Cox v Ministry of Justice

“The law of vicarious liability is on the move.” So Lord Phillips said, in the last judgment which he delivered as President of this court, in Various Claimants v Catholic Child Welfare Society [2012] UKSC 56; [2013] 2 AC 1 (“the Christian Brothers case”), para 19. It has not yet come to a stop. This appeal, and the companion appeal in Mohamud v WM Morrison Supermarkets plc [2016] UKSC 11, provide an opportunity to take stock of where it has got to so far.

The scope of vicarious liability depends upon the answers to two questions. First, what sort of relationship has to exist between an individual and a defendant before the defendant can be made vicariously liable in tort for the conduct of that individual? Secondly, in what manner does the conduct of that individual have to be related to that relationship, in order for vicarious liability to be imposed on the defendant? Although the answers to those questions are inter-connected, the present appeal is concerned with the first question, and approaches it principally in the light of the judgment in the Christian Brothers case, where the same issue was considered. The appeal in the case of Mohamud is concerned with the second question, and approaches it principally in the light of the historical development of this branch of the law. As will appear, the present judgment also seeks to relate the approach adopted to the first question to ideas which have long been present in the law. The two judgments are intended to be complementary.

The first question arises in this case in relation to a public authority performing statutory functions for the public benefit, on the one hand, and an individual whose activities form part of the means by which the authority performs its functions, on the other hand. Specifically, the question is whether the prison service, which is an executive agency of the appellant, the Ministry of Justice, is vicariously liable for the act of a prisoner in the course of his work in a prison kitchen, where the act is negligent and causes injury to a member of the prison staff.

At the material time the respondent, Mrs Cox, worked as the catering manager at HM Prison Swansea. She had day to day charge of all aspects of catering at the prison, including the operation of the kitchen, where meals were produced for the prisoners. She was in charge of four members of staff. There were also about 20 prisoners who worked in the kitchen and came under her supervision.

On 10 September 2007 Mrs Cox was working in the kitchen with a catering assistant and about 20 prisoners. Some kitchen supplies were delivered to the ground floor of the prison, and Mrs Cox instructed four prisoners to take them upstairs to the kitchen stores. During the course of this operation, a sack of rice was dropped by one of the prisoners and burst open. Mrs Cox bent down to prop it up and prevent spillage. While she was bent over, another prisoner, Mr Inder, attempted to carry two sacks past her, lost his balance, and dropped one of the sacks on to Mrs Cox’s back, causing her injury. It is accepted that Mr Inder was negligent.

Rule 31(1) of the Prison Rules 1999 (SI 1999/728) provides that a convicted prisoner shall be required to do useful work for not more than ten hours a day. In terms of rule 31(3), no prisoner shall be set to do work not authorised by the Secretary of State. Those provisions apply to prisoners detained in privately operated prisons as well as to those operated by the prison service.

The Ministry’s current policy in relation to prisoners’ work is explained in the Green Paper, “Breaking the Cycle: Effective Punishment, Rehabilitation and Sentencing of Offenders” (2010) (Cm 7972). According to that document, the Ministry wants prisons to use the discipline and routine of regular working hours to instil an ethos of hard work into prisoners. Prison should be a place where work is central to the regime, and where offenders learn vocational skills in environments organised to replicate, as far as practical and appropriate, real working conditions. The document speaks of the “working prison”, where prisoners work a full working week, and education is geared primarily to providing prisoners with skills enabling them to perform work effectively and to get a job on release. It is said that, in public prisons, 9,000 prisoners are employed in prison workshops, with many more doing essential jobs to help prisons run smoothly. Work may be provided either by the prison or by external providers in the private, voluntary and community sectors. Prisoners may work either inside or outside the prison. In the latter situation, they may undertake voluntary or charitable work, or may undertake paid work for outside employers.

Work within a prison kitchen, in particular, is regarded as providing a good working environment, with regular hours and the ability to gain nationally recognised vocational qualifications. Prisoners can apply to work in the kitchen, and a selection is made after relevant assessments have been carried out, including a risk assessment considering such matters as the prisoner’s temperament, potential security implications, any relevant medical or hygiene problems, and the need for any relevant training in relation to such matters as manual lifting or other skills.

