Jeff King: What Next? Legislative Authority for Triggering Article 50

ucl-profile-photoThe Supreme Court judgment in Miller v Secretary of State for Exiting the European Union on the morning of 24 January 2017 made it clear that an Act of Parliament is required for a notice under article 50(2) of the Treaty of the European Union. The Government and Opposition should now state their positions clearly on the form of such an Act without delay.  So far, there has been little such discussion. This post (originally published on 8 November 2016 and reposted here for convenience) suggests form, content and conditions for such legislation that neither challenge the result of the 23 June 2016 referendum nor the Government’s stated timelines for giving notice.

The Form of Legislation

Bringing forward the Great Repeal Bill and trying to pass it before 31 March 2017 would be an awful idea, because it would embroil the issue of notice in a much greater challenge. As Sionaidh Douglas-Scott has clarified, its task is enormously complex and unresolved, and the use of Henry VIII powers (granting executive authority to repeal primary legislation) and strains on the devolution settlement might even provoke a crisis. What is rather needed is a bill limited to granting the authority to give notice and, as I argue here, framework conditions primarily concerning Parliament’s role in the negotiation period.

To achieve this goal, something such as the (hypothetical) EU Withdrawal (Exit Negotiations) Bill containing a simple clause will start the parliamentary discussion.   In such a bill, section 1 might read ‘Notwithstanding the provision of any Act to the contrary, the Prime Minister shall have the authority to issue notice of the United Kingdom’s intent to withdraw from the European Union under article 50(2) of the Treaty of the European Union.’

An Act containing only that section would be sufficient to grant authority to give notice without undermining the ECA 1972, whose ongoing force will be required in the negotiation period. That would be the minimalist option. Yet this bill would be the occasion for Parliament to secure a right to exercise genuine input into the manner in which negotiations for withdrawal shall take place. That would be the conditions option. The imposing of any such conditions has been derided by some commentators as an exercise in scuppering Brexit, and hence the will of the people. Whether or not this is the intention of some MPs, it is clearly not the intention of the vast majority as repeatedly made clear in official statements. Further, it is not only possible but eminently reasonable to accept the referendum result and insist on a role for Parliament during negotiations. The referendum decided whether to withdraw, not how or on what conditions. As I make clear below, no serious person – and above all not the Minister for Exiting the EU (Mr. David Davis MP) – doubts that Parliament should be involved. So, is insisting on it legally a mistake? Is the actual extent of parliamentary involvement somehow inappropriate for parliamentary determination? Clearly not. Expecting the proper scheme of consultation to follow from executive grace is pious at best, foolish at worst.

What Kinds of Conditions?

The crucial question now is, therefore, what conditions can Parliament legitimately impose without denying the significance of the referendum result? Simply blocking the bill at second reading would in my view be illegitimate. Indeed, for what it’s worth, my own opinion is that having held a referendum on the question, there is a political duty upon Government and Parliament to commence exit negotiations in good faith. However, that leaves many possibilities open.

Consistently with that view, four principles ought to be respected in any parliamentary imposition of conditions. First, Parliament must be given clear rights to notice, to comment on key negotiating positions and draft agreement text, and to a response from the Government to its comments. Second, the devolved governments and legislatures should enjoy formal participation in the consultation process in rough parity with the Westminster Parliament. Third, the Government’s stated notice deadline of 31 March 2017 should be respected (assuming it is not itself to blame for further delay through bullish behaviour). Fourth, there should be no attempt to load the bill with a variety of veto points that would have the effect of destroying the possibility of Brexit during negotiations on the exit agreement. If Parliament really wants a U-turn, it must do so openly rather than insidiously.

These four principles are observed in the discussion that follows.

Parliamentary Oversight of Treaty Negotiations

As a preliminary matter, it must be acknowledged that in most constitutional democracies, the executive branch conducts treaty negotiations largely free from parliamentary oversight. Yet the reason for this is that most constitutional democracies provide for legislative ratification or approval of the final agreement. Since the Constitutional Reform and Governance Act 2010, the UK also has a procedure that legally allows the House of Commons to vote down the possibility of treaty ratification by restarting ad infinitum a 21 day consultation period. This form of final control has traditionally been considered an adequate constitutional safeguard for treaty-making.

