England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE CARDIFF COUNTY COURT
HIS HONOUR JUDGE SEYS LLEWELLYN QC

		Royal Courts of Justice Strand, London, WC2A 2LL
		23/05/2018

Lord Justice Hamblen:

Introduction

1. The appellant, Mr Philip Clay, brought a claim for damages for personal injury against the respondent, TUI UK Limited. The appellant's injury was suffered when he fell from a balcony at the hotel Guayarmina Princess, Tenerife ("the hotel"), where he and his family had gone on a package holiday booked with the respondent.
2. The appellant's claim was dismissed following a trial on liability before HHJ Seys Llewellyn QC in Cardiff County Court. In his reserved judgment dated 7 April 2016 the judge found that the appellant's claim failed as a matter of causation. It is against that finding that the appellant appeals. There is also an issue between the parties as to whether the judge made a finding of breach of duty and, if he did, the respondent challenges that finding by Respondent's Notice.

Factual and Procedural Background

3. In July 2011 the appellant, together with his family including his wife, their two sons (then aged 11 and 14), and his parents, went on a package holiday to the hotel provided by the respondent.
4. The appellant together with his wife and children occupied Room 358 which adjoined his parents' Room 357. Each room had its own balcony, accessible from the rooms via a sliding door which could be locked. These balconies were offset from each other. Underneath each balcony was a ledge. Between Room 358 and 357 there was a gap of 78 cm between the ledges, a gap of more than 123 cm between the exterior of the balconies, and a considerably larger gap between the inside to the handrails of the balconies. Rooms 357 and 358 were two storeys up with a drop of around 20 feet to the terrace below.
5. At around midnight on 20 July 2011, the appellant returned from dinner with his family and settled his children in bed for the night. At about 1am, he and his wife joined his parents on their balcony for a drink. Before leaving, they told their elder child where they were going and left the door to the balcony of Room 358 closed, but not locked, in case the children needed them.
6. After joining his parents on their balcony, the appellant went back into their room to use the bathroom. Upon returning to the balcony, he closed the door in order to prevent insects entering the room. The family described hearing a 'clicking' sound as the door closed and realised that it had inadvertently locked, trapping them on the balcony.
7. After trying without success to attract attention for about 30 minutes, the appellant endeavoured to step across from the balcony of Room 357 to that of Room 358. In doing so, he stood on the ledge underneath the balcony. The ledge gave way and he fell to the terrace below and was seriously injured, fracturing his skull. Fortunately, he has made a good recovery since.
8. The appellant brought a claim for personal injury against the respondent, relying on the the Package Travel, Package Holidays and Package Tours Regulations 1992. Under these regulations the respondent was liable to the appellant for the proper performance of the obligations under the contract, irrespective of whether those obligations were to be performed by the package travel holiday company, or by the hotel in Tenerife.
9. There was a 3 day trial before the judge in February 2016, at which both parties were represented by leading counsel.
10. At the trial the judge received factual evidence by witness statement and orally from the appellant, his wife Valerie Clay, and his parents Kenneth and Patricia Clay; and on behalf of the respondent factual witness evidence by witness statement and orally from Jose Luis Lima Dorta (Assistant Reception Manager at the hotel), Juan Carlos Garcia Martin (an architect/building surveyor/engineer for Princess Hotels in the Canary Islands), and Jesus Daniel Fuentes (Head of Maintenance at the hotel). Witness statement evidence was also received from the following witnesses whom the appellant did not require to be called: Maria De La Paloma Juliana Campomanes (hotel receptionist at the hotel that night); Juan Jesus Valencia Barrios and Jose Angel Gonzales Flores (security guards working at the hotel that night), and Enrique Aleman (a valet at the hotel in 2011).
11. The judge also received expert evidence on technical matters in written reports, a joint written statement and orally from Mr D Guillermo Mesas, instructed by the appellant, and from Mr Alvaro Montoya instructed by the respondent.
12. It was agreed that the respondent's liability fell to be judged by reference to standards local to the hotel in Tenerife – see, for example, Lougheed v On the Beach Limited [2014] EWCA Civ 1538.
13. The claim was brought on the basis that the respondent was liable for breaches of local standards by the hotel, namely: "(i) by failure to maintain the facilities, and in particular to maintain the lock to the sliding door, in its original and proper condition; (ii) in that the ledge on which the Claimant stepped was part of the overhanging balcony but was insufficient to hold his weight; and/or (iii) in that the hotel did not inform the Claimant of a potential risk within the premises, and about the safety measures adopted" [8].
14. The respondent denied breach of duty and contended that the appellant's act in climbing over the balustrade of the balcony of Room 357 so as to step across the gap to the balcony of Room 358 was "so unexpected and/or foolhardy as to be a novus actus interveniens" [19].

The judgment

15. The judge found that there was no breach of local standards in relation to the ledge on to which the appellant stepped. It was not required by applicable regulations or local construction practice to be weight bearing. The judge found at [76] that:

"….The standard here in question does not recognise expressly or impliedly that compliance with it may be insufficient. There is no other evidence that the construction or weight bearing capacity of this cornice failed to meet local standards. Just as there was no evidence of prior incident similar to that which befell the Claimant, whether at this hotel or at another in the Princess group or in the Canary Islands or Spain more widely, so also there was no evidence of a prevailing practice to construct a cornice with reinforced concrete such as would bear the weight of a person, or even illustration of it at other hotels or buildings…"

This finding is not challenged on appeal.

16. The judge also found that there was no breach of local standards in failing to give a warning that the ledge was not weight bearing or that guests should not attempt to jump or step across from one balcony to another. He found as follows at [121]-[122]:

"121. The case for the hotel is that there was no evidence of this being attempted in this large hotel of over 500 rooms, over 22 years of operation, or in any other hotel in the group, and that it was such a foolhardy act in the eyes of any reasonable hotelier that it was beyond sensible contemplation, or beyond contemplation as a risk which required warning: 'the hotel could not reasonably have foreseen that anyone would attempt to walk on those ledges'.

122. In my judgment the "sooner or later" argument would be easier to present in like circumstances in the case of a claim in negligence at common law in England and Wales. First, I accept above that it was not unusual in the Canary Islands to find decorative features to balconies such as the present. Second, I have no evidence other than statement of his opinion by Mr Mesas (without supporting evidence) that failure to give such warning in relation to the ledge or cornice was inconsistent with local practice or local recognition of that which was required. Third, in the present case, the expert evidence of Mr Montoya runs against recognition of an obligation to warn in respect of the cornice, and he was easily the more impressive of the two expert witnesses whom I heard. Fourth, and contrary to my extreme sympathy for the Claimant, I find it difficult to accept that those responsible at a hotel in the Canary Islands should have foreseen that a guest would climb onto the outer ledge. For these reasons above, I find myself unable to conclude that under local standards that warning in respect of the fragility of the cornice was reasonably demanded under Article 18 of the Ordinance."

This finding is also not challenged on appeal.

17. On the appellant's case the judge did, however, find that the locking mechanism of the sliding door in Room 357 was defective in breach of local standards and that the respondent was accordingly in breach of duty. This is disputed by the respondent.
18. On any view, the judge found that the claim failed as a matter of causation and this is the central issue on the appeal.
19. The judge described the novus actus interveniens defence in the following terms at [19]:

"The Defence pleads the exception to liability under the Regulations by Regulation 15(2)(c), that the travel company shall not be liable if any failure to perform the contract is due to unusual and unforeseeable circumstances beyond the control of the travel company, or an event which the travel company or the hotel even with all due care could not foresee or forestall. At trial, in skeleton argument and in submissions, this is put in more general terms that even if breach of contract or negligence were proved, the action the Claimant elected to take should be regarded as a novus actus interveniens.

20. He identified the relevant issue at [18] as being:

"(iii) Was the act of the Claimant, in climbing to the other side of the balustrade and preparing to jump or step across the gap to his own balcony, so unexpected and/or foolhardy as to be a novus actus interveniens?"

21. The judge addressed the issue of causation at [106]-[113] of his judgment as follows:

"106. Causation. There is no doubt the door clicked shut so as to lock out the Claimant and his wife and parents. On my findings above, it is possible to categorise this as a defect and thus to constitute a departure from the standard that premises and services must be kept at the standards required to obtain authorisation for 'touristic' activity (albeit so might a failure to replace furniture which had become shabby and threadbare). It requires independent consideration whether any defect in the locking mechanism can be said to have caused the accident itself. The Defendant submits that it was not, and that the accident injury was the product solely of the Claimant's own actions.

107. On the one hand, the Claimant and his parents and wife were impressively ordinary and considerate people, and of apparently careful background.

108. The Claimant himself was at the time of the accident employed as a security officer/fire office at an oil refinery in Milford Haven, he had prior to the accident had frequent health and safety training, and had been safety conscious on arrival at the hotel. Because it was something of a maze, he advised his children that if ever there were a fire they would have to make their way down from the balcony rather than risk trying to find their way through the hotel itself. The Claimant's wife was a nurse, who told me that she dealt with life and death, and would never have contemplated letting her husband attempt this if she had thought that he might fall.

109. His father was prior to retirement a truck operator instructor and also a health and safety representative for Esso. His first act on arriving at room 358 was to read the safety notice on the back of the hotel door and look for an escape route in case of an emergency, saying "this has been instilled in me over the years with my health and safety training" (witness statement paragraph 19 Bundle 2 page 528). The Claimant's mother was a retired nurse.

110. As individuals, they were decent people, with children/grandchildren, not single young men out on a spree. There was no relevant inebriation. The distance itself at the narrowest point between the exterior cornices of the respective balconies was 2 foot 7 inches and so if this had been at ground level or only a few feet above the ground it would have been a simple step, easily in the compass of someone 6 feet 2 as was the Claimant.

111. However any defect here in question, namely a readiness of the lock to snap closed in the locked position, was not a direct danger to those on the balcony, (unlike fragility of the ledge if this had been in breach of local standard). It was disquieting to be locked out and have no-one heed cries for attention. But there was no fire. There was no emergency. The minimum temperature that night was 19 degrees centigrade. The Claimant did not know and could not know that it was safe to stand on the ledge. The risk of injury, if it was not safe to stand on the ledge was obvious, and at two storeys up, life threatening. Unsurprisingly, the Claimant agreed that he would not have done this if he were on the top of a skyscraper.

112. Once the Claimant stepped on the ledge outside the balustrade, its fragility for his weight gave way under him like a trap door so as to plummet him to serious injury. It was this which caused the accident fall and injury. Whilst it may be that if there was breach of standard in relation to the weight bearing capacity of this ledge, this would not bar recovery, I consider that in relation to any defect in the lock of the sliding door, it was a strikingly new and independent act on the part of the Claimant. Even allowing for the open textured and pragmatic tests of causation in the law of England and Wales, I am driven to the conclusion that this was so new and independent an act, in circumstances which presented no emergency or threat, that it could not be said the locking out was a sufficiently proximate cause of the accident, as opposed to being part of the history and background to it. Alternatively stated, any defect in the lock having the potential to lock spontaneously did not render the premises unsafe by local standards.

113. It seems to me that the Claimant can legitimately argue that fragility of the ledge, if in breach of standards, caused his very serious injury, but that breach of standards in a lock which permitted the door to be inadvertently locked closed was itself not a sufficiently proximate cause of his injury."

The Grounds of Appeal

22. The appellant appeals this decision on the grounds that:

(1) The judge misdirected himself as to the appropriate test of remoteness.

(2) The judge failed to consider relevant evidence and/or the judge's conclusion that the defect in the locking mechanism was not causative of the accident was wrong.

23. In support of these grounds of appeal, Mr Robert Weir QC, who did not represent the appellant at trial, contends, in particular, as follows:

(1) The starting point is that a defendant is liable for a consequence of a kind which is reasonably foreseeable, unless the court finds that the damage was caused by a novus actus interveniens or unreasonable conduct on the part of the claimant, even if it was reasonably foreseeable: Simmons v British Steel plc [2004] UKHL 20, [2004] ICR 585 at [67].

(2) The judge should have started by making an assessment as to whether it was reasonably foreseeable that the appellant would try to escape from the confines of the balcony, rather than whether it was reasonably foreseeable that he would attempt to do so by crossing between the balconies: Hicks v Young [2015] EWHC 1144(QB) per Edis J at [33].

(3) As set out by Morris LJ in Sayers v Harlow UDC [1958] 1 WLR 623 at p630, it is: "…the most natural and reasonable action on the part of someone who finds herself undesignedly confined is to seek the means of escape." Had the judge applied the proper test, he would or should have found that it was reasonably foreseeable that the appellant (or a member of the trapped group) would seek to escape.

(4) When a claimant is injured seeking to escape from being locked up by the defendant's breach of duty, resulting in inconvenience rather than imminent danger, the injury will not necessarily be too remote: see Sayers. It is submitted that this case establishes that the correct approach is to weigh the degree of inconvenience to which the claimant had been subjected with the risks involved in trying to escape. The judge failed to apply this test or undertake the required balancing exercise. He also did not address whether the conduct was sufficiently unreasonable to amount to a novus actus interveniens.

(5) When addressing the level of the risk faced by the appellant, it was insufficient simply to describe it as "obvious" and "life threatening" [111]. The judge needed to return to the evidence on the risk that the appellant and his family considered he was taking when stepping over the balustrade.

(6) The judge had made findings of fact that the appellant and his family were all careful people and were not inebriated. The decision to cross between the balconies was not a rash one, but was carefully considered. The risk of injury was clearly not obvious to the appellant or his family at the time. The judge wrongly equated the appellant's lack of knowledge that the ledge would take his weight with an assessment that the appellant took an obvious risk [111].

(7) The appellant's mistake was to believe that the ledge would take his weight; this was not one which was so unreasonable that it should break the chain of causation, particularly when balanced against those factors set out above going to the inconvenience faced by the appellant and his family. Rather, his mistake was one which is amenable to be addressed as an issue of contributory fault.

Ground (1) – Whether the judge misdirected himself as to the appropriate test of remoteness.

24. We were referred to a number of cases in relation to the appropriate test of remoteness or causation in law. In addition to those referred to above, reference was made to Overseas Tankship (U.K.) Ltd and The Miller Steamship Co. Pty (The Wagon Mound No.2) [1967] AC 617; McKew v Holland & Hannon & Cubitts (Scotland) Ltd [10969] 3 All ER 1621; Emeh v Kensington and Chelsea and Westminster Area Health Authority and Others [1985] 1 QB 1012; Jolley v Sutton London Borough Council [2000] 1 WLR 1082; Elizabeth Anne Webb v Barclays Bank Plc & Portsmouth Hospitals NHS Trust [2001] EWCA Civ 1141; Tomlinson v Congleton Borough Council [2003] UKHL 47, [2004] 1 AC 46; Attorney General of the British Virgin Islands v Hartwell [2004] UKPC 12, [2004] 1 WLR 1273; Corr v IBC Vehicles Ltd [2008] UKHL 13, [2008] 1 AC 884; Spencer v Wincanton Holdings Limited [2009] EWCA Civ 1404 and Debell v Rochester Cathedral [2016] EWCA Civ 1094. We have also considered Clerk & Lindsell on Torts (22^nd ed.) at 2-105 to 2-130 and McGregor on Damages (19^th ed.) at 8-033 to 8-077.
25. As the appellant submits, there may be a threshold question of whether the consequence is of a kind which is reasonably foreseeable – see Lord Rodger's summary of the approach to remoteness of damage in Simmons v British Steel at [67].
26. If the consequence is of a kind which is not reasonably foreseeable then it will be too remote. If it is reasonably foreseeable then it may be necessary to consider whether the damage is too remote because it has been caused by a novus actus interveniens.
27. Determining whether there has been a novus actus interveniens requires a judgment to be made as to whether, on the particular facts, the sole effective cause of the loss, damage or injury suffered is the novus actus interveniens rather than the prior wrongdoing, and that the wrongdoing, whilst it might still be a "but for" cause and therefore a cause in fact, has been eclipsed so that it is not an effective or contributory cause in law.
28. As Aikens LJ observed in Spencer v Wincanton at [45], where the line is to be drawn is not capable of precise definition. Various considerations may, however, commonly be relevant. In a case involving intervening conduct, these may include:

(1) The extent to which the conduct was reasonably foreseeable – in general, the more foreseeable it is, the less likely it is to be a novus actus interveniens.

(2) The degree of unreasonableness of the conduct – in general, the more unreasonable the conduct, the more likely it is to be a novus actus interveniens and a number of cases have stressed the need for a high degree of unreasonableness.

(3) The extent to which it was voluntary and independent conduct – in general, the more deliberate the act, the more informed it is and the greater the free choice involved, the more likely it is to be a novus actus interveniens.

29. The first legal criticism made by Mr Weir QC of the judgment is that when considering the issue of causation in law the judge failed to address the issue of reasonable foreseeability. It is submitted that the relevant kind of consequence to be foreseen was incurring personal injury in seeking to escape from the balcony, not the precise means by which such injury occurred, and that such injury was reasonably foreseeable. As to that:

(1) It was not submitted before the judge that this was an issue which the judge needed to address when considering causation or foreseeability, nor was it submitted that the relevant kind of consequence was any personal injury, as opposed to injury resulting from attempting to walk on the balcony ledges.

(2) The appellant cannot point to any finding made by the judge as to reasonable foreseeability on this wider basis and it is not for this court to second guess what findings might have been made or to make its own findings. Foreseeability was an issue addressed by evidence and had the case been put on this wider basis it is likely that it would have been the subject of evidence. As such, it is difficult to see how it can now be open to the appellant to put the case in this different and wider way.

(3) The issue of reasonable foreseeability was addressed by the judge when considering whether there was a breach of duty by failing to give an appropriate warning. In this connection the judge accepted the respondent's argument that attempting to walk on the ledges was "such a foolhardy act in the eyes of any reasonable hotelier that it was beyond sensible contemplation, or beyond contemplation as a risk which required warning", finding that: "I find it difficult to accept that those responsible at a hotel in the Canary Islands should have foreseen that a guest would climb onto the outer ledge".

30. In these circumstances, in my judgment the judge cannot fairly be criticised for failing expressly to address the issue of reasonable foreseeability when dealing with causation, or failing to do so in the manner now contended for. Further, even if one assumes it is open to the appellant to advance this wider case, he does not have the findings to support it and it is not for this court to make such findings. Yet further, on the basis of the kind of consequence which was treated as being relevant at trial, he has found that personal injury was not reasonably foreseeable. Indeed, this may be said to be reflected in his alternative finding on causation at [112] that any defect in the lock "did not render the premises unsafe by local standards".
31. The second legal criticism made by Mr Weir QC of the judgment is that the judge failed to follow the approach taken by the court in Sayers to a case where a claimant is injured seeking to escape from being locked up by the defendant's breach of duty, resulting in inconvenience rather than imminent danger. When assessing whether the breach of duty is a legal cause of the claimant's injury in this context it is submitted that the court should follow the approach set out by Lord Evershed MR in Sayers at 626, namely:

"…to balance the risk taken against the consequences of the breach of duty; in other words, … to weigh the degree of inconvenience to which the plaintiff had been subjected with the risks that she was taking in order to try and do something about it."

32. That case concerned a woman who was locked in a public lavatory. Having failed to attract anyone's attention for 10-15 minutes she thought she could get out by climbing over the door and stood with her right foot on the lavatory seat and her left foot on the toilet roll and its attachment. Having levered herself up she decided she could not get over the door and in climbing down the toilet roll revolved causing her to slip, fall and sustain injury. The Court of Appeal reversed the trial judge's decision that the damage suffered was too remote and found the defendant to be liable, subject to 25% contributory negligence.
33. This case was not cited to the judge and I would deprecate any suggestion that the legal approach to the issue of novus actus interveniens varies as between different factual categories of case. In any event, I agree with the respondent that in essence this was the approach taken by the judge. The judge contrasted the fact that there was no danger, emergency or threat with the obvious risk of life threatening injury involved in the course of action the appellant chose to take. He was thereby weighing "the degree of inconvenience to which the plaintiff had been subjected with the risks" taken "in order to try and do something about it", thereby balancing "the risk taken against the consequences of the breach of duty". It should also be noted that in his judgment in Sayers Lord Evershed MR found that the plaintiff was not "engaged in a hazardous enterprise" or doing something "at all unreasonable" (see his judgment at p629), in obvious contrast to the judge's findings in this case.
34. It is also to be observed that the court cited with apparent approval the following passage (at p628) from the judgment of Montague-Smith J in Adams v Lancashire and Yorkshire Ry Co. (1869) LR 4 CP 739:

"….I quite agree that, if the negligence of a railway company puts a passenger in a situation of alternative danger, that is to say, if he will be in danger by remaining still, and in danger if he attempts to escape, then, if he attempts to escape, any injury that he may sustain in so doing is a consequence of the company's negligence; but if he is only suffering some inconvenience, and, to avoid that, he voluntarily runs into danger, and injury ensues, that cannot be said to be the result of the company's negligence. It is hardly necessary to say, that though I use the words 'danger' and 'inconvenience,' yet, if the inconvenience is very great and the danger run in avoiding it very slight, it may not be unreasonable to incur that danger."

35. The court in Sayers considered it to be a case where the inconvenience was great and the danger slight; hence no novus actus interveniens. On the judge's findings, the present case was one of some inconvenience and the danger obvious and life threatening; hence a novus actus interveniens.
36. The third legal criticism made by Mr Weir QC of the judgment is that the judge failed to have any or any sufficient regard to the fact that generally a high degree of unreasonableness is required for the claimant's conduct to amount to a novus actus interveniens. In my judgment, the judge did have appropriate regard to the degree of unreasonableness required. This is inherent in the balancing exercise which he carried out which contrasted the lack of threat or danger to the appellant and his family with the obviously life threatening risk which the appellant chose to take. He plainly regarded that "strikingly new and independent act" as being a highly unreasonable action to take in all the circumstances. It is also clear that the judge recognized the need to consider whether the novus actus interveniens eclipsed the prior wrongdoing. At [106] he identified the question as being whether the injury was "solely" as a result of the appellant's actions, and at [112] he found that those actions rendered the locking out "part of the history and the background" to the accident.
37. In my judgment, the appellant cannot show that the judge has misdirected himself in law as to the appropriate test for remoteness. At trial it does not appear that the judge was referred to any of the authorities now relied upon by the appellant and the issue was treated as one for the judge to determine on the facts. This is not surprising. The judge is very experienced in personal injury litigation, as were both parties' leading counsel. He has to deal with issues of causation on a regular basis. Not surprisingly, it was realistically recognised that the judge should be able to reach his decision on causation without the need for detailed consideration of authority or the law.

Ground (2) - The judge failed to consider relevant evidence and his conclusion that the defect in the locking mechanism was not causative of the accident was wrong.