At Swansea in particular, prisoners were assessed for their suitability to work in the kitchen by the Inmate Regime Employment Board, a panel which carried out risk assessments and decided where prisoners should work around the prison. Those selected for work in the kitchen numbered about 20, out of a total of about 400 prisoners. They received instruction and training in relation to such matters as food hygiene, the safe use of kitchen equipment and other aspects of safety at work. Each prisoner had a training record to show what instruction he had received. The prisoners worked alongside civilian catering staff as part of the team comprising the catering department, and were accountable to the catering manager and the other civilian staff. They were subject to day-to-day supervision by the catering staff, and could be removed from the catering department in the event that their performance was unsatisfactory. Mr Inder’s instructions required him to work a six day week, from 8.30 am to 5 pm each day, with a break for lunch.

Under rule 31(6) of the Prison Rules, prisoners may be paid for their work at rates approved by the Secretary of State. It is the Ministry’s policy, as set out in Prison Service Order No 4460 (“the PSO”), entitled “Prisoners’ Pay”, that all prisoners who participate in purposeful activity must be paid. The purpose of paying prisoners is explained as being to encourage and reward their constructive participation in the regime of the establishment. Prisoners doing work in pursuance of prison rules are expressly excluded from the scope of the national minimum wage: National Minimum Wage Act 1998, section 45. At the time of the accident, Mr Inder was paid £11.55 per week. If prisoners did not work in the catering department, additional costs would have to be incurred in employing staff or engaging contractors.

The PSO also states that prison governors are legally required to deduct national insurance contributions and income tax from the earnings of prisoners whose wages exceed the thresholds, and to pay employer’s national insurance contributions. Notwithstanding the terms of the PSO, it was the Ministry’s position in this appeal that there was no liability to tax or national insurance on the earnings of prisoners working within prisons under prison rules. That was disputed on behalf of Mrs Cox, but it is unnecessary to resolve the issue. Whether vicarious liability should be imposed does not depend on the classification of the relationship for the purposes of taxation or national insurance.

It is also relevant to note the legislative provisions concerned with the provision of meals in prisons. In terms of rule 24 of the Prison Rules, no prisoner shall be allowed to have any food other than that ordinarily provided, subject to specified exceptions. Prisoners therefore depend on the prison service to be fed. Section 51 of the Prison Act 1952 provides that all expenses incurred in the maintenance of prisoners (an expression which is defined by section 53(2) as including all necessary expenses incurred for food) shall be defrayed out of moneys provided by Parliament.

The claim was heard by His Honour Judge Keyser QC in the Swansea County Court. In a judgment given on 3 May 2013, he found that the accident occurred because Mr Inder had failed to take reasonable care for Mrs Cox’s safety, but dismissed the claim on the basis that the prison service was not vicariously liable for Mr Inder’s negligence. On the basis of a careful review of the law on vicarious liability, as stated in particular at paras 35 and 47 of Lord Phillips’s judgment in the Christian Brothers case, he focused on the question whether the relationship between the prison service and Mr Inder was akin to that between an employer and an employee. He concluded that it was not. Although he accepted that there were some respects in which the prison service’s relationship with Mr Inder resembled employment, he considered that there was a crucial difference. Employment was a voluntary relationship, in which each party acted for its own advantage. The employer employed the employee as the means by which the employer’s undertaking or enterprise was carried on and furthered. The position regarding prisoners at work was quite different. The prison authorities were legally required to offer work to prisoners. They were required, by the policy set out in the Prison Service Order, to make payment for that work. Those requirements were not a matter of voluntary enterprise but of penal policy. The provision of work was a matter of prison discipline, of prisoners’ rehabilitation, and possibly of discharging the prisoners’ obligations to the community. Although the work done by prisoners might contribute to the efficient and economical operation of the prison, the situation was not one in which prisoners were furthering the business undertaking of the prison service.

An appeal against that decision was allowed by the Court of Appeal: [2014] EWCA Civ 132; [2015] QB 107. McCombe LJ, giving a judgment with which Beatson and Sharp LJJ agreed, focused like the judge on paras 35 and 47 of the Christian Brothers case, and in particular on the five features listed by Lord Phillips in para 35. He observed that the work performed by prisoners in the kitchen was essential to the functioning of the prison, and if not done by prisoners would have to be done by someone else. It was therefore done on behalf of the prison service and for its benefit. It was part of the enterprise or business of the prison service in running the prison. In short, the prison service took the benefit of this work, and there was no reason why it should not take its burdens. Although the relationship differed from a normal employment relationship in that the prisoners were bound to the prison service not by contract but by their sentences, and also in that the prisoners’ wages were nominal, those differences rendered the relationship if anything closer than one of employment: it was founded not on mutuality but on compulsion.

