Gordon Campbell on the Supreme Court review of the Peter Ellis case

ellis-imageThere is a sense of inevitability about the Supreme Court being asked to review the Peter Ellis case.

Last October the Criminal Case Review Commission Bill passed its first reading. Eventually this legislation will set up a formal mechanism for reviewing convictions (in those cases it chooses to accept) after all the usual appeal channels have been exhausted. Importantly, this Commission will not make a finding of guilt or innocence. Where appropriate, it will merely send the proceedings back to the courts for them to take another look at the case, in order to either confirm the original conviction or overturn it.

In all likelihood, the Ellis case would have been among those first in line for this review process – and there are reasons pro and con, as to whether this case would have been the best way of launching the work of such a Commission. Last week the courts saved the Commission the trouble, and the case has now been sent back to the Supreme Court for reconsideration on the grounds as to whether a miscarriage of justice occurred, chiefly with regard to how a lot of the crucial evidence was obtained, and interpreted.

What’s At Stake

It is not only (a) Ellis and his supporters and (b) the victims and their families who have a huge stake in the outcome. The legal process which delivered the original convictions – and which confirmed them right on through the appeal system, and via a subsequent official inquiry – also has its credibility on the line. Either way, the outcome will have repercussions for how juries will choose to regard the credibility of child witnesses in paedophilia cases in future. The outcome will also affect how the legal system treats the testimony of “expert” witnesses – who serve as competing guns for hire for both the prosecution and the defence, with the result that juries (and the judges as well) can often be left none the wiser as to which brand of the expert testimony they should be regarding as being more credible.

Luckily or otherwise, one of the justices central to the Ellis review will be Supreme Court judge Susan Glazebrook, who has written and commented extensively in recent years on the problems the courts face in managing the testimonies of competing expert witnesses. This 2018 article in the Victoria University Law Review Journal( VULRJ) reflects Glazebrook’s latest thinking on the subject. Among other things, she points to the dangers of ‘confirmation bias’ – whereby experts see what they expect to see and find what they have been hired to find, in situations where they are routinely schooled as to what is being expected of them by the prosecution or defence teams that hired them. A certain ‘team spirit’ can readily come to prevail. Being hired for similar work in future may also hinge on how successful the expert in question has proved to be in convincing judges and juries of the rightness of his or her opinion. All these factors can motivate the experts to exaggerate the certainty of their findings. As Glazebrook says:

There will…be “adversarial bias” arising from an expert’s involvement in the adversarial system. Some commentators have said that it is almost an “inevitable” consequence of the appointment of experts by partisans. This adversarial bias may be present due to a number of factors and lead to an expert moulding his or her evidence to fit appointing counsel’s case. This may be an unconscious bias through being part of a “team” and only getting one side of the story. In addition to being part of a side, cross-examination could lead an expert to become defensive and to take more extreme positions than they would otherwise adopt. Another more general bias, known as “forensic confirmation bias”, refers to the “class of effects through which an individual’s pre-existing beliefs, expectations, motives, and situational context influence the collection, perception, and interpretation of evidence during the course of a criminal case”. All involved in the justice system need to remain aware of this.

Indeed they do, starting at the top. Clearly then, the problems of differing expert approaches and interpretations and the varying types of evidentiary “contamination” are not unique to the Ellis case. Just as clearly, Glazebrook will be a central figure in how the Ellis appeal unfolds. According to last week’s Supreme Court press release, the arrangements for the appeal (eg its scope, likely parameters of evidence and timetable) will be worked out shortly in a telephone conference between Glazebrook and the counsel for both sides.

Given her central role, it may be of concern to the Crown prosecution team and to the families of the creche victims that Glazebrook began her recent VULRJ article by citing a series of infamous overseas examples where expert testimony resulted in serious miscarriages of justice. Conversely, Glazebrook’s VULRJ article may well hearten Ellis’ supporters, who have long seen his case in that same light, and who have repeatedly criticised the expert testimony (notably by Karen Zelas) relied on by the prosecution at the original trial. In the past, Ellis supporters have also criticised the expert advisory role played by Justice Ministry official Val Sim in the Eichelbaum inquiry into the Ellis case. Experts have readily been seen as the heroes or villains of the partisans ranged on either side.

Last week, the Supreme Court indicated that the affidavits filed on Ellis’ behalf by experts Dr. Thelma Patterson and Professor Harlene Hayne had been central to its decision to grant the appeal. Hayne’s research team, the Supreme Court noted, recently concluded their work in applying what Hayne claims to be an “empirical methodology” to the original interviews with the children, and with respect to the potential for contamination of their evidence. At para 17 of their ruling, the Court says:

In our view the affidavits of Professor Hayne and Dr Patterson raise issues of general and public importance and significant issues specific to Mr Ellis’ case. The interest of justice requires that these issues be ventilated on appeal, despite the length of time since the second Court of Appeal decision.

As yet, the shape of the likely Crown response to the arguments being advanced by Hayne and Patterson remains an unknown. The original trial had also involved competing expert testimony – by Karen Zelas for the Crown, by Keith Le Page for Ellis – and there is every likelihood that competing experts will again be summoned to the witness stand for the new appeal. While claims of ‘Satanic panic’ have been a prominent theme among the Ellis supporters in casting doubt on how the testimony of the child victims was gathered and interpreted, there has been a parallel development in the opposite direction over the past 20 years with respect to the paedophilia accusations involving the Catholic Church.

In that realm, child testimony and recovered memory have been increasingly relied upon by juries as being valid, in their core elements. Presumably, the Supreme Court will be seeking to avail itself of expert advice on the evidentiary advances being made on that front, too.

One can only hope so. In the recent conviction of Archbishop George Pell in Australia for instance, the jury reached a unanimous verdict – now under appeal – that relied entirely upon the sole testimony of the victims, despite the lack of corroboration and despite the obvious potential for evidential contamination, given the time elapsed and given the publicity given to similar cases of priestly abuse.

The Lundy Precedent

Obviously, it would be foolish to try to second guess how the Supreme Court might proceed in the Ellis appeal. All the same, the Court of Appeal ruling last October (by a judicial bench that included the new Supreme Court chief justice Helen Winkelman) on an appeal brought by the double murderer Mark Lundy provides some insights into the thinking of our higher courts in situations where doubt has been cast on the expert testimony originally relied on to obtain a conviction. In the Lundy appeal last year, the Court of Appeal decided the fresh doubts had substance, but were insufficient to justify overturning the conviction.

One relevant point to the Ellis case being… obviously, the Court of Appeal didn’t divorce its re-consideration of the expert testimony from the wider body of evidence that had been presented at trial. While the Supreme Court may not want to re-litigate the entire Ellis case, it is very hard to see how it can legitimately fence off particular aspects of the case (eg the Zelas testimony, the Police interview techniques) and conclude that any errors detected were sufficient to outweigh the evidence it didn’t/couldn’t reconsider, but which went before the original jury. (Second-guessing a jury 26 years after the fact would seem to be a perilous enterprise.)

Here’s how the Court of Appeal dealt in general with the issue of what happens when the experts contradict each other on the witness stand:

…The robustness of a methodology cannot legitimately be established by an inexpert judge and jury. The essential work of validation most occur before the courtroom is entered…a track record by a body of scientific opinion… needs to have been established.”

That’s the ideal, but good luck with that. Even on the rare occasions when the science speaks with relative unanimity, “the body of scientific opinion” routinely changes its position over time. Briefly, what was at issue in the Lundy appeal last year was the new empirical methodology used at trial with regard to messenger RNA – the relevant test was called “ brain plex” – that had claimed to be able to establish with certainty the nature and origins of residues found on Lundy’s shirt. In its 2018 ruling, the Court of Appeal decided that this expert evidence should have been inadmissible, partly because no scientific consensus existed on its validity, and partly because the jury wouldn’t have been able to understand the technical claims and counter claims being made about its reliability, anyway.

Interestingly, if the Lundy legal team hadn’t been able to find and fund an expert witness of its own, the “brain plex” argument would have gone out before the jury, uncontested. Even more to the point though, the Court of Appeal ruled that the Lundy conviction would still stand, regardless – even though the technical test now ruled to be inadmissible had (at the original trial) been the platform from which the prosecution was able to make the devastating argument to the jury that Lundy “had his wife’s brains on his shirt”. (Juries can understand and do internalise statements like that.)

