TY - JOUR
AU - Turton, Gemma
AB - This is an excellent collection that draws together varied perspectives from leading academics on causation. Recent years have seen rapid and controversial developments in the approach to factual causation in negligence, in the UK and beyond, so the book naturally focuses more on negligence law although it is certainly of value to those interested in causation more broadly. The structure of the book is well considered, dividing the contributions into three broad themes: legal doctrine, scientific evidence, and legal theory. The short introductory chapter by Lord Hoffmann sets out the challenge facing the remaining authors.1 He observes that causation is one area where courts have failed to take note of academic opinion (p. 3). Despite the problems arising from the ‘but for’ test for causation, he says that judges have not heard of the ‘necessary element of a sufficient set’ (NESS) test developed by Wright (p. 3).2 He also maintains that judges do not approach causation in two stages, separating factual causation from legal causation, and is critical of academics' insistence on a strict understanding of causation as a factual concept. Indeed, ‘[i]t is this concept of something having to be really a cause according to criteria lying outside the law which puzzles lawyers’ (p. 5). He asks why academics insist on a philosophical concept of causation and highlights the fact that academics are not agreed on what actually amounts to being a factual cause (p. 4). Instead, he prefers the notion of ‘causal requirement’, and the causal requirement of any legal rule reflects the legal policy behind that rule (p. 4). This is not a convincing argument. While the decision to include a causal element in a legal rule may be a matter of legal policy, causation itself is a natural concept, so the notion of factual causation cannot depend on the purposes of the legal rule for its meaning and content. Lord Hoffmann's introduction, therefore, presents a challenge for the contributors to this collection; to justify the place of factual causation and to propose how the law should respond to problematic cases. The remaining chapters present an excellent analysis of the subject which not only looks at the existing law, but also considers what kinds of challenges the law is likely to encounter in the near future and how it might respond. The first section addresses the law relating to causation, and the focus is largely on the tort of negligence. Writing about causal doctrine in Scots law, Hogg's contribution addresses many of the concerns that were raised by Lord Hoffmann in his introduction.3 Hogg observes the lack of judicial engagement with academic writing on causation and considers whether there are still too many competing theories and classifications of causation, often expressed in highly complicated ways, to give the judiciary sufficient confidence to abandon established modes of thinking in favour of something new (p. 44). Hogg highlights the issue of causal terminology, sensibly arguing that the judiciary will resist changes unless the academic community can reach a consensus on the terminology surrounding causation (p. 45). He also explains the advantages of adopting Wright's NESS test instead of the traditional ‘but for’ test and the ‘Wardlaw’ test of ‘material contribution to harm’.4 Hogg seeks to dispel the perception among lawyers and judges that the NESS test is complicated to apply, and shows that it is a ‘more comprehensive test’ (p. 47), which improves on the ‘but for’ test. Hogg provides a variation on the classic ‘double hit hunters’ scenario to illustrate how each test responds to duplicated causation. He explains that where two vehicles each strike a pedestrian, killing him, and each blow on its own would have been fatal, the ‘but for’ test leads to the ‘bizarre conclusion that neither vehicle caused the accident (because the impact of the other vehicle would have caused the death in any event), [but] the NESS test correctly identifies each vehicle as a cause’ (p. 47). Hogg shows that the NESS test is straightforward to apply, explaining that [i]f one wants to see if V1 is a cause of P's death, one removes V2 from the list of conditions and one still has a set minimally sufficient for the outcome; however removing V1 means that one no longer has such a set, so V1 was a cause of the death. The same result is reached if one separately tests V2 as a possible cause using the NESS test: V2 is also shown to be a cause of the death (p. 48). He takes a similarly clear approach to cases of pre-emptive