England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
HEARING THE MATTER OF A JUDICIAL REVIEW

		Royal Courts of Justice Strand, London, WC2A 2LL
		15/03/2021

Covid-19 Protocol: This judgment was handed down remotely by circulation to the parties' representatives by email, release to BAILII and publication on the Courts and Tribunals Judiciary website. The date and time for hand-down is deemed to be 10:30am on 15 March 2021.

LORD BURNETT OF MALDON CJ:
 


Introduction

1. This is the judgment of the court on an application for judicial review being heard at first instance in the Court of Appeal. The claimant is a campaigning organisation with deep concerns about the low and declining rate of prosecutions for rape and serious sexual offences when compared with the number of alleged offences reported to the police. There is a widespread and legitimate interest in the underlying question which has animated these proceedings, shared by successive Directors of Public Prosecutions, the wider legal community, the judiciary and many more. A Government review led by the Ministry of Justice is underway looking at this issue with wide-ranging participation.
2. The President of the Queen's Bench Division identified the fundamental complaint in these proceedings when giving the judgment of the Divisional Court refusing permission to bring judicial review proceedings, [2020] EWHC 929 (Admin):

"The fundamental complaint made by the claimant is that the defendant has changed the policy of the Crown Prosecution Service ("the CPS") in relation to the prosecution of rape and other sexual offences since about late 2016. The claimant submits that there has been a change of policy from the merits-based approach to the bookmaker's approach, which is unlawful. In the alternative, the claimant contends that the defendant has taken steps which amount to a change in practice. Even if not in policy. The defendant denies that there has been any change of policy or practice." (para 2)

3. The terms 'merits-based approach' ("MBA") and "bookmaker's approach" are found in the judgment of Toulson LJ, with whom Forbes J agreed, in R (FB) v. DPP [2009] EWHC 106 (Admin); [2009] 1 WLR 2072 a judgment we will discuss in more detail. They contrast two different ways in which a prosecutor might determine whether a prosecution is more likely than not to succeed. In short, the first is an objective approach, the second a predictive approach based upon experience of juries. Importantly, the second would consider perceived prejudices of juries arising from so-called rape myths as relevant to whether to prosecute. It is no longer submitted that there has been a substantive change to the bookmaker's approach, but rather that the change in policy created a risk that prosecutors would understand that they should apply the bookmaker's approach.
4. The claimant sought permission to appeal the order of the Divisional Court. The application was adjourned to an oral hearing and came before the Court on 30 July 2020. When considering such an application the Court of Appeal has wide powers, including the power to grant permission to apply for judicial review (rather than permission to appeal); and, if it grants permission, it may retain the substantive claim for judicial review and determine that claim itself rather than remit the case to the High Court: see CPR 52.8(1), (5) and (6). The court granted permission to apply for judicial review and retained the claim in the Court of Appeal. In the result the suggested change of policy or practice is now challenged on five grounds:

a) it was irrational.

b) It led to a risk of systemic illegality because it created confusion in the mind of prosecutors about the test they should apply in making decisions whether or not to prosecute.

c) It was unlawful because the DPP should have consulted 'stakeholders' before making the decision, as, by reason of the DPP's previous practice of consultation, they had a legitimate expectation that they would be consulted.

d) The decision was made in breach of section 149 the Equality Act 2010 ('the 2010 Act').

e) The decision was a breach of the DPP's duty of transparency to publish her policies.

5. The parties agree that prosecutors must apply the Code for Crown Prosecutors in deciding whether to prosecute. It has a two-part test for deciding whether a criminal prosecution should be brought. That test applies to every offence. This case is about the first part of the test which deals with evidence, or "the evidential limb", of that test rather than the second, which is concerned with the public interest. For convenience, we will refer to this as "the full Code test". There is common ground on one important point. The parties agree that the MBA and the full Code test "are the same thing".
6. The claimant's challenge is to "the abrupt change in policy or practice by the defendant, pursuant to which the CPS ceased to apply the 'merits-based approach' to prosecuting cases of sexual violence, in particular cases of rape and the introduction of a new policy or practice to that effect". Mr Little QC accepted that when she was DPP, Dame Alison Saunders DCB decided in the summer of 2016 that prosecutors should receive further training about the test to be applied when an allegation of rape or a serious sexual offence had been made. These are known as "RASSO" cases. That training was given in the course of "RASSO roadshows". She also decided that the term "merits-based approach" should be excised from training materials and guidance, and that a separate guidance document entitled "Code for Crown Prosecutors Test – Merits Based Approach" ("the MBA Guidance") should be withdrawn.
7. That "new policy or practice" (if such it is) has now been superseded by further interim guidance, which is the subject of a current consultation. The claimant is content with the terms of this interim guidance, which, in part, at least, restore some of the text of the MBA Guidance although without reference to the MBA itself.
8. The parties to the application therefore agree that the DPP made a relevant decision (or decisions). We will use the word "decision". They do not agree about the effect and implications of that decision, or about whether it was lawful. The central dispute is whether the DPP changed the policy or practice in the way which the claimant alleges. The premise of many of the grounds of challenge is that there was such a change; and of the defence to those grounds, that there was not. This dispute, which turns on a comparison of the meaning and legal effect of the old guidance and the new guidance respectively, lies at the heart of this claim. A central plank of the DPP's case is that the full Code test has always applied and that this has never changed.

The functions of the Director of Public Prosecutions

The statutory provisions

9. Section 1 of the Prosecution of Offences Act 1985 ("the 1985 Act") establishes the CPS. The DPP is the head of the CPS (section 1(1)(a)). The CPS also includes Chief Crown Prosecutors ("CCPs") who are responsible to the DPP for supervising the operation of the CPS in each of their areas, and other staff appointed by the DPP (section 1(1)(b) and (c)). The DPP may designate Crown Prosecutors from members of the CPS (section 1(3)). The DPP must divide England and Wales into areas (section 1(4)). Every Crown Prosecutor has all the powers of the DPP "as to the institution and conduct of proceedings but shall exercise those powers under the supervision of" the DPP (section 1(6)).
10. The DPP is appointed by the Attorney General (section 2(1)). He or she must make annual reports to the Attorney General, which the Attorney General lays before Parliament (section 9).
11. Section 10 is headed "Guidelines for Crown Prosecutors". Section 10(1) requires the DPP to issue a Code for Crown Prosecutors:

"giving guidance on general principles to be applied by [prosecutors]

(a) in determining in any case –

(i) whether proceedings for an offence should be instituted, or where proceedings have been instituted, whether they should be discontinued; or

(ii) what charges should be preferred; and

(b) in considering, in any case, representations to be made by them to any magistrates' court about the mode of trial suitable for that case."

