England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HASTINGS COUNTY COURT
HIS HONOUR JUDGE HOLLIS
HS101225

		Royal Courts of Justice Strand, London, WC2A 2LL
		17/10/2008

LORD JUSTICE MAY:

Introduction

1. It is a commonplace that, if a passenger is injured in a collision between two motor vehicles when each driver's negligence was a material contributing cause of the collision and therefore the injury, the passenger can recover the full amount of his loss from either of the drivers, provided that he does not recover in total more than the full amount of the loss. The two drivers are left to sort out the appropriate contribution between each other under the Civil Liability (Contribution) Act 1978. This is a common illustration of how uncontentious law treats the necessary causation element of the tort of negligence. Causation can, however, sometimes gives rise to difficulties when it is suggested that the injury for which the claimant claims damages resulted from, as the defendant may contend, more than one cause operating cumulatively or consecutively. The defendant may then contend, as in the present case, that he should only be liable for a proportion of the claimant's loss apportioned to the cause for which he is liable. It may be noted in passing that, if one driver in the collision case is injured and claims damages against the other driver, the claimant driver's damages may be reduced for his own contributory negligence. But we are not concerned with contributory negligence in the present appeal.

Facts

2. The claimant, Christopher Ellis, was born on 26^th October 1967. He worked for the defendants, the Environment Agency, as a plant machinery driver. On 9^th June 1998, he had an accident in the course of his employment. He was preparing to drive a loading shovel close to the place at Pevensey Bay in East Sussex where William the Conqueror landed with his army in 1066. Mr Ellis needed to remove a protective shutter from the front of the machine. While he was doing this, a gust of wind caught the shutter so that he fell 7 or 8 feet to the ground. As a result he injured his back. In proceedings begun in June 2001, he claimed that this accident was caused by the negligence and breach of statutory duty of the defendants. The defendants by their defence admitted that, if the accident had taken place in the manner alleged, it had been caused by their negligence. They contended that Mr Ellis' own negligence had contributed to the accident. But HH Judge Hollis, in a judgment given in the Hastings County Court on 30^th March 2007, held that the defendants had not proved any contributory negligence on Mr Ellis' part. This appeal is brought against other parts of the judge's decision, but his rejection of the case of contributory negligence is not challenged.
3. An event which in part gives rise to problems with causation in this case happened on 30^th April 2000. On that day, Mr Ellis had a fall on his staircase at home because his back gave way. The main enduring injury from this fall was to his right knee, where there was a complete rupture of the anterior cruciate ligament and a tear of the medial meniscus. This required, among other things, an operation under general anaesthetic in March 2001. Mr Ellis remained in employment with the defendants until 2004, but did not work after March 2001. He has suffered continuing problems with his knee and other debilitating consequences, including bouts of depression.
4. Mr Ellis' case was that the April 2000 fall was a consequence of the June 1998 lumbar spine injury. The defendants by their defence denied this. They pleaded that in so far as the April 2000 fall was in any way related to Mr Ellis' back condition it was caused by accidents which Mr Ellis sustained after the June 1998 accident. They referred in particular to another accident which Mr Ellis had at work on 5^th May 1999, when he missed his footing on a ladder, after which he was off work for 5 months. The defence also refers to other incidents in 1999 which did not feature greatly in the causation debate other than as narrative. They do not affect the point of principle which this appeal raises. I understand that Mr Ellis brought a claim for the May 1999 accident which was struck out on a limitation issue. Thus any consequences of this accident are not the responsibility of the defendants.
5. The yet further complication is that upon orthopaedic examination Mr Ellis was shown to have had in 1998 a degree of undiagnosed and symptomless spinal degeneration to a greater extent than is usual in 30 year old men. This would have developed quite apart from the June 1998 accident, so that Mr Ellis would not have been able to undertake heavy work for his full working life. His case was that the June 1998 accident brought about symptoms in his back some 10 years before they would have occurred but for that accident. We are told that the judge in essence accepted this when he quantified Mr Ellis' damages on this reduced basis.

The judge's decision

6. In section D of his judgment, the judge considered the extent of damage for which the defendants should be responsible. He said that the central question was the extent to which the June 1998 accident caused Mr Ellis' present condition and the extent to which the April 2000 fall down the stairs was a consequence of the injury Mr Ellis received when he fell off the shovel. The judge noted that the knee injury from the April 2000 fall was that which had had the most serious effect. The judge's conclusion, in paragraph D(vi) of his judgment, included that there was

"… a continuum … of significant back problems … arising from the serious fall which constituted the 1998 accident … which ultimately led to the 2000 fall down the stairs, and that but for that accident the fall would not have happened."

The judge thus rejected the defendants' case that the June 1998 accident had not been causative of the April 2000 fall. So far as it goes, there is no appeal against that finding. But the defendants say that there were other causes of the April 2000 fall as well.