causation (p. 48). Later chapters on legal theory engage with a high-level defence of the NESS test,5 and it is vital to remember that, as Hogg shows, although the NESS test sounds more complex than the ‘but for’ test, its adoption would simplify matters because it is able to address a wider range of causal problems than the ‘but for’ test. A key concern throughout the first section is the coherence of the law, and whether the exceptional approaches to causation adopted by the courts have a principled basis. Both Morgan and Sanders consider ways in which the courts have tackled the difficulties arising from the evidentiary gap relating to causation of mesothelioma. Morgan focuses on the ‘Fairchild’ exception,6 which imposed liability on the basis that the defendant's negligence had ‘materially increased the risk of harm’.7 Sanders takes a comparative approach contrasting Fairchild with the American decision in Rutherford v Owens-Illinois Inc.,8 where the court similarly determined causation of asbestos-related lung cancer in terms of risk rather than physical injury; what Sanders labels a ‘risk rule’ (p. 11).9 Both authors note the potential for the Fairchild exception to apply outside mesothelioma cases, explaining that there is no principled reason for restricting its application to this disease. However, Sanders advocates restricting its application in this way, because it ‘buys us very little at the cost of considerable uncertainty concerning the nature of the causal question in an array of toxic tort cases and perhaps beyond’ (p. 39). Morgan focuses on the wider lessons about legal reasoning and the roles of courts and legislators we can learn from Fairchild, and argues that if a court's decision cannot be reconciled with general legal principles and is, instead, confined as being an unprincipled exception, ‘this violates every precept of common law reasoning, and brings the judicial development of tort law into disrepute’ (p. 91). The Fairchild decision is widely seen as a policy-based exception to established rules relating to causation. Morgan accepts that policy considerations may validly affect liability decisions, but he emphasises that those policy arguments must be ‘convincingly weighed and applied to the facts of the case’ (p. 61). He clearly shows that the policy arguments raised in Fairchild fail to justify the court's decision. Morgan argues that the House of Lords should have dismissed the claims in Fairchild leaving the task to Parliament to create an exceptional solution. He explains that it is essential to the development of common law that judges maintain general principles, contrasting this with legislation which can create robust, isolated exceptions. Indeed, he argues that confusion can arise if the legislator goes beyond creating clear rules and seeks to base a rule on wider principle, citing as an example the Scottish legislation dealing with pleural plaques. Pleural plaques are an asymptomatic disease that the House of Lords in Rothwell held does not constitute actionable damage,10 but which are actionable in Scotland due to subsequent legislation to reverse the effects of the decision in Rothwell.11 Hogg has, elsewhere, criticised the Scottish legislation, arguing that it lacks a principled basis.12 In contrast, Morgan argues that the advantage of creating exceptions through legislation rather than common law is precisely that the legislature can create robust, isolated exceptions to the existing legal principle. In seeking to provide the pleural plaques legislation with a principled basis by reversing the reasoning as well as the outcome of the Rothwell decision, the Scottish Parliament ‘create[d] legal uncertainty which might open the door for the legislation to be applied by analogy’ (p. 90). Taken together, these chapters provide stimulating analyses of the existing law as well as the lessons which can be learnt when addressing future causal problems. The second section focuses on scientific evidence of causation and the question of what will be considered to satisfy the balance of probabilities standard of proof, both of general causation (can the substance to which the defendant negligently exposed the claimant cause the injury suffered?) and specific causation (did the negligent exposure to the harmful substance cause this claimant's injury?). Goldberg identifies the most pertinent issue as ‘the lack of clarity in being able to determine at what point the balance of probabilities standard (legal) and the standard for epidemiology (science) intersect’ (p. 150). This is especially important in the