Friday, October 03, 2014

Book Review: Peter Gill, "Misleading DNA Evidence: Reasons for Miscarriages of Justice"

Every criminal lawyer needs to read Peter Gill’s “Misleading DNA Evidence: Reasons for Miscarriages of Justice” (Academic Press, 2014).

Judges should read it too. It is admirably clear and no expertise is required of the reader. If you have some experience with DNA evidence and Bayesian reasoning you will still find the book useful for its descriptions of how things can go wrong.

Peter Gill is one of the world’s leading experts on forensic DNA. I say ‘one of’ just to be polite to any other contender, for Dr Gill is really the leader.

Things can go wrong (as I so eloquently put it above) because DNA detection is now very sensitive and DNA is everywhere. A typical crime scene will have detectable amounts of DNA from many sources. Partial DNA profiles can be detected too, and this complicates the evaluation of mixed samples. Investigators can contaminate a crime scene, transferring DNA from one thing to another, with even latex gloves being vehicles for transfer. Detection of DNA does not mean that it is necessarily present because it was in a body fluid present at a crime scene. Samples collected for analysis must be sealed, and gloves changed, so that no transfer should occur between them, and in the laboratory the risk of transfer between suspect and scene samples must be minimised. There should be studies of the risks of contamination within a laboratory, and these should be done without the knowledge of those who work in that laboratory.

Courts tend to leap to conclusions that are not scientifically sound. Inferences that could be tested experimentally are often drawn without that evidence. A common example may be the assumption that a defendant’s DNA on a murder weapon found at a crime scene was only on the weapon and was not also on a  number of other items which would innocently explain why it was also on the weapon; what the media refers to as the Amanda Knox case is an illustration that Dr Gill uses to make this point: the defendant’s DNA on a knife, allegedly the murder weapon, in a kitchen drawer, was given significance it may not have deserved in the absence of evidence of whether her DNA was also on other items in the drawer (which it surely would have been as she used the kitchen).

I don’t want to spoil your enjoyment of this book by saying more. The Kindle app version, giving coloured text, has links that are easy to navigate, including links to some of the cited technical papers that are available online. It is a book that should be discussed with expert witnesses, whether they are relied on by the prosecutor or by the defendant.

Sunday, August 24, 2014

Wrapping up

And that was ten years of commentary. Just a consequence of having the radio on while working at my desk on 25 August 2004, and hearing a computer person being interviewed about how to start a blog.

600 commentaries. Ovid’s self-rebuke comes to mind:

Cum relego, scripsisse pudet, quia plurima cerno, me quoque qui feci judice, digna lini.


Saturday, August 16, 2014

A defendant's liability for Parliament's mistake

A couple of statutory interpretation questions were answered in Beezadhur v The Independent Commission against Corruption & Anor (Mauritius) [2014] UKPC 27 (7 August 2014), one unanimously, the other not.

The relevant provisions were s 5 of the Financial and Anti-Money Laundering Act 2002 [Mauritius], which prohibited transactions in cash above a fixed amount but subject to exemptions, and s 2 which defined exempt transactions in terms which, relevantly here, included the phrase “lawful business activities”. This legislation has, since the result of this case in the Mauritius Supreme Court, been amended by omitting the reference to “business” [4].

The Board agreed that the defendant had the burden of proof at trial on the issue of whether a financial transaction came within a statutory exception to liability, applying the long-recognised [26] policy that this should be no particular hardship because the information would be readily available to the defendant [32], applying R v Johnstone [2003] 1 WLR 1736 para 50 per Lord Nicholls.

The other issue, the meaning of “business” in the relevant statutory exemption, in particular the phrase “the transaction does not exceed an amount that is commensurate with the lawful business activities of the customer”, drew a dissent from Lord Kerr. He favoured a liberal construction, to give a wider exemption, consistently with the shift of the burden of proof to the defendant [39], [40]. The majority held that the policy of the legislation required exemptions from liability to be construed narrowly [22], [23].

This raises, we might think, a more general question: should legislative action, taken before a case has completed its appeal process, to remedy what has been revealed in the lower court to be an unintended consequence of the original enactment, be taken into account in interpreting the original provision for the purposes of the final appeal?

It was accepted [37] that the defendant’s (appellant’s) four deposits, which were retirement savings, were from legitimate sources and his one withdrawal was also for legitimate purposes, and the Board noted that there was no reason to believe that he thought he had broken the law. The legislation was new and it was not clear to what extent it had received publicity. The Bank, it seemed, should have alerted him to the requirements of the Act, and for unknown reasons the Bank had not also been prosecuted. Although there was no appeal against sentence (which had been fines on each of the five counts), the Board indicated that a non-penal disposal could properly have occurred.

The limitation on the Board’s powers, confined here to the appeal against conviction, meant that the defendant had received an unduly harsh sentence (I say “harsh” relative to a discharge; the total fines here amounted to only 1.68 percent of the total monies involved in the charges). Where the penalty imposed is harsher, the jurisdictional limit may be worked around, so to speak: compare Ramdeen v The State (Trinidad and Tobago) [2014] UKPC 7 noted here on 28 March 2014, where the Board riled against a jurisdictional limitation and ordered commutation of a death sentence to life imprisonment. I try not to be sarcastic, but must ask: is it only murderers who get the Board's extended efforts to achieve a just result?

Thursday, August 14, 2014

More on expert evidence

Two decisions of the High Court of Australia yesterday don’t develop the law but they do illustrate how careful everyone should be about expert evidence.

Is the evidence based on specialised knowledge?

In Honeysett v The Queen [2014] HCA 29 (13 August 2014) an expert was called to compare images of the offender obtained from CCTV footage with photographs of the defendant (appellant) taken at a police station. The expert noted some points of similarity. Although in the High Court the respondent suggested that this evidence was adduced only to show that the defendant “could not be excluded” as the offender, the reality was that at trial it had been advanced as supporting an inference of identity. The issue was whether the witness was giving expert testimony.

It seems obvious, now that we know the result, that this sort of comparison could be made by members of the jury for themselves; it did not require any expertise. The limited nature of the evidence in this case was such that it did not involve use of the witness’s expertise in anatomy. Also, the respondent resiled from what in the lower courts had been advanced as engaging expertise: the witness’s experience in viewing CCTV images and making comparisons with suspects; only expertise in anatomy was now relied on [39].

The non-engagment with the witness’s expertise was apparent from a passage in the court below (Macfarlan JA, with whom the other members of the NSW CCA agreed, in upholding the conviction), and quoted by the High Court at [37]:

“In addition to his formal qualifications in anatomy, Professor Henneberg is a person of extensive practical experience in examining CCTV footage, with all its deficiencies, and attempting to identify characteristics of persons depicted in it. The view he expressed on this topic is necessarily subjective and not amenable to elaboration beyond the reasons he gave, or to measurement and calculation.”

(emphasis added by me). This, correctly understood, supported the conclusion that the evidence was not based on specialised knowledge [43]:

“Professor Henneberg's opinion was not based on his undoubted knowledge of anatomy. Professor Henneberg's knowledge as an anatomist, that the human population includes individuals who have oval shaped heads and individuals who have round shaped heads (when viewed from above), did not form the basis of his conclusion that Offender One and the appellant each have oval shaped heads. That conclusion was based on Professor Henneberg's subjective impression of what he saw when he looked at the images. This observation applies to the evidence of each of the characteristics of which Professor Henneberg gave evidence.”

The evidence was not admissible. The Court did not have to decide whether, as submitted by the appellant, to be based on specialised knowledge an opinion must be able to be independently validated [42].

Mere use of technical language does not mean that an opinion is based on specialised knowledge [45]:

“Professor Henneberg's evidence gave the unwarranted appearance of science to the prosecution case that the appellant and Offender One share a number of physical characteristics. [footnote: HG v The Queen [1999] HCA 2; (1999) 197 CLR 414 at 429 [44] per Gleeson CJ; Morgan v The Queen [2011] NSWCCA 257; (2011) 215 A Crim R 33 at 61 [145] per Hidden J.] Among other things, the use of technical terms to describe those characteristics – Offender One and the appellant are both ectomorphic – was apt to suggest the existence of more telling similarity than to observe that each appeared to be skinny.”

What is the witness actually saying?

The other decision, Fitzgerald v The Queen [2014] HCA 28 (13 August 2014) concerns a prosecution case that, it now seems, could hardly have been weaker: a trace of DNA from the defendant (appellant) was found on a didgeridoo located at the scene of a violent attack. The prosecutor, held the High Court, could not exclude innocent explanations for the DNA’s presence and the conviction was quashed and a verdict of acquittal entered.

The error in the courts below was essentially that the expert’s evidence had been misunderstood. In evaluating whether the prosecutor had proved guilt beyond reasonable doubt the High Court referred to the following points [36]:

“On Dr Henry's evidence ... the prosecution's main contention, that the appellant's DNA in Sample 3B derived from the appellant's blood, was not made out beyond reasonable doubt. Secondly, Dr Henry's evidence was not that secondary transfer of DNA was "rare"; rather, she said that a primary transfer is a much more likely source of contact or trace DNA than a secondary transfer, but that nevertheless a secondary transfer of contact or trace DNA is possible. There was no conflict in the evidence that there were at least two distinct occasions ... on which a secondary transfer of the appellant's DNA to the didgeridoo may have occurred. Thirdly, the recovery of the appellant's DNA from the didgeridoo did not raise any inference about the time when or circumstances in which the DNA was deposited there. For those reasons, it could not be accepted that the evidence relied on by the prosecution was sufficient to establish beyond reasonable doubt that the appellant was present at, and participated in, the attack. The jury, acting reasonably, should have entertained a reasonable doubt as to the appellant's guilt. [footnote: M v The Queen (1994) 181 CLR 487 at 493-494.] Alternative hypotheses consistent with the appellant's innocence, in particular the hypothesis that Sumner [co-defendant] transferred the appellant's DNA to the didgeridoo on Sumner's first visit to the house on the day in question, were not unreasonable and the prosecution had not successfully excluded them. As the evidence was not capable of supporting the appellant's conviction for either offence, no question of an order for a new trial arose.”

Other cases on expert evidence

See also: Dasreef Pty Ltd v Hawchar [2011] HCA 21 (22 June 2011), discussed here on 22 August 2011, and on various types of expert evidence: fingerprints here on 27 January 2006, drug couriers here on 25 February 2014, hypnosis here on 2 February 2007, blood alcohol here on 18 April 2008, sexual assault injuries here on 18 June 2010, the dangers of manipulation of expert evidence here on 26 June 2009, sounds on a recording here on 12 June 2009, unfair criticism of an expert by counsel here on 28 May 2009, and (a brief mention, see the case for details) qualifying as an expert here on 23 January 2006.

Wednesday, August 13, 2014

Photos from abroad

In taking photographs of a girl performing oral sex on another man, the defendant did an act “with or on” the girl so as to commit an offence as a principal offender under s 132(3) of the Crimes Act 1961 [NZ] (following Y (SC40/2013) v R [2014] NZSC 34 (3 April 2014), noted briefly here on 4 April 2014). He was therefore liable in New Zealand, even though this all occurred overseas, because of the extension of jurisdiction by s 144A of that Act.

That is how the Court reasoned in LM v R [2014] NZSC 110 (13 August 2014).

In addition to accepting that liability could arise in that way, two of the five judges found another path. They considered that the defendant was liable as a secondary party under s 66(1)(b)-(d) of the Act, regardless of where the principal offender committed the offence, and that therefore, since he committed the acts of secondary participation overseas, Mr LM came within the s 144A extended jurisdiction. On this approach (as I understand it) one pretends, when asking whether the defendant was aiding or abetting, inciting counselling or procuring the commission of an offence, that the offence was committed in New Zealand by the principal, just for the sake of establishing that what was done was an offence under New Zealand law.

William Young J for the majority on this aspect of the case (with Elias CJ and McGrath J) was “unable” to construe ss 144A and s 66(1) as imposing liability on someone who is a secondary party to a substantive “offence” which occurred overseas but which, because the principal party was a foreigner, was not an offence recognised in terms of s 144A [24]. In this case the other man’s acts done overseas were not an offence under New Zealand law because he was not a New Zealand citizen and was not ordinarily resident in New Zealand.

Putting aside the “with or on” aspect of the actus reus, which enabled the court to find that Mr LM could be liable as a principal party, more generally there are three permutations of circumstances illustrating potential liability. If both men had been New Zealanders, both would have been liable in New Zealand, one as a principal and one as a secondary party [17]. If the principal party had been a New Zealander, but not the other, the principal would be prosecutable in New Zealand, but as the legislation does not specifically impose liability on secondary parties who are New Zealanders, a foreign secondary party’s liability would be less certain, and the better view [20] is that only New Zealanders could be liable as secondary parties under s 144A (the minority seem to agree: [40, footnote 26]). And if the principal was a foreigner and the so-called secondary party a New Zealander, the New Zealander would not be liable as there was no offence, over which there was jurisdiction in New Zealand law, to which he could have been a secondary party [23]-[24] (disagreeing with the minority).

The majority suggest that s 144A should be revised to specifically refer to secondary liability, as in the corresponding provisions in Australia and the United Kingdom, to meet the concerns raised by this analysis of the permutations [25]. There seems, as the minority point out [40], no logical reason for the policy to be to exclude the secondary liability of a New Zealander merely because the principal party happens to be a foreigner.

