HIGH COURT OF AUSTRALIA FRENCH CJ, HAYNE, CRENNAN, KIEFEL, BELL AND GAGELER JJ ROSEANNE BECKETT APPLICANT AND THE STATE OF NEW SOUTH WALES RESPONDENT Beckett v New South Wales [2013] HCA 17 8 May 2013 S144/2012 ORDER 1. Special leave to appeal granted. 2. The appeal be treated as instituted and heard instanter and allowed with costs. 3. Set aside that part of order (b) of the Court of Appeal of the Supreme Court of New South Wales made on 2 May 2012 dismissing the appeal with costs and, in lieu thereof, order that: (a) appeal allowed with costs; and (b) the answer of Davies J to question A of the respondent's notice of motion filed on 16 May 2011 be set aside and, in lieu thereof, question A be answered "No". 4. The respondent pay the appellant's costs of the separate determination before Davies J. On appeal from the Supreme Court of New South Wales Representation G O'L Reynolds SC with S M Nixon and G R Rubagotti for the applicant (instructed by Turner Freeman Lawyers) B W Walker SC with W G Roser SC and P J Saidi for the respondent (instructed by Crown Solicitor (NSW)) Notice: This copy of the Court's Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports. CATCHWORDS Beckett v New South Wales Torts – Malicious prosecution – Elements – Whether proof of innocence required where proceedings terminated by entry of nolle prosequi – Whether entry of nolle prosequi terminates proceedings in favour of accused – Whether Davis v Gell (1924) 35 CLR 275 should be followed – Whether direction that no further proceedings be taken against person under s 7(2)(b) of Director of Public Prosecutions Act 1986 (NSW) equivalent to termination by entry of nolle prosequi. Words and phrases – "favourable termination of the prosecution", "malicious prosecution", "nolle prosequi". Director of Public Prosecutions Act 1986 (NSW), s 7(2). FRENCH CJ, HAYNE, CRENNAN, KIEFEL AND BELL JJ. This appeal is concerned with proof of the tort of malicious prosecution in a case in which the criminal proceeding giving rise to the claim is terminated by a direction from the Director of Public Prosecutions ("the Director") that no further proceedings be taken against the person. With one exception, the plaintiff's guilt or innocence of the criminal charge is not an issue in the action for malicious prosecution. The exception, allowed in Commonwealth Life Assurance Society Ltd v Smith on the authority of Davis v Gell, requires the plaintiff to prove his or her innocence at the trial of the civil action where the prosecution was terminated by the entry of a nolle prosequi by the Attorney-General ("the Davis exception"). The Attorney-General, or a person acting under the authority of the Attorney-General, may enter a nolle prosequi at any time after the indictment is signed and before the return of the verdict. The entry of the nolle prosequi brings the proceedings to a halt without determination of guilt. It does not bar the subsequent prosecution of the accused on the same charge. Section 7(2)(b) of the Director of Public Prosecutions Act 1986 (NSW) ("the DPP Act") confers the same functions on the Director as the Attorney-General with respect to directing that there be no further proceedings against a person who has been committed for trial. The two questions raised in the appeal are whether the Davis exception is good law and, if it is, whether the termination of a prosecution in the exercise of the Director's statutory power is within it. The tort of malicious prosecution The wrong for which the tort provides redress is the malicious instigation or maintenance of the prosecution of the plaintiff without reasonable and probable cause. The elements of the tort are set out in A v New South Wales. In summary, the plaintiff must prove four things: (1) the prosecution was initiated by the defendant; (2) the prosecution terminated favourably to the plaintiff; (3) the defendant acted with malice in bringing or maintaining the prosecution; and (4) the prosecution was brought or maintained without reasonable and probable cause. A v New South Wales considered the third and fourth of those elements. One aspect of that consideration which assumes importance in this appeal is the discussion of the temporal dimension of the tort: proof of the absence of reasonable and probable cause directs attention to the state of affairs at the time the defendant is alleged to have instigated or maintained the prosecution. Evidence bearing on the existence of reasonable and probable cause is confined to the material available to the defendant at the time the prosecution was commenced or maintained. The second element of the tort is a requirement of policy. Differing accounts of the rationale for the requirement are found in the early cases. It is said that a person should not be permitted to allege that a pending proceeding is "unjust", and that the possibility of a conflict in judicial decisions should not be allowed. The rationales for the rule evince the concern of the law with the consistency of judicial determinations, a concern that is distinct from proof of actual innocence or guilt: a plaintiff who is wrongfully convicted of an offence cannot maintain an action for malicious prosecution notwithstanding that he or she may possess irrefutable proof of innocence. The requirement that the prosecution has terminated avoids the possibility of conflict in the decisions of the court trying the criminal charge and the court trying the civil action. Any termination that does not result in conviction is favourable to the plaintiff for the purposes of the civil