Whenever the new Ellis appeal hearing occurs, one can safely predict that – just as at the original trial – the experts will disagree as to (a) the validity of the children’s testimonies and the expert interpretation placed on them (b) whether the risks of evidence contamination was competently managed and (c) whether any shortcomings were of such magnitude as to overturn the conviction. This year, part of the Ellis defence team’s argument for the right to appeal was that “the evidential interviews fell far short of best practice (even at the time) and there was a strong possibility of contamination of the evidence.”

Not only will the Supreme Court have to rule on these matters but – as the higher courts did with Lundy – it will also have to decide what weight to give to any shortcomings that it detects. It promises to be a very tricky exercise. With the passage of time, the wisdom of hindsight and given the subsequent improvements in evidence-gathering procedures, many prior convictions in New Zealand courts might look less robust today than they once did. (In the wake of the David Tamihere case, we all know a lot more now about the perils of relying on evidence solicited from jailhouse snitches.)

As mentioned, the Court of Appeal concluded that the potential for error in the Lundy case did not reach the threshold required to overturn the conviction. As for the Ellis case… to be kind, it seems clear that interviewing methods have not been a pressing area of reform for the authorities in this country. As the Supreme Court noted when granting the Ellis appeal, the first comprehensive study of present day interview standards in New Zealand was not published until 2016.

Panics past, and present

Routinely, the Ellis case has been linked to the ‘Satanic panic’ outbreaks that occurred at daycare centres in the US during the 1980s, such as in the McMartin Preschool case and the Kelly Michaels case, which resulted in the then 23 year old Michaels spending the next seven years in prison before all charges against her were dismissed. Certainly, both these cases included allegations of the same bizarre elements and fantastical behaviours that cropped up during the child testimonies in the Ellis case. Despite those parallels, the contention that the Ellis case was purely a by-product of these overseas cases and the related expert-driven fixations – with no assistance from the bantering, fantastical culture that was reportedly a commonplace at the creche – is something that the Supreme Court will now have to assess afresh.

In the meantime, the possibility that a Satanic panic led to the Ellis prosecution and conviction is at risk of being superseded by an equally unquestioning assumption of his innocence, which has been fuelled – in part at least – by a conspiratorial view of the allegedly collusional failings of the justice system. Whatever the outcome of the Ellis case, it is crucial that the treatment of paedophilia (and rape) accusations by the Police and by the courts is actually improved, and not set back into the previous dark ages of scepticism and denial. As the US academic Lynn Henderson pointed out some years ago:

The McMartin and Michaels cases preserved the “innocence” of children, in that neither case really centred on claims that children were purposely lying; rather, the problems were with the intimidation, coaching, and manipulation of child witnesses by adult authority figures. Nevertheless, suspicions about the credibility of child witnesses in other sexual abuse cases increased dramatically as a result of these and other legal disasters.

Popular culture and the media moved from shocked accounts of horrible abuse and – as is typical in media coverage of criminal cases – rapid conclusions that those accused were guilty, to outraged accounts of innocent people railroaded by false accusations, overzealous prosecutors, and over-reactive adults. As a result, any progress made toward understanding child sexual abuse appears to have been stalled, if not reversed, and attempts to prove its existence have become extremely difficult again.

Henderson was writing before the revelations about paedophilia within the Catholic Church gave fresh impetus and renewed credence to the testimony of children, and to the recovered memories of adults who experienced abuse as children. Currently, we have a Royal Commission underway into the abuse of children in state care. There has to be a reason why paedophilia and child abuse accusations have become associated with institutions like state care, religious bodies, and creches. Perhaps it is because in such contexts, the children’s care is being exercised in conditions where their powerlessness is extreme, and where their voices have routinely not been heard, or believed.

One can only hope that Glazebrook and her colleagues on the Supreme Court will approach their task with an open mind as to whether the Ellis case belongs with those other instances of expert-driven injustice that Glazebrook cited as cautionary tales in her VULRJ article.

Footnote One: For a thorough, balanced account of the problems associated with the legal system’s over-reliance on expert witnesses, see the article by Thomas Kearney (“The Unresolved Problem of Expert Evidence”) in the Australian Law Journal (2018) 92 ALJ 127. Rather than abandon expertise to the vagaries ( and to the varying-sized cheque books) of the adversarial system, Kearney discusses (among other things) the argument that perhaps the courts themselves should hire and manage the relevant experts, given that judges and juries rely so heavily on them for guidance.

