This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.

In the court OF APPEAL of the republic of singapore
[2018] SGCA 61
Civil Appeal No 189 of 2017
Between
BOI
… Appellant
And
BOJ
… Respondent
In the matter of HCF/Divorce Transfer No 6179 of 2013
Between
BOI
… Plaintiff
And
BOJ
… Defendant
grounds of decision
[Courts and Jurisdiction] — [Judges] — [Recusal]
This judgment is subject to final editorial corrections approved by the court and/or redaction pursuant to the publisher’s duty in compliance with the law, for publication in LawNet and/or the Singapore Law Reports.
BOI

v

BOJ
[2018] SGCA 61
Court of Appeal — Civil Appeal No 189 of 2017
Sundaresh Menon CJ, Andrew Phang Boon Leong JA and Judith Prakash JA
26 March 2018
4 October 2018
Andrew Phang Boon Leong JA (delivering the grounds of decision of the court):
Introduction
1 The field of litigation is a potential hotbed for excessive friction and, on the rare occasion, even the flaring of tempers and the engendering of ill-will. Fortunately, this is rare. Unfortunately, the present application allegedly involves one such occasion. It concerns a broken and severely fractured relationship between former spouses. Counsel for the Appellant (the ex-wife), Mr Thomas Sim, described it as a “high-conflict case”. But, indeed, he went further in this application: he argued that the High Court judge (“the Judge”) did not give his client a fair opportunity to put forward her case through her then counsel (his colleague). The Appellant therefore wanted the Judge to recuse herself from the case. An application was brought after consultation with her lawyers. The Judge heard and rejected the application. That led to the present appeal. We dismissed the appeal and now give the detailed grounds for our decision.
2 At the outset, two general but important observations, which also apply to the present case, are in order.
3 First, counsel are not the mere “mouthpieces” of their clients. They are not mere automatons, executing every instruction of the client, especially where the client wants each and every point to be taken in order to inflict maximum “damage” on the other party, and where the taking of such points is – in a word – pointless and would not only engender a wastage of the other party’s, but also the court’s, time and resources. There is a reason why lawyers are also known as “counsel” – in such situations, lawyers must counsel their clients and apprise them of what is permissible and what is not. We operate within an adversarial system. However, as the learned Lord Chief Justice Cockburn observed in an extra-judicial address (see George P Costigan Jr, “The Full Remarks on Advocacy of Lord Brougham and Lord Chief Justice Cockburn at the Dinner to M Berryer on November 8, 1864” (1931) 19 Cal L Rev 521 at p 523), which our courts have endorsed and recapitulated on several occasions (most recently by this Court in Goh Seng Heng v Liberty Sky Investments Ltd and another [2017] 2 SLR 1113 at [62]):
My noble and learned friend, Lord Brougham ... said that an advocate should be fearless in carrying out the interests of his client; but I couple that with this qualification and this restriction—that the arms which he wields are to be the arms of the warrior and not of the assassin. It is his duty to strive to accomplish the interests of his clients per fas, but not per nefas; it is his duty, to the utmost of his power, to seek to reconcile the interests he is bound to maintain, and the duty it is incumbent upon him to discharge, with the eternal and immutable interests of truth and justice. [emphasis added]
4 In order to wield the “correct” set of arms, the lawyer, whilst owing a duty to his or her client, obviously cannot be the mere conduit pipe of that client. He also owes a duty – and indeed a paramount one – to the court. A lawyer has to tread a fine line when adhering to these occasionally inconsistent duties in practice, but that is the very essence of being a legal professional.
5 On a related note, the lawyer also needs to guard against his or her own bias, which may be subconscious and insidious. In particular, the lawyer must guard against taking on, especially subconsciously, the persona of his or her client. This is all the more so in cases where the parties’ emotions run high. There is sound logic in this because a lawyer who takes on the persona of his or her client is also likely to lose his or her objectivity and sense of perspective. And such loss often results (ironically) in advice and possibly even outcomes that are ultimately to the client’s detriment.
6 Secondly, whilst patience and rectitude ought to be the ideal at all times, we are all only human. Hence, there must be the proverbial give and take on the part of both judge and counsel alike. This relates to the issue of perspective, and the following oft-cited advice from a father to his daughter in a famous novel ought to be noted (see Harper Lee, To Kill A Mockingbird (William Heinemann Ltd, 1960; reprinted in the New Windmill Series, 1966) at p 35):
First of all, … if you can learn a simple trick, Scout, you’ll get along a lot better with all kinds of folks. You never really understand a person until you consider things from his point of view … until you climb into his skin and walk around in it.
7 This sage advice was quoted by the Court of Three Judges in Singapore Medical Council v Wong Him Choon [2016] 4 SLR 1086 (at [4]) in relation to the need for a doctor to be cognisant of the patient’s perspective as well, but it embodies a universal truth that applies perhaps a fortiori to the present case. In our view, both judge and counsel could have been more patient with each other had they stepped into each other’s shoes even if for a brief moment – with the lawyers concerned simultaneously realising that, if they had inadvertently stepped into their client’s shoes, they ought to have stepped out of them with some haste.
8 That having been said, impatience has never been a ground for recusal of a judge. The hope is that everyone is an angel but the truth is that we are all only human and ought to bear, as far as possible, with each other’s foibles and weaknesses. One cannot help but feel that had there been more give and take in the present case, it would not have escalated into the proceedings that we now have before us. Perhaps some good can come out of this unfortunate situation if judges and lawyers (and, indeed, all who are involved in the discipline of law) learn the lesson, if not of perfect patience, then at least of the need for perspective as well as for self-restraint that has the quest for patience at its foundation, especially when every fibre within oneself seeks (or even cries out) to conduct oneself otherwise. Notwithstanding the fact that it is often bandied about out of context, there is much truth in the old adage that one ought to at least disagree without being disagreeable.
9 With these observations in mind, we now turn to the case proper.
The factual backdrop
10 The parties to the present appeal are former spouses. The Appellant commenced divorce proceedings on 23 December 2013 and interim judgment was granted on 4 February 2014.
11 The ancillary matters were scheduled to be heard before the Judge on 9 November 2016, 8–9 February, 18 April, 21 April, and 20–21 June 2017. A further full day hearing date was fixed on 11 July 2017.
12 On 7 July 2017, the Appellant filed Summons No 240 of 2017 (“SUM 240”) in the High Court seeking the recusal of the Judge from hearing the ancillary matters, and for another judge to hear the same. The application for recusal was premised on the Judge’s alleged conduct at the hearings on 18 April, 21 April, 20 and 21 June 2017.
13 On 2 October 2017, the Judge heard the parties and dismissed SUM 240.
14 In the parties’ submissions below, the Appellant’s main ground for recusal was that the Court had been infected by apparent bias. The Appellant elaborated that this was because her counsel had been unduly hampered in the presentation of her case before the Judge; the Judge had favoured the Respondent by giving his counsel more time and greater leeway; and the Judge had prejudged the case against the Appellant without having heard the parties fully.
15 The Respondent submitted that apparent bias was not made out on the facts. The Respondent added that the recusal application was a backdoor appeal prematurely brought before any order was made by the Judge because the Appellant was unhappy with the Judge’s factual findings. Further, as a matter of public policy, the recusal application should not be allowed as it would condone “judge shopping”.
16 At the time this appeal was heard, the hearing of the ancillary matters in the divorce suit had not concluded and no orders had been made.
The decision below
17 The Judge saw no reason to recuse herself. In her unpublished grounds of decision, she began by describing the general procedure in ancillary matters hearings. Unlike trials, ancillary matters hearings do not involve cross-examination of parties and witnesses. Instead, evidence for such hearings would comprise affidavits of assets and means as well as joint summaries of relevant information. The evidence is supplemented by written submissions. All the court is concerned with is the objective evaluation of what has been placed before it. The court can therefore adopt a more interventionist judge-led approach. This, according to the Judge, is reinforced by r 22 of the Family Justice Rules 2014 (GN No S 813/2014) (“the FJR”), which provides as follows:
Power to make orders and give directions for just, expeditious and economical disposal of proceedings
22.—(1) Despite anything in these Rules, the Court, when dealing with any cause or matter, is to adopt a judge‑led approach —
(a) to identify the relevant issues in the cause or matter; and
(b) to ensure that the relevant evidence is adduced by the parties to the cause or matter.
(2) In adopting a judge‑led approach, the Court may, at any time after the commencement or at the hearing of any proceedings, of its own motion or on an application by any party to the proceedings, direct any party or parties to those proceedings to appear before it, for the Court to make such order or give such direction as it thinks fit, for the just, expeditious and economical disposal of the cause or matter.
…
The Judge added that the court has wide powers to manage the flow of the proceedings, including the power to limit time for oral arguments.
18 Next, the Judge rejected the allegation of prejudgment on her part. She explained that after she had reviewed each issue with counsel, she would sometimes indicate her finding or likely findings. In any case, there were still issues that had not been covered, and she could not have prejudged the outcome of the case since the factual matrix would have to be viewed in its entirety in order to arrive at a decision on the division of matrimonial assets and maintenance.
19 The Judge also made some observations on counsel’s conduct of the case. Acknowledging that she had been less than patient with lead counsel for the Appellant below, Ms Linda Joelle Ong (“Ms Ong”), she also commented that Ms Ong tended to prevaricate and not give direct answers to questions or issues posed. Additionally, Ms Ong frequently engaged in whispered exchanges with her assistants and spent time looking for documents and affidavits. On the other hand, counsel for the Respondent below, Ms Judy Ang (“Ms Ang”), was direct in answering questions and upfront in acknowledging any point that she had not covered or was not sure about.
20 The Judge further explained that her approach was to pose questions in order to furnish counsel the opportunity to specifically address her concerns, and to provide clarification whenever such clarification was required.
The issues
21 We first set out the gist of the parties’ cases on appeal. We will delve into further details where appropriate.
22 The Appellant’s case was essentially two-pronged. First, the Appellant argued that the Judge was apparently biased against her and that she had wrongly applied the “real danger” or “real likelihood” test, as opposed to the “reasonable suspicion” test, in determining whether apparent bias had been made out.
23 Secondly, the Appellant reiterated largely the same arguments on apparent bias as she did before the Judge below:
(a) The Judge had unduly obstructed and hampered Ms Ong from effectively presenting the Appellant’s case.
(b) The Judge had treated the parties unequally by giving Ms Ang more leeway and time.
(c) The Judge had prejudged the case in a manner adverse to the Appellant and was not open to hearing evidence that did not conform to her pre-formed views. On one occasion, the Judge led Ms Ang to make certain arguments that purportedly reached the conclusion that the Judge had already made.
24 The Respondent refuted the above arguments, reiterating in particular the policy argument that the Appellant’s unfounded allegations of bias should not be allowed as they would create doubt in the public’s mind about the impartiality of the judiciary and hinder the effective administration of justice.
25 We will address three grounds for recusal, although we note from the outset that the Appellant subsumed the second and third grounds under an overarching argument in relation to apparent bias:
(a) whether the doctrine of apparent bias was made out;
(b) whether the Judge had predetermined the issue; and
(c) whether the Judge interfered excessively in the proceedings, thus denying the Appellant a fair opportunity to present her case.
26 Given that these grounds overlap to some degree and, as counsel for the Appellant noted in his oral submissions, the arguments made were interrelated points, we will begin by setting out the applicable legal principles before applying them to the facts.
The applicable principles
Apparent bias
27 The Appellant’s first argument that the Judge applied the wrong test for apparent bias can be summarily dismissed as the Judge explicitly referred to and applied the “reasonable suspicion” test. It is well-established that this is the applicable test in determining whether the doctrine of apparent bias is made out, ie, whether “there are circumstances which would give rise to a reasonable suspicion or apprehension in a fair-minded reasonable person with knowledge of the relevant facts that the [decision-maker] was biased” (see the Singapore High Court decision of Re Shankar Alan s/o Anant Kulkarni [2007] 1 SLR(R) 85 (“Re Shankar Alan”) at [91]).
28 Be that as it may, the implicit premise of the Appellant’s first argument is that there is a distinction between the two tests that have traditionally been applied for apparent bias – the “reasonable suspicion of bias” test on the one hand and the “real likelihood of bias” test on the other. As to whether there exists any material difference between the two tests, there have been two local High Court decisions that seemingly conflict: Tang Kin Hwa v Traditional Chinese Medicine Practitioners Board [2005] 4 SLR(R) 604 (“Tang Kin Hwa”) and Re Shankar Alan. Given the apparent divergence in local jurisprudence, we consider it appropriate to take this opportunity to address this apparent conflict that has been in existence for over a decade.
29 In essence, the conflict between Tang Kin Hwa and Re Shankar Alan is more apparent than real if one considers that the “real likelihood” test examined in Tang Kin Hwa is not the test set out in R v Gough [1993] AC 646 (“Gough”) but, rather, the test set out in Porter v Magill [2002] 2 WLR 37 (“Porter”) – bearing in mind the fact that Re Shankar Alan disapproved of the test set out in Gough. As we will elaborate on below, Gough stands for the proposition that apparent bias is to be ascertained from the perspective of the court and that it is unnecessary to look at a particular matter through the eyes of a reasonable man. On the other hand, in Porter, the test for apparent bias is whether the fair-minded and informed observer, having considered the facts, would conclude that there was a real possibility that the tribunal was biased. As such, as between the test in Porter (which was endorsed in Tang Kin Hwa) and the “reasonable suspicion” test (endorsed in Re Shankar Alan), there is a fundamental commonality in perspective: under both tests, the court is to adopt the vantage point of a reasonable observer. In our view, once it is appreciated that there is this commonality in perspective, there is no longer a significant difference between the two tests (see generally [90]−[95] below).
30 We will begin by summarising Tang Kin Hwa and Re Shankar Alan. Thereafter, we will undertake a brief survey of the respective legal positions in various Commonwealth jurisdictions in order to understand the relevant legal background against which the two conflicting local decisions were handed down and, more importantly, why this conflict is, in the final analysis, more apparent than real. This survey will also be helpful in elaborating on how the reasonable observer construct is to be applied, and we will conclude by summarising and restating the applicable position in Singapore today.
Tang Kin Hwa and Re Shankar Alan
31 In Tang Kin Hwa, the High Court referred to both the “reasonable suspicion of bias” test and the “real likelihood of bias” test. It recognised that the prevailing test in Singapore was whether “a reasonable and right-thinking person sitting in court and knowing the relevant facts would have any reasonable suspicion that [in that particular case] a fair trial for the appellant was not possible” (at [16]), although it had been left open in Tang Liang Hong v Lee Kuan Yew [1997] 3 SLR(R) 576 (“Tang Liang Hong”) whether the allegedly more stringent “reasonable likelihood of bias” test should apply (at [18]). Since neither test was satisfied on the facts of Tang Kin Hwa, the court did not need to take a definitive view on which test should be preferred (at [20] and [46]).