12. Section 10(2) gives the DPP power to make alterations to the Code. The provisions of the Code are to be set out in the DPP's report to the Attorney General (section 10(3)), as is any alteration to the Code (section 10(3)).

The merits-based approach

13. The phrase "the merits-based approach" was used in the judgment of Toulson LJ in FB. FB was a challenge to a decision of the DPP not to prosecute a suspect in respect of an allegation of an assault contrary to section 18 of the Offences Against the Person Act 1861. The complainant alleged that he had been attacked by an assailant, whom he named. There was no doubt that the complainant's ear had been bitten off. The CPS decided not to prosecute the suspect because concerns about the complainant's mental state meant that there was no realistic prospect of conviction. The Divisional Court quashed the decision on the grounds that it was irrational.
14. Toulson LJ quoted para 5 of the 2004 edition of the Code, which was headed "The Full Code Test". He rejected the claimant's submission that the evidential test was met if the evidence would survive a submission of no case to answer at the end of the prosecution case. The Code required the prosecutor to look at all the evidence, whereas a judge at half-time only considers the prosecution evidence.
15. At para 49 Toulson LJ noted a topic of discussion at the hearing, namely whether a prosecutor should answer the question whether there was a realistic prospect of conviction by taking a "bookmaker's approach" (as it was referred to in argument) or "should imagine himself to be the fact finder and ask himself whether, on balance, the evidence was sufficient to merit a conviction taking into account what he knew about the defence case". The DPP submitted that the latter approach was correct and the bookmaker's approach wrong. The claimant made no submissions. Toulson LJ agreed with the DPP.
16. In para 50 of the judgment, he gave "date rape" cases as examples in which it is "notorious that convictions are hard to obtain, even though the officer in the case and the Crown prosecutor may believe that the complainant is truthful and reliable". If the prosecutor were to apply

"a purely predictive approach based on past experience of similar cases (the bookmaker's approach) he might well feel unable to conclude that a jury was more likely than not to convict the defendant. But for a Crown prosecutor effectively to adopt a corroboration requirement in such cases, which Parliament has abolished, would be wrong. On the alternative "merits based" approach the question whether the evidential test was satisfied would not depend on statistical guesswork."

17. FB was not argued or decided on the basis that the prosecutor had misdirected himself in law, but on the footing that the decision not to prosecute was irrational. There was no issue in the case about whether the bookmaker's approach or the MBA was the right one. The observations in FB confirmed that the MBA and the full Code test were the same.

Introduction to the facts

18. Subject to a small number of exceptions which have no bearing on this case, the court in judicial review proceedings is neither concerned nor equipped to resolve issues of fact. The public authority's evidence of the facts will be accepted. There is a consistent line of authority to that effect starting with R v. Board of Visitors of Hull Prison ex p St. Germain (No. 2) [1979] 1 WLR 1401 at 1410 H and more recently encapsulated in para 135 of R (Watkins-Smith) v. Aberdare Girls High School, [2008] EWHC 1865 (Admin); [2008] FCR 203). The court:

"must proceed on the factual basis put forward by the defendant or resolve any disputes of fact in the defendant's favour. This principle has been frequently applied."

It has particular salience in this case because the DPP has put forward a multiplicity of witness evidence, coupled with contemporaneous notes, to rebut the case advanced by the claimant.

19. The then DPP and Mr McGill (the Director of Legal Services at the CPS) have explained the decision and the background to it. Senior and experienced prosecutors have also provided witness statements. Much of the evidence concerns the assessment of trends by senior prosecutors who were responsible for overseeing decision making in RASSO cases. This evidence supports the central contention of the DPP that in 2016 there was a problem with decision making in RASSO cases. A significant proportion of cases were being prosecuted that did not satisfy the full Code test. Kevin McGinty, Chief Inspector of Her Majesty's Crown Prosecution Service Inspectorate ("the Inspectorate"), and another in that organisation, have made statements to explain their findings set out in reports from 2016 and 2019. In 2016 the DPP's concern, shared by senior prosecutors in the CPS and supported by the findings of the Inspectorate, was that the MBA was being interpreted and applied in a way which impermissibly lowered the threshold for prosecution and that steps were needed to stop that happening.
20. The claimant has served four reports from Professor Adams, an expert in the analysis of statistics. She has looked at the statistics available which detail charge and conviction rates for rape cases straddling the period when the changes to the documentation were made. We have considered her reports without prejudice to whether they are properly admissible in evidence in these proceedings. There is no doubt that there was a significant decline in the volume and percentage of rape allegations which led to a charge in 2017/18, in 2018/19 and in 2019/20. That fall is worrying, especially since the reporting of rape allegations has increased greatly during that time. The DPP is as concerned about it as the claimant. The issue which Professor Adams was asked to consider in her reports was whether that fall was "consistent with a change in…CPS practise [sic] toward the charging of rape…" (first report, para 6). She was not asked to consider what caused the fall, and, rightly, accepts that she cannot say. There are several possible causes, which could be operating singly or concurrently. She does not, of course, make the mistake of assuming that an association between two events shows that one was caused by the other.
21. In R (Law Society) v. Lord Chancellor [2018] EWHC 2094 (Admin); [2019] 1 WLR 1649 the Divisional Court explained that because of the nature of the issues on an application for judicial review, expert evidence will only rarely be admitted, and only where such evidence was "reasonably required to resolve the proceedings". Nonetheless, it admitted the evidence in that case, because the Lord Chancellor's decision was based on a technical error which would not have been obvious to the court but could be shown by an appropriately qualified person to be "incontrovertible".
22. The evidence of Professor Adams does not meet those tests. It was (self-evidently) not before the DPP when she made the decision; nor does it show that technical evidence which was before the DPP is incontrovertibly wrong. The purpose of the evidence is to support an argument that the decision led to a risk of systemic illegality. But the statistics do not show such risk. What they show, at their highest, is that, if there was a risk, that risk might have materialised or that the fall in prosecution rates is consistent with an impact of the RASSO roadshows and change made to written guidance. In our view, the evidence is not reasonably required to resolve these proceedings. We are not in a position to determine the cause or causes of the fall in prosecution rates in the relevant period which, as Mr Little submitted, were informed by a wealth of factors. We refuse the claimant's application to rely on Professor Adams' reports.

