How Not to do Devolution: Wales and the Problem of Legislative Competence

Posted in: UK politics

Dr David Moon is Lecturer in Politics and Tomos Evans is a postgraduate research student in the Department of Politics, Languages & International Studies at the University of Bath.

There is agreement that the reserved powers model – on which Scotland and Northern Ireland operate – is a better way of devolving power than the conferred powers model used by Wales. But the Wales Act 2017, which will move Wales to a reserved powers model, will still not solve many of the problems that the National Assembly for Wales faces.

Devolution in the UK is still in its teenage years and while both Scotland and Northern Ireland have experienced little change in the structures underpinning their devolution settlements, Wales’ settlement is about to undergo its third major change since 1998, thanks to the Wales Act 2017.

The story of Welsh constitution-building has been an example of how the process should not be done: for every step forward, it has seemed as if two were taken back. The new Act does not mark an end to this story; in March 2017, Richard Rawlings – who serves on the Welsh advisory committee of the Law Commission – claimed that the Wales Act 2017 “carries the seeds of its own destruction”. Understanding why this is the case and the lessons that others can learn from the Welsh example, it is important to focus upon the question of legislative competence; the different ways of devolving powers; and the torturous route taken to our current situation.

After a wafer-thin majority in favour of establishing a ‘Welsh Assembly’ in the 1997 referendum, the first National Assembly for Wales met on 12 May 1999. The National Assembly’s structures and powers were, at this point, reminiscent of a county council, not a parliament, being unable to pass primary legislation or raise any taxes. Marking the first major reform of the system, the Government of Wales Act 2006 provided the National Assembly with the means to pass ‘Measures’ – essentially Acts, but by a different name. This ability to pass Measures was closely guarded by Westminster, with the Assembly only able to gain Measure-making powers via an Act of Parliament or a ‘Legislative Competence Order’ (LCO), and with the 20 areas devolved to the Assembly being incrementally populated with ‘matters’ over which Measures could be passed.

This LCO system for legislating, which was in place between 2007 and 2011, has been rightly criticised as cumbersome, opaque, and bureaucratic. Yet, putting those issues to one side, the constitutional cleverness of the system yielded certain positives that only really became clear in retrospect, with the apparent step forward taken through the shift in 2011 to a system with direct, primary law-making powers for the Assembly.

The positives of the LCO system, outlined below, were directly related to the conferred powers model upon which Wales’ devolution settlement was founded. A conferred powers model of devolution provides a list of things you can do; in contrast, reserved powers models, like Scotland and Northern Ireland’s, provides a list of things you cannot do.

Since 2011, still operating within the constraints of a particular list of specific subjects upon which it could legislate, Wales’ settlement regularly falls victim to what might be labelled the ‘grey spots of GoWA’, the ‘silent subjects’, or to use the terminology of the UK’s Changing Union Partnership, ‘areas in limbo’. These are policy areas where there is a lack of clarity over who has power over certain broadly defined subject areas. The lack of clarity in the Welsh devolution settlement has created problems when formulating policy and legislating.

Two Acts of the Assembly – including the first ever – have been referred to the Supreme Court by the UK Government as they believed them to be outside of the Assembly’s competence. The Agricultural Sector (Wales) Bill ended up in the Supreme Court because of a major disagreement stemming from the abolition of the Agricultural Wages Board in 2013. The case is a prime example of the ‘grey spots of GoWA’: The Supreme Court held that the legislation was within the competence because ‘agricultural wages’ was not explicitly outside of the Assembly’s powers.

The most recent example of the lack of clarity affecting governing in Wales is around the UK Government’s Trade Union Act 2016. As the Bill made its way through the Westminster Parliament, the Welsh Government contested that some of its provisions should not extend to devolved public services and vowed to bring forward its own legislation to disapply those provisions in Wales at the earliest opportunity. In January 2017, the Welsh Government introduced the Trade Union (Wales) Bill to the National Assembly to repeal certain provisions of the Trade Union Act 2016 as they apply in Wales. The UK Government argues that trade union law is non-devolved and so it is highly likely that a third piece of Welsh legislation will end up in the Supreme Court via a UK Government referral.

Faced with this situation, it is with hindsight that the positives provided by the LCO system’s convoluted processes have become visible, through its otherwise overarching flaws:

(i) it provided clarity over legislative competence;

(ii) there was a presumption in favour of increasing the powers of the Assembly;

(iii) inter-governmental disagreements were dealt with at the beginning of the policy or law making process.

What the system did, in a conferred powers model, was provide the clarity which has been lacking and created problems since 2011 – avoiding subsequent trips to the Supreme Court.

The lack of clarity in the current system has been widely acknowledged by academics, civil society, and policymakers alike. But rather than return to the inherently flawed LCO system, there has been general agreement on the need to move from a conferred to a reserved powers model of devolution. The result is that the Wales Act 2017 will overhaul the Welsh devolution settlement and put it on the same foundations as Scotland and Northern Ireland – a reserved powers model. This reform was recommended by the Silk Commission in 2014 as a way of reducing the chronic uncertainties in Wales’ devolution settlement.

In its previous form the draft Wales Bill manged to make a fudge of the move to a reserved powers model: although it would have introduced more clarity, it would have created more hoops for the National Assembly to jump through to legislate, thus making it even more complex than the current system. As Robert Thomas’ comprehensive briefing on the matter highlights, the proposals solved one problem but created another, potentially worse one. One step forward, had become two steps back.

Revised and reformed, the reserved powers model that the Wales Act 2017 institutes is a superior way of devolving power, but it does not provide complete clarity or eliminate the scope for disagreement entirely. The reserved powers model is the best way to devolve power, but as the draft Wales Bill proves, the model in itself is no panacea.

Resultantly, although the Wales Act 2017 is supposed to provide a “strong and lasting” devolution settlement, it has only settled the conferred versus reserved powers model debate. Discussions over whether powers should sit in Cardiff or London – especially over policing and the judiciary – will continue. So it is that, as in 1998, 2007, and 2011, the 2017 Act creates as much as it conquers problems; Wales, once again, continues to provide a prime example to policymakers of what not to do.

This post, which was originally published on LSE's British Politics and Policy blog, draws upon arguments outlined in greater depth in the authors’ recent article “Welsh devolution and the problem of legislative competence”, published in British Politics.

Posted in: UK politics


  • (we won't publish this)

Write a response