The Assembly’s legislative limbo land

Emyr Lewis explains the why the uncertain legal boundaries between Westminster and Cardiff Bay are causing problems


The current Welsh devolution settlement has created a set of complex and uncertain legal boundary lines between England and Wales. That leads to tensions between administrations and legislatures in Cardiff and London. The first ever Bill passed by the Assembly has already been the subject of litigation in the Supreme Court (see here), and the second may also be as well.

Perhaps the most fundamental of those boundary lines marks out the subjects about which the Assembly can make laws, its ‘subject-matter competence’. The way in which the boundary is drawn creates some unexpected and rather messy results. As drafted, the settlement means that the Assembly’s power to legislate includes areas which are intuitively UK matters. Conversely, this also leaves unnecessary gaps in the devolved powers of the Assembly.

Briefly, an Assembly law must relate to one of a list of subjects in Schedule 7 of the Government of Wales Act 2006, and must not fall within any exceptions in Schedule 7. In effect this creates two lists. The first is the list of subjects, and the second is the list of exceptions. This means that there are some subjects which are not contained in either list.  This is hardly surprising, given that:

  1. It would be an impossible task to describe comprehensively in two lists all aspects of the world, or even Wales, about which legislation might be made.
  2. The categories into which we classify things, the ways in which we choose to describe the world, are not closed (100 years ago, for instance, there was no such category as ‘broadcasting’).
  3. Because there are different ways in which we classify things, for example according to what colour or size they are, subjects can belong in more than one category.

There is therefore a wide range of subjects which is neither expressly included in, nor expressly excluded from, the Assembly’s subject-matter competence. They are in limbo.The Assembly can, however, make laws about something which is in limbo if it relates to one or more of the subjects in Schedule 7.

There has been more than one example of people failing to grasp this concept. It has been observed, for instance, that the Assembly cannot legislate about the criminal law – for example, in the context of banning the smacking of children – because it is not listed in Schedule 7. That is wrong. It can, because the criminal law is not excluded.

There are some surprising consequences. In some cases the Assembly’s powers are broader than was perhaps intended, and in some cases narrower.

For example, the Assembly has the power to legislate in relation to the “protection and wellbeing of children and young people”. Does this mean that the Assembly could pass a law banning the recruitment in Wales of young people between the ages of 16 and 18 into the Armed Forces? The law would ‘relate to’ the protection and wellbeing of children, and (unlike the situation in Scotland or Northern Ireland), there is no exception covering ‘the armed forces’ or ‘defence’. So the answer is probably Yes, even though many would assume that this wasn’t intended.

By contrast, Schedule 7 does not expressly give the Assembly the power to make laws to combat violence against women.  Some laws could be made to protect some women in all circumstances, and all women in some circumstances. However, for a comprehensive Act to protect all women, special pleading would be needed to fit this within the Schedule 7 powers. If this were to happen it would become the kind of case which is ripe for disagreement between Cardiff and London.

The Welsh devolution model contrasts with the models for Scotland and Northern Ireland. While there are differences between the dispensations in Northern Ireland and Scotland, the basic model is the same. That is to say, there is no list of subjects on which laws can be made.  Rather, the model sets out what is excluded. There are lists in Schedules of excluded matters, as well as exceptions to those exclusions. The starting point is that laws can be made about anything, unless excluded.

In this model, multiple categories and overlapping categories are less problematic.  There can be no subjects in limbo under this model, because the default position is that all subjects are included, unless expressly excluded. The Scottish Parliament would have no problem in passing a comprehensive Act for protecting women from violence, since it would not have to pick over and construe a list of detailed conferred powers, worrying how far it could go.  It would have to ensure that a provision did not relate to an excluded matter, but that is a far less precarious starting point than in Wales.

Furthermore, in both Scotland and Northern Ireland, the phrase ‘relating to’ is used to define what is excluded from subject-matter competence rather than what is included within it. So, the Scottish Parliament would be unable to legislate to ban the recruitment of children into the armed forces, because such a provision would ‘relate to’ something which is excluded.

There are various possible ways of resolving these difficulties.

  • Exclude from the law-making powers everything which is not expressly conferred. The problems with this approach are clear. Quite apart from the fact that it would remove a huge amount of the power for which the people of Wales voted in the 2011 referendum, this would make it impossible for instance for the Assembly to legislate in a way which created criminal sanctions to protect elderly people, because that would ‘relate to’ the criminal law as well as to the protection of elderly people. In almost any case where there are multiple possible categories and overlapping categories, the Assembly would be unable to legislate.
  • Extend the categories of exceptions, so that they cover at least those areas currently excluded from the subject-matter competence of the Scottish Parliament. While this solution deals with the problem of powers which are possibly too broad, such as the recruitment of children, it does not deal with those which are possibly too narrow, such as violence against women, because the list of conferred subjects is not changed, and there will still be limbo areas.
  • Combine the previous method with a more generous list of conferred subjects on which the Assembly could legislate. This would give the Welsh Assembly greater flexibility, but there would still be limbo areas.
  • Introduce the Scotland / Northern Ireland model into the Wales devolution settlement. This would do away with limbo areas and resolve the issues of both too narrow and too broad competence. It would mean much more certainty about the basic subject-matter competence of the Assembly, saving much work for Welsh Ministers, the Wales Office, the Assembly Commission and all their staff. In addition there would be less pettifogging turf warfare between the administrations and legislatures in London and Cardiff. It would also create greater parity between the devolved legislatures.

Emyr Lewis is a Partner with Morgan Cole LLP and a Senior Fellow at the Wales Governance Centre, Cardiff University. This article is a synopsis of preliminary evidence the Changing Union project is giving to the Silk Commission ahead of its examination of Part 2 of its remit on the powers, accountability and scrutiny capacity of the National Assembly.