Eyewitness identification was crucial in the infamous New Zealand missing-persons-turned-murder case – that of Olivia Hope and Ben Smart in the Marlborough Sounds in the first days of 1998.
Drifter Scott Watson was convicted of their murders and sentenced to life imprisonment with a minimum non-parole period of 17 years. Watson, who has always claimed innocence, appealed against his conviction in 2000 and again last year. The Court of Appeal is expected to deliver its verdict on the second appeal very soon. In the meantime, Watson, whose latest parole hearing was cut short in March and is yet to be resumed, waits in jail.
During the 11-week trial in 1999, the jury heard from 480 witnesses, including two prisoners who claimed Watson had confessed to the murders while on remand in Christchurch’s Addington Prison. Subsequently, one said he’d thought police would reward him and had pressured him to lie; the other admitted he had never been in the same cell as Watson. Overall, the prosecution case relied strongly on eyewitness and circumstantial evidence – the physical evidence was weak.
The defence argued Watson’s boat did not resemble a murder scene when it was examined. It did not even resemble the boat where the young friends, Smart, 21, and Hope, 17, were last seen.
Hope and Smart had been celebrating New Year’s Eve at Furneaux Lodge, in Endeavour Inlet, along with at least 1500 other people. Around 4am, they boarded a water taxi and headed off to sleep on a boat that Hope, her elder sister, Amelia, and some friends had chartered. When they arrived, Hope discovered all the bunks were taken.
On board the water taxi was another passenger who came to be known during the trial as the “mystery man”. He offered them beds on his boat. Olivia thanked him, and they decided to accept the offer. The taxi then headed to the mystery man’s boat. Once on board his two-masted ketch, Hope and Smart were never seen again.
Police questioned partygoers at Furneaux Lodge but no one knew the mystery ketch-owning water taxi passenger. He had been seen at the lodge, and bar manager Roz McNeilly worked with police to produce an identikit picture of him. In it, he had stubble and long hair. The water taxi driver, Guy Wallace, told police the man who boarded the ketch with Hope and Smart had wavy brown hair and two days’ facial hair growth and wore a green Levi’s shirt. Another passenger on the water taxi, Hayden Morrissey, described the man as having straggly, shoulder length hair.
Scott Watson, 26, had also been seen at the lodge. A photo taken that evening shows Watson had very short hair, was dressed in a blue shirt and was clean-shaven.
Watson was known to a number of party guests. He had built his own steel-hulled yacht, the Blade, in his parents’ backyard in Picton, and had sailed it to the inlet, attaching the single-masted sloop to that of some friends with whom he had shared a few drinks.
Watson had left school early, didn’t have a regular job, and lived on his boat. He had a criminal record, most of it from his teens. It included burglary, theft and possession of cannabis. It did not include any sexual offences or major crimes.
By his own admission, he was drunk at the party – he was known to be obnoxious when drunk – and partygoers recalled he’d unsuccessfully propositioned young women.
Police soon decided he was the mystery man and had most likely murdered Hope and Smart. The resulting prosecution hinged on eyewitness evidence.
Police spokesman Matthew Gittos, from National Forensic Services, refused to comment on procedures, nor on whether methods used in gathering eyewitness evidence have changed in the quarter century since Watson was charged. Gittos referred to the 2006 Evidence Act, which prescribes a few basic rules: the person carrying out the procedure should not indicate who the suspect is, and “the eyewitness should be informed the suspect may not be among the persons in the procedure”.
The act does allow for exemptions from this procedure. They include “a substantial change in appearance of the suspect after the alleged offence occurred”. It was established, however, that Watson did not have long hair before he came to Furneaux Lodge that New Year’s Eve.
Best practice
Assessing the reliability of eyewitness evidence is a specialist area of research for psychologists. Gary Wells is one of the most prominent in the field. In 2020, the professor emeritus at Iowa State University and other colleagues updated best practice for eyewitness evidence in a scientific review for the American Psychological Society. Wells is currently advising New Zealand’s Criminal Cases Review Commission for its report on eyewitness ID procedures.
Wells told the Listener the first thing that should be done is to obtain a written description of the person the eyewitness saw as soon as possible. There should be no showing of pictures of the suspect nor any preliminary unrecorded interviews: everything should be on the record.
