Lipton and another v BA Cityflyer Ltd

INTRODUCTION

Mr and Mrs Lipton were booked onto a flight operated by the appellant (“Cityflyer”) scheduled to depart from Milan Linate Airport at about 5 pm on 30 January 2018. They were supposed to arrive at London City Airport just after 6 pm local time. The flight was cancelled because the pilot reported that he was unwell shortly before the flight and it was not possible to find a replacement in time. The Liptons were rebooked onto another flight and arrived at London City Airport at about 8:45 pm that evening, just over 2.5 hours later than they expected. The Liptons put in a claim for compensation for €250 (about £220) but Cityflyer has refused to pay.

From this all too familiar tale of frustration and annoyance arise two important legal issues which have made their way up through the several tiers of the court system to occupy a full day of argument before this court.

One of those legal issues, raised by Ground 1 of this appeal, is directly related to the facts of this case. The Liptons’ claim was made under Regulation (EC) 261/2004 of 11 February 2004 (“Regulation 261”) which confers a right on passengers to compensation for cancelled flights. Regulation 261 provides the airline with a defence if it can show that the cancellation was the result of “extraordinary circumstances”. Ground 1 raises the question of what is meant by that term and whether the pilot’s illness which caused the cancellation of the Milan flight counts as an “extraordinary circumstance” or not.

The second issue is potentially of much broader significance. Between the date of the cancelled flight and the date of the hearing of the Liptons’ claim before the Court of Appeal, the United Kingdom left the European Union. The second issue is what effect, if any, that had on the Liptons’ ability to recover compensation by pursuing the right that was conferred on them by Regulation 261 whilst the UK was still an EU Member State. That issue forms Ground 2 of the appeal now before the court.

THE FACTS IN MORE DETAIL

Cityflyer is an airline. It is an “operating air carrier” within the meaning of Regulation 261.

The Liptons were booked on the Cityflyer flight due to depart from Milan at 17.05 local time on 30 January 2018 and to arrive the same day at London City Airport at 18.05 (local time).

At 16.05 local time in Milan the captain of the aircraft for the flight reported that he was unwell. At that time he was not yet on duty and was not at his place of work. He was told to speak to a medical services consultancy (Medaire), who determined that he was not fit to fly until signed off by one of their doctors. The cause of the captain’s illness was not related to his work.

The flight could not be operated without the captain. As a result of his illness the flight had to be cancelled as there was no replacement pilot available in Milan to operate it within a reasonable time.

The Liptons were re-booked onto another flight the same day. They arrived at London City Airport at 20.41, which was 2 hours and 36 minutes after their original scheduled arrival time.

THE IMPLEMENTATION OF BREXIT IN THE UNITED KINGDOM

Brexit and the implementation period

The moment that the United Kingdom ceased to be a member of the European Union is identified with great precision as 11 pm GMT on 31 January 2020. However, the process of withdrawal, in particular as regards its legal consequences, has been a complex and gradual one. Many of those consequences were dealt with by the provisions of the European Union (Withdrawal) Act 2018 (“the Withdrawal Act 2018”) which received Royal Assent on 26 June 2018 having completed its passage through both Houses of Parliament a few days earlier.

The date set for the United Kingdom to cease to be a member by the Withdrawal Act 2018, as it originally entered into force, was 29 March 2019. The term “exit day” used in that Act was defined in section 20(1) as 29 March 2019 at 11 pm (that being midnight Central European Time). Section 1 of the Withdrawal Act 2018 provided simply that “The European Communities Act 1972 is repealed on exit day”. The European Communities Act 1972 (“the ECA 1972”) had been enacted to implement the United Kingdom’s accession to the European Economic Community (as it then was) on 1 January 1973.

Section 2(1) of the ECA 1972 as originally enacted is regarded by many as a peerless example of the Parliamentary drafters’ skill. It encapsulated the effect of this country’s accession in a single sentence. It appears to have been unproblematic in achieving whatever needed to be achieved between 1 January 1973 and Brexit despite the seismic changes in EU law over that period. In its original form it said:

“2(1) 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; and the expression ‘enforceable Community right’ and similar expressions shall be read as referring to one to which this subsection applies.”

That provision therefore was one of the provisions repealed by section 1 of the Withdrawal Act 2018.

At the point when the Withdrawal Act 2018 gained Royal Assent, the arrangements for an orderly transition in terms both of future relations between the UK and the EU and in terms of the corpus of law applicable in the UK had not yet been settled. The United Kingdom and the EU concluded a treaty, called the Agreement on the Withdrawal of the United Kingdom of Great Britain and Northern Ireland from the European Union and the European Atomic Energy Community, on 24 January 2020 (“the Withdrawal Agreement”). This was signed the day after the text of the draft treaty was approved by the UK Parliament. The Withdrawal Agreement set out the agreement between the EU and the UK that the UK’s exit would be followed by a time limited transition period which would last until 11 pm on 31 December 2020. Article 127(3) of the Withdrawal Agreement provided that during that transition period, EU law “shall produce in respect of and in the United Kingdom the same legal effects as those which it produces within the Union and its Member States, and shall be interpreted and applied in accordance with the same methods and general principles as those applicable within the Union.” This meant that during that 11 month period, the EU Treaties and other EU law would continue to apply in the UK by way of transitional provision.

The Withdrawal Agreement between the EU and the UK was implemented in the United Kingdom by the European Union (Withdrawal Agreement) Act 2020 which gained Royal Asset on 23 January 2020 (“the Withdrawal Agreement Act 2020”). That Act made extensive amendments to the Withdrawal Act 2018. Further, the European Union (Withdrawal) Act 2018 (Exit Day) (Amendment) (No 3) Regulations 2019 (SI 2019/1423) amended the definition of exit day in section 20 of the Withdrawal Act 2018 so that it read 31 January 2020. This meant that 11 pm on 31 January 2020 was therefore the date and time of Brexit and the date and time when the United Kingdom became a non-Member State.

However, to reflect what had been agreed with the EU in the Withdrawal Agreement, the Withdrawal Agreement Act 2020 introduced the concept of the implementation period after Brexit, and of “IP completion day” which would mark the end of that period. In effect, therefore, much of the legal landscape would stay the same during the 11 months between Brexit actually occurring on 31 January 2020 and the completion of the implementation period on 31 December 2020. As the Explanatory Notes to the Withdrawal Agreement Act 2020 put it, the effect of the ECA 1972, as modified to give effect to the Withdrawal Agreement, was saved for the time limited implementation period: para 21. Further:

“22.

The [Withdrawal Agreement Act 2020] also modifies the saved ECA [1972] provisions to reflect the fact that the UK has left the EU, and that the UK’s relationship with EU law during this period is determined by the UK’s obligations under the Withdrawal Agreement, rather than as a Member State. The Act will also make sure that existing legislation continues to operate properly during the implementation period, despite the fact that the UK is no longer a Member State. As such, the Act will provide glosses to make clear how EU terms on the UK statute book should be read during the implementation period. …

23.

EU rules and regulations will continue to apply in the UK during the implementation period. The Act, therefore, amends the EU (Withdrawal) Act 2018 so that the conversion of EU law into ‘retained EU law’ and the domestication of historic Court of Justice of the European Union (CJEU) case law can take place at the end of the implementation period rather than on ‘exit day’. The Act defines this point in time as ‘IP completion day’ at section 39 [that is 31 December 2020].”

Section 1 of the Withdrawal Agreement Act 2020 therefore inserted a new section 1A into the Withdrawal Act 2018. This saved and amended the ECA 1972 for the purpose of giving effect to the Withdrawal Agreement. As the Explanatory Notes put it (para 76):

“Until ‘exit day’, the ECA’s purpose is to implement EU law as required by the UK’s membership of the EU; during the implementation period, by contrast, the modified and repurposed 1972 Act will implement EU law as set out in the Withdrawal Agreement.”

If the law which formed part of domestic law because of the UK’s membership of the EU had simply ceased to have effect on IP completion day, that would have left large gaps in our legal system dealing with many important aspects of our lives. Sections 2, 3 and 4 of the Withdrawal Act 2018, as amended by the Withdrawal Agreement Act 2020, dealt with carrying forward EU enactments and rights into domestic law after IP completion day. These provisions are at the heart of Ground 2 of this appeal and are set out in paras 53 to 58 below. For the moment it is enough to summarise them as follows:

Those provisions were an interim solution until the suitability of each EU enactment could be assessed in a domestic context and it could either be amended, revoked or continued in force in the form of domestic primary or secondary legislation. Whilst that detailed assessment was carried out in slower time, the EU enactments as they applied pre-Brexit needed not only to be carried forward but to be amended in the interim if they did not make sense after Brexit. For example, the term “Member State” frequently used in directly applicable EU instruments no longer included the United Kingdom. To deal with this, section 8 of the Withdrawal Act 2018 provided that a Minister of the Crown could make regulations as appropriate “to prevent, remedy or mitigate” any failure of retained EU law to operate effectively, or any other deficiency in retained EU law arising from the withdrawal of the UK from the EU. Section 8(7) restricted the power to make regulations by prohibiting regulations which made retrospective provision, or created certain criminal offences or established a public authority.

Many regulations, including those at issue in the current appeals, were expressed to take effect on exit day because they had been introduced in the period before the Withdrawal Act 2018 was amended to make IP completion day the critical date. But para 1 of Schedule 5 to the Withdrawal Agreement Act 2020 provided that, where any subordinate legislation made before exit day under a power in the Withdrawal Act 2018 provided that it would come into force on exit day, that provision was to be read instead as providing for it to come into force on IP completion day. Henceforth in this judgment the references to the Withdrawal Act 2018 are to that Act as revised after the amendment to change “exit day” to IP completion day.

Some EU laws were expressly excluded from the scope of what was carried forward by the Withdrawal Act. For example,

Another key aim of Brexit was to reduce the influence of the case law of the Court of Justice of the European Union (“CJEU”) on domestic law. This was achieved by section 6 of the Withdrawal Act 2018 which is discussed in detail below. Section 6 is headed “Interpretation of retained EU law”. It provides:

Developments since IP completion day

Further very substantial amendments have been made to the Withdrawal Act 2018 by later legislation, in particular the Retained EU Law (Revocation and Reform) Act 2023 (“the REUL Act 2023”). The purpose of these further amendments includes, amongst other things:

The version of the Withdrawal Act 2018 which the parties to this appeal agreed was the version applicable to the consideration of the Liptons’ claim is the version after it was amended by the Withdrawal Agreement Act 2020 but before it was amended by the REUL Act 2023. We have not received submissions on the effect of the REUL Act 2023 or taken its provisions into account in this judgment. Where we refer below to the provisions of the Withdrawal Act 2018 and set them out, the text is the text as amended prior to the amendments made by the REUL Act 2023. We will use the pre-2024 nomenclature to refer to the concepts introduced by the Withdrawal Act 2018. Where we refer to “Brexit” or “pre-Brexit” or “post-Brexit” we are generally referring to the completion of the implementation period and pre- and post- IP completion day (that is 31 December 2020) rather than pre- and post- exit day which was 11 months earlier on 31 January 2020.

The Trade and Cooperation Agreement and the Future Relationship Act 2020

Following Brexit, the EU and the UK as a non-Member State negotiated and concluded a treaty to govern their new relationship going forward. The EU-UK Trade and Cooperation Agreement was signed on 30 December 2020 and applied provisionally as from the end of the implementation period, in effect from 1 January 2021, coming fully into force on 1 May 2021.

The Trade and Cooperation Agreement was implemented in the United Kingdom by the European Union (Future Relationship) Act 2020 (“the Future Relationship Act 2020”). That also came into force at the end of the implementation period, on 31 December 2020. That is relevant to part of the judgment of the Court of Appeal in this case.

AIR PASSENGER COMPENSATION

Regulation 261

Regulation 261 establishes the minimum rights for passengers who are denied boarding against their will, or where their flight is delayed or cancelled. According to article 3 it applies, broadly, to passengers departing from an airport located in a Member State and to passengers arriving in a Member State from a third country if the flight is operated by an airline which is licensed in a Member State.

A passenger’s rights when their flight is cancelled are set out in article 5. The passenger must be offered assistance in the form of a “right to care” described in article 9, including meals and refreshments free of charge and hotel accommodation if it becomes necessary for the passenger to stay overnight. They must also be offered a choice between either reimbursement together with a flight back to their point of departure if the onward flight no longer serves the purpose for which they originally planned to travel or alternatively re-routing to their final destination.

By virtue of article 7 the passenger is also entitled to compensation depending on the geographical length of the flight. The compensation is €250 for all flights of up to 1,500 km, €400 for all intra-Community flights of more than 1,500 km and all other flights between 1,500 km and 3,500 km and €600 for all other flights. This compensation is halved if the re-routing flight they are offered arrives not long after their scheduled arrival time, the length of delay depending on the distance flown. The Liptons’ flight in the present case was less than 1,500 km.

Article 5(3) and (4) is central to Ground 1 of this appeal. That provides:

“3.

An operating air carrier shall not be obliged to pay compensation in accordance with Article 7, if it can prove that the cancellation is caused by extraordinary circumstances which could not have been avoided even if all reasonable measures had been taken.

4.

The burden of proof concerning the questions as to whether and when the passenger has been informed of the cancellation of the flight shall rest with the operating air carrier.”

For passengers affected by delay rather than cancellation the air carrier may be required to offer assistance in accordance with articles 8 and 9, depending on the flight distance and the length of the delay.

Article 15 provides that the air carrier cannot limit or exclude its obligations to passengers, for example by including a waiver in the contract of carriage, and article 14 provides that the air carrier must ensure that at check-in a clearly legible notice alerting passengers to their rights is displayed.

The Air Passenger Regulations 2019

One of the sets of regulations made under the power in section 8 of the Withdrawal Act 2018 (amongst other powers) was the Air Passenger Rights and Air Travel Organisers’ Licensing (Amendment) (EU Exit) Regulations 2019 (SI 2019/278) (“the Air Passenger Regulations 2019”). Part 4 amended several retained directly effective EU legislative instruments. Regulation 8 contained various amendments to the EU instrument, Regulation 261. Some of these amendments were needed because, for example, the phrase “intra-Community flights” used in Regulation 261 would no longer catch flights between the UK and airports in the EU and an air carrier licensed by the UK authority would no longer fall within the term “Community carrier”.

The Air Passenger Regulations 2019 also changed the amounts of compensation payable under article 7 of Regulation 261 when a flight is cancelled or delayed by converting them from euros to sterling and removing the distinction between intra-Community flights and flights from elsewhere. The compensation is now:

THE PROCEEDINGS BELOW

The Liptons’ claim was heard before Deputy District Judge Printer in the County Court at Portsmouth. The judge dismissed the claim in a judgment handed down on 28 June 2019, holding that the “extraordinary circumstances” defence under article 5(3) of Regulation 261 was established.

The Liptons appealed. Their appeal was heard by His Honour Judge Iain Hughes QC sitting in the County Court at Winchester in December 2019. He handed down judgment on 11 February 2020, dismissing the appeal.

In his judgment Judge Hughes dealt fully with the defence of “extraordinary circumstances” and with the CJEU case law we discuss below. He stated at para 48 that when considering whether there were extraordinary circumstances for the purposes of article 5(3), the court has to focus on the source of the material events. The court must then determine “whether the airline was in control of that cause or source”. He referred to CJEU case law which held that a delay caused by a bird strike was caused by an extraordinary circumstance but that delay caused by defective aircraft components was not. He held that the CJEU’s reasoning makes plain that if the cause was in fact external to the airline there would be no control within the meaning of the Regulation:

“57.

It must follow that where the absence of a crew member is caused by events which satisfy neither the inherency test (is that type of problem inherent to the airline’s operations?) nor the control test (is that type of problem within the airline’s control?), that absence must be an extraordinary circumstance.”

He therefore rejected the Liptons’ argument that because all employees go sick from time to time, managing sickness falls within the normal activities of an airline: para 68. He upheld the Deputy District Judge’s decision.

There was no discussion in either Deputy District Judge Printer’s or Judge Hughes’ judgments of which version of Regulation 261 applied – the text of the instrument promulgated by the EU institutions (“the EU text”) or the text of the instrument as amended by the Air Passenger Regulations 2019 (“the amended version”). The former judgment refers to the compensation claimed as the euro amount as appears in the EU text and this is not queried in the latter judgment.

The appeal to the Court of Appeal lodged by the Liptons in February 2020 raised two grounds, both related to the “extraordinary circumstances” defence. The day before the hearing in the Court of Appeal, the parties to the appeal lodged an agreed statement regarding the post-Brexit position. They had concluded that:

The Court of Appeal handed down judgment on 30 March 2021: [2021] EWCA Civ 454; [2021] 1 WLR 2545 (“Lipton”). There were two substantive judgments, by Coulson LJ and by Green LJ. Haddon-Cave LJ agreed with both judgments. The judgment delivered by Coulson LJ dealt with whether Cityflyer had made out a case of extraordinary circumstances. At the outset of his judgment, he referred to the question “as to the status” of Regulation 261 noting that counsel agreed that it formed part of domestic law by virtue of the Withdrawal Act 2018. He said at para 4 that he agreed with the path by which Green LJ arrived at that conclusion so that “this court has jurisdiction to decide this appeal”.

Coulson LJ disagreed with the conclusions of the courts below as to the meaning of the term “extraordinary circumstances”. We examine his reasons for doing so in our consideration of Ground 1 of the appeal in Section 7 of this judgment.

Green LJ agreed with the judgment of Coulson LJ on the meaning of “extraordinary circumstances”. He started his analysis of the approach to be adopted following Brexit by noting at para 53 that:

“Submissions and argument advanced to us during the appeal proceeded very much as it would have done in 2019, when the UK was a member of the EU, or even in 2020 when the transitional period (“the Transitional Period”) governing the extraction of the UK from the EU was still in force (until 11pm on 31st December 2020). However, the hearing took place in February 2021 when the Transitional Period had expired. As at this point in time a new set of legal arrangements are in place which governed the relationship of the UK to EU law. The Court cannot therefore assume that the old ways of looking at EU derived law still hold good. We must apply the new approach. There is much that is familiar but there are also significant differences.”

He then described the Withdrawal Act 2018, the Withdrawal Agreement Act 2020 and the Trade and Cooperation Agreement. The last of these was, he said, “incorporated into domestic law” by the Future Relationship Act 2020: para 56. He referred to the general principles of EU law and to section 6 of the Withdrawal Act 2018 concerning the interpretation of retained EU law.

Green LJ then referred to Regulation 261 and the Air Passenger Regulations 2019 which came into force on 31 December 2020. He referred to regulation 8 of those Regulations and said: “The cumulative effect is that the present governing law is Regulation 261/04 as amended”. He set out the text of the amendments made to Regulation 261 by the Air Passenger Regulations 2019.

He then referred to various articles of the Trade and Cooperation Agreement, as discussed further below. He said, contrary to the view shared by the parties in their note to the court, that “the [Trade and Cooperation Agreement] has something of relevance to say about the subject matter of the present dispute”. He considered that the task of the court in construing and applying Regulation 261 was to check whether the amended text of that Regulation, which he had concluded was the governing text, was consistent with the UK’s obligations under the Trade and Cooperation Agreement. If it was not, then he held that the combined effect of the terms of the Trade and Cooperation Agreement and section 29 of the Future Relationship Act 2020 was that the amended text was automatically modified without the need for further intervention by Parliament: para 78. He concluded that the amendments made by the Air Passenger Regulations 2019 as construed by Coulson LJ did not conflict with the provisions of the Trade and Cooperation Agreement which he cited. There had therefore been no automatic modification of that text pursuant to section 29 of the 2020 Act.

Although the meaning of the phrase “extraordinary circumstances” formed Ground 1 of the Airline’s grounds of appeal, challenging the Court of Appeal’s conclusion that the pilot’s illness did not amount to extraordinary circumstances, it makes sense to address Ground 2 first as that Ground raises the prior question of what law applies to the determination of the issue raised by Ground 1.

