Category Archives: Wales

Ann Sherlock: Supreme Court ruling on Welsh legislation.

AnnOn 9 July 2014, the Supreme Court delivered its unanimous ruling that the Agricultural Sector (Wales) Bill was within the legislative competence of the National Assembly for Wales.

The Bill had been referred to the Supreme Court in August 2013 by the Attorney General for England and Wales under section 112(1) of the Government of Wales Act 2006 (hereafter GWA 2006). This provision, whose equivalents in the Scottish and Northern Irish devolution legislation have yet to be used, allows for the referral of a Bill passed by the Assembly if the Attorney General or the Counsel General, the Welsh Government’s law officer, considers that it goes beyond the Assembly’s legislative competence.

This is the second occasion on which a Bill has been referred: the first concerned the Local Government Byelaws (Wales) Bill. The Attorney General argued that the Bill exceeded the Assembly’s competence in that it flouted a general restriction on the Assembly’s competence by removing or modifying a function of a Minister of the Crown. In the event, the Supreme Court ruled unanimously that, while the Bill did remove some of the Secretary of State’s functions, that removal was saved by the exception in the GWA 2006 which permits the removal of a function as long as it is ‘incidental to, or consequential on, any other provision contained in the Act of the Assembly.’ In the case of the Agricultural Sector (Wales) Bill, the issue related to whether the legislation went outside the subject areas in which the Assembly has legislative competence.

The Assembly’s legislative competence

As will be known, unlike in Scotland and Northern Ireland where all power is devolved unless expressly reserved (or excepted) to the UK Parliament, the GWA 2006 uses a conferred powers model under which the Assembly may legislate only on those subjects enumerated in the Act. Since 2011, those subjects are set out in Schedule 7 of the GWA 2006. Section 108(4) of that Act provides that an Assembly Act will be within its competence if ‘it relates to one or more of the subjects listed under any of the headings in Part 1 of Schedule 7’ and does not fall within any of the exceptions set out under any of the headings in that Part of the Schedule. (Additional limits on competence, such as those requiring compatibility with EU law and the ‘Convention rights’, were not relevant here.) Section 108(7) of the GWA 2006 states that the meaning of the term ‘relates to’ is to be ‘determined by reference to the purpose of the provision, having regard (among other things) to its effect in all the circumstances.’

The Bill under review

The Agricultural Sector (Wales) Bill was passed in order to provide for a scheme to regulate agricultural wages in Wales following the abolition of the Agricultural Wages Board for England and Wales under the (UK) Enterprise and Regulatory Reform Act 2013. Until its demise, the Agricultural Wages Board set minimum wages for workers employed in agriculture, and other terms and conditions of employment. The Welsh Government wished to retain a system for regulating agricultural wages within Wales and sought to do this in the legislation under review. In general terms, the Bill preserved a statutory regime for workers in the agriculture sector which acknowledged the distinctiveness of this sector, and sought to safeguard a succession of skilled workers, with provisions for apprentices and trainees. It preserved the level of statutory protections in the Agricultural Wages Order of 2012 which, without the provision of this Bill, would have been revoked from October 2013. That Order recognised different categories of worker based on qualifications, competence, experience and levels of responsibility: all of these grades were above the current national minimum wage. The Bill provided for the establishment of an Agricultural Advisory Panel for Wales which would carry out similar but modified functions to those undertaken by the Agricultural Wages Board.

The Attorney General questioned the competence of the Assembly to make this legislation on the basis that it dealt with employment matters and industrial relations rather than agriculture. The Counsel General submitted that the Bill related to agriculture and on that basis came within the Assembly’s legislative competence.

The Court’s ruling

In reaching its decision, a number of matters to which the Attorney General referred the Court were ruled to be irrelevant to the interpretation of Schedule 7. The Court held that a ministerial statement in Parliament regarding the purpose of the GWA 2006, as being to ‘deepen’ rather than to ‘broaden’ devolution, was too general and ambiguous to be of assistance in interpreting the GWA. It also ruled that it would be inappropriate to consider correspondence which took place prior to the introduction of the Government of Wales Bill in 2005 between the Wales Office, the Welsh Government and Parliamentary Counsel: this correspondence was said to set out the views of the two executives on the scope of the subject of ‘agriculture’ and whether it should include specific references to competence in relation to the Agricultural Wages Board. Since this was correspondence which was never referred to in Parliament or made public, the Court held that it would be inconsistent with transparency and the democratic process to take it into account. Finally, the Court held that the fact that a power had not been transferred under the first or second phases of devolution was irrelevant to the position pertaining under the third, and current, phase of devolution for Wales.

As to how the GWA 2006 should be interpreted, the Court referred to the general principles developed in the previous Welsh Byelaws case, namely that:

  1. whether the provision was outside the Assembly’s competence must be determined by the rules laid down in section 108 and Schedule 7;
  2. the GWA 2006 should be interpreted in the same way as any other statute and its description as ‘an Act of great constitutional significance’ could not be taken, in itself, as a guide to its interpretation;
  3. when enacting the GWA 2006, ‘[t]he aim was to achieve a constitutional settlement’ and it was proper to have regard to that purpose in determining the meaning of words.

The Court examined the subjects listed in Schedule 7, noting the enumeration of agriculture as an area of competence and the exceptions to that particular subject (which relate to hunting with dogs, regulation of experiments on animals, import and export controls and regulation of the movement of animals, and authorisations of veterinary medicines and medicinal products.)   Since an exception will be relevant wherever it appears in Schedule 7, the Court examined the other subject headings and the exceptions listed under each of these. The Court noted the areas listed under the heading ‘Economic development’ (which includes economic regeneration and development and promotion of business and competiveness as areas of competence) and the exceptions listed under that heading. In particular, it noted that occupational and personal pension schemes were exceptions to the Assembly’s competence: this exception related to specific aspects of employment but Schedule 7 did not include any general exception in respect of employment or remuneration of employees.