action. Prosecutions may terminate in a number of ways without verdict: the magistrate may not commit for trial; the Director may not find a bill of indictment; the Director may direct that no further proceedings be taken after a bill has been found; or the Attorney‑General may enter a nolle prosequi. The plaintiff has no control over the termination of the proceedings in any of these ways and in those circumstances it would be unjust to deprive him or her of the ability to recover for the tort. As Professor Salmond explained it: "What the plaintiff requires for his action is not a judicial determination of his innocence, but merely the absence of any judicial determination of his guilt." As will appear, the Davis exception was allowed in Smith because of uncertainty concerning whether the entry of a nolle prosequi terminates the prosecution, and not because the innocence of the plaintiff was an issue in the civil action. At this juncture it is convenient to turn to the decisions in Davis and Smith. Davis The prosecution in Davis was terminated by the entry of a nolle prosequi after Mr Gell entered a plea of not guilty to the presentment in the Supreme Court of Victoria. At the trial of Mr Gell's subsequent action for damages for malicious prosecution, the jury were directed to assume his innocence because the prosecution had terminated in his favour. The jury returned a verdict for Mr Gell. Mr Davis applied for a new trial contending that the direction to assume Mr Gell's innocence was wrong. He appealed from the dismissal of his application to the Full Court of the Supreme Court of Victoria. The Full Court dismissed the appeal, holding that any error had not occasioned a miscarriage of justice. From that decision Mr Davis appealed to this Court. Isaacs ACJ said that an element of the tort is that "[t]he prosecution must have been groundless". In this respect, his Honour relied upon the statements of Cleasby B in Johnson v Emerson that the prosecution must be "wholly unfounded" and "really without foundation". Isaacs ACJ explained that groundlessness in this context "means that the plaintiff in the civil action is innocent, because, the prosecution being groundless, there was, when all the circumstances are known, no real cause for it". The entry of a nolle prosequi established the favourable termination of the prosecution, but it remained for the plaintiff to prove his innocence to succeed in his action. In the case of a prosecution which terminated in the plaintiff's acquittal, the policy against permitting conflicting decisions gave rise to an irrebuttable presumption of innocence at the trial of the civil action. However, in the case of a prosecution which terminated favourably to the plaintiff without verdict, it remained for the plaintiff to prove his innocence in accordance with the ordinary rules of evidence. Gavan Duffy J agreed that proof of innocence was an element of the tort. The trial judge had been wrong to direct that the plaintiff's innocence was to be assumed from the entry of the nolle prosequi. His Honour found it unnecessary to consider the position in the case of a prosecution that terminated in the plaintiff's acquittal. Starke J held that innocence was an issue in the civil action on the authority of Cox v English, Scottish, and Australian Bank Ltd and Crowley v Glissan (No 2). These approved a statement made by Bowen LJ in Abrath v North Eastern Railway Co that "in an action for malicious prosecution the plaintiff has to prove, first, that he was innocent and that his innocence was pronounced by the tribunal before which the accusation was made". Starke J considered that public policy was against re-litigation of the issue of innocence in the civil action in a case in which the prosecution terminated in acquittal. Proceedings terminated by the entry of a nolle prosequi, the ignoramus of a grand jury, the refusal of a justice to commit for trial, some want of jurisdiction or technical defect in the indictment were all forms of termination favourable to the plaintiff, but none sufficed to prove innocence. At around the time of the decision in Davis, a court in India held that the plaintiff was required to prove his innocence in an action for malicious prosecution. On appeal to the Privy Council, Lord Dunedin characterised this requirement as one that was "quite erroneous". His Lordship observed that the action required proof "[t]hat the proceedings complained of terminated in favour of the plaintiff if from their nature they were capable of so terminating". Smith Mr Smith was committed for trial on a charge of defrauding the Commonwealth Life Assurance Society Ltd. The prosecution was terminated by the decision of the Attorney-General not to file an indictment. In a subsequent action for damages for malicious prosecution, Mr Smith sought to adduce evidence of his innocence. The defendant objected on grounds of relevance. The trial judge, on the authority of Davis, admitted the evidence. On appeal in this Court, the question squarely raised was whether the plaintiff's guilt or innocence is in issue in the civil action. The majority held it was not. In so holding, Rich, Dixon, Evatt and McTiernan JJ concluded that Bowen LJ's dictum in his ex tempore reasons in Abrath was intended to state no more than that the plaintiff must prove a favourable termination of the prosecution in order to maintain the action. Their Honours observed that none of the pleading precedents and none of the text-writers suggested that proof of innocence was required in the civil action. They concluded: "Except