Introduction
‘Occupiers’ liability’ may sound like an arcane area to the torts non-specialist, but its principles are regularly invoked by victims of a serious injury who are searching for a possible remedy. In one year alone, the higher jurisdictions decided the claims of someone who became quadriplegic after falling off a bridge in a park,Footnote 1 a tourist who fell into the moat of a castle,Footnote 2 and a prisoner who was injured during an electricity power outage.Footnote 3 Coupled with more recent and older case law, this indicates a field prone to considerable litigation and judicial disagreements. At its core sits the question: did the occupier of the premises where an incident took place breach the duty of care owed to all visitors to these premises?
The legislation in England and Wales which currently governs occupiers’ liability to visitors is the Occupiers’ Liability Act 1957 (OLA 1957). The intention behind this statutory provision had been to simplify a legal area which had been developing under the common law and was criticised for being in a ‘confused state’.Footnote 4 It was also a response to the inconsistency and harshness of the common law, where the scope and extent of the duty owed depended on the type of lawful entrant to premises. This statute reformed the law by providing that occupiers owe a common duty of care to all visitors to premises and that the duty is ‘to take such care as in all the circumstances of the case is reasonable to see that the visitor will be reasonably safe in using the premises for the purposes for which he is invited or permitted by the occupier to be there’.Footnote 5 It was recognised that the contours of breach of duty drafted in such general terms would have to be defined more precisely by the courts in the course of time.Footnote 6
Commentators generally welcomed the legislation.Footnote 7 One dissenting voice was Diplock QC, who thought that the codification would cause ‘for a considerable period of years until the new case law has been settled, uncertainty over a wide field of legal rights and obligations which affect every member of the public in his daily life’.Footnote 8 Echoing this scepticism, Payne foresaw extended litigation, as ‘the greater the discretion conferred on the court, the more uncertain the outcome of a case will be, and the higher will be the proportion of cases that go to trial instead of being settled out of court’.Footnote 9
These remarks were prescient. In the six decades since the OLA 1957 came into force, the factors judicially considered in determining breach of duty have multiplied: was the injury foreseeable or was the danger so obvious that a reasonable person would have been expected to avoid it? Should there have been a warning notice, a fence to close off the dangerous area, or some other protective measure? Would the latter have made a difference or would visitors have ignored it? Might it have represented a disproportionate cost for the occupier? And so on and so forth.
The factors taken into account vary from case to case, and often within one case from one level of jurisdiction to the next. Uncertainty as to which factor will be deemed crucial (or even simply relevant) in any one case puts legal representatives in a difficult position, as well as judges who do not direct the arguments submitted to them. What cannot be contested is that competing interests are at stake and have to be balanced against each other: ‘on the one hand, there is the interest of the land occupier to have untrammelled use and control of his property; on the other hand, and in opposition to this, there is the interest of society in ensuring the physical safety of all its members, and when a member is injured the interest to see that he is compensated’.Footnote 10 A compromise must be reached between conflicting and evolving values, such as the need to rein in the excesses of a growing ‘compensation culture’Footnote 11 and the importance of continuing to protect members of society, especially when they are vulnerable.Footnote 12
After over 60 years of judicial decisions an analysis of the common themes identified by the courts in determining breach of duty under the OLA 1957 is timely. Based on a critical review of the case law, this paper seeks to expose inconsistencies in judicial adjudication and to identify the central factors that should guide the courts in determing breach of duty. This approach offers a more structured legal analysis of the myriad factual scenarios that may be presented to the courts and is intended to clarify how guidelines developed by the courts under the jurisdiction of the OLA 1957 should be applied.
Liability under OLA 1957 obviously requires the Act to apply. For this pre-condition to be fulfilled, the injury must have occurred because of a danger due to the state of the premises. A threshold test is therefore in order, which allows ‘occupancy duties’ to be distinguished from ‘activity duties’.Footnote 13 If the injury is due to the activities which took place on the premises, then the claim falls outside the field of occupiers’ liability. An example of this would be the attempt to slide down an ornate banister when drunk (unsafe activity on safe premises).Footnote 14
If an injury has been caused by a defect or danger on the premises, a breach of occupiers’ duty may have occurred. The central argument of this paper is that this would be best determined by recourse to a three-stage test. The test's first stage relates to foreseeability. If the risk of injury was not foreseeable to a reasonable person, the occupier cannot be liable; no further discussion is needed to conclude the case on occupier's liability (though note that there may be liability in common law negligence).Footnote 15 If the risk is deemed to have been foreseeable, the second stage of the analysis is to consider whether the occupier could have reasonably been expected to take steps in order to mitigate the risk. If this test is answered in the affirmative, a third question arises: was any remedial action taken by the occupier appropriate? Only if this answer is negative will there be a breach of OLA 1957's duty of care.
The application of our test should assist the development of a clearer body of case law, at the same time as eliminating the need for discussions which currently seem to occupy much judicial time despite being unnecessary within our scheme (because they pertain to an issue which need not have been addressed). The test should also equip occupiers with a better understanding of what the law requires of them, thereby inciting them to keep their premises in a safe state or, when incidents take place, to accept settlement at an early stage.Footnote 16 Last but not least, the clarity gained through the application of a more systematic and logical test should enhance the transparency of judicial reasoning. This should make more explicit the policy choices that underpin the way the judiciary exercises its discretion in this area, thereby facilitating debate in society as to the values that the law is protecting.
1. The legal framework
Before turning to the case law, it is important to circumscribe the scope of OLA 1957.
The term ‘occupier’ refers to any person who, under common law rules, occupies premises. The legislative intention was and remains to have the liability fall on the person most likely to have been in a position to prevent the harm from occurring. This may or may not be the owner; it is anyone who occupies the premises, ie who has control over them.Footnote 17
The term ‘premises’ can cover all kinds of things, including railway lines, aircrafts, ships and other vehicles, the sea bed in a harbour, pubs, schools, fairground attractions and even ladders.Footnote 18 Section 1(1) of the 1957 Act states that the occupiers’ duty of care arises in respect of ‘dangers due to the state of the premises or to things done or omitted to be done on them’.Footnote 19 If the injury was due to activities taking place on premises, the case falls outside the material scope of the OLA 1957 and is governed by negligence rules under the common law. For example, in 2017, an injury caused by a fall from a bouldering wall was quickly found to be attributable to the undertaking of an inherently dangerous activity rather than to the dangerous state of the premises.Footnote 20
As alluded to in our introduction, pre-1957 common law classified visitors in different categories on a sliding scale, with some visitors needing to be protected against unusual danger of which the occupier knew or ought to have known, but others only having to be warned of any concealed danger (or trap) of which the occupier knew.Footnote 21 The distinction between ‘unusual’ and ‘concealed’ dangers was not obvious; this is one of the reasons why the OLA 1957 introduced a common duty of care, owed to all visitors.Footnote 22
The statutory duty contained in the OLA 1957, s 2(2) explicitly establishes that the occupier must ensure the individual visitor is reasonably safe,Footnote 23 and so the vulnerability of the entrant (including their specific mental or physicial attributes) is a relevant factor when examining whether the occupier has met the appropriate standard of care.Footnote 24 This focus on the safety of the visitor, not the premises, means that a warning of a danger or a barrier surrounding the danger may in some circumstances be sufficient for the occupier to satisfy the standard of care even if the danger has not been completely eliminated. Furthermore, the section is clear that the occupier's duty extends only to protecting the visitor ‘for the purposes for which he is invited or permitted by the occupier to be there’. Thus, the occupier is not expected to protect the visitor from dangers due to misbehaviour on the premises.
OLA 1957 only gives limited guidance as to what factors should be used to assess whether the occupier has breached their duty. Section 2(3)(a) provides that an occupier should be prepared for children to be less careful than adults. Section 2(3)(b) adds that an occupier may expect professional visitors to guard themselves, ‘in the exercise of their calling’, against any special risk ordinarily incident to the job they have been called to perform. Section 2(4)(a) also specifies that any warning given by the occupier to visitors is not to be treated as absolving the occupier from liability, unless it is enough to enable the visitor to be reasonably safe. Beyond this, OLA 1957 simply states that ‘regard is to be had to all the circumstances’.Footnote 25
Occupiers of premises have been held liable for various acts and omissions, including: polishing a floor to such an extent as to make it dangerous,Footnote 26 omitting to light stairs adequately,Footnote 27 or failing to remove hazards likely to injure playing children.Footnote 28 The case law also contains numerous examples of situations where an occupier has been held not liable, for example when adequate warning notice had been given,Footnote 29 where a hotel guest had slipped on a pool of water in the vicinity of a jacuzzi,Footnote 30 or a teenager had dived into a paddling pool.Footnote 31
2. Multiple factors
Before turning to our test, it is worth explaining what prompted us to devise it. Despite contrary opinion,Footnote 32 we find the case law concerning OLA 1957 surprisingly abundant. This may be attributable to the absence of precise legislative guidelines to determine how to assess whether an occupier has breached the duty of care owed to visitors, which gives the judiciary freedom to take into consideration a multitude of ever increasing factors. The problem with this approach is that judicial reasoning then fails to identify which factor must be considered key, nor does it indicate how the various factors logically relate to each other, creating uncertainty in the law.
Legislative and judicial precepts make it possible to draw a long list of possibly relevant factors, as follows:
#1 foreseeability of injury;
#2 obviousness of the danger;
#3 absence/presence of warnings;
#4 whether adequate lighting was provided;
#5 whether fencing was erected;
#6 age range of expected visitors;
#7 purpose of their visits;
#8 conduct that could reasonably be expected of them;
#9 whether the injured visitor was acting in the exercise of their calling;
#10 the extent of precautions the occupier could be expected to take (less for a householder, more for a professional);
#11 whether the occupier knew or should have known about the danger;
#12 how difficult and expensive it would have been to remove the danger;
#13 the general practice of occupiers in the relevant field;
#14 the existence of official or semi-official safety rules;
#15 whether the occupier had acted on professional or semi-professional advice;
#16 whether a risk assessment had been performed;
#17 the likelihood or otherwise of the danger materialising;
#18 the desirability of keeping the amenity in an unaltered state;
#19 aesthetic matters;
#20 the desirability of leaving the premises in their original conditions;
#21 the seriousness of the injury;
#22 a (lack of) history of past incidents;
#23 how much time had passed since the danger was identified;
#24 whether additional protective measures would have had any effect;
#25 whether the visitor would have utilised additional safety measures;
#26 whether repairs were regularly carried out;
#27 what is happening on the premises (eg the use of explosives requires a high degree of care);Footnote 33
#28 the desirability (social value) of the activity performed on the premises;
#29 a recognition that those who take risks of their own free will should not expect compensation for injury;
#30 the undesirability of living in a society where all risks have been eliminated.
This indicative list cannot be comprehensive: new circumstances will always lead to the identification of additional potentially relevant factors. What our discussion will highlight is that it is not immediately clear why some factors rather than others are considered in any one case. Without a ‘control’ system determining when a specific range of factors should be applied, the approach appears ad hoc and unable to guarantee against inconsistent judicial decision making and unfairness. We shall demonstrate this through a review of five cases (four of which were decided in the same year). Any number which appears in parentheses refers to the list above.
English Heritage concerned a visitor who had sustained a serious head injury after falling into the moat of a castle on the Isle of Wight. The first instance judge and the Court of Appeal agreed that occupier's liability was incurred, but arguably focused on different reasons. Emphasis at first instance was on the lack of warnings (#3), with the judge finding that additional warnings should have been in place to ensure the safety of the claimant, given the sheer drop of the moat.Footnote 34 The Court of Appeal confirmed the liability of English Heritage, but seemed to put the stress on whether there had been an obvious danger (#2), given there is no need for an occupier to protect irresponsible visitors against perfectly obvious dangers.Footnote 35 Here the danger was found not to have been obvious.
In Rochester Cathedral, differences of judicial opinion as to which factors should be considered produced different outcomes. In this case, a man tripped on a small lump of concrete which was protruding from the base of a traffic bollard in the precints of Rochester Cathedral. The court at first instance found for the claimant, whereas the Court of Appeal held the Cathedral not to be liable. The key factor at first instance seems to have been that there was a foreseeable risk of injury (#1). Although the Court of Appeal also considered the foreseeability of injury, it was swayed by other factors such as the difficulty and expense of removing the danger (#12), in particular the time and money that occupiers would have to spend identifying and remedying any and every fault of this nature;Footnote 36 and the likelihood of the risk of injury (#17) (it was unlikely a pedestrian would walk so close to the bollard).Footnote 37
Edwards v Sutton Borough Council illustrates that similar factors can be considered at different levels of jurisdiction, but to different effect. A visitor to a park had fallen from a small ornamental bridge with a low parapet onto a rock in the water below, resulting in a spinal cord injury. He sued the park's occupier, Sutton Borough Council. To find the council liable, the first instance judge referred to the presence of a danger due to the state of the premises (our initial threshold test);Footnote 38 the failure to have undertaken a formal risk assesment (#16);Footnote 39 the foreseeability of the claimant tripping (#1);Footnote 40 the fact that warnings could have been installed at no significant cost (#12);Footnote 41 the ‘catastrophic’ nature of the injury if it were to eventuate (#21);Footnote 42 and the absence of warning about the ‘dangerously low’ parapet (#3).Footnote 43 The Court of Appeal discussed most of these factors, but put them in a different context, thus finding the council not liable. It observed that it was important to first of all identify the danger (and admonished the trial judge for not having considered this issue adequately).Footnote 44 The danger would have been obvious, the court stressed, making the potential for injury obvious too (#2). The Court stated that occupiers are under no duty to protect or even warn against obvious dangers. The danger had been ‘extremely unlikely’ to materialise in this case (#17), and one cannot guard against the many foreseeable risks that are extremely unlikely to happen.Footnote 45 ‘[A]ny warning or sign would not have told the claimant anything that he did not already know’.Footnote 46 Thus, ‘there was no need for a risk assessment as it would not have lessened the accident risk’ (#16).Footnote 47 The social value of walking over the bridge (#28) and its amenity (#18) as well as the absence of previous incidents (#22) were also mentioned.Footnote 48
To close this review, we move to two cases which tragically involved children. Bourne Leisure (the one case decided prior to 2016 discussed in this section) concerned a two-year-old who had drowned whilst his family was holidaying at a caravan park.Footnote 49 At first instance, Bourne Leisure was held liable mainly due to the lack of warnings the parents had received (#3). In the judge's view, the holiday park had been under ‘a high duty to inform, clearly and unequivocally, the parents fully of the location and means of access to such ponds or lakes as were present on the site’;Footnote 50 merely providing parents with a map of the site without drawing their attention to the danger was insufficient. To quote, ‘I am satisfied on the balance of probabilities that sensible information as to the location and easy access to the pond [from the claimant's caravan] would, without any dire over-worrying warnings as to the height of fences, have made every difference’.Footnote 51 The last remark alludes to a fence that had been installed after a four-year-old had nearly drowned at the same pond. This fence was not of the height recommended for domestic and/or school ponds by the Royal Society for the Prevention of Accidents (RSoPA) (#14). However, the judge did not find this circumstance critical, since the holiday park was neither a family home nor a school, and advice had been taken after the earlier incident from environmental health officers.Footnote 52 For the judge, the key issue was the absence of warning. On appeal, the holiday park was held not to be liable. At this level, the main issue appears to have been that the danger was obvious (#2). To quote, ‘the danger of the lake to a small child, should that child in fact stray, was obvious’.Footnote 53 In the view of the Court of Appeal, any warning would not have told the parents anything they did not already know. ‘Such warnings were irrelevant in a case such as this where the parents were quite aware of the need to accompany their young children and of the dangers at the water's edge, of which they needed no reminder’.Footnote 54
In the Scottish case of Anderson v Imrie, a child had sustained serious skull and brain injuries.Footnote 55 The occupier of the land where this happened was found liable under the Occupiers’ Liability (Scotland) Act 1960, which is similar to OLA 1957. The defendant lived on a 100-acre farm and had arranged for an eight-year-old boy to come and play with her five-year-old son. She had told the children they could play in the farmhouse and in the courtyard but that they must not go into the race (a type of livestock crush). She was going back and forth between the farmhouse, where her mother was taking care of her baby, and the courtyard, where she was dressing her horse. While she went to fetch something for the horse into the stable, the visiting child climbed over the gate separating the courtyard from the race. Once in the race he climbed on to a stock gate attached to a barrier. He lifted the chain off, causing the gate to become detached from the barrier and over-balance on top of him, causing him to fall back and strike his head against the concrete surface of the race. For the court, taking into account the age of the visitor (#6) and finding that the race and the gate would have been irresistible to a child, the injury was foreseeable (#1): ‘he must have been out of her sight for at least several minutes … that was dangerously long … a foreseeable risk that within such a timeframe the pursuer would suffer an accident’.Footnote 56 The judge found the mother liable.