The facts

23. We were provided with twelve files of documents in this case, including the pleadings and skeleton arguments, which comfortably exceed 5000 pages. These included many witness statements with extensive exhibits.
24. The starting point for the claimant is the notes used by Alison Levitt QC in 2009 for training prosecutors about the merits-based approach. She was the legal adviser to the DPP. In paragraph 1 of her notes, Ms Levitt described four objections to the MBA.

a) It ran counter to common sense to prosecute cases which experienced prosecutors knew they were "bound to lose".

b) It was not in the interests of rape victims to get their hopes up "when we know that the outcome will be unpleasant [cross examination] followed by an acquittal".

c) It meant that prosecutors were applying a lower standard to rape cases than to other offences.

d) It was potentially unfair to defendants to put them through a trial which "we know" will result in an acquittal especially when the complainant is protected but the defendant ends up all over the newspapers.

She hoped to show that "some of these criticisms…are actually wrong" and "even those that are correct are outweighed by the advantages" of the merits-based approach.

25. Ms Levitt touched on FB. The test for rape prosecutions was the same as for other offences. It must be more likely than not that there would be a conviction. If there was no realistic prospect of conviction, no prosecution could be brought. She quoted paragraph 50 of FB (see para 16 above).
26. She explained that

"In the context of sexual offences, what this means is that even though past experience might tell a prosecutor that juries can be unwilling to convict…where, for example, there has been a lengthy delay in reporting the offence or the complainant had been drinking at the time the rape was committed, these sorts of prejudices against complainants should be ignored for the purposes of deciding whether or not there is a realistic prospect of conviction.

In other words, the prosecutor should proceed on the basis of a notional jury which is wholly unaffected by myths or stereotypes of the type which, sadly, still have a degree of prevalence in some quarters.

Instead of asking necessarily what is the LIKELIHOOD of conviction we should ask ourselves, what are the MERITS of a conviction - taking into account what we know about the defence case.

May sound like a subjective approach, even a morality judgment.

But it is not; the merits-based approach simply reminds prosecutors of how to approach the evidential stage of the Full Code Test in tricky cases.

It does not establish a different standard for sexual offences."

27. Ms Levitt's notes suggested that some jurors "think that a woman who has been drinking only has herself to blame if she is raped" or that "a woman who was really raped would of course report it immediately and delay …IN AND OF ITSELF means that she is likely not to be telling the truth'. She added 'WE KNOW THAT BOTH OF THESE ARE UNTRUE" and asked why the CPS "should refuse to prosecute simply because we MIGHT get a juror who believes something that is prejudiced, irrational and untrue" (original emphasis).
28. In para 27 she noted that the public, and "politicians ... are of course hugely critical of the number of sexual offences prosecutions which do not result in conviction". In para 28 she listed four advantages of prosecuting "safer" cases. She discounted the interests of defendants by saying "what is good for defendants is not our problem – they have their own lawyers". If there was a realistic prospect of conviction on the basis of a notional unbiased jury, the prosecutor's duty was to "the victims". She said that the issue of publicity was for Parliament, not for the CPS. It was not the job of the CPS to decide whether avoiding the trauma of a trial was in the best interests of "victims" or not. That was patronising.
29. She made clear that "we want to see the volume of prosecutions go up! We will accept that it will cost more and we accept that possibly initially we will lose more cases. But we think it is worth it. We want to see the merits-based approach applied in all the Areas and ... would accept a short-term increase in the attrition rate". She gave three reasons for this. It was "morally right". It was the "intellectually rigorous approach to the full Code test". Finally, "…by clever and sensitive prosecuting, we can actually change attitudes". She asked her audience to go back to their prosecutors "and help them to understand the approach, help them implement it" by being "bold ... brave and … creative". They should only use counsel who "understand and subscribe to the merits-based approach". She ended her talk by saying that she hoped that she had persuaded her audience of the advantages of the MBA, "the Director has made clear it is a priority".
30. The problem Ms Levitt was dealing with was a concern that, contrary to the full Code test, some prosecutors might in fact be applying the bookmaker's approach to decisions to prosecute in RASSO cases. The training was designed to refocus attention onto the proper test.
31. By 2016, Ms Levitt herself recognised that there was a problem with the way in which the MBA was being applied. She was an editor of Rook and Ward on Sexual Offences: Law and Practice, 6^th ed and wrote at para 17.16:

"Decisions in rape cases are made by specialist prosecutors who operate in [RASSOs] and have received extensive training on how to eliminate so-called myths and stereotypes from their analysis of the evidence. That being said, there is evidence that some degree of oversteering may have taken place since the judgment in FB. Some prosecutors appear to have interpreted the guidance as meaning that all complaints should result in prosecution even when this involves ignoring obvious flaws in the case."

The Merits-based Approach Guidance

32. In April 2015 the DPP published the MBA Guidance which was in force until 2017. It referred to FB, and to the contrast between the MBA and the 'bookmaker's approach'. It is in sexual offences that

"there is the greatest risk that myths and stereotypes will influence a jury and in which, therefore, an assessment based on a predictive or bookmaker's [sic] is most likely to involve a failure to apply the Code. Decisions should not be based on perceptions of how myths and stereotypes might lead a particular jury to reach a particular conclusion. The [merits-based approach] is closely linked to the CPS's determination to avoid flawed review decisions. The word "approach" does not indicate any change to what is always required when applying the [Code], which calls on prosecutors to assess whether a particular outcome is more or less likely, assuming that the case will be considered by 'an objective impartial and reasonable jury or bench of magistrates or judge hearing the case alone, properly directed and acting in accordance with the law."