38. This is essentially an attack on the factual findings made by the judge. This is a heavy burden to discharge and requires demonstrating that the decision was one which no reasonable judge could have reached on the evidence.
39. Mr Weir QC seeks to avoid the difficulty this creates for the appeal by submitting that even if the judge applied the correct legal test, he did not appreciate how it should be applied and so fell into legal error in its application.
40. In particular, Mr Weir QC submits that the judge failed to pay any or any sufficient regard to the evidence of the appellant, his parents and his wife as to why they considered that it was safe to step on to the balcony ledge. As the judge recognised, these were reasonable and responsible people and it is submitted that great weight should have been given to their assessment of the situation and of the danger involved. Mr Weir QC also criticizes the judge for applying the test he had set out by way of stating conclusions rather than conducting any detailed analysis.
41. It is correct that the judge's reasoning in relation to the issue of causation is condensed. It is essentially set out in [111] and [112] of the judgment. The core of his reasoning is nevertheless clear. The judge found that the appellant "did not know and could not know that it was safe to stand on the ledge". In those circumstances, viewed objectively, the risk of injury in standing on the ledge was "obvious" and "life threatening". The great and obvious danger involved so far outweighed the inconvenience with which the appellant and his family were faced that voluntarily running into that danger was a new and independent act which eclipsed the prior breach of duty.
42. In arriving at this conclusion, there is no reason to doubt that the judge had full regard to the evidence of the appellant, his parents and his wife. He recites that evidence at some length earlier in his judgment and, in addressing causation, he makes reference to how they were "impressively ordinary and considerate people and of apparently careful background" and to their knowledge and experience of health and safety and of life and death issues. The matters now urged upon this court were stressed strongly in submissions made to the judge at trial. In any event, it is not necessary for a judge to set out every evidential factor considered in reaching an evaluative judgment of this kind.
43. As the judge made clear, the fact of the matter is that the appellant and his family assumed that the ledge was safe. It was dark. They had not previously paid any close attention to the ledge. It was below and not part of the balcony. There was no basis for the assumption made and it was accepted in evidence that there was no reason to suppose that the ledge was safe. As the judge found, the appellant "did not know and could not know that the ledge was safe". This may be a conclusion, but it was a conclusion firmly founded on the evidence. Unless the ledge was safe, what was being attempted was obviously foolhardy and highly dangerous, as the judge found.
44. There is equally no force in the criticism made by Mr Weir QC that the judge failed to take into account the degree of inconvenience which the appellant and his family faced. The factors relied upon were urged upon the judge in submissions. Many of them are referred to in the judge's rehearsal of the witnesses' evidence. The judge was entitled to find that none of the circumstances relied upon, either individually or collectively, gave rise to more than inconvenience, let alone any situation of emergency. Nor is it correct that the family were facing the prospect of remaining on the balcony for several hours. Although they had made some attempts to attract the attention of passers-by on the roads below, they had not done so for more than a short time, nor had they shouted loudly, because they did not wish to disturb people, as Mrs Clay stated in evidence. After the accident other hotel occupants reported the disturbance to hotel staff who arrived in about 20 minutes.
45. In my judgment, there are accordingly no grounds for challenging the decision reached by the judge on the evidence or contending that this involved any error of law.
46. Finally, if regard is had to the considerations set out in paragraph 28 above, then, on the basis of the judge's findings:

(1) The appellant's conduct, which was the only conduct which was said to be relevant to foresight at trial, was not reasonably foreseeable.

(2) The conduct was unreasonable to a high degree given that the appellant and his family were faced with inconvenience rather than any danger, emergency or threat and the obvious risk of life threatening injury involved in the course of action which the appellant chose to take.

(3) The conduct was voluntary. It was both considered and deliberate. There was no necessity for the appellant to take any risk, but he nevertheless chose to expose himself to real danger and to an obvious risk of death or serious personal injury.

In these circumstances, a finding that there was a novus actus interveniens is clearly justifiable.

47. Moylan LJ has reached a different conclusion as, in his view, the appellant's conduct was not of the degree of unreasonableness required to make it an intervening event. This was a matter for the judge to evaluate and determine and in my judgment he did so in a manner which neither calls for nor justifies intervention by this court.

Conclusion

48. For the reasons outlined above, I would dismiss the appeal against the judge's findings on causation. In those circumstances it is not necessary to consider the Respondent's Notice and the issue of whether a finding of breach of duty was or should have been made.
49. In my judgment the appeal should be dismissed.

Lord Justice Moylan:

50. I find myself unable to agree with the judgment of Hamblen LJ for reasons which I endeavour to explain below. In doing so, I gratefully adopt the facts of the case and the summary of the appellant's submissions as set out in his judgment.

The Judgment

51. The judge identified the following issues as requiring determination:

"(i) Was the locking mechanism of the sliding door in room 357 defective in breach of local standards?

(ii) Did the ledge on which the Claimant stood and which gave way beneath him form part of the balcony, so as to be in breach of local standards?

(iii) Was the act of the Claimant, in climbing to the other side of the balustrade and preparing to jump or step across the gap to his own room, so unexpected or foolhardy as to be a novus actus interveniens?

(iv) If otherwise the Claimant would be entitled to succeed, was there contributory negligence and if so what deduction should be made?"

52. There is an issue as to whether the judge found that the locking mechanism was defective and in breach of local standards. If he did it is argued, by way of a Respondent's Notice, that he erred in so finding. The focus of the latter is that the judge should have found that there was no breach of local standards. I address this below.
53. Although the judge set out the issues which required determination as referred to above, his judgment followed a different pattern.
54. The judge went through each of the "asserted breaches of local standards". He first rejected a number of alleged breaches in respect of the lock on the sliding door leaving until later in the judgment the breach based on the "asserted failure of the lock on the present occasion or on repeated occasions at this hotel".
55. He then addressed the ledge and, as set out above in Hamblen LJ's judgment, at [15], he found that there was no requirement for it to be weight bearing. The ledge had been constructed by applying a cement coating over an expanded polystyrene mould. The judge found that this was not a breach of local standards because the ledge was not to be treated as forming part of the balcony which, obviously, was required to be weight bearing.
56. The judge next dealt with whether "the lock was defective or prone to shut locked". I return to this below.
57. Having dealt with the lock, the judge next considered the issue of causation in respect, and only in respect, of the lock. This is where he addressed the issue of novus actus. I also return to this below.
58. The judge next dealt with the allegation that a failure to warn was a breach of local standards. He found that it was not. In the context of the need to warn the judge addressed the character of the appellant's actions in seeking to step from one balcony to another. He found that this was not sufficiently foreseeable to require a warning of the "fragility" of the ledge: see Hamblen LJ's judgment at [16].
59. Finally, the judge addressed contributory negligence. He did so only in respect of the ledge. If the "fragility" of the ledge had been in breach of local standards, the judge assessed "the relative causation of breach of standard and the act of climbing over as equal, and the respective blameworthiness of each as essentially equal but fractionally trimmed on the Claimant's side", leading to the appellant being 45% contributory negligent.
60. I have puzzled over this conclusion. I find it difficult to see why "the act of climbing over" would lead, in those circumstances, to this finding of contributory negligence when the only reason for the appellant to have done this was because he was locked out. Yet, when considering the lock alone, the judge had found a break in the chain of causation. The breach of standards in respect of the ledge only became relevant because the appellant had climbed onto it. I appreciate that it only became an operative breach once the appellant stood on it but, if it was "beyond sensible contemplation" that anyone would climb onto the ledge, why was the chain of causation between the breach and the appellant's injuries not broken as it was in respect of the lock?
61. This conundrum, as I see it, is evident when the judge said:

"It seems to me that the Claimant can legitimately argue that fragility of the ledge, if in breach of standards, caused his serious injury, but that breach of standards in a lock which permitted the door to be inadvertently locked closed was itself not a sufficiently proximate cause of his injury".

