Federal Court of Australia

GetSwift Limited v Webb [2021] FCAFC 26

ORDERS

THE COURT ORDERS THAT:

1.    The appeal be allowed.

2.    Order 1 of the primary judge made on 9 September 2020 in proceeding NSD580 of 2018 be set aside and, in lieu thereof, proceeding NSD580 of 2018 be referred to the National Operations Registrar for reallocation to a judge in the Commercial and Corporations National Practice Area.

3.    There be no order as to costs, other than the parties to the appeal (including Mr Webb) and the Australian Securities and Investment Commission are to each pay one quarter of the costs of the contradictor to be fixed at $40,000 plus GST.

Note:    Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

REASONS FOR JUDGMENT

THE COURT:

INTRODUCTION

1    The ultimate issue in this appeal is whether the primary judge should have disqualified himself from hearing the trial in the class action proceeding below (the ‘Webb proceeding’) where he has refused to do so: see Webb v GetSwift Limited (No 6) [2020] FCA 1292 (reference to which will be made as ‘J’).

2    The appeal raises questions concerning the knowledge to be attributed to the hypothetical observer (the fair-minded lay observer) and the extent to which extraneous information in the mind of a fact finding judge which is to be discarded might still have a subconscious effect on the decision to be made by that judge.

3    We have come to the view that the primary judge should have disqualified himself, and referred the Webb proceeding to the National Operations Registrar for reallocation to a judge in the Commercial and Corporations National Practice Area. In our view, the primary judge failed (in the circumstances confronting him) to appreciate or take into account the hypothetical observer’s understanding that whilst a judge has an ability to discard irrelevant information, the ‘human frailty’ of the decision maker and the subconscious effect the extraneous information might have in influencing the primary judge’s final decisions nevertheless continue to be material.

4    We should say at the outset that the primary judge carefully considered the various issues in deciding not to disqualify himself. He appreciated the “unusual need for caution” (J [48]), and in fact gave leave to appeal appreciating that the decision he made gave rise to issues of some general importance. However, this appeal shows that it may be more prudent for an independent mind (or minds) to consider disqualification applications on some occasions. This approach may assist to promote confidence in the legal system, which after all is a key rationale for the apprehended bias rule. As the primary judge understood, despite any countervailing issue of public policy and case management, if “the law requires disqualification then so be it” (J [52]).

5    The primary judge was properly and selflessly motivated by a desire to achieve great economies if one docket judge could case manage and hear both regulatory proceedings and a class action when they involve consideration of the same underlying facts. The primary judge was mindful of the circumstance where the trial judge in Kirby v Centro Properties Ltd (No 2) [2011] FCA 1144; (2011) 202 FCR 439 (‘Centro (No 2)’) disqualified himself from hearing the class action after delivery of his judgment in the regulatory proceeding, a circumstance the primary judge wanted to avoid. The fact that the primary judge was concerned that he would find himself in the position of the trial judge in Centro (No 2) says something about the gravity of his concern.

6    The other matter to observe is that no party involved in the regulatory proceeding or the Webb proceeding desires that the primary judge adopt the course of case management he has in fact foreshadowed. Accepting that appropriate case management may have a significance transcending the desires of the parties to litigation, the fact that all the parties to the regulatory proceeding and the Webb proceeding either support or do not oppose the application to disqualify the primary judge does suggest caution should be exercised in not acceding to the application for disqualification.

BACKGROUND

7    In the Webb proceeding, the Respondent in this appeal, Mr Raffaele Webb, makes allegations against the First Appellant, GetSwift Limited (‘GetSwift’), of continuous disclosure contraventions under s 674(2) of the Corporations Act 2001 (Cth) (the ‘Corporations Act’), and of false or misleading statements and misleading and deceptive conduct in contravention of ss 1041E and 1041H of the Corporations Act, s 12DA of the Australian Securities and Investments Commission Act 2001 (Cth) (‘ASIC Act’) and s 18 of Sch 2 to the Competition and Consumer Act 2010 (Cth) (‘Australian Consumer Law’). Against the Second Appellant, Mr Joel MacDonald (the Managing Director of GetSwift), Mr Webb alleges contraventions of ss 1041E and 1041H of the Corporations Act, s 12DA of the ASIC Act and s 18 of the Australian Consumer Law, and also alleges contraventions under s 647(2A) of the Corporations Act by reason of knowing involvement in GetSwift’s s 647(2) continuous disclosure contraventions. The allegations arise out of announcements issued by GetSwift to the Australian Securities Exchange (‘ASX’) regarding 16 agreements entered into between GetSwift and various customers.

8    Civil penalty proceedings have also been brought by the Australian Securities and Investments Commission (‘ASIC’) against GetSwift and Mr Macdonald, as well as Mr Bane Hunter (GetSwift’s executive chairman) and Mr Brett Eagle (legal counsel and director of GetSwift) (‘ASIC proceeding’). Against GetSwift, ASIC alleges contraventions of ss 674(2) and 1041H of the Corporations Act and s 12DA of the ASIC Act. Against Mr Macdonald, ASIC alleges contraventions of ss 647(2A) and 1041H of the Corporations Act and s 12DA of the ASIC Act, as well as s 180(1) of the Corporations Act. As the primary judge accepted, the “allegations raised by ASIC and by Mr Webb are largely the same, save that in the class action the allegations relate to GetSwift’s contractual obligations and dealings with a total of 16 customers, comprising 13 customers referred to in the ASIC proceeding plus an additional three customers”: J [5]. Again as the primary judge accepted, he will be dealing with “largely the same course of events in the class action as in the ASIC proceeding”: J [46].

