Commonwealth Director of Public Prosecutions (CDPP)

Prosecutorial discretion in Australia today

Date of Publication: 
18 April 1996

Author: Rozenes, M. Date: 18/04/1996 Venue: Prosecuting justice, 18 and 19 April 1996, Melbourne [Australian Institute of Criminology Conference]

Any discussion of prosecutorial discretion today cannot sensibly be confined to an examination of how it is decided who gets prosecuted and, if a prosecution is commenced, how it is conducted. An important consideration in any understanding of how prosecutorial discretion is exercised in Australia today is the arrangements or system if you like for conducting prosecutions; in short who makes the decisions whether and how a prosecution should proceed. Accordingly, before discussing how prosecution decisions are made, may I first deal with who in fact makes those decisions. In a passage which has since been referred to many times the Philips Commission in its 1981 Report stated that a prosecution system should be judged by broad standards of fairness, openness and accountability, and efficiency: "Is the system fair; first in the sense that it brings to trial only those against whom there is an adequate and properly prepared case and who it is in the public interest should be prosecuted ..., and secondly in that it does not display arbitrary and inexplicable differences in the way that individual cases or classes of case are treated locally or nationally? Is it open and accountable in the sense that those who make the decisions to prosecute or not can be called publicly to explain and justify their policies and actions as far as that is consistent with protecting the interests of suspects and accused? Is it efficient in the sense that it achieves the objects that are set for it with the minimum use of resources and the minimum delay? Each of these standards makes its own contribution to what we see as being the single overriding test of a successful system. Is it of a kind to have and does it in fact have the confidence of the public it serves?1 Judged against such standards, how do the prosecution arrangements in Australia fare? This is the second conference that the Australian Institute of Criminology has organised which is concerned with examining aspects of the prosecution process in Australia. The first conference was held in 1984 and a number of the papers delivered at that conference referred to a passage from the ALRC's Report on Sentencing which had been released a few years earlier in which the Commission stated: "The process of prosecutions in Australia at both State and Federal level is probably the most secretive, least understood and most poorly documented aspect of the administration of criminal justice."2 As a general comment on the prosecution system applying in Australia at the time, it was perhaps not that wide of the mark. At that stage only two jurisdictions, Victoria and the Commonwealth, had established the Office of Director of Public Prosecutions, and for the most part the arrangements applying in this country were little different to those which had evolved in the last century, with the police having the conduct of most prosecutions before the summary courts and Crown law authorities only becoming involved in indictable matters, and then only once a committal order had been obtained. Since then, of course, the Office of DPP has been established in all Australian jurisdictions, and the emergence of that Office as the principal prosecuting authority has undoubtedly resulted in many beneficial changes in the prosecution process in Australia. There has been a "depoliticisation" of prosecutorial decision-making, at least in relation to those matters for which the DPPs are responsible, and there has been increased professionalism in the conduct of cases before the superior courts. The arrangements for conducting prosecutions in this country are now fairer and more open than was the case in the early 1980s. In particular, each DPP office now operates under a set of published guidelines which are to be followed in the making of prosecution decisions. The main purpose of those guidelines is to promote consistency in the making of the various decisions which arise in the prosecution process. It is unavoidable that DPP prosecutors will have considerable scope for the exercise of discretion at various stages of the prosecution process, and accordingly it is vital that they have clear guidance in making those decisions. However, those guidelines also serve the important purpose of informing the public generally of the principles on which the statutory functions of the DPP are exercised. Unfortunately, the emergence of the Office of DPP throughout Australia as the principal prosecuting authority was not accompanied by any real change in the basic structure of prosecution arrangements in this country. With the exception of the ACT and the Commonwealth, and increasingly New South Wales, we still do not have a system of public prosecutions in this country. By that I mean arrangements for the prosecution of offences which involve a clear separation of the investigation and prosecution functions, with professional prosecutors who are independent of the police having complete control over the conduct of prosecutions, including possibly the initial decision to institute a prosecution. It continues to be the case that the majority of prosecutions in this country are instituted without there having been any prior reference to a DPP, and more importantly many of those will be resolved without any DPP involvement. In the same decade that most Australian jurisdictions established the Office of DPP the prosecution arrangements for England and Wales underwent a fundamental re-organisation with the establishment of the Crown Prosecution Service.3 The main objective in establishing the CPS in 1985 was to remove police control and indeed influence over the conduct of prosecutions. Although the police would continue to have responsibility for the initial decision whether to charge, once charges were laid the future course of the prosecution, including the decision whether to continue with the prosecution and, if so, how it should be conducted, would be matters for the CPS to decide. In contrast, for most of the Australian jurisdictions the decision to establish the Office of the DPP was not motivated by any desire to remove the police from the conduct of prosecutions. The new Office of the DPP was not seen as providing a professional prosecuting service which had primary responsibility for the prosecution process as a whole. Rather the main responsibilities of the new Office would be confined in the main to the prosecution of offences on indictment. In most jurisdictions the main motivating factors in establishing the Office of DPP were, first, a desire to remove the decision-making process from the political arena, with responsibility for day-to-day control of prosecutions on indictment passing from the Attorney-General to the Director and, secondly, to revitalise and generally update the arrangements for the prosecution of offences on indictment. Indeed, in some instances the enabling legislation placed some restrictions on the authority of the DPP to become involved in the conduct of prosecutions before the summary courts. In any event, even if a DPP was minded to become involved in prosecutions before the summary courts, often the Office simply did not have had the funds to do so. There are compelling reasons of both efficiency and public policy why the police should not be involved in the conduct of prosecutions save possibly for very minor summary offences. If the police must retain the right to institute a prosecution, once proceedings have commenced the further conduct of the matter should be the responsibility of an independent prosecuting service. Decisions relating to the conduct of a prosecution must be made dispassionately, and it may be asking too much of human nature for a police prosecutor, who is part of the same organisation that investigated the offence, always to achieve the necessary degree of detachment no matter how honest and conscientious he or she may be. Further, police prosecutors are unlikely to be legally qualified, and therefore will be ill-equipped to deal with the complex issues of fact and law that can arise before the summary courts with their ever increasing jurisdiction. Any claim that the continued use of police prosecutors in the summary courts can be justified on the grounds of financial expediency is, in my view, illusionary. From my own experience as Commonwealth DPP I have no doubt that the involvement of an independent prosecuting authority from the outset makes it far more likely that the unpromising cases will be weeded out at an early stage. It is far easier to explain to the police why a prosecution should not be instituted than it is to justify a decision to discontinue one that they commenced, particularly if the police can point to the fact that a magistrate considered there was sufficient evidence to warrant making a committal order. The involvement of the DPP at the earliest possible stage will lead to not only an overall saving in the resources that must be devoted to the prosecution of offences but it will lessen the burden on already over stretched court resources.4 In the last two or three decades there have been numerous calls in Australia for an end to police involvement in the conduct of prosecutions in the summary courts.5 Until recently, however, most Australian jurisdictions appeared to be content for their prosecution arrangements to involve a "mixed" system, with the responsibilities of the DPP confined in the main to the conduct of prosecutions on indictment and related proceedings. One State where significant moves have been made towards establishing a public prosecution system is New South Wales. Not long after the office of DPP was established in that State it took over the decision making process where applications were made to the police to terminate the prosecution of indictable offences before a committal order had been obtained.6 This was followed by the New South Wales DPP taking over the conduct of all committal proceedings which had previously been the responsibility of police prosecutors. While police prosecutors in NSW still conduct summary prosecutions, earlier this year the Wood Royal Commission stated that it intends in its final report to make recommendations on whether the conduct of summary prosecutions instituted by the police should be assigned to the New South Wales DPP,7 and last month the Attorney-General for New South Wales announced that this year the New South Wales DPP will conduct a six month pilot scheme in which certain summary offences will be prosecuted by solicitors employed by the DPP.8 As the remaining Australian jurisdictions move, however slowly, towards establishing a public prosecution system there may be some who will argue that the assumption by the DPP of the conduct of all prosecutions will not go far enough in distancing the police from the prosecution function; that the police will still be able to commence a prosecution and they will still retain the important discretions to decide to take no action, to caution etc. I have no difficulty with the police retaining the right at law to institute a prosecution. In arrest matters, for example, it would be neither practicable nor appropriate to require that the decision to charge must be made by a prosecutor. The decision to arrest is properly a function of the police - not the prosecutor. Although arrest and charge are distinct stages in law, in practice they are part of the same process. For the prosecutor to be involved in the decision whether to charge following arrest would entail an unacceptable risk of the prosecutor becoming embroiled in the investigation. In any event, unless the prosecutor had been involved in the investigation it is unlikely that in arrest matters the prosecutor would have the necessary familiarity with the available evidence to be able to decide whether to charge, or what charges to lay.9 What is important is that there be procedures in place which will enable the prosecutor to screen the charges laid by the police at the earliest possible stage.10 Of course, not all prosecutions are instituted by way of arrest and charge, and indeed the majority of them are not. In summons matters it may be practicable for there to be prior consultation with the DPP.11 This is in fact what happens in summons matters prosecuted by my Office, with the result that in such matters it is my Office which in practice makes the decision whether to institute a prosecution and, if so, on what charges.12 I have already referred to the differing perceptions of what is involved in an "independent" prosecution service. In the United Kingdom this is seen as independence from the police in the conduct of prosecution. In Australia, on the other hand, the concept of an "independent" Director of Public Prosecutions is seen as removing the prosecution process from the political arena, that is, independence from government. While I have perhaps already intruded into the matters to be discussed in the next session of the conference, may I make some brief observations on this concept of an "independent" Director of Public Prosecutions. One of the more important issues that must be addressed in establishing a statutory office of the Director of Public Prosecutions is how to define the relationship between the DPP and his or her Attorney-General. This essentially involves reconciling the potentially conflicting ideals of independence and impartiality in prosecutorial decision making with the principle of ministerial accountability. On the one hand, there is the concern to ensure that prosecution decisions are made in a fair and even-handed manner, without even the perception of bias or political interference. However, if the responsibility for making prosecution decisions is to be placed in the hands of a public official, then our system of government requires that that person should be ultimately accountable to the Attorney-General, who in turn is answerable in the Parliament for the official's actions. In the case of a number of jurisdictions it would appear that those responsible for the preparation of the relevant legislation took the view that the ideals of independence and accountability could not both be satisfied, or at least could not be satisfied completely. For example an Attorney-General may be empowered to issue directions or guidelines relating to the circumstances in which the DPP should institute or carry on prosecutions but be precluded from furnishing directions or guidelines in relation to any particular case.13 Further, in the case of the Victorian legislation it would appear that the Victorian Attorney-General is even precluded from giving general directions or guidelines.14 This may be contrasted with the position under the Commonwealth Director of Public Prosecutions Act. In preparing that legislation it was considered that adherence to the ideal of independence in prosecutorial decision making could not be taken so far as to place ultimate authority for the conduct of prosecutions in the hands of the Director of Public Prosecutions. It was considered that the dictates of ministerial accountability require that, while the DPP would be in charge of the machinery of prosecutions, he or she must in law be ultimately accountable to the Attorney-General. Accordingly, under the Commonwealth Act ultimate responsibility for the prosecution function remains with the Attorney-General. Not only does the Commonwealth Attorney-General retain all of his or her powers with respect to the institution and conduct of prosecutions but, more significantly, the Attorney-General is empowered by s8 of the Commonwealth Act to give directions to the Director, and those directions may relate to the conduct of particular cases.15 The relationship between the Attorney-General and his or her DPP reflected in the Commonwealth legislation is, in my view, the appropriate one. The principle of ministerial accountability requires in my view that any legislation establishing the statutory office of DPP must recognise the ultimate authority of the Attorney-General in the prosecution process. On the other hand, although the Commonwealth legislation recognises the final authority of the Attorney-General, clearly that does not involve a concomitant obligation on the Attorney-General to become involved in the handling of particular cases as a matter of course. The authority to give directions to a DPP, even in relation to particular cases, is a necessary constitutional safeguard. However, unless and until the occasion arises for that power to be invoked, the Attorney-General should refrain from becoming involved in day-to-day decision making. To do otherwise would render the establishment of the office of DPP a quite meaningless exercise. Under the Commonwealth Act the expectation that the DPP will be independent in practice in the making of prosecution decisions is reinforced by the conditions which attach to the issue of a direction. Not only must a direction be preceded by consultation with the Director and be in writing, but more significantly the direction must be published in the gazette and tabled in each House of the Parliament within 15 sitting days.16 The conditions which attach to the issue of a direction in practice ensure that the power to give directions will only be resorted to where the Attorney-General is convinced that it is his or her public duty to intervene. May I turn now to a discussion of how prosecutorial discretion is exercised in Australia today. Those responsible for the conduct of a prosecution may be called on to exercise a wide variety of discretions during the course of the prosecution. I propose, however, to confine my remaining comments to three of the more important discretions - the initial decision whether or not to institute a prosecution, the discretion to discontinue a prosecution and the discretion to "indemnify" a witness.