GROUND 2: WHAT LAW APPLIES AND WHY?

The delay to the Liptons’ flight took place before Brexit took effect, at a time when the EU text of Regulation 261 applied.

Subject to Cityflyer’s possible defence of “extraordinary circumstances”, the Liptons had an accrued cause of action entitling them to the compensation provided for in Regulation 261 when they landed at London City Airport. They could have asserted that claim immediately after they got back to England and, if the airline had not raised the defence of extraordinary circumstances, no doubt the compensation would have been paid and the matter resolved before Brexit. The cause of action arose because of the direct applicability and direct effect of Regulation 261. Direct applicability means that the regulation takes effect in the domestic law of the Member States of the EU without the need for any domestic law measure of implementation or transposition. Provisions have direct effect if they confer rights directly on individuals which those individuals can enforce in the domestic courts, again without the need for domestic implementing measures. The Liptons’ rights under Regulation 261, therefore, arose directly from the adoption of that Regulation under what is now article 288 TFEU, and were, on the date of the Liptons’ flight, “recognised and available in law” according to section 2(1) of the ECA 1972.

It is common ground, and rightly so, that the delay caused by the matter becoming contentious and needing to be resolved through litigation cannot deprive the Liptons of whatever cause of action they had at the time of the relevant events which gave rise to it. The question is: what is it that prevented their pre-Brexit accrued cause of action being extinguished by the UK’s exit from the EU and the repeal of section 2(1) of the ECA 1972?

The relevant provisions of the Withdrawal Act 2018

As we have already stated, section 1 of the Withdrawal Act 2018 repealed the ECA 1972 as at exit day but section 1A continued the effect of section 2(1) of the ECA 1972 until the 31 December 2020, being IP completion day.

Sections 2, 3 and 4 then dealt with what was to happen after IP completion day. So far as material they provide as follows:

“2 Saving for EU-derived domestic legislation

(1)

EU-derived domestic legislation, as it has effect in domestic law immediately before IP completion day, continues to have effect in domestic law on and after IP completion day.

...

(3)

This section is subject to section 5 and Schedule 1 (exceptions to savings and incorporation) and section 5A (savings and incorporation: supplementary).

3 Incorporation of direct EU legislation

(1)

Direct EU legislation, so far as operative immediately before IP completion day, forms part of domestic law on and after IP completion day.

(2)

In this Act "direct EU legislation" means—

(a)

any EU regulation, EU decision or EU tertiary legislation, as it has effect in EU law immediately before IP completion day and so far as—

(ai) it is applicable to and in the United Kingdom by virtue of Part 4 of the withdrawal agreement,

(bi) it neither has effect nor is to have effect by virtue of section 7A or 7B,

(i)

it is not an exempt EU instrument (for which see section 20(1) and Schedule 6) and

(iii)

its effect is not reproduced in an enactment to which section 2(1) applies,

…

(3)

For the purposes of this Act, any direct EU legislation is operative immediately before IP completion day if—

(a)

in the case of anything which comes into force at a particular time and is stated to apply from a later time, it is in force and applies immediately before IP completion day,

(b)

in the case of a decision which specifies to whom it is addressed, it has been notified to that person before IP completion day, and

(c)

in any other case, it is in force immediately before IP completion day.

(4)

This section—

(a)

brings into domestic law any direct EU legislation only in the form of the English language version of that legislation, and

(b)

does not apply to any such legislation for which there is no such version,

but paragraph (a) does not affect the use of the other language versions of that legislation for the purposes of interpreting it.

(5)

This section is subject to section 5 and Schedule 1 (exceptions to savings and incorporation) and section 5A (savings and incorporation: supplementary).

Section 4 Saving for rights etc. under section 2(1) of the ECA

(1)

Any rights, powers, liabilities, obligations, restrictions, remedies and procedures which, immediately before IP completion day-

(a)

are recognised and available in domestic law by virtue of section 2(1) of the European Communities Act 1972, and

(b)

are enforced, allowed and followed accordingly,

continue on and after IP completion day to be recognised and available in domestic law (and to be enforced, allowed and followed accordingly).

(2)

Subsection (1) does not apply to any rights, powers, liabilities, obligations, restrictions, remedies or procedures so far as they—

(a)

form part of domestic law by virtue of section 3

…

(b)

arise under an EU directive (including as applied by the EEA agreement) and are not of a kind recognised by the European Court or any court or tribunal in the United Kingdom in a case decided before IP completion day (whether or not as an essential part of the decision in the case).

(3)

This section is subject to section 5 and Schedule 1 (exceptions to savings and incorporation) and section 5A (savings and incorporation: supplementary).”

Sections 5 and 5A made provision for certain exceptions to savings and incorporation of EU related legislation and rights. So far as material section 5 provides:

“5 Exceptions to savings and incorporation

(1)

The principle of the supremacy of EU law does not apply to any enactment or rule of law passed or made on or after IP completion day.

(2)

Accordingly, the principle of the supremacy of EU law continues to apply on and after IP completion day so far as relevant to the interpretation, disapplication or quashing of any enactment or rule of law passed or made before IP completion day.

…

(4)

The Charter of Fundamental Rights is not part of domestic law on or after IP completion day.

(5)

Subsection (4) does not affect the retention in domestic law on or after IP completion day in accordance with this Act of any fundamental rights or principles which exist irrespective of the Charter …

(6)

Schedule 1 (which makes further provision about exceptions to savings and incorporation) has effect.

…”

Section 6 dealt with the binding nature, or lack of it, of judgments of the CJEU on cases being heard after IP completion day. The key provisions of section 6 are as follows:

“6 Interpretation of retained EU law

(1)

A court or tribunal—

(a)

is not bound by any principles laid down, or any decisions made, on or after IP completion day by the European Court, and

(b)

cannot refer any matter to the European Court on or after IP completion day.

(2)

Subject to this and subsections (3) to (6), a court or tribunal may have regard to anything done on or after IP completion day by the European Court, another EU entity or the EU so far as it is relevant to any matter before the court or tribunal.

(3)

Any question as to the validity, meaning or effect of any retained EU law is to be decided, so far as that law is unmodified on or after IP completion day and so far as they are relevant to it—

(a)

in accordance with any retained case law and any retained general principles of EU law, and

(b)

having regard (among other things) to the limits, immediately before IP completion day, of EU competences.

(4)

But—

(a)

the Supreme Court is not bound by any retained EU case law,

(b)

[the High Court of Justiciary is not bound by any retained EU case law in certain circumstances]

(ba) a relevant court or relevant tribunal is not bound by any retained EU case law so far as is provided for by regulations under subsection (5A), and

(c)

no court or tribunal is bound by any retained domestic case law that it would not otherwise be bound by.

(5)

In deciding whether to depart from any retained EU case law by virtue of subsection (4)(a) or (b), the Supreme Court or the High Court of Justiciary must apply the same test as it would apply in deciding whether to depart from its own case law.”

Subsections (5A) to (5D) of section 6 were inserted by the Withdrawal Agreement Act 2020 and provided for a Minister by regulation to enable certain other courts or tribunals to have power to depart from retained EU case law applying a test specified in such regulations. Regulations were made under that power: the European Union (Withdrawal) Act 2018 (Relevant Court) (Retained EU Case Law) Regulations 2020 (SI 2020/1525) (“the Relevant Court Regulations”). Those Regulations came into force on IP completion day and added a number of courts to the class of courts which could depart from retained EU case law under section 6(4). These included the Court of Appeal in England and Wales, the Inner House of the Court of Session and the Court of Appeal in Northern Ireland. The test that they were required to apply in deciding whether to depart was the same as the test in section 6(5).

Section 6(6) provided:

“Subsection (3) does not prevent the validity, meaning or effect of any retained EU law which has been modified on or after IP completion day from being decided as provided for in that subsection if doing so is consistent with the intention of the modifications.”

Section 6(7) then set out the definitions of the key concepts introduced in section 6 and used throughout the Withdrawal Act 2018:

“6(7) In this Act—

"retained case law" means—

(a)

retained domestic case law, and

(b)

retained EU case law;

"retained domestic case law" means any principles laid down by, and any decisions of, a court or tribunal in the United Kingdom, as they have effect immediately before IP completion day and so far as they—

(a)

relate to anything to which section 2, 3 or 4 applies, and

(b)

are not excluded by section 5 or Schedule 1,

(as those principles and decisions are modified by or under this Act or by other domestic law from time to time);

"retained EU case law" means any principles laid down by, and any decisions of, the European Court, as they have effect in EU law immediately before IP completion day and so far as they—

(a)

relate to anything to which section 2, 3 or 4 applies, and

(b)

are not excluded by section 5 or Schedule 1,

(as those principles and decisions are modified by or under this Act or by other domestic law from time to time);

"retained EU law" means anything which, on or after IP completion day, continues to be, or forms part of, domestic law by virtue of section 2, 3 or 4 or subsection (3) or (6) above (as that body of law is added to or otherwise modified by or under this Act or by other domestic law from time to time);

"retained general principles of EU law" means the general principles of EU law, as they have effect in EU law immediately before IP completion day and so far as they—

(a)

relate to anything to which section 2, 3 or 4 applies, and

(b)

are not excluded by section 5 or Schedule 1,

(as those principles are modified by or under this Act or by other domestic law from time to time).”

The Interpretation Act 1978

So far as is relevant, section 16 of the Interpretation Act 1978 (“the Interpretation Act”) provides:

“16.

General savings

(1)

… where an Act repeals an enactment, the repeal does not, unless the contrary intention appears,—

…

(b)

affect the previous operation of the enactment repealed or anything done or suffered under that enactment;

(c)

affect any right, privilege, obligation or liability acquired, accrued or incurred under that enactment;

…

(e)

affect any investigation, legal proceeding or remedy in respect of any such right, privilege, obligation, [or] liability …

and any such investigation, legal proceeding or remedy may be instituted, continued or enforced … as if the repealing Act had not been passed.”

The two rival analyses

The parties put forward two alternative analyses to explain how the Liptons’ cause of action arising from their cancelled flight was carried forward post-Brexit. In summary they are as follows:

Why does the choice of analysis matter? There are many respects in which the choice between the two analyses does not matter, but in some circumstances it may do. The questions that need to be addressed in respect of a pre-Brexit accrued claim like that of the Liptons are as follows:

We now consider each of these points in more detail.

Which version of Regulation 261 applies to pre-Brexit accrued causes of action?

As we have described earlier, many domestic instruments were made under section 8 of the Withdrawal Act 2018 making amendments to the EU texts of directly applicable EU legislation to come into effect on IP completion day. There is no doubt that if a flight was cancelled shortly after IP completion day, then the version of Regulation 261 that would apply to that claim would be the version as amended by the Air Passenger Regulations 2019. That makes a small difference to all such claims because the compensation is now expressed in sterling whereas the compensation under the EU instrument is expressed in euros. Other domestic regulations made under section 8 of the Withdrawal Act 2018 may have made more substantial changes to the existing EU text of whatever EU regulation they were addressing. It is important, therefore, for the courts to be clear which version they should be applying to any particular set of facts.

The Court of Appeal held that the amended version of Regulation 261 was the “present governing law”: see Lipton, para 72 per Green LJ, referring to the time the case was heard in that court. Although the circumstances giving rise to the Liptons’ claim occurred at a time when the EU text of Regulation 261 was applicable, the court applied the amended version. Since that judgment was handed down, courts have regarded themselves as bound by that to apply the amended version of Regulation 261 when considering cases where the cause of action accrued before Brexit. For example, in Air Canada v Varano [2021] 2 Lloyd’s Rep 372, Geraint Webb QC, sitting as a Deputy High Court Judge was considering the application of Regulation 261 in respect of a flight which was delayed in 2016. An issue arose about how the Regulation applied where the carrier was a non-EU carrier and the delayed flight was the second leg of a flight booked in a single booking, Heathrow to Toronto to Austin, Texas. The hearing before Judge Webb took place on 3 March 2021. He recorded at para 67 of his judgment that the parties had proceeded at the hearing before him on the basis that it was Regulation 261 that applied in its unamended form and not the amended version. On 30 March 2021, the Court of Appeal’s judgment in the present case was handed down and the judge invited submissions on its implications.

In their post-hearing submissions, the claimant, Ms Varano, confirmed that, applying Lipton, it must be the amended version of Regulation 261 which applied but that there was no substantive difference between them for the purposes of her claim. Air Canada, however, respectfully disagreed with Green LJ’s analysis in Lipton. They argued that it was the EU text which was the governing law. They also put forward what we have called the Interpretation Act analysis though they agreed with the claimant that it made no difference to the claim before the court. The judge referred to the effect of the analysis set out by Green LJ as “potentially far reaching” but concluded that he was bound to treat the claim as brought under the amended version of the Regulation. That version did not contain modifications of substance relevant to the claim.

Later in 2021, the Court of Appeal had occasion to revisit this issue in Chelluri v Air India Ltd [2021] EWCA Civ 1953; [2022] Bus LR 286. That was another case concerning a single booking of connecting flights taking the claimant from the USA to India, one of which legs – from London to Mumbai - was substantially delayed. The flights took place in May 2019, before Brexit, and the claim came before the Portsmouth County Court in January 2021. The Court of Appeal upheld the dismissal of the claim for compensation under Regulation 261 relying on the CJEU authority of Wegener v Royal Air Maroc SA (Case C-537/17) [2018] Bus LR 1366 (“Wegener”). Coulson LJ said:

“16.

Following Green LJ’s summary in Lipton, in Varano v Air Canada [2021] EWHC 1336 (QB), [2021] 2 Lloyd’s Rep 372 … Geraint Webb QC (sitting as a Deputy High Court Judge) held that it was the amended Regulation which fell to be applied, even to pending claims which were commenced prior to 31 December 2020. Neither counsel in the present case expressly disagreed with that approach, although I would not want to approve it without having heard full argument on the point. In any event, both accepted that the amendments made no substantive difference to this appeal.”

In our judgment the Court of Appeal in Lipton fell into error in holding that it was the amended version which governs the Liptons’ claim. This is contrary to a basic principle of the rule of law which Parliament must be taken to respect, according to which it is the law in place at the time the material events occur which applies, rather than some different version introduced at a later date. To analyse the position as the Court of Appeal did would produce strange results and would undermine the important value of finality in litigation. It would mean that the relevant law applicable to two identical cases which occurred on the same date might be different, depending on the time at which the relevant claims were brought and the vicissitudes of listing hearings in the respective courts in which the proceedings were commenced. It might also encourage parties to continue litigation even if the court at first instance had been completely correct in understanding the law which it was its task to apply to the case and had committed no legal error.

The approach of the Court of Appeal also ignores the wording of section 3 of the Withdrawal Act 2018 which, in the case of an EU regulation, expressly refers to the text of the instrument as that which is operative immediately before IP completion day. No doubt Parliament was well aware that on IP completion day a myriad of statutory instruments would come into force. Those instruments did not purport to apply those changes to fact patterns that arose before the changes made in the amending instrument came into force. Section 3 makes clear that what is carried forward as the instrument which governs such fact patterns is the EU text version in each case.

The parties did not draw our attention to any statutory instrument coming into force on IP completion day which purports to make a retrospective amendment to an EU instrument but nor could they rule out the possibility that such a provision exists somewhere. It could not have been made under section 8, which expressly excludes retrospective amendments, but there are other powers to make regulations which might perhaps not be so restricted and we were not taken on an exhaustive tour through them. The validity and effect of such a provision, if it indeed exists somewhere and is relevant to a cause of action accruing pre-Brexit, will need to be considered in the context in which it arises.

The role of the Trade and Cooperation Agreement

The Court of Appeal were clearly uncomfortable with the consequences of their conclusion, with which we disagree, that the Liptons’ cause of action accruing pre-Brexit was now governed by a different version of Regulation 261 from the one which would have applied if their claim had been resolved before Brexit. If the amendments made to Regulation 261 by the Air Passenger Regulations 2019 had, for example, significantly cut back the amount of compensation or removed the right completely in some circumstances, this would seem to erode or extinguish a valid accrued claim.

It seems that Green LJ’s reference to the Trade and Cooperation Agreement and the Future Relationship Act 2020 was aimed at mitigating this difficulty. Green LJ’s reasoning proceeded as follows. He cited article AIRTRN.22 of the Trade and Cooperation Agreement headed “Consumer protection”. That provides that the UK and EU “share the objective of achieving a high level of consumer protection and shall cooperate to that effect”. It provides further that the parties must ensure that “effective and non-discriminatory measures” are taken to protect the interests of consumers in air transport “including … compensation in case of … cancellation or delays”.

Green LJ then set out an article which comes from the Common Provisions part of the Trade and Cooperation Agreement, that is to say, provisions which are not related to any particular sector of economic activity:

“Article COMPROV.16: Private rights

1.

… nothing in this Agreement or any supplementing agreement shall be construed as conferring rights or imposing obligations on persons other than those created between the Parties under public international law, nor as permitting this Agreement or any supplementing agreement to be directly invoked in the domestic legal systems of the Parties.”

He concluded from this (para 75) that nothing in the Trade and Cooperation Agreement is to be construed as conferring or imposing rights or obligations on individual citizens and nothing in it can be directly invoked in UK domestic law: “In other words the [Trade and Cooperation Agreement] does not have direct effect”.

Green LJ then turned to how the Trade and Cooperation Agreement had been implemented into domestic law by the Future Relationship Act 2020. He noted that there is nothing in that Act which specifically implements Article AIRTRN.22, but this did not mean that it had not been implemented. He then set out section 29 of the 2020 Act which provides so far as relevant:

“29 General implementation of agreements

(1)

Existing domestic law has effect on and after the relevant day with such modifications as are required for the purposes of implementing in that law the Trade and Cooperation Agreement …. so far as the agreement concerned is not otherwise so implemented and so far as such implementation is necessary for the purposes of complying with the international obligations of the United Kingdom under the agreement.”

Green LJ’s conclusion as to the effect of these provisions was set out in para 78 of his judgment. He concluded that section 29 transposed the Trade and Cooperation Agreement into domestic law, “implicitly changing domestic law in the process”. Existing domestic law therefore “now means what the [Trade and Cooperation Agreement] says it means, regardless of the language used”.

He said (para 81) that courts and tribunals will in due course “confront many situations where they must interpret and apply the [Trade and Cooperation Agreement] in order to find out what the effect of domestic law is”. At para 82 he described three steps that a court or tribunal should take when so confronted. The first was to identify the relevant domestic law, which in this case he considered was the amended version of Regulation 261. The second was to determine “whether the domestic law is the same as the corresponding provisions of the [Trade and Cooperation Agreement]”. The third was that:

“… If there is inconsistency, daylight or a lacuna then the inconsistent or incomplete provision is amended or replaced, and the gap is plugged. As to this the [Trade and Cooperation Agreement] imposes a duty on the parties, in pursuit of a principle of consumer protection, to “ensure” that “effective” measures are taken to protect consumers in the field of transport including in relation to compensation for denied boarding and with ensuring “efficient handling complaint handling procedures.”

Green LJ held that the interpretation of the amended version of Regulation 261 as construed by Coulson LJ achieved the degree of consumer protection required by article AIRTRN.22 so no modification was necessary.

In Air Canada v Varano (see paras 63 and 64, above) the judge felt obliged by Lipton not only to treat the amended Regulation 261 as the governing law but to include the following paragraph at the end of his judgment:

“93 In light of Lipton, at para 82, I have considered the corresponding provisions of the [Trade and Cooperation Agreement] and am satisfied that the above construction of the [amended version of Regulation 261] accords with the requirements of the [Trade and Cooperation Agreement] and provides appropriate protection to consumers in the context of the matters under consideration in that case.”

If Green LJ is right as to the effect of section 29 of the Future Relationship Act 2020 on the body of “existing domestic law”, it is a most extraordinary statutory provision. Professor Simon Whittaker has written helpfully about the legal effects of Brexit and the Lipton decision: “Retaining European Law in the United Kingdom” (2021) 137 LQR 477. He said of section 29 as construed by Green LJ (p 489):

“…it goes much further than either the ‘reading down’ of provisions under s.3 of the Human Rights Act or the special approach required of national courts under the EU principle of conforming interpretation, both of which are subject to a condition of possibility, that is, that the words of the legislation in question are capable of being read in a way compatible with, respectively, Convention rights or EU law. There is no similar condition of possibility in s.29 and this would mean that, where there is no specific implementation of the [Trade and Cooperation Agreement], UK domestic law having effect before IP completion day must be read so as to conform to that agreement’s provisions even if this conflicts directly with the terms of that law by way of what is often called in the European context, interpretation contra legem.”

In our judgment section 29 does not have the effect Green LJ suggests in relation to provisions in the Trade and Cooperation Agreement which are expressed in such general and aspirational terms as article AIRTRN.22. The UK and EU expressly agreed in article COMPROV.16 that nothing in the Trade and Cooperation Agreement permits that agreement to be directly invoked in the domestic legal system of the UK and that it does not confer rights on individuals. Given that the aim of Brexit was to remove the influence of EU law from our domestic law, it would be entirely inconsistent with that aim and with the wording of article COMPROV.16 to interpret section 29 as having such an intrusive and automatic effect in our law.

There may be a future case arising from facts occurring post-Brexit where a party wishes to put forward an argument that the effect of section 29 is to modify the wording of some existing domestic law on which that party relies. If such a submission is made, that court or tribunal should consider submissions as to the meaning and effect of section 29 afresh. They should not have regard to what has been said by the Court of Appeal in this case. Certainly, no court or tribunal should consider itself required, as Deputy High Court Judge Webb thought he was required, to conduct its own assessment as to whether any “automatic” modification has taken place when the parties are not inviting the court to recognise any such modification.

Bringing forward an accrued cause of action

The question of why pre-Brexit accrued causes of action are not extinguished by the repeal of section 2(1) of the ECA 1972 has been considered in a number of cases and academic articles. This topic has been bedevilled by unhelpful labelling of the different choices, in particular treating “accrued EU rights” as being one option as contrasted with “retained EU rights”. There is no doubt that, putting aside Cityflyer’s defence of extraordinary circumstances, what the Liptons had as at IP completion day was an accrued cause of action arising from the EU text of Regulation 261. That remains the case regardless of the route by which that accrued cause of action can be enforced post-Brexit even though section 2(1) of the ECA 1972 has been repealed. Describing the cause of action as an “accrued EU law right” does not therefore help to answer the question.

We preface the following analysis by saying that in the light of our conclusion on the proper construction of “extraordinary circumstances” discussed below, the adoption of the Complete Code analysis or the Interpretation Act analysis does not make any difference to the outcome of the present case. What follows is therefore strictly obiter, like the discussion of this issue in the Court of Appeal. But we recognise that courts and tribunals at all levels of the judiciary are increasingly going to be grappling with cases like the Liptons’ where causes of action under EU law accruing pre-Brexit are being adjudicated upon post-Brexit. It is important that they and the parties to those disputes know the status of CJEU case law.

Our conclusion is that the Complete Code analysis and not the Interpretation Act analysis is the right one, according to the true construction of the Withdrawal Act 2018. Section 3 of the Withdrawal Act 2018 is effective not only to bring forward into domestic law as “retained EU law” the text of Regulation 261 itself as it was “operative immediately before IP completion day”, that is to say in the form of the EU text of the Regulation. It also brings forward accrued causes of actions such as the Liptons’ arising under direct EU legislation within the meaning of section 3.

Regulation 261 is an EU regulation and so, according to section 3(2)(a), “as it has effect in EU law immediately before IP completion day” it falls within the definition of “direct EU legislation”. According to EU law, EU regulations have direct effect without the need for a Member State to enact any implementing legislation. Regulation 261 “has effect in EU law” in the period up to immediately before IP completion day in two ways: (i) by stipulating the law to be applied to any new fact situations arising between the date the Regulation comes into force and the precise point of time referred to; and (ii) by conferring causes of action and requiring any causes of action accruing under that Regulation by reason of fact situations arising in that period to be recognised and enforced by the domestic courts.

Section 3(1) provides that, as direct EU legislation, “so far as operative immediately before IP completion day”, Regulation 261 forms part of domestic law on and after that day. Section 3(3) defines what it means for direct EU legislation to be “operative” at that time. Subparagraph (a) has no application in this case. It contemplates that there may be a small and specific subset of EU regulations which may be “in force” (in the language of subparagraph (a)) and “[have] effect in EU law” immediately before IP completion day (so as to qualify as direct EU legislation under section 3(2)(a)), in the sense that they have been validly promulgated by that time by the EU legislator, but which are not operative at that time because they are “stated to apply” only after that time. EU regulations in that specific category do not qualify as being operative at that time and so are not given continued effect in domestic law after IP completion day under section 3(1); and it should be noted that there could be no question of anyone having accrued a cause of action by the time of IP completion day under such an instrument precisely because it has no application before then. But Regulation 261 is not in that category. Subparagraph (b) also has no application in this case. It refers to relevant decisions which have effect in EU law and stipulates that they only qualify as “operative” immediately before IP completion day, so as to be given continued effect in domestic law under section 3(1), if they had actually been notified to the addressee by that time; otherwise, they have no such continued effect. Subparagraph (c) is the provision which applies in the present case. It aligns the meaning of “operative” with what it means for an EU instrument to be “in force immediately before IP completion day” in a straightforward sense, without the complication of a later stipulated date of application as provided for in subparagraph (a). An EU instrument is “in force” before IP completion day to the extent that it has effect in EU law and hence in domestic law (by virtue of section 2(1) of the ECA 1972), that is as explained in para 84, above.

Regulation 261 was therefore operative immediately before IP completion day in two ways which correspond with the ways in which it had effect in EU law at that time as set out in para 84, above: (i) by stipulating the law to be applied to any new fact situations which happened to arise at that time; and (ii) by requiring any causes of action which had accrued under that Regulation by reason of fact situations arising in the period of its application up to and including that time to be recognised and enforced by the domestic courts. If, notionally, the Liptons had brought their claim and it had been determined at the point in time immediately before IP completion day, Regulation 261 would have been operative to require recognition and enforcement of their accrued cause of action under that Regulation. Therefore section 3(1) has the effect that both (i) and (ii) continue to form part of domestic law after IP completion day. It follows from point (i) that the EU text of Regulation 261 continues to have prospective effect to govern new fact situations arising on and after IP completion day, subject of course to any changes introduced by domestic regulations such as the Air Passenger Regulations 2019. It follows from point (ii) that causes of action which have arisen under Regulation 261 by IP completion day continue to form part of domestic law – and so are required to be enforced – on and after that day.

In addition to the natural meaning of the language used in section 3 there are several indications that the Complete Code analysis is the correct construction of the statute. The first is that it respects what we see as the fundamental purpose of the Withdrawal Act 2018, which is to provide comprehensively for the post-Brexit legal landscape. In particular this construction makes clear how a court dealing with an accrued cause of action should apply CJEU case law, because under the Complete Code analysis the accrued cause of action counts as “retained EU law” within the meaning of section 6.

Proponents of the Interpretation Act analysis argue that the existence of section 16 of the Interpretation Act means that it was not necessary for sections 2, 3 or 4 of the Withdrawal Act 2018 to provide for the post-Brexit enforcement of pre-Brexit accrued causes of action. Hence it should be construed as not so providing. In our judgment this ignores the political as well as the legal imperatives behind the enactment of the Withdrawal Act 2018. The Withdrawal Act 2018, by repealing the ECA 1972 and providing for the incorporation of all EU law into domestic law, signalled a break with the past and a wholesale shift of all relevant legal regimes into a purely domestic framework. The Interpretation Act analysis would mean that a very extensive and valuable component of that new legal order – the ongoing domestic enforcement of pre-Brexit accrued causes of action under EU regulations – was not expressly covered by the regime at all. It was instead silently left to the application of a very different enactment which itself had nothing to do with Brexit. We do not see that as being in keeping with the drafting or the tenor of the Withdrawal Act 2018.

Further, there is a lack of clarity about the precise effect of section 16 of the Interpretation Act in relation to the repeal of an unusual provision like section 2(1) of the ECA 1972. Section 2(1) does not authorise the EU institutions to do anything; it simply provides for recognition in domestic law of legal acts and instruments of those institutions in so far as they have effect under EU law. Is a right acquired under an EU regulation which has effect in EU law and receives recognition in domestic law by virtue of section 2(1) a right acquired “under that enactment” (that is, under section 2(1))? It is not acquired under that enactment in the straightforward sense that a right directly created by a domestic statutory provision is created by that provision. Arguments could be presented for and against the application of section 16 in relation to EU rights arising under EU law given recognition by section 2(1). If a court had to resolve that question, it could and would; and the heading of section 4 of the Withdrawal Act 2018 (set out at para 52, above) might provide a clue, although the text of that provision uses different language. But it is not plausible that the drafter of the Withdrawal Act 2018 would have left such an important matter as the continued effect in domestic law of accrued EU rights up in the air and open to debate in this way.

It is true that the extensive Explanatory Notes accompanying the Withdrawal Act 2018 do not in terms mention the post-Brexit enforcement of accrued causes of action when describing the functions of the Withdrawal Act 2018 and the effect of section 3. They refer instead (para 10) to the principal purpose of the Act as being “to provide a functioning statute book on the day the UK leaves the EU. As a general rule, the same rules and laws will apply on the day after exit as on the day before”. But Parliament was clearly aware of the issue of the future enforcement of accrued causes of action given the provisions limiting pre-Brexit rights to Francovich damages discussed above. Subject to that point, the concept of a fully functional statute book naturally extends to the continued effect of rights acquired as a result of legislative provisions recognised under section 2(1) of the ECA 1972 up to IP completion day. Rights arising as a consequence of directly applicable EU regulations are the paradigm and most obvious form of right arising under EU law. It would be very odd for this potentially very large category of accrued causes of action, especially of such an obvious type, to be left outside the scope of the principal provisions of the Withdrawal Act 2018 and the basic regime for managing the Brexit transition which that Act was intended to establish; and to do so without any mention of the Interpretation Act in the legislation or the Explanatory Notes. Section 4 of the Withdrawal Act 2018 is headed “Saving for rights etc under section 2(1) of the ECA [1972]” and is a provision which contemplates the continuation of rights accrued under section 3 but makes no reference to the Interpretation Act: see below.

When the discussion of section 4 in the Explanatory Notes (paras 92-98) is brought into account, the Notes make clear, in our judgment, that sections 2, 3 and 4 of the Withdrawal Act 2018 were intended to cover everything that needed to be covered in order to provide for legal continuity after Brexit as a complete code. The Explanatory Notes indicate that section 4 is intended to catch any rights that are not covered by sections 2 and 3. See in particular para 92 which says “Section 4 ensures that any remaining EU rights and obligations which do not fall within sections 2 and 3 continue to be recognised and available in domestic law after exit. This includes, for example, directly effective rights contained within EU treaties”.

Proponents of the Interpretation Act analysis argue that various elements in the wording of section 3 are inapposite to bring forward accrued causes of action under direct EU legislation in addition to bringing forward the direct EU legislation itself. However, we do not consider that any of these points undermine the Complete Code analysis.

First, one can compare the opening words of section 3(1) with the opening words of section 4(1). Such a comparison indicates, it is argued, that section 3(1) relates only to the “direct EU legislation” instrument itself and not to rights and obligations arising under it. Section 3(2) defines that term as meaning “any EU regulation, EU decision or EU tertiary legislation, as it has effect in EU law immediately before IP completion day”. The rest of section 3 is directed at an enactment and not to causes of action arising under that enactment. This can be contrasted, it is said, with section 4(1) which expressly refers not to enactments or legislation or instruments but to “[a]ny rights, powers, liabilities, obligations, restrictions, remedies and procedures”. That wording may be effective to bring forward accrued rights under, for example, a directly effective provision of an EU Treaty or under one of the general principles but it is much wider than the more limited wording of section 3.

We do not agree that a comparison of the language of section 3(1) and section 4(1) indicates that section 3(1) does not cover accrued causes of action under direct EU legislation. The context and purpose of the two provisions is materially different. Section 3(1) is directed to achieving continuity of effect of direct EU legislation such as EU regulations in two respects (prospective application and application through recognition of accrued causes of action) and uses language which is apt to cover both: see paras 83-86, above. Section 4(1), read in conjunction with section 4(2), is concerned only with recognition of accrued causes of action under other instruments and uses language appropriate for that more limited purpose.

Reference to section 4(2) indicates that Parliament legislated on the basis that section 3(1) covered application of direct EU legislation through recognition of accrued causes of action as well as giving it prospective effect. Section 4(2) carves out of section 4(1) “any rights, powers, liabilities, obligations, restrictions, remedies or procedures so far as they form a part of domestic law by virtue of section 3”. This shows first that section 3 and section 4(1) are intended to be mutually exclusive, indicating in line with the Explanatory Notes that together they operate as a complete code. But, secondly, it shows that section 3 brought forward in domestic law precisely the same range of accrued rights and liabilities arising under direct EU legislation as were brought forward by section 4(1) in so far as they arose under other EU law instruments such as the TFEU itself. If the Interpretation Act analysis were right, one would have expected section 4(2) to refer to rights and liabilities “recognised and available under direct EU legislation” indicating that such rights and liabilities were being carved out of section 4(1) but not brought forward by section 3 – leaving them to be carried forward instead by the Interpretation Act. Also, since according to its heading, section 4 is a specific saving provision and section 16 of the Interpretation Act is a general saving provision, one would have expected the interaction with section 16 (if there was one) to have been expressly spelled out.

Furthermore, we do not consider that it makes conceptual sense in this context to distinguish between legislation on the one hand and the rights and obligations that are created by the legislation on the other. The purpose of legislation is to impose obligations and confer rights. This is particularly so here in view of the breadth of the definition given to “direct EU legislation” so as to include in particular an EU decision. The purpose of an EU decision will often be to impose a liability on a person specified in the decision and in some cases a corresponding right on someone else. So when section 3 brings forward operative direct EU legislation in domestic law after Brexit, it is natural to read that as covering causes of action which have accrued under such legislation (as defined) and are in existence immediately before IP completion day.

We do not agree, therefore, that the Withdrawal Act 2018 is silent about preserving rights and liabilities arising under direct EU legislation, leaving them to be dealt with by section 16 of the Interpretation Act. Section 4(2) of the Withdrawal Act 2018 refers to them and states that they are covered by section 3. Section 4(2) indicates that the reason these rights and liabilities are not recognised and available in domestic law by virtue of section 4(1) is because they have already been dealt with under section 3.

In addition, the phrase “which continues to be, or forms part of, domestic law by virtue of section 2, 3, 4, or 6(3) or (6)” is important in other parts of the Withdrawal Act 2018. This confirms, in our view, that this phrase is intended to be exhaustively comprehensive of the entire body of EU rights carried forward under the Act, including all accrued causes of action. For example, para 37 of Schedule 8 to the Act provides that “anything done” or “in the process of being done” in connection with “anything which continues to be, or forms part of, domestic law by virtue of section 2, 3, 4 or 6(3) or (6)”, if in force or effective immediately before IP completion day, continues to be in force or effective afterwards. This is subject to a list of exceptions. The complexity and lack of clarity involved in trying to work out how that provision, and the exceptions listed, would apply to a substantial body of accrued causes of action which are not part of domestic law by virtue of sections 2, 3 or 4 suggests that there is no such substantial body. Every cause of action accrued under EU law by IP completion day is covered by those sections. In a short Annex to this judgment we set out the detail of provisions in the Schedules to the Withdrawal Act 2018 which in our view constitute further indications in support of the Complete Code analysis interpretation of that Act.

Similarly, section 4(2)(b) of the Withdrawal Act 2018 carves out of the rights carried forward in section 4(1) any rights that would otherwise have arisen as a result of a post-Brexit CJEU or UK court ruling which declares that a particular directive confers a directly effective right. Paragraph 38 of Schedule 8 to the Withdrawal Act 2018 disapplies this so that if a UK court or tribunal issues a decision post-Brexit in a case started pre-Brexit, and holds in its decision that there is a new directly effective right, then that is carried forward by section 4(1) despite section 4(2)(b). There is no such exception from the carve out for decisions of the CJEU handed down post-Brexit in a case which started pre-Brexit. Again, we consider this level of detail as to what is brought forward, and how, is inconsistent with a construction which relies on the Interpretation Act to achieve so much of what was clearly recognised as needing to be addressed.

Parliament has made elaborate provision in the Withdrawal Act 2018 regarding how various aspects of EU law are to apply pre- and post-Brexit. Section 5 (para 53, above) shows that Parliament very deliberately intended to address in that Act the availability and enforceability of rights arising under EU law in the period before Brexit as well as afterwards. In the light of this approach, it is natural to infer that Parliament’s object was to make comprehensive provision to govern the novel legal situation which Brexit created and to remove and resolve so far as possible the legal uncertainties to which it would give rise. For example, section 5(4) provides that the Charter of Fundamental Rights is not part of domestic law after IP completion day, recognising thereby that it was part of domestic law before IP completion day. In EU law the Charter of Fundamental Rights may affect the interpretation of EU instruments which give rise to enforceable rights for individuals and may itself create such rights, so Parliament clearly contemplated that individuals might have accrued causes of action before Brexit affected by the Charter and made provision in respect of this. We address section 6 of the Withdrawal Act 2018 below and refer in the Annex to this judgment to provisions in the Schedules to that Act which demonstrate that this was Parliament’s approach. This reinforces our interpretation of the Withdrawal Act 2018 in line with the Complete Code analysis.

Another difficulty said to arise for the Complete Code analysis from the wording of section 3 is that causes of action may accrue before IP completion day under a version of direct EU legislation which is an earlier version than that which is in force immediately before IP completion day. Suppose, for example, Regulation 261 had been amended by an EU regulation between the date of the Liptons’ cancelled flight and Brexit in a way which reduced the levels of compensation payable for passengers whose flights were cancelled after the amendment came into force. The amending EU regulation would make clear that the change operated prospectively only and did not affect the compensation payable to the Liptons whose flights had been cancelled before the amendment came into effect. Clearly, the Liptons’ accrued cause of action would confer on them a right to the original higher level of compensation. If the version of Regulation 261 which is “operative immediately before IP completion day” (section 3(1)) is only the amended version with the lower compensation levels, then how can the Liptons’ full accrued cause of action be vindicated post-Brexit? Section 3(3) provides that any direct EU legislation is operative immediately before IP completion day if “it is in force immediately before IP completion day”.

We do not regard this as creating a difficulty in the circumstances posited. Many pieces of legislation, whether domestic or EU, are “operative”, “[have] effect” and are “in force” (to use the language of section 3(1), 3(2) and 3(3), respectively) at any particular moment in a number of versions depending on whether prospectively applicable amendments apply on the date of the relevant fact pattern to which they are being applied. Indeed, the Withdrawal Act 2018 itself currently exists in several versions: its original version as enacted, the version as amended by the Withdrawal Agreement Act 2020 and the version as further amended by the REUL Act 2023. Amendments made by a later enactment may be brought into force at various times as the amending legislation is commenced. Lawyers are very familiar with this process, made easier in recent years by the greater accessibility online of different versions of the same enactment as it was in force at any particular date. The parallel existence of legislation in different versions is a consequence of the principle that legislative change only operates prospectively unless an enactment clearly says otherwise. In each case the critical point is that the version of the legislation which applies is that which was applicable at the time the relevant facts arose, and in so far as its application to the facts means that a cause of action accrues that version continues to be operative, to have effect and to be in force until such time as the cause of action is satisfied: see paras 83-86, above.