Vicarious liability in tort is imposed upon a person in respect of the act or omission of another individual, because of his relationship with that individual, and the connection between that relationship and the act or omission in question. Leaving aside other areas of the law where vicarious liability can operate, such as partnership and agency (with which this judgment is not concerned), the relationship is classically one of employment, and the connection is that the employee committed the act or omission in the course of his employment: that is to say, within the field of activities assigned to him, as Lord Cullen put it in Central Motors (Glasgow) Ltd v Cessnock Garage & Motor Co 1925 SC 796, 802, or, adapting the words of Diplock LJ in Ilkiw v Samuels [1963] 1 WLR 991, 1004, in the course of his job, considered broadly. That aspect of vicarious liability is fully considered by Lord Toulson in the case of Mohamud.

It has however long been recognised that a relationship can give rise to vicarious liability even in the absence of a contract of employment. For example, where an employer lends his employee to a third party, the third party may be treated as the employer for the purposes of vicarious liability. In recent years, the courts have sought to explain more generally the basis on which vicarious liability can arise out of a relationship other than that of employer and employee.

The general approach to be adopted in deciding whether a relationship other than one of employment can give rise to vicarious liability, subject to there being a sufficient connection between that relationship and the tort in question, was explained by this court in the Christian Brothers case, in a judgment given by Lord Phillips with which the other members of the court agreed. That judgment was intended to bring greater clarity to an area of the law which had been unsettled by a number of recent decisions, including those of the House of Lords in Lister v Hesley Hall Ltd [2001] UKHL 22; [2002] 1 AC 215 and Dubai Aluminium Co Ltd v Salaam [2002] UKHL 48; [2003] 2 AC 366.

The case concerned the question whether the Institute of the Brothers of the Christian Schools, an international unincorporated association whose mission was to provide children with a Christian education, was vicariously liable for the sexual abuse of children by members of the institute, otherwise known as brothers, who taught at an approved school. Another organisation managed the school and employed the brothers as teachers. It had been held to be vicariously liable for the abuse. The issue was whether the institute was also vicariously liable. The Supreme Court held that it was. Vicarious liability was thus imposed on a body which did not employ the wrongdoers, in circumstances where another body did employ them and was also vicariously liable for the same tort.

At para 35 of his judgment, Lord Phillips stated:

The five factors which Lord Phillips mentioned in para 35 are not all equally significant. The first - that the defendant is more likely than the tortfeasor to have the means to compensate the victim, and can be expected to have insured against vicarious liability - did not feature in the remainder of the judgment, and is unlikely to be of independent significance in most cases. It is, of course, true that where an individual is employed under a contract of employment, his employer is likely to have a deeper pocket, and can in any event be expected to have insured against vicarious liability. Neither of these, however, is a principled justification for imposing vicarious liability. The mere possession of wealth is not in itself any ground for imposing liability. As for insurance, employers insure themselves because they are liable: they are not liable because they have insured themselves. On the other hand, given the infinite variety of circumstances in which the question of vicarious liability might arise, it cannot be ruled out that there might be circumstances in which the absence or unavailability of insurance, or other means of meeting a potential liability, might be a relevant consideration.

The fifth of the factors - that the tortfeasor will, to a greater or lesser degree, have been under the control of the defendant - no longer has the significance that it was sometimes considered to have in the past, as Lord Phillips immediately made clear. As he explained at para 36, the ability to direct how an individual did his work was sometimes regarded as an important test of the existence of a relationship of master and servant, and came to be treated at times as the test for the imposition of vicarious liability. But it is not realistic in modern life to look for a right to direct how an employee should perform his duties as a necessary element in the relationship between employer and employee; nor indeed was it in times gone by, if one thinks for example of the degree of control which the owner of a ship could have exercised over the master while the ship was at sea. Accordingly, as Lord Phillips stated, the significance of control is that the defendant can direct what the tortfeasor does, not how he does it. So understood, it is a factor which is unlikely to be of independent significance in most cases. On the other hand, the absence of even that vestigial degree of control would be liable to negative the imposition of vicarious liability.

The remaining factors listed by Lord Phillips were that (1) the tort will have been committed as a result of activity being taken by the tortfeasor on behalf of the defendant, (2) the tortfeasor’s activity is likely to be part of the business activity of the defendant, and (3) the defendant, by employing the tortfeasor to carry on the activity, will have created the risk of the tort committed by the tortfeasor.

These three factors are inter-related. The first has been reflected historically in explanations of the vicarious liability of employers based on deemed authorisation or delegation, as for example in Turberville v Stampe (1698) 1 Ld Raym 264, 265 per Holt CJ and Bartonshill Coal Co v McGuire (1858) 3 Macq 300, 306 per Lord Chelmsford LC. The second, that the tortfeasor’s activity is likely to be an integral part of the business activity of the defendant, has long been regarded as a justification for the imposition of vicarious liability on employers, on the basis that, since the employee’s activities are undertaken as part of the activities of the employer and for its benefit, it is appropriate that the employer should bear the cost of harm wrongfully done by the employee within the field of activities assigned to him: see, for example, Duncan v Findlater (1839) 6 Cl & Fin 894, 909-910; (1839) MacL & Rob 911, 940 per Lord Brougham and Broom v Morgan [1953] 1 QB 597, 607-608 per Denning LJ. The third factor, that the defendant, by employing the tortfeasor to carry on the activities, will have created the risk of the tort committed by the tortfeasor, is very closely related to the second: since the risk of an individual behaving negligently, or indeed committing an intentional wrong, is a fact of life, anyone who employs others to carry out activities is likely to create the risk of their behaving tortiously within the field of activities assigned to them. The essential idea is that the defendant should be liable for torts that may fairly be regarded as risks of his business activities, whether they are committed for the purpose of furthering those activities or not. This idea has been emphasised in recent times in United States and Canadian authorities, sometimes in the context of an economic analysis, but has much older roots, as I have explained. It was reaffirmed in the cases of Lister and Dubai Aluminium. In the latter case, Lord Nicholls of Birkenhead said at para 21:

Lord Phillips’s analysis in the Christian Brothers case wove together these related ideas so as to develop a modern theory of vicarious liability. The result of this approach is that a relationship other than one of employment is in principle capable of giving rise to vicarious liability where harm is wrongfully done by an individual who carries on activities as an integral part of the business activities carried on by a defendant and for its benefit (rather than his activities being entirely attributable to the conduct of a recognisably independent business of his own or of a third party), and where the commission of the wrongful act is a risk created by the defendant by assigning those activities to the individual in question.

Lord Phillips illustrated the approach which I have described by considering two earlier cases in the Court of Appeal. He discussed first its decision in Viasystems (Tyneside) Ltd v Thermal Transfer (Northern) Ltd [2005] EWCA Civ 1151; [2006] QB 510. That case concerned a situation of a kind which commonly arises in modern workplaces. Employees of the third defendants were supplied to the second defendants on a labour-only basis, under a contract between the two companies, and worked under the supervision of a self-employed person also working under a contract with the second defendant. The question was whether the second defendant, as well as the third, was vicariously liable for the negligence of the employees in the course of their employment. The Court of Appeal agreed that it was, but for different reasons: May LJ considered that the imposition of vicarious liability depended on who had the right to control the employees’ activities, whereas Rix LJ formulated a test which was based not on control, but on the integration of the employees into the employer’s business enterprise. He stated that vicarious liability was imposed because the employer was treated as picking up the burden of an organisational or business relationship which he had undertaken for his own benefit. Accordingly, what one was looking for was “a situation where the employee in question, at any rate for relevant purposes, is so much a part of the work, business or organisation of both employers that it is just to make both employers answer for his negligence”: p 537. Lord Phillips endorsed the approach of Rix LJ.

Lord Phillips next considered the decision of the Court of Appeal in E v English Province of Our Lady of Charity [2012] EWCA Civ 938; [2013] QB 722. In that case, a diocesan trust, treated as being equivalent to the diocesan bishop, was held to be vicariously liable for sexual abuse committed by a Roman Catholic priest when visiting a children’s home in the diocese, on the basis that the relationship between the priest and the Roman Catholic Church was akin to employment. Lord Phillips summarised Ward LJ’s approach as asking “whether the workman was working on behalf of an enterprise or on his own behalf and, if the former, how central the workman’s activities were to the enterprise and whether these activities were integrated into the organisational structure of the enterprise”. Ward LJ found it possible to describe the relationship between the bishop and the priest as being akin to employment, as Lord Phillips put it, “by treating the ministry of the Roman Catholic Church as a business carried on by the bishop, by finding that the priest carried on that business under a degree of control by the bishop and by finding that the priest was part and parcel of the organisation of the business and integrated into it”: [2013] 2 AC 1, paras 49, 54.

