Esso Australia Resources Ltd v Federal Commissioner of Taxation

HIGH COURT OF AUSTRALIA

GLEESON CJ,
GAUDRON, McHUGH, GUMMOW, KIRBY AND CALLINAN JJ

ESSO AUSTRALIA RESOURCES LIMITED  APPELLANT

AND

THE COMMISSIONER OF TAXATION
OF THE COMMONWEALTH OF AUSTRALIA    RESPONDENT

Esso Australia Resources Limited v The Commissioner of Taxation
[1999] HCA 67
21 December 1999
M53/1999

ORDER

1. Appeal allowed with costs.

2. Set aside the orders of the Full Court of the Federal Court of Australia made on 22 December 1998, and in place thereof, order that:

i)the appeal to that Court be allowed with costs; and

ii)the respondents pay the costs of the proceedings before Foster J.

3. Questions of law raised for decision for Foster J answered:

a) The correct test is the dominant purpose test, which is the common law test for claiming legal professional privilege.

b)Does not arise.

On appeal from the Federal Court of Australia

Representation:

B J Shaw QC with J W de Wijn QC for the appellant (instructed by Clayton Utz)

2.

D M J Bennett QC, Solicitor-General for the Commonwealth with C M Maxwell QC and M M Gordon for the respondent (instructed by Australian Government Solicitor)

Notice:  This copy of the Court’s Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports.

CATCHWORDS

Esso Australia Resources Limited v The Commissioner of Taxation

Practice and procedure – Discovery – Legal professional privilege – Whether test for privilege is "dominant purpose" test or "sole purpose" test.

Evidence – Legal professional privilege – Client legal privilege – Test for – Whether "dominant purpose or sole purpose".

Statutes – Influence on common law – Evidence Act 1995 – Applicable in three Australian jurisdictions – Whether enactments affect development of common law – Significance of non-enactment in other jurisdictions – Relevance (if any) of Act to application to review past authority of High Court.

Words and phrases – "client legal privilege".

Evidence Act 1995 (Cth), ss 118, 119.
Federal Court Rules, O 15 r 15.

1. GLEESON CJ, GAUDRON AND GUMMOW JJ.   The central issue in this appeal concerns the test for determining whether what is usually called legal professional privilege, (an expression which might suggest erroneously that the privilege is that of the lawyer), or what is called in the Evidence Act 1995 (Cth) client legal privilege[1], applies in relation to discovery and inspection of confidential written communications between lawyer and client.

   [1]The Evidence Act also confers a similar privilege on unrepresented litigants:  s 120.

2. The test in the Evidence Act is whether the communication was made, or the document was prepared, for the dominant purpose of the lawyer providing legal advice or legal services[2].  The dominant purpose test accords with the common law test now adopted in England[3], New Zealand[4], Ireland[5], and most Canadian Provinces[6].  It was favoured by Barwick CJ in the leading Australian case on the subject, Grant v Downs[7].  However, a majority in that case (Stephen, Mason and Murphy JJ) preferred a sole purpose test.  Hence, since 1976, courts in Australia have applied the common law of legal professional privilege on the basis that privilege will only attach to a confidential communication, oral or in writing, made for the sole purpose of obtaining or giving legal advice or assistance or of use in legal proceedings.

   [2]Evidence Act, ss 118, 119.

   [3]Waugh v British Railways Board [1980] AC 521.

   [4]Guardian Royal Exchange Assurance of New Zealand Ltd v Stuart [1985] 1 NZLR 596.

   [5]Silver Hill Duckling Ltd v Minister for Agriculture [1987] IR 289.

   [6]Levin v Boyce [1985] 4 WWR 702; Milton Farms Ltd v Dow Chemical Canada Inc (1986) 13 CPC (2d) 174; Doiron v Embree (1987) 16 CPC (2d) 70; Ed Miller Sales & Rentals Ltd v Caterpillar Tractor Co (No 1) (1988) 22 CPR (3d) 290.

   [7](1976) 135 CLR 674.

3. The difference between the Evidence Act test and what has, since Grant v Downs, been accepted in Australia as the common law test, has given rise to a number of problems.  The Evidence Act only applies in proceedings in a federal court or an Australian Capital Territory court.  New South Wales has enacted legislation in the same terms for that State, but no other jurisdiction has done so.  Moreover, even in a jurisdiction where the Evidence Act applies, the relevant provisions relate only to the adducing of evidence.  The reason for this was explained in the report of the Australian Law Reform Commission which proposed the legislation, and which said[8]:  "The Terms of Reference limit the Commission to considering the application of the privilege in the courtroom where evidence is sought to be given."

   [8]Australian Law Reform Commission, Evidence, Report No 38, (1987), par 199.

4. As was pointed out in Mann v Carnell[9], the circumstances in which legal professional privilege may apply are not limited to the adducing of evidence.  As in the present case, the privilege may be invoked in other circumstances, such as discovery and inspection of documents.  Documents may be discoverable, or the subject of a demand for inspection, even though they are not admissible in evidence.  They may be significant, for example, because they open up a line of inquiry.  Furthermore, in this country, the application of the privilege is not confined to judicial or quasi-judicial proceedings[10].  On any view, the ambit of the common law doctrine of legal professional privilege in Australia exceeds that of the relevant provisions of the Evidence Act. Given the specific and limited heads of legislative power in s 51 of the Constitution, there also may be questions as to the extent of the legislative power of the Parliament to deal with the privilege, apart from its operation in relation to judicial proceedings under Ch III.

   [9][1999] HCA 66.

   [10]Baker v Campbell (1983) 153 CLR 52.

5. The Australian Law Reform Commission was aware of the problem, and adverted to it in its report.  After referring to the limitation in its terms of reference, the Commission said[11]:

   "Situations may arise where a party obtains access to documents outside the courtroom which are protected in the courtroom by the proposed privilege.  Under the proposal, the privilege will still apply in the courtroom unless the client voluntarily disclosed the document.  Having wider access on discovery or under a search warrant is usual.  Access is not determined by the rules of admissibility such as relevance and hearsay.  It is not unreasonable to have wider access in the investigative stage."

   [11]Australian Law Reform Commission, Evidence, Report No 38, (1987), par 199.

6. Some judges have disagreed with the Commission's views as to what was not unreasonable.  They have seen the co-existence of two different tests for privilege as anomalous.  A deal of ingenuity has been devoted to developing arguments which could overcome the lack of congruence between the statute and the common law.  In this Court, however, the appellant has relied principally upon an argument that was not available in any of the other courts that have so far considered the question.  The appellant has invited the Court to reconsider Grant v Downs, and to declare that the dominant purpose test now represents the common law of Australia.  In that event, the common law of this country would be in line with that of England, Canada, New Zealand and Ireland, and with the legislation which now applies in federal courts and in New South Wales and the Australian Capital Territory.  The inconsistencies resulting from the limited reach of the statutory provisions would not be eliminated, because there are other respects in which the statute differs from the common law, but they would be reduced in one significant respect.

   The present litigation

7. In 1996 the appellant commenced proceedings in the Federal Court of Australia, appealing against amended assessments of income tax for the years ended 31 December 1987 to 31 December 1992.  General orders for discovery were made in October 1996.  The rules and practice of the Federal Court as to discovery and inspection of documents, and as to the making of directions with respect to the conduct of the proceedings, were summarised by Burchett J in Murex Diagnostics Australia Pty Ltd v Chiron Corporation[12].  The provisions of Orders 10 and 15 of the Federal Court Rules are of particular relevance.  In June 1997, the appellant filed and served a list of documents verified by affidavit.  Privilege was claimed in respect of 577 documents.  Disagreement concerning the claims for privilege arose.  After an exchange of correspondence, the area of disagreement narrowed.  It was accepted that, in relation to some of the documents, which were said to have been brought into existence for the sole purpose of giving or receiving legal advice, there was no dispute.  However, it appeared that, in relation to many of the documents in question, the appellant's contention was that "their disclosure would result in disclosure of a confidential communication made between [the appellant] and a lawyer for the dominant purpose of the lawyer … providing legal advice to [the appellant]".  The descriptions of the purposes of the communications varied slightly, but in all cases where the claim was disputed it was based upon an assertion of dominant purpose.  In October 1997 the respondent filed Notices of Motion seeking orders that the appellant produce for inspection the documents in respect of which the claim for privilege was disputed.

   [12](1995) 55 FCR 194 at 198-199.

8. The applications came for hearing before Foster J.  Pursuant to O 29 r 2, Foster J ordered that there be decided separately the following questions of law:

   "(a)     Whether the correct test for claiming legal professional privilege in relation to the production of discovered documents is the 'sole purpose' test as formulated by the High Court in Grant v Downs (1976) 135 CLR 674 or the 'dominant purpose' test as set out in s 118 and s 119 of the Evidence Act 1995 (Cth)?

   (b)      Whether the Court has power pursuant to Order 15 rule 15 of the Federal Court Rules to make an order excluding from production discovered documents on the basis that such documents meet the 'dominant purpose' test as set out in s 118 and s 119 of the Evidence Act 1995 (Cth)?"

9. Foster J answered the questions as follows:

   "In respect of question (a), [declare] that the correct test for claiming legal professional privilege in relation to the production of discovered documents is the 'sole purpose' test as formulated by the High Court in Grant v Downs. 

   In respect of question (b), [declare] that the question be answered in the negative."

10. The appellant appealed to the Full Court of the Federal Court.  Because judges of the Federal Court, and of other courts, had expressed differing opinions on the point, and because it appeared that there would be an invitation to reconsider the earlier decision of the Full Court of the Federal Court in Adelaide Steamship Co Ltd v Spalvins[13], which had been followed by the Court of Appeal of New South Wales in Akins v Abigroup Ltd[14], a specially constituted court of five members sat.  By majority (Black CJ, Sundberg and Finkelstein JJ, Beaumont and Merkel JJ dissenting) the Full Court substantially upheld the decision of Foster J[15].  The decisions in Adelaide Steamship Co Ltd v Spalvins and Akins v Abigroup Ltd were not followed.  The Full Court agreed with the answer given by Foster J to the first question.  The answer given by Foster J to the second question was varied to read:

    "Yes, but to exclude from production discovered documents for the sole reason that they meet the 'dominant purpose' test in ss 118 and 119 would not be a proper exercise of the power."

    [13](1998) 81 FCR 360.

    [14](1998) 43 NSWLR 539.

    [15](1998) 83 FCR 511.

11. The appellant now appeals to this Court.  As was noted, a submission is advanced which was not available in the Federal Court, where the arguments and reasons for judgment were constrained by the authority of Grant v Downs.  The appellant argues that this Court should declare, contrary to what was said by three members of the Court in Grant v Downs, but in accordance with what was said by Barwick CJ in that case, that at common law in Australia the dominant purpose test applies.  If that argument were accepted, the first of the questions asked by Foster J would be answered by declaring that the correct test is that of dominant purpose, and it would be unnecessary to answer the second question.  Alternatively, the appellant repeats the arguments which it advanced in the courts below and which, if accepted, would lead to a similar result.

12. It is convenient first to consider the arguments dealt with in the courts below, and then to come to the appellant's new argument.

    The arguments considered by the Full Court

13. In the Full Court, the appellant relied upon three alternative methods by which, it was contended, one could arrive at the conclusion that, notwithstanding Grant v Downs, in a case such as the present a claim for privilege in relation to discovery and inspection of documents was to be resolved by the application of the dominant purpose test.  First, it was argued that, upon its true construction, notwithstanding the apparently restricted terms in which it was expressed, the Evidence Act establishes such a test as applicable to discovery and inspection.   Secondly, the appellant invited the Full Court to follow a line of reasoning which had prevailed in some earlier cases[16], to the effect that, even if the provisions of the Evidence Act did not directly apply to claims for privilege made in relation to discovery and inspection of documents (and in other circumstances not involving the adducing of evidence), the common law, by analogy or derivation, should be treated as modified to accord with the statutory test "at least in the jurisdictions where the Act applies". (As will appear, the qualification reflects a significant problem in the argument.) Thirdly, the appellant relied upon the discretionary power, in O 15 r 15, as to ordering, or refusing to order, production of discovered documents as a basis upon which courts could, and should, make the test on discovery and production conform to that to be applied in adducing evidence.