9    The status of each proceeding is as follows:

(a)    the ASIC proceeding was heard by the primary judge from 15 June 2020 to 15 July 2020, followed by closing submissions on 14 August 2020 and 30 September 2020. Judgment is reserved;

(b)    the Webb proceeding was due to commence on 15 September 2020 before the primary judge, but the hearing was vacated in the light of the disqualification application made on 11 August 2020;

(c)    the primary judge intends that judgment in each proceeding would be based on (and only based on) the evidence adduced in, and argument advanced in, each proceeding (that is, without regard to the evidence adduced in, and argument advanced in, the other proceeding);

(d)    the primary judge has indicated that he will not deliver judgment in the ASIC proceeding until he has heard the Webb proceeding, and will then deliver judgment contemporaneously in both proceedings;

(e)    when it comes time to hear and determine the Webb proceeding, the primary judge will have heard all the evidence and argument on liability in the ASIC proceeding. Likewise, when it comes time to deliver judgment on liability in the ASIC proceeding, and to hear and determine any penalty, the primary judge will have heard all the evidence and argument on the common issues in the Webb proceeding.

10    The following is the position confronting the primary judge:

(a)    the primary judge has received a substantial body of evidence and submissions over the course of a lengthy hearing of the ASIC proceeding;

(b)    during that time, the primary judge will have assessed and analysed that evidence, and has been asked by the parties to draw certain inferences from that material;

(c)    the primary judge has expressed the view that, in both cases, it is “highly likely” that the factual issues will be determined by reference to the documentary evidence that is common to both the Webb and ASIC proceeding and inferences drawn from it: J [28]; and

(d)    the primary judge would have already formed some views about the documentary evidence adduced in the ASIC proceeding.

11    As we have indicated, on 9 September 2020, the primary judge dismissed an interlocutory application by GetSwift and Mr MacDonald that the proceedings be referred to the National Operations Registrar for reallocation to a different judge. The primary judge also granted leave to appeal. The appeal is brought by GetSwift and Mr Macdonald. ASIC supports the appeal but seeks an alternative form of relief. Mr Webb does not oppose the relief sought by GetSwift and Mr Macdonald, although does oppose the alternative relief sought by ASIC. In the circumstances, the Court appointed a contradictor to assist the Court in its deliberations.

12    The Notice of Appeal relevantly provides the following grounds of appeal:

1.    The primary judge erred in concluding that, in all the circumstances, he should not disqualify himself from hearing the trial of proceeding NSD580 of 2018 (Webb proceeding) by reason of the existence of a reasonable apprehension of bias.

 2.    Further to ground 1 above:

(a)    The primary judge erred in determining the application on the basis that the relevant question was whether the reasonable observer might apprehend that “the court might not move its mind from the position reached on one set of materials even if different materials were presented at trial”: J [38]-[39].

(b)    The primary judge should have determined the application on the basis that the relevant question was whether the reasonable observer might apprehend that the Court might not be able to disregard the evidence adduced in proceeding VID 146 of 2019 (ASIC proceeding) in deciding the Webb proceeding and vice versa.

(c)    The primary judge erred in dismissing the application without having identified any reason for concluding that the reasonable observer might not apprehend that the Court might not be able to disregard the evidence adduced in the ASIC proceeding in deciding the Webb proceeding and vice versa.

(d)    [not pressed]

(e)    The primary judge erred in holding that the evidence of the witnesses called in the ASIC proceeding was not of such importance as to constitute extraneous information the receipt of which gave rise to a reasonable apprehension of bias in respect of hearing the Webb proceeding, or could be disregarded for that purpose because the “most important” material is documentary: J [28], [29], [33].

(f)    The primary judge erred in holding that the reasonable observer would be taken to know that the core documentary material in evidence would be likely to be decisive in the determination of both the ASIC proceeding and the Webb proceeding: J [23], [33].

(g)    The primary judge erred in holding that the possibility that the reasonable observer might apprehend that the Court might not bring an impartial mind to the determination of the Webb proceeding would be reduced by the fact that, although the court had fully heard the ASIC proceeding, it had not yet delivered judgment in that proceeding.

(h)    The primary judge erred in failing to apply the precautionary principle in circumstances where there was at least a serious question as to whether the reasonable observer might apprehend that the Court might not bring an impartial mind to the determination of the Webb proceeding or, if the Court proceeded to hear the Webb proceeding, to the ASIC proceeding.

13    In essence, GetSwift submits that the primary judge erred in failing to conclude that a fair-minded lay observer might reasonably apprehend that the primary judge might not bring an impartial mind to the resolution of the Webb proceeding by reason of his having heard evidence and argument in the ASIC proceeding. That submission has two aspects:

(a)    a fair-minded lay observer might reasonably apprehend that the primary judge, consciously or subconsciously, might be influenced by extraneous information from the ASIC proceeding (‘extraneous information ground’);

(b)    further or alternatively, a fair-minded lay observer might reasonably apprehend that, in hearing, considering, and forming views about the material in the ASIC proceeding, the primary judge might have prejudged the resolution of issues common to both proceedings (‘prejudgment ground’).

14    In respect of the prejudgment ground, GetSwift does not rely upon any particular feature in the primary judge’s conduct or comments, which could give rise to an apprehension of bias. The real issue depends upon the hypothetical observer’s understanding of the primary judge’s ability to discard extraneous information. We take the view that the extraneous information ground raised by GetSwift is the critical matter for our consideration.

15    Mr MacDonald adopted and sought to supplement GetSwift’s submissions on both grounds. ASIC raised similar and additional grounds to GetSwift, which to the extent necessary we will mention in our consideration of the grounds of appeal raised by GetSwift.

16    We briefly identify the outcomes contended for by the various interested parties:

(a)    GetSwift and Mr MacDonald seek that the primary judge be disqualified from hearing the Webb proceeding;

(b)    Mr Webb seeks that the Webb proceeding be listed for hearing, but considers that in practice this outcome is most likely to be achieved by assigning the Webb proceeding to the docket of a new judge, who has capacity to set the matter down for trial in the near short term; and

(c)    ASIC is not opposed to the primary judge ultimately hearing and determining the Webb proceeding but considers that the primary judge should in the first instance deliver judgment in the ASIC proceedings so that all parties to the Webb proceeding can make an informed decision as to whether there would be a reasonable apprehension of bias if the primary judge were then to proceed to hear and determine the Webb proceeding.