Lord Phillips then considered the facts of the Christian Brothers case itself. In the context of vicarious liability, the relationship between the institute and the brothers had all the essential elements of the relationship between an employer and employees. The institute was subdivided into a hierarchical structure and conducted its activities as if it were a corporate body. The teaching activity of the brothers was undertaken because the local administration of the institute directed the brothers to undertake it. It was undertaken by the brothers in furtherance of the objective, or mission, of the institute. The manner in which the brothers were obliged to conduct themselves as teachers was dictated by the institute’s rules. The relationship between the brothers and the institute differed from that between employer and employee in that the brothers were bound to the institute not by contract but by their vows, and in that, far from the institute paying the brothers, the brothers were obliged to transfer all their earnings to the institute. Neither of these differences was material. Indeed, they rendered the relationship between the brothers and the institute closer than that of an employer and its employees. The relationship was therefore sufficiently akin to that of employer and employee to be capable of giving rise to vicarious liability.

The three cases which I have discussed illustrate the general approach set out by Lord Phillips at paras 35 and 47 of the Christian Brothers case. It may be said that the criteria are insufficiently precise to make their application to borderline cases plain and straightforward: a criticism which might, of course, also be made of other general principles of the law of tort. As Lord Nicholls observed in Dubai Aluminium at para 26, a lack of precision is inevitable, given the infinite range of circumstances where the issue arises. The court has to make a judgment, assisted by previous judicial decisions in the same or analogous contexts. Such decisions may enable the criteria to be refined in particular contexts, as Lord Phillips suggested in the Christian Brothers case at para 83.

It is important, however, to understand that the general approach which Lord Phillips described is not confined to some special category of cases, such as the sexual abuse of children. It is intended to provide a basis for identifying the circumstances in which vicarious liability may in principle be imposed outside relationships of employment. By focusing upon the business activities carried on by the defendant and their attendant risks, it directs attention to the issues which are likely to be relevant in the context of modern workplaces, where workers may in reality be part of the workforce of an organisation without having a contract of employment with it, and also reflects prevailing ideas about the responsibility of businesses for the risks which are created by their activities. It results in an extension of the scope of vicarious liability beyond the responsibility of an employer for the acts and omissions of its employees in the course of their employment, but not to the extent of imposing such liability where a tortfeasor’s activities are entirely attributable to the conduct of a recognisably independent business of his own or of a third party. An important consequence of that extension is to enable the law to maintain previous levels of protection for the victims of torts, notwithstanding changes in the legal relationships between enterprises and members of their workforces which may be motivated by factors which have nothing to do with the nature of the enterprises’ activities or the attendant risks.

It is also important not to be misled by a narrow focus on semantics: for example, by words such as “business”, “benefit”, and “enterprise”. The defendant need not be carrying on activities of a commercial nature: that is apparent not only from the cases of E and the Christian Brothers, but also from the long-established application of vicarious liability to public authorities and hospitals. It need not therefore be a business or enterprise in any ordinary sense. Nor need the benefit which it derives from the tortfeasor’s activities take the form of a profit. It is sufficient that there is a defendant which is carrying on activities in the furtherance of its own interests. The individual for whose conduct it may be vicariously liable must carry on activities assigned to him by the defendant as an integral part of its operation and for its benefit. The defendant must, by assigning those activities to him, have created a risk of his committing the tort. As the cases of Viasystems, E and the Christian Brothers show, a wide range of circumstances can satisfy those requirements.

The other lesson to be drawn from the cases of Viasystems, E and the Christian Brothers is that defendants cannot avoid vicarious liability on the basis of technical arguments about the employment status of the individual who committed the tort. As Professor John Bell noted in his article, “The Basis of Vicarious Liability” [2013] CLJ 17, what weighed with the courts in E and the Christian Brothers case was that the abusers were placed by the organisations in question, as part of their mission, in a position in which they committed a tort whose commission was a risk inherent in the activities assigned to them.

