Sunday, March 06, 2005

Equality of arms

The New Zealand Court of Appeal, in Brown v Attorney-General 3/3/05, CA39/03, has, without having to decide the question, touched upon the topic of whether compensation is an appropriate remedy where there has been a breach of the right to a fair trial. This topic has recently received attention in these blogs (Monday, 21 February 2005, see the discussion of the House of Lords decision in R v Secretary of State for the Home Department, ex parte Greenfield [2005] UKHL 14 (16 February 2005)).

Five judges heard the appeal in Brown. All agreed in the result, four judges jointly, and William Young J delivered a separate judgment in which he preferred to decide the matter of the availability of compensation rather than, as the other judges had, leave it open. The appeal from the refusal of compensation failed.

The joint judgment, delivered by Chambers J, held that the State (in the form of a District Legal Aid Sub-committee, referred to as the DLSS) did not breach Mr Brown’s right to a fair trial. This was so because the decision it reached, not to cover the costs of a DNA analysis that the defence sought to have conducted in Australia, was reasonable on the information before the DLSS at the time the decision was made.

Two concerns might be entertained about this approach. First, was the DLSS obliged to give reasons for its refusal to allow the costs? It is not clear from the judgment whether the DLSS did seek additional information or advise counsel for the defendant of what might have strengthened the claim. Much of the joint judgment is concerned with the inadequacy of the information placed before the DLSS by counsel for Mr Brown. A failure to give reasons would, of course, only be significant if it would have led to a different decision. This brings us to the second concern.

Were the results of the tests that were sought to be conducted likely to be of significance for the defence? The joint judgment (para 64, 85, 90, 98) assumes that the results would not have been significant. Essentially, the defence sought to establish that the accused’s DNA was not present on a shirt worn by a robber, and that therefore he was not the robber. The Crown accepted that there was no DNA which could be proved to have come from the accused on the shirt. The joint judgment concluded that it was unlikely that further testing would have produced any results, let alone decisive results (para 85, points (b) and (c)), saying that the absence of Mr Brown’s DNA took the matter nowhere.

Well, one might respectfully wonder about that. If the most sensitive testing then available could not find evidence that should be there if Mr Brown had worn the shirt, that fact would have significant probative value in favour of the defence. This can be seen from an application of Bayes’ Theorem. The critical point is, how likely is it that the robber’s DNA would have been detectable on the shirt? Scientists have an obligation to consider matters such as that (see Robertson and Vignaux, Interpreting Evidence, 1995, pp 76 – 77). If the robber’s DNA should have been there, and especially if DNA was there but it wasn’t Mr Brown’s, then the likelihood of Mr Brown being the robber was diminished. [Update: in dismissing the application for leave to appeal this decision, Brown v Attorney-General [2005] NZSC 59, 29 August 2005, the Supreme Court noted that the evidence was that a significant proportion of the population do not secrete DNA in their sweat. The SC held that there was no evidence that the absence of the tests that the defence had requested caused any loss to the appellant.]

Applying Bayes’ Theorem involves comparing the probability of obtaining the evidence (that Mr Brown’s DNA was not detected) assuming that he was guilty, to the probability of obtaining the same evidence given that he was not guilty. The ratio of these probabilities (the former divided by the latter) is called the likelihood ratio. If the accused was not guilty, then the probability of the "none of his DNA" result would be 1. The denominator of the likelihood ratio would be 1. The numerator is therefore the critical thing: what is the probability that the robber would not leave detectable DNA on the shirt? That is the critical point on which the DLSS should have been informed (or, on which it should have sought information).

If there was a chance of, say, one in ten that the robber could have worn the shirt without leaving DNA on it, then the likelihood ratio would affect the likelihood of guilt compared to the likelihood of innocence, by a tenth. That, one would have thought, should have been sufficient to raise a reasonable doubt and thus be grounds for a verdict of not guilty.

A weakness of the joint judgment in Brown is, therefore, its emphasis on the results of the proposed tests being unlikely to assist the defence.

This is not to say that the Court was wrong to reject compensation as a remedy for breach of the right to a fair trial. While the joint judgment leaves the point open, it does acknowledge the matters referred to by William Young J at paras 126 to 142 of his judgment. Even if there had been a breach of his right to a fair trial, Mr Brown’s remedy was through the appeal process (obtaining a quashing of his conviction and, perhaps, a retrial). The most persuasive reason for this conclusion is the difficulty in distinguishing cases of ordinary error at trial, cured by appeal, from errors that merit the constitutional remedy of compensation from the Crown. After all, there will be overlaps. An erroneous direction to the jury on the applicable law, curable by appeal, deprives the accused of the right to a fair trial. And misconduct by agents of the State is normally remedied by exclusion of evidence or stay of proceedings, not by compensation. There is therefore a question over the future of Simpson v Attorney-General (Baigent’s Case) [1994] 3 NZLR 667 (CA), at least to the extent that it might have been thought to be applicable, beyond its facts, to cases where the rights that are breached belong to a person who is, consequently, charged with an offence.

Two aspects of William Young J’s agreement on the issue of breach of rights leave cause for concern. At para 124 he assumes that the accused would not have himself paid the costs of the tests, had he had the funds to do so. A person of even ordinary financial means would probably think anything around $10,000 (the tests here were likely to have cost much less) to be a wise investment, in view of the prospect of 9 years’ imprisonment (the sentence Mr Brown received). Secondly, William Young J says (para 125) that the appellant’s approach to the equality of arms argument was selective: results favourable to the prosecution should have been disclosable if the tests had been allowed. This, with respect, is a dangerous suggestion. If the tests had been privately funded, any results unfavourable to the defence would certainly not have been disclosed to the Crown, so why should an impecunious defendant be in any different position just because the State is funding his defence?

No comments: