R (Miller) v Secretary of State for Exiting the EU

On 1 January 1973, the United Kingdom became a member of the European Economic Community (“the EEC”) and certain other associated European organisations. On that date, EEC law took effect as part of the domestic law of the United Kingdom, in accordance with the European Communities Act 1972 which had been passed ten weeks earlier. Over the next 40 years, the EEC expanded from nine to 28 member states, extended its powers or “competences”, merged with the associated organisations, and changed its name to the European Community in 1993 and to the European Union in 2009.

In December 2015, the UK Parliament passed the European Union Referendum Act, and the ensuing referendum on 23 June 2016 produced a majority in favour of leaving the European Union. UK government ministers (whom we will call “ministers” or “the UK government”) thereafter announced that they would bring UK membership of the European Union to an end. The question before this Court concerns the steps which are required as a matter of UK domestic law before the process of leaving the European Union can be initiated. The particular issue is whether a formal notice of withdrawal can lawfully be given by ministers without prior legislation passed in both Houses of Parliament and assented to by HM The Queen.

It is worth emphasising that nobody has suggested that this is an inappropriate issue for the courts to determine. It is also worth emphasising that this case has nothing to do with issues such as the wisdom of the decision to withdraw from the European Union, the terms of withdrawal, the timetable or arrangements for withdrawal, or the details of any future relationship with the European Union. Those are all political issues which are matters for ministers and Parliament to resolve. They are not issues which are appropriate for resolution by judges, whose duty is to decide issues of law which are brought before them by individuals and entities exercising their rights of access to the courts in a democratic society.

Some of the most important issues of law which judges have to decide concern questions relating to the constitutional arrangements of the United Kingdom. These proceedings raise such issues. As already indicated, this is not because they concern the United Kingdom’s membership of the European Union; it is because they concern (i) the extent of ministers’ power to effect changes in domestic law through exercise of their prerogative powers at the international level, and (ii) the relationship between the UK government and Parliament on the one hand and the devolved legislatures and administrations of Scotland, Wales and Northern Ireland on the other.

The main issue on this appeal concerns the ability of ministers to bring about changes in domestic law by exercising their powers at the international level, and it arises from two features of the United Kingdom’s constitutional arrangements. The first is that ministers generally enjoy a power freely to enter into and to terminate treaties without recourse to Parliament. This prerogative power is said by the Secretary of State for Exiting the European Union to include the right to withdraw from the treaties which govern UK membership of the European Union (“the EU Treaties”). The second feature is that ministers are not normally entitled to exercise any power they might otherwise have if it results in a change in UK domestic law, unless statute, ie an Act of Parliament, so provides. The argument against the Secretary of State is that this principle prevents ministers withdrawing from the EU Treaties, until effectively authorised to do so by a statute.

Most of the devolution issues arise from the contention that the terms on which powers have been statutorily devolved to the administrations of Scotland, Wales and Northern Ireland are such that, unless Parliament provides for such withdrawal by a statute, it would not be possible for formal notice of the United Kingdom’s withdrawal from the EU Treaties to be given without first consulting or obtaining the agreement of the devolved legislatures. And, in the case of Northern Ireland, there are certain other arguments of a constitutional nature.

The main issue was raised in proceedings brought by Gina Miller and Deir dos Santos (“the applicants”) against the Secretary of State for Exiting the European Union in the Divisional Court of England and Wales. Those proceedings came before Lord Thomas of Cwmgiedd LCJ, Sir Terence Etherton MR and Sales LJ. They ruled against the Secretary of State in a judgment given on 3 November 2016 - R (Miller) v The Secretary of State for Exiting the European Union [2016] EWHC 2768 (Admin) . This decision now comes to this Court pursuant to an appeal by the Secretary of State.

The applicants are supported in their opposition to the appeal by a number of people, including (i) a group deriving rights of residence in the UK under EU law on the basis of their relationship with a British national or with a non-British EU national exercising EU Treaty rights to be in the United Kingdom, (ii) a group deriving rights of residence from persons permitted to reside in the UK because of EU rights, including children and carers, (iii) a group mostly of UK citizens residing elsewhere in the European Union, (iv) a group who are mostly non-UK EU nationals residing in the United Kingdom, and (v) the Independent Workers Union of Great Britain. The Secretary of State’s case is supported by Lawyers for Britain Ltd, a group of lawyers.

Devolution arguments relating to Northern Ireland were raised in proceedings brought by Steven Agnew and others and by Raymond McCord against the Secretary of State for Exiting the European Union and the Secretary of State for Northern Ireland. Those arguments were rejected by Maguire J in a judgment given in the Northern Ireland High Court on 28 October 2016 - Re McCord, Judicial Review [2016] NIQB 85. On application by the Attorney General for Northern Ireland, Maguire J referred four of the issues in the Agnew case to this court for determination. Following an appeal against Maguire J’s decision, the Northern Ireland Court of Appeal has also referred one issue to this Court.

The Attorney General for Northern Ireland supports the Secretaries of State’s case that no statute is required before ministers can give notice of withdrawal. In addition, there are interventions on devolution issues by the Lord Advocate on behalf of the Scottish government and the Counsel General for Wales on behalf of the Welsh government; they also rely on the Sewel Convention (as explained in paras 137 to 139 below). They support the argument that a statute is required before ministers can give notice of withdrawal, as do the advocates for Mr McCord and for Mr Agnew.

We are grateful to all the advocates and solicitors involved for the clarity and skill with which the respective cases have been presented orally and in writing, and for the efficiency with which the very substantial documentation was organised. We have also been much assisted by a number of illuminating articles written by academics following the handing down of the judgment of the Divisional Court. It is a tribute to those articles that they have resulted in the arguments advanced before this Court being somewhat different from, and more refined than, those before that court.

As mentioned in paras 7 and 9 above, the appellant in the English and Welsh appeal is the Secretary of State for Exiting the European Union, and the Northern Irish proceedings were brought against the Secretary of State for Exiting the European Union and the Secretary of State for Northern Ireland. For the sake of simplicity, we will hereafter refer to either or both Secretaries of State simply as “the Secretary of State”.

From about 1960, the UK government was in negotiations with the then member states of the EEC with a view to the United Kingdom joining the EEC and associated European organisations. In October 1971, when it had become apparent that those negotiations were likely to be successful, and following debates in each House, the House of Lords and the House of Commons each resolved to “approve … Her Majesty’s Government’s decision of principle to join the European Communities on the basis of the arrangements which have been negotiated”. In the course of the debate in the House of Commons, the Prime Minister, Mr Heath, said that he did not think that “any Prime Minister has … in time of peace … asked the House to take a positive decision of such importance as I am asking it to take”, and that he could not “over-emphasise tonight the importance of the vote which is being taken, the importance of the issue, the scale and quality of the decision and the impact that it will have equally inside and outside Britain”. In a debate in the House of Commons in January 1972, in which the earlier resolution was effectively re-affirmed, Mr Rippon, the Chancellor of the Duchy of Lancaster, said “I do not think Parliament in negotiations on a treaty has ever been brought so closely into the process of treaty-making as on the present occasion”, adding that “we all accept the unique character of the Treaty of Accession”.

On 22 January 1972, two days after that later debate, ministers signed a Treaty of Accession which provided that the United Kingdom would become a member of the EEC on 1 January 1973 and would accordingly be bound by the 1957 Treaty of Rome, which was then the main treaty in relation to the EEC, and by certain other connected treaties. As with most international treaties, the 1972 Accession Treaty was not binding unless and until it was formally ratified by the United Kingdom.

A Bill was then laid before Parliament, and after it had been passed by both Houses, it received Royal assent on 17 October 1972, when it became the European Communities Act 1972 . The following day, 18 October 1972, ministers ratified the 1972 Accession Treaty on behalf of the United Kingdom, which accordingly became a member of the EEC on 1 January 1973.

The long title of the 1972 Act described its purpose as “to make provision in connection with the enlargement of the European Communities to include the United Kingdom …”. Part I of the 1972 Act consisted of sections 1 to 3, which contained its “General Provisions”, and they are of central importance to these proceedings.

Section 1(2) of the 1972 Act contained some important definitions. “The Communities” meant the EEC and associated communities (now amended to “the EU” meaning the European Union). And “the Treaties” and “the Community Treaties” (now amended to “the EU Treaties”) were the treaties described in Schedule 1 (which were the existing treaties governing the rules and powers of the EEC at that time), the 1972 Accession Treaty itself, and “any other treaty entered into by any of the Communities, with or without any of the member States, or entered into, as a treaty ancillary to any of the Treaties, by the United Kingdom”. The use of a capital T in “the Treaties” and in “the EU Treaties” was significant. It meant that future treaties which were concerned with changing the membership or redefining the rules of the EEC could only become “Treaties” and “EU Treaties” and have effect in UK law as such if they were added to section 1(2) by an amending statute. By contrast, “ancillary” treaties covered other treaties entered into by the European Union or by the United Kingdom as a treaty ancillary to the EU Treaties. By virtue of section 1(3) , even such an ancillary treaty did not take effect in UK law unless and until it was declared to do so by an Order in Council which had first to be “approved” in draft form “by resolution of each House of Parliament”.

Section 2 of the 1972 Act was headed “General Implementation of Treaties”. Section 2(1) of the 1972 Act was in these terms:

“All such rights, powers, liabilities, obligations and restrictions from time to time created or arising by or under the Treaties, and all such remedies and procedures from time to time provided for by or under the Treaties, as in accordance with the Treaties are without further enactment to be given legal effect or used in the United Kingdom shall be recognised and available in law, and be enforced, allowed and followed accordingly …”

Section 2(2) of the 1972 Act provided that “Her Majesty may by Order in Council, and any designated Minister or department may by regulations, make provision” (a) “for the purpose of implementing any Community [now EU] obligation of the United Kingdom” (which is defined as any obligation “created or arising by or under the Treaties”) or “enabling any rights … enjoyed … by the United Kingdom under or by virtue of the Treaties to be exercised”, and (b) for ancillary purposes, including “the operation from time to time of sub section (1 )”. Sub section (2 ) has since been amended, but nothing hangs on the amendments for present purposes. Schedule 2 to the 1972 Act contained “Provisions as to Subordinate Legislation” in relation to the powers conferred by section 2(2) .

Section 2(4) provided as follows:

“The provision that may be made under sub section (2 ) above includes ... any such provision (of any such extent) as might be made by Act of Parliament, and any enactment passed or to be passed, other than one contained in this Part of this Act , shall be construed and have effect subject to the foregoing provisions of this section …”

Section 3 of the 1972 Act provided, among other things, for any question as to the meaning and effect of the Treaties, or as to the validity, meaning or effect of any “Community instrument” (now “EU instrument”) to be treated as a question of EU law by the UK courts, and it further provided for such determination to be made in accordance with principles laid down by the European Court of Justice (“the Court of Justice”) or, if necessary, to be referred to the Court of Justice.

Part II of the 1972 Act , which contained sections 4 to 12, and incorporated Schedules 3 and 4, set out a number of statutory repeals and amendments which were needed to enable UK domestic law to comply with the requirements of EU law, that is the law from time to time laid down in the EU Treaties, Directives and Regulations, as interpreted by the Court of Justice.

Following a manifesto commitment made during a general election in 1974, the UK government decided to hold a referendum on whether the United Kingdom should remain in the EEC. To that end, it laid a Bill before Parliament which was duly enacted as the Referendum Act 1975 . The referendum pursuant to that Act took place on 5 June 1975, and a majority of those who voted were in favour of remaining in the EEC.

In the past 40 years, over 20 treaties relating to the EEC, the European Community and the European Union were signed on behalf of the member states, in the case of the United Kingdom by ministers. After being signed, each such treaty was then added to the list of “Treaties” in section 1(2) of the 1972 Act through the medium of an amendment made to that statute by a short appropriately worded statute passed by Parliament, and the treaty was then ratified by the United Kingdom. Some of these Treaties were concerned with redefining and expanding the competences of the EEC, the European Community and the European Union and changing the constitutional role of the European Parliament within the European Community or Union. They included the Single European Act signed in 1986, Titles II, III and IV of the Maastricht Treaty on European Union of 7 February 1992 (“the TEU”), the 1997 Amsterdam Treaty, the 2001 Treaty of Nice, and the Treaty of Lisbon amending the TEU and the Treaty on the Functioning of the European Union (“TFEU”), both signed in Lisbon on 13 December 2007 - see respectively section 1(2) (j), added by the European Communities (Amendment) Act 1986 ; section 1(2) (k), added by the European Communities (Amendment) Act 1993 ; section 1(2) (o), added by the European Communities (Amendment) Act 1998 ; section 1(2) (p), added by the European Communities (Amendment) Act 2002 ; and section 1(2) (s), added by the European Union (Amendment) Act 2008 .

The Treaty of Lisbon introduced into the EU Treaties for the first time an express provision entitling a member state to withdraw from the European Union. It did this by inserting a new article 50 into the TEU. This article (“article 50”) provides as follows:

“1. Any member state may decide to withdraw from the Union in accordance with its own constitutional requirements.

A member state which decides to withdraw shall notify the European Council of its intention. In the light of the guidelines provided by the European Council, the Union shall negotiate and conclude an agreement with that state, setting out the arrangements for its withdrawal …

The Treaties shall cease to apply to the state in question from the date of entry into force of the withdrawal agreement or, failing that, two years after the notification referred to in paragraph 2, unless the European Council, in agreement with the member state concerned, unanimously decides to extend this period. …”

In these proceedings, it is common ground that notice under article 50(2) (which we shall call “Notice”) cannot be given in qualified or conditional terms and that, once given, it cannot be withdrawn. Especially as it is the Secretary of State’s case that, even if this common ground is mistaken, it would make no difference to the outcome of these proceedings, we are content to proceed on the basis that that is correct, without expressing any view of our own on either point. It follows from this that once the United Kingdom gives Notice, it will inevitably cease at a later date to be a member of the European Union and a party to the EU Treaties.

After 1975, in addition to the amending statutes referred to in para 24 above, statutes were enacted to give effect to changes in the way that members of the European Parliament were selected. The first was the European Assembly Elections Act 1978 , which contained in section 6 a stipulation that no new treaty providing for an increase in the powers of the European Assembly (as it then was) should be ratified unless approved by an Act of Parliament. This provision was re-enacted as section 12 of the European Parliamentary Elections Act 2002 . Section 1 of the 2002 Act provided for a specific number of Members of the European Parliament (“MEPs”) for specified regions of the United Kingdom. Section 8 of the 2002 Act stated that a person was entitled to vote in elections to the European Parliament if he or she satisfied certain residence requirements, and section 10 identified the (very limited) categories of people who were disqualified from standing as MEPs.

In addition to adding the Treaty of Lisbon and the TFEU to section 1(2) of the 1972 Act , the 2008 Act , referred to at the end of para 24 above, contained certain restrictions on the UK government’s agreement to changes in the rules of the European Union. Section 5 provided that any treaty which amended the TEU or the TFEU by altering the competences of the European Union, or which altered the decision-making processes of the European Union or its institutions in such a way as to dilute the influence of individual member states, should not be ratified by ministers “unless approved by Act of Parliament”. Section 6 of the 2008 Act stated that ministers should not support any decision under certain specified articles of the TEU and of the TFEU unless both Houses of Parliament had approved a motion sanctioning that course.

Subject to an immaterial exception, the European Union Act 2011 repealed and replaced sections 5 and 6 of the 2008 Act . Part I of the 2011 Act included section 1 which was “Introductory”, sections 2 to 10, which imposed “Restrictions” both “relating to amendments of TEU and TFEU” and “relating to other decisions under TEU and TFEU”, and sections 11 to 13, which related to the conduct of referendums. Sections 2 to 5 imposed restrictions on the ratification by the United Kingdom of any treaty which amended or replaced TEU or TFEU, and also on ministers approving certain specified types of EU decisions under the so-called simplified revision procedure. Those restrictions were that (a) a statement relating to the treaty or decision had to be laid before Parliament, (b) the treaty or decision had to be approved by statute, and, (c) in broad terms, where the treaty or decision increased the competences of the European Union, it had to be approved in a UK-wide referendum. Section 6 provided that ministers should not, without prior approval both in a statute and in a UK-wide referendum, vote in favour of certain decisions, including those which resulted in a dilution in the influence of individual member states in relation to a number of different articles of the TEU and TFEU including in particular article 50(3). Sections 7 to 10 of the 2011 Act contained further restrictions on ministers voting in favour of certain measures under the TEU and TFEU without the prior approval of Parliament.

Section 18 of the 2011 Act provided as follows:

“Directly applicable or directly effective EU law (that is, the rights, powers, liabilities, obligations, restrictions, remedies and procedures referred to in section 2(1) of the European Communities Act 1972 ) falls to be recognised and available in law in the United Kingdom only by virtue of that Act or where it is required to be recognised and available in law by virtue of any other Act.”

Following a manifesto commitment in the 2015 general election to hold a referendum on the issue of EU membership, the UK government laid before Parliament a Bill which became the 2015 Act. Section 1 provided that “[a] referendum is to be held” on a date no later than 31 December 2017 “on whether the United Kingdom should remain a member of the European Union”, and it specified the question which should appear on the ballot papers. The remaining sections were concerned with questions such as entitlement to vote, the conduct of the referendum, rules about campaigning and financial controls.

The referendum duly took place throughout the United Kingdom on 23 June 2016, and it resulted in a majority in favour of leaving the European Union. Ministers have made it clear that the UK government intends to implement the result of the referendum and to give Notice by the end of March 2017.

On 7 December 2016, following a debate, the House of Commons resolved “[to recognise] … that this House should respect the wishes of the United Kingdom as expressed in the referendum on 23 June; and further [to call] on the Government to invoke article 50 by 31 March 2017”.

The Secretary of State’s case is based on the existence of the well-established prerogative powers of the Crown to enter into and to withdraw from treaties. He contends that ministers are entitled to exercise this power in relation to the EU Treaties, and therefore to give Notice without the need for any prior legislation. Following the giving of Notice by the end of March 2017, ministers intend that a “Great Repeal Bill” will be laid before Parliament. This will repeal the 1972 Act and, wherever practical, it will convert existing EU law into domestic law at least for a transitional period. Under article 50, withdrawal will occur not more than two years after the Notice is given (unless that period is extended by unanimous agreement among the other member states), and it is intended that the Great Repeal Bill will come into force at that point.

As was made clear by Lord Browne-Wilkinson in R v Secretary of State for the Home Department, Ex p Fire Brigades Union [1995] 2 AC 513 , 552, ministers’ intentions are not law, and the courts cannot proceed on the assumption that they will necessarily become law. That is a matter for Parliament to decide in due course. The issues before us must be resolved in accordance with the law as it stands, as the Secretary of State rightly accepted.

The applicants’ case in that connection is that when Notice is given, the United Kingdom will have embarked on an irreversible course that will lead to much of EU law ceasing to have effect in the United Kingdom, whether or not Parliament repeals the 1972 Act . As Lord Pannick QC put it for Mrs Miller, when ministers give Notice they will be “pulling … the trigger which causes the bullet to be fired, with the consequence that the bullet will hit the target and the Treaties will cease to apply”. In particular, he said, some of the legal rights which the applicants enjoy under EU law will come to an end. This, he submitted, means that the giving of Notice would pre-empt the decision of Parliament on the Great Repeal Bill. It would be tantamount to altering the law by ministerial action, or executive decision, without prior legislation, and that would not be in accordance with our law.

Following opening remarks made by HM Attorney General, Mr Eadie QC in his submissions on behalf of the Secretary of State, did not challenge much if any of the factual basis of these assertions, but he did challenge the conclusions that were said to derive from them. He argued that the fact that significant legal changes will follow from withdrawing from the EU Treaties does not prevent the giving of Notice, because the prerogative power to withdraw from treaties was not excluded by the terms of the 1972 Act , and that, in any event, “acts of the government in the exercise of the prerogative can alter domestic law”. More particularly, he contended that the 1972 Act gave effect to EU law only insofar as the EU Treaties required it, and that that effect was therefore contingent upon the United Kingdom remaining a party to those treaties. Accordingly, he said, in the 1972 Act Parliament had effectively stipulated that, or had sanctioned the result whereby, EU law should cease to have domestic effect in the event that ministers decided to withdraw from the EU Treaties.

Mr Eadie also relied on the fact that, while statutes enacted since 1972 have imposed Parliamentary controls over the exercise of prerogative powers in relation to the EU Treaties, they have not touched the prerogative power to withdraw from them. Implicitly, therefore, he contended, Parliament has recognised that the power to withdraw from such treaties exists and is exercisable without prior legislation. Mr Eadie also suggested that the 2015 Act was enacted on the assumption that the result of the referendum would be decisive. Mr Eadie’s reliance on the legislation since 1972 was largely for the purpose of supporting his argument on the effect of the 1972 Act , but he did raise an argument that the legislation from 1972 to 2015 should be looked at as a whole. Also, in answer to a question from the Court, he adopted a suggestion that, even if Parliamentary authority would otherwise have been required, the 2015 Act and the subsequent referendum dispensed with that requirement, but he did not develop that argument, in our view realistically.

Before addressing these arguments, it is right to consider some relevant constitutional principles and in particular the Royal prerogative.

Unlike most countries, the United Kingdom does not have a constitution in the sense of a single coherent code of fundamental law which prevails over all other sources of law. Our constitutional arrangements have developed over time in a pragmatic as much as in a principled way, through a combination of statutes, events, conventions, academic writings and judicial decisions. Reflecting its development and its contents, the UK constitution was described by the constitutional scholar, Professor AV Dicey, as “the most flexible polity in existence” - Introduction to the Study of the Law of the Constitution (8th ed, 1915), p 87.

Originally, sovereignty was concentrated in the Crown, subject to limitations which were ill-defined and which changed with practical exigencies. Accordingly, the Crown largely exercised all the powers of the state (although it appears that even in the 11th century the King rarely attended meetings of his Council, albeit that its membership was at his discretion). However, over the centuries, those prerogative powers, collectively known as the Royal prerogative, were progressively reduced as Parliamentary democracy and the rule of law developed. By the end of the 20th century, the great majority of what had previously been prerogative powers, at least in relation to domestic matters, had become vested in the three principal organs of the state, the legislature (the two Houses of Parliament), the executive (ministers and the government more generally) and the judiciary (the judges). It is possible to identify a number of seminal events in this history, but a series of statutes enacted in the twenty years between 1688 and 1707 were of particular legal importance. Those statutes were the Bill of Rights 1688/9 and the Act of Settlement 1701 in England and Wales, the Claim of Right Act 1689 in Scotland, and the Act s of Union 1706 and 1707 in England and Wales and in Scotland respectively. (Northern Ireland joined the United Kingdom pursuant to the Act s of Union 1800 in Britain and Ireland).

The independence of the judiciary was formally recognised in these statutes. In the broadest sense, the role of the judiciary is to uphold and further the rule of law; more particularly, judges impartially identify and apply the law in every case brought before the courts. That is why and how these proceedings are being decided. The law is made in or under statutes, but there are areas where the law has long been laid down and developed by judges themselves: that is the common law. However, it is not open to judges to apply or develop the common law in a way which is inconsistent with the law as laid down in or under statutes, ie by Acts of Parliament.

This is because Parliamentary sovereignty is a fundamental principle of the UK constitution, as was conclusively established in the statutes referred to in para 41 above. It was famously summarised by Professor Dicey as meaning that Parliament has “the right to make or unmake any law whatsoever; and further, no person or body is recognised by the law as having a right to override or set aside the legislation of Parliament” - op cit , p 38. The legislative power of the Crown is today exercisable only through Parliament. This power is initiated by the laying of a Bill containing a proposed law before Parliament, and the Bill can only become a statute if it is passed (often with amendments) by Parliament (which normally but not always means both Houses of Parliament) and is then formally assented to by HM The Queen. Thus, Parliament, or more precisely the Crown in Parliament, lays down the law through statutes - or primary legislation as it is also known - and not in any other way.

In the early 17th century Case of Proclamations (1610) 12 Co Rep 74, Sir Edward Coke CJ said that “the King by his proclamation or other ways cannot change any part of the common law, or statute law, or the customs of the realm”. Although this statement may have been controversial at the time, it had become firmly established by the end of that century. In England and Wales, the Bill of Rights 1688 confirmed that “the pretended power of suspending of laws or the execution of laws by regall authority without consent of Parlyament is illegall” and that “the pretended power of dispensing with laws or the execution of laws by regall authoritie as it hath beene assumed and exercised of late is illegall”. In Scotland, the Claim of Right 1689 was to the same effect, providing that “all Proclamationes asserting ane absolute power to Cass [ie to quash] annull and Dissable lawes … are Contrair to Law”. And article 18 of the Act s of Union of 1706 and 1707 provided that (with certain irrelevant exceptions) “all … laws” in Scotland should “remain in the same force as before … but alterable by the Parliament of Great Britain”.

The Crown’s administrative powers are now exercised by the executive, ie by ministers who are answerable to the UK Parliament. However, consistently with the principles established in the 17th century, the exercise of those powers must be compatible with legislation and the common law. Otherwise, ministers would be changing (or infringing) the law, which, as just explained, they cannot do. A classic statement of the position was given by Lord Parker of Waddington in The Zamora [1916] 2 AC 77 , 90:

“The idea that the King in Council, or indeed any branch of the Executive, has power to prescribe or alter the law to be administered by Courts of law in this country is out of harmony with the principles of our Constitution. It is true that, under a number of modern statutes, various branches of the Executive have power to make rules having the force of statutes, but all such rules derive their validity from the statute which creates the power, and not from the executive body by which they are made. No one would contend that the prerogative involves any power to prescribe or alter the law administered in Courts of Common Law or Equity.”

It is true that ministers can make laws by issuing regulations and the like, often known as secondary or delegated legislation, but (save in limited areas where a prerogative power survives domestically, as exemplified by the cases mentioned in paras 52 and 53 below) they can do so only if authorised by statute. So, if the regulations are not so authorised, they will be invalid, even if they have been approved by resolutions of both Houses under the provisions of the relevant enabling Act - for a recent example see R (The Public Law Project) v Lord Chancellor [2016] AC 1531 .

The Royal prerogative encompasses the residue of powers which remain vested in the Crown, and they are exercisable by ministers, provided that the exercise is consistent with Parliamentary legislation. In Burmah Oil Co (Burma Trading) Ltd v Lord Advocate [1965] AC 75 , 101, Lord Reid explained that the Royal prerogative is a source of power which is “only available for a case not covered by statute”. Professor HWR Wade summarised the position in his introduction to the first edition of what is now Wade on Administrative Law (1961), p 13:

“[T]he residual prerogative is now confined to such matters as summoning and dissolving Parliament, declaring war and peace, regulating the armed forces in some respects, governing certain colonial territories, making treaties (though as such they cannot affect the rights of subjects), and conferring honours. The one drastic internal power of an administrative kind is the power to intern enemy aliens in time of war.”

Thus, consistently with Parliamentary sovereignty, a prerogative power however well-established may be curtailed or abrogated by statute. Indeed, as Professor Wade explained, most of the powers which made up the Royal prerogative have been curtailed or abrogated in this way. The statutory curtailment or abrogation may be by express words or, as has been more common, by necessary implication. It is inherent in its residual nature that a prerogative power will be displaced in a field which becomes occupied by a corresponding power conferred or regulated by statute. This is what happened in the two leading 20th century cases on the topic, Attorney General v De Keyser’s Royal Hotel Ltd [1920] AC 508 and Fire Brigades Union cited above. As Lord Parmoor explained in De Keyser at p 575, when discussing the prerogative power to take a subject’s property in time of war:

“The constitutional principle is that when the power of the Executive to interfere with the property or liberty of subjects has been placed under Parliamentary control, and directly regulated by statute, the Executive no longer derives its authority from the Royal Prerogative of the Crown but from Parliament, and that in exercising such authority the Executive is bound to observe the restrictions which Parliament has imposed in favour of the subject.”

In Burmah Oil cited above, at p 101, Lord Reid described prerogative powers as a “relic of a past age”, but that description should not be understood as implying that the Royal prerogative is either anomalous or anachronistic. There are important areas of governmental activity which, today as in the past, are essential to the effective operation of the state and which are not covered, or at least not completely covered, by statute. Some of them, such as the conduct of diplomacy and war, are by their very nature at least normally best reserved to ministers just as much in modern times as in the past, as indeed Lord Reid himself recognised in Burmah Oil at p 100.

Consistently with paras 44 to 46, and the passage quoted from Professor Wade in para 47 above, it is a fundamental principle of the UK constitution that, unless primary legislation permits it, the Royal prerogative does not enable ministers to change statute law or common law. As Lord Hoffmann observed in R (Bancoult) v Secretary of State for Foreign and Commonwealth Affairs (No 2) [2009] AC 453 , para 44, “since the 17th century the prerogative has not empowered the Crown to change English common or statute law”. This is, of course, just as true in relation to Scottish, Welsh or Northern Irish law. Exercise of ministers’ prerogative powers must therefore be consistent both with the common law as laid down by the courts and with statutes as enacted by Parliament.

Further, ministers cannot frustrate the purpose of a statute or a statutory provision, for example by emptying it of content or preventing its effectual operation. Thus, ministers could not exercise prerogative powers at the international level to revoke the designation of Laker Airways under an aviation treaty as that would have rendered a licence granted under a statute useless: Laker Airways Ltd v Department of Trade [1977] QB 643 - see especially at pp 718-719 and 728 per Roskill LJ and Lawton LJ respectively. And in Fire Brigades Union cited above, at pp 551-552, Lord Browne-Wilkinson concluded that ministers could not exercise the prerogative power to set up a scheme of compensation for criminal injuries in such a way as to make a statutory scheme redundant, even though the statute in question was not yet in force. And, as already mentioned in para 35 above, he also stated that it was inappropriate for ministers to base their actions (or to invite the court to make any decision) on the basis of an anticipated repeal of a statutory provision as that would involve ministers (or the court) pre-empting Parliament’s decision whether to enact that repeal.

The fact that the exercise of prerogative powers cannot change the domestic law does not mean that such an exercise is always devoid of domestic legal consequences. There are two categories of case where exercise of the prerogative can have such consequences. The first is where it is inherent in the prerogative power that its exercise will affect the legal rights or duties of others. Thus, the Crown has a prerogative power to decide on the terms of service of its servants, and it is inherent in that power that the Crown can alter those terms so as to remove rights, albeit that such a power is susceptible to judicial review: Council of Civil Service Unions v Minister for the Civil Service [1985] AC 374 . The Crown also has a prerogative power to destroy property in wartime in the interests of national defence (although at common law compensation was payable: Burmah Oil cited above). While the exercise of the prerogative power in such cases may affect individual rights, the important point is that it does not change the law, because the law has always authorised the exercise of the power.

The second category comprises cases where the effect of an exercise of prerogative powers is to change the facts to which the law applies. Thus, the exercise of the prerogative to declare war will have significant legal consequences: actions which were previously lawful may become treasonable (as in Joyce v Director of Public Prosecutions [1946] AC 347 ), and some people will become enemy aliens, whose property is liable to confiscation. Likewise, in Post Office v Estuary Radio Ltd [1968] 2 QB 740 the Crown’s exercise of its prerogative to extend UK territorial waters resulted in the criminalisation of broadcasts from ships in the extended area, which had previously been lawful. These are examples where the exercise of the prerogative power alters the status of a person, thing or activity so that an existing rule of law comes to apply to it. However, in such cases the exercise has not created or changed the law, merely the extent of its application.

The most significant area in which ministers exercise the Royal prerogative is the conduct of the United Kingdom’s foreign affairs. This includes diplomatic relations, the deployment of armed forces abroad, and, particularly in point for present purposes, the making of treaties. There is little case law on the power to terminate or withdraw from treaties, but, as a matter of both logic and practical necessity, it must be part of the treaty-making prerogative. As Lord Templeman put it in JH Rayner (Mincing Lane) Ltd v Department of Trade and Industry [1990] 2 AC 418 , 476, “[t]he Government may negotiate, conclude, construe, observe, breach, repudiate or terminate a treaty”.

Subject to any restrictions imposed by primary legislation, the general rule is that the power to make or unmake treaties is exercisable without legislative authority and that the exercise of that power is not reviewable by the courts - see Civil Service Unions case cited above, at pp 397-398. Lord Coleridge CJ said that the Queen acts “throughout the making of the treaty and in relation to each and every of its stipulations in her sovereign character, and by her own inherent authority” - Rustomjee v The Queen (1876) 2 QBD 69 , 74. This principle rests on the so-called dualist theory, which is based on the proposition that international law and domestic law operate in independent spheres. The prerogative power to make treaties depends on two related propositions. The first is that treaties between sovereign states have effect in international law and are not governed by the domestic law of any state. As Lord Kingsdown expressed it in Secretary of State in Council of India v Kamachee Boye Sahaba (1859) 13 Moo PCC 22, 75, treaties are “governed by other laws than those which municipal courts administer”. The second proposition is that, although they are binding on the United Kingdom in international law, treaties are not part of UK law and give rise to no legal rights or obligations in domestic law.

It is only on the basis of these two propositions that the exercise of the prerogative power to make and unmake treaties is consistent with the rule that ministers cannot alter UK domestic law. Thus, in Higgs v Minister of National Security [2000] 2 AC 228 , 241, Lord Hoffmann pointed out that the fact that treaties are not part of domestic law was the “corollary” of the Crown’s treaty-making power. In JH Rayner cited above, at p 500, Lord Oliver of Aylmerton put it thus:

“As a matter of the constitutional law of the United Kingdom, the Royal Prerogative, whilst it embraces the making of treaties, does not extend to altering the law or conferring rights upon individuals or depriving individuals of rights which they enjoy in domestic law without the intervention of Parliament. Treaties, as it is sometimes expressed, are not self-executing. Quite simply, a treaty is not part of English law unless and until it has been incorporated into the law by legislation. So far as individuals are concerned, it is res inter alios acta [ie something done between others], from which they cannot derive rights and by which they cannot be deprived of rights or subjected to obligations; and it is outside the purview of the court not only because it is made in the conduct of foreign relations, which are a prerogative of the Crown, but also because, as a source of rights and obligations, it is irrelevant.”

It can thus fairly be said that the dualist system is a necessary corollary of Parliamentary sovereignty, or, to put the point another way, it exists to protect Parliament not ministers. Professor Campbell McLachlan in Foreign Relations Law (2014), para 5.20, neatly summarises the position in the following way:

“If treaties have no effect within domestic law, Parliament’s legislative supremacy within its own polity is secure. If the executive must always seek the sanction of Parliament in the event that a proposed action on the international plane will require domestic implementation, parliamentary sovereignty is reinforced at the very point at which the legislative power is engaged.”

While ministers have in principle an unfettered power to make treaties which do not change domestic law, it had become fairly standard practice by the late 19th century for treaties to be laid before both Houses of Parliament at least 21 days before they were ratified, to enable Parliamentary objections to be heard. In 1924, following an indication by the previous government that it did not regard itself as bound by the practice, Arthur Ponsonby, the Parliamentary Under-Secretary of State for Foreign Affairs, assured the House of Commons that ministers would in future adhere to this practice, which became known as the Ponsonby Convention. The convention was superseded and formalised by section 20 of the Constitutional Reform and Governance Act 2010 . However, by virtue of section 23(1) of that Act , this section does not apply to new EU Treaties, because they are governed by the more specific statutory controls discussed in paras 28 and 29 above.

With that background, we turn to analyse the effect of the 1972 Act and the arguments as to whether, in the absence of prior authority from Parliament in the form of a statute, the giving of Notice by ministers would be ineffective under the United Kingdom’s constitutional requirements, as it would otherwise impermissibly result in a change in domestic law.

Many statutes give effect to treaties by prescribing the content of domestic law in the areas covered by them. The 1972 Act does this, but it does considerably more as well. It authorises a dynamic process by which, without further primary legislation (and, in some cases, even without any domestic legislation), EU law not only becomes a source of UK law, but actually takes precedence over all domestic sources of UK law, including statutes. This may sound rather dry or technical to many people, but in constitutional terms the effect of the 1972 Act was unprecedented. Indeed, it is fair to say that the legal consequences of the United Kingdom’s accession to the EEC were not fully appreciated by many lawyers until the Factortame litigation in the 1990s - see the House of Lords decisions in R v Secretary of State for Transport, Ex p Factortame Ltd (No 2) [1991] 1 AC 603 and (No 5) [2000] 1 AC 524 . Of course, consistently with the principle of Parliamentary sovereignty, this unprecedented state of affairs will only last so long as Parliament wishes: the 1972 Act can be repealed like any other statute. For that reason, we would not accept that the so-called fundamental rule of recognition (ie the fundamental rule by reference to which all other rules are validated) underlying UK laws has been varied by the 1972 Act or would be varied by its repeal.

In one sense, of course, it can be said that the 1972 Act is the source of EU law, in that, without that Act , EU law would have no domestic status. But in a more fundamental sense and, we consider, a more realistic sense, where EU law applies in the United Kingdom, it is the EU institutions which are the relevant source of that law. The legislative institutions of the EU can create or abrogate rules of law which will then apply domestically, without the specific sanction of any UK institution. It is true that the UK government and UK-elected members of the European Parliament participate in the EU legislative processes and can influence their outcome, but that does not diminish the point. Further, in the many areas of EU competence which are subject to majority decision, the approval of the United Kingdom is not required for its legislation to take effect domestically. It is also true that EU law enjoys its automatic and overriding effect only by virtue of the 1972 Act , and thus only while it remains in force. That point simply reflects the fact that Parliament was and remains sovereign: so, no new source of law could come into existence without Parliamentary sanction - and without being susceptible to being abrogated by Parliament. However, that in no way undermines our view that it is unrealistic to deny that, so long as that Act remains in force, the EU Treaties, EU legislation and the interpretations placed on these instruments by the Court of Justice are direct sources of UK law.

The 1972 Act did two things which are relevant to these appeals. First, it provided that rights, duties and rules derived from EU law should apply in the United Kingdom as part of its domestic law. Secondly, it provided for a new constitutional process for making law in the United Kingdom. These things are closely related, but they are legally and conceptually distinct. The content of the rights, duties and rules introduced into our domestic law as a result of the 1972 Act is exclusively a question of EU law. However, the constitutional processes by which the law of the United Kingdom is made is exclusively a question of domestic law.

Under the terms of the 1972 Act , EU law may take effect as part of the law of the United Kingdom in one of three ways. First, the EU Treaties themselves are directly applicable by virtue of section 2(1) . Some of the provisions of those Treaties create rights (and duties) which are directly applicable in the sense that they are enforceable in UK courts. Secondly, where the effect of the EU Treaties is that EU legislation is directly applicable in domestic law, section 2(1) provides that it is to have direct effect in the United Kingdom without the need for further domestic legislation. This applies to EU Regulations (which are directly applicable by virtue of article 288 of the TFEU). Thirdly, section 2(2) authorises the implementation of EU law by delegated legislation. This applies mainly to EU Directives, which are not, in general, directly applicable but are required (again by article 288) to be transposed into national law. While this is an international law obligation, failure of the United Kingdom to comply with it is justiciable in domestic courts, and some Directives may be enforced by individuals directly against national governments in domestic courts. Further, any serious breach by the UK Parliament, government or judiciary of any rule of EU law intended to confer individual rights will entitle any individual sustaining damage as a direct result to compensation from the UK government: Brasserie du Pêcheur SA v Germany; R v Secretary of State for Transport (Ex p Factortame Ltd) (No 4) (Joined Cases C-46/93 and C-48/93) [1996] QB 404 (provided that, where the breach consists in a court decision, the breach is not only serious but also manifest: Köbler v Austria ( Case C-224/01 ) [2004] QB 848 ).

Thus, EU law in EU Treaties and EU legislation will pass into UK law through the medium of section 2(1) or the implementation provisions of section 2(2) of the 1972 Act , so long as the United Kingdom is party to the EU Treaties. Similarly, so long as the United Kingdom is party to the EU Treaties, UK courts are obliged (i) to interpret EU Treaties, Regulations and Directives in accordance with decisions of the Court of Justice, (ii) to refer unclear points of EU law to the Court of Justice, and (iii) to interpret all domestic legislation, if at all possible, so as to comply with EU law (see Marleasing v La Comercial Internacional de Alimentacion SA ( Case C-106/89 ) [1990] ECR I-4135). And, so long as the United Kingdom is party to the EU Treaties, UK citizens are able to recover damages from the UK government in cases where a decision of one of the organs of the state based on a serious error of EU law has caused them loss.

In our view, then, although the 1972 Act gives effect to EU law, it is not itself the originating source of that law. It is, as was said on behalf of the Secretary of State echoing the illuminating analysis of Professor Finnis, the “conduit pipe” by which EU law is introduced into UK domestic law. So long as the 1972 Act remains in force, its effect is to constitute EU law an independent and overriding source of domestic law.

Section 18 of the 2011 Act , set out in para 30 above, was enacted in order to make it clear that the primacy of EU law over domestic legislation did not prevent it being repealed by domestic legislation. But that simply confirmed the position as it had been since the beginning of 1973. The primacy of EU law means that, unlike other rules of domestic law, EU law cannot be implicitly displaced by the mere enactment of legislation which is inconsistent with it. That is clear from the second part of section 2(4) of the 1972 Act and Factortame Ltd (No 2) [1991] 1 AC 603 . The issue was informatively discussed by Laws LJ in Thoburn v Sunderland City Council [2003] QB 151 , paras 37-47.

The 1972 Act accordingly has a constitutional character, as discussed by Laws LJ in Thoburn cited above, paras 58-59, and by Lord Reed and Lords Neuberger and Mance in in R (Buckinghamshire County Council) v Secretary of State for Transport [2014] 1 WLR 324 , paras 78 to 79 and 206 to 207 respectively. Following the coming into force of the 1972 Act , the normal rule is that any domestic legislation must be consistent with EU law. In such cases, EU law has primacy as a matter of domestic law, and legislation which is inconsistent with EU law from time to time is to that extent ineffective in law. However, legislation which alters the domestic constitutional status of EU institutions or of EU law is not constrained by the need to be consistent with EU law. In the case of such legislation, there is no question of EU law having primacy, so that such legislation will have domestic effect even if it infringes EU law (and that would be true whether or not the 1972 Act remained in force). That is because of the principle of Parliamentary sovereignty which is, as explained above, fundamental to the United Kingdom’s constitutional arrangements, and EU law can only enjoy a status in domestic law which that principle allows. It will therefore have that status only for as long as the 1972 Act continues to apply, and that, of course, can only be a matter for Parliament.

We should add that, for these reasons, we do not accept the suggestion that, as a source of law, EU law can properly be compared with, delegated legislation. The 1972 Act effectively operates as a partial transfer of law-making powers, or an assignment of legislative competences, by Parliament to the EU law-making institutions (so long as Parliament wills it), rather than a statutory delegation of the power to make ancillary regulations - even under a so-called Henry the Eighth clause, as explained in the Public Law Project case, cited above, paras 25 and 26. The 1972 Act cannot be said to constitute EU legislative institutions the delegates of Parliament: they make laws independently of Parliament, and indeed they were doing so before the 1972 Act was passed. If EU law had the same status in domestic law as delegated legislation, the Factortame litigation referred to above would have had a different outcome. A statutory provision which provides that legislative documents and decisions made by EU institutions should be an independent and pre-eminent source of UK law is thus quite different from a statutory provision which delegates to ministers and other organs of the executive the right to make regulations and the like. The exceptional nature of the effect of the 1972 Act is well illustrated by the passages quoted by Lord Reed in para 182 below from the decisions of the Court of Justice in Van Gend en Loos ( Case C-26/62 ) [1963] ECR 1, 12 and Costa v ENEL ( Case C-6/64 ) [1964] ECR 585, 593. They demonstrate that rules which would, as Lord Reed says, normally be incompatible with UK constitutional principles, became part of our constitutional arrangements as a result of the 1972 Act and the 1972 Accession Treaty for as long as the 1972 Act remains in force.

Rights capable of replication in UK law;

Rights derived by UK citizens from EU law in other member states;

Many current EU rights fall within the first category. They include, for instance, the rights of UK citizens to the benefit of employment protection such as the Working Time Directive, to equal treatment and to the protection of EU competition law, and the right of non-residents to the benefit of the “four freedoms” (free movement of people, goods and capital, and freedom to provide services). Some of these rights have already been embodied in UK law by domestic legislation pursuant to section 2(2) of the 1972 Act , and they will not cease to have effect upon the United Kingdom’s withdrawal from the European Union (unless the domestic legislation giving effect to them is repealed in accordance with the law), although the Court of Justice will no longer have any binding role in relation to their scope or interpretation. Other rights, arising under EU Regulations or directly under the EU Treaties, will cease to have effect upon withdrawal (save in relation to rights and liabilities already accrued), but many could be replicated in a new statute - eg the proposed Great Repeal Bill. But, as the Divisional Court pointed out, the need for such replication would only arise because withdrawal from the EU Treaties would have abrogated domestic rights created by the 1972 Act of effect, and again the Court of Justice would no longer have any binding role in relation to them.