As to the meaning of ‘agriculture’, which is not defined in the GWA 2006, the Court concluded that ‘agriculture’ could not be intended to refer only to ‘the cultivation of the soil or the rearing of livestock’. Rather, it needed to be understood ‘in a broader sense as designating the industry or economic activity of agriculture in all its aspects, including the business and other constituent elements of that industry’. This view was supported by the broad definition that had been given in Schedule 5 to the ‘red meat industry’, the only area of agriculture in which the Assembly had legislative competence prior to 2011.

With agriculture thus defined, the Court had little difficulty in concluding that the Bill was ‘aptly classified’ as relating to agriculture: ‘the purpose and effect of such a regime are to operate on the economic activity of agriculture by promoting and protecting the agricultural industry in Wales.’ However, the Attorney General submitted that the Bill would have an effect on employment and industrial relations, neither of which was listed as a subject on which the Assembly had legislative competence. However, the Court observed that neither were these matters specified anywhere in the Act as exceptions to the Assembly’s competence: as noted earlier, certain aspects of employment are listed as exceptions but in the Court’s view the specifying of these particular aspects suggested that there was no intention to create a more general limitation on the Assembly’s competence.

The Court accepted the Attorney General’s submission that the Bill might be characterised as relating to employment and industrial relations. This made it necessary to consider whether a Bill relating to a listed area of competence might still be regarded as falling outside competence if it also related to an area which was not listed as devolved. The Court considered that this issue would not arise very frequently given the relatively extensive list of exceptions set out in Schedule 7: this case arose because, despite not being devolved, employment and industrial relations were not stated to be exceptions to those areas which were explicitly devolved.

The crux of the Attorney General’s argument was that, in reality, this Bill did not relate to agriculture but to employment and industrial relations and should be characterised in that way. He contended that the Court should determine the ‘real’ purpose and objective effect of the legislation. The Court refused. It accepted that, as in this case, there might be more than one way of characterising the purpose and effect of a Bill: a Bill establishing a scheme for regulating agricultural wages could ‘in principle reasonably be classified either as relating to agriculture or as relating to employment and industrial relations. Which classification is the more apt depends on the purpose for which the classification is being carried out, and on the classifactory scheme which has to be employed.’ In the Court’s view, the rules in section 108 and Schedule 7 had to be interpreted according to the ordinary meaning of the words used: doing so would achieve a ‘coherent, stable and workable outcome’. In most cases, an explicit exception to a devolved subject area would resolve a question about competence. However, when, as here, no exception to the devolved subject was stated, section 108 still provided the test: provided that a Bill ‘fairly and realistically’ satisfied the test set out in section 108(4) and (7) and did not fall within an exception, it came within the Assembly’s competence. It did not matter that it might also be capable of being classified as relating to a subject which had not been devolved, as long as the latter had not been explicitly excepted. To agree to the Attorney General’s submission would be to add exceptions to those specified in the GWA 2006 and would give rise to uncertainty and to scheme that was ‘neither stable nor workable.’ Accordingly, a Bill which undoubtedly related to a devolved subject would be within the Assembly’s competence even if it could also be characterised as a Bill relating to a non-devolved matter which was not explicitly excepted in the GWA 2006.

General comments

The ruling in this case makes a significant clarification in relation to the competence of the Assembly. The Supreme Court sets out a straightforward approach to determining whether there is competence, which is grounded in the terms of the GWA 2006: as long as a Bill ‘fairly and realistically’ relates to a subject which is listed in Schedule 7, this being determined by reference to its effect and purpose, it will be within the Assembly’s competence unless it falls within an exception listed in Schedule 7 or elsewhere in the Act.

While most cases will be determined by the express grants and exceptions in Schedule 7, there will be other cases where there is less certainty. One such case would arise if the Assembly were to enact legislation providing for a general prohibition on smacking children and young people, by removing the defence of ‘reasonable chastisement’: this was a commitment of earlier Welsh Governments but was not included in the Social Services and Well-being (Wales) Act 2014 and when the issue was raised by the Assembly’s Health and Social Care Committee during the passage of the Bill (18 April 2013), the Deputy Minister expressed concerns that if such a prohibition were included in the Bill, there would be a challenge to the legislation from the UK Government regarding the Assembly’s competence ( Under Schedule 7 is it is clear that that the Assembly has competence in relation to protecting and promoting the well-being of children and young people. Criminal law on the other hand is not listed among the devolved subjects. However, section 108(5) provides that an Act will be within the Assembly’s competence if it is to enforce a provision of legislation that is within the Assembly’s competence or is otherwise incidental or consequential on such a provision.) The Welsh Government has stated that it has no plans to legislate on this issue during the current Assembly term: if such legislation is put forward at a later stage, we can expect another reference to the Supreme Court and a further clarification of the Assembly’s competence.

There is one further reference to the Supreme Court in the pipeline, concerning the Recovery of Medical Costs for Asbestos Diseases (Wales) Bill. This Bill allows for the recovery of costs incurred by the NHS in Wales in providing care and treatment to victims of an asbestos-related disease. Interestingly, this Bill has been referred by the Welsh Government’s own law officer, the Counsel General. His statement to the Assembly made clear that he considered the Bill to be within the Assembly’s competence but, aware of the fact that the insurance industry had disputed this throughout the Bill’s passage, wished the matter to be determined by the Supreme Court before its entry into force rather than waiting for what he considered an inevitable challenge afterwards which would be more time-consuming and more expensive. Were the Bill to be found to be outside the Assembly’s competence, this ‘pre-emptive challenge’ avoids the possible need for unpicking transactions made under it. The Counsel General considered it ‘very unlikely’ that such home-grown references would be made very often, although he was reluctant to describe this approach as ‘exceptional’.

While many in Wales consider that a move to a reserved powers model would greatly improve the clarity of the devolution settlement for Wales, some disputes will arise whatever the model. Nonetheless, the particular conferred powers model in Wales with its very specific grants and exceptions, and, as in this case, issues which are not mentioned explicitly as exceptions to devolved subjects, does not help. Accordingly, the recommendation in the Silk Part 2 Report for a reserved powers model was generally welcomed in Wales. However, the current arrangements are likely to be in place for some time still – even if the UK Government were to follow the Silk recommendations, the Silk report does not envisage an Assembly operating under the new system until 2021. In the meantime, and for those disputes which concern ‘borderline’ areas under any model, the clarification provided by the Supreme Court is valuable in improving the workability of the current arrangements.


Ann Sherlock, Centre for Welsh Legal Affairs, Aberystwyth University


(Suggested citation: Ann Sherlock, ‘Supreme Court ruling on Welsh legislation’ U. K. Const. L. Blog (30th July 2014) (available at:


Filed under Devolution, Wales

Alan Trench: Welsh byelaws and the UK Supreme Court

At the end of July, we learned that the UK Attorney-General has referred the first Act of the National Assembly passed since the March 2011 referendum to the UK Supreme Court. There’s news coverage of this from the BBC here, and a good analysis from Toby Mason of BBC Wales here.  This is the first time that any devolved legislation has been referred to the Supreme Court before receiving royal assent; it is the first time the UK Government’s law officers have challenged the legal competence of devolved legislation; and it is the first time that legal challenges involving Welsh legislation have been brought.  Any one of those would make it a noteworthy event indeed.  To find all three of them in one place – coupled with a significant point of law – makes it a case of rare interest.  The latest information is that there will be a hearing in early October.

This post will discuss what the case is about, how it comes before the court and what the constitutional political issues at stake are, as well as the black-letter legal ones.  It is a case of interest that goes far beyond Wales, because although there are significant differences between the Welsh arrangements and those for Scotland or Northern Ireland, the case also raises some rather broader questions about the legal working of devolution.

This is not the first time Wales has broken new ground in challenging how devolution works, of course.  Most notably, Wales was the first jurisdiction to refuse consent to Westminster legislation affecting a devolved function under the Sewel convention – aspects of the Police Reform and Social Responsibility Act 2011.  Perhaps the more provisional, evolutionary nature of the Welsh arrangements mean that it is more prone to test the legal aspects of its devolution arrangements than Scotland or Northern Ireland, where they are more clearly established.

The legal background 

The bill that the Attorney General is challenging is the Local Government Byelaws (Wales) bill, which was passed by the National Assembly on 3 July 2012.  The Byelaws bill is a pretty dull piece of legislation.  Its main effect is to enable Welsh local authorities to make byelaws under a number of specified powers for places like public conveniences, sea-side promenades or private swimming pools, without requiring that these be ‘confirmed’ by an approving authority.  Up to now, and still in England, byelaws have to be made under specific statutory powers, with procedural requirements including local advertising.  They must then be approved by a confirming authority, historically the Home Secretary, before they can come into effect.  To secure that approval, it is usually necessary to use established model byelaws, or else to make a strong argument why something different is needed and why it is appropriate in a particular case.  Welsh byelaws would still need to be limited to specified purposes, be consulted on, and advertised in the local area; but many of them would not need approval by another level of government.

In the case of the Byelaw bill, the power to make byelaws under the various statutes involved – the Public Health Amendment Act 1907, the Public Health Act 1936, the Highways Act 1980 and more – was passed to the National Assembly by the ‘jumbo’ transfer of functions order in 1999.  These powers were passed unconditionally, and by virtue of paragraph 30 of Schedule 11 to the 2006 Act have now passed to the Welsh Government.  However, the power for local authorities to make byelaws, which is set out in section 236 of the Local Government Act 1972, provides that confirmation of byelaws is a matter for concurrent action by both to be the Assembly (meaning now the Welsh Ministers) and the Secretary of State.  Thus, to the extent that the Byelaws bill alters the way section 236 LGA 1972 works, it affects a function of the Secretary of State.

Schedule 7 to the Government of Wales Act 2006 provides that devolved legislation can only affect pre-commencement functions of UK ministers like the section 236 bye-law confirming power if it does so only incidentally or consequentially to its main purpose, or if the Secretary of State gives her consent.  Consent has not been given in this case.  In the Attorney General’s view, clauses 6 and 9 of the Byelaws bill therefore transgress on UK ministers’ powers, so are beyond the powers of the National Assembly.

Referring a bill to the Supreme Court

The Attorney General has never before used his power to refer a devolved bill to the UK Supreme Court, though he has a similar one under each of the devolution settlements (under section 33 of the Scotland Act 1998 and section 11 of the Northern Ireland Act 1998, as amended by sections 27-8 of and Schedule 7 to the Justice (Northern Ireland) Act 2002) .  This is therefore an unprecedented step, and takes us into unknown territory.  When issues have arisen with devolved legislation in the past, they have been resolved quietly behind the scenes – perhaps by amending the legislation, perhaps by adjusting the devolution settlement to ensure the bill is within competence.  As all devolved legislation needs a declaration whether it is within the legislative powers of the parliament involved from both the minister or other member responsible for the bill, and the presiding officer, it is highly unusual for a bill to reach this stage with any major question about its legal competence.

In this case, both the presiding officer and the Minister for Local Government and Communities stated that they considered the bill to be within the Assembly’s legislative competence.  The explanatory memorandum makes no reference to the issue about confirming powers at all, and indeed implies that confirmation was simply a matter for Welsh Ministers, with no obligation for them to act in conjunction with UK ones at all.

A reference means that the case goes straight to the Supreme Court, without any earlier hearing in the High Court or Court of Appeal.  In the Welsh case, the procedure is set out in Schedule 9 of the 2006 Act.  The procedure bears some resemblances to the procedure for an advisory opinion of the European Court of Justice, which is quite a novelty for lawyers from the UK, and has never been used before.  But the lawyers will need to work out what to do quite quickly, as a hearing is apparently due to start on 9 October.

Functions of Ministers of the Crown

It has been increasingly clear for some time that the protection of the position of Ministers of the Crown was likely to be a serious brake on the exercise of the National Assembly’s legislative powers under Part 4 of the 2006 Act.  Many executive functions were transferred to the National Assembly in 1999 (by what is known as the ‘jumbo’ transfer of functions order), and a few others have been since.  Those functions were then transferred to the Welsh Ministers when the 2006 Act came into effect (by paragraph 30 of Schedule 11 to the 2006 Act).  But there remain a good many residual UK functions, reflecting the history of administrative entanglement between England and Wales – and the defensiveness of some Whitehall departments about devolution.  Those arguments were re-fought when the 2006 Act was being framed, and one key safeguard was the block on legislation affecting functions of ‘ministers of the Crown’.

One of many problems with that safeguard is that no-one knows quite how far it reaches.   There is no list of functions of ministers of the Crown, and producing one would be a huge task that would inevitably produce mistakes.  But that history of administrative entanglement goes back a long way (remember: the Welsh Office was only created in 1964), and has a powerful effect.  And UK Departments have been keen to maintain control over important functions like planning for big energy projects or defence installations, as well as safeguarding water supplies, business regulation and the common law of England and Wales.  That has usually been done by express provision, rather than by relying on a general saving clause as here.

It is worth noting that the position for Wales is rather different from that for Scotland.  There, following an opinion of the Attorney General, the UK Government has accepted that it is within the power of the Scottish Parliament to legislate for matters that affect UK ministers, provided the legislation is within Holyrood’s legislative competence – i.e., that the legislation does not ‘relate to’ a reserved matter.  There is no general saving for UK ministerial functions, though there are rather fewer of them relating to devolved functions in Scotland.  This is documented in Devolution Guidance Note 15, which also emphasises the need for consultation before Holyrood legislates in ways that affect UK ministers.

It is also worth noting that a ‘reserved powers’ model of conferring legislative powers on the National Assembly would not solve this problem.  Nor would creating a separate Welsh legal jurisdiction (though that might have an effect on it).  The problem arises because of the relationship between UK executive functions and devolved legislative powers, not the scope of those powers.

The politics of this case

In some circumstances, constitutional litigation can happen more or less by accident.  Even years after legislation is passed, a private litigant can raise the question of whether the Act was in fact legally competent.  That does not appear to be the case here.  I understand that the Wales Office tried repeatedly to draw the Welsh Government’s attention to the problem, and to suggest what changes they needed for the Secretary of State to be able to give her consent (which she was willing in principle to do).  The Welsh Government failed to respond, in circumstances that suggest this cannot have been accident or oversight.  But legally speaking the Welsh Government’s position here is weak, if not hopeless; the legislation affects a function of a minister of the Crown, consent has not been given, therefore the legislation is beyond the Assembly’s powers.  The Welsh Government seems to be walking into a judgment that what would have been the Assembly’s first Act is void.   The only hope for its argument might be if it could show that the formal protection of the UK Minister’s confirming power had in fact fallen into disuse, but even if this can be proven it is hard to see how it helps given the clear position under statute.

Equally, this analysis means that the UK Attorney General had little choice but to refer the bill to the Supreme Court.  The legal position is quite clear.  If he failed to do so, he would have been in dereliction of his duty; he would, effectively, be conniving at the National Assembly passing legislation it had no power to pass.  His legal role means that not only does he have the power to make such a reference, but in these circumstances he has to do so.  Even if he did not, it would not solve the problem – the legal flaw in the legislation would remain, and (for example) any person aggrieved at his or her conviction under a bye-law made under the new system would be able to challenge their conviction.  In that sense, the power the Attorney General has (and his exercise of it) are ways of bringing matters to a head sooner rather than later, and his use of it is an appropriate response to pretty aggressive tactics from the Welsh Government.  The only surprise here is that the Attorney General waited until the last day he had to bring this challenge (there is a 28-day limit), rather than doing so promptly after the bill was passed.

Why might the Welsh Government have drafted and promoted legislation that was beyond the Assembly’s law-making power, and handled matters so as to invite such a legal challenge?  That is a question best directed at the First Minister or Counsel General, of course.  Two reasons suggest themselves why they might do so, though.

The first is that they think there is just a chance that the Supreme Court will find in their favour, and want to see how much leeway or support they can get from the court.  In the Axa case, Lord Hope – inaccurately, in my view – said obiter that the ‘essential nature of the legislatures created’ for Wales, Scotland and Northern Ireland were the same in each case (at paragraph 43).  That may have given them an (unjustified) degree of optimism in what the Court might say, given that these constraints do not apply under the Scotland or Northern Ireland settlements.  It would certainly look like an avenue worth exploring, even if the chance of success is limited.