in the case of a nolle prosequi covered by the decision in Davis v Gell, we are of opinion that the guilt or innocence of the plaintiff is not an issue going to the cause of action in malicious prosecution." Turning to the decision in Davis, their Honours observed that in Balbhaddar Singh v Badri Sah Lord Dunedin had adverted to the position of the nolle prosequi as "possibly unsettled". They quoted the statement in Goddard v Smith that the entry of a nolle prosequi "only puts the defendant without day", and said that there was "some uncertainty as to the sufficiency of a nolle prosequi". The uncertainty was with respect to whether the entry of a nolle prosequi stayed proceedings permitting "fresh process [to] be awarded in the same indictment". It was uncertainty as to the termination of the prosecution. Starke J adhered to the view he had expressed in Davis, that proof of the falsity of the charge is an essential element of the cause of action. Balbhaddar affirmed that termination of a prosecution by the refusal to commit for trial (as had occurred in that case) conclusively established the falsity of the charge. Turning to other forms of termination, his Honour concluded that it would not be consistent with principle or public policy that the decisions of magistrates, grand juries or the Attorney-General should be tried again on the merits and "blowed off by a side wind". It followed that Mr Smith's innocence was conclusively presumed from proof that the Attorney-General had decided not to find a bill of indictment against him. The evidence adduced to establish Mr Smith's innocence at the trial of the civil action was irrelevant. Smith held that the statements in Davis, that innocence was an issue in the civil action requiring proof in cases in which the prosecution had not been terminated by acquittal, were not to be followed. Davis had been decided on the effect of a termination by nolle prosequi, and on principle and on the authority of Balbhaddar it was not to be extended further. Procedural history At this point there should be some reference to the history of the present proceedings. The appellant was arrested by members of the New South Wales Police Force and charged with a number of offences against her husband. She was committed to stand trial in the Supreme Court of New South Wales. A bill of indictment charging the appellant with nine counts was found and she was arraigned upon it. The eighth count was preferred ex officio. At the conclusion of the appellant's trial on 11 September 1991 the jury returned verdicts of guilty on counts 1, 2, 3, 4, 6, 7 and 9, and on an alternative charge to the offence charged in count 5. A verdict of not guilty was returned respecting the offence charged in count 8. In October 1991, the appellant was sentenced to a term of imprisonment of twelve years and three months with a non-parole period of ten years and three months. She appealed unsuccessfully against her convictions and sentence to the New South Wales Court of Criminal Appeal. In 2001, the appellant petitioned the Governor seeking a review of her convictions. The Attorney-General referred the application to the Court of Criminal Appeal. The Court of Criminal Appeal remitted the determination of a number of factual questions to Acting Judge Davidson. Following the delivery of Davidson ADCJ's findings, on 17 August 2005 the Court of Criminal Appeal allowed the appeal in relation to counts 1, 2, 5, 6, 7 and 9 and quashed each conviction. The Court entered a verdict of acquittal on count 9. A new trial was ordered on counts 1, 2, 5, 6 and 7. The appellant's appeal against her convictions for the offences charged in counts 3 and 4 was dismissed. On 22 September 2005, the Director directed that there be no further proceedings against the appellant on the outstanding charges that were the subject of the Court of Criminal Appeal's order for a new trial. On 26 September 2005, a document communicating the Director's determination was forwarded to the Registry of the Court of Criminal Appeal. The civil proceedings – the determination of two separate questions On 15 August 2008, the appellant instituted proceedings against the respondent in the Common Law Division of the Supreme Court claiming damages for malicious prosecution on the basis that the respondent was vicariously liable for the conduct of the police officers who instigated the prosecution. The respondent filed a notice of motion on 16 May 2011 seeking a separate determination in relation to two questions: "A. With respect to each of the counts 1, 2, 5, 6 and 7 for which the plaintiff was tried: i. Accepting that the proceedings terminated in favour of the plaintiff, to the extent that the plaintiff's claim for malicious prosecution is based upon each of these counts, does the plaintiff need to prove her innocence in relation to each count to succeed? B. With respect to count 9 for which the plaintiff was tried: i. To the extent that plaintiff's the claim [sic] for malicious prosecution is based upon this count does the plaintiff need to prove her innocence of the charge?" The primary judge (Davies J) agreed to the separate determination of the two questions. His Honour said that the order quashing the appellant's convictions and directing a new trial on the specified counts meant that the issues "raised by the indictment upon which those counts were tried will remain justiciable". He held that the indictment on which the appellant had been tried was extant. His Honour concluded that the notification to the Registry of the Court of Criminal Appeal