Rather belatedly, perhaps, I’ve been trying to digest the judgment of the UK Supreme Court in the case of Martin and Miller v. Lord Advocate,  UKSC 10 (available here). This was decided back in March, so I’m some way behind the times with this, but the judgment makes for fascinating reading and needs serious consideration. Rather surprisingly, there’s been relatively little discussion of it so far. Apart from some news coverage (such as the Scotsman’s report, here), about the only person to comment on the case is Aidan O’Neill QC, whose 2-part post on the UK Supreme Court blog is available here and here. The lack of comment is surprising, and I’m afraid I find O’Neill’s philosophical approach to the case obscures rather than highlights the real issues.
The case is important for several reasons. It’s the first proper devolution case concerning the division of powers to reach the Supreme Court (or for that matter the House of Lords or Judicial Committee of the Privy Council). There have been plenty of cases concerning ‘devolution issues’, but these all concerned the applicability of the European Convention on Human Rights to Scottish criminal prosecutions. (Often, they didn’t even concern the verdict of the court, but whether it was proper for the prosecution to have been brought, particularly if delay was involved.) This case concerned the key issue in any system of multi-level legislative powers: whether legislation passed by the Scottish Parliament was within the legislative competence of the Parliament or not. In these cases, the question related to sentences for driving while disqualified, which is a reserved matter, set out in the Road Traffic Act 1988 and the Road Traffic Offenders Act 1988. But of course the criminal courts and criminal justice more generally are within the competence of the Scottish Parliament, and by the Criminal Proceedings etc (Reform) (Scotland) Act 2007 (passed in the Parliament’s second session – when Labour and the Liberal Democrats were in government) Holyrood sought to change the arrangements for sentencing of this particular offence. Section 45 of the 2007 Act was at issue here, and it extended the sentence that a sheriff could impose for the offence of driving while disqualified when dealing with the matter summarily (rather than by jury trial), from six months to twelve months. The maximum sentence was unchanged; what was altered were the sentencing powers of the sheriff sitting alone, rather than with a jury. This followed a careful review by the Scottish Executive of how best to dispose of the courts’ business, and how to reduce the demands placed on the higher courts.
The legal issue concerned whether the Scottish Parliament’s provision was within legislative competence or not. The fundamental rule under the Scotland Act 1998 is that the Parliament may not legislate for reserved matters, and if it does so that legislation is a nullity and has no effect. There are a number of procedural mechanisms to ensure that legislation considered at Holyrood is within competence – the Presiding Officer has to certify legislation to that effect, and any government legislation has to be similarly certified by the minister introducing it. The Act also contains a number of provisions that have the effect of ensuring that devolved legislation is regarded as within competence, even if it touches on reserved matters. These include a form of the ‘pith and substance’ test in section 29(3), and the saving for modifications of Scots criminal law applying to reserved matters which have the purpose of making the law apply consistently to reserved and non-reserved matters in section 29 (4). Schedule 4 to the Act also contains provisions, in paragraphs 2 and 3, which save legislation touching on reserved matters, if that happens only incidentally or when the effect is not ‘special’ to a reserved matter. The ‘pith and substance’ test was first developed by the Judicial Committee of the Privy Council when dealing with Canadian federalism cases under the British North America Act 1867, in the late nineteenth and early twentieth centuries. It was expressly applied to devolution to the Stormont Parliament in Northern Ireland in a 1937 House of Lords case, Gallagher v. Lynn  A.C. 863 (HL). It holds that where the ‘pith and substance’ of legislation by legislature concerns a matter within its competence, the validity of the legislation is not affected by an incidental effect on matters beyond its competence. Gallagher v. Lynn illustrates this well: the case concerned legislation to safeguard public health, by requiring all suppliers of milk sold in Northern Ireland to be licensed. That was within Stormont’s competence. The legislation had an incidental effect on cross-border trade, which was not – the case involved the prosecution of a seller of milk in Derry, who was based across the border in the Irish Republic. Despite this, as the principal purpose of the legislation was within legislative competence and the effect on a non-devolved matter incidental and irrelevant.
What’s interesting about the case is the different approaches taken by the judges, and particularly the two Scottish judges (Lord Hope and Lord Rodger) who sit on the Supreme Court. Lord Hope considered that the matter related to a non-reserved matter (Scots criminal law), and so was within competence. It was also to make the law apply consistently to reserved and non-reserved matters, and was not ‘special’ to a reserved matter, so on those grounds as well it was within the Parliament’s legislative competence. Although he started his opinion with a careful discussion of the pith and substance doctrine (both in the abstract and in its statutory form), he did not end up needing to apply it at all. Lord Hope was joined by two English judges (Lord Walker and Lord Brown – formerly Lord Justice Simon Brown), so by a majority the court determined that the legislation was within the Parliament’s legislative competence and upheld the convictions.