The second is that they know they are going to lose, and want to use the Court’s judgment to underline to the wider public just how limited the new legislative powers of the National Assembly are.  There could be no more powerful sign of that than that the first, innocuous Act of the Assembly is in fact blocked.  After all, there are more contentious bills in the pipeline – notably the Human Tissue bill.  That in turn may serve a wider political purpose, of making it hard politically for the Secretary of State to refuse consent to legislation in future.  It may even secure some guidelines from the Court about how the Secretary of State should approach the giving or withholding of consent, rather than simply leave it to the Secretary of State’s untrammelled discretion.

In that context, it is worth looking at the forebears of the provision that legislation affecting ministers of the Crown needs consent.  Two spring readily to mind.  The Northern Ireland Act 1998 creates three sorts of legislative powers: ‘excepted’ and ‘reserved’ matters, which are beyond the Northern Ireland Assembly’s competence, and ‘transferred’ ones, which are the rest.  Reserved matters can be devolved by a relatively simple order, so they are candidates for future devolution (though that has not happened in practice), and they may also be the subject of devolved legislation with the Secretary of State’s consent.  In the early years of the Assembly, this power was used relatively often, mainly when criminal offences were being created in connection with devolved functions (something not needed under Part 4 of the 2006 Act; see section 108(5)).  With the devolution of justice and policing, the need for that seems to have diminished considerably.

The other model is the more general inspiration for the 2006 Act – the Scotland Act 1978.  That never came into effect, of course, but it provided for a ‘defined powers’ model of legislative devolution.  It also gave extensive powers to the Secretary of State for Scotland (part of the UK Government, not the devolved administration) to approve legislation passed by the Scottish Assembly before that legislation came into effect.  That would have turned the Secretary of State into a Viceroy in all but name.  Such powers would have been very problematic – had the 1978 Act ever come into effect, it would either have stopped devolution working, if used, or otherwise fallen into disuse.  The danger in Wales is that the power of the Secretary of State to consent to legislation affecting ministers of the Crown is sufficiently broad that, although in form it safeguards the legal functions of UK minsters, in practice it operates as a discretionary veto on what the National Assembly may do.

Conclusion: a bad place to be

The fundamental issue here is that the March 2011 referendum creates a strong public expectation that the National Assembly will have power to legislate for the twenty devolved ‘subject areas’ set out in Schedule 7.  Formal niceties about protecting UK ministers’ functions were no part of that.  Undermining the effective working of that model of devolution will risk de-legitimising that model of devolution, and raise questions about what should happen instead.  Given the strength of public support for devolution, that is a dangerous course for all involved.  Even if the Supreme Court’s decision about the Byelaws will does no more than confirm that the Act is beyond the Assembly’s powers for want of the Secretary of State’s consent, it will both increase political tensions between Cardiff and London, and make a significant step forward in relying on the judicial resolution of difficult points in the devolution arrangements.  There are serious reasons to doubt whether the constitutional framework of devolution as it presently operates can withstand that.

Alan Trench is honorary fellow at the University of Edinburgh, honorary senior research fellow at the Constitution Unit, University College London, and author of the blog Devolution Matters.

Suggested citation: A. Trench, ‘Welsh byelaws and the UK Supreme Court’ UK Const. L. Blog (17 September 2012)(available at available at


Update (4th October 2012): 

Since I wrote the post above, the Welsh Government has released a collection of papers and correspondence regarding the Byelaws bill and the court challenge.  They go back to September 2011 and show that the Welsh Government considers that the National Assembly has power to remove functions of ministers of the crown through its power to alter such functions when that is incidental to and or consequential on other, intra vires, provisions in Acts of the Assembly.  We will see what the Supreme Court makes of that line of reasoning.  As the removal was central rather than incidental to the statutory scheme in the bill, it is hard to see how the argument is likely to succeed.

The papers released also show that, although the Welsh Government was well aware of the UK Government’s view that the Assembly had no powers to legislate without the Secretary of State’s consent, it did not seek to make the Assembly aware of that problem.  Nor did it ask for consent from the Secretary of State, even when that was offered subject to limited amendments to the bill.

The documents released can be found on the Welsh Government’s website here, and I’ve written a more detailed post about it on Devolution Matters, here.


Filed under Devolution, Wales

Tarun Khaitan: How to interpret Constitutional Statutes?

Most of us will be aware of the famous remarks of Lord Justice Laws in Thoburn v Sunderland City Council (2002) that constitutional statutes are immune from implied repeal, and therefore somewhat entrenched against Parliament. The issue of the entrenchment of the UK constitution against non-sovereign legislatures, such as the devolved legislatures and the European Union, has received relatively less attention.

In this post, I will highlight a purpose other than that of entrenchment for which certain statutes are being characterised as ‘constitutional’. In the following cases, the devolution statutes, namely the Scotland Act 1998, the Northern Ireland Act 1998, the Government of Wales Acts 1998 and 2996, have been so characterised in order to justify the adoption of special interpretive approaches towards these statutes.

At least two broad, and apparently inconsistent, interpretive rules can be seen to be emerging. On the one hand, we have cases which suggest that constitutional statutes, like canonical constitutional codes in other jurisdictions, should be interpreted in a generous and purposive manner. On the other hand, some judges have held that constitutional statutes require literal interpretation, with especial fidelity to the text. Their argument is that Parliament has chosen a precise set of words while being fully cognisant of the constitutional importance of the Bill under consideration. As such, they call for strong judicial deference.