In the present case, the requirements laid down in the Christian Brothers case are met. The prison service carries on activities in furtherance of its aims. The fact that those aims are not commercially motivated, but serve the public interest, is no bar to the imposition of vicarious liability. Prisoners working in the prison kitchens, such as Mr Inder, are integrated into the operation of the prison, so that the activities assigned to them by the prison service form an integral part of the activities which it carries on in the furtherance of its aims: in particular, the activity of providing meals for prisoners. They are placed by the prison service in a position where there is a risk that they may commit a variety of negligent acts within the field of activities assigned to them. That is recognised by the health and safety training which they receive. Furthermore, they work under the direction of prison staff. Mrs Cox was injured as a result of negligence by Mr Inder in carrying on the activities assigned to him. The prison service is therefore vicariously liable to her.

A number of arguments were advanced against that conclusion on behalf of the Ministry. First and foremost, it was argued, on a number of grounds, that the relationship between the prison service and prisoners working in a prison is fundamentally different from that between a private employer and its employees. The primary purpose of the prison service, in setting prisoners to work in prison, is not to advance any business or enterprise of the prison, but to support the rehabilitation of the prisoners as an aim of penal policy. It does not seek to make a profit, but acts in the public interest. Unlike employees, the prisoners have no interest in furthering the objectives of the prison service. Even in the Christian Brothers case, the interests of the institute and the brothers were in alignment.

I am unable to accept this argument. It is true that the prison service seeks to rehabilitate prisoners, and that setting them to work is one of the means by which it attempts to achieve that objective. Rehabilitation is, however, not its only objective: it has also been an aim of penal policy since at least the nineteenth century to ensure, as it was put in a 1991 White Paper, “that convicted prisoners contribute to the cost of their upkeep by helping with the running and maintenance of the prison and by providing goods and services in prison industries and on prison farms”: “Custody , Care and Justice: The Way Ahead for the Prison Service in England and Wales” (1991) (Cm 1647), para 7.22. More importantly, when prisoners work in the prison kitchen, or in other workplaces such as the gardens or the laundry, they are integrated into the operation of the prison. The activities assigned to them are not merely of benefit to themselves: a benefit which is, moreover, merely potential and indirect. Their activities form part of the operation of the prison, and are of direct and immediate benefit to the prison service itself.

As for the other points, I have already explained that it is not essential to the imposition of vicarious liability that the defendant should seek to make a profit. Nor does vicarious liability depend upon an alignment of the objectives of the defendant and of the individual who committed the act or omission in question. It would be as naïve to imagine that all employees are subjectively committed to the interests of their employer as to imagine that no prisoner working in a prison kitchen derives any satisfaction from doing his job well or from obtaining the vocational qualifications available to him. The fact that a prisoner is required to serve part of his sentence in prison, and to undertake useful work there for nominal wages, binds him into a closer relationship with the prison service than would be the case for an employee. It strengthens, rather than weakens, the case for imposing vicarious liability.

Secondly, other aspects of the relationship between the prison service and prisoners were said to differ from the characteristics of an ordinary employment relationship. The prison service was under a duty to provide useful work for prisoners. Its choice of workers was restricted to the prisoners who happened to be held there. In that regard, it was pointed out that the courts had not imposed vicarious liability in respect of compulsory pilotage, where the master of the ship was compelled to surrender the navigation of his vessel to a pilot and had no power of selection. Furthermore, the prisoners’ pay was not a commercial wage, but a payment intended to motivate them.

These differences do not lead to the conclusion that vicarious liability should not be imposed, applying the approach approved in the Christian Brothers case. The fact that the incentive payments made to prisoners are below the level of a commercial wage reflects the context in which prisoners work, but does not entail that vicarious liability should not be imposed. The Christian Brothers case demonstrates that the payment of a wage is not essential.

