Key Points:

The "not guilty" verdicts in the recent trial of Clint Rickards, Brad Shipton and Bob Schollum reveal the difficulties the public faces in interpreting criminal trial outcomes. A verdict of "not guilty" is not a verdict of innocence.

Indeed the term "not guilty" itself is not a correct use of the English language. A verdict of not guilty means that the jury does not find the accused "guilty". It does not mean that the jury finds the accused "not guilty". An acquittal is a non-verdict. It covers three logical possibilities: that the accused is innocent, that the outcome is inconclusive, or that the accused is guilty "on the balance of probability" but not "beyond reasonable doubt".

This is an age-old problem central to the scientific method of knowledge acquisition. The best a forensic scientist can do is to disprove a hypothesis (such as the hypothesis of innocence), beyond reasonable doubt. In many cases, such disproof is extremely difficult. That means for example in economics and other social sciences we continue to treat a number of hypotheses that are most likely false as if they were true.

The science of statistics attempts to assign probabilities to verdicts. Hence a popular standard of proof "beyond reasonable doubt" might mean that, for every 100 guilty verdicts, 99 accused persons really are guilty and that only one in reality is innocent.

A different standard of proof "on the balance of probability" is used to gain protection orders necessary to evict people accused of domestic violence from their homes. This may mean that, out of 100 persons with protection orders granted against them, 90 really were a danger to their families and 10 were not a danger. In this case great weight is given to the safety of children in an allegedly violent household, at some cost to the rights of the accused. Likewise, the decision whether to fire an accused person from their job arguably requires a different standard of proof than does the decision by a jury whether to convict.

In the case of the courts, the myth we most commonly subscribe to is that an acquitted person has been cleared of wrong-doing, and should be regarded as innocent, despite no finding of innocence by a judge or jury. That is an unreasonable expectation to place on the public.

A popular form of criminal defence in New Zealand is for lawyers to simply create doubt in the minds of a jury, given that is all that is required to gain a "not guilty" verdict. While such a defence may achieve its immediate objectives, it does little to assure the public that such an accused is innocent.

In Scotland the jury can opt for a "not proven" verdict instead of a "not guilty" verdict. This means that a person found "not guilty" is really believed to be innocent by the jury. This contrasts with the New Zealand case where all "not guilty" verdicts really are "not proven" verdicts. The semantics are very important to an accused who really is innocent.

Our juries should be asked to deliberate on an accused person's innocence as well as on their guilt. The reporting procedure could be as follows. The judge would ask the jury if the accused has been found guilty of the charge. The foreman of the jury would simply say "yes" or "no". If the answer is "no", then the judge would ask if the accused has been found innocent of the charge. Once again, the answer would be "yes" or "no". Finally, in the event of another "no" answer, the jury should be asked if they believe that the accused is "probably guilty".

A person who is found "probably guilty" will suffer no formal retribution such as imprisonment, but will face a degree of social opprobrium. More importantly, to be found to be innocent will free accused persons of the public perception that they only "got off" because they had a "clever lawyer".

By reporting verdicts in this way, there will be more onus on defence lawyers to attempt to establish the innocence of their clients, and more accused will choose to testify rather than relying on their barristers to make a case based on doubt. This procedure would place more emphasis on the finding of the truth. And less emphasis would be placed on the trial as a contest with a winner and a loser.

One final benefit would be that the public understanding of many wider issues would be enhanced. For example the battles by war veterans who have been exposed to chemicals (such as Agent Orange in the Vietnam War) would be understood more clearly. In many of these cases, the war veterans' illnesses (and the genetic problems faced by their children) may be shown to be caused by these chemicals "on the balance of probability" though not "beyond reasonable doubt".

Going back to Rickards, Shipton and Schollum, decisions about the employment future of Rickards in particular might be better informed if we knew whether in fact he was found by a jury to be innocent.

* Keith Rankin teaches economics and statistics at the Unitec Business School