This is slightly amended version of an article that appeared in the summer 2011 edition of the Institute of Welsh Affairs’ Agenda magazine.
The legislative powers conferred on the National Assembly thanks to the Yes vote in the March 2011 referendum are much more ample, much clearer and much more legally certain than those it had under before the referendum. Part 4 of the Government of Wales Act 2006 is a much better foundation for law-making than Part 3. However, the Yes vote is by no means the end of the story of Wales’s constitutional development either. Part 4 also has some serious legal flaws, and making it work is not going to be a smooth ride.
Some of the reasons have little to do with Wales. The main one is the fact that devolved legislation is increasingly finding its way into the courts, so the question of how courts approach devolved laws is important in a way it was not in the first decade of devolution. Unsurprisingly, it is Scottish legislation that has been in the lead here. While the bulk of the ‘devolution issues’ cases considered by the Judicial Committee of the Privy Council or the UK Supreme Court in the first decade of devolution concerned human rights issues and the treatment of those charged with a criminal offence or in prison, that is now changing. The cases before the courts now involve issues of purely domestic law, not the application of the European Convention on Human Rights, and raise wide-ranging questions of principles about how devolution works.
One case, Martin and Miller v. Lord Advocate, decided in the summer of 2010 (available here), asked the Supreme Court to consider how it should approach the task of judging whether devolved legislation was within the powers of the Scottish Parliament. The issue here was not directly relevant to Wales – it was about the sentencing powers of a Sheriff sitting alone on summary business (not with a jury), when dealing with a driving offence. Criminal justice, including organisation of the courts and sentencing policy, is a devolved matter in Scotland, but road traffic offences are reserved to Westminster. What’s important is the way the court looked at the question. Should the courts look at the general purpose of the legislative provision to decide whether it was within devolved competence? Or should they establish this by means of detailed verbal analysis, the traditional way of understanding what statute law means? In the Martin and Miller case, the judges split by three to two. The majority took a particularly broad-brush approach and concluded the legislation was within devolved powers. The minority took an ‘analytical’ approach, and found it was not. So, while the legislation was upheld, the Supreme Court set a two-pronged problem. On one hand, a differently composed court – and this was just five judges, of its 12 members – could easily have reached a different view. On the other, the minority’s proposed solution – for the Scottish Government to have negotiated and obtained wider powers from Westminster to cover the legislation it proposed to make, using the order-making powers in the Scotland Act 1998 – would, for many practical purposes, turn the Scottish settlement into something very like Part 3 of the Government of Wales Act 2006. (See HERE for my more detailed discussion of that case and its implications.)
There has not yet been a judgment in the second major Supreme Court case, Axa General Insurance v. Lord Advocate. This was heard by the court in June 2011, and the Welsh Government formally intervened in proceedings both on appeal in Scotland and at the Middlesex Guildhall. We can expect a judgement fairly soon, probably in mid-October. It concerns legislation passed by the Scottish Parliament (and the Northern Ireland Assembly) to allow claims for damages for ‘pleural plaques’, a condition caused by exposure to asbestos so common in areas where there used to be heavy industry. The House of Lords decided in 2004 that this condition (which doesn’t directly affect the lives of those who have it) did not found a claim in damages. The Scottish Government and Northern Ireland Executive considered this unacceptable, and proposed legislation to reinstate claims for damages. A number of insurance companies, led by Axa, challenged this legislation. Central to their arguments is the view that devolved legislation cannot exceed certain common-law limits of rationality, and that by overturning a decision of the House of Lords this legislation was irrational. While it’s hard in principle to question the idea that devolved legislatures – which legally are creatures of Westminster, even if they were endorsed by referendums and have their own electoral mandates – are subject to some such limit. The real question is how that is applied. Even local authorities benefit from what is known as the ‘Wednesbury’ test, which holds that a decision is only unlawful if it is ‘so unreasonable that no reasonable authority could have reached it.’ That sounds tautological (though it isn’t), and it creates a pretty high threshold for unreasonableness to be proved. The threshold should be higher than that for a devolved legislature, and the Supreme Court is unlikely to encourage further challenges, but we must wait for the Supreme Court judgment to see what it says.
The question for Wales is how devolved Welsh law-making will be able to operate in this environment. How well would Acts of the National Assembly be able to stand up to the sort of scrutiny that the courts are subjecting Scottish legislation to? After all, there are some big differences in how Scottish and Welsh devolution work – the difference between the ‘reserved powers’ and ‘defined powers’ model. The Scottish Parliament can pass any laws it wishes, unless they concern matters reserved to Westminster. The National Assembly can only pass laws on subjects where it expressly has powers. The fact that the Assembly has powers over ‘health and health services’ doesn’t mean it can do anything it likes that affects health. Legislation must fall within the definition set out in Schedule 7 to the 2006 Act, subject 9 of which (omitting the exceptions) covers:
Promotion of health. Prevention, treatment and alleviation of disease, illness, injury, disability and mental disorder. Control of disease. Family planning. Provision of health services, including medical, dental, ophthalmic, pharmaceutical and ancillary services and facilities. Clinical governance and standards of health care. Organisation and funding of national health service.
That looks comprehensive, but is it? Do those powers include the power to set up a system of settling potential claims against the NHS in such a way as to limit complainants’ access to the courts? (That is what the NHS Redress (Wales) Measure 2008 does.) It’s clearly not to do with promoting health or treating or preventing disease. Such legislation probably falls within the general category of ‘organisation and funding of the national health service’, but one cannot be certain – and the issue is not whether readers think this, or even AMs think it, but ultimately whether a court does.
That is where the real difficulties may arise. Scottish legislation may face serious problems if the courts do take an ‘analytical’ approach to considering it, but Welsh legislation would be in much greater difficulties. The idea that every clause of every Welsh bill should be subject to detailed scrutiny, to ensure that it directly relates to the powers set out in Schedule 7 of the 2006 Act, could make legislating very difficult indeed. It would clearly run against the current of what people thought they were voting for at the March referendum, as well as creating a good deal of practical difficulty.
Making the ‘Part 4’ system work will call for give and take in a number of quarters. The Welsh Government and the Assembly are going to have to be canny in deciding how far their powers go. Even then, those powers may need to be tweaked before legislation can be passed. Schedule 7, setting out those detailed law-making powers, has already been amended twice, before it even comes into use. It is likely to require further amendment in the coming years, to ensure that the National Assembly is able to legislate properly on the twenty subject areas devolved to it. Section 109 of the 2006 Act provides for this to be done by orders in council, which have to be approved by both Houses of Parliament at Westminster and the National Assembly. However, it doesn’t address the question of who initiates such an order or create the more defined procedure that applied for legislative competence orders under Part 3. This isn’t a case of ‘creeping devolution’, so much as simply making it work, and indeed Schedule 7 wouldn’t be a very appropriate legislative vehicle to confer whole new areas of the competence anyway. If Schedule 7 is regarded as being set in stone, that is likely to hamstring the Assembly, to no-one’s real benefit – most certainly not that of the Welsh public. So Welsh law-making will in turn need ongoing cooperation between the Welsh Government and the Wales Office, and between the National Assembly and UK Parliament. There’s nothing new about this; the number of orders made under the Scotland Act 1998, to adjust law either side of the border following legislation made at Holyrood, is pretty large – roughly half the Acts of the Scottish Parliament necessitate some sort of tweak somewhere. The much greater entanglement of government and public services between England and Wales only increases the need for that for Wales.
March’s referendum result may create a much more powerful legislature than Wales has known since before the middle ages, but that legislature remains part of a complicated system of devolved governance. Making this new system work in the interests of the people of Wales will be a complex task, and there will be some rough patches on the way. But managing the complex relationship will be central to making the new phase of Welsh devolution work. Law-making powers aren’t the end of the story, but the start of a new one.