Section 3(4) is consistent with the Complete Code analysis. This provides that only the English language version of the legislation is brought forward by section 3 and if there is no English version then it is not covered by that provision. The chances of someone seeking to enforce an accrued cause of action in the UK courts arising under an EU instrument of which there is no English version seem to us to be extremely remote and the provision is essentially included out of an abundance of caution. In so far as such an instrument is ever identified, subsection (4) avoids the practical difficulties which would exist for domestic courts applying legislation in a foreign language prospectively as part of our own domestic statute book. In so far as someone did have an accrued cause of action immediately before IP completion day under such a foreign language instrument, then since by virtue of section 3(4) that would not form part of domestic law by virtue of section 3, section 4(2)(a) would not have the effect of disapplying section 4(1) and it is that provision which would therefore be the governing provision. It would allow enforcement of that accrued cause of action. This simply means that section 4 would operate as the sweeper provision to pick up stray cases not covered elsewhere, as indicated in the Explanatory Notes, and as would be natural for such an extremely unusual type of case as this. Section 3(4) is careful to provide that this provision “does not affect the use of other language versions of that legislation for the purposes of interpreting” the English version. This expressly recognises but rules out any risk that the Liptons might be disadvantaged post-Brexit by being unable to rely on a more favourable nuance in, say, the French version of Regulation 261 when enforcing their claim.

Lord Lloyd-Jones in his judgment relies on articles 4 and 127 of the Withdrawal Agreement in support of the Interpretation Act analysis. With respect, we do not consider that these provisions can bear the weight he places on them. Neither party relied on them in argument. As explained at paras 14-17, above, the primary purpose of the Withdrawal Agreement was to provide for the continued implementation of EU law during a transitional period while the future relations between the UK and the EU were finalised. In other words, rather than accrued rights under EU law being transposed into retained EU law on exit day, the purpose and effect of the Withdrawal Agreement was for such rights instead to be transposed into retained EU law on IP completion day. It was not intended that causes of action accruing during the transitional period would be treated differently from causes of action accruing prior to Brexit. The Complete Code analysis treats all such causes of action in the same way.

As mentioned at para 88, above, the Withdrawal Act 2018 signalled a break with the past and a wholesale shift of all relevant legal regimes into a purely domestic framework. As amended, this included bringing to an end the temporary legal commitments made under the Withdrawal Agreement during the transitional period that sought to reflect substantively the legal relationship between the UK and the EU while the UK was still a Member State. Indications that such a break with the past was to be brought about include section 5(4) (disapplication of the Charter of Fundamental Rights), paragraph 1(2)-(4) to Schedule 1 and paragraph 39(1) to Schedule 8 of the Withdrawal Act 2018. It is also significant that article 4(5) of the Withdrawal Agreement uses the phrase “have due regard” to describe the approach required of UK courts to pre-Brexit CJEU case law. This is consistent with and reflects the wording of section 6(2) of the Withdrawal Act 2018.

Looking at the matter the other way round, we think that the Interpretation Act analysis generates a number of serious problems with the application of the Withdrawal Act 2018 and runs counter to the clear purpose of the Act.

The most significant of these is the difficulty of applying section 6 to the consideration by courts and tribunals of pre-Brexit CJEU cases. The heading of section 6 indicates that it governs the interpretation of retained EU law. If the Liptons’ cause of action counts as “retained EU law” as defined in section 6(7), as it would do according to the Complete Code analysis, then the provisions of section 6 apply without difficulty in determining the Liptons’ claim:

Under the Interpretation Act analysis, by contrast, it is not at all clear how these provisions apply. That is because according to that analysis the Liptons’ accrued cause of action does not fall within the definition of “retained EU law”, as it is not something which “continues to be, or forms part of, domestic law by virtue of section 2, 3 or 4” or section 6(3) or (6), but only by virtue of section 16 of the Interpretation Act. If that were right:

Overall, therefore, section 6 makes the best sense both linguistically and in terms of the purpose of the Withdrawal Act 2018 by adopting the Complete Code analysis we have set out. According to that analysis, pre-Brexit causes of action accrued pursuant to EU Regulations are within the scope of section 3 and hence form part of “retained EU law”. The heading of section 6 indicates that the regime in section 6 applies to them, as does the text of section 6 itself. According to this analysis Parliament has prescribed the effect to be given to CJEU case-law in relation to those accrued causes of action in the same way as it has prescribed the effect to be given to such case-law in relation to all other matters involving EU law, as one would expect.

The Interpretation Act analysis, by contrast, would mean that the ability of this court and other courts covered by section 6(4)(b) and (ba) to depart from pre-Brexit CJEU case law will in effect be postponed for many years post Brexit until the many pre-Brexit causes of action accrued under EU law and not provided for by the Withdrawal Act 2018 itself have been disposed of. It bears repeating that since there are many EU Regulations which apply across wide fields of the law and since the effect of many such Regulations under EU law was to create directly effective rights of action for individuals, the potential size of the category of such causes of action is very large. This is not a negligible matter which the drafter of the Act might have overlooked.

The proponents of the Interpretation Act analysis recognise the uncomfortable combination of the continued binding effect of both pre- and post-Brexit CJEU case law for several years with the undoubted inability to make references to the CJEU under the TFEU (see section 6(1)(b) of the Withdrawal Act 2018). Such control as the UK courts previously had to challenge binding case law by referring a question in effect inviting the CJEU to depart from one of its own decisions has been lost. According to that analysis, therefore, the UK courts would be bound by the CJEU’s decisions more tightly than they were when the UK had the power to influence the development of the law by making references.

The proponents of the Interpretation Act analysis attempt to answer this by arguing that section 6(1) applies to all accrued causes of action and one should ignore the wording of the heading: see para 108(a), above. However, if that were right then section 6 would create a peculiar and anomalous juridical animal that does not otherwise exist. As a matter of general principle decisions of the CJEU regarding the proper interpretation of EU instruments expound the meaning of those instruments as they apply at all times and are declaratory in nature (see eg Meilicke v Finanzamt Bonn-Innenstadt (Case C-292/04) [2008] STC 2267, para 34). The legal meaning of EU instruments, as authoritatively interpreted by the CJEU, is taken to be stable across time in the same way as the meaning of a domestic statute as interpreted by this court. That is subject to the ability of the CJEU (like this court) to depart from its own case-law if it concludes that its previous interpretation was mistaken (see eg SA CNL-SUCAL NV v Hag GF AG (Case C-10/89) [1990] ECR I-3711). Where the CJEU does so, then, subject to any specific direction given by the CJEU, the later judgment has declaratory effect regarding the meaning of the relevant instrument as it applies at all times. But the effect of section 6(1), according to the Interpretation Act analysis, would be to disrupt this ordinary and juristically sound picture. Although amendments made by a new EU Regulation to the text of an existing EU Regulation will generally have prospective effect only, the CJEU’s construction of the text takes effect for the benefit of all those who might, in light of that construction, have a cause of action arising from events which occurred before the CJEU’s decision. The following examples illustrate the point.

We discuss below in relation to Ground 1 the decision of the CJEU in Sturgeon v Condor Flugdienst GmbH (Joined Cases C-402/07 and C-432/07) [2010] Bus LR 1206 (“Sturgeon”). In that case the CJEU decided, somewhat unexpectedly, that the right to compensation conferred by Regulation 261 on passengers whose flights were cancelled must also be paid to passengers whose flights were delayed for a certain time. There is no doubt that this decision conferred causes of action on many passengers who had experienced a delayed flight prior to that decision but who had not thought that they were entitled to claim. Subject to any limitation period points, they could rely on the declaratory effect of that judgment by bringing a claim after Sturgeon was handed down in respect of the delay of their pre-Sturgeon flight. Suppose that the Liptons’ flight had been delayed rather than cancelled and they had made their claim based on Regulation 261 as construed by the CJEU in Sturgeon. Suppose further that after Brexit but before their claim came before the county court, the CJEU reconsidered its earlier decision and overruled Sturgeon, holding that no compensation is available for delay. Under the Complete Code analysis, by virtue of section 6(1) the new decision would not be binding but it would be open to the airline to invite this court to exercise its power under section 6(4) to depart from Sturgeon. Under the Interpretation Act analysis, the court is bound by Sturgeon and although in theory it can “have regard” to the post-Brexit ruling under section 6(2), it is difficult to see that as conferring an unconstrained power to depart from Sturgeon given that no test is envisaged (by contrast with the detailed regime in subsections (4) and following).

The Interpretation Act analysis in relation to section 6(1) would therefore create a cause of action arising from an enactment where the enactment must be construed in accordance with all declaratory case law up to a certain date but where the claim cannot be affected by any later case law declaring the true construction of the same enactment.

We recognise that the effect even of the Complete Code analysis is that a court considering the Liptons’ claim post-Brexit is in a different position from the court considering such a claim now if the UK had not left the EU. If the UK had not exited the EU then all CJEU case law would be binding. But a fundamental object of Brexit was to remove the supremacy of the CJEU. In any event, the fact that the cause of action accrued before IP completion day may be a factor that a court can take into account when deciding how to “have regard to” a post-Brexit CJEU judgment under section 6(2) of the Withdrawal Act 2018.

The issue of how section 6 of the Withdrawal Act 2018 applies to pre-Brexit causes of action was considered by the Competition Appeal Tribunal (“the CAT”) in Umbrella Interchange Fee Claimants v Umbrella Interchange Fee Defendants [2023] CAT 49, [2023] 5 CMLR 24 (“Umbrella Interchange”). That case is part of the ongoing litigation arising out of the interchange fees charged by the entities issuing Mastercard and Visa credit cards. The issue that gave rise to this judgment was the applicable limitation period and in particular the effect, if any, of the post-Brexit ruling of the CJEU in Volvo AB v RM (Case C-267/20) [2023] 5 CMLR 17 handed down on 22 June 2022 (“Volvo”). The claimants asserted that the Volvo decision meant that, as a matter of EU law, the limitation period for their claims under the competition provision in article 101 of the TFEU only started to run once the defendants’ infringement of that provision came to an end. That would mean that no part of their claim (all of which arose pre-Brexit) was time barred. The claimants argued further that this overrode any limitation provision in domestic law to the contrary. The CAT was unanimous in holding that properly understood, the CJEU had not come to that conclusion in Volvo. But they also addressed the question of whether section 6 of the Withdrawal Act 2018 would operate in relation to such a post-Brexit ruling if Volvo had been authority for that proposition.

This required the tribunal to address whether the claimants’ cause of action amounted to “retained EU law”, in which case section 6(1) of the Withdrawal Act 2018 would apply. If that were correct, the CAT was not bound by Volvo but according to subsection (2) could have regard to it. The alternative argument put forward by the claimants relied on what we have called the Interpretation Act analysis. The claimants argued that their cause of action was not “retained EU law” but was brought forward by section 16 of the Interpretation Act such that it was not within any part of section 6 of the Withdrawal Act 2018. The consequence was, it was argued, that the post-Brexit Volvo decision was fully binding on the CAT. The majority of the CAT (Marcus Smith J (President) and Ben Tidswell) held as follows at para 68:

The majority regarded the contention that section 6(1) did not apply to the adjudication of the claimants’ claim as unsustainable. The claimants’ reliance on the Interpretation Act analysis would, the majority said, “cut across the whole purpose of the 2018 Act and to create a regime which is both unreasonably complex and uncertain, which we anticipate Parliament would have wished to avoid.”: para 71. If the Volvo decision had had the effect on the limitation period for which the claimants contended, then the CAT would not be bound by it because of section 6(1) of the Withdrawal Act 2018. They could still have regard to it in accordance with section 6(2) but they would have decided not to apply it (if it meant what the claimants said it meant): it would be inappropriate and unnecessary to adjust the limitation period applicable in England.

The majority referred to the Court of Appeal’s judgment in the present case and to other authorities to which we were also taken including Balogun v Secretary of State for the Home Department [2023] EWCA Civ 414 and News Corp UK and Ireland Ltd v Revenue and Customs Comrs [2023] UKSC 7, [2024] AC 89. We agree with their view that these cases do not provide relevant guidance. We particularly agree that the obiter dicta in the latter case (at para 7) should not be read as saying anything which is inconsistent with the Complete Code analysis set out in this judgment.

Roth J, the third member of the CAT in Umbrella Interchange, agreed with the majority that the Volvo ruling did not say what the claimants asserted: para 84. He expressed the critical question at para 100 as whether rights which arose before IP completion day which were at that time determined and governed by EU law become converted by the Withdrawal Act 2018 into rights under retained EU law, and thus rights under domestic law. He agreed with the majority that the Withdrawal Act 2018 converted or adopted directly applicable EU law into “retained EU law” and so covered accrued rights: (para 102):

“the statutory scheme, considered as a whole, indeed has, in effect, converted or ‘translated’ rights which had accrued under EU law into rights under retained EU law, save insofar as the 2018 Act otherwise expressly provides. Put another way, as from IP completion day, the 2018 Act becomes the basis upon which such rights may be pursued or asserted in the United Kingdom.”

He disagreed with the majority in considering that section 16 of the Interpretation Act would be engaged by the repeal of section 2(1) ECA 1972 unless its operation was expressly excluded. However, on his analysis of the Withdrawal Act 2018, section 16 had been excluded, by amongst other provisions, section 6 of the 2018 Act. He therefore concluded that rights which accrued under EU law prior to IP completion day became rights under retained EU law for the purposes of the Withdrawal Act 2018 so that by virtue of section 6(2) any post-Brexit CJEU rulings were not binding. He referred to the decision of the Court of Appeal in Tower Bridge GP Ltd v Revenue and Customs Comrs [2022] EWCA Civ 998; [2022] STC 1324, a VAT case where the transactions took place in 2009, long before Brexit. The Court of Appeal in that case had to consider two post-Brexit rulings of the CJEU and declined to follow one of them on the basis that it was at odds with the previous jurisprudence of the CJEU: para 119. Roth J in Umbrella Interchange noted that it appeared that Tower Bridge had proceeded on the common assumption that the case was to be determined on the basis of retained EU law so that section 6(2) of the Withdrawal Act 2018 applied. A similar assumption was made in Warner Music UK Ltd v TuneIn Inc [2021] EWCA Civ 441; [2021] Bus LR 1119, described by Roth J at para 126 of his judgment.

The judgments in Umbrella Interchange and of the Court of Appeal in the present case have been the subject of some academic discussion.

In the article by Professor Whittaker referred to above he does not directly address the choice between the Complete Code analysis and Interpretation Act analysis but discusses what he describes as the “temporal effect” of a decision of this court if it departed from pre-Brexit CJEU case law pursuant to the power in section 6 of the Withdrawal Act 2018: see (2021) 137 LQR 477, 486-488. He recognises the difficulty that the new interpretation of the EU instrument would operate retrospectively given the declaratory nature of such rulings. His solution to this is to suggest that a decision by this court to depart from pre-Brexit CJEU case law has retrospective declaratory effect but only back to IP completion day and not beyond. He regards the category of “retained EU law” as something different from what he calls “European EU law” and notes that section 6 only applies to the interpretation of retained EU law. If this court decided to depart from pre-Brexit CJEU case law in its construction of Regulation 261, then that is only the court’s interpretation of “retained” EU case law and not its interpretation of “European” EU case law. That interpretation could not, therefore, affect any claims in respect of flights cancelled before IP completion day because those are governed by “European” EU law and not by “retained” EU law. We recognise that this might have been an innovative solution to the perceived problem. But if the provisions should be construed as creating a new and unusual kind of court judgment which is declaratory of the law but only back to a certain date, we would have expected to see that made express.

In proposing that solution, Professor Whittaker disagrees with Green LJ’s conclusion that it is the amended version of Regulation 261 which applies to the Liptons’ appeal: see footnote 74. But in that same footnote Professor Whittaker notes the “different and more difficult issue” about how section 6(1), which appears to apply more generally, would apply to the application of post-Brexit CJEU rulings to pre-Brexit fact patterns – the issue that arose in Umbrella Interchange. He notes that “[i]t may be thought that a UK court should follow this case law where the UK court applies legislation to cases arising before IP completion day on the basis that this case law provides the best evidence of the proper interpretation of EU law at the earlier time”. But he seems to accept that section 6(1) applies more generally to both “retained” EU law and “European” EU law.

The two possible analyses we have discussed are considered more directly in blog-posts by Jack Williams. In his fourth post on the topic (5 January 2022) he supports the Interpretation Act analysis. He argues that as there is no suggestion that the repeal of the ECA 1972 was intended to apply retrospectively, accrued rights will continue by virtue of section 16 of the Interpretation Act. He therefore appears to regard the decision of the Court of Appeal in Tune-In as having been mistaken in debating whether to depart from retained EU case law. Because the facts arose before IP completion day it seems he would regard the Court of Appeal as being bound by pre-Brexit CJEU case law. Mr Williams describes the three delayed flight cases (Lipton, Varano and Chelluri, above) as examples of cases “where things have unfortunately not worked as they should have”. He also disagrees with the CAT’s decision in Umbrella Interchange because the tribunal rejected the Interpretation Act analysis: see his blog of 5 September 2023 “The CAT goes astray on accrued EU law rights”. The reason Mr Williams gives for adopting the Interpretation Act analysis is that section 4 of the Withdrawal Act 2018 brings forward only the EU instrument and not the rights accrued under the instrument.

However, with respect, we find this difficult to understand. We do not find the distinction between the EU instrument and the rights accruing under it persuasive in this context (see para 96, above) and it seems clear that according to its ordinary meaning, section 4(1) covers such a case. Where a cause of action for breach of a Treaty right has accrued prior to Brexit, there is no question of disapplication of section 4(1) by virtue of section 4(2). In the language of section 4(1), the cause of action is a “right” of the claimant (with a corresponding “liability” and “obligation” on the part of the defendant). Relevant “remedies” in respect of that right were recognised and available in domestic law by virtue of section 2(1) of the ECA 1972, and “enforced, allowed and followed accordingly”. That would have been the position until IP completion day. Section 4(1) thus provides that those matters “continue … to be recognised and available in domestic law (and to be enforced, allowed and followed accordingly)” on and after IP completion day.

It is also relevant to note that since EU Regulations are often in a form which puts determinate flesh on the bones of Treaty rights, it is difficult to infer that the drafter of the Withdrawal Act 2018 intended that the approach to their interpretation and application, in the light of any relevant CJEU case law, should be any different. That would be confusing and would undermine the legal certainty which the Withdrawal Act 2018 was supposed to promote. This is a further reason which reinforces the Complete Code analysis of the interpretation of that Act which we have set out above.

We have considered carefully the points raised in these thoughtful and interesting commentaries but they do not in our view undermine the conclusion that looking at the Withdrawal Act 2018 as a whole and against the background of the UK’s withdrawal from the EU, the better construction of that Act is that the Complete Code analysis applies as set out above. We conclude, therefore that the Liptons’ cause of action accruing on the day their flight was cancelled is enforceable post-Brexit pursuant to section 3 of the Withdrawal Act and therefore forms part of “retained EU law”. To summarise our discussion in this section, that construction of the Withdrawal Act is to be preferred for the following reasons.

First, the legal and political imperatives behind the enactment of the Withdrawal Act 2018 point in favour of sections 2 to 6 bringing about a complete shift of all relevant EU legal regimes into the new purely domestic framework. There is no distinction between the enactment brought forward and the causes of action which have accrued under that enactment – that is clear from the wording of section 4 which makes clear that all rights under EU law are brought forward by one or other of sections 2, 3 and 4.

Secondly, the lack of clarity about whether and how section 16 of the Interpretation Act would apply to different causes of action militates against the drafter having left this to be dealt with by legislation which is not referred to in the Withdrawal Act 2018 or its Explanatory Notes. Further, the detailed provisions throughout the Act dealing with when and to what extent particular accrued causes of action, for example claims relying on rights under the Charter or claims for Francovich damages indicate that the Act is comprehensive in its treatment of accrued rights.

Thirdly, the Interpretation Act analysis creates serious difficulties with the application of section 6 because under that analysis accrued rights would not amount to “retained EU law”. That analysis would postpone for a considerable time the ability of courts empowered to depart from retained EU law to do so, despite having removed the ability of those courts to make a reference to the CJEU for a preliminary ruling.

We therefore address Ground 1 of Cityflyer’s appeal on the basis that:

GROUND 1: THE MEANING OF “EXTRAORDINARY CIRCUMSTANCES”

The test for identifying “extraordinary circumstances”

Ground 1 of the appeal is concerned with the meaning of the phrase “extraordinary circumstances” in article 5(3) of Regulation 261 (para 30, above). That phrase is not defined in the Regulation. It falls to be interpreted in the light of the general object of the Regulation and having regard to the specific object of article 5(3).

As regards the object of the Regulation, in its judgment in Wallentin-Hermann v Alitalia (Case C-549/07), [2009] Bus LR 1016 (“Wallentin-Hermann”), at para 18, the Grand Chamber of the CJEU said that the purpose of Regulation 261 was to ensure “a high level of protection for passengers and take account of the requirements of consumer protection in general, in as much as cancellation of flights causes serious inconvenience to passengers.” In Sturgeon, Advocate General Sharpston, at point 44 of her opinion, appropriately summarised the effect of previous CJEU authority as requiring that the Regulation had to be examined, not from the air carrier’s perspective, but from the perspective of the passenger; and see Sturgeon, paras 44-60 of the judgment.

Article 5(3) reflects recitals 14 and 15 to Regulation 261, which state:

“(14)

As under the Montreal Convention [the Convention for the unification of certain rules for international carriage by air (1999)], obligations on operating air carriers should be limited or excluded in cases where an event has been caused by extraordinary circumstances which could not have been avoided even if all reasonable measures had been taken. Such circumstances may, in particular, occur in cases of political instability, meteorological conditions incompatible with the operation of the flight concerned, security risks, unexpected flight safety shortcomings and strikes that affect the operation of an operating air carrier.

(15)

Extraordinary circumstances should be deemed to exist where the impact of an air traffic management decision in relation to a particular aircraft on a particular day gives rise to a long delay, an overnight delay, or the cancellation of one or more flights by that aircraft, even though all reasonable measures had been taken by the air carrier concerned to avoid the delays or cancellations.”

The lists in recital 14 and recital 15 give an indication of the type of events which the drafter understood would constitute extraordinary circumstances. They assist to some degree in working out just how unusual an event has to be to constitute something “extraordinary”.

The case-law of the CJEU concerning the interpretation and application of article 5(3) is mature and settled. The basic principles to be applied have been articulated and there are a number of practical examples provided by individual authorities which fall on one side or the other of the “extraordinary circumstances” line.

Clearly, in applying article 5(3) the court has to seek to follow and apply the basic principles. Individual cases may provide analogies, but there is a limit to which they can provide assistance. Since Regulation 261 exists for the protection of consumers and gives rise to many small value claims usually handled by lower courts with minimal involvement of lawyers, it is appropriate to identify the patterns of cases in terms of broad categories which are straightforward and easy to apply.

The language employed in article 5(3) is important. Clearly, in light of recitals 14 and 15, full value has to be given to the word “extraordinary”.

Further, since article 5(3) derogates from the purpose of Regulation 261 and the normal rules it lays down providing for rights to receive compensation, the concept of “extraordinary circumstances” has to be read according to a strict interpretation: Wallentin-Hermann, para 20; Siewert v Flugdienst GmbH (Case C-394/14), judgment of 14 November 2014, (“Siewert”), para 16; Krüsemann v TUIfly GmbH [2018] Bus LR 1191 (“Krüsemann”), para 36.

Naturally enough, Mr Akhil Shah KC for Cityflyer focused in his submissions on particular situations addressed in the case-law which appeared to offer some basis for argument by analogy which could support his contention that the sudden illness of the captain in this case constituted extraordinary circumstances in the requisite sense. For example, he particularly relied on Pešková v Travel Service as (Case C-315/15) [2017] Bus LR 1134 (“Pešková”) concerning delay caused by a bird-strike and Moens v Ryanair Ltd (Case C-159/18) [2019] Bus LR 2041 (“Moens”), in which it was held that a delay caused by closure of a runway owing to a petrol spill was an extraordinary circumstance, where the petrol in question did not emanate from an aircraft of the carrier that operated the flight. Conversely, Mr Michael Rawlinson KC for the Liptons referred to other illustrations which appeared to offer potential analogies to support their case that pilot illness is not an extraordinary circumstance for the purpose of article 5(3).

However, references in this way to single instances of application of article 5(3) are not very illuminating save insofar as they might illustrate an underlying pattern which may provide guidance. In our view, the appropriate course is to begin the analysis, as the Court of Appeal did, with reference to the basic principles underlying the application of article 5(3) as explained by the CJEU. The approach to be followed is summarised by the CJEU in Wallentin-Hermann at para 17:

“It is settled case law that the meaning and scope of terms for which Community law provides no definition must be determined by considering their usual meaning in everyday language, while also taking into account the context in which they occur and the purposes of the rules of which they are part. Moreover, when those terms appear in a provision which constitutes a derogation from a principle or, more specifically, from Community rules for the protection of consumers, they must be read so that that provision can be interpreted strictly: see, to that effect, easyCar (UK) Ltd v Office of Fair Trading (Case C-336/03) [2005] ECR I-1947, para 21 and the case law cited. Furthermore, the Preamble to a Community measure may explain the latter’s content: see, to that effect, inter alia, R (International Air Transport Association) v Department for Transport (Case C-344/04) [2006] ECR I-403, para 76.”

The CJEU referred to the list in recital 14 and stated the test for “extraordinary circumstances” under article 5(3) as follows:

“23.

Although the Community legislature included in that list ‘unexpected flight safety shortcomings’ and although a technical problem in an aircraft may be amongst such shortcomings, the fact remains that the circumstances surrounding such an event can be characterised as ‘extraordinary’ within the meaning of article 5(3) … only if they relate to an event which, like those listed in recital 14 …, is not inherent in the normal exercise of the activity of the air carrier concerned and is beyond the actual control of that carrier on account of its nature or origin.

24.

In the light of the specific conditions in which carriage by air takes place and the degree of technological sophistication of aircraft, it must be stated that air carriers are confronted as a matter of course in the exercise of their activity with various technical problems to which the operation of those aircraft inevitably gives rise. It is moreover in order to avoid such problems and to take precautions against incidents compromising flight safety that those aircraft are subject to regular checks which are particularly strict, and which are part and parcel of the standard operating conditions of air transport undertakings. The resolution of a technical problem caused by failure to maintain an aircraft must therefore be regarded as inherent in the normal exercise of an air carrier’s activity.

25.

Consequently, technical problems which come to light during maintenance of aircraft or on account of failure to carry out such maintenance cannot constitute, in themselves, ‘extraordinary circumstances’…

26.

However, it cannot be ruled out that technical problems are covered by those extraordinary circumstances to the extent that they stem from events which are not inherent in the normal exercise of the activity of the air carrier concerned and are beyond its actual control. That would be the case, for example, in the situation where it was revealed by the manufacturer of the aircraft comprising the fleet of the air carrier concerned, or by a competent authority, that those aircraft, although already in service, are affected by a hidden manufacturing defect which impinges on flight safety. The same would hold for damage to aircraft caused by acts of sabotage or terrorism.

…

36.

As was stated at para 27 of this judgment, it is for the referring court to ascertain whether the technical problems cited by the air carrier in question in the main proceedings stem from events which are not inherent in the normal exercise of its activity and are beyond its actual control. It is apparent from that that the frequency of the technical problems experienced by an air carrier is not in itself a factor from which the presence or absence of ‘extraordinary circumstances’ within the meaning of article 5(3) … can be concluded.”

There are two limbs of the test to be applied: (i) determination whether the relevant event is inherent in the normal activity of the carrier and (ii) determination whether the carrier has a requisite degree of control in relation to the occurrence of that event and its consequences. We agree with the analysis of Elias LJ in Jet2.com Ltd v Huzar [2014] EWCA Civ 791; Bus LR 1324 (“Huzar”). That case concerned a delay caused by a wiring defect in the aircraft which could not have been prevented by prior maintenance or inspection. Applying the test in Wallentin-Hermann, the court held that this did not constitute an extraordinary circumstance. Difficult technical problems, including unforeseeable ones, arose as a matter of course as part of, and inherent in, the carrier’s ordinary activity. As to whether the CJEU was setting out a single or a dual test, Elias LJ said:

“47.

In my judgment, therefore, for all these reasons the appeal fails even on the assumption that the concept of extraordinary circumstances should be defined by reference to a single composite test and not two distinct conditions. If the appellant is right about there being a single composite test, then in my judgment it is essentially as the respondent described it. The second limb will take its meaning from the first rather than vice versa. The event causing the technical problem will be within the control of the carrier if it is part of the normal everyday activity which is being carried on and will be beyond the carrier’s control if it is not.

48.

I am inclined to think that this is indeed the correct analysis. I recognise that it can be said to render the second limb redundant. But it does not in my view strip the limb of all significance. It helps identify the parameters of those acts which can properly be described as inherent in the carrier’s normal activities and those which cannot; and it also chimes with the examples of events identified in recitals (14) and (15) as being potentially capable of constituting extraordinary circumstances. It makes it clear that events which are beyond the control of the carrier because caused by the extraneous acts of third parties, such as acts of terrorism, strikes or air traffic control problems, or because they result from freak weather conditions, cannot be characterised as inherent in the normal activities of the carrier. It is not fanciful to suggest that there may otherwise be an argument that they can be so described; indeed, [counsel for the carrier] advanced that very argument in the course of his submissions. So on this analysis the second limb is intended to help elucidate the scope of the first but is not intended to establish a distinct and independent condition.”

As Elias LJ explained, although the test laid down in Wallentin-Hermann has two limbs, the principal question is the first (which for shorthand, like Coulson LJ, we will call the test of “inherency”).

The case law of the CJEU on causes of cancellation and delay

Turning to particular instances relied on by the parties regarding the scope and operation of article 5(3), Coulson LJ helpfully grouped them under three general headings: mechanical defects in the aircraft; external or one-off events; and staff absence. We agree that this provides a helpful scheme of analysis and draw on his judgment in what follows.

Mechanical defects in the aircraft

We were shown no case in which a carrier has successfully maintained a submission that a mechanical defect in the aircraft itself amounted to an extraordinary circumstance. In Wallentin-Hermann the CJEU concluded at para 25 (para 143, above) that mechanical problems coming to light during maintenance or arising because of a failure to carry out maintenance were not extraordinary circumstances. See also Sturgeon, para 70, to similar effect. In Huzar the technical problem was held not to amount to extraordinary circumstances, even though it was unforeseeable and was not preventable. At para 36 Elias LJ noted that “difficult technical problems arise as a matter of course in the ordinary operation of the carrier’s activity. Some may be foreseeable and some not but all are, in my view, properly described as inherent in the normal exercise of the carrier’s activity. They have their nature and origin in that activity; they are part of the wear and tear.” We agree.

The same approach has been adopted in more recent cases. In van der Lans v Koninklijke Luchtvaart Maatschappij NV (Case C-257/14) [2015] Bus LR 1107, the CJEU held that, whilst a breakdown due to the premature malfunction of certain components could be described as an unexpected event, air carriers were confronted with such problems as a matter of course, and the malfunction was therefore inherent in the normal exercise of the air carrier’s activity. In A v Finnair Oyj (Case C-832/18) [2020] Bus LR 1002, the same result was arrived at, even though the flight was delayed by the failure of an “on condition” part (that is, one which was only due to be replaced when it became defective).

External or one-off events

The mere fact that the cancellation was caused by an external event (including an event perpetrated by a third party) does not necessarily comprise an extraordinary circumstance within the meaning of article 5(3). But some other one-off events have been so categorised, by application of the inherency test.

In Siewert the flight was the subject of a lengthy delay when the aircraft was damaged by a set of mobile boarding stairs in the course of a preceding flight. The stairs were “operated by the airport and not the airline”. The CJEU rejected the submission that this constituted extraordinary circumstances, observing at para 19:

“However, as regards a technical problem resulting from an airport’s set of mobile boarding stairs colliding with an aircraft, it should be pointed out that such mobile stairs or gangways are indispensable to air passenger transport, enabling passengers to enter or leave the aircraft and accordingly, air carriers are regularly faced with situations arising from their use. Therefore, a collision between an aircraft and any such set of mobile boarding stairs must be regarded as an event inherent in the normal exercise of the activity of the air carrier. Furthermore, there is nothing to suggest that the damage suffered by the aircraft which was due to operate the flight at issue was caused by an act outside the category of normal airport services (such as an act of sabotage or terrorism) [so as to fall within Wallentin-Hermann, para 26, as would be necessary to qualify as an extraordinary circumstance].”

By contrast, in Pešková the CJEU, despite the contrary opinion of the Advocate General, held that a delay due to a bird strike amounted to extraordinary circumstances. The reasoning of the CJEU is brief:

“23.

Conversely, it is clear from the court’s case law that the premature failure of certain parts of an aircraft does not constitute extraordinary circumstances, since such a breakdown remains intrinsically linked to the operating system of the aircraft. That unexpected event is not outside the actual control of the air carrier, since it is required to ensure the maintenance and proper functioning of the aircraft it operates for the purposes of its business....

24.

In the present case, a collision between an aircraft and a bird, as well as any damage caused by that collision, since they are not intrinsically linked to the operating system of the aircraft, are not by their nature or origin inherent in the normal exercise of the activity of the air carrier concerned and are outside its actual control. Accordingly, that collision must be classified as ‘extraordinary circumstances’ within the meaning of Article 5(3)…”

The reasoning of the CJEU does not demonstrate any intention to depart from the basic principles of application of article 5(3) as laid down in Wallentin-Hermann; nor are other cases on the application of article 5(3) disapproved or taken to be inconsistent with it. Clearly, in applying those principles, some individual cases may fall closer to the dividing line than others and Pešková appears to be one of these. In our view it does not illustrate a general pattern of application of that provision in a manner which provides material support for Cityflyer’s submissions.

Other examples of one-off events giving rise to a finding of extraordinary circumstances can be found in Moens; Germanwings GmbH v Pauels (Case C-501/17) [2019] Bus LR 1122, where it was held that a tyre damaged by a foreign object on the runway was not something inherent in the normal exercise of the activity of an air carrier; and LE v Transport Aereos Portugueses SA (Case C-74/19) [2020] Bus LR 1503, where it was held that the presence of a seriously unruly passenger which justified the pilot in diverting the flight to have them removed was not inherent in the activity of the carrier. It therefore constituted an extraordinary circumstance (although this was accompanied by the observation that, since such behaviour was within the control of the carrier, such a finding would be precluded if the carrier contributed to the occurrence of the unruliness or failed to take appropriate measures early enough). The same comment applies as in relation to Pešková, above.

In a similar category are cases concerned with unusual weather events such as the Icelandic volcano dust cloud (McDonagh v Ryanair Ltd (Case C-12/11) [2013] 2 All ER (Comm) 735) and air traffic control decisions to close particular routes or airports (Blanche v Easyjet Airline Co Ltd [2019] EWCA Civ 69: [2019] Bus LR 1258). These are further instances of extraordinary circumstances. Such events are clearly outside the control of air carriers, are highly unusual and indeed are outside the ordinary working of the general air transportation system as a whole. 

Staff absence

At the time of the hearing in the Court of Appeal, as Coulson LJ pointed out, there were very few cases concerned with the application of article 5(3) to situations where staff absence has caused the cancellation of or significant delay to a flight and no authority at all dealing with staff illness. The authorities under this heading are principally concerned with non-attendance due to strike action. To these has now been added the post-Brexit CJEU judgment in TAP Portugal v Flightright GmbH (Joined Cases C-156/22 to C-158/22) [2023] Bus LR 875 (“TAP Portugal”).

Finnair Oyj v Timy Lassooy (Case C-22/11) [2013] 1 CMLR 18 (“Finnair”) concerned a case of “denied boarding”, as covered by article 4 of Regulation 261. A strike by staff at Barcelona airport on 28 July 2006 meant that a scheduled flight to Helsinki that day had to be cancelled. This led the carrier to use a flight scheduled for the morning of 29 July 2006 to transport the passengers on the cancelled flight in order to minimise their delay. This had the effect that the claimant with a ticket for the 29 July flight was denied boarding on that flight and was taken to Helsinki on another flight on 30 July 2006. The CJEU held that this constituted “denied boarding” for which compensation had to be paid under article 7. Further, the defence in article 5(3) could not be made out in the case of denial of boarding an aircraft scheduled two days after the strike in question. The CJEU contrasted this with the reference to “the impact of an air traffic management decision in relation to a particular aircraft on a particular day” in recital 15: para 37. The CJEU seems to have assumed that the strike itself might amount to extraordinary circumstances, but there was no analysis in the judgment of why that might be. So, again, this is another instance which provides little guidance for the application of article 5(3) in the situation with which we are concerned.

In Krüsemann there was unofficial “wildcat” strike action by airline staff triggered by the announcement of a corporate restructuring process. This caused cancellations and delays. The airline claimed that this was an extraordinary circumstance. The CJEU disagreed, on the basis that such strikes were inherent in the normal carrying out of the activity of the air carrier. It observed (paras 40-42) that “the restructuring and reorganisation of undertakings are part of the normal management of those entities” in relation to which carriers “may, as a matter of course, when carrying out their activity, face disagreements or conflicts” with staff. This meant that “the risks arising from the social consequences that go with such measures must be regarded as inherent in the normal exercise of the activity of the carrier concerned”. Also, the strike was not regarded as outside the carrier’s control: para 43.

In TAP Portugal the carrier cancelled a flight from Stuttgart to Lisbon less than two hours before the scheduled departure time when the entire crew declared themselves unfit to fly following the sudden death of the co-pilot. The CJEU held that this did not fall within the concept of “extraordinary circumstances” in article 5(3). The management of unexpected absence of staff, due to illness or death of a staff member whose presence was essential to the operation of a flight, was intrinsically linked to the question of crew planning and staff working hours and was therefore something that was inherent in the carrier’s activities; accordingly it was not an extraordinary circumstance. It said:

“20. It is appropriate to begin by determining whether the unexpected absence—due to illness or death of a crew member whose presence is essential to the operation of a flight—which occurred shortly before the flight’s scheduled departure, is capable of constituting, by its nature or origin, an event which is not inherent in the normal exercise of the activity of the operating air carrier.

21.

In that regard, it must be held that measures relating to the staff of the operating air carrier fall within the normal exercise of that carrier’s activities. That is true of measures relating to the working conditions and remuneration of the staff of such a carrier (Airhelp Ltd v Scandinavian Airlines System Denmark – Norway – Sweden [(Case C-28/30) EU:C:2021:226; [2021] Bus LR 674, GC], para 29), which includes measures relating to crew planning and staff working hours.

22.

Therefore, operating air carriers may, as a matter of course, be faced, in the exercise of their activity, with the unexpected absence, due to illness or death, of one or more members of staff whose presence is essential to the operation of a flight, including shortly before the departure of that flight. Accordingly, the management of such an absence remains intrinsically linked to the question of crew planning and staff working hours, with the result that such an unexpected event is inherent in the normal exercise of the operating air carrier’s activity.

23.

It should be pointed out that where, as in the present case, the absence is due to the unexpected death of a member of staff whose presence is essential to the operation of a flight and which occurred shortly before the departure of that flight, such a situation, however tragic and final it may be, is no different, from a legal point of view, from that in which a flight cannot be operated because such a member of staff has unexpectedly fallen ill shortly before the departure of the flight. Thus, it is the very absence, due to illness or death, of one or more crew members, even if it was unexpected, and not the specific medical cause of that absence that constitutes an event inherent in the normal exercise of that carrier’s activity, with the result that the carrier must expect such unforeseen events to arise in the context of planning its crews and the working hours of its staff.