This traditional position and its rationale are not adequate for the present situation for one crucial reason: the effects of non-ratification are radically different than the usual scenario, where the effect of denying ratification is that the status quo remains the same. This explains the common mantra that ratification of treaties by the Crown produces no domestic effect without an Act of Parliament. In the scenario under contemplation, a failure to ratify leads to exit without any agreement and hence to a potential economic disaster. More importantly, the legal integration of both orders is so comprehensive that in reality the current situation is more akin to a secession or dissolution of a state than it is to an ordinary treaty ratification or withdrawal. It is entirely new territory, and there is no argument from custom that should discourage Parliament from exercising its undoubted legal power to impose conditions on the treaty negotiation process.

Substantive Conditions

The precise nature of the conditions that parliamentarians might see fit to impose are a matter for greater discussion than can be afforded here. But I will nevertheless sketch out a survey of some of the options that ought to be contemplated (even if rejected).

  1.      A Right to Notice, Comment and Response

EU law provides a clear example of legally secured rights of consultation of the European Parliament vis-a-vis the European Commission in treaty negotiations. These are detailed in the Report of the EU Select Committee, Brexit: Parliamentary Scrutiny (4th Report of Session 2016-17, HL Paper 50), and at length in the remarkably good and comprehensive written evidence provided by Prof Derek Wyatt QC. This procedure provides that the European Commission must:

  • present draft negotiating guidelines to the European Parliament if it does so for the European Council;
  • take due account of the Parliament’s comments throughout the negotiations;
  • promptly and regularly inform Parliament and explain whether and how its comments were incorporated into the text ‘and if not why’;
  • allow Parliament, for agreements that require its consent, to see ‘draft amendments to adopted negotiating directives, draft negotiating texts, agreed articles, agreed date for initialing the agreement and text of the agreement to be initialed.’; and
  • establish appropriate procedures for forwarding confidential information.

These procedures amount at their core to a right to notice, comment, and a Government response to the comment. The EU Committee concluded that this level of scrutiny ‘provides a baseline against any arrangements agreed in the United Kingdom Parliament must be measured; it would be unacceptable for the European Parliament to have greater rights of scrutiny over the negotiations on Brexit than Westminster.’ ([58]). The Brexit Secretary, Mr. David Davis MP claimed ‘We will certainly match, and, hopefully, improve on what the European Parliament sees.’ [47].

Notably, however, the EU Committee rejects the option of a ‘scrutiny reserve’ policy, under which a matter under committee scrutiny is one upon which the Government may not ordinarily agree a final position internationally. ([54]-[57], [61]). The procedure is used by the EU Committee under the Scrutiny Reserve Resolution, and a similar one is employed by the House of Commons. The Committee and the Secretary of State (Mr. Davis) rejected the need for such an option, and Prof Wyatt was quoted as doubting whether it was necessary. The Secretary of State claimed (and the Committee accepted) that the Government would have to be ‘nimble, fast, and responsive’ in negotiations. Yet the Secretary of State said he would consider the issue further and the Committee indicated they would keep an open mind and the Committee said it was not needed ‘at this stage’ [61]. In my view, the reasoning offered by the Committee and Secretary of State on this point does not sufficiently explain the difference between those cases where scrutiny reserve is used presently from the significant issues that will arise in Brexit negotiations. Professor Wyatt did offer good reasons in his written evidence to the Committee (namely that parallel bilateral negotiations will make the terms of reference of an overarching scrutiny reserve unworkable), but in that same evidence he also recommended adopting a tailored form of scrutiny reserve under which the Government would inform the EU Committee of ‘ its negotiating objectives, and of the contents of negotiating documents to be tabled in negotiations with the EU.’ (Written Evidence of Prof Derek Wyatt, [51]) Furthermore, the existing scrutiny reserve procedure for the EU Committee allows executive override, and this is exercised very frequently. In 2015, for example, there were 143 uses of the override – see here and here). So further examination of the scrutiny reserve option seems more than warranted in parliamentary debate on this matter.

There may be a particularly apt role for a scrutiny reserve procedure where it is known in advance that a particular issue or set of issues is crucial or acute. A clear candidate would be that of acquired rights of EU citizens residing in the UK and British citizens residing in other EU member states. Other plausible candidates would include environmental policy, successor arrangements to the Common Agricultural Policy, and energy policy. In other words, and at the very least, it may be appropriate to recognize a two-track scrutiny function, one for general matters, and an enhanced procedure for acute or fundamental ones. The possibility of certifying matters for enhanced scrutiny should also be considered.

  1.      Rights of Devolved Legislatures and Governments

Notwithstanding the fact that international relations is a reserved rather than devolved matter, it is not constitutionally appropriate to marginalize the devolved legislatures and governments in the discussion. Withdrawal from the EU will require amendments to the devolution statutes, possibly the Belfast Agreement (which plainly envisages ongoing membership of the EU for Northern Ireland), and it has enormous political and economic consequences for matters within devolved competence. It does no good to say that their citizens are represented in the Westminster Parliament. Their institutions are the sites for a distinct political culture and membership of the EU is especially important at the national level for each of them. In a genuinely federal system, constitutional amendment tends to require special approval from a supermajority of the federal sub-units (provinces, states, Bundesländer). While the UK is not a federal state, the devolution arrangements have a quasi-federal character.

The importance of inter governmental relations has been neglected but is becoming increasingly apparent (see Constitution Committee of the House of Lords, Inter Governmental relations in the United Kingdom (11th Report, HL Paper 146, March 2015). The involvement of the devolved governments/administrations in EU policy and international relations of the United Kingdom are specified in the Memorandum of Understanding and Supplementary Agreements between the United Kingdom Government, Scottish Ministers, Welsh Executive and Northern Ireland Executive relating to Devolution, which includes various concordats. The MoU and concordats are ‘binding in honour’ rather than enforceable.

Two such sets of concordats address cooperation on EU policy and on international relations. They specify a general role for the devolved governments in EU relations: that they receive ‘full and comprehensive information,’ and should be ‘fully involved in the UK’s policy position on all issues’ within devolved competence, including ‘upstream opportunities to influence EU proposals before they emerge as well as the period after formal proposals are made…’ ([B.43]).   There is further provision for the actual representation of the devolved governments in UK-led negotiating teams before European Council of Minister meetings ([B.413]). For international relations, there is a similar framework concerning the exchange of information, formulation of UK policy and conduct of international relations, and international representation as part of the UK negotiating team ‘may be appropriate.’ ([D.4] generally).

It may be that the devolved governments believe this would be sufficient despite its non-binding character. Or perhaps a new, Brexit-specific set of concordats might be agreed. But assuming a more formal and binding arrangement were sought, including rights for the devolved legislatures instead of only for the governments (see further below), there would be at least two ways to proceed using Westminster legislation. One is to insert a legal requirement in the notice bill that legislative consent motions be obtained for certain critical junctures: e.g. either the set of procedures I specified above, or for key negotiating offers or (the most extreme) on the ratification of the final exit agreement. The effect of such a provision, however, would be to seal the fate of Brexit going ahead due to the clear opposition of Scotland to the result. Such motions would also rarely obtain cross-community support as is likely to be required for such matters in Northern Ireland under the petition of concern system, and the mere exercise could further destabilise politics and community relations there. It would amount to unacceptable veto points in the process.

Another and in my view clearly acceptable way is to provide legislatively that the devolved governments and legislatures receive parity of treatment in access to information and notice and comment that the Westminster Parliament obtains. The existing framework for EU and international relations in the concordats envisages something related but less precisely defined, though in some ways it goes beyond it. Arrangements could further be made that the Joint Ministerial Committee or one of its sub-committees should meet on a specified schedule to give formal occasion for airing differences and seeking common agreement rather than providing input bilaterally to individual UK departments or the UK Government. Similar provision for consulting the North South Ministerial Council and the British Irish Council would be important for Northern Ireland. Notably, however, it would be crucial that the legislatures themselves enjoy similar access. The Northern Ireland Executive, for example, is unlikely to be able to present a common view on the Brexit question. Neither do the Scottish or Welsh Governments represent the full spectrum of opinion in the respective nations. If voice and participation are the aim, then respecting the diversity of opinion will require looking past the governments.