33. The MBA Guidance required prosecutors to recognise that not all points made by the defence will be good ones. "Cases do not fail the code test merely because they are difficult". If the factors for and against the prosecution have been "properly considered and are not objectively undermining, the prosecution should be robust in the face of challenges". But they should not ignore factors that are objectively undermining. "Prosecuting flawed cases undermines public confidence, raises unrealistic expectations for complainants and diverts resources from other case [sic], which are delayed as a result".
34. There was a list of six bullet points.

The MBA is not a different test, merely the approach we must take in applying the Code test.

The MBA does not change, or differ from, the Code test.

It is not new. It is just a different way of expressing what has always been there.

It applies in all cases, not just rape cases.

It reflects the requirement to assume that every case will be considered by an objective, impartial and reasonable tribunal, properly directed and acting in accordance with the law, who will decide cases in the light of the evidence they have heard in court without being influenced by anything else.

It requires an objective assessment of the factors which potentially undermine the case for the prosecution or assist the case for the defence. It does not involve suspending judgement but it does require prosecutors to take objective decisions which are fair and reasonable. Any decision which takes into account subject matters like myths and stereotypes, preconceptions and predications based on previous cases cannot be an objective decision

35. The MBA Guidance then reminded prosecutors that they must not introduce a requirement for corroboration in the review process. One person's word can be, and often is, enough. What mattered was the quality of the evidence.
36. There followed a list of questions and answers that repeated that the MBA does not change the Code test. It is "best understood as an explanation of the correct principles for decision making under the Code". The tests in the Code were repeated. The MBA "makes it clear that the proper application of the Code test should reflect" the principles expressed in the evidential stage of the Code test, and the principle that prosecutors "should assume that the tribunal will approach its task in that way and will decide the case in the light of the evidence heard in court, without being influenced by anything heard, read or seen elsewhere". Prosecutors must make their decisions objectively, impartially and reasonably, according to the evidence, having regard to any defence and any other information the suspect has put forward or on which he or she might rely. They must not allow themselves to be influenced by myths or stereotypes or by predictions based on the outcomes of previous similar cases or by anything they have heard, read or seen elsewhere.
37. The MBA Guidance explained that predictions based on past outcomes are "inherently flawed since we cannot know why a particular decision was reached". It emphasised the importance of applying the Code test correctly and to record reasoning clearly. "Applying the Code test correctly necessarily involves taking the [MBA]".
38. There was a list of questions prosecutors should ask when assessing the likely impact of each piece of information or evidence, and a detailed consideration of other evidential questions, such as relevance and admissibility. Factors which undermine a particular aspect of the evidence should always be considered in the context of the evidence as a whole. Prosecutors should assume that juries will be properly directed on myths and stereotypes. Prosecutors should consider the likely impact of evidential inconsistencies. Significant inconsistencies should not be ignored. Where "a factor or a combination of factors, properly considered and assessed, leads to the conclusion that there is not a realistic prospect of conviction, then the case does not pass the evidential stage of the full Code test and should not proceed". The Guidance stated that "the [MBA] does not involve suspending all judgment but it does require prosecutors to take decisions which are fair and reasonable".
39. The MBA was also referred to in the DPP's RASSO Guidance and Child Sexual Abuse Guidelines.

The decision under challenge

40. In August 2016, Dame Alison Saunders was the DPP. Her statement explains that she was appointed DPP in 2013 and had worked for the CPS since it was established. For four years immediately before her appointment as DPP she was Chief Crown Prosecutor for London. She is familiar both with the development of policy and practice in prosecution work and "with the reality of prosecuting rape cases", including the adverse impact on complainants of "bad decisions" by prosecutors.
41. She explains in her witness statement that the CPS was the first government department to produce a strategy for tackling violence against women and girls ("VAWG") in 2008. This led to the publication of guidance or policy and specific training for prosecutors in domestic violence and rape. It later included child abuse, so-called "honour killing" and female genital mutilation. By 2013, the CPS was treating serious sexual offences as a priority. She was able to influence the training of prosecutors so that they should understand vulnerable victims, consent, myths and stereotyping, as well as how to build a case. She had frequent meetings with "stakeholders and third sector groups" on the policy of the CPS and its performance in VAWG cases, including rape and serious sexual assaults.
42. As DPP, she knew about general trends in prosecutions. She also spoke to CPS staff in the different CPS areas and in RASSO units. The CPS had ways of analysing and learning from its work, "particularly in rape cases which ended in an acquittal". Prosecutors had to submit reports to the relevant CPS Area summarising the case so that if improvements could be made and lessons learned they would be. It was her responsibility to balance the competing rights at issue in prosecutions. In the case of serious sexual offences, she considered that it was "fundamental" to ensure that "the pendulum did not swing too far so as to compromise the interests of either complainants or defendants".
43. She had some general concerns that the approach advocated by Ms Levitt had led to some "misunderstanding". It appeared that some prosecutors understood the MBA to mean that if a complainant said something and the CPS could not gainsay it, it meant that there had to be a prosecution. This meant that prosecutors were not applying the full Code test because they were not going on to evaluate and to assess whether there was a realistic prospect of a conviction. She explained:

"[My] experience was that some prosecutors understood the merits-based approach was encompassed within the evidential test in the Code…but that others thought the merits-based approach implied a presumption of prosecution. There was a concern on my part that there was inconsistency across the country. That there was an issue is borne out by the work done by Mr McGill in 2016."