It seems strange that the race in Anderson was found irresistable to a child but not the pond in Bourne Leisure (despite the previous incidence of a near-drowning). It is also noteworthy that the occupier in Bourne Leisure was a professional contractor deriving income from the visits that were its raison d’être, whilst in Anderson it was a mother who had had a child around to play at home. The difference in outcome between the two cases will be discussed further below.
3. Which factors apply in any one case?
Can judges simply determine the factors which to them seem most relevant to the circumstances of the case before them, without having to justify their selection or demonstrating consistency with previous judgments? Certainly, the case law is not explicit as to why some factors are chosen over others and/or become key.
Tomlinson v Congleton is a leading House of Lords’ case in the field of occupiers’ liability. Ostensibly about the duty owed to trespassers,Footnote 57 their Lordships nonetheless opined at length on matters relevant to the duty owed to visitors and it is cited in numerous cases concerning the OLA 1957. Tomlinson confirmed that the right approach is to start by identifying a danger due to the state of the premises. Not all cases do this, however. For example, in a case brought against the London School of Economics by a student who had slipped in her university-owned shower room, a county court discussed the absence of previous accidents (#22), the fact that slip resistant flooring had been installed which accorded with relevant official safety advice for universities (#14), as well as the very small risk that the danger would have materialised (#17) before dismissing the claim.Footnote 58 This meticulous analysis was arguably superfluous: early in the judgment the possibility of the standard shower room having been a ‘trap’ had been eliminated. Yet more puzzling is another county court's decision which found a holiday resort incorporating a bar not liable after a customer had attacked another with a bottle of wine. The reason given was that the customer had not acted in a way that could have been foreseen (#1).Footnote 59 It would surely have been simpler to establish at the outset that there had been no danger due to the state of the premises.
In Tomlinson, Lord Hoffmann highlighted four major factors to determine breach of duty in occupiers’ liability: the foreseeability of injury (#1), the seriousness of injury (#21), the social value of the activity (#28) and the cost of preventative measures (#12). The question of whether people should accept responsibility for the risks they choose to run (#29) was a fifth factor whose saliency derives from concerns about the real or perceived dangers of a ‘compensation culture’.Footnote 60
These five factors remain regularly cited in the case law, as a review of three cases decided in 2017 may serve to illustrate. Cook concerned a fall on black ice in an unmanned car park.Footnote 61 To conclude that the city council had not breached its duty, the Court of Appeal relied on the first four factors, alongside the obviousness of the danger (#2): gritting the car park on a Saturday would have been ‘disproportionate’ to the risk and ‘would have diverted such resources from situations where attention was more urgently required’.Footnote 62Singh produced a first instance judgment on similar lines. Walking home along a path, the claimant slipped down into a brook and remained there overnight. The High Court reasoned that the injury was severe and the cost of fencing low but that the risk was low (no previous accident reported) and there was a real social value in allowing people to walk there; decisively, the visitor had willingly accepted the risk.Footnote 63 In Robinson,Footnote 64 about a two-and-a-half-metre fall from a terrace, the county court mentioned the first four factors, but ended up rejecting the county's liabiity on the basis that the claimant had been drinking (#8).Footnote 65 The drop was obvious,Footnote 66 and thus no duty arose as there had been no danger related to the state of the premises,
However, many cases concentrate on altogether different issues. This seems particularly the case when compliance with official guidance (#14) is at issue. In such cases, whether guidance was followed looms large. In the shower room case, the judge was clearly influenced by the fact that the University had addressed the risk of injury by installing slip-resistant flooring which accorded with the relevant Code of Practice.Footnote 67 Another example concerns a resurfaced balcony's balustrade which was below the height recommended by British Standards Institutions, which seems to have been determinant to find the Ritz hotel liable.Footnote 68 In yet another case involving a fall, a handrail had been below standard. This was key, since the occupier – licensed premises – must have expected inebriated customers like the claimant to rely on handrails when taking stairs.Footnote 69
In many cases, whether a risk assessment has been carried out (#16) is important, such as when a man fell from a hospital's roof,Footnote 70 whereas in other cases it is felt an assessment would not have added anything at all.Footnote 71 An absence of previous incidents (#22) sometimes appears determinant, such as in Edwards (fall from a low parapet bridge) for a finding of non liability. Lack of warning (#3) can be decisive for concluding that the occupier is liable,Footnote 72 but is dismissed as irrelevant in other cases where the risk is considered to be of an obvious nature.Footnote 73 In Rochester Cathedral, the outcome ended up resting on the difficulty and cost of identifying and remedying in a systematic manner apparently small dangers such as the protruding concrete (#12).
Clearly it is the function of the judge to determine which facts are legally relevant and how they should be qualified and interpreted. In this sense, for different judges to examine arguably similar cases (or the same case at various levels) differently is not per se disturbing; this is inherent to the judicial function. What concerns us, however, is that occupier's liability appears to be a field where it is hardly possible to discern consistency in judicial reasoning and where an abstract, logical appreciation of how the factors to be taken into consideration interrelate seems to be absent.
4. Factors lacking in clarity
This state of affairs is accompanied by a lack of clarity regarding two factors which are arguably relevant to any case having to do with occupier's liability, whatever their precise circumstances, and which can therefore be considered fundamental to this area of the law. This is the foreseeability of the injury (#1 in our list) and the obviousness of the danger (#2).
Foreseeability heads our list because it is regularly discussed in the case law, and appears quite early in most judicial reasoning. It was vividly discussed in Rochester Cathedral. The judge at first instance framed the main question as being the existence of a foreseeable risk, which it found established.Footnote 74 However, the Court of Appeal admonished the lower court for not having applied this criterion properly: ‘The judge did not apply the foreseeability test in the appropriate way … There is no recognition in the judgment that not all foreseeable risks give rise to the duty to take remedial action’.Footnote 75 The Court of Appeal explained that the test must be considered met ‘only where there is a real source of danger which a reasonable person would recognise as obliging the occupier to take remedial action’,Footnote 76 with anything else liable ‘to result in too onerous a standard of care’.Footnote 77
A second factor which also lacks clarity despite being regularly discussed relates to the obvious character of the danger (#2). Settled law has it that an occupier is under no duty to protect against dangers which are obvious to a visitor – who is expected to take reasonable care for their own safety. This principle derives from numerous cases, including one where a fall into a ‘ha-ha’ (a ditch) was found not to engage the liability of the occupier as the danger was obvious,Footnote 78 another where the college's duty of care did not extend to protecting someone of full age and capacity from obvious risks such as diving into a small inflatable pool,Footnote 79 and other cases already discussed above such as English Heritage and Bourne Leisure.
North explains: ‘It is not every danger in relation to the state of the premises which may give rise to a duty of care owed by the occupier. He will not be liable for dangers stemming from obvious risks’.Footnote 80 Although this sounds wise, the precept should be refined in view of how, for example in Bourne Leisure, the danger of water would have been clear to parents, but not to the young children who could be expected to visit the caravaning site. As Stuart-Smith LJ said in an earlier case: ‘the nature of and extent of what it is reasonable to expect of the occupier varies greatly depending on whether the [visitor] is very young or very old and so may not appreciate the nature of the danger which is or ought to be apparent to an adult’.Footnote 81 Lord Scott of Foscote seems also to have suggested in Tomlinson that it is insufficient to phrase the obviousness test in terms of what a reasonable person would or should be aware of.Footnote 82
Admittedly, any factor in our list of 30 will always lend itself to being further refined as practice throws slightly new circumstances that call to be accommodated. For example, the ‘difficulty and expense of removing the danger’ might have appeared a precise factor – until Rochester Cathedral highlighted that it should include not only the difficulty and expense of removing the danger, but also ‘the cost in terms of time and money of having to identify … faults of this nature’.Footnote 83 Such tweaking is not intellectually difficult to accomplish and can easily be smoothly integrated as and when circumstances change. However, what this section has sought to highlight is of a different order, namely, that there is a lack of conceptual clarity in the way the fundamental factors at play in the area of occupier's liability should be understood and how they relate to each other.
5. A revised test
To sum up, our review of the case law has identified three problems with the current judicial approach to occupiers’ liability. First, cases are sometimes fully heard that should arguably have been found to fall outside occupiers’ liability because they did not relate to ‘a danger due to the state of the premises’. Secondly, the judicial approach to determining breach of duty appears to satisfy itself with selecting in each case the factors that will be taken into consideration without explaining why the emphasis is put on some factors to the neglect of others. Thirdly, there is a lack of clarity as to the meaning of some factors, despite them appearing as cornerstones of this area of the law. What is needed, therefore, is a more systematic approach capable of achieving a clearer and more logical integration between the multitude of possibly relevant factors. Our proposed test assembles the factors, identified through 60 years of case law, via a formal analytical process able to establish breach of duty.
Although logically apart from our breach-of-duty test, it is worth noting that the judge must initially check that the case falls within the scope of occupiers’ liability. This ‘threshold’ determination must result in a yes/no decision, which is not to say that this is a straightforward exercise.Footnote 84 For example, should the fall of an 11-year-old trying to climb the underside of a fire escape be found to be due to the ‘state of the premises’ or to the claimant's conduct? The Court of Appeal decided it was the latter in Keown.Footnote 85 By contrast, in the arguably similar case of Spearman, which saw a vulnerable patient leaving the emergency department of a hospital, climbing five flights of stairs to a flat roof, going over a protective barrier and falling into a courtyard suffering serious injuries, the court determined that there had been a danger due to the state of the premises (rejecting the hospital's claim that the patient had gone where he was not permitted and was thus a trespasser).Footnote 86 The threshold test is too often ignored by the judiciary or at least not spelt out at the start of the judicial reasoning. That this be done is nonetheless essential, as recognised by Lord Hoffmann in Tomlinson. Footnote 87
There are cases where this issue could have been – but was not – contested.Footnote 88 Other cases go to the trouble of finding a danger – for example identifying ‘surface integrity issues’ with the low-parapet bridge in Edwards – even though this seems a bit far fetched. At first sight, one might associate a judicial willingness to discern a danger due to the state of the premises with a wish to protect the visitor, and a refusal to discern such a danger with a reluctance to impose liability. However, given the way the judiciary often muddles issues that we argue would best be kept logically separated, hasty conclusions should not be drawn. The important point in the context of this paper is that, if the answer given to the initial threshold test is negative, then there is no need to proceed further, as the case is determined to fall outside the ambit of occupier's liability. By contrast, if the answer is positive, then whether a breach of duty has occurred is best determined by applying a test made of up to three consecutive stages.
(a) Stage 1: foreseeability
In its first stage, the test asks: Was the danger identified in the initial threshold test actually presenting a foreseeable risk of injury?
This corresponds to the first of the four factors highlighted by Lord Hoffmann in Tomlinson and is also factor #1 in our list. Stage 1 must be distinguished from the initial threshold test: it is not because the premises are subsequently recognised to have been dangerous (thus fulfilling the requirement that it engages the field of occupiers’ liability) that the defect should necessarily be recognised to have constituted a foreseeable risk of injury. Foreseeability is a key consideration in all cases. It cannot be approached as simply one factor amongst many. It requires its own stage of assessment.
There are two rather different situations where one must conclude that the risk of injury was unforeseeable. The first is when the occupier did not know about the danger and could not have been expected to have known about it (#11). From the occupier's perspective, the harm was therefore utterly unforeseeable. The second situation where a conclusion of unforeseeability must be reached is when the danger due to the state of premises was known (or ought to have been known) but the risk of an incident giving rise to the injury actually materialising is considered to have been entirely minimal. This may be for a multitude of reasons. One could be that the premises were not expected to be visited.Footnote 89 Alternatively, foreseeability is not demonstrated where a reasonable person would not have predicted that an incident such as happened would ever take place. This is because the risk was so negligible that it was not foreseeable. The judiciary has used various terminology to try to capture the threshold at which the foreseeability of risk of injury bites.Footnote 90 We concur with those judgments that identify that the risk must have been more than ‘slight and remote’, ‘small’ or ‘minimal’.Footnote 91 Thus, more than a minimal or negligible risk of injury is required to satisfy the test of foreseeability.
Stage 1 includes two factors from our long list: the foreseeability of an injury being sustained if an accident were to happen (#1) and the likelihood of this accident actually happening (#17).Footnote 92 Other factors which may need to be considered at this stage include the age of the expected visitors (#6), their expected conduct (#8), the purpose of their visits (#7) and the nature of the activities performed on the premises (#27). The more that similar accidents have been known to have happened in the past, the more likely it will be to conclude that the risk would have been either foreseeable or likely.Footnote 93 A history of accidents (#22) thus normally indicates that the foreseeability test is passed and that one should proceed to stage 2.
(b) Stage 2: duty to act
At stage 2, the test becomes: Would a reasonable person have expected the occupier to take remedial action against the foreseeable risk identified at stage 1 of our test?
Stage 2 asks whether there existed a particular duty to act on the part of the occupier in respect of the particular danger identified in the circumstances of this particular case. This stage is less straightforward or objective than the previous one, for it requires balancing various interests.Footnote 94 At the heart of stage 2 lay policy choices: should the stress be on protecting members of the public or on giving occupiers unfettered liberty?Footnote 95 To what extent does society want its members to be kept free from risk, thus imposing safety standards on everyone? Taking the debate in a different direction, how many resources should be devoted to safety, thereby diverting some resources from the pursuit of other social goods? In Tomlinson, Lord Hoffmann remarked that an occupier of land is not under a duty to prevent a claimant from freely choosing to engage in dangerous pastimes at their own risk. In addition to asserting this principle of individual responsibility, Lord Hoffmann also noted the negative impact overbearing safety requirements might have on the organisation of community events and other social and leisure activities and on the resources of public authorities.Footnote 96
Illustrating this last point is the Court of Appeal's judgment in Keown (child climbing the underside of a fire escape), which concluded that the hospital was not liable after acknowledging that limits to funding forced the NHS to make choices:Footnote 97 prioritising patient care over the maintenance of premises was a legitimate choice.Footnote 98 By contrast, the first instance judge had found there was a danger due to the lack of notices, warnings, barriers, fencing or security. Indeed, one might ask why the NHS would not be put in a position to attend both to patient care and to maintaining their premises in an adequately secure state. This seems to have been the perspective adopted by Spearman (fall from hospital roof), which identified – in a somewhat clunky terminology – the most vulnerable patient as the ‘lowest common denominator’ to be used when assessing the duty of care owed by a hospital to its visitors.Footnote 99
In general, it can be said that the more serious the foreseeable injury (#21), the higher the duty to do something about it. This is even more so when the chances of the risk materialising – identified at stage 1 – are high, thus possibly necessitating urgent remedial action. Inversely, the less serious the foreseeable injury, and the less likely the chances of it materialising, the less pressing the duty to do something about it now or in the future. As Lord Hoffmann remarked, ‘it may lead to the conclusion that even though injury is foreseeable … it is still in all the circumstances reasonable to do nothing about it’.Footnote 100
Deciding whether society (‘a reasonable person’) would have expected the occupier to act can turn out to be a judgment call. In some cases, the response is nonetheless straightforward. Dividing stage 2 into three successive sub-questions helps to make this clear.
(i) Do legal standards or general guidance indicate that society had expected the occupier to act?
This question ascertains whether applicable guidelines, including relevant professional or semi-professional standards (#14), or recognised general good practice exist in the field (#13). If such standards exist, the presumption should be that if the occupier fails to have regard to them the occupier was not satisfying the duty to protect lawful visitors. If Question (i) is answered positively, the relevance of Questions (ii) and (iii) is highly diminished; in most cases, the stage 2 test will already be passed.
(ii) Had the occupier satisfactorily considered the risk to the visitor?
This question first examines whether the danger was obvious (#2). If it was, the occupier is less likely to have been under a duty to take remedial action. It is important to ask whether the danger was obvious not just to a reasonable person but to persons who could be expected to visit (#7).Footnote 101 How the visitors can be expected to conduct themselves (#8) is an important consideration. So can their age (#6) be, as being young and unable to identify an obvious danger, or old and frail and less likely to be able to navigate it, militates towards finding that the occupier was under a duty to act. The same is true of other factors of vulnerability. Thus, a prison's delay in restoring electricity after a power failure to the cell of a prisoner whom G4S knew, or should have known, was suffering mobility problems, was found to have breached their occupier's duty of care.Footnote 102 At the opposite end of this spectrum, the duty to act is diminished where visitors are present in the exercise of their calling (#9), as per OLA 1957's specific terms – s 2(3)(b).Footnote 103 The purpose of the visits (#7), what is happening on the premises (#27), the desirability of the activity performed on the premises (#28), the risks visitors decide to take (#29) and the value society puts on not eliminating all risks (#30) are other potentially relevant factors.Footnote 104
(iii) Were the circumstances of the occupier such that he was actually not expected to act, despite the risk to the visitor?
