represents the notion of contribution. This is known as the NESS test: 54 a factor, such as a single vote in our club example, 'contributes' to the outcome if it is a Necessary Element for the Sufficiency of a Subset of the facts. This algorithm identifies the contribution of the vote of committee member number one to the passage of the motion to expel: this vote is necessary for the subset (consisting of the vote of number one along with the votes of number two and number three) to be sufficient for passage.
Even in the typical case where the plaintiff's claim of factual cause rests on the claim that the breach was necessary for the outcome, proof of factual causation clearly involves establishing both facts about the actual past and speculation about what would have happened in the past in some hypothetical world where (inter alia) the breach had not occurred. 55 Orthodoxy requires both to be established on the balance of probabilities. 56 (These issues are often hotly contested requiring the decision maker to make careful evaluations of complex evidence.) So, for example, where hypothetical conduct is in issue the relation of interest to the law is what the particular individual would probably have done. The Ipp Report correctly emphasised that the relation of interest is what the relevant individual would have done, not what some normative creation, such as a 'reasonable person', would have done: thus, in Commissioner of Main Roads v Jones^57 the issue was what this driver would probably have done had there been relevant warning and speed signs, not what a reasonable driver would have done. In their post Ipp Report civil liability reform legislation 58 five States explicitly reiterated this rule in relation to the hypothetical conduct of the injured person.
But the Ipp Report and the responding legislation are significantly incomplete: the appropriate subjectivity rule is a general one and applies to the hypothetical conduct, not only of the injured person, but also of the defendant and third parties. Fortunately the High Court has been alive to the need to supplement the post Ipp Report civil liability legislation with common law principles. Thus in Adeels Palace Pty Ltd v
(^53) See, eg, Seltsam v McGuiness (2000) 49 NSWLR 262 (is inhalation of asbestos capable of
contributing to the contraction of renal cell carcinoma?). (^54) See Stapleton, 'Choosing', above n 15, 444, 459 and 471ff. (^55) See, eg, Naxakis v Western General Hospital (1998) 197 CLR 269; Gates v City Mutual Life
Assurance Society Ltd (1986) 160 CLR 1. (^56) These issues must be distinguished from the contingency that another factor which did not
operate might have done so and caused the phenomenon in issue (such as an injury): see Malec v JC Hutton Pty Ltd (1990) 169 CLR 638. (^57) (2005) 215 ALR 418. (^58) Civil Liability Act 2003 (Qld) s 11(3)(a); Civil Liability Act 2002 (NSW) s 5D(3)(a); Wrongs Act
1958 (Vic) s 51(3); Civil Liability Act 2002 (Tas) s 13(3)(a); Civil Liability Act 2002 (WA) s 5C(3)(a).
480 Federal Law Review Volume 38
Moubarak^59 when two patrons who had been shot by another patron sued a restaurant for providing inadequate security, the High Court correctly inquired into whether the evidence of the specific assailant, 'a determined person armed with a gun and irrationally bent on revenge', 60 supported the plaintiffs' claim that additional security staff would have succeeded in deterring him.
VI SPECIAL RULES CONCERNING PROOF OF FACTUAL
CAUSATION
Plaintiffs can face insuperable 61 evidentiary gaps when trying to establish either the relevant past facts or what would have happened in a hypothetical past situation. Of course usually such a plaintiff fails 62 but on occasion courts or legislatures have, in response to a particular policy concern, crafted a special rule allowing the plaintiff to leap the evidentiary gap and establish factual cause.
Before analysing these, however, it is worth noting that in rare cases a special rule of proof of factual cause has been created which makes the plaintiff's task more difficult.
Self serving testimony
In some contexts the concern that a plaintiff will give self serving testimony as to what they would have done in the relevant hypothetical past situation has triggered a rule barring such testimony. The Ipp Report recommended such a rule and four States have included it in their reform legislation. 63
What these moves overlook is that self serving testimony may be proffered by any party. Consider a medical negligence case where the alleged breach was a failure to attend to the patient. If the plaintiff can prove that, had the defendant attended, the defendant would, on the balance of probabilities, have taken steps that would have prevented the deleterious outcome, factual cause will be established. But what if one possible reasonable treatment would not have achieved such prevention: as was the case in Bolitho v City and Hackney Health Authority.^64 The defendant has a real incentive to give the self serving testimony that, had she attended the patient, she would have chosen the treatment regime that would have made no difference to the outcome.
Heeding presumption
Much more commonly courts and legislatures have crafted a special rule concerning the proof of factual cause to assist the plaintiff. Sometimes such a rule relates to proof of what would have happened in the relevant hypothetical past world had there been no breach. A striking example was created in the US when plaintiffs brought defective
(^59) (2009) 239 CLR 420. (^60) Ibid 441 [49]. (^61) Of course many evidentiary gaps can be bridged by legitimate inference as in TNT
Management Ltd v Brooks (1979) 23 ALR 345. (^62) See, eg, West v Government Insurance Office of NSW (1981) 148 CLR 62. (^63) Civil Liability Act 2003 (Qld) s 11(3)(b); Civil Liability Act 2002 (NSW) s 5D(3)(b); Civil
Liability Act 2002 (Tas) s 13(3)(b); Civil Liability Act 2002 (WA) s 5C(3)(b). (^64) [1998] AC 232. For an especially insightful discussion see Michael Jones, Medical Negligence
(4 th^ ed, 2008) 445–6.
482 Federal Law Review Volume 38
injury as if it was indivisible. In the United States courts commonly recognise such an 'indivisibility of injury' fiction in relation to water pollution and asbestosis. 70
While the result in Bonnington Castings Ltd v Wardlaw^71 is also consistent with the 'indivisibility of injury' fiction, it is unclear whether the House of Lords intended to create such a special rule of proof of factual cause because the case revolved around a more fundamental issue, namely whether a defendant could be liable at all if his breach had only caused part of a total injury. 72 When the point came for resolution in other cumulative disease cases in the UK 73 no special rule was applied: the plaintiff was required to show, on a rough and ready basis, how much of the total injury the breach had factually caused.
The material contribution/exposure to risk doctrine: unknown mechanism; indivisible injury
Much more radical is the rule created in some jurisdictions to allow plaintiffs to establish that one source of risk among many was a factual cause of an indivisible medical condition when the mechanism of that condition is unknown, as is the case with mesothelioma. A victim of such a condition could not succeed on orthodox principles because it is not possible to make a 'robust and pragmatic' 74 inference that the individual source of risk contributed to the condition: an impossibility elegantly exposed in Fairchild v Glenhaven Funeral Services Ltd. 75 Nor is it possible to apply the special Summers v Tice proof rule of alternative liability which requires all risk creators to be culpable and before the court; nor can the plaintiff seek to use the 'indivisibility of injury' fiction which requires that it is known that the total injury resulted from cumulative injuries from every source of risk.
The radical rule of proof adopted in these unknown mechanism cases, described as the 'material contribution to risk' doctrine in the UK or the 'exposure to risk' doctrine in the US, does present significant challenges, for example: what is its incidence (eg does the rule only apply to 'single agent' conditions 76 )? Does the rule shift the burden of persuasion? Does the rule support in solidum or joint and several liability? And is the rule appropriately characterised only as a special rule of proof of factual causation or is it preferable to see it as recognising exposure to risk (ie loss of a chance) as actionable
(^70) See Stapleton, 'The Two Explosive Proof of Causation Doctrines', above n 6. Both the
'indivisibility of injury' rule and the 'alternative liability' rule have been restated in s 28 of the Restatement (Third) of Torts: Liability for Physical and Emotional Harm (2010). The former had been covered by s 433B(2) of the Second Restatement while the latter had been covered by s 433B(3). (^71) [1956] AC 613. (^72) Jane Stapleton, 'Lords a'leaping Evidentiary Gaps' (2002) 10 Torts Law Journal 276, 283. (^73) Such as Thompson v Smiths Shiprepairers (North Shields) Ltd [1984] 1 QB 405 (deafness) and
Holtby v Brigham & Cowan (Hull) Ltd (2000) 3 All ER 421, 428 (asbestosis). (^74) Wilsher v Essex Area Health Authority [1988] AC 1074, 1090 per Lord Bridge in a doomed
attempt to fit McGhee v National Coal Board [1973] 1 WLR 1 into orthodox principles. (^75) [2003] 1 AC 32. See, eg, Lord Bingham at 57–8 [22]. (^76) See Stapleton, 'The Two Explosive Proof of Causation Doctrines', above n 6. Contrast Barker
v Corus UK Ltd [2006] 2 AC 572 with Rutherford v Owens Illinois Inc, 941 P2d 1203 (Cal, 1997).
2010 Factual Causation 483
damage in the limited area of incidence of the rule? 77 Nevertheless, turning a blind eye to the dilemma that such diseases present to orthodoxy leads to the sort of incoherence that bedevils the case law on asbestos cancers in Australia which currently proceeds on the basis of unacknowledged and conflicting fictions about aetiology. 78
The Ipp Report
We are now in a position to consider the Ipp Report recommendation in relation to proof of factual causation that 'in appropriate cases, proof that the negligence materially contributed to the harm or the risk of the harm may be treated as sufficient to establish factual causation' 79 so long as the normative nature of such a special rule of proof is acknowledged and openly analysed. The recommendation is to be welcomed in as much as it will bolster the separation of the factual question of whether a factor is a cause from the normative issue of whether this particular class of plaintiff should be afforded a special rule of proof on this factual issue.
Five States have enacted a provision broadly in line with this recommendation but South Australia and the ACT limited their provision to cases where the plaintiff 'has been negligently exposed to a similar risk of harm by a number of different people (the defendants) and it is not possible to assign responsibility for causing the harm to any one or more of them'. 80 The latter provision is clearly not wide enough to authorise special rules in one wrongdoer cases such as a heeding presumption or the approach adopted in cases such as McGhee v National Coal Board. 81 This limited provision may not even authorise a rule purporting to cover situations where not all risk creators need be joined as defendants such as the market share doctrine, the 'indivisibility of injury' rule and the material contribution/exposure to risk doctrine.
But such developments may still occur at common law. Apart from the entrenchment of the rule that the burden of proof rests on the plaintiff, there is nothing in the post Ipp Report civil liability legislative provisions dealing with either factual causation or scope of liability that inhibits the future recognition of special rules of proof or the enunciation of guiding principles governing the scope of liability. In relation to both, the legislation is best seen as reporting the legitimate law making role of the courts, 82 save with the caveat that courts must enunciate the considerations that have been taken into account. Beyond that the legislation properly 'offers no further guidance about how the task is to be performed.' 83 In these areas, and despite
(^77) Note the problematic characterisation by Lord Hoffmann in Barker v Corus UK Ltd [2006] 2
AC 572 on which see Stapleton, 'Occam's Razor Reveals an Orthodox Basis for Chester v Afshar ', above n 18, 448 fn 77. (^78) See Jane Stapleton, 'Factual Causation and Asbestos Cancers' (2010) 126 Law Quarterly
Review 351 (a note on Amaca Pty Ltd v Ellis (2010) 240 CLR 111). (^79) Commonwealth of Australia, above n 22, 118 (Recommendation 29(d)). (^80) Civil Law (Wrongs) Act 2002 (ACT) s 45(2); Civil Liability Act 1936 (SA) s 34(2). (^81) [1973] 1 WLR 1. (^82) Jane Stapleton, 'The Golden Thread at the Heart of Tort Law: Protection of the Vulnerable'
in Peter Cane (ed), Centenary Essays for the High Court of Australia (2004) 242, 244: A 'vital secret of our constitutional arrangements is the close union of the judicial and legislative powers in the court of ultimate appeal and that our common law legal systems embrace a form of the separation of powers doctrine that accommodates this substantial law making capacity.' (^83) Adeels Palace Pty Ltd v Moubarak (2009) 239 CLR 420, 443 [54].