7. The judge held that Mr Ellis was entitled to recover 90% of his loss. He directed himself (paragraph D(i) of the judgment) that Mr Ellis had to prove that but for the June 1998 accident the April 2000 fall would not have taken place and that there was a direct causative link between the two events. He found, in the passage in paragraph D(vi) to which I have already referred, that Mr Ellis had proved this. He further held (paragraph D(vii)) that it was axiomatic that a tortfeasor is only responsible for the consequences of the tortious act and that the compensation should be appropriately adjusted to take account of the medical evidence that but for the injury Mr Ellis would in any event have encountered lumbar problems after about 10 years and serious disability after 20 years. As I have said, the judge's quantification of the damage took account of this. But there was an intervening event, that is the May 1999 accident, which was not the defendants' responsibility and which was, on Mr Staniforth's evidence, a further contributing event in the deterioration of Mr Ellis' lumbar spine, and hence a "contributory intervening factor" in his fall down the stairs in April 2000. On the basis of Mr Staniforth's assessment, the judge believed it would be right to discount the total award by 10% to take into account that "further contributory factor".
8. The amount which, after a further quantum hearing, the judge by order dated 19^th November 2007 awarded to Mr Ellis was £290,404 plus interest. This represented 90% of the full award. The order is expressed to be made upon the parties having reached agreement, on the basis of the judge's findings at the trial and the further hearing, that the damages were properly calculated as at 25^th July 2007. We were told that elements of the £290,404 for which judgment was given took account of the judge's finding in paragraph E(iii) of his March 2007 judgment that Mr Ellis, with his wife's support, should be expected to be in some form of employment by 12 months from the final disposal of the case.

One ground of cross-appeal

9. The relevance of this last point is that Mr Ellis, by Respondent's Notice, seeks to have elements of the quantified damages increased because this appeal has deferred the final disposal of the case. It is convenient to deal with this point at this stage. Mr Leonard submits that it is difficult to imagine anything less conducive to the settlement of Mr Ellis' condition in preparation for his return to the job market than having the appeal hanging over him. The damages were eventually agreed, as the order recites, on the basis of the judge's findings as at 25^th July 2007. The final disposal of the case should be taken, as thing have turned out, as the final disposal of this appeal. The 12 months to which the judge referred should therefore run from 29^th July 2008 (when this court announced in anticipation what the main result of this appeal would be) in substitution for 25^th July 2007.
10. There is, in my view, no proper basis for this contention. Mr Ellis does not, so it appears, seek to appeal any decision which the judge made which this court might hold to be wrong, because the relevant elements of the damages calculation were agreed between the parties; and Mr Leonard accepted that the agreement contained no reservation that the damages should be recalculated if there was an appeal which delayed the final disposal of the case. In short, the judge indicated how loss of earnings and similar heads of loss should be addressed. The parties then made an agreement as to the date from which calculations should be made. Those calculations found their way into the order. Neither the agreement nor the order reserved the possibility of recalculation. I would reject this part of the cross-appeal.

Grounds of appeal and cross appeal

11. The defendants appeal against the judge's apportionment of 90% of the loss. Smith LJ gave hesitant permission to appeal on a renewed oral application. Mr Ellis cross-appeals contending that he should have been awarded 100% of his loss. In summary, the respective cases of the parties before this court are as follows.
12. Mr Leonard, on behalf of Mr Ellis, contends that the June 1998 accident, for which the defendants accept liability, was a sufficient material cause of the April 2000 fall, and that Mr Ellis is entitled to 100% of his loss consequent upon both accidents. The May 1999 accident was no more than part of the chain of causation which on the evidence had no causative effect. That loss would embrace his back problems from June 1998 and the damage to his right knee in April 2000. The quantification of the loss should take account of Mr Ellis' symptomless pre-existing back degeneration which would have resulted in back problems anyway at an earlier age than normal.
13. Mr Alliott, on behalf of the defendants, submits that the award of damages to Mr Ellis should take account of three causes and that there should be an apportionment between each of them. The three causes here are the pre-existing spinal degeneration, the June 1998 accident and the May 1999 accident. As between these three causes, 70% should be attributed to the pre-existing degeneration, 20% to the June 1998 accident, and 10% to the May 1999 accident. These percentages derive from a part of the evidence of Mr Staniforth, the parties' single joint orthopaedic consultant expert, whose evidence I shall refer to later in this judgment. Mr Alliott has a fall back submission that, if the damage should not be proportionately reduced on account of the pre-existing spinal degeneration, there should be apportionment of two-thirds to the June 1998 accident and one-third to the May 1999 accident. These proportions derive from the 20% and 10% of the apportionment in Mr Alliott's main submission. Thus Mr Alliott's main submission is that Mr Ellis should recover only 20% of his loss; his fallback submission is that he should recover two-thirds of the loss.

