BAILII [Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

England and Wales Court of Appeal (Civil Division) Decisions

You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> The Civil Aviation Authority v Jet2.Com Ltd, R. (on the Application of) [2020] EWCA Civ 35 (28 January 2020)
Cite as: [2020] WLR(D) 51, [2020] EWCA Civ 35, [2020] QB 1027, [2020] 4 All ER 374, [2020] 2 WLR 1215

[New search] [View without highlighting] [Printable PDF version] [View ICLR summary: [2020] WLR(D) 51] [Buy ICLR report: [2020] QB 1027] [Buy ICLR report: [2020] 2 WLR 1215] [Help]

Neutral Citation Number: [2020] EWCA Civ 35
Case No: C1/2019/0621


[2018] EWHC 3364 (Admin) and [2019] EWHC 336 (Admin)

Royal Courts of Justice
Strand, London, WC2A 2LL

B e f o r e :





- and –



- and -




Sam Grodzinski QC, Tamara Oppenheimer and Anna Medvinskaia
(instructed by Mayer Brown International LLP) for the Appellant
Charles Béar QC and Nicolas Damnjanovic (instructed by Norton Rose Fulbright LLP)
for the Respondent
David Pievsky (instructed by The Law Society of England and Wales)
(written submissions only) for the Intervener

Hearing dates: 4 and 5 December 2019



Crown Copyright ©

    Lord Justice Hickinbottom :


  1. In considering legal advice privilege ("LAP") in 1881, Bacon VC said this
  2. "This subject is always a difficult one. On the one hand I have to consider the right of the Plaintiff to discovery, and on the other hand, to consider what are the rights of the Defendants to protect themselves against disclosing anything that has taken place in the course of confidential communications" (Wheeler v Le Marchant (1881) 17 Ch D 675 at page 677).

    In the event, the Vice Chancellor held that letters which passed between solicitors and surveyors in relation to the grant of a lease, sent with no litigation active or contemplated, were privileged from disclosure in later proceedings for the specific performance of the lease. Four days later, the Court of Appeal disagreed concluding that the letters were disclosable, and overturned his judgment. Since then, the subject has not become any more straightforward. Indeed, given the more complex arrangements that now exist for commercial transactions and the obtaining of legal advice, including new modes of communication between those involved in such activities, the difficulties have been compounded.

  3. This appeal raises important issues concerning LAP, notably:
  4. i) whether, for a communication to fall within the scope of that privilege, it must have had the dominant purpose of seeking or giving legal advice; and

    ii) in the light of the answer to (i), the proper approach to determining the privileged status of email communications between multiple parties where one of the senders or recipients is a lawyer.

    It also potentially raises issues concerning the proper approach to the collateral waiver of privilege in respect of documents otherwise non-disclosable, as the result of the voluntary disclosure of other privileged documents.

  5. The issues arise in the context of judicial review proceedings issued on 12 April 2018, brought by the Respondent ("Jet2"), a company operating flights to and from the United Kingdom, against the Appellant ("the CAA"), the UK aviation industry regulator, challenging the lawfulness of the CAA's decisions (i) to issue a press release in December 2017 and (ii) subsequently to publish correspondence between the CAA and Jet2 in February 2018 including the provision of such correspondence to the Daily Mail. Both the press release and the CAA correspondence criticised Jet2's refusal to participate in an alternative dispute resolution scheme for the resolution of consumer complaints which the CAA had promoted and in which almost all other large domestic airlines, and a substantial number of non-domestic airlines flying into the UK, had chosen to participate ("the ADR Scheme"). The grounds of challenge to those decisions relevant to this appeal are that the CAA had no power to make the publications or alternatively, if it had such power, it exercised the power for unauthorised and improper purposes namely to damage Jet2's trading interests, to punish Jet2 for its decision not to join the ADR Scheme and to put pressure on Jet2 to join the voluntary scheme.
  6. On 16 January 2018, before the publication of the correspondence (and, of course, well before the issue of proceedings), Jet2 wrote to the CAA complaining about the press release ("the 16 January 2018 letter"); to which the CAA responded on 1 February 2018 ("the 1 February 2018 letter"). Given the grounds of challenge, Jet2 made an application in the judicial review claim for disclosure of several categories of document, including (e) all drafts of the 1 February 2018 letter and (f) all records of any discussions of those drafts. Morris J concluded that all of those documents should be disclosed. Following a further hearing, he held that, even if he had found that those documents were privileged, that privilege was waived by the CAA in respect of all of those documents by the disclosure of an email dated 24 January 2018 from Matthew Buffey, the CAA's Head of Consumer Enforcement Department, to several CAA employees including Serena Lim, a Principal Legal Adviser. In this appeal, the CAA contend that the judge erred in both judgments, and in ultimately concluding that all drafts of the 1 February 2018 letter and records of discussion of the drafts should be disclosed.
  7. Before us, Sam Grodzinski QC, Tamara Oppenheimer and Anna Medvinskaia of Counsel, appeared for the CAA; and Charles Béar QC and Nicolas Damnjanovic of Counsel for Jet2. We also had the benefit of written submissions from David Pievsky for the Law Society of England and Wales as Intervener. At the outset, I thank them all for their contribution to the debate.
  8. The Factual Background

  9. For many years, consumer groups and governments, both national and European, have been anxious to increase protection for consumers, including ensuring prompt and proportionate disposal of consumer complaints.
  10. Directive 2013/11/EU of the European Parliament and of the Council of 21 May 2013 on alternative dispute resolution for consumer dispute requires Member States to ensure that consumers can, on a voluntary basis, access ADR processes for disputes concerning contractual relations between consumers and traders. In the UK, that Directive is enforced through the Alternative Dispute Resolution for Consumer Disputes (Competent Authorities and Information) Regulations 2015 (SI 2015 No 542) ("the 2015 Regulations"), which provide for an ADR scheme which is bilaterally voluntary, i.e. trader and consumer are each able to elect whether or not to adopt it.
  11. So far as the air passenger industry is concerned, the CAA has for many years funded a service, the Passenger Advice and Complaints Team, which has a scheme to mediate consumer complaints ("the PACT Scheme"). However, under the 2015 Regulations, the CAA is the designated competent authority; and it is a vigorous proponent of the new ADR Scheme, expressing support for primary legislation to make participation in the scheme mandatory for the air passenger industry. In the meantime, it has the express objective of obtaining full participation in the scheme, and closing down the PACT Scheme. But, at present, the ADR Scheme is still voluntary; and Jet2 has chosen not to participate in it. Jet2 continues to rely on the PACT scheme.
  12. As part of its promotion of the new ADR Scheme, on 27 December 2017, the CAA published a policy document, "ADR in the aviation sector – a first review" (CAP 1602) ("the Review"), together with a press release headed "Thousands more airline passengers are now receiving compensation thanks to [ADR]" ("the Press Release") criticising airlines who had opted not to participate in the new scheme.
  13. The Press Release gave the number of airlines which had signed up to the new scheme; and identified others by name which had not, urging them to do so. It singled out Jet2, the largest UK airline not to have signed up, for particular criticism. It said:
  14. "… Jet2, the Leeds-based airline, one of the largest UK airlines, has 'inexplicably and persistently' refused to sign up – denying its customers access to a fair arbitration service, which can legally resolve disputed complaints fairly and efficiently".

    The internal quotation was a reference to observations by Andrew Haines, the CAA's Chief Executive Officer, whom the Press Release quoted more fully, as follows:

    "… ADR is good for UK consumers, which is why it is extremely disappointing that Jet2, one of the UK's largest airlines, has so far inexplicably and persistently refused to sign up, denying their passengers access to an independent arbitration service.
    Clearly this decision puts Jet2's customers, and those of other airlines that haven't yet signed up, at a distinct disadvantage, and in many cases, could mean their passengers are denied the fundamental rights they are entitled to.
    I am therefore calling on Jet2 and other airlines including Aer Lingus and Emirates to commit to ADR in the interests of their passengers".
  15. Jet2 considered that these comments were false (or, at least, misleading) and unfair. On 16 January 2018, its Executive Chairman, Philip Meeson, wrote to Mr Haines, complaining about the tone and content of the Press Release and giving reasons why Jet2 had not signed up for the ADR Scheme: it considered ADR untried and untested in this context, and unsuited to the resolution of delay and cancellation claims which formed over 90% of the complaints made and which could, in Jet2's view, better be resolved by other means. On 18 January 2018, Jet2 issued its own press release; and, at some point, it put its 16 January 2018 letter onto its own website.
  16. Following receipt of the 16 January 2018 letter and Jet2's press release, the CAA considered an appropriate response. In particular, in an internal email dated 18 January 2018 to Richard Moriarty (then Group Director of CAA's Consumer and Markets Group, and later Mr Haines's successor as Chief Executive Officer), Richard Stephenson (CAA's Communications Director), Andrew McConnell (CAA's Senior Communications Adviser) and Mr Buffey, Mr Haines said:
  17. "We should develop a narrative around Jet2. They have been (one of) the most litigious airline disputing 261, threatening legal action against the CAA. References to their billionaire chairman might not go amiss in the process. We could share it with Jet2 as our rebuttal of any continuation of such misleading information.
    Attack dogs please Lord S"

    "261" is a reference to Regulation No 261/2004 of the European Parliament and of the Council of 11 February 2004 establishing common rules on compensation and assistance to passengers in the event of denied boarding and of cancellation or long delay of flights. "Lord S" is a reference to Mr Stephenson, the inference of the last line of the email being that Mr Haines wished to see vigorous positive media publicity in response to Jet2's stance.

  18. On 24 January 2018, Mr Buffey circulated a first draft response to the 16 January 2018 letter. The draft was sent under a covering email, addressed to Mr Moriarty, Jackie Knight (a CAA Consumer Enforcement Department Manager) and Serena Lim (a Principal Legal Adviser with the CAA), in which Mr Buffey said
  19. "I wouldn't quite call it 'attack dog' style. More of a cranky alpaca.
    Anyway, see what you think. I'd like to get it to [Mr Haines] and [Mr Stephenson] by cop Thursday if possible."

    That is the only draft of that letter that was before Morris J (or is now before us); although, as will shortly be apparent, there were other drafts and internal communications within the CAA before the response was finalised.

  20. Mr Haines eventually responded to Mr Meeson on 1 February 2018 in a letter which, whilst making it clear that it was not suggested that Jet2 provided a poor overall service, further criticised it for the stance it had taken in relation to the ADR Scheme. In particular, it said:
  21. "Your letter was surprising and extremely disappointing on two fronts; your apparent disregard for the rights of customers when your levels of service fall below that which you say you aspire to and secondly the poor and inconsistent case you make in seeking to defend, what I regard, as your indefensible position.
    It is unfortunate that you chose to put forward such a transparently narrow and self-interested set of arguments against ADR but, more importantly, the arguments are redundant for the reason I set out below".

    In dealing with the arguments Mr Meeson had put forward in his earlier letter, Mr Haines repeated the criticism of denying customers their fundamental rights and asserted that, in publishing the Review and Press Release, the CAA was pursuing a proper purpose by drawing attention to Jet2's ongoing failure to participate in the ADR Scheme. Mr Haines concluded by reserving the right to publish the correspondence between the CAA and Jet2 on the issue.

  22. That correspondence was in fact provided by the CAA to the Daily Mail which, on 6 February published an on-line article and the following day an article in the Money Mail section of the newspaper (headed, "In a leaked letter, Jet2 boss reveals why it won't join new flight delay compensation scheme… Too many customers will win their money back!"), with a follow-up article on 19 February 2018. The articles referred in detail to the correspondence, including quotations, as well as to the Press Release. They repeated some of the criticisms of Jet2 made in the 1 February 2018 letter, and generally adopted the CAA stance on the issue of participation in the new scheme.
  23. The Proceedings

  24. On 12 April 2018, Jet2 issued judicial review proceedings challenging the CAA's decision to publish the Press Release and the post-Press Release correspondence to the press, on four grounds.
  25. i) The publications by the CAA were ultra vires. Whilst section 83 of the Civil Aviation Act 1982 gives the CAA the power to publish information and advice for the purpose of assisting consumers to compare services and facilities used in connection with the use of air transport services or with a view to improving the standards of such services, it does not give it the power publicly to criticise Jet2 for choosing not to participate in the new ADR Scheme. Thus, the CAA had no power to issue the Press Release or to publicise the correspondence as it did.

    ii) In publishing the Press Release and the correspondence, the CAA acted for unauthorised purposes, namely to damage Jet2's trading interests and reputation by singling it out for severe criticism and thereby punishing Jet2 for its decision not to join the new scheme and to put pressure on Jet2 to take the voluntary step of joining the scheme. Alternatively, by acting in such a way, the CAA took into account such matters, which were irrelevant.

    iii) The publications were made in breach of the duty of procedural fairness.

    iv) The decision to publish was irrational.

    This appeal particularly concerns (i) and (ii).

  26. On 8 June 2018, Turner J granted permission to proceed on the basis that there was an arguable case.
  27. On 3 August 2018, the CAA served Detailed Grounds of Opposition which, so far as relevant, contended that (i) publication of the Press Release and correspondence fell within section 83, and (ii) there was no improper purpose, the purpose of publication being to promote the interests of consumers by making them aware of which airlines had not signed up to the new scheme which (the CAA considered) benefited customers.
  28. That formal response to the claim was accompanied by a witness statement of Mr Moriarty dated 3 August 2018. Mr Moriarty made the following observations about Mr Haines' email of 18 January 2018 quoted above (see paragraph 12). He said (at paragraph 27):
  29. "… The phrase 'billionaire chairman' in the first paragraph was a reference to Mr Meeson. I consider that it was inappropriate to suggest that any part of the CAA's 'narrative' should be making reference to any individual in this way, and Mr Haines also acknowledges that it was inappropriate to have made such a reference. The CAA does not and should not make negative personal comments about individuals within businesses, and in fact, no comment about Mr Meeson personally was made externally by the CAA. The phrase 'attack dogs' in the second paragraph was a reference to pressing on with the CAA's media publicity. 'Lord S' is a reference to Mr Stephenson. Mr Haines' passion for consumer rights did, on occasion, lead him to express himself in colourful terms. This email was not reflective of any part of the approach taken by the CAA. This can be seen from an email sent to Mr Haines by one of my colleagues, Mr Buffey, and from the content of the material that we published."

    The email to which Mr Moriarty refers was exhibited to his statement. It was not in fact to Mr Haines, but was the email of 24 January 2018 to Mr Moriarty, Ms Knight and Ms Lim, quoted at paragraph 13 above.

  30. On 26 October 2018, Jet2 made an application for specific disclosure of several categories of document including all drafts of the 1 February 2018 letter and all CAA records of any discussions concerning those drafts, said to be necessary in order to understand the CAA's reasons and purpose behind the publication of the 1 February 2018 letter and therefore relevant to the "improper purposes" ground of challenge.
  31. In response, the CAA relied upon a statement of Ms Lim dated 13 November 2018, which confirmed that there were further drafts of the 1 February 2018 letter and there were internal discussions about those drafts by several people within the CAA including Mr Haines. However, she said (at paragraph 13) that either she or Dilsha Caldera (another in-house Legal Adviser at the CAA) "were involved in those discussions and gave advice in relation to the various drafts, the content of which advice is privileged and the CAA does not waive privilege in that advice".
  32. The disclosure application came before Morris J who, in a judgment dated 10 December 2018 ([2018] EWHC 3364 (Admin)) ("the December 2018 judgment"), found that, so far as the preparation of the response to the 16 January 2018 letter was concerned, the CAA lawyers "became involved for the purpose of giving legal advice; and were not involved merely as members of the in-house team of executives providing commercial advice" (see [98]). That finding remains unchallenged.
  33. In terms of legal principle, he held as follows.
  34. i) Claims to LAP are subject to a dominant purpose test. In respect of that proposition, Morris J cited and relied upon the following:

    a) Three Rivers Council v The Governor and Company of the Bank of England (No 5) [2002] EWHC 2730 (Comm); [2003] CP Rep 34 (the first instance judgment of Tomlinson J ("Three Rivers (No 5) (Comm Ct)"), itself citing Hellenic Mutual War Risks Association (Bermuda) Limited v Harrison (The Sagheera) [1997] 1 Lloyd's Rep 160 ("The Sagheera") at pages 167-8);
    b) Three Rivers Council v The Governor and Company of the Bank of England (No 5) [2003] EWCA Civ 474; [2003] QB 1556 (the judgment of this court (Lord Phillips of Matravers MR, Sedley and Longmore LJJ) on appeal from Tomlinson J ("Three Rivers (No 5)") at [32], as commended by Lord Carswell in Three Rivers Council v The Governor and Company of the Bank of England (No 6) [2004] UKHL 48; [2005] 1 AC 610 ("Three Rivers (No 6)") at [70]);
    c) the judgment of Moore-Bick J in United States of America v Philip Morris Inc [2003] EWHC 3028 (Comm) ("Philip Morris (Comm Ct)") at [38]; and
    d) the textbook "Documentary Evidence" by Charles Hollander QC (13th Edition) ("Hollander") at paragraph 17-16.

    From these authorities, the judge concluded (at [95(4)]) that:

    "Whilst I am aware of academic commentaries suggesting that the point is not free from doubt, in my judgment, on the current state of the authorities and obiter observations in the Court of Appeal, claims for [LAP] are, in principle, subject to a dominant purpose test, namely whether the communication or document was brought into existence with the dominant purpose of it or its contents being used to obtain legal advice."

    ii) Therefore (also at [95(4)]):

    "… [I]n normal cases of an email sent to an external lawyer, the issue of dominant purpose is unlikely to arise…. However, the issue may be more acute where material is sent to in-house lawyers, who may have a dual role in the company. Lawyers, particularly in-house solicitors, may often take part in general business discussions which do not involve legal advice. Where the in-house lawyer is clearly being asked for legal advice, privilege is likely to attach. However, where the in-house lawyer is being consulted also as an executive about a largely commercial issue, then the dominant purpose test will fall to be applied."

    iii) With regard to communications sent to multiple addressees, some of whom are lawyers and some of whom are not, he said that the position was not established by authority; but, he continued (at [95(5)]):

    "In my judgment, if the dominant purpose of the email is to seek advice from the lawyer and others are copied in for information only, then the email is privileged, regardless of who it is sent to. If on the other hand, the dominant purpose of the email is to seek commercial views, and the lawyer is copied in, whether for information or even for the purpose of legal advice, then the email, in so far as it is sent to the non-lawyer, is not privileged. Further, if sent to the non-lawyer for a commercial comment, but sent to the lawyer for legal advice, then, in my judgment, the email is not protected by privilege, unless it or the non-lawyer's response discloses or might disclose the nature of the legal advice sought and given."
  35. Applying those principles to this case, the judge found as follows.
  36. "99. Against this background and applying the above principles, any draft of the 1 February letter created before the [CAA's] in-house lawyers were consulted or created without any involvement of in-house lawyers is not privileged. That is the case, even if it were known that in due course legal advice would be taken on the draft, unless the dominant purpose of the person creating the draft was to seek legal advice on it.
    100. Further drafts of the 1 February letter are not covered by privilege unless specifically drafted by the lawyers or for the dominant purpose of obtaining legal advice. Such drafts do not subsequently attract privilege when they were shown to the in-house lawyers. However if a particular draft was created by the in-house lawyers, or by another specifically for the purpose of seeking or giving legal advice then that draft will be privileged.
    101. On the basis that the [CAA's] in-house lawyers were instructed for the purposes of obtaining legal advice, then any communication with those lawyers (to and fro) and including comments and advice on the draft letter (whether on the document itself or in a covering communication) are covered by [LAP]. Moreover any further communication between non-lawyer executives which discloses or might disclose or concerns comments and advice from the in-house lawyers in relation to the draft of the 1 February letter is also covered by legal advice privilege.
    102. Where a draft of the 1 February letter (or even discussion about such a draft) was sent in one email to both in-house lawyers and other non-lawyer personnel within the [CAA] (such as the email of 24 January which has already been voluntarily disclosed), then, even assuming that in so far as the email was sent to the in-house lawyer it is privileged, in so far as it is also sent to a non-lawyer, neither the email nor the response of the non-lawyer is protected by [LAP], unless the content of the email, or the response from the non-lawyer, discloses or is likely to disclose the nature and content of the legal advice sought and obtained. If the email to the non-lawyer clearly seeks, and the response provides, commercial views, with no connection to the legal advice, then it is not covered by [LAP]; here the dominant purpose of the email, as sent to the non-lawyer and any enclosed draft was to obtain commercial views. The email of 24 January falls into this category. (I add that if, contrary to the foregoing, a multi-addressee email of this type is in principle covered by privilege, there would be a strong argument that, assuming Ms Lim was copied in for the purpose of seeking legal advice, in any event by disclosing the email of 24 January the [CAA] has, in this case, waived privilege in this class of document)."
  37. The CAA indicated an intention to appeal against the order for disclosure; and, when directed by the judge to reconsider the privilege status of documents falling within categories (e) and (f), it did not disclose any further documents on the basis that all documents were covered by LAP either as found by the judge or as claimed by it pending an appeal.
  38. At a further hearing on 19 December 2018, the CAA was directed to provide a more detailed witness statement in relation to privilege, identifying separately each email and each attachment over which privilege was claimed and explaining which documents it considered would be privileged under the December 2018 judgment and which would not be privileged under that judgment but over which the CAA still claimed privilege and why.
  39. The CAA responded by way of a witness statement of Imogen Brooks (another Principal Legal Adviser with the CAA), dated 22 January 2019, which considered each of the relevant documents in an annex, Annex A. In the light of the requirements of the 19 December 2018 Order, she identified and considered emails and any attachment(s) discretely (see paragraphs 17-18 of her statement). On the basis of the December 2018 judgment, as set out in paragraph 10 of her statement, in relation to each email and attachment, she applied the following approach:
  40. "(a) First, one asks: is the dominant purpose of the multi-addressee email or an attachment to a multi-addressee email to seek or give legal advice? If so, then the email or the attachment to that email, as the case may be, is privileged.
    (b) If the answer to subparagraph (a) above is no, then one asks: does the email or the attachment disclose; or is it 'likely' to disclose; or 'might' it disclose, the nature and content of the legal advice sought from, or given by, the in-house lawyer? If so, the email or attachment to that email is privileged. I have understood the word 'might', as connoting a realistic possibility."

    In my view, Ms Brooks was correct to construe Morris J's judgment as applying to documents and communications which would, or might realistically, disclose legal advice. It seems to me that he was using "likely" in that sense (rather than as meaning more likely than not).

  41. However, Ms Brooks also applied this approach in a way consistent with Morris J's finding (at [102] of the December 2018 judgment) that the 24 January 2018 email (see paragraph 13 above) was not protected from disclosure by LAP: contrary to the CAA's own view that that email was covered by privilege (and, by disclosing it, the CAA did not waive privilege over more than that email itself), the judge held that that email and attachment (i) was not prepared for the dominant purpose of obtaining legal advice and was not copied to the other (non-lawyer) individuals for information only, and (ii) did not disclose the nature of the legal advice being sought nor might it do so (see paragraphs 10-16 of her statement).
  42. Applying this approach, in respect of eleven documents, Ms Brooks was unable to determine whether, on the basis of the December 2018 judgment, they were privileged or not, because of two particular problems, namely, in determining whether a document disclosed (or might disclose) the nature of legal advice, sought or given, it was unclear from the judgment whether (i) the particular document should be looked at discretely or in the context of the communications which preceded and followed it, and (ii) it is necessary for legal advice to be specifically requested (paragraphs 23-26). (In the event, at the 4 February 2019 hearing, on the basis of proportionality and pragmatism, without conceding that the documents were in fact not privileged, the CAA indicated that it would treat these documents as disclosable under the December 2018 judgment.)
  43. Otherwise, on the basis of the approach she adopted, in respect of most of the other documents she listed, Ms Brooks accepted that it could not be said that either (i) the dominant purpose was to seek legal advice, or (ii) the email/attachment disclosed or might disclose the nature of the legal advice being sought from or given by the in-house lawyer. Such documents would therefore not be covered by LAP on the basis of the December 2018 judgment. However, in respect of these documents, she said:
  44. "CAA claims privilege on the basis that the email forms part of the continuum of communications between a client and lawyer, for the purposes of seeking or receiving legal advice".