The more time has passed since the danger was identified (#23), the more the occupier will be presumed to have been in violation of the duty owed to visitors. This presumption is nonetheless rebuttable. One factor that may offset the duty to act is the costs of identification and removal of the danger (#12), if these are disproportionately high. A householder should normally not be expected to take as many precautions as a professional or someone who derives economic benefit from the visits paid to their premises (#10).Footnote 105 Other elements mitigating against the obligation to take protective measures are a concern for the desirability of keeping the amenity in an unaltered state (#18), aesthetics matters (#19), or the wish to keep the premises in their original condition (#20). In a different perspective, the positive or on the contrary illegitimate or unlawful nature of the activity which is being pursued on the premises (#28) may be highly relevant. Linked to this, respecting the individual freedom to take risks (#29)Footnote 106 should also be considered, however much our society has become risk averse (#30). To paraphrase Edwards, not all bridges should have a high railing.
Deciding whether the occupier should have acted to protect a visitor from a particular danger is different from assessing whether there was a foreseeable risk of injury (stage 1) arising from a danger due to the state of the premises (threshold test). It may well be that even though there is a danger due to the state of the premises and a foreseeable risk of injury, there is no requirement on the occupier to act. Only if the answer to stage 2 is affirmative should the case proceed to stage 3.
(c) Stage 3: effectiveness of protection
Stage 3 completes the test by asking: was any action adopted by the occupier, prior to the injury occurring, appropriate?
This question examines how effective the protection put it place by the occupier was, after it is accepted that protection was owed. Judicial authority makes it clear that the occupier should not necessarily be expected to have totally eliminated the risk of injury. As North observes, the duty is not one of perfection and the occupier should not be regarded as an insurer of the premises.Footnote 107 Whether any remedial actions taken by the occupier are deemed appropriate depends on all the circumstances of the case, with the key test being that the protective measures must be proportionate but no more than proportionate.
It may be conceptually useful to distinguish between different sub-tests which arise within stage 3. It would make sense to start by checking whether any applicable regulations, guidelines (#14) or good practice (#13) were correctly applied. If they were not, it will need to be assessed whether their being respected would have been likely to prevent the accident (#24). If there were no relevant guidelines, the next test would relate to the (general) attitude of the occupier. Relevant questions include: was the occupier acting on professional/semi-professional advice (#15)? Should a risk assessment have been conducted, and if it was, was it acted upon (#16)? Were the premises generally well maintained and repairs regularly carried out as fitted the nature of the premises (#26)? Finally, the details of the concrete protective measures which were actually taken or could have been expected to be adopted will need to be scrutinised. For example, if warnings (#3) are relevant to the case, were there any and if so were they appropriate? If not, would the visitor have acted upon a proper warning or would they have ignored it (#25)? The same type of questions will arise in respect to lighting (#4), fencing (#5), and whatever other measure might be thought to be relevant to the circumstances of the case.
All the above questions will need to be examined by reference to the context of the case, eg the age range of the visitors who could be expected to visit the premises (#6), the conduct which could be expected of them (#8); what was happening on the premises (#27); the nature of the activities visitors were performing (#28); the private or profitable purpose of the visits (#10) as well as the occupier's status and possibly even resources.Footnote 108
If, at this final stage, the assessment is that the measures adopted by the occupier were not appropriate, then the occupier is to be found in violation of the duty owed to visitors.Footnote 109 If, by contrast, in all the circumstances of the case, they were appropriate, then the occupier is not in breach of duty.
By way of concluding this section, it is worth pointing out that a judge minded to dismiss a case can do so at any stage of our test. To take the example of the student who slipped in a shower room, the judge could have said the shower room was in order so that there was no danger on the premises (initial threshold test), or that the slip was unforeseeable as there had been no previously reported accidents (stage 1), or that the University was not under a duty to have taken any remedial action against the remaining risk of a slip (stage 2), or that the remedial action taken by the University in the form of a slip-resistant floor was perfectly adequate (stage 3). As long as the judge carefully distinguishes the various stages of the test and examine these in the right order, clarity and transparency of judicial reasoning will be assured.
6. Revisiting the case law
Our test makes it possible to understand why different cases put the emphasis on different factors. This can be illustrated by revisiting three cases, all of which happen to involve falls. Robinson stressed the claimant's drunken state surely because this factor more than anything else explained in the eyes of the judge the fall from the terrace; the judge must have thought there would never have been any real (‘more than minimal’) risk of injury in normal circumstances. In other words, Robinson is a stage 1 case. (Admittedly, it could also have been disposed of by saying the terrace was not inherently defective so that the threshold test was not met). If Ritz Hotel asked whether the balustrade to the balcony respected British Standards Institutions’ recommended height, this is presumably because the case would have sailed through the first stages of our test, so to speak, with the important discussion pertaining to stage 3. It also makes sense that Spearman focused on risk assessment as it concerned hospital premises visited by vulnerable patients. In sum, the insights gained from our consecutive-stage analysis reveal that the seemingly haphazard approach we criticised in section 3 of this paper may well have a logic in its selection.
Applying our test to a case often produces the same outcome as that reached by the judiciary. For example, in circumstances such as arose in Rochester Cathedral, one can accept there was a danger due to the state of the premises. At stage 1, one may conclude that the risk was unlikely to happen, as confirmed by the lack of previous incidents. Even if the case is allowed to progress to stage 2, the case goes no further on the basis that not all risks give rise to a duty to take remedial action. Taking into account, as the Court of Appeal did, the cost of identifying and removing such faults as the protruding concrete, the outcome is that there was no duty to take remedial action in this instance, the more so since the unlikelihood of serious injury if the risk materialised and the social value of the activity (allowing people to walk in the precincts of a historical cathedral) mitigated against finding liability.
