Facts

Matthews claimed that he had sustained personal injury caused by exposure to asbestos while he was serving in the Royal Navy between 1955 and 1968. But until 1987 the Crown enjoyed an immunity under s. 10 Crown Proceedings Act 1947 from liability in tort for things done by members of the armed forces and for the nature or condition of land, premises, ships, aircraft or vehicles used for their purposes. (The repeal of this provision in 1987 did not help Matthews because it only made tort claims possible for subsequent events.) Matthews argued that this immunity was incompatible with Article 6 of the European Convention of Human Rights because it deprived him of the right to have his civil rights determined by an independent and impartial tribunal. Against this the Ministry of Defence argued that the immunity merely reflected the fact that Matthews had no relevant civil right under English law, so Article 6 was not relevant. In the House of Lords the barrister representing Matthews put particular emphasis on the fact that the immunity under s. 10 was conditional on the Secretary of State issuing a certificate stating that the claimant’s injury would be treated as attributable to his military service for the purpose of qualifying for a disablement award or pension. Thus he argued that this indicated that after 1947 Matthews had a civil right to go to court to claim damages unless and until the Secretary of State acted to bar him from doing so.

Held

The House of Lords decided that the claimant did not have any relevant right to claim damages from the Ministry of Defence under the substantive law of England, because the true effect of s 10 was that preserved part of the ancient substantive law that “the King can do no wrong”. Consequently Article 6 was not relevant. The provision relating to certification was explained as intended to assist claimants in obtaining disablement awards and pensions rather than to allow the Secretary of State to thwart the judicial enforcement of civil rights.

Comment

When there is a dispute about whether a person has a particular civil right, and it is at least arguable that the person has such a right, then Article 6 maintains that he ought to be allowed access to an independent and impartial tribunal which will determine the question. Some procedural limits on this right of access can be justified, for instance in the form of rules requiring the giving of security for costs, or providing a time limit for commencing such claims, but the European Court of Human Rights has held that such procedural limits will only be compatible with Article 6 if they are proportionate to a pressing social need. Further, the European Court of Human Rights has repeatedly stated that Article 6 does not oblige contracting states to confer any particular substantive rights on any class of individuals. So far the principles are clear. But what if a contracting state tries to avoid a judicial assessment of the proportionality of a procedural limit by presenting it as a substantive limit on the claimant’s civil rights? For instance, say that the state has a procedural concern about tort litigation by persons who cannot afford legal fees and responds by passing a law stating that ‘as a matter of law persons with savings below £50,000 are not owed duties of care by occupiers, drivers, employers … etc’. The obvious reaction is that such subterfuge cannot be allowed and this restriction on access to court should be assessed for proportionality even if it is dressed up as a substantive rule. But this answer means that it is necessary to have some test for determining whether a limit is really substantive, or is procedural-disguised-as-substantive. And if we start looking for such a test we soon realize that there is no easily identifiable natural borderline between really procedural and really substantive.

In Matthews the House of Lords decided that s 10 of the Crown Proceedings Act 1947 involved the opposite of the problem just mentioned: the immunity, because of the certification provision, looked somewhat procedural, but it was really substantive. The importance of the case for tort students, however, is in what their Lordships said about how to distinguish the really procedural from the really substantive. This is important because any rule applying in tort cases which is classified as really procedural will have to be assessed for proportionality.

Broadly, there are two possible approaches to distinguishing the really procedural from the really substantive. The first method might be called the “purposive approach” and was most clearly supported by Lords Hoffmann and Hope. The idea behind this approach is that the best way of drawing the distinction is to reflect on the underlying purpose for having such a distinction. For Lord Hoffmann that underlying purpose is to identify those cases where there is a true human rights issue involved, that is those case where courts might have to be able to intervene “to maintain the rule of law and the separation of powers” (para [29]). Thus Lord Hoffmann suggested that a provision might be subject to scrutiny under Article 6 if its effect was “to give the government an arbitrary power to stop the proceedings” (para [35]) or to create “executive power to encroach upon the functions of the judicial branch” (para [39]).

The second approach might be called the “analytical approach”. The idea behind this approach is that careful analysis of the claim being made and the rule which is preventing that claim will reveal if the rule is really procedural or really substantive. Lords Bingham and Millett seem to have supported the “analytical approach”, though neither found it easy to explain exactly what an analyst should look for. It appears, however, that there are at least three relevant matters for an analyst to take into account. First, analysis should concentrate on practical reality not form or technicality. Second, temporal sequence can be important. In other words, if a right to claim came into existence and was later restricted by a rule, then the rule is more likely to be found to be really procedural that if the restricting rule was created simultaneously with the right to claim (or before the right to claim). Third, the relationship between the reasons for the rule which prevents the claim and the reasons for the rule on which the claim is founded is important. A restricting rule is more likely to be really procedural if the reasons for it are unrelated to the reasons for the rule on which the claim is founded. Similarly, if the reasons for the restricting rule are reasons which directly deny or negate the reasons for the rule on which the claim is founded then it is more likely that the restricting rule is substantive. Support for the relevance of these three factors can be found in para [79] of Lord Millett’s speech where he said: “It is best to avoid a formalistic approach and inquire whether the rule which bars the claim is of general application and is independent of the facts which found the claim. Often the answer can be found by tracing the history of the domestic rule and examining the underlying policy to which it gives effect. It is, perhaps illogically, easier to treat restrictions on a newly created legal right as limitations of substantive law than to accord the same treatment to the withdrawal of existing legal rights.” Lord Bingham’s support for these factors is harder to locate and has to be gleaned from the way in which he analyses cases previously classified as involving procedural bars by the European Court of Human Rights (see especially paras [17] and [18]).