The fact that the prison service, and the operators of contracted-out prisons, are under a statutory duty to provide prisoners with useful work, is not incompatible with the imposition of vicarious liability. The legislation does not itself exclude the imposition of vicarious liability. Nor is it argued that any distinct point arises under section 2(1)(a) of the Crown Proceedings Act 1947, in terms of which the Crown is subject to all those liabilities in tort to which, if it were a private person of full age and capacity, it would be subject in respect of torts committed by its servants or agents. Authorities concerned with compulsory pilotage are not in point: the prison service is not required to provide particular types of employment, or to allocate particular prisoners to particular activities. In practice, prisoners can be allocated to a variety of workplaces both inside and outside prisons, having regard to the relevant risks. More particularly, the prison service is not compelled to employ prisoners in the kitchen, and has a meaningful power of selection in respect of the prisoners it chooses to employ there. It appears from the evidence that the prison service takes particular care when selecting prisoners who are suitable to work in the kitchen, having regard to the risks involved in that setting. A restricted pool from which to select a workforce was a feature of the Christian Brothers case, and is not uncommon even in ordinary cases of employment: an employer can only select from those who apply for appointment, and may often have a small pool from which to choose.

Thirdly, it was argued that to hold the prison service vicariously liable for the acts of a prisoner would be a major development of the common law, which should be developed by the courts only cautiously. It does not appear to me that this case involves a major development of the law. The conclusion which I have reached follows from the application of the approach laid down in the Christian Brothers case.

Fourthly, it was argued that it was always necessary to ask the broader question whether it would be fair, just and reasonable to impose vicarious liability. In that regard, reliance was placed on the fact that the prison service acts for the benefit of the public, and on the fact that any liability would have to be met out of scarce public funds. It was also argued that there was no justification for imposing vicarious liability on the prison service in addition to its common law duty of care towards Mrs Cox, and its various statutory duties.

I do not consider that it is always necessary to ask the broader question. The criteria for the imposition of vicarious liability listed by Lord Phillips in the Christian Brothers case are designed, as he made clear at paras 34, 35 and 47, to ensure that it is imposed where it is fair, just and reasonable to do so. That was the whole point of seeking to align the criteria with the various policy justifications for its imposition. As I have explained, the criteria may be capable of refinement in particular contexts. But in cases where the criteria are satisfied, it should not generally be necessary to re-assess the fairness, justice and reasonableness of the result in the particular case. Such an exercise, if carried out routinely, would be liable to lead to uncertainty and inconsistency.

At the same time, the criteria are not to be applied mechanically or slavishly. As Lady Hale rightly observed in Woodland v Swimming Teachers Association [2013] UKSC 66; [2014] AC 537 at para 28, the words used by judges are not to be treated as if they were the words of a statute. Where a case concerns circumstances which have not previously been the subject of an authoritative judicial decision, it may be valuable to stand back and consider whether the imposition of vicarious liability would be fair, just and reasonable. The present appeal is such a case. On considering the matter, however, I do not regard the conclusion which I have reached as unreasonable or unjust. Those adjectives might more aptly describe a situation in which Mrs Cox’s ability to obtain compensation for the injury she suffered at work depended entirely on whether the member of the catering team who dropped the bag of rice on her back happened to be a prisoner or a civilian member of staff. For the prison service to be liable to compensate a victim of negligence by a member of the prison catering team appears to me to be just and reasonable whether the negligent member of the team is a civilian or a prisoner.

Finally, like the Fat Boy in The Pickwick Papers, counsel sought to make our flesh creep. It was argued that, if the present claim succeeded, there would be similar claims arising from the other activities undertaken by prisoners with a view to their rehabilitation, such as educational classes or offending behaviour programmes. There was also a risk of fraudulent claims being made for prisoner on prisoner incidents. A finding of vicarious liability might lead the prison service to adopt an unduly cautious approach to the type of tasks which prisoners were given the opportunity to do, given the potential impact on scarce financial resources.

I am not persuaded by these apprehensions. It is true that prisoners who participate in educational classes or offending behaviour programmes contribute towards their own rehabilitation, and in that sense may be said to be acting in furtherance of one of the aims of the prison service. But there is an intelligible distinction between taking part in activities of that kind and working as an integral part of the operation of the prison and for its benefit. As for the risk of fraudulent claims, that risk is inherent in the law relating to compensation for personal injuries, and employers, insurers and the courts are all experienced in guarding against it. As for the risk of an unduly cautious approach being adopted by the prison service, that risk is entirely speculative, and is based on a consideration only of the costs potentially resulting from the imposition of vicarious liability, without taking account of the costs which would result from a decision to cease employing prisoners and instead to employ civilian staff or external contractors at market rates of pay.

I would dismiss this appeal.