    [16]eg Telstra Corporation v Australis Media Holdings [No 1] (1997) 41 NSWLR 277, Adelaide Steamship Co Ltd v Spalvins (1998) 81 FCR 360 and Akins v Abigroup Ltd (1998) 43 NSWLR 539.

14. In the Full Court of the Federal Court, the majority rejected all three arguments.  They were correct to do so.

15. The provisions of the Evidence Act which deal with the subject of client legal privilege are ss 118 and 119.  They are in the following terms:

    "118.  Evidence is not to be adduced if, on objection by a client, the court finds that adducing the evidence would result in disclosure of:

    (a)a confidential communication made between the client and a lawyer; or

    (b)a confidential communication made between 2 or more lawyers acting for the client; or

    (c)the contents of a confidential document (whether delivered or not) prepared by the client or a lawyer;

    for the dominant purpose of the lawyer, or one or more of the lawyers, providing legal advice to the client.

    119.  Evidence is not to be adduced if, on objection by a client, the court finds that adducing the evidence would result in disclosure of:

    (a)a confidential communication between the client and another person, or between a lawyer acting for the client and another person, that was made; or

    (b)the contents of a confidential document (whether delivered or not) that was prepared;

    for the dominant purpose of the client being provided with professional legal services relating to an Australian or overseas proceeding (including the proceeding before the court), or an anticipated or pending Australian or overseas proceeding, in which the client is or may be, or was or might have been, a party."

16. Those sections appear in Ch 3 (ss 55-139) of the Act which, save for presently immaterial exceptions, is concerned with the admissibility of evidence. This limitation in the language of the statutory provisions was noted, and was essential to the decision of this Court, in Northern Territory v GPAO[17]. The explanation of the wording of the Act is set out above. The statutory language is clear. It deals with the adducing of evidence. That would cover adducing evidence in interlocutory proceedings as well as at a final hearing, or on an appeal, but it does not cover all the circumstances in which a claim for privilege might arise. To take the most obvious example, it would not cover the circumstances considered in Baker v Campbell[18].

    [17](1999) 73 ALJR 470; 161 ALR 318.

    [18](1983) 153 CLR 52.

17. The claim in contention in the present case was not a claim that certain evidence could, or could not, be adduced.  It was a claim that the appellant was not obliged to make certain written communications available for inspection by the respondent.  Sections 118 and 119 are expressed in language which does not address that claim.  The refusal of the majority in the Full Court to apply the statutory provisions otherwise than in accordance with their terms was right in principle, and was consistent with the decision of this Court in Northern Territory v GPAO.

18. The second argument was not at the forefront of the appellant's oral submissions in this Court and reliance is placed upon the written submissions.  The interrelation and interaction between common law and statute may trigger varied and complex questions requiring full argument in cases where they arise.

19. Significant elements of what now is regarded as "common law" had their origin in statute or as glosses on statute or as responses to statute.  For example, in Peters v The Queen[19], McHugh J explained the derivation of the criminal law of conspiracy from statutes enacted in the thirteenth century.  The doctrine of part performance is expressed in three centuries of case law which has the effect of allowing specific performance of a contract which on its face the Statute of Frauds renders unenforceable.  The Statute of Limitations in its terms does not operate directly upon equitable remedies, but, as Dixon J put it in Cohen v Cohen[20], "such remedies are barred in Courts of equity by analogy to the statute".  On the other hand, the courts did not refuse to enforce rights arising under a contract or trust merely because the trust or contract is associated with or in furtherance of a purpose rendered illegal by a statute which applied to the relevant parties[21].

    [19](1998) 192 CLR 493 at 513-515.

    [20](1929) 42 CLR 91 at 100.

    [21]Nelson v Nelson (1995) 184 CLR 538; Fitzgerald v F J Leonhardt Pty Ltd (1997) 189 CLR 215.

20. In the Chorley Lecture 1984, titled "Common Law and Statute Law"[22], Professor Atiyah referred to such matters and continued[23]:

    "If, as I have suggested, statute law and common law do, at least in many areas, work together in some kind of legal partnership, creating sometimes amalgams of law of various kinds, [is it] possible for the courts to take account of statute law, in the very development of the common law itself?  Can the courts, for instance, use statutes as analogies for the purpose of developing the common law?  Can they justify jettisoning obsolete cases, not because they have been actually reversed by some statutory provision, but because a statute suggests that they are based on outdated values?  Could the courts legitimately draw some general principle from a limited statutory provision, and apply that principle as a matter of common law?  It must be clear that using statutes in this way is fundamentally different from any process of construction, however benevolent or liberal that might be.  Construction, as a matter of theory at least, requires the court to give effect to what it thinks the legislation actually enacts.  Using statutes by way of analogy quite clearly involves using them to produce results which the legislation does not enact."

    [22](1985) 48 Modern Law Review 1.

    [23](1985) 48 Modern Law Review 1 at 6.

21. It is in the area identified in this passage that the appellant's second argument lies.  Its origin is to be found in the judgment of McLelland CJ in Eq in Telstra Corporation v Australis Media Holdings [No 1][24].  That was a case concerning a claim to inspect documents produced on subpoena.  The documents were said to be the subject of legal professional privilege, and a question arose as to the test to be applied to determine that contention.  McLelland CJ in Eq noted the difference between the statutory test and the common law test.  He also said that, as a matter of statutory construction, ss 118 and 119 of the Evidence Act did not apply to ancillary process, and that no question of adducing evidence arose.  However, his Honour regarded it as anomalous and "verging on the absurd"[25] that different tests should apply to a claim for privilege made in an ancillary process and a claim made at the stage of adducing evidence.  Clearly, he did not agree with the Australian Law Reform Commission's view that this was not unreasonable.  He pointed out that, historically, the doctrine of legal professional privilege was established in the context of testimonial compulsion, although, now, in Australia, it operated in a wider context.  He said[26]:

    "In this sense the principles of legal professional privilege applicable to testimony at a trial provide the paradigm, and the extension of the same principles to ancillary processes was derivative in nature.  Accordingly, any change to the paradigm should rationally be reflected in the derivatives."

    [24](1997) 41 NSWLR 277.

    [25](1997) 41 NSWLR 277 at 279.

    [26](1997) 41 NSWLR 277 at 279.

22. This approach was taken up by the Full Court of the Federal Court in Adelaide Steamship Co Ltd v Spalvins[27], which, in turn, was followed by the Court of Appeal of New South Wales in Akins v Abigroup Ltd[28].  The Full Court of the Federal Court, in Adelaide Steamship, approving what McLelland CJ in Eq had said, referred also[29] to the principles concerning the analogical use of statutes in developing common law principles and concluded that the Evidence Act had "created an entirely new setting to which the common law must now adapt itself"[30].

    [27](1998) 81 FCR 360.

    [28](1998) 43 NSWLR 539. But not, however, by the Court of Criminal Appeal of New South Wales in R v Young [1999] NSWCCA 166.

    [29](1998) 81 FCR 360 at 373.

    [30](1998) 81 FCR 360 at 373.

23. As was pointed out by the Full Court of the Federal Court in the present case, there is a fundamental difficulty with this line of reasoning.  The legislation in question does not apply throughout Australia.  At present, it applies only in federal courts, and in the courts of New South Wales and the Australian Capital Territory.  In Lange v Australian Broadcasting Corporation[31] this Court said that "[t]here is but one common law in Australia which is declared by this Court as the final court of appeal".  Certain legislatures in Australia have enacted legislation concerning privilege which differs in a number of respects from the common law principles.  One respect concerns whether the test to be applied for determining privilege is the sole purpose or the dominant purpose test.  There are other differences, which are not material to the present case, but which should not be overlooked.  As was observed in Mann v Carnell[32], for example, the rules relating to loss of privilege are also different in some respects.  Other legislatures have not enacted similar legislation.  Furthermore, the legislation, even in the jurisdictions where it applies, in its terms leaves untouched certain areas in which the privilege may operate.  In such a setting, there is no consistent pattern of legislative policy to which the common law in Australia can adapt itself.  The fragmentation of the common law implicit in the qualification that such adaptation should occur only in those jurisdictions in which the Evidence Act applies is inconsistent with what was said in Lange, and is unacceptable.

    [31](1997) 189 CLR 520 at 563.

    [32][1999] HCA 66.

24. In Warnink v J Townend & Sons (Hull) Ltd[33] Lord Diplock said:

    "Where over a period of years there can be discerned a steady trend in legislation which reflects the view of successive Parliaments as to what the public interest demands in a particular field of law, development of the common law in that part of the same field which has been left to it ought to proceed upon a parallel rather than a diverging course."

    Subsequently, in Photo Production Ltd v Securicor Transport Ltd[34], Lord Wilberforce, in supporting the discarding by the House of Lords of the "fundamental breach doctrine", referred to the legislative intention manifested in the Unfair Contract Terms Act 1977 (UK) that, consumer contracts apart, the parties to commercial contracts be free to apportion the risks as they see fit.

    [33][1979] AC 731 at 743.

    [34][1980] AC 827 at 843.

25. Their Lordships were speaking in the context of a nation with a single Parliament.  What has occurred in Australia in relation to the legislation here in question cannot be said to reflect a consistent legislative view of what the public interest demands in relation to the law of legal professional privilege.  Most Australian legislatures have not adopted the Evidence Act, and those which have adopted it have limited its application to part only of the field in which the privilege operates.  In Moorgate Tobacco Co Ltd v Philip Morris Ltd [No 2][35], Deane J spoke against a background of federal law, namely the provisions of Pts IV and V of the Trade Practices Act 1974 (Cth), when rejecting the existence of a common law action for "unfair competition" or "unfair trading". His Honour, speaking for the Court, said[36]:

    "Those limits, which define the boundary between the area of legal or equitable restraint and protection and the area of untrammelled competition, increasingly reflect what the responsible Parliament or Parliaments have determined to be the appropriate balance between competing claims and policies."

    Subsequently, in R v L[37], the Court rejected the proposition that it was part of the common law of Australia that, by marriage, a wife gave irrevocable consent to sexual intercourse with her husband.  Mason CJ, Deane and Toohey JJ referred to the uniform pattern of legislation in five States and added that the notion of irrevocable consent was out of keeping[38] "with recent changes in the criminal law of this country made by statute, which draw no distinction between a wife and other women in defining the offence of rape".

    [35](1984) 156 CLR 414. See also Perre v Apand Pty Ltd (1999) 73 ALJR 1190 at 1223-1224; 164 ALR 606 at 654-655.

    [36](1984) 156 CLR 414 at 445.

    [37](1991) 174 CLR 379.

    [38](1991) 174 CLR 379 at 390.

26. In Moragne v States Marine Lines Inc[39], the question before the United States Supreme Court was whether a decision of 1886, which held that the principles of maritime law which applied in the federal courts did not afford a cause of action for wrongful death, should no longer be regarded as acceptable in the light of what by 1970 was the enactment in every State of the Union of a wrongful-death statute.  The Supreme Court decided that, although no State legislation applied to the case in hand, the federal rule should adapt by analogy to the position established in the various States.

    [39]398 US 375 (1970).

27. The situation in these authorities, where the analogy is drawn from federal statute law, or from a consistent pattern of State legislation, markedly differs from the situation presented here.

28. In South Pacific Manufacturing Co Ltd v New Zealand Security Consultants & Investigations Ltd, Cooke P, speaking in a unitary system, observed that "[t]he analogy of a statute may properly influence the development of the common law"[40].  However, whatever may be involved in this doctrine of analogy, as to which it is unnecessary now to venture any further opinion, the situation in the present litigation cannot provide an occasion for its application.

    [40][1992] 2 NZLR 282 at 298.

29. The members of the Full Court of the Federal Court were in substantial agreement upon the third argument relied upon by the appellant.

30. Order 15 r 15 of the Federal Court Rules provides:

    "The Court shall not make an order under this Order for the filing or service of any list of documents or affidavit or other document or for the production of any document unless satisfied that the order is necessary at the time when the order is made."

31. The necessity referred to is necessity for the fair disposition of the case[41].

    [41]Burmah Oil Co Ltd v Bank of England [1980] AC 1090 at 1141-1142.

32. Discovery and inspection of documents may be onerous and oppressive, and unless kept within the bounds of necessity may add greatly to the expense and delay involved in litigation.  The rule confers a salutary discretionary power, but its purpose is not to enable the court to subvert, or circumvent, the rules which determine the existence of privilege.

33. In a particular case, the circumstance that a document is one which could not be tendered in evidence because, for example, it was covered by client legal privilege under the Evidence Act, might possibly be a consideration relevant to a decision as to whether it was necessary that it be produced for inspection even though at the stage of discovery it was not covered by privilege.  However, as Finkelstein J pointed out[42], in that respect a document produced for a dominant purpose related to legal advice or assistance may be in a position no different from any other document.  What was held in the Full Court, correctly, was that the question is one of necessity, and that this is to be determined in the light of the facts and circumstances of the individual case.

    [42](1998) 83 FCR 511 at 568.

34. The Full Court was right to reject the arguments advanced by the appellant in that court.  It is necessary now to turn to the appellant's invitation to reconsider Grant v Downs.

    The common law of legal professional privilege

35. Legal professional privilege (or client legal privilege) protects the confidentiality of certain communications made in connection with giving or obtaining legal advice or the provision of legal services, including representation in proceedings in a court.  In the ordinary course of events, citizens engage in many confidential communications, including communications with professional advisers, which are not protected from compulsory disclosure.  The rationale of the privilege has been explained in a number of cases, including Baker v Campbell[43], and Grant v Downs itself.  The privilege exists to serve the public interest in the administration of justice by encouraging full and frank disclosure by clients to their lawyers.  In Waterford v The Commonwealth[44], Mason and Wilson JJ explained that legal professional privilege is itself the product of a balancing exercise between competing public interests and that, given the application of the privilege, no further balancing exercise is required.  As Deane J expressed it in Baker v Campbell[45], a person should be entitled to seek and obtain legal advice in the conduct of his or her affairs, and legal assistance in and for the purposes of the conduct of actual or anticipated litigation, without the apprehension of being prejudiced by subsequent disclosure of the communication.  The obvious tension between this policy and the desirability, in the interests of justice, of obtaining the fullest possible access to the facts relevant to the issues in a case lies at the heart of the problem of the scope of the privilege.  Where the privilege applies, it inhibits or prevents access to potentially relevant information.  The party denied access might be an opposing litigant, a prosecutor, an accused in a criminal trial, or an investigating authority.  For the law, in the interests of the administration of justice, to deny access to relevant information, involves a balancing of competing considerations.  This Court is now asked to reconsider the balance that was struck in Grant v Downs.

    [43](1983) 153 CLR 52.

    [44](1987) 163 CLR 54 at 64-65. See also Carter v Northmore Hale Davy & Leake (1995) 183 CLR 121 at 128 per Brennan J, 134 per Deane J, 147 per Toohey J, 163 per McHugh J.

    [45](1983) 153 CLR 52 at 114.

36. Like the present case, and the leading English case of Waugh v British Railways Board[46], Grant v Downs was about discovery and inspection of documents in pending litigation.  Although privilege, where it applies, attaches to communications, not to pieces of paper, discovery is concerned with documents, and privileged communications are frequently in writing.  If a written communication is made for the sole purpose of seeking or giving legal advice, or obtaining or providing legal services, the problem of present concern does not arise.  It arises where the documentary communication comes into existence for some purpose or purposes in addition to the legal purpose.

    [46][1980] AC 521.

37. As the facts of the cases illustrate, this is not an unusual situation.  In Grant v Downs, the inmate of a public psychiatric hospital died in circumstances which gave rise to an action by his widow against the New South Wales Government for damages under the Compensation to Relatives Act 1897 (NSW). In accordance with standard Departmental practice, reports had been made about the occurrence. Upon discovery it was claimed that the reports were privileged. They were said to have been 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 case of any coronial or civil proceedings. Such a multiplicity of purposes is commonplace, especially in large corporations or bureaucracies, which will often have their own internal legal staff, who are amongst those to whom such reports will be directed. In Waugh, an employee of a railway board was killed in a collision between locomotives.  His widow sued the board.  There was an internal inquiry into the accident, resulting in a report.  The report was prepared for two purposes:  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.

38. In both cases, the claims for privilege were disallowed.  In neither case was the obtaining of legal advice or assistance the dominant, let alone the sole, purpose of bringing the documents into existence.  It may be added that the conditions of s 118 and s 119 of the Evidence Act would not have been satisfied in either case.

39. At the time Grant v Downs was decided at first instance, the law, both in Australia and England, as to the test to be applied in such cases had not been determined by any court of ultimate authority, but the prevailing view was that it was sufficient to attract privilege to such reports if one purpose of their preparation was to obtain legal advice or assistance.  In some cases it was said that the purpose had to be substantial, or appreciable, but the weight of authority was against the view that the existence of another purpose, or other purposes, in addition to obtaining legal advice or assistance, resulted in loss of the privilege.  It was well settled that it was the purpose of the report, not the motive of the individual who made it, that mattered.  In many cases the reports would result from established corporate or bureaucratic procedures, and the individual who made the report would simply be following instructions.  It may be necessary to understand the internal procedures, or the objectives of some person of higher authority, in order to identify the purpose or purposes for which reports were prepared. 

40. The generally accepted view, however, was that, if there were multiple purposes, it was sufficient to attract privilege that one, not insubstantial, purpose was that of obtaining legal advice or assistance.  It was to the correctness of that view that the judgments in Grant v Downs were primarily directed.  Rath J, in the Supreme Court of New South Wales, held that the reports were privileged.  The Court of Appeal refused leave to appeal from his decision[47].  Special leave to appeal to this Court was granted.  The decision of Rath J was reversed.  Similarly in Waugh, both Donaldson J and the Court of Appeal in England upheld the claim for privilege[48], but the House of Lords reversed their decisions.  Thus, in both Grant v Downs and in Waugh, this Court and the House of Lords narrowed the scope of the privilege.  It was in relation to the extent of the narrowing that the cases differed.

    [47](1976) 135 CLR 674 at 675-676.

    [48][1980] AC 521.

41. Although the judgment of Barwick CJ in Grant v Downs is sometimes referred to as a dissenting judgment, that is not strictly accurate.  All five members of the Court agreed in the result.  They were all of the opinion that the test applied by Rath J, (that a purpose of obtaining legal advice or assistance was sufficient, even though there were other purposes), should no longer represent the common law in Australia.

42. It is in the joint judgment of Stephen, Mason and Murphy JJ that the sole purpose test emerged.  In the light of subsequent developments in England and other common law jurisdictions, it is to be noted that nowhere in their reasons did their Honours expressly consider a dominant purpose test as an alternative possibility, or give reasons for rejecting such a test.  The reasons they gave were advanced as reasons for rejecting the prevailing test, which had been applied by Rath J.  An examination of the transcript of the argument in the case shows that the question whether, if the prevailing test were rejected, the new test should be a sole purpose or a dominant purpose test, was not debated.  Neither party to the appeal had an interest in that question.  It was sufficient for the appellant's purposes that either a dominant purpose or a sole purpose test be adopted.  The respondent, to succeed, had to contend for the test applied by Rath J.  It did not matter to either party whether, if a test stricter than that applied by Rath J were adopted, it was a sole purpose or a dominant purpose test.

43. A reading of the joint judgment shows that a reason which influenced the decision was a concern that, in large corporations and public authorities, especially those with internal legal officers, routine reports and other documents prepared by subordinates for the information of their superiors would also, in the ordinary course, be provided to lawyers for the purpose of obtaining legal advice or assistance.  It was regarded as unacceptable, and contrary to the interests of justice, that such documents should be privileged merely because one of their intended destinations was the desk of a lawyer.

44. Their Honours said[49]:

    "It is difficult to see why the principle which lies behind legal professional privilege should justify its extension to material obtained by a corporation from its agents with a double purpose.  The second purpose, that of arming central management of the corporation with actual knowledge of what its agents have done, is quite unconnected with legal professional privilege; it is but a manifestation of the need of a corporation to acquire in actuality the knowledge that it is always deemed to possess and which lies initially in the minds of its agents.  That cannot itself be privileged; quite the contrary.  If the party were a natural person or, more accurately, an individual not acting through servants or agents, it would be precisely that knowledge which would be discoverable and the party cannot be better off by being a corporation.  The fact that a second purpose may also be being served, a purpose to which the privilege would extend, does not cover with that privilege information which would otherwise be discoverable.

    …

    All that we have said so far indicates that unless the law confines legal professional privilege to those documents which are brought into existence for the sole purpose of submission to legal advisers for advice or for use in legal proceedings the privilege will travel beyond the underlying rationale to which it is intended to give expression and will confer an advantage and immunity on a corporation which is not enjoyed by the ordinary individual.  It is not right that the privilege can attach to documents which, quite apart from the purpose of submission to a solicitor, would have been brought into existence for other purposes in any event, and then without attracting any attendant privilege."

    [49](1976) 135 CLR 674 at 687-688.

45. That reasoning, if accepted, explains why the prevailing test, applied by Rath J, should be rejected, but it does not necessarily demand rejection of a dominant purpose test.  The premise that it is unsatisfactory that one purpose of obtaining legal advice or assistance is sufficient does not sustain a conclusion that it is necessary that such a purpose be the only purpose.  The fact that a report which is prepared for a dominant purpose, which is a legal purpose, and for a subsidiary purpose as well, does not necessarily mean that, if the dominant purpose did not exist, the report would nevertheless still have come into existence.  To use the language of Jacobs J, to which further reference will be made below, it might be the dominant purpose which alone accounts for the existence of the report.

46. Leaving to one side whatever room for evaluation or judgment might have resulted from qualifying adjectives such as "substantial" or "appreciable" under the previous test, the joint judgment appeared to substitute one bright-line test for another.  Previously, in the case of a multiplicity of purposes, it sufficed if one purpose was submission to legal advisers or use in legal proceedings.  Now the test was whether that was the sole purpose.  The reasons for the joint judgment addressed those alternatives.  The other two members of the Court also addressed intermediate possibilities.  Jacobs J distinguished between communications to obtain advice or action in litigation which is pending or in fact expected, and communications where litigation is merely a contingent possibility.  In the latter case, his Honour said, the test is whether the purpose of supplying the material to the legal adviser accounts for the existence of the material[50].  Barwick CJ, in a passage that proved influential in other jurisdictions, said[51]:

    "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."

    [50](1976) 135 CLR 674 at 690, 692.

    [51](1976) 135 CLR 674 at 677.

47. The Chief Justice explained that he thought a sole purpose test too narrow, but was unable to accept a lesser requirement than dominant purpose.

48. As the judgment of Jacobs J illustrates, there are other possible ways of formulating a suitable test without reference to the concept of purpose, whatever qualifying adjective is employed.  In many areas of the law, references to purpose bring their own potential for uncertainty and argument.  However, as appears from the detailed examination of the earlier authorities on the subject made by Havers J in Seabrook v British Transport Commission[52], for more than a century courts have expressed the principles relating to the privilege in terms of purpose, and it would not contribute to certainty to depart from that term at this stage.

    [52][1959] 1 WLR 509; [1959] 2 All ER 15.