    The concept of a "continuum of communications between a client and lawyer" derives from Balabel v Air India [1988] Ch 317 ("Balabel"), to which I shall return shortly (see paragraph 62 and following below).

  45. A further hearing took place on 4 February 2019 to determine the CAA's application for permission to appeal. In response to that application, in an alternative to its primary submission that none of these communications was privileged, if and insofar as they were privileged, Jet2 submitted for the first time that an appeal would have no real prospect of success because, amongst other things, by disclosing the 24 January 2018 email, the CAA had waived privilege in all communications concerning the draft 1 February 2018 letter.
  46. In a further judgment handed down on 5 February 2019 ([2019] EWHC 336 (Admin)) ("the February 2019 judgment"), Morris J accepted that submission. In respect of the scope of the "transaction" over which there had been a waiver of privilege by disclosure of the 24 January 2018 email, he said this (at [22]):
  47. "As regards the 'transaction', the issue between the parties is whether the transaction is limited to the email itself or whether it is the course of correspondence and discussion. Whilst in some cases the transaction may be limited to the document disclosed itself, I do not accept that that is the position here. The transaction must, at the very least, include the preceding 'attack dogs' email from Mr Haines. Indeed, the transaction or issue in respect of which the disclosure was made by Mr Moriarty is the approach taken by the CAA in response to [the] 16 January letter. That involved a single process of internal discussion, which does not have discrete parts, and the email of 18 January and the email of 24 January form part of that single process. The fact that the process of discussion took place by way of emails is not of itself a basis for distinguishing it from circumstances where the discussion may have taken place in an oral conversation at one and the same time."

    He consequently concluded that the transaction in question comprised all drafts of the 1 February 2018 letter, and emails and internal discussion about those drafts, in the period between the 16 January letter to the publication of the Daily Mail article on 7 February 2018; and fairness required disclosure of the entire chain of discussion whether or not individual documents include legal advice (see [23]-[24]).

  48. In two appeals, the CAA sought permission to appeal to this court, submitting that, in holding that the drafts and communications about the drafts were not privileged (or, if privileged, that privilege had been waived), the judge made four errors of law.
  49. Ground 1: He erred in holding that claims for LAP are in principle subject to a dominant purpose test, i.e. that the privilege will only apply where the document or other communication was brought into existence with the dominant purpose of it or its contents being used to give or seek legal advice.

    Ground 2: Particularly as the result of the dominant purpose test which he applied, he erred with respect to the proper approach to be adopted when considering whether multi-addressee communications (notably emails from or to both lawyers and non-lawyers) are protected by LAP.

    Ground 3: He erred in holding that an assessment of whether an email and any attachment must be carried out discretely and without reference to any attachment or covering email respectively.

    Ground 4: In a separate appeal in respect of the February 2019 judgment (which is contingent upon the CAA succeeding, under Grounds 1-3, in showing that the drafts etc were privileged), the CAA contend that the judge erred in his approach to collateral waiver. In particular, the judge incorrectly found that the relevant "transaction" for the purposes of collateral waiver extended to all emails and internal discussions in the period from the 16 January 2018 letter up to the publication of the Daily Mail article on 7 February 2018, in that the voluntary disclosure of the 24 January 2018 email resulted in the collateral waiver of privilege in respect of all those documents.

  50. Morris J gave permission to appeal in respect of Ground 4; and, on 7 June 2019, Haddon-Cave LJ granted permission to appeal on all other grounds.
  51. Legal Advice Privilege


  52. In line with the principle that the best probative evidence should be available to the court, admissible and relevant communications are generally disclosable in legal proceedings even if they are confidential. However, there is a rule that evidence which falls within the scope of legal professional privilege does not have to be disclosed, even if it is admissible and relevant to the issues in the litigation.
  53. As originally formulated by the courts, the privilege covered only confidential communications made between a lawyer and his client, or a lawyer or client and a third party, which came into existence for the purposes of litigation ("litigation privilege"). The rationale of the privilege was said to be that "it is an absolute necessity that a man, in order to prosecute his rights or defend himself from an improper claim, should have resource to the assistance of professional lawyers" and that "he should be able to make a clean breast of it to the gentlemen whom he consults" in the sure knowledge that his communications to and from the lawyer will be "kept secret" unless disclosed with his consent (Anderson v Bank of British Columbia (1876) 2 Ch D 644 ("Anderson") at page 649 per Sir George Jessel MR); and that a party in adversarial proceedings had no right to have access to his opponent's brief nor any obligation to disclose any part of that brief.
  54. However, in Greenough v Gaskell (1833) 1 My & K 98, Lord Brougham LC held that the privilege applied to all confidential communications between a lawyer and his client made for the purposes of giving or obtaining legal advice, whether or not there were existing or contemplated proceedings (i.e. LAP). This appeal concerns LAP.
  55. LAP involves the right to withhold disclosure of relevant, and possibly crucial, evidence from legal proceedings; and consequently it potentially detracts from the fairness of those proceedings. Such a right requires powerful justification. Having comprehensively reviewed the authorities, the rationale for LAP was described by Lord Taylor of Gosforth CJ (with whom the rest of the Judicial Committee of the House of Lords agreed) in R v Derby Magistrates' Court ex parte B [1996] AC 487 ("ex parte B") at page 507C-D, as follows:
  56. "The principle which runs through all these cases, and the many other cases which were cited, is that a man must be able to consult his lawyer in confidence, since otherwise he might hold back half the truth. The client must be sure that what he tells his lawyer in confidence will never be revealed without his consent. Legal professional privilege is thus much more than an ordinary rule of evidence, limited in its application to the facts of a particular case. It is a fundamental condition on which the administration of justice as a whole rests."
  57. Legal professional privilege is regarded of such importance that it has been described as "absolute"; but, like most rights, it is not absolute in the true sense of that word. As I describe below (paragraphs 74 and following), where a document is produced for the purpose of litigation, it will nevertheless be disclosable if that is not its dominant purpose. Furthermore, the right of the client to withhold disclosure of the evidence cannot be relied upon where to do so would further fraud or crime. Nevertheless, the vital nature of the privilege, and its underlying rationale, has been regularly emphasised by the courts. Two examples will suffice.
  58. First, Lord Hoffman in R (Morgan Grenfell & Co Limited) v Special Commissioner of Income Tax [2002] UKHL 21; [2003] AC 563 (with whom the rest of the Judicial Committee agreed) at [7], said this:
  59. "… [Legal professional privilege] is a fundamental human right long established in the common law. It is a corollary of the right of any person to obtain skilled advice about the law. Such advice cannot be effectively obtained unless the client is able to put all the facts before the adviser without fear that they may afterwards be disclosed and used to his prejudice. The cases establishing this principle are collected in the speech of [Lord Taylor in ex parte B]. It has been held by the European Court of Human Rights to be part of the right of privacy guaranteed by article 8 of the [European Convention on Human Rights] (Campbell v United Kingdom (1992) 15 EHRR 137; Foxley v United Kingdom (2000) 31 EHRR 637) and held by the European Court of Justice to be part of Community law (A M & S Europe Limited v Commission of the European Communities (Case 155/79) [1983] QB 878)."
  60. Second, Lord Scott of Foscote in Three Rivers (No 6) at [34], having considered the relevant authorities, continued:
  61. "None of these judicial dicta tie the justification for [LAP] to the conduct of litigation. They recognise that in the complex world in which we live there are a multitude of reasons why individuals, whether humble or powerful, or corporations, whether large or small, may need to seek the advice or assistance of lawyers in connection with their affairs; they recognise that the seeking and giving of this advice so that the clients may achieve an orderly arrangement of their affairs is strongly in the public interest; they recognise that in order for the advice to bring about that desirable result it is essential that the full and complete facts are placed before the lawyers who are to give it; and they recognise that unless the clients can be assured that what they tell their lawyers will not be disclosed by the lawyers without their (the clients') consent, there will be cases in which the requisite candour will be absent. It is obviously true that in very many cases clients would have no inhibitions in providing their lawyers with all the facts and information the lawyers might need whether or not there were the absolute assurance of non-disclosure that the present law of privilege provides. But the dicta to which I have referred all have in common the idea that it is necessary in our society, a society in which the restraining and controlling framework is built upon a belief in the rule of law, that communications between clients and lawyers, whereby the clients are hoping for the assistance of the lawyers' legal skills in the management of their (the clients') affairs, should be secure against the possibility of any scrutiny from others, whether the police, the executive, business competitors, inquisitive busybodies or anyone else (see also paragraphs 15.8-15.10 of Zuckerman's Civil Procedure (2003) where the author refers to the rationale underlying legal advice privilege as 'the rule of law rationale'). I, for my part, subscribe to this idea. It justifies, in my opinion, the retention of [LAP] in our law, notwithstanding that as a result cases may sometimes have to be decided in ignorance of relevant probative material."
  62. The original formulation of LAP as set out in Greenough v Gaskell essentially continues to apply. Lord Rodger of Earlsferry put it thus in Three Rivers (No 6) at [50]:
  63. "… [LAP] attaches to all communications made in confidence between solicitors and their clients for the purpose of giving or obtaining legal advice even at a stage when litigation is not in contemplation".

    As Mr Grodzinski submitted, it is therefore necessary for the party seeking to rely on the right to withhold evidence to satisfy the criteria of each of four elements: there must be (a) a communication (whether written or oral); (b) between a client and a lawyer, or a lawyer and his client; (c) made in confidence; (d) for the purpose of giving or obtaining legal advice.

  64. The focus of this appeal (and, especially, Grounds 1 and 2) is whether the "purpose" in (d) must be the dominant purpose; but some aspects of the other elements are also pertinent to the issues before us. The following five propositions, relevant to this appeal, arise from the authorities. The first three propositions are uncontroversial; the other two require some greater consideration.
  65. Proposition 1

  66. Although the older cases (decided at a time when legal advice was generally obtained from or through solicitors in private practice) concern external lawyers, LAP applies to communications, not only with a lawyer in independent practice, but also with an in-house lawyer (see, e.g., Alfred Crompton Amusement Machines Ltd v Customs & Excise Commissioners (No 2) [1972] QB 102, and Financial Services Compensation Scheme Limited v Abbey National Treasury Services Limited [2007] EWHC 2868 (Ch) at [9]).
  67. Proposition 2

  68. Although the privilege attaches to communications between a lawyer and his client, the law recognises that legal advice is not given for hypothetical purposes, but to be considered and (insofar as accepted) applied by the client. It is therefore well-established that it covers, not only a document from the lawyer containing advice and the client's own written record of advice (whether given in writing or orally), but also any communication (again, whether written or oral) passing on, considering or applying that advice internally (Bank of Nova Scotia v Hellenic Mutual War Risks Association (Bermuda) Limited (The Good Luck) [1992] 2 Lloyd's Rep 540 ("The Good Luck") at pages 540-1 per Saville J, and USP Strategies Plc v London General Holdings Limited [2004] EWHC 373 (Ch) ("USP Strategies") at [19(c)] per Mann J). Indeed, there are circumstances in which the privilege will attach to the dissemination of advice to third parties (USP Strategies and Gotha City v Sotheby's [1998] 1 WLR 114). Equally, LAP attaches to communications from a lawyer to a third party containing information provided by the client to the lawyer which is covered by LAP and which the client has given the lawyer authority to disclose (Raiffeisen Bank International AG v Asia Coal Energy Ventures Limited and Ashurst LLP [2020] EWCA Civ 11 at [63]).
  69. Proposition 3

  70. The privilege applies to communications only for the purpose of obtaining or giving legal advice, and not (e.g.) other professional or commercial advice. Wheeler v Le Marchant (cited in paragraph 1 above) concerned the disclosure of various documents passing between a client and his surveyor, which appear to have included not only primary factual information but also surveying advice and opinion. Cotton LJ said (at pages 684-5):
  71. "It is said that as communications between a client and his legal advisers for the purpose of obtaining legal advice are privileged, therefore any communication between the representatives of the client and the solicitor must be also privileged. That is a fallacious use of the word 'representatives'. If the representative is a person employed as an agent on the part of the client to obtain the legal advice of the solicitor, of course he stands in exactly the same position as the client as regards protection, and his communications with the solicitor stand in the same position as the communications of his principal with the solicitor. But these persons were not representatives in that sense. They were representatives in this sense, that they were employed on behalf of the clients, the defendants, to do certain work, but that work was not the communicating with the solicitor to obtain legal advice. So their communications cannot be protected on the ground that they are communications between the client by his representatives and the solicitor. In fact, the contention of the Respondents comes to this, that all communications between a solicitor and a third person in the course of his advising his client are to be protected. It was conceded there was no case that went that length, and the question is whether, in order fully to develop the principle with all its reasonable consequences, we ought to protect such documents. Hitherto such communications have only been protected when they have been in contemplation of some litigation, or for the purpose of giving advice or obtaining evidence with reference to it. And that is reasonable, because then the solicitor is preparing for the defence or for bringing the action, and all communications he makes for that purpose, and the communications made to him for the purpose of giving him the information, are, in fact, the brief in the action, and ought to be protected. But here we are asked to extend the principle to a very different class of cases, and it is not necessary, in order to enable persons freely to communicate with their solicitors and obtain their legal advice, that any privilege should be extended to communications such as these."

