Police reform and Wales’ legislative consent

Alan Trench on a little-known principle that Westminster legislation affecting devolved matters needs the consent of the legislature concerned

An interesting row has started brewing in Wales about the legislative consent motion for the Westminster Police Reform and Social Responsibility bill. Actually, there were two LCMs for this – one relating to the provisions concerning seizure powers in byelaws, the other concerning the new elected Police Commissioners and particularly the Police and Crime Panels. While the first motion was approved with only one vote against, the other produced a split vote: the parties in the Coalition at Westminster supported the motion (and bill), but it was opposed by members of the One Wales coalition, and therefore defeated. There’s BBC News coverage of it, and Lord Elis-Thomas’s forceful assertion that the legislation needs Cardiff Bay’s consent, here, and discussion of it by Matt Withers from the Western Mail here and the BBC’s Betsan Powys here.

These motions are needed by virtue of the Sewel convention – the principle that Westminster legislation affecting devolved matters needs the consent of the legislature affected. As well as consent for legislation affecting substantive devolved matters like health or education, in Scotland the convention also clearly applies to provisions that affect the legislative powers of Scottish Parliament and the executive ones of the Scottish Government. Westminster legislation adding, or removing, devolved functions needs Holyrood’s consent.  (It’s a curious feature of Welsh devolution that it doesn’t work like that there.) It’s a well-established feature of devolution in Scotland, but a comparatively new and little-used one in Wales.  The UK Government’s understanding of how the Convention is meant to work is set out in Devolution Guidance Note 9 on Post Devolution Primary Legislation Affecting Wales (available here; see particularly paragraph 17).

The legislative consent memorandum has a couple of curious features. It does not authorise particular legislative provisions – it just approves the consideration of them by the UK Parliament. This is an odd choice of wording, which doesn’t ensure that the Assembly gives its consent in a clear way (and can rescind it if need arises).  It also relates to rather a technical aspect of the new Police and Crime Panels – that they will be constituted as committees or joint committees of local authorities, when local government is a legislatively devolved matter. They don’t relate to the fact that the Assembly Government is to nominate a member of each new panel in Wales (so it confers a new executive function on WAG).  The latter would require an LCM in Scotland as well, but not it seems in Wales.

The refusal of legislative consent creates quite a jam for the Home Office. As far as I can recall, this is unprecedented; certainly no legislation affecting Scotland has ever been denied consent following a vote at Holyrood. (Provisions have never seen the light of day because opposition was signalled privately, but that’s another matter.) If the point in contention were just WAG’s power to nominate members of the panels, that power could easily be removed. But as the panels are meant to be local authority committees, there’s no such way out. It would be possible to use a drafting fix, and deem the panels not to be committees. But would that apply just in Wales, or across the whole of England and Wales? Creating them as local authority committees means that a corpus of law (and practice) applies automatically to them, which it would be necessary to write out at great length if they were to be something different. And doing that just for Wales would show that it was simply trying to get round the Assembly’s powers, not respect them in the way the Conservatives were keen to say they would before and immediately after the UK election last year.

DGN 9 makes it clear that these sorts of matters are supposed to be resolved by consultation between the UK Government department involved and the Assembly Government. It’s worth asking what happened in that consultation process. In his speech during the Assembly’s debate on the LCM, Carl Sargeant (the Welsh Minister for Social Justice and Local Government) emphasised that the Assembly Government opposed the provisions of the UK bill, and has long done so. Was that communicated to the Home Office? Did the Home Office proceed regardless? If so, it’s the author of its own misfortune. If the Assembly Government didn’t make its position clear, it has some grounds to feel aggrieved, but not much. It could have designed this policy to achieve the same objective in a way that didn’t affect devolved matters, after all.  Designation as a local authority committee is more a piece of legislative shorthand than a fundamental feature of the (non-devolved) policy.

What is important, though, is that the Home Office must respect the view of the National Assembly now it has voted, however inconvenient that may be.  Either it should find a way of dis-applying this legislation in Wales, or it should rewrite the bill so that the legislation only affects non-devolved matters and not devolved ones.  There are two reasons why it should.  One is practical.   The point of the Sewel convention is to manage the interface between devolved and non-devolved legislative powers. The key question is ‘would this be within the legislative competence of the National Assembly, if it were contained in a proposed Assembly Measure?’ The answer here appears to be yes. If the Westminster legislation were to over-rule the views of the Assembly, it would be open to the Assembly simply to repeal those provisions after they were enacted – and there would be nothing Westminster or the Home Office could do about that.

The other reason is to do with constitutional principle.  The Sewel convention is a constitutional cornerstone of how devolution works. If the UK Government is able to overrule devolved decisions whenever it wishes to, it turns the devolved governments and legislatures from entities with a distinct constitutional status and authority into simple agents of the UK Government. For devolution to work, it can’t just mean that; it has to mean the UK Government accepting that devolved governments will want to do things differently and respecting their right to do so.  The safeguard for Westminster’s sovereignty has to be regarded as an absolutely last resort (personally, I think it’s legally redundant).  When devolved and non-devolved functions are as entangled as they are, that may cause practical problems, but those should not be used to subvert the wider constitutional principle.

What Home Office does now is hugely important; this is a litmus test for what devolution really means, in Scotland and Northern Ireland as well as Wales.   And in deciding what to do, the Home Office needs to be conscious that its action will set a precedent for an even more important legislative consent motion that will shortly come to Westminster – on the Scotland bill. If Westminster fails to respect the views of Holyrood on that bill (flawed as both bill and process may be), it risks undermining not just this particular policy, but a much wider part of the UK’s  constitution. This isn’t just a technical issue about Wales, but a vital constitutional principle where the stakes are high.

This post originally appeared on Alan Trench’s blog, Devolution Matters.

Alan Trench is an honorary senior research fellow at the Constitution Unit, University College London, and is also associated with the School of Social and Political Studies at the University of Edinburgh.

Also within Politics and Policy