The Criminal Procedures Bill is about to begin its path through Parliament (Double Jeopardy Debate Begins). Among other things this bill will allow for retrial of aquitted defendants under some circumstances. Civil libertarians may be concerned about this change but knee-jerk responses need to be avoided.
I have always felt the more important issue was whether juries correctly apply the standard of "proof beyond reasonable doubt" in the first place. In scientific investigations a trial investigating (for example) the efficacy of a new drug is considered "significant" only if the statistical probability that it's results could have arisen by chance is greater than 1 in 20, and the drug's efficacy would be accepted only when confirmed by significant results from a number of independent studies conducted by multiple researchers under varying circumstances. Even then few scientists would say that the efficacy of the drug was proven "beyond reasonable doubt".
Juror's are denied the luxury of waiting for further results. It is the prosecution's responsibility to investigate thoroughly before the case comes to trial. A thoughtful juror who judges that the defendant is "90% probably guilty" should demand an aquittal but will come under enormous pressure from other jurors who don't know the meaning of the word, "doubt". This is not helped by the knowledge that a wrongful acquittal can never be remedied coupled with the (largely erroneous) belief that a wrongful conviction can. It would help if all juries were more thoroughly instructed on the meaning of reasonable doubt (any doubt that a reasonable person might seriously entertain) and specifically warned against applying the lesser standard of "balance of damage" ("better to acquit ten guilty men than to convict one innocent man"). It would also help if a "not guilty" verdict were called "not proven" (which is what it is).
Actually I quite like the idea of a two-stage process for criminal proceedings. The first stage would be conducted as at present but the jury would decide whether each charge was "proven" or "not proven". Majority verdicts might be allowed. A verdict of "proven" would then proceed to argument on sentence (just as a guilty plea does now). A verdict of "not proven" would lead to a second stage in which both counsel and the trial judge could resummarise their cases (but not present new evidence). The jury could then (after deliberation) bring in a new verdict of "not guilty" (if they were satisfied that the accused was innocent on the balance of probabilities) or leave the verdict as "not proven". A "not guilty" verdict would then be a positive finding of innocence and would normally entitle the accused to costs and to protection against double jeopardy. The Crown would be able to appeal a "not proven" verdict (and possibly gain a retrial) under more or less the same conditions as a defendant can appeal a "proven" verdict but could appeal a "not guilty" verdict under only very limited conditions.
I'd be interested to hear other bloggers thoughts.