    Therefore, as Longmore LJ said in Three Rivers (No 5) (at [18]):

    "This case thus makes clear that [LAP] does not extend to documents obtained from third parties to be shown to a solicitor to advise."

    Proposition 4

  72. The second proposition to which I referred concerns the extent to which legal advice, privileged when given, can be disseminated internally and externally without the loss of privilege, where the law has taken a flexible and realistic approach, reflecting the realities of modern corporate and commercial arrangements. However, the law has taken a somewhat different approach to the collection of material, internally and externally, for the purpose of obtaining legal advice, based upon the principle set out in Wheeler v Le Marchant. The fourth proposition derived from the authorities is that material collected by a client (or by his lawyer on his behalf) from third parties or independent agents for the purposes of instructing lawyers to give advice is not covered by LAP; and, further, where the relevant client is a corporation, documents or other materials between an employee of that corporation and a co-employee or the corporation's lawyers, even if required or designed to equip those lawyers to give legal advice to the corporation, do not attract LAP unless the employee was tasked with seeking and receiving such advice on behalf of the company (Three Rivers (No 5) at [8] and following, and the judgment of this court (Sir Brian Leveson PQBD, Sir Geoffrey Vos C and McCombe LJ) in Director of the Serious Fraud Office v Eurasian Natural Resources Corporation Limited [2018] EWCA Civ 2006; [2019] 1 WLR 791 ("Eurasian") especially at [79]-[81] and [123]-[130]).
  73. On this issue, Three Rivers (No 5) is binding upon this court (see Eurasian at [130]); but, like the constitution of this court in Eurasian, I find parts of the judgment, including this part, difficult.
  74. The background to the Three Rivers litigation is both well-known and helpfully set out in Eurasian (at [67]-[81]). For the purposes of this appeal, I can be brief. The claimants were the liquidators and creditors of Bank of Credit and Commerce International SA ("BCCI"), who sued the Bank of England ("the Bank") for misfeasance in public office in respect of its supervision of BCCI following its commercial collapse. The claimants sought disclosure of various documents sent by the Bank to solicitors instructed to advise on the preparation and presentation of the Bank's evidence before an enquiry into the supervision of BCCI chaired by Bingham LJ, namely Freshfields Bruckhaus Deringer LLP ("Freshfields").
  75. The Bank instructed, and only communicated with, Freshfields through something called "the Bingham Inquiry Unit" ("the BIU"), comprising three of the Bank's officials. The issue concerned the extent to which documents sent by the BIU to Freshfields were privileged from disclosure in the litigation. Although a number of sub-issues concerning disclosure arose, all of the relevant documents were sent to Freshfields. The role of that firm was the subject of scrutiny; but no issue arose in respect of emails or other communications sent simultaneously to lawyers and non-lawyers for their respective advice/input, as it does in this case.
  76. The inquiry was not adversarial, and so litigation privilege did not apply. The issue in Three Rivers (No 5) was whether the Bank could claim LAP, despite the fact that many of the documents sent to Freshfields were not prepared by the BIU but by other employees of the Bank. At first instance, Tomlinson J held that LAP applied. On appeal, this court disagreed. Permission to appeal to the House of Lords was refused (see [2003] QB 1556 at page 1583G). In Three Rivers (No 6), the House of Lords declined to determine or express views on the issue; and thus "the guiding precedent on the issue [continues] to be the Court of Appeal judgment in Three Rivers (No 5)" (see [46]-[48] per Lord Scott, with whom Lord Rodger (at [49]), Baroness Hale of Richmond (at [61]) and Lord Brown of Eaton-under-Heywood (at [119]) agreed; confirmed in Eurasian at the paragraphs cited above).
  77. In Three Rivers (No 5), on this issue, citing 19th century authorities (including the passage from the judgment of Cotton LJ in Wheeler v Le Marchant quoted above: paragraph 46), and passages from the judgments of Mellish LJ and Baggallay JA in Anderson (cited at paragraph 36 above)), Longmore LJ said (at [19]):
  78. "By the end of the 19th century it was, therefore, clear that [LAP] did not apply to documents communicated to a client or his solicitor for advice to be taken upon them but only to communications passing between that client and his solicitor (whether or not through any intermediary) and documents evidencing such communications."
  79. Longmore LJ then went on to hold that, for these purposes, agents or employees who were not instrumental in instructing the lawyers could be equated with third parties; so that documents or other information given by an employee to an employer or a fellow-employee did not attract LAP, even though on the facts it was intended to be shown to a lawyer (see [22]-[31]). Thus, on the facts of the case, he found that the Bank was not able to rely on LAP in respect of any of the documents prepared by its employees for the purposes of instructing Freshfields.
  80. I do not find that analysis, or conclusion, easy; nor, it seems, did the constitution of this court in Eurasian, in which one of the questions for the court was: "What did Three Rivers (No 5) actually decide?" (see [124]-[130]). Having carefully considered the judgment in Three Rivers (No 5) and found that this was an essential part of Longmore LJ's reasoning (so that the court should not depart from it, even if on one view it could be argued that it was obiter: see [77]), Sir Geoffrey Vos C, giving the judgment of the court in Eurasian, concluded as follows (at [13]):
  81. "We can fully accept that the Court of Appeal could have decided Three Rivers (No 5) on the simple basis that Freshfields' client was the BIU (not the Bank), and the documents had been prepared by the Bank (not the BIU), so that the position of the particular Bank employee who had prepared them was irrelevant to the question of [LAP]. We do not, however, think that, fairly read, that was the Court of Appeal's reasoning. As we have explained, it seems to us that Longmore LJ reasoned that, because agents and employees, on authority, stood in the same position in relation to legal professional privilege, once it was established that only communications between the lawyer and the client, and not between the lawyer and an agent of the client, could attract [LAP], communications between a lawyer and an employee of the client (other than employees specifically tasked with seeking and receiving legal advice) could also not be privileged. As we have said, we are not sure that it is necessary for us to determine whether this reasoning was the ratio decidendi, but if that did have to be decided, we would hold that it was."
  82. The constitution of this court in Eurasian indicated that, if it had been open to it to depart from Three Rivers (No 5) on this issue, it would have done so. In addition to the preference for common law jurisdictions to be in step on such issues as this – Three Rivers (No 5) being out of step with overseas common law on this issue – the court considered that the foundation of Longmore LJ's judgment (i.e. an analysis of 19th century authorities) was unsafe, because those cases were decided at a time when the distinction between litigation privilege and LAP was "very much in its infancy", and without any of the principled analysis of LAP which has subsequently taken place (see paragraphs 38 and following above). Having quoted from Lord Scott's speech in Three Rivers (No 6) (including the passage from [34], quoted at paragraph 41 above), the Chancellor continued (at [127]):
  83. "This last passage makes clear that large corporations need, as much as small corporations and individuals, to seek and obtain legal advice without fear of intrusion. If legal advice privilege is confined to communications passing between the lawyer and the 'client' (in the sense of the instructing individual or those employees of a company authorised to seek and receive legal advice on its behalf), this presents no problem for individuals and many small businesses, since the information about the case will normally be obtained by the lawyer from the individual or board members of the small corporation. That was the position in most of the 19th century cases. In the modern world, however, we have to cater for legal advice sought by large national corporations and indeed multinational ones. In such cases, the information upon which legal advice is sought is unlikely to be in the hands of the main board or those it appoints to seek and receive legal advice. If a multi-national corporation cannot ask its lawyers to obtain the information it needs to advise that corporation from the corporation's employees with relevant first-hand knowledge under the protection of [LAP], that corporation will be in a less advantageous position than a smaller entity seeking such advice. In our view, at least, whatever the rule is, it should be equally applicable to all clients, whatever their size or reach. Moreover, it is not always an answer to say that the relevant subsidiary can seek the necessary legal advice and, therefore, ask its own lawyers to secure the necessary information with the protection of legal advice privilege. In a case such as the present, there may be issues between group companies that make it desirable for the parent company to be able to procure the information necessary to obtain its own legal advice."
  84. I respectfully agree. In addition:
  85. i) Three Rivers (No 5) does not appear to allow for any caveats to the proposition that material sent by a third party/agent/employee to a lawyer (and vice versa) is not covered by LAP. However, where lawyers are instructed, the individual within a corporation instructing them must be able to ensure that the instructions are in accordance with the wishes of the senior executives in the company, which may involve input from more junior employees who are knowledgeable about the relevant issues. Internal communications settling instructions must be covered by LAP. It is unclear to me how the proposition in Three Rivers (No 5) quite allows for that.

    ii) For no obvious reason, the law in relation to LAP as set out in Three Rivers (No 5) in respect of collection of information for the instruction of lawyers appears to be out of line with the law in respect of the dissemination of advice from lawyers, once received (i.e. Proposition 2, as described in paragraph 45 above).

  86. For those reasons, like the constitution of the court in Eurasian, on the basis of both principle and practical application, I respectfully doubt both the analysis and conclusion of this court in Three Rivers (No 5) on this issue; and, had it been in this court's power, I too would be disinclined to follow it.
  87. But, as I have indicated, we do not have that power. In Three Rivers (No 5), this court held that communications between an employee of a corporation and the corporation's lawyers does not attract LAP unless that particular employee was tasked with seeking and receiving such advice on behalf of the client; and, as confirmed in Three Rivers (No 6) at [47] per Lord Scott and Eurasian, that is binding on this court. As Hildyard J succinctly put it in The RBS Rights Issue Litigation [2016] EWHC 3161 (Ch); [2017] 1 WLR 1991:
  88. "… [T]here can be no real doubt as to the present state of the law in this context…: Three Rivers (No 5) confines legal advice privilege to communications between lawyer and client, and the fact that an employee may be authorised to communicate with the corporation's lawyer does not constitute that employee the client or a recognised emanation of the client."
  89. However, the facts of this case – and the issue to which they give rise – are significantly different from those in Three Rivers (No 5). In that case, the relevant lawyers (Freshfields) were external; and the issue was whether, in advising on the presentation of evidence to the Bingham Inquiry, they were involved qua lawyers; and therefore whether the communications concerning that advice could be in respect of "legal advice" so that they were covered by LAP. The relevant lawyers in this case were in-house; and Morris J found that the in-house lawyers were involved in the internal correspondence qua lawyers rather than merely as executives providing commercial advice (see paragraph 22 above). The relevant non-lawyers were all relatively senior executives. There appears to be no evidence suggesting that any of those involved in the relevant internal correspondence did not have the ability to seek legal advice from those lawyers, or that, for these purposes, they were not "an emanation of the client". It seems to me that, on the evidence, in this case each of the non-lawyers involved fell within the scope of "client" so far as the lawyers involved were concerned. Therefore, leaving aside for the moment the question of whether the purpose has to be "dominant", LAP will attach to any confidential communication between a lawyer and a non-lawyer in this case, made for the purpose of giving or obtaining legal advice. As I understood his submissions, Mr Béar did not suggest the contrary.
  90. Proposition 5