The second category may appear to be irrelevant for present purposes as the rights within it arise from the incorporation of EU law into the law of other member states, and not from UK legislation. However, some rights falling within one category may be closely linked with rights falling within another category. For example, the rights under Council Regulation (EC) No 2201/2003, concerning jurisdiction and the recognition and enforcement of judgments in matrimonial matters and the matters of parental responsibility (known as “Brussels II Revised”), would be undermined if a domestic judgment governing the residence of a child could not be enforced outside the UK.

The rights in the third category will cease when the United Kingdom is no longer a member of the European Union, as they are by their very nature dependent on continued membership. They include the right to stand for selection or later election to the European Parliament, and the right to vote in European elections, as well as the right to invite the Commission to take regulatory action. However, they have the character of what Mr Eadie described as “club membership rights”, and are of a different nature from the other more “freestanding” rights in the first and second categories.

Given that it is clear that some rights in the first category will be lost on the United Kingdom withdrawing from the EU Treaties, it is unnecessary to consider whether, for the purpose of their present arguments, the applicants can rely on the loss of rights in the second and third categories. If they cannot succeed in their argument based on loss of rights in the first category, then invoking loss of rights in the other categories would not help them; and if they can succeed on the basis of loss of rights in the first category, they would not need to invoke loss of rights in the other categories.

While accepting that some rights will be lost on withdrawal from the EU Treaties, the Secretary of State’s case is that the loss of these rights in such circumstances is provided for, and has therefore effectively been sanctioned, by Parliament in the 1972 Act itself. In this connection, Mr Eadie pointed out that the 1972 Act does not simply incorporate the EU Treaties into UK law in the same way as, say, the Carriage of Goods by Sea Act 1971 incorporates the Hague Rules. By contrast, he said, section 2 of the 1972 Act is “ambulatory”: in other words, it gives effect to whatever may from time to time be the international obligations of the United Kingdom under or pursuant to the EU Treaties.

He pointed out that changes in EU law are brought into domestic law through the medium of section 2 of the 1972 Act and that, once the United Kingdom ceases to be bound by the EU Treaties, there will be no rights and remedies etc to which section 2(1) can apply, and no EU obligations which require delegated legislation under section 2(2) , and that the possibility of withdrawal from EU Treaties is therefore effectively provided for in the wording of section 2 . It is his case that, by providing that EU law rights, remedies etc “from time to time provided for by or under the Treaties” were “to be given effect or used in the United Kingdom”, section 2(1) accommodated the possibility of ministers withdrawing from the Treaties without Parliamentary authority. He also contended that it was self-evident that Parliament cannot have intended that the variable content of EU law should continue to be part of domestic law after the UK withdraws from the EU Treaties.

We accept the proposition that the ambit of the rights and remedies etc which are incorporated into domestic law through section 2 of the 1972 Act varies with the United Kingdom’s obligations from time to time under the EU Treaties. This proposition is reflected in the language of sub sections (1 ) and (2) of section 2 , which are quoted in paras 18 and 19 above. However, this proposition is also limited in nature. Thus, the provisions of new EU Treaties are not automatically brought into domestic law through section 2 : only once they have been statutorily added to “the Treaties” and “the EU Treaties” in section 1(2) can section 2 give effect to new EU Treaties. And section 2 can only apply to those rights and remedies which are capable of being “given legal effect or used” or “enjoyed” in the United Kingdom.

We also accept that Parliament cannot have intended that section 2 should continue to import the variable content of EU law into domestic law, or that the other consequences of the 1972 Act described in paras 62 to 64 above should continue to apply, after the United Kingdom had ceased to be bound by the EU Treaties. However, while acknowledging the force of Lord Reed’s powerful judgment, we do not accept that it follows from this that the 1972 Act either contemplates or accommodates the abrogation of EU law upon the United Kingdom’s withdrawal from the EU Treaties by prerogative act without prior Parliamentary authorisation. On the contrary: we consider that, by the 1972 Act , Parliament endorsed and gave effect to the United Kingdom’s membership of what is now the European Union under the EU Treaties in a way which is inconsistent with the future exercise by ministers of any prerogative power to withdraw from such Treaties.

In short, the fact that EU law will no longer be part of UK domestic law if the United Kingdom withdraws from the EU Treaties does not mean that Parliament contemplated or intended that ministers could cause the United Kingdom to withdraw from the EU Treaties without prior Parliamentary approval. There is a vital difference between changes in domestic law resulting from variations in the content of EU law arising from new EU legislation, and changes in domestic law resulting from withdrawal by the United Kingdom from the European Union. The former involves changes in EU law, which are then brought into domestic law through section 2 of the 1972 Act . The latter involves a unilateral action by the relevant constitutional bodies which effects a fundamental change in the constitutional arrangements of the United Kingdom.

So far as the interpretation of sub sections (1 ) and (2) of section 2 of the 1972 Act are concerned, any right available under EU law to the United Kingdom to withdraw from the EU Treaties does not, as Mr Eadie rightly accepted, fall within the subsection, as it is not one which would be given “legal effect or used in”, or which would be “enjoyed by the United Kingdom”. Further, the fact that section 2(1) envisages EU law rights and procedures applying “as in accordance with the Treaties” “from time to time” and “without further enactment” takes matters no further. Sub section 2(1) and (2) are concerned to ensure that the variable content of EU law as it stands from time to time, is given effect in domestic law, and there was no practical alternative to such an arrangement in a dualist system. However, it does not follow from this that prerogative powers may be used to withdraw from the Treaties and so cut off the source of EU law entirely.

One of the most fundamental functions of the constitution of any state is to identify the sources of its law. And, as explained in paras 61 to 66 above, the 1972 Act effectively constitutes EU law as an entirely new, independent and overriding source of domestic law, and the Court of Justice as a source of binding judicial decisions about its meaning. This proposition is indeed inherent in the Secretary of State’s metaphor of the 1972 Act as a conduit pipe by which EU law is brought into the domestic UK law. Upon the United Kingdom’s withdrawal from the European Union, EU law will cease to be a source of domestic law for the future (even if the Great Repeal Bill provides that some legal rules derived from it should remain in force or continue to apply to accrued rights and liabilities), decisions of the Court of Justice will (again depending on the precise terms of the Great Repeal Bill) be of no more than persuasive authority, and there will be no further references to that court from UK courts. Even those legal rules derived from EU law and transposed into UK law by domestic legislation will have a different status. They will no longer be paramount, but will be open to domestic repeal or amendment in ways that may be inconsistent with EU law.

Accordingly, the main difficulty with the Secretary of State’s argument is that it does not answer the objection based on the constitutional implications of withdrawal from the EU. As we have said, withdrawal is fundamentally different from variations in the content of EU law arising from further EU Treaties or legislation. A complete withdrawal represents a change which is different not just in degree but in kind from the abrogation of particular rights, duties or rules derived from EU law. It will constitute as significant a constitutional change as that which occurred when EU law was first incorporated in domestic law by the 1972 Act . And, if Notice is given, this change will occur irrespective of whether Parliament repeals the 1972 Act . It would be inconsistent with long-standing and fundamental principle for such a far-reaching change to the UK constitutional arrangements to be brought about by ministerial decision or ministerial action alone. All the more so when the source in question was brought into existence by Parliament through primary legislation, which gave that source an overriding supremacy in the hierarchy of domestic law sources.

The point may be illustrated by reference to the formula which Lord Reed uses to make the argument about the variable content of EU law. That formula is “All such [members of a specified category] as [satisfy a specified condition] shall be [dealt with in accordance with a specified requirement]”. In the present case, the “specified condition” is a continuing obligation under the EU Treaties, and it must be satisfied by EU laws, which are the relevant “members of [the] specified category”, in order for the “specified requirement”, namely that those EU laws are binding domestically, to apply. The membership of the specified category has a variable content which is contingent on the decisions of non-UK entities, and the contingency may change that content. That much may well be accommodated by the 1972 Act . But the very formula is not itself variable: it is a fixed rule of domestic law, enacted by Parliament. It is nothing to the point that there was, for UK purposes, no content in the specified category until the 1972 Accession Treaty was ratified (on the day after the 1972 Act received the royal assent). As mentioned in para 77 above, by the 1972 Act , Parliament endorsed and gave effect to the UK’s future membership of the European Union, and this became a fixed domestic starting point. The question is whether that domestic starting point, introduced by Parliament, can be set aside, or could have been intended to be set aside, by a decision of the UK executive without express Parliamentary authorisation. We cannot accept that a major change to UK constitutional arrangements can be achieved by ministers alone; it must be effected in the only way that the UK constitution recognises, namely by Parliamentary legislation. This conclusion appears to us to follow from the ordinary application of basic concepts of constitutional law to the present issue.

While the consequential loss of a source of law is a fundamental legal change which justifies the conclusion that prerogative powers cannot be invoked to withdraw from the EU Treaties, the Divisional Court was also right to hold that changes in domestic rights acquired through that source as summarised in para 70 above, represent another, albeit related, ground for justifying that conclusion. Indeed, the consequences of withdrawal go further than affecting rights acquired pursuant to section 2 of the 1972 Act , as explained in paras 62 to 64 above. More centrally, as explained in paras 76 to 79 above, section 2 of that Act envisages domestic law, and therefore rights of UK citizens, changing as EU law varies, but it does not envisage those rights changing as a result of ministers unilaterally deciding that the United Kingdom should withdraw from the EU Treaties.

Mr Eadie also argued that exercise of prerogative powers can change domestic law. While there are circumstances (as described in paras 52 and 53 above) where the exercise of prerogative powers can affect domestic legal rights, they plainly do not apply in the present case. The rights which flow through the conduit pipe of the 1972 Act are contingent on the possibility of their being removed or changed in accordance with decisions taken by EU institutions, as is recognised by the expression from “time to time” in section 2(1) . However, as implied in para 79 above, far from helping the Secretary of State’s case, the presence of those words in section 2 highlights their absence from the definition of “Treaties” and “EU Treaties” in section 1(2) . When one reads the two subsections together, the clear implication is that the continued existence of the conduit pipe, as opposed to the contents which flow through it, can be changed only if Parliament changes the law.

In the course of his attractively presented submissions, Mr Eadie sought to meet these points with the argument that the 1972 Act (as amended from time to time) effectively incorporates the EU Treaties, and that the applicants cannot point to any provision in the Act which states that the prerogative powers in relation to those treaties are to be abrogated. Given that there is nothing in the 1972 Act which expressly or by necessary implication abrogated ministers’ prerogative powers to withdraw from the Treaties to which it applied, he contended that it followed that the prerogative to withdraw from the EU Treaties was not precluded by the 1972 Act . In this connection, he relied on dicta in De Keyser cited above (including Lord Parmoor’s reference to “directly regulated by statute” in the passage quoted in para 48 above) which suggested that prerogative powers should not be treated as abrogated unless a statute expressly, or by necessary implication, provided for their abrogation. Mr Eadie also relied on R v Secretary of State for Foreign and Commonwealth Affairs, Ex p Rees-Mogg [1994] QB 552 , in which the Court of Appeal held that ministers could ratify a protocol to the TEU without Parliamentary approval. In the course of his reasons for rejecting an argument based on the proposition that prerogative powers could not be used to alter the law, Lloyd LJ at p 567H appears to have concluded that ministers’ prerogative powers exist generally in relation to the EU Treaties, apparently on the basis that a prerogative power can be fettered by statute only in express terms.

However, as explained above, the EU Treaties not only concern the international relations of the United Kingdom, they are a source of domestic law, and they are a source of domestic legal rights many of which are inextricably linked with domestic law from other sources. Accordingly, the Royal prerogative to make and unmake treaties, which operates wholly on the international plane, cannot be exercised in relation to the EU Treaties, at least in the absence of domestic sanction in appropriate statutory form. It follows that, rather than the Secretary of State being able to rely on the absence in the 1972 Act of any exclusion of the prerogative power to withdraw from the EU Treaties, the proper analysis is that, unless that Act positively created such a power in relation to those Treaties, it does not exist. And, once one rejects the contention that section 2 accommodates a ministerial power to withdraw from the EU Treaties (as to which see paras 79 and 84 above), it is plain that the 1972 Act did not create such a power of withdrawal, as the Secretary of State properly accepts.

We accept, of course, that it would have been open to Parliament to provide expressly that the constitutional arrangements and the EU rights introduced by the 1972 Act should themselves only prevail from time to time and for so long as the UK government did not decide otherwise, and in particular did not decide to withdraw from the EU Treaties. But we cannot accept that the 1972 Act did so provide. As Lord Hoffmann explained in R v Secretary of State for the Home Department, Ex p Simms [2000] 2 AC 115 , 131, “the principle of legality means that Parliament must squarely confront what it is doing and accept the political cost”, and so “[f]undamental rights cannot be overridden by general … words” in a statute, “because there is too great a risk that the full implications of their unqualified meaning may have passed unnoticed in the democratic process”. Had the Bill which became the 1972 Act spelled out that ministers would be free to withdraw the United Kingdom from the EU Treaties, the implications of what Parliament was being asked to endorse would have been clear, and the courts would have so decided. But we must take the legislation as it is, and we cannot accept that, in Part I of the 1972 Act , Parliament “squarely confront[ed]” the notion that it was clothing ministers with the far-reaching and anomalous right to use a treaty-making power to remove an important source of domestic law and important domestic rights.

In our judgment, far from indicating that ministers had the power to withdraw from the EU Treaties, the provisions of the 1972 Act , particularly when considered in the light of the unusual nature of those Treaties and the Act ’s unusual legislative history, support the contrary view. As the Divisional Court said, the long title of the 1972 Act stated that its purpose was to make provision in connection with the “enlargement” of what is now the European Union, which is not easy to reconcile with a prerogative power to achieve the opposite. Similarly, the side-note to section 2 , “General implementation of Treaties”, points away from a prerogative to terminate any implementation. In addition, there is the fact that the 1972 Act required ministers not to commit the United Kingdom to any new arrangement, whether it increased or decreased the potential volume and extent of EU law, without first being approved by Parliament - by statute in the case of a new EU Treaty and by an approved Order in Council in the case of a treaty ancillary to any existing EU Treaty. It would scarcely be compatible with those provisions if, in reliance on prerogative powers, ministers could unilaterally withdraw from the EU Treaties, thereby reducing the volume and extent of EU law which takes effect domestically to nil without the need for Parliamentary approval.

For these reasons, we disagree with Lloyd LJ’s conclusion in Rees-Mogg in so far as he held that ministers could exercise prerogative powers to withdraw from the EU Treaties. It is only right to add that his ultimate decision was nonetheless correct for the reason he gave on p 568, namely that ratification of the particular protocol in that case would not in any significant way alter domestic law.

The EU Treaties as implemented pursuant to the 1972 Act were and are unique in their legislative and constitutional implications. In 1972, for the first time in the history of the United Kingdom, a dynamic, international source of law was grafted onto, and above, the well-established existing sources of domestic law: Parliament and the courts. And, as explained in paras 13-15 above, before (i) signing and (ii) ratifying the 1972 Accession Treaty, ministers, acting internationally, waited for Parliament, acting domestically, (i) to give clear, if not legally binding, approval in the form of resolutions, and (ii) to enable the Treaty to be effective by passing the 1972 Act . Bearing in mind this unique history and the constitutional principle of Parliamentary sovereignty, it seems most improbable that those two parties had the intention or expectation that ministers, constitutionally the junior partner in that exercise, could subsequently remove the graft without formal appropriate sanction from the constitutionally senior partner in that exercise, Parliament.

The improbability of the Secretary of State’s case is reinforced by the point that, if, as he contends, prerogative powers could be invoked in relation to the EU Treaties despite the provisions of the 1972 Act , it would have been open to ministers to take such a course on or at any time after 2 January 1973 without authorisation by Parliament. It would also follow that ministers could have taken that course even if there had been no referendum or indeed, at least in theory, even if any referendum had resulted in a vote to remain. Those are implausible propositions.

To meet this criticism, it was suggested that, if ministers had invoked their prerogative powers to withdraw from the EU Treaties in such circumstances, their decision may have been judicially reviewable. That is rather a bold suggestion, given that it has always been considered that, because they only operate on the international plane, prerogative treaty-making powers are not subject to judicial review - see para 55 above. It was also suggested that it should not cause surprise if ministers could exercise prerogative powers to withdraw from the EU Treaties, as they would be accountable to Parliament for their actions. This seems to us to be a potentially controversial argument constitutionally. It would justify all sorts of powers being accorded to the executive, on the basis that ministers could always be called to account for their exercise of any power. There is a substantial difference between (i) ministers having a freely exercisable power to do something whose exercise may have to be subsequently explained to Parliament and (ii) ministers having no power to do that thing unless it is first accorded to them by Parliament. The major practical difference between the two categories, in a case such as this where the exercise of the power is irrevocable, is that the exercise of power in the first category pre-empts any Parliamentary action. When the power relates to an action of such importance to the UK constitution as withdrawing from the Treaties, it would clearly be appropriate for the power to be in the second category. The fact that ministers are free to issue a declaration of war without first obtaining the sanction of Parliament does not assist the Secretary of State’s case. Such a declaration, while plainly of fundamental significance in practice, does not change domestic laws or domestic sources of law, although it will lead to new laws - provided Parliament decides that it should.

Thus, the continued existence of the new source of law created by the 1972 Act , and the continued existence of the rights and other legal incidents which flow therefrom, cannot as a matter of UK law have depended on the fact that to date ministers have refrained from having recourse to the Royal prerogative to eliminate that source and those rights and other incidents.

The Secretary of State relied on the fact that it was inevitable that Parliament would be formally involved in the process of withdrawal from the European Union, in that primary legislation, not least the Great Repeal Bill referred to in para 34 above, would be required to enable the United Kingdom to complete its withdrawal in an orderly and coherent manner. That seems very likely indeed, but it misses the point. If ministers give Notice without Parliament having first authorised them to do so, the die will be cast before Parliament has become formally involved. To adapt Lord Pannick’s metaphor, the bullet will have left the gun before Parliament has accorded the necessary leave for the trigger to be pulled. The very fact that Parliament will have to pass legislation once the Notice is served and hits the target highlights the point that the giving of the Notice will change domestic law: otherwise there would be no need for new legislation.

It was also argued on behalf of the Secretary of State that, when ministers are participating in EU law-making processes and are therefore involved in making EU law, and hence domestic law, they are thereby exercising prerogative powers, and that the giving of Notice would be an equally legitimate exercise of those powers. We readily accept, without formally deciding, that ministerial activity in the EU law-making process is effected under the Royal prerogative. However, it does not follow from this that ministers should be entitled to exercise a prerogative power to leave the European Union. When taking part in EU decision-making, UK ministers are carrying out the very functions which were envisaged by Parliament when enacting the 1972 Act . Withdrawing from the EU Treaties involves ministers doing the opposite, namely, unilaterally dismantling the very system which they set up in a co-ordinated way with Parliament, as explained in paras 13 to 15 above. Consistently with this, article 16 of TEU stipulates that “a representative of each member state at ministerial level” can commit member states by voting on the European Council, whereas article 50 provides that withdrawal must be effected by a member state “in accordance with [its] constitutional requirements”.

It was further pointed out that unilateral actions by other member states could remove EU law-based rights enjoyed by EU nationals (including UK citizens) living in the United Kingdom - eg if another member state withdrew from the European Union. We agree, but cannot accept that it has any relevance to the present dispute, which concerns the domestic constitutional arrangements which apply if the UK government wishes to withdraw from the EU Treaties. The fact that it is inevitable that to the extent that they depend on a particular foreign government, EU rights can be abrogated by the withdrawal from EU Treaties by that foreign government gives no guidance as to what is required by the United Kingdom’s constitutional arrangements before ministers can cause the United Kingdom to withdraw from those Treaties.

Mr Eadie identified two instances which, he contended, showed that there were circumstances in which the UK government could withdraw from treaties without prior Parliamentary sanction, even if such withdrawal changed domestic law. The first was the United Kingdom’s withdrawal in 1972 from the European Free Trade Agreement, EFTA. That is of no assistance to the Secretary of State. For, in stark contrast with UK membership of the European Union as a result of the 1972 Act , no directly effective rights had been created as a result of UK membership of EFTA. Moreover, the decision to withdraw from EFTA was an inevitable corollary of joining the EEC, and the formal notice withdrawing from EFTA was only served after both Houses of Parliament had “approve[d]” the “decision of principle to join the European Communities” as explained in para 13 above; it was thus an aspect of the exercise which the Prime Minister and the Chancellor of the Duchy of Lancaster respectively described in the House of Commons in October 1971 and January 1972.

The second instance given by Mr Eadie was that of bilateral double taxation treaties (“DTTs”), which were entered into with other states by the UK government under section 788 of the Income and Corporation Taxes Act 1988 (“ section 788 ”), now replaced by section 2 of the Taxation (International and Other Provisions) Act 2010 (“TIOPA”). This point was hardly mentioned in the oral argument before us, perhaps because discussions in some of the articles referred to in para 11 above have shown that the DTTs are an unsatisfactory analogy. By section 788 and now by TIOPA, Parliament provided in primary legislation that arrangements agreed by ministers in a DTT at international level will have effect in national law, but only if those arrangements are specified in an Order in Council which is approved by the House of Commons. Thus, unlike EU law which becomes part of UK law automatically as a result of the 1972 Act , the arrangements under a DTT do not take effect automatically as a result of section 788 or, now, TIOPA, but only through a specific Order in Council which has to be approved by Parliament. The conduit pipe metaphor which applies to the 1972 Act in relation to EU law is inapposite for section 788 and TIOPA in relation to DTTs.

Before concluding on the effect of the 1972 Act , it is worth mentioning two points. First, eminent judges have taken it for granted that it is a matter for Parliament whether the United Kingdom withdraws from the EU Treaties. In Blackburn v Attorney General [1971] 1 WLR 1037 , 1040, Lord Denning MR said that “[i]f her Majesty’s Ministers sign this treaty and Parliament enacts provisions to implement it” he did “not envisage that Parliament would afterwards go back on it and try to withdraw from it”, but “if Parliament should do so” then the courts would consider it. In Macarthys Ltd v Smith [1981] ICR 785 , 789, Lord Denning (albeit in a dissenting judgment) made “a constitutional point”, and referred to the possibility of “our Parliament deliberately pass[ing] an Act with the intention of repudiating the Treaty”. In Pham v Secretary of State for the Home Department [2015] 1 WLR 1591 , para 80, having stated that “EU law [is] part of domestic law because Parliament has so willed”, Lord Mance said that “[t]he question how far Parliament has so willed is thus determined by construing the 1972 Act ”. In R (Shindler) v Chancellor of the Duchy of Lancaster [2016] 3 WLR 1196 , para 58, Lord Dyson MR said that “Parliament agreed to join the EU by exercising sovereign powers untrammelled by EU law and I think it would expect to be able to leave the EU in the exercise of the same untrammelled sovereign power”.

Secondly, if, as the Secretary of State has argued, it is legitimate to take account of the fact that Parliament will, of necessity, be involved in its legislative capacity as a result of UK withdrawal from the EU Treaties, it would militate in favour of, rather than against, the view that Parliament should have to sanction giving Notice. An inevitable consequence of withdrawing from the EU Treaties will be the need for a large amount of domestic legislation. There is thus a good pragmatic argument that such a burden should not be imposed on Parliament by exercise of prerogative powers and without prior Parliamentary authorisation. We do not rest our decision on that point, but it serves to emphasise the major constitutional change which withdrawal from the European Union will involve, and therefore the constitutional propriety of prior Parliamentary sanction for the process.

Accordingly, we consider that, in light of the terms and effect of the 1972 Act , and subject to considering the effect of subsequent legislation and events, the prerogative could not be invoked by ministers to justify giving Notice: ministers require the authority of primary legislation before they can take that course.

We turn, then, to deal with the impact of legislation and events after 1972.

With one exception, the legislation and events between 1973 and 2014 were relied on in argument by the Secretary of State rather than by the applicants. We will first discuss the Secretary of State’s points in this connection and we will then turn to the applicants’ point.

We start by addressing the fact that the EU Treaties contained no provision entitling a member state to withdraw at the time of the 1972 Act , and that such a provision, article 50, was introduced by the TFEU in 2008. Although its invocation will have the inevitable consequence which Lord Pannick described (as mentioned in para 36 above), article 50 operates only on the international plane, and is not therefore brought into UK law through section 2 of the 1972 Act , as explained in para 79 above. Accordingly, the Secretary of State can derive no domestic authority from the fact that the EU Treaties now include provision for unilateral withdrawal. In any event, article 50 only entitles a member state to withdraw from the EU Treaties “in accordance with its own constitutional requirements”, which returns one to the issue in the current proceedings.

It was suggested that, by incorporating the TFEU, including its introduction of article 50, into section 1(2) of the 1972 Act in 2008, it cannot have been the intention of Parliament to “strip” ministers of their ability to exercise their powers under article 50. That is not the issue. Nobody doubts but that, under the TFEU and the TEU, ministers can give Notice under article 50(2); the question we have to decide is whether they can do so under prerogative powers or only with Parliamentary authority.

So far as the 2008 Act and the 2011 Act are concerned, Mr Eadie rightly did not go so far as to suggest that they conferred power on ministers to withdraw if that power did not exist under the 1972 Act . More subtly, he submitted that these later statutes implicitly, but clearly, recognised the existence of the prerogative power to withdraw from the EU Treaties, unconstrained by Parliamentary control.

He pointed out that the two statutes specified in detail the prerogative powers which Parliament intended to control in relation to the EU Treaties, and that they did not include the power to withdraw from those treaties under article 50(2). That omission was said to be particularly striking because, as explained in para 29 above, the 2011 Act covered another aspect of article 50, as it required legislation and a referendum before ministers could vote in favour of a decision under article 50(3) to depart from the need for unanimity in any decision to extend the two-year period in the event of another member state seeking to withdraw from the EU Treaties. But it did not seek to control the giving of notice by ministers under article 50(2), for all its fundamental and irreversible consequences.

We do not accept this argument. The fact that a statute says nothing about a particular topic can rarely, if ever, justify inferring a fundamental change in the law. As explained in Ex p Simms [2000] 2 AC 115 , 131 cited in para 87 above, “[f]undamental rights cannot be overridden by general … words” in a statute, “because there is too great a risk that the full implications of their unqualified meaning may have passed unnoticed in the democratic process.” If this is true of general expressions in a statute it must a fortiori be a principle which applies to omissions in a statute.

Even if this principle admits of exceptions, they must be rare, and there is no justification for the view that the absence of any reference to article 50(2) in the 2008 and 2011 Acts is such an exception. Those statutes were not attempting to codify the legislative restrictions on the use of the prerogative in relation to the EU Treaties. The restrictions imposed by the two statutes were largely prompted by the fact that the TFEU had both increased the competences of the EU and included provisions which enabled EU institutions to short-circuit some of the EU’s decision-making processes by replacing some of the previous requirements for unanimity or consensus with majority voting or involvement of the European Parliament. (It is fair to add that the restrictions also applied to certain policy issues such as the inclusion of the UK in the Schengen area and the UK’s adoption of the Euro, but that does not undermine the point).

As explained in paras 5 and 6 of the Explanatory Notes to the 2011 Act , Part 1 of that Act was intended to impose specific restrictions, which in summary terms were as follows. It required “a referendum [to] be held before the UK could agree to an amendment” of TEU or TFEU, and “before the UK could agree to certain decisions already provided for by TEU and TFEU … if these would transfer power or competence from the UK to the EU”. Further, a referendum and “[i]n addition, … an Act of Parliament would be required before the UK could agree to a number of other specified decisions provided for in TEU and TFEU”. Also, “certain other decisions would require a motion to be agreed … in both Houses of Parliament before the UK could vote in favour of specified decisions in [EU institutions]”.

In other words, expressed in broad terms, Part 1 of the 2011 Act was aimed at preventing ministers, without prior Parliamentary approval (plus, in many cases prior approval in a referendum), from supporting any decisions made by the European Union or its institutions which would extend EU competences and the like, or which would dilute the effect of UK voting rights in the EU or any EU institutions. It cannot be inferred from the fact that it was thought necessary to deal with such issues that Parliament intended or assumed that there were no legal limits to the prerogative powers that ministers might exercise in other types of case. Part 1 of the 2011 Act was concerned with decisions of EU institutions in which ministers played a part, not with unilateral decisions of ministers. More broadly, the absence of any Parliamentary controls on article 50(2) in the 2011 Act is entirely consistent with the notion that Parliament assumed that ministers could not withdraw from the EU Treaties without a statute authorising that course - and that if and when Parliament had to consider the issue, it would decide whether and if so on what terms, if any, to give such authorisation.

If prerogative powers are curtailed by legislation, they may sometimes be reinstated by the repeal of that legislation, depending on the construction of the statutes in question. But if, as we have concluded, there never had been a prerogative power to withdraw from the EU Treaties without statutory authority, there is nothing to be curtailed or reinstated by later legislation. The prerogative power claimed by the Secretary of State can only be created by a subsequent statute if the express language of that statute unequivocally shows that the power was intended to be created - see per Lord Hobhouse of Woodborough in R (Morgan Grenfell & Co Ltd) v Special Commissioner of Income Tax [2003] 1 AC 563 , para 45. Mr Eadie was right to concede that, however one approaches them, the 2008 and 2011 Acts did not show that.

Mr Eadie further submitted that, rather than looking at the question whether ministers could give Notice without statutory authorisation in historical terms starting in 1972, it should be addressed by viewing the effect of the present state of the legislation as a whole, without regard to what the position might have been at some earlier stage. We do not agree. A statute cannot normally be interpreted by reference to a later statute, save in so far as the later statute intends to amend the earlier statute or the two statutes are in pari materia , ie they are given a collective title, are required to be construed as one, have identical short titles, or “deal with the same subject matter on similar lines” - see Bennion on Statutory Interpretation (6th ed, 2013) section 28(13). None of these tests can possibly be said to be satisfied by the 2008 Act or the 2011 Act in relation to the 1972 Act , not least because the later statutes are concerned with a different issue from the 1972 Act . In any event, even if the two later statutes were in pari materia with the 1972 Act , for the reasons given in paras 110 to 112 above we do not consider that they would together yield the interpretation for which the Secretary of State contends.

The one feature of the post-1972 history on which the applicants relied was the effect of the 2002 Act . As explained in para 27 above, that Act gave most people of the United Kingdom the right to vote in elections for MEPs, and (albeit by inference) the right to stand for election as an MEP. On the face of it, these are free-standing rights outside the ambit of the 1972 Act , in that they are domestically granted in primary legislation passed by Parliament. The Secretary of State cannot argue that these rights are in any sense ambulatory. And they are rights which will inevitably be lost if the United Kingdom withdraws from the EU Treaties and ceases being a member of the European Union.

There is therefore some force in the argument that, even if formal Parliamentary sanction to the giving of Notice was not needed on the grounds discussed in paras 74 to 101 above, it would nonetheless be needed because withdrawal from the EU Treaties would deprive UK citizens of the rights given them by Parliament through the 2002 Act . However, there is also force in the Secretary of State’s response that the rights given by the 2002 Act are simply rights of institutional participation which are contingent on continued UK membership of the European Union. The same sort of arguments might perhaps arise in relation to statutory provisions such as section 4(2) of the Communications Act 2003 , which requires OFCOM, the UK telecommunications regulator, to carry out its statutory functions “in accordance with the six Community requirements”, which are set out in the ensuing subsections and give effect to, and are mandated by, an EU Directive. Given our conclusion that, in the light of the terms and effect of the 1972 Act , ministers cannot give Notice without prior sanction from the UK Parliament, we can limit ourselves to saying that we consider that the arguments based on the 2002 Act do nothing to undermine and may be regarded as reinforcing that conclusion.

We turn to the 2015 Act and the ensuing referendum. The Attorney General submitted that the traditional view as to the limits of prerogative power should not apply to a ministerial decision authorised by a majority of the members of the electorate who vote in a referendum provided for by Parliament. In effect, he said that, even though it was Parliament which required the referendum, the response to the referendum result should be a matter for ministers, and that it should not be constrained by the legal limitations which would have applied in the absence of the referendum.

The referendum is a relatively new feature of UK constitutional practice. There have been three national referendums: on EEC membership in 1975, on the Parliamentary election voting system in 2011 and on EU membership in 2016. There have also been referendums about devolution in Scotland, Wales and Northern Ireland and about independence in Scotland. In 2000, it was considered worth having a legislative framework for the conduct of referendums “held in pursuance of any provision made by or under an Act of Parliament” - see Part VII of the Political Parties, Elections and Referendums Act.

The effect of any particular referendum must depend on the terms of the statute which authorises it. Further, legislation authorising a referendum more often than not has provided for the consequences on the result. Thus, the authorising statute may enact a change in the law subject to the proviso that it is not to come into effect unless approved by a majority in the referendum. The Scotland Act 1978 provided for devolution, but stipulated that the minister should bring the Act into force if there was a specified majority in a referendum, and if there was not he was required to lay an order repealing the Act . The Parliamentary Voting System and Constituencies Act 2011 had a provision requiring the alternative vote system to be adopted in Parliamentary elections, but by section 8 stated that the minister should bring this provision into force if it was approved in a referendum, but, if it was not, he should repeal it. Section 1 of the Northern Ireland Act 1998 (“the NI Act”) provided that if a referendum were to result in a majority for the province to become part of a united Ireland, the Secretary of State should lay appropriate proposals before Parliament.

All these statutes stipulated what should happen in response to the referendum result, and what changes in the law were to follow, and how they were to be effected. The same is true of the provisions in Part 1 of the 2011 Act . By contrast, neither the 1975 Act nor the 2015 Act, which authorised referendums about membership of the European Community or European Union, made provision for any consequences of either possible outcome. They provided only that the referendum should be held, and they did so in substantially identical terms. The way in which the proposed referendum was described in public statements by ministers, however, differed in the two cases. The 1975 referendum was described by ministers as advisory, whereas the 2016 referendum was described as advisory by some ministers and as decisive by others, but nothing hangs on that for present purposes. Whether or not they are clear and consistent, such public observations, wherever they are made, are not law: they are statements of political intention. Further, such statements are, at least normally, made by ministers on behalf of the UK government, not on behalf of Parliament.

It was suggested on behalf of the Secretary of State that, having referred the question whether to leave or remain to the electorate, Parliament cannot have intended that, upon the electorate voting to leave, the same question would be referred straight back to it. There are two problems with this argument. The first is that it assumes what it seeks to prove, namely that the referendum was intended by Parliament to have a legal effect as well as a political effect. The second problem is that the notion that Parliament would not envisage both a referendum and legislation being required to approve the same step is falsified by sections 2 , 3 and 6 of the 2011 Act , which, as the Explanatory Notes (quoted in para 111 above) acknowledge, required just that - albeit in the more elegant way of stipulating for legislation whose effectiveness was conditional upon a concurring vote in a referendum.

Where, as in this case, implementation of a referendum result requires a change in the law of the land, and statute has not provided for that change, the change in the law must be made in the only way in which the UK constitution permits, namely through Parliamentary legislation.

What form such legislation should take is entirely a matter for Parliament. But, in the light of a point made in oral argument, it is right to add that the fact that Parliament may decide to content itself with a very brief statute is nothing to the point. There is no equivalence between the constitutional importance of a statute, or any other document, and its length or complexity. A notice under article 50(2) could no doubt be very short indeed, but that would not undermine its momentous significance. The essential point is that, if, as we consider, what would otherwise be a prerogative act would result in a change in domestic law, the act can only lawfully be carried out with the sanction of primary legislation enacted by the Queen in Parliament.

This is why the Secretary of State rightly accepted that the resolution of the House of Commons on 7 December 2016, calling on ministers to give Notice by 31 March 2017, cannot affect the legal issues before this court. A resolution of the House of Commons is an important political act. No doubt, it makes it politically more likely that any necessary legislation enabling ministers to give Notice will be enacted. But if, as we have concluded, ministers cannot give Notice by the exercise of prerogative powers, only legislation which is embodied in a statute will do. A resolution of the House of Commons is not legislation.

Thus, the referendum of 2016 did not change the law in a way which would allow ministers to withdraw the United Kingdom from the European Union without legislation. But that in no way means that it is devoid of effect. It means that, unless and until acted on by Parliament, its force is political rather than legal. It has already shown itself to be of great political significance.

It is instructive to see how the issue was addressed in ministers’ response to the 12th Report of Session 2009-10 of the House of Lords Select Committee on the Constitution ( Referendums in the United Kingdom ). The Committee included the following recommendation in para 197:

“[B]ecause of the sovereignty of Parliament, referendums cannot be legally binding in the UK, and are therefore advisory. However, it would be difficult for Parliament to ignore a decisive expression of public opinion.”

Does any provision of the NI Act, read together with the Belfast Agreement and the British-Irish Agreement, have the effect that primary legislation is required before Notice can be given?

If the answer is “yes”, is the consent of the Northern Ireland Assembly required before the relevant legislation is enacted?

Does section 75 of the NI Act prevent exercise of the power to give Notice in the absence of compliance by the Northern Ireland Office with its obligations under that section?

Does the giving of Notice without the consent of the people of Northern Ireland impede the operation of section 1 of the NI Act?

Following the hearing, our attention was drawn to the decision of the Northern Irish Court of Appeal in Lee v McArthur and Ashers Baking Co Ltd (No 2) handed down on 22 December 2016. That decision suggests that the High Court may not have had jurisdiction to have made the reference in these proceedings as sought by the Attorney General for Northern Ireland. Given that the issues raised in that reference were fully debated, and that no party to these proceedings has sought belatedly to rely on the decision of the Court of Appeal, we think it appropriate to deal with the reference.

The NI Act is the product of the Belfast Agreement and the British-Irish Agreement, and is a very important step in the programme designed to achieve reconciliation of the communities of Northern Ireland. It has established institutions and arrangements which are intended to address the unique political history of the province and the island of Ireland. Yet there is also a relevant commonality in the devolution settlements in Northern Ireland, Scotland and Wales (i) in the statutory constraint on the executive and legislative competence of the devolved governments and legislatures that they must not act in breach of EU law (“the EU constraints”); and (ii) in the operation of the Sewel Convention. (The EU constraints are to be found in sections 29(2) (d), 54 and 57(2) of the Scotland Act 1998 ; sections 108(6) (c) and 80(8) of the Government of Wales Act 2006 ; and sections 6(2) (d) and 24(1) of the NI Act).

Because we have concluded that primary legislation is required to authorise the giving of Notice, the third question is superseded. The first question is for a similar reason less significant than it otherwise might have been but we address it briefly. When enacting the EU constraints in the NI Act and the other devolution Acts, Parliament proceeded on the assumption that the United Kingdom would be a member of the European Union. That assumption is consistent with the view that Parliament would determine whether the United Kingdom would remain a member of the European Union. But, in imposing the EU constraints and empowering the devolved institutions to observe and implement EU law, the devolution legislation did not go further and require the United Kingdom to remain a member of the European Union. Within the United Kingdom, relations with the European Union, like other matters of foreign affairs, are reserved or excepted in the cases of Scotland and Northern Ireland, and are not devolved in the case of Wales - see section 30(1) of, and paragraph 7(1) of Schedule 5 to, the Scotland Act 1998 ; section 108(4) of, and Part 1 of Schedule 7 to, the Government of Wales Act 2006 ; and section 4(1) of, and paragraph 3 of Schedule 2 to, the NI Act.

Accordingly, the devolved legislatures do not have a parallel legislative competence in relation to withdrawal from the European Union. The EU constraints are a means by which the UK Parliament and government make sure that the devolved democratic institutions do not place the United Kingdom in breach of its EU law obligations. The removal of the EU constraints on withdrawal from the EU Treaties will alter the competence of the devolved institutions unless new legislative constraints are introduced. In the absence of such new restraints, withdrawal from the EU will enhance the devolved competence. We consider the effect of the alteration of competence in our discussion of the Sewel Convention in paras 136 to 151 below.

Mr Scoffield QC, who appeared for Mr Agnew, is unquestionably right, however, to claim that the NI Act conferred rights on the citizens of Northern Ireland. Sections 6(2) (d) and 24(1), in imposing the EU constraints, have endowed the people of Northern Ireland with the right to challenge actions of the Executive or the Assembly on the basis that they are in breach of EU law. A recent example of the exercise of such a right is found in the case of Re JR65’s Application [2016] NICA 20, where the lifetime ban on men who have had sex with other men from giving blood in Northern Ireland was challenged as being contrary to EU law.

As already explained, it is normally impermissible for statutory rights to be removed by the exercise of prerogative powers in the international sphere. It would accordingly be incongruous if constraints imposed on the legislative competence of the devolved administrations by specific statutory provisions were to be removed, thereby enlarging that competence, other than by statute. A related incongruity arises by virtue of the fact that observance and implementation of EU obligations are a transferred matter and therefore the responsibility of the devolved administration in Northern Ireland. The removal of a responsibility imposed by Parliament by ministerial use of prerogative powers might also be considered a constitutional anomaly. In light of our conclusion that a statute is required to authorise the decision to withdraw from the European Union, and therefore the giving of Notice, it is not necessary to reach a definitive view on the first referred question. The EU constraints and the provisions empowering the implementation of EU law are certainly consistent with our interpretation of the 1972 Act but we refrain from deciding whether they impose a discrete requirement for Parliamentary legislation.

Section 75(1) of the NI Act obliges a public authority in carrying out its functions in relation to Northern Ireland to “have due regard to the need to promote equality of opportunity”. By section 75(2), this duty includes an obligation to have regard to the desirability of promoting good relations between persons of different religious belief, political persuasion or radical group. Section 75(3) defines “public authority” for the purpose of the section and, unlike section 76(7), does not include within the definition a Minister of the Crown. Thus, the Secretary of State does not fall within its ambit. Further, in our view, and in agreement with the Attorney General for Northern Ireland, the decision to withdraw from the European Union and to give Notice is not a function carried out by the Secretary of State for Northern Ireland in relation to Northern Ireland within the meaning of section 75. Because we have held that there is no prerogative power to give Notice, the fourth question is superseded. But in so far as the Secretary of State may have a role in the measures taken by the UK Parliament to give Notice, we are satisfied that section 75 imposes no obligation on him in that context.

We also answer the fifth question in the negative. Section 1 of the NI Act is headed “Status of Northern Ireland” and it provides:

“(1) It is hereby declared that Northern Ireland in its entirety remains part of the United Kingdom and shall not cease to be so without the consent of a majority of the people of Northern Ireland voting in a poll held for the purposes of this section in accordance with Schedule 1.

But if the wish expressed by a majority in such a poll is that Northern Ireland should cease to be part of the United Kingdom and form part of a united Ireland, the Secretary of State shall lay before Parliament such proposals to give effect to that wish as may be agreed between Her Majesty’s Government in the United Kingdom and the Government of Ireland.”

In our view, this important provision, which arose out of the Belfast Agreement, gave the people of Northern Ireland the right to determine whether to remain part of the United Kingdom or to become part of a united Ireland. It neither regulated any other change in the constitutional status of Northern Ireland nor required the consent of a majority of the people of Northern Ireland to the withdrawal of the United Kingdom from the European Union. Contrary to the submission of Mr Lavery QC for Mr McCord, this section cannot support any legitimate expectation to that effect.

That leaves the second question, which raises in substance the application of the Sewel Convention. The convention was adopted as a means of establishing cooperative relationships between the UK Parliament and the devolved institutions, where there were overlapping legislative competences. In each of the devolution settlements the UK Parliament has preserved its right to legislate on matters which are within the competence of the devolved legislature. Section 5 of the NI Act empowers the Northern Ireland Assembly to make laws, but sub section (6 ) states that “[t]his section does not affect the power of the Parliament of the United Kingdom to make laws for Northern Ireland”. Section 28(7) of the Scotland Act 1998 provides that the section empowering the Scottish Parliament to make laws: “does not affect the power of the Parliament of the United Kingdom to make laws for Scotland”. Substantially identical provision is made for Wales in section 107(5) of the Government of Wales Act 2006 .

The practical benefits of achieving harmony between legislatures in areas of competing competence, of avoiding duplication of effort, of enabling the UK Parliament to make UK-wide legislation where appropriate, such as establishing a single UK implementing body, and of avoiding any risk of legal challenge to the vires of the devolved legislatures were recognised from an early date in the devolution process. The convention takes its name from Lord Sewel, the Minister of State in the Scotland Office in the House of Lords who was responsible for the progress of the Scotland Bill in 1998. In a debate in the House of Lords on the clause which is now section 28 of the Scotland Act 1998 , he stated in July 1998 that, while the devolution of legislative competence did not affect the ability of the UK Parliament to legislate for Scotland, “we would expect a convention to be established that Westminster would not normally legislate with regard to devolved matters in Scotland without the consent of the Scottish Parliament”. That expectation has been fulfilled.

The convention was embodied in a Memorandum of Understanding between the UK government and the devolved governments originally in December 2001 (Cm 5240). Para 14 of the current Memorandum of Understanding, which was published in October 2013, states:

“The UK Government will proceed in accordance with the convention that the UK Parliament would not normally legislate with regard to devolved matters except with the agreement of the devolved legislature. The devolved administrations will be responsible for seeking such agreement as may be required for this purpose on an approach from the UK Government.”

Thus, the UK government undertook not to seek or support relevant legislation in the UK Parliament without the prior consent of the devolved legislature. That consent is given by a legislative consent motion which the devolved government introduces into the legislature. Para 2 of the Memorandum of Understanding stated that it was a statement of political intent and that it did not create legal obligations.

Over time, devolved legislatures have passed legislative consent motions not only when the UK Parliament has legislated on matters which fall within the legislative competence of a devolved legislature, but also when the UK Parliament has enacted provisions that directly alter the legislative competence of a devolved legislature or amend the executive competence of devolved administrations. Thus, as the Lord Advocate showed in a helpful schedule, legislative consent motions were passed by the Scottish Parliament before the enactment of both the Scotland Act 2012 and the Scotland Act 2016 . Similarly, the Welsh Assembly passed a legislative consent motion in relation to the Wales Act 2014 , and in November 2016 the Welsh government laid a legislative consent motion before the Assembly in relation to the current Wales Bill 2016. But legislation which implements changes to the competences of EU institutions and thereby affects devolved competences, such as the 2008 Act which incorporated the Treaty of Lisbon amending the TEU and the TFEU into section 1(2) of the 1972 Act , has not been the subject of legislative consent motions in any devolved legislature.

Before addressing the more recent legislative recognition of the convention, it is necessary to consider the role of the courts in relation to constitutional conventions. It is well established that the courts of law cannot enforce a political convention. In Re Resolution to Amend the Constitution [1981] 1 SCR 753, the Supreme Court of Canada addressed the nature of political conventions. In the majority judgment the Chief Justice (Laskin) and Dickson, Beetz, Estey, McIntyre, Chouinard and Lamer JJ stated at pp 774 to 775:

“The very nature of a convention, as political in inception and as depending on a consistent course of political recognition by those for whose benefit and to whose detriment (if any) the convention developed over a considerable period of time is inconsistent with its legal enforcement.”

In a dissenting judgment on one of the questions before the court, the Chief Justice and Estey and MacIntyre JJ developed their consideration of conventions at p 853:

“[A] fundamental difference between the legal, that is the statutory and common law rules of the constitution, and the conventional rules is that, while a breach of the legal rules, whether of statutory or common law nature, has a legal consequence in that it will be restrained by the courts, no such sanction exists for breach or non-observance of the conventional rules. The observance of constitutional conventions depends upon the acceptance of the obligation of conformance by the actors deemed to be bound thereby. When this consideration is insufficient to compel observance no court may enforce the convention by legal action. The sanction for non-observance of a convention is political in that disregard of a convention may lead to political defeat, to loss of office, or to other political consequences, but will not engage the attention of the courts which are limited to matters of law alone. Courts, however, may recognise the existence of conventions …”

Martland, Ritchie, Dickson, Beetz, Chouinard and Lamer JJ made the same point at pp 882 to 883:

“It is because the sanctions of convention rest with institutions of government other than courts … or with public opinion and ultimately, with the electorate, that it is generally said that they are political.”

Attempts to enforce political conventions in the courts have failed. Thus in Madzimbamuto v Lardner-Burke [1969] 1 AC 645 , the Judicial Committee of the Privy Council had to consider a submission that legal effect should be given to the convention which applied at that time that the UK Parliament would not legislate without the consent of the government of Southern Rhodesia on matters within the competence of the Legislative Assembly. In its judgment delivered by Lord Reid the Board stated at p 723 that:

“That is a very important convention but it had no legal effect in limiting the legal power of Parliament. It is often said that it would be unconstitutional for the UK Parliament to do certain things, meaning that the moral, political and other reasons against doing them are so strong that most people would regard it as highly improper if Parliament did these things. But that does not mean that it is beyond the power of Parliament to do such things. If Parliament chose to do any of them the courts could not hold the Act of Parliament invalid.”

While the UK government and the devolved executives have agreed the mechanisms for implementing the convention in the Memorandum of Understanding, the convention operates as a political restriction on the activity of the UK Parliament. Article 9 of the Bill of Rights, which provides that “Proceedings in Parliament ought not to be impeached or questioned in any Court or Place out of Parliament”, provides a further reason why the courts cannot adjudicate on the operation of this convention.

Judges therefore are neither the parents nor the guardians of political conventions; they are merely observers. As such, they can recognise the operation of a political convention in the context of deciding a legal question (as in the Crossman diaries case - Attorney General v Jonathan Cape Ltd [1976] 1 QB 752 ), but they cannot give legal rulings on its operation or scope, because those matters are determined within the political world. As Professor Colin Munro has stated, “the validity of conventions cannot be the subject of proceedings in a court of law” - (1975) 91 LQR 218, 228.

The evolving nature of devolution has resulted in the Sewel Convention also receiving statutory recognition through section 2 of the Scotland Act 2016 , which inserted sub- section (8 ) into section 28 of the Scotland Act 1998 (which empowers the Scottish Parliament to make laws). Thus sub sections (7 ) and (8) now state:

“(7) This section does not affect the power of the Parliament of the United Kingdom to make laws for Scotland.

But it is recognised that the Parliament of the United Kingdom will not normally legislate with regard to devolved matters without the consent of the Scottish Parliament.”

As the Advocate General submitted, by such provisions, the UK Parliament is not seeking to convert the Sewel Convention into a rule which can be interpreted, let alone enforced, by the courts; rather, it is recognising the convention for what it is, namely a political convention, and is effectively declaring that it is a permanent feature of the relevant devolution settlement. That follows from the nature of the content, and is acknowledged by the words (“it is recognised” and “will not normally”), of the relevant subsection. We would have expected UK Parliament to have used other words if it were seeking to convert a convention into a legal rule justiciable by the courts.

In the Scotland Act 2016 , the recognition of the Sewel Convention occurs alongside the provision in section 1 of that Act . That section, by inserting section 63 A into the Scotland Act 1998 , makes the Scottish Parliament and the Scottish government a permanent part of the United Kingdom’s constitutional arrangements, signifies the commitment of the UK Parliament and government to those devolved institutions, and declares that those institutions are not to be abolished except on the basis of a decision of the people of Scotland voting in a referendum. This context supports our view that the purpose of the legislative recognition of the convention was to entrench it as a convention.

The Lord Advocate and the Counsel General for Wales were correct to acknowledge that the Scottish Parliament and the Welsh Assembly did not have a legal veto on the United Kingdom’s withdrawal from the European Union. Nor in our view has the Northern Ireland Assembly. Therefore, our answer to the second question in para 126 above is that the consent of the Northern Ireland Assembly is not a legal requirement before the relevant Act of the UK Parliament is passed.

In reaching this conclusion we do not underestimate the importance of constitutional conventions, some of which play a fundamental role in the operation of our constitution. The Sewel Convention has an important role in facilitating harmonious relationships between the UK Parliament and the devolved legislatures. But the policing of its scope and the manner of its operation does not lie within the constitutional remit of the judiciary, which is to protect the rule of law.

Accordingly, (i) we dismiss the Secretary of State’s appeal against the decision of the English and Welsh Divisional Court, (ii) we invite the parties to the reference from the Northern Irish Court of Appeal to agree or, failing agreement, to make written submissions as to the order to be made on the appeal from that Court, and (iii) we answer the second and fifth questions referred by the courts of Northern Ireland as indicated respectively in paras 150 and 134 above, and we do not answer the first, third and fourth questions as they have been superseded.

Article 50 of the Treaty of European Union (“TEU”) provides:

“1. Any member state may decide to withdraw from the Union in accordance with its own constitutional requirements.

A member state which decides to withdraw shall notify the European Council of its intention. In the light of the guidelines provided by the European Council, the Union shall negotiate and conclude an agreement with that state, setting out the arrangements for its withdrawal, taking account of the framework for its future relationship with the Union …

The Treaties shall cease to apply to the state in question from the date of entry into force of the withdrawal agreement or, failing that, two years after the notification referred to in paragraph 2, unless the European Council, in agreement with the member state concerned, unanimously decides to extend this period …”

The cases before the court arise from disputes as to the “constitutional requirements” which govern a decision by the United Kingdom to withdraw from the European Union under article 50(1): a decision which must be taken before notification can be given under article 50(2). In the case brought by Mrs Miller and Mr Dos Santos (whom I shall refer to as the Miller claimants), the Miller claimants maintain that the Crown cannot lawfully give notification under article 50(2) unless an Act of Parliament authorises it to do so. The Secretary of State for Exiting the European Union, on the other hand, maintains that the decision is one which can lawfully be taken by the Crown in the exercise of prerogative powers. The Divisional Court decided the case in favour of the Miller claimants, and the case now comes before this court as an appeal against that decision.

A number of interested parties and interveners have taken part in the Miller appeal. They include the Lord Advocate and the Counsel General for Wales, who as well as presenting arguments in support of those advanced by the Miller claimants, have also argued that, in the event that an Act of Parliament is required, the consent of the Scottish Parliament and the National Assembly for Wales is also required, in accordance with a convention known as the Sewel Convention.

Two other cases are also before the court. In the first, an application for leave to apply for judicial review brought by Mr Agnew and others, a number of devolution issues have been referred to this court by the High Court of Northern Ireland. Put shortly, the court is asked to decide whether provisions of the Northern Ireland Act 1998 (“the Northern Ireland Act”) have the effect that an Act of Parliament is required before notification is given under article 50(2); if so, whether the consent of the Northern Ireland Assembly is required before such an Act of Parliament is enacted, in accordance with the Sewel Convention; and, in any event, whether the Northern Ireland Act prevents or constrains the exercise of the power to give notice.

In the second case, an application for leave to apply for judicial review brought by Mr McCord, another devolution issue has been referred to this court by the Court of Appeal of Northern Ireland. The court is asked to decide whether the giving of notification under article 50(2) in the exercise of prerogative powers, without the consent of the people of Northern Ireland, would impede the operation of section 1 of the Northern Ireland Act, which provides that Northern Ireland shall not cease to be part of the United Kingdom without the consent of a majority of the people of Northern Ireland.

I shall begin by considering the Miller appeal.

Each side of the argument in the Miller appeal is based on a principle of the British constitution. Counsel on each side cited a library’s worth of authority, but I need mention only a few of the most important cases, as the essence of the relevant principles is clear and well-known. The Secretary of State relies on the principle that, as a matter of law, the conduct of the UK’s foreign relations falls within the prerogative power of the Crown, advised by its Ministers. This prerogative power includes the power to negotiate international treaties, to amend them, and to withdraw from them. The exercise of that treaty-making power is not justiciable by the courts, unless statute has made it so. As Lord Oliver of Aylmerton said in the Tin Council case ( JH Rayner (Mincing Lane) Ltd v Department of Trade and Industry [1990] 2 AC 418 , 499:

“On the domestic plane, the power of the Crown to conclude treaties with other sovereign states is an exercise of the Royal Prerogative, the validity of which cannot be challenged in municipal law: see Blackburn v Attorney General [1971] 1 WLR 1037 . The Sovereign acts ‘throughout the making of the treaty and in relation to each and every of its stipulations in her sovereign character, and by her own inherent authority; and, as in making the treaty, so in performing the treaty, she is beyond the control of municipal law, and her acts are not to be examined in her own courts:’ Rustomjee v The Queen (1876) 2 QBD 69 , 74, per Lord Coleridge CJ.”

The compelling practical reasons for recognising this prerogative power to manage international relations were identified by Blackstone:

“This is wisely placed in a single hand by the British constitution, for the sake of unanimity, strength, and despatch. Were it placed in many hands, it would be subject to many wills: many wills, if disunited and drawing different ways, create weakness in a government; and to unite those several wills, and reduce them to one, is a work of more time and delay than the exigencies of state will afford.” ( Commentaries on the Laws of England (1765-1769) , Book I, Chapter 7, “Of the King’s Prerogative”)

Confiding foreign affairs to the Crown, in the exercise of the prerogative, does not, however, secure their effective conduct at the expense of democratic accountability. Ministers of the Crown are politically accountable to Parliament for the manner in which this prerogative power is exercised, and it is therefore open to Parliament to require its exercise to be debated and even to be authorised by a resolution or legislation: as it has done, for example, in relation to the ratification of certain treaties under the European Union Amendment Act 2008, the Constitutional Reform and Governance Act 2010 and the European Union Act 2011 . The Crown can, in addition, seek Parliamentary approval before exercising the prerogative power if it so chooses. There is however no legal requirement for the Crown to seek Parliamentary authorisation for the exercise of the power, except to the extent that Parliament has so provided by statute: that follows from the general principle set out in Blackburn v Attorney General and the Tin Council case. Since there is no statute which requires the decision under article 50(1) to be taken by Parliament, it follows that it can lawfully be taken by the Crown, in the exercise of the prerogative. There is therefore no legal requirement for an Act of Parliament to authorise the giving of notification under article 50(2). So runs the Secretary of State’s argument.

In support of this argument, the Secretary of State points out that there has been considerable Parliamentary scrutiny of Ministers’ conduct and their plans in relation to article 50. That scrutiny has included inquiries by the House of Commons Select Committee on Exiting the EU and by the House of Lords European Union Committee, as well as Parliamentary questions and debates. The latter have included a debate in the House of Commons on 7 December 2016, following which the following motion was agreed:

“That this House recognises that leaving the EU is the defining issue facing the UK; notes the resolution on parliamentary scrutiny of the UK leaving the EU agreed by the House on 12 October 2016; recognises that it is Parliament’s responsibility to properly scrutinise the Government while respecting the decision of the British people to leave the European Union; confirms that there should be no disclosure of material that could be reasonably judged to damage the UK in any negotiations to depart from the European Union after article 50 has been triggered; and calls on the Prime Minister to commit to publishing the Government’s plan for leaving the EU before article 50 is invoked, consistently with the principles agreed without division by this House on 12 October; recognises that this House should respect the wishes of the United Kingdom as expressed in the referendum on 23 June; and further calls on the Government to invoke article 50 by 31 March 2017.”

The Secretary of State also emphasises, in response to the argument of the Miller claimants, that the giving of notification under article 50(2) does not in itself alter any laws in force in the UK: it merely initiates a process of negotiation. If, at the end of those negotiations, a withdrawal agreement is reached, the procedures for Parliamentary approval laid down in the Constitutional Reform and Governance Act 2010 are likely to apply. Parliament will in any event be invited to legislate before the EU treaties cease to apply to the UK, so as to address the issues then arising in relation to rights and obligations under EU law which are currently given effect in the UK through the European Communities Act 1972 as amended (“ the 1972 Act ”).

The Miller claimants, on the other hand, rely on decided cases concerned with the use of prerogative powers in other situations. They argue that those cases establish the existence of legal constraints on the exercise of those powers, and that those constraints are applicable in the admittedly different situation with which we are now concerned. They argue that the effect of those constraints is that Ministers cannot lawfully give notification under article 50(2) unless an Act of Parliament authorises them to do so.

The starting point of this argument is the Case of Proclamations (1611) 12 Co Rep 74, which concerned the question whether James I could, by proclamation, prohibit the construction of new buildings in and around London, and prohibit the manufacture of starch from wheat. Coke CJ stated that “the King by his proclamation or other ways cannot change any part of the common law, or statute law, or the customs of the realm” (p 75). Those three categories were exhaustive of English law: “the law of England is divided into three parts, common law, statute law, and custom; but the King’s proclamation is none of them” (ibid). It followed that “the King cannot create any offence by his prohibition or proclamation, which was not an offence before, for that was to change the law” (ibid).

The same approach can be seen in more recent cases. For example, in The Zamora [1916] 2 AC 77 an issue arose as to whether the courts were bound, by an Order in Council made under prerogative powers, to decide that a neutral ship found during wartime to have a contraband cargo on board, while ostensibly bound for a neutral port, was lawful prize: an issue which, under established legal principles, depended on whether the ship or its cargo was in reality destined for the enemy. Lord Parker of Waddington stated:

“The idea that the King in Council, or indeed any branch of the Executive, has power to prescribe or alter the law to be administered by Courts of law in this country is out of harmony with the principles of our Constitution ... No one would contend that the prerogative involves any power to prescribe or alter the law administered in Courts of Common Law or Equity.” (p 90)

These cases were not concerned with the prerogative power to conduct foreign relations. It is however consistent with those cases that, although the Crown can undoubtedly enter into treaties in the exercise of prerogative powers, it cannot, by doing so, alter domestic law. That is known as the dualist approach to international law, in distinction to the monist approach adopted by many other countries, under which treaties automatically take effect in the domestic legal system. In support of the principle that treaties cannot alter domestic law, the Miller claimants rely on the explanations of the relationship between international and domestic law given by Lord Templeman and Lord Oliver in the Tin Council case. The case concerned the question whether a Minister of the Crown was liable under English law for the debts of an international organisation which had been established by a treaty to which the UK was party. Rejecting the contention that the Minister was liable, Lord Templeman said:

“A treaty is a contract between the governments of two or more sovereign states. International law regulates the relations between sovereign states and determines the validity, the interpretation and the enforcement of treaties. A treaty to which Her Majesty’s Government is a party does not alter the laws of the United Kingdom. A treaty may be incorporated into and alter the laws of the United Kingdom by means of legislation. Except to the extent that a treaty becomes incorporated into the laws of the United Kingdom by statute, the courts of the United Kingdom have no power to enforce treaty rights and obligations at the behest of a sovereign government or at the behest of a private individual.” (pp 476-477)

The principle that the Crown cannot alter the common law or statute by an exercise of the prerogative was developed in the case of Attorney General v De Keyser’s Royal Hotel Ltd [1920] AC 508 , which concerned the requisitioning of a hotel during the First World War for use as the headquarters of the Royal Flying Corps. After the war, a dispute arose over the basis on which the compensation to be paid to the owners should be assessed. There was a statutory scheme for requisitioning, which included a statutory right to compensation, but Ministers argued that the Crown was in any event entitled to requisition the hotel under prerogative powers, in which event compensation was payable ex gratia rather than being assessed in accordance with the statutory scheme. That argument was rejected by the House of Lords on the basis that “if the whole ground of something which could be done by the prerogative is covered by the statute, it is the statute that rules” (per Lord Dunedin at p 526). As Lord Dunedin reasoned:

“Inasmuch as the Crown is a party to every Act of Parliament it is logical enough to consider that when the Act deals with something which before the Act could be effected by the prerogative, and specially empowers the Crown to do the same thing, but subject to conditions, the Crown assents to that, and by that Act , to the prerogative being curtailed.” (p 526)

In the light of these decided cases, and others to the same effect, the Miller claimants argue that giving notification under article 50(2) will alter domestic law and destroy statutory rights. That is because it will result in the EU treaties ceasing to apply to the UK, in accordance with article 50(3), from the date of the entry into force of the withdrawal agreement or, failing that, from the expiry of a period of two years after notification, or any longer period which may be agreed with the European Council. Since the EU treaties have been given effect in domestic law by the 1972 Act , so as to create rights enforceable before our national courts, it would offend against the principle established in the Case of Proclamations, and explained more recently in the Tin Council case, for that alteration in domestic law to be effected under the prerogative. This argument assumes that, once notification is given under article 50(2), the process of withdrawal from the EU cannot be stopped. It is common ground in all the cases before the court that it should proceed on that assumption. In any event, even if the process might be stopped, it is common ground that Ministers’ power to give notice under article 50(2) has to be tested on the basis that it may not be stopped. In those circumstances, that is the basis on which this court is proceeding.

Furthermore, since the 1972 Act makes provision for the effect of the EU treaties in domestic law, and notification under article 50(2) will sooner or later result in the treaties ceasing to have effect in domestic law, it is argued that there is a conflict between the exercise of the prerogative to give notification and the statutory scheme. Following De Keyser, that conflict should be resolved in favour of the statute, by holding that the prerogative must be constrained.

Both sides of the argument proceed on the basis that the referendum on membership of the EU, held under the European Union Referendum Act 2015 (“ the 2015 Act ”), which resulted in a vote to leave the EU, does not provide the answer. The Secretary of State’s argument proceeds on the basis that the Crown has taken the decision under article 50(1), accepting the result of the referendum. The Miller claimants argue that only Parliament can take that decision. Both the Secretary of State and the Miller claimants proceed on the basis that the referendum result was not itself a decision by the UK to withdraw from the EU, in accordance with the UK’s constitutional requirements, and that the 2015 Act did not itself authorise notification under article 50(2). In these circumstances, there is no issue before the court as to the legal effect of the referendum result. Nor is this an appropriate occasion on which to consider the implications for our constitutional law of the developing practice of holding referendums before embarking on major constitutional changes: a matter on which the court has heard no argument.

In addition to the arguments advanced by the parties to the Miller appeal, the court also has before it the submissions presented on behalf of the interested parties and interveners. They largely provide further elaboration of the arguments presented on behalf of the principal parties. Without intending any discourtesy, I do not think it is necessary to set out their arguments in full, and would generally wish only to acknowledge the assistance which they have provided. It is however appropriate to note the submissions made by the Lord Advocate (which share common ground with those of the first interested party and the fourth interveners), and by the Counsel General for Wales.