light of recent attempts, highlighted by Goldberg, Feldschreiber and colleagues, and Wright (in the first of his chapters), wrongly to conflate the balance of probabilities with statistical or epidemiological evidence of a doubling of the risk of harm.13 These authors explain that statistical or epidemiological evidence concerning the ‘relative risk’ is proof of association rather than causation. Goldberg, therefore, argues that insistence on a doubling of risk would demand both too much and too little of a claimant; it should not constitute a minimum threshold because a causal relationship can exist below this level of relative risk, and it should not be considered sufficient as a proof of causation in an individual case (pp. 155–8). The chapters proceed in different directions from this basis. As Wright explains, the balance of probabilities is a standard of persuasion rather than a mere probabilistic concept. He argues that statistical and epidemiological evidence on its own is incapable of proving specific causation because it tells us nothing about the individual, being population-based instead (pp. 206–7). He suggests that while statistical probabilities might enable us to place a bet on what happened, they do not allow us to form a rational belief in what actually happened in any particular instance (p. 207). For this, Wright argues, particularistic evidence is required. Goldberg similarly says that a statistical chance of causation ‘has no compensatory value, until the data is “personalised”’ (p. 161). Wright's approach to particularistic evidence should be contrasted, however, with the function it has in Goldberg's approach. Goldberg favours the use of Bayes' theorem to ‘personalise’ the chance (pp. 162–63). This takes data specific to the individual claimant to refine the statistical probability of causation in the particular case.14 Wright, however, rejects such an approach, maintaining that probabilistic evidence cannot tell us what happened in a particular case. He argues, instead, that the role of particularistic evidence is to enable the fact-finder to develop a belief as to which of the competing causal generalisations actually occurred. So, the reliability of any particularistic evidence will affect the assessment of the degree of fit, but it will do so only by affecting our judgment of the overall coherence of the particularistic evidence with the competing causal stories rather than through some (e.g. Bayesian) modification of the ex ante causal probabilities (p. 210). The solution seems to require a balancing of both authors' approaches. Wright's rejection of what he terms ‘naked statistics’ (p. 210) is unconvincing in its extremism, because it seems conceivable that in rare cases an especially high statistical probability could enable a fact-finder to form a rational belief in causation. But, as he argues, emphasising the qualitative nature of the balance of probabilities may enable the court to make a fuller use of particularistic evidence than if it were confined to a purely statistical role (pp. 211–2). Related issues are also raised by Brown, who examines the place of inference-drawing in determinations of causation.15 While there is a danger that inference-drawing should not be used as a cover for courts to infer causation despite a lack of a scientific basis, and judgments should not be based simply on common sense, Brown's contribution identifies qualitative assessments that can be made of the available evidence in order to draw rational conclusions from it. These chapters on the place of statistical and epidemiological evidence are complemented by chapters exploring the fields of statistics and epidemiology. Seeking to familiarise lawyers with statistical techniques, Dawid explains some of the finer details of statistical calculations, flagging up potential pitfalls to be avoided.16 Feldschreiber, Mulcahy, and Day focus on epidemiology and biostatistics. They recap the Bradford Hill criteria used in modern epidemiology to evaluate causal relationships,17 showing clearly that statistical probabilities are only one element in epidemiology. They also explain how epidemiologists address the potential for errors such as bias and confounding, as well as the calculation of confidence intervals. The authors note that lawyers sometimes assume that while the standard of proof in negligence is the balance of probabilities, the standard observed by scientists is set much higher at 95% (p. 190). Feldschreiber and colleagues explain that this 95% confidence standard (or a p-value of 0.05) only tells us about