Perhaps the differences between these approaches will be made clearer by considering a hypothetical case. Imagine that C, a victim of a serious crime committed by X, attempts to sue a police force in negligence for carelessly investigating previous crimes committed by X with the result that X was left at liberty to commit the later crime. Imagine further that at all times further crimes by X were reasonably foreseeable and the likely future victims of X were an identifiable class, for instance, owners of small Post Offices within 25 miles of Oxford. Given all this most tort lawyers would probably predict that a trial judge, bound by Hill v Chief Constable of West Yorkshire [1989] AC 53, would hold that no duty of care was owed by the police force to C, so it would be unnecessary for a court to consider whether the police were in fact grossly negligent in their investigation. Would the rule preventing C’s claim going further be really substantive or really procedural?

If we apply the “purposive approach” then we have to ask if the rule poses a threat to the rule of law and separation of powers. According to Lord Hoffmann (para [43]) the obvious answer is that it does not. It is, for him, a rule that raises “issues of …fairness, but not of human rights.” The rule does not give the executive any power to stifle litigation, and does not prevent the courts from examining the legality of police conduct which directly restricts liberty. Thus using the “purposive approach” the rule is one which should be classified as really substantive, and consequently should not be assessed for proportionality under Article 6.

Using the “analytical approach” we find that the temporal sequence factor is problematic. It requires us to ask, “was there was a pre-existing right to claim against all professional officials in negligence and then Hill put a limit on that, or was there never any liability in such circumstances, so Hill merely confirmed the status quo?” The answer is not easy: some lawyers believe that duties of care are islands of potential liability in an ocean of no liability, whilst others believe that there is a general presumption of a duty unless and until a limit is judicially proclaimed. But probably, after the overturning of the two-stage test from Anns, we should accept that Hill did not cut back on the scope of potential liability but merely refused to make an incremental extension.

Turning to the reasons for the rules factor, we have to compare the reasons for the (supposed) rule that claimants should be able to sue professional officials in negligence with the reasons for the rule that potential victims of future crime are not owed a duty of care by police officers investigating past crimes. Are the reasons for the two rules “independent” or “unrelated”? The answer is again not easy, but on balance it is probably “no, the reasons are related”. The reasons for the “no duty rule” are reasons which negate the reasons for recognizing a duty of care: the main reasons to be urged in favour of a duty might be that it would protect C’s interests from unnecessary harm, but the reasons given by Lord Keith for opposing such a duty were that it would not actually improve detection rates and might harm the interests of other potential victims as police resources were diverted into form-filling and case preparation. These two sets of reasons seem related in a way that reasons for obviously procedural rules, like limitation rules or security for costs rules, are not related to the reasons for creating duties of care. Thus the “analytical approach” also points towards the rule being really substantive.

In future cases we might expect English courts to use the “purposive approach” and the “analytical approach” in parallel with each used to confirm the result reached with the other. But what is less clear, as Lord Walker in particular pointed out, is what approach, if any, the European Court of Human Rights will adopt. Although in Z v UK (2001) 10 BHRC 384 the Strasbourg court drew back from some of its bolder statements in Osman v UK (1998) 5 BHRC 293, the Z case divided the court (12-5) and the dicta from the previous cases which formed the foundation for the Osman ruling were not overruled. Indeed the Strasbourg court has continued to quote these dicta in cases since Z. Thus, as Lord Walker said at para [141], “The uncertain shadow of Osman v UK still lies over this area of the law.”

It may be worth ending by reflecting a little further on the suggestion that what Article 6 should potentially cover is only rules which threaten the rule of law, for instance by giving “the government an arbitrary power to stop the proceedings”. But what if the government did not have to seek such a power, because the courts were themselves so anxious to protect governmental interests that they made rules preventing such proceedings? Would such a position be any better from the perspective of the rule of law? Fortunately, whilst the English courts (following the guidance of the House of Lords in Barrett v Enfield LBC [2001] 2 AC 550) are weighing governmental arguments about over-deterrence and diversion of resources carefully, the answers are not an immediate practical concern.

This is a draft of a comment to be published in the

Student Law Review