  91. The focus therefore turns to the scope of "legal advice" for these purposes, and the fifth proposition namely that, for LAP to apply, the communication must be made "in a legal context" (the first limb), but otherwise "legal advice" is widely defined (the second limb).
  92. As I have described, LAP applies to a confidential communication between a client and a lawyer for the purpose of giving and obtaining legal advice. Whilst this appeal focuses on that purpose, many of the authorities – even those upon which the parties rely as providing some assistance to the "dominant purpose" issue – primarily concern the relationship between lawyer and client, and the scope of "legal advice" in that context.
  93. In Balabel, this court considered the issue of whether LAP extends only to communications particularly seeking or conveying legal advice or to all that passes between lawyer and client on matters within the ordinary business of a lawyer. Taylor LJ, having reviewed the authorities (which, he accepted, were divergent in respect of the scope of the privilege), continued (at page 330D-331A):
  94. "Although originally confined to advice regarding litigation, the privilege was extended to non-litigious business. Nevertheless, despite that extension, the purpose and scope of the privilege is still to enable legal advice to be sought and given in confidence. In my judgment, therefore, the test is whether the communication or other document was made confidentially for the purposes of legal advice. Those purposes have to be construed broadly. Privilege obviously attaches to a document conveying legal advice from solicitor to client and to a specific request from the client for such advice. But it does not follow that all other communications between them lack privilege. In most solicitor and client relationships, especially where a transaction involves protracted dealings, advice may be required or appropriate on matters great or small at various stages. There will be a continuum of communication and meetings between the solicitor and client. The negotiations for a lease such as occurred in the present case are only one example. Where information is passed by the solicitor or client to the other as part of the continuum aimed at keeping both informed so that advice may be sought and given as required, privilege will attach. A letter from the client containing information may end with such words as 'please advise me what I should do'. But, even if it does not, there will usually be implied in the relationship an overall expectation that the solicitor will at each stage, whether asked specifically or not, tender appropriate advice. Moreover, legal advice is not confined to telling the client the law; it must include advice as to what should prudently and sensibly be done in the relevant legal context.
    It may be that applying this test to any series of communications might isolate occasional letters or notes which could not be said to enjoy privilege. But to be disclosable such documents must be not only privilege-free but also material and relevant. Usually a letter which does no more than acknowledge receipt of a document or suggest a date for a meeting will be irrelevant and so non-disclosable. In effect, therefore, the 'purpose of legal advice' test will result in most communications between solicitor and client in, for example, a conveyancing transaction being exempt from disclosure, either because they are privileged or because they are immaterial or irrelevant."
  95. Similarly, in Three Rivers (No 6) (at [111]), Lord Carswell, having quoted from Balabel, said that it did not disturb or modify the principle affirmed in Minter v Priest [1930] AC 558 (in which it was held that conversations between a solicitor and client relating to the business of obtaining a loan for the deposit for the purchase of land were privileged, as the business was professional business within the ordinary scope of a solicitor's employment), namely:
  96. "… [A]ll communications between a solicitor and his client relating to a transaction in which the solicitor has been instructed for the purpose of obtaining legal advice will be privileged, notwithstanding that they do not contain advice on matters of law or construction, provided that they are directly related to the performance by the solicitor of his professional duty as legal adviser of his client."
  97. Those cases, therefore, confirmed both limbs of Proposition 5.
  98. In respect of the first limb, I can be brief, because Morris J found that the lawyers engaged in the internal correspondence here were indeed involved qua lawyers, and so communications made to them were generally made in an appropriate "legal context". However, there was some discussion before us about whether the dominant purpose test applied to the legal context, in the sense of whether the dominant purpose of the role or retainer of (or instructions to) the relevant lawyer had to be for giving or seeking legal advice. In The Sagheera (at page 168), Rix J suggested that in practice it might, so that, if the dominant purpose of the retainer is the obtaining and giving of legal advice, "although it is in theory possible that individual documents may fall outside that purpose, in practice it is unlikely", whereas, if the dominant purpose of the retainer is some business purpose, the documents will not be privileged unless specific legal advice is given and/or sought. Property Alliance Group Limited v Royal Bank of Scotland plc [2015] EWHC 3187 (Ch); [2016] 1 WLR 992, to which we were referred, is an example of a case in which the scope of the retainer of (in that case, external) lawyers, coupled with a properly broad (rather than nit-picking) approach to the continuum of communications, resulted in all of the memoranda for and minutes of meetings prepared by the lawyers in the context of a regulatory investigation being covered by LAP.
  99. In this context, it is worthy of note that all of their Lordships in Three Rivers (No 6) emphasised the need for a "relevant legal context" (see, e.g., Lord Scott at [38]) but all (except, perhaps, Lord Carswell: see, e.g., [84]) declined to import a dominant purpose test for legal context or the lawyer's brief, although Sir Sydney Kentridge QC (for the Law Society, which intervened) had expressly argued for such (see page 632E-G). For example, Lord Rodger referred to a requirement that the lawyer is instructed "qua lawyer" (at [58]); or whether, in the circumstances of that particular case, Freshfields were being asked "to put on legal spectacles when reading, considering and commenting on the drafts" of material for presentation to the Bingham Inquiry (at [60]).
  100. In his submissions, in considering whether documents are the subject of LAP, Mr Grodzinski emphasised the importance of the purpose of the relevant lawyer's retainer and role. I accept that, despite some suggestions to the contrary in previous cases, the general instructions to and role of the relevant lawyer make a good starting point, particularly given the court's acceptance that the breadth of "legal advice" in this context includes a lawyer giving advice with the benefit of his skills as a lawyer or "through a lawyer's eyes". However, it is not determinative of the different question as to whether a specific document passing between them is subject to LAP. It is clear from the authorities that, even if the dominant purpose of a lawyer's retainer (or his "dominant role") is related to that of giving legal advice, that is not conclusive on the question of whether LAP applies to a particular communication between lawyer and client (United States of America v Philip Morris Inc [2003] EWCA Civ 330; [2004] 1 CLC 881 at [79] per Brooke LJ with whom Scott Baker and Chadwick LJJ agreed, on appeal from Moore Bick J referred to in paragraph 23(i)(c) above). Therefore:
  101. "If part of a solicitor's duties embrace the giving of legal advice on his client's rights, liabilities and obligations, and a further significant part of his duties relate to activities which cannot be so characterised, then the two recent decisions of this court in the Three Rivers cases [i.e. Three Rivers (No 5) and Three Rivers (No 6) in the Court of Appeal [[2004] EWCA Civ 218; [2004] QB 916]: it had not yet then reached the House of Lords] show that it is too simplistic to refer to a dominant purpose of the original retainer, or to try and identify the dominant purpose of a role assumed over a number of years, involving the solicitor in many different activities." (ibid).

    Three Rivers (No 5) also said that authority did not justify a shift of focus from the dominant purpose for which a document is prepared to the "dominant purpose of the retainer" (see [28]). I will return to what Three Rivers (No 5) and Three Rivers (No 6) did say about "dominant purpose" shortly (see paragraphs 84 and following below).

  102. However, as to the second limb of Proposition 5, it is now well-established and uncontroversial that LAP covers more than just communications between lawyers and clients with regard to what the law is. Balabel indicated that the scope of LAP in this regard is wide, in two ways. First, it includes advice on the application of the law or, as Taylor LJ put it, "legal advice is not confined to telling the client the law; it must include advice as to what should prudently and sensibly be done in the relevant legal context". That was endorsed in Three Rivers (No 6) (see [62] per Lord Rodger, expressing the agreed view of the Judicial Committee). Second, once a legal context is established, LAP applies, not just to those communications which expressly seek or give legal advice, but also to the "continuum of communications" between a lawyer and client aimed at "keeping both informed so that advice may be sought and given as required". That again was endorsed in Three Rivers (No 6) (see, e.g., [111] per Lord Carswell, quoted at paragraph 63 above, with whom Lord Rodger, Baroness Hale and Lord Brown agreed). In arguing that dominant purpose does not apply to LAP, Mr Pievsky for the Law Society relied on the fact that Balabel did not expressly require a document to have such a purpose for LAP to apply: but that does not seem to me to add force to his argument. Balabel primarily concerned the scope of "legal advice". Whether or not that test applied to LAP was not at issue in that case.
  103. Therefore, summarising the position as indicated by the authorities (and still leaving aside for the time being the issue of whether the relevant purpose has to be "dominant"):
  104. i) Consideration of LAP has to be undertaken on the basis of particular documents, and not simply the brief or role of the relevant lawyer.

    ii) However, where that brief or role is qua lawyer, because "legal advice" includes advice on the application of the law and the consideration of particular circumstances from a legal point of view, and a broad approach is also taken to "continuum of communications", most communications to and from the client are likely to be sent in a legal context and are likely to be privileged. Nevertheless, a particular communication may not be so – it may step outside the usual brief or role.

    iii) Similarly, where the usual brief or role is not qua lawyer but (e.g.) as a commercial person, a particular document may still fall within the scope of LAP if it is specifically in a legal context and therefore, again, falls outside the usual brief or role.

    iv) In considering whether a document is covered by LAP, the breadth of the concepts of legal advice and continuum of communications must be taken into account.

    v) Although of course the context will be important, the court is unlikely to be persuaded by fine arguments as to whether a particular document or communication does fall outside legal advice, particularly as the legal and non-legal might be so intermingled that distinguishing the two and severance are for practical purposes impossible and it can be properly said that the dominant purpose of the document as a whole is giving or seeking legal advice.

    vi) Where there is no such intermingling, and the legal and non-legal can be identified, then the document or communication can be severed: the parts covered by LAP will be non-disclosable (and redactable), and the rest will be disclosable (see, e.g., Curlex Manufacturing Pty Limited v Carlingford Australia General Insurance Limited [1987] Qd R 335 and GE Capital Corporate Finance Group Limited v Bankers Trust Company [1995] 1 WLR 172).

    vii) A communication to a lawyer may be covered by the privilege even if express legal advice is not sought: it is open to a client to keep his lawyer acquainted with the circumstances of a matter on the basis that the lawyer will provide legal advice as and when he considers it appropriate.

    Ground 1: Legal Advice Privilege and Purpose

  105. That leads me to the first question which has to be considered in this appeal, at the heart of Grounds 1 and 2, namely does a claim for LAP require the proponent to show that the relevant document or communication was created or sent for the dominant purpose of obtaining legal advice?
  106. Although it has been said that litigation privilege and LAP are "integral parts of the same privilege" (Re L (A Minor) (Police Investigation: Privilege) [1997] AC 16 at page 33E per Lord Nicholls of Birkenhead), it is now generally accepted that they are two distinct limbs of legal professional privilege with different characteristics (see, e.g., Waugh v British Railways Board [1980] AC 521 at page 541G-H per Lord Edmund-Davies, Three Rivers (No 6) at [103] per Lord Carswell, and Eurasian at [63]). One obvious difference is that established by Wheeler v Le Marchant (see paragraphs 1 and 46 above): whilst LAP is restricted to communications between lawyer and client (and vice versa), litigation privilege can attach to communications with other individuals. The categories are therefore distinct; but, of course, overlapping; as communications between solicitor and client concerning litigation may fall into both categories.
  107. Despite their differences, each limb of the privilege is defined in part by reference to the purpose of the relevant communication: in respect of litigation privilege, the purpose of pursuing or defending actual or proposed litigation and, in respect of LAP, the purpose of giving or seeking legal advice.
  108. There consequently arose an issue as to whether, in circumstances in which a communication had more than one purpose, it was sufficient for the privilege to attach for one purpose to be for litigation or for obtaining/seeking legal advice, or whether that had to be the appreciable, dominant or even sole purpose.
  109. The issue first arose in Australia. In Grant v Downs (1976) CLR 674, a widow sued the New South Wales Government after her husband died in a psychiatric hospital. The widow considered the hospital to blame. In accordance with standard practice, the relevant government department obtained reports on the death. When the widow sought discovery of the reports, litigation privilege was claimed. It was said that the reports were prepared for a number of purposes: to assist in determining whether there had been a breach of staff discipline, to detect whether there were any faults in the hospital's systems and procedures, and to enable the department to obtain legal advice as to its possible liability and to obtain legal representation in the event of any coronial or civil proceedings.
  110. A majority of the High Court of Australia adopted a sole purpose test for both limbs of the privilege; but, in a dissenting judgment, Barwick CJ preferred a dominant purpose test. He said (at page 677):
  111. "Having considered the decisions, the writings and the various aspects of the public interest which claim attention, I have come to the conclusion that the court should state the relevant principle as follows: a document which was produced or brought into existence either with the dominant purpose of its author, or of the person or authority under whose direction, whether particular or general, it was produced or brought into existence, of using it or its contents in order to obtain legal advice or to conduct or aid in the conduct of litigation, at the time of its production in reasonable prospect, should be privileged and excluded from inspection" (emphasis added).

    That test was ultimately adopted in Australia, for both limbs of the privilege, in Esso Australia Resources Limited v Commissioner of Taxation [1999] HCA 67; 201 CLR 49 ("East Australia Resources"). In Australia, it is now well-established that the dominant purpose test applies to LAP (see, e.g., AWB Limited v Cole [2006] FCA 571).

  112. The dominant purpose issue first arose in this jurisdiction in Waugh, the facts of which did not substantially differ from those of Grant v Downs. A railway employee died in a collision between locomotives.  His widow sued the railway board.  There was an internal inquiry into the accident, resulting in a report which was prepared for two purposes which were (the court held) of equal rank and weight:  to assist the board to decide whether there was a need to revise safety and operational procedures, and to obtain legal advice in anticipation of litigation. Discovery of the report was sought in the claim. The House of Lords, overruling previous authority to the effect that the privilege applied where one substantial purpose for bringing the document into existence was that it should be available for legal advice in anticipation of litigation and approving the passage of the judgment of Barwick CJ in Grant v Downs quoted above, held that the report was disclosable because its dominant purpose was not to obtain legal advice for litigation.
  113. Waugh was a case which proceeded on the basis that the effective limb of legal professional privilege, if any, was litigation privilege; and it is uncontroversial that it established that, for litigation privilege to be claimed, the relevant communication must have had the dominant purpose of pursuing or defending actual or contemplated litigation. The question raised by Ground 1 is whether, in this jurisdiction, the same principle applies to LAP: must the relevant communication have had the dominant purpose of giving or obtaining legal advice?
  114. Mr Grodzinski for the CAA, and Mr Béar for Jet2, agreed that there was no authority directly on point; but each contended, as his primary submission, that the preponderance of authority was in his favour.
  115. Mr Grodzinski's "core submission" was that the application of the dominant purpose test to LAP is inconsistent with Three Rivers (No 6). He submitted that it was telling that, in authoritatively defining the test for LAP, Lord Carswell in Three Rivers (No 6) at [111] did not refer to "dominant purpose" as being an element of that test (nor, indeed, did Taylor LJ in Balabel). Other authorities which suggest that it is, such as The Sagheera, Philip Morris and Three Rivers (No 5) (although Mr Grodzinski submitted that this last judgment, when read as a whole, did not in any event support the proposition that the dominant purpose test applies to LAP) were decided before Three Rivers (No 6) and therefore have to be read in the light of the authoritative definition set out in that case. In any event, he said, the issue was recently considered in Eurasian, in which this court squarely addressed the issue and firmly concluded (at [132]) that the dominant purpose test did not apply to LAP. Because of its findings in relation to other issues (see paragraph 55 and following above), that was formally obiter; but nevertheless, he submitted, the issue had been fully argued before the court, and it was obiter of substantial persuasive authority. This court should follow it.
  116. Mr Béar submitted, to the contrary, that The Sagheera, Three Rivers (No 5), Three Rivers (No 6) and Philip Morris all held or accepted that the dominant purpose test was applicable. That was in general line with both overseas authority, and the leading text books. The observations in Eurasian were clearly obiter, and were made in the entirely different context of a case in which the relevant lawyers were external solicitors retained to provide legal advice, so that a dominant purpose test was not pertinent in practice. Furthermore, the court failed to grapple with the earlier authorities. Eurasian could not dislodge the preponderance of authorities to the effect that dominant purpose is required for LAP to apply.
  117. Below, Morris J held that the balance of authority was in favour of the dominant purpose test; and, whilst Eurasian expressed a contrary view, it was obiter and was insufficient to undermine the preponderant view of the earlier authorities. I agree.
  118. In respect of authorities, the generally held view is that Waugh did not purport to say anything about the position with regard to LAP (see, e.g., Hollander at paragraph 17-16, Privilege by Colin Passmore (3rd edition) ("Passmore") at paragraph 2-099 and The Law of Privilege edited by Bankim Thanki QC (3rd edition) ("Thanki") at paragraph 2.177). Indeed, neither party before us submits that Waugh is directly applicable to the case before us, or that any indication in that case could be more than obiter in respect of LAP, the case being determined on the basis of litigation privilege criteria.
  119. However:
  120. i) In Waugh, on the basis of the defendant's contentions, both litigation privilege and LAP applied, the issue being whether the report had been prepared for the purpose of procuring legal advice in respect of anticipated litigation.

    ii) In Waugh, the majority of their Lordships approved the relevant passage from Barwick CJ's judgment in Grant v Downs: see Lord Wilberforce at pages 532-3, Lord Simon of Glaisdale at pages 534 and 537, and Lord Edmund-Davies at pages 543-4.

    iii) In respect of the relevant purpose test, Grant v Downs did not draw any distinction between litigation privilege and LAP. In my view, that is clear from the words used by Barwick CJ; but that view has been confirmed in The Sagheera at page 167, Three Rivers (No 5) (Comm Ct) at [24] and [26]-[27], and Slade LJ in the leading judgment in Guinness Peat Properties Limited v Fitzroy Robinson Partnership [1987] 1 WLR 1027 at pages 1043F-1035A. On one reading of the judgment of Longmore LJ in Three Rivers (No 5) at [24]-[25], he considered that that interpretation was incorrect, and, when properly read, Barwick CJ was restricting himself to a principle that applied to litigation privilege only; but:

    a) in my respectful view, the words used by Barwick CJ were unambiguous; on their face, they were intended to apply to both limbs of legal professional privilege;
    b) if they were intended to apply to both limbs of legal professional privilege, I do not see how that is in any conflict with Wheeler v Le Marchant which concerned a different aspect of LAP (i.e. whether it applied to communications that did not concern legal advice);
    c) the interpretation of Barwick CJ's words adopted in Three Rivers (No 5) fails to take into account the judgment of the majority (at page 682), which applied the sole purpose test to both limbs of the privilege;
    d) as I have described, it is not the interpretation put on those words by other courts in this jurisdiction; and
    e) importantly, in the face of those observations in Three Rivers (No 5), the Federal Court of Australia has expressly denied that that interpretation is correct, stating that the principle formulated by Barwick CJ "was intended to be declaratory of the law for the future; was stated compendiously; and had equal application to both manifestations of the privilege" (Pratt Holdings Pty Limited v Commissioner of Taxation [2004] FCAFC 122 at [16] per Finn J).

    iv) Whilst I accept that Waugh was decided on the basis that litigation privilege applied, none of their Lordships suggested that there was a distinction between litigation privilege and LAP in this regard. Indeed, in applying Grant v Downs, if anything, it seems to me that overall they suggested that there was no distinction (see, e.g., at page 532H per Lord Wilberforce, and at page 543B per Lord Edmund-Davies). In my view, there is force in the observation of Tomlinson J in Three Rivers (No 5) (Comm Ct) at [26], where he said:

    "I do not regard it as plausible to suggest that their Lordships did not appreciate that the test which they were approving was stated in terms which apparently embraced legal professional privilege in both its manifestations, [LAP] and litigation privilege. It would I think be surprising if their Lordships uncritically adopted a passage of such apparent generality if they thought that a different approach was called for in relation to [LAP] to which the passage apparently, or at any rate on one obvious reading, made reference."
  121. Turning to the Three Rivers cases, in respect of the appeal from Tomlinson J (i.e. Three Rivers (No 5) in the Court of Appeal) on this issue, again, I do not find the judgment as clear as it might have been. Whether the dominant purpose test applies to LAP was in issue before Tomlinson J; and the Court of Appeal judgment appears to record that Tomlinson J held that it did (see [7]). However, by the time the application came before the Court of Appeal, it seems that that issue was not further contested (see the record of argument, especially at page 1559F-G, where the Bank appears to accept that the dominant purpose test does apply), the core issue being, as I have described, whether that test applied to documents emanating from employees of the Bank. But the judgment at least assumed that that is the test. Notably, there are references throughout the judgment to "dominant purpose" without adverse comment; and there is a lengthy passage (at [32]-[37]) on the issue of whether, had the resolution of the earlier issues required it, "the internal documentation of the Bank, which came into existence after the setting up of the Bingham Inquiry, was indeed prepared with the dominant purpose of obtaining [legal] advice". Mr Grodzinski could not explain why the court embarked on such an exercise if, as a matter of principle, the dominant purpose test could not even contingently be applicable.
  122. Mr Pievsky forcefully submitted that the House of Lords in Three Rivers (No 6) did not refer to "dominant purpose" in the context of LAP, suggesting that their Lordships did not consider that it applied. However, the Judicial Committee did not expressly consider the dominant purpose point in that case: their Lordships did not have to deal with it in the context of that appeal, and clearly declined so to do. However, the Court of Appeal (the constitution of which included both the Master of the Rolls and Longmore LJ, who had also sat on Three Rivers (No 5)) again at least appears to have assumed that the dominant purpose test applied. The issue was not whether that test applied, but whether it had been satisfied (see, e.g., [6]). In their Lordships' speeches on appeal, there are numerous references to the test, again without adverse comment. Indeed, Lord Carswell (at [70]) appears clearly to support the proposition that the test applied, expressly referring to Tomlinson J's reasoning in regard to the test in Three Rivers (No 5) (Comm Ct) and saying that he considered it had "considerable force".
  123. In Philip Morris (Comm Ct), Moore-Bick J seems to have considered (at [38]) that Three Rivers (No 5) held that it was a condition of LAP applying that the relevant communication "must be for the dominant purpose of obtaining or giving legal advice…"; which the Court of Appeal confirmed without adverse comment (at [77]).
  124. In each of these authorities, therefore, it seems to have been at least accepted by the court without any adverse comment that the dominant purpose test applies to LAP. Insofar as the Court of Appeal in Three Rivers (No 5) and the House of Lords in Three Rivers (No 6) accepted the test applied, then acceptance (even if made on the basis of a concession by the proponent of LAP) was made by the highest courts in circumstances in which, in my view, it is inconceivable that an acceptance or assumption would have been made without adverse comment unless it was considered that the test was applicable. On that issue, as I have indicated, I agree with the sentiments of Tomlinson J in Three Rivers (No 5) (Comm Ct) (quoted at paragraph 83(iv) above).
  125. The only authority to suggest that the dominant purpose test does not apply is Eurasian. Because the court considered it was bound by Three Rivers (No 5) to proceed on the basis that communications between an employee of a corporation and the corporation's lawyers could not attract LAP unless that employee was tasked with seeking and receiving such advice on behalf of the client (which was determinative in the Eurasian case), it did not have to determine the issue of dominant purpose. However, it said:
  126. "131. The SFO submitted that it should in any event be held that, if information passed to a company's lawyers by employees who were not authorised to seek and receive legal advice could be the subject of legal advice privilege, a further qualification should be added, namely that the information must be shown to have been obtained for the dominant purpose of obtaining legal advice. This, submitted Mr James Segan for the SFO, was established by a line of cases including, for example, The Sagheera... at page 168, Three Rivers (No 5) [(Comm Ct)]… at [20], [21], [26] and [30], and Philip Morris [in the Court of Appeal] at [43] and [77].
    132. In the light of the approach that we have adopted thus far to legal advice privilege, it would not be appropriate to reach any final conclusion on this submission. In our judgment, however, it is hard to see why the suggested additional qualification is necessary, when the privilege can, by definition, only be claimed when legal advice is being sought or given. It is one thing to say that litigation privilege can only be claimed where the communication is created for the dominant purpose of the litigation, but entirely another to say that legal advice privilege can only be claimed where the communication is created for the dominant purpose of seeking legal advice. The second is tautologous."
  127. However:
  128. i) The observations made by the court were clearly obiter: indeed, the court expressly indicated that it was inappropriate to reach a conclusion on the issue in that case.

    ii) The court did not consider any of the authorities on the issue, domestic or overseas. Particularly given the preponderance of authority in favour of the dominant purpose test applying, I consider that this substantially detracts from the persuasive weight of these observations.

    iii) The court considered that the qualification that the purpose had to be "dominant" was unnecessary and "tautologous" because LAP can "only be claimed when legal advice is being sought or given". However, litigation privilege can only apply when litigation is pending or in reasonable contemplation and, in terms of rationale, there appears to me to be no relevant difference in principle between the two limbs of the privilege. Eurasian concerned external lawyers, whose role was to provide legal advice; and the court does not appear to have considered the possibility of a document being created, as here, partly for the purpose of obtaining legal advice but partly for some other reason such as obtaining non-legal advice. This possibly reflects other cases which appear to fail to take into account the possibility that a communication to a lawyer to obtain legal advice may be part of a multi-addressee email which also seeks advice/input from non-lawyers, which is the case before us.

