In R (Black) v Secretary of State for Justice, the Supreme Court was required to consider when Acts of Parliament are binding on the Crown. It found that the existing law was inconsistent and unsatisfactory. It then made things worse.
The UK Supreme Court’s year in 2017 was framed by two constitutional cases rooted in the medieval history of the British monarchy.
At the beginning of the year, in R (Miller) v Secretary of State for Exiting the EU, the court was concerned with what remains, in the early Twenty-First Century, of the scope of the royal prerogative. And at the year’s end, in R (Black) v Secretary of State for Justice, it had to consider the circumstances in which a contemporary Act of Parliament is binding on the Crown.
The first of these cases, which was related to the legal mechanism for delivering Brexit, achieved wide publicity. The second did not. But the judgment in Black is revealing as to the state of the UK constitution as it enters 2018, and of practical application in a wide range of cases. It tells us something about why the Crown still matters in UK law.
In constitutional law, ‘the Crown’ is not a reference to the Queen in her personal capacity, but to the institutions and agencies of government carried on in her name.
Historically, the Crown was immune from civil liability, but this rule was abolished by the Crown Proceedings Act 1947. However, general immunity from criminal prosecution survives – ‘one of the few remaining bastions of the Crown’s ancient privileges‘ (Sunkin,  PL 716).
The persistence of Crown immunity in criminal cases is ‘exceptional and incongruous‘ (Sunkin again), difficult to justify in a modern constitution, and frequently embarrassing when relied on in high profile cases. Moreover, there is no constitutional need for it, and Parliament can, and does, remove the immunity in the case of selected statutory offences – see, for instance, section 11 of the Corporate Manslaughter and Corporate Homicide Act 2007. But these removals are piecemeal, and do nothing to undermine the immunity as it applies more widely in respect of common law (non-statutory) offences.
Black is not about Crown immunity, but relates to a rule of statutory interpretation that springs from essentially the same soil. This is that Acts of Parliament are to be treated by default as non-binding on the Crown. The rule has its origins in medieval conceptions of the relationship between a sovereign and his subjects – because a statute was a directive addressed by the ruler to those who were ruled, ‘a general expression in a statute such as “any person”, descriptive of those on whom the statute imposes obligations or restraints, is not to be read as including the ruler himself‘ (Diplock LJ in BBC v Johns  Ch 32 at 78).
These historical underpinnings are now outmoded in every respect, but they have given rise to a principle that still survives in modern UK law. Again, this is not because there is any constitutional necessity for it. Parliament can make exceptions to the general rule whenever it passes individual statutes, and it frequently does. Indeed, prior to Black, the rules for recognising when Parliament had done this were already well-understood and passably clear.
An Act will be treated as binding on the Crown where it says so either expressly or by necessary implication. ‘Expressly‘ means what it says – there will be a section somewhere in the Act which states that the legislation (or a part of it) binds the Crown. ‘Necessary implication‘ means that the courts recognise cases in which, even when an Act remains silent on the subject, its terms as a whole, interpreted in context, require the conclusion that Parliament intended the Crown to be bound.
Clearly, there may be room for argument about whether a ‘necessary implication’ arises in any given case. In the Court of Appeal judgment in Black, there was a useful summary of the law as it stood. Necessary implication must be seen as a ‘strict test‘ , not least because Parliament knows very well how to indicate that the Crown is bound by an Act. ‘In ascertaining the legislative intention, the court should presume that Parliament is aware of the test and that, if it intends the Crown to be bound by a statute, it is very easy to say so‘ . Consequently, it ‘is not sufficient that the implication would be desirable, reasonable or sensible‘; it is important not to ‘whittle down‘ the requirement of necessity . The type of case in which a necessary implication arises is typically one in which ‘the statutory purpose would be wholly frustrated were the Crown not to be bound‘ .
A rare recent example of a necessary implication arising in just these circumstances is R (Commissioners of HM Revenues and Customs) v HM Coroner for the City of Liverpool. In that case, the Coroners and Justice Act 2009 allowed coroners to obtain information of relevance to their inquiries. But Parliament had failed to state that the Crown was bound to provide this information. The court found that a fundamental purpose of the Act was to empower coroners to conduct effective investigations, and so ensure that the UK was compliant with Article 2 of the European Convention on Human Rights. If coroners had no right to demand information from the Crown, that purpose would be frustrated. The court was therefore willing to imply into the legislation, of necessity, a provision that it was binding on the Crown.