The initial decision to prosecute

Clearly the most important discretion exercised by a prosecutor is the initial decision whether or not to institute a prosecution. As the Commonwealth's Prosecution Policy states: "In every case great care must be taken in the interest of the victim, the suspected offender and the community at large to ensure that the right decision is made. A wrong decision to prosecute or, conversely, a wrong decision not to prosecute, both tend to undermine the confidence of the community in the criminal justice system."17 It is now generally accepted that there are two fundamental considerations involved in any decision whether to prosecute. First, is the evidence sufficient to justify a prosecution. Secondly, is a prosecution required in the public interest. It is the first of these considerations that has occasioned prosecuting authorities the most difficulty in articulating a test which is principled but at the same time workable. How much evidence must there be before a prosecution will be justified; is a prima facie case sufficient, or is more required and if so, what? The difficulty in devising a workable but principled test is perhaps indicated by the fact that on each occasion the Commonwealth's guidelines have been revised some changes have been made to the test of evidential sufficiency. While a bare "prima facie" test was fashionable until well into the 1980s, all DPPs in Australia now agree that, by itself, a prima facie case is an inadequate standard of the sufficiency of evidence. Rather the standard, no matter how it is expressed, must have regard to the prospect of securing a conviction. In the first version of the Commonwealth's Prosecution Policy, which was issued in 1982, the standard of sufficiency of evidence was, first, that there should be a prima facie case and, secondly, that "a prosecution should not normally proceed unless there is a reasonable prospect of conviction". 18 This was equated with it being "rather more likely than not that the prosecution will result in a conviction".19 Thus, although the test of sufficiency of evidence was a reasonable prospect of securing a conviction, it was apparently considered that that of itself was an insufficiently precise standard by which to determine sufficiency of evidence. It was thought to require elaboration, and understandably it was that elaboration - that a conviction be more likely than not - that became the test that was applied in practice. The "more likely than not" test (or the "51% rule" as it soon became known) was criticised, however, as setting too high a standard. In deciding whether a conviction was more likely than not, the prosecutor is no longer concerned with whether the arbiter of fact could convict on the evidence to be adduced by the prosecution, but whether, at the end of the day, the arbiter of fact is likely to convict. The prosecutor must not only make an assessment of the strength of the prosecution case on paper but also make some prediction as to the likely impact that the prosecution witnesses will have on a jury. The prosecutor must also take into account the impact of any likely defence. In some cases the prosecutor, no matter how experienced he or she is and no matter how dispassionately he or she approaches this task, will simply be unable to say whether a conviction or an acquittal is the more likely result. Should a prosecution proceed in such a case? The 1982 guidelines were equivocal. In stating that a prosecution should not "normally" proceed unless there was a reasonable prospect of securing a conviction the 1982 guidelines appeared to acknowledge that in some (unspecified) circumstances a prosecution would be justified although, strictly speaking, the "51% rule" was not satisfied. It was this deficiency that the 1986 version of the Commonwealth's guidelines sought to address. In drafting the 1986 version of the guidelines it was considered that public interest factors may be crucial in deciding whether a prosecution will be justified in those cases where, on the best judgment one can make, it is impossible to say whether a conviction or an acquittal is the more likely result. It was for this reason that the main test of sufficiency of evidence in the 1986 guidelines was subsumed within the public interest consideration so that whether a conviction was the more likely result became the dominant factor in determining whether the public interest required a prosecution. In the "hard cases" it might still be appropriate to proceed if there were public interest factors in favour of a prosecution provided a conviction was reasonably open on the evidence. Nevertheless, this approach did not prove to be satisfactory. The incorporation of the sufficiency of evidence test within the public interest consideration is somewhat artificial, for the latter has traditionally been regarded as separate from considerations relating to the sufficiency of evidence. However, perhaps the main criticism of the 1986 version of the test was that it was simply too difficult to apply in practice. For the 1990 version of the Guidelines it was decided to revert to the "reasonable prospect of conviction" test but this time shorn of any reference to that being equated with a conviction being more likely than not. In doing so we adopted a recommendation made in 1985 by the Victorian Shorter Trials Committee.20 In assessing the strength of the prosecution case the prosecutor will be required to take into account not just the weight that is likely to be placed on the prosecution's evidence, but also any potential for certain evidence to be ruled inadmissible. For example, a record of interview containing admissions may have been obtained in breach of the requirements for the conduct of such interviews. The breach on the part of the investigating police may have been quite blatant, and it may be quite clear that a trial court is unlikely to admit the record of interview into evidence. In such a case the assessment of the sufficiency of evidence should proceed on the basis that the prosecution will not be able to rely on the admissions. On the other hand, if there is an arguable case that the trial court should exercise its discretion in favour of admitting the record of interview into evidence then the appropriate course will usually be to assess "reasonable prospect" on the assumption that the record will not be ruled inadmissible. I should emphasise that the "reasonable prospect" test is an objective one. In assessing the strength of the prosecution case the prosecutor should not take into account any perceived potential for a jury to have regard to what are essentially extraneous factors in reaching its verdict. In assessing "reasonable prospect" the prosecutor is to proceed on the assumption that "the jury will act in an impartial manner in accordance with its instructions."21 This ensures, amongst other things, that what is a weak case does not satisfy the "reasonable prospect" test simply because there are extraneous factors which may motivate a jury towards conviction. This is not to say that any potential for a court or jury to approach a particular defendant or type of case in a particular way must be excluded altogether from the decision whether to prosecute. If, for example, it is considered that a jury is likely to regard the prosecution of a particular defendant as oppressive, and as a result may be motivated towards an acquittal despite the strength of the prosecution case, that may be a very relevant factor in deciding whether a prosecution is warranted in the public interest. While I do not anticipate any changes being made to the Commonwealth DPP's basic test of evidential sufficiency, one area where perhaps some refinement is necessary is in the application of this test at the initial stages of a prosecution. The Commonwealth guidelines suggest that if at any stage the evidence available to the prosecutor does not satisfy the "reasonable prospect" test, a prosecution will not be instituted or continued as the case may be. In some cases, however, it may be quite unrealistic to expect the evidence