random sampling errors in the data, but does not address other issues such as bias (p. 184). The value of these chapters, which introduce statistical and epidemiological techniques to lawyers, is highlighted by Goldberg who is critical of the court's reluctance to engage fully with the epidemiological evidence in McTear18 (pp. 157–8). He convincingly argues that Lord Nimmo Smith's reluctance to be persuaded by the epidemiological evidence led him to conclude wrongly that general causation could not be established in that case. Thus, the burden should not rest on the claimant to educate the court about epidemiological methodologies; instead, the court should take an active role in understanding epidemiology since ‘there is a clear societal function of a judge to resolve these matters to the satisfaction of both parties’ (p. 160). To reconcile scientific and legal standards, Feldschreiber, Mulcahy, and Day highlight the fact that courts not only need to learn about the methodologies of epidemiology, but must also understand its role in decision-making across various disciplines (pp. 189–90). The scientific standard of proof is generally considered to be much higher than the civil law standard of the balance of probabilities, but in this chapter the authors explain that, in the regulatory context, medical scientists may make decisions based on ‘plausible causality’ (p. 191). They explain that ‘plausible causality sufficient to ground regulatory action may be determined at a level of mere possibility rather than high probability’ (p. 191). They contrast the ‘prospective forecasting’ nature of regulatory decision-making with the retrospective assessment of evidence that takes place in a negligence claim (pp. 189–90). The regulator will take a more precautionary approach, possibly based on less evidence than is available by the time of any negligence action, so ‘it cannot be assumed that because regulatory action has been taken, causation can be proved’ (p. 190). They provide the example of the decision in XYZ v Schering Health Care Ltd,19 where it could not be established in the negligence action that the third-generation combined oral contraceptive pill created a greater risk of thromboembolism than the second-generation pill, even though the third-generation pill had previously been subject to regulatory action by the Committee on the Safety of Medicines. Another issue highlighted by Feldschreiber and colleagues is the distinction between ‘divisible’ (dose-related) and ‘indivisible’ (‘all or nothing’) injuries (p. 193). This distinction is widely used in caselaw, yet the authors suggest that it is ‘a distinction which is not known to science and which we would predict is likely to prove increasingly problematic’ (p. 193). This distinction does call for closer scrutiny, and the later chapter by Cranor provides useful insight into the processes of disease development.20 He explains biological effects, processes such as initiation, promotion, and progression, and how repeated exposures to a toxic substance can have a cumulative effect on the body (pp. 268–80). For example, where a threshold level of exposure is needed to cause illness or where a substance weakens the immune system leaving a person at risk of other diseases. He also explains that one harmful substance may combine with another, possibly naturally occurring, substance to ‘contribute to disease, accelerate the onset of an illness or worsen it’ (p. 274), as well as genetic and multi-generational effects of harmful substances. Although labels such as ‘divisible’ and ‘indivisible’ disease have an attraction of simplicity, Cranor's chapter highlights the need to develop a fuller understanding in law of how substances interact to trigger disease, as well as those factors that affect the severity of the disease. The final section turns to legal theory and the NESS test for causation. In his second contribution, Wright defends the NESS test against criticisms,21 as does Miller in a chapter that also seeks to explain the NESS test in a way that is readily accessible.22 In addressing a range of criticisms here, however, one development risks reducing the value of NESS as a workable test. Wright explains that in any particular instance a condition is a cause if it is ‘part of (rather than being necessary for) the instantiation of one of the abstract conditions in the completely instantiated antecedent of a causal law’ (p. 291). This risks minimising the role of necessity within the NESS test to such an extent that NESS may lose practical value as a test because it is so inclusive. Wright