Although it is never expressed in these terms, the use of necessary implication in this type of case is rather like rectification in private law. Parliament really intended that the Act would bind the Crown. It both could and should have written that fact down, but somehow neglected to do so. This was a mistake. Exceptionally, the courts will correct this error where the unpalatable alternative is that the entire purpose of the legislation would otherwise be frustrated.
Difficulties – Anomalies – Absurdities
In all other cases, the rule remains that where an Act is silent on the subject, it will not be binding on the Crown. Black exemplifies why this rule – even when it is mitigated by the ‘necessary implication’ principle – leads to a range of deeply unsatisfactory outcomes.
The appellant, Mr Black, was a long-term prisoner with a range of health problems. Some of those problems were exacerbated by tobacco smoke, and smoking occurred widely in the common areas of his prison, although it was officially prohibited. Mr Black argued that the smoking ban provisions in Part 1 of the Health Act 2006, supported as they were by criminal sanctions and local authority enforcement, applied to the jail and could be relied on by him whenever the prison governor did not adequately enforce its official no smoking policy.
In response, the Secretary of State for Justice relied on the general rule of interpretation. He pointed out that Part 1 of the Health Act 2006 did not say that it was binding on the Crown. What’s more (a point Mr Black in the end could not overcome) other parts of the same legislation expressly did bind the Crown. So the approach taken by Part 1 cannot simply have been in error; the omission in that Part must be treated as intentional, the result of a deliberate decision by Parliament. And since Her Majesty’s Prison Service is a Crown body (the clue being in the name) it was not bound by the smoking ban found in that Part of the Act.
If this argument was correct, as the Supreme Court found it was, it would only prove that the rule of interpretation gives rise to three significant anomalies.
First, some pieces of legislation are said to be binding on the Crown, others not. It is very hard to discern any pattern or coherence in Parliament’s choice of what is, and what is not. For instance, as the Supreme Court noted, the Health and Safety at Work Act 1974 and the Food Safety Act 1990 both bind the Crown . So it is unlawful for Mr Black to be poisoned by fumes in the prison laundry or food bought from the prison shop, but not by second hand tobacco smoke in the prison canteen. Why? There are no good answers. An equivalent set of anomalies could be listed across the range of statute law.
Second, in the UK’s fragmented political economy, with its mix of both public and private provision of services, the application of the rule creates sharp divisions that have no good reason to exist. As the Supreme Court recognised, Part 1 of the Health Act 2006 may not be binding in relation to prisons run by the Crown, but it is binding on companies who run private prisons, even though they take prisoners under contract with the Crown . So the reason why Mr Black was not protected by the statutory smoking ban was that he was held in custody in a State-run prison. If he had been sent instead to a private prison paid for by the State, the Health Act 2006 would have applied in its entirety. Why? Again there is no good answer. The application of the law in equivalent cases should not depend on essentially random matters of resource allocation between public and private sectors.
Third, and most fundamentally, the rule only exacerbates inherent anomalies within the UK State itself. While the Supreme Court acknowledged the first and second problems, it did not take note of this one (although it has been touched on by others, most notably in the Liverpool Coroner case).
The problem arises because not all UK public bodies perform their functions on behalf of the Crown. Diplock LJ said in BBC v Johns that the Crown ‘today [he was speaking in 1964] personifies the executive government of the country‘. But this was not accurate at the time, and is certainly far from accurate now. It is quite possible in the UK to be an emanation of the State without being an emanation of the Crown.
Central government, the armed forces and State-run prisons are all emanations of the Crown. Local government, the police and State-run hospitals are not. These distinctions are broadly explicable – albeit far from obvious to anyone who did not already know – and there are valid historical reasons for them.
However, when it comes to those public bodies that have more recently been created by Parliament – a considerable part of the modern State – it is harder to find any underlying coherence. In these cases, the answer turns on whether Parliament has included a line in the statute stating that the relevant body acts on behalf of the Crown. This often appears to be no more than an arbitrary outcome of the legislative process.
Take for example the UK’s extensive regulatory State, much of it created by Parliament in the last 30 years or so, and ask whether or not a sample of its various avatars perform their functions on behalf of the Crown.
The answers appear almost entirely random. The water regulator does; the environment regulator does not. The rail regulator does; the aviation regulator does not. The energy regulator does; the communications regulator does not. The medicines regulator does; the health services regulator does not. The qualifications regulator does; the new higher education regulator does not. The schools inspectorate does; the hospitals inspectorate does not. And so on.