17    In addition, the contradictor submits that the appeal be dismissed and the primary judge’s proposed approach of deferring judgment in the ASIC proceeding until after the hearing and determination of the Webb proceeding be maintained.

ISSUES AND EVIDENCE IN EACH PROCEEDING

18    The primary judge accepted that in both proceedings he will be “required to make findings on factual matters in contention common to both proceedings” (J [25]), including:

(1)    the state of affairs of agreements and dealings between GetSwift and its clients;

(2)    the materiality of certain information;

(3)    the beliefs that would be held by a reasonable shareholder regarding GetSwift’s statements in relation to the company’s business model;

(4)    the beliefs that would be held by a reasonable shareholder regarding GetSwift’s statements in relation to the status of its agreements with certain customers;

(5)    the basis for various statements alleged to have been made the company; and

(6)    the relevant involvement of GetSwift’s directors in the company’s conduct in releasing certain announcements.

19    The documentary evidence adduced in the ASIC proceeding, and proposed to be adduced in the Webb proceeding, is substantially the same (J [8]), with the result that there will be a “core body of documentary material” common to each proceeding: J [28].

20    However, there are differences (whether substantial or not needs not be determined, but the differences are not minimal) as to the lay and expert evidence in each proceeding. In the ASIC proceeding, ASIC relied upon a total of 37 lay witnesses, including 19 witnesses from customers of GetSwift; four witnesses who are former associates of GetSwift; 10 witnesses from ASIC, the ASX and Chi-X Australia (a securities and derivatives exchange); and four witnesses from organisations who were large investors in GetSwift. Twenty-four of those witnesses were cross-examined. ASIC also relied upon expert opinion evidence from Mr Andrew Molony as a “professional investor” who was also cross-examined. The Defendants did not give evidence or call any witnesses.

21    In the Webb proceeding, Mr Webb intends to give evidence and to call seven out of the 19 customer witnesses called in the ASIC proceeding. All of those witnesses were cross-examined and are intended to be required for cross-examination in the Webb proceeding. Three of the witnesses will only be called if Mr Webb’s efforts to secure their attendance remotely from overseas are successful. Mr Webb intends to rely on two experts (Dr Ramsay Zein and Mr Robert Turner) as does GetSwift (Dr Sanjay Unni and Mr Martin Duncan).

22    As the primary judge accepted, when he comes to hear the Webb proceeding the Court “will be in possession of information extraneous to the evidence adduced in the class action when making findings in the class action” (J [26]): namely the written and oral evidence of 30 lay witnesses from the ASIC proceeding (or perhaps 33 lay witnesses, if the attendance of the three overseas witnesses cannot be secured); the opinion evidence of Mr Molony; and written and oral submissions.

23    Likewise, for the purposes of determining the ASIC proceeding, and if necessary any penalty, the primary judge will be in possession of extraneous information in the form of at least four experts from the Webb proceeding, and the parties’ written and oral submissions.

24    We should mention one other important factor. Mr Molony was a significant witness relied upon by ASIC on the issue of materiality. His expert opinion evidence was said to be unchallenged. No expert evidence was called by the Defendants. In the Webb proceeding, by way of example as to the significance of different evidence being adduced, Dr Zein in his expert evidence details an analysis of materiality which is contrary to that of Mr Molony in the ASIC proceeding. So the primary judge will be in the position (if he proceeds as envisaged by him) where he will be considering the detailed submissions in the ASIC proceeding divorced from any other analysis, and will then have to consider (as a separate exercise) the position put forward, for example, by Dr Zein, who takes a different approach on materiality, being the very same issue to be considered in both proceedings.

25    We would interpolate that we are not suggesting that the hypothetical observer will be aware of this exact scenario but consider that the hypothetical observer would be aware of the commonality of issues and the difference in the evidence to be called in the two proceedings.

PRINCIPLES OF LAW

26    We do not consider that there was any real dispute before us (or the primary judge) as to the applicable principles of law.

27    The test for apprehended bias is whether a fair-minded lay observer might reasonably apprehend that the judge might not bring an impartial mind to the resolution of the question he or she is required to decide: Ebner v Official Trustee in Bankruptcy [2000] HCA 63; (2000) 205 CLR 337 (‘Ebner’) at [6] (Gleeson CJ, McHugh, Gummow and Hayne JJ), applied in CNY17 v Minister for Immigration [2019] HCA 50; (2019) 375 ALR 47 (‘CNY17’) at [17]-[18] (Kiefel CJ and Gageler J); [50] (Nettle and Gordon JJ); and [132] (Edelman J). The bias rule is concerned as much to preserve the public appearance of independence and impartiality as it is to preserve the actuality: CNY17 at [18] (Kiefel CJ and Gageler J). It also reflects a precautionary approach: “In a case of real doubt, it will often be prudent for a judge to decide not to sit in order to avoid the inconvenience that could result if an appellate court were to take a different view”: Ebner at [20]. The application of the rule involves two steps: first, identification of the factor which it is said might lead the judge to decide the case otherwise than on its legal and factual merits; and, second, an articulation of the logical connection between that factor and the feared deviation from the course of deciding the case impartially on the merits: Ebner at [8]; CNY17 at [21]; cf Isbester v Knox City Council [2015] HCA 20; (2015) 255 CLR 135 (‘Isbester’) at [59] (Gageler J) where three steps are articulated. The connection must be assessed objectively: see Michael Wilson & Partners v Nicholls [2011] HCA 48; (2011) 244 CLR 427 at [67] (Gummow ACJ, Hayne, Crennan and Bell JJ). The conclusion of apprehended bias is “largely a factual one”: CNY17 at [93] (Nettle and Gordon JJ).

