Abolishing the "not proven" verdict

The Scottish Parliament's Justice Committee is currently considering the Criminal Verdicts (Scotland) Bill, the latest attempt to abolish Scotland's peculiar "not proven" verdict. We have submitted evidence to the committee in support of abolishing that verdict, in the following terms.

Justice Committee

Criminal Verdicts (Scotland) Bill

Written submission from James Chalmers and Fiona Leverick

1. Thank you for the invitation to submit evidence on the Criminal Verdicts (Scotland) Bill.

2. We submitted a written response to Mr McMahon’s consultation in 2012, and would repeat the general point which we made then. We support a two verdict system. There is, quite simply, no merit in having two different verdicts of acquittal, when each verdict has exactly the same practical consequence and the distinction between them is not well understood. We can see no defensible case for the current system. In particular, we support the argument that it is wrong for a verdict of acquittal to carry any implication of stigma. What follows is intended to provide some background information which may be of assistance to the Committee in considering the Bill.

3. Earlier this year, we edited (along with Alasdair Shaw, then a research assistant at the University of Glasgow School of Law) the Report of the Academic Expert Group for Lord Bonomy’s Post-Corroboration Safeguards Review. That report includes a detailed chapter on jury majority, size and verdicts (chapter 10) which itself contains a discussion of the not proven verdict (pages 157-160). We draw on that work here, and reproduce (with modification) some of the text of the report, but draw the Committee’s attention to the report for more detailed discussion.

4. We would also draw the committee’s attention to the research paper by Lorraine Hope and others, “A third verdict option: exploring the impact of the not proven verdict on mock juror decision making” (2008) 32 Law and Human Behaviour 241-252, which is summarised in that chapter.

5. The chapter discussing juries in the Report of the Academic Expert Group suggested that a number of tentative conclusions might be drawn from the research carried out by Hope et al. We summarise these conclusions, along with the relevant aspects of that research paper, as follows:

• In theory, the existence of the not proven verdict should have no effect on the rate of acquittals in criminal cases. Jurors should decide first, whether the case has been proven beyond a reasonable doubt and only then consider which verdict of acquittal is appropriate. However, Hope et al’s research suggests that the availability of the verdict may have some effect on the rate of acquittals.
• The not proven verdict may help to prevent wrongful convictions, because it was in the cases described by Hope et al as involving “moderately strong” evidence – those where most caution would seem to be required – that it appeared to be particularly attractive to mock jurors.
• 37 per cent of the mock jurors who had considered the case on the basis of three verdicts believed that a not proven verdict permitted the retrial of the accused at a later date, despite written instructions which they had received to the contrary. The extent to which jurors may wrongly believe that a not proven verdict permits a subsequent retrial, despite having received direct instructions to the contrary, is particularly concerning.
• The not proven verdict may inhibit thorough deliberation by jurors; Hope et al noted that deliberations between mock jurors appeared to “dry-up” once the possibility of a not proven verdict was raised.

6. The not proven verdict, then, might offer a degree of protection against wrongful conviction by rendering juries less likely to convict in marginal cases. It does so, however, at a cost: jurors may misunderstand the consequences of the verdict, its availability as a compromise verdict may inhibit deliberation (something which could increase the likelihood of both wrongful conviction and wrongful acquittal in different cases), and it is undesirable in principle to have two different verdicts of acquittal when the difference between them cannot properly be articulated.

7. It is tempting to assume that any problems with the not proven verdict – particularly any misunderstanding on the part of jurors – should be capable of being cured by appropriate direction from the trial judge. However, if a criminal justice system employs two different verdicts of acquittal, it is rational to assume that there is some meaningful difference between them, whether in meaning, effect, or both. The notion that there is no meaningful difference – or, at least, none which can be safely spelt out – is at first sight so irrational that it is plausible that a juror may fail properly to understand a direction to that effect.

8. We hope that this information is of use, and would conclude by noting two points of distraction which may arise during discussions on this Bill.

9. Retaining not proven and removing not guilty. When this debate has arisen in the past, it has sometimes been suggested that – if Scotland were to move to a two-verdict system – it would be appropriate for “not guilty” to be the verdict which disappears, on the basis that all a jury ever does is to decide whether a charge is proven or not. This is an unhelpful distraction which proceeds on a failure to appreciate the consequences of the presumption of innocence. It is correct to say that all a jury can ever do is ask whether a charge is "proven" or "not proven". But that is true of any jury in any system. Because all persons are innocent until proven guilty, any person against whom a charge is not proven is entitled to be declared innocent in law. The presumption of innocence means that a not guilty verdict is the logical consequence of a charge not being proven.

10. Waiting for jury research. It might be suggested that it would be better to postpone consideration of this issue pending jury research. The Committee should, however, note the limitations of such research. Research might, for example, show that mock juries asked to view simulated trials are either more or less likely to convict when presented with two possible verdicts rather than three. There would, however, be no means of establishing for the purposes of such a study what the correct conviction rate was, and so the research would not establish which of a three or two verdict system was “better”. Ultimately, the not proven verdict raises questions of principle which must be confronted directly and cannot be evaded by calls for further empirical research.

James Chalmers Regius Professor of Law, University of Glasgow

Fiona Leverick Professor of Criminal Law and Criminal Justice, University of Glasgow