32 The court nevertheless expressed a “tentative view” that there was no real conflict between the two tests (at [24]), and that the “common substance” of both tests was “whether or not there was a perception on the part of a reasonable person that there would be a real likelihood of bias” (at [39]). The court’s reasoning may be further broken down into four parts.
33 First, the focus in Gough was on the concept of “possibility” rather than “probability” (at [26]), and this standard of “possibility” was reflected as well in the “reasonable suspicion” test (at [39]). Secondly, both tests were premised on an objective basis (at [36]). As Leggatt J observed in the English High Court decision of Cook International Inc v B V Handelmaatschappij Jean Delvaux and Braat, Scott and Meadows [1985] 2 Lloyd’s Rep 225 (“Cook International”) at 231, the true contrast was between “reasonable suspicion of bias” and “the appearance of a real likelihood of bias”, both of which were concerned with apparent rather than actual bias; when the “real likelihood” test was expressed in such a fashion, there was little (if any) difference between the two tests (Tang Kin Hwa at [37] and [39]). Thirdly, one ought not to draw a sharp distinction between the court’s perspective and that of the public. Both were “integral parts of a holistic process” since the court would have to ascertain the perspective of the public and, to that extent, “personifie[d]” the reasonable man (at [40]). Equally, the court could not replace the reasonable man and should not inappropriately rely upon special knowledge, the minutiae of court procedure, or other matters outside the ken of the ordinary, reasonably well-informed member of the public (at [40], citing the English Court of Appeal decision of Locabail (UK) Ltd v Bayfield Properties Ltd and another [2000] QB 451 (“Locabail”)).
34 The court in Tang Kin Hwa also cautioned that there was a need to guard against gratuitous semantic confusion, and that a practical substance-over-form approach must be taken that takes into account not only the possible meanings of the words and phrases in question but also the context in which they appear (at [39] and [43]).
35 On the other hand, in Re Shankar Alan, a different member of the High Court distinguished the “reasonable suspicion of bias” test from the “real likelihood of bias” test articulated in Gough and stated that Gough was not the law in Singapore (at [53]). The court considered that Gough had “shifted” the inquiry “from one directed at how it might all appear to a reasonable man to whether the judge thinks there is in fact a sufficient possibility of bias” (at [62]). The emphasis of the “real likelihood of bias” test was “whether the court thinks there is a real likelihood … that the tribunal is biased” (at [66]), which went towards the issue of actual bias instead of apparent bias. This was a “very significant point of departure”, even assuming that “likelihood” was to be equated with “possibility” (at [62] and [69]).
36 The court in Re Shankar Alan therefore disagreed with the observation in Tang Kin Hwa that there was no practical difference between the two tests. That court highlighted two differentiating aspects: of perspective (ie, whether bias was assessed from the viewpoint of the court or that of a reasonable member of the public) and of substance (ie, whether the test concerned the degree of possibility of actual bias or the appearance of bias to an observer). The “real likelihood of bias” test would be met as long as a court was satisfied of a sufficient degree of possibility of bias. In contrast, the “reasonable suspicion of bias” test would be met as long as a reasonable member of the public could harbour a reasonable suspicion of bias, even if there was no real danger of actual bias on the facts. The former test was directed at mitigating the difficulty of proving bias, whereas the latter test was driven by the public interest in ensuring public confidence in the administration of justice (at [74]–[75]).
37 The court in Re Shankar Alan then expressed its preference for the latter test, and the reasons provided for its preference include judicial support for the “reasonable suspicion” test; the fact that the “reasonable suspicion” test made no adverse finding on whether the tribunal in question was actually biased (thereby avoiding the possibility that the “real likelihood” test could create public disquiet in the administration of justice); the imprecision inherent in the “real likelihood” test (which the court found to be impossibly subjective); and the fact that the “reasonable suspicion” test focuses on the real mischief and inquiry, that is, whether the fair-minded member of the public would reasonably entertain a suspicion or apprehension of bias (at [81], [83] and [84]).
38 As evident from the above account, the two cases differ in respect of whether there is any meaningful contrast to be drawn between the “reasonable suspicion” and “real likelihood” tests. However, as already alluded to above (at [29]), this difference may be more apparent than real – a point to which we return to elaborate upon in much greater detail (see generally [90]−[95] below) after having first surveyed the test for apparent bias in the various Commonwealth jurisdictions. And it is to this last-mentioned task that our attention now turns.
A brief survey of Commonwealth jurisdictions
(1) Australia
39 The “reasonable suspicion of bias” test is well-established in Australia: see, eg, the High Court of Australia decisions of Dickason v Edwards and others (1910) 10 CLR 243 at 256–257 (per O’Connor J) and at 258 (per Isaacs J); as well as R v Commonwealth Conciliation and Arbitration Commission; Ex parte Angliss Group (1969) 122 CLR 546 at 553–554.
40 In the landmark decision of The Queen v Watson; Ex parte Armstrong (1976) 136 CLR 248 (“Armstrong”), a majority of the High Court of Australia noted (at 258) a “difference of opinion” on whether the “reasonable suspicion of bias” test or the “real likelihood of bias” test governed the doctrine of apparent bias. In a joint judgment by Barwick CJ, Gibbs, Stephen and Mason JJ, the majority reviewed the Australian cases cited earlier at [39] and undertook a comprehensive review of all the authorities (including those of the Supreme Courts of the various States in Australia) up to the time of the decision (at 260–262). Having done so, the court took the position that it would suffice to show a reasonable suspicion of bias against a judge, and that this was correct in principle as it was of fundamental importance that the public have confidence in the administration of justice (at 262–263).
41 In the subsequent decision of Livesey v The New South Wales Bar Association (1983) 151 CLR 288 (“Livesey”), the High Court of Australia affirmed its earlier decision of Armstrong and observed (at 293–294) that:
It was common ground between the parties to the present appeal that the principle to be applied in a case such as the present is that laid down in the majority judgment in Reg v Watson; Ex parte Armstrong. That principle is that a judge should not sit to hear a case if in all the circumstances the parties or the public might entertain a reasonable apprehension that he might not bring an impartial and unprejudiced mind to the resolution of the question involved in it ...
42 Subsequently, in the High Court of Australia decision of Vakauta v Kelly (1989) 167 CLR 568 (“Vakauta”), Dawson J cited Armstrong and Livesey and observed as follows (at 575):
The relevant principle is that laid down in Reg v Watson; Ex parte Armstrong, and applied in Livesey v New South Wales Bar Association, namely, that a judge should not sit to hear a case if in all the circumstances the parties or the public might entertain a reasonable apprehension that he might not bring an impartial and unprejudiced mind to the resolution of the question involved in it …
43 In Johnson v Johnson (2000) 201 CLR 488 (“Johnson”), Gleeson CJ, Gaudron, McHugh, Gummow and Hayne JJ in the High Court of Australia issued a joint judgment endorsing the position that the court had to be satisfied that “a fair-minded lay observer might reasonably apprehend that the judge might not bring an impartial and unprejudiced mind to the resolution of the question the judge is required to decide” (at [11]). The court went on to elaborate on what characteristics this “lay observer” would bear (at [12]−[13]):
12 … At the same time, two things need to be remembered: the observer is taken to be reasonable; and the person being observed is “a professional judge whose training, tradition and oath or affirmation require [the judge] to discard the irrelevant, the immaterial and the prejudicial”.
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 … Judges, at trial or appellate level, who, in exchanges with counsel, express tentative views which reflect a certain tendency of mind, are not on that account alone to be taken to indicate prejudgment. Judges are not expected to wait until the end of a case before they start thinking about the issues, or to sit mute while evidence is advanced and arguments are presented. …
44 In the same case, Kirby J also observed as follows, which we do not consider substantively dissimilar from the position taken by the majority (at [48]–[49]):
48 However, the test which the law calls for interposes a fictitious bystander. It hypothesises a person whose judgment in the matter is taken to be determinative. One might consider that the fiction should not be taken too far. Indeed, it is important to reserve to the appellate court a capacity to review the facts and the complaints having regard to the “serious and sensitive issues” raised by an allegation of prejudgment.
49 Nevertheless, the interposition of the fictitious bystander and the adoption of a criterion of disqualification expressed in terms of possibilities rather than “high probability” are both intended to serve an important social interest which must be restated in disposing of this appeal. Each of these considerations lays emphasis on the need to consider the complaint made ultimately, not by what adjudicators and lawyers know, but by how matters might reasonably appear to the parties and to the public. That is why one does not attribute to the fictitious bystander highly specialised knowledge, such as that known perhaps to only some lawyers concerning the supposed inclinations and capacities of a particular adjudicator. It is also why it would be a mistake for a court simply to impute all that was eventually known to the court to an imaginary reasonable person because to do so would be only to hold up a mirror to itself.
[emphasis added]
Kirby J also elaborated on the attributes that the fictitious bystander ought to take (at [53]):
53 The attributes of the fictitious bystander to whom courts defer have therefore been variously stated. 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. The bystander would be taken to know commonplace things, such as the fact that adjudicators sometimes say, or do, things that they might later wish they had not, without necessarily disqualifying themselves from continuing to exercise their powers. The bystander must also now be taken to have, at least in a very general way, some knowledge of the fact that an adjudicator may properly adopt reasonable efforts to confine proceedings within appropriate limits and to ensure that time is not wasted. The fictitious bystander will also be aware of the strong professional pressures on adjudicators (reinforced by the facilities of appeal and review) to uphold traditions of integrity and impartiality. Acting reasonably, the fictitious bystander would not reach a hasty conclusion based on the appearance evoked by an isolated episode of temper or remarks to the parties or their representatives, which was taken out of context. Finally, a reasonable member of the public is neither complacent nor unduly sensitive or suspicious. [emphasis added in italics and bold italics]
45 In Ebner v Official Trustee in Bankruptcy (2000) 205 CLR 337 (“Ebner”), the High Court of Australia refined its approach and recognised two steps to the test for apparent bias: first, identifying what it was said might lead the judge in question to decide the case on a basis other than its factual and legal merits; and second, articulating the logical connection between the matter and the feared deviation from the course of deciding the case on its merits (at [8] and [72]). The “reasonable suspicion of bias” test, as elaborated on in Johnson and Ebner, was subsequently applied in more recent decisions of the High Court of Australia, such as Concrete Pty Ltd v Parramatta Design & Developments Pty Ltd and another (2006) 229 CLR 577 at [110] and British American Tobacco Australia Services Limited v Laurie and others (2011) 242 CLR 283 (“Laurie”) at [37].
46 Of special relevance, however, to the specific issue of the relationship between the “reasonable suspicion of bias” test and the “real likelihood of bias” test is the decision of the High Court of Australia in Webb v The Queen (1994) 181 CLR 41 (“Webb”) where the court (not surprisingly) endorsed the “reasonable apprehension of bias” test and where Mason CJ and McHugh J, in a joint judgment, observed thus (at 50−51):
In considering the merits of the test to be applied in a case where a juror is alleged to be biased, it is important to keep in mind that the appearance as well as the fact of impartiality is necessary to retain confidence in the administration of justice. Both the parties to the case and the general public must be satisfied that justice has not only been done but that it has been seen to be done. Of the various tests used to determine an allegation of bias, the reasonable apprehension test of bias is by far the most appropriate for protecting the appearance of impartiality. The test of “reasonable likelihood” or “real danger” of bias tends to emphasize the court’s view of the facts. In that context, the trial judge’s acceptance of explanations becomes of primary importance. Those two tests tend to place inadequate emphasis on the public perception of the irregular incident. [emphasis added in italics and bold italics]
47 Deane J also observed the difference in perspective, holding that the standard to be observed was that of a hypothetical fair-minded and informed lay-person (at 68). He added that the adoption of a “real likelihood” or “real danger” test would, in effect, substitute the apparent bias test with a test of actual bias attenuated by a modified standard of proof (at 71).
48 On the issue of perspectives, the court in Laurie, citing Johnson, commented as follows (at [48], per French CJ (dissenting, albeit not on this particular point)):
The interposition of the fair-minded lay person could never disguise the reality that it is the assessment of the court dealing with a claim of apparent bias that determines that claim. As Professor Olowofoyeku says:
In the end, despite the pitch on objectivity and the view that the apprehensions of bias must have an objective basis, it is the opinion of the reviewing court on this issue that matters.
Professor Olowofoyeku has expressed the view that the judicial construct of the informed observer no longer provides a reliable guide to decision-making on the issue of apparent bias. However, the utility of the construct is that it reminds the judges making such decisions of the need to view the circumstances of claimed apparent bias, as best they can, through the eyes of non-judicial observers. In so doing they will not have recourse to all the information that a judge or practising lawyer would have. It requires the judges to identify the information on which they are to make their determinations. …
[emphasis in bold italics]
49 The references above to the concept of perspectives (see, in particular, the observations by Kirby J in Johnson, as quoted above) are important, as this was one of the key points which formed the basis of the apparent conflict between the two Singapore cases as to what was the precise relationship between the “reasonable suspicion of bias” test on the one hand and the “real likelihood of bias” test on the other (see Re Shankar Alan at [65]).
(2) New Zealand
50 The position in New Zealand is that there is no practical difference between the two tests. In Auckland Casino Ltd v Casino Control Authority [1995] 1 NZLR 142 (“Auckland Casino Ltd”), the New Zealand Court of Appeal downplayed the conflict between the English case of Gough (which stood for the real likelihood of bias test) and Webb (which stood for the reasonable suspicion of bias test) (at 149):
The approach that has been adopted in this Court in recent years, however, has been to emphasise that there is little if any practical difference between the two tests. See E H Cochrane Ltd v Ministry of Transport [1987] 1 NZLR 146, 153; R v Te Pou [1992] 1 NZLR 522, 527; Matua Finance Ltd v Equiticorp Industries Group Ltd [1993] 3 NZLR 650, 654. References to earlier New Zealand cases will be found in the three cases cited. In some of them the possibility of a genuine distinction has been recognised. But once it is granted that the hypothetical reasonable observer must be informed, so that as indicated by the House of Lords in Gough at pp 664 and 673 R v Sussex Justices, ex parte [McCarthy] [1924] 1 KB 256 is dubious authority, the distinction becomes very thin. If a reasonable person knowing all the material facts would not consider that there was a real danger of bias, it would seem strained to say that nevertheless he or she would reasonably suspect bias. One must query whether the law should countenance such refinements. In the result we accept the real danger test as satisfactory. [emphasis added in bold italics]
The court in Auckland Casino Ltd therefore took the position that there was little separating the two tests when one considered that the reasonable observer had to have knowledge of all material facts, and it was on that basis that the court in Auckland Casino Ltd concluded that the two tests were closely similar (see also the Supreme Court of New Zealand decision of Saxmere Company Ltd v Wool Board Disestablishment Company Ltd [2009] 1 NZLR 35 (“Saxmere”) at [70]).
51 In Man O’War Station Ltd v Auckland City Council [2001] 1 NZLR 552, the New Zealand Court of Appeal affirmed the test set out in Auckland Casino Ltd and Gough, holding that the material question was whether a fully informed court personifying the reasonable observer would perceive a real danger that the judge unfairly regarded a party’s case with disfavour (at [13], [19] and [29]). The appellants then appealed to the Privy Council, where the appellants urged the Board to adopt the modified Gough test enunciated by Lord Hope of Craighead in Porter ([29] above) which emphasised the perspective of the fair-minded observer rather than the view of the court. The Privy Council declined to do so on the basis that it did not have the benefit of the New Zealand Court of Appeal’s opinion on this, but went on to comment that any distinction was a “fine one”, since the application of the original Gough test would undoubtedly require the court to take into account public perception and confidence (see Man O’War Station Ltd v Auckland City Council (Judgment No 1) [2002] 3 NZLR 577 at [10]).
52 In Muir v Commissioner of Inland Revenue [2007] 3 NZLR 495, the New Zealand Court of Appeal noted the messy state of the law in the area (at [56]), and sought to “extinguish the tenuous hold on existence the Gough test has had in New Zealand”, preferring the two-step test set out in the Australian case of Ebner (at [60], [62] and [94]; for the two-step test, see [45] above).
53 Subsequently, the Supreme Court of New Zealand in Saxmere expressed the view that after some “semantic differences” (ie, any differences before that point were semantic), the English position (citing Porter) and the Australian position (citing Ebner) had become essentially the same (at [3]). The court cited (at [68]) the excerpt in Auckland Casino Ltd cited above at [50] and went on to note that there was no significant difference, in reality, between the test of a real possibility of bias (the test in England), and a reasonable apprehension of bias (the test in Australia), given that they both emphasised the need for the fair-minded observer to be fully informed of all relevant circumstances, including knowledge of the specialist adversarial context of actual litigation, the role played by judges and barristers, as well as the “legal traditions and culture of [the] jurisdiction” (at [75]–[76]). At this particular juncture, it is important to note that the court in Saxmere cited Porter as representing the English position. This harks back to our earlier observation (at [29]) that there is a fundamental commonality in perspective as between the test in Porter (as opposed to that in Gough) on the one hand and the “reasonable suspicion” test on the other – a point that is of critical importance in understanding why the controversy noted earlier in the Singapore context is more apparent than real.
54 Having expressed this view, the court in Saxmere then went on to unequivocally accept the “reasonable apprehension of bias” test as articulated in Ebner. The court considered that this test was sufficiently wide in scope and would provide appropriate protection for, and was also convenient for, the administration of justice (at [90]); was aligned with the principles of natural justice and international human rights jurisprudence (at [91]); and gave significant weight to the need for public confidence in the integrity of the judicial system (at [92]). The court contrasted the “reasonable apprehension of bias” test to its main alternative, the “real danger of actual bias” test. In the court’s view, the latter test was highly difficult for the judge in question to apply to his or her own situation and it carried the risk of unfair damage to the judge’s reputation if he or she was reversed on appeal (at [90]).
55 Saxmere was later relied upon in the Supreme Court of New Zealand decisions of Siemer v Heron [Recusal] [2011] 1 NZLR 293 at [11] and in A v The Queen [2016] NZSC 31 at [16] as authority for applying the reasonable apprehension of bias test as set out in Ebner.
(3) Canada
56 The “reasonable apprehension of bias” test appears to be the applicable test in Canada. In the oft-cited dissenting judgment of Martland, Judson and de Grandpré JJ in the Supreme Court of Canada decision of Committee for Justice & Liberty v Canada (National Energy Board) [1978] 1 SCR 369, delivered by de Grandpré J, it was held (at [40]) that:
The proper test to be applied in a matter of this type was correctly expressed by the Court of Appeal. … [T]he apprehension of bias must be a reasonable one, held by reasonable and right minded persons, applying themselves to the question and obtaining thereon the required information … the test is “what would an informed person, viewing the matter realistically and practically – and having thought the matter through – conclude …”
We note that this test laid down by de Grandpré J was, in fact, also endorsed by the majority of the court as well and, as we shall see below, has been endorsed and applied since.
57 Notably, de Grandpré J went on to comment that he saw “no real difference” between the expressions “reasonable apprehension of bias”, “reasonable suspicion of bias” and “real likelihood of bias” (at [41]). Grounds for the apprehension had to be “substantial”, although de Grandpré J does not seem to have addressed the ostensible difference in perspectives underlying the two approaches.
58 The “reasonable apprehension of bias” test set out by de Grandpré J was affirmed in the subsequent Supreme Court of Canada decision of R v S (RD) [1997] 3 SCR 484 (“R v S (RD)”) (see the decision of L’Heureux-Dubé and McLachlin JJ at [31] and the judgment of Cory J (with which Iacobucci J concurred) at [111]). It bears mention that Cory J agreed (at [112]−[113]) with the observation by de Grandpré J that there was “no real difference” between the tests of “real likelihood of bias” and “reasonable apprehension of bias”:
112 The appellant submitted that the test requires a demonstration of “real likelihood” of bias, in the sense that bias is probable, rather than a “mere suspicion”. This submission appears to be unnecessary in light of the sound observations of de Grandpré J in Committee for Justice, supra, at pp 394–95:
I can see no real difference between the expressions found in the decided cases, be they ‘reasonable apprehension of bias’, ‘reasonable suspicion of bias’, or ‘real likelihood of bias’. The grounds for this apprehension must, however, be substantial and I agree with the Federal Court of Appeal which refused to accept the suggestion that the test be related to the “very sensitive or scrupulous conscience”.
Nonetheless the English and Canadian case law does properly support the appellant’s contention that a real likelihood or probability of bias must be demonstrated, and that a mere suspicion is not enough. See R v Justice of Camborne [1954] 2 All ER 850 (Eng QB); R v London Rent Assessment Panel Committee [1968] 3 All ER 304 (Eng CA); R v Gough (1992), [1993] 2 WLR 883 (Eng CA); [R v Bertram (December 5, 1989), Doc Toronto RE 1411/89 (Ont HC)], at p 53; [R v Stark (February 23, 1994) Doc 7270/92 (Ont Gen Div)], at para 74; [R v Gushman (April 22, 1994) Doc U1382/93 (Ont Gen Div)], at para 30.
113 Regardless of the precise words used to describe the test, the object of the different formulations is to emphasize that the threshold for a finding of real or perceived bias is high. It is a finding that must be carefully considered since it calls into question an element of judicial integrity. Indeed an allegation of reasonable apprehension of bias calls into question not simply the personal integrity of the judge, but the integrity of the entire administration of justice … Where reasonable grounds to make such an allegation arise, counsel must be free to fearlessly raise such allegations. Yet, this is a serious step that should not be undertaken lightly.
In relation to the traits that the reasonable man ought to possess, the court in R v S (RD) noted that the reasonable person must be an informed person with knowledge of all the relevant circumstances, including “the traditions of integrity and impartiality that form a part of the background and apprised also of the fact that impartiality is one of the duties the judges swear to uphold”, as well as the social reality that forms the background to a particular case, such as societal awareness and acknowledgement of the prevalence of racism and gender bias in a particular community (at [111]).
59 In Arsenault-Cameron v Prince Edward Island [1999] 3 SCR 851 (“Arsenault”), Bastarache J in the Supreme Court of Canada, citing [112]–[113] of R v S (RD) which we reproduced above, held that “[a] real likelihood or probability of bias must be demonstrated” for apparent bias to be made out (at [2]).
60 It would appear from the excerpts above that the Canadian courts have approached the alleged conflict between the “real likelihood of bias” test and the “reasonable suspicion of bias” test in terms of the threshold for a finding of apparent bias, rather than in terms of (substantive) perspective as the Australian courts have.
61 In Roberts v R [2003] 2 SCR 259 at [60], the Supreme Court of Canada again endorsed the “reasonable apprehension of bias” test (see also at [73] and [82]). This remains the applicable test to date (see, eg, the Supreme Court of Canada decision of Yukon Francophone School Board, Education Area #23 v Attorney General of the Yukon Territory [2015] 2 SCR 282 (“Yukon”) at [20]; the Court of Appeal for British Columbia decision of Lucas Siemens v Graham Alexander Howard [2018] BCCA 197 at [45]; and the Nova Scotia Court of Appeal decision of R v Keats [2018] NSCA 15 at [28]). It should be noted that in Yukon, the Supreme Court of Canada, citing Arsenault, reiterated its position that the test for a reasonable apprehension of bias required a “real likelihood or probability of bias” (at [25]).
(4) South Africa
62 The South African courts have consistently applied the “reasonable suspicion” or “reasonable apprehension” test. The Supreme Court of South Africa (Appellate Division) in BTR Industries South Africa (Pty) Ltd and others v Metal and Allied Workers’ Union and another (1992) 3 SA 673, after considering the “real likelihood of bias” and “reasonable suspicion of bias” tests, preferred the latter, which the court found was “less exacting” (at 690). The court stated that as a matter of policy, it was crucial for the public to have confidence in the courts, and that the requirement of “an appearance of a real likelihood of bias” would cut at the root of the principle that justice must be seen to be done and impede the due administration of justice (at 694).
63 The “reasonable suspicion of bias” test was then applied in a number of subsequent cases, including the Supreme Court of South Africa (Appellate Division) decision of Moch v Nedtravel (Pty) Ltd t/a American Express Travel Service (1996) 3 SA 1 at 8, as well as the Constitutional Court decisions of President of the Republic of South Africa and others v South African Rugby Football Union and others (1999) 4 SA 147 (“South African Rugby Football Union”) at [48] and S v Basson (2005) 12 BCLR 1192 (CC) (“Basson”) at [29] and [31].
64 In Basson, the court also made the observation that judges were not “silent umpires” and should be allowed to ask questions and ensure that the parties conduct their cases properly (at [35]). Further, inappropriate behaviour by a judge would not ordinarily give rise to a reasonable apprehension of bias, unless it clearly arose not from irritation or impatience with the way a case had been litigated, but from what might reasonably be perceived to be bias (at [36]).
(5) Scotland
65 The “reasonable suspicion of bias” test is also well-established in Scotland.
66 In the High Court of Justiciary decision of Bradford v McLeod 1986 SLT 244 (“Bradford”), the Lord Justice-Clerk (Ross) and Lord Hunter considered and adopted (at 247–248) as the law of Scotland the approach taken in the English decision of Law v Chartered Institute of Patent Agents [1919] 2 Ch 276 (at 289) that “if there are circumstances so affecting a person acting in a judicial capacity as to be calculated to create in the mind of a reasonable man a suspicion of that person’s impartiality, those circumstances are themselves sufficient to disqualify although in fact no bias exists” [emphasis added in italics and bold italics].
67 Bradford was applied by the same court in Doherty v McGlennan 1997 SLT 444 (at 445–446). Reference may also be made to Hoekstra v HM Advocate (No 1) 2000 SLT 602 at [4] as well as the Privy Council decision of Millar v Dickson 2001 SLT 988 at [64].
68 In the First Division decision of Secretary of State for Work and Pensions v Gillies 2004 SLT 14 (“Gillies (First Division)”), the Lord President (Cullen), who delivered the opinion of the court, affirmed the requirement that there be an “apprehension of bias”, further noting that Porter was a departure from Gough (at [22]):
For a number of years the test in England appears to have differed from that adopted in Scotland and in the Commonwealth. As was pointed out by Lord Hope of Craighead in Porter v Magill at para 100, the “reasonable likelihood” and “real danger” tests which Lord Goff of Chievely described in R v Gough at [1993] AC, p 670 were criticised in the High Court of Australia in Webb v The Queen at (1994) 181 CLR, p 50, on the ground that they tended to emphasise the court’s view of the facts and to place inadequate emphasis on the public perception of the irregular incident. In the event the Appellate Committee in Porter v Magill unanimously agreed with the suggestion of Lord Hope (in para 103) that the test for apparent bias in R v Gough should be adjusted in the light of the jurisprudence of the European Court of Human Rights. … [emphasis added in bold italics]
The Lord President observed thus (at [23]):
We would add this: looking at the matter from the viewpoint of the fair minded and informed observer assists in concentration on the question whether the circumstances are such as to give rise to an apprehension of bias, as opposed to whether there is actual bias.
69 This decision was later endorsed by the House of Lords in Gillies v Secretary of State for Work and Pensions [2006] 1 WLR 781 (“Gillies”), where the earlier decision of the House in Porter was applied. It is evident, therefore, that Porter marked the convergence of both the English and Scottish positions. The House in Gillies perceived no contradiction and appeared to suggest, instead, that the “reasonable suspicion of bias” test hitherto applied in the context of Scotland was the same as the test set out in Porter (Gillies at [38]). It would be appropriate at this juncture to turn our attention to the present English position.
(6) England and Wales
70 The test for apparent bias has had a more chequered history in English jurisprudence.
71 The “reasonable suspicion of bias” test under English law has been applied in a number of earlier cases (see, for example, the observations by Lord Esher MR in the English Court of Appeal decision of Allinson v General Council of Medical Education and Registration [1894] 1 QB 750, especially at 759; see also at 762 per Lopes LJ, as well as Lord Hewart CJ’s statement in R v Sussex Justices, ex parte McCarthy [1924] 1 KB 256 at 259 that “[n]othing is to be done which creates even a suspicion that there has been an improper interference with the course of justice”).
72 But, as we have alluded to above at [29], it may be that there is a commonality of perspective (that of the reasonable observer) for both the “real likelihood of bias” and “reasonable suspicion of bias” tests. Support for this may arguably be drawn from the following statements by Lord Denning MR in the English Court of Appeal decision of Metropolitan Properties Co (FGC) Ltd v Lannon and others [1969] 1 QB 577 (“Metropolitan Properties”) (at 599):
... [I]n considering whether there was a real likelihood of bias, the court does not look at the mind of the justice himself or at the mind of the chairman of the tribunal, or whoever it may be, who sits in a judicial capacity. It does not look to see if there was a real likelihood that he would, or did, in fact favour one side at the expense of the other. The court looks at the impression which would be given to other people. Even if he was as impartial as could be, nevertheless if right-minded persons would think that, in the circumstances, there was a real likelihood of bias on his part, then he should not sit. And if he does sit, his decision cannot stand ... There must be circumstances from which a reasonable man would think it likely or probable that the justice, or chairman, as the case may be, would, or did, favour one side unfairly at the expense of the other. The court will not inquire whether he did, in fact, favour one side unfairly. Suffice it that reasonable people might think he did. The reason is plain enough. Justice must be rooted in confidence: and confidence is destroyed when right-minded people go away thinking: “The judge was biased.” [emphasis added in italics and bold italics]
