This is a reply to Iain Jamieson’s post which was itself a response to a post of ours discussing the significant changes to Scots landlord and tenant law contained in provisions of the ( then) Immigration Bill. Royal Assent to the Immigration Act 2016 (c.19)( the 2016 Act) has since been granted, and so the way is clear for the provisions under discussion to be implemented. Iain Jamieson’s post proceeds on the assumption that we consider that the enactment of these provisions without the consent of the Scottish Parliament constitutes a breach of the Sewel convention as currently understood. In fact, we accept that those provisions of the 2016 Act relate to reserved matters and so do not amount to legislation on devolved matters, but we may be responsible for the confusion by failing to make our position clear. To clarify matters, our position is that the UK Government’s approach to enacting the 2016 Act illustrates a problem with the operation of the convention which ought to be addressed by revising and refining the content of the convention. The revision we proposed was that the scope of the convention be extended so that it would apply, not only to provisions of UK Bills which relate to devolved matters, but also to provisions which, while designed to advance policy objectives in reserved areas, also have substantial effects on devolved areas of law and policy. So, we agree with Iain Jamieson that, as currently operated, the Sewel convention does not apply to those provisions and this reply is concerned solely with the question of whether the scope of the convention ought to be extended.
Iain Jamieson also argues against this extension of the convention, saying that it would be clearly constitutionally inappropriate on the basis that the Scottish Parliament has no power or responsibility over reserved matters and it would therefore be inappropriate to invite that Parliament to consent to provisions in a UK Bill which are outside its competence. He suggests instead that where UK legislation on reserved matters does make substantial inroads into devolved matters, that it should be properly consulted upon under the usual machinery of inter-governmental relations, and it should be expected that the UK Government would consult the Scottish Government both before and during the Bill’s progress through the UK Parliament and the Scottish Government might in turn consult the Scottish Parliament.
We make two points. The first is that the usual machinery seems to have broken down in this case. The correspondence released, although it does indicate that discussions have taken place between civil servants, suggests that UK Ministers have given little serious consideration to the Scottish Government’s views. James Brokenshire’s (Immigration Minister) reply to Margaret Burgess (Minister for Housing and Welfare at the time) of 13/10/15 explained the immigration control rationale of the measures but did not address the constitutional question and he refused to meet her on the grounds that, as immigration was a reserved matter and Scottish Parliament consent was not required, a meeting would serve no purpose. It should be obvious that the proposed legislation did require discussion with Scottish Ministers given the nature of its effects on otherwise devolved areas. This reluctance to engage may have much to do with the fractious politics of immigration control, but if the political sensitivity of a topic provides a reason for UK ministers to proceed unilaterally where there are matters of shared concern that does not bode well for the future.
The second is that, while the trend for immigration legislation to reach into areas of life beyond those associated with immigration control as traditionally envisaged is not new, this trend has intensified recently. Our post focused on landlord and tenant law, but the 2016 Act’s intrusions into devolved areas ( such as licensing and social care) go much wider and were a cause for comment by the House of Lords Constitution Committee among others. As such it goes further in its effect on generally devolved areas of law than previous UK legislation; it is pushing the envelope of what the UK government can do without the Scottish Parliament’s consent. On the evidence of this case, the current government seems to consider that the fact legislation is intended to advance a reserved purpose is by itself sufficient to justify proceeding without consent regardless of the effects of that legislation on otherwise devolved areas of law. As we said in our previous post this approach, if generalised, would allow major incursions into devolved policy areas in a wide range of areas to the detriment of the purpose of devolved government.
In support of his argument, Iain Jamieson quotes Lord Rodger in Martin v Most  UKSC 10, 2010 SC (UKSC) 40:
“Parliament cannot have intended to remove from the United Kingdom government the power to take all the necessary decisions on asylum matters.” (Martin v Most at para 87)
But, it is important to note that Lord Rodger was talking about legal competence under the Scotland Act and, as we argued in our previous post, the scope of convention need not rigidly track legal competence. It may be the case now that matters which in law are reserved matters do not fall within the scope of the convention, but that need not always be the case in future.
The rationale for extending the scope of the convention is that the current understanding does not adequately reflect the complexity of multi-level government in the UK. Let us assume that the Scotland Act 1998 was intended to create a workable constitutional settlement that satisfied Scottish aspirations for self-government. The settlement must, therefore, have nuanced ways of dealing with the multitude of ways in which reserved and devolved responsibilities interact. Provisions which, whilst intended to advance objectives of reserved policy areas, also have substantial impact on policy areas which are generally devolved are clearly a distinct case. They are provisions in which the Scottish Government and Parliament have a legitimate interest to which the UK Government and Parliament ought to give substantial consideration. As we noted previously, the 2016 Act is striking for the extent to which it seeks to enlist private sector actors as agents of immigration control and the consequence is that it cannot be treated purely as an immigration control measure; it is both an immigration measure and a housing measure, as well as a licensing, social care and labour market measure.
Would there be any adverse consequences of extending the convention? Of course, it creates the possibility that the Scottish Parliament might refuse to consent to a change proposed by the UK Government and effectively gives that Parliament a veto over UK legislation. We must be clear that there is no legal veto. Whatever the Sewel convention says, the UK Parliament retains power to legislate for Scotland on any matter, devolved or reserved. But it can be argued that there is a political veto. That in turn depends on the response of the UK Parliament. It could choose to legislate anyway but to do that would undoubtedly cause constitutional controversy, the threat of which might deter the UK government from going ahead. Would that be constitutionally illegitimate? That depends on one’s understanding of our constitutional arrangements. If we assume that a UK mandate always trumps a devolved mandate that would suggest that the convention should not be developed in the manner we suggest. But we can question that assumption and argue that our constitution has evolved to the point where some matters should be seen as shared responsibilities rather than as exclusively for devolved or UK government. Those matters include where measures are proposed to advance reserved purposes but which have substantial effects in devolved areas. Just as the reach of immigration and asylum legislation into devolved areas is not new – the UK Government’s Immigration and Asylum Act 1999 instituted a housing and support regime for destitute asylum seekers along similar lines to that discussed by Lord Rodger in Martin v Most –stirrings of institutional unease about the impact of immigration legislation in devolved areas are not a new experience either, and we can point to initiatives such as the Scottish Executive’s former Fresh Talent Initiative encouraging certain third country national students to remain in Scotland after graduation, or the Scottish Government’s response to asylum seekers and refugees on integration matters such as health as divergences in policy and practice on immigration which have been negotiated since devolution.
These are matters on which negotiation and compromise are appropriate, but the current approach puts no obligation on the UK government to negotiate and seems to give it little incentive to do so. We think that greater incentive to compromise on both sides is needed; this is particularly so when the UK and Scottish Governments are not politically aligned.
So, we are not convinced that the ‘clash of mandates’ argument makes the proposed extension to the convention constitutionally inappropriate. There is a more pragmatic argument which is that to do this might be to endanger the union by increasing the prospects of damaging conflict between Scotland and the UK Government and Parliament which might inflame nationalist sentiment. But the current approach of blithely invading devolved areas without seeking agreement also carries that risk. Moreover, it might be no bad thing from the perspective of preserving the Union if future UK governments had to be more considerate of Scottish sensibilities when suggesting policy change affecting devolved areas.
This will be all the more important in the light of the additional responsibilities conferred by the Scotland Act 2016, particularly those relating to social security which seem to increase the possibilities for confusion about where responsibility lies, a situation which will require careful political management by both Scottish and UK Governments.
Tom Mullen is a Professor and Sarah Craig a Senior Lecturer at Glasgow University School of Law