In summarising this case I have used the terms “principal party” and “secondary party” in the sense that is consistent with s 66(1): a principal party actually commits the offence and is liable under s 66(1)(a), and a secondary party does any of the acts described in s 66(1)(b)-(d). The meaning in the judgments is the same but the tendency is to use the expression “party liability” to mean “secondary party liability”, see, for example [23], and so to distinguish a principal from a party (meaning a principal party from a secondary party) as in [19].

Monday, August 04, 2014

The embarrassing c.v.

This month marks the 10th anniversary of the start of this site. I try to be modest but the facts make that unconvincing. My normally fetching diffident modesty could easily be misinterpreted as obnoxious condescending pomposity. Not that many of my colleagues will notice, as they seem to regard the internet as almost exclusively an outlet for their obsessive prostate therapy.

And here, for the less onanistic, we have an exciting new decision from the Supreme Court of Canada: R v Hart, 2014 SCC 52 (31 July 2014).

At least it is exciting for lawyers who encounter occasions of police operations that in Canada are called “Mr Big” operations. Broadly, and the details are in the decision, this involves undercover police officers masquerading as members of a gang that the defendant wants to join. They require the defendant to detail the offences that form a sort of curriculum vitae in the application for membership, and those details are later offered to the court as evidence on relevant charges. When are such confessions admissible?

The majority, in a joint judgment delivered by Moldaver J (Cromwell J separately concurring on the test for admissibility, Karakatsanis J dissenting on that) laid down a “new” test. I explain the quote marks in a moment. The test has two prongs, as the Court called them. First, there is a balancing of probative value against prejudicial effect, and secondly there is consideration of abuse of process. The prongs need not be considered in that order, because if there would be an abuse of process in admitting the evidence then it is excluded, or in an extreme case the prosecution is stayed, without the need for consideration of the first prong [89].

My quote marks are because these prongs are not new. The test is only new in the sense that evidence obtained in the context of one of these operations is presumptively inadmissible on the first prong [10], [85]. The prosecutor must prove that, on the balance of probabilities, the probative value of the evidence outweighs the prejudicial effect of admitting it [89]. But the defendant still has the traditional burden if reliance is placed on abuse of process as the ground for exclusion [11].

While this is a special rule for special facts, the joint judgment includes explanations of the balancing exercise and the abuse of process decision that could have general application.

If one wants to find disappointment, one should contemplate the missed opportunity to sort out the relationship between the probative value/prejudicial effect balancing and trial fairness. As is generally accepted – and obscurely referred to at [109] - a problem with describing the probative value/prejudicial effect decision as a balancing exercise is that it suggests that high probative value can only be outweighed by a high level of risk of prejudicial effect, and this in turn suggests a high tolerance of risk of trial unfairness. The only solution offered here is “trial judges will have to lean on their judicial experience” in difficult cases [109].

A better requirement would have been that the first prong would focus on trial fairness: would admission of the evidence create a real risk that the trial would be unfair because it would endanger the impartial determination of the facts. There would be no balancing, just an assessment of this risk. “Impartial” here is used in the sense that it emerges from trial fairness jurisprudence.

Still, Canadians must work with the prongs as established in this case. There are valuable comments at [95] – [105] on how probative value should be assessed, addressing both the circumstances in which the confession was made, and the credibility of the confession itself. And prejudicial effect is addressed at [106] – [107].

As far as abuse of process – the second prong – goes, the joint judgment acknowledges that this has not hitherto provided an effective remedy in this context, and recognises that it has to be “reinvigorated” [114], mainly through enhanced judicial sensitivity to the risk that the circumstances of a given case may amount to coercion [114] – [118].

One should ask whether the new test for evidence obtained in the context of Mr Big operations provides adequate protection against self-incrimination, which was the basis for exclusion that Karakatsanis J would have preferred [170], [180]-[185]. Moldaver J’s reasons for disagreeing are at [124]-[125], essentially they are that the way the principle against self-incrimination would provide a remedy here would have to be worked out, adapting rules of evidence and prevention of abuse of process (illustrations of similar workings out are the confessions rule and the right to silence), and the two-pronged approach does that.

Karakatsanis J’s reason for dissent was that the two-pronged rule does not adequately take into account broader concerns like human dignity, personal autonomy, and the administration of justice [167]. The focus should be on three “vital concerns”: the reliability of the evidence, the autonomy of suspects, and the potential for abuse of state power [168]. There is established case law on the principle against self-incrimination and there is no need for a new rule [181].