One argument advanced by the Lord Advocate and by Ms Mountfield QC on behalf of the first interested party is that the UK’s withdrawal from the EU will alter the UK’s rule of recognition: that is to say, the rule which identifies the sources of law in our legal system and imposes a duty to give effect to laws emanating from those sources. The status of the EU institutions as a recognised source of law will inevitably be revoked, sooner or later, following notification under article 50(2). Since that will be a fundamental alteration in the UK’s constitution, it can only be effected by Parliamentary legislation. An Act of Parliament is therefore argued to be necessary before notification can be given.

The Lord Advocate also cites material from Scottish sources which is consistent with the principle derived by the Miller claimants from English case law, such as the Case of Proclamations and the Tin Council case. This includes the provision of the Claim of Right Act 1689:

“… That all Proclamationes asserting ane absolute power to Cass annull and Dissable lawes… are Contrair to Law.”

The Lord Advocate and the Counsel General for Wales have also advanced submissions concerning the Sewel Convention. That convention was originally stated by Lord Sewel, when Parliamentary Under-Secretary of State at the Scottish Office, in the House of Lords during the passage of the Scotland Bill. He said that “we would expect a convention to be established that Westminster would not normally legislate with regard to devolved matters in Scotland without the consent of the Scottish Parliament”: Hansard (HL Debates), 21 July 1998, col 791. The convention was later embodied in a Memorandum of Understanding between the UK Government and the devolved governments (Cm 5240, 2001). Para 14 of the current Memorandum of Understanding, which was published in October 2013, states:

“The United Kingdom Parliament retains authority to legislate on any issue, whether devolved or not. It is ultimately for Parliament to decide what use to make of that power. However, the UK Government will proceed in accordance with the convention that the UK Parliament would not normally legislate with regard to devolved matters except with the agreement of the devolved legislature.”

This section does not affect the power of the Parliament of the United Kingdom to make laws for Scotland.

But it is recognised that the Parliament of the United Kingdom will not normally legislate with regard to devolved matters without the consent of the Scottish Parliament.”

It may be helpful to summarise at this stage the conclusions which I have reached in relation to the Miller appeal, before explaining the reasons why I have arrived at them.

I entirely accept the importance in our constitutional law of the principle of Parliamentary supremacy over our domestic law, established in the Case of Proclamations, the Tin Council case, and other similar cases such as The Zamora. That principle does not, however, require that Parliament must enact an Act of Parliament before the UK can leave the EU. That is because the effect which Parliament has given to EU law in our domestic law, under the 1972 Act , is inherently conditional on the application of the EU treaties to the UK, and therefore on the UK’s membership of the EU. The Act imposes no requirement, and manifests no intention, in respect of the UK’s membership of the EU. It does not, therefore, affect the Crown’s exercise of prerogative powers in respect of UK membership. For essentially the same reason, the supposed analogy with De Keyser appears to me to be misplaced. Further, since the effect of EU law in the UK is entirely dependent on the 1972 Act , no alteration in the fundamental rule governing the recognition of sources of law has resulted from membership of the EU, or will result from notification under article 50. It follows that Ministers are entitled to give notification under article 50, in the exercise of prerogative powers, without requiring authorisation by a further Act of Parliament.

Given that conclusion, the argument in relation to the Sewel Convention does not arise: the convention concerns Parliamentary legislation, not the exercise of prerogative powers.

The issue which lies at the heart of these cases is the effect of the 1972 Act , as amended. Section 2(1) provides:

“All such rights, powers, liabilities, obligations and restrictions from time to time created or arising by or under the Treaties, and all such remedies and procedures from time to time provided for by or under the Treaties, as in accordance with the Treaties are without further enactment to be given effect or used in the United Kingdom shall be recognised and available in law, and be enforced, allowed and followed accordingly …”

The expression “the Treaties” is defined by section 1(2) . Put shortly, it includes the pre-accession treaties (described in Part 1 of Schedule 1), taken with other treaties listed in section 1(2) , and “any other treaty entered into by the EU ... with or without any of the member States, or entered into, as a treaty ancillary to any of the Treaties, by the United Kingdom”. In relation to the treaties in the latter categories, section 1(3) lays down a procedure to be followed:

“If Her Majesty by Order in Council declares that a treaty specified in the Order is to be regarded as one of the EU Treaties as herein defined, the Order shall be conclusive that it is to be so regarded; but a treaty entered into by the United Kingdom after the 22nd January 1972, other than a pre-accession treaty to which the United Kingdom accedes on terms settled on or before that date, shall not be so regarded unless it is so specified, nor be so specified unless a draft of the Order in Council has been approved by resolution of each House of Parliament.”

Section 1(2) is prospective in scope: it is not confined to treaties existing when the 1972 Act was originally enacted, but envisages treaties being entered into in the future. At the time of accession, the Treaties were relatively few in number, and included the Treaty of Rome. Since then, many other treaties, including the Maastricht Treaty and the Treaty of Lisbon, have been added, either by the amendment of section 1(2) so as to add to the list of specified treaties, or by the making of Orders of Council approved by resolutions of both Houses, under section 1(3) .

Returning to section 2(1) , it is important to understand why it was necessary. It follows from the UK’s dualist approach to international law that the Treaties could only be given effect in our domestic law by means of an Act of Parliament. This was so notwithstanding the doctrine of EU law, established by the European Court of Justice in Van Gend en Loos ( Case C-26/62 ) [1963] ECR 1, 12, that the Treaty of Rome was “more than an agreement which merely creates mutual obligations between the contracting states”, and that “independently of the legislation of member states, Community law therefore not only imposes obligations on individuals but is also intended to confer upon them rights which become part of their legal heritage.” This doctrine was reiterated in Costa v ENEL ( Case C-6/64 ) [1964] ECR 585, 593:

“By contrast with ordinary international treaties, the EEC Treaty has created its own legal system which, on the entry into force of the Treaty, became an integral part of the legal systems of the member states and which their courts are bound to apply.”

This doctrine is incompatible with the dualist approach of the UK constitution, and ultimately with the fundamental principle of Parliamentary sovereignty. This was explained by Lord Denning MR in two cases decided around the time when the UK joined the European Communities. The first, Blackburn v Attorney General [1971] 1 WLR 1037 , was as explained earlier an attempt to prevent the Crown from acceding to the Treaty of Rome by signing the Treaty of Accession. Having been referred to Costa v ENEL, the Master of the Rolls observed:

“Even if a treaty is signed, it is elementary that these courts take no notice of treaties as such. We take no notice of treaties until they are embodied in laws enacted by Parliament. and then only to the extent that Parliament tells us.” (p 1039)

Considering section 2(1) in greater detail, it is a long and densely-packed provision, whose syntax is complex, and whose meaning is not immediately clear. It requires to be read with care. Its essential structure can be expressed in this way:

All such [ members of a specified category] as [satisfy a specified condition] shall be [dealt with in accordance with a specified requirement ].

Two features of such rules should be noted. First, the rule is conditional in nature: the application of the requirement which it imposes depends on there being members of the specified category that satisfy the relevant condition. In the examples just given, for example, the relevant conditions are being charged with conduct contrary to good order and discipline; and displaying symptoms of ebola. Secondly, although a rule in that form contemplates the possibility that the condition may be satisfied, the form of the rule does not convey any intention that the condition will be satisfied. In the examples just given, for example, the rule does not convey an intention that there will be prisoners who are charged, or passengers who display symptoms of ebola. The intention of the rule-maker, so far as it can be derived from the rule, would not therefore be thwarted or frustrated if, either immediately, or at some point in the future, there were no members of the relevant category which satisfied the relevant condition.

In section 2(1) , the relevant category is:

“ rights, powers, liabilities, obligations and restrictions from time to time created or arising by or under the Treaties, and ... remedies and procedures from time to time provided for by or under the Treaties. ”

This is relevant in the present context, since it demonstrates that Parliament has recognised that rights given effect under the 1972 Act may be added to, altered or revoked without the necessity of a further Act of Parliament (something which is also apparent from section 1(3) ). In response to this point, the majority of the court draw a distinction, described as “a vital difference”, between changes in domestic law resulting from variations in the content of EU law arising from new EU legislation, and changes resulting from withdrawal by the UK from the European Union. There is no basis in the language of the 1972 Act for drawing any such distinction. Under the arrangements established by the Act , alterations in the UK’s obligations under the Treaties are automatically reflected in alterations in domestic law. That is equally the position whether the alterations in the UK’s obligations under the Treaties result from the Treaties’ ceasing to apply to the UK, in accordance with article 50, or from changes to the Treaties or to legislation made under the Treaties. The Act simply creates a scheme under which the effect given to EU law in domestic law reflects the UK’s international obligations under the Treaties, whatever they may be. There is nothing in the Act to suggest that Parliament’s intention to ensure an exact match depends on the reason why they might not match.

The requirement imposed by section 2(1) is:

“shall be recognised and available in law, and be enforced, allowed and followed accordingly .”

The condition which must be satisfied, in order for that requirement to apply, is set out in the following phrase:

“All such ... as in accordance with the Treaties are without further legal enactment to be given legal effect or used in the United Kingdom .”

The words “without further enactment” reflect the EU law concept of direct effect, established by Van Gend en Loos and Costa v ENEL as explained above (and, in so far as it may be regarded as distinct, the concept of direct applicability, established by article 189 of the Treaty of Rome and now stated in article 288 of the Treaty on the Functioning of the European Union (TFEU): see section 18 of the European Union Act 2011 ). Accordingly, where “in accordance with the Treaties”, rights, powers and so forth are to be directly applicable or directly effective in the law of the UK, section 2(1) achieves that effect. But there is no obligation “in accordance with the Treaties” to give effect in the UK to EU rights, powers and so forth merely because they are directly effective under EU law: such an obligation arises only if and for so long as the Treaties apply to the UK. The extent to which the effect given by section 2(1) to rights, powers and so forth arising under EU law is dependent on the Treaties cannot therefore be confined to the question whether the rights, powers and so forth are, under the Treaties, directly effective: it also depends, more fundamentally, on whether the Treaties impose any obligations on the UK to give effect to EU law.

Whether rights, powers and so forth are to be given legal effect in the UK, in accordance with the Treaties, therefore depends on whether the Treaties apply to the UK. As the majority of the court state at para 77, “Parliament cannot have intended that section 2 should continue to import the variable content of EU law into domestic law, or that the other consequences of the 1972 Act described in paras 62 to 64 above should continue to apply, after the United Kingdom had ceased to be bound by the EU Treaties.” If the Treaties do not apply to the UK, then there are no rights, powers and so forth which, in accordance with the Treaties, are to be given legal effect in the UK.

This point is illustrated by the fact that, when the 1972 Act came into force on 17 October 1972, the Treaty of Accession had not yet been ratified or entered into force, with the consequence that the Treaties did not apply to the UK. In consequence, section 2(1) initially had no practical application, there being at that time no rights, powers and so forth which, in accordance with the Treaties, were to be given legal effect in the UK. It was not until 1 January 1973, when the Treaty of Accession came into force, following its ratification by the Crown in the exercise of its prerogative powers, that the condition to which section 2(1) subjected the domestic effect of EU law was satisfied.

The Miller claimants respond to this point by arguing that the effect of the 1972 Act was to require the Crown to ratify the Treaty of Accession. This is not, in the first place, an answer to the point that the effect of section 2(1) was contingent on the Treaty’s entering into force. Furthermore, although it is fair to say that the 1972 Act was enacted in anticipation that ratification was likely to occur that is far from saying that ratification was required by statute.

In the first place, as explained in para 159 above, it is a basic principle of our constitution that the conduct of foreign relations, including the ratification of treaties, falls within the prerogative powers of the Crown. That principle is so fundamental that it can only be overridden by express provision or necessary implication, as is accepted in the majority judgment at para 48. No such express provision exists in the 1972 Act . Nor do its provisions override that principle as a matter of necessary implication. As Lord Hobhouse of Woodborough explained in R (Morgan Grenfell & Co Ltd) v Special Commissioner of Income Tax [2002] UKHL 21 ; [2003]1 AC 563, para 45:

“A necessary implication is not the same as a reasonable implication … A necessary implication is one which necessarily follows from the express provisions of the statute construed in their context. It distinguishes between what it would have been sensible or reasonable for Parliament to have included or what Parliament would, if it had thought about it, probably have included and what it is clear that the express language of the statute shows that the statute must have included. A necessary implication is a matter of express language and logic not interpretation.”

Secondly, it is not difficult to contemplate circumstances in which ratification might not have occurred. The passage of the 1972 Act was hard fought (as the former minister Ken Clarke’s memoir, Kind of Blue (2016), pp 66ff, makes clear), and the possibility of a future Labour Government taking the UK out of the European Communities was apparent. When the Labour Government subsequently came to power, in 1974, it proceeded to hold a referendum in accordance with its manifesto commitment. If the Conservative Government had fallen and the Opposition had come to power while the Treaty of Accession remained unratified, the incoming Labour Government would have been unlikely to ratify it without holding a referendum. Indeed, the Opposition continued to oppose ratification following the Parliamentary passage of the 1972 Act , using an adjournment debate on the date of Royal Assent to criticise ratification as being against the wish of the British people (Hansard (HC Debates), 17 October 1972, cols 58-59). The Government won the division by 31 votes; but if it had lost it, would it have been acting unlawfully if it had decided to respect the will of the House of Commons by not ratifying the treaty? Would it have been legally bound by the 1972 Act to ratify the treaty regardless? These questions can only be answered in the negative. The point can also be illustrated by considering what would have happened if some crisis had occurred in the UK’s diplomatic relations with one of its intended partners in the European Communities. If, for example, some dispute comparable in gravity to the then current dispute with Iceland, or the subsequent dispute with Argentina, had occurred with one of the other parties to the Treaty of Rome or the Treaty of Accession, is it likely that the UK would then have ratified the Treaty of Accession?

The seemingly less ambitious suggestion in the majority judgment at para 78, that it was not “contemplated”, when the 1972 Act was being passed, that Ministers would not ratify the Treaties or that, having ratified them, would at some point repudiate them, meets the same objection. That ratification was contemplated is clear, but that tells you nothing about whether the operation of the 1972 Act is conditional on continued membership. What individual members of Parliament contemplated, or expected to happen, is on ordinary principles not relevant to the construction of the Act . In any event it is likely to have varied a good deal. The possibility of the UK being taken out of the European Communities if there were a change of government was apparent.

Referring to the structure of section 2(1) of the 1972 Act as set out at para 184 above, it is said at para 82 of the majority judgment that “the membership of the specified category [viz, the rights, powers and so forth arising under EU law to which domestic effect must be given] has a variable content which is contingent on the decisions of non-UK entities”. Section 2(1) says nothing, however, which either expressly or impliedly limits the contingency, to which the duty to give domestic effect to EU law is subject, to decisions by non-UK entities. The contingency is that the rights, powers and so forth are “such ... as in accordance with the Treaties are without further legal enactment to be given legal effect or used in the United Kingdom”. It follows from that contingency that the effect given to EU law in our domestic law is conditional on the Treaties’ application to the UK. That condition was not satisfied when the Act came into force, because the Treaties did not then apply to the UK. The content of the specified category was therefore zero. The satisfaction of the condition, some months later, depended on the decision of a UK entity: it depended on the Crown’s exercise of prerogative powers. The content would return to zero if the condition ceased to be satisfied as the result of the UK’s invoking article 50. That would be so whether the decision to invoke article 50 had, or had not, been authorised by an Act of Parliament. It is, indeed, accepted by the majority that the condition would cease to be satisfied if the Crown invoked article 50 after being authorised to do so by statute. So the contingency cannot be limited to decisions by non-UK entities. The only issue in dispute is whether the action by the Crown, as a result of which the contingency will cease to be satisfied, must be authorised by an Act of Parliament. On that issue, section 2(1) is silent. Neither expressly nor by implication does it require such action to be authorised by Parliament. The fact that section 2(1) is itself a fixed rule of domestic law enacted by Parliament does not affect that conclusion, since a fixed rule which is conditional will necessarily operate only for as long as the condition is satisfied. Nor does it support a conclusion that Parliament has, by necessary implication, deprived the Crown of its prerogative powers: from what words, one might ask, is that implication derived?

I have discussed the position as it stood in 1972. But the real question in the Miller appeal concerns the position following the signing of the Treaty of Lisbon in 2007, and its entry into force in 2009. That is because it was the Treaty of Lisbon which inserted article 50 into the TEU.

Parliament addressed the Treaty of Lisbon in the European Union Amendment Act 2008 (“ the 2008 Act ”). Section 2 of that Act provides:

“At the end of the list of treaties in section 1(2) of the European Communities Act 1972 (c 68) add; and

the Treaty of Lisbon Amending the Treaty on European Union and the Treaty Establishing the European Community signed at Lisbon on 13 December 2007 (together with its Annex and protocols), excluding any provision that relates to, or in so far as it relates to or could be applied in relation to, the Common Foreign and Security Policy;”

It follows that the words “such ... as in accordance with the Treaties are without further legal enactment to be given legal effect or used in the United Kingdom”, in section 2(1) of the 1972 Act , must be read as meaning “such ... as in accordance with the Treaties, including article 50 TEU, are without further legal enactment to be given legal effect or used in the United Kingdom”. The contingency to which the effect of EU law in our domestic law has been subject since the amendment of the 1972 Act by the 2008 Act therefore includes the potential operation of article 50. It is entirely “in accordance with the Treaties” for article 50 to operate, with the result that, when a withdrawal agreement comes into force, or the time allowed under article 50(3) expires, there may be no rights which, “in accordance with the Treaties”, are to be given legal effect in the UK.

This conclusion is not inconsistent with the statement by the majority of the court, at para 104, that article 50 is not given effect in domestic law by section 2 of the 1972 Act . The majority may be right about that, although the point has not been argued, and the opposite view may be arguable (see, for example, Robert Craig, “Casting Aside Clanking Medieval Chains: Prerogative, Statute and Article 50 after the EU Referendum”, (2016) MLR 1041, where it is argued that section 2(1) of the 1972 Act has given article 50 domestic effect as a power exercisable by Ministers, superseding the prerogative but also supplying the Parliamentary authorisation desiderated by the Miller claimants). Whether article 50 has direct effect in domestic law does not however affect the question whether its operation forms part of the contingency on which the direct effect given to EU law by the 1972 Act is dependent.

The result of section 2 of the 2008 Act is thus that the effect given by section 2(1) of the 1972 Act to EU law, which was always conditional on the Treaties’ applying to the UK, is now subject to the exercise of the power conferred by article 50 to initiate a particular procedure under which the Treaties will cease to apply to the UK.

The Miller claimants respond to these points by arguing that section 2(1) of the 1972 Act impliedly requires the power of withdrawal under article 50 to be exercised by Parliament. In so far as that argument is based on the common law principles established by such authorities as the Case of Proclamations, The Zamora, the Tin Council case and the De Keyser case, I shall discuss those principles later. One can however note at present that, as previously mentioned, there is nothing in section 2(1) which demonstrates that Parliament intended to depart from the fundamental principle that powers relating to the UK’s participation in treaty arrangements are exercisable by the Crown. As the majority of the court rightly state at para 108, the fact that a statute says nothing about a particular topic can rarely, if ever, justify inferring a fundamental change in the law. Nor would withdrawal under article 50 be inconsistent with the 1972 Act , any more than a failure to ratify the Treaty of Accession. The result would simply be that there were no rights answering to the description in section 2(1) : there would be no rights “such ... as in accordance with the Treaties are without further legal enactment to be given legal effect or used in the United Kingdom”.

This is a point of general importance. If Parliament chooses to give domestic effect to a treaty containing a power of termination, it does not follow that Parliament must have stripped the Crown of its authority to exercise that power. In the present context, the impact of the exercise of the power on EU rights given effect in domestic law is accommodated by the 1972 Act : the rights simply cease to be rights to which section 2(1) applies. Withdrawal under article 50 alters the application of the 1972 Act , but is not inconsistent with it. The application of the 1972 Act after a withdrawal agreement has entered into force (or the applicable time limit has expired) is the same as it was before the Treaty of Accession entered into force. As in the 1972 Act as originally enacted, Parliament has created a scheme under which domestic law tracks the obligations of the UK at the international level, whatever they may be.

Other post-1972 legislation is of only secondary importance. It is however relevant in so far as it demonstrates, first, that Parliament has legislated on the basis that the 1972 Act did not restrict the exercise of the foreign affairs prerogative in relation to other aspects of the EU treaties, and secondly, that Parliament is perfectly capable of making clear its intention to restrict the exercise of the prerogative when it wishes to do so.

Several examples can be given. The earliest is section 6(1) of the European Parliamentary Elections Act 1978 (as amended by section 3 of the European Communities (Amendment) Act 1986 ), which provided:

“No treaty which provides for any increase in the powers of the European Parliament shall be ratified by the United Kingdom unless it has been approved by an Act of Parliament.”

A further example is the 2008 Act , which imposed numerous restrictions on the exercise of prerogative powers in relation to provisions of the Lisbon Treaty. Section 5 is particularly significant. It provided:

“(1) A treaty which satisfies the following conditions may not be ratified unless approved by Act of Parliament.

the Treaty on European Union (signed at Maastricht on 7 February 1992),

the Treaty on the Functioning of the European Union (the Treaty establishing (what was then called) the European Economic Community, signed at Rome on 25 March 1957 (renamed by the Treaty of Lisbon)), or

the Treaty establishing the European Atomic Energy Community (signed at Rome on 25 March 1957).

Condition 2 is that the treaty results from the application of article 48(2) to (5) of the Treaty on European Union (as amended by the Treaty of Lisbon) (Ordinary Revision Procedure for amendment of founding Treaties, including amendments affecting EU competence).”

Article 48 TEU was a provision introduced by the Lisbon Treaty to provide a simplified procedure for the conclusion of treaties amending the TEU or the TFEU. Paragraphs (2) to (5) provided, so far as material:

“2. The Government of any member state, the European Parliament or the Commission may submit to the Council proposals for the amendment of the Treaties. These proposals may, inter alia, serve either to increase or to reduce the competences conferred on the Union in the Treaties ...

If the European Council, after consulting the European Parliament and the Commission, adopts by a simple majority a decision in favour of examining the proposed amendments, the President of the European Council shall convene a Convention composed of representatives of the national Parliaments, of the Heads of State or Government of the member states, of the European Parliament and of the Commission ... The Convention shall examine the proposals for amendments and shall adopt by consensus a recommendation to a conference of representatives of the governments of the member states as provided for in paragraph 4.

A conference of representatives of the governments of the member states shall be convened by the President of the Council for the purpose of determining by common accord the amendments to be made to the Treaties.