One might say that we ought not to formalize inter governmental relations in this way, and work through more flexible ad hoc arrangements. Yet it is precisely the absence of formality that has been a consistent complaint about the irregular nature of the system: see Report of the Smith Commission, [28]; Lords Constitution Committee Report, Intergovernmental Relations, [77]-[86]; Lords Constitution Committee Report, The Union and Devolution (10th Report of Session 2015-16,, HL Paper 149 (25 May 2016), [285]-[291]. In a moment such as this, hard schedules and clearly delineated and binding rights and duties, tailored for the Brexit process, are exactly what is needed.

  1.     Referendum on the Exit agreement

Many MPs and Peers have called for a referendum on the exit agreement.   The reasoning runs as follows: if we had a referendum on the general question, why not have one on the particulars when they are ready for public dissemination? Were the United Kingdom actually allowed to revoke notice and commence negotiations again, under article 50, this might be an acceptable proposition. It could give notice, bargain for an agreement, hold a referendum, and then revoke notice if the people say no. The next day it could give notice again and repeat the process until the correct result is obtained. That scenario, though, is exactly why I think notice is not revocable (though some distinguished jurists do, including Professors Paul Craig and Derek Wyatt, and Lord Kerr).

If such a power does not exist, then a referendum rejecting the exit agreement would mean that we have no deal at the expiration of the two-year period.   Given the narrowness of the negotiating time window, and the time that holding a referendum would take, and the inevitable absence of a plan B that is a feature of British politics, we can rest assured that such a referendum would take place right up against the two year timeline. To extend it, we would need the unanimous consent of the European Council.   One single-hold out would be sufficient to say no. In his evidence to the EU Committee, Prof Wyatt expressed the view that given the importance of the UK to the EU, ‘an extension is likely.’ (Brexit: Parliamentary Scrutiny, above, [44]). Perhaps. But it would be an enormous gamble. The risk that it won’t be extended is so great that a referendum may have the perverse result that either (a) it is either a rubber stamp on the agreement or we are kicked out without a deal, or, (b) worse, that people are forced to accept the agreement because the consequences of rejecting it are so great. And in that latter case, the whole affair might even legitimize a bad exit agreement. None of these options is acceptable, and this does not even address the enormous social upheaval that might result from the disaffection of those who voted Leave in June 2016. For these reasons, a second referendum on the final package is likely to be problematic and could even derail the best deal we can get. In such a case, it seems best not to insert such a requirement into the notice bill and to rely at least for the time being on Parliament’s power to reject the agreement under the Constitutional Reform and Governance Act 2010 (though see s.22 for a potential wrinkle, which could be disapplied in the notice bill).

  1.    Parliamentary Control of the Notice Date

Another technique Parliament could use to pass the legislation without conferring immediate authority to give notice, is to provide that the date upon which notice will be given must be set out in an Order in Council and laid before the Commons or both Houses and approved by affirmative resolution. This would ensure that Parliament could insist that all other arrangements, not worked out in the notice legislation itself, can be in place before the notice is finally issued. This would function in a manner similar to commencement orders for statutes, but in this case would be subject to an affirmative resolution procedure.

In reality, this one clause, when paired with the confirmation of ministerial power to give notice, would contain precisely the parliamentary power required to obtain everything else mentioned above without having to include any of it in the statute. However, my sense is that such a procedure on its own would merely be to delay the needed discussions raised by a scheme of consultation. Such discussions ought to begin happening within the structure of legislative scrutiny rather than in the more disorganized manner that would inevitably follow outside of it. Within that process, some general principles may be identified and the Government can indicate willingness to follow through at a later time when drafting the precise terms of the scheme. It may be conceivable that Parliament would wish to control the date of notice over and above the other controls outlined below.

