Public Authority Liability for Careless Failure to Protect from Harm

3.1 Substantive Justifications of the Restrictive Equality Principle

Arguments (1)–(3) – no right, arbitrary limits, and absence of proof of beneficial effect – can be understood as substantive, rather than focussed on a concern about which institutions may develop the law in this area.

No right . It seems fair to say that the majority in Michael does not provide much by way of justification of the absence of a right to careful protection of one’s life when one’s life is imminently endangered. In major part, Lord Toulson merely asserts that equality principle should govern the case, with the result that no right can exist beyond those applicable to private individuals.

As a general principle – applicable to all duties of the state and private individual, and correspondingly liabilities – the equality principle is difficult to defend. Part of the moral justification of the state is that there are moral demands that must be met through collective action, but which cannot permissibly be imposed on a particular individual alone. Given my status a free person, it would normally be impermissible to require me, absent my choice, to devote my life to acting as a police service for the community. However, there is a moral need for a police service in order to prevent right violations. Consequently, so long as the burdens are not too onerous, each of us can reasonably be required to contribute to the existence of a protective institution, such as the police, which protects rights. Such a protective institution, in order to fulfil the moral need, while respecting each person’s freedom, must be tasked with duties that are not duties owed by ordinary individuals. That is the moral raison d’etre of the institution. In short, a significant part of the justification of public authorities entails a rejection of the negative equality principle.

So any justification of the equality principle which seeks to show that generally the state, or its constituent public authorities, should be subject only to the same duties as private individuals will prove too much: it will show that most of the legal framework governing modern liberal states is unjustified. This might be true, but we’d need a compelling case for believing it. If the majority’s position in Michael is to be defensible, what is needed is a justification of the equality principle which is local to tort law, or private law.

Here, then, is a more plausible version of the equality principle:

EP1. The rights and duties which a person owes in tort law should not depend upon whether that person is a private or public person.

There is no defence of EP1 in Michael. Here is one. It begins with the idea that private law is the law that applies to people simply in virtue of their status as legal person. Tort law is the area of law, then, which demarcates what we owe to each other as persons in virtue of being persons.^[26] If this view of tort law is correct, then special duties owed by virtue of the public status of the duty-bearer are necessarily external to tort law. In this way, this view of tort law supports EP1: by virtue of its nature, tort law is only concerned with what people owe each other as people; public officials are still people, and owe duties as such, and tort law enforces these duties, absent some immunity rule. Further, if public authorities owe special duties, they do not owe them in virtue of their status as a person, and so these special duties should be enforced by other areas of law. Let’s call this the ‘conceptual argument’ for EP1.

The conceptual argument has some appeal, but the account of tort law underpinning it is problematic. Tort rights are not only rights one has simply by virtue of being a person. Property rights are acquired rights and protected in tort. Similarly, tort law also protects certain rights acquired by acts of choice on the part of others, rights that arise by virtue of an ‘assumption of responsibility’.^[27] Without more, then, the idea that tort law is a law of rights and duties owed simply in virtue of being persons is difficult.^[28]

Even if we were persuaded by the conceptual argument for EP1, it supports a limited conclusion. It supports the conclusion that, if special duties are recognised on public authorities, civilly actionable by compensation claims, then these duties should not be called ‘tort’ duties. It is an argument that supports a conclusion about conceptual categorisation. It has no resources to reject the rejection of such special duties if we were to call them ‘public tort duties’.^[29] As such, the conceptual defence of EP1 is normatively inert in relation to the question of whether a person should have a judicially-created compensation claim for breach of duty owed it to by a public body in virtue of the latter’s public status.

In saying this, I am not denying the value in having words which clearly separate between the enforceable duties of individuals and the enforceable duties of public institutions. A central normative objection to the ‘constitutionalisation’ of private law is that private individuals are not under enforceable duties to protect the rights of everyone, because they have limited duties to make use of their own bodies and resources for the benefit of others.^[30] Having clear demarcations between what is owed by virtue of one’s status as an individual, and what is owed by some other status is useful: it mitigates the risk of duties that are properly owed only by public bodies creeping onto private individuals.

No Right from No Right . Consider again the following line of thought:

if a person is not entitled to the establishment of a protective institution, then the person cannot be entitled to careful performance of the protective institution’s role.

To illustrate the idea, consider:

Storm. I see you drowning in stormy waters. Despite the extraordinary danger I face, I decide to rescue you in my boat. I have not filled up the boat with enough fuel – which was readily available – and so the boat runs out fuel before I get to you. You drown. Had I filled up with fuel, I would have saved you.

