Bechara v Rex

1. [1]

   McHUGH JA: This is an entirely singular case. The applicant, Mr Bechara, seeks leave pursuant to s 5F(3)(a) of the Criminal Appeal Act 1912 (NSW) to appeal against an interlocutory decision given by the primary judge, Bennett ADCJ, refusing an application that his Honour recuse himself on the ground of apprehended bias. In the exceptional circumstances referred to below, leave to appeal should be granted and the appeal should be allowed. However, it should be made clear at the outset that the outcome of the appeal is in no way a reflection upon the primary judge, and should not be understood to lend any credence to the allegations against him referred to below.

2. [2]

   Although the question of leave to appeal logically precedes the determination of the appeal itself, Mr Bechara’s application for leave depends heavily on the strength of his argument that there was error in the primary judge’s decision not to recuse himself. That being so, in order to understand the basis upon which leave should be granted, it is necessary first to consider Mr Bechara’s arguments going to error and the context in which they are made.

The charges and Mr Bechara’s substantive application for a stay or s 138 exclusion

1. [3]

   Mr Bechara is to be tried on indictment in the District Court of New South Wales in respect of a number of charges alleging his involvement in the supply of prohibited drugs over the period November 2022 through to February 2023. The principal charges are in summary:

   1. (1)

      Count 1: Ongoing supply of a prohibited drug between 14 November 2022 on 13 December 2022 contrary to s 25A(1) of the Drug Misuse and Trafficking Act 1985 (NSW) (DMTA).

   2. (2)

      Counts 2-7 (in the alternative to Count 1): Supply prohibited drug contrary to s 25(1) DMTA.

   3. (3)

      Count 8 (in the alternative to Count 1): Supply prohibited drug (commercial quantity) contrary to s 25(2) DMTA.

   4. (4)

      Counts 9, 14, 15, 16: Supply prohibited drug contrary to s 25(1) DMTA.

   5. (5)

      Count 10: Supply prohibited drug (large commercial quantity) contrary to s 25(2) DMTA.

   6. (6)

      Counts 11, 12, 13: Supply prohibited drug (commercial quantity) contrary to s 25(2) DMTA.

2. [4]

   The Crown case is that Mr Bechara supplied quantities of 1,4-butanediol, methylamphetamine, methylone and 3,4-methylenedioxymethylamphetamine to an undercover police operative (UCO); that the UCO was operating under authority of a controlled operation authorised under Part 2 of the Law Enforcement (Controlled Operations) Act 1997 (NSW); and that a civilian participant (CP) who introduced Mr Bechara to the UCO was also authorised under the controlled operation.

3. [5]

   Mr Bechara applied to have the proceedings against him permanently stayed or, in the alternative, to have the evidence constituting the charges excluded pursuant to s 138 of the Evidence Act 1995 (NSW). The grounds relied upon were misconduct by the police and by the CP. Mr Bechara challenges the “legality and/or propriety” of the controlled operation and the legality of the CP’s conduct. Mr Bechara alleges that the CP offered to introduce him to a “legitimate” buyer (i.e., the UCO) if Mr Bechara was sexually intimate with the CP; that that offer amounted to a fraudulent inducement; and that the CP thereby committed an offence against s 61I of the Crimes Act 1900 (NSW). The Court was also informed that Mr Bechara may be expected to advance other allegations of misconduct in either the controlled operation or the prosecution. Of potential relevance to the issues discussed below, Senior Counsel for Mr Bechara informed the Court that he could not exclude the possibility that an allegation of fabricating evidence might be made.

4. [6]

   Mr Bechara submits that issues in his application will include the reliability and credibility of police officers; alleged police misconduct; and alleged illegal conduct by the CP. It is contemplated that Mr Bechara may give evidence in the case, in which case the primary judge may need to make determinations about his reliability and credibility.

5. [7]

   The application was listed before the primary judge, commencing on 24 March 2026. On the first day of the hearing, Mr Bechara applied for an order that the primary judge disqualify himself. The primary judge heard that application on 25 March 2026 and indicated that he would not disqualify himself. His Honour delivered written reasons on 27 March 2026: Rex v Bechara [2026] NSWDC 66.

6. [8]

   Mr Bechara’s application for a permanent stay or for the exclusion of evidence pursuant to s 138 is now listed before the primary judge on 21 September 2026.

The circumstances giving rise to the application

1. [9]

   The application for recusal arose out of the following circumstances.

   1. (1)

      The Honourable RA Hulme AJ has been holding an inquiry pursuant to s 79(1)(a) of the Crimes (Appeal and Review) Act 2001 (NSW) (Inquiry) into the convictions, more than four decades ago, of men known as the Croatian Six for offences of conspiracy to bomb and possession of explosives.

   2. (2)

      The primary judge, who was then a Detective Senior Constable with the NSW Police, had been involved in the investigation of the Croatian Six. One of the petitioners at the Inquiry, by his Senior Counsel, has alleged serious misconduct on the part of the primary judge when he was a police officer. In particular, the argument in this Court proceeded on the basis that it has been expressly submitted to Hulme AJ that the primary judge fabricated evidence for use in the prosecution of the Croatian Six, and that he was aware that one of the Six had been beaten in an interview room at the CIB but did nothing about it.

   3. (3)

      The primary judge was a witness at the Inquiry, where he is legally represented. He denied the allegations. He was cross-examined. Although the evidence before this Court as to the submissions made at the Inquiry is incomplete, the argument was conducted on the basis that the petitioner has submitted to Hulme AJ that the primary judge lied on oath at the Inquiry.

   4. (4)

      The report of the Inquiry has not yet been delivered, and it is unknown when that will occur.

Applicable principles: apprehended bias

1. [10]

   The basic principles as to apprehended bias are well-established. The test was articulated by Gleeson CJ, McHugh, Gummow and Hayne JJ as follows in Ebner v Official Trustee in Bankruptcy (2000) 205 CLR 337; [2000] HCA 63 at [6]:

2. [11]

   As their Honours said at [7], “The question is one of possibility (real and not remote), not probability.” Edelman J described the test as relatively undemanding in QYFM v Minister for Immigration, Citizenship, Migrant Services and Multicultural Affairs (2023) 279 CLR 148; [2023] HCA 15.

3. [12]

   Applying the “double might” test nevertheless requires a degree of analytical specificity. Thus, it was said in Ebner at [8]:

4. [13]

   Gageler CJ, Gleeson, Jagot and Beech-Jones JJ described the three stages required by the Ebner test in Director of Public Prosecutions v Smith (2024) 419 ALR 212; [2024] HCA 32 at [92] as follows:

5. [14]

   See also Sunshine Loans Pty Ltd v Australian Securities and Investments Commission [2026] HCA 8 at [1] per Gageler CJ and Gleeson J.

6. [15]

   It is convenient to defer discussion of the principles relevant to the third stage — the reasonableness assessment — until after Mr Bechara’s argument has been explained.

Mr Bechara’s argument

1. [16]

   Mr Bechara identified “the question[s]” which the primary judge might “resolve other than on [their] legal and factual merits” as the reliability and credibility of police officers and the CP (and, if he gives evidence, of Mr Bechara); alleged police misconduct; alleged illegal conduct by the CP; and, if there was such misconduct, an evaluative judgment in order to determine whether its relative seriousness was enough to warrant the relief sought (a stay or s 138 exclusion).

2. [17]

   As to the first stage referred to in Smith, Mr Bechara identified “the factor which it is said might lead [the] judge to resolve the question other than on its legal and factual merits”, in different ways at different times. Overall, there were two clearly distinct bases on which Mr Bechara put his case.

3. [18]

   What emerged in argument as the primary “factor” relied on was the fact that the primary judge was facing unresolved allegations of serious misconduct as a police officer, as well as an allegation that he had lied on oath. That is, it was said that the fact of the allegations itself might in the circumstances lead his Honour to depart from resolving the issues before him on their merits, whether or not his Honour had engaged in any actual misconduct.

4. [19]

   As to the second stage identified in Smith, the logical connection between the fact that the judge was facing the allegations and the feared departure from impartial determination may be articulated as follows.

   1. (1)

      The allegations made against the primary judge of police misconduct, which included fabricating evidence, were very serious. As the primary judge himself appears to have accepted in the course of argument (at least with respect to hearing criminal proceedings), if the allegations were established his position as a judge would be untenable: Tcpt 24 March 2026 at 3.14.