That said, it is not precisely clear what Lord Denning MR meant. For instance, it has been postulated in Grant Hammond, Judicial Recusal: Principles, Processes and Problems (Hart Publishing, 2009) at p 36 that he was in fact espousing Lord Hewart’s formulation that a reasonable suspicion would suffice; but the author also acknowledged that his “language may have also been a little loose, and it left some room for doubt as to just what the test was”. On the other hand, Lord Goff of Chieveley in Gough took the view that Lord Denning had applied an adaptation of the “real likelihood of bias” test (at 667F).
73 Further, although the result in Metropolitan Properties was a unanimous one in so far as the court found that apparent bias was made out, the other two judges appeared to prefer the “reasonable suspicion” test (at 602C–D (per Danckwerts LJ) and at 606 (per Edmund Davies LJ)), although Danckwerts LJ found that there were some unresolved doubts as to the applicable test (at 601). In the subsequent English Divisional Court decision of R v Altrincham Justices, Ex parte N Pennington [1975] QB 549, Lord Widgery CJ too found that it was “not altogether clear … how the matter was left in [Metropolitan Properties] so far as which of those tests was the correct one to apply in a given case” (at 553G).
74 In the English Court of Appeal decision of Hannam v Bradford Corporation [1970] 1 WLR 937 (“Hannam”), Sachs LJ doubted (at 942C) “whether in practice materially different results are produced by the ‘real likelihood of bias’ test” as opposed to the “reasonable suspicion of bias” test, although the learned judge did express a preference for the latter test for reasons provided by Lord Denning in Metropolitan Properties (although, as already noted above, the precise legal test utilised by Lord Denning was by no means clear). Cross LJ (as he then was) was also of the view (at 949C) that there was “little (if any) difference between the two tests”. It bears setting out his observations in full:
To my mind, there really is little (if any) difference between the two tests which are propounded in the cases which have been cited to us. If a reasonable person who has no knowledge of the matter beyond knowledge of the relationship which subsists between some members of the tribunal and one of the parties would think that there might well be bias, then there is in his opinion a real likelihood of bias. Of course, someone else with inside knowledge of the characters of the members in question might say: “Although things don’t look very well, in fact there is no real likelihood of bias.” That, however, would be beside the point, because the question is not whether the tribunal will in fact be biased, but whether a reasonable man with no inside knowledge might well think that it might be biased. [emphasis added in italics, bold italics and underlined bold italics]
75 The substance of Cross LJ’s observations in Hannam is, in our view, of the first importance. Even though the concept of “a real likelihood of bias” is relevant, the focus in these observations is on the perspective of the reasonable person. In our view, this last-mentioned perspective must surely be correct, given the fact that the court is not concerned with actual bias as such but, rather, with apparent bias, viz, how the situation concerned would appear to a reasonable person. It is also important to note that Cross LJ also defined “a reasonable person” as a person with no inside or specialist knowledge as such (see also Johnson at [48]–[49] per Kirby J, quoted above at [44], and Gillies at [39] per Baroness Hale, quoted below at [87]).
76 Cross LJ’s observations in Hannam were also endorsed by the English Divisional Court in R v Liverpool City Justices, Ex parte Topping [1983] 1 WLR 119. Ackner LJ (as he then was), delivering the judgment of the court, was of the view that the test to be applied was as follows (at 123; reference may also be made to the English High Court decision of Cook International ([33] above) at 231):
Would “a reasonable and fair-minded person sitting in court and” knowing all the relevant facts have a “reasonable suspicion that a fair trial for” the applicant “was not possible”?
77 However, Lord Goff’s observations in the House of Lords decision of Gough subsequently introduced what was effectively a new approach in the law of apparent bias. In Gough, Lord Goff, who delivered the leading judgment, preferred the term “real danger of bias” [emphasis added] as opposed to the term “real likelihood of bias” (at 670E–F), although he saw “no practical distinction” between these two ways of stating the test (at 668D). The learned Law Lord was also of the view that the focus here was on the concept of “possibility” as opposed to probability (at 665F–H and 668B–D); indeed, he also referred (at 668C) to “a real possibility of bias” as well [emphasis added]. This particular focus will also be significant for reasons that will become apparent in a moment. In the meantime, it is useful to set out Lord Goff’s very helpful and perceptive summary of the law which, as we mentioned, introduced in effect a new approach to the doctrine of apparent bias (at 670C–F):
In conclusion, I wish to express my understanding of the law as follows. I think it possible, and desirable, that the same test should be applicable in all cases of apparent bias, whether concerned with justices or members of other inferior tribunals, or with jurors, or with arbitrators. Likewise I consider that, in cases concerned with jurors, the same test should be applied by a judge to whose attention the possibility of bias on the part of a juror has been drawn in the course of a trial, and by the Court of Appeal when it considers such a question on appeal. Furthermore, I think it unnecessary, in formulating the appropriate test, to require that the court should look at the matter through the eyes of a reasonable man, because the court in cases such as these personifies the reasonable man; and in any event the court has first to ascertain the relevant circumstances from the available evidence, knowledge of which would not necessarily be available to an observer in court at the relevant time. Finally, for the avoidance of doubt, I prefer to state the test in terms of real danger rather than real likelihood, to ensure that the court is thinking in terms of possibility rather than probability of bias. Accordingly, having ascertained the relevant circumstances, the court should ask itself whether, having regard to those circumstances, there was a real danger of bias on the part of the relevant member of the tribunal in question, in the sense that he might unfairly regard (or have unfairly regarded) with favour, or disfavour, the case of a party to the issue under consideration by him … [emphasis added in italics and bold italics]
78 As is evident from the above extract, Lord Goff was not setting out a higher standard that was – if at all – embodied within the “real likelihood of bias” test. While he preferred the terminology of a “real danger of bias” as opposed to a “real likelihood of bias”, we would suggest that there is, in substance, no change in the latter test as such because Lord Goff’s intention was simply to stress that the relevant standard under either approach was that of a “possibility” and not a “probability”. However, Lord Goff was nevertheless of the view (at 668C–D) that the criterion of a “real possibility of bias” sufficiently served the principle that justice must manifestly be seen to be done, and rendered it “unnecessary … to have recourse to a test based on mere suspicion, or even reasonable suspicion” [emphasis added].
79 It is suggested that the novelty of the approach in Gough consists in the fact that Lord Goff had wholly (and radically) transformed the test for apparent bias under English law by reconceptualising the “reasonable person”. In particular, he was of the view that it was unnecessary to require that the court look at the matter through the eyes of a reasonable man, as the court, in such cases, personifies the reasonable man (at 670; also reproduced above at [77]). Indeed, on a literal reading, one might even state that Lord Goff had, in substance, eradicated the concept of “the reasonable person” altogether, replacing it with “the court”. Not surprisingly, as our discussion above has already alluded to, Gough was, as a result, criticised and not followed in many Commonwealth jurisdictions (though cf the legal position in Malaysia and New Zealand which were summarised in Tang Kin Hwa at [29] and [32]).
80 The critique of Gough and the consequent need for possible review of that particular decision in the English context was recognised by Lord Browne-Wilkinson in the House of Lords decision of R v Bow Street Metropolitan Stipendiary Magistrate and others, Ex parte Pinochet Ugarte (No 2) [2000] 1 AC 119 at 136A–C. However, the learned Law Lord found such a review to be unnecessary for the determination of that particular case and therefore preferred not to express a view.
81 However, in the English Court of Appeal decision of Locabail ([33] above), the court, after acknowledging (at [17]) the fact that Gough had “not commanded universal approval elsewhere” (and citing decisions from Scotland, Australia and South Africa, some of which have been referred to earlier in this judgment), was of the view that it “need not debate whether the substance of the two tests is different” and did not need to “consider whether application of the two tests would necessarily lead to the same outcome in all cases” (ibid). It was of the view that “whatever the merit of the reasonable suspicion or apprehension test, the test of real danger or possibility has been laid down by the House of Lords [in Gough] and is binding on every subordinate court in England and Wales” (ibid). The court then proceeded to observe thus (ibid):
In the overwhelming majority of cases we judge that application of the two tests would anyway lead to the same outcome. Provided that the court, personifying the reasonable man, takes an approach which is based on broad common sense, without inappropriate reliance on special knowledge, the minutiae of court procedure or other matters outside the ken of the ordinary, reasonably well-informed member of the public, there should be no risk that the courts will not ensure both that justice is done and that it is perceived by the public to be done. [emphasis added in italics and bold italics]
82 The passage just quoted in the preceding paragraph has, we suggest, not only added a gloss on Lord Goff’s observation on the concept of “the reasonable person” in Gough but has also (and more importantly) stripped it of the principal element that made it controversial to begin with. Put simply, if the court steps into the shoes of “the reasonable person” without any specialist knowledge (see also the views in the cases cited above at [43], [44], [74] and [75]), then the court is, in substance and effect, precisely the same as “the reasonable person” as originally conceived pursuant to the “reasonable suspicion of bias” test. Put even more simply, the court in Locabail has, in substance and effect, restored the “reasonable suspicion of bias” test, whilst simultaneously paying “lip service”, as it were, to Lord Goff’s view in Gough (probably because, as the court itself acknowledged, it was technically bound by Gough). In this regard, it should also be noted that the High Court of Australia, in Johnson, after having noted (at [12]) that “a differently expressed test [ie, different from the “reasonable apprehension of bias” test] has been applied in England”, very interestingly appended a footnote in which Locabail was cited by way of contrast.
83 Indeed, in the subsequent English Court of Appeal decision of In re Medicaments and Related Classes of Goods (No 2) [2001] 1 WLR 700 (“In re Medicaments”), Lord Phillips of Worth Matravers MR, who delivered the judgment of the court, observed thus (at [85]):
When the Strasbourg jurisprudence is taken into account, we believe that a modest adjustment of the test in R v Gough is called for, which makes it plain that it is, in effect, no different from the test applied in most of the Commonwealth and in Scotland. The court must first ascertain all the circumstances which have a bearing on the suggestion that the judge was biased. It must then ask whether those circumstances would lead a fair-minded and informed observer to conclude that there was a real possibility, or a real danger, the two being the same, that the tribunal was biased. [emphasis added in italics, bold italics and underlined bold italics]
84 Subsequently, in the important House of Lords decision of Porter, Lord Hope also acknowledged (at [100]) the critique of Gough and the debate as well as controversy that had occurred as a result. In his view, however, “it is now possible to set this debate to rest” (at [102]). The learned Law Lord first cited (at [102]) the very passage from In re Medicaments that was quoted in the preceding paragraph, and then proceeded to state what he understood to be the law as follows (at [103]):
I respectfully suggest that your Lordships should now approve the modest adjustment of the test in R v Gough set out in that paragraph. It expresses in clear and simple language a test which is in harmony with the objective test which the Strasbourg court applies when it is considering whether the circumstances give rise to a reasonable apprehension of bias. It removes any possible conflict with the test which is now applied in most Commonwealth countries and in Scotland. I would however delete from it the reference to “a real danger”. Those words no longer serve a useful purpose here, and they are not used in the jurisprudence of the Strasbourg court. The question is whether the fair-minded and informed observer, having considered the facts, would conclude that there was a real possibility that the tribunal was biased. [emphasis added in italics, bold italics and underlined bold italics]
85 As we have noted at [69] above, after looking at Gillies, the key point to note is that, after Porter, the test for apparent bias under English law is no longer that propounded by Lord Goff in Gough but is, instead, the same test that obtains in most of the Commonwealth (including Scotland), viz, the “reasonable suspicion of bias” test. Whilst it is true that the House in Porter did not expressly depart from Lord Goff’s approach in Gough, this last-mentioned approach has, with respect, been all but superseded (in substance).
86 Indeed, a quick survey of the subsequent case law confirms that the Porter test remains the position in England to date (see, for example, the subsequent House of Lords decision of Helow v Secretary of State for the Home Department and another [2008] 1 WLR 2416 (“Helow”) at [14]; the English Court of Appeal decisions of Taylor and another v Lawrence and another [2003] QB 528 at [60], Flaherty v National Greyhound Racing Club Ltd [2005] EWCA Civ 1117 at [26], and National Assembly for Wales v Elizabeth Condron and another [2006] EWCA Civ 1573 at [38]−[40]; and, more recently, in the context of arbitration, the English High Court decisions of Cofely Ltd v Bingham and another [2016] EWHC 240 (Comm) (“Cofely”) at [72]; H v L and others [2017] 1 WLR 2280 (“H v L”) at [16]).
87 In addition, it should be noted that guidelines were given as to the qualities of the “fair-minded and informed observer” in Gillies, where Baroness Hale of Richmond observed thus (at [39]):
The “fair-minded and informed observer” is probably not an insider (ie another member of the same tribunal system). Otherwise she would run the risk of having the insider’s blindness to the faults that outsiders can so easily see. But she is informed. She knows the relevant facts. And she is fair-minded. She is, as Kirby J put it in Johnson v Johnson 201 CLR 488, para 53, “neither complacent nor unduly sensitive or suspicious”.
Further, in Helow, Lord Hope elaborated on what being “fair-minded” and “informed” meant (at [2]–[3]):
2 The observer who is fair-minded is the sort of person who always reserves judgment on every point until she has seen and fully understood both sides of the argument. She is not unduly sensitive or suspicious, as Kirby J observed in Johnson v Johnson (2000) 201 CLR 488, 509, para 53. Her approach must not be confused with that of the person who has brought the complaint. The “real possibility” test ensures that there is this measure of detachment. The assumptions that the complainer makes are not to be attributed to the observer unless they can be justified objectively. But she is not complacent either. She knows that fairness requires that a judge must be, and must be seen to be, unbiased. She knows that judges, like anybody else, have their weaknesses. She will not shrink from the conclusion, if it can be justified objectively, that things that they have said or done or associations that they have formed may make it difficult for them to judge the case before them impartially.