In these circumstances, it is initially obscure why, if I owed you no duty to rescue you, I would then have a duty to do so carefully. In short, if A does not have a right against B that B do x, then A does not have right that, in doing x, B take reasonable care to do x.^[31]

In fact, though, a person can have a right to the careful performance of x, even if the person had no right that another undertake x at all. One reason is that the rights to assistance one has depend, in part, on the costs of assistance. If a person has done something which reduces the cost of future assistance then this may require them to do acts which they were not previously obligated to do.^[32] Consider a variant of Storm:

Storm 2. I see you drowning in stormy waters. Despite the extraordinary danger I face, I decide to rescue you in my boat. When I reach you, I have passed the worst part of the storm; from here on in, it is plain sailing and I could rescue you at no risk to myself.

In Storm 2, I am not morally obligated to commence rescuing you, given the extraordinary danger. The expected personal cost of the rescue permits me not to help. However, when I arrive beside you, and there is no longer any danger from the storm, I am morally obligated to rescue you. This is because the costs of assistance for me have changed because of the commencement of the rescue.

Suppose it is accepted arguendo that a person has no right to the establishment of a police service. If that is true, presumably it is because while individuals are important, one individual alone does not justify the enormous cost of the complex set of institutions and coercive apparatus that constitute the police. However, once that apparatus is in place, the cost of a particular act of assistance may be such that the institution is under a moral duty to provide it – in the same way that my previous actions might make me duty-bound to provide assistance that would previously have been supererogatory.

It may be objected that the law has rejected the existence of enforceable duties to provide low-cost assistance, and so the fact that the state is often positioned such that assistance is low-cost is neither here nor there. If there is no right to an easy rescue, then the fact that the state is often an easy rescuer does not justify a duty of assistance on the part of the state.

Of course, it is true that common law jurisdictions overwhelmingly reject the existence of a right to be rescued at low cost between private individuals. Responding to the ‘no right from no right’ objection does not require us to accept, however, the justifiability of legal rights to be rescued by other private individuals. The easy rescue example is merely illustrative of a more general point: as a normative matter, one may go from a situation of ‘no right’ to ‘right’ because of a change in the costs of intervention. Even if one is not entitled to the establishment of a protective institution, the existence of the institution may confer rights that previously did not exist.^[33]

Arbitrary limits . Consider, again, Lord Toulson’s concern:

If the foundation of a duty of care is the public law duty of the police for the preservation of the Queen’s peace, it is hard to see why the duty should be confined to potential victims of a particular kind of breach of the peace. Would a duty of care be owed to a person who reported a credible threat to burn down his house?

This is a reasonable concern. If the rationale of the duty to preserve the Queen’s peace is to protect an environment in which rights generally may be enjoyed by all, it is unclear why recovery should be restricted to particular categories of right. The answer to Lord Toulson’s question, in principle, is ‘Yes’. This would, however, be a wide-ranging expansion of the actionable positive duties of the police. The concern that this would be an illegitimate judicial expansion of the common law is briefly considered below.

No proven beneficial effect. If the evidential foundation for the arguments against police liability for pure omission is flimsy, it would be unconvincing to rest a new positive duty upon an equally flimsy intuition that it would improve police performance. However, it may be doubted whether one needs a positive consequence-based case for such a duty in order for the duty to be justifiable. If the law of assault prevented no assaults, it would still be justified to impose legal duties not to commit assaults, and to require those who assault others to provide compensation. In short, the law may justifiably enforce a morally justified legal duty, even if that duty does not have a beneficial impact upon conformity to the required conduct.^[34]

3.2 An Institutional Argument

Institutional justifications do not purport to provide a substantive justification of the negative, or limiting, aspect of the equality principle – the aspect which precludes more extensive duties than those owed by private individuals. Consequently, they do not need to reject, or call in question, the justifiability of the statutory examples of public body positive duties that give rise to rights to compensation – for instance the statutory right to compensation for a public utility supplier’s failure to maintain a dwelling’s connection to the water supply,^[35] the statutory right to compensation for harm suffered due to failure to repair a highway by a highway authority,^[36] or the various state positive duties that are rendered civilly actionable under the Human Rights Act 1998.

It is this kind of argument which I find most persuasive against the recognition of new positive, common law, duties to take reasonable care to protect from harm. It is one that is presented almost as an after-thought in the majority’s judgment, after the substantive justifications above are adduced. But for anyone who accepts the substantive justifiability of (1) positive duties on the state going beyond those owed by individuals, and (2) those duties giving rise to compensation claims in the event of breach, this is a more attractive possibility for justifying the decision in Michael.

Essentially, the institutional argument is that the recognition of such a duty would not be an incremental development of the law that is within the proper scope of judicial authority. Two considerations might be adduced to support this.

First, nowhere else in the tort of negligence is there a special common law duty rule geared towards public authorities specifically.

Second, it is likely that the introduction of a special positive duty upon the police would require modification of other legal doctrines. In particular, it would require a modification of the doctrine of vicarious liability. A public tortfeasor is personally liable for their wrong, albeit that it is typically the case that their liability is not enforced. This personal liability seems potentially unfair in cases of negligent wrongs when the duty being discharged is essentially a collective duty, not one owed personally by the tortfeasor in virtue of their humanity, but a duty that is being discharged on behalf of everyone.

There are forceful rejoinders to these considerations, however. First, the common law has developed public-authority-specific torts in the past: the tort of misfeasance in public office. Second, there are immunity rules created at common law in the context of certain inherently public activities: the combat immunity rule pares back the liability of those engaged in war and the immunity of superior court judges is judge-created. Third, the unfairness in untrammelled personal liability of public office-holders for negligent wrongs arguably infects positive act cases, too, and yet this has not prevented the development of duties of care in positive act cases.^[37]

4.1 Ordinary Duties

There is a simple, and plausible, line of thought which can be offered to justify, as a default, personal liability of officials for ordinary right violations. It begins with the observation that certain duties are owed, and enforceable, simply by being a member of the moral community. Call the subset of these duties which are enforceable by others, ‘basic duties’.^[38] For example, a person’s duty not to defame another does not arise because of any special relation or role in which the person stands to others. The moral case for the duty can be made independently of such facts. If the moral case for the duty can be made independently of such facts – if it holds simply in virtue of being a moral agent – then it is initially obscure why one’s choice to occupy some special role or relation could defeat the duty or render it unenforceable. John Gardner expressed this idea as follows:

Although police officers as such are indeed in special moral positions, there is no distinct ‘political’ morality applicable to them that displaces ordinary moral judgment. Morality is just morality, and it applies to people. It applies to public officials (judges, soldiers, parliamentarians, police officers, local authority librarians, etc.) because they are people. They do not stop being people and hence do not stop being bound by morality when they put on their uniforms, or otherwise go on duty.^[39]

Gardner immediately goes on to observe that occupying a role might alter one’s circumstances in morally relevant ways, and thus alter one’s moral duties and permissions. By assuming the role of police officer, for instance, one incurs a moral duty to uphold the rule of law, and a moral duty to protect people from (certain) harms, which non-role-holders generally do not have.

Implicit in Gardner’s discussion, however, is the idea that merely occupying the role of police officer does not itself confer any new moral permission. It might indirectly alter the non-normative facts in ways that permit one to do acts that others are not permitted to do. For instance, if one’s being a police officer involves having specialist driving skills, that may make it morally permissible to drive at a higher speed than others since the risk one imposes in doing so is not unreasonable. But that is not the same as one’s role directly affecting one’s moral permissions. And this chimes with our intuitive reactions to some cases in which a person assumes a role. If I assume the role of Secretary of the International Biang Biang Noodle Society, that doesn’t affect my basic moral duties to everyone else; my unilateral choice to occupy a role has no such moral impact.

Yet there is a risk here of running together two separate issues: (1) is there a distinctive ‘political’ morality? (2) are there facts which apply typically to public officials which generate special moral permissions or immunities? One can consistently deny (1), as Gardner does, while accepting (2). Gardner’s discussion does not rule out the possibility that a public official’s role may be correlated with facts that generate special moral permissions or immunities.

Daniel Viehoff has recently made an argument of this kind.^[40] His view is that by virtue of acting on behalf of persons in a jurisdiction, in a specific sense, public officials enjoy a limited moral immunity for certain right-infringing mistakes against those persons. One acts on behalf of another, in the relevant sense, if one is guided solely by their interests in one’s deliberation, and one’s being so guided in one’s action is likely in expectation to benefit the other. The occupation of certain public offices may involve acting on behalf of citizens in this sense. This is plausibly true of judicial office-holders. If so, when a judge honestly and mistakenly imprisons another based on a reasonable assessment of the evidence, the judge is not, according to Viehoff, morally liable to defensive harm, nor liable to have their resources extracted for purposes of compensation, even though a private individual acting on their own account would be morally liable to defensive harm and to compensatory cost-bearing.^[41] Shifting the cost of right-infringing conduct to the public official is unfair because the public official is not acting on their own account, pursuing their own ends, but on behalf of citizens. This is not to say that there is a distinctive political morality. A private individual who acts on behalf of another, according to Viehoff, is similarly not morally liable to defensive harm and compensatory cost-bearing for the benefit of that other when certain conditions are met.

The only point I wish to make here is that the ‘liability for ordinary violations’ limb of state liability is not established merely by arguing that there is no distinctive political morality. If the fact that a person is acting for others is morally significant – and other aspects of tort liability, such as the contrasting effects of public versus private necessity – suggest that the law’s view is that it is – then there is more to be said in order to defend the full thrust of the positive limb of the equality principle.

4.2 Special Duties

Special duties of officials arise from the occupation of their roles. This tells us something about the nature of the duty: that it is not one owed generally, but instead it arises in virtue of the occupation of a position. But what justifies the existence of the role of police officer and which duties properly attach to it?

Part of the answer is that we owe it to each other to contribute to the existence of a system which provides adequate protection of each others’ fundamental interests. Given the intrinsic value of a human being, and such a being’s vulnerability to extinction or serious harm by the actions of others and natural events, it is morally required of each of us to make contributions to the existence of a system which contributes to the preservation of life and the preservation of basic rights. It would not constitute discharge of this moral requirement if we collectively established a system which conferred a discretion to save a life that is imminently endangered when there is no significant cost to protecting that life. The duty would only be discharged by the creation of a system of roles that imposed duties owed to others to take steps to protect life in such circumstances. In essence, each of us has a duty to contribute to the establishment of a system of role duties that are necessary to the protection of fundamental interests from risks that cannot be or are not appropriately managed by individuals alone. When a person agrees to serve in such a role, that person becomes morally bound to the role duties, and so far as the role duties are correlated with rights, those right-holders become morally entitled to demand performance from the role-holder.

If there is a duty to contribute to the setting up and maintenance of a protective institution that assigns role-duties to protect endangered life at reasonable cost, is there a duty to compensate for violations of one of the role duties? The issue here is essentially whether there is an important distinction between the entitlement to compensation for breach of ordinary duties and an entitlement to compensation for breach of special duties.

Although there is personal liability in English law for violations of ordinary duties (subject to the immunity rules mentioned), it is the state, by virtue of the doctrine of vicarious liability, which foots the bill, and so, indirectly, all of us. If citizens are required to foot the bill for official wrongs that constitute the breach of ordinary duties, the pertinent question is whether citizens can reasonably object to being required to foot the bill for official wrongs that only constitute the breach of special duties.^[42]

It might be answered that vicarious liability for breach of ordinary duties is justified because, in these situations, the operation of the state has harmed a person; it has not merely failed to protect from harm. Being the victim of harm resulting from the operation of the state, it might be said, justifiably gives one a special claim to compensation from the state. By contrast if a public official wrongfully fails to protect from harm, this gives one no special claim to compensation since there are many instances of state failure to alleviate harm. Consider two persons, P1, and P2, who suffer identical harms at the hands of two separate criminals, but in P1’s case the police wrongfully failed to protect P1. Why prioritise P1 by giving P1 a tort(-like) right to compensation over P2, who will need to rely upon a state compensation scheme, if there is one?^[43]

Notice that this is not simply a version of a somewhat crude ‘distributive justice’ objection to tort law. It is sometimes said that tort law itself arbitrarily distinguishes between those who suffer harm to due to the wrongful agency of another, and those who innocently suffer harm. This objection is unpersuasive because it would be impermissible to impose a duty to compensate upon any particular individual for another’s harm, if the latter was not individually responsible for that harm. While individuals have limited moral duties of assistance to each other, it would generally be wrong to impose full compensatory burdens upon one individual for another individual’s harm absent responsibility for the harm. Consequently there is no moral arbitrariness or injustice in protecting non-wrongdoers (compared to wrongdoers) from compensatory duties to other individuals.

But this distributive objection has more bite when we are considering the use of state resources to compensate. Here the question is essentially: when should citizens collectively pay (through the doctrine of vicarious liability, and taxation) for injuries for which they are not individually responsible? Perhaps they should only pay for compensation for those who are worsened in relation to their right-protected interests by the operation of the state, but not for those whose position is not improved by the operation of the state. In these circumstances, it might be said that compensation is a ‘cost of doing state business’, and each us is required to contribute to that cost in so far as the state is reasonably just.

Ultimately, however, this seems unpersuasive. Each of us may reasonably be required to contribute towards remedying the wrongs done by public officials who act on our behalf. If it is accepted that such persons may commit wrongs simply by failing to assist, because sometimes such persons may have a role-duty to assist, then it should also be accepted that such wrongs are ones for which all of us may be called upon to contribute towards remedying. If each of us is required to contribute to the protection of each others’ fundamental interests, then this requires both compensation or welfare schemes for non-wrongfully caused serious injury and compensation for breach of official duties to take reasonable care to protect from right-protected harm.