The orthopaedic evidence

14. Mr Staniforth was the single joint orthopaedic expert. As things turned out, his evidence taken as a whole was not entirely satisfactory. This arose largely from the nature of what he was expected to do. He wrote no less than four substantial reports dated 5^th March 2001, 16^th May 2002, 6^th April 2004 and 27^th July 2005. He also wrote at least four letters, the later reports and the letters being written in part in response to letters or questions from the parties' solicitors. He then gave oral evidence, was cross-examined and re-examined. Taken as a whole, this body of evidence was not entirely consistent, such that each party is able to point to passages in it in support of their respective cases. For example, in his first report, he expressed the opinion that on the basis of the evidence he had seen and on the balance of probabilities, the accident in June 1998 resulted in pain and insecurity in the lower back and legs which ultimately led to Mr Ellis falling downstairs in April 2000. Part of the information then given to Mr Staniforth was inaccurate because, as the judge was to hold, Mr Ellis was mistakenly attributing to the June 1998 accident some of the time off work and other problems which medical notes showed happened in 1999 or which were not attributable to his back.
15. The details of the evidence of Mr and Mrs Ellis on this topic are in paragraph C(ii) of the judge's judgment, where the judge's conclusion was that he did not consider that Mr Ellis was inaccurate in describing his work pattern as being "on and off" and that this was partly attributable to the June 1998 back injury. The judge accepted Mrs Ellis' evidence that it was at this time that Mr Ellis would on occasions return from work, have to lie flat on the floor due to his back discomfort, and would be "up and down for a lot of the night", although he would have some good days. She was adamant, said the judge, that she was not confusing the two accidents. The judge also in paragraph D(i) accepted the central veracity of Mr and Mrs Ellis in their accounts of Mr Ellis' condition between June 1998 and May 1999.
16. In his second report of May 2002, Mr Staniforth said that there was no new evidence which affected his opinions concerning the causation of the two injuries sustained in separate accidents. There were then two letters in September and October 2002, which the judge referred to in some detail in paragraphs D(iii) and D(iv) of the judgment, in which Mr Staniforth apparently qualified his opinion to the effect that lack of reference to back problems in the GP notes made it more difficult to substantiate his previously held opinion; and that after reconsideration he would attribute 70% of the cause of the April 2000 accident to the disease process in Mr Ellis' lumbar spine to which the June 1998 accident added 20% and the "last straw" May 1999 accident a further 10%. The judge recorded Mr Staniforth's feeling that the apportionment exercise he was asked to consider was an artificial one.
17. Mr Staniforth was then to write in his third report (paragraph 50) after further questioning that on the balance of probabilities Mr Ellis' back would not have given way resulting in the right knee injury in April 2000 if he had not had the accident in June 1998; and in a letter of 21^st July 2004 (page 6) that on the balance of probabilities the incident of June 1998 caused significantly more damage/mechanical derangement than that which occurred in May 1999, and that the June 1998 back injury rendered Mr Ellis more vulnerable to sustaining a further back injury. Mr Staniforth further expressed the belief (page 7) that the May 1999 injury would not of itself have caused any significant symptoms in the lower back if Mr Ellis had not previously been injured; that if he had not sustained a back injury in June 1998, the jarring injury of May 1999 would have been relatively minor and would not have contributed anything to the knee injury of April 2000; that the contributing effect of wear and tear and accumulation of spinal derangement from May 1999 onwards would have been minimal compared with what they were if Mr Ellis had not sustained the June 1998 injury; and that if he had not sustained that injury, none of the subsequent events would have occurred especially the April 2000 knee injury. Mr Alliott says that answers given by Mr Staniforth in cross-examination helped the defendants' case. Mr Leonard says that answers given by Mr Staniforth in re-examination helped Mr Ellis' case.
18. The judge was, in my view, right to say in paragraph D(v) of the judgment that clearly Mr Staniforth was having problems in his efforts to assist the parties and the court in expressing an opinion on the difficult questions of causation and contribution bearing in mind the nature of the underlying problem, the nature and immediate effect of each accident and the relative violence of each accident. He was an experienced and respected expert, and the judge felt he was justified in finding the exercise an artificial one. The judge was also, I think, right to say, in paragraph D(vi), that of course the answers that Mr Staniforth was giving were couched in terms that reflected the nature of the questions that were being asked.
19. The judge was also, in my judgment, fully entitled on the evidence before him (subject to questions of law which I discuss below) to reach his central conclusion of fact in paragraph D(vi) to which I have already briefly referred. In so far as Mr Alliott was looking to chip away at the factual conclusions, he was not, I think, properly able to do so in this court. The factual conclusions were the product of an evaluation by the judge of written and oral expert evidence which he heard and this court has not. On the face of it, the evidence, taken as a whole, was capable of sustaining the conclusion. This court should not interfere with the product of such an evaluation unless it was plainly wrong, which in my judgment it was not.