44. She knew that acquittal rates were going up. She was concerned about that, and about its effect on the confidence which complainants had in the system. Publicity about failed prosecutions and high acquittal rates could deter complaints. There was also an issue for suspects if the CPS was prosecuting cases when it should not have been. She recognised that an acquittal does not necessarily show that the prosecution should not have been brought, but that if rates were rising, it was necessary to find out why. There was negative publicity about the CPS because of a "handful" of cases in 2016. She wanted to understand the background so would often ask for a report about such cases. She wanted to know whether mistakes had been made in the decision to prosecute or in the conduct of the prosecution.
45. She met Mr McGill and Mr Moore to discuss four cases on 31 August 2016. Mr Moore was her legal advisor and made decisions in some cases. Mr McGill was responsible for "the quality of legal decision-making and case progression in the whole of the CPS". Mr McGill summarises the four cases in his second witness statement. One concerned a failure in disclosure. In two cases, juries had acquitted the defendants after deliberating for very short periods: 20, and 26 minutes. Her general recollection of the meeting is that they "would have discussed the issues and debated the solutions to be considered".
46. Mr McGill presented a paper to the Senior Leadership Group on 16 September 2016. The meeting decided that Mr McGill and Mr Moore should go into the Areas and "do some refresher training on decision making in RASSO cases". These were the RASSO roadshows. Dame Alison accepts that the adverse publicity about specific cases was part of the background but the decision to do the roadshows was based on wider concerns about decisions by the CPS in cases involving serious sexual offences. She considered and approved the actions which were agreed at the meeting. The reason for the roadshows was to clarify that prosecutors should apply the Code test, "which encompassed the merits-based approach". The policy was clear but "needed to be reinforced by direct discussions and debate". She explains what the roadshows were intended to achieve. Among other things, the MBA was not "a separate, different or lesser test than the evidential test and did not need to be articulated in a way that suggested that it was". For "precisely the same reason" the term MBA was to be removed from training materials and guidance, "as was the discrete merits-based approach guidance". That did not detract from

"the core message that RASSO prosecutors were to approach the evidential test disregarding myths and stereotypes and were not to use the bookmaker's approach".

47. Dame Alison did not think that this was a change of policy and so did not consider that it was necessary to consult on it.
48. Mr McGill has been Director of Legal Services at the CPS since 2016. Having qualified as a solicitor in 1987, he worked as a crown prosecutor from 1991 to 2002. He joined the CPS again when it merged with the Revenue and Customs Prosecutions Office, for which he had been working since 2005.
49. He explains in his witness statement that, in addition to the Code, the DPP publishes internal Legal Guidance on the CPS Infonet (repeated on the CPS website). The CPS publishes extensive guidance on Rape and Sexual Offences ("the RASSO Guidance") which is intended to help prosecutors. There is a chapter on Case Building. Until November 2017, that Guidance had a paragraph on the MBA which had a hyperlink to the MBA Guidance. Chapter 21 of the RASSO Guidance is headed "Societal Myths". It explains that myths and stereotypes about rape and sexual violence can be an obstacle to getting convictions. They should be recognised and challenged. The Guidelines on Prosecuting Cases of Child Sexual Abuse were finalised in 2013 after consultation. Prosecutors should consider the credibility of an allegation, not just the credibility of the complainant. These Guidelines contained two paragraphs about the MBA which were removed in July 2017.
50. In February 2016, the Inspectorate published a report about CPS RASSO units. In 10% of the 98 cases sampled the full Code test was not applied correctly when the decision to charge was made. This reflected "poor quality decision making where decisions should be made by trained specialist lawyers working in a specialist unit…" There was evidence that some prosecutors "in a limited number of areas" were applying the merits-based approach far too vigorously and cases were charged that did not have a realistic prospect of conviction. Inspectors were also aware of cases in which the MBA was seen as separate from the Code, rather than integral to it. That led to poor decisions, an increase in acquittals, and "ultimately, a poor service to victims".
51. We note that in 2019 the Inspectorate returned to the issue in its report and concluded that decision making in RASSO cases had improved since 2016 and the steps taken by the DPP with which this claim is concerned had not driven aberrant decision making.
52. Mr McGill holds Area Performance Reviews three or four times a year when he meets the Chief Crown Prosecutor. Key performance indicators are thoroughly reviewed, including conviction rates to see whether there are any underlying issues. In 2016, partly because of the Inspectorate's report, and partly because of his investigation of data via the reviews, he began to look at performance data for rape VAWG cases. He was especially concerned about the performance of three CPS Areas. He thought that the data corroborated the concerns of the Inspectorate. Chief Crown Prosecutors generally agreed that the way decisions were made needed to be examined.
53. Mr McGill became concerned that the conviction rates for RASSO cases were particularly low. In 2015/16, he introduced a conviction rate of 60% as an "ambition", a "performance indicator" or a "benchmark by which to measure our performance". It was not "a target" although we consider the distinction a fine one. It was transient and abandoned in 2017/18. Mr McGill recognised that it had not been as helpful as had been hoped.
54. If there was "a consistent pattern of juries acquitting in the majority of rape cases, when the Crown Prosecutors had judged that the evidential test was met" that raised a question. The CPS has a duty to prosecute only where the full Code test is met because it is only by applying that test properly that it can balance the rights of the complainant and of the suspect. A key function of the CPS is to ensure that balance. From his discussions with prosecutors about failed prosecutions, it became clear to him that some of the CPS lawyers

"felt that there was (a) an expectation that they should be prosecuting cases of rape without properly applying the evidential test and (b) there was confusion as to how to properly apply the full Code test."

55. He explains that the Senior Leadership Group met every six weeks. The Chief Crown Prosecutors (members of that group) were responsible for casework in their Areas. Many members of the group had personal experience of making decisions in RASSO cases. The Senior Leadership Group enabled Chief Crown Prosecutors "to influence the development of operational strategy, practice and performance".
56. He also explains why he presented a paper to that group on 16 September 2016. He was concerned about whether the full Code test was being applied properly in RASSO cases, and whether the MBA was being misapplied. He referred to the falling conviction rate in rape trials (down from 52.4% in 2010/11 to 45% in 2016/17). It was necessary to get better outcomes. That could be done by getting 197 more convictions or by not charging 350 weak cases; because a small number of cases could have a big impact on overall performance figures. He confirms that the group agreed that he and Mr Moore would do some refresher training.