Legislative Specification of the Conditions

The European Parliament’s rights to information and consultation are legally founded on general wording of article 218(10) of the Treaty of the European Union , which provides that the European Parliament must be ‘immediately and fully informed at all stages of the procedure’, as well as article 295 which provides that the European Parliament, Council and Commission shall ‘shall consult each other and by common agreement make arrangements for their cooperation.’ Pursuant to article 295, the Framework Agreement on Relations between the European Parliament and European Commission (OJL 204/47) fleshes out the nature of the arrangements. This provides an obvious precedent for putting the consultation rights on a formal basis. Yet it is likely that in British constitutional arrangements, and above all in the present political climate, that we would wish to avoid putting wording so vague into a statute. It would open the possibility of judicial review of the adequacy of information at heated moments.

What would be a superior approach, then, is to provide legislatively that a scheme of consultation must be laid before both Houses of Parliament and approved by affirmative resolution. For example, statutory wording of the following sort might do: ‘The Secretary of State shall lay a scheme before each House setting out a code of consultation governing the rights of each House [and any members of the public] to receive information relating to the Government’s negotiating strategy, including but not limited to the following: texts of any draft agreements or proposed amendments to existing agreements or draft agreements; procedures relating to the treatment of confidential information; and procedures for the Government’s consideration and response to the views expressed by either House or its Committees; and for the Committees or Houses’ timely rejoinder to such responses. Such a scheme must be approved by affirmative resolution of both Houses.’   Such a provision would create a scheme, but one which can be amended from time to time and hence preserve the needed flexibility. And the terms are so clear and non-evaluative that the adequacy of the code would not be judicially reviewable. Whether compliance with the code itself would be so reviewable is another question, but not likely a significant issue if it were well-drafted with specific terms and clear rules instead of vague principles.

A separate provision of the bill would need to address devolution and inter-governmental arrangements, but could do so in a similar fashion. The provision could provide that the Secretary of State shall lay a scheme before the Westminster Parliament outlining a consultation scheme on rough parity with what is provided above and a role for formal recourse to the Joint Ministerial Committee (and related Councils for Northern Ireland) or similar but bespoke intergovernmental relations framework. That the scheme would be laid before the Westminster Parliament for approval is in some ways not ideal. The devolved administrations were part of the negotiation of the Memorandum of Understanding and various concordats, a notable difference to what is proposed here. One way to square the circle in this process is to seek legislative consent motions on the notice bill, and if there is an impasse in negotiations, to proceed without them. The Sewell and related conventions are just that – a convention, one that ‘normally’ obtains: s.2 Scotland Act 2016. The present situation is distinguishable from the norm on several levels. The scheme proposed here would clearly build on the existing framework and be the only way to guarantee fleshed out and legally secure rights to notice, comment and response short of putting a veto power into the hands of the devolved governments or legislatures. If politics is the art of compromise, and if the ongoing power of the Westminster Parliament to conduct international relations for the United Kingdom is given credence, and that we do in fact have a devolved rather than federal system, it seems that this is the best solution broadly considered.

A crucial question remains. In what way would such requirements to obtain approval of the Houses to the consultation and devolution schemes relate to giving authority to the Government to give notice under Article 50? The duty to have such schemes approved in this way could be expressed either as (1) conditions to be satisfied prior to the giving of notice, or (2) as obligations to be discharged ‘prior to the commencement of formal exit negotiations with the European Union.’   The first of these options provides an iron clad guarantee to Parliament, but could have the unintended effect of rushing negotiations over the content of the schemes in order to meet the Government’s promised deadline for giving notice. Hence the latter option is likely preferable for providing greater flexibility, by allowing notice to issue by the 31 March 2017 deadline but barring negotiations until the consultation schemes are agreed. It seems plain to me that there is a difference between giving notice on the one hand, and commencing negotiations on the other. But were there any doubts about this it would be clarified in the statute.

The Role of the House of Lords

In respect of this particular bill, all Peers will rightly recognize that whatever the importance of constitutionally appropriate procedures governing parliamentary scrutiny of Brexit, some restraint will be appropriate. Such is a matter of plain democratic politics, but there is another reason of a more constitutional nature: the Salisbury-Addison Convention. The convention is ‘an understanding that a “manifesto” Bill, foreshadowed in the governing party’s most recent election manifesto and passed by the House of Commons, should not be opposed by the second chamber on Second or Third Reading.’ It also is often interpreted to include a prohibition on wrecking amendments to any such bills. (Joint Committee on Conventions, First Special Report (HL Paper 189, HC 1151, 2005–06 (25th May 2006), pp. 4–5; see further background on the convention here).

The Election Manifesto of the Conservative Party in 2015 declares the following: ‘We will honour the result of the referendum, whatever the outcome.’ (p.73). It is true that the same manifesto declares: ‘We say: yes to the Single Market. […] Yes to a family of nation states, all part of a European Union – but whose interests, crucially, are guaranteed whether inside the Euro or out.’ (p.72) One might argue that the messages are mixed. But the interpretation that makes them consistent is that it was a party campaigning to remain but committing to leave if the Leave vote won the referendum. In my view, the Salisbury-Addison Convention is engaged.

What implications does this have? The Lords can still amend the bill, provided they do not block it or insist on wrecking amendments. This is fully compatible with its role as a revising chamber and it would be absurd to exclude the chamber’s expertise and committees from the process.

Belief that reliance on the Parliament Acts 1911 and 1949 can be sufficient for securing the primacy of the Commons would not be appropriate in my view. Under these acts, the Lords may obstruct the bill for up to one year, roughly speaking, after which the Commons could present it for Royal signature and it would become an Act of Parliament. Yet this would put the Government in an unacceptable position due to the delays, and would arguably put the UK in breach of its international obligations vis-a-vis the EU. The House of Lords should regard them as off the table.

The Impropriety of Fast-Track Legislation

The term ‘fast-track legislation’ is not official but is recognised in reports of the House of Lords Constitution Committee, (‘Fast Track Legislation: Constitutional Implications and Safeguards’, 15th Report of 2008-2009 (17 June 2009); The Process of Constitutional Change, 15th Report of 2010-2011 (6 July 2011). It refers to the expedited passage of a bill, often taking all stages in one day or more than one stage on a single day. The ordinary course of legislative scrutiny is severely curtailed. It would be common ground that constitutional legislation, whatever it may be, ought not to be subject to such a procedure. Indeed, the Constitution Committee finds in its 2011 report that ‘parliamentary scrutiny of such [constitutional] bills should not be rushed unless there are justifiable reasons for fast-tracking them’ ([199]). It found that ordinarily there should be a green paper and then white paper, and that if the Government departs from this practice it should set out a ministerial statement explaining why. It envisages departure ‘only in exceptional circumstances and where there are clearly justifiable reasons for so doing.’ ([73]).

The question is therefore whether this particular notice bill could be regarded as a constitutional bill or bill having constitutional implications, and if so, whether there are clearly justifiable reasons for fast tracking. The answer to the former question is yes. It is far from the Great Repeal Bill in its significance, however it will trigger a potentially irreversible withdrawal on a tight timeframe and could significantly affect intergovernmental relations within the United Kingdom.   However, the protracted public discussion both before and after the referendum, during which time there was significant opportunity for public and parliamentary comment, is a relevant consideration. The notice legislation is only part of the package.

The more important point is that there is simply no need whatsoever to expedite the legislation in the way seen with other fast tracked legislation. There is no emergency. There is time for proper if somewhat rapid legislative scrutiny, including even committee hearings and reports. The matter is one that all can expect to be given priority over other business and it should be noticed by all that the 31 March 2017 deadline is one that suits the EU as much as Prime Minister May’s Government. It will thus be crucial that the Government and Opposition put a bill before the House of Commons at the earliest opportunity to enable this more appropriate form of scrutiny to take place.

Jeff King is a Professor of Law at the Faculty of Laws, University College London. He wishes to thank Nick Barber, Rory O’Connell, Gavin Phillipson, Maurice Sunkin, Stephen Tierney, and above all, Julika Erfurt, for feedback and support.

(Suggested citation: J. King, ‘What Next? Legislative Authority for Triggering Article 50’ U.K. Const. L. Blog (8th Nov 2016) (available at