What is it that unites Ofwat, ORR, Ofgem, the MHRA, Ofqual and Ofsted and distinguishes them from the Environment Agency, CAA, Ofcom, NHS Improvement, OfS and the CQC? Why are parts of statute law not binding on the former, though they are binding on the latter? An enquiring observer will look in vain for any coherent or compelling answer to these questions.
The Supreme Court
Even without considering the range of anomalies generated by this third point, it is plain that the Supreme Court in Black did not much like the rule of interpretation that results in so much legislation failing to bind the Crown. It summarised the opinion of Professor Paul Craig of Oxford University – ‘In his view, careful thought is not always given to whether the Crown should be bound, which may be overlooked or receive scant attention when legislation is drafted‘ . Quite.
None of these concerns could give Mr Black the result he wanted. If an Act of Parliament expressly states that some of its parts are binding on the Crown, it is an uphill struggle to argue that other parts of the same Act are equally binding, even though Parliament was silent about them. The Supreme Court held, albeit ‘not without considerable reluctance‘ , that his case had to fail.
But this did not stop the court considering what should be done more generally with the rule of interpretation. It weighed up two options. The first, proposed by Paul Craig and others, was to reverse the presumption underlying the rule – the Crown should be bound unless legislation expressly says that it should not. The second, suggested by the current author of Bennion on Statutory Interpretation, was to abolish the rule – as with all other legislation, the only question should be what Parliament intended, with no presumption one way or the other.
The Supreme Court appeared attracted by the first option, but did not feel that it could simply reverse a rule which Parliament must be taken to have had in mind when passing so much legislation that is currently on the statute books . It briefly toyed with the intriguing possibility of changing the rule on a forward-looking basis only. However, this would be ‘wholly exceptional‘, and the case for doing so had not been made .
Equally, the court did not (or at least not quite) go down the Bennion route of abandoning the rule altogether and reverting to a general concept of Parliamentary intention – a good outcome for the authors of books on statutory interpretation; perhaps less so for legal certainty.
It did, however, move in the Bennion direction, formally retaining the rule but offering a ‘clarification‘ . For all practical purposes this entails relaxing the rule but maintaining the pretence of not doing so. From now on, while the test notionally remains the same – the Crown may be bound in a statute either expressly or by necessary implication [36(1)] – it is no longer a ‘strict‘ test . In considering whether a necessary implication arises, one must seek to discover Parliamentary intention, ‘to be gathered from the words used by Parliament, considered in the light of their context and their purpose‘ [36(4)].
In consequence, it is no longer necessary to establish that the purpose of the legislation would be wholly frustrated if the Crown was not bound. It is sufficient that a purpose of the statute would be frustrated [36(6)].
The new test is, in summary: ‘whether, in the light of the words used, their context and the purpose of the legislation, Parliament must have meant the Crown to be bound‘ .
One of the most notable features of the UK constitution is its evolutionary development. Ancient concepts such as monarchical power have over time been gradually re-purposed – without (for the most part) the need for revolutionary upheaval – to render them fit for continuing use, or at least capable of doing no harm, in the modern world. There is a lot to be said for this adaptive, careful and conservative approach to constitutional change.
However, it is liable to leave ragged edges – odd constitutional survivals that could never rationally have been designed that way. So far as these are confined to the ceremonial aspects of UK political life (what Bagehot famously called the ‘dignified’ elements of the constitution) they are tolerable. But sometimes they intrude on unavoidably real world questions, such as why it is unlawful to smoke in the common parts of a hospital, but not a prison, or in a private prison, but not a public one.
At these moments, a reality check and a serious tidying up is called for. The current state of affairs in which certain public bodies are classified as Crown bodies and others are not, while certain laws apply to Crown bodies and others do not, is arbitrary and unsatisfactory. Instead of taking a broom to the situation, the Supreme Court somehow persuaded itself that it would be useful to adopt a ‘clarification’ that only makes things more messy – leaving it even less certain, and more arguable, than before whether any given Act of Parliament is likely to bind that part of the State which we still call ‘the Crown’.
Lady Hale, current President of the Supreme Court and formerly a Law Commissioner, thinks that the Law Commission should look at how the situation could be rationalised . That is a recipe for glacial change. Nonetheless, in this, at least, she was not wrong.