Causation

20. A fuller version of paragraph D(vi) of the judge's judgment is as follows:

"[Mr Staniforth] accepted that his original conclusion was based on the account that he had had from Mr Ellis. … Although the record shows that what actually happened between 6/98 and 5/99 was not as bad as was indicated in those accounts, and did not include any referral beyond one to the osteopath, I have found that overall Mr Ellis has given an honest account of his position. It was Mr Staniforth's clear evidence that the "ratcheting effect" was initiated by the 6/98 accident and he accepted too that the record of days off may not make much difference in assessing the back condition at the time unless there is objective evidence of a complete recovery. But Mr Ellis was having days off and has always said that he was. Having reviewed all the evidence I am satisfied on the balance of probabilities that there was a continuum (to use Mr Leonard's term) of significant back problems in this case arising from the serious fall which constituted the 1998 accident and which continued after his two visits in 7/98 to the osteopath and which ultimately led to the 2000 fall down the stairs, and that but for that accident the fall would not have happened. This was a man who had not experienced back problems until 6/98, and then did so. These worsened after the, apparently more minor, accident of 5/99 and led in time to the twinge which caused his fall down the stairs in 2000. I have no doubt that had Mr Ellis been arguing for 100% liability on the basis of the 5/99 accident alone then the defendant would, and with justification, have been making much of the evidence of a pre-existing problem, arising from the 1998 accident."

21. The judge's decision, as expressed in paragraph 20, is an orthodox application of the usual rule for causation that, in order to recover damages for negligence, a claimant must prove that but for the defendant's wrongful conduct he would not have sustained the harm or loss in question. Recent clear authority for this may be found in Fairchild v Glenhaven Funeral Services [2003] 1 AC 32, [2002] UKHL 22, where the general rule was not in doubt, but the question was whether in the exceptional circumstances of that case there should be a variation or relaxation of the rule so as to depart from what Lord Nicholls of Birkenhead referred to in paragraph 41 as the usual threshold "but for" test of causal connection – for the general rule, see for example paragraphs 8 and 9 of the opinion of Lord Bingham of Cornhill and paragraphs 37, 38, 41, 43 and 45 of the opinion of Lord Nicholls. The problem in Fairchild was that claimants who developed mesothelioma caused by exposure at work to asbestos dust had been exposed to asbestos dust during periods of employment with more than one employer. They were unable to prove on the balance of probabilities that the onset of their mesothelioma was attributable to any particular or cumulative wrongful exposure. The House of Lords held that a modified approach to causation was justified. Among the many cases which the House of Lords considered were Bonnington Castings v Wardlaw [1956] AC 613 and McGhee v National Coal Board [1973] 1 WLR 1, from which claimants have regarded themselves as able to argue that it is sufficient to prove that a defendant's negligence materially contributed to the claimant's loss; and Wilsher v Essex Area Health Authority [1988] AC 1074, in which the House of Lords upheld the dissenting judgment of Sir Nicholas Browne-Wilkinson V-C that the baby claimant's claim could not succeed where the defendants failed to take reasonable precautions to prevent one of five possible causative agents, but no one could tell that that agent was more likely than any of the other four to have caused the baby's condition. As Lord Bingham said in paragraph 22 of Fairchild, it is one thing to treat an increase in risk as equivalent to the making of a material contribution where a single noxious agent is involved, but quite another where any one of a number of noxious agents may equally probably have caused the damage.
22. I take Fairchild thus briefly because, apart from containing clear expressions of the normal rule of causation, it is not directly relevant to or illustrative of the facts of the present case. What it does show is that departure from the normal rule is exceptional and to be approached with caution; and that a modified approach to proof of causation was justified to assist a highly deserving claimant who from the nature of the facts had evidentiary difficulty in proving that the negligence of one of two or more possible tortfeasors, whose negligence may have operated successively, satisfied on the balance of probabilities the "but for" test. The present case is not concerned with two or more successive tortfeasors, nor essentially with helping Mr Ellis to establish that, but for the defendant's negligence, the April 2000 accident would not have happened. On the judge's findings, Mr Ellis has established all this. The question is whether Mr Ellis' damages should be reduced because, it is said, there were two other contributory causes for neither of which was a tortfeasor (or Mr Ellis) responsible.
23. Before considering the two principal authorities on which Mr Alliott relies, it is instructive to look at the recent decision of this court in Clough v First Choice Holidays [2006] EWCA Civ 15. In that case, the 26 year old appellant suffered catastrophic injury on holiday when he slipped from a wall and broke his neck in a swimming pool. The horizontal surface of the pool wall was not coated with non-slip paint, but with a coating which was less effective to minimise the risk of slipping. The first instance judge had held that the failure to use non-slip paint was a negligent breach of duty by the respondents for the purpose of the contractual arrangements between them and the appellant, and a breach of contract. A physical barrier should have been provided around the wall for the protection, not of adults, but of children. The judge further held in summary that the claimant had decided to go to the pool wall, for whatever reason, quite significantly affected by drink and that he had to take the risks of doing so on his own shoulders. He slipped on a surface which should have been, but was not, painted with non-slip paint. His feet were wet, but by walking where he did and being where he was, he was not doing anything abnormal nor prohibited by the rules of the complex. The claim failed at first instance on causation, the judge concluding that the negligence and breach of duty lacked causative potency. In short, he reached this conclusion by reference to a number of features of the evidence including the claimant's intoxication and a conclusion that it was virtually inevitable that the best quality of non-slip paint would not have made the surface of the wall completely non-slip. The judge held that a relaxed approach to causation such as in Fairchild could not apply to the case before him, and that it was impossible to conclude that but for an essentially unquantifiable increased risk of slipping from not using non-slip paint, the claimant's slip would not have occurred.
24. It was submitted in this court on behalf of the appellant claimant that exceptionally the omission to use non-slip paint should be treated as having made a material contribution to the accident, and further that a material contribution to the risk of slipping should be sufficient where as a matter of policy the court might exceptionally find causation proved even though, due to the limits of current knowledge, the evidence did not permit a factual inference that the increase in risk materially contributed to the damage. These submissions were made with reference to Bonnington Castings, Wilsher, McGhee, Fitzgerald v Lane [1987] 3 WLR 249 and Fairchild, all of which cases this court examined in detail. The single substantive judgment is that of the President of the Queen's Bench Division, with which Richards and Hallett LJJ both agreed. The President's judgment has this in paragraphs 43 to 45:

"The authorities to which Mr Burton drew attention establish that the "but for" test, applied in its full rigour, should no longer be treated as a single, invariable test applicable to causation issues, in whatever circumstances they may arise. The question in the present appeal is whether Fairchild and the series of decisions developing the law of which it represented the culmination (subject of course to subsequent developments) have any application here. In my judgment, in agreement with the judge, they do not. On any view, it would be absurd to describe this unfortunate accident as exceptional. Accidents like this happen all too frequently, and even though negligence by an identified tortfeasor is established, the question still remains whether the negligence caused the claimant's injuries. A successful claim for damages for personal injuries consequent on negligence or breach of duty requires the court to be satisfied that the injuries were indeed consequent on the defendant's negligence. Even if it may have some application in different situations, the distinction sought to be drawn by Mr Burton between material contribution to damage and material contribution to the risk of damage has no application to cases where the claimant's injuries arose from a single incident. In this Court any modification of the principles relating to causation in the context of claims for damages for personal injury must be approached with the greatest caution. Certainly, however the law of causation may develop, save in the House of Lords, it cannot develop in a way which revives or is dependent on the approach adopted by Mustill LJ in Wiltshire, and subsequently twice rejected in the House of Lords. That route is closed. In reality, for the purposes of cases like this, trite law is unchanged.

The breadth of Mr Burton's submission may have distracted attention from a single but potent consideration. In the context of causation, the two words "but for" are shorthand. They encapsulate a principle understood by lawyers, but applied literally, or as if the two words embody the entire principle, the words can mislead. They may convey the impression that the claimant's claim for damages for personal injuries must fail unless he can prove that the defendant's negligence was the only, or the single, or even, chronologically the last cause of his injuries. The authorities demonstrate that such an impression would be incorrect. The claimant is required to establish a causal link between the negligence of the defendant and his injuries, or, in short, that his injuries were indeed consequent on the negligence. Although, on its own it is not enough for him to show that the defendant created an increased risk of injury, the necessary causal link would be established if, as a matter of inference from the evidence, the defendant's negligence made a material contribution to the claimant's injuries. As Lord Rodger explained and demonstrated in Fairchild, there was "nothing new" in Lord Reid's comment in Bonnington that what was required was for the plaintiff to make it appear at least "that, on a balance of probabilities, a breach of duty caused, or materially contributed to his injury". Lord Rodger observed that there was ample authority for the proposition in English and Scots law, both before and after Lord Reid had, in effect, treated it as so elementary that it required no support from authority.

This, as it seems to me, was precisely reflected in the approach taken by the judge. In paragraph 74, he referred to the "but for" principle, but accepted, alternatively, that it would be sufficient for the claimant to prove "on the balance of probabilities, that the absence of proper non-slip paint caused or materially contributed to his slip and subsequent fall". In short the judge correctly identified the appropriate principle, and rightly rejected the suggestion made to him that Fairchild had any application to the present case."

25. Thus Clough again shows that the shorthand "but for" test, which Mr Alliott would have us depart from in the present appeal, is only not applied in exceptional cases (which Clough decidedly was not). It is also an example of an attempt, which in that case failed, to relax the test for causation where a highly deserving claimant had evidential difficulties in establishing causation to the normal standard. Furthermore, Clough was a single accident case, not one in which the negligence of more than one successive tortfeasor may (or may not) have caused the claimant's loss.
26. The two authorities on which Mr Alliott chiefly relies to persuade the court to move away from the "but for" test are Holtby v Brigham & Cowan (Hull) Ltd [2000] ICR 1086 and Allen v British Rail Engineering Ltd [2001] ICR 942.
27. In Holtby, the claimant was diagnosed in 1996 as suffering from asbestosis, having been exposed to asbestos dust at work for many years from 1942 to 1981. For approximately half this time he was employed by the defendants, for the remainder with various other employers. The extent of exposure to asbestos dust was similar throughout the period. The judge found that the defendants had been negligent and had materially contributed to the claimant's asbestosis, but he reduced the claimant's general damages by 25% because the exposure to asbestos dust was cumulative and the defendants were only to be liable for that proportion of the damage which they had caused. This court dismissed the claimant's appeal, the majority (Stuart-Smith and Mummery LJJ) holding that the defendants were only liable to the extent of their quantifiable contribution to the claimant's asbestosis.
28. Stuart-Smith LJ, in paragraph 1 of the judgment, expressed the issue, which surprisingly did not appear to have been considered previously by this court, as this:

"Where a claimant suffers injury, in this case asbestosis, as a result of exposure to a noxious substance by two or more persons, is the defendant who is one of those persons whose tortious act has made a material contribution to the injury, liable in respect of the whole resulting disability, subject only to such rights as he has against other tortfeasors? Or is he liable only to the extent that he has contributed towards the disability?"