   2. (2)

      The allegations were live issues for determination in a s 79(1)(a) judicial inquiry on foot, the outcome of which was not yet known.

   3. (3)

      The primary judge had been cross-examined on his denial of police misconduct at the Inquiry. The petitioner’s attack on his credit raised a further serious allegation: that his Honour had lied on oath to Hulme AJ.

   4. (4)

      In those circumstances, a fair-minded lay observer would (or at least might) expect even a professional judge to be highly sensitive to the allegations of police misconduct against him. A fair-minded lay observer might apprehend that his Honour might be unable to put to one side his own experience as the subject of allegations of police misconduct. That might be so in either of two ways.

   5. (5)

      First, his Honour might identify, even if only subconsciously, with the police officers against whom misconduct was alleged in Mr Bechara’s case. On that basis, a fair-minded lay observer might apprehend that his Honour might be unable to bring an impartial mind to the question whether the officers were guilty of misconduct and, if so, to assessing the seriousness and significance of that misconduct. The apprehension of bias in that case would be one that favoured the prosecution.

   6. (6)

      Alternatively, a fair-minded lay observer might apprehend that, in light of the serious allegations made against him, his Honour might be predisposed, again even subconsciously, to resolve the questions before him in a way that demonstrated that he would not tolerate police misconduct. Although the apprehension of bias in that case would be one that favoured the defence, that makes no difference as a matter of principle.

5. [20]

   There remains the third stage identified in Smith: whether, from the perspective of a fair-minded lay observer, the apprehension would be reasonable. Gageler CJ and Gleeson J said in Sunshine Loans at [40]:

6. [21]

   The knowledge of the legal, statutory and factual context to be attributed to the fair-minded lay observer may be extensive: see, for example, Smith at [95]. However, the significance of the professional judge’s “training, tradition and oath or affirmation” should be understood as having limits. The words quoted at [40] of Sunshine Loans were those of McHugh JA in Vakauta v Kelly (1988) 13 NSWLR 502 at 527. His Honour did not there suggest that a fair-minded lay observer would expect that a professional judge would always succeed in “discard[ing] the irrelevant, the immaterial and the prejudicial”. The point that his Honour made in the balance of the sentence from which those words are quoted was less absolute: that in the case of a professional judge, “a conclusion that there is a reasonable apprehension that he is biased should not be drawn lightly.” That is consistent with the observation in Ebner at [8]: “The apprehension of bias principle admits of the possibility of human frailty. Its application is as diverse as human frailty.”

7. [22]

   A fair-minded lay observer would understand that the primary judge, as a professional judge, was required to put aside, and would generally succeed in putting aside, his personal views and anything irrelevant or prejudicial arising from his ordinary life experiences. If an apprehension that a professional judge might not bring an impartial mind to bear was based only on such matters, the apprehension would generally not be reasonable. The range of ordinary experiences and circumstances is very broad. Although some personal views and ordinary life experiences may be felt more strongly than others, mere differences of degree are not enough. It will thus be a rare case in which some aspect of a judge’s personal experience alone (as opposed to some conduct or communication involving the judge) affords a reasonable basis for an apprehension of bias.

8. [23]

   The present situation, however, involves a difference in kind. Mr Bechara did not rely on any ordinary life experience. He relied on a wholly extraordinary one: a sitting judge, whose integrity must be unimpeachable, faces allegations of serious misconduct when he was a police officer; the allegations are heard at a judicial inquiry directed under statute; a finding of misconduct would render the judge’s position untenable; the judge is cross-examined about the allegations; his evidence is submitted to be false; and the inquirer’s decision is not yet known. Crucially, this was not an experience receding into the past. The allegations were hanging over the judge’s head at the very time he was required to put them from his mind in order to rule on allegations of misconduct made against serving police officers who were themselves to be cross-examined before him.

9. [24]

   In the singular circumstances described above, and particularly given the currency and circumstances of the unresolved allegations against the judge, a fair-minded lay observer might reasonably apprehend that even the most robust of professional judges might not bring an impartial mind to the resolution of the question the judge was required to decide.