of the Director's decision to take no further proceedings against the appellant was the equivalent of the entry of a nolle prosequi. He held that he was bound to apply the Davis exception. He answered the questions as follows: "A. Yes. B. No." The appellant appealed to the Court of Appeal of the Supreme Court of New South Wales (Beazley and McColl JJA and Tobias AJA) against the answer to question A and the respondent cross-appealed against the answer to question B. The Court of Appeal agreed with the primary judge that the direction under s 7(2)(b) of the DPP Act constituted the entry of a nolle prosequi. It followed that the primary judge had been right to conclude that the Davis exception applied. The appeal and the cross-appeal were dismissed. The appellant applied for special leave to appeal. On 5 October 2012, Gummow, Hayne and Heydon JJ referred the application into an enlarged Full Court for hearing as on appeal. For the reasons to be given, special leave should be granted and the appeal allowed. It is convenient in these reasons to refer to the application as an appeal and to the applicant as the appellant. The grounds of challenge The appellant submits that there is no principled basis for distinguishing proceedings terminated by the entry of a nolle prosequi from those terminated in any of the other ways favourable to the plaintiff. She seeks leave to re-open Davis. Alternatively, she contends that the Davis exception was wrongly applied in the trial of her action because the prosecution of the charges against her was terminated under s 7(2)(b) of the DPP Act and not by the entry of a nolle prosequi by the Attorney-General. The appellant's submissions in support of her alternative ground departed from the submissions on which she relied below. In this Court she did not maintain that the indictment was spent at the time the s 7(2)(b) determination was communicated to the Registry of the Court of Criminal Appeal. Her case is that the Davis exception is confined to "the special case" of proceedings terminated by nolle prosequi. The termination of proceedings by the Director under statutory power is said to be of a different character: the source of the power is different and the statutory power is wider. The functions of the Attorney-General with respect to the prosecution of criminal offences derive from Imperial statutes. However, it may be accepted that the purpose of the provision of the Australian Courts Act 1828 (Imp), which conferred power on the Attorney-General to prosecute offences on ex officio indictment, was to arm the Attorney-General with a power in all respects similar to that enjoyed by the Attorney-General in England. That power included the entry of a nolle prosequi. The importance of the functions that are incidents of the office of the Attorney-General is recognised in s 38 of the Constitution Act 1902 (NSW). The Director's functions with which the appeal is concerned are set out in the DPP Act as follows: "7 Principal functions (1) The principal functions and responsibilities of the Director are: (a) to institute and conduct, on behalf of the Crown, prosecutions (whether on indictment or summarily) for indictable offences in the Supreme Court and the District Court, ... (2) The Director has the same functions as the Attorney General in relation to: (a) finding a bill of indictment, or determining that no bill of indictment be found, in respect of an indictable offence, in circumstances where the person concerned has been committed for trial, (b) directing that no further proceedings be taken against a person who has been committed for trial or sentence, and (c) finding a bill of indictment in respect of an indictable offence, in circumstances where the person concerned has not been committed for trial." The Director may not delegate the exercise of the function conferred on him under s 7(2)(b) save to a Deputy Director. The Director may not, without the consent of the Attorney-General, exercise a function in a manner that is inconsistent with the manner in which the Attorney-General has already exercised a function in the same matter. The Attorney-General's power to enter a nolle prosequi is unaffected by the DPP Act. The appellant relies on the decision of the Court of Criminal Appeal of New South Wales in GKA for the proposition that the Director's power to exercise the function under s 7(2)(b) is wider than the function of entering a nolle prosequi. The holding in GKA was that once the Director's direction under s 7(2)(b) was communicated to the court, the court was deprived of power to proceed further "upon the current indictment". The passage on which the appellant relies is set out below: "The power to direct a nolle prosequi is the same as one power referred to in ss 7(a) and 27(a) of the DPP Act. A determination of a 'no bill of indictment' would not prevent the bringing of a further indictment. The substance of the power contemplated by ss 7(2)(b) and 27(b) is wider than a nolle prosequi because it constitutes a direction that no further proceedings be taken against a person who has been committed for trial or sentence. Nonetheless, the power conferred by s 7(2)(b) includes a power to require entry of a nolle prosequi. It is not necessary in this proceeding to further define the scope of the power conferred by s 7(2)(b)." It is not clear in what respect the Court in GKA considered the power under s 7(2)(b) to be wider than the power to enter a nolle prosequi. The functions in s 7(2)(a) and (b) respecting the termination of proceedings in the case of a person who has been committed for trial are those of determining that no bill of indictment be found (par (a)) and directing that no further proceedings be taken against the person (par (b)). Each terminates the prosecution without barring the subsequent prosecution of the person for the same offence. The power to enter a nolle prosequi is not engaged until a bill is found and the indictment is signed. While the power under par (b) is not confined to a direction after a bill has been found, when the two paragraphs are read together it is apparent that the power to terminate proceedings by declining to find a bill of indictment is found in par (a). The appellant submits that the reference to the width of the statutory power in GKA recognises that a direction under s 7(2)(b) brings proceedings on an indictment to an end, whereas the entry of a nolle prosequi merely stays the proceedings on the indictment sine die. The appellant relies on the statements in Smith that the nolle prosequi "does no more than bring the trial to an end" and that "fresh process may be awarded in the same indictment and the prisoner again put on his trial". Nolle prosequi – a stay of proceedings sine die? It will be recalled that, in Smith, it was uncertainty concerning the sufficiency of a nolle prosequi to terminate the prosecution that was held to justify the Davis exception. In their joint reasons, Rich, Dixon, Evatt and McTiernan JJ quoted Professor Winfield's A Text-Book of the Law of Tort: "The effect of a nolle prosequi (staying by the Attorney-General of proceedings on an indictment) is open to question. An old case indicates that it is not a sufficient ending of the prosecution because it still leaves the accused liable to be indicted afresh on the same charge. But this seems inconsistent with the broad interpretation put upon 'favourable termination of the prosecution' which signifies, not that the accused has been acquitted, but that he has not been convicted." Their Honours also noted Professor Winfield's reference to American authorities, which were to the same effect as the decision of the New South Wales Supreme Court in Gilchrist v Gardner. In Gilchrist, Darley CJ said the entry of a nolle prosequi "puts an end to that prosecution, though [the Attorney‑General] may afterwards cause a fresh prosecution to be commenced". Windeyer J in the same case said that the entry of a nolle prosequi "put an end altogether to the prosecution", although it may be commenced anew on a fresh indictment. It was not necessary in Smith to resolve the controversy concerning the effect of the entry of a nolle prosequi. It was sufficient to note that uncertainty in this respect had led to proceedings terminated by nolle prosequi constituting a "special case". This category of special case was "covered" by the decision in Davis. In Question of Law Reserved on Acquittal (No 3 of 1995), Debelle J considered the power to enter a nolle prosequi. In question in that case was the power of the court to refuse to accept the entry of a nolle prosequi. His Honour observed that in England the form of endorsement of a nolle prosequi is of a stay postponing the proceedings sine die. He said that it "seems that it is possible for the Crown later to proceed on the original indictment or information", although his Honour noted that the usual practice is to issue a fresh indictment or information. These statements were based on Professor Edwards' account of the incidents of a nolle prosequi in The Attorney General, Politics and the Public Interest: "[T]he effect of a nolle prosequi is neither a bar to a fresh indictment nor a discharge of the original offence. What it does is to postpone sine die the prosecution. Should the Attorney General decide at a later date to reopen the original charges he can, theoretically speaking, reactivate the earlier indictment that was placed in suspension when the nolle prosequi was filed in the court's records. Alternatively, fresh proceedings leading to a new indictment can be commenced to which the accused will be precluded from raising a plea of autrefois acquit on the basis of the nolle prosequi." Three authorities are cited for these propositions: Goddard v Smith, R v Ridpath and R v Allen. Each is authority for the proposition first stated. There is, however, tension between Goddard and Allen with respect to the proposition that the indictment may be reactivated. Early text-writers differed on the question. Chitty's Criminal Law stated that the effect of a nolle prosequi was to permit the defendant to be re-indicted and "even upon the same indictment fresh process may be awarded". The authority given for the proposition was Goddard. Archbold's Practice of the Crown Office stated that the entry of a nolle prosequi "has the effect of putting an end to the prosecution altogether". Goddard was decided in 1704 and the several reports of the decision are not easy to reconcile. It was an action on the case for conspiracy maliciously to cause the plaintiff to be indicted for barratry. The plaintiff's declaration pleaded that the indictment had been "according to law in a due and lawful manner thereof discharged". At the trial the plaintiff produced a nolle prosequi by the Attorney-General. The question reserved for the consideration of the Judges of the King's Bench was whether proof of the nolle prosequi was sufficient to maintain the declaration. In the course of argument Holt CJ said: "that the entering a nolle prosequi was only putting the defendant sine die, and so far from discharging him from the offence, that it did not discharge any further prosecution upon that very indictment, but that, notwithstanding, new process might be made out upon it; and sure it is hard to allow a man who gets off by a nolle prosequi to maintain an action for a malicious prosecution." Goddard held that production of the nolle prosequi did not prove the material part of the declaration. It is less clear that it is authority for the proposition that the entry of a nolle prosequi stays proceedings, permitting the prosecutor to proceed upon the same indictment at a later date. A later report of the decision records that "the nonpross is not a discharge of the crime, but only of the indictment". Mr Harcourt, the Master of the Crown Office, appears to have informed the Court of the practice respecting the plea of nolle prosequi. The report concludes: "[A]nd they went upon Mr Harcourt's report, that they used to indict them again, and not to proceed upon the same indictment." In Allen, Crompton J observed that "the nolle prosequi being on the record, there is an end of this prosecution; but the question remains whether that is final or not". His Lordship considered that Archbold was correct and that the nolle prosequi "has the effect of putting an end to the prosecution altogether". He thought that Goddard only decided that the entry of a nolle prosequi is not a decision on the merits, observing of the report of the decision in volume six of the Modern Reports: "[T]he Court, in the course of the argument, said that the Attorney General might issue new process upon the indictment; but, as I have said, I rather think the nolle prosequi puts an end to the prosecution." It is far from clear when the older authorities speak of the entry of a nolle prosequi as "putting the defendant sine die" that more is being said than that it does not bar a subsequent prosecution. It is rare for proceedings to be revived after termination by nolle prosequi, and there is an absence of authority on the point. The preferable view, which accords with practice, is that stated in Allen: the entry of a nolle prosequi brings proceedings on the indictment to an end without barring a subsequent prosecution on a fresh indictment. The prosecution of indictable offences in all Australian jurisdictions is now conferred on a statutory office holder, the Director of Public Prosecutions. In each jurisdiction, the Director of Public Prosecutions has power to terminate the prosecution of proceedings on indictment. In each jurisdiction, the Attorney-General retains the power to enter a nolle prosequi, although in the majority of them the power is now sourced in statute. In New South Wales, Victoria and South Australia the Attorney-General's power in this respect continues as an incident of office. As a matter of practice, the occasions on which the prosecution of a person is terminated by the Attorney-General entering a nolle prosequi are likely to be rare. The appeal should not be decided on the narrow footing that the appellant's prosecution was terminated under the statutory power, and not in the exercise of the Attorney-General's prerogative power. The power under s 7(2)(b) is in substance and effect the same as the power to enter a nolle prosequi. A principled distinction between termination under s 7(2)(b) and other forms of termination? The joint reasons in Smith drew a distinction between the functions of the Attorney-General when declining to find a bill following a committal for trial and when terminating proceedings by the entry of a nolle prosequi. The same distinction is evident in the division of functions in s 7(2)(a) and (b). The respondent submits that the distinction provides a principled basis for the Davis exception, at least in some cases terminated by entry of a nolle prosequi or by direction under s 7(2)(b). The respondent's argument is that the determination of whether or not to find a bill of indictment involves an assessment of the legal merit of the prosecution. By contrast, the respondent points out that the decision not to take further proceedings against a person may be unconnected to the strength of the prosecution case. The decision to take no further proceedings against the appellant is said to be such a case: the appellant had served all but a few months of the non-parole period of the sentence imposed on her in 1991 at the time the Court of Criminal Appeal quashed her convictions and ordered a new trial. The inference is open that the Director determined to take no further proceedings against her for utilitarian reasons having nothing to do with the cogency of the prosecution case. The respondent says that it would be a scandal in the administration of justice to permit recovery of damages for malicious prosecution in circumstances in which a nolle prosequi has been entered on a strong prosecution case. The respondent calls in aid decisions of the Supreme Court of Illinois, in which the circumstances surrounding the entry of a nolle prosequi have been examined to determine whether its entry is "not indicative of the innocence of the accused". However, the decisions are concerned with proof of the termination of the prosecution. In the same connection, the Supreme Court of Canada in Miazga v Kvello Estate commented that "a live issue may arise whether the termination of the proceedings was 'in favour' of the plaintiff" in the case of a termination that is not an adjudication on the merits, such as a settlement or a plea bargain. In Miazga the elements of the tort were stated consistently with the statement in A v New South Wales. Decisions in other common law jurisdictions also accord with A v New South Wales in describing the tort of malicious prosecution as comprising the four elements summarised earlier in these reasons. The respondent's argument did not identify to which of those elements proof of the plaintiff's innocence is relevant in a case terminated for any reason by the entry of a nolle prosequi (or under the equivalent statutory power). The respondent's submission wrongly assumes that other forms of termination favourable to the plaintiff incorporate an element of "merit assessment". The termination of a prosecution may be for a technical reason that is unconnected to the strength of the prosecution case. The termination is nonetheless one favourable to the plaintiff such as to maintain the civil action. For example, the decision not to find a bill of indictment may be taken for reasons which are not connected to the strength of the prosecution case. The requirement that the plaintiff prove favourable termination, as earlier explained, is concerned with consistency of judicial decisions. Proof of favourable termination does not involve an inquiry into the underlying merits of the prosecution. The respondent was right to acknowledge in drafting the separate question that the prosecution had terminated in favour of the appellant. The respondent's submissions are apt to overlook two things. First, the appellant must prove the absence of reasonable and probable cause before she can recover in the civil action. Secondly, for whatever reason, the appellant has not been convicted of the offences charged in counts 1, 2, 5, 6 and 7 of the indictment. In the event that the appellant is able to prove that her prosecution by persons for whom the respondent is vicariously liable was instigated or continued maliciously and without reasonable and probable cause, her recovery in the civil action would not scandalise the administration of justice. The circumstances in which this Court will depart from a previous decision were the subject of recent consideration in Wurridjal v The Commonwealth. The appellant is right to say that the reasoning in Davis is undermined by subsequent authority. As has been observed, the conclusion that Mr Gell was required to prove his innocence did not stem from doubts that the entry of the nolle prosequi was a favourable termination of the prosecution. All three Justices accepted that it was. Their reasoning, that proof of innocence was an element of the tort, cannot stand with Smith and A v New South Wales. There is no principled reason to distinguish a prosecution terminated by the entry of a nolle prosequi by the Attorney-General or a direction by the Director under the statutory power from other forms of termination falling short of acquittal. The Davis exception produces an anomalous outcome. Were the decision of the Court of Appeal to stand, the appellant would be entitled to lead evidence at the trial of matters tending to establish her innocence that were unknown to the respondent at the time the prosecution was commenced and maintained. However, material that has come to light since the prosecution was commenced and maintained tending to establish her guilt would not be admissible on the issue of reasonable and probable cause. The unsatisfactory nature of that outcome was noted in Smith: "In the course of proving facts on which he based the prosecution, the defendant may sometimes succeed in raising a doubt of the plaintiff's innocence. When this happens an absence of reasonable and probable cause is hardly likely to be found. But it would be surprising if a defendant could go into the guilt or innocence of the plaintiff as a separate issue, though on the issue of reasonable and probable cause he is not permitted to prove facts which he did not know at the time of the prosecution even when the facts amount to the highest degree of objective cause for the prosecution, namely, proof of the real guilt of the accused." Davis should not be followed. Orders There should be orders as follows: 1. Special leave to appeal granted. 2. The appeal be treated as instituted and heard instanter and allowed with costs. 3. Set aside that part of order (b) of the Court of Appeal of the Supreme Court of New South Wales made on 2 May 2012 dismissing the appeal with costs and, in lieu thereof, order that: (a) appeal allowed with costs; and (b) the answer of Davies J to question A of the respondent's notice of motion filed on 16 May 2011 be set aside and, in lieu thereof, question A be answered "No". 4. The respondent pay the appellant's costs of the separate determination before Davies J. GAGELER J. The majority holding in Commonwealth Life Assurance Society Ltd v Smith was that "[e]xcept in the case of a nolle prosequi covered by the decision in Davis v Gell … the guilt or innocence of the plaintiff is not an issue going to the cause of action in malicious prosecution". The question is whether the exception stated in that holding should be maintained. It should not. Davis should be overruled. The common law of Australia should be declared to be that the guilt or innocence of the plaintiff is never an issue going to the cause of action in malicious prosecution. The reasoning in Davis was founded on the proposition that it was an element of every cause of action in malicious prosecution that the plaintiff was innocent. The reasoning was that termination of a prosecution by conviction or acquittal created a res judicata foreclosing any issue of guilt or innocence. Termination by other means left innocence to be proved by the plaintiff. Entry of a nolle prosequi terminated the prosecution, but did not create a res judicata. In the case of a prosecution terminated by entry of a nolle prosequi, it therefore remained for the plaintiff in an action for malicious prosecution to lead evidence to establish the fact of innocence. No part of that reasoning survived Smith. The proposition that it was an element of every cause of action in malicious prosecution that the plaintiff was innocent was rejected. But, it was said, "[t]here was controversy" about whether entry of a nolle prosequi was sufficient to establish another element of the cause of action in malicious prosecution. That other element was that the prosecution proceedings must have terminated in the plaintiff's favour. The controversy was "as to what terminated proceedings, as, eg, whether a nolle prosequi … was a termination". It was seen not to be necessary to resolve that controversy in Smith because the case was not one where the prosecution proceedings had been terminated by a nolle prosequi. The case was one where the proceedings had been terminated by refusal to file an indictment. Three reasons were nevertheless identified for the rule that the prosecution proceedings must have terminated in the plaintiff's favour. The first was to prevent the collateral questioning of a conviction. The second was to prevent imputations in one proceeding against the justice of another proceeding still pending. The third, "from which the conclusion in Davis … was deduced", was that only a terminated proceeding could be shown to be without foundation. It was sufficient for the purpose of the decision in Smith to note that those three reasons for the rule did "not affect the nature and application of the rule itself" and did not have "an independent and further operation in imposing some additional condition as a necessary element in the cause of action for malicious prosecution". It was held that, contrary to the conclusion deduced in Davis, the third reason for the rule did not turn on equating acquittal, or any other form of termination, with the establishment of innocence. The decision in Davis "covered" only a case of a nolle prosequi and, on principle, was not to be "extended further". Contemporaneous commentary described the attitude of the majority in Smith to the decision in Davis as "that of those who come neither to praise nor yet to bury". The time to bury Davis is now. The controversy referred to in Smith about the effect of a nolle prosequi has now long been resolved. At common law, the entry of a nolle prosequi terminates proceedings on an indictment even though it does not prevent new proceedings being brought on a new indictment. Termination of prosecution proceedings by entry of a nolle prosequi should be held to be sufficient to establish the element of the cause of action in malicious prosecution that requires prosecution proceedings to have terminated in the plaintiff's favour. The Full Court of the Supreme Court of New South Wales properly so held, after Smith, in Mann v Jacombe. To treat termination of proceedings by entry of a nolle prosequi as a termination in the plaintiff's favour for the purpose of an action in malicious prosecution is consistent with the first two reasons identified in Smith for the rule that prosecution proceedings must have terminated in the plaintiff's favour. It is no less consistent with the third reason than was the form of termination in Smith itself, being refusal to file an indictment. Principle as articulated in Smith told against the decision in Davis being extended. It tells equally against the decision in Davis being maintained. Even if entry of a nolle prosequi were insufficient to establish the element of malicious prosecution that the prosecution proceedings have been terminated in the plaintiff's favour, the principled result would not be to apply Davis. The principled result would rather be that no action for malicious prosecution would lie. There is no principled reason why the absence of one element of the cause of action in malicious prosecution should be capable of being remedied by the importation of another element that otherwise forms no part of the cause of action. In the present case, Ms Beckett brings an action in malicious prosecution against the State of New South Wales. At the request of the State of New South Wales questions were reserved for separate determination in the Supreme Court of New South Wales. The questions expressly accepted that, insofar as prosecution proceedings against Ms Beckett had terminated by direction of the Director of Public Prosecutions under s 7(2)(b) of the Director of Public Prosecutions Act 1986 (NSW) ("the DPP Act"), those prosecution proceedings had terminated in her favour. The questions asked whether Ms Beckett needs to prove her innocence. The answer is that Ms Beckett does not need to prove her innocence. The primary judge (Davies J) and the Court of Appeal (Beazley and McColl JJA and Tobias AJA) gave a different answer because, being bound by Davis, they held that no principled distinction could be drawn between a case of proceedings terminated by the entry of a nolle prosequi and a case of proceedings terminated by direction under s 7(2)(b) of the DPP Act. The overruling of Davis makes any question about whether such a distinction can or should be drawn redundant. Without error on the part of the Court of Appeal, special leave to appeal should be granted and the appeal should be allowed. For these reasons, I agree with the orders proposed in the joint reasons for judgment.