- Covers judge’s direction to the jury in R v Mohammad.
- Applying too high or low a standard is an inherent problem with ‘sure’.
- A comparison with making an important decision in life dilutes the criminal standard.
The direction on the criminal standard of proof is short and relies on the nebulous concept of ‘sure’. It seems almost designed to set judges up for what could be described as ‘dreaded questions’ from juries (R v Majid [2009] EWCA Crim 2563). The ‘example’ direction, which is found in the Crown Court Compendium at 5-3, states: ‘The prosecution will only succeed in proving that D is guilty if you have been made sure of D’s guilt. If, after considering all of the evidence, you are sure that D is guilty, your verdict must be “Guilty”. If you are not sure that D is guilty, your verdict must be “Not Guilty”.’
The dreaded questions which juries ask are mostly about how sure they need to be when deciding whether D is guilty. They ask these questions in different ways using different reference points. Often, they will use percentages and sometimes they will use the old description of the criminal standard, ‘beyond reasonable doubt’ because they may have heard advocates mention it in speeches (this is permissible) or they may simply be aware of it because it is time-honoured. When judges are faced with such questions, the Compendium’s guidance is that they should be answered as shortly as possible (5-2, para 6), presumably to avoid answers which might make the concept even more difficult to understand and apply.
Relying on R v JL [2017] EWCA Crim 621, the Compendium gives judges the following guidance as to what the jury may be told (5-2, para 7): (i) they can be told that they are not required to be 100% certain if they specifically ask, ‘do we need to be 100% certain?’; (ii) if they ask about beyond reasonable doubt, they may be told (bizarrely) that ‘sure’ and ‘beyond reasonable doubt’ mean the same thing; and (iii) if the jury ask what a reasonable doubt is, they may be told that it is the sort of doubt that might affect their minds if they were making decisions in matters of importance in their own affairs.
A dubious comparison
In the recent case of R v Mohammad [2022] EWCA Crim 380, the jury asked the judge: ‘We are directed to be “sure” of guilt. How sure do we have to be? Do we have to be 100% with no doubt? Would 99% sure be acceptable for example?’
The judge’s answer was: ‘In order to prove the case against a defendant, the prosecution would have to make sure you were satisfied so that you were sure of guilt. If you were less than sure, then the case would not have been proven… The simple answer to your question is no, you are not required to be 100% sure with no doubt. The courts do not place percentages on the word “sure”. What I can say is that you should use “sure” in your deliberations as you would in your day-to-day lives, when making decisions in matters of importance in your own affairs, in your own lives or those of your loved ones.’
The direction was challenged on the basis the answer lowered the standard of proof, and to support this contention, the appellant sought to introduce an in-depth report by two experts who the court recognised as eminent in the field of forensic linguistics. In their report, Dr Chris Heffer and Professor Malcolm Coulthard concluded there were major problems involved in the use of ‘sure’ and the current law and practice on the standard of proof ‘was wrong or unsatisfactory and should be changed’ (at [19]).
Drawing on the report, one of the arguments advanced concerned the direction the judge had given to the jury about using ‘sure’ as they would in their day to day lives when making decisions of importance in their own affairs. Telling juries to make this kind of comparison has already been disavowed by appellate courts in New Zealand (R v Wanhalla [2006] NZCA 229) and Canada (R v Layton (2009) SCC 36). It is a comparison that is fraught with difficulty and tends to dilute the standard because people will take risks when making important decisions in life which simply should not be taken when deciding guilt, where liberty may be at stake and the consequences of a wrong decision will be catastrophic. However, the court held that the way the judge had explained the standard of proof was not contrary to the law of England and Wales. As to the expert evidence, while recognising it might be relevant to policy issues in other contexts, the court held it was not admissible as evidence in the appeal or indeed in any appeal.
It is a pity the court left this comparison intact. It creates a real risk of both lowering the standard and creating inconsistency between individual jurors as they attempt to understand the standard and apply it. There are degrees of ‘importance’ and an important decision in life will vary between jurors according to their age, life experience and circumstances. Some jurors might be young and feel they have not yet taken a truly important decision in their lives, and so have no frame of reference. Older jurors might have made several important decisions and might have become hardened to risk and less affected by doubt over time, or more averse and careful. Higher or lower standards might also be applied depending on a person’s general disposition. There might be jurors who have been gung-ho about important decisions in their lives, and who might be robust and persuasive in the jury room.
The problem with ‘sure’
The problem inherent in the use of ‘sure’ is that jurors may apply too high or too low a standard of proof. Because the concept is vague, jurors may struggle with issues such as whether they need absolute 100% certainty (they don’t) or how much doubt they are able to live with and still be legally sure (fanciful doubts may be disregarded), or what kind of doubt means that they cannot be sure (reasonable doubts). In R v Mohammad, the court observed that the jury’s question meant they were operating within a narrow field of extremely high probability, between 99%-100% and there was no risk of the standard having been reduced. However, while this jury may have been applying a standard of 99%-100% probability when reaching its majority verdict, academic research has shown that people’s understanding of how much probability is needed to be legally ‘sure’ can be highly variable and, if cast in percentage terms, can be much less than 99%.
For example, in separate articles in NLJ two years ago, Professor Zander and Professor Keane (replying to Professor Zander) explored this. Professor Zander revisited interesting research he had undertaken 20 years previously, when he had asked members of the public and magistrates to state ‘sure’ as a percentage (The criminal standard of proof: how sure is sure? 150 NLJ 1517; The criminal standard of proof: how sure is sure? Pt 2 170 NLJ 7888 29, p 18). Three quarters of respondents stated that ‘sure’ meant 90% or more, showing that most take the duty very seriously and set the standard high. However, Professor Keane (Not so sure about the criminal standard of proof, 170 NLJ 7890, p15) pointed out that one quarter of respondents cast it as less than 90%, which must be too low, and alarmingly, a minority of 5% cast it as 75%, which is far too low. This kind of variability is consistent with findings in other academic studies which Professor Keane referred to in his article.
Missed opportunity
It is hard to disagree with expert voices from the field of forensic linguistics which say the law and practice relating to the criminal standard of proof is unsatisfactory and needs to be changed. Jury questions reveal a problem of comprehension and show that juries do need more help. Is the law and practice on the standard of proof likely to change? The short answer is, ‘don’t hold your breath’. The Court of Appeal’s role is to apply the law and to consider whether the conviction is safe, and it is disinclined to get too involved with problems inherent in the use of ‘sure’. This is unsurprising, but in R v Mohammad there was an opportunity to at least discourage judges and advocates from drawing the dubious comparison between being sure when deciding whether a person has committed a criminal offence and being sure when making an important decision in your own affairs. This would have been a small step to take and one which would be consistent with appellate authority in other common law jurisdictions. At the end of its judgment the court in R v Mohammad said this was not ‘a suitable case, if ever there might be, for a wholesale and fundamental review of authorities or the established practice in England and Wales in relation to the standard of proof’ (at [34]). The words ‘if ever there might be’ mean the door to a fundamental review is not entirely closed. But it will be exceedingly difficult to get through.
Paul McKeown, Associate Professor of Law, City, University of London, and author, with Emeritus Professor of Law Adrian Keane, City, University of London, of The Modern Law of Evidence, 14th ed, OUP.