Investigators, he says, should also “instruct the witnesses to not discuss the event with other co-witnesses. The entire interview should be video recorded.”
Wells insists neither the interviewer nor the eyewitness should know which of the mugshots shown to them, if any, is the suspect, because the way an identification procedure is conducted will influence the reliability of the evidence.
None of the interviews in the days following Hope and Smart’s disappearance met this requirement.
Watson’s photo had been released to the media and was shown by police to the eyewitnesses.
The prosecution’s expert witness for last year’s appeal, Margaret Bull Kovera, also thought the identification evidence was unreliable. Yet the crown opposed having her report read out to the Court of Appeal. Kovera is professor of psychology at John Jay College of Criminal Justice, City University of New York.
A report prepared by the Independent Police Conduct Authority in May 2010 after a complaint by journalist Keith Hunter and Watson’s father, Chris, concluded police procedures “fell short of best practice” and were “highly undesirable”. Hunter has long believed Watson was wrongly convicted, a belief shared by Chris Watson. Olivia Hope’s father, Gerald, also expressed concern about the trial, telling journalist Mike White he felt “very uncomfortable about the way the crown ran the case”.
At first sight
Wells is also firm about the use of the witness’s “statement of confidence” – that is, how sure the witness is at the first identification, which should be the only identification. Wells insists a later statement of confidence is less reliable and should not be used in court because it can be subject to manipulation and contamination.
At Watson’s trial, water taxi skipper Guy Wallace said he was “pretty definite” Watson was the man on his boat who was last seen with Hope and Smart. But in the transcript of his initial police interview, Wallace did not ID Watson. With best practice, the court should have heard only what Wallace said at that time he first made a formal identification in an interview with Detective Sergeant Tom Fitzgerald.
Fitzgerald: “The mystery man you described to me … he doesn’t exist either. Perhaps you haven’t been telling the truth Guy. I know what happened.”
Wallace: “Well … do you wanna know what I think? I think you guys can’t find them, you’re looking for a reason so you’re gonna put it on me.”
Fitzgerald: “Tell me the truth, I don’t want to have to talk to you again.”
Wallace: “You’re just trying to put words in my mouth, it might be police procedure mate but it sucks.”
After the trial, Wallace retracted his identification of Watson and maintained for the rest of his life he had been pressured into making a false statement. In 2003, Wallace recorded an interview with Keith Hunter and said Watson was the wrong man. Wallace committed suicide in 2021.
By 2012, there were indications of a change in intervewing practices. The police staff magazine, Ten One, reported that senior officer-turned-academic Nina Westera was working with expert psychologist Matthew Gerrie on a new training programme.
“If you go in looking for a confession or admission, you tend to become focused on that outcome,” the article ran. “The model we’re teaching – called conversation management – is designed specifically to eliminate these issues. We want to get as much information as possible so we can either eliminate them from our inquiry or tie them into their story and then challenge them appropriately.”
Later on, an interviewing technique developed by Fitzgerald came in for criticism. In 2022, a judge faulted the police interviews in the case of Lois Tolley who was shot in 2016. The detectives involved were using a model developed by Fitzgerald. Justice Simon France said the interviews “weren’t a pursuit of a ‘neutral truth’” but “a sustained pursuit of a particular ‘truth’”. The model was later dropped.
Led by mugshots
Another golden rule from the psychologists is that mugshots should always match descriptions from eyewitnesses. Seven photo IDs were shown to witnesses in the Watson case: all looked like Watson to varying degrees. None resembled the long-haired, unshaven man described by Wallace and water taxi passenger Hayden Morrissey.
Dr Adele Quigley-McBride, assistant professor in psychology at Simon Fraser University in Canada, was an expert witness for the 2024 appeal, along with Gary Wells. In an email to the Listener, she listed a number of concerns. One was that the police failed to report non-identification of Watson by eyewitnesses.
“This is important exonerating evidence. Prosecutions must share exonerating evidence with the defence.”
A second concern was that Morrissey did not identify Watson, and a third that Wallace’s ID of Watson showed signs of inaccuracy. “Slow, not very confident, he wavered between two photos. This is particularly strange given his memory had been so contaminated – he should have been quick and sure given how much suggestion he had been exposed to.”