29. It was submitted on behalf of the claimant that, if he proved that the defendant's conduct made a material contribution to his disease, he was entitled to recover all his loss from that defendant, notwithstanding that others may have contributed as well. Reliance was placed on Bonnington Castings and McGhee. But Stuart-Smith LJ said of each of these cases that the House of Lords did not consider what the extent of the defendants' liability was. The claimant had also referred to Milner v Humphreys and Glasgow (an unreported decision of Longmore J, as he then was, on 24^th November 1998). In Milner the claimant contracted asbestosis after an overall exposure of 4 years and 10 months, with six different employers, only 7 months of which were with the defendant. The same claimant's counsel who appeared in Holtby had advanced the same argument as was for consideration in Holtby. Longmore J said this:

"Many diseases depend on cumulative exposure; many diseases may have more than one cause; some causes may be tortious in origin and some may be non-tortious. It may be difficult to separate the two. Where, as in Thompson v Smiths Shiprepairers (North Shields) Ltd [1084] ICR 236, the operations of a single employer or the same operations of two consecutive employers may have been tortious only after a certain date justice may require and the facts may permit an apportionment to be made so that the tortious employer is not held liable for the consequences of non-tortious conduct of himself or [the conduct, tortious or otherwise of] another. Such cases do, however, present quite serious factual difficulties and the law has been concerned to ensure that a meritorious plaintiff does not fail for want of proof.

…

I do not think that this argument can be accepted in the broad form in which it was put. As Mustill J pointed out in Thompson's case, at p. 273, the Bonnington principle is but a variant of the principle that, where an injury is indivisible, any tortfeasor whose act has been a proximate cause must compensate for the whole injury, leaving the tortfeasor to sort out with other possible tortfeasors any appropriate claim for contribution: see Dingle v Associated Newspapers Ltd [1961] 2 QB 162, 188, per Devlin LJ. Where there are causes concurrent in time, the likelihood is that a resulting injury will be indivisible; but where causes are sequential in time, it is not likely that an injury will be truly indivisible especially if (as I do not think [Dr Howard] contested) the injury is a disease which can get worse with cumulative exposure. The Bonnington Castings case can, nevertheless, assist the plaintiff to this extent; the principle, as formulated by Mustill J, is that where it is proved that a wrongful act has made a material contribution to the plaintiff's injury, the law regards this as sufficient discharge of the plaintiff's burden of proof on causation to render the defendant liable for the injury in full. That does not mean that no question of apportionment can ever arise but it does, in my judgment, mean that, unless the defendant pleads and proves facts which justify apportionment, the plaintiff can recover in full."

30. Stuart-Smith LJ in Holtby did not accept the submissions on behalf of the claimant. He said in paragraph 20:

"In my judgment, as the passages cited from the three House of Lords decisions show, the onus of proving causation is on the claimant; it does not shift to the defendant. He will be entitled to succeed if he can prove that the defendant's tortious conduct made a material contribution to his disability. But strictly speaking the defendant is liable only to the extent of that contribution. However, if the point is never raised or argued by the defendant, the claimant will succeed in full as in Bonnington Castings Ltd. V Wardlaw [1956] AC 613 and McGhee v National Coal Board [1973] 1 WLR 1. I agree with Judge Altman that strictly speaking the defendant does not need to plead that others were responsible in part. But at the same time I certainly think it is desirable and preferable that this should be done. Certainly the matter must be raised and dealt with in evidence, otherwise the defendant is at risk that he will be held liable for everything. In reality I do not think that these cases should be determined on onus of proof. The question should be whether at the end of the day, and on consideration of all the evidence, the claimant has proved that the defendant is responsible for the whole or a quantifiable part of his disability. The question of quantification may be difficult and the court only has to do the best it can using its common sense, as Lord Salmon said in the passage cited. Cases of this sort, where the disease manifests itself many years after the exposure, present great problems, because much of the detail is inevitably lost. I can see that in Borel's case, 439 F.2d 1076 where the defendants were manufacturers as opposed to employers the position may be particularly difficult. But, in my view, the court must do the best it can to achieve justice, not only to the claimant but also to the defendant, and among defendants."

Stuart-Smith LJ said that this had been the approach of Mustill J in Thompson v Smith's Shiprepairers, an industrial deafness case in which the defendant employers were not guilty of negligence until 1963. A substantial part of the claimant's hearing impairment took place before the defendants were in breach. Mustill J saw no reason why the present impossibility of making a precise apportionment should in justice lead to the result that the defendants were adjudged to be liable to pay in full when it was known that only part of the damage was their fault. The court should make the best estimate that it can making the fullest allowance in favour of plaintiffs for the uncertainties known to be involved in any apportionment. Stuart-Smith LJ said in paragraph 22 that evidential difficulties of apportionment in Thompson were not present in Holtby since for asbestos the progression was linear depending on the amount of dust inhaled. All dust contributed to the final disability.

31. Clark LJ (as he then was) dissented in part in Holtby but agreed that the appeal should be dismissed. He did not consider that justice demanded that the burden of establishing an appropriate apportionment in cases of successive exposure should be on the claimants. He also considered that it was at least arguable on the basis of previous decisions that in a case where the claimant proves that two employers had made a material contribution to his condition, he is entitled to judgment in full against each, leaving them to contest issues of contribution between themselves. "That would certainly be the case where the injury was truly indivisible so that each made a material contribution to the same damage, as in a case of damage caused by, say, a collision." The present case is essentially, in my view, an indivisible single injury case. Apart from the immediate back pain following the June 1988 accident and before the April 2000 accident, the main loss which Mr Ellis has suffered is, and is only, that which followed and resulted from the April 2000 accident.
32. In Allen v British Rail Engineering Limited, the claimants worked for the defendants using percussive tools, the first claimant from the late 1950s. In consequence, from 1968 the first claimant suffered from "vibratory white finger". In 1987 his employment with the defendants came to an end, but he continued in employment which involved vibrating tools and which caused further damage. The claimants made claims in negligence against the defendants, in which the judge found that, in the light of their knowledge in 1973, the defendants should have carried out surveys to ascertain the incidence of vibratory white finger and should have warned the claimants of the risk. She also held that by 1976 the defendants should have found alternative less damaging work for the first claimant, or should have reduced the time he spent exposed to vibration. She assessed the appropriate compensation for the whole of the first claimant's injury at £11,000 but awarded him only £4,000, deducting from the total £1,500 for the period before 1976, £1,500 for the period after 1987, and £4,000, that is half the remaining £8,000, for the period from 1976 to 1987 to take account of the damage already suffered and the fact that, on the evidence, if the defendants had complied with their duty, the claimant's exposure could have been reduced by half. The first claimant accepted the deductions for damage suffered before 1973 and after 1987, but disputed the further deduction of £4,000.
33. Schiemann LJ giving the judgment of the court, of which Latham LJ and I were the other members, expressed in paragraph 8 the issue of principle. It was contended for the first claimant that, once it had been shown that the defendants' negligence made a material contribution to the injury suffered by the claimant, the defendants were liable for the whole of the claimant's injury except in so far as it was shown or accepted that this had been aggravated by non-negligent exposure before 1973 or after 1987 by the first claimant's new employers. It was submitted on behalf of the defendants that in principle an employer was only liable for that part of the harm suffered by the employee which was attributable to the employer's negligence.
34. Having referred to the principal authorities considered in Holtby and to Holtby itself, Schiemann LJ concluded in paragraphs 20 and 21 as follows:

"In our judgment the case law as it now stands establishes five propositions of which the first is concerned with liability and the others with quantifying damages. (i) The employee will establish liability if he can prove that the employer's tortious conduct made a material contribution to the employee's disability. (ii) There can be cases where the state of the evidence is such that it is just to recognise each of two separate tortfeasors as having caused the whole of the damage of which the claimant complains; for instance where a passenger is killed as the result of a head-on collision between two cars each of which was negligently driven and in one of which he was sitting. (iii) However in principle the amount of the employer's liability will be limited to the extent of the contribution which his tortious conduct made to the employee's disability. (iv) The court must do the best it can on the evidence to make the apportionment and should not be astute to deny the claimant relief on the basis that he cannot establish with demonstrable accuracy precisely what proportion of his injury is attributable to the defendant's tortious conduct. (v) The amount of evidence which should be called to enable a judge to make a just apportionment must be proportionate to the amount at stake and the uncertainties which are inherent in making any award of damages for personal injury.

The application of those propositions should lead to a just and principled result. We mention by way of coda that this approach seems to accord with the view of the authors of the American Law Institute Restatement of the Law, Torts, 2d (1965), section 433A(e):

"Apportionment may also be made where a part of the harm caused would clearly have resulted from the innocent conduct of the defendant himself, and the extent of the harm has been aggravated by his tortious conduct.""

35. Mr Alliott submits that Holtby and Allen should be applied in the present case. The evidence established that Mr Ellis' injuries following the April 2000 accident resulted from three causes – the pre-existing spinal degeneration, the June 1998 accident and the May 1999 accident. The defendants are responsible for the second of these. Mr Staniforth's evidence sufficiently enabled the court to make an apportionment attributing 20% to the June 1998 accident. The "but for" test applied by the judge was inappropriate and produced an unjust result. Mr Alliott refers in support of this submission to paragraph 2-28 of Clerk & Lindsell on Torts (19^th Edition, 2006) for the proposition that

"Where it is possible to identify the extent of the contribution that the defendant's wrong made to the claimant's damage then the defendant is liable only to that extent, and no more."

Thompson v Smiths Shiprepairers is given as an example of this and there is then an extended discussion of Holtby and the main cases referred to in Stuart-Smith LJ's judgment in Holtby.

36. I have no hesitation in rejecting Mr Alliott's submission in so far as it seeks to attribute 70% of Mr Ellis' loss to his pre-existing undiagnosed back condition. This was not on the evidence properly to be seen as a cause of the April 2000 accident at all. For it was Mr Staniforth's evidence that, if Mr Ellis had not sustained the June 1998 accident, none of the subsequent events, especially the knee injury which occurred in April 2000 would have occurred. More significantly perhaps, the pre-existing spinal degeneration was a condition which affected Mr Ellis, not an event for which somebody (including Mr Ellis himself if he were careless) might or might not be responsible. It was not a cause in the sense that careless driving or negligent or non-negligent exposure to vibration may be a cause of damage. On the contrary, it was a condition akin to that which underlies the so called eggshell skull principle – see Smith v Leech Brain [1962] 2 AB 405 at 414 and Page v Smith [1996] AC 155 at 182, 189 – which Mr Alliott does not seek to challenge. Further, in so far as exceptional departures from the "but for" rule seek to do justice to both claimants and defendants, there is no injustice to the defendants in rejecting this submission, since Mr Ellis' damages fall to be reduced (and were reduced) because his back would have deteriorated progressively anyway apart from the June 1998 and April 2000 accidents. His damages were calculated on the basis that the June 1998 accident brought forward that deterioration by 10 years, thus taking the back condition into account in the defendants' favour. Yet further, Mr Leonard is correct to submit that the judge made no finding that but for the pre-existing condition the April 2000 accident would not have happened, nor did he adopt Mr Staniforth's 70%.
37. The question then is whether the judge was correct to apportion 10% of Mr Ellis' loss to the May 1999 accident for which the defendants are not responsible. The judge, who clearly intended to apply the "but for" test and in the main did so, gave no detailed reason in law for making this apportionment, although the 10% obviously came from Mr Staniforth's artificial apportionment. As my discussion has sufficiently shown, an application of the "but for" test would not result in any deduction. If that test is properly to be applied, the May 1999 accident, if it had any causative effect, would no more reduce Mr Ellis' damages than would the negligence of a second driver when the injured passenger claimed his full loss against a negligent first driver. As the President indicated in Clough, if the "but for" test is to be applied, the claimant who satisfies that test, as Mr Ellis on the judge's findings has, does not have to prove additionally that the defendants' negligence was the only, or the single, or even, chronologically the last cause of his injuries.
38. The judge held that the May 1999 accident was a "further contributory event" and hence a "contributory intervening factor". Mr Alliott rightly disclaims any suggestion that the May 1999 accident was a full blown intervening event of the kind referred to by Lord Bingham of Cornhill in Corr v IBC Vehicles [2008] 2 WLR 499 at paragraph 15. If it had been, the causative potency of the June 1998 accident would have ceased to have any effect, and Mr Ellis would have recovered no damages at all. There is, in my judgment, no free-standing principle which would give apportioning effect to a contributory intervening event. The expression appears to have overtones of contributory negligence which is not in point in the present appeal. The single question therefore is whether this case is one where exceptionally the Holtby and Allen principles should apply. In my judgment, they should not apply for the following reasons.
39. The normal rule for causation in personal injury negligence cases is the "but for" rule. The Holtby/Allen principles are an exception to the general rule intended to do justice in a particular class of case. Although at the fringes the delineation of the class of case may be debateable, in the main it has been applied and in this court at least should be limited to, industrial disease or injury cases where there has been successive exposure to harm by a number of agencies, where the effect of the harm is divisible, and where it would be unjust for an individual defendant to bear the whole of a loss when in commonsense he was not responsible for all of it. That is simply not Mr Ellis' case. His case is no more exceptional than was Clough. It is essentially a single accident case, where the main injury of April 2000, and that to which the May 1999 accident is said to have had a contributory effect, was a single indivisible event. The fact that Mr Staniforth was led to attribute artificially a small causative percentage to the May 1999 accident does not make the April 2000 accident divisible in the way that successive causes over a number of years in industrial cases may be capable of division. If the present case properly called for apportionment, which in my judgment it does not, the commonplace road accident, where each of two drivers is in part to blame, would also require apportionment, not only between defendant drivers but also between each of those drivers and the innocent claimant passenger. The law might in theory so provide, but it does not.
40. For these reasons, I would dismiss the appeal and allow the cross-appeal to the extent of removing the judge's 10% reduction of Mr Ellis' damages, so that he now becomes entitled to 100% in place of 90%.

Hallett LJ: I agree.

Lawrence Collins LJ: I also agree.