The most famous case adopting the first position is the judgment of the House of Lords in Robinson v Secretary of State for Northern Ireland (2002). The case concerned the validity of the election of the First Minister and his Deputy by the Northern Ireland Assembly two days after a six-week deadline prescribed by the Northern Ireland Act 1998 for such election. A majority in the House of Lords upheld the election as valid. Lord Bingham, speaking for the majority, held that:

The 1998 Act does not set out all the constitutional provisions applicable to Northern Ireland, but it is in effect a constitution. So to categorise the Acts is not to relieve the courts of their duty to interpret the constitutional provisions in issue. But the provisions should, consistently with the language used, be interpreted generously and purposively, bearing in mind the values which the constitutional provisions are intended to embody. [11]

He suggested that the purposes of the Northern Ireland constitution included ensuring ‘that there be no governmental vacuum’, attempting ‘to end decades of bloodshed’ and facilitating ‘participation by the unionist and the nationalist communities in shared political institutions … [which] had to have time to operate and take root.’[10] The rationale for the six-week deadline, Lord Hoffmann explained in his concurring opinion, had ‘been to induce a willingness to compromise on the part of the members of the Assembly by the prospect of having to fight a new election.’[28] Giving the requirement of the deadline a rigid interpretation to invalidate the election held after it had passed, he argued, would be contrary to the most fundamental purpose of the Belfast Agreement which the 1998 Act was clearly seeking to implement: ‘namely to create the most favourable constitutional environment for cross-community government.’[30]

Lord Hoffmann’s reliance on the Belfast Agreement is particularly interesting. He justified this reliance by suggesting that the Agreement, along with the political context surrounding it, formed ‘part of the admissible background for the construction of the Act just as much as the Revolution, the Convention and the Federalist Papers are the background to construing the Constitution of the United States.’[33] In doing so, he borrowed from the interpretive traditions usually applied in the context of short, general and vague constitutional texts, citing the paradigm example of constitutionalism of this variety, namely the United States.

In Imperial Tobacco Limited (2010), Lord Bracadale expressly followed this ‘purposive and generous’ approach in Robinson to hold that ‘The court should endeavour to find in the Scotland Act a constitutional settlement which is coherent, stable and workable.’[3]

Similarly, the High Court in R (Governors of Brynmawr Foundation School) v The Welsh Ministers (2011) also cited Robinson to hold that the Government of Wales Acts (1998 and 2006) were constitutional statutes, and therefore ‘in applying the rules of statutory construction in order to determine the scope of the powers conferred on the Welsh Ministers or the Assembly by GOWA 2006, the court will take into account its constitutional status.’[73] Mr Justice Beatson adopted a generous approach and held that ‘Given the constitutional status of GOWA 2006, the court is reluctant to read implied limitations into it by reference to legislation which is not of a “constitutional” nature.’[87]

Apparently endorsing this approach, the Supreme Court said recently in AXA General Insurance v The Lord Advocate (2011) that ‘The carefully chosen language in which [certain provisions of the Scotland Act] are expressed is not as important as the general message that the words convey. The Scottish Parliament takes its place under our constitutional arrangements as a self-standing democratically elected legislature.’ [46]

On the other hand, there are cases which, while they agree that a special interpretive approach is warranted for constitutional settlements contained in devolution statutes, adopt an interpretive approach that is quite the opposite of the one just described. Thus, in Mills v HM Advocate (No 2) (2001) the High Court of Justiciary said that

‘There is also, in our view, force in the argument that the particular and detailed provisions dealing with devolution issues are part of the constitutional settlement embodied in the Scotland Act and that requirement should not therefore be avoided or circumvented. If the effect of the provisions is that appeals are open to the Privy Council on matters involving questions of Scots criminal law, that, in our view, must simply be accepted. It does not provide any reason to reject the argument based on the plain terms of the legislation.’ [19]

Soon after Mills, the Privy Council held in “R” v HM Advocate [2002] that ‘The Scotland Act is a major constitutional measure which altered the government of the United Kingdom’.  In this case, Lord Rodger suggested that when Parliament had consciously enacted ‘a constitutional settlement of immense social and political significance’, courts must be especially deferential: they ‘must loyally give effect to the decision of Parliament on this sensitive matter, even if – or perhaps especially if – there are attractions in a different solution’.[155]

Unlike the first set of cases, these two cases suggest that the proper way to interpret constitutional statutes is to do so literally rather than purposively. Indeed, they demand a literal application of even the mundane or ordinary provisions contained in constitutional statutes (after all, not all provisions in a constitutional statute are ‘constitutional’). The rationale seems to be that Parliament has in its wisdom settled these mundane details while being sensitive to the constitutional nature of the overall project. This context implies that the importance of the overall project rubs off to some extent on all provisions contained in a constitutional statute. The room for judicial manoeuvre is small, and a literal interpretation that is warranted. The second set of cases seems to better recognise that UK style constitutional statutes (at least those containing the devolution settlements), although ‘constitutional’, remain statutes. They are drafted differently from canonical constitutional codes, and tend to be very detailed, delving into the minutiae of governmental functioning.

One may think that these two interpretive approaches can be reconciled with each other, inasmuch as they (one may argue), apply to different types of constitutional provisions. On this argument, one could say, that a generous and purposive interpretive approach is appropriate for provisions which are framed in a general and vague language, or which embody broad legal principles normally found in preambles to constitutions and Bills of Rights. On the other hand, a literal approach is best for those constitutional provisions which embody a detailed rule where the scope for indeterminacy is minimal. Such a distinction is surely plausible, except that it cannot be supported on the facts of the cases described above. The provision being interpreted in Robinson was fairly clear, specifying a fixed time period within which the elections of the Ministers was to take place. Indeed, most of the aforementioned cases dealt with relatively precise and detailed provisions in constitutional statutes. Which of these two approaches ultimately finds favour with the courts remains to be seen.

Tarunabh Khaitan is a Fellow in Law, Christ Church, Oxford.


Filed under Devolution, Judicial review, Northern Ireland, Scotland, Wales

Christine Bell: Bills of Rights and Devolution: From the Universal to the Particular.

‘To produce one Bill of rights may be regarded as a misfortune.  To produce eight, looks like carelessness’.

This blog picks up on Nicholas Barber’s blog of September 11, 2011.  There he sketched the complicated options for taking a human rights case with respect to the Human Rights Act, the European Convention on Human Rights and the European Court of Justice.  He opined: whilst one bill of rights shows caution, three suggests panic.’

In this blog I examine the complications of the similar ‘disorder of legal orders’ (to use Neil Walker’s great phrase), emerging on the domestic rights front.  A stochastic set of Bill of Rights initiatives and Human Rights Act devolution dilemmas are driving multiple processes of reform in different directions, towards a wonderfully chaotic panoply of unintended consequences.

From three to eight?

 Within the UK three Bills of Rights processes are in train, in addition to Nicholas Barber’s three.

4. A Northern Ireland Bill of Rights.  The Belfast / Good Friday Agreement provided for a bill of rights in Westminster legislation providing for ‘rights supplementary to those in the European Convention on Human Rights, to reflect the particular circumstances of Northern Ireland, drawing as appropriate on international instruments and experience.’ (See further, advice given by Northern Ireland Human Rights Commission and post by Colin Harvey earlier this year).

5. An all-Ireland Charter of Rights.  This was to be considered in a joint initiative of the Northern Irish and Republic of Ireland Human Rights Commissions as also required by the Belfast/Good Friday Agreement, and underwritten by a British-Irish Treaty, although little progress has been made.

6. A UK Bill of Rights.  As provided for by the coalition  government, a Commission has been tasked with investigating a UK Bill of Rights with terms of reference that are remarkably similar to the NI Bill of Rights process namely, to fashion ‘add-ons’ to the ECHR to form a UK Bill of Rights. (And without complicating things too much, even should this come to nothing AXA General Insurance Limited and others v The Lord Advocate and others (Scotland), [2011] UKSC 46 offers the prospect of building ‘rule of law’ judicial review which, in theory, could begin re-incorporating a range of rights were the HRA dismantled see Adam Tomkins’ blog on the case.)

7. Scottish Rights for Scotland. Regarding a seventh rights framework, we could further add the outcome of a process, one suspects still to run, in Scotland, regarding when and how rights frameworks should apply in this devolved context.  The Scottish Human Rights Commission talk of a ‘National Human Rights Action Plan’ rather than a bill of rights (and opposes a UK Bill of Rights, preferring the Human Rights Act).  Meanwhile, the Scottish government has focused serious and sustained attention on how to limit the UK Supreme Court’s HRA review of criminal cases – this jurisdiction having resulted as anomalous  by-product of the shift of Privy Council ultra vires jurisdiction to the UK Supreme Court (see Walker Report 2010, Advocate General Expert Group Report 2010, McCluskey Report 2011, and Scotland Bill).  At present proposals are limited to a new appeals mechanism involving a reference procedure for cases involving convention rights, but underlying this reform lurks continuing pressure for a distinctive Scottish human rights regime policed by Scottish courts.  Such impetus is likely to be galvanized if – as they must – UK Bill of Rights proposals require amendment of the Scotland Act 1998 (where the definition of ultra vires incorporates the Human Rights Act).  SNP rhetoric with reference to the UK Supreme Court decision of Cadder v Her Majesty’s Advocate [2010] UKSC 43 appears to lean towards mediating and limiting the direct application of European Court of Human Rights’ jurisprudence in Scotland.  However, SNP’s independence agenda suggests a more extensive rights regime to perhaps include socio-economic rights.  Scotland’s devolved governments have repeatedly taken a different path to England and Wales with regard to socio-economic issues and the recently produced independence document of the SNP talks of new constitutional rights protection for Scotland and provides socio-economic commitments with a ‘rights’ flavour, all of which have relevance to ‘independence lite/devo-max’ outcomes as well as fully blown independence.

8. Welsh rights for Wales? For number eight we must turn to Wales where it is easy to imagine that with the Welsh Parliament which has recently increased its powers, moving towards its own bill of rights (see Liberal Democrat community bill of rights proposal), or national action plan for Wales.  The equality duty in the Equality Act 2010 already has a distinctive Welsh dimension and indicates that devolved diversity in rights is already with us but where equality is concerned has managed to come in under the radar (Northern Ireland of course also having an already-differentiated equality duty, and conditioning the powers of the NI Assembly not just on the HRA but on equality).

Although I too paraphrase Oscar Wilde I should make it clear that I do not view bills of rights negatively.  However, I would view as misfortune a single UK bill of rights, which watered down the Human Rights Act and negated ten years of Bill of Rights debate and a sensible proposal in Northern Ireland.  And the UK Bill of Rights process does seem somewhat careless and unfortunate from whatever political point of view one comes from.  Seized on by the Conservative Party apparently to limit judicially protected rights, to ‘get out’ of unpopular ECHR and HRA rulings,  and perhaps even strengthen the (UK) union, the terms of reference they cobbled together in coalition contemplate extending EHCR rights and show little thought to devolution, much less any evaluation of whether a bill of rights process will be a force for unity or galvanize (already existing) commitments to devolved rights regimes.  Were I an anti-HRA/ECHR conservative I would view such this result as very careless.

Plus, it all seems a quite careless anyway.  The Commission is to consult on ‘which rights’, but not on process or enforcement; has a mandate and membership which ‘forgets’ devolution, does not mention or draw on expertise of identical devolved initiatives, and then tacks on a few devolved ‘advisors’ using a mysterious appointments process; apparently uses ‘polarized views on Bills of Rights’ as its key membership criteria; is given an almost non-existent website (buried as a section of the ‘Justice’ department); establishes a once-off time-limited written consultation process with no education dimension and little public purchase (despite lack of public ‘ownership’ being a key rationale for moving from the Human Rights Act, see  Grieve); and manifests its divisions on its first, less controversial, task of advising the government on reforms to the European Court of Human Rights, where the Commission’s official advice was supplemented/counteracted by a letter to Ministers from the Chair, both of which were then unilaterally denounced in the press by one of the members (see  UK Human Rights Blog).  Whew.

Lost in this Bill of Rights mess, it is easy to miss the quite fundamental underlying negotiations that beget and sustain the mess.  Two key negotiations in particular are worth noting.

Plural rights for constitutional pluralism, or rights hierarchy for a unitary state?

One submerged negotiation is over a unitary or pluralist conception of the UK current constitutional arrangements.  For better or worse, the UK seems to be tracked into a process of incremental constitutional reform that includes on-going negotiation of devolution, and on-going negotiation of rights frameworks and the balance between legislature and judges at both national and devolved levels.  It is unreaslistic to expect such a process to deliver order, a common rights framework, and a neat rights hierarchy from devolved regions and courts, to the ‘UK’ and the UK Supreme Court, all embraced by a ‘we the people’ consensus. The current ‘Bills of Rights’ debates are  a by-product of incremental piece-meal reform, but also expose the limitations of such a process.  It is difficult to draft ‘we the people’ documents and broker foundational political compromises as to centre v. periphery, legislatures v judges, through processes that pretend to be minor tinkering at the edges.  We are indeed a big and complex society or even constellation of societies, and we need a ‘big conversation’.

Navigating the Universal and the Particular

The second negotiation is one over where and how universal human rights are best fashioned and applied.  All efforts to implement international human rights standards into domestic law and practice involve an attempt to provide for a particular application of universal rights.  The general wording of international human rights standards and the jurisprudence of their implementation bodies all leave considerable room for translating the universal into the particular.  In the translation, however, arguments arise as to what constitutes ‘translation’ and what constitutes re-writing.

Underlying the debate as to where and how to protect rights as between European, UK-wide and devolved fora, is a political debate as to who it is will really uphold ‘universal values’ in a locally appropriate way, and who has an agenda to ‘particularize’ rights to their own imperialist/ partisan political image.  Thus conservative politicians charge the ECHR as foreign and imperialist in terms of a European paradigm, and assert a UK Bill of Rights as a legitimate ‘more appropriate’ alternative.  However, proponents of the HRA suspect a conservative particularist/imperialistproject at play in the UK Bill of Rights debate, and an attempt at ‘watering down’ or even jettisoning the ECHR for narrow political gain.

Meanwhile, the regions (and others) also charge the process as an imperialist conservative party and unionist project from an internal perspective; while attempts to construct differentiated regional rights provision are counter-alleged by the centre to be ‘too locally driven’, not ‘proper rights’, or not rights which can be delivered to devolved regions alone (see Labour government response to Northern Ireland Bill of Rights Advice).  Behind these counter-charges lies the suggestion that devolved rights agendas also hide imperialist ambitions, this time of devolution nationalists.

Reframing the issue

So what do I say to this?  First of all, I point it out to clarify debates.  But to go further I suggest that the following are useful to bear in mind.

1. Attempted imposition of ‘neat’ rights hierarchies will not make the difficulties of ‘what rights,  where?’ go away.  What is is, and it would be a mistake to assume that a constitutional default position exists.  Let’s embrace the negotiation which is an important one over where and how best to protect rights, and what the relationship is and should be between all the UK’s courts and legislatures.  If this is the debate let us be honest and explicit about it rather than pretending it can be ‘won’ by setting up or running a circumscribed process.

2.     We can never assume or take for granted who best protects  and develops universal human rights.  There is a tendency to think and argue that universality lies in a similar hierarchy to courts of appeal and legislatures themselves, and that we move from more particularized to more universal as we move from local, to national, to regional, to international legislatures and courts.  However, international law makers, courts and tribunals can have their own agendas just as surely as national ones.  While it is true that the expression of rights is likely to move from the more specific to the more generalized and abstract as we move from local to national to international, this does not necessarily equate to a movement from the particular to the universal.

3.     Judges best win jurisdiction over issues by the ways in which they reason from the universal to the particular, rather than by asserting a place in a legal hierarchy.  If local courts are to take themselves seriously and clamour for jurisdiction, they must show themselves capable of engaging with and reaching to universal values when they decide particular cases.  There is a difference between local application of universal values, and ignoring those values or seeking to attenuate them.  Similarly, international judges have most power to affect the implementation of rights where they take local context seriously.  The universal and particular are linked – and a court’s legitimacy in the arena of rights often depends on how it frames its decision in terms of navigating between the two.

4.     Most esoterically, but perhaps most importantly, the difficulty of navigating the universal and the particular is always with us.  Translating the universal to the particular, and using the particular to build, apply, and even establish the universal is an eternal dialectic.  Constant challenge and negotiation is less a problem, and more an important part of how we actually negotiate and come to understand values as universal in the first place.

5.     Therefore we should be suspicious of a priori claims to the Universal as linked automatically to the ‘level’ of the rights initiative, and seek to understand how challenge and counter-challenge work.  That is the end to which this blog has been written, and so I close.

Christine Bell is Professor of Constitutional Law at the University of Edinburgh.


Filed under Constitutional reform, Devolution, Human rights, Northern Ireland, Scotland, UK Parliament, Wales