  129. In the circumstances, I do not consider that Eurasian significantly undermines the authorities which either hold or accept that the dominant purpose test applies to LAP.
  130. We were also referred to three legal textbooks, with which I should briefly deal.
  131. i) Passmore (at paragraph 2-100) accepts that the dominant purpose test as an element of LAP "received a measure of approval" in Three Rivers (No 5); but submits that "the approach of the House of Lords to the identification of legal advice [in Three Rivers (No 6)] is such that the dominant purpose test has no obvious role to play in determining claims for [LAP] under the law of England and Wales, which is now concerned only with establishing a 'relevant legal context' in which legal advice is sought or given". I agree that Three Rivers (No 5) gives support to the inclusion of the test; but, for the reasons I have given, I do not agree that the requirement for the establishment of a "relevant legal context" avoids the need for consideration of the purpose element.

    ii) Thanki suggests "tentatively" that the House of Lords in Three Rivers (No 6) did not consider that dominant purpose had any application to "pure" lawyer-client communications (paragraph 2.177); but accepts that, even if that be the case, "where a communication is not purely between lawyer and client but between lawyer and client and another party or parties (for example a group of professional advisers) [i.e. this case], the dominant purpose test must apply" (paragraph 2.184).

    iii) Hollander submits that there is no suggestion in Three Rivers (No 6) that dominant purpose has any role to play in LAP (paragraph 17-16); but, in respect of documents sent to multiple addressees, submits that these should be treated as separate communications as between the various participants (rendering the principle of dominant purpose in this context redundant). For the reasons I have given, it seems to me that, in most cases, this approach will, if applied correctly, result in the same result as applying a dominant purpose test; and, in any event, on that basis, the relevant emails passing between non-lawyers will be disclosable.

  132. I accept that this court may not be formally bound by the authorities which are considered above; but each party contested the appeal on the basis that these authorities, whilst not binding, were important, if not crucial, in giving guidance. Insofar as we are not formally bound, I consider the authorities of considerable persuasive weight, and we should follow them unless there is good reason not to do so.
  133. Mr Grodzinski's submissions on this ground were heavily reliant upon the authorities which, he contended, were adverse to dominant purpose being an element of LAP. However, in support of his submissions, he relied upon three other matters.
  134. i) He stressed the importance of the privilege to the rule of law: he submitted that, if the dominant purpose test applies, then some communications between client and lawyer which have as a purpose the giving or obtaining of legal advice will be disclosable, undermining the whole purpose of the privilege. However, legal professional privilege is an important but, as I have indicated (paragraph 39 above), not an absolute principle. As I have indicated, where the material over which privilege is claimed is crucial to an issue in proceedings, it potentially undermines the fairness of a trial. The common law is not bound to acknowledge a right to withhold evidence that would otherwise be disclosable simply because the relevant material has, as simply one, minor purpose, the obtaining of legal advice, without consideration of the respective weight of purpose. It is entitled to balance the public interest in these respective principles, and draw a line between them. Otherwise, as Mr Béar submitted, swathes of internal and external material could be excluded from disclosure simply because a lawyer had been copied in and asked for his legal advice as and when he considered it appropriate to give it.

    ii) Mr Grodzinski submitted that the logic of privilege in a communication being lost by reason of it being copied to non-lawyers is contrary to the principle that a privileged document in the form of legal advice may be circulated internally (and even, in some circumstances, to third parties) without that privilege being lost. I accept that that is the case; but, as I have described, that is the result of Three Rivers (No 5) taking the view it did take over the collection of material for the purposes of briefing lawyers (see paragraphs 47 and following above).

    iii) Mr Grodzinski accepts that LAP requires some control mechanism; but submits that it is found in the "definition" of LAP found in Lord Carswell's speech in Three Rivers (No 6) (see paragraph 63 above), namely in the requirement that the lawyer is sent the communication for the purposes of giving legal advice. He submitted that the inclusion of a dominant purpose criterion into LAP is unnecessary, and would mean that requests for advice and input made simultaneously to lawyers and other, non-lawyer executives would effectively result in the loss of LAP in the communications with the lawyers. Properly, he submitted, all such communications should be regarded as part of the "continuum of communication" between client and lawyers, and so be covered by LAP. However, for the reasons I set out under Ground 2, I do not consider this point to have any real force. In particular, it is to be noted that, in Three Rivers (No 6) at [111], Lord Carswell was especially addressing the issue of the ambit of "legal advice"; and, as Lord Carswell himself observed (at [110]), in construing judicial pronouncements, context is vital. In particular, he clearly did not have in mind the possibility of (e.g.) simultaneous communications by a single email to lawyers and non-lawyers. I was unimpressed by the submission that the result I favour will make life difficult for those who wish to obtain legal and non-legal advice simultaneously, because (a) there is no indication that that causes a problem in other Commonwealth countries, such as Australia and Singapore, which have adopted the dominant purpose test for LAP, and (b) LAP is a privilege, and those who wish to take advantage of it should be expected to take proper care when they do so.

  135. For those reasons, I do not consider that there is any good ground for not following the preponderance of authority which supports the inclusion of a dominant purpose criterion into LAP.
  136. Further, in my view there are good grounds for including such a criterion.
  137. i) Although they do have some different characteristics, litigation privilege and LAP are limbs of the same privilege, legal professional privilege. It is uncontroversial that the dominant purpose test, grown out of Grant v Downs, applies to litigation privilege. For the reasons I have given, I am unpersuaded that Eurasian is correct to consider the limbs as fundamentally different with regard to purpose. In my view, there is no compelling rationale for differentiating between limbs of the privilege in this context. The "dominant purpose" test in litigation privilege fixed by Waugh derives from Australian jurisprudence, which has since Grant v Downs treated the purpose test (whatever it might be) as applying to both limbs of the privilege.

    ii) Whilst I accept that the position is not uniform, generally the common law in other jurisdictions has incorporated a dominant purpose test in both limbs of legal professional privilege, e.g., in addition to Australia above (considered above: see paragraphs 74-75), Singapore (Skandinaviska Enskilda Banken AB v Asia Pacific Breweries [2007] 2 SLR 367) and Hong Kong (Citic Pacific Limited v Secretary of Justice [2016] 1 HKC 157 at [51]-[62]). This not only suggests that such a test is able to work in practice; but this is a legal area in which there is advantage in the common law adopting similar principles.

  138. For those reasons, whilst I readily accept that the jurisprudence is far from straightforward and the authorities do not speak with a single, clear voice, I consider Morris J was correct to proceed on the basis that, for LAP to apply to a particular communication or document, the proponent of the privilege must show that the dominant purpose of that communication or document was to obtain or give legal advice. I would dismiss the appeal on Ground 1.
  139. Ground 2: Legal Advice Privilege and Multi-addressee Communications

  140. Most of the documents which are the subject of the disclosure application giving rise to this appeal are emails sent to a number of addressees, one or more of whom were in-house lawyers (who, Morris J found, were acting qua lawyers) but one or more of whom were non-lawyers.
  141. Of these emails in the December 2018 judgment, he held:
  142. i) The dominant purpose criterion applied; so that, if the dominant purpose of the email was to obtain legal advice from an in-house lawyer, then it would be privileged, even if it also at the same time sought the commercial views of others. However, if its dominant purpose was to seek commercial views, then the email would not be privileged even if it was contemporaneously sent to a lawyer for the purpose of giving legal advice (see [95(5)] and [102]).

    ii) However, even if the dominant purpose is not in respect of obtaining legal advice, it may still be privileged if it "discloses or is likely to disclose the nature and content of the legal advice sought and obtained" (see [102]), or if it "might disclose" such advice (see paragraph [101]).

  143. As I have found in respect of Ground 1, the dominant purpose criterion does apply to LAP. Of most of the emails in respect of which disclosure is sought, the CAA accept that it cannot be said that (i) the dominant purpose was to seek legal advice, or (ii) the email/attachment might realistically disclose the nature of the legal advice being sought from, or given by, the in-house lawyer (see paragraph 30 above). To a large extent, Ground 1 would appear effectively to resolve the issues in Ground 2.
  144. In respect of these emails, I generally agree with the approach applied by Morris J in paragraphs 99-102 of the December 2018 judgment. In my view, the following is the appropriate approach to multi-addressee emails such as those of which Jet2 seek disclosure in this case.
  145. i) As I have indicated, the dominant purpose test applies to LAP. As I have indicated (paragraph 67 above), although the general role of the relevant lawyer may be a useful starting point (and may, in many cases, in practice be determinative), the test focuses on documents and other communications and has to be applied to each such.

    ii) In respect of a single, multi-addressee email sent simultaneously to various individuals for their advice/comments, including a lawyer for his input, the purpose(s) of the communication need to be identified. In this exercise, the wide scope of "legal advice" (including the giving of advice in a commercial context through a lawyer's eyes) and the concept of "continuum of communications" must be taken fully into account. If the dominant purpose of the communication is, in substance, to settle the instructions to the lawyer then, subject to the principle set out in Three Rivers (No 5) (see paragraphs 47 and following above), that communication will be covered by LAP. That will be so even if that communication is sent to the lawyer himself or herself, by way of information; or if it is part of a rolling series of communications with the dominant purpose of instructing the lawyer. However, if the dominant purpose is to obtain the commercial views of the non-lawyer addressees, then it will not be privileged, even if a subsidiary purpose is simultaneously to obtain legal advice from the lawyer addressee(s).

    iii) The response from the lawyer, if it contains legal advice, will almost certainly be privileged, even if it is copied to more than one addressee. Again, whilst the dominant purpose test applies, given the wide scope of "legal advice" and "continuum of communications", the court will be extremely reluctant to engage in the exercise of determining whether, in respect of a specific document or communication, the dominant purpose was the provision of legal (rather than non-legal) advice. It is difficult to conceive of many circumstances in which such an exercise could be other than arid and unnecessary.

    iv) There was some debate before us – as there is in the textbooks (e.g. in Hollander (see paragraph 91(iii) above)) – as to whether multi-addressee communications should be considered as separate bilateral communications between the sender and each recipient, or whether they should be considered as a whole. My preferred view is that they should be considered as separate communications between the sender and each recipient. LAP essentially attaches to communications. Where the purpose of the sender is simultaneously to obtain from various individuals both legal advice and non-legal advice/input, it is difficult to see why the form of the request (in a single, multi-addressee email on the one hand, or in separate emails on the other) in itself should be relevant as to whether the communications to the non-lawyers should be privileged. That is not to say, of course, that the form may not in some cases reveal the true purpose of the communication, e.g. it may appear from the form of the email that the dominant purpose of the email is to settle the instructions to the lawyer who has merely been copied in by way of information, or to the contrary that the dominant purpose of sending the email to the non-lawyers is to obtain their substantive (non-lawyer) input in any event.

    v) In my view, there is some benefit in taking the approach advocated by Hollander (at paragraph 17-17), namely to consider whether, if the email were sent to the lawyer alone, it would have been privileged. If no, then the question of whether any of the other emails are privileged hardly arises. If yes, then the question arises as to whether any of the emails to the non-lawyers are privileged, because (e.g.) its dominant purpose is to obtain instructions or disseminate legal advice.

    vi) However, whether considered as a single communication or separate communications to each recipient, and whilst there may perhaps be "hard cases", I doubt whether in many cases there will be any difference in consequence, if the correct approach to LAP is maintained. Where there is a multi-addressee email seeking both legal advice and non-legal (e.g. commercial) advice or input, if regarded as separate communications, those to and from the lawyer will be privileged: otherwise, they will not be privileged, unless the real (dominant) purpose of a specific email to/from non-lawyers is that of instructing the lawyer. If it is not for that purpose, in most cases, the email as a whole will clearly not have the dominant purpose of obtaining legal advice.

    vii) I agree with Morris J, that, where a communication might realistically disclose legal advice (in the sense of there being a realistic possibility of it disclosing such advice), then that communication will in any event be privileged (see paragraph 27 above). However, in respect of the relevant documents in this case, on the basis of that test, as I understand it, Ms Brooks appears to have considered that none would or might disclose such advice (see paragraph 28 above).

    viii) Mr Grodzinski suggested that this approach would cause difficulties in terms of meetings (including records of meetings), attended by non-lawyers and lawyers, at which commercial matters were discussed with the lawyer adding legal advice and input if and when required. The whole of what transpires at such a meeting, he submitted, should be the subject of LAP. However, I disagree; and consider the same principles set out above as applying to documents and other communications are applicable. Legal advice requested and given at such a meeting would, of course, be privileged; but the mere presence of a lawyer, perhaps only on the off-chance that his or her legal input might be required, is insufficient to render the whole meeting the subject of LAP so that none of its contents (including any notes, minutes or record of the meeting) are disclosable. If the dominant purpose of the meeting is to obtain legal advice (or, subject to the principle set out in Three Rivers (No 5) (see paragraphs 47 and following above), to settle instructions to a lawyer), unless anything is said outside that legal context, the contents of the meeting will be privileged. If the dominant purpose of the discussions is commercial or otherwise non-legal, then the meeting and its contents will not generally be privileged; although any legal advice sought or given within the meeting may be. It is likely that, where not inextricably intermingled, the non-privileged part will be severable (and, on disclosure, redactable) (see paragraph 69 above).