Then there’s the boat. Psychology professor and boatie Barry Kirkwood and private investigator Mike Kalaugher were among those who considered the mystery ketch was further away than Wallace had thought. Fitzgerald, in his interview with Wallace, thinks his description is wrong, too, but instead of considering Wallace may have been confused between the many trips he took that evening, goes a step further to imply Wallace did not see the ketch, and that he was lying.
Whoever the “mystery man” was, by January 2 he and his boat were no longer in the Sounds. Initially treated as a missing persons case, by January 7 the disappearance of Ben Smart and Olivia Hope had become a potential murder inquiry.
Watson has said police found him an easy target as he had arrived and left the lodge party alone.
The police investigation was led by Inspector Rob Pope, who said he felt Watson had “the right sort of agenda and pedigree” and started to “stick out like dog’s balls”.
Keith Hunter examined the police investigation in his 2003 documentary, Murder on The Blade?, and a book, Trial by Trickery. He concluded the case against Watson was not only unconvincing, but inexplicable.
Unreliable witnesses
Eyewitnesses have a poor track record at identifying real culprits. Wells and his colleagues surveyed the relevant scientific literature in 1998. In examining 40 cases where the accused had been exonerated in court, they found all had involved mistaken eyewitness identifications.
“Eyewitness identification evidence is among the least reliable forms of evidence yet persuasive to juries,” they wrote in their findings published in the US Journal of Applied Psychology. “Judges, like attorneys, are not particularly adept at evaluating eyewitness identification procedures.”
In 2001, a year before Watson’s unsuccessful appeal, some real police mugshot montages were tested to see how reliable they were. As part of her PhD thesis, psychology student Kelly Fitzmaurice had been allowed to test 12 photographic montages of mugshots used in real cases by Wellington police. She also had the original eyewitness written descriptions of the alleged offender.
Fitzmaurice asked her participants to identify the suspect in each montage from only the written eyewitness description. In one montage, all but one of the participants picked the suspect. If the suspect in this real case had been convicted, then his identification from this montage could not be relied upon because the identified person was the only one closely resembling the description. Fitzmaurice also referenced studies showing witnesses confidently identifying an innocent person because that person most closely resembled the description.
Review of procedures
The Criminal Cases Review Commission Te Kāhui Tātari Ture, launched in 2020, has already sent two cases back to appellate courts due to concerns over visual identification evidence.
It is now conducting an inquiry into “factors which could impact on the reliability of eyewitness identification evidence such as cross-racial identification, pre- and post-event influence, social media contamination and confidence”. It will also consider whether police practice relating to eyewitness identification evidence are in line with international standards and best practice. This may lead to changes to the 2006 Evidence Act and lead to fewer wrongful convictions.
In a 1999 conviction, Stephen Stone went to prison for the 1989 double murder of Deane Fuller-Sandys and Leah Stephens. The convictions were quashed last October and he was acquitted in April after the crown announced it would not re-try the case. Like Watson, Stone had spent 25 years in jail. Private investigator and former police officer Tim McKinnel has said there were many things seriously wrong with the investigations. It was “best described in my view as a corrupt investigation, and the fact it’s taken this long to remedy is a tragedy”.
McKinnel, who has also followed Watson’s case, for one believes the Evidence Act is overdue for revision. “I suspect that a fair proportion of the resistance faced in some of the really obvious wrongful convictions cases we have worked on, from police, Crown Law and Ministry of Justice, has less to do with the facts and concern for the victims, and more to do with institutional protectionism,” he says.
“Powerful state institutions are acutely political, and I think because of this, their default position on many of these cases is more about themselves and appeasing their political masters than the community and the victims of the miscarriages. And by victims, I mean the innocent who are sent to prison as well as the victims of the original crimes.
“Aside from the Alan Hall case [another conviction based on wrong identification], where a principled Crown Law lawyer acted quickly and courageously, my experience is that many of the officials play an asinine game where they pretend not to be able to see the obvious, and resist and squirm for as long as they can, before scurrying off into the shadows of the bureaucracy when the house of cards finally falls.
“My wish is for a little more courage and a little more humanity in the parts of the system that have a role in rectifying these terrible cases. We’d all be better off for it.”