10. [25]

    That is so despite the fact that the specific police misconduct alleged against the primary judge (fabricating evidence and condoning the bashing of a suspect) differs significantly from the central allegation in the stay / s 138 application before him (that the CP committed an offence against s 61I when he fraudulently induced Mr Bechara to be sexually intimate with him in return for the introduction to the UCO). (It should be noted that, while Senior Counsel could not exclude the possibility that Mr Bechara may also at some stage allege that evidence was fabricated, it would be impermissible speculation to determine the matter on the basis of mere possibilities.)

11. [26]

    The Ebner test acknowledges that a lay observer might have more than one reasonable response to the identified “factor”: “The question is one of possibility (real and not remote), not probability”: at [7]. It is true that a fair-minded lay observer might reasonably consider that the specific misconduct alleged against the police officers in the stay / s 138 application was sufficiently distinguished from that alleged against the primary judge that there was no real risk that his Honour would bring an impartial mind to the question before him. But that would not be the only reasonable response.

12. [27]

    It is in the nature of the asserted logical connection here that a fair-minded lay observer also might reasonably apprehend that there was a real possibility that the feared departure from deciding the case on its merits might operate at a somewhat less specific level of generality. In both cases, a police officer (or, in the judge’s case, a former officer) faces allegations of serious misconduct in the discharge of their duty as police. In both cases, the officer (or former officer) is cross-examined about their conduct. In both cases, there is a real prospect that their evidence will be criticised (in the primary judge’s case, that has already occurred).

13. [28]

    Those matters were sufficient for present purposes. The fair-minded lay observer might reasonably apprehend that his Honour might not bring an impartial mind to the resolution of the questions before him concerning alleged police misconduct.

14. [29]

    It is necessary to emphasise that the mere making of an unsubstantiated allegation against a sitting judge will not generally give rise to any apprehension of bias, still less a reasonable one. The fair-minded lay observer is taken to understand that professional judges are well capable of putting such matters from their minds. To take any other view would undermine the rationale for the apprehension of bias principle itself: “the requirement that justice should both be done and be seen to be done” (Ebner at [6]). That rationale compels the conclusion that the mere making of an allegation against a judge not suffice to require disqualification. Justice would not “be seen to be done” if a litigant could veto the judge to whom their matter was allocated by the simple expedient of making an unsubstantiated allegation, or procuring a third party to do so. That would amount to permitting litigants to choose their own judges.

15. [30]

    The application made to the primary judge was that he disqualify himself for apprehended bias. That was an issue going to jurisdiction: see QYFM at [26]-[27] per Kiefel CJ and Gageler J; [92] per Gordon J; [121] per Edelman J; cf [311] per Jagot J. The particular argument which has succeeded in this Court was not advanced in precisely that form before his Honour. But it is not suggested that if the argument had been articulated before the primary judge as it was in this Court, the matter could have been conducted in some different way that might have any bearing on the conclusion reached above. That being so, the fact that the particular argument was not put to the primary judge is not dispositive in this Court.

16. [31]

    There are statements in some authorities to the effect that the determination of questions of apprehended bias can involve evaluations of degree on which reasonable minds might differ: see, e.g., Livesey v New South Wales Bar Association (1983) 151 CLR 288; [1983] HCA 17 at 294, cited in QYFM at [31] per Kiefel CJ and Gageler J. There was some suggestion in the Crown’s submissions (Tcpt 6 May 2026 at 12.34-50) that the applicable standard of review, at least on a s 5F appeal, might be whether the decision was “reasonably open”, rather than the correctness standard (as to which, see GLJ v Trustees of the Roman Catholic Church for the Diocese of Lismore (2023) 280 CLR 442; [2023] HCA 32 at [16]-[17], [23] and [26] per Kiefel CJ, Gageler and Jagot JJ). The Crown submitted that it was unnecessary to resolve that issue because, on the Crown’s argument, the primary judge’s “decision was both reasonably open and correct”. The first part of that submission may strictly be right; the latter part is not.