For a long discussion about why the Criminal Case Review Commission is a really good idea, here’s a Werewolf interview I did with the guy who headed the UK version of it, including during one of the infamous cases referred to by Glazebrook.

Footnote Two: In rape and paedophilia cases, notions like ‘corroboration’ and ‘contamination’ are often relative terms, not absolutes. Rape, for instance, is commonly committed without corroborating witnesses being present. Additionally in paedophilia cases, some degree of contamination of evidence can be inevitable. However ideal it may be that parents should not ‘contaminate’ the testimony of the child victims, it would be a rare parent indeed who – when it is possible that their child has been molested – refrained from questioning their child any further, lest that should imperil a prosecution that might eventuate, at some future date.

Similarly, while having each child interviewed only once may be the ideal, this is far easier to achieve when there are only one or two complainants. In the Ellis case, there were at least 13 original complainants, with convictions involving seven of them being eventually confirmed. In such circumstances it is very hard to imagine how the “one interview per child” ideal could have been met by Police, given that they were faced with cross-checking and re-checking a large and mounting number of allegations.

Researchers have long been aware of the fraught dynamics in the multi-victim contexts common to daycare centre child abuse allegations. On the other hand, the same body of research also indicates that pre-schoolers are more susceptible than school-age children to suggestive questioning that offers few open-ended questions, and that relies heavily instead upon closed-ended formulations consisting of yes/no questions and forced choices. Presumably, the Hayne research will shed more light on the structure of the evidential interviews that were conducted in the Ellis case.

Obviously, the inherent challenges involved in multi-victim abuse cases are not an argument for anyone – parents, investigators or experts – being allowed to wilfully compound the contamination. Yet in the Ellis case and other paedophilia cases, it is worth keeping in mind that a risk of some level of concerned parental “contamination” is inevitable, and may even be desirable for the ultimate wellbeing of the children involved. With both rape and paedophilia accusations, the inherent problems of corroboration and contamination should not automatically be treated as grounds for exoneration. By the same token…in the decades since the early 1990s, there is also no doubt that investigators have learned a lot more about the need to tread with particular care when interviewing multiple numbers of pre-schoolers, and interpreting the results.

Somehow, the Supreme Court is going to have to devise some kind of (non-subjective) rule for deciding how much contamination is to be regarded as too much contamination – and how much suggestive questioning is to be regarded as too much, in the light of current knowledge. Like earlier efforts, it is likely to be an imperfect exercise, carried out in good faith.

Media, WTF?

The reaction clip by Democratic presidential candidate (and El Paso resident) Beto O’Rourke to a typically braindead media query about what the President can do to combat racism is the sort of thing that should be run forever in Journalism Class 101:

Among other things, O’Rourke is telling the media not to wake up each morning as if its brain has been erased of all previous knowledge. We’re not living in a Memento-like movie reality, even though it can feel like it on occasions like this. As Nate Silver has indicated, even the New York Times coverage of Trump’s post El Paso/post Dayton responses has been credulously and cretinously devoid of context:

There’s a good summary here in the Columbia Journalism Review of the media’s inability to contextualise Trump’s response to the shootings, including a reference to the fury of NYT staffers at their own paper’s ineptitude.

Lana Del Rey, El Paso, Dayton

Talking of such things, Lana Del Rey wrote a song over the weekend about her response to the mass shootings in the US. Her engineer Laura caught her on video, working at it. Del Rey posted the unfinished, unpolished result immediately on her Instagram account. You can access it here.

The song is called “Looking For America” …There’s a line in it about how we used to worry about the children only in the hours after dark. She adds: “I’m still looking for my own version of America, one without the gun… used to go to drive-ins and listen to the blues, so many things that I think twice about before I do.”

True enough. Americans now have to think about the possible consequences before doing utterly mundane things like going to a supermarket in Texas, or a nightclub in Ohio. Like a lot of Americans, Del Rey finds herself looking for a better form of America, in the knowledge that the country she once knew no longer exists, even for the most privileged.