Hugh McLachlan: A mistake to put previous convictions on trial

The Scottish Law Commission has recently proposed that juries in Scotland should be made aware of the previous convictions of the accused in criminal trials. We should be resistant to this proposal, particularly if it is motivated by a desire to increase the conviction rate.

Evidence and information that might be relevant for pre-trial enquiry will not necessarily be appropriate for use at a trial. For instance, previous convictions of a person can be reasonable grounds for suspecting him of the commission of some particular subsequent similar offences and good grounds for investigating his movements at the times of their suspected occurrence. However, it might not be appropriate to cite past convictions in an actual trial.

Whether or not such previous convictions are evidence that an accused person has committed the particular subsequent crime at issue, it could be unfair to cite them. It could be prejudicial to a fair trial. There is a danger that the convictions would be regarded as more powerful evidence than they reasonably should be. They might have a blinding effect on a jury.

If we know that someone has previously committed, say, several rapes or murders, we would be more likely to believe that he is guilty on subsequent particular occasions than if we believed that he had, hitherto, led a blameless life. However, it does not follow that we are justified in our belief.

It is the nature of the justification of the beliefs of members of a jury concerning the innocence or guilt of accused people, rather than the intensity with which they are held, that is crucially important for the delivery of fair, reasonable and just verdicts.

Furthermore, it could well be argued that knowledge concerning previous convictions is not evidence of any sort – not even weak evidence – with regard to the question of whether or not an accused person committed a particular subsequent offence.

Someone who has never consumed alcohol is less likely than a habitual drunkard to be guilty of drink driving next week. However, it is not clear that, or how, such a consideration should influence a jury.

Suppose that two people are charged with drink driving and that the case against each of them is identical, apart from the fact that one of them has a previous conviction for drink driving and is a reputed drunkard. With regard to the specific accusations against them, should the courts not treat them both the same? I think so.

The principle is the same no matter what the nature of the alleged offence.

Suppose that members of a jury are aghast and groan audibly when they discover that someone whom they have acquitted of, say, rape or murder has previously been convicted for having committed such an offence. The inference that might be drawn is that, had they known of such convictions before they gave their verdict, they would not have acquitted him.

This is a good reason for concealing such information about previous convictions from the jury.

If they thought that the case presented in the court failed to establish beyond all reasonable doubt that the accused person had committed the crime in question, then they were right to acquit him.

If the jury thinks that the case which was presented has not established beyond all reasonable doubt that the defendant is guilty of the crime in question, it is difficult to see how the addition of the information concerning previous convictions could render any lingering doubt unreasonable.

If there is not sufficient evidence to convict someone without knowledge of his past convictions, there is, I suggest, insufficient evidence to convict him fairly, justly and reasonably.

There is a difference between using statistical unlikelihood in a general, forward-looking way, and using it retrospectively with regard to particular instances.

The odds against winning the lottery are vast. That is a good reason for thinking that Mr X of Plymouth will not win the lottery next week. However, it is not evidence that he did not win it last week. Things which are very highly improbable do happen.

Similarly, with regard to both horses and people, previous form is a reasonable basis for predicting future behaviour and for calculating the probabilities of future actions. However, it should not be confused with evidence pertaining to particular actions.

For instance, suppose that we backed a horse at the odds of, say, 100/1. After the race was run, we would not accept the bookie’s claim that the odds against our horse’s winning are good evidence that it did not win. They are not even weak evidence. If the question of what horse actually won the race were to be subject to an inquiry, those required to assess the relevant evidence should disregard the previous form of the horses and riders concerned.

Jurors should take a similar approach when they are asked to consider whether or not a particular person performed a particular action at a particular time and place.

It is not the purpose of criminal courts to secure convictions but to give those who are accused of committing particular crimes fair, just and reasonable trials. If the rules and procedures of the courts are fair, reasonable and just, we should accept the outcomes whatever they might happen to be.

Low conviction rates are not in themselves a bad thing, any more than high conviction rates are in themselves a good thing. For instance, there was a high rate of convictions in Scotland in the 16th and 17th centuries in witchcraft trials. That was not a sign that the courts were functioning well. Similarly, that there is a low rate of convictions in, for instance, rape cases is not in itself a sign that the courts are functioning badly.

There is a danger that the proposed change would make Scottish criminal courts less fair, less reasonable and less just than they are. We should be wary of making known to them the previous convictions of defendants lest juries are irrationally or unduly influenced by such information.

• Hugh McLachlan is Professor of Applied Philosophy at Glasgow Caledonian University