17. [32]

    While the fact that the appeal is brought pursuant to s 5F means that the requirements for leave must be satisfied (as to which, see below), the procedural mechanism by which an appeal is brought does not alter the applicable standard of review once leave is granted. The presence of a reasonable apprehension of bias deprives the court of jurisdiction: see QYFM at [26]-[27], [92], [121]; cf [311]. The existence of jurisdiction is not a question about which the law will tolerate a diversity of reasonable views. To adopt the language used in GLJ, whether there is a reasonable apprehension of bias is a question which “admit[s] of but one uniquely right answer”: at [17]. The standard of review on appeal is thus the correctness standard. See, e.g., Frigger v State of Western Australia [2024] WASCA 159 at [60].

18. [33]

    But even if the question on appeal were whether the primary judge’s decision was reasonably open, the result would be the same, and for substantially the same reasons. Applying the Ebner test to the facts of this case, it was not reasonably open to conclude that no fair-minded lay observer might reasonably conclude that the primary judge might not bring an impartial mind to the questions before him.

19. [34]

    For the reasons given above, his Honour should have disqualified himself at the time the application was determined.

20. [35]

    It follows that, if leave to appeal were granted, the appeal would be allowed.

21. [36]

    Since Mr Bechara’s primary case of apprehended bias has been made out, it is unnecessary to determine whether his alternative case is established. However, for completeness, in summary, Mr Bechara’s alternative argument was as follows.

22. [37]

    Unlike Mr Bechara’s primary case (where the “factor” relied on is that the judge was facing serious allegations), his alternative argument depends on the possibility that the primary judge may be guilty of the misconduct alleged against him. It was expressed as follows in writing:

23. [38]

    That is to say, the “factor” relied on is actual misconduct. The logical connection between the alleged misconduct and “the apprehended deviation from deciding th[e] question[s] on [their] merits” is said to lie in the possibility that the primary judge’s determinations might be “influenced by his own experience and engagement in misconduct”.

24. [39]

    That gives rise to an immediate issue: the alleged misconduct is denied. The proper approach where the factor on which an application for disqualification for apprehended bias is based is disputed was discussed in CUR24 v Director of Public Prosecutions (2012) 83 NSWLR 385; [2012] NSWCA 65 at [38]-[52] per Meagher JA, Whealy JA agreeing. The context was an alleged out of court statement by a judge concerning a class of criminal behaviour; the statement was said to give rise to a reasonable apprehension of bias in the sense of prejudgment against those accused of crimes in the class. It was said:

25. [40]

    Taking that approach, the threshold question would be whether there was plausible evidence of misconduct by the primary judge. “[Taking] account of the fact of the dispute” (i.e., that the primary judge denied any misconduct), it would then be necessary to determine whether, from the fair-minded lay observer’s point of view, the alleged misconduct gave rise to a reasonable apprehension of bias. That analysis would involve identification of the logical connection between the misconduct and the feared departure from impartial adjudication.

26. [41]

    It is not only unnecessary to resolve those questions; it is also undesirable to do so, for the reasons which follow.

27. [42]

    As to the threshold question (plausible evidence of misconduct), Meagher JA in CUR24 said that “[t]he circumstances to which the fair-minded observer must be taken to have regard include those possibilities or likelihoods which will reflect the plausibility of the material relied upon to establish the relevant circumstances” at [44]. As has been seen, the fair-minded lay observer is taken to know a good deal about the proceeding in which the application for disqualification is made. But it is less clear what the fair-minded lay observer will be taken to know about the circumstances (including the forum) in which the allegation of misconduct is made. For the purposes of the plausibility assessment, is the fair-minded lay observer taken to know not only that the allegation has been made, but also by whom and in what circumstances? How far into the circumstances is the fair-minded lay observer taken to descend?

28. [43]

    For example, it was submitted for Mr Bechara that it is highly relevant that the allegations were made at the Inquiry by senior counsel. It was submitted that the fair-minded lay observer would understand the ethical constraints on counsel, including the provisions of the professional conduct rules applicable to barristers which require a proper basis before counsel may make allegations of serious misconduct. The submissions for Mr Bechara went still further, and contemplated (contrary to the Crown’s submissions) that the eminence of the particular senior counsel would be relevant to the plausibility assessment.

29. [44]

    For its part, the Crown submitted that it was relevant that it was only the petitioner at the Inquiry, who was plainly self-interested, who had made the allegations. The Crown emphasised that counsel assisting had not adopted them. Mr Bechara submitted that those matters were irrelevant.

30. [45]

    The answer to these questions may be expected to be found in the rationale of the apprehended bias rule and the purposes which the construct of the fair-minded lay observer serves. The ultimate purpose is maintaining public confidence in the administration of justice: QYFM at [45] per Kiefel CJ and Gageler J. It may be doubted that public confidence would be enhanced by a process of reasoning that required judges to assess a barrister’s reputation, competence or indeed compliance with the Bar Rules. But it would be preferable to leave those questions for resolution in a case in which it was necessary to do so. The questions which it is unnecessary to determine include:

    1. (1)

       what matters the fair-minded lay observer would be entitled to consider when assessing the plausibility of the allegations of misconduct made against the primary judge in this case;

    2. (2)

       in light of (1), whether the fair-minded lay observer might apprehend that, because of his “own [alleged] experience and engagement in misconduct”, the primary judge might not bring an impartial mind to the resolution of the questions before him;

    3. (3)

       in light of (2), whether any such apprehension would be reasonable.

31. [46]

    But it is worth noting the artificiality in Mr Bechara’s alternative argument. On the hypothesis that the judge was guilty of the “engagement in misconduct” alleged against him — for example, fabricating evidence for use in the prosecution of the Croatian Six or giving false evidence at the Inquiry — it is difficult to see how he could continue serving as a judge at all. But if that were the case, there would be no occasion for him to resolve the questions arising in Mr Bechara’s case which (Mr Bechara submits) the primary judge’s “own experience and engagement in misconduct” would lead him to deviate from deciding on their merits. His Honour acknowledged as much in the course of argument, as noted above.

Leave to appeal: applicable principles

1. [47]

   Mr Bechara’s appeal lies to this Court, if at all, only by leave pursuant to s 5F(3)(a). Much of the Crown’s argument was directed to the leave question.

2. [48]

   As entered in JusticeLink, the order the primary judge made (referring to himself as the Judge) was as follows:

3. [49]

   That is substantially similar to the order made by Pritchard J in Maules Creek Coal Pty Ltd v Environment Protection Authority (2023) 112 NSWLR 507; [2023] NSWCCA 275 (see at [3]):

4. [50]

   In that case a five-member bench (Leeming, Payne and Kirk JJA, Wilson and Fagan JJ) unanimously held that a decision to grant or refuse an application for disqualification based on bias is an interlocutory order capable of being the subject of an application under s 5F of the Criminal Appeal Act: [72]. Accordingly, subject to obtaining leave, Mr Bechara may appeal against the refusal of the disqualification application.

5. [51]

   However, there are strong reasons of policy for taking a strict approach to the grant of leave. Thus it was said in Maules Creek at [73]-[77]:

6. [52]

   There was some debate before us about whether it was necessary in every case to point to an error of principle or a matter of general importance before leave may (or should) be granted pursuant to s 5F. The Crown particularly relied on what the Chief Justice said in Koschier v R (2024) 113 NSWLR 491; [2024] NSWCCA 24 (Harrison CJ at CL and Chen J agreeing) at [45]-[51]:

7. [53]

   The Crown relied particularly on the words, “the need for ‘demonstration of error of principle … as well as the possibility or likelihood of substantial injustice’”. The Crown pointed out that similar formulations have been used in later cases.

8. [54]

   In Sayer-Jones v R [2024] NSWCCA 73, after referring at [80]-[81] to Koschier, N Adams J (Leeming JA and Campbell J agreeing) said at [83]: “The starting point is that leave would only be granted if the applicant can demonstrate an error of principle and the possibility of substantial injustice” (emphasis in original).

9. [55]

   In Burton v R [2024] NSWCCA 213, Yehia J (Kirk JA and Rothman J agreeing) noted at [29] that the “formulation” in Koschier — “the need for a demonstration of error of principle … and the possibility or likelihood of substantial injustice” — had been applied in Sayer-Jones. However, her Honour went on at [30]-[40] to consider a number of the same decisions discussed in Koschier, as well as others. These included, at [35], R v Matovski (1989) 15 NSWLR 720 (which had been the subject of the Bell CJ’s observation at [49] of Koschier). In Matovski, Gleeson CJ said at 723 (Lee CJ at CL and Grove J agreeing):

10. [56]

    Yehia J concluded her review of the authorities at [41]:

11. [57]

    In explaining why leave had been refused, her Honour said at [73]: “Not only is there no issue of principle raised by Ground 1, but there is no issue or error of principle that carries with it the possibility or likelihood of substantial injustice.” The Crown described that as “again … the application of the Koschier test”: Tcpt, 6 May 2026 at 15.46.

12. [58]

    Finally, in JK v R [2025] NSWCCA 44, Davies, Sweeney and Huggett JJ had referred at [38]-[39] to the parties’ submissions, which the Crown characterised as having been made by reference to “the Koschier test”: Tcpt 16.2. At [49], the Court explained that it had granted leave expressly “[a]pplying the test reaffirmed in Koschier”.

13. [59]

    The emphasis on this aspect of Koschier might have suggested that the Crown’s position was that there was a more or less rigid “test” for the grant of leave under s 5F(3)(a). Under that test (and putting to one side cases involving a matter of public importance or a conflict in prior decisions), it would be necessary to show both an error of principle and the possibility or likelihood of substantial injustice. The Crown did not, however, ultimately make that submission. Instead, consistently with what the Chief Justice had said at [49] of Koschier, the Crown acknowledged that it was not “suggesting that the discretion is fettered”: Tcpt, 6 May 2026 at 13.45. It was correct to do so.

14. [60]

    Although the language of s 5F(3)(a) is not confined, it is subject to limitations arising out of its scope, object and purpose. The ultimate purpose of the provision is to facilitate the administration of justice. But justice has many dimensions. The consequences for prosecution and defence, for victims, witnesses, and members of the community, and for the system of administration of justice as a whole, of permitting criminal proceedings to become fragmented and delayed are aspects of the administration of justice that cannot be ignored. There are thus powerful reasons of policy, derived from the nature of criminal proceedings and from what is necessary to do practical justice to those involved in and affected by them, for keeping a tight rein on interlocutory appeals in criminal cases. One purpose of s 5F(3)(a), embodied in the requirement for leave (what was described in Koschier as the “filter”), is to give effect to that policy. At the same time, another purpose of s 5F(3)(a), embodied in the provision of a mechanism for appeal, is to ensure that justice may be done in the relatively rare cases in which interlocutory appellate intervention is required. Bearing in mind that what justice may demand in a particular case is almost infinitely protean, giving effect to that purpose requires some degree of flexibility in the principles governing the grant of leave.

15. [61]

    When asked whether, if the Court was of the view that the interests of justice in the particular circumstances of a case required the grant of leave, the Court should nevertheless not grant leave unless the other criteria were satisfied, the Crown submitted:

16. [62]

    So much may be accepted; it follows from the existence of the leave requirement itself that mere error is insufficient. But to acknowledge that is not to deny that the interests of justice may be sufficiently engaged by error, when combined with other circumstances in a particular case, to justify the grant of leave.

17. [63]

    Nevertheless, it is abundantly clear on the authorities that, other than in unusual cases, leave to appeal should not be granted unless the case raises a question of principle or a matter of general importance.

Leave to appeal: application

1. [64]

   The point just considered — that leave to appeal may be granted even in the absence of an error of principle or of a matter of general importance — is arguably itself a question of principle warranting the grant of leave: cf Maules Creek at [74(1)].

2. [65]

   This is in any event an unusual case. Leave should be granted irrespective of any question of error of principle, on the ground that in the particular circumstances the interests of justice warrant the Court’s intervention.

3. [66]

   Where leave is sought on that basis, the strength of an applicant’s argument for error in the decision below will be highly material. Here, the substantive issue was fully argued, and the Court has come to the clear conclusion that the Ebner test is met. The question of leave to appeal should thus be approached on the footing that there was error in the primary judge’s decision not to recuse himself, and that his Honour lacked jurisdiction to hear Mr Bechara’s stay / s 138 application.

4. [67]

   Against that background, it is instructive to consider the other three reasons for granting leave pursuant to s 5F given in Maules Creek at [74] (remembering that it was decided on 3 November 2023):

5. [68]

   The first of those reasons (numbered 2) applies with even greater force in the present case. It is true that, having dismissed the recusal application, the primary judge did not go on to rule on the substantive stay / s 138 application. That means that no substantive decision has yet been affected by a reasonable apprehension of bias. But his Honour considers himself to be part-heard (the correctness of that view not being an issue before this Court), and the substantive application has been listed before him in September. The presence of a reasonable apprehension of bias is jurisdictional. On the hypothesis that the primary judge embarked on hearing the substantive stay / s 138 application so as to be part-heard, his Honour lacked jurisdiction to do so. If his Honour were to go on to determine the substantive stay / s 138 application, the want of jurisdiction to begin hearing the matter is an egg that cannot be unscrambled now, whatever else might occur before his Honour gave judgement.

6. [69]

   (Moreover, it is unknown when the Inquiry’s report will be delivered. But at least until that occurs, the circumstances giving rise to a reasonable apprehension of bias will persist on the basis of the primary “factor” on which Mr Bechara relies (to say nothing of his alternative case, which it has not been necessary for this Court to determine). Even putting to one side that his Honour had no jurisdiction to become part-heard, it would be highly undesirable to have that uncertainty hanging over the proceeding.)

7. [70]

   Given the conclusion reached above (that the primary judge should have disqualified himself at the time the application was determined), if leave to appeal were now refused, and in September his Honour went on to determine the substantive stay / s 138 application adversely to Mr Bechara, the proceedings would continue in circumstances in which all parties knew that there was error in the primary judge’s decision which would taint the whole proceeding. On a conviction appeal Mr Bechara would have a certain ground of appeal. For the matter to proceed on that footing would be worse than undesirable. That state of affairs can be avoided now by the simple expedient of granting leave to appeal.

8. [71]

   As to the second reason in Maules Creek (numbered 3), there is no trial underway; no date for trial has even been fixed. The substantive stay / s 138 application is not listed until September 2026. That leaves a large amount of time for another judge to hear the substantive application, without in any way affecting the progress of the proceeding more generally. There is no suggestion that granting leave now will in any practical sense fragment or delay the proceeding. While the delay in the proceedings until September is regrettable, it was the primary judge himself who adjourned the proceedings in March after announcing his decision, and that has already occurred. His Honour adjourned the matter after being informed that Mr Bechara proposed to seek leave to appeal the decision, on the basis that the “substantive application could not be heard further and an adjournment was necessary”: J[12]. The adjournment had been sought by Mr Bechara, but given his Honour’s ruling, it is not clear why the substantive application could not be heard further. In any event, it was the making of the s 5F application itself in March, rather than any leave to appeal granted by this Court in May, which caused the stay / s 138 application to be ultimately listed in September.

9. [72]

   As to the third reason (numbered 4), the concern in Maules Creek was to dispel any doubt about whether the trial judge should continue to hear the matter. That occurred in circumstances in which the Court went on to dismiss the appeal. The case for resolving the issue is even more compelling given the view reached above that the Ebner test has been satisfied and that any appeal should be allowed.

10. [73]

    There does not appear to be any countervailing consideration, other than the general need to exercise caution before granting leave to appeal from interlocutory decisions in criminal proceedings. But the force of the reasons that give rise to the need for caution is very substantially reduced in this case. No substantive decision has yet been made that would have to be remade. There is no reason to think that granting leave and allowing the appeal would fragment the proceeding or lead to any further delay.

Conclusion and orders

1. [74]

   Leave to appeal should be granted and the appeal allowed. The orders I propose are as follows.

   1. (1)

      Grant leave to appeal pursuant to s 5F(3)(a) of the Criminal Appeal Act 1912 (NSW).

   2. (2)

      Appeal allowed.

   3. (3)

      Vacate the order of his Honour Acting Judge Bennett of the District Court refusing the applicant’s application that he disqualify himself from hearing the matter.

   4. (4)

      Remit the proceedings to the District Court of New South Wales to be heard by a Judge other than his Honour Acting Judge Bennett.

2. [75]

   SWEENEY J: I agree with McHugh JA.

3. [76]

   EMMETT J: I agree with McHugh JA.