makes the bold claim that the NESS test ‘captures the essence of causation and gives it a comprehensive specification and meaning’ (p. 322). This can be contrasted with Miller who insists that the aims of the NESS test are ‘modest’ (p. 323); to offer a coherent solution in situations where the ‘but for’ test fails. He maintains that the NESS test does not claim to offer a complete account of causation, noting that this is a task that defeats philosophers. Indeed, by adopting more modest aims, the NESS test seems more readily defensible against criticisms. As noted above, Hogg explains that the NESS test has many advantages over the ‘but for’ test because it is able to address a range of causal scenarios that the ‘but for’ test cannot resolve. Its widespread adoption would therefore surely be a positive legal development. In his earlier chapter, Hogg noted that ‘it is a fruitless endeavour seeking to identify a single reported Scots judgment of recent years where the court has referred to academic writing on the subject of causation’ (p. 43), and that there are no references to the NESS test of causation. If the NESS test is to overcome judicial reticence, then it must not only be robust but also be accessible, and taken together the chapters in this book address the finer details whilst also making the NESS test more readily understandable. There are many other interesting chapters in this collection, addressing different aspects of causation, and drawing lessons from other jurisdictions, notably Fairgrieve and G'Sell Macrez's contribution on causation on French law.23 It is clear that proof of causation will continue to present new challenges, and by drawing together analysis from a range of perspectives this collection makes an invaluable contribution to understanding those challenges and to developing solutions for them. 1 Rt Hon Lord Hoffman, ‘Causation’, Ch 1. 2 See R Wright, ‘Causation in Tort Law’ (1985) 73 California L Rev 1735; R Wright, ‘Causation, Responsibility, Risk, Probability, Naked Statistics, and Proof: Pruning the Bramble Bush by Clarifying the Concepts’ (1987) 73 Iowa L Rev 1001. 3 M Hogg, ‘Developing Causal Doctrine’, Ch 3. 4 Bonnington Castings v Wardlaw [1956] AC 613 (HL). 5 R Wright, ‘The NESS Account of Natural Causation: A Response to Criticisms’, Ch 14; C Miller, ‘NESS for Beginners’, Ch 15. 6 Fairchild v Glenhaven Funeral Services (t/a GH Dovener & Son) [2002] UKHL 22, [2003] 1 AC 32. 7 J Morgan, ‘Causation, Politics and Law: The English – and Scottish – Asbestos Saga’, Ch 4. 8 Rutherford v Owens-Illinois Inc 941 P2d 1203 (Cal 1997). 9 J Sanders, ‘Risky Business: Causation in Asbestos Cancer Cases (And Beyond?)’, Ch 2. 10 Rothwell v Chemical and Insulating Co Ltd and another [2007] UKHL 39, [2007] 3 WLR 876. 11 Damages (Asbestos-related Conditions) (Scotland) Act 2009. 12 Scots Parliament Justice Committee, Official Report, 2 September 2008, col 1065–1074; M Hogg, ‘Asbestos Related Conditions and the Idea of Damage in the Law of Delict’ (2008) 31 Scottish L Times 207. 13 R Goldberg, ‘Using Scientific Evidence to Resolve Causation Problems in Product Liability: UK, US and French Experiences’, Ch 8; P Feldschreiber, L-A Mulcahy and S Day ‘Biostatistics and Causation in Medicinal Product Liability Suits’, Ch 9; RW Wright, ‘Proving Causation: Probability versus Belief’, Ch 10. 14 See further R Goldberg, Causation and Risk in the Law of Torts: Scientific Evidence and Medicinal Product Liability (Hart Publishing: Oxford, 1999) 38–43. 15 R Brown, ‘Inferring Cause in Fact and the Search for Legal “Truth”’, Ch 5. 16 AP Dawid, ‘The Role of Scientific and Statistical Evidence in Assessing Causality’, Ch 7. 17 AB Hill, ‘The Environment and Disease: Association or Causation’ (1965) 58 Proc R Soc Med 295. 18 McTear v Imperial Tobacco Ltd 2005 2 SC 1. 19 XYZ v Schering Health Care Ltd & others [2002] EWHC 1420, (2002) 70 BMLR 88 (QB). 20 CF Cranor, ‘The Challenge of Developing Science for the Law of Torts’, Ch 13. 21 RW Wright, ‘The NES Account of Natural Causation: A Response to Criticisms’, Ch 14. 22 C Miller, ‘NESS for Beginners’, Ch 15. 23 D Fairgrieve and F G'Sell-Macrez, ‘Causation in French Law: Pragmatism and Policy’, Ch 6. © The Author [2013]. Published by Oxford University Press; all rights reserved. For Permissions, please email: journals.permissions@oup.com
TI - Perspectives on Causation
JF - Medical Law Review
DO - 10.1093/medlaw/fwt016
DA - 2014-01-01
UR - https://www.deepdyve.com/lp/oxford-university-press/perspectives-on-causation-Jlhj1EYUC5
SP - 131
EP - 137
VL - 22
IS - 1
DP - DeepDyve
ER -