Well, as I indicated above, the new bit is really the presumption that evidence obtained in a Mr Big context will be more prejudicial than probative and that it should be excluded for that reason alone. Whether that shift of burden to the prosecutor is really significant should be assesed by reference to how difficult it was for the defendant to put the question of admissibility in issue. Usually, as judges tend to say, it is only in marginal cases that it matters who has the burden of persuasion (see for example Lord Brown in O v Crown Court at Harrow [2006] UKHL 42 at [35], noted here on 31 July 2006). Trial judges have, however, been given a nudge towards being very careful to avoid improper prejudice, and to be especially sensitive as to abuse of process, in cases where this new rule applies.

Cromwell J would have left application of the new rule to the trial court in the event that the prosecutor decided to pursue a new trial, whereas the majority ruled the evidence inadmissible in this case [152] - [163].

The case also illustrates another point: the trial judge should have allowed the defendant to give evidence in the absence of the public (who could have been accommodated in another courtroom to view the proceedings by closed-circuit TV), because of the special vulnerability of this defendant [42], [48], [51]-[55].

Monday, July 21, 2014

When is fairness intolerable?

The right to legal advice was central to R v Taylor, 2014 SCC 50 (18 July 2014). The defendant had requested legal advice but this request was not acted on during the collection of a blood sample which was of central importance on an alcohol-related driving offence. On conducting the balancing exercise in accordance with R v Grant (discussed here on 19 July 2009) the Court held that the improperly obtained evidence was inadmissible.

The right to legal advice is closely associated with the right to silence. A motive for refusal of access to a lawyer, it may be reasonably be conjectured, could be that the police do not want the suspect to exercise the right to silence. And the right to silence is of fundamental importance: it is a corollary of the obligation on the prosecutor to prove the charge, and of the need to do so without the assistance of the defendant.

The right to legal advice has previously been the subject of commentary here: see Cadder v HM Advocate [2010] UKSC 43 (here on 27 October 2010), R v Sinclair [2010] SCC 35 (here on 15 October 2010), Salduz v Turkey [2008] ECtHR 1542 (here on 3 December 2008).

The central issue highlighted in those comments is whether a breach of the right to legal advice raises fair trial issues, in which case a balancing exercise is not appropriate (but an exclusionary rule is), or whether it raises issues of public policy, where balancing of competing interests is appropriate.

It is easy to forget history and to say that the issue of the admissibility of evidence obtained in breach of the right to legal advice is an issue of policy. The Birmingham Six abuses of police power, and the manufacture of false confessions, came as a shock to British justice, and drove home the importance of the right to silence and the vulnerability of people who are held in custody.

In his wonderful article in the London Review of Books, vol 15 no 21 (4 November 1993) “A sewer runs through it”, Alastair Logan (a solicitor whose clients included the Guildford Four and the Maguire Seven) noted that research presented to the Royal Commission on Criminal Justice

“shows that 42 percent of those who were arrested and detained in police stations during the currency of the Commission were educationally subnormal or of borderline intelligence; another 7 percent were suffering from defined mental illnesses. The average IQ of detained persons was 82. One third were intellectually impaired, and 35 percent were not in a normal mental state due to extreme distress, mental disorder or drugs. Twenty percent were suffering from an unusually high level of anxiety and distress. About 20 percent required the assistance of an adult to safeguard them and their rights, though the police identified only 4 percent of that number as requiring such protection. The police commonly failed to recognise that detained persons suffering from depression were vulnerable. There is no systematic training available to police officers to enable them to identify vulnerable suspects or mental disorder. The removal of the right to silence attacks the vulnerable and the disorientated, who massively outnumber the terrorists and the professional criminals, in or out of uniform.”

It is very easy, in cases like Taylor where the alleged offending is not of the most serious kind, for courts to conclude that a balancing exercise favours exclusion of evidence obtained in breach of the right to legal advice. Indeed, in Taylor the Court was not concerned to explore what advice the defendant could have been given [41], or in what way absence of legal advice may have prejudiced his defence. For more serious allegations the balancing exercise may well include such considerations. Yet it is when being held in custody on more serious allegations that a person will be most in need of the protection of the right to legal advice, and a court would have to consider whether use of evidence improperly obtained from a vulnerable defendant really promotes respect for the law.

Support for determination of admissibility of improperly obtained evidence by a balancing of policy interests is found in the seemingly ever-present fear that those defendants who face the most serious charges will have to have fair trials.