If, two years after the signature of a treaty amending the Treaties, four fifths of the member states have ratified it and one or more member states have encountered difficulties in proceeding with ratification, the matter shall be referred to the European Council.”

The TFEU establishes numerous rights which are given effect in the UK by section 2(1) of the 1972 Act . Those rights could be altered by a treaty concluded by the UK Government and the governments of the other member states, under article 48(2) TFEU. Section 5 of the 2008 Act required an Act of Parliament before such a treaty could be ratified. If the Miller claimants’ arguments are correct, an Act of Parliament was already necessary before the UK Government could exercise the treaty-making prerogative so as to alter those rights. Section 5 of the 2008 Act was, however, understood as introducing a requirement for legislation where none previously existed: that was the mischief intended to be addressed. For example, the House of Lords Select Committee on the Constitution stated:

“Clause 5 of the Bill seeks to create a new requirement for prior parliamentary authorisation of ratification. It would apply to amendments of the founding treaties - the Treaty on European Union, the Treaty on the Functioning of the European Union and the Treaty Establishing the European Atomic Energy Community - when those amendments are made by the ‘ordinary revision procedure’.

It is also relevant to note section 6(1) of the 2008 Act , which imposed restrictions on the UK’s participation in several procedures laid down in the Lisbon Treaty:

“A Minister of the Crown may not vote in favour of or otherwise support a decision under any of the following unless Parliamentary approval has been given in accordance with this section ...”

The Constitutional Reform and Governance Act 2010 (“ the 2010 Act ”) is also relevant. It codifies the previous Ponsonby Rule (a convention that treaties, with limited exceptions, would be laid before Parliament before they were ratified), and sets out detailed procedures for Parliamentary scrutiny of new treaties. It does not apply to treaties which are covered by section 5 of the 2008 Act or by the European Union Act 2011 (“ the 2011 Act ”), to which I turn next. A withdrawal agreement under article 50(3) would be likely to fall within its scope, but it would have no application to a decision to withdraw from a treaty or to commence the process of withdrawal.

The 2011 Act repealed section 12 of the 2002 Act and sections 5 and 6 of the 2008 Act (subject to an immaterial exception), replacing them with a more elaborate system of Parliamentary control. The evident aim was to introduce stronger Parliamentary controls, in relation to matters falling within the scope of the legislation, than were present under the existing law. The power to amend article 50(3), concerning the extension of the two year period for negotiation, or to adopt the ordinary legislative procedure in relation to that provision, was brought within the scope of these controls by sections 4 and 6, read with Schedule 1. Article 50(1) and (2), concerning the decision to withdraw and its notification, were not.

As explained earlier, section 5 of the 2008 Act was enacted on the basis that the Crown could exercise its treaty-making power so as to alter EU rights given effect in domestic law by the 1972 Act , without necessarily requiring further authorisation by an Act of Parliament. One can also infer from this body of legislation, as the Divisional Court did in the case of R v Secretary of State for Foreign Affairs, Ex p Rees-Mogg [1994] QB 552 , discussed in paras 235-237 below, that since Parliament has repeatedly placed express restrictions on the exercise of the prerogative in relation to the EU treaties, the absence of a particular restriction in the 1972 Act tends to support the conclusion that no such restriction was intended to arise by implication.

It is also necessary to consider the 2015 Act . For the reasons explained in para 171 above, I do not propose to consider the legal implications of the referendum result. It is, however, proper to take note of the judgment of Lord Dyson MR, with whom the other members of the court agreed, in R (Shindler) v Chancellor of the Duchy of Lancaster [2016] EWCA Civ 419 ; [2016] 3 WLR 1196 . That was a case in which a challenge was brought to the franchise rules applicable to the referendum. Having referred to the provision in article 50(1) that any member state may decide to withdraw from the EU “in accordance with its own constitutional requirements”, the Master of the Rolls stated:

“ The 2015 Act contains part of the constitutional requirements of the UK as to how it may decide to withdraw from the EU ... In short, by passing the 2015 Act , Parliament decided that one of the constitutional requirements that had to be satisfied as a condition of a withdrawal from the EU was a referendum.” (paras 13 and 19)

In the light of the foregoing discussion, one can return to the arguments advanced by the Miller claimants on the basis of authorities concerned with the common law limits of prerogative powers. The first argument, summarised at paras 165-167 and 169 above, is that the giving of notification under article 50(2) will result in the alteration of the law and the destruction of statutory rights, and therefore cannot be effected in the exercise of prerogative powers, applying the principles established in such cases as the Case of Proclamations, The Zamora, the Tin Council case, and Higgs v Minister of National Security, and reflected also in the Bill of Rights and the Claim of Right.

The argument that the 1972 Act created statutory rights which cannot be taken away without a further Act of Parliament starts from a premise which requires examination. The 1972 Act did not create statutory rights in the same sense as other statutes, but gave legal effect in the UK to a body of law now known as EU law. As explained at paras 186-187 above, section 2(1) recognises that the rights arising under that body of law can be altered from time to time, as a result of changes to the Treaties or to the laws made under the procedures laid down in the Treaties, without the necessity of a further Act of Parliament. Such alterations result not only in the creation of EU rights which are consequently given effect in domestic law by the 1972 Act , but also in the repeal and restriction of EU rights previously created, and given effect under domestic law. The successive regulations imposing fishing quotas are an example. To give another example, if Greece were to decide to leave the EU while the UK remained a member, the Treaties would cease to apply to Greece either when a withdrawal agreement entered into force, or in any event after two years had expired. Greek citizens living in the UK would then cease to enjoy the EU rights which continued to be enjoyed here, for example, by French citizens. As these examples illustrate, rights given direct effect by section 2(1) of the 1972 Act are inherently contingent, and can be altered without any further Act of Parliament. This is a very different situation from any contemplated by the judges in the cases relied on, or by the Scottish and English Parliaments at the time of the Glorious Revolution or the Act s of Union.

As noted earlier, the majority of the court respond to this point by drawing a distinction between changes which result from the UK’s giving notice under article 50, for which a further Act of Parliament is argued to be necessary, and changes which result from any other alteration in the Treaties or in the instruments made under the Treaties, for which no further Act of Parliament is necessarily required. That distinction cannot be derived from the principle established by the Case of Proclamations. It has to be based on an interpretation of the 1972 Act : the matter which was discussed at paras 179-214 above. For the reasons there explained, I see no basis in the 1972 Act for drawing any such distinction. The Act simply creates a scheme under which domestic law reflects the UK’s international obligations, whatever they may be.

It is equally questionable whether notification under article 50 will alter “the law of the land”, in the sense in which judges have used that expression. That can be illustrated by reflecting on the effect of notification, and on the ability of Parliament to maintain in force the EU rights currently given effect under section 2(1) of the 1972 Act . The giving of notification does not in itself alter EU rights or the effect given to them in domestic law. Nor does it impinge on Parliament’s competence to enact legislation during the intervening period before the treaties cease to have effect. Parliament can enact whatever provisions it sees fit in order to address the consequences of withdrawal from the EU, including provisions designed to protect rights which are currently derived from EU law. Parliament cannot, however, replicate EU law. It cannot establish those elements of it which involve reciprocal arrangements with the other member states, or which involve the participation of EU institutions. Nor can it create rights which have the distinguishing characteristics of EU rights, such as priority over subsequent legislation, and authoritative interpretation by the Court of Justice. The fact that notification alters no law, and that Parliament retains full competence to legislate so as to protect rights before withdrawal occurs, illustrates how different this situation is from those addressed in the cases relied upon. Equally, the fact that the enactment of EU law lies beyond the ability of Parliament illustrates how different it is from “the law of the land” as usually understood.

More fundamentally, however, the argument that withdrawal from the EU would alter domestic law and destroy statutory rights, and therefore cannot be undertaken without a further Act of Parliament, has to be rejected even if one accepts that the 1972 Act creates statutory rights and that withdrawal will alter the law of the land. It has to be rejected because it ignores the conditional basis on which the 1972 Act gives effect to EU law. If Parliament grants rights on the basis, express or implied, that they will expire in certain circumstances, then no further legislation is needed if those circumstances occur. If those circumstances comprise the UK’s withdrawal from a treaty, the rights are not revoked by the Crown’s exercise of prerogative powers: they are revoked by the operation of the Act of Parliament itself.

In so far as the Miller claimants place reliance on rights under EU law as given effect in the legal systems of other member states, such as the right of UK citizens to live and work in Greece, there is no rule which prevents prerogative powers being exercised in a way which alters rights arising under foreign law.

In so far as the Miller claimants place reliance on statutes creating rights in respect of EU institutions, such as the right to vote in elections to the European Parliament under the European Parliamentary Elections Act 2002 , such statutory rights are obviously conditional on the UK’s continued membership of the EU. Parliament cannot have intended them to operate on any other basis. If they cease to be effective following the UK’s withdrawal from the EU, that is inherent in the nature of the right which Parliament conferred. The only logical alternative is to hold that Parliament has created a right to remain in the EU, and none of the arguments goes that far.

As explained at para 173 above, it is argued that the 1972 Act created “an entirely new, independent and overriding source of domestic law” (as it is put in the majority judgment at para 80). Since the identification of a country’s sources of law is one of the most fundamental functions of its constitution, it follows that the Crown cannot lawfully revoke a source of law in the exercise of prerogative powers. So runs the argument.

As put by counsel, this argument is based on the concept of the rule of recognition: that is to say, the foundational rule in a legal system which identifies the sources of law in that system and imposes a duty to give effect to laws emanating from those sources. The Lord Advocate and Ms Mountfield QC argue that the rule would be altered by withdrawal from the EU, and therefore, sooner or later, by the giving of notification under article 50.

The UK’s entry into the EU did not, however, alter its rule of recognition, and neither would its withdrawal. That is because EU law is not a source of law of the relevant kind: that is to say, a source of law whose validity is not dependent on some other, more fundamental, source of law, but depends on the ultimate rule of recognition. The true position was explained by Lord Mance in Pham v Secretary of State for the Home Department [2015] UKSC 19 ; [2015] 1 WLR 1591 , para 80:

“For a domestic court, the starting point is, in any event, to identify the ultimate legislative authority in its jurisdiction according to the relevant rule of recognition. The search is simple in a country like the United Kingdom with an explicitly dualist approach to obligations undertaken at a supranational level. European law is certainly special and represents a remarkable development in the world's legal history. But, unless and until the rule of recognition by which we shape our decisions is altered, we must view the United Kingdom as independent, Parliament as sovereign and European law as part of domestic law because Parliament has so willed. The question how far Parliament has so willed is thus determined by construing the 1972 Act .”

As Lord Mance rightly explained, it follows from the UK’s dualist approach to international law that EU law is not one of the sources of law identified by the UK’s rule of recognition. That was recognised in the cases of Blackburn v Attorney General and McWhirter v Attorney General, as explained in para 183 above. As a source of law, EU law, like legislation enacted by the devolved legislatures, or delegated legislation made by Ministers, is entirely dependent on statute (which is not, of course, to say that EU law has the same effects, as devolved or delegated legislation). It derives its legal authority from a statute, which itself derives its authority from the rule of recognition identifying Parliamentary legislation as a source of law. The recognition of its validity does not alter any fundamental principle of our constitution.

The fact that the 1972 Act has a prospective effect, in giving effect to laws made from time to time by the EU institutions, does not affect this analysis. Nor does the limited primacy given to EU law by the 1972 Act alter the position, since that primacy itself derives from the 1972 Act . That was recognised by Lord Bridge of Harwich in R v Secretary of State for Transport, Ex p Factortame Ltd (No 2) [1991] 1 AC 603 :

“ Under the terms of the Act of 1972 it has always been clear that it was the duty of a United Kingdom court, when delivering final judgment, to override any rule of national law found to be in conflict with any directly enforceable rule of Community law.” (p 659: emphasis supplied)

Parliament has itself made it clear that EU law has not altered the UK’s rule of recognition. Section 18 of the 2011 Act provides :

“Directly applicable or directly effective EU law (that is, the rights, powers, liabilities, obligations, restrictions, remedies and procedures referred to in section 2(1) of the European Communities Act 1972 ) falls to be recognised and available in law in the United Kingdom only by virtue of that Act or where it is required to be recognised and available in law by virtue of any other Act.”

This understanding underpins the discussion of the constitutional status of EU law in R (Buckinghamshire County Council) v Secretary of State for Transport [2014] UKSC 3 ; [2014] 1 WLR 324 . The issue raised by a conflict between an EU directive and long-established constitutional principles of domestic law was identified as “the extent, if any, to which these principles may have been implicitly qualified or abrogated by the European Communities Act 1972 ” (para 78). The issue, in other words, was one of domestic law, turning on the interpretation of the 1972 Act . It was said:

“Contrary to the submission made on behalf of the claimants, that question cannot be resolved simply by applying the doctrine developed by the Court of Justice of the supremacy of EU law, since the application of that doctrine in our law itself depends upon the 1972 Act . If there is a conflict between a constitutional principle, such as that embodied in article 9 of the Bill of Rights, and EU law, that conflict has to be resolved by our courts as an issue arising under the constitutional law of the United Kingdom.” (para 79)

The majority of the court respond that this analysis is unrealistic. Although it is accepted that the effect of EU law in domestic law is dependent on the 1972 Act , they argue that for EU law to cease to have effect in our domestic law would be a major change in the UK’s constitution. As I understand it, the argument is concerned with the effect of the 1972 Act . Whether the 1972 Act has that effect depends on its interpretation, which simply takes one back to the issues discussed at paras 179-214 above.

A further reason for rejecting the argument that the 1972 Act created a new source of law, which cannot be revoked without further legislation, is one that applies even if it is accepted that the 1972 Act created a new source of law (in some sense or other). Since the 1972 Act gives effect to EU law only as long as the Treaties apply to the UK, as explained at paras 189-204 above, that source of law is inherently contingent on the UK’s continued membership of the EU. EU law’s ceasing to have effect as a result of the UK’s withdrawal from the Treaties is something which follows from the 1972 Act itself, and does not require further legislation.

Although the majority judgment does not adopt the Miller claimants’ argument based on a supposed analogy with the De Keyser case, it is nevertheless necessary to address it. As explained earlier, that case established that where Parliament has regulated a matter by statute, the Crown cannot have recourse to a prerogative power in respect of the same matter. The argument by analogy asserts that, since notification under article 50 will eventually render the 1972 Act redundant, it follows that notification cannot be given in the exercise of prerogative powers. I am unable to accept that argument, for a number of reasons.

First, the De Keyser principle denies that prerogative power can be exercised where a parallel statutory scheme exists. It does not follow from that principle that a prerogative power cannot be exercised where the eventual consequence will be that a statutory provision will cease to have a practical application. The latter proposition cannot be derived from De Keyser, but must be derived from some other source. The only obvious candidate is Parliament’s intention in enacting the statutory provisions in question: an intention, it has to be argued, to impose a limitation on the exercise of the prerogative power. That simply takes one back to the argument as to whether an intention to strip the Crown of its prerogative powers in respect of adherence to the EU treaties can be derived from the 1972 Act : an argument which was addressed at paras 201-204 above.

Secondly, the 1972 Act does not, in any event, regulate withdrawal from the EU: it recognises the existence of article 50, as explained in paras 198-202 above, but it says nothing about how or by whom a decision to invoke article 50 should be taken.

Thirdly, the difference between the present situation and that with which the De Keyser principle is concerned is also evident at the level of remedies. In De Keyser itself, the remedy was a declaration that the owners were entitled to compensation under the statutory scheme. The remedy flowed from the logic of the principle: Ministers were obliged to comply with the statutory scheme. No comparable remedy can be granted in the present case, since there is no statutory scheme governing the operation of article 50.

Finally, in relation to the Miller claimants’ arguments, it should be noted that this is not the first time that the courts have had to address these arguments. In R v Secretary of State for Foreign Affairs, Ex p Rees-Mogg , one of counsel’s arguments in support of a challenge to the ratification of the Maastricht Treaty was recorded by Lloyd LJ as follows:

“He submits that by ratifying the Protocol on Social Policy, the Government would be altering Community law under the EEC Treaty ... It is axiomatic that Parliament alone can change the law. Mr Pannick accepts, of course, that treaties are not self-executing. They create rights and obligations on the international plane, not on the domestic plane. He accepts also that the treaty-making power is part of the Royal Prerogative ... But the EEC Treaty is, he says, different. For section 2(1) of the European Communities Act 1972 provides ...

The Divisional Court rejected this argument:

“We find ourselves unable to accept this far-reaching argument. When Parliament wishes to fetter the Crown’s treaty-making power in relation to Community law, it does so in express terms, such as one finds in section 6 of the Act of 1978. Indeed, as was pointed out, if the Crown’s treaty-making power were impliedly excluded by section 2(1) of the Act of 1972, section 6 of the Act of 1978 would not have been necessary. There is in any event insufficient ground to hold that Parliament has by implication curtailed or fettered the Crown’s prerogative to alter or add to the EEC Treaty.” (p 567)

I agree with the Divisional Court’s reasoning in the passage which I have cited, and in particular with the final sentence: even apart from the inference which might be drawn from examples of express provisions restricting the exercise of prerogative powers in relation to EU law, there is in any event insufficient ground to hold that Parliament has by implication curtailed or fettered the Crown’s prerogative powers in relation to the Treaties.

Finally, in relation to the Miller appeal, it is argued by the majority at para 91 that the Secretary of State’s contentions cannot be correct since, if they were, it would have been open to Ministers to invoke prerogative powers to withdraw from the EU even if there had been no referendum, or indeed even if any referendum had resulted in a vote to remain.

There are two answers to this point. First, it does not necessarily follow from my conclusions that Ministers could properly have invoked article 50 whenever they pleased, or, more specifically, in the event of a vote to remain. As Lord Carnwath makes clear at para 266 below, there has been no discussion in this appeal of the question whether there might be any circumstances in which the exercise of the prerogative power in question might be open to review, such as if the referendum held under the 2015 Act had resulted in a vote to remain, and I express no view on that point.

Secondly, and more fundamentally, controls over the exercise of ministerial powers under the British constitution are not solely, or even primarily, of a legal character, as Lord Carnwath explains in his judgment. Courts should not overlook the constitutional importance of ministerial accountability to Parliament. Ministerial decisions in the exercise of prerogative powers, of greater importance than leaving the EU, have been taken without any possibility of judicial control: examples include the declarations of war in 1914 and 1939. For a court to proceed on the basis that if a prerogative power is capable of being exercised arbitrarily or perversely, it must necessarily be subject to judicial control, is to base legal doctrine on an assumption which is foreign to our constitutional traditions. It is important for courts to understand that the legalisation of political issues is not always constitutionally appropriate, and may be fraught with risk, not least for the judiciary.

For all the foregoing reasons, I would have allowed the Secretary of State’s appeal in the Miller case.

Given my disagreement with the decision of the majority of the court as to the necessity for an Act of Parliament before article 50 can be invoked, it follows that I would also have dealt with the devolution issues raised in the Northern Irish cases differently. So far as those cases raise issues which are distinct from those arising in the Miller appeal, however, I agree with the way in which the majority have dealt with them. Nothing in the Northern Ireland Act bears on the question whether the giving of notification under article 50 can be effected under the prerogative or requires authorisation by an Act of Parliament. More specifically, neither section 1 nor section 75 of the Northern Ireland Act has any relevance in the present context. Nor does a political convention, such as the Sewel Convention plainly is in its application to Northern Ireland, give rise to a legally enforceable obligation.

For the reasons given by Lord Reed, I would have allowed the appeal by the Secretary of State in the main proceedings. In view of the importance of the case, and the fact that we are differing from the Divisional Court and the majority in this court, I shall add some comments of my own from a slightly different legal perspective. I agree with the majority judgment in respect of the Northern Irish cases and the other devolution issues.

At the heart of the case is the classic statement of principle by Lord Oliver in the Tin Council case ( JH Rayner (Mincing Lane) Ltd v Department of Trade and Industry [1990] 2 AC 418 ):

“… as a matter of the constitutional law of the United Kingdom, the Royal Prerogative, whilst it embraces the making of treaties, does not extend to altering the law or conferring rights upon individuals or depriving individuals of rights which they enjoy in domestic law without the intervention of Parliament …” (Lord Oliver pp 499E-500D)

In the Tin Council case Lord Oliver was speaking only of the “making of treaties”, not withdrawal. Lord Templeman had earlier made clear that the prerogative enables the Government to “negotiate, conclude, construe, observe, breach, repudiate or terminate a treaty …” (p 476F-H). However, there was no discussion of how the classic statement might need modification or development in the context of termination or withdrawal. In principle the same basic rule should apply. Just as the Executive cannot without statutory authority create new rights or obligations in domestic law by entering into a treaty, so it cannot by termination of a treaty take away rights or obligations which currently exist. However, that tells one nothing about the process by which this result is to be achieved, nor at what stage of that process the intervention of Parliament is required.

Precedents are hard to find. Counsel have taken us on an interesting journey through cases and legal sources from four centuries and different parts of the common law world. The only example we were shown of withdrawal from a treaty was a recent decision of the Canadian Federal Court: Turp v Ministry of Justice & Attorney General of Canada 2012 FC 893. That was an unsuccessful challenge by the executive to the use of its prerogative powers to withdraw from the Kyoto Protocol on Climate Change, against the background of a statute (passed against the opposition of government) requiring the preparation of plans giving effect to the Protocol. On its face it is a striking example of the use of the prerogative to frustrate the apparent intention of Parliament as expressed in legislation. However, the authority is of limited assistance in the present context, since it had been held in a previous case ( Friends of the Earth v Canada (Governor in Council) , 2008 FC 118) that the obligations under the statute were not justiciable in the domestic courts.

In the end the search through the authorities tells one little that is not sufficiently expressed by the classic rule. It also confirms the lack of any direct precedent for withdrawal from a treaty previously given effect in domestic law, let alone one which has played such a vital part in the development of our laws over more than 40 years. However, lack of precedent is not a reason for inventing new principles, nor is there a need to do so. The existing principles correctly applied provide a clear and coherent framework for effective resolution of all the competing considerations, including the referendum result.

In considering that framework it is important to recognise the sensitivity in our constitution of the balance between the respective roles of Parliament, the Executive and the courts. The Divisional Court saw this principally in binary terms: the Executive versus Parliament. Under the general heading, “the sovereignty of Parliament and the prerogative powers of the Crown”, they referred on the one hand to “the most fundamental rule that the Crown in Parliament is sovereign” (para 20), and on the other to the “general rule” that “the conduct of international relations and the making and unmaking of treaties” are “matters for the Crown in the exercise of its prerogative powers” (para 30), the balance between the two being as explained by Lord Oliver in the Tin Council case (paras 32-33).

Although the Tin Council principles as such are not in doubt, they are only part of the story. It is wrong to see this as a simple choice between Parliamentary sovereignty, exercised through legislation, and the “untrammelled” exercise of the prerogative by the Executive. Parliamentary sovereignty does not begin or end with the Tin Council principles. No less fundamental to our constitution is the principle of Parliamentary accountability. The Executive is accountable to Parliament for its exercise of the prerogative, including its actions in international law. That account is made through ordinary Parliamentary procedures. Subject to any specific statutory restrictions (such as under the Constitutional Reform and Governance Act 2010 ), they are a matter for Parliament alone. The courts may not inquire into the methods by which Parliament exercises control over the Executive, nor their adequacy.

Defining the proper boundaries between the respective responsibilities of Parliament, the Executive and the courts lay at the heart of the dispute in the FBU case ( R v Secretary of State for the Home Department Ex p Fire Brigades Union [1995] 2 AC 513 ). That case concerned statutory provisions (under the Criminal Justice Act 1988 ) providing for compensation for criminal injuries, intended to replace a previous non-statutory scheme established under the prerogative. Section 171 provided that the new scheme should come into force on “such day as the Secretary of State may by order … appoint”. No such date was appointed, but instead after some years the Secretary of State announced that a new non-statutory scheme would be introduced, which was inconsistent with the scheme provided for by the Act . The House of Lords held by 3-2 that this action was an abuse of power and so unlawful. In the leading judgment Lord Browne-Wilkinson noted that the new scheme was to be brought into effect -

“… at a time when Parliament has expressed its will that there should be a scheme based on the tortious measure of damages, such will being expressed in a statute which Parliament has neither repealed nor (for reasons which have not been disclosed) been invited to repeal.

In a recent article ( A dive into deep constitutional waters: article 50, the Prerogative and Parliament (2016) 79(6) MLR 1064-1089), Professor Gavin Phillipson considers some lessons from that decision for the present case. As he points out ( ibid p 1082), the apparently fundamental difference of approach between majority and minority came down ultimately to a narrow issue of statutory construction of section 171: whether the section imposed no duty owed to the public (p 544F per Lord Keith), or rather, as the majority thought (p 551D, per Lord Browne Wilkinson), it imposed a continuing obligation on the Secretary of State to consider whether to bring the statutory scheme into force, which was frustrated by implementation of the inconsistent non-statutory scheme.

Professor Phillipson also draws attention to the important observations by Lord Mustill on the balance between the three organs of the state, and in particular the means by which Parliament exercises control of the Executive, not restricted to legislative control. Although stated in a minority judgment, the underlying principles are not I believe controversial. Lord Mustill said:

“It is a feature of the peculiarly British conception of the separation of powers that Parliament, the executive and the courts each have their distinct and largely exclusive domain. Parliament has a legally unchallengeable right to make whatever laws it thinks right. The executive carries on the administration of the country in accordance with the powers conferred on it by law. The courts interpret the laws, and see that they are obeyed. This requires the courts on occasion to step into the territory which belongs to the executive, not only to verify that the powers asserted accord with the substantive law created by Parliament, but also, that the manner in which they are exercised conforms with the standards of fairness which Parliament must have intended. Concurrently with this judicial function Parliament has its own special means of ensuring that the executive, in the exercise of delegated functions, performs in a way which Parliament finds appropriate . Ideally, it is these latter methods which should be used to check executive errors and excesses; for it is the task of Parliament and the executive in tandem, not of the courts, to govern the country …” (p 567D-F, emphasis added)

Lord Mustill went on to comment on the development over the previous 30 years of court procedures to fill gaps where the exercise of such “specifically Parliamentary remedies” has been perceived as falling short, and “to avoid a vacuum in which the citizen would be left without protection against a misuse of executive powers”. He thought these judicial developments were welcome but not without risks:

“As the judges themselves constantly remark, it is not they who are appointed to administer the country. Absent a written constitution much sensitivity is required of the parliamentarian, administrator and judge if the delicate balance of the unwritten rules evolved (I believe successfully) in recent years is not to be disturbed …” (p 567H)

Professor Phillipson comments:

“… the British constitution works most effectively when parliamentary and judicial forms of control and accountability, rather than being framed as antagonistic alternatives, or mutually exclusive directions of travel, work together, but with clearly defined, differentiated and mutually complementary roles.” (p 1089)

That observation is particularly pertinent having regard to the debate which took place on the opposite side of Parliament Square on the last day of the hearing in the Supreme Court. That led to the motion, passed by a large majority of the House of Commons, the terms of which have been set out by Lord Reed (para 163). In particular, it recognised that it is “Parliament’s responsibility to properly scrutinise the Government while respecting the decision of the British people to leave the European Union”, and ended by “call(ing) on the Government to invoke article 50 by 31 March 2017”. Of course the House of Commons is not the same as “the Queen in Parliament”, whose will is represented exclusively by primary legislation. However, the motion lends support to the view that, at least at this initial stage of service of a notice under article 50(2), the formality of a Bill is unnecessary to enable Parliament to fulfil its ordinary responsibility for scrutinising the government’s conduct of the process of withdrawal.

The logical starting point for consideration of the present case is the power which is in issue: that is, the power under article 50 of the Lisbon treaty to initiate the procedure for withdrawal by a decision in accordance with our “constitutional requirements”, followed by service of a notice. The existence of that power in international law is not in doubt. The issues for the court are, first, who has the right under UK constitutional principles to exercise it, and, secondly, subject to what constitutional requirements. As to the first, under Tin Council principles the position is clear. In the absence of any statutory provision to the contrary, the power to make or withdraw from an international treaty lies with the Executive, exercising the prerogative power of the Crown. As to the second, it is necessary to consider whether that power is subject to any restrictions by statute, express or implied, or in the common law.

In agreement with Lord Reed, and for the same reasons, I find no such restrictions in the EU statutes. I agree with Mr Eadie that this issue must be considered by reference to the statutory scheme as it exists at the time the power in question is to be exercised. The 1972 Act of course provided the framework for what followed. But I find it illogical to search in that Act for a presumed Parliamentary intention in respect of withdrawal, at a time when the treaty contained no express power to withdraw, and there was no reason for Parliament to consider it. The 1972 Act did not remove the Crown’s treaty-making prerogative in respect of European matters, whether expressly or by implication (as under the De Keyser principle: majority judgment para 48). No-one doubts the power of the Executive in 2008 to enter into the Lisbon Treaty, including article 50.

The critical issue is how Parliament dealt with that matter for the purposes of domestic law. In the 2008 Act Parliament recognised the Lisbon treaty (including article 50) by its inclusion in the treaties listed in section 1 of the 1972 Act . Thereafter it became (by virtue of 1972 Act section 2(1) ) part of the statutory framework “in accordance with” which, and therefore subject to which, any rights and obligations derived from EU law by virtue of that Act were to be enjoyed in domestic law. Unlike other powers in the treaty, the 2008 Act did not impose any restriction on the exercise of article 50 by the Executive. That position was confirmed by the 2011 Act , which made specific reference to article 50(3) but placed no restriction on article 50(2). There the matter rests today.

Turning to the common law, the Tin Council rule is simple and uncontroversial: the prerogative does not extend “to altering the law or conferring rights upon individuals or depriving individuals of rights which they enjoy in domestic law without the intervention of Parliament”. Judged by that test the answer again is clear. Service of an article 50(2) notice will not, and does not purport to, change any laws or affect any rights. It is merely the start of an essentially political process of negotiation and decision-making within the framework of that article. True it is that it is intended to lead in due course to the removal of EU law as a source of rights and obligations in domestic law. That process will be conducted by the Executive, but it will be accountable to Parliament for the course of those negotiations and the contents of any resulting agreement. Furthermore, whatever the shape of the ultimate agreement, or even in default of agreement, there is no suggestion by the Secretary of State that the process can be completed without primary legislation in some form.

This analysis was in substance adopted by Maguire J in the McCord proceedings, in line with the submissions of the Attorney General for Northern Ireland (repeated in this court). He said:

“In the present case, it seems to the court that there is a distinction to be drawn between what occurs upon the triggering of article 50(2) and what may occur thereafter. As the Attorney General for Northern Ireland put it, the actual notification does not in itself alter the law of the United Kingdom. Rather, it is the beginning of a process which ultimately will probably lead to changes in United Kingdom law. On the day after the notice has been given, the law will in fact be the same as it was the day before it was given. The rights of individual citizens will not have changed - though it is, of course, true that in due course the body of EU law as it applies in the United Kingdom will, very likely, become the subject of change. But at the point when this occurs the process necessarily will be one controlled by parliamentary legislation, as this is the mechanism for changing the law in the United Kingdom.” (para 105)

The Divisional Court (para 17) took a different approach. They in effect adopted the analysis proposed by Lord Pannick, taking account of the agreed (albeit possibly controversial) assumption that the article 50(2) notice is irrevocable. On that footing, even if it has no immediate effect, it will lead inexorably to actual withdrawal at latest two years later (subject to agreement to defer). Lord Pannick drew the analogy of a trigger being pulled (written case para 11-12):

“… it is the giving of the notice which triggers the legal effects under article 50(3). Those effects are that once notification is given, ‘[t]he Treaties shall cease to apply to the state in question’, from the date of a withdrawal agreement, or - if no such agreement is reached - at the latest within two years from notification, unless an extension of time is unanimously agreed by the European Council and the member state concerned. Notification is … the pulling of the trigger which causes the bullet to be fired, with the consequence that the bullet will hit the target and the Treaties will cease to apply.”

Lord Pannick’s trigger/bullet analogy is superficially attractive, but (with respect) fallacious. A real bullet does not take two years to reach its target. Nor is its progress accompanied by an intense period of negotiations over the form of protection that should be available to the victim by the time it arrives. The treaties will indeed cease to apply, and domestic law will change; but it is clearly envisaged that the final form of the changes will be governed by legislation. As the Secretary of State has explained, the intention is that the legislation will where possible reproduce existing European-based rights in domestic law, but otherwise ensure that there is no legal gap.

Although there is no evidence from any government witness on the intended role of Parliament, we were shown without objection or contradiction the statement made by the Secretary of State to Parliament on 10 October 2016 (Hansard Vol 615). Having described the “mandate” for Britain to leave the European Union as “clear, overwhelming and unarguable”, he explained the government’s plans for a “great repeal Bill”:

“We will start by bringing forward a great repeal Bill that will mean the European Communities Act 1972 ceases to apply on the day we leave the EU …

On the assumption that such a Bill becomes law by the time of withdrawal, there will be no breach of the rule in its classic form. The extent to which existing laws are changed or rights taken away will be determined by the legislation. Ultimately of course that result depends on the will of Parliament; it is not in the gift of the executive. But there is no basis for making the opposite assumption. Lord Pannick’s argument in effect requires the classic rule to be reformulated: “the prerogative does not extend to any act which will necessarily lead to the alteration of the domestic law, or of rights under it, whether or not that alteration is sanctioned by Parliament”. We were shown no authority to support a rule as so stated, nor any principled basis for the court to invent it. In any event, that process, like the service of the article 50 notice, will be subject to Parliamentary scrutiny in whatever way Parliament chooses. It will be for Parliament and the Executive acting in partnership to determine the timing and content of the legislative programme.

One possible answer to the analysis in the previous paragraph is that it would involve the Executive unlawfully “frustrating” or “pre-empting” the will of Parliament. This point is touched on in the majority judgment by reference in particular to the Lord Browne-Wilkinson’s statements in the FBU case (see para 250 above). They are said to establish the principle that ministers cannot “frustrate” the purpose of a statute “for example by emptying it of content or preventing its effectual operation”; and that it is -

“… inappropriate for ministers to base their actions (or to invite the court to make any decision) on the basis of an anticipated repeal of a statutory provision as that would involve ministers (or the court) pre-empting Parliament’s decision whether to enact that repeal.” (majority judgment para 51)

As I understand the majority judgment, however, this line of argument does not ultimately form part of their reasoning, in my view rightly so. In the first place, the FBU case was about abuse, not absence, of power. There was no doubt as to the existence of the prerogative power. But it was held to be an abuse to use it for a purpose inconsistent with the will of Parliament, as expressed in a statute which it had neither repealed nor been invited to repeal. Such issues do not arise in this case. The Miller respondents base their case unequivocally on absence of a prerogative power to nullify the statutory scheme set up by the 1972 Act , rather than abuse (see Lord Pannick’s response to Lord Reed: Day 2 Transcript, p 158, lines 8-25).

Further, Lord Browne-Wilkinson was not purporting to lay down any general principle about the relevance of future legislation in relation to the exercise of the prerogative. His comments were directed to the facts of the particular case, in which the new scheme was being introduced without any reference at all to Parliament. Similar arguments in the present case would have to be seen in a quite different context, which (as Lord Pannick accepts) would include the 2015 Act and the referendum result. It is one thing, as in the FBU case, to use the prerogative to introduce a scheme which is directly contrary to an extant Act, and which Parliament has had no chance to consider. It is quite another to use it to give effect to a decision the manner of which has been determined by Parliament itself, and in the implementation of which Parliament will play a central role. In such circumstances talk of frustrating or pre-empting the will of Parliament would be wide of the mark. Conversely, it would be wrong to assume (as the majority appears to do: para 91) that the courts would necessarily have been powerless in the (politically inconceivable) event of the Executive initiating withdrawal entirely of its own motion, or even in defiance of a referendum vote to remain.

I would not wish to leave the case without acknowledging the important submissions made by the other respondents and interveners, particularly as to the scale and significance of the interests which will be affected by withdrawal. It is not clear, however, how a requirement for statutory authority for the article 50(2) notice will do anything to safeguard those interests, nor indeed to advance the process of Parliamentary scrutiny which will ultimately be critical to their protection.

I take as representative the cases for the third and fourth respondents, presented by Ms Mountfield QC and Mr Gill QC. Their submissions provide vivid illustrations of the variety of ways in which individual and group interests will be profoundly affected by implementation of the decision to leave the EU. Ms Mountfield for example provides a detailed breakdown of “fundamental” and “non-replicable” EU citizenship rights. The list starts with the “fundamental status” of EU citizenship (Citizens’ Directive 2004/38/EC preamble), leading to more specific rights, such as the right to move, reside, work and study throughout the member states, the right to vote in European elections, the rights to diplomatic protection, and the right to equal pay, and to non-discriminatory healthcare free at the point of use. She categorises the government’s case as an assertion of -

“untrammelled prerogative power to do away with the entire corpus of European law rights currently enjoyed under UK law, and render a whole suite of constitutional statutes meaningless, without any Parliamentary authority in the form of a statute.”

Similar arguments are made by Mr Gill for the fourth respondents (the AB parties). They are representative, among others, of the very large numbers of EEA nationals and their children living in this country, whose rights to continued residence will be threatened unless adequate arrangements are made to protect them. Mr Gill refers in particular to the important right under the Citizens’ Directive for those who have lived in the UK for five years to apply for citizenship in the following year, a right which will be lost on withdrawal. Section 7 of the Immigration Act 1988 provides that a person shall not require leave to enter or remain in the United Kingdom “in any case in which he is entitled to do so by virtue of an enforceable EU right”.

Typical is Mrs KK, a Polish national resident and working here since 2014, married to a third country national, with a Polish national child born in the UK in 2015. She feels “in a complete state of limbo” having received no assurance from the Secretary of State as to what her status will be during and after the withdrawal negotiations, nor how her husband and child will be affected. Such people, says Mr Gill, will have made life-changing decisions and moved permanently to the UK with the ultimate intention of acquiring permanent residence. They may also find themselves exposed to criminal liability under the Immigration Act 1971 if their status is removed. Mr Gill recognises that Parliament may prior to actual withdrawal put in place a statutory protection mechanism; but that depends on the will of Parliament, which, he says, the Secretary of State cannot lawfully pre-empt. It is, he submits, a misuse of the prerogative to “foist” such a situation on Parliament; the rights to remain “must be addressed by Parliament before the giving of the article 50(2) notice.”

There are two problems with that submission. First, it is difficult to talk of the Executive “foisting” on Parliament a chain of events which flows directly from the result of the referendum which it authorised in the 2015 Act . Secondly, however desirable it would be for issues of detail such as those affecting his clients to be addressed at this stage, it is wholly inconsistent with the structure of article 50. That assumes the initiation of the process by a simple notice under article 50(2), to be followed by detailed negotiations leading if possible to an agreement on the terms of withdrawal. The details of the protections available for Mr Gill’s clients must depend, at least in part, on the outcome of those negotiations.

No doubt for this reason such an extreme argument is not adopted by the other respondents. They accept that, at this stage of the article 50 process, they cannot reasonably expect anything more than bare statutory authorisation for the service of the notice. That is realistic. But it also underlines the point that successful defence of the Divisional Court’s order will do nothing to resolve the many practical issues which will need to be addressed over the coming period, nor to protect the rights of those directly affected. Those problems, and the need for Parliament to address them, will remain precisely the same with or without statutory authorisation for the article 50 notice. If that is what the law requires, so be it. But some may regard it as an exercise in pure legal formalism.

Shortly after the 1972 Act came into force, Lord Denning famously spoke of the European Treaty as “like an incoming tide. It flows into the estuaries and up the rivers. It cannot be held back …” ( Bulmer Ltd v Bollinger [1974] Ch 401 , 418F). That process is now to be reversed. Hydrologists may be able to suggest an appropriate analogy. On any view, the legal and practical challenges will be enormous. The respondents have done a great service in bringing these issues before the court at the beginning of the process. The very full debate in the courts has been supplemented by a vigorous and illuminating academic debate conducted on the web (particularly through the UK Constitutional Law Blog site). Unsurprisingly, given the unprecedented nature of the undertaking there are no easy answers. In the end, in respectful disagreement with the majority, I have reached the clear conclusion that the Divisional Court took too narrow a view of the constitutional principles at stake. The article 50 process must and will involve a partnership between Parliament and the Executive. But that does not mean that legislation is required simply to initiate it. Legislation will undoubtedly be required to implement withdrawal, but the process, including the form and timing of any legislation, can and should be determined by Parliament not by the courts. That involves no breach of the constitutional principles which have been entrenched in our law since the 17th century, and no threat to the fundamental principle of Parliamentary sovereignty.

Some observers, who have not been provided with the very detailed arguments which have been debated before us (or the something over 20,000 pages of documents which supported those arguments) might easily think that the principal question in this case is: “Does the 2016 referendum result not conclude the issue, and mean that the country is bound to leave the EU?” In fact, that is not the principal question. No-one suggests that the referendum by itself has the legal effect that a Government notice to leave the EU is made lawful. Specifically, that is not the contention of the Government, speaking through the Secretary of State for exiting the EU. The referendum result undoubtedly has enormous political impact, but it is not suggested by the Government that it has direct legal effect.

The principal question in this case is not whether the UK ought or ought not to leave the EU. That is a matter for political judgment, which is where the referendum comes in. Courts do not make political judgments. The question in this case is not whether , but how , the UK may lawfully set about leaving the EU, if that is the political decision made. It is about the legal mechanics of leaving.

As the foregoing judgments show, this case is capable of stimulating discussion on a number of legally interesting topics. There are also supplementary questions arising out of the legal positions of Scotland, Northern Ireland and Wales. But, at some risk of over-simplifying, the main question centres on two very well understood constitutional rules, which in this case apparently point in opposite directions. They are these:

Rule 1

Nobody questions either of these two rules. Mrs Miller relies on the first. The government relies on the second. The government contends that Rule 2 operates to recognise its power, as the handler of foreign relations, to unmake the European Treaties. Mrs Miller contends that Rule 1 shows that the power to handle foreign relations stops short at the point where UK statute law is changed.

Mrs Miller’s case is that because there was an Act of Parliament (the European Communities Act 1972 ) to give effect to our joining the (then) EEC and to make European rules part of UK law, there has to be another Act of Parliament to authorise service of notice to leave. This is the effect, she says, of Rule 1. Thus, she says, Rule 2 is true, but does not apply.

The government’s case is that the European Communities Act 1972 , which did indeed make European rules into laws of the UK, will simply cease to operate if the UK leaves. The Act was only ever designed to have effect whilst we were members of the EU. It agrees that as a government it cannot alter the law of the UK which statute has made, but it says that if it serves notice to leave the EU, and in due course we leave, it would not be altering the statute; the statute would simply cease to apply because there would no longer be rules under treaties to which the UK was a party. Thus, it says, Rule 1 does not apply and Rule 2 does.

Which of these arguments is correct depends in the end on the true reading of the European Communities Act 1972 . Clearly, either reading is possible. The majority judgment gives cogent expression to the conclusion that it is Mrs Miller’s reading which is correct. For my part, for the reasons which Lord Reed very clearly sets out, I would have preferred the view that this Act was only ever to be operative for so long as the UK was a member of (first) the EEC, and now the EU. It is not helpful, particularly because this is a minority view, to repeat the analysis which Lord Reed expounds. I agree with his judgment. In short, because of Rule 1 the Act was necessary to convert the UK’s international obligations under the various European treaties into law with domestic effect. Without the Act , those European rules would have had effect between States at the international level but would not have been part of domestic UK law and so would not have bound UK citizens individually. But the Act is couched in terms which give legal effect to the obligations and rules which arise under the treaties. If the UK leaves the EU, there are no longer any treaties to which this country is a party. It seems to me to follow that the Act will cease to import any of the rules which presently it does. The Act is not changed; it does, however, cease to operate because there are no longer any treaty rules for it to bite upon.

Thus I would, for myself, have allowed the appeal of the Secretary of State from the decision of the (English) Divisional Court. I agree that on either view of the principal Miller appeal, the devolution questions raised should all be answered “no”, for the reasons set out in the majority judgment. I likewise agree with the majority’s treatment of the Sewel convention.

It remains only to add that the arguments before us made it clear that whatever the outcome of the Miller appeal, much the same legislative programme will be required in Parliament, upon the UK’s departure from the EU, to deal with the multifarious legal rules presently operative via the 1972 Act . The issues before this court do not touch this exercise, which will be a matter in any event for Parliament. The court is concerned only with the necessary procedure for the service of an article 50 notice to leave.