To give a second example, Edwards v Sutton (fall from the low parapet bridge) could pass the initial test, in that there arguably was a danger due to the state of the premises. However, it would not necessarily progress beyond stage 1, for the most logical decision appears to be that there was no foreseeable risk of injury (given the absence of previous incidents). Had an accident happened in the past, this could militate for a finding of foreseeability and translate at stage 2 into an initial finding that there was a need to remedy a danger which was greater than first met the eye. At this stage, however, the desirability of not attending to every possible danger and of leaving individuals at liberty to take the risks they wish, as well as the amenity value of keeping a park without high fences everywhere, to which could be added the social value of the activity undertaken (enjoying a walk in a park that feels open rather than restricted) could lead to the case being closed for lack of a duty to take remedial action. Even if the case were to progress to stage 3, the finding of non liability by the Court of Appeal could still be confirmed by determining that there had been appropriate and proportionate protection, in that the low parapet would have properly indicated the border of the bridge.
If our test generally confirms the outcome reached by the courts, this is not true of the two cases reviewed above which involved children. Applying our test to their circumstances reinforces our feeling, already alluded above, that their outcome is rather incomprehensible.
Bourne Leisure (drowning of a two-year-old) meets the initial threshold test, with an expanse of water by nature a danger in a campsite.Footnote 110 Stage 1 is met too: the risk of injury is foreseeable; it is neither minimal nor unlikely to materialise (per the nature of ponds, and as confirmed by the previous near-drowning). Stage 2 is more complex, making it worthwhile to go through its three sub-tests. Was there relevant guidance such that Bourne Leisure would have been expected to act to reduce or eliminate the identified danger? The answer to question (i) appears positive, but as the judge found the RSoPA guidance not applicable, we shall proceed with asking question (ii) – had the occupier properly considered the risk to the visitors? The latter included young children who are known to go unaccompanied on caravaning sites for short periods of time. Their vulnerability should arguably have provided ‘the denominator’ (in Spearman’s language) used to assess whether the holiday park had breached its duty of care. To these children, the danger, far from being obvious, was an ‘alluring trap’.Footnote 111 This is the more so since from the occupier's perspective the visit was commercially profitable.Footnote 112 All this indicates a duty to act, which is the more pressing given the seriousness of the injury if the risk materialises. Question (iii) nonetheless needs to be raised to check whether there might have been elements counteracting this duty. The holiday park's large capacity (4,500 guests) suggests that it could have supported the cost of installing effective protective measures. Submerged pond guards would have been unlikely to diminish the aesthetic appeal of the setting. The case should thus have proceeded to stage 3 and the appropriateness of any measure taken by the occupier examined. Fencing was below the height recommended by RSoPA for domestic and school ponds. The judge found this irrelevant (despite occupiers sometimes being found liable for failing to respect non legally binding guidelines).Footnote 113 Even if the fencing was not ruled inappropriate, the appropriateness of warnings should have been examined more closely than it was.Footnote 114
The case of Anderson v Imrie, where a child was injured by a falling gate whilst playing at his friend's house, clearly met the initial test. Whether stage 1 was met is arguably more difficult to decide in that there may not have been a foreseeable risk of injury (given the absence of past incident and the fact that the children had not been expected to go into the race). If stage 1 is found to be met, stage 2 arises. The occupier either knew or ought to have known about the danger. Given the age of the visitor and the purpose of his visit, there was a duty on the part of the occupier to act. Proceeding to stage 3, warning the children not to go there, guarding against the risk by fixing the gate with a chain, and keeping an eye on the children for all but a few minutes appear to us to have been proportionate actions to the personal situation of the occupier, making the finding of liability difficult to justify.
We have already noted how occupiers’ liability is not a neutral area, but one which involves values and policy choices. One uncontroversial principle is that in general responsibility for little children rests with their parents.Footnote 115 Whilst we accept it, we suspect that Bourne Leisure and Anderson may reflect an attitude towards mothers to which we would object. This would be that mothers who fail to successfully protect their child (or a child under their supervision) against dangers which are retrospectively determined to have been, or ought to have been, foreseeable to them, deserve to be legally sanctioned.Footnote 116 We are concerned that this view – though not shared by the whole judiciaryFootnote 117 – seems to explain the outcomes in Bourne Leisure and Anderson. One could surmise that the courts are failing to put appropriate emphasis on the statutory requirement found in the OLA 1957, s 2(3)(a) regarding the obligation for occupiers to safeguard children on their land.Footnote 118
Conclusion
The OLA 1957 has failed in its attempt to make the law on occupiers’ liability to visitors simpler, or at least clearer. Higher courts frequently admonish lower courts for their misunderstandings of the law; commentators regularly note its confusion.Footnote 119 Even outcomes which are instinctly persuasive seem to rest on the application of factors which could appear to have been selected out of judicial whim rather than because this was logically compelling. Such a situation does not favour early settlement, and translates into extended litigation. Moreover, it leaves occupiers in the dark as to the steps they are legally required to take in order to ensure their visitors are reasonably safe. One benefit of our proposed test is that occupiers of land should have a clearer understanding of what OLA 1957 requires of them.
Drawing on judicial edicts, we have attempted to construct a breach of duty test which could achieve the clarity which has so far proved elusive. The test applies after it has been determined that there was a danger due to the state of the premises. Is is made of three consecutive stages which concern: (1) the foreseeability of the risk of injury and the likelihood that this risk would materialise; (2) a reasonable expectation that the occupier was under a duty to have taken remedial action; and (3) the question of whether this duty was discharged in an appropriate way. All stages involve judicial discretion but stage 2 is often the most complex, involving the implicit or explicit consideration of social and political values. These include, amongst others, the handling of risks, the assessment of the value of activities, and the role of compensation.
Adjudication is a matter of interpretation, which it is the responsibility of the judges to exercise. Our test, although linear, is not mechanical.Footnote 120 Retaining flexibility, it nonetheless brings about a more systematic approach to adjudication.Footnote 121 Greater systematisation and transparency in the way the OLA 1957 is applied should be of benefit to everyone in society, given we all assume the role of occupiers and visitors multiple times on any single day.
