The Sewel convention has rightly come to be seen as key to the working of devolution in the United Kingdom. It may have first been envisaged as a way of enabling Westminster to continue to legislate for devolved matters and maintaining something like the practical pre-devolution status quo in policy-making, when convenient and politically acceptable, but it was quickly understood to mean more than that.
One reason may be that devolved legislative powers are more far-reaching than was at first appreciated. More important, though, is the emergence of the ‘constitutional’ dimension of the convention. The wording used in the Memorandum of Understanding (first agreed in 1999 and not changed since then) may refer to ‘the UK Parliament … not normally legislat[ing] with regard to devolved matters except with the agreement of the devolved legislatures’, but Devolution Guidance Note 10 on Post Devolution Primary Legislation regarding Scotland has been clear that consent is also required where there are changes to the functions of the Scottish Executive/Government or Parliament. This means that functions cannot be removed from the devolved tier of government without its consent. It also means functions cannot be added without consent, meaning that the UK tier cannot get rid of inconvenient functions, or transfer them without adequate funding, if a devolved legislature objects.
How the convention applies to Wales has always been less clear than its application to Scotland. The formal basis is the same for all three devolved governments; that statement in the Memorandum of Understanding. The National Assembly for Wales had very limited legislative powers until 2007, of course, and in that circumstance had no formal right to approve changes to its functions, merely a generally accepted right to be consulted about changes made by Westminster legislation (which also explains why the Secretary of State for Wales had the right to attend and participate in Assembly proceedings and receive documents about those). With the assumption of law-making powers it was clear that the Sewel convention applied to Westminster legislation that affected substantively devolved functions and the Assembly began to consider legislative consent motions, and sometimes to reject them.
However, the Assembly was initially reluctant to engage with the ‘constitutional dimension’. This was the case even when the UK Parliament conferred functions directly on Welsh ministers without any engagement with the Assembly – a matter that drew the attention of the Assembly’s Constitutional and Legislative Affairs Committee in 2011-12. Another odd feature of the 2007-11 term was that Assembly consent was required for changes to the Assembly’s law-making powers when made by legislative competence order, but not if that was done by provision in Westminster primary legislation (so-called ‘framework’ powers). A reading of the current (2012) version of Devolution Guidance Note 9 on Parliamentary and Assembly Primary Legislation Affecting Wales (available here in PDF format) confirms that the UK Government understands that the convention applies to changes to devolved powers and functions. (The Note is hardly an example of clear drafting, but does note in paragraph 14 that the convention is triggered, inter alia, by provisions ‘likely to have an impact on the Assembly’s legislative competence or the Welsh Ministers’ executive powers’.)
Perhaps the key test regarding whether the convention applies in its ‘constitutional’ dimension applies in Wales arose with proceedings on the Wales Act 2014. The National Assembly considered the bill in some detail, with inquiries by both the Finance Committee and Constitutional and Legislative Affairs Committee. Its legislative consent motion was approved on 1 July 2014, after completion of consideration of the bill in the Commons and formal introduction into the Lords but before Lords second reading. The Assembly has certainly asserted its right to consent (or not to consent) to Westminster legislation that changes its powers or those of the Welsh Government, and it is hard to see how that power can be denied given the general applicability of the commitment in the Memorandum of Understanding and the Assembly’s exercise of that right. It’s clearly recognised in the explanatory notes to the draft bill, paragraph 10 of which notes the need for a legislative consent motion for the bill.
This has two implications for the current debate. First, when the Wales bill (published as a draft last week) is finalised and introduced into Parliament, this will require the legislative consent of the National Assembly as well. In effect, the National Assembly will be a kind of third chamber of Parliament, albeit without the detailed process of second reading, committee and report stages – just as the Scottish Parliament was for the Scotland Act 2012 and is for the Scotland bill it is presently considering. If the Assembly does not assent to the bill, the bill should not pass.
If the bill were to be enacted without consent, it would lead to a major constitutional crisis. In these circumstances, there is no emergency that might trigger the limitation requiring devolved consent only in ‘normal’ circumstances, so the UK Government would be asserting Westminster’s sovereignty even when faced with opposition from the National Assembly. That amounts to tearing up the constitutional fabric of devolution – and the very entrenchment of devolved rights that was agreed in the St David’s Day process and which the draft bill seeks to implement. Worse, such a move would make it clear that very different rules apply to Scotland and to Wales, with Wales plainly and visibly being a second-class devolved nation. As a result, arguments about the UK being a diverse multinational union which accepts and even finds unity in that diversity, as we sought to argue in the Bingham Centre Devolution review, would be very hard to sustain.
The second implication is for the requirement for a referendum on the partial devolution of income tax set out in the Wales Act 2014. Any change to that would similarly need the consent of the National Assembly, even if the referendum has not yet been held. The powers conferred on the Welsh Government and the National Assembly (to trigger a referendum and call for it by a two-thirds majority) are clearly Assembly functions, and cannot be altered without the Assembly’s consent. Any idea of a ‘quickie’ referendum without the Assembly’s approval is similarly, constitutionally, a non-starter.
As the Trade Union bill shows, the requirement for devolved legislative consent is a complex area, made more complex by different frameworks for devolved powers. The complexity is likely to increase if it is put on a statutory basis, as the Scotland bill and draft Wales bill propose, as it will then become a possible subject for litigation. (That commitment will prove to be hugely problematic; the courts are profoundly ill-equipped to consider such matters, and the commitment runs directly contrary to the guarantee of Parliamentary immunity from having its proceedings questioned in any court under the 1689 Bill of Rights.) But the Sewel convention has huge value, to the point where it can be regarded as a piece of constitutional magic. It puts in place a powerful safeguard for devolved interests, but one which is consistent with the UK’s unwritten, political constitution. If devolved and UK interests are able to agree to a change in the framework, it can be made to happen; if they cannot, it will not. In the absence of a written constitution (with all the rigidities that brings), legislative consent provides as effective a solution to the question of constitutional protection and entrenchment as one is likely to find. It’s incumbent on UK Government and Parliament to ensure it is recognised and complied with.