3 Then there is the attribute that the observer is “informed”. It makes the point that, before she takes a balanced approach to any information she is given, she will take the trouble to inform herself on all matters that are relevant. She is the sort of person who takes the trouble to read the text of an article as well as the headlines. She is able to put whatever she has read or seen into its overall social, political or geographical context. She is fair-minded, so she will appreciate that the context forms an important part of the material which she must consider before passing judgment.
Apart from providing guidance on what the application of the reasonable observer construct entails, the comments by Baroness Hale and Lord Hope firmly establish the central role that the “fair-minded and informed observer” construct plays post-Porter.
The position in Singapore
88 The Singapore position on apparent bias is well-established, although (as alluded to) there is at present an apparent conflict of authority in so far as the relationship between the “reasonable suspicion of bias” test and the “real likelihood of bias” test is concerned.
89 At the outset, it is important to reiterate that, in the Singapore context, the test that applies is the “reasonable suspicion of bias” test. Indeed, this was the situation both prior to Tang Kin Hwa and Re Shankar Alan (see, for example, the Straits Settlements Court of Appeal decision of Alkaff & Company v The Governor-in-Council and others [1937] MLJ 211, this Court’s decisions in Jeyaretnam Joshua Benjamin v Lee Kuan Yew [1992] 1 SLR(R) 791 at [83] and Tang Liang Hong ([31] above) at [46], as well as the Singapore High Court decision of De Souza Lionel Jerome v Attorney-General [1992] 3 SLR(R) 552 at [35]–[39]; though cf the earlier Singapore decisions of Re Singh Kalpanath [1992] 1 SLR(R) 595 and Re Chuang Wei Ping [1993] 3 SLR(R) 357 where the “reasonable suspicion of bias” and the “real likelihood of bias” tests appear to have been perceived as being interchangeable) and since (see, eg, the decision of this Court in Manjit Singh s/o Kirpal Singh and another v Attorney-General [2013] 2 SLR 844 at [74]; Sim Yong Teng and another v Singapore Swimming Club [2016] 2 SLR 489 (“Sim Yong Teng”) at [62]). In respect of Tang Liang Hong, it should be noted that as Gough had already been decided by the time the proceedings in that case arose, this Court did refer to it, only to observe (at [48]) that for the purpose of that appeal, it was not material whether the “reasonable suspicion of bias” test or the “real danger of bias” test was applied because all counsel were relying on the “reasonable suspicion of bias” test, and because it appeared that “whichever of the tests the court [applied], the court must ascertain the relevant facts and circumstances on which the alleged apparent bias [was] founded”.
(1) Explaining the apparent inconsistency
90 In order to explain the apparent inconsistency between Tang Kin Hwa on the one hand and Re Shankar Alan on the other, it would be helpful to recapitulate that under English law at the time of these two decisions, the “real likelihood of bias” test was no longer embodied by Lord Goff’s articulation in Gough but, rather, had been substantively modified in Porter. Indeed, as we noted above, the test laid down in Porter is, in both substance and effect, the “reasonable suspicion of bias” test – or at least something that closely approximates it. Indeed, after the “modest adjustment” of the Gough test in Porter, the English position has been described as being no different from the test applied in the Commonwealth and in Scotland (In re Medicaments at [85] and Porter at [103] (see [83]–[85] above), both of which were also cited in Tang Kin Hwa at [30] and [31]; see also Saxmere at [3] (see [53] above) and Gillies (First Division) at [22] (see [68] above)).
91 It will be recalled that the court in Re Shankar Alan was of the view that the principle in Gough did not represent the law in Singapore (see [35] above). To the extent that Gough has indeed been departed from and the current English position is now represented by the decision in Porter, it could be said that there is little – if any – difference between the approach adopted in Re Shankar Alan and that adopted in Porter.
92 Turning to Tang Kin Hwa, whilst the court in that case did comment (at [40]) that “one ought not to draw a sharp distinction between the court’s perspective on the one hand and that of the public on the other”, it also noted that the court’s role is not to replace the reasonable man, but to ascertain the perspective of the public – and to that extent (and to that extent only) “personify” the reasonable man. Consistent with this approach, the court also drove home the point (at [41]) that the “reasonable suspicion of bias” test must be applied without attributing any specialist knowledge to the reasonable person (citing with approval at [40] the observations by the English Court of Appeal in Locabail at [17]). All this is, of course, contrary to the approach laid down in Gough and consistent with that laid down in Re Shankar Alan which (it bears reiterating) rejected the approach laid down in Gough. However, looked at in this light, the court in Tang Kin Hwa might have gone a step too far in so far as it simultaneously sought (at [40]) to at least implicitly downplay the criticism in Webb against Gough. It could also have characterised the actual decision in Porter more accurately (cf the characterisation at [33]). Unfortunately, these two last-mentioned points probably contributed to a perceived inconsistency between Tang Kin Hwa and Re Shankar Alan when there was (in substance at least) none to begin with. Perhaps most importantly it should be noted that in Tang Kin Hwa, the court was at pains to emphasise (at [37]) that the “real likelihood of bias” is indeed concerned with the appearance of bias (and not actual bias), citing Cook International at 231 (also referred to above at [33]). The court went on to observe (ibid) that:
… the oft-cited words of Lord Hewart CJ in the English Divisional Court decision of The King v Sussex Justices, ex p McCarthy [1924] 1 KB 256 at 259 that “it is not merely of some importance but is of fundamental importance that justice should not only be done, but should manifestly and undoubtedly be seen to be done” therefore apply not only to the “reasonable suspicion of bias” test but also to the “real likelihood of bias” test as well. [emphasis added in bold italics]
Evidently, both Tang Kin Hwa and Re Shankar Alan (see especially the latter at [70]–[72]) are entirely consistent in terms of the importance accorded to the appearance of justice being done.
93 Finally, we note that, on a substantive level, there is no material difference between the “reasonable suspicion of bias” test and the “real likelihood of bias” test. Indeed, it is difficult to see how a reasonable person might reasonably suspect bias without considering that there was a real likelihood of bias (see Auckland Casino Ltd at 149, quoted at [50] above). In this regard, the English decisions of Hannam (in particular, Cross LJ’s remarks at 949, reproduced at [74] above) and Locabail (at [17], see [81] above) also support the view that there is “little (if any) difference” between the two tests and that they would lead to the same outcome in the “overwhelming majority of cases”. We highlight again the observations in Porter, In re Medicaments, Saxmere and Gillies (see [68] above) that the Porter test (notwithstanding the usage of the phrase “real possibility” in its formulation) aligned the English position with the positions in the Commonwealth and Scotland, ie, the “reasonable suspicion of bias” test. Reference, in this regard, may also be made to the relevant academic literature (see, for example, Lionel Leo and Chen Siyuan, “Reasonable Suspicion or Real Likelihood: A Question of Semantics?” [2008] Sing JLS 446 (“Reasonable Suspicion or Real Likelihood”) at p 449 and Tham Lijing, “The Rule Against Apparent Bias – Language Disguising Thought”, Singapore Law Gazette (July 2014), p 21 (“The Rule Against Apparent Bias – Language Disguising Thought”) at p 23).
94 However, there is nevertheless both conceptual and practical merit in adopting a standard terminology (see also Reasonable Suspicion or Real Likelihood at p 454). We therefore take this opportunity to state that the appropriate test henceforth is the “reasonable suspicion of bias” test. Indeed, this is (as we have seen) also entirely consistent with the test adopted generally across the Commonwealth.
95 For the reasons set out above, we hold that Re Shankar Alan and Tang Kin Hwa do not conflict with each other and (most importantly) that the test that henceforth ought to be applied in respect of the issue of apparent bias is the “reasonable suspicion of bias” test.
(2) The attributes of the reasonable observer
96 Given the central importance of the perspective of the reasonable observer in the test for apparent bias pursuant to the “reasonable suspicion of bias” test, it would be useful at this juncture to pay closer scrutiny to what this perspective entails.
97 As we have earlier noted, the extent of the observer’s knowledge is key – as the High Court of Australia held in Laurie (cited at [48] above), the utility of the “reasonable observer” construct is that it reminds the judges that they are to embark on the inquiry of apparent bias from the perspective of a non-judicial observer, and consequently, there is a need to identify the information on which they make their determinations (see also Chen Siyuan and Kenny Lau Hui Ming, “The test for apparent bias: Lawrence Khong Kin Hoong v Singapore Polo Club [2014] SGHC 82” in Singapore Law Watch Commentary Issue 1/May 2014 at p 5, where it is posited that the key question is what the judge must exclude from his mind when personifying the reasonable man; the same point is made in Reasonable Suspicion or Real Likelihood at p 452). It should also be noted that the exclusion of extraneous considerations has gained barely any attention in local jurisprudence (likely because the point has never been raised or determinative of the case). Courts applying the “reasonable suspicion of bias” test in Re Shankar Alan have generally referred to how the fair-minded reasonable observer is to have knowledge of the “relevant facts”, and stopped short of addressing what ought to be regarded as irrelevant (see, eg, the Singapore High Court decisions of Kempinski Hotels SA v PT Prima International Development [2011] 4 SLR 633 at [66] (appeal allowed, PT Prima International Development v Kempinski Hotels SA and other appeals [2012] 4 SLR 98 (but not on this particular point)); TMM Division Maritima SA de CV v Pacific Richfield Marine Pte Ltd [2013] 4 SLR 972 at [122]; JRP & Associates Pte Ltd v Kindly Construction & Services Pte Ltd [2015] 3 SLR 575 at [53]; AMZ v AXX [2016] 1 SLR 549 at [94]; TOW v TOV [2017] 3 SLR 725 (“TOW v TOV”) at [31]; Metropole Pte Ltd v Designshop Pte Ltd [2017] 4 SLR 277 at [47]; and UES Holdings Pte Ltd v KH Foges Pte Ltd [2018] 3 SLR 648 at [29]).
98 We touched on this issue at [92] above by referring to [40] and [41] of Tang Kin Hwa and to what the English Court of Appeal stated in Locabail. In essence, the reasonable observer should not be taken to have detailed knowledge of the law or court procedure, or insider knowledge of the character, inclinations or ability of the members of the relevant court (see Livesey ([41] above) at 299; Gillies at [39]; Vakauta ([42] above) at 585; Johnson at [13] and [49]; Hannam at 949 per Cross LJ; and Re Singh Kalpanath at [82]). As the court in Locabail put it (at [17]), the court must not inappropriately rely on “special knowledge, the minutiae of court procedure or other matters outside the ken of the ordinary, reasonably well-informed member of the public” (see also The Rule Against Apparent Bias – Language Disguising Thought at p 22: “so long as the Court takes a broad common sense approach without inappropriate reliance on specialised knowledge, the reasonable man is personified by the Court” [emphasis added]). On the flipside, as Kirby J noted in Johnson, it would be a mistake for the court to impute all that what was eventually known to the court to the imaginary observer because it would only be holding up a mirror to itself (at [49]).
99 The reasonable observer is also informed. Although the observer is not a lawyer, he or she is not wholly uninformed and uninstructed about the law in general or the issue to be decided: Johnson at [53]. Further, the observer will take the trouble to inform himself or herself on all relevant facts that are capable of being known by members of the public generally, and he or she would also be able to consider what has been seen or read together with its proper context (Gillies at [17] and [39]; Helow at [3]; H v L at [16]; Cofely at [72]).
100 The reasonable observer is also taken to know of the traditions of integrity and impartiality that administrators of justice in general have to uphold. Hence, he or she would not jump to hasty conclusions based on isolated episodes of temper or remarks taken out of context. On a related note, the observer may also be taken to know generally that the court (or an adjudicator) may adopt reasonable efforts to confine proceedings within appropriate limits and to ensure that time is not wasted (Johnson at [53]; R v S (RD) ([58] above) at [111]; Basson ([63] above) at [35]–[36]).
101 The last point to be made in respect of the reasonable observer (which is unrelated to the question of knowledge) is that he or she is fair-minded. The observer is “neither complacent nor unduly sensitive or suspicious” (Johnson at [53]; Gillies at [39]). The observer is the sort of person who always reserves judgment on every point until he or she has seen and fully understood both sides of the argument (Helow at [2]). He or she will be able to distinguish between what is relevant and what is irrelevant, and to decide what weight should be given to the facts that are relevant when exercising his or her judgment (Gillies at [17]). The reasonable observer should not be confused with the person who has brought the complaint. There must be this measure of detachment. The assumptions that the complainant makes are not to be attributed to the observer unless they can be justified objectively (Helow at [2]). As the court in H v L noted (at [16]), litigation is highly stressful and expensive, and litigants would likely oppose anything which they perceive would imperil their prospects of success; hence, the complainant is likely to lack objectivity and be far from dispassionate.
102 But the observer is not a complacent one either. He or she knows that fairness requires that a judge must be, and must be seen to be, unbiased. He or she knows that judges, like anybody else, have their weaknesses. He or she will not shrink from the conclusion, if it can be justified objectively, that things that they have said or done or associations that they have formed may make it difficult for them to judge the case before them impartially (Helow at [2]).
(3) Restatement of the applicable principles
103 This is an appropriate juncture to restate the applicable principles in relation to the doctrine of apparent bias in Singapore:
(a) The applicable test is whether there are circumstances that would give rise to a reasonable suspicion or apprehension of bias in the fair-minded and informed observer. We do not think this is any different from asking whether the observer would conclude that there was a “real possibility” of bias, bearing in mind that the word “real” here refers to the degree of likelihood of bias (which must be substantial and not imagined), and not whether the bias is actual (see Tang Kin Hwa at [39] and Re Shankar Alan at [50]).
(b) As the test for apparent bias involves a hypothetical inquiry into the perspective of the observer and what the observer would think of a particular set of circumstances, the test is necessarily objective (see Tang Kin Hwa at [36]; see also the Singapore High Court decisions of Petroships Investment Pte Ltd v Wealthplus Pte Ltd (in members’ voluntary liquidation) (Koh Brothers Building & Civil Engineering Contractor (Pte) Ltd and another, interveners) and another matter [2018] 3 SLR 687 at [161] and TOW v TOV at [32] and [34]).
(c) A reasonable suspicion or apprehension arises when the observer would think, from the relevant circumstances, that bias is possible. It cannot be a fanciful belief, and the reasons for the suspicion must be capable of articulation by reference to the evidence presented (see Re Shankar Alan at [49]). But adopting a standard of possibility rather than probability furthers the vital public interest of ensuring that the administration of justice is beyond reproach from the perspective of reasonable members of the public (Re Shankar Alan at [64] and [75]; Tang Kin Hwa at [37]).
(d) In establishing whether the observer would harbour a reasonable suspicion of bias, the court must be mindful not to supplant the observer’s perspective by assuming knowledge outside the ken of reasonably well-informed members of the public (ie, detailed knowledge of the law and court procedure, or insider knowledge of the inclinations, character or ability of the members of the court or adjudication body) (see [98] above). The observer would be informed – that is, he or she would be apprised of all relevant facts that are capable of being known by members of the public generally (see [99] above). The observer would also be fair-minded; he or she would be neither complacent nor unduly sensitive and suspicious. He or she would know the traditions of integrity and impartiality that administrators of justice have to uphold, and would not jump to hasty conclusions of bias based on isolated episodes of temper or remarks taken out of context (see [100]–[102] above).
(e) In line with (d), the relevant circumstances which the court may take into account in finding a reasonable suspicion of bias would be limited to what is available to an observer witnessing the proceedings. Such circumstances might include, for example, the demeanour of the judge and counsel, the interactions between the court and counsel, and such facts of the case as could be gleaned from those interactions and/or known to the general public.
(4) Other issues
104 There are two other penumbral issues that we will briefly discuss. First, in TOW v TOV, Aedit Abdullah JC (as he then was) considered whether there is also a separate requirement for a nexus between the lack of impartiality and the issues requiring determination in the future if the judge does not recuse himself or herself (at [33], citing Murray & Tomas and another [2011] Fam CAFC 81; this corresponds to the second step of the test laid out in Ebner ([45] above)). The learned judge took the view that it would not be appropriate to do so given that the “reasonable suspicion of bias” test is aimed at considering the perception of bias through a hypothetical observer. This being the case, whether or not a nexus exists would seem immaterial. This was not a point that the parties addressed, and we will consider it in the future if the need arises.
105 Second, it has also been said that the phrase “reasonable apprehension” ought to be preferred to “reasonable suspicion” given the potentially inappropriate connotations the latter might engender: see, eg, South African Rugby Football Union ([63] above) at [38] and Livesey at 294. However, it should be recalled that in Re Shankar Alan, the court noted that the “reasonable suspicion” test made no adverse finding on whether the court or tribunal in question was affected by actual bias (at [81]; see [37] above); we agree with that holding and we therefore do not think there is any unintended nuance of meaning that the phrase “reasonable suspicion”, as opposed to “reasonable apprehension”, would convey.
106 Having said the above, we reject the first prong of the Appellant’s case (see [22] above), not only because we find that the Judge had correctly identified the right test to apply, but also because we find that in substance, there is no material difference between the two tests. We now turn to the applicable principles relating to prejudgment and excessive judicial interference.
Prejudgment
107 The rule against prejudgment prohibits the decision-maker from reaching a final, conclusive decision before being made aware of all relevant evidence and arguments which the parties wish to put before him or her. The primary objection against prejudgment is the surrender by a decision-making body of its judgment such that it approaches the matter with a closed mind (Sim Yong Teng at [50]).
108 While it has been said that prejudgment is distinct from (though related to) apparent bias (see the Singapore High Court decision of Public Prosecutor v Chua Siew Wei Kathleen [2016] 2 SLR 713 (“Kathleen Chua”) at [24]), the preponderance of authority has referred to prejudgment as something that amounts to apparent bias. Locally, in Sim Yong Teng, the appellants’ argument on prejudgment or predetermination was characterised as an apparent bias argument (at [47]). This Court agreed (at [63]–[64]) that any reasonable, fair-minded and fully informed observer looking at the circumstances of the case would have formed the view that there was prejudgment amounting to apparent bias. In addition, in Kay Swee Pin v Singapore Island Country Club [2008] 2 SLR(R) 802, this Court regarded prejudgment as a form of apparent bias (at [65]). Prejudgment has also been classified as a form of apprehended bias in most other jurisdictions: see Johnson, and in particular [21], where South African Rugby Football Union, Auckland Casino Ltd, R v S (RD) and Locabail are cited; see also the recent High Court of Australia decision of Laurie (where the High Court found that the reasonable observer would apprehend that the judge might not move from the position expressed in an earlier interlocutory judgment) and the English Court of Appeal decision of Otkritie International Investment Management Ltd and others v Urumov [2014] EWCA Civ 1315 at [1].
109 To establish prejudgment amounting to apparent bias, therefore, it must be established that the fair-minded, informed and reasonable observer would, after considering the facts and circumstances available before him, suspect or apprehend that the decision-maker had reached a final and conclusive decision before being made aware of all relevant evidence and arguments which the parties wish to put before him or her, such that he or she approaches the matter at hand with a closed mind.
110 To this, we would add the following. Judges are obliged to make the effort to prepare for a hearing beforehand and inevitably, provisional views and conclusions would be formed during such preparations: see the decision of this Court in Prometheus Marine Pte Ltd v King, Ann Rita and another appeal [2018] 1 SLR 1 (“Prometheus”) at [39]. The court in Prometheus also noted that an open mind does not mean an empty mind and it is consistent with the judicial function to pose provisional views and concerns to counsel for them to be addressed. Indeed, as the court went on to observe, citing Re Shankar Alan, counsel are often assisted by the chance to peek into the judicial mind as they then have the opportunity to persuade the court to come to a different view if they so wish. The same view was expressed in Johnson at [13]. The High Court of Australia noted that the reasonableness of any apprehension of bias was to be considered in the context of the exigencies of modern-day practice and the need for active case management. Judges could not be expected to sit mute until the presentation of the parties’ cases was over, and prejudgment could not be made out solely because tentative views reflecting a certain tendency of mind were expressed during exchanges with counsel.
Excessive judicial interference
111 Quite apart from apparent bias, there is also the separate ground of whether the Judge excessively interfered with the proceedings. This Court dealt with this ground at length in Mohammed Ali bin Johari v Public Prosecutor [2008] 4 SLR(R) 1058 (“Mohammed Ali bin Johari”) at [117] et seq, and it would suffice for present purposes to set out the summary of principles at [175] of the judgment, which represents the current state of the law (see Lim Choo Suan Elizabeth and others v Goh Kok Hwa Richard and others [2009] 4 SLR(R) 193 at [173]; Nim Minimaart (a firm) v Management Corporation Strata Title Plan No 1079 [2010] 2 SLR 1 (“Nim Minimaart”) at [10]; and Sinnappan a/l Nadarajah v Public Prosecutor [2018] SGCA 21 at [3]):
(a) The system the courts are governed by under the common law is an adversarial (as opposed to an inquisitorial) one and, accordingly, the examination and cross-examination of witnesses are primarily the responsibility of counsel.
(b) It follows that the judge must be careful not to descend (and/or be perceived as having descended) into the arena, thereby clouding his or her vision and compromising his or her impartiality as well as impeding the fair conduct of the trial by counsel and unsettling the witness concerned.
(c) However, the judge is not obliged to remain silent, and can ask witnesses or counsel questions if (inter alia):
(i) it is necessary to clarify a point or issue that has been overlooked or has been left obscure, or to raise an important issue that has been overlooked by counsel; this is particularly important in criminal cases where the point or issue relates to the right of the accused to fully present his or her defence in relation to the charges concerned;
(ii) it enables him or her to follow the points made by counsel;
(iii) it is necessary to exclude irrelevancies and/or discourage repetition and/or prevent undue evasion and/or obduracy by the witness concerned (or even by counsel);
(iv) it serves to assist counsel and their clients to be cognisant of what is troubling the judge, provided it is clear that the judge is keeping an open mind and has not prejudged the outcome of the particular issue or issues (and, a fortiori, the result of the case itself).
The judge, preferably, should not engage in sustained questioning until counsel has completed his questioning of the witness on the issues concerned. Further, any intervention by the judge during the cross-examination of a witness should generally be minimal. In particular, any intervention by the judge should not convey an impression that the judge is predisposed towards a particular outcome in the matter concerned (and cf some examples of interventions which are unacceptable which were referred to in [Regina v Valley (1986) 26 CCC (3d) 207] (“Valley”)).
(d) What is crucial is not only the quantity but also the qualitative impact of the judge’s questions or interventions. The ultimate question for the court is whether or not there has been the possibility of a denial of justice to a particular party (and, correspondingly, the possibility that the other party has been unfairly favoured). In this regard, we gratefully adopt the following observations by Martin JA in Valley (reproduced above at [138]):
Interventions by the judge creating the appearance of an unfair trial may be of more than one type and the appearance of a fair trial may be destroyed by a combination of different types of intervention. The ultimate question to be answered is not whether the accused was in fact prejudiced by the interventions but whether he might reasonably consider that he had not had a fair trial or whether a reasonably minded person who had been present throughout the trial would consider that the accused had not had a fair trial … [emphasis added in bold italics]
(e) Mere discourtesy by the judge is insufficient to constitute excessive judicial interference, although any kind of discourtesy by the judge is to be eschewed.
(f) Each case is both fact-specific as well as context-specific, and no blanket (let alone inflexible) rule or set of rules can be laid down.
(g) The court will only find that there has been excessive judicial interference if the situation is an egregious one. Such cases will necessarily be rare. It bears reiterating what we stated earlier in this judgment (at [125] above):
[T]he argument from judicial interference cannot – and must not – become an avenue (still less, a standard avenue) for unsuccessful litigants to attempt to impugn the decision of the judge concerned. This would be a flagrant abuse of process and will not be tolerated by this court. Parties and their counsel should only invoke such an argument where it is clearly warranted on the facts …
[emphasis in original]
112 Although the facts of any given case may give rise to both apparent bias and a finding of excessive judicial interference, it is important to note the difference between the two grounds. The “excessive judicial interference” ground guards against the risk of a fair trial being compromised because of the failure of a decision-maker to observe his proper role and his duty not to descend into the arena (Re Shankar Alan at [110]), and is borne out of the fact that the system of justice in Singapore is founded on an adversarial model rather than an inquisitorial model (Re Shankar Alan at [107]). The mischief arises from the decision-maker taking up a position and pursuing it with the passion of an advocate, thereby slipping into the perils of self-persuasion (Re Shankar Alan at [115]). And as the learned judge in that case noted at [117(b)], the resolution of a complaint of excessive judicial interference depends not on appearances or what impressions a fair-minded observer might be left with, but rather on whether the reviewing court is satisfied that the manner in which the challenged tribunal or judge acted was such as to impair its ability to evaluate and weigh the case presented by each side.
113 We would add that another mischief that the “excessive judicial interference” ground guards against is impeding a party’s presentation of its case (as noted in Mohammed Ali bin Johari at [175(b)]). In the Ontario Court of Appeal decision of Regina v Valley (1986) 26 CCC (3d) 207 (“Valley”) (cited in Mohammed Ali bin Johari at [138]), the situation “[w]here the interventions have made it really impossible for counsel for the defence to do his or her duty in presenting the defence” was raised as one of the principal types of intervention that would lead to the quashing of criminal convictions (at 231). The case of Nim Minimaart is a useful example of how such excessive interference impinges on a party’s ability to prove his case. There, the counsel bringing the complaint deposed in his affidavit that the remarks made by the judge led him to feel traumatised and to think that he would be in contempt of court if he continued, and he therefore decided to settle (at [17]–[18]). In Kathleen Chua, both the hampering of a party’s conduct of its case as well as the “descent into the arena” to pursue a particular position were listed as grounds of challenge (at [24]), and both were made out on the facts of the case (at [32] and [40]). Notably, the court identified how the Prosecution’s presentation of their case was hampered by identifying lines of potentially relevant questioning that had been prematurely stopped by the judge below (at [28]).
114 Having set out the applicable legal principles, we turn now to the facts of the case, beginning with the circumstances the Appellant relies on to make out her argument on apparent bias.
Our findings
Apparent bias
115 The Appellant raises the following arguments to support her case on apparent bias:
(a) First, the Appellant argues that Ms Ong was unduly hampered in the presentation of the Appellant’s case because she was allegedly interrupted mid-sentence and berated constantly by the Judge. The “harsh comments” and “hostile environment” made it “impossible for [Ms Ong] to effectively present the [Appellant’s] case to the Court”. According to the Appellant, the Judge repeatedly raised her voice, used harsh rebukes such as “Stop”, “Enough”, “No”, “Never mind”, and “Don’t bother”, and, on at least two occasions, slammed her hand on the table.
(b) Second, the Appellant argues that the Judge pre-judged certain issues, including the purpose of moneys that the Appellant loaned from her sister.
(c) Third, the Appellant argues that the Judge treated the parties unequally, and this can be seen from (a) the wide latitude given to the Respondent to introduce new arguments and/or allegations that were unsupported by affidavit evidence; (b) the relatively less intense scrutiny given to the Respondent’s case; and (c) the Judge’s inclination to find in favour of the Respondent on issues in the absence of documentary proof from the parties.
116 We disagree with the Appellant’s arguments, and we find that the circumstances would not give rise to a reasonable suspicion or apprehension of bias on the part of the fair-minded and informed observer. We make five points in relation to this.
117 But before we elaborate on these points, we make a preliminary observation. The recusal application was brought based on what transpired during the final ancillaries hearings, and the proceedings were captured by Appellant’s counsel in an unofficial transcript that the Appellant has relied heavily upon. The Respondent has not objected to the accuracy of this transcript. We note, however, the Judge’s comment that she did not accept it wholesale though it did cover the salient aspects of the proceedings. Indeed, we find the transcript to be tinted with a certain view of the Judge and the proceedings. This is evident from the various editorial adjustments to the transcript, such as the use of uppercase, multiple exclamation marks and the inclusion of the recorder’s comments in parentheses.
118 For instance, during the proceedings on 18 April 2017, it was recorded that the Judge went “on rant for 2 minutes about tables and their importance”. Further, the following exchanges between the court, Ms Ang, and Ms Ong (both counsel were referred to as “JA” and “LO” respectively in the transcript) transpired on 18 April 2017:
Court:  Cut to the chase. Show me that
JA: This is at Vol 2 (Part A), Tab 12, page 48. End of policy year 8 is 14 November 2014, as YH can see the policy inception date ([Judge] reads document with no objection-asks for a letter from [Ms Ang]))
LO:  YH we have…
Court: MS.ONG I’m recording this, please! (she not typing)
…
JA:  Ok the later affidavit is Tab…
Court:  They obviously consider it VERY important (so sarcastic/snide)
LO:  No YH, my point wa…
[emphasis added in bold italics]
119 In our view, the heavy editorial hand that ran throughout the transcript demonstrated – quite ironically − a loss of objectivity and a subconscious bias against the Judge. This is the very pitfall that we cautioned against at [5] above, and it is indicative that the Appellant’s counsel has failed to take a dispassionate view of the proceedings. It also brings to the fore the importance of bearing in mind that the reasonable observer is not to be confused with the person who has brought the complaint (see [101] above). We are satisfied that even if we were to wholly accept the transcripts (including what the use of uppercase was intended to convey as well as the aforesaid parenthetical remarks), there would still not have been a reasonable suspicion or apprehension of bias.
120 The first reason for this is that the manner in which the proceedings panned out required the Judge to play a more active role in case management. As the parties had said, the case was a “high conflict” one, and parties therefore had little success in narrowing the disputed issues and facts down to a more limited scope. In fact, we had the impression that counsel were instructed to challenge every last point, and this unnecessarily prolonged the proceedings to a great extent.
121 For instance, one of the issues that caused annoyance was an asset referred to as monies in a Thinkorswim account. The Judge thought that there were many aspects of this account that the parties could have reached agreement on – such as the date on which the account was opened, the date on which the account was closed and the closing balance. Yet, a considerable amount of time was expended on these easily ascertainable (and objective) facts, as the Appellant sought to argue that she had suffered a net loss in relation to that account, which was a wholly separate substantive point.
122 Shortly thereafter, the parties dealt with another asset, a fixed deposit account, which the wife had labelled as “unknown” in the parties’ joint summary. The Judge thought that the account should be excluded from the matrimonial pool because the account had been closed on 31 May 2008, many years before the divorce proceedings. The Appellant explained that she had labelled the account as “unknown” because the Respondent had claimed, in an earlier affidavit, that the account was closed on 1 June 2007 instead, and because they were not aware of the closing balance of the account. Again, a lot of time was expended, even though it would not have mattered whether Ms Ong was right about the date or not. As the Judge noted (based on the Appellant’s transcript), after listening to what counsel had to say:
There was an email of 1 February 2016 from the same person at DBS saying that the account was closed as of 1 June 2007, which was exhibited in the husband’s affidavit of 12 February 2016 at page 17, Tab 55, Vol 13. As it would seem that this account was closed by June 2007 or May 2008.
This account, as earlier stated will not be included in the pool.
All this time wasted. …
123 This was, in fact, the basic manner in which the proceedings unfolded, as the Judge trudged through the items on the parties’ summary of assets, repeatedly imploring both counsel to narrow the scope of their dispute. Neither budged very much, purportedly on their clients’ instructions. Matters were made worse by the fact that the joint summary prepared by the parties did not state clearly what the dispute was, and after some questioning by the Judge, it became clear that, in respect of numerous issues, there was no real dispute to begin with and the parties could have come to an agreement between themselves. This led the Judge to tell the parties the following:
… I told lawyers before as well as parties, it’s not the job of the court for both counsels or both parties to simply disagree. And then I run through and say after going through all of this don’t you agree that its very obvious? This is this, so why can’t you concede so that we can save time. So this approach I prefer to take. …
…
[In relation to the fixed deposit account earlier mentioned:] All I’m saying is that if it was factual, when the account was opened, when the account was closed, it’s factual. Agree or not? Cause if I were to look at it I can’t see anything to disagree. There is nothing for me to decide. So let’s not disagree for the sake of disagreement. …
124 The present case was therefore one that cried out for judicial intervention. The Judge had to keep a tight rein on the proceedings and steer them in the direction of a resolution of the final ancillary matters. We have some sympathy for the Judge because of the unenviable task that she faced. While she had to intervene and pose questions to the parties frequently, this was necessitated by how the proceedings were progressing (or, rather, not progressing).
125 It also did not help that the Appellant could not furnish information for matters that would ordinarily be within her knowledge. For instance, the Appellant, despite being the owner of the Thinkorswim account, did not know when the account was opened. Further, in relation to a car owned by the Appellant, the Appellant claimed to not have records of the value of the car at the time of the interim judgment because it was driven by the Respondent. The Judge offered a solution – to take the value of the Certificate of Entitlement at the time of the interim judgment and then to make adjustments based on that. She sought information in respect of this. Ms Ong agreed to speak to the client during the lunch break. When she returned after lunch, Ms Ong essentially repeated her client’s position that they had no evidence pertaining to the value of the car as of the interim judgment date, and that her client “leaves it to the court” to decide on the value of the car.
126 In our judgment, the fair-minded observer would have concluded that the Judge was not affected by apparent bias, but was determined to expedite the proceedings instead. As we have mentioned above, the observer may be taken to know generally that the court would adopt reasonable efforts to confine proceedings within appropriate limits and to ensure that time is not wasted (see [100] above), even if the observer is not to be taken to know of the minutiae of court procedure such as the judge-led approach provided for under r 22 of the FJR (cf the Judge’s reasoning at [17] above).
127 On a related note, our second reason as to why the observer would not have reasonably apprehended or suspected bias is the fact that the Judge treated Ms Ang in a similar manner. For instance, in relation to the value to be ascribed to the Appellant’s car mentioned at [125] above, Ms Ang was admonished for not providing a proper basis of valuation:
Court:  What value do you then propose? You cannot just disagree. When you do a Joint Summary, you take a position on it.
JA: Estimated value of $25,000
Court: And how did you work that out? Based on?
JA:  Based on my client’s car value, half of that YH.
Court: I don’t think it helps me. I don’t know what your client’s car is. And half of that? Of what? Look this is a Court of Law it’s not a …
128 The Judge’s annoyance with the snail’s pace at which the proceedings were going was expressed towards both parties and their respective counsel. For instance, Ms Ang was chided for not promptly providing the relevant information to the Judge:
Court:  So Ms Ang the next time you give me dates that would sort it out
JA: Sorry YH
Court:  If the wife or the husband takes issue with everything under the sun, right? Whether it is the wife or the husband then you point out to me all relevant information so I don’t spend 10 minutes or 20 minutes going through the dates!!
JA:  Sorry YH
And when Ms Ang unnecessarily wasted the court’s time, she too was admonished by the Judge. In particular, the Judge (understandably) grew annoyed when Ms Ang informed her belatedly of her client’s intention to exclude certain insurance policies reflected in the Joint Summary. The following exchange is recorded in the Appellant’s transcript of the proceedings:
Court:  4F!!!! You mean we wasted all that time for a policy that was going to be excluded?? I cannot believe this!!!! I’m going to record this, I just cannot believe this! So which other policies are we excluding please?
…
Court:  … Counsel then informed the Court that 12 policies were excluded, including policy (f) for which we had spent time at the hearing to determine if the bonus was payable at time of IJ, when such an exercise was wholly unnecessary as this policy had been agreed for exclusion. Directions had even been given. Directions given for documentary evidence to be produced by Prudential are therefore revoked.
I don’t wish to belabour the point, I just do this to illustrate all I’m asking for… look just be aware, WHY did we spend time on that policy?! And why was the comment made about that policy? Alright?
JA: Do we still have to produce a document?
Ct: NO! We don’t disagree for everything simply because, you know.
These exchanges demonstrate very clearly that the Judge’s patience was being tested by both counsel’s inability to cut down on the issues of dispute, the unnecessary wastage of time, and the “high conflict” nature of the case.
129 This takes us to our third point: even if the Judge expressed her annoyance and frustration and was not a model of patience with the parties, that would not amount to apparent bias. Discourtesy is generally to be discouraged – but try as judges may, they will not always maintain a perfect judicial temperament, especially in a case like the present. We have touched on this above at [7]–[8].
130 During the oral hearing of this appeal, counsel for the Appellant suggested that the question as to when discourtesy crosses into bias is a question of degree. His case on apparent bias is premised on the sustained, multiple occasions on which the Judge allegedly spoke to Ms Ong harshly. But we do not think that the test of apparent bias can be reduced to the number and severity of instances when the judge may be said to have been discourteous towards counsel. There is no magical threshold at which a judge is so discourteous that there is a reasonable suspicion of bias. The question is simply whether the fair-minded observer would reasonably apprehend or suspect bias, viewing the alleged instances of discourtesy in their respective contexts, and also taking into account other relevant circumstances (for example the judge’s demeanour towards opposing counsel). As we have earlier mentioned, the reasonable observer is taken to know of the traditions of integrity and impartiality that courts have to uphold. He or she will not jump to hasty conclusions of bias simply based on isolated episodes of temper or remarks taken out of context. Given the context earlier described, we do not think that apparent bias is made out at all.
131 The fourth point relates to the Appellant’s argument on prejudgment. The issue that the Appellant claims the Judge prejudged is whether loans taken by the Appellant from her sister were wholly for her legal bills (and should therefore be disregarded) or whether they went towards meeting family and household expenses. The Appellant’s argument on this issue is that the Judge focused solely on the evidence that supported the view that the loaned moneys were used for the Appellant’s legal bills and was not receptive to evidence that supported the other view, which Ms Ong highlighted the following day.
132 We do not agree with the Appellant. In the first place, the Judge refrained from making any final orders in relation to the division of assets until the outcome of this appeal. Putting that to one side, even if we accept that the issue was discrete, we do not think that the Judge pre-judged the matter. When this issue was first raised before the Judge on 20 June 2017, Ms Ong did not even put any contrary views across. During the hearing on 21 June 2017, Ms Ong then decided to highlight statements made by the Appellant in her affidavit which supported the Appellant’s position that the loaned sums were used for household expenses instead. The Judge responded with “Fine I will record it. I have gone through this but I can” and “Let me record so I don’t miss it”. She proceeded to note the citations given by Ms Ong. She then mentioned that the Appellant had made statements to that effect multiple times. She even invited Ms Ang to pick up the affidavits to look at what Ms Ong was referring to. This sequence of events detracts from there being prejudgment on the part of the Judge in two main respects. In the first place, the Judge had already seen and considered the evidence that Ms Ong wanted to raise. All Ms Ong did on 21 June 2017 was to highlight what the Judge already had sight of. Hence, even if the Judge had reached a firm conclusion after reading the affidavits, and Ms Ong did not have the opportunity to orally reiterate what was in the affidavits, that would not amount to prejudgment. More importantly, the Judge’s mind was not closed to what Ms Ong was raising. She wanted to record what Ms Ong said so that she would not miss it, only stopping when she realised that Ms Ong was raising something that she had already read. In our view, what the Appellant has raised is far removed from being a case of prejudgment.
133 Finally, we find that the Appellant’s argument on unequal treatment (which, in a nutshell, is based on the Judge’s preference for the Respondent’s case and the greater leeway that the Judge purportedly showed to the Respondent in respect of documentary proof) should more properly be regarded as a matter for appeal, not recusal.
Excessive judicial interference
134 The circumstances supporting the Appellant’s argument on apparent bias overlap with the basis on which the Appellant purports to demonstrate excessive judicial interference by the Judge, and we reiterate what we have set out at [120]–[130] above.
135 As this Court held in Mohammed Ali bin Johari at [175(d)], what is crucial is the qualitative (not only the quantitative) impact of the judge’s interventions. The ultimate question to be answered, paraphrasing what was held in Valley, is whether a reasonably minded person present throughout the ancillaries hearing would consider that the Appellant had not had a fair hearing. We answer this question in the negative and find that the Appellant was not hampered in the presentation of her case.
136 Unlike Nim Minimaart and Kathleen Chua (see [113] above), the Appellant has not demonstrated how her case was hampered in any way. In addition, unlike a trial where evidence is to be elicited from the witness orally, Ms Ong was given ample opportunity to tender written submissions and joint summaries, which the Judge took note of. For these reasons, Ms Ong was not unduly hampered in the presentation of the Appellant’s case.
137 The Appellant argues that the Judge descended into the arena by asking leading questions to counsel for the Respondent in order to reach a conclusion that she had already made. The sole basis for this argument is the following exchange between the Judge and counsel for the Respondent on 20 June 2017:
JA:  This was the 4th tranche
Ct:  So you are saying that the evidence could have been obtained well before?
JA:  Yes. I note that the Wife’s request to DBS for this confirmation was made on 2 December 2016, with the confirmation given on the same day
Ct:  So what does that imply to you?
JA: It implies to me she could have made the request and obtained it on the same day before the AM was filed.
Ct:  Why could she have produced it on the same day? What could she have produced? She could have produced her bank statement, or anything. Is there anything you wish to imply?
JA:  It implies that she could have gotten the information before the filing the 2nd AM, with the cheque no. and cheque date. I take that this is an application for them to admit this evidence, and if it is admitted I will accept that this is the Wife’s contributions
[emphasis in original]
We reject this argument. The exchange is short and one-off, and on this sole basis, it seems too much of a stretch to draw the conclusion that the Judge descended into the arena and was pursuing a particular position as an advocate. We agree with the Respondent that the Judge was merely expanding upon and clarifying the point that the Respondent was making.
Conclusion
138 For the reasons set out above, we find that apparent bias was not made out and that the Judge did not interfere excessively in the proceedings. In the circumstances, the appeal is dismissed.
139 We would like to conclude much as we began. First, as we emphasised right at the outset of this judgment (at [3] above), counsel are not the mere “mouthpieces” of their clients. They have to be especially aware of this in family proceedings (such as the present) because such proceedings are frequently driven by emotional and even bitter clients. To the extent that they do not counsel (and/or even, if necessary, admonish) their clients and scrupulously observe their duty to the court, lawyers fail to live up to their calling as members of a profession whose hallmarks are honour and justice.
140 We also emphasised the importance of perspective (at [6]−[7] above). It bears reiterating that had all concerned stepped into the other’s shoes, if even for a brief moment, the present situation might well have been averted. As we also said, however, counsel should also avoid stepping into their client’s (ill-fitting) shoes and should, if they have inadvertently done so, step out of them with some haste. There is some evidence that at least suggests that counsel were, in fact, guilty of this in the present case. Be that as it may, the perspective required here is one characterised by patience and empathy – the complete antithesis of what happened in this case.
141 Finally, we cannot emphasise enough how extremely serious allegations of judicial bias are. Indeed, such allegations can be utilised not only as a weapon of abuse by disgruntled litigants but also waste valuable court time and resources in the process. We would imagine that, by their very nature, such allegations would be rare in the extreme. Should such proceedings arise before the court in the future and be found to be unmeritorious, there may be serious consequences.
Sundaresh Menon  Andrew Phang Boon Leong Judith Prakash
Chief Justice   Judge of Appeal  Judge of Appeal
Sim Yuan Po Thomas, Linda Joelle Ong and Lim Xiao Wei Charmaine (Engelin Teh Practice LLC) for the appellant;
N Sreenivasan SC and Lim Shu Fen (Straits Law Practice LLC) for the respondent.

SUPREME COURT OF SINGAPORE

4 October 2018

Case summary

BOI v BOJ [2018] SGCA 61

Civil Appeal No 189 of 2017

----------------------------------------------------------------------------------------------------------------

Decision of the Court of Appeal (delivered by Judge of Appeal Andrew Phang Boon Leong):

The Court of Appeal provides guidance in relation to the test of apparent bias, and affirms the High Court Judge’s dismissal of the recusal application brought against her.

Background

1.The parties to the appeal are former spouses. At the time of the recusal application, they were in the midst of divorce proceedings. The ancillary matters were scheduled to be heard by the High Court Judge (“the Judge”) between November 2016 and July 2017. Before the conclusion of the ancillary matters, and before the issuance of any decision in respect of these matters, the ex-wife made an application for the Judge to recuse herself on grounds of apparent bias.

2. The Judge saw no reason to recuse herself. She referred to the Family Justice Rules 2014 (GN No S 813/2014) which provided that she was entitled to adopt a judge-led approach. She dismissed any allegation of prejudgment as there were still issues that had not been covered, and the factual matrix would have to be viewed in its entirety in order to arrive at a decision on the division of matrimonial assets and maintenance. The ex-wife appealed, arguing that the Judge applied the wrong test (the “real danger” or “real likelihood” test), and that apparent bias was made out on the facts.

Decision on appeal

3. The Court of Appeal restated the test for apparent bias in Singapore. The question to be asked is whether there are circumstances that would give rise to a reasonable suspicion or apprehension of bias in the fair-minded and informed observer. The test is objective. A reasonable suspicion or apprehension arises when the observer would think, from the relevant circumstances, that bias is possible. In establishing whether the observer would harbour a reasonable suspicion of bias, the court must be mindful not to supplant the observer’s perspective by assuming knowledge outside the ken of reasonably well-informed members of the public (ie, detailed knowledge of the law and court procedure, or insider knowledge of the inclinations, character or ability of the members of the court or adjudication body). The observer would be informed – that is, he or she would be apprised of all relevant facts that are capable of being known by members of the public generally. The observer would also be fair-minded; he or she would be neither complacent nor unduly sensitive and suspicious. He or she would know the traditions of integrity and impartiality that administrators of justice have to uphold, and would not jump to hasty conclusions of bias based on isolated episodes of temper or remarks taken out of context. ([103])

4. The Court of Appeal found that apparent bias was not made out. First, the parties were unable to narrow the disputed issues and facts to a more limited scope, and were pursuing every last point. The Judge therefore had to play a more active role in case management, and the fair-minded observer would have concluded that the Judge was not affected by bias, but was determined to expedite the proceedings instead. Second, the Judge’s annoyance with the snail’s pace at which the proceedings were going was expressed towards both parties and their respective counsel. Third, even if the Judge expressed her annoyance and frustration and was not a model of patience with the parties, that would not necessarily amount to apparent bias. Fourth, the Judge had not prejudged the specific issue complained about. Fifth, the argument on unequal treatment (which was based on the Judge’s preference for the Respondent’s case and the greater leeway that the Judge purportedly showed to the Respondent in respect of documentary proof) was a matter for appeal, not recusal. ([120], [126], [128], [129], [132], [133])

5. The Court of Appeal also found that the Judge had not excessively interfered with the proceedings. Counsel for the ex-wife was not unduly hampered in the presentation of the ex-wife’s case; on the contrary, ample opportunity was afforded to tender written submissions and joint summaries. The argument that the Judge had descended into the arena was based on one brief exchange, and it was too much of a stretch to conclude that the Judge had descended into the arena and pursued a particular position as an advocate when all the Judge had been doing was merely expanding upon and clarifying the point that the ex-husband was making. ([136], [137])

 

This summary is provided to assist in the understanding of the Court’s grounds of decision. It is not intended to be a substitute for the reasons of the Court. All numbers in bold font and square brackets refer to the corresponding paragraph numbers in the Court’s grounds of decision.

Back to Top

This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.

Version No 1: 27 Oct 2020 (22:40 hrs)