That’s not the end of the matter, though. The other two judges – Lord Rodger and Lord Kerr (formerly Lord Chief Justice of Northern Ireland) disagreed. They found that section 45 of the 2007 was outwith the powers of the Scottish Parliament. In reaching this conclusion, they approached the question rather differently, and it’s those differences of outlook and approach that are interesting and important.
The majority applied a pretty broad-brush analysis, rooted in the purpose of the provision in question, the policy behind it and its legislative history, even citing speeches from debates in the Scottish Parliament to ascertain these. To summarise it crudely, the legislation related to a non-reserved matter, and to the extent it did not was within the various saving provisions of the Scotland Act. It was in their view pretty plainly within devolved legislative competence.
The minority differed in two important respects. First, they considered the nature of the devolution settlement. Instead of highlighting the scope of devolved autonomy, Lords Rodger and Kerr emphasised its limitations, and the extent to which reserved and non-reserved matters interact. They also pointed out the solution to this problem – co-operation with the UK level, through the various mechanisms in the Scotland Act to enable Holyrood to legislate where the Act itself did not empower it to do so. In this case, they drew attention to section 104 of the Scotland Act, and the way this could be used to extend the competence of the Parliament if needed. Second, their approach can be described as analytical, if not reductive; the very fine-grained approach of verbal analysis that is customary in the UK’s higher courts, and which establishes very precisely the meaning of each particular clause or statutory provision. In essence, the majority has applied a purpose test, and treated the Scotland Act as needing to be construed in a different way to ordinary statutes; the minority has treated it as an exercise in conventional statutory interpretation.
The dissent is certainly right on one point. The legal division of powers between UK and devolved levels (for Wales as well as Scotland – and if not in Northern Ireland, then for reasons rooted more in administrative history than in constitutional structures) is entangled and intertwined, and in reality any policy adopted by the devolved level will need some form of assent of the UK level. That may well be negative, to ensure that no objection is raised or create an appropriate space in Westminster for devolved legislation, but some form of interaction is key to making devolution work (I developed this argument at length in chapter 3 of Devolution and Power in the United Kingdom). Of course, for Wales, such interaction is built into the arrangements of Part 3 of the Government of Wales Act 2006. However, if one needs to seek consent for a particular action, there is a possibility that it will be denied. It would certainly be wrong in principle to expect the UK Government or Parliament simply to rubber-stamp any such application made by the devolved bodies. Experience with the operation of the Government of Wales Act 2006 illustrates how fraught the process of obtaining approval from Westminster and Whitehall can be.
The problem is that this does not accord with public expectations or understandings of devolution. The wider belief is that Scotland is autonomous when it comes to devolved matters – not that it is subject to obtaining Westminster’s consent for (in practice) anything it wants to do, when that crosses the boundary of reserved matters. If one were to ask the general public whether they thought the Scottish Parliament should have had the power to pass the legislation at issue here, one would expect a huge majority to think it should (assuming they understood or could be persuaded to answer the question). Certainly the Scottish Executive and the Parliament’s legal advisers thought so – they certified the legislation as being within competence. A 3:2 split decision means, of course, that with a different composition of the court, the result could have gone the other way. It must be said that the minority’s way of thinking is perhaps more traditionally used by the higher judiciary than the approach of the majority, so that possibility is a very real one – and the minority are quite right about both the legal framework of devolution, and the existence of mechanisms to resolve them.
What this case puts on the table, but does not resolve, is the sort of approach that the Supreme Court will take to dealing with devolution issues. There are sure to be more of these in coming years – for example, a third-party challenge to Scottish legislation controlling the sale of tobacco has already been heard at first instance in the Court of Session. The approach of the majority may accord with wider public expectations, but does not fully address key features of the devolution settlement and failed to convince two senior judges on this occasion. The approach of the minority offers a route to greater legal certainty and sit better with traditions of common-law jurisprudence, but is likely to provoke intergovernmental disagreement and wider public confusion – as evidenced by the response in the Journal of the Law Society of Scotland (here). Moreover, the path to the legal certainty that it ultimately offers will entail greater uncertainty, in the amount of litigation needed to establish where the boundary between reserved and non-reserved lies, in the amount of negotiation that will be necessary to put the secondary legislation in place, and in the amount of secondary legislation that will be necessary. It will also involve a significant shift in the role of that secondary legislation, from being a useful form of flexibility in the devolution settlement, used when needed, to a central feature of its day-to-day operation.
Perhaps it’s useful that the Supreme Court has put these two alternative approaches on the table. However, they involve radically different understandings of what devolution is for and how it is meant to work, and they are fundamental to its operation. Early clarification of them, perhaps by an enlarged court, is something that is badly needed.
(A slightly different version of this post is on the UK Supreme Court blog here.)