28    Whilst a precautionary approach is to be observed, the cases emphasise that an allegation of apprehension of bias must be “firmly established”: see, eg, Reece v Webber [2011] FCAFC 33; (2011) 192 FCR 254 at [45] (Jacobson, Flick and Reeves JJ) citing Re JRL; Ex parte CJL (1986) 161 CLR 342 (‘Re JRL’) at 352 (Mason J). The reference to “firmly established” originated in R v Commonwealth Conciliation and Arbitration Commission; Ex parte Angliss Group (1969) 122 CLR 546 at 553 (Barwick CJ, McTiernan, Kitto, Taylor, Menzies, Windeyer and Owen JJ). A conclusion of apprehended bias “is not to be reached lightly”: see CNY17 at 61 [56] (Nettle and Gordon JJ) citing Re JRL at 371 (Dawson J).

29    There is a variety of ways in which the impartiality of a court may be or may appear to be compromised. In Webb v The Queen (1994) 181 CLR 41 at 74, Deane J, who was not dissenting on this point, identified four of them as “distinct, though sometimes overlapping, main categories of case”. They were:

(1)    interest – where the judge has an interest in the proceedings, whether pecuniary or otherwise, giving rise to a reasonable apprehension of prejudice, partiality or prejudgment;

(2)    conduct – where the judge has engaged in conduct in the course of, or outside, the proceedings, giving rise to such an apprehension of bias (including prejudgment);

(3)    association – where the judge has a direct or indirect relationship, experience or contact with a person or persons interested in, or otherwise involved in, the proceedings; and

(4)    extraneous information – where the judge has knowledge of some prejudicial but inadmissible fact or circumstance giving rise to the apprehension of bias.

30    We are primarily concerned with the fourth category of extraneous information.

31    It is convenient to make some observations on the extent of the knowledge attributable to the hypothetical observer for the purpose of determining whether that observer would reasonably apprehend bias. That knowledge does not extend to a knowledge of the law or any detailed knowledge of the evidence relied upon or to be relied upon by the fact-finding judge.

32    The question was discussed in Johnson v Johnson [2000] HCA 48; (2000) 201 CLR 488 (‘Johnson v Johnson’), where the plurality (Gleeson CJ, Gaudron, McHugh, Gummow and Hayne JJ) said at [13]:

Whilst the fictional observer, by reference to whom the test is formulated, is not to be assumed to have a detailed knowledge of the law, or of the character or ability of a particular judge, the reasonableness of any suggested apprehension of bias is to be considered in the context of ordinary judicial practice. The rules and conventions governing such practice are not frozen in time. They develop to take account of the exigencies of modern litigation. At the trial level, modern judges, responding to a need for more active case management, intervene in the conduct of cases to an extent that may surprise a person who came to court expecting a judge to remain, until the moment of pronouncement of judgment, as inscrutable as the Sphinx.

(Footnote omitted.)

Justice Kirby also discussed the attributes of the fictitious bystander at [53]:

Such a person is not a lawyer. Yet neither is he or she a person wholly uninformed and uninstructed about the law in general or the issue to be decided. Being reasonable and fair-minded, the bystander, before making a decision important to the parties and the community, would ordinarily be taken to have sought to be informed on at least the most basic considerations relevant to arriving at a conclusion founded on a fair understanding of all the relevant circumstances.

(Footnotes omitted.)

And further at [53]: “a reasonable member of the public is neither complacent nor unduly sensitive or suspicious.” (Footnote omitted.)

33    It is probably fair to conclude that the hypothetical observer today is more aware of the court processes than, say, a few decades ago. Knowledge about courts has become more accessible through the media, and the courts are more accountable in the conduct of judicial functions. It is also appropriate to conclude that the hypothetical observer would, before forming a view about the existence of a reasonable apprehension of bias, take the trouble to inform himself or herself to the extent necessary to make a fair judgment of what might occur in the process confronting a fact-finding judge.

34    The hypothetical observer is to be attributed with knowledge of the nature of the decision, the context in which it is made, and the circumstances leading to it: Isbester at [23] (Kiefel, Bell, Keane and Nettle JJ). Nevertheless, it is always to be kept in mind that the observer is a layperson and not a lawyer.

35    The hypothetical observer is taken to understand how a judge is capable of putting irrelevant and immaterial matters to one side as part of the assumed abilities of a judge. In this regard a number of observations have been made by the courts:

(a)    a judge as a professional decision-maker can ordinarily be expected to be capable of discarding “the irrelevant, the immaterial and the prejudicial”: see CNY17 at [28] (Kiefel CJ and Gageler J) citing Johnson v Johnson at [12] (Gleeson CJ, Gaudron, McHugh, Gummow and Hayne JJ) quoting Vakauta v Kelly (1988) 13 NSWLR 502 at 527;

(b)    a judge is “equipped by training, experience and their oath or affirmation to decide factual contests solely on the material that is in evidence”: British American Tobacco Australia Services Ltd v Laurie & Ors [2011] HCA 2; (2011) 242 CLR 283 (‘BAT v Laurie’) at [140] (Heydon, Kiefel and Bell JJ)); and furthermore is aware of “the possibility of the evidentiary position changing”: BAT v Laurie at [145] (Heydon, Kiefel and Bell JJ); Centro (No 2) at [20] (Middleton J); GlaxoSmithKline Australia Pty Ltd v Reckitt Benckiser Healthcare (UK) Limited [2013] FCAFC 150 at [38] (Allsop CJ, Middleton and Katzmann JJ);

(c)    “[a] judge will be assumed to have a capacity to put from his or her mind evidence of a prejudicial kind which has been heard or seen but is not relevant to the determination of the question before the Court”: see R v Burrell [2007] NSWCCA 79; (2007) 175 A Crim R 21 at [7] (McClellan CJ at CL, Sully and James JJ agreeing). See also State of Victoria v Australian Building Construction Employees and Builders Labourers Federation (1982) 152 CLR 25 at 58 (Gibbs CJ), 76 (Stephen J); and

(d)    judges are capable of impartially reconsidering matters which have previously been considered or which may even have been pronounced upon by that particular judge – subject always to the nature of the findings: see, eg, R v Commonwealth Conciliation and Arbitration Commission; Ex parte Angliss Group (1969) 122 CLR 546 at 554 (Barwick CJ, McTiernan, Kitto, Taylor, Menzies, Windeyer and Owen JJ); Reece v Webber [2011] FCAFC 33; (2011) 192 FCR 254 at 272 [52] (Jacobson, Flick and Reeves JJ); and Centro (No 2) at [57] and [60] (Middleton J).

36    We agree that there is a degree of artificiality about this attribution, and indeed some scepticism has been expressed about its intellectual coherence: see Commonwealth Bank of Australia v Jackson McDonald (a firm) [2014] WASC 301 at [24] (Martin CJ). The primary judge himself has made similar remarks including that the test may obscure “normative standards of behaviour determined by the Court itself”: McKenzie v Cash Converters International Ltd (No 3) [2019] FCA 10 at [28]. See also J [30]; Webb v GetSwift Ltd (No 5) [2019] FCA 1533 at [27].

37    The test for apprehended bias is the same wherever it arises, although the context in which it falls to be applied will clearly affect how the test is applied: Cabcharge Australia Ltd v Australian Competition and Consumer Commission [2010] FCAFC 111 (Kenny, Tracey and Middleton JJ) at [25].

38    A reasonable apprehension of bias may arise from the decision-maker receiving extraneous information, including knowledge of some prejudicial but inadmissible fact or circumstance. This ground is not dependent on showing that the decision-maker might have prejudged the issues by making particular findings or rulings on the extraneous information. It is also not necessary to show that the information has in fact worked to the prejudice of the applicant – it is enough that it might do so: Re JRL at 349 (Gibbs CJ) citing Kanda v Government of Malaya [1962] A.C. 322 at 337-338.

39    Importantly, and we think determinatively in this appeal, it is to be recalled that even where a decision-maker has not consciously considered the extraneous information, a reasonable apprehension of bias can arise because of its “subconscious” influence: CNY17 at [27]-[28] (Kiefel CJ and Gageler); [51], [92], [97] (Nettle and Gordon JJ); [111] (Edelman J). Where there is a risk of subconscious bias “that risk cannot be cured by putting the information aside”: CNY17 at [97] (Nettle and Gordon JJ). Because “reason cannot control the subconscious influence of feelings of which it is unaware [where] there is ground for believing that such unconscious feelings may operate in the ultimate judgment, or may not unfairly lead others to believe they are operating, judges recuse themselves”: Public Utilities Commission of the District of Columbia v Pollak (1952) 343 US 451 (‘Pollak’) at 466-467. In CNY17, Kiefel CJ and Gageler J (at [27]) set out the following extract from Pollak (at 466-467):

…The fact is that judges do lay aside private views in discharging their judicial functions. This is achieved through training, professional habits, self-discipline and that fortunate alchemy by which men are loyal to the obligation with which they are entrusted. But is also true that reason cannot control the subconscious influence of feelings of which it is unaware. When there is ground for believing that such unconscious feelings may operate in the ultimate judgment, or may not unfairly lead others to believe they are operating, judges recuse themselves. They do not sit in judgment. They do this for a variety of reasons. The guiding consideration is that the administration of justice should reasonably appear to be disinterested as well as be so in fact.

40    Chief Justice Kiefel and Gageler J then paraphrased that reasoning as follows (at [24]):

The fair-minded lay observer would recognise that although the Authority is not a court and although a Reviewer is not necessarily a lawyer, the Authority as constituted by a Reviewer is a professional decision-making body that can ordinarily be expected to be capable of discarding ‘the irrelevant, the immaterial and the prejudicial’. But, the fair-minded lay observer must also be taken to recognise that even a professional decision-maker is not a ‘passionless thinking machine’ and that information consciously and conscientiously discarded might still sometimes have a subconscious effect on even the most professional of decision-making

(Footnotes omitted.)

41    In a separate judgment, Edelman J (at [136]) indicated that a “more robust approach” might be taken to the possibility of a judge, as opposed to a member of the executive, being influenced by extraneous information. However, his Honour nonetheless accepted that the principles governing the test for apprehended bias are based on “the recognition of human nature” and “human frailty”: at [132] citing BAT v Laurie at [139] and Ebner at [8]; and the hypothetical observer’s presumed knowledge that “in adjudication, as in life generally, the mental plasticity of human decision making is subject to the unconscious”: at [133]. In this regard, the difference between a judge or a member of the executive in the mind of the hypothetical observer will be limited.

42    In a recent article by Professor Gary Edmond and Associate Professor Kristy A Martire, ‘Just Cognition: Scientific Research on Bias and Some Implications for Legal Procedure and Decision‐Making’ (2019) 82(4), The Modern Law Review 633, it was observed (at 646):

From a cognitive science perspective biases are predispositions and preferences that affect judgment and decision-making. They can be thought of as the cognitive equivalent of a reflexive knee-jerk; they occur quickly, effortlessly and automatically. Biases are strategies that ‘are highly economical and usually effective, but they lead to systematic and predictable errors’. That is, biases are decision-making styles that do not necessarily incorporate what might be understood as rational (or legally normative) approaches to relevant admissible evidence, and they can influence how information is processed, prioritised and evaluated. Decades of research has identified not only a seemingly endless array of different types of bias, but confirmed their ubiquity and influence irrespective of a person’s profession, experience or intelligence.

(Footnotes omitted.)

43    In the book How Judges Judge, Empirical Insights into Judicial Decision-Making (New York, NY: Routledge, 2021), Dr Brian M Barry said (at pp 68-69):

Where judges are susceptible to the influence of biasing but inadmissible evidence, is motivated reasoning at play? Wistrich, Guthrie and Rachlinski suggested and speculated other psychological effects may be at work. For example, they referred to psychological reactance (a variation on what is commonly known as “reverse psychology”), ironic process theory (the difficulty people have ignoring thoughts they are trying to suppress), or mental contamination (the idea that misleading information persists in contaminating decision-making, even after someone is aware it is misleading). There is some degree of overlap between these psychological phenomena and motivated reasoning. Whatever the case, results from these studies suggest that judges sometimes seemed motivated to reason towards a particular result relying on inadmissible evidence, even though they knew to suppress and ignore it.

(Footnotes omitted.)

And earlier (at p 16) it was stated that:

Judges are aware of their obligations to use information cautiously and even-handedly. As far back as 1660, English Chief Justice Matthew Hale drafted a sort of early self-help guide for the judicial profession, which he called “things necessary to be continually bad in remembrance.” One of his resolutions was “that I suffer not myself to be prepossessed with any judgment at all, till the whole business and both parties be heard.” Simply put, judges should consciously resist the urge to prejudge a case before all evidence is aired. Nevertheless, research shows that, in fact, judges form impressions and make preliminary decisions on cases at an early stage in proceedings. Two questions arise: are judges susceptible to confirmation bias in how they use information, and if so, what does this mean for judicial outcomes?

(Footnotes omitted.)

44    Whilst then a judge is understood by the hypothetical observer to be able to discard the irrelevant, the immaterial and the prejudicial, and decide cases solely on the material that is in evidence, there still remains the possibility of apprehension of bias in respect of judicial officers seized of extraneous information. Otherwise, there would be no place for the recognition of human frailty and human nature which are accepted aspects of the approach of the fair-minded lay observer in considering apprehension of bias.

45    It follows that there is a need to recognise “human frailty” such that, even where a decision-maker has expressly disavowed consideration of certain material, it may be that the particular nature and relevance of that material can still give rise to the risk of a subconscious influence.

CONSIDERATION

46    In our view, the hypothetical observer would recognise that judges are human, not a “passionless thinking machine” (CNY17 at [28] (Kiefel CJ and Gageler J) quoting Re J P Linahan Inc (1943) 138 F 2d 650 at 653) or robot just assessing information. In the circumstances before the primary judge, there is a real difficulty of disregarding the extraneous information. The primary judge is not being asked to disregard the evidence of one witness, on a discrete topic, heard at a remote point in time. This case is removed from the “everyday duty of a judge at first instance to sift inadmissible from admissible material”: R v Judge Leckie; ex parte Felman (1977) 18 ALR 93 (‘R v Judge Leckie’) at 99 (Gibbs J). In contrast to the present case, when a judge rules evidence inadmissible, he or she is generally not required to return to it again.

47    For the purposes of determining the Webb proceeding, the primary judge is required to ignore the evidence of 30 to 33 lay witnesses, an expert, and the submissions made by the parties in the ASIC proceeding, while at the same time reviewing that material in the ASIC proceeding. Likewise, for the purposes of determining the ASIC proceeding, the primary judge is required to ignore the evidence of four experts, and the submissions made by the parties in the Webb proceeding, while at the same time reviewing that material in drafting judgment in the Webb proceeding.

48    The test does not focus on the particular characteristics of the judge. The hypothetical observer looking at the reality of the process might apprehend that it might be difficult for any person, even a professional judge, confronted with different and potentially conflicting evidence and submissions in different proceedings (albeit proceedings directed to the same underlying transactions and issues) to decide the Webb proceeding without the contamination of the extraneous information. As a result the hypothetical observer might reasonably apprehend that the judge might be influenced subconsciously by the extraneous information in deciding the issues in the Webb proceeding.

49    The primary judge accepted that the evidence of the 30 to 33 lay witnesses and of Mr Molony could not be dismissed “as being marginal or insignificant”: J [28]. The primary judge has proceeded on the basis that, notwithstanding the significance of the witness evidence, the fair-minded lay observer would recognise that the documentary evidence was “the most important evidentiary material”: J [33]. We would not assume that the hypothetical observer would expect that evidence to be necessarily preferred insofar as there was a conflict between the documentary and witness evidence. In some instances at least, the witness evidence might be preferred. The choice between documentary and witness evidence is not always clear cut. Documentary evidence, such as email correspondence for instance, can be ambiguous or confusing and witness evidence can be necessary to understand it and can influence the meanings attributed to it. Further, witness evidence can address matters that are not addressed by documentary evidence, as the primary judge recognised was the case with ASIC’s expert evidence on materiality: see J [33].

50    As we have observed, Mr Webb intends to call two experts in the Webb proceeding. Mr Webb has served four reports setting out the experts’ opinions as to the materiality of the information the subject of the Webb proceeding, which included some of the information that ASIC alleges, in the ASIC proceeding, to be material and required to be disclosed by GetSwift. The hypothetical observer would expect that there might be substantial cross-examination of those experts. The court that hears the Webb proceeding will hear different expert evidence in relation to some of the information alleged by ASIC to be material and required to be disclosed by GetSwift. The opinions admitted into evidence, the cross-examination and the submissions will inevitably differ between the two proceedings. It is clear that the opinion evidence as to materiality adduced in the Webb proceeding will be extraneous information for the purposes of the ASIC proceeding.

51    In this respect, the primary judge concluded (at J [48]) that:

In the course of the initial trial of the class action, some extraneous information to that received in the ASIC proceeding will be received. In particular, expert evidence in the class action will likely be received from four experts who will address matters which are in issue in the ASIC proceeding, particularly in connexion with the question of materiality of alleged nondisclosures. This includes substantive issues on which the Court has already received detailed submissions in the ASIC proceeding, a number of which were not the subject of extensive expert evidence. I accept the question of materiality is a key issue in the ASIC proceeding, but for reasons substantially similar to those that I have already explained in relation to the class action, I do not consider that a reasonable bystander knowing all the circumstances of the case might reasonably apprehend that the judge might not bring an impartial mind to the determination of the ASIC proceeding.

(Emphasis added.)

52    As to what the primary judge had “already explained in relation to the class action”, the primary judge said earlier (at J [33]):

As to the expert evidence of Mr Molony (such as was ultimately admitted), it could only ever provide some assistance in determining the objective, predictive and (one might say commonsense) exercise that s 674 of the [Corporations Act], read together with s 677, requires; specifically, whether the pleaded (omitted) information would, or would be likely to, influence persons who commonly invest in securities in deciding whether to acquire or dispose [of] the relevant shares.

53    The primary judge has taken the view that the expert evidence in general is of insufficient importance to the “key issue” of materiality such that the disparity in the expert evidence admitted in the ASIC proceeding and that proposed to be adduced in the Webb proceeding is insufficient to create an apprehension of bias in relation to his determination of the ASIC proceeding.

54    At a general level, we do not accept that the hypothetical observer would understand any expert evidence to be so unimportant or of little weight. In the present context, the fair-minded lay observer would understand that the question of materiality is a “key issue” in both proceedings and the primary judge considers that to be the case, that expert evidence as to materiality is relevant and would assist a fact-finding judge, and that ASIC and Mr Webb both sought to adduce a substantial volume of expert evidence on the topic of materiality.

55    The hypothetical observer would be unlikely to believe that the expert evidence was of slight or trivial importance. The hypothetical observer would consider that the expert evidence on the “key issue” of materiality might be a substantial aspect of the determination of the largely overlapping issues in dispute in each proceeding. This in itself might lead the fair-minded lay observer to apprehend that the primary judge might be unable to keep the expert evidence compartmentalised so as to not influence his decision, at least subconsciously, in either proceeding.

56    A further reason for caution is that, if the primary judge hears the Webb proceeding, then that may also bear influence on the ASIC proceeding. As we have already observed, the primary judge accepted that “in the course of the initial trial in the class action, some extraneous information to that received in the ASIC proceeding will be received”: J [48]. In particular, in the Webb proceeding the primary judge will receive evidence from four experts (two economists and two professional investors) who did not give evidence in the ASIC proceeding. Further, the primary judge will have heard evidence and submissions in the Webb proceedings on three additional customer agreements which are not the subject of the ASIC proceeding and which the hypothetical observer might reasonably apprehend will influence the primary judge (at least subconsciously), particularly in the event that it is necessary to determine penalties in the ASIC proceeding. If the primary judge proceeds to hear the Webb proceeding and is ultimately disqualified from the ASIC proceeding for that reason, there will obviously be a waste of costs as a result.

57    The contradictor correctly points out that concurrent and consecutive trials are envisaged by the Federal Court Rules 2011 (Cth) and are not uncommon. In this regard, we should say that there is a difference between a concurrent trial and a consecutive trial. In a concurrent trial, the evidence is generally the same in both proceedings (albeit that some limited evidence may not be admissible as against a particular party), and, accordingly, the concern raised in this appeal would be unlikely to arise. In respect of a consecutive trial, we make the following comments.

58    The contradictor contends that there are sufficient legal protections in place to ensure that consecutive trials do not carry the posited risk of a judge departing from the judicial oath, and points to the following:

(a)    the obligation to accord procedural fairness, which is an “essentially practical” (HT v R [2019] HCA 40; (2019) 374 ALR 216 at [18] (Kiefel CJ, Bell and Keane JJ)) and flexible concept, requiring at a minimum that a court “provide each party to proceedings before it with an opportunity to be heard, to advance its own case and to answer, by evidence and argument, the case put against it”: International Finance Trust Company Ltd v New South Wales Crime Commission [2009] HCA 49; (2009) 240 CLR 319 at [54] (French CJ);

(b)    the usual public and professional scrutiny, which is the rationale of the open court principle: Commissioner of Australian Federal Police v Zhao [2015] HCA 5; (2015) 255 CLR 46 at [44] (French CJ, Hayne, Kiefel, Bell and Keane JJ);

(c)    the duty to give reasons, including the requirement to refer to material evidence and make findings about material issues (Waterways Authority v Fitzgibbon; Mosman Municipal Council v Fitzgibbon; Middle Harbour Yacht Club v Fitzgibbon [2005] HCA 57; (2005) 221 ALR 402 at [130]-[131] (Hayne J)) and to provide reasons for making relevant findings of fact: Beale v Government Insurance Office of New South Wales (1997) 48 NSWLR 430 at 443-444 (Meagher JA);

(d)    the right to a full appellate rehearing (see Minister for Immigration and Multicultural Affairs v Jia Legeng [2001] HCA 17; (2001) 205 CLR 507 at [75] (Gleeson CJ and Gummow J) and R v Judge Leckie at 99 (Gibbs J), saying that “if the judge did decide otherwise than in accordance with the evidence, his decision could be reversed on appeal”) which will allow a full assessment against the transcript, evidence and judgment of how the judge has come to the result in each case.

59    The contradictor submits that it is within the framework of these protections that the judge exercises his or her known ability to consider and ultimately decide each case solely on the evidence and argument led in that proceeding. So at the level of appearance, the community ought not to be regarded as so lacking in a proper understanding of the judicial process as to perceive a real and substantial risk of a judge deceiving him or herself into departure from the oath in the posited manner, when no such risk exists in actuality.

60    We accept that there are institutional protections that guard against the risk of a reasonable apprehension of bias and which might be imputed to the reasonable fair-minded lay observer, including the judicial oath and the parties’ rights to a subsequent full appellate hearing. However, these matters neither avert the risk of a reasonable apprehension of bias nor the inconvenience of the interested parties being left to the cost and burden of an appellate process to cure error. This is particularly the case where is a real danger of subconscious bias in the way we have described.

61    The contradictor then made the following written submission:

The absolutist approach is myopic in its focus on only one aspect of justice: the perceived purity of having a fresh judge deal with each separate manifestation of the same controversy. The absolutist approach ignores other fundamental normative priorities, for example: avoiding undue delay, considerations of efficiency and cost, and avoiding inconsistent outcomes on the same or similar questions.

(Footnotes omitted.)

62    This submission introduces notions of case management principles. However, as the primary judge himself recognised, the principle of impartiality will override any case management consideration. Whilst a pragmatic or cost-benefit approach to the work of a judge in both managing cases and making decisions is sometimes to be encouraged, it must be tempered by the rule of law and the importance of upholding confidence in the administration of justice.

63    The second aspect of GetSwift’s submission, as adopted by Mr MacDonald (and ASIC’s similar submission), is that an apprehension of bias arises by reason of prejudgment of issues in the Webb proceeding.

64    We do not delay in considering this aspect in light of our conclusions on the primary argument of GetSwift. However, we do not accept there has been any prejudgment shown to exist at this stage.

65    Where apprehended bias is said to arise because a judge has prejudged matters in a controversy, there is a need to firmly establish a reasonable basis for thinking that the decision maker “might be so committed to a conclusion as to be incapable of persuasion to a different view”: see Cabcharge Australia v Australian Competition and Consumer Commission [2010] FCAFC 111 at [25] (Kenny, Tracey and Middleton JJ) citing Re JRL at 352 (Mason J); Laws v Australian Broadcasting Tribunal (1990) 170 CLR 70 at 100 (Gaudron and McHugh JJ).

66    At the moment there is no previous decision by the primary judge, no articulation of any preliminary views, and no indication of adversity to any one of the parties. The primary judge did acknowledge the possibility that he would form “preliminary views” based upon the ASIC proceeding as to the “likely factual findings to be made in the Webb Proceeding” (J [38]). However, at this stage, and as observed by the primary judge, the hypothetical observer would appreciate that those views “would be preliminary” (J [35]).

67    To the extent GetSwift relies on Centro (No 2), the situation there was described by the trial judge (at [60]) as “unusual”. The trial judge concluded (at [64]) that the reasonable observer might apprehend that he would “find it difficult to ‘start afresh’” in that case, because of the extent and significance of the factual findings actually made in the prior proceeding which were adverse to the interests of PricewaterhouseCoopers and PricewaterhouseCoopers Securities Ltd (together, ‘PwC’) in the later proceeding, and which PwC wished to challenge. This was particularly so, as the trial judge noted, given that a theme of the judgment in the prior proceeding (to which PwC was not a party) had been “the obviousness of the errors” in financial statements that involved PwC as auditor: Centro (No 2) at [63]. The facts of the present situation are quite different.

68    We should record that the contradictor submits Mr Macdonald is wrong to rely on the apprehended bias that would arise in the determination of the ASIC proceeding if the primary judge heard the Webb proceeding. In other words, they contend that it is only the apprehension of bias that arises in the determination of the Webb proceeding that is relevant for present purposes. This is significant in relation to the evidence and submissions that are likely to be received in the Webb proceeding that go beyond what was received in the ASIC proceeding, including the expert evidence. The contradictor’s submission is put on the basis that no application was brought in the ASIC proceeding for the judge to disqualify himself and Mr Macdonald’s submission is premature.

69    In view of our conclusions, it is unnecessary to say anything further on this particular submission.

70    The order sought before the primary judge was that the Webb proceeding be referred to the National Operations Manager for the allocation of a new trial judge, thereby curing the apprehension of bias that would otherwise arise were the same judge to determine each proceeding. That approach was preferable to seeking that the primary judge disqualify himself from the ASIC proceeding after he had heard the trial and reserved judgment.

COSTS

71    As to costs of the appeal, GetSwift, Mr Macdonald and ASIC agree that, whatever the result, there should be no order as to costs.

72    As to costs of the contradictor, GetSwift, Mr Macdonald and ASIC had been proceeding on the following basis:

(a)    there was no order in relation to costs of the contradictor in the orders of 26 November 2020 appointing the contradictor (as would usually be the case);

(b)    no approach had been made by the contradictor to them concerning the payment of costs prior to the commencement of the appeal; and

(c)    GetSwift, Mr Macdonald and ASIC therefore understood that the contradictor were appearing pro bono.

73    In those circumstances, GetSwift, Mr Macdonald and ASIC had not expected to have to meet the costs of the contradictor and are not in a position to consent to do so. They have, however, enquired as to amount of those costs and been informed that the contradictor incurred $72,000 plus GST in costs, but that in the circumstances only $40,000 plus GST will be charged.

74    On that basis, GetSwift, Mr Macdonald and ASIC would not seek to be heard further if the Court was minded to order that the parties to the appeal (including Mr Webb) and ASIC each pay one quarter of the costs of the contradictor. This would accord with the agreed approach to costs of the appeal itself (which is that each party or interested person should bear its own costs) and reflect the fact that each of the parties and ASIC adopted a materially similar position on the outcome of the appeal.

75    The position of Mr Webb is that, whatever the result, there should be no order as to costs.

76    As to costs of the contradictor, Mr Webb submits that he had been proceeding on the same basis as GetSwift, Mr Macdonald and ASIC. In those circumstances, Mr Webb had not expected to have to meet the costs of the contradictor and is not in a position to consent to do so.

77    Mr Webb additionally submits that, if a costs order is to be made in relation to the costs of the contradictor, he should not be liable for any portion of the costs, given he did not appeal the judgment and nor did he oppose the appeal. Mr Webb also notes that the contradictor’s written submissions did not address any contention put by Mr Webb, whose own submissions were programmed to be, and were, filed on the same day as those of the contradictor.

78    In this regard, while Mr Webb consents to an order that he bear his own costs, he resists an order that he bear any part of the costs of the contradictor.

79    We consider that all the participants should bear equally the costs of the contradictor despite the expectations of the parties, and particularly Mr Webb. It is true that Mr Webb neither opposed nor supported the appeal. However, Mr Webb did file written submissions in support of a particular outcome (namely, assigning the Webb proceeding to the docket of a new judge) and opposing the alternative relief sought by ASIC, and oral submissions were also made to this effect.

80    In essence, all the interested parties did not support the primary judge’s approach and adopted a materially similar position on the outcome of the appeal. We will make no order as to costs other than that the parties to the appeal (including Mr Webb) and ASIC each pay one quarter of the costs of the contradictor set at $40,000 plus GST.

Associate:

Dated:    5 March 2021