available to the prosecutor to satisfy that test at the initial stage of the matter. Where the prosecution is instituted by way of arrest and charge often the police will not be in a position to supply the prosecutor with a full brief of evidence until some later stage in the prosecution, and particularly at the stage of the defendant's first appearance in court the material provided by the police to the prosecutor may consist of little more than inadmissible observations. Statements may still have to be obtained by the police from certain witnesses, and indeed the police may still have to complete certain lines of inquiry. It would be absurd to require the prosecutor to discontinue the prosecution if at this initial stage the material available to the prosecutor in the form of admissible evidence is insufficient to satisfy the "reasonable prospect" test. What would seem to be required is an acknowledgment that in some cases it may not be practicable for the police to provide the prosecutor at the initial stage of the prosecution with sufficient evidence which satisfies that test. Until the prosecutor is in a position to make an assessment based on the evidence in admissible form collected during the course of the investigation the prosecutor must be satisfied that there exists evidence to justify the continued remand of the defendant, and that in due course there is likely to be sufficient evidence to satisfy that test. The second consideration in the decision whether to prosecute is whether a prosecution is warranted in the public interest. However, this consideration does not arise unless the prosecutor is first satisfied that there is a reasonable prospect of securing a conviction on the available evidence. No matter how compelling the public interest factors in favour of proceeding with a prosecution, they are irrelevant unless there is sufficient evidence to justify a prosecution. This is the dominant consideration in the decision whether to prosecute in the sense that, although there is sufficient evidence to support a prosecution, nevertheless a prosecution should not proceed unless that is warranted in the public interest. On the other hand, it is important to keep the public interest consideration in perspective. First, it should not be seen as consisting only of those factors which may tend against proceeding with a prosecution although there is sufficient evidence available. There are public interest factors which favour proceeding with a prosecution. The fact that a weapon was used, the alleged offender was in a position of authority or trust, and that the alleged offender has prior convictions of a relevant nature, are obvious examples.22 Secondly, as a matter of practical reality in many instances the appropriate decision will be to proceed with a prosecution notwithstanding that in the circumstances of the particular case there are considerations militating against a prosecution. In this regard, "generally speaking the more serious the offence the less likely it will be that the public interest will not require that a prosecution be pursued".23 However, unless the offence is so serious as to require a prosecution irrespective of any public interest factors militating against that course being adopted, the prosecutor "should always apply his or her mind to whether the public interest requires a prosecution".24 The Commonwealth's guidelines list some 20 factors which in the circumstances of a particular case may arise for consideration in determining whether a prosecution is required in the public interest. May I deal briefly with just one of those factors - the likely length and expense of a trial25 - as I believe this is likely to become increasingly important in determining whether the public interest requires a prosecution. The plain fact of the matter is that, as our criminal justice system has become more just, it has also become exceedingly expensive and rather inefficient. As one commentator has put it: "For too long those involved in the [criminal justice] system have been wedded to the idea that justice is such a sacrosanct commodity that it has virtually no price. One result of this attitude has been, until recently, little demonstration of any serious concern about the economics of the criminal trial system."26 While no doubt many factors have contributed to the expense and general inefficiency of our present criminal justice system, in my view the main culprits are the outdated procedural and, to a lesser extent, evidentiary rules that govern the conduct of criminal proceedings in this country. Unless new rules are devised which will make our criminal justice system more efficient, but no less just, I fear there will soon come a time when we will be unable to prosecute certain cases that should be prosecuted for the only reason that it would be simply too costly to do so.27

Discontinuing a prosecution

Once a prosecution has commenced the matter must always be kept under review, and the strength of the prosecution case reassessed where appropriate. If the DPP did not make the initial decision to institute the prosecution then of course the first review should occur as soon as practicable after the matter has been referred to the DPP. However, even where the DPP made the initial decision to prosecute, new evidence or information can become available which makes it no longer appropriate for the prosecution to proceed. In so far as Commonwealth matters are concerned, so long as the prosecution is before a court of summary jurisdiction, whether the proceedings are for summary conviction or committal for trial, it is the established practice that neither the Director nor a member of the staff at the Office is the informant in the prosecution; rather the informant is usually an officer from the agency that investigated the offence, with the DPP carrying on the prosecution. Initially the Director of Public Prosecutions Act 1983 (Commonwealth) drew a distinction between the situation where the DPP was carrying on a prosecution instituted by another, and where such a prosecution had been taken over in the exercise of the power under s9(5) of the Act.28 In the case of the latter, the DPP's authority to determine whether the prosecution should proceed is clear. However, it was considered to be otherwise where the Office was carrying on a prosecution instituted by another. Given that the Director is not the informant when carrying on a prosecution it was considered that the role of the informant in the prosecution strictly speaking involved something more than the nominal one of lending his or her name to the title of the proceedings. Thus, while it was for the Office of the DPP to determine, broadly speaking, how a prosecution should proceed (for example, what evidence to call) so long as the Office was carrying on a prosecution instituted by another it could not be discontinued without the concurrence of the informant. In the event the informant did not agree, it was necessary for the Director to first exercise the power under s9(5) to take over the prosecution before the Director could then discontinue it. It is doubtful whether this rather curious state of affairs was intended by those responsible for the preparation of the legislation. They may well have assumed that, where a Commonwealth agency referred a prosecution to the Director, the conferral on the Director of the function of carrying on the prosecution would also confer on the Director the power to discontinue it if the Director saw fit. Only in the case of a "private prosecution" would it be necessary for the Director to first assume the role of the informant before the Director would be in a position to discontinue it. This, of course, assumes there is a distinction between an informant who is a private individual and one who, in instituting a prosecution, is acting in the course of a public office or duty. While there is of course such a distinction in practice, strictly speaking both are private individuals in the eyes of the court. This rather curious state of affairs was corrected in 1991 with an amendment to the Commonwealth Act with the result that where the Director is carrying on a prosecution instituted by another the Director may decline to carry on that prosecution without first taking it over.29 Although the informant in a prosecution before a court of summary jurisdiction now performs only a purely formal role in the prosecution, and it is for the DPP to decide whether the prosecution should be continued, that does not of course mean that the DPP should discontinue the prosecution without first consulting with the agency that investigated the matter. The need for consultation in such circumstances is dealt with thus in the Commonwealth's guidelines: "Whenever the DPP is contemplating discontinuing a prosecution the practice is for the DPP to first consult the AFP or responsible department or agency. In this regard, the independence of the DPP in the prosecution process does not mean that those who investigated the matter should be excluded from the decision-making process. Indeed, where the DPP is contemplating discontinuing a prosecution close liaison is vital to the maintenance of a harmonious relationship between the Office and the relevant Commonwealth agency. Of course, the extent of that consultation will depend on the circumstances of the case in question, and in particular on the reasons why the DPP is contemplating discontinuing the prosecution. If it is considered the available evidence is insufficient, it can be expected the AFP or responsible department or agency will accept the DPP's assessment of the evidence, and the consultation will be largely confined to the prospects of obtaining additional evidence. On the other hand, the AFP or responsible department or agency can legitimately expect to have its views taken into account if discontinuance on public interest grounds is contemplated. The more finely balanced the factors involved, the greater is the need for discussion."30 In indictable matters there is of course an obvious point where the DPP must assess the initial decision to institute or proceed with a prosecution, and that is after a committal order has been obtained and the decision must be made whether to present an indictment. In this instance the prosecutor may have the advantage of being able to assess the sufficiency of the available evidence on the basis of how the prosecution's witnesses actually stood up under cross examination. It is of course at this stage that applications are not frequently received from the defence not to proceed with a trial on indictment. However, often prosecutions will be discontinued at this stage without there having been any prior representations from the defence. In the last financial year, for example, some 40 per cent of the "no bills" were entered without prior representations from the defence.31 Very occasionally where a "no bill" is entered the Office may receive a request for the reasons for that decision. I have earlier mentioned that one of the standards by which a prosecution system should be judged is the extent to which those who make prosecution decisions are accountable. The need for accountability is particularly acute where a magistrate has determined in open court that there is a case fit to go to a jury yet the DPP has decided not to proceed with a trial on indictment. Silence on the part of the prosecution as to the reasons for such a decision may spawn inaccurate speculation or give rise to suspicion that the prosecution was discontinued for improper reasons, particularly where the matter concerned is one that has excited public interest. There may also be persons apart from the particular government agency that investigated the matter who have a clear interest in being informed of the reasons for the decision. In any event, public confidence in the administration of the criminal justice system can only be enhanced where those responsible for the making of such decisions are seen to do so in a proper manner. It is for those reasons that my Office is quite willing to provide short reasons for such decisions. While it is not our practice to issue reasons in the form of a media release, nevertheless we have done so in some cases where the matter involved has been one of considerable public concern and interest.

Indemnification of Witnesses

The final discretion I wish to deal with is the discretion to indemnify a person in order to secure that person's evidence for the prosecution. Upon receipt of the brief of evidence it may be clear to the DPP that, without the evidence of one or more persons who also participated in the criminal activity which is the subject of the charge against the defendant, the prosecution simply has no case, or at best a very weak one. However, unless the accomplice has been dealt with in respect of his or her own involvement in the offence the accomplice will be in a position to claim the privilege against self-incrimination in respect of the very matter the prosecution wishes to adduce into evidence. Alternatively, it may be quite clear that the accomplice will refuse to co-operate with the prosecution if he or she is also prosecuted. Although an accomplice who has been dealt with in respect of his or her own involvement in the offence will no longer be able to avoid testifying by relying on the privilege against self-incrimination, usually there will be little to be gained, and perhaps much to be lost, by calling an accomplice who is determined not to assist the prosecution. Thus, if the prosecution is to succeed it may be necessary to provide that accomplice with some concession in order to secure his testimony for the prosecution. Of course, the ideal situation is to secure an accomplice's testimony without having to offer any concession, save possibly agreeing to proceed on lesser charges than would otherwise be appropriate. Upon pleading guilty the accomplice who is prepared to co-operate in the prosecution of former colleagues can expect to receive a substantial reduction in the sentence that would otherwise have been appropriate. In conjunction with any concession that may be offered to the accomplice as to the charges to be proceeded with, an accomplice will often have a powerful incentive to agree to co-operate with the prosecution on such a basis, particularly if the accomplice's bargaining position is weak. In any event, the prosecuting authority should always explore the possibility of securing an accomplice's testimony at this far cheaper price, even if it is prepared to make greater concessions if necessary. The indemnification of a witness must always be regarded as a course of last resort. However, this option may not be possible in some cases. As mentioned earlier, the accomplice may make it quite clear that he or she will not co-operate with the prosecution if he or she is also prosecuted. Alternatively, there may be insufficient admissible evidence to support charges against the accomplice.32 Thus, the question may arise whether it is in the overall interests of justice that the opportunity to prosecute the accomplice be forgone in order to secure that person's testimony in the prosecution of another. In determining where the balance lies, the Commonwealth's guidelines state that account should be taken of the following matters: "(a) the degree of involvement of the accomplice in the criminal activity in question compared with that of the defendant; (b) the strength of the prosecution evidence against the defendant without the evidence it is expected the accomplice can give and, if some charge or charges could be established against the defendant without the accomplice's evidence, the extent to which those charges would reflect the defendant's criminality; (c) the extent to which the prosecution's evidence is likely to be strengthened if the accomplice testifies - apart from taking into account such matters as the availability of corroborative evidence, and the weight that the arbiter of fact is likely to give the accomplice's testimony, it will also be necessary to consider the likely effect on the prosecution case if the accomplice does not come up to proof; (d) the likelihood of the weakness in the prosecution case being strengthened other than by relying on the evidence the accomplice can give (for example, the likelihood of further investigations disclosing sufficient independent evidence to remedy the weakness); (e) whether there is or is likely to be sufficient admissible evidence to substantiate charges against the accomplice, and whether it would be in the public interest that the accomplice be prosecuted but for his or her preparedness to testify for the prosecution if given an undertaking under the Act; and (f) whether, if the accomplice were to be prosecuted and then testify, there is a real basis for believing that his or her personal safety would be at risk while serving any term of imprisonment."33 Under s9(6D) - (6F) of the Director of Public Prosecutions Act 1983 (Commonwealth) specific provision is made for the DPP to undertake to a person that he or she will not be prosecuted for a specified offence against Commonwealth law, or in respect of specified acts or omissions that constitute, or may constitute, an offence. Such an undertaking may be subject to conditions and, once given, precludes a prosecution being instituted by any person in respect of the offence or acts or omissions specified. While a number of the other DPPs have been specifically empowered to grant immunity from prosecution34 the absence of an express power will, ordinarily, not preclude a DPP from granting what amounts to immunity from prosecution. Not only is a DPP able to give an assurance to a prospective prosecution witness that no proceedings will be instituted against the witness either by the DPP or on his or her behalf (which merely involves the exercise of a power which is incidental to the general power of the DPP to institute a criminal prosecution) but more importantly the DPP has the necessary statutory powers to intervene in the event the witness is the subject of a private prosecution.35 Curiously, however, that is not the case in so far as the New South Wales DPP is concerned, for s19 of the Director of Public Prosecutions Act 1986 (NSW) provides that, while the DPP may request the Attorney-General to grant "indemnity from prosecution", the DPP may not grant such an indemnity.36 For my part I can see no good reason for such matters to be reserved for the decision of the Attorney-General. The involvement of the Attorney-General in such matters would seem to be quite inconsistent with the concept of an independent Director of Public Prosecutions. The grant of immunity from prosecution for the purpose of securing the evidence of an accomplice essentially involves choosing the lesser of two evils, and there is an understandable reluctance to take the step of effectively absolving an accomplice for past misdeeds as the price for securing that person's evidence for the prosecution. That is perhaps the main reason why it is now more usual to secure the testimony of an accomplice by means of an undertaking which precludes the subsequent use of the witness' evidence against him but does not in terms involve an undertaking that the witness will not be prosecuted - the so-called "use" or "use/derivative use" indemnity. Under the Director of Public Prosecutions 1983 (Commonwealth), for example, the DPP is empowered to give to a person an undertaking the effect of which is that the person's testimony, and any evidence derived either directly or indirectly from that testimony, cannot be used against the person in any subsequent proceedings, save in respect of the falsity of the evidence given. Once such an undertaking is given to a prospective witness that person will not be able to rely on the privilege against self-incrimination.37 At a superficial level a use/derivative use indemnity would seem to be a more attractive option for law enforcement authorities than a conferral of immunity from prosecution. However, at least in so far as the Commonwealth provisions are concerned, the claimed advantages of a use/derivative use indemnity over a grant of immunity from prosecution are more notional than real, and in practice there is little difference between the two. Unless the prosecution is in possession of sufficient independent evidence to support the prosecution prior to the grant of a use/derivative use indemnity such an indemnity in practice may approach a grant of immunity from prosecution. In Kastiger v. US 406 US 441 (1972) at 461 the US Supreme Court referred to the "heavy burden" on the prosecution to establish that the evidence it proposed to rely on was derived from a source which was wholly independent of the compelled testimony. This may for all practical purposes be impossible if that evidence was obtained or provided by persons who were exposed to the compelled testimony. In this regard, it is worth noting that in so far as the Commonwealth is concerned there would appear to be no case where the recipient of a s9(6) undertaking has subsequently been prosecuted in respect of any offence touched on by that person in the course of giving evidence. On the other hand, a use/derivative use indemnity, at least in its Commonwealth form, has one clear disadvantage in that the protection afforded by such an indemnity cannot be confined to evidence dealing only with specified criminal activity. Thus, there will always be the risk that, for example, cross examination of the witness could illicit a disclosure implicating the witness in some other criminal activity of which the prosecution was not aware. Any subsequent prosecution of the witness for such an offence will of course depend on whether the prosecution can establish that the evidence it proposes to rely on was derived from a source which is wholly independent of the protected testimony. On the other hand, a grant of immunity from prosecution under the Commonwealth Act enables the DPP to confine the protection afforded by the Commonwealth provision to specified criminal conduct.38 Thus, while a use/derivative use indemnity has the theoretical advantage that the prosecution is not precluded from prosecuting the accomplice in respect of his or her own involvement, this advantage is effectively cancelled out by the considerable difficulties that are likely to confront the authorities in practice in obtaining sufficient evidence against the accomplice which is wholly independent of the protected testimony. Accordingly, the practical effect of granting a use/derivative use indemnity will be to confer something approaching immunity from prosecution in respect of any offence disclosed by the accomplice in the course of giving evidence if, prior to the accomplice testifying, the prosecution is not in possession of sufficient evidence to support a prosecution in respect of that offence. Despite all this, prosecuting authorities now have a distinct preference for the use/derivative use indemnity as the means of securing the evidence of an accomplice. It would seem that we are more comfortable with providing an accomplice with an undertaking that does not in terms confer actual immunity from prosecution. Although it would be quite unrealistic to suggest that use/derivative use indemnities are relied on with the objective of preserving the option of subsequently prosecuting the accomplice, and indeed the expectation is that the accomplice will not be prosecuted, nevertheless a use/derivative use indemnity enables the prosecution to avoid taking that final and irrevocable step of conferring on the accomplice a complete guarantee that he or she will not be prosecuted. I have no difficulty with this provided the court, and particularly the jury, is not mislead as to the basis upon which the accomplice is testifying for the prosecution. If there have been prior negotiations between the accomplice and the DPP, as a result of which it is agreed that the accomplice will not be prosecuted in respect of his or her own involvement then, irrespective of the exact terms of the undertaking given to the accomplice, it should be made quite clear to the court and the jury that the accomplice is testifying for the prosecution on the basis that he or she will not be prosecuted.

Footnotes 1 The Royal Commission on Criminal Procedure, Report (1981) Cmnd 8092, p127-8. 2 Australian Law Reform Commission, "Sentencing of Federal Offenders" Report No. 15, p61 3 Prior to the establishment of the CPS the institutional arrangements for the prosecution of offences in England and Wales in practice were very similar to those which still apply in the majority of Australian jurisdictions, with those who investigated the offence having the responsibility for any ensuing prosecution. Although the pre-CPS Director of Public Prosecutions was empowered to assume control of the more important or difficult prosecutions, that right of intervention was exercised only rarely, and prior to the establishment of the CPS the vast bulk of prosecutions instituted by the police, including most prosecutions on indictment, remained in their hands. In contrast, the police in Australia have never had a role in the conduct of prosecutions on indictment. Upon the foundation of the colony of New South Wales it was considered that the common law system of initiating criminal proceedings by the preferring of a bill of indictment by a grand jury and then trial by petty jury was unsuitable for a penal colony. Accordingly, by the Australian Courts Act 1828 it was provided that, until provision was made for presentment by a grand jury and trial by a petty jury, indictment for all offences might be presented by the Attorney-General or a person appointed by the Governor (the latter eventually evolving into the office of Crown Prosecutor). Although it was intended that this would only be an interim measure pending the establishment of grand juries, in fact the grand jury system was only ever introduced in Victoria. Even when the system of preliminary examination as we now know it was adopted in the middle of last century, the Attorney-General and Crown Prosecutors retained their sole authority to present an indictment. 4 "A minimum amount of attention paid to increased efficiency at the earlier stages of the ... process raises the important prospect that much time will be saved in its later stages." AG's Advisory Committee, "Charge Screening, Disclosure and Resolution Discussions" (Ontario, 1993), referred to in Mack and Roach Anleu "Pleading Guilty : Issues and Practices" (AIJA, Melbourne 1995) at p77. 5 See, for example, Report of the Commission to Inquire into NSW Police Administration (the "Lusher Report"), 1981, p238-258; Report of the Royal Commission of Inquiry into Certain Committal Proceedings against K E Humphries (the "Street Report"), 1983, Appendix 14, p99; Second Report of the Criminal Law and Penal Methods Reform Committee of South Australia : "Criminal Investigation", 1974, p168-176. 6 Annual Report of the Director of Public Prosecutions (NSW) (1987-88), p10. 7 The Royal Commission into the New South Wales Police Service, Interim Report, February 1996, paragraph 6.46. 8 Sydney Morning Herald, Tuesday, March 12, 1996, p7. The New South Wales Attorney-General is reported as stating that "I would be reasonably confident that the results [of the pilot scheme] will be positive and that in the longer term we will be taking the role of police prosecutor away from the police service and giving it to the Director of Public Prosecutions." 9 Further, so long as the right of any person to commence a prosecution is retained it would be curious indeed if those whose public duty it is to enforce the criminal law were to be denied that right. 10 It would necessarily follow from the retention by the police of the right to institute a prosecution that they may decide to take no action, or to caution etc an offender. I have no difficulty with the police making such decisions without reference to the DPP provided such decisions are made in accordance with the relevant prosecution policy guidelines and the police are prepared to consult with the DPP in cases of doubt. 11 Although the prosecution function of a DPP may be confined in the main to the prosecution of offences on indictment and the conduct of related proceedings it is now more common for at least specialist police units to regularly consult with the DPP before laying charges. 12 Prior to the establishment of the Commonwealth DPP, officers within the Commonwealth Attorney-General's Department conducted on behalf of the police all prosecutions before the summary courts which had been instituted by the police. It was the established practice in other than arrest matters for the police, and indeed other agencies which investigated offences against Commonwealth law, to forward a brief of evidence to the Department for "advice" whether a prosecution should be instituted and, if so, on what charge. 13 See, for example, s26, Director of Public Prosecutions Act 1986 (NSW); s27, Director of Public Prosecutions Act 1991 (WA). 14 See Edwards, J.LI. "Walking the Tightrope of Justice : An examination of the Office of Attorney-General in Canada with Particular Regard to its Relationships with the Police and Prosecutors and the Arguments for Establishing a Statutorily independent Director of Public Prosecutions", a series of opinion papers prepared for the Royal Commission on the Donald Marshall, Jr., Prosecution (1989) Nova Scotia, Canada, p58 15 S8, Director of Public Prosecutions Act 1983 (Commonwealth). 16 S8(5) of the Director of Public Prosecutions Act 1983 provides, however, that publication may be delayed if publication proximate to the issue of the direction would prejudice pending criminal proceedings. 17 Prosecution Policy of the Commonwealth, Commonwealth Director of Public Prosecutions, AGPS, Canberra (1992), paragraph 2.2. 18 Prosecution Policy of the Commonwealth, Attorney-General's Department, AGPS, Canberra (1982), paragraph 16. 19 ibid. 20 That Committee considered that "reasonable prospect of conviction" is "not to be equated with a 51% chance of conviction being sustained. It may be something less than that and is not appropriately expressed in mathematical terms" (Shorter Trials Committee, "Report on Criminal Trials", AIJA, 1985, paragraph 3.53). Another factor in our decision to adopt a bare "reasonable prospect of conviction" test was that that was the test that was applied by both the Victorian and New South Wales DPPs, and at a meeting of DPPs and Senior Crown Prosecutors in 1989 there was general agreement both as to the desirability of each jurisdiction operating under the same test, and that that test should be the one recommended by the Shorter Trials Committee. It would appear that the CPS has also experienced similar difficulty in arriving at a workable but principled test. In 1983 the UK Attorney-General issued guidelines which stated that the standard of sufficiency of evidence was "whether there is a reasonable prospect of conviction; or put another way, whether conviction is more likely than an acquittal before an impartial jury properly directed in accordance with law." S10 of the Prosecution of Offences Act 1985, which established the Crown Prosecution Service, requires the DPP to issue a Code for Crown Prosecutors dealing with, amongst other things, the criteria governing the decision to prosecute. The first Code issued pursuant to s10 replaced the UK Attorney-General's criteria with a "realistic prospect of conviction" test. However, in the first major revision of the CPS Code (released in June 1994) it would seem that the evidential test under which the CPS now operates has for all intents and purposes gone full circle, for paragraph 5.2 of the revised Code now elaborates on what is meant by "realistic prospect of conviction": "5.2 A realistic prospect of conviction is an objective test. It means that a jury or bench of magistrates, properly directed in accordance with the law, is more likely than not to convict the defendant of the charge alleged." (emphasis added) This paragraph also restored the "objective" nature of the UK Attorney-General's 1983 guidelines. It was arguable that the bare "realistic prospect" test in the initial version of the Code would have permitted the prosecutor in assessing the sufficiency of the evidence to take into account any perceived potential for the arbiter of fact to have regard to extraneous factors in reaching its decision. 21 Prosecution Policy of the Commonwealth (1992), paragraph 2.5. 22 This is now recognised in the 1994 version of the CPS Code (see CPS Code (June 1994), paragraph 6.4). There was little recognition in earlier versions of that Code that there are public interest factors in favour of proceeding with a prosecution which must be balanced against those militating against a prosecution in determining whether a prosecution is warranted in the public interest. 23 Prosecution Policy of the Commonwealth (1992), paragraph 2.9. 24 Ibid, paragraph 2.11. 25 Prosecution Policy of the Commonwealth (1992), paragraph 2.10(p). 26 Shorter Trials Committee, "Report on Criminal Trials", AIJA, 1985, paragraph 1.53. 27 The problem of costs is likely to be exacerbated if the recommendations made by the ALRC in its report No. 75, "Cost Shifting - Who Pays for Litigation?" concerning the award of costs in indictable matters are adopted. On the other hand, it is difficult to argue against the proposition that in principle a successful defendant in a prosecution on indictment should be entitled to recover his or her legal costs as of right. 28 "For the purposes of the performance of his functions, the Director may take over a proceeding that was instituted or is being carried on by another person, being a proceeding: (a) for the commitment of a person for trial in respect of an indictable offence against a law of the Commonwealth; or (b) for the summary conviction of a person in respect of an offence against a law of the Commonwealth; and where the Director takes over such a proceeding, he may decline to carry it on further." 29 S9(5A), inserted by Act No. 28 of 1991, provides: "Where the Director is carrying on a proceeding instituted by another person, being a proceeding of the kind mentioned in paragraph (5)(a) or (b), the Director may decline to carry it on further even if the Director has not taken it over under s(5)." 30 Prosecution Policy of the Commonwealth (1992), paragraph 4.6. 31 Commonwealth Director of Public Prosecutions, Annual Report 1994/95, page 7. In that financial year there were 37 no bill applications received from the defence of which 16 were granted and 21 refused. A further 23 "no bills" were entered on the basis of a recommendation from a regional office without prior representations from the defence. Of the total of 39 "no bills" which were entered, it was considered that in 28 cases the available evidence was insufficient to support a prosecution. 32 On the other hand, sometimes this will be because the police never viewed the accomplice as a potential defendant; rather they concentrated on establishing a case against the other offenders around the evidence it was expected the accomplice could give. By the time the matter is referred for prosecution the DPP may be presented with what amounts to a fait accompli - either the remaining offenders can be prosecuted by relying on the evidence the accomplice can give, or there will be insufficient admissible evidence to support the prosecution of anyone, or at least the major players. It is for such reasons that the Commonwealth guidelines counsel the police to consult with the DPP if practicable before adopting such a course. That will minimise the potential for an otherwise meritorious prosecution being abandoned because it is eventually decided that it would not be in the interests of justice to indemnify the accomplice (see Prosecution Policy of the Commonwealth (1992), paragraph 5.8). 33 Prosecution Policy of the Commonwealth (1992), para 5.6. 34 For example, the ACT, South Australia and Western Australia. S9(7) - (9) of the Director of Public Prosecutions Act 1990 (ACT) are in very similar terms to the provisions in the Commonwealth Act. In contrast, in the Spartan style now favoured by drafters of South Australian legislation paragraph 7(1)(f) of the Director of Public Prosecutions Act 1991 (SA) merely confers on the DPP the power "to grant immunity from prosecution in appropriate cases". 35 In this regard, in the two English cases of Turner v. DPP (1979) 68 Cr App R 70 and Raymond v. Attorney-General (1982) QB 839 it was held that no objection could be sustained to the exercise by the DPP of his powers to intervene in a private prosecution to bring such a prosecution to an end if it had been instituted to either punish an accomplice for having testified for the prosecution or to prevent the accomplice from doing so. 36 The New South Wales provision also precludes the New South Wales Director from giving a "use" indemnity to a witness. 37 See s9(6), Director of Public Prosecutions Act 1983. Although this subsection does not in terms provide that an undertaking under the subsection overrides the recipient's right to claim the privilege against self-incrimination, there can be no doubt that it has that effect. Where the recipient of an undertaking testifies as to his or her own involvement in the criminal activity the subject of the charge against the defendant the witness may still of course be prosecuted in respect of his or her own involvement. However, by reason of the prohibition on the use not only of the witness' testimony but also any information derived from that testimony the witness will not be exposed to any greater risk of being prosecuted by being required to answer. In BTR Engineering (Australia) v. Patterson (1990) 20 NSWLR 724 Giles J stated, at 730, that "the purpose of the privilege is to protect the witness against the jeopardy of criminal charges, and the purpose will not be fulfilled where the answer to the question will not add to the jeopardy in which he already stands." In its original form s9(6) only protected the witness' actual testimony ie it did not also protect any evidence derived either directly or indirectly from the protected testimony. Accordingly, as the protection afforded by an undertaking under s9(6) in its original form was not co-extensive with the privilege against self-incrimination, my Office took the view that it did not override an otherwise allowable claim of privilege. See Sorby v. The Commonwealth (1983) 57 ALJR 248, per Gibbs CJ and Murphy J; Registrar of the Court of Appeal v. Craven (1994) 126 ALR 668, per Meagher JA at 695. 38 In this regard, s9(6D) may be contrasted with provisions such as s87-89 of the Evidence Act (Tas). Where a certificate is granted to a witness under s89 of that Act the witness may not be prosecuted in respect of any offence touched on by the witness in the course of giving evidence. It is provisions such as s87-89 of the Tasmanian Act which are the true "immunity bath". Prior to amendments in 1990, s11-13 of the Evidence Act (WA) were to the same effect. However, as a result of those amendments a certificate issued by the court under the WA provisions will only confer immunity from prosecution where the witness was compelled to give evidence in revenue cases. In other cases the certificate will only provide the witness with a use indemnity.