The RASSO roadshows

57. Mr McGill explains that the roadshows were intended to encourage discussion about compliance with the Code. A deliberate decision was made not to provide formal training materials. The format changed over time. Mr McGill used the statistics in the paper he had presented to the Senior Leadership Group. The CPS was getting fewer convictions in rape trials. He wanted prosecutors to be sure that they were properly applying the Code and only prosecuting cases which meet the full Code test. If prosecutors were doing that, they should carry on.
58. Mr McGill would explain that he understood from his discussions with prosecutors that they felt they had to prosecute nearly every RASSO case because the CPS expected that. He would say that it was right not to prosecute if the full Code test was not met and could not be with further investigation. He described his view that the MBA could sometimes be confusing for prosecutors. Decisions had to be made without the influence of myths and stereotypes. It was clear to him that some prosecutors were only asking "What is the merit of this prosecution?" Some prosecutors referred to the MBA as the "merits-based test". This was where the confusion lay. He would support prosecutors in making difficult decisions where the full Code test was not met. Mr Moore went through a case study with prosecutors. Both Mr McGill and Mr Moore wished to remove references to the MBA from training materials. They were satisfied it was properly reflected in the full Code test.
59. They received positive feedback. Some prosecutors were relieved because they had felt under pressure to prosecute when they felt the full Code test was not met. One prosecutor said she had thought there were two tests under the Code: one for general crime and a lower test for RASSO prosecutions. It was more than once described as "the most relevant legal discussion that they had received".
60. In a second witness statement Mr McGill provides more evidence about the roadshows given by Mr Moore. Mr Moore drafted a case study and charging advice. This was intended to illustrate how not to make a charging decision. He wanted prosecutors to think about what weight should be given to different strands of evidence to avoid misapplying myths and stereotypes. For example, it is a myth that complainants always remember a sexual assault consistently, but that does not mean that all inconsistencies in an account should be ignored.
61. Mr Moore prepared a script to assist his presentation at the Roadshow meetings. That script was augmented by a document headed "RASSO Roadshow Lines". The script refers to a charging advice which rehearses all the weaknesses in the evidence of the complainant yet concludes that "applying the MBA" its author finds that there is a realistic prospect of conviction. It goes on to ask, what is wrong with that? It gives the answer that there was little analysis of the weaknesses, and that was because the MBA "is being used as a broom to sweep away all weaknesses, come what may". Mr Moore and Mr McGill "are here to 'nudge the tiller' and perhaps tack a slightly different course – to correct that oversteer".
62. The script immediately acknowledges that there is a danger of going too much in the opposite direction: "we don't want to do that". Whether there is an oversteer which needs a nudge on the tiller depends on

"whether you are off course in the first place. So I hope to reiterate the approach we should be taking and give some practical advice on how to apply it – if that is the course you are already steering, perfect – don't change a thing. If not, then I hope what I say helps"

The aim was to help prosecutors to feel "more confident" when making decisions and to "put the merits-based approach in its proper place…"

63. The script then deals with FB, pointing out that the MBA was "hardly argued" and was "a perhaps incidental point in the case but in which a phrase was born which has taken on a life of its own and become a bit of a monster". In reality, it means no more than "looking at the merits of the case on the available evidence…rather than adopting a predictive or bookmaker's approach based on past experience." The MBA adds nothing to the Code. It is "not a device to sweep away all the weaknesses in the case. Still less is it a reason to accept what C says, come what may".
64. There is then a detailed section about how to make decisions applying the Code, some by reference to the prepared case study. The script also explains that the MBA does not require prosecutors to turn a stereotype on its head in order to ignore potential weaknesses in evidence. An objective analysis of all the evidence is needed.
65. The Lines document notes that there have been some recent cases "where the decision to charge might be described as somewhat ambitious. Some of these cases have attracted unfavourable media attention". It suggests that the trends in acquittal rate statistics "tend to support a suggestion of overcharging". The views of prosecutors for a number of Areas about the causes are summarised in three bullet points.

The belief that the MBA permits or even requires them to charge despite factors which undermine the evidence pointing to guilt;

The 'enduring impression' from the lectures delivered by Ms Levitt that charging was to be encouraged in, for instance, a "one person's word" case; and

A concern/fear that their decision might be overturned on [victim's right of review].

66. The Lines document then summarises the purpose and content of the Roadshows. The main purpose is to "give prosecutors the confidence to get charging decisions right first time and to convey the message that we will support those decisions" (original emphasis). It states, "the aim is not to go back to charging habits of old but to ensure we have not gone too far back in the other direction" (original emphasis). The style is to be an "informal lecture style but with plenty of opportunity for audience questions and interaction, which is encouraged".

Changes to the CPS Guidance

67. In his first witness statement, Mr McGill describes the changes to CPS Guidance. He explains that the CPS does in-house training. In May 2017 he said in an email that all references to the MBA in RASSO training materials should be removed and replaced with references to the full Code test. He further explains that the Prosecution Policy Unit produces legal guidance and ensures that it is up to date. At some point in 2016 the CPS started a full review of all its Legal Guidance. The RASSO guidance was being reviewed. It has taken longer than expected. The MBA legal guidance was removed from the CPS internal and external websites on 3 November 2017. The references to the MBA in the RASSO Guidance and Child Sex Abuse Guidelines were removed on 22 November 2017.

The live grounds of challenge

Discussion

The legal effect of the decision

68. The DPP may promulgate legal guidance about aspects of the work of prosecutors, and about how the Code is to be interpreted and applied. Mr Little accepts that decisions of the DPP are amenable to judicial review. If Guidance issued by the DPP is wrong in law, then the Court should say so.
69. The first issue, therefore, is whether the decision was a decision to adopt guidance which was wrong in law. Such a decision (it would follow) would have been a change from the previous, lawful, policy.
70. It is necessary to distinguish between two senses in which the MBA can be understood. The first is the meaning as it was used in FB. The second entails a group of inaccurate secondary concepts with which, in the view of the DPP, Mr McGill and Mr Moore (supported by the independent Inspectorate) the phrase had become associated in the years since FB. There are many examples of such secondary concepts in the evidence. Dame Alison Saunders was concerned that some prosecutors understood the MBA to mean that if a complainant said something and the prosecution could not gainsay it, there had to be a prosecution. Some prosecutors thought that the MBA was a presumption in favour of prosecution, while others understood the full Code test properly. Mr McGill describes how the MBA was associated with an over-zealous approach to prosecuting cases which did not have a realistic prospect of success, and with confusion about how to apply the full Code test. It was seen by some as different from the full Code test, and to be applied in RASSO cases only. Prosecutors felt under pressure to prosecute nearly every RASSO case and that the CPS expected them to do so. Some prosecutors thought the MBA was a distinct test, and some, instead of applying the full Code test, were asking themselves, "what is the merit of this prosecution?"
71. Used in the first sense, the MBA is simply a shorthand for applying the full Code test. The reasoning in FB is that the MBA was expressed in the 2004 version of the Code. The 2010 edition of the Code did not change the law but elaborated on the test in the light of the obiter reasoning in FB. It is important to remember that the DPP has never suggested that the bookmaker's approach is correct, which is why there was no issue about it in FB. It is not and has never been the case, that there are two possible tests: the MBA or the bookmaker's approach. The position has always been that there is one test, which is expressed in the Code. The MBA is another way of expressing that test. It is not an alternative to, or a substitute for it.
72. Those two aspects of the term MBA must inform any analysis of the decision. It is clear from the evidence which we have summarised that in the summer of 2016 senior leaders of the CPS were concerned about a fall in conviction rates, and about four high profile cases and the bad publicity which they had attracted. There is nothing irrational or unlawful in such concerns. They are proper concerns for the CPS, and for its senior leaders, all of whom had extensive practical experience of prosecuting cases. It is unrealistic to suggest that the CPS should not be seriously concerned about a perceived fall in conviction rates, and with the possible causes and implications of such a fall. If conviction rates are falling that is and should be a cause for concern. There are at least three obvious reasons why. First, cases might have been prosecuted that did not meet the full Code test; secondly, because of its impact on complainants present and future; and thirdly, because of its impact on suspects.
73. A reasonable decision maker could conclude that there was evidence that the Code test was not being understood or applied correctly by prosecutors on the ground with the result that cases were being prosecuted which did not meet the full Code test. A reasonable decision maker could have concluded that "a nudge on the tiller" was needed and that prosecutors should be reminded that they should only be applying the full Code test; and that they would be supported if they did so. We have carefully considered the evidence about the RASSO roadshows. They do not show that prosecutors were being told to use the bookmaker's approach. They were told, instead, with practical examples, how to apply the full Code test. There was nothing in the roadshow materials which supports a suggestion that the DPP, through Mr McGill and Mr Moore, was promulgating an unlawful approach to prosecutorial decision making.
74. For similar reasons, we do not consider that it was unlawful to decide to remove references to the MBA from the DPP's legal guidance. Stripped of references to the MBA, the remaining guidance is not unlawful. It tells prosecutors about the full Code test and how to apply it. The absence of references to the MBA does not signal expressly, or by implication, that the bookmaker's approach should be applied. The decision was a decision to change the language of the legal Guidance but that was not a change of legal substance. The position throughout the relevant period has been that prosecutors must apply the full Code test. The material from the RASSO roadshows does not undermine that conclusion. That material repeats that the test is the full Code test.
75. In summary, we conclude that the decision did not adopt guidance which was wrong in law nor was there a change in substance to the previous policy. We consider the discrete grounds in that context.

Was the decision irrational?

76. This ground suggests that the decision of the DPP to make the revisions to the written Guidance and to embark on the RASSO roadshows was irrational in the sense explained in Associated Provincial Picture Houses Ltd. v. Wednesbury Corporation [1948] 1 KB 223, in that the decision was so unreasonable that no reasonable decision maker would ever consider making it; or as Lord Diplock put it in CCSU v. Minister for the Civil Service [1985] AC 374 at 410G the decision was "so outrageous in its defiance of logic or of accepted moral standards that no sensible person who had applied his mind to the question to be decided could have arrived at it".
77. At the heart of Ms Kauffman QC's submissions on behalf of the claimant is the proposition that the decision was not based on adequate evidence. She submits that changing the materials involved risk in the sense that it might result in prosecutors adopting an over-cautious, risk averse approach to prosecutions (oversteer). She poses the rhetorical question for the DPP: "Do we have enough evidence to take this obvious risk?" She submits that there was no need to take that risk and that the DPP should have pursued a course of refresher training first to see whether that had any effect before removing the reference to MBA. She also submits that the course adopted by the DPP left prosecutors confused (or at least risked doing so).
78. Ms Kaufman went through the contemporary records of the meetings and discussions in 2016 to develop a submission that the decision making was founded on inadequate information. This submission betrayed a misunderstanding of the nature of institutional decision making and approached it as if the DPP, her senior lieutenants and the Senior Leadership Group, should make decisions as if they were adjudicating judicially, starting with a blank evidential canvas. The reality is that they all brought to bear their accumulated knowledge and experience in addition to raw data. The DPP and her colleagues had ample material, including from the Inspectorate, to conclude that there was a problem which needed resolution.
79. Whilst there can be no doubt that the DPP might have dealt differently with the problem that she and her senior colleagues had identified, including by adjusting training whilst leaving the content of the various documents alone (at least for a while), that does not begin to sustain an argument that not to do so was irrational.
80. The suggestion that the roadshows and change in the documentation sowed confusion in the minds of RASSO prosecutors is not supported by the evidence. Moreover, it overlooks the fact that the guidance is directed to experienced prosecutors. The members of the Senior Leadership Group were concerned that references to the MBA were, themselves, the cause of damaging confusion. The purpose of the decision was to minimise that confusion by removing references to the MBA (the source of the confusion) and by reinforcing, in the RASSO roadshows, that the task of prosecutors was to apply the full Code test. That was a test with which they were all very familiar. The decision was to change the language of the DPP's guidance to remove references to, and elaborations of, a phrase, which, in the informed view of senior leaders in the DPP, was confusing and contributing to poor prosecutorial decisions, without changing the legal effect of that guidance. The decision is not irrational.
81. The claimant developed an argument that the DPP failed to have regard to a number of relevant reports into sexual offending against women (to one of which Dame Alison wrote a foreword and another she jointly commissioned) in making the decision. It was irrational not to have regard to them, submits Ms Kaufmann. We have not overlooked this submission (which was not advanced in the grounds) but neither legally nor factually, in our view, does it have any merit.
82. We received extensive written and oral argument on the question whether the circumspection with which a court of review generally approaches a decision of the DPP or CPS (see e.g. R (Corner House Research) v. Director of the Serious Fraud Office [2008] UKHL 60; [2009] 1 AC 756 at [30]) has a bearing on the rationality argument. Given that this ground fails comprehensively on the evidence, it is unnecessary to explore the issue.

Did the decision create a risk of systemic illegality?

83. It is unnecessary to consider the nature of the correct legal test for intervention by a court in the face of an argument that a decision (here to remove text which did not undermine the legal accuracy of the guidance in question, coupled with statements made in the RASSO roadshows) gave rise to a risk of systemic illegality. The illegality is said to be that prosecutors would fail to apply the full Code test. This ground fails on the facts. It is linked to the assertion that the specialist RASSO prosecutors were left in a state of confusion. We reject the submission that the decision created any risk of systemic illegality. Properly analysed, the decision was a decision to remove unhelpful references to the MBA and to emphasise the primacy of the full Code test. We are unable to accept that it gave rise to any risk that prosecutors would wrongly apply the bookmaker's approach to decision making, or otherwise fail to apply the full Code test.

Should the DPP have consulted before making the decision?

84. Dame Alison fairly accepted in her witness statement that if she had thought that the decision was a change of policy, she would have consulted on it. She did not consult because she did not consider that it was a change of policy. We have held that, although the decision involved the excision of a phrase "the MBA" it was not a decision to adopt an unlawful policy, or to change the existing, lawful, policy. It was, instead, a decision that the existing, lawful, policy should be expressed in different terms. We do not consider that the DPP was under a legal duty to consult stakeholders on such a change. Consultation is designed to inform policy decisions of substance. Moreover, we are doubtful whether by 2016 and 2017 there was a legal duty to consult on changes of substance to Guidance or the content of training materials (which for these purposes the RASSO roadshows would resemble). Relevant guidance was not consulted on in 2015 although there was consultation on the Child Abuse Guidance and on the changes proposed in 2020.

Did the DPP breach section 149 of the Equality Act 2010?

85. The Claimant relies on para 26 of the judgment of McCombe LJ in R (Bracking) v. Secretary of State for Work and Pensions [2013] EWCA (Civ) 1345 for four propositions which are said to derive from section 149 of the 2010 Act about which there was no argument (see the second sentence of para 25 of the judgment). In para 44 of Powell v. Dacorum Borough Council [2019] EWCA (Civ) 23, McCombe LJ said that the previous decisions about section 149 must be taken in their contexts. The way in which section 149 will apply on the facts will be different in each case, depending on what function is being exercised. The judgments, including the judgment in Bracking, must not be read as if they were statutes. He referred, with approval, to a similar statement by Briggs LJ in para 41 of Haque v. Hackney London Borough Council [2017] EWCA (Civ) 4.
86. Section 149 of the 2010 Act applies to a public authority when it exercises its functions (see section 149(1)). It requires a public authority to give the equality needs which are listed in section 149 the regard which is 'due' in the particular context. It does not dictate a particular result. It does not require an elaborate structure of secondary decision making every time a public authority makes any decision which might engage the listed equality needs, however remotely. The court is not concerned with formulaic box-ticking, but with the question whether, in substance, the public authority has complied with section 149. A public authority can comply with section 149 even if the decision maker does not refer to section 149 (see, for example, Hottak v. Southwark London Borough Council [2015] UKSC 30; [2016] AC 811).
87. The claimant argues that the DPP failed to have due regard to the need to "eliminate discrimination, harassment, victimisation and any other conduct that is prohibited by or under any this Act". These are the factors identified in section 149(1)(a). In her written argument Ms Kaufmann focussed on the need to avoid discrimination by encouraging the application of the bookmaker's approach – in other words failing to apply the full Code test in RASSO cases, where complainants are disproportionately women. In oral argument she identified the risk of "oversteer" as being something that the DPP should have been alive to.
88. The decision in question in these proceedings was taken to deal with the specific problem of prosecutions being initiated when the evidence did not satisfy the full Code test. The decision was concerned to ensure that the full Code test was applied in all RASSO cases, as it should always have been, in accordance with the law and without any discrimination between different classes of complainant. The decision was made by the DPP in an area with which she was very experienced and knowledgeable. She had concerned herself intimately with all issues surrounding the prosecution of sexual offences precisely because she was determined to secure better outcomes for complainants, in accordance with the law and her duties. Mr Little submits that on the facts the duty under section 149(1) did not arise because the decision did not give rise to any real question engaging "the need to eliminate discrimination". It involved a reinforcement of the only lawful approach to prosecution. We agree with that submission; but, in any event, the whole underlying aim of the change was to improve decision making and conviction rates in an environment where most complainants were women. The gender context was always patent and well understood. To the extent that there was a need to have regard to the fact that complainants were disproportionately women, in our judgment, on the facts of this case, the material in the witness statement of Dame Alison Saunders satisfies us that she had the regard to the relevant equality needs which was due.

Did the DPP breach the duty of transparency?

89. This ground is based on the principle that if a public authority has a policy which it applies in the exercise of its public functions, that policy must be published. The reason for this is to enable those affected by the policy to tailor their representations to the public authority in the light of that policy and thus ensure that their representations hit the right target. There is a related principle, which is that if an authority publishes a policy, it is unlawful for it make decisions in the territory covered by the policy in accordance with a secret policy which is different from its published policy (see R (Lumba) v. Secretary of State for the Home Department [2011] UKSC; [2012] 1 AC 245. That second aspect is not engaged. Neither, in our judgment, is the first engaged. The written policies in question are not directed at the public, in order to help the public to understand how to influence decisions affecting them, but are directed, instead, at prosecutors, in order to help them to comply with their legal duties. The critical "policy" is the guidance found in the Code which is a public and transparent document

Conclusion

90. In giving the judgment of the Divisional Court Dame Victoria Sharp P observed:

"...the claimant's grounds of challenge all, with the possible exception of [transparency], rest upon a factual foundation which is fundamentally disputed by the [DPP]."

She explained why in law the evidence filed for the DPP should be preferred and then why the claim was bound to fail.

91. We have had the benefit of much fuller argument, both written and oral, than was deployed before the Divisional Court and been provided with substantial evidential material. We have set out in some detail the summary of the evidence deployed by the DPP which comprehensively undermines the grounds advanced by the claimant. The changes in language in the guidance did not change its effect. The full Code test remained. Our more detailed consideration of the materials has resulted in our coming to the same conclusion as did the Divisional Court in essence for the same reasons.
92. In the result we dismiss this application for judicial review.