24.

Furthermore, the fact that such an unexpected absence occurred even though the crew member concerned had fully completed the regular medical examinations prescribed by the applicable legislation cannot call into question the conclusion set out in para 22 of the present judgment. Any person, including those who have successfully undergone regular medical examinations, may, at any time, unexpectedly fall ill or die.”

Analysis in light of Wallentin-Hermann and the pattern of the case-law

The fact that TAP Portugal was decided post-Brexit led Mr Shah to point out that it is not a binding decision. He submitted that this court should not find it persuasive pursuant to section 6 of the Withdrawal Act 2018 for the purposes of the interpretation of article 5(3). We do not agree. In our view, the reasoning and the result in TAP Portugal is well in line with the principles set out in Wallentin-Hermann and with the general pattern of cases which had already emerged in the CJEU’s jurisprudence prior to Brexit. Having regard to this degree of “fit” with that previous jurisprudence, we consider that TAP Portugal is persuasive authority which strongly supports the contention of the Liptons that the illness of the captain of their flight is a matter which the airline would be expected to cope with in the ordinary course of managing its business, as a matter inherent in the conduct of that business. Therefore, Cityflyer cannot satisfy the inherency test limb of the approach to interpretation of “extraordinary circumstances” in article 5(3) as laid down in Wallentin-Hermann.

But we should add that we would have reached the same conclusion without the assistance to be derived from TAP Portugal, for the same reasons set out by Coulson LJ in the Court of Appeal.

It is Cityflyer who has the burden of proving that the captain’s non-attendance due to illness fell within the concept of “extraordinary circumstances” in article 5(3). We agree with Coulson LJ that for the six inter-linked reasons he gives Cityflyer fails to show that the situation in this case falls within that concept. As he observed, the non-attendance of the captain due to illness was an inherent part of Cityflyer’s activity and operations as an air carrier and could in no way be categorised as extraordinary. 

Ordinary meaning of the words

The CJEU emphasised in Wallentin-Hermann, para 17, that the expression “extraordinary circumstances”’ in article 5(3) must be given its usual meaning in everyday language. It means something out of the ordinary: see Sturgeon and Huzar. Staff illness, and the need to accommodate such illness on a daily basis, is a commonplace for any business. As Coulson LJ points out, it is a mundane fact of commercial life which is in no way out of the ordinary. To use the terminology deployed in Wallentin-Hermann at para 44(1), the possibility of the captain’s absence was, by its nature and origin, inherent in the normal exercise of the activity of Cityflyer. It was part of its operating system.

That reading of Regulation 261 in accordance with its ordinary meaning is consistent with its purpose, which is to provide a high level of protection for consumers: see para 134, above. Moreover, it interprets article 5(3) strictly, because it derogates from this purpose: see para 140, above.

Consistency with authorities in respect of staff absence

That interpretation is consistent with the authorities in respect of staff absence. Staff absence is not one of the factors identified in recital 14 as indicative of “exceptional circumstances”. The only factor listed there that is potentially referable to staff absence is strike action, which is much more likely to be out of the ordinary than a sick member of staff. In the Court of Appeal Coulson LJ relied on the fact that there was no authority anywhere to support the proposition that staff absence due to illness is an extraordinary circumstance. Now TAP Portugal provides strong support for the conclusion that it is not.

The absence of airport staff (not airline staff) was apparently assumed to be an extraordinary circumstance in Finnair. In Krüsemann, on the other hand, a strike by the air carrier’s staff was said not to be extraordinary, but was instead found to be an inherent part of the carrier’s activity and operations. Since it is the case (as noted in Krüsemann, para 41) that carriers may, “as a matter of course when carrying out their activity” face disagreements or conflicts with all or part of their members of staff culminating in absences due to strikes, it can also be said with force that as a matter of course when carrying out their activity carriers have to take account of the potential absence of some of their staff at any given time due to illness, bereavement or the like. 

Consistency with authorities in respect of technical defects

The interpretation as proposed by the Liptons is also consistent with the authorities concerned with technical defects. As explained above, defects in the aircraft (what was called mechanical “wear and tear” in Huzar) have regularly been held to be an inherent part of an air carrier’s activity and not an extraordinary circumstance. We agree with Coulson LJ that these cases strongly suggest a similar answer to this appeal. A carrier’s operation depends on two principal resources: its people and its aircraft. Wear and tear of the aircraft and its component parts is not extraordinary. The wear and tear on people, manifesting itself in illness, should not be regarded as any different. As Coulson LJ put it, the captain is just as much part of “the operating system” (Pešková) as the mechanical components of the aircraft. The illnesses of staff and the wearing out of parts of the aircraft are ultimately no different when considering inherency, because they both need to be allowed for in the air carrier’s operating system. 

Consistency with authorities in respect of external or one-off events

In our view, the interpretation proposed by the Liptons is also consistent with this category of authorities. The petrol spillage that closed a runway, the foreign object on the runway, and the unruly passenger whose conduct was so bad that the pilot had to divert the flight were all one-off events which were not matters that fell within the carrier’s normal everyday activity. Such examples are very different from the mundane circumstances of the present case which, unlike them, involved the carrier’s own employee. On any view, those authorities concerned rare or infrequent events.

Siewert shows that an event can be external to the carrier but still be inherent to the carrier’s operation. Thus, even if it could be said that the captain’s illness in this case was in some way external (because it happened when he was off duty), that would not take it outside the inherency test. The mobile stairs in Siewert were described as being “indispensable to air passenger transport” and it was said that air carriers were “regularly faced with situations arising from their use”. Similarly, captains and crew are indispensable to air passenger transport and air carriers are regularly faced with situations arising from their non-attendance, for whatever reason. There is again no material difference for these purposes. 

Inherency and the relevance of off duty events

Coulson LJ pointed out that the pilot of an aircraft is critical to the carrier’s activity and operations and correctly observed that this was a strong indication regarding the proper conclusion according to the inherency test. The captain’s attendance for work is an inherent part of the carrier’s operating system. If he failed to attend work due to illness, that non-attendance is “inherent in the normal exercise of the activity of the air carrier concerned” (Krüsemann). 

From that perspective it does not matter whether the captain of this flight happened to fall ill an hour before he clocked on for work, rather than half an hour afterwards. It is unrealistic to say that the captain was only an inherent part of the airline’s operation when he had clocked on for work, and that in the minutes leading up to that point, he was somehow irrelevant to the air carrier’s activity. 

As further support for this realistic conclusion, Coulson LJ referred to rules governing the activities of airline pilots outside their working hours, such as the law that a pilot cannot drink alcohol in the 24 hours before he or she flies. If the pilot drinks so as to be unfit to report for work and the flight is cancelled, then the reason for the cancellation is inherent in the airline’s activity and operations. The same is also true of the need for the captain and other cabin crew to ensure that they are properly rested during stopovers. They have numerous obligations both to their employers and to the public during those periods. These are all inherent in the carrier’s activity and operations and if, for whatever reason, they are unable to attend for work as a result of something going awry during those rest periods (whether it is their fault or not), that failure to attend is not an extraordinary circumstance within article 5(3). 

Avoidance of excessively granular investigations

A further reason for concluding that precisely when, why or how the staff member in question fell ill is irrelevant to the proper operation of article 5(3) arises from the nature of Regulation 261 itself. It is designed to provide a standardised, if modest, level of compensation to those who suffer the inconvenience of cancelled or delayed flights. The exception at article 5(3) has to be considered in that light: see para 138, above. The vast bulk of claims arising under the Regulation should be capable of being determined on the papers. In those circumstances, as Coulson LJ points out, it is contrary to the scheme of the Regulation to allow the carrier to embark on a complex analysis of precisely when, why or how a staff member became ill so as to explain their absence and the subsequent cancellation of the flight.

In any event, there are obvious difficulties in identifying precisely when, why or how someone first falls ill. Is it when they first exhibit the symptoms? Or is it when they are first exposed to the infection? Why are they unwell? How has that happened? If a crew is on a particularly tight schedule, with a meal then a flight, then a rest and then a repeat for the return flight, how can it be safely worked out when, why or how the crew member actually fell ill, and whether that happened, as Cityflyer would have it, on their own time or the carrier’s time? The scheme under the Regulation is not designed to require investigation of these questions.

CONCLUSION

For the reasons set out above, we would dismiss this appeal.

Introduction

I agree with the judgment of Lord Sales and Lady Rose on ground 1. The illness of the captain of the flight did not fall within the defence of “extraordinary circumstances” under article 5(3) of Regulation (EC) No 261/04 (“Regulation 261”). It follows that I agree that the appeal should be dismissed and I say nothing further about ground 1.

I also agree with Lord Sales and Lady Rose on ground 2 (ie which law applies and why?). However, given their disagreement with Lord Lloyd-Jones on ground 2, this judgment sets out in my own words why I consider that Lord Sales and Lady Rose are correct. I agree that what they call the “Complete Code analysis” (but which I shall refer to as the “retained EU law analysis”) is to be preferred to what they refer to as the “Interpretation Act analysis” (but which I shall refer to as the “not retained EU law analysis”). At root, my reasoning is that the former analysis is to be preferred because, after IP completion day (31 December 2020), accrued EU law rights fall within what is referred to as “retained EU law” in the European Union (Withdrawal) Act 2018 as amended by the European Union (Withdrawal Agreement) Act 2020 (which I shall refer to compendiously as the “EU (Withdrawal) Act 2018”). Accrued EU law rights continue to exist, and have not been retrospectively removed, but do so as part of retained EU law. Accrued EU law rights do not exist as a valid body of law separate from retained EU law.

I am most grateful to Lord Sales and Lady Rose for having set out the relevant facts and the relevant legislative provisions and case law, which I do not need to repeat. I also gratefully adopt what Lord Sales and Lady Rose have said, at paras 23-24, about the Retained EU Law (Revocation and Reform) Act 2023, the provisions of which have not been taken into account in this judgment.

Accrued EU law rights form part of retained EU law

It is common ground that accrued EU law rights remain valid and have not been retrospectively removed by any of the legislation implementing the withdrawal of the UK from the EU. However, it is of central importance – and represents the essential reason why I disagree with the judgment of Lord Lloyd-Jones – that the correct explanation for the continued validity of accrued EU law rights is that they fall within the category of retained EU law.

Section 6(7) of the EU (Withdrawal) Act 2018 defines “retained EU law” as meaning, so far as relevant, “anything which, on or after IP completion day, continues to be, or forms part of, domestic law by virtue of section … 3 or 4 … (as that body of law is added to or otherwise modified by or under this Act or by other domestic law from time to time”. I agree with Lord Sales and Lady Rose, essentially for the reasons they give, that the provision by which accrued EU law rights have in this case become retained EU law is section 3 of the EU (Withdrawal) Act 2018.

At first sight, section 4 of the EU (Withdrawal) Act 2018, with its heading of “Saving for rights etc under section 2(1) of the [European Communities Act 1972]” may appear to be the more natural route by which the accrued EU law rights have here become retained EU law. But overall I accept that the correct interpretation is that section 3 incorporates into domestic law “direct EU legislation” (which must be “operative immediately before IP completion day” the meaning of which is explained in section 3(3)) not merely prospectively but also in respect of accrued rights (eg accrued causes of action) that have been triggered by that legislation. Moreover, by section 4(2), if rights form part of domestic law by reason of section 3, they fall outside section 4(1). Section 4 covers rights, including, in my view, accrued rights, conferred by, for example, treaties that are not defined as “direct EU legislation” and are therefore outside section 3. But in so far as there would otherwise be overlap between section 3 and section 4(1), any such overlap is avoided because section 4(2) makes clear that section 3 is the applicable section. In other words, section 3 and section 4(1) are mutually exclusive with section 3 being the dominant provision.

Although the section 3 route to the incorporation of accrued EU rights into retained EU law is here to be preferred to the section 4 route, in practice it may not make any significant difference in this case which of those two routes applies. The central point is that both reach the same conclusion that accrued EU law rights are retained EU law.

The important consequence of accrued EU law rights being retained EU law is that the interpretation of accrued EU law rights is straightforwardly governed by section 6 of the EU (Withdrawal) Act 2018 which is headed “Interpretation of retained EU law” (emphasis added). That in turn has three main practical consequences for accrued EU law rights (as for other retained EU law).

First, an English court is not bound by any decision of the Court of Justice of the European Union (“CJEU”) made after IP completion day (section 6(1)(a)).

Secondly, no references to the CJEU can be made by a domestic court or tribunal after IP completion day (section 6(1)(b)).

Thirdly, accrued EU law rights are to be interpreted in accordance with EU law as it was prior to IP completion day (section 6(3)), subject to the power of the UK Supreme Court (and, because of subsequent domestic regulations, some other appellate courts: see para 55 above) to depart from any retained EU case law applying the equivalent of the 1966 Practice Statement, [1966] 1 WLR 1234 (section 6(4)). It should be noted that, in contrast to section 6(1)(a) and (b), the actual text of section 6(3) and (4) – and not just the heading – refers to retained EU law. It is therefore indisputable that section 6(3) and 6(4) would not apply to accrued EU law rights unless they are retained EU law.

This analysis of accrued EU law rights as being part of retained EU law is in line with the purpose of the EU (Withdrawal) Act 2018 which was to retain as domestic law, after IP completion day, EU law as it was before IP completion day, subject to specified exceptions. Put another way, and subject to specified exceptions, a snapshot of EU law as it applied in the UK was taken as at IP completion day and that state of the law was to be carried forward as valid retained EU law (ie as domestic law) after IP completion day. That that was the purpose of the EU (Withdrawal) Act 2018 is clear from the provisions of the legislation and from the explanatory notes. This analysis is also in line with what Lord Sales and Lady Rose have explained was the apparent purpose of the EU (Withdrawal) Act 2018 as being to provide a comprehensive code for UK law, in respect of EU law, after IP completion day.

Applying this analysis of the law to the facts of this case leads to the following conclusions:

Reasons why it is incorrect to treat accrued EU law rights as a valid body of law separate from retained EU law.

It follows from what has been said above that the correct interpretation of the EU (Withdrawal) Act 2018 is that accrued EU law rights are retained EU law. However, some commentators (in particular, Jack Williams in a series of carefully and tightly reasoned blogs on these issues: see, especially, “Accrued EU law rights: a guide for the perplexed” posted 5/1/2022) have argued that accrued EU law rights remain valid (and have not been swept away) but are not covered by the EU (Withdrawal Act) 2018 and are therefore not retained EU law. That is because, so it is argued, there was no need for them to be covered: without any express mention, they simply fall within section 16(1) of the Interpretation Act 1978 which lays down that, without clear provision to the contrary, legislation (here the EU (Withdrawal Act) 2018) does not take away accrued rights. On this “not retained EU law analysis”, accrued EU law rights are therefore not retained EU law. By reason of section 16(1) of the Interpretation Act 1978, they validly exist independently of the EU (Withdrawal Act) 2018. This “not retained EU law analysis” is also the one favoured by Lord Lloyd-Jones in his judgment on ground 2.

With respect, that analysis is, in my view, unconvincing. That is not only because, for the reasons already given, there is a better interpretation of the status of accrued EU law rights but also because there are a number of specific objections to, or problems with, the “not retained EU law analysis”.

First, the “not retained EU law analysis” would undermine what Lord Sales and Lady Rose have explained was the apparent purpose of the EU (Withdrawal) Act 2018 of providing a complete legislative code covering the relationship between EU and UK law after the withdrawal of the UK from the EU. For the entire suite of relevant legislation concerned with Brexit not to mention anywhere an important area of EU law (ie accrued EU law rights) that continues to be valid and binding as pure EU law would be most odd when an aim of that legislation was clarity and certainty and, on the face of it, comprehensivity.

Secondly, the “not retained EU law analysis” would appear to give alarming or, at best, unclear or illogical answers to several fundamental practical questions about the status of accrued EU law rights. In particular, and following from paras 182-184 above, one can articulate four such questions.

In summary, the “not retained EU law analysis” gives alarming or, at best, uncertain or illogical answers to those four questions. In contrast, if one adopts the “retained EU law analysis” set out above, the answers to those questions are clear and non-problematic. They are to be answered, respectively, “no”, “no”, “yes” and “no”.

Conclusion

For all those reasons, I therefore reject the “not retained EU law analysis” and accept the “retained EU law analysis”. Indeed, even if both analyses were possible (which I do not accept), there appears to be nothing to be gained and, on the contrary, significant practical disadvantages if one denies that retained EU law can be, and should be, interpreted as embracing accrued EU law rights.

I agree with the majority on the substantive issue of whether the appellant can raise a defence of extraordinary circumstances, for the reasons given by Lord Sales and Lady Rose in their joint judgment. I also agree with the majority that the Court of Appeal was in error in concluding that the claim in the present proceedings is governed by Regulation 261/04 as amended by the Air Passenger Rights and Air Travel Organisers’ Licencing (Amendment) (EU Exit) Regulations 2019. However, I take a different view from the majority on the issue of which law does apply. The majority consider the applicable law is retained EU law and on this basis they conclude that it is the unamended version of the retained EU regulation, Regulation 261/04, which applies. I consider that the applicable law is EU law, as this was the law in force when the claim accrued in 2018, and that on this basis Regulation 261/04 itself applies. The essential point of difference between us is whether the legislative provisions on retained EU law have only prospective effect. I agree, however, with the majority that this difference between us is not decisive in these proceedings and the various views expressed are therefore obiter.

I consider that when their flight from Milan Linate Airport to London City Airport on 30 January 2018 was cancelled Mr and Mrs Lipton (the respondents) became entitled to compensation under EC Regulation 261/2004 in the version then in force, subject to a possible defence under article 5(3). That entitlement arose in EU law which was given effect in domestic law within the United Kingdom by the European Communities Act 1972.

The majority consider that on IP completion day (ie 11 pm on 31 December 2020) those rights of action were transposed by the provisions of the European Union (Withdrawal) Act 2018 (“the 2018 Act”) into retained EU law, a new body of domestic law, and were thereafter enforceable in the United Kingdom as retained EU law. I respectfully disagree. I consider that they remain accrued rights in EU law enforceable as such in the United Kingdom because the 2018 Act left such accrued rights undisturbed. The repeal of the European Communities Act 1972 by section 1 of the 2018 Act did not affect accrued rights in EU law, save where the 2018 Act expressly provided to the contrary.

My starting point is the general principle of the common law that conduct and events are normally governed by the law in force at the time at which they took place. As a result, subsequent legislative changes in the law are not generally given retrospective effect. Evidence of a clear contrary intention would be required before they could be given retrospective effect, for example by disturbing accrued rights. There is a general presumption at common law that legislation is not retrospective in the sense that it alters the legal consequences of things that happened before it came into force (Chitty on Contracts, 35^th ed (2023), para 1-031A; Bennion, Bailey and Norbury on Statutory Interpretation, 8^th ed (2020), paras 7.13, 7.14). This general rule reflects public expectations and notions of fairness and legal certainty. It is apparent, for example, in the judgment of the majority of the Supreme Court in R (Miller ) v Secretary of State for Exiting the European Union [2017] UKSC 5; [2018] AC 61 (“Miller 1”) where it was clearly contemplated that the disengagement of the United Kingdom from EU law would, in general, leave accrued rights undisturbed. At para 70 the majority referred to the rights in EU law enjoyed in the United Kingdom that were “capable of replication in UK law”. It noted that some of these rights had already been embodied in UK law by domestic legislation pursuant to section 2(2) of the 1972 Act (for example UK regulations passed under that subsection to implement Directives). It then stated that:

“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.”

In this passage the majority acknowledged that claims which have already accrued under directly effective EU law would be preserved following the withdrawal of the United Kingdom from the EU.

The passage also helpfully distinguishes between a body of rules derived from EU law which is to be given prospective effect in domestic law following withdrawal and accrued rights and liabilities which have already arisen under EU law. The transposition of the first category was subsequently achieved by the provisions of the 2018 Act establishing a body of retained EU law. In this way it was intended to secure continuity and legal certainty and to avoid a vacuum in the UK statute book. The 2018 Act lays down with prospective effect what rules derived from EU law will apply in the United Kingdom following withdrawal from the EU. It addresses the legal situation as from IP completion day. The question for consideration here is whether it was also intended to make provision retrospectively in respect of rights which had already accrued under EU law before withdrawal. I differ from my colleagues in the majority in considering that it was not intended to have this effect. My reasons for coming to this conclusion are set out in the following paragraphs and may be summarised as follows:

Before turning to these matters in detail, I also note that my view is consistent with the approach followed in a number of decided cases, although I accept that the issue may not have been fully canvassed there (R (Fratila) v Secretary of State for Work and Pensions [2021] UKSC 53; [2022] PTSR 448, para 1; News Corp UK and Ireland Ltd v HMRC [2023] UKSC 7; [2024] AC 89, para 7; R (Open Rights Group Ltd) v Secretary of State for the Home Department [2021] EWCA Civ 800; [2021] 1 WLR 3611, paras 12–13; Revenue and Customs Comrs v Perfect [2022] EWCA Civ 330; [2022] 1 WLR 3180; Anan Kasei Co Limited, Rhodia Operations SAS v Neo Chemicals & Oxides (Europe) Ltd [2023] EWCA Civ 11; [2023] FSR 14, para 85.) (The decision of the Competition Appeal Tribunal in Umbrella Interchange Fee Claimants v Umbrella Interchange Fee Defendants [2023] CAT 49 is to the opposite effect.) Similarly, this view has support in academic writing (Chitty on Contracts, 35^th ed, paras 1-031–1-034; Whittaker (2021) 137 LQR 477 at 486-8; Halford and Whitehouse (2021) Tax Journal, Issue 1533 at 8-11; Williams, “Accrued EU law rights: a guide for the perplexed” in EU Relations Law (blog) 5 January 2022 at https://eurelationslaw.com/blog/accrued-eu-law-rights-a-guide-for-the-perplexed).

First, there was no need to make provision in the 2018 Act for the preservation of accrued rights and obligations arising in EU law because there were already in place statutory provisions of general effect which achieve precisely that result. In this jurisdiction the general principle of non-retroactivity of legislation is given effect by section 16(1) of the Interpretation Act 1978 which provides:

“Without prejudice to section 15, where an Act repeals an enactment, the repeal does not, unless the contrary intention appears –

(a)

revive anything not in force or existing at the time at which the repeal takes effect;

(b)

affect the previous operation of the enactment repealed or anything duly done or suffered under that enactment;

(c)

affect any right, privilege, obligation or liability acquired, accrued or incurred under that enactment;

(d)

affect any penalty, forfeiture or punishment incurred in respect of any offence committed against that enactment;

(e)

affect any investigation, legal proceeding or remedy in respect of any such right, privilege, obligation, liability, penalty, forfeiture or punishment;

and any such investigation, legal proceeding or remedy may be instituted, continued or enforced, and any such penalty, forfeiture or punishment may be imposed, as if the repealing Act had not been passed.”

This provision creates a presumption that, in the absence of a clear contrary intention, legislation will not be read as disturbing established rights and obligations. Silence in subsequent legislation is consistent with the application of the presumption under section 16 (Aitken v South Hams District Council [1995] 1 AC 262 per Lord Woolf at p 272; R (York City Council) v Secretary of State for Housing, Communities and Local Government [2018] EWHC 2699 (Admin); [2019] PTSR 471 per Kerr J at para 40).

In this case, Mr and Mrs Lipton sought to rely on rights which had accrued under a directly effective EU regulation. While the United Kingdom was a Member State, such rights were given direct effect in the United Kingdom by section 2(1) ECA 1972. I reject the submission on behalf of the respondents that section 16(1) has no application to the repeal of section 2(1) of the 1972 Act. First, as I understand it, it is submitted that section 16(1)(c) has no application because rights were not acquired “under” the 1972 Act which simply acted as a conduit pipe for the transposition of directly effective rights in EU law into domestic law. This submission is founded on an over-technical reading of the word “under” in section 16(1)(c). The reality was that rights were enjoyed in domestic law by virtue of the 1972 Act. Secondly, it is submitted that section 16(1)(c) is not engaged here because, while the 2018 Act formally repeals the 1972 Act, it does not repeal or in any way cancel the operative rights thereunder, but instead provides a means for their implementation in domestic law, which was previously not required because of the conduit pipe provided by the 1972 Act. This submission assumes what it seeks to establish, namely that accrued rights are preserved by the 2018 Act. It also proceeds on a false premise. The presumption under section 16(1)(c) is not only engaged when an enactment “repeals or in any way cancels” rights accrued under the repealed enactment. It is engaged where the repeal may, absent the presumption, “affect” any right accrued under the repealed enactment. Rights accrued under directly effective EU legislation would, on the respondents’ case, be affected by the repeal of section 2 of the 1972 Act because the law applicable to those rights would have changed, in this case from EC Regulation 261/2004 to the retained version of the regulation. Even more fundamentally, the source and constitutional character of the law applicable to those rights would have changed from EU law to domestic law. Such a change of applicable law in itself would be sufficient to engage the presumption under section 16(1)(c), even if the substantive content of the new law were identical to the law which applied when the right in question accrued. (There is here an analogy with the repeal of a previous Act of Parliament and its re-enactment by a consolidating Act. Although there may be no change to the substantive rules applicable, section 16(1)(c) would be engaged because the change of applicable law would otherwise affect the rights accrued under the repealed statute. In such a case the general rule would remain that any claims should be pleaded under the statute which applied when the relevant cause of action arose.)

In any event, section 16(1)(b) of the 1978 Act creates a presumption against affecting “the previous operation of the enactment repealed or anything done or suffered under that enactment”, which must be engaged here even if the limb under section 16(1)(c) were not. Two consequences follow. First, there is therefore a presumption against the 2018 Act having retrospective effect or affecting accrued rights and obligations. Secondly, if section 16(1) of 1978 Act applies, there would be no need for the 2018 Act to preserve accrued rights and obligations as they would be preserved by the automatic operation of section 16(1), although it would, of course, be open to Parliament, if it chose to do so, to preserve accrued rights by alternative means. There can, however, be no doubt that section 16, if applicable, would be effective to preserve accrued rights.

Secondly, the text of the 2018 Act and the Explanatory Notes to the Bill provide no support for the view that it was intended to preserve rights and obligations derived from EU law which had already accrued as EU rights in domestic law. There is no express indication in the 2018 Act that it is intended to make general provision for the survival of accrued rights and obligations by converting them into retained EU law. If that was the legislative intention it is astonishing that it does not say so expressly.

Similarly, contrary to the assertion of Lord Sales and Lady Rose at para 91, there is nothing in the Explanatory Notes which indicates that the provisions of the 2018 Act are intended to operate as a complete code in the sense for which they contend. Although it is possible to read certain passages in the Explanatory Notes as wide enough to include accrued rights, those passages are at least equally consistent with a reading limited to giving the new body of retained EU law prospective effect. The Explanatory Notes to the Bill explain the intended approach of the 2018 Act in the following terms (at para 10):

“The principal purpose of the Bill is to provide a functioning statute book on the day the UK leaves the EU. As a general rule, the same rules and laws will apply on the day after exit as on the day before. It will then be for Parliament and, where appropriate, the devolved legislatures to make any future changes.”

This is clearly focussed on the prospective application of EU law within the United Kingdom in the form of retained EU law. The Notes then go on to identify the four main functions which the 2018 Act was intended to perform.

“The Bill performs four main functions. It:

•

repeals [the 1972 Act];

•

converts EU law as it stands at the moment of exit into domestic law before the UK leaves the EU;

•

creates powers to make secondary legislation, including temporary powers to enable corrections to be made to the laws that would otherwise no longer operate appropriately once the UK has left the EU and to implement a withdrawal agreement; and

•

maintains the current scope of devolved decision making powers in areas currently governed by EU law.”

It is highly significant for present purposes that the Explanatory Notes do not suggest that the purpose of the 2018 Act is to keep causes of action accrued under EU law alive and actionable post-withdrawal. If the Act was thought to be needed to keep these causes of action alive, this would plainly have been a main function of the Act and would have been identified as such in the Explanatory Notes.

Furthermore, it is highly relevant that where the 2018 Act intends that certain EU law should not apply to matters arising before IP completion day, ie where the 2018 Act is intended to operate with retroactive effect in relation to causes of action which accrued before IP completion day, it makes express provision to that effect. For example, it provides that the EU Charter of Fundamental Rights, general principles of EU law (unless recognised by the Court of Justice of the European Union in a case decided before IP completion day) and the right to Francovich damages do not apply to events occurring before IP completion day (section 5(4), paragraph 1(2)-(4) of Schedule 1 and paragraph 39(1) of Schedule 8). In these specific instances the 2018 Act does operate to exclude the operation of section 16 of the Interpretation Act 1978 and to have retroactive effect. It achieves this by expressly stating that the relevant provisions of the 2018 Act “apply in relation to anything occurring before IP completion day (as well as anything occurring on or after IP completion day)” (paragraph 39(1) of Schedule 8). The existence of these exceptions provides support for the view that the 2018 Act was not intended to operate retroactively, so as to affect causes of action which accrued before IP completion day, save where it made express provision to that effect. (See Chitty, para 1-031A.)

The 2018 Act must be read subject to the presumption against retroactive effect established by section 16(1) Interpretation Act 1978. Subject to the specific exceptions such as those referred to above, the 2018 Act contains no indication that it is intended to affect accrued rights and obligations, let alone one sufficient to rebut the presumption.

Thirdly, the language of sections 2, 3 and 4 of the 2018 Act is not apt to achieve the transposition of accrued rights and obligations into retained EU law, by the different routes for which the members of the majority variously contend. Retained EU law is defined for the purposes of the 2018 Act by section 6(7) as follows:

“‘retained EU law’ means anything which, on or after IP completion day, continues to be, or forms part of, domestic law by virtue of section 2, 3 or 4 or subsection (3) or (6) above (as that body of law is added to or otherwise modified by or under this Act or by other domestic law from time to time)”

If it was intended that the 2018 Act should transpose accrued EU causes of action into domestic law as retained EU law, this would have to be inferred from sections 2, 3 or 4 of the Act. To my mind, however, the more natural reading of these sections is that they were intended to put on the UK statute book a snapshot of EU law as it stood on IP completion day with prospective effect and with no effect on pre-existing causes of action.

Section 2(1) provides:

“EU-derived domestic legislation, as it has effect in domestic law immediately before IP completion day, continues to have effect in domestic law on and after IP completion day.”

“EU-derived domestic legislation” is defined by section 1B(7):

“In this Act "EU-derived domestic legislation" means any enactment so far as—

(a)

made under section 2(2) of, or paragraph 1A of Schedule 2 to, the European Communities Act 1972,

(b)

passed or made, or operating, for a purpose mentioned in section 2(2)(a) or (b) of that Act,

(c)

relating to—

(i)

anything which falls within paragraph (a) or (b), or

(ii)

any rights, powers, liabilities, obligations, restrictions, remedies or procedures which are recognised and available in domestic law by virtue of section 2(1) of the European Communities Act 1972, or

(d)

relating otherwise to the EU or the EEA,

but does not include any enactment contained in the European Communities Act 1972 or any enactment contained in this Act or the European Union (Withdrawal Agreement) Act 2020 or in regulations made under this Act or the Act of 2020.”

The language of section 2(1) and section 1B(7) is not apt to include causes of action accrued under EU-derived domestic legislation as it stood prior to IP completion day. Take, for example, a claim for a hypothetical breach of the Working Time Regulations 1998 which occurred in 2019. The cause of action which would have accrued at the date of breach is neither “legislation” within section 2(1) nor an “enactment” within section 1B(7). Similarly, there is nothing in the Explanatory Notes in relation to section 2 which supports the view that rights which accrued before IP completion day under EU-derived domestic legislation are to be treated as retained EU law.

Section 3(1), on which the majority rely to transpose accrued EU causes of action into domestic law as retained EU law, provides:

“Direct EU legislation, so far as operative immediately before IP completion day, forms part of domestic law on and after IP completion day.”

“Direct EU legislation” is then defined in section 3(2) as including, most notably, EU regulations and EU decisions. As in the case of section 2, the language of section 3 is inapt to describe the translation into retained EU law of causes of action arising from direct EU legislation in addition to the legislation itself. Causes of action which have arisen by virtue of legislation are not the same as legislation. This reading of section 3 is reinforced by section 3(4) which provides that section 3 brings into domestic law any direct EU legislation only in the form of the English language version of that legislation and does not apply to any such legislation for which there is no such version. It is not possible to have an English language version of a cause of action. Similarly, discussing clause 3 of the EU Withdrawal Bill, the Explanatory Notes state (at para 80):

“Where legislation is converted under this clause, it is the text of the legislation itself which will form part of domestic legislation. This will include the full text of any EU instrument (including its recitals).”

This all points to a reading of section 3 which limits its function to the transposition of direct EU legislation onto the UK statute book. In this regard I note that the majority in Umbrella Interchange Fee Claimants v Umbrella Interchange Fee Defendants [2023] CAT 49; [2023] 5 CMLR 24 (Marcus Smith J and Mr Tidswell) considered (at para 44(2)(iii)(c)-(d)) that, first, the transposition effected by section 3(1) is limited to direct EU legislation and does not extend, for example, to a cause of action that might have accrued out of the rights conferred by direct EU legislation and, secondly, that section 3(1) is expressly prospective in effect.

Section 4 might, at first sight, be considered a better prospect for those contending for such a transposition of accrued rights into retained EU law. (It should be noted that while section 4 has now been repealed by the Retained EU Law (Revocation and Reform) Act 2023, this repeal did not apply in relation to anything occurring before the end of 2023. See sections 2 and 22(5).) However, once again the section read in context is inapt to achieve that result.

For the purposes of this appeal, section 4(1) provides:

“Any rights, powers, liabilities, obligations, restrictions, remedies and procedures which, immediately before IP completion day–

(a)

are recognised and available in domestic law by virtue of section 2(1) of the European Communities Act 1972, and

(b)

are enforced, allowed and followed accordingly,

continue on and after IP completion day to be recognised and available in domestic law (and to be enforced, allowed and followed accordingly).”

Section 4(1) is then qualified by sections 4(2) and 4(3). In particular, section 4(2)(a) provides:

“Subsection (1) does not apply to any rights, powers, liabilities, obligations, restrictions, remedies or procedures so far as they–

(a)

form part of domestic law by virtue of section 3 …”

I note that the majority do not rely on section 4 as transposing the cause of action relied upon by Mr and Mrs Lipton into retained EU law. In my view, they are correct in not doing so, for the following reasons.

The wording of section 4(1) is taken from section 2(1) of the 1972 Act. It is clear that the word “rights” in section 2(1) of the 1972 Act refers to rights held by individuals by virtue of directly effective EU legislation. Such rights could arise, for example, from treaty provisions, regulations, decisions or, in certain circumstances, directives. On any view, the primary purpose of section 4(1) must be to transpose rights in that sense into retained EU law. Repeating the language of section 2(1) of the 1972 Act would be the obvious way of achieving this purpose. This alone cannot be taken to evince an intention to use this section to transpose, in addition, causes of action.

There is an important distinction between a right conferred prospectively under EU law and a cause of action arising pursuant to that right. It is the difference between a right to receive compensation if one’s flight is cancelled in certain circumstances and the right of Mr and Mrs Lipton to receive compensation for the cancellation of Flight BA 7304 from Milan Linate Airport to London City Airport on 30 January 2018. The former is a rule. The latter is a cause of action arising pursuant to the rule.

The Explanatory Notes to the EU Withdrawal Bill state that clause 4 (which became section 4) ensures that remaining EU rights and obligations which do not fall within clauses 2 and 3 (which became sections 2 and 3) continue to be recognised and available in domestic law after exit. These will be principally Treaty rights and some directly effective rights arising under directives if they have been recognised in cases decided before IP completion day (see section 4(2)(b)). The Explanatory Notes make no reference to the transposition of causes of action. On the contrary, they are entirely consistent with section 4 merely transposing rights under directly effective EU legislation. They address how such rights will be brought into domestic law. Thus in relation to directly effective treaty rights, they state (at para 88):

“Where directly effective rights are converted under this clause, it is the right which is converted, not the text of the article itself.”

The Notes then set out (at para 89) a non-exhaustive list of articles of the Treaty on the Functioning of the European Union that the Government considered “contain directly effective rights which would be converted into domestic law as a result of this clause”. The Notes explain (at para 91) that any directly effective rights converted into domestic law as a result of section 4 would be subject to amendment or repeal via statutory instrument made under clause 7, which became section 8 of the 2018 Act. The Notes reinforce the view that all section 4 was intended to do was to transpose with prospective effect the directly effective rights to which it relates and not causes of action which may have accrued under them.

The most natural reading of section 4, in light of the structure of the 2018 Act, is that it is intended to perform the same function as sections 2 and 3, namely to transpose different categories of EU law onto the UK statute book prospectively. If section 4 were to be read as transposing causes of action, it would be doing something fundamentally different from sections 2 and 3 which, as we have seen, are expressly limited to transposing “legislation”. That these three sections are intended to perform the same function is also demonstrated by section 4(2)(a) which provides that section 4(1) does not apply to any rights, powers, liabilities, obligations, restrictions, remedies or procedures so far as they form part of domestic law by virtue of section 3. This provision is necessary because directly effective rights arising under EU regulations would otherwise fall within section 4(1) as they previously had effect in the United Kingdom by virtue of section 2(1) of the 1972 Act. There is a need for mutually exclusive categories because the transposition of the different rules into retained EU law is achieved in different ways here. In the case of EU regulations the legislation is transposed “in the form of the English language version of that legislation” (section 3(4)(a)); in the case of directly effective Treaty rights under section 4 it is the right itself which is transposed and not the text. Once again this points to the conclusion that section 4 is concerned with the prospective transposition of rules into retained EU law.

Parliament cannot have intended that section 4(1) should operate as a mechanism to transpose all accrued causes of action arising in EU law into retained EU law. Section 4(1) can only be read as applying to directly effective rights. It could not apply to a cause of action accrued under UK legislation which transposed an EU directive because that would not be a right “recognised and available in domestic law by virtue of section 2(1)” of the 1972 Act.

Sections 2, 3 and 4 of the 2018 Act cannot, for these reasons, be interpreted as demonstrating a legislative intention to transpose accrued causes of action founded on EU-derived rights into retained EU law. On the contrary, they were intended to transpose a snapshot of EU law as it stood on IP completion day onto the UK statute book on a prospective basis in order to avoid a legislative vacuum. Causes of action which accrued prior to IP completion day were preserved not by the 2018 Act but by default under section 16(1) of the Interpretation Act 1978. This follows from the status of those causes of action as rights in domestic law by virtue of the 1972 Act (see Miller 1 at paras 70 and 83). Subject to the express exceptions established by the 2018 Act (referred to at para 204 above), courts deciding cases involving rights which accrued before IP completion day under EU law should apply EU law as it stood at the time the cause of action accrued.

Fourthly, a substantial obstacle in the path of the reading favoured by the majority is that it does not necessarily result in the application to the claim of the law in force at the date the cause of action accrued.

In the present appeal we are concerned with causes of action which are alleged to have arisen under Regulation 261/2004 on 30 January 2018. On first principles, such causes of action should be governed by the version of that Regulation which was in force on that date. On the accrued rights approach, which I consider correct, that result follows naturally. The matter is governed by EU law as it was in force at that date and this is not affected retroactively by the 2018 Act. However, on the retained EU law analysis the position is much more complicated. The majority consider that it is retained EU law which applies in such a case, ie the retained EU law version of the Regulation 261/2004 as saved by section 3(1) of the 2018 Act. Section 3(1) provides that what forms part of domestic law on and after IP completion day is “[d]irect EU legislation, so far as operative immediately before IP completion day”. However, the version of an EU Regulation in force immediately before IP completion day may not be the version which was operative when the relevant cause of action arose.

A hypothetical example may make this clearer. Assume that after Mr and Mrs Lipton’s flight was cancelled on 30 January 2018, amendments were made by the EU to Regulation 261/06 in June 2018 so as to double the amount of compensation for cancellation payable under article 7(1) and that the amendments did not have retroactive effect. Those experiencing a similar cancellation on 30 July 2018 would each have been entitled to compensation of €500 instead of €250. Clearly Mr and Mrs Lipton’s award for the cancellation on 30 January 2018 should remain €250 each. However, the version of the EU Regulation which would be transposed into retained EU law with prospective effect would be the amended version because that was the version which was operative immediately before IP completion day.

The majority seek to circumvent this difficulty, as I understand it, by giving a particularly wide meaning to “operative” in section 3(1) so as to permit the transposition into retained EU law of not only the version of the EU Regulation in force immediately before IP completion day but also a series of shadow versions corresponding to the changes made during the legislative history of the EU Regulation. Not only is this a contorted reading of section 3(1), but it is also inconsistent with the definition of “operative”. Section 3(3)(c) provides that, subject to two special cases in section 3(3)(a) and (b) which need not detain us, for the purposes of the 2018 Act any direct EU legislation is operative immediately before IP completion day if, “…it is in force immediately before IP completion day”. The version of an EU Regulation in force at the date a cause of action arose but which was subsequently amended prior to IP completion day cannot possibly be described as “in force” immediately prior to IP completion day. Similarly, the version of an EU Regulation in force at the date a cause of action arose but which was subsequently repealed prior to IP completion day cannot be described as “in force” immediately prior to IP completion day.

Such a reading is also inconsistent with the statutory scheme of the 2018 Act. Section 3(4) provides that section 3 brings into domestic law any direct EU legislation only in the form of the English language version of that legislation. If section 3 really were to have the effect of transposing into retained EU law shadow versions of direct EU legislation, in the various amended versions in force prior to IP completion day, it would introduce contradictory provisions and the statute book would be unintelligible.

Article 4 of the Withdrawal Agreement provides:

Article 4

Methods and principles relating to the effect, the implementation and the application of this Agreement

1.

The provisions of this Agreement and the provisions of Union law made applicable by this Agreement shall produce in respect of and in the United Kingdom the same legal effects as those which they produce within the Union and its Member States.

Accordingly, legal or natural persons shall in particular be able to rely directly on the provisions contained or referred to in this Agreement which meet the conditions for direct effect under Union law.

2.

The United Kingdom shall ensure compliance with paragraph 1, including as regards the required powers of its judicial and administrative authorities to disapply inconsistent or incompatible domestic provisions, through domestic primary legislation.

3.

The provisions of this Agreement referring to Union law or to concepts or provisions thereof shall be interpreted and applied in accordance with the methods and general principles of Union law.

4.

The provisions of this Agreement referring to Union law or to concepts or provisions thereof shall in their implementation and application be interpreted in conformity with the relevant case law of the Court of Justice of the European Union handed down before the end of the transition period.

5.

In the interpretation and application of this Agreement, the United Kingdom's judicial and administrative authorities shall have due regard to relevant case law of the Court of Justice of the European Union handed down after the end of the transition period.

Article 127 of the Withdrawal Agreement provides in material part:

Article 127

Scope of the transition

1.

Unless otherwise provided in this Agreement, Union law shall be applicable to and in the United Kingdom during the transition period…

…

3.

During the transition period, the Union law applicable pursuant to paragraph 1 shall produce in respect of and in the United Kingdom the same legal effects as those which it produces within the Union and its Member States, and shall be interpreted and applied in accordance with the same methods and general principles as those applicable within the Union. …

The provisions of the Withdrawal Agreement and the provisions of EU law made applicable by that Agreement which meet the conditions for direct effect are directly effective in domestic law by virtue of section 7A of the 2018 Act. As a result, domestic enactments including an enactment contained in the 2018 Act must be read and given effect in a manner that is compatible with those provisions of the Withdrawal Agreement (section 7A(3)).

The Withdrawal Agreement required that EU law be applicable “to and in” the United Kingdom during the transition period, subject to certain irrelevant exceptions (article 127(1)). The provisions of EU law which applied in the United Kingdom by virtue of the Withdrawal Agreement were required to produce the same legal effects as they produced within the EU (article 4(1); article 127(3)).

In AT v Secretary of State for Work and Pensions [2023] EWCA Civ 1307; [2024] 2 WLR 967 Green LJ observed (at para 85) that “article 4(1) imposes a duty of reciprocal and identical effect which is intended to ensure that UK citizens and EU citizens working and residing in the EU and UK respectively are treated in the same way.” He went on to adopt the submission of counsel, Ms Marie Demetriou KC, that “[a]rticle 4(1) is on its face a mandatory, outcome driven rule: it requires that the provisions of the [Withdrawal Agreement] produce the same outcome in the UK and the EU Member States.”

Neither party relied in argument on the provisions of the Withdrawal Agreement and accordingly it would not be appropriate to express any concluded view on the relevance of articles 4 and 127 to the matter in hand. Nevertheless, it is difficult to reconcile articles 4 and 127 of the Withdrawal Agreement with the retained EU law analysis favoured by the majority on this appeal. Article 127(1) of the Withdrawal Agreement expressly provides that EU law shall apply in the United Kingdom during the transition period. However, on the retained EU law analysis, the effect of the 2018 Act is to alter retrospectively the legal regime which applied during the transition period (and, indeed, before that period) with the result that it is retained EU law which applies to causes of action accruing during the transition period. However, retained EU law is not EU law but domestic law. It would appear that it was not open to the United Kingdom, consistently with article 127(1) of the Withdrawal Agreement, to legislate retrospectively to provide that EU law did not apply during the transition period. Furthermore, not only is there a presumption of statutory interpretation that legislation intended to give effect to a treaty obligation should be read consistently with that obligation (Kabab-Ji SAL v Kout Food Group [2021] UKSC 48; [2021] Bus LR 1717 per Lord Hamblen and Lord Leggatt at para 31; Argentum Exploration Ltd v Republic of South Africa [2024] UKSC 16; [2024] 2 WLR 1259, per Lord Lloyd-Jones and Lord Hamblen at para 90; Bennion, Bailey and Norbury on Statutory Interpretation at para 24.16) but section 7A(3) of the 2018 Act requires that all domestic enactments must be read so as to conform with the directly effective provisions of the Withdrawal Agreement. If this interpretation of the Withdrawal Agreement is correct, the retained EU law analysis must be rejected. By contrast, on the accrued rights analysis effect is given to causes of action which accrued during the transition period under EU law, in conformity with the apparent intention of the Withdrawal Agreement and section 7A of the 2018 Act. While Mr and Mrs Lipton’s rights accrued in 2018, prior to the transition period, there is no reason why a different approach should be followed for claims which accrued while the UK was a Member State.

Finally, it is necessary to address the suggestion that the accrued rights analysis, which I consider to be correct, creates difficulties for the application of section 6 of the 2018 Act and must, therefore, be rejected. This suggestion, which appears to be a driving force behind the interpretation of the majority, concerns two matters: (1) the status of any principles laid down or any decisions made on or after IP completion day by the CJEU (sections 6(1) and (2)); and (2) the power of certain domestic courts in the United Kingdom to depart from any retained EU case law (sections 6(3) and (4)). These matters were not fully argued before us at the hearing and, as a result, it would be inappropriate to express concluded views in this judgment. Nevertheless, I am not persuaded of any difficulties which would justify the majority’s interpretation or invalidate the accrued rights analysis.

Decisions of the CJEU made on or after IP completion day

Section 6 of the 2018 Act appears under the heading “Interpretation of retained EU law”. Sections 6(1) and (2) provide:

“(1)

A court or tribunal—

(a)

is not bound by any principles laid down, or any decisions made, on or after IP completion day by the European Court, and

(b)

cannot refer any matter to the European Court on or after IP completion day.

(2)

Subject to this and subsections (3) to (6), a court or tribunal may have regard to anything done on or after IP completion day by the European Court, another EU entity or the EU so far as it is relevant to any matter before the court or tribunal.”

The accrued rights analysis has been criticised on the ground that courts would be unable to rely on section 6(1) and (2) when deciding cases involving causes of action which accrued prior to IP completion day unless such accrued rights had effect as retained EU law and that, as a result, courts would be bound by post-IP completion day decisions of the CJEU. However, I consider the concern unfounded as the operation of sections 6(1) and (2) is not limited to retained EU law. As a result, on either of the competing views, UK courts are not bound by any principles laid down or any decisions made on or after IP completion day by the CJEU.

The natural reading of sections 6(1) and (2) is that they are of general application in the sense that they are instructions to courts in the United Kingdom dealing with any case. In this regard they should be contrasted with sections 6(3), (4) and (5), considered below, which expressly refer to “retained EU law” and “retained EU case law” (or now for matters arising after the end of 2023, “assimilated law” and “assimilated EU case law”). The suggestion that sections 6(1) and (2) should be read subject to the implication of words such as “when deciding any question as to the validity, meaning or effect of any retained EU law” is founded on the heading of section 6: “Interpretation of retained EU law”. Such headings can be an aid to interpretation where the language of a statute is ambiguous (R v Montila [2004] UKHL 50; [2004] 1 WLR 3141). However, as external aids to interpretation they cannot displace the meaning of words in a statute which are clear and unambiguous in their wider context and which do not produce absurdity (R(O) v Secretary of State for the Home Department [2022] UKSC 3; [2023] AC 255). The editors of Craies on Legislation 12^th ed (2020), para 26.1.9, warn against treating headings as if they were part of the text of the Act. They point out that the authorities of both Houses do not regard them as part of the Bill before the House, and therefore permit their informal amendment more or less at the will of the drafter. They conclude that:

“It would therefore be quite wrong to regard a heading as part of the text that expresses the settled will and intention of Parliament.”

Similarly, the editors of Bennion, Bailey and Norbury on Statutory Interpretation, para 16.7, state:

“Where a heading differs from the material it describes, this puts the court on inquiry, but it is most unlikely to be right to allow the plain meaning of the words to be overridden purely by reason of a heading.”

In the present case, the heading to section 6 cannot bear the weight of a conclusion so fundamental to the operation of the 2018 Act and cannot overcome the problems with the retained EU law analysis identified earlier in this judgment. On the contrary, the broad reading of sections 6(1) and (2) is consistent with the purpose of the legislation. As explained above, there is nothing in the provisions of the 2018 Act or in any relevant external material to support the view that its purpose was to substitute retained EU law for the EU law applicable to accrued causes of action at the time they arose. If, however, it was the intention of Parliament that accrued causes of action should continue to be governed by EU law as it stood at the time they arose, there is no reason why it should not have been intended that sections 6(1) and 6(2) should apply to such cases. On the contrary, the policy underlying the 2018 Act was, in part, to extricate the United Kingdom from the EU legal system and to withdraw from the supervision of the CJEU.

A broad reading of sections 6(1) and (2) would be consistent with the declaratory theory of judicial decisions. When the causes of action relied upon by Mr and Mrs Lipton accrued, EU law applied in the United Kingdom through the operation of section 2 of the 1972 Act and domestic courts were charged with the interpretation and application of EU law. Prior to IP completion day the CJEU was the ultimate interpretative authority on EU law. On IP completion day, Parliament replaced the CJEU with the domestic court system, with prospective effect, as the ultimate interpretative authority on EU law as it still applies within the United Kingdom. The status of the CJEU in this regard was inevitably terminated as a result of the withdrawal of the United Kingdom from the EU legal system.

Departure from retained EU case law

The power of certain domestic courts in the United Kingdom to depart from any retained EU case law appears in sections 6(3) and (4) of the 2018 Act. The power to depart has been amended by the Retained EU Law (Revocation and Reform) Act 2023, but the relevant amendments to section 6 either have not been brought into force yet or have prospective effect only. For present purposes it is convenient to address this issue by reference to section 6 as it stood immediately following IP completion day until the 2023 Act amendments came into force on 1 January 2024.

During that period, section 6(3) and (4) provided:

“(3)

Any question as to the validity, meaning or effect of any retained EU law is to be decided, so far as that law is unmodified on or after IP completion day and so far as they are relevant to it—

(a)

in accordance with any retained case law and any retained general principles of EU law, and

(b)

having regard (among other things) to the limits, immediately before IP completion day, of EU competences.

(4)

But—

(a)

the Supreme Court is not bound by any retained EU case law,

…

(ba) a relevant court or relevant tribunal is not bound by any retained EU case law so far as is provided for by regulations under subsection (5A), and

(c)

no court or tribunal is bound by any retained domestic case law that it would not otherwise be bound by.”

The effect of section 6(4) is that, in certain circumstances, certain courts in the United Kingdom are not bound by “any retained EU case law”. “Retained EU case law” is defined by section 6(7):

““retained EU case law” means any principles laid down by, and any decisions of, the European Court, as they have effect in EU law immediately before IP completion day and so far as they—

(a)

relate to anything to which section 2, 3 or 4 applies, and

(b)

are not excluded by section 5 or Schedule 1,

(as those principles and decisions are modified by or under this Act or by other domestic law from time to time)”

On the retained EU law analysis preferred by the majority of the court in this appeal, causes of action accruing under EU law prior to IP completion day are transposed into retained EU law within section 6(3) and therefore the power conferred by section 6(4) to depart from any retained EU case law would apply in proceedings where such rights were in issue. The consistent availability of a power to depart whether the cause of action arose before or after IP completion day might be considered a point in favour of the retained EU law analysis. On the other hand, on the accrued rights analysis, which I consider correct, the rights would remain governed by EU law as opposed to retained EU law and proceedings in which they were in issue would appear to fall outside the power conferred by section 6(4) to depart from any retained EU case law. This, it might be suggested, demonstrates that the accrued rights analysis must be incorrect because it could not have been intended that the power to depart from retained EU case law should not apply in cases concerning rights which accrued in EU law prior to IP completion day. However, this is not necessarily correct.

We have seen (paras 223-226 above) that article 127(1) of the Withdrawal Agreement provides that EU law shall be applicable to and in the United Kingdom during the transition period, unless otherwise provided by the Withdrawal Agreement. Article 4(4) provides that provisions of the Withdrawal Agreement referring to EU law (which would appear to include article 127) shall in their interpretation and implementation be interpreted in conformity with the relevant case law of the CJEU handed down before the end of the transition period. Article 127(3) provides that during the transition period the provisions of EU law made applicable by the Withdrawal Agreement shall produce in respect of and in the United Kingdom the same legal effects as those which it produces within the EU and its Member States, and this would also appear to be required by article 4(1). These provisions are given effect within the United Kingdom by section 7A of the 2018 Act which gives domestic courts within the United Kingdom the power to disapply domestic legislation, including the 2018 Act, insofar as it is incompatible with the Withdrawal Agreement.

These provisions strongly suggest that the power to depart from retained EU case law conferred by section 6(4) of the 2018 Act only applies in cases where the cause of action accrued after IP completion day. Consider an appeal coming before the Supreme Court after IP completion day where the claimant’s cause of action accrued during the transition period. The defendant invites the court to depart from a pre-IP completion day decision of the CJEU which would, if followed, determine the matter in favour of the claimant. If the court were to refuse to apply the relevant CJEU case, there would appear to be at least a serious risk that the United Kingdom would be in breach of articles 4 and 127 of the Withdrawal Agreement. First, it would arguably breach article 127(1) read with article 4(4) which requires judges in the United Kingdom to interpret and apply the EU law in force in the United Kingdom during the transition period “in conformity with the relevant case law of the [CJEU] handed down before the end of the transition period”. Secondly, it would arguably breach article 127(1) read with article 127(3) and article 4(1) because the EU law which was supposed to apply during the transition period would not be producing the same legal effects within the United Kingdom as within Member States. If the power to depart applied in such a case, it would appear to create the prospect that there would be a different outcome for an alleged breach of EU law arising in the United Kingdom during the transition period compared with a materially identical alleged breach arising in a Member State in the same period. It is difficult to see how this could be reconciled with the requirements of the Withdrawal Agreement. (See AT v Secretary of State for Work and Pensions, considered at para 226 above.)

If it is correct that the power to depart from retained EU case law conferred by section 6(4) of the 2018 Act only applies in cases where the cause of action accrued after IP completion day, this may have an important bearing on the present issue. If, as the majority maintain, rights arising before IP completion day are transposed into retained EU law, the power to depart conferred by section 6(4) could, at least on its face, be exercised in any case involving EU-derived rights whenever the cause of action accrued, including cases where the exercise of the power is seemingly prohibited by articles 4 and 127 of the Withdrawal Agreement and section 7A of the 2018 Act. This cannot have been the intention. By contrast, the accrued rights analysis is consistent with the provisions of the Withdrawal Agreement and section 7A. First, on the accrued rights approach the courts apply EU law to causes of action which accrued during the transition period, in compliance with article 127(1). Secondly, in such cases courts are bound by CJEU decisions handed down before IP completion day with no power to depart, in compliance with articles 4(1), 4(4) and 127(3). Thirdly, by virtue of section 6(2) of the 2018 Act, courts may take into account post-IP completion day case law of the CJEU, which allows judges to comply with the apparent obligation under article 4(5) to have due regard to the relevant case law of the CJEU handed down after the end of the transition period.

The obligations imposed on the United Kingdom considered above apply to the transition period and not to the earlier period when the United Kingdom was a Member State. However, an approach which allowed UK courts to depart from CJEU case law when deciding claims which accrued when the United Kingdom was a Member State, but not when deciding claims which accrued during the transition period would make no sense.

These considerations would appear, therefore, to point towards the conclusion that the power to depart conferred by section 6(4) relates only to matters governed by retained EU law (or its successor, assimilated law). In general, these will be matters arising after IP completion day. They lend strong support to the view that retained EU law within the 2018 Act is a body of rules brought into existence on IP completion day for the purpose of applying thereafter with prospective effect and has nothing to do with rights accrued in EU law before IP completion day. This is the view expressed by the editors of Chitty on Contracts. Having referred to the general principle that the overruling of judicial decisions is retroactive in its effect, they continue (at para 1-034):

“However, the courts’ powers to depart from retained EU case law in s. 6(3)-(6) of the 2018 Act clearly apply only to the interpretation of “retained EU law” and this category of law came into existence only on IP completion day. It is submitted, therefore, that this means that where the Supreme Court or a listed appellate court departs from retained EU case law, its decision should be seen as retrospective but only as regards matters governed by retained EU law and, therefore, at least in principle, arising after IP completion day.” (original emphasis). (See also Whittaker (2021) 137 LQR 477, 486-7.)

Accordingly, it is far from clear that the accrued rights analysis, which I consider correct, is inconsistent with section 6 of the 2018 Act and must, therefore, be rejected. On the contrary, it seems to me that the retained EU law analysis, preferred by the majority, is difficult to reconcile with the operation of section 6(4) of the 2018 Act.

For these reasons I consider that the law applicable to Mr and Mrs Lipton’s claim is EC Regulation 261/2004, the very same law which Deputy District Judge Printer had to apply when he heard their claim at first instance in June 2019. The causes of action on which Mr and Mrs Lipton rely in these proceedings, which arose in EU law and which were given effect in domestic law by the European Communities Act 1972, were not transposed into retained EU law by the 2018 Act but were preserved as rights in EU law by section 16(1) of the Interpretation Act 1978.

On the substantive issue, I consider that the defence under article 5(3) fails for the reasons given by Lord Sales and Lady Rose in their joint judgment. I would therefore dismiss the appeal.