49. In Waugh, the House of Lords, after an examination of the previous English authorities, agreed that the reports in that case were not privileged, and that the pre-existing rule, applied in the Court of Appeal, which was to the same effect as that applied by Rath J in Grant v Downs, should be changed.  However, in considering Grant v Downs, their Lordships unanimously preferred the test expounded by Barwick CJ to that of the joint judgment.  They acknowledged that a dominant purpose test was less clear than a sole purpose test, but they found the latter unduly restrictive.  They pointed out that dominant purpose is a concept well known, and frequently applied, in other areas of legal discourse.

50. Upon one point, which was the point of decision in each case, all members of the High Court and the House of Lords were agreed:  if the most that could be said of the reports in question was that the purposes for which they came into existence included a purpose of obtaining legal advice or assistance, then privilege would not apply.

51. In 1985, in Guardian Royal Exchange Assurance of New Zealand Ltd v Stuart[53] the same question arose in the Court of Appeal of New Zealand.  In that country, the prevailing authority supported a test which asked whether the legal purpose was an appreciable purpose of a report or communication[54].  The Court of Appeal preferred a narrower test, and adopted the dominant purpose test.  Richardson J said[55]:

    "… I am satisfied that we should move to a dominant purpose test.  First, a more restrictive test than appreciable purpose is called for in balancing the relevant public interest considerations. …  Second, in terms of ease of application a dominant purpose test is both familiar to lawyers and more straightforward in its application.  As [has been] observed … the phrase 'appreciable purpose' is not precise and in some instances an appreciable purpose test would be much more difficult to apply than a sole or dominant purpose test.  And a dominant purpose test is a familiar concept in other branches of the law, notably insolvency and taxation.  Finally, it holds the scales in even balance, whereas at the other extreme, unless read down by refusing to rank as a 'purpose' any considerations other than submission to legal advisers which were in mind, a sole purpose test would provide extraordinarily narrow support for the privilege."

    [53][1985] 1 NZLR 596.

    [54]Konia v Morley [1976] 1 NZLR 455.

    [55][1985] 1 NZLR 596 at 605.

52. Although it has no direct bearing upon the issue now under consideration, reference should be made to a point emphasised both in the joint judgment in Grant v Downs[56] and by Cooke J in Guardian Royal Exchange Assurance of New Zealand Ltd v Stuart[57].  A claim for privilege is not conclusively established by the use of a verbal formula.  A court has power to examine documents in cases where there is a disputed claim, and it should not be hesitant to exercise such a power[58].  In appropriate cases, there is also power to allow cross-examination of a deponent of an affidavit claiming privilege[59].

    [56](1976) 135 CLR 674 at 689.

    [57][1985] 1 NZLR 596 at 599.

    [58]See also Trade Practices Commission v Sterling (1979) 36 FLR 244.

    [59]National Crime Authority v S (1991) 29 FCR 203.

53. The dominant purpose test has been taken up in Ireland and in Canada.  In Australia, when the Parliament enacted the Evidence Act, it adopted the dominant purpose test for the areas in which the Act was to apply, and in this respect it was followed by the Parliament of New South Wales.

54. The appellant does not invite this Court to refuse to follow the actual decision in Grant v Downs, which was unanimous.  It established that the previously accepted test, by reference to whether the legal purpose was one purpose, was inappropriate.  This Court is not invited to reconsider Grant v Downs with a view to deciding that the common law in Australia is now even further out of line with the common law in other countries, and with the Evidence Act, than was thought.  Nor are we invited by the appellant to declare that what Jacobs J said in Grant v Downs states the test, although in practice it may not be very different from the dominant purpose test.  Whatever its merits, it has not been adopted in other jurisdictions.  The submission is that we should reconsider the point upon which the judgment of Barwick CJ differed from the joint judgment; a point which was not the subject of argument in the case and which was not critical to the decision.

55. Although what is proposed does not involve an overruling of a previous decision of the Court, nevertheless the question whether to reconsider the reasoning of Stephen, Mason and Murphy JJ, and to refuse to follow it if we disagree with it, should be decided by reference to considerations of the kind discussed by Gibbs CJ in The Commonwealth v Hospital Contribution Fund[60].  These considerations were applied in John v Federal Commissioner of Taxation[61] when overruling Curran v Federal Commissioner of Taxation[62].  Grant v Downs has, for more than 20 years, been accepted in Australia as authority for the sole purpose test of legal professional privilege, and it has been consistently followed in later decisions and has been applied in this Court in National Employers' Mutual General Insurance Association Ltd v Waind[63] and Waterford v The Commonwealth[64].  The power to disturb settled authority is, as Gibbs CJ said, one to be exercised with restraint, and only after careful scrutiny of the earlier course of decisions and full consideration of the consequences.

    [60](1982) 150 CLR 49 at 55-58.

    [61](1989) 166 CLR 417 at 438-439, 450-453. See also Northern Territory v Mengel (1995) 185 CLR 307 at 338.

    [62](1974) 131 CLR 409.

    [63](1979) 141 CLR 648.

    [64](1987) 163 CLR 54.

56. The sole purpose test enunciated by Stephen, Mason and Murphy JJ did not rest upon a principle that had been worked out in a succession of cases.  On the contrary, it overturned what was, until then, accepted principle.  Insofar as the question was whether there should be a sole purpose or a dominant purpose test, that question was not important to the parties to the appeal, and was not the subject of argument save to the extent that what was said about the point in issue in the case, which was whether the pre-existing test should prevail, indirectly reflected on the matter.  The reasons given in the joint judgment for rejecting the pre-existing test do not, as a matter of logic or of policy, require a preference for the sole purpose test over the dominant purpose test, and nowhere do those reasons address a possible choice between those two tests.  The House of Lords in England, and the Court of Appeal in New Zealand, with the benefit of the reasoning in Grant v Downs available to them, subsequently preferred the dominant purpose test, and the law in Australia is now out of line with other common law jurisdictions.  The parliaments of the Commonwealth and New South Wales have adopted the dominant purpose test for their Evidence Acts.  All those circumstances, in combination, lead to the conclusion that this Court should now reconsider the matter.

57. The search is for a test which strikes an appropriate balance between two competing considerations:  the public policy reflected in the privilege itself, and the public policy that, in the administration of justice and investigative procedures, there should be unfettered access to relevant information.  Additionally, whatever test is adopted must be capable of being applied in practice with reasonable certainty and without undue delay and expense in resolving disputed claims.

58. At first sight, sole purpose appears to be a bright-line test, easily understood and capable of ready application.  Many disputes as to its application could be resolved simply by examining the documents in question.  However, there is reason to believe that the position is not quite as it appears.  The main objection to the test is what was described in the Court of Appeal in New Zealand as its extraordinary narrowness.  If it is to be taken literally, one other purpose in addition to the legal purpose, regardless of how relatively unimportant it may be, and even though, without the legal purpose, the document would never have come into existence, will defeat the privilege.  This has led some judges to apply the Grant v Downs test in a manner which might suggest that it is not to be taken literally.  For example, in Waterford v The Commonwealth[65], Deane J said the test of whether a document is to be protected is whether "the cause of its existence, in the sense of both causans and sine qua non, must be the seeking or provision of professional legal advice".  That may be closer to dominant purpose than sole purpose.  At the least, it seems to involve a reformulation aimed at avoiding the use of "purpose" and also at avoiding the conclusion that the existence of any purpose in addition to the legal purpose, albeit minor and subsidiary, will mean that no privilege attaches.  In argument in the present case, counsel for the respondent endeavoured to explain the meaning of the sole purpose test in a manner that equated it with the test expounded by Jacobs J in Grant v Downs.  Whilst seeking to uphold a sole purpose test, they submitted that "if a document is created for the purpose of seeking legal advice, but the maker has in mind to use it also for a subsidiary purpose which would not, by itself, have been sufficient to give rise to the creation of the document, the existence of that subsidiary purpose will not result in the loss of privilege".  That appears close to a dominant purpose test.  If the only way to avoid the apparently extreme consequences of the sole purpose test is to say that it should not be taken literally, then it loses its supposed virtue of clarity.

    [65](1987) 163 CLR 54 at 85.

59. One of the considerations prompting rejection of the pre-existing test was that it was unduly protective of written communications within corporations and bureaucracies.  The sole purpose test goes to the other extreme.  Such organisations necessarily conduct a large proportion of their internal communications in writing.  If the circumstance that a document primarily directed to lawyers is incidentally directed to someone else as well means that privilege does not attach, the result seems to alter the balance too far the other way.  This may be the kind of result Deane J was intending to avoid in his reformulation of the privilege, but it seems to follow unless one puts a gloss upon the sole purpose test.

60. A dominant purpose test was sufficient to defeat the claims for privilege in Grant v Downs, and Waugh.  The reason why Barwick CJ, the House of Lords, and the New Zealand Court of Appeal preferred that test was that they were unable to accept, as either necessary or desirable, the apparent absoluteness and rigidity of a sole purpose test.  If the only way to avoid that absoluteness and rigidity is to water down the sole purpose test so that, in its practical application, it becomes more like the dominant purpose test, then it should be abandoned.  Either the test is too strict, or it lacks the clarity which the respondent claims for it.

61. It would be possible to seek to formulate a new test, such as that adopted by Jacobs J in Grant v Downs, or Deane J in Waterford, in a further attempt to adjust the necessary balance of competing policies.  To do so, however, would produce only confusion.  As a practical matter, the choice presently confronting this Court is between sole purpose and dominant purpose.  The dominant purpose test should be preferred.  It strikes a just balance, it suffices to rule out claims of the kind considered in Grant v Downs and Waugh, and it brings the common law of Australia into conformity with other common law jurisdictions.

    Conclusion

62. The appeal should be allowed with costs.  The orders of the Full Court of the Federal Court of Australia should be set aside.  In their place, the appeal to that court from Foster J should be allowed.  The questions of law raised for decision by Foster J should be answered:

    (a)The correct test is the dominant purpose test, which is the common law test for claiming legal professional privilege.

    (b)Does not arise.

63. The respondent should pay the appellant's costs of the proceedings before Foster J and of the appeal to the Full Court of the Federal Court.

64. McHUGH J.   In my opinion this appeal should be dismissed.  The facts and issues are set out in the joint judgment of Gleeson CJ, Gaudron and Gummow JJ.  I agree with their Honours that the Full Court of the Federal Court was correct in holding that the Evidence Act 1995 (Cth) does not apply to the discovery of confidential written communications made between lawyer and client, either directly or by a derivative modification of the common law. I also agree with their Honours that the discretionary power in O 15 r 15 of the Federal Court Rules should not be exercised for the purpose of subverting or circumventing the rules "which determine the existence of privilege"[66].

    [66]Esso Australia Resources Ltd v The Commissioner of Taxation of the Commonwealth of Australia [1999] HCA 67 at [32].

65. However, I am unable to accept the proposition that the Court should now overrule the ratio decidendi of Grant v Downs[67] and substitute a dominant purpose test of privilege for the sole purpose test laid down in that case.  Two reasons lead me to that conclusion.  First, it would extend the area of privilege with the result that a party to litigation, and the court, would have less access to relevant material.  Second, it would impose a test that is not easy of application and which seems inconsistent with the rationale of legal professional privilege.  Furthermore, a dominant purpose test is one that must lead to extensive interlocutory litigation because there seems to be a growing acceptance, contrary to earlier authority and former practice, that the person claiming privilege can be cross-examined on the affidavit claiming privilege[68].

    [67](1976) 135 CLR 674.

    [68]For the earlier reluctance to allow cross-examination, see Brambles Holdings Ltd v Trade Practices Commission (No 3) (1981) 58 FLR 452 at 454 per Franki J; Fruehauf Finance Corporation Pty Ltd v Zurich Australian Insurance Ltd (1990) 20 NSWLR 359 at 366 per Giles J; and cf Morgan v Shaw (1819) 4 Madd 54 [56 ER 629] (a solicitor's statement on oath that he cannot answer a question without breaching privilege is conclusive unless it appears from the question that no issue of privilege could arise). The more permissive approach is seen in National Crime Authority v S (1991) 29 FCR 203 at 211 per Lockhart J; Hartogen Energy Ltd (in liq) v Australian Gas Light Company (1992) 36 FCR 557 at 561 per Gummow J; and Re Compass Airlines Pty Ltd (1992) 35 FCR 447 at 462 per Beaumont and Gummow JJ.

    In New South Wales, Pt 36 r 13(3) of the Supreme Court Rules confers power on the Court to permit, if it thinks fit, cross-examination on any affidavit used in support of a claim of privilege.

66. The decision of this Court in Grant v Downs[69] in 1976 came as a surprise to me, as I am sure it did to most practising lawyers of those days[70].  Even the minority view of Barwick CJ – who would have applied the dominant purpose test – had little, if any, support in dicta[71].  Before Grant, the accepted view was that legal professional privilege applied if one purpose of a confidential communication was to obtain legal advice or assistance or for use in litigation.  Although there was no decision that established this proposition, my view, and I think the view of many practising lawyers, was that the judgment of Buckley LJ, concurred in by Vaughan Williams LJ, in Birmingham and Midland Motor Omnibus Co Ltd v London and North Western Railway[72] correctly stated the law.  In that case, Buckley LJ referred to what was required in an affidavit claiming legal professional privilege for a communication between lawyer and client.   His Lordship said:

    "It is not I think necessary that the affidavit should state that the information was obtained solely or merely or primarily for the solicitor, if it was obtained for the solicitor, in the sense of being procured as materials upon which professional advice should be taken in proceedings pending, or threatened, or anticipated."

    [69](1976) 135 CLR 674.

    [70]See the article by Professor Pearce "Legal Professional Privilege – Sole or Dominant Purpose" in [1979] Australian Current Law DT 281 where he states that the decision in Grant was "greeted with disbelief by some practitioners."

    [71]The very experienced senior counsel for the appellant agreed with a suggestion that I put to him in argument that "most members of the profession would have thought at the time that both the majority and the minority [in Grant] got it wrong": transcript of proceedings, 28 September 1999 at 6.

    [72][1913] 3 KB 850 at 856.

67. On this view, legal professional privilege protected a communication if it could be said that it was a communication made for the purpose of obtaining or giving legal advice or for use in litigation which was on foot or might reasonably be expected or anticipated.  If the communication was made for that purpose, it was immaterial that it was also made for some other purpose or purposes.  But that view of the scope of legal professional privilege changed with Grant v Downs[73].  After the decision in Grant, it was no longer sufficient that one of the purposes of the communication was obtaining legal advice or assistance.  It had to be the sole purpose of the communication.

    [73](1976) 135 CLR 674.

68. The decision in Grant brought about such a change in practice that it was not surprising that it would come under criticism.  The criticism was not long in coming.  In the next Australian edition of Cross on Evidence, which was published in 1979, the learned authors wrote[74]:

    "It is apparent that the majority decision will expose to production a great number of documents which have been traditionally supposed to be immune, such as routine reports following accidents or even loss assessors' reports to insurance companies which have among their purposes that of informing the underwriters of the justice of a claim by the insured.  It is submitted that a too rigid application of the principle in Grant v Downs will lead to an undesirable reluctance on the part of such persons to express opinions which might subsequently be used against their principals.  Moreover, in practice, it is likely to lead to unnecessary refinements in the concept of 'purpose', as it is used in the rule, with a consequent loss of certainty among litigants as to the precise status of documents in their possession."

    [74]2nd Aust ed, par 11.27.

69. The criticism has continued[75].  Furthermore, no other final appellate court in the English-speaking world has adopted the sole purpose test, and the federal and New South Wales parliaments have adopted the dominant purpose test for the adduction of evidence[76].

    [75]Australia, Senate Legal and Constitutional Legislation Committee, Evidence Bill 1993 Final Report, December 1994, par 1.36.

    [76]Evidence Act 1995 (Cth) and Evidence Act 1995 (NSW), ss 118, 119.

70. These seem powerful reasons for reviewing a decision which established a principle which had no support in the previous case law or professional practice and which, as the majority judgment in this case points out[77], was not even the subject of argument in this Court or at first instance or on appeal in the courts of New South Wales.  But that said, I am not convinced that we should now substitute the dominant purpose test for the sole purpose test.

    [77]Esso Australia Resources Ltd v The Commissioner of Taxation of the Commonwealth of Australia [1999] HCA 67 at [42].

71. First, the dominant purpose test will inevitably restrict the amount of information that a person can be required to disclose on discovery or in answer to subpoenas.  Courts will have less information before them.  How much less is impossible to tell.  But applying a dominant purpose test must often mean that a party to litigation will not have access to documents containing information that is relevant to the litigation.  It is not merely that the party will be deprived of material that can be tendered in evidence.  Discovery and subpoenas often bring to light material that points the way to obtaining relevant and admissible information not in the possession of the party issuing a subpoena or seeking discovery.  A document is discoverable if it "may fairly lead ... to a train of inquiry" which might either advance the case of the party seeking discovery or damage the case of the party resisting it[78].  If dominant purpose becomes the test of legal professional privilege, a party will have less access to information than he or she would if Grant v Downs stated the relevant law.  Because that is so, courts may sometimes have to make decisions contrary to what they would have made if the sole purpose test was the governing criterion.

    [78]Compagnie Financiere du Pacifique v Peruvian Guano Co (1882) 11 QBD 55 at 63.

72. "Legal professional privilege is itself the product of a balancing exercise between competing public interests"[79], those competing interests being the common law's pursuit of truth and the need to give effect to the rationales underlying legal professional privilege[80].  In the age of the Internet and freedom of information legislation, the line which strikes the balance should not be changed to restrict the volume of information available to decision makers.  It seems contrary to the spirit of the times for a common law court in 1999 to change a point of balance struck in 1976 when the change will result in judges and juries and other decision makers having less information available to them than they would have had under the rule in Grant v Downs.

    [79]Commissioner of Australian Federal Police v Propend Finance Pty Ltd (1997) 188 CLR 501 at 583.

    [80]For discussion of these rationales, see my comments in Carter v Northmore Hale Davy & Leake (1995) 183 CLR 121 at 160-161.

73. Second, a dominant purpose test is much harder to apply than the sole purpose test and its use must increase the volume of interlocutory litigation[81].  When "sole" purpose or "a" purpose is the criterion for existence of the privilege, seldom can it be necessary to go beyond the contents of the document and the identity of its recipient to determine whether privilege protects the document.  But if "dominant" purpose becomes the test, it will often be necessary to examine the state of mind of the person creating the document.  Cross-examination of the deponent to the affidavit claiming privilege will be inevitable in many cases, particularly in relation to what the second Australian edition of Cross on Evidence called "routine reports following accidents or even loss assessors' reports to insurance companies ... of a claim by the insured."[82]  If deponents can be cross-examined, the cross-examiner's side must be able to adduce evidence to prove that legal advice or assistance or use in litigation was not the dominant purpose of the communication.

    [81]See my comments in Perre v Apand Pty Ltd (1999) 73 ALJR 1190 at 1205; 164 ALR 606 at 628-629, in relation to the costs of litigation and the need for certainty.

    [82](1979), par 11.27.

74. It needs to be kept in mind that the application of the different tests for legal professional privilege – "a purpose", "sole purpose" or "dominant purpose" – brings about a different result only when a communication is made to the same person for more than one purpose.  If a loss assessor sends his or her report to the solicitor for a party and a copy to the insurer of a party "for information", the latter communication is not privileged unless it also was sent for a purpose connected with the legal aspects of the litigation.  Without that purpose, the copy is not privileged whatever test is applied – "sole", "dominant" or "a" purpose.

75. But if dominant purpose becomes the test of legal professional privilege, production will often depend on whether one or two communications are made.  If the loss assessor sends the report to the insurer for a purpose such as settling a claim and also for the purpose of being used in litigation if it ensues, how will any court realistically determine whether one or other purpose was dominant?  Will the issue turn on an investigation of the likelihood of litigation being commenced or legal advice being sought?  Claims of privilege for loss assessors' reports are likely to be a fertile field of dispute.  But such disputes can arise whenever a communication is sent or received with a legal purpose and a non-legal purpose in mind.

76. To take another example, if a company secretary writes a letter, explaining the company's version of an event, to a director who is also the company's solicitor and does so for the purpose of keeping the director informed of company affairs and also for the purpose of seeking legal advice, privilege will depend on what was the dominant purpose of the secretary.  It seems odd that privilege should exist if, on balance, the dominant purpose of the communication was to obtain or receive legal advice but is lost if neither purpose dominated.  If the secretary had sent two separate letters, one addressed to the recipient in her capacity as a solicitor and the other in her capacity as a director, no privilege could be claimed for the latter letter.  On no basis could it be claimed that the latter communication was privileged.  The dominant purpose test, therefore, produces a result that depends on form.  That cannot happen with the "a purpose" or "sole purpose" tests.  Under those tests, claims of legal professional privilege do not depend on whether there are two communications or one communication with two purposes.

77. There are also practical difficulties with the dominant purpose test.  Sometimes, the contents of the document may indicate that one or other purpose was dominant.  But often it may not.  Moreover, frequently the issue of dominant purpose will turn on the surrounding facts and circumstances, particularly previous dealings[83] between the parties.  A test that requires the party claiming privilege to examine surrounding facts and circumstances seems ill suited to the discovery and subpoena processes.  A person who resists the production of a document on the ground of privilege should not be required to examine collateral circumstances to make a decision concerning the existence of privilege.  In commercial litigation, discovery often requires the examination of hundreds of thousands of documents and takes months to complete.  It is often a very expensive process and it should not be made more expensive by a party or its employees and their lawyers being forced to investigate the circumstances surrounding the making of a communication to ascertain what was the dominant purpose of the communication.

    [83]Ankin v London and North Eastern Railway Co [1930] 1 KB 527; Abbott v Brisbane City Council [1941] QWN 44.

78. Furthermore, legal professional privilege exists so that persons can obtain legal advice and assistance without fear of their confidences being disclosed.  It does not exist to protect communications for non-legal purposes.  It is contrary to the rationale of the privilege that communications made for non-legal purposes should be able to free-ride on the protected purpose and obtain protection.  If a person would disclose information for a non-legal purpose in any event, why should he or she get the benefit of a privilege that exists only because it is thought necessary to have it so that people will communicate freely with their lawyers?  No doubt in many cases, the same facts will be directed to all purposes of a communication.  In others, they will not.  In some cases, it may be possible to uphold privilege for some facts and not others.  But it will not always be possible to sever facts and matters directed to a non-legal purpose from the facts and matters going to the relevant legal purpose.  Because that is so, information not directed to the privileged purpose will obtain protection from disclosure, although it would not be protected if communicated separately.  In any event, issues of severance can only make the process of discovery or the answering of subpoenas more complex than it has ever been in this country.  Whether or not the facts of a communication can be categorised into those relating to legal and those relating to non-legal purposes, the rationale of the privilege provides no ground for extending it to facts, matters or information that are disclosed for non­legal purposes.

79. I am afraid that I do not understand the contention that in practice the sole purpose test is never applied and that a dominant purpose test is effectively applied.  This contention and much of the criticism of the sole purpose test seem to spring from two errors:  first, failing to keep in mind that the privilege attaches to communications and not to documents; second, failing to keep in mind that the privilege belongs to the client, not the legal adviser[84].  In some circumstances, the legal adviser may waive the privilege because of ostensible authority even when he or she has been expressly told not to waive the privilege[85].  But subject to instructions to the contrary, the legal adviser has no actual authority to waive privilege or to act in a way that is inconsistent with the privilege.

    [84]R v Davies (1921) 21 SR (NSW) 311.

    [85]Great Atlantic Insurance Co v Home Insurance Co [1981] 1 WLR 529 at 539-540 per Templeman LJ; [1981] 2 All ER 485 at 492-494.

80. Where the communication is constituted by or recorded in a document, the document is merely evidence of the communication.  Thus, notes of a solicitor's interview with a client are privileged because they record the communication[86].  As I pointed out in Commissioner of Australian Federal Police v Propend Finance Pty Ltd[87], "[l]egal professional privilege is concerned with communications, either oral, written or recorded, and not with documents per se."  When privilege is claimed for a document, it is because it records or constitutes a communication prepared, given or received for the purpose of obtaining legal advice or assistance.  Any document brought into existence for the purpose of recording information that is to be submitted to a solicitor is privileged[88].  The communication need not come from the client; it may be a communication from a third party to a solicitor providing information at the request of the solicitor or the client[89].  In some cases, privilege may exist for the communication even where the third party does not anticipate litigation, provided that the client does[90].

    [86]Ainsworth v Wilding [1900] 2 Ch 315.

    [87](1997) 188 CLR 501 at 552.

    [88]The Theodor Körner (1878) 3 PD 162; Southwark Water Co v Quick (1878) 3 QBD 315; Ankin v London and North Eastern Railway Co [1930] 1 KB 527; Seabrook v British Transport Commission [1959] 1 WLR 509; [1959] 2 All ER 15.

    [89]Tooheys Ltd v Housing Commission of New South Wales (1953) 53 SR (NSW) 407.

    [90]Di Pietrantonio v Austin Hospital [1958] VR 325.

81. If the sole purpose of the communication is to obtain or give legal advice or assistance, privilege exists under Grant v Downs.  If there was some other purpose for the communication, privilege does not exist.  But it is the purpose of the communication that is decisive, not the purpose in making the document (for example, to have a record) or any copies of the document.  If six copies of a communication are made because they may later be useful, they are all privileged if the communication was privileged.  That is because they evidence a privileged communication.  Thus, even an entry in a bill of costs may be privileged because it records a communication[91].

    [91]Ainsworth v Wilding [1900] 2 Ch 315.

82. If a copy of a document recording a privileged communication is subsequently used as the basis of a new communication, whether that new communication is privileged will depend on the purpose for which it is made.  The copy may not be privileged because the purpose for which it is used is not within the protection of the privilege.  The result may be that the privilege attaching to the original communication is lost.  But the privilege in the original communication is not lost because the legal adviser made more copies of it than were necessary or because he or she wanted to keep a record of the communication for his or her own purposes.  Those states of mind of the legal adviser do not bear upon the purpose of the communication that attracts the privilege.

83. In my opinion, we should not substitute a dominant purpose test for a sole purpose test.  No doubt in the courts of New South Wales and in courts exercising federal jurisdiction, some practitioners and judges may feel uncomfortable about using a dominant purpose test for the adduction of evidence and a sole purpose test for the production of documents.  But even in those jurisdictions the worst that can be said is that the different tests may lead to the production of documents that cannot be tendered in evidence.  That is nothing new – often documents which must be produced on discovery cannot be tendered in evidence because they merely record inadmissible hearsay, opinion material or other inadmissible matter which "may fairly lead ... to a train of inquiry" which will advance the case of the party seeking discovery or which will damage the case of the party resisting it[92].  But whatever the disadvantages of using the sole purpose test it has one great advantage over the dominant purpose test:  it has a greater potential to lead to the production of documents that lead to other forms of evidence that will be admissible.  Add to that advantage, the inevitable cost and expense of applying a dominant purpose test, and the case for overruling Grant v Downs is not persuasive.

    [92]Compagnie Financiere du Pacifique v Peruvian Guano Co (1882) 11 QBD 55 at 63.

84. The appeal should be dismissed.

85. KIRBY J.   This appeal[93] concerns the scope of the common law rule governing legal professional privilege in Australia ("the privilege").  The arguments of the parties are finely balanced.  However, in my view the appeal should be dismissed.  The present authority of this Court[94] should be maintained.

    [93]From the Full Court of the Federal Court of Australia.  Esso Australia Resources Ltd v Commissioner of Taxation (1998) 83 FCR 511 affirming the decision of Foster J: (1997) 150 ALR 117.

    [94]Grant v Downs (1976) 135 CLR 674.

86. In Commissioner of Australian Federal Police v Propend Finance Pty Ltd[95], after referring to the undesirability of impeding the beneficial operation of pre­trial facilities such as discovery and subpoenas, which had "contributed to a tendency in Australia to define narrowly the applicability of legal professional privilege"[96], I suggested[97]:

    "… a brake on the application of legal professional privilege is needed to prevent its operation bringing the law into 'disrepute'[98], principally because it frustrates access to communications which would otherwise help courts to determine, with accuracy and efficiency, where the truth lies in disputed matters".

    I remain of that view.  This appeal is an attempt to change the common law in Australia.  It should be rejected.

    [95](1997) 188 CLR 501.

    [96](1997) 188 CLR 501 at 581 citing Ligertwood, Australian Evidence, 2nd ed (1993) at 226-227, 228-229; Cross and Tapper on Evidence, 8th ed (1995) at 472; cf ReUnited States of America v Mammoth Oil Co [1925] 2 DLR 966.

    [97](1997) 188 CLR 501 at 581.

    [98]Zuckerman, "Legal Professional Privilege and the Ascertainment of Truth", (1990) 53 Modern Law Review 381 at 381.

    Facts, issues and common ground

87. The facts and issues in the appeal are stated in the reasons of Gleeson CJ, Gaudron and Gummow JJ.  Certain basic matters are not in dispute.  They can be stated briefly to provide the setting for the fundamental question which must be answered. 

88. First, whilst called "legal professional privilege" in common law decisions the description of the privilege in the Evidence Act 1995 (Cth)[99] ("the Act") as "client legal privilege" is now more accurate.  The privilege belongs to the client not the lawyer.  This feature, which involved a "change of hands" in the 18th century[100], derives from the fundamental contemporary character and purpose of the privilege.

    [99]ss 118, 119.  See also Evidence Act 1995 (NSW), ss 118, 119.

    [100]Desiatnik, Legal Professional Privilege in Australia (1999) at 10. 

89. Secondly, in practical terms, whether in interlocutory proceedings (such as discovery or the answer to subpoenas) or at a trial (as in the tender of, or attempts to elicit, evidence) claims for the privilege usually concern documents.  However, the privilege actually attaches to communications[101] – oral, written, electronic or by signs.  This is also a feature of the privilege that derives from the purpose which it upholds to permit a person, with an actual or potential legal problem, to communicate with a legal practitioner with complete candour, being able to "bare his breast"[102] so as to obtain the full benefit of professional advice and assistance[103].

    [101]Odgers, Uniform Evidence Law, 3rd ed (1998) at 341.

    [102]Waugh v British Railways Board [1980] AC 521 at 531 per Lord Wilberforce; Anderson v Bank of British Columbia (1876) 2 Ch D 644 at 649; cf Baker v Campbell (1983) 153 CLR 52 at 74; Upjohn Co v United States 449 US 383 at 389 (1981).

    [103]Grant v Downs (1976) 135 CLR 674 at 685. See also Waterford v The Commonwealth (1987) 163 CLR 54 at 62; Carter v Northmore Hale Davy & Leake (1995) 183 CLR 121 at 126-128; Commissioner of Australian Federal Police v Propend Finance Pty Ltd (1997) 188 CLR 501 at 551-552, 583-584.

125. In Telstra Corporation v Australis Media Holdings [No 1][167] McLelland CJ in Eq accepted that although the Act does not apply of its own force to ancillary processes it does have:

     "an indirect or flow-on effect, in the application of equivalent principles to all ancillary processes".

     [167](1997) 41 NSWLR 277 at 279.

126. McLelland CJ in Eq was impressed by the reasoning in Trade Practices Commission v Port Adelaide Wool Co Pty Ltd[168], in which Branson J concluded that it was arguable that the Act applied derivatively to privilege claims in respect of ancillary processes.

     [168](1995) 60 FCR 366 at 369.

127. In Towney v Minister for Land and Water Conservation(NSW)[169], Sackville J adopted and applied these two decisions, saying, in doing so, that the reasoning of McLelland CJ in Eq was "convincing". 

     [169](1997) 76 FCR 401 at 412.

128. But there is a category of cases in which a different conclusion has been reached.  Akins v Abigroup Ltd[170] and Zemanek v Commonwealth Bank of Australia[171] are instances.

     [170](1998) 43 NSWLR 539.

     [171]Unreported, Federal Court of Australia, 2 October 1997.

129. In the present case, Foster J decided not to follow Telstra and Towney. His Honour's view was that in terms, the Act applied only in respect of the adducing of evidence, and that the common law, as settled by this Court in Grant v Downs applied in other circumstances, including the present.

130. There is a third category of cases in which consideration has been given to an argument that O 15 r 15 of the Federal Court Rules authorises a Court, in effect, to circumvent Grant v Downs by refusing to make orders for discovery on the basis of a sole purpose test.

131. The rule provides:

     "The Court shall not make an order under this Order for the filing or service of any list of documents or affidavit or other document or for the production of any document unless satisfied that the order is necessary at the time when the order is made."

132. In BT Australasia Pty Ltd v New South Wales[172], Sackville J (following the reasoning of Branson J in Trade Practices Commission v Port Adelaide Wool Co Pty Ltd[173]) was of the view that the Court could, and would ordinarily, exclude from production, documents which do not meet the dominant purpose test, on the basis that the fact that neither the documents nor evidence of them might be adduced on trial was relevant to the issue of "necessity" with which the rule is concerned[174]. 

     [172](1996) 140 ALR 268.

     [173](1995) 60 FCR 366.

     [174](1996) 140 ALR 268 at 271.

133. Branson J thought it anomalous that a party might have access to a wider class of documents at an early stage of the proceedings than would be admissible at trial.  Foster J pointed out however that the scope of discovery is not limited to documents which would be admissible in evidence[175].

     [175]See Compagnie Financiere et Commerciale du Pacifique v The Peruvian Guano Co (1882) 11 QBD 55 at 62-63; Mulley v Manifold (1959) 103 CLR 341 at 345; Trade Practices Commission v CC (NSW) Pty Ltd (1995) 58 FCR 426 at 436.

134. Foster J, in disposing of the premise upon which the expansive view of the Federal Court Rules was said to be justified, held that the taking of a procedural step to widen privilege would be to accord a privilege expressly denied by the High Court in Grant v Downs[176].

     [176](1997) 150 ALR 117 at 125.

135. The questions of law were answered by Foster J in this way[177]:

     "In respect of question (a), that the correct test for claiming legal professional privilege in relation to the production of discovered documents is the 'sole purpose' test as formulated by the High Court in Grant v Downs.

     In respect of question (b), that the question be answered in the negative."

     [177]Esso Australia Resources v Federal Commissioner of Taxation (1997) 150 ALR 117 at 126.

136. The appellants appealed to the Full Court of the Federal Court, comprised of a bench of five judges (Black CJ, Beaumont, Sundberg, Merkel and Finkelstein JJ)[178].  By a majority (Black CJ and Sundberg J; Finkelstein J) the appeal was allowed in part, but only in respect of the answer given to question 1(b) which was amended as follows[179]:

     "Yes, but to exclude from production discovered documents for the sole reason that they meet the 'dominant purpose' test in ss 118 and 119 would not be a proper exercise of the power."

     [178]Esso Australia Resources v Commissioner of Taxation (1998) 83 FCR 511.

     [179](1998) 83 FCR 511 at 527.

     Construction of the Act

137. Black CJ and Sundberg J were of the view that the Act did not apply either directly or indirectly to the production of documents in pre-trial settings. Their Honours rejected the argument that the Act had a derivative effect on the common law; that the common law must adapt itself to include the Act as part of its fabric; or that the common law was indirectly modified by the Act. They held that Adelaide Steamship Co Ltd v Spalvins[180] was incorrectly decided. They said that the modification view, although it had received approval in the United States, has not been embraced in this country. In any event, their Honours thought this case would be an inappropriate one in which to apply such a principle (even if it were available), because so few jurisdictions in this country have enacted the Act: the United States cases which applied the modification principle were all cases in which there had been all but overwhelming adoption, by valid legislation of the provisions exerting the influence[181]. Finkelstein J was of a similar mind. There were only two bases in his Honour's view upon which ss 118 and 119 could be construed to produce any derivative effect upon the common law: first if to do so would promote the purpose and object of the Act, and, secondly, if the failure to adopt that construction would lead to unfair or absurd consequences.

     [180](1998) 81 FCR 360.

     [181]See Moragne v States Marine Lines Inc 398 US 375 (1970); cf Mobil Oil Corporation v Higginbotham 436 US 618 (1978).

138. In answer to the first proposition, Finkelstein J concluded that a literal reading of the sections (118 and 119) was appropriate, and that they have no application to pre-trial processes.  In relation to the second, his Honour rejected the view that prevailed in many of the cases: that apparent asymmetry or inconsistency in the tests applicable at pre-trial and trial stages was sufficient to justify a reading of the sections contrary to their express and literal meaning.  Then his Honour made two observations[182]:

     "Only one object of discovery, and in many cases not the principal object, is to obtain evidence.  Often the principal object is to obtain information that will throw light on the case … including information that would, directly or by train of enquiry, advance a party's own case or disadvantage that of his or her adversary."

     [182](1998) 83 FCR 511 at 566.

139. And he concluded the point by saying[183]:

     "Further, in the preparation of a case for trial it is commonplace, and it has been for over 150 years, for a party to obtain possession of a large quantity of documents many of which are not capable of being tendered in evidence for one reason or another … Thus, rather than creating confusion and disorder, the ascertainment of facts and information from documents not themselves admissible is often likely to lead to a just determination of a cause."

     [183](1998) 83 FCR 511 at 567.

140. The minority (Merkel J, with whom Beaumont J, in brief reasons, agreed) held that there had been no modification of the common law other than that part of the common law directly altered by the Act. Merkel J also referred to what was said by Gaudron, McHugh, Gummow and Hayne JJ in Garcia v National Australia Bank Ltd:  that it is for the High Court and not other courts in this country to depart from, or overrule its own decisions[184].

     [184](1998) 194 CLR 395 at 403.

141. Their Honours differed from the majority on the construction of ss 118 and 119 of the Act: they thought that on its proper construction it did apply to ancillary processes. Otherwise, they said, the Act would frustrate and defeat the object attributed to the Statute: to protect client legal privilege in proceedings in the federal courts.

142. On the other point, that of the use to which O 15 r 15 of the Federal Court Rules could be put, there was unanimity.  Their Honours concluded that to use the rule as a basis to accord privilege to a party when, on the authority of Grant v Downs, none existed, would not be a proper exercise of power. 

     The appeal to this Court

143. On some of the matters raised I can state my conclusions shortly. 

144. I would reject the theory that the Act operates to alter the common law, so as in some way to make its provisions applicable to circumstances other than the adducing of evidence. The United States "modification" theory has not received any acceptance so far in this Court[185].  Abstinence from legislation on a matter can on occasions be, as telling as legislation on it, or, as here, upon a closely related matter. 

     [185]Lamb v Cotogno (1987) 164 CLR 1 at 11-12.

145. I would also reject the argument that it may, and ordinarily would not be "necessary" within the meaning of O 15 r 15 to order discovery of documents before a hearing if they, or their contents could not be adduced in evidence, over an objection on the ground of client legal privilege during a hearing.  The purpose of the rule is to control oppressive and unnecessary obligations of discovery, and more particularly perhaps, to prevent obfuscation by excessive discovery.  Its purpose is not to permit modification of the law of privilege.

146. A third argument of the appellant was that on its proper construction the Act did not confine the dominant purpose test to the adducing of evidence: indeed, that as a matter of statutory construction, ss 118 and 119 apply to ancillary processes. The appellant submits that a proper, purposive construction of the Act compels that conclusion[186].  In support of this submission the appellant pointed to a quantity of extraneous material, the reports of the Australian Law Reform Commission and the deliberations and reports of the relevant Committee of the Senate[187].  The submission was that par 199 of the Final Report of the Australian Law Reform Commission could not be imputed to Parliament itself.  Paragraph 199 was in these terms:

     "The Terms of Reference limit the Commission to considering the application of the privilege in the courtroom where evidence is sought to be given.  Situations may arise where a party obtains access to documents outside the courtroom which are protected in the courtroom by the proposed privilege.  Under the proposal, the privilege will still apply in the courtroom unless the client voluntarily disclosed the document.  Having wider access on discovery or under a search warrant is usual.  Access is not determined by the rules of admissibility such as relevance and hearsay.  It is not unreasonable to have wider access in the investigative stage." (footnotes omitted)

     [186]Parramore v Duggan (1995) 183 CLR 633 at 644, 651; Mills v Meeking (1990) 169 CLR 214 at 233, 242-243; Saraswati v The Queen (1991) 172 CLR 1 at 21-23; Corporate Affairs Commission (NSW) v Yuill (1991) 172 CLR 319 at 345-348; Pyramid Building Society (In Liq) v Terry (1997) 189 CLR 176 at 195; Newcastle City Council v GIO General Ltd (1997) 191 CLR 85.

     [187]Evidence Bill 1993, Interim Report by the Senate Standing Committee on Legal and Constitutional Affairs, June 1994; Evidence Bill 1993, Final Report by the Senate Legal and Constitutional Legislation Committee, December 1994.

147. The Evidence Bill 1993 in its original form contained cll 118 and 119 which did not adopt the Law Reform Commission Proposal and were in quite different terms from the Commission's draft Bill.  It provided for a "sole purpose" test.  The Bill was referred to the Senate Standing Committee on Legal and Constitutional Affairs on 9 February 1994.  It was the Report of the Committee, which, it may be inferred, the appellant argues, influenced the Senate initially and ultimately the Parliament to adopt the dominant purpose test.

148. The recommendation of the Committee was as follows[188]:

     "The Committee recommends that clients should be able to claim legal privilege for confidential communications made and confidential documents prepared by or for a lawyer acting for the client, for the dominant purpose of providing legal advice to the client, or where the dominant purpose is to provide the client with professional legal services in connection with litigation, or anticipated litigation involving, or possibly involving, the client.  A dominant purpose test should also apply to confidential communications between, or confidential documents prepared by or for, unrepresented parties and their advisers."

     [188]Evidence Bill 1993, Final Report by the Senate Legal and Constitutional Legislation Committee, December 1994 at 13.

149. I do not propose to deal any further with this argument. Recourse to the extraneous material is not warranted. The language of ss 118 and 119 is clear. The Act is silent as to any situation other than the adducing of evidence. There is no reason to suppose that that silence on the part of the Parliament was other than deliberate. The intention of the Parliament must be taken to have been that the common law would govern the discovery, production and inspection of documents in situations other than the adducing of evidence.

150. There was a related, fourth argument which may also be shortly disposed of.  It was that the discovery, production and inspection of documents were in truth a way of adducing evidence.  It was pointed out that historically discovery was not available at common law.  Access to documents could only be ordered by Courts of Equity in suits in equity, seeking, as final relief, a Bill of Discovery, which, if granted would enable a party to gain access to documents (and information in the possession of another party), for use in common law proceedings[189].  Notwithstanding the origins of the process of discovery, production and inspection, the rules in relation to them have long since outgrown those origins, and are now the subject of other detailed rules and many decisions intended to adapt them to the efficient conduct of modern litigation.  All courts in this country today recognise the distinction between the adducing of evidence at a hearing, and the discovery, production and inspection of material otherwise. 

     [189]Bray, The Principles and Practice of Discovery (1885) at 4, 5. 

151. The appellant's major argument is that the decision in Grant v Downs should be re-opened, and a test of dominant purpose substituted by this Court.

152. It has been held that this Court does have "power to review and depart from its previous decisions".  This must, with respect, be so.  Although "such a course is not [to be] lightly undertaken"[190] there is "no very definite rule as to the circumstances in which [the Court] will reconsider an earlier decision"[191].  Dixon J made the observation that the rigid (different) rule adopted by the Court of Appeal in England in Young v Bristol Aeroplane Co Ltd[192] was "incompatible with the practice of the [High] Court and is inappropriate"[193].

     [190]John v Federal Commissioner of Taxation (1989) 166 CLR 417 at 438.

     [191]   Attorney-General for NSW v Perpetual Trustee Co (Ltd) (1952) 85 CLR 237 at 243­244.

     [192][1944] KB 718.

     [193]Attorney-General for NSW v Perpetual Trustee Co(Ltd) (1952) 85 CLR 237 at 244.

153. In John v Federal Commissioner of Taxation[194] the Court had regard to four matters or conditions which Gibbs CJ thought relevant in The Commonwealth v Hospital Contribution Fund[195] to justify a departure from an earlier decision.  The appellant submits that each of those conditions is satisfied in this case.

     [194](1989) 166 CLR 417 at 438.

     [195](1982) 150 CLR 49 at 56-58.

154. The appellant first submits that the decision in Grant v Downs did not rest upon a principle carefully worked out in a significant succession of cases.  As Barwick CJ said[196]:

     "There is no such statement of authority binding the courts in Australia.  The matter has been discussed in cases decided in England and in articles in legal journals.  But no authoritatively accepted statement of principle has emerged."

     [196](1976) 135 CLR 674 at 676. See also Birmingham and Midland Motor Omnibus Co Ltd v London and North Western Railway Co [1913] 3 KB 850 (produced for the solicitor in relation to proceedings pending, threatened or anticipated); Woolley v North London Railway Co (1869) LR 4 CP 602 (obtained with a view to litigation); Fenner v London and South Eastern Railway Co (1872) LR 7 QB 767 (obtained for the purpose of litigation); Southwark and Vauxhall Water Co v Quick (1878) 3 QBD 315 (per Cotton LJ, for the purpose of being communicated to the solicitor with the object of obtaining his advice); City of Baroda (1926) 134 LT 576 (for the purpose of being communicated to the solicitor); Seabrook v British Transport Commission [1959] 1 WLR 509; [1959] 2 All ER 15 (wholly or mainly for obtaining for and furnishing to the solicitor evidence and information where there is litigation contemplated); Longthorn v British Transport Commission [1959] 1 WLR 530; [1959] 2 All ER 32 (substantial purpose).

155. The submission is correct.  That this is so appears also from the reasons for decision of the majority in Grant v Downs who dwelt at length upon policy considerations (particularly in relation to corporations)[197] as a justification for the rule which they proposed to adopt.

     [197](1976) 135 CLR 674 at 686-688.

156. In Grant v Downs Jacobs J did not adopt the sole purpose test.  His Honour preferred a test which looked to the reason for the existence of the material and posed the question in the following, perhaps, deceptively simple terms[198]:

     "does the purpose of supplying the material to the legal adviser account for the existence of the material?"

     [198](1976) 135 CLR 674 at 692.

157. Later, in Waterford v The Commonwealth[199] Deane J used this language in stating the test his Honour then thought appropriate: 

     "For the document to be protected, the cause of its existence, in the sense of both causans and sine qua non, must be the seeking or provision of professional legal advice."

     [199](1987) 163 CLR 54 at 85.

158. The second matter upon which the appellant relies in seeking a review of Grant v Downs is that the stating of a sole purpose test by the majority was not necessary for the decision.  All Justices allowed the appeal and ordered production and inspection:  although the joint judgment considered in general terms documents brought into existence for several purposes they did not weigh up and consider the competing advantages and disadvantages of a dominant purpose test, as opposed to a sole purpose test.  As the distinction between the two tests had no relevance to the outcome of the case it is likely, it was put, that there was little by way of submission which would have assisted the Court in striking the correct balance.  In substance this submission also is correct.  

159. The third matter to which the appellant points is the inconvenience which, it contends, has flowed from the application of a sole purpose test.  Although it may have been no understatement for Dennis Pearce in an article "Legal Professional Privilege – Sole or Dominant Purpose"[200] to say that the decision in Grant v Downs was "greeted with disbelief by some practitioners" and that the adoption of the sole purpose test may have confined the privilege too narrowly, it is certainly true to say that it is the experience of practitioners and courts that the sole purpose test has proved no less fertile a ground of controversy and uncertainty than any that it replaced.

     [200][1979] Australian Current Law 281.

160. The authors of Cross on Evidence in 1979[201] made the following comments and criticisms:

     "It is apparent that the majority decision will expose to production a great number of documents which have been traditionally supposed to be immune, such as routine reports following accidents or even loss assessors' reports to insurance companies which have among their purposes that of informing the underwriters of the justice of a claim by the insured.  It is submitted that a too rigid application of the principle in Grant v Downs will lead to an undesirable reluctance on the part of such persons to express opinions which might subsequently be used against their principals.  Moreover, in practice, it is likely to lead to unnecessary refinements in the concept of 'purpose', as it is used in the rule, with a consequent loss of certainty among litigants as to the precise status of documents in their possession."

     [201]2nd Aust ed, par 11.27.

161. I agree with those observations.

162. Just as it is important to be alive to, and to avoid, so far as can be, the placing of corporations in favoured positions, there is no reason why they should be placed in a disadvantageous position by comparison with natural persons.  The application of a "sole purpose" test may in practice discriminate unfairly against corporations.  The employees of natural persons, partners, officials in bureaucracies and employees and directors of corporations may, and often must, out of practical necessity communicate internally by written report.  A corporation "cannot … think or write or act except by certain machinery, which is, so to speak, extraneous of itself"[202]. 

     [202]Mayor & Corporation of Bristol v Cox (1884) 26 Ch D 678 at 682.

163. The fourth matter referred to by Gibbs CJ in Hospital Contribution Fund was whether the earlier decision had been independently acted on "in a manner which militated against reconsideration"[203].  It is not entirely clear what acting independently on an earlier decision in a manner which militated against reconsideration would involve in every case.  One matter which could clearly militate against reconsideration is the extent to which people may have arranged their affairs on the basis of a well settled understanding of the law.  There is no doubt that Grant v Downs has been applied extensively throughout the Commonwealth including by this Court since it was decided[204].  Application may however be one thing, the organisation of affairs in a certain way, another. 

     [203]See John v Federal Commissioner of Taxation (1989) 166 CLR 417 at 438-439.

     [204]Waterford v The Commonwealth (1987) 163 CLR 54; National Employers' Mutual General Insurance Association Ltd v Waind (1979) 141 CLR 648.

164. The presence of any one or more of the matters to which Gibbs CJ referred does not mean that this Court must review an earlier decision even if the Court is persuaded that the earlier decision is one with which the later Court does not agree.  Nor should it be assumed that the category of criteria identified by Gibbs CJ should be regarded as necessarily closed.  Should this Court take the view, for example, that a decision reached by a majority of three to two should command the same weight and respect as a decision reached by a majority of all the Justices of the Court?  Another question which may arise is whether the decision of a bench which itself may have overturned what had for a long time been regarded as settled legal orthodoxy should have a monopoly on the thinking on the topic in question for all time?  If the answer to this last question is an affirmative one it would mean that those who support change of this kind would be able to entrench their changes by capitalizing on the caution of those who favour an incrementalist approach.  These questions do not need to be answered in this case because there is other, good reason to overrule Grant v Downs.  In Astley v Austrust Ltd, I referred to the disadvantage to people, particularly litigants, who have acted on a perceived, settled state of the law, when the law is restated in a quite different way[205].  Lord Browne-Wilkinson in Kleinwort Benson Ltd v Lincoln City Council also recently pointed to the anomalous position of a party who had acted on the basis that the law precluded reliance on mistake of law to ground a claim, when the House of Lords decided to change the law to make such a claim then, and in those proceedings maintainable[206].  Legislators can, and usually do enact transitional provisions when they change the law.  The courts have so far found and provided no like means of cushioning the impact of decisions which effect significant changes.  It may ultimately turn out to be an inescapable concomitant of any role that a final court may arrogate to itself to change the common law markedly, that it do so only in a way which is sensitive to the affairs and expectations of those who have acted upon the basis of what they reasonably took to be the legal status quo.  If the proposition that judges do not change the law is to be acknowledged as a fiction, then something may have to be done to displace the effect of the other legal fiction, that the law as found by the Court has always been so, and those who may have acted upon a different understanding in the past are nonetheless bound by the Court's most recent exposition of the law.  Merely to state the problems is to expose the difference between the legislative and curial roles.  Certainty, predictability, the desirability of a gradual and incremental development of the common law only, and respect for the knowledge, wisdom and experience of those who made the earlier decision are very important considerations.  The last of these matters will always however invite the question whether those who made the decision under challenge themselves paid due deference to those who in the past held a different opinion.

     [205](1999) 73 ALJR 403 at 434; 161 ALR 155 at 197.

     [206][1998] 3 WLR 1095 at 1100-1101; [1998] 4 All ER 513 at 518.

165. A change in the present circumstances is unlikely to cause any serious inconvenience to anyone.  A change in the rules relating to discovery is quite different from the creation of a new or different cause of action, the creation of a new defence, or the abolition of a pre-existing one.  Those who satisfy a test of sole purpose should certainly be able to satisfy any lesser test.  People are unlikely to have arranged their affairs on the basis that they or others may have brought into existence documents with an eye to a sole purpose test.  The only likely inconvenience is in the case of pending actions in which discovery and inspection have already been given.  In relation to those situations I think that there may be a great deal to be said for the view that Lord Browne-Wilkinson took in Kleinwort, effectively that the decision (in this case the decision of this Court) should in all respects be taken to operate prospectively only[207].

     [207]cf Bropho v Western Australia (1990) 171 CLR 1 at 22-23.

166. I do not think that I would have decided Grant v Downs in the way in which the majority did.  I do not, with respect, regard it as stating a convenient test, or a wholly fair one in accordance with the underlying rationale for legal professional privilege, of candour by clients in communications with legal advisers, or one which necessarily emerged as a result of full and considered argument by the parties.  I think that it may have a tendency to discriminate against corporations and other large organisations.  It was based to some extent on policy considerations upon which minds have much differed.

167. In my opinion therefore Grant v Downs should be overruled.  When it was decided it did represent a departure from the generally prevailing, if not universal view of the obligation of discovery as expressed by Buckley LJ in Birmingham and Midland Motor Omnibus Co Ltd v London and North Western Railway Co[208]:

     "It is not I think necessary that the affidavit should state that the information was obtained solely or merely or primarily for the solicitor, if it was obtained for the solicitor, in the sense of being procured as materials upon which professional advice should be taken in proceedings pending, or threatened, or anticipated.  If it was obtained for the solicitor, as above stated, it is none the less protected because the party who has obtained it intended if he could to settle the matter without resort to a solicitor at all."

     [208][1913] 3 KB 850 at 856.

168. The policy considerations which weighed with the majority in Grant v Downs are capable of pulling two ways.  The sole purpose test has not proved more convenient, less productive of controversy or productive of some higher order of justice.  It is also of some relevance that other common law countries have not adopted a sole purpose test[209].  The decision in Grant v Downs was not unanimous, and one of the judges, Jacobs J, who rejected a dominant purpose test, stated a test, that did not use the word "sole" and which could operate differently from the sole purpose test of the majority. 

     [209]In the United Kingdom, see Waugh v British Railways Board [1980] AC 521. For Canadian authority, see Levin v Boyce [1985] 4 WWR 702; Werner v Warner Auto-Marine Inc (1990) 44 CPC (2d) 175 (HC).  In Ireland see Silver Hill Duckling Ltd v Minister for Agriculture [1987] IR 289 and in New Zealand, Guardian Royal Exchange Assurance of NZ Ltd v Stuart [1985] 1 NZLR 596. In the United States the matter is one for each of the States and otherwise is likely to be affected by the Fifth Amendment.

169. I did give consideration to the possibility that a different test from either dominant or sole purpose might be formulated.  For example, a test of "a substantial purpose" has its attractions.  Dominant purpose is, however, by now a well understood test by reason of its adoption elsewhere.  And in any event, as the parties presented their arguments, there were effectively only two contenders, "sole" or "dominant".

170. In Waugh v British Railways Board[210] the House of Lords was asked to adopt a sole purpose test for the United Kingdom as stated by the majority in Grant v Downs.  Their Lordships declined the invitation.  Their reasoning is compelling.  Lord Simon of Glaisdale referred in pragmatic language to the competing considerations[211]:

     "The issue exemplifies a situation which frequently causes difficulties – where the forensic situation is covered by two valid legal principles which point each to a different forensic conclusion.  Here, indeed, both principles subserve the same legal end – the administration of justice.  The first principle is that the relevant rules of law should be applied to the whole body of relevant evidence – in other words, in principle all relevant evidence should be adduced to the court.  The report in question in this appeal undoubtedly contains information relevant to the matters in issue in the litigation here.  The first principle thus indicates that it should be disclosed, so that the appellant may make use of it if she wishes. 

     The second general principle arises out of the adversary (in contradiction to the inquisitorial) system of administration of justice.  Society provides an objective code of law and courts where civil contentions can be decided.  But it contents itself with so providing a forum and a code (and nowadays some finance for those who could not otherwise get justice).  Having done so much, society considers that it can safely leave each party to bring forward the evidence and argument to establish his/her case, detaching the judge from the hurly-burly of contestation and so enabling him to view the rival contentions dispassionately."

     [210][1980] AC 521.

     [211][1980] AC 521 at 535.

171. Later his Lordship stated his conclusion in terms with which I respectfully agree[212]:

     "Your Lordships will therefore, I apprehend, be seeking some intermediate line which will allow each of the two general principles scope in its proper sphere.  Various intermediate formulae as a basis for the privilege have been canvassed in argument before your Lordships, most based on some authority – the obtaining of legal advice was 'an appreciable purpose'; 'a substantial purpose'; 'the substantial purpose'; it was 'wholly or mainly' for that purpose; that was its 'dominant' purpose; that was its 'primary' purpose.

     Some of these are in my view too vague.  Some give little or no scope to the principle of open litigation with the minimum exclusion of relevant evidence.  The one that appeals to me is 'dominant' purpose, as it did to Barwick CJ in Grant v Downs[213].  It allows scope to each of the governing principles.  It seems to me less quantitative than 'mainly'; and I think it would be easier to apply – the law is already cognisant of the concept of a dominant purpose – in the law of conspiracy, for example (see Crofter Hand Woven Harris Tweed Co Ltd v Veitch)[214] and in the law as to fraudulent preference in bankruptcy[215]."

     [212][1980] AC 521 at 537.

     [213](1976) 135 CLR 674.

     [214][1942] AC 435 esp at 445 per Viscount Simon LC, 452 per Viscount Maugham.

     [215]See Halsbury's Laws of England, 4th ed, vol 3, at 496, 499, pars 908, 913.

172. Whether a purpose is a dominant purpose, is, in my view, a matter to be objectively determined[216] but the subjective purpose will always be relevant and often decisive.

     [216]Guinness Peat Ltd v Fitzroy Robinson [1987] 1 WLR 1027; [1987] 2 All ER 716.

173. I would answer the first stated question as follows:

     "The appropriate test is the dominant purpose test at common law."

174. The second question perhaps should have incorporated a reference to O 15 r 15 of the Federal Court Rules as the matter originally arose on a directions hearing pursuant to it.  But because of the conclusion I have reached on the first question there is no need to consider the implications, if any, of this.  It is therefore unnecessary to answer the second stated question.

175. The matter should be remitted to the Federal Court to deal with the applications in accordance with the judgments of this Court.  The Solicitor­General offers no serious resistance to an order that costs should follow the event in this Court.  Accordingly the appeal is allowed with costs.  The respondent should pay the appellant's costs of the proceedings before Foster J and of the appeal to the Full Court of the Federal Court.