Grounds of Appeal

62. The grounds of appeal are:

(1) The judge was wrong to find that the respondent's breach of duty was not a cause of the accident. The judge misdirected himself as to the appropriate test to apply and/or made an assessment as to the impact of the appellant's conduct on the chain of causation which was wrong;

(2) The judge's determination in respect of the appellant's contributory negligence was wrong.

The appellant's submissions are summarised by Hamblen LJ at [23] above.

Discussion and Determination

63. In order for the appellant to have succeeded at trial he had to establish a breach of duty; that the injury he sustained was reasonably foreseeable; and, in the circumstances of this case, that his own actions were not a novus actus interveniens.
64. Dealing first with the breach of duty. As referred to in Hamblen LJ's judgment, at [15]-[16], the judge rejected the alleged breaches of duty in respect both of the ledge and of the failure to warn. As Hamblen LJ also says, at [17], the parties are not agreed as to whether the judge found a breach in respect of the locking mechanism of the door.
65. It has to be said that the judgment is not as clear on this issue as it might be. However, I am satisfied that the judge found that the lock was defective. As referred to above, the judge dealt separately with the breach "based on the asserted failure of the lock". At the beginning of the part of the judgment addressing this issue the judge framed the issue of, "Whether the lock was defective or prone to shut locked" by asking the question, "Did the Claimant inadvertently slide the door to, when the lock was in the closed position?" This would strongly suggest that the judge considered that the answer to this question would determine whether the lock was defective.
66. The judge answered the question in the negative because he was satisfied that the lock "could inadvertently and without undue difficulty … be left at an intermediate position which a guest might assume to be fully opened but which permitted the latch to snap shut when … the sliding door was closed with any force against the door frame".
67. Having made this finding, the judge next considered whether this was a breach of local standards by reference to a local Ordinance which provided:

"The quality of premises and services. Accommodation establishments must ensure that their premises and services are kept, at all times and as a minimum, at the standards required to obtain any authorisation to carry out their touristic activity."

The judge then said,

"During the trial it was for practical purposes assumed, as I understand it, that if there were a defect in the lock there would be a breach of this standard. Since the evidence of the Defendant is … that Regulations would have applied in respect of balcony doors and the balcony construction in the late 1980s when it was constructed, that there would have been approval of the plans and visit by building inspectors to check that there was compliance with the Building Regulations, and that only once those checks had been completed and everything approved and agreed that the "Apertura" or operating licence be issued, this appears to be the case".

The words, "this appears to be the case", mean, it seems to me, that the presence of the defect constituted a breach of the local standard and so confirmed the assumption which had been made during the trial.

68. The judge next addressed the issue of "Causation". When doing so he said, as set out at more length above, at [21]:

"There is no doubt the door clicked shut so as to lock out the Claimant and his wife and parents. On my findings above, it is possible to categorise this as a defect and thus to constitute a departure from the" required standard.

There was some debate during the hearing of the appeal as to whether these words meant that the judge had not found a breach because of the use of the word "possible". That is clearly one interpretation. However, it is not consistent with what the judge had said in the previous paragraph nor his conclusion as to how the door came to be locked shut.

69. I am fortified in my conclusion by what the judge said in response to the application for permission to appeal. In response to the ground asserting that "the defect in the lock was causative of the accident because there was a foreseeable risk that a guest trapped on a balcony would or might look for a route off the balcony", the judge did not say that he had found that there was no defect nor that it was not a breach of local standards. What he said was, "The defect in the lock itself did not itself present danger of physical injury" (my emphasis). He then said that it was "the fragility of … the cornice … which was the immediate and direct cause of the accident".
70. It is, therefore, clear to me that the judge found that the lock was defective and that this was a breach of local standards for which the respondent was liable.
71. In a Respondent's Notice it is argued that the judge erred in finding a breach of duty. This is said to have been incompatible with some of his findings as to the door's locking mechanism. It is also said that judge was wrong to find that there had been a breach of duty because he should have found that the lock was not defective so as to be in breach of local standards.
72. I am not persuaded by the matters raised in the Respondent's Notice that the judge was wrong in this respect. The judge was plainly entitled to find that the lock was defective. He carefully analysed all the evidence before reaching this conclusion. I also consider that it was plainly open to the judge to find that this constituted a breach of local standards for which the respondent was liable. Indeed, as referred to above, the judge had understood that if there was a defect it was "for practical purposes assumed" that it would be a breach of local standards.
73. I next turn to the issue of foreseeability or remoteness. This was not one of the three issues referred to in the judgment as set out above. The judge addressed the issue of "causation" because the defence asserted that the appellant's conduct broke "the chain of causation" but did not raise the issue of remoteness. Nor did the latter feature in the "Schedule of Agreed Facts and Issues" prepared for the trial. Again, all that featured was "Claimant's Alleged Conduct and Causation".
74. In this case it could be said that the issues of remoteness and causation overlap. However, as Mr Weir submitted, there is an important distinction in that remoteness is concerned with the "kind" of damage (see below).
75. Having said that the primary focus was on the issue of causation, I acknowledge that the judge set out that he would have to "explore whether simply locking a guest out on the balcony so as to require assistance to open the door from within was foreseeably likely to cause anything other than inconvenience or delay as opposed to injury". However, although the judge set out that issue it is not easy, in my view, to see that he answered it, at least not in the negative.
76. The focus of the judge's legal, as opposed to factual, determination was, as set out in [111]-[112] (see [21] above), the issue of causation. He decided that the appellant's actions were "a strikingly new and independent act" such that "it could not be said the locking out was a sufficient proximate cause of the accident". This is not answering the question of whether injury was foreseeable. Indeed, the only reference to injury was at [113] when the judge stated that "breach of standards in a lock which permitted the door to be inadvertently locked closed was itself not a sufficiently proximate cause of his injury". I find it difficult to interpret this as a determination of foreseeability of injury.
77. Further, the paragraphs at [121]-[122] of the judgment, on which Mr Walker relies as having addressed the issue of foreseeability, are in that part of the judgment when the judge is dealing with the issue of whether it had been a breach of local standards for there to be no warning. It is in this context, and this context alone, that the judge concluded that the hotel could not "have foreseen that a guest would climb onto the outer ledge".
78. As Mr Weir submitted during the hearing, this was not a finding which dealt with remoteness of damage. The issue was not whether it was foreseeable that the appellant might be injured by trying to get from one balcony to another but whether he might be injured as a result of being trapped on the balcony. For example, in Simmons v British Steel plc, Lord Rodger said at [67]:

"(1) The starting point is that a defender is not liable for a consequence of a kind which is not reasonably foreseeable: McKew v Holland & Hannen & Cubitts (Scotland) Ltd 1970 SC (HL) 20, 25 per Lord Reid; Bourhill v Young [1943] AC 92, 101 per Lord Russell of Killowen; Allan v Barclay 2 M 873, 874 per Lord Kinloch …

… if the pursuer's injury is of a kind that was foreseeable, the defender is liable, even if the damage is greater in extent than was foreseeable or it was caused in a way that could not have been foreseen: Hughes v Lord Advocate [1963] AC 837, 847."

In Corr v IBC Vehicles Ltd, Lord Bingham said at [13]:

"The Court of Appeal majority were right to uphold the claimant's submission that it was not incumbent on her to show that suicide itself was foreseeable. But, as Lord Pearce observed in Hughes v Lord Advocate [1963] AC 837, 857, "to demand too great precision in the test of foreseeability would be unfair to the pursuer since the facets of misadventure are innumerable". That was factually a very different case from the present, but the principle that a tortfeasor who reasonably foresees the occurrence of some damage need not foresee the precise form which the damage may take in my view applies."

79. What must have been reasonably foreseeable was that the appellant might sustain personal injury as a result of being trapped on the balcony.
80. The point which could be advanced as indicating that the judge had decided personal injury was not foreseeable was what the judge said when dismissing the application for permission to appeal. As referred to above, the judge said that the defect in the lock did not "itself present danger of physical injury". However, I do not consider that this is sufficient to alter the whole focus of the judgment which was on the way in which the appellant was injured and not on whether he might sustain injury.
81. Mr Walker also referred to the judge's conclusion on causation and whether the appellant's actions were a novus actus. The judge considered whether the "locking mechanism can be said to have caused the accident" and decided that "the locking out" was not a "sufficiently proximate cause of the accident". He then said, "Alternatively stated, any defect in the lock having the potential to lock out spontaneously did not render the premises unsafe by local standards". In my view one cannot draw from these any separate consideration of the issue of foreseeability. They are different questions from whether the risk of personal injury was foreseeable because they are dealing with the cause of the accident and whether the premises were unsafe. On this issue, I respectfully disagree with Hamblen LJ's observations at [29]-[30] above.
82. In my view, if the judge had expressly answered this issue, he would have concluded that it was foreseeable that the appellant might sustain personal injury from being trapped on the balcony because it was foreseeable that he might try to escape and might sustain some injury as a result. This is particularly so when one of the Agreed Issues, which the judge was being asked to determine, was whether the appellant failed "to take the reasonable step of breaking the glass in the balcony door to gain access to the room".
83. I next turn to causation and novus actus interveniens.
84. I recognise, of course, that an appeal from the judge's conclusion that the appellant's actions were sufficient to break the chain of causation can only succeed if the judge made an error of law or if his conclusion was plainly wrong. His conclusion demands the respect that is always accorded to an evaluative determination made by the trial judge.
85. Mr Weir submits that the judge approached the issue of novus actus from too narrow a perspective. The judge focused on whether there was a "fire" or an "emergency" or a "threat" rather than undertaking a broader assessment to determine whether the appellant's actions were sufficient to amount to a novus actus.
86. Mr Weir relies on a number of the judge's findings. The appellant and his family were "moderate" and "sensible" people. The appellant was employed as a security office/fire officer. The appellant's wife said that she "would never have contemplated letting her husband attempt this if she had thought that he might fall". The appellant and his father had "a good look at the ledge and the distance between the balconies and decided it would be possible to step from one balcony to the other using the ledge on the outer edge of the balustrade". This was after they had tried but failed to open the door by "pushing and pulling it with some force" or to attract help by shouting for at least 20 minutes. The appellant's wife "began to feel anxious" (in particular, about one of the children) and "began to need the toilet". Mr Weir also pointed to the fact that the ledge looked to be of the same construction as the balcony. The appellant and his father had concluded that the ledge "looked wide and strong".
87. The judge found that the attempt to cross from one balcony to the next was made after at least half an hour had passed from when the door was found to be locked. There was no relevant inebriation. He also found that the gap between the balconies was small so that if it had been only a few feet above the ground "it would have been a simple step".
88. Mr Walker submits that the judge made an evaluation that was open to him and that he asked himself the right question, namely whether the appellant's conduct was "so unexpected and/or foolhardy as to be a novus actus interveniens".
89. We were taken through an array of authorities on this aspect of the case as referred to at [24] above. To the passages quoted above, at [28]. [31] and [34], I would add the following.
90. In Emeh v Kensington and Chelsea and Westminster Area Health Authority [1985] 1 QB 1012 Waller LJ phrased the relevant question as being (at p. 1018H):

"Can it be said that the plaintiff's conduct was so unreasonable as to eclipse the defendants' wrongdoing?"

He then quoted from McKew v Holland before saying that "the degree of unreasonable conduct which is required is, on Lord Reid's view, very, very high" (p. 1019C). Lord Reid had used the expression "utterly unreasonable" which Waller LJ adopted when considering the plaintiff's conduct (at p. 1019E).

91. In Webb v Barclays Bank the court addressed the effect of the intervening negligence of a doctor. Henry LJ gave the judgment of the court which determined that the chain of causation had not been broken. This was because of several factors including that "the original wrong-doing remained a causative factor" because of its effect on the claimant and because the doctor had been negligent "not grossly negligent", [56].
92. In Corr v IBC Vehicles Ltd Lord Bingham identified the "rationale of the principle that a novus actus interveniens breaks the chain of causation is fairness. It is not fair to hold the tortfeasor liable, however gross his breach of duty may be, for damage caused to the claimant not by the tortfeasor's breach of duty but by some independent, supervening cause … for which the tortfeasor is not responsible", at [15].
93. In my view the authorities demonstrate that a broad evaluation is required when determining whether a claimant's conduct is sufficiently unreasonable to break the chain of causation by eclipsing the causative effect of the defendant's wrong-doing. This involves consideration of all facts relevant to that question. Further, as was pointed out by Aikens LJ in Spencer v Wincanton Holdings Limited, at [44] and [45], the conduct must be such as to take it beyond that which would be within the scope of contributory negligence.
94. Did the judge apply the right test and/or was his evaluation flawed? In my view, although the judge referred to the issue as being whether the appellant's actions were "so unexpected and foolhardy" as to break the chain of causation, when he came to determine that issue he did not undertake the broad evaluation required to determine whether those actions were sufficiently unreasonable to eclipse the causative effect of the respondent's wrong-doing. Accordingly, he either applied the wrong test or reached a flawed evaluation.
95. Although the judge referred to a number of factors at [107]-[113] (see [21] above), he ultimately undertook what I consider to be an unduly narrow evaluation. This can be seen from [111]-[112] where he set out his assessment of the appellant's conduct. The judge focused on whether the defect in the lock was a "direct danger" causing an "emergency or threat". Balanced against this the judge referred only to the air temperature and to his assessment that, because the appellant "did not know and could not know that it was safe to stand on the ledge" (my emphasis), the risk of life threatening injury was "obvious".
96. This analysis was too narrow because a proper assessment of whether the appellant's conduct was sufficiently unreasonable was not confined to whether the lock created a direct danger (in the sense of creating an emergency or threat) nor to whether the appellant did not "know" that the ledge was safe to stand on. The absence of danger was relevant but so were the other matters referred to in the evidence which led the appellant to respond to being trapped on the balcony by trying to get to the other balcony. I have also emphasised the word "know" because, in my view, this was not the appropriate assessment and suggests that the judge applied the benefit of hindsight.
97. I agree with Mr Weir's submission that the judge's analysis in these critical paragraphs ignored his assessment that the appellant and his family were moderate and sensible people and ignored the fact that they had carried out what they thought was a sufficient analysis of the ledge to conclude that it was "wide and strong". They were wrong in this analysis but that is a different point. The judge should have considered whether their subjective conclusion, which led the appellant to seek to step from one balcony to the other, was objectively so unreasonable a response to being trapped on the balcony that it eclipsed the causative effect of the defect which had led to them being trapped.
98. In summary, and somewhat repetitively, in my view the judge failed to apply the analysis required properly to answer the question of whether the appellant's response to being trapped on the balcony with the other members of his family, as a result of the respondent's wrong-doing, was so unreasonable as to mean that the reason for them being trapped was no longer an operative cause of the accident.
99. If the judge had carried out such an analysis in my view he would have come to the conclusion that the appellant's conduct was not such as to "eclipse" the causative effect of the defective lock. The family had sought to procure their release by other means for at least half an hour. It was very late at night. The appellant's wife began to feel anxious as referred to above. The appellant with, in particular, his father had carried out a considered analysis and had come to the conclusion that he could escape from the balcony by stepping onto what appeared to be part of the balcony because it was covered with the same material, namely concrete. In my view, that this analysis was flawed and that there was no emergency should not be treated as sufficient to make his actions an intervening event which broke the chain of causation. The defect in the lock remained a causative factor because the appellant's response to being trapped was not of the degree of unreasonableness required to make it an intervening event.
100. I finally turn to the issue of contributory negligence. I have already expressed my response to the judge's determination of 45% if the ledge had been in breach of local standards. Despite my conclusions as to other aspects of the judge's decision, I see no reason to depart from this assessment when applied to the defect in the door. In other words, if the judge had not found there to be a break in the chain of causation, I see no reason why he would not also have reached the same conclusion in respect of contributory negligence. The same act was being assessed, namely the appellant "climbing over", in the context of the respondent's breach having created the situation with which the appellant was confronted.
101. In conclusion, for the reasons set out above, I would have allowed the appeal and substituted a finding of liability with contributory negligence of 45%.

Lord Justice Kitchin:

102. I agree with Hamblen LJ that this appeal should be dismissed.
103. The critical issue which divides my Lords, Hamblen LJ and Moylan LJ, is whether the judge fell into error in finding that any defect in the locking mechanism to the balcony door of room 357 was not the cause of the accident that befell Mr Phillip Clay.
104. In my view the judge carried out precisely the evaluation that the law requires. He made express findings that Mr Clay, his wife and his parents were careful and considerate people; that Mr Clay was a security and fire officer with health and safety experience; that Mrs Clay was a nurse who "dealt with life and death, and would never have contemplated" letting Mr Clay attempt what he did if she had thought he might fall; and that Mr Clay's parents were, prior to their retirement, also engaged in jobs in which they had to undertake considerable responsibility for safety. The judge must also have had well in mind his earlier findings that these members of the family had been locked onto the balcony for around 30 minutes; that it was now in the early hours of the morning and getting chillier; that they had concerns about the children asleep in the next door room; and that Mrs Clay had begun to "need the toilet".
105. Nevertheless, the judge also found that these members of the family were in no direct danger; that there was no emergency or threat; and that the temperature was still 19ºC. Moreover and importantly, Mr Clay did not know and had no way of knowing whether it was safe to stand on the ledge outside the balustrade and, the room and balcony being two storeys up, if the ledge was not safe, the risk of life threatening injury was obvious.
106. The judge then balanced the degree of inconvenience to which Mr Clay and his family were subjected against the risk of injury were he to attempt to jump from the one balcony to the other, and he found that Mr Clay's action in climbing over the balustrade and putting his weight on the outside ledge constituted a strikingly new and independent act such that it became, in law, the effective cause of his injury.
107. It is true that the judge's reasoning is concise but I am wholly unpersuaded that it betrays any error of law or that he has failed to have proper regard to any material aspect of the evidence or that he has arrived at a conclusion which is plainly wrong. He found, in substance, that the actions of Mr Clay eclipsed the wrongdoing of the defendant and constituted a novus actus interveniens. Put another way, his injuries were the result of the risk to which he exposed himself. That is a conclusion which was properly open to him upon the evidence and it is one with which we should not interfere.
108. For these reasons and those given by Hamblen LJ, I would dismiss this appeal.