  146. As I have described (see paragraph 29 above), on considering whether each document satisfied the criteria (i.e. whether (i) the dominant purpose was to seek legal advice, or (ii) the email/attachment might realistically disclose the nature of the legal advice being sought from, or given by, the in-house lawyer), Ms Brooks identified two matters which, she considered, disenabled her from determining whether the document was or was not privileged. Neither of these went to dominant purpose. However, she considered that, in determining whether a document disclosed (or might disclose) the nature of legal advice, sought or given, it was unclear from the December 2018 judgment whether (i) it is necessary for legal advice to be specifically requested, and (ii) the particular document should be looked at discretely or in the context of the communications which preceded and followed it. In respect of (i), it is not necessary for advice to have been specifically requested: as has been clear since Balabel, it is sufficient for a legal adviser to have been sent material upon which to advise if he or she considered advice necessary or appropriate. As to (ii), in considering whether any document might disclose legal advice, of course it has to be considered in context: only in the context of communications which preceded and followed it can it be determined whether a document might disclose legal advice.
  147. Finally, to deal with a discrete point, before Morris J it was contended on behalf of the CAA that the 24 January 2018 email was covered by privilege. In my view, the judge was right to conclude that it was not: although the email was sent to three people including a lawyer (Ms Lim), it was clearly not its dominant purpose to obtain legal (as opposed to non-legal commercial) advice.
  148. For those reasons, Ground 2 fails.
  149. Ground 3: Separate Consideration of Emails and Attachments

  150. The effect of the December 2018 judgment and the Order which followed it was to require the CAA to assess whether each email and each attachment, considered separately, was privileged in accordance with the judgment.
  151. Mr Grodzinski submitted that that approach was wrong, because LAP protects communications (or records of communications) and an attachment to an email cannot be regarded as a separate communication from the email itself. The difference in approach may have some practical consequences. He posed the example of a senior executive sending an email to multiple recipients, including a lawyer, attaching a draft document prepared by the lawyer; and, in the covering email, suggesting some amendments and asking for the views of his (non-lawyer) colleagues. It was submitted that, looked at as a whole, both email and attachment would be privileged as they would reveal the nature of the legal advice given; but, looked at separately, the attachment would be privileged, but the email would not.
  152. This ground did not feature heavily before us. I do not consider that it raises any point of principle, nor does it appear to bear upon whether any of the relevant documents in this case are privileged or not. I can deal with it shortly.
  153. It is well-established that a document which is not privileged does not become so simply because it is sent to lawyers, even as part of a request for legal advice (Ventouris v Mountain [1991] 1 WLR 607 at page 616F, and Imerman v Tchenguiz [2009] EWHC 2901 at [14]). In giving disclosure, some separate consideration of substantive documents and attachments therefore has to be undertaken. Whilst an email and attachment can be regarded as a single communication, separate consideration will need to be given to the attachment, given that it will have been received or created by the sender, and therefore may require discrete consideration. I do not find Mr Grodzinski's example of assistance. In that case, it is likely that the email, considered separately, will be privileged as inevitably disclosing legal advice received (in the form of the draft letter).
  154. I do not consider that Morris J erred in requiring the separate consideration of emails and attachments for the purposes of identifying documents that are covered by LAP and those which are not. I would refuse the appeal on this ground.
  155. Ground 4: Waiver

  156. In the February 2019 judgment, Morris J held that if, contrary to his primary conclusion that they were not, the relevant documents were privileged, then privilege in all the documents was waived by the CAA by its disclosure of the 24 January 2018 email (referred to in paragraph 12 above). Mr Grodzinski submitted the judge was wrong to do so, because (i) he wrongly identified the relevant "transaction" in respect of which privilege had been waived, and (ii) he was wrong to conclude that, the 24 January 2018 email having been disclosed, fairness required the disclosure of all the drafts of the 1 February 2018 letter and internal correspondence about those drafts.
  157. As I have concluded that the relevant documents are not privileged, this ground has become academic. I will deal with it shortly.
  158. The relevant principles are uncontroversial. Although the voluntary disclosure of a privileged document may result in the waiver of privilege in other material, it does not necessarily have the result that privilege is waived in all documents of the same category or all documents relating to all issues which the disclosed document touches. However, voluntary disclosure cannot be made in such a partial or selective manner that unfairness or misunderstanding may result (Paragon Finance plc v Freshfields [1999] 1 WLR 1183 at page 1188D per Lord Bingham CJ).
  159. Collateral waiver of privilege allows for documents and other material that would otherwise be non-disclosable on public interest grounds, to be required to be disclosed even though the individual with the right to withhold disclosure objects. The courts have therefore imposed certain constraints on collateral waiver.
  160. The starting point is to ascertain "the issue in relation to which the [voluntarily disclosed material] has been deployed", known as the "transaction test" (General Accident Fire and Life Assurance Corporation Limited v Tanter [1984] 1 WLR 100 at 113D per Hobhouse J), waiver being limited to documents relating to that "transaction" subject to the overriding requirement for fairness. The "transaction" is not the same as the subject matter of the disclosed document or communication, and waiver does not apply to all documents which could be described as "relevant" to the issue, in the usual, Peruvian Guano sense of the term as used in disclosure (Compagnie Financière et Commerciale du Pacifique v Peruvian Guano Company (1882) 11 QBD 35).
  161. In Fulham Leisure Holdings Limited v Nicholson Graham & Jones [2006] EWHC 158 (Ch); [2006] 2 All ER 599 at [18], having reviewed the relevant authorities, Mann J described the approach thus:
  162. "18. What those citations show is that it is necessary to bear in mind two concepts. First of all, there is the actual transaction or act in respect of which disclosure is made. In order to identify the transaction, one has to look first at what it is in essence that the waiving party is seeking to disclose. It may be apparent from that alone that what is to be disclosed is obviously a single and complete 'transaction' – for example, the advice given by a lawyer on a given occasion…. [O]ne is in my view entitled to look to see the purpose for which the material is disclosed, or the point in the action to which it is said to go…. Mr Croxford [Counsel for the claimant, which sought to rely on LAP] submitted that the purpose of the disclosure played no part in a determination of how far the waiver went. I do not agree with that; in some cases it may provide a realistic, objectively determinable definition of the 'transaction' in question. Once the transaction has been identified, then those cases show that the whole of the material relevant to that transaction must be disclosed. In my view it is not open to a waiving party to say that the transaction is simply what that party has chosen to disclose (again contrary to the substance of a submission made by Mr Croxford). The court will determine objectively what the real transaction is so that the scope of the waiver can be determined. If only part of the material involved in that transaction has been disclosed then further disclosure will be ordered and it can no longer be resisted on the basis of privilege.
    19. Once the transaction has been identified and proper disclosure made of that, then the additional principles of fairness may come into play if it is apparent from the disclosure that has been made that it is in fact part of some bigger picture (not necessarily part of some bigger 'transaction') and fairness, and the need not to mislead, requires further disclosure. The application of this principle will be very fact sensitive, and will therefore vary very much from case to case…."

    The purpose of the voluntary disclosure, which has prompted the contention that privilege in other material has been collaterally waived, is therefore an important consideration in the assessment of what constitutes the relevant "transaction" (see also Dore v Leicestershire County Council [2010] EWHC 34 (Ch) at [18]-[19] also per Mann J).

  163. In this case, the 24 January 2018 email was voluntarily disclosed by the CAA as an attachment to Mr Moriarty's statement served in support of its defence to the claim (see paragraphs 19 above). Mr Moriarty explained (at paragraph 27 of his statement) that the purpose of that disclosure was to show that the language used by Mr Haines in his email of 18 January 2018 was "not reflective of any part of the approach taken by the CAA". Mr Moriarty said that:
  164. "This can be seen from an email sent to Mr Haines by one of my colleagues, Mr Buffey [attached as an exhibit],… and from the content of the material that we published."
  165. Morris J accepted that that was the purpose of disclosure (February 2019 judgment at [21]). However, he concluded that the relevant "transaction" extended to all emails and internal discussions in the period between the 16 January 2018 letter and the publication of the Daily Mail article on 6/7 February 2018, i.e. all of the documents of which disclosure was sought under categories (e) and (f).
  166. Mr Béar submits that the judge was entitled to draw that conclusion. He referred to the relevant authorities, and the correct principles (as briefly described above); and he was entitled to make (as he did), (i) the factual finding that the approach of the CAA to Jet2 at the time was "a single process of internal discussion, which does not have discrete parts", and (ii) the assessment that fairness required the disclosure of the whole chain of correspondence, without which disclosure of the 24 January 2018 email alone would be misleading. Mr Béar referred us to the unreported judgment of this court in Tanap Investments (UK) Limited v Tozer, in which judgment was given on 11 October 1991, which (rightly) emphasised that this court will only interfere with an exercise of discretion by the court below where the judge misapplied the principles or was plainly wrong.
  167. Mr Grodzinski submitted that the judge erred in law in proceeding on the basis that the relevant "transaction" extended to everything that was relevant to the response to the 24 January 2018 email (notably the drafts of the 1 February 2018 letter), including all drafts and all email discussions of the drafts. Having regard to the purpose of the voluntary disclosure (which did not involve disclosing any legal advice, or anything to do with the request or provision of legal advice), the relevant transaction should have been limited to the 24 January 2018 email itself; nor did fairness require any wider disclosure.
  168. On this issue, I prefer the submissions of Mr Grodzinski. In my view, the purpose and nature of the voluntary disclosure are crucial; and, in this case, I consider Morris J unfortunately failed properly to take these matters into account. As the judge accepted, the disclosure was not in respect of any legal advice; and so it could not be said that there was any risk of the email presenting a partial or selective disclosure of legal advice and thus there was no risk of unfairness that might have been caused by such partial disclosure. The purpose of the email was modest: it was intended to show that (in Mr Moriarty's words) the language used by Mr Haines in his email of 18 January 2018 was "not reflective of any part of the approach taken by the CAA"; or, perhaps more accurately, that not all of the executives at CAA shared the approach suggested by Mr Haines' earlier email. It cannot be right that such a modest voluntary disclosure could result in the collateral waiver (and thus the forced disclosure by the CAA) in respect of all the internal communications relating to the drafting of the 1 February 2018 letter, including those that expressly reveal legal advice from the CAA's lawyers; nor is that what the law (or fairness) requires.
  169. In my view, the relevant transaction so far as the voluntary disclosure is concerned is restricted to the 24 January 2018 email itself. Fairness does not require more.
  170. For those reasons, had I concluded that the relevant documents are privileged, I would have held that that privilege was not waived in respect of any of them by the voluntary disclosure of the 24 January 2018 email.
  171. Conclusion

  172. For those reasons, I find none of the grounds made good; and, subject to my Lords Patten and Peter Jackson LJJ, I would dismiss the appeal.
  173. Lord Justice Peter Jackson :

  174. I agree.
  175. Lord Justice Patten :

  176. I also agree.

BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII