David Farrar: The Polkinghorne verdict

I wrote this post before the jury delivered a verdict, but after the conclusion of evidence. It will only be published after the verdict is announced (which was just announced as not guilty).

Like many I have followed the media reports of the death of Pauline Hanna since it occurred. It is worth reflecting that regardless of the cause of death of Hanna, it is a very sad and tragic story. You had an incredibly competent professional woman in a top job, earning a high salary – yet she was in a relationship which saw her apologising for not paying enough for groceries, not always looking “good” enough, trying to please her husband through group sex etc. It was so sad having her life and insecurities exposed through her death. Regardless of the verdict, she is a tragic victim.

To a degree Polkinghorne is also a figure of sadness. He was one of the most respected eye surgeons in New Zealand, if not outside. He was earning $750,000 a year. He had an amazing wife and house. He was popular and had a wide circle of friends. It’s a life many would aspire to. But he was using P, addicted to sex, worrying about money and most sadly of all fell in love with his escort, mistaking a transactional relationship for a true relationship.

So before we get to the criminal side, it is worth remembering it is also a strategic story of two unhappy individuals, and how so many people have lives that appear perfect on the surface, but are deeply flawed underneath.

When Hanna first died, it was difficult to reach conclusions on what happened. The Police seemed to be very effective at leaking evidence to the media that portrayed Polkinghorne is a bad light. The rushing off to the arms of Madison Ashton made him look guilty, and the rumoured affairs and drug use made it look like a typical he killed the wife to get her out of the way.

But then more stories came out that Hanna knew of his outside relationships, and even took part in some of them. It didn’t appear she was going to leave him. Motive was less.

When the trial started I thought it was probably murder on the balance of probabilities. There didn’t seem to be much of a motive though. And then the opening statement for the prosecution was that he had pleaded guilty to P charges. To me this filled in a missing link. People on P can do insane things, go into a rage etc. It seemed very possible that he snapped under the influence of P. So at the beginning of the trial, I thought it was very very likely a murder, but not quite at the level of reasonable doubt. The rope evidence also pointed to a staged scene.

But as the prosecution went on, the case was underwhelming. Admittedly Ron Mansfield is stunningly good defence lawyer (can someone send me his cellphone number in case I ever get charged with murder), but the evidence was not overly strong. The lack of significant injuries on Hanna was a major factor. Now it is possible that Polkinghorne managed to kill her in a way that very cleverly left no evidence and made suicide plausible – but if he was high on P at the time and in a rage, how likely is it that he was coherent enough to cover it up so well.

The tech experts introduced more doubt, and the evidence on Hanna’s mental health was heartbreaking but made suicide more plausible. And the two Crown pathologists said they could not rule out suicide, which by itself seemed to be reasonable doubt.

So by the end of the (reporting of) the prosecution case, I was in the not guilty beyond reasonable doubt space – before even hearing the defence’s case. I just didn’t think the prosecution had eliminated reasonable doubt. I still thought on balance of probabilities he may have done it, but a long way from beyond reasonable doubt.

Then we had the defence case, and the evidence from the two defence pathologists was very powerful. Basically the only thing the four pathologists agreed on was that it could have been suicide, and they were split 2/2 on whether it could have been homicide. Again hard to see this being beyond reasonable doubt.

The defence also resonated with the lack of evidence for homicide compared to the evidence for suicide such as decades of anti-depressants, a previous failed attempt, the massive work pressures etc.

The summings up went the other way to how I saw the evidence. I thought the crown did a much better summing up than the defence. The defence was too focused on outrage that there was even a trial, and not enough on the numerous areas of reasonable doubt.

Anyway by the end of the trial, I now shifted on the balance of probabilities to it being suicide. I will be very surprised if he is convicted, but of course respect the jury’s decision as they heard all the evidence, and I have relied on media reporting.

By coincidence Martin Lally has done an intriguing bayesian analysis of the trial, and he concludes:

Fourthly, when applying a Bayesian analysis to the case, there are plausible combinations of

views about the evidence that lead to a high probability that the case is a murder, of 87% in

Table 11, but there are also plausible combinations of views about the evidence that lead to a

very low probability that the case is a murder, of 0.7% in Table 10. If one took the view that

the parameter values underlying Table 11 were correct, one would be at or close to the point

where reasonable doubt about guilt did not exist. Alternatively, if one took the view that both

tables 10 and 11 deserved some weight, and further weight on other possibilities between them,

such a person would have reasonable doubt about guilt.

I’ve never seen this approach to trial evidence before, but it is sort of what juries are meant to do – look at the evidence as a whole, not just as individual strands. Lally finds the most favourable interpretation of the evidence is an 87% chance of murder (which would still be reasonable doubt) and the least favourable is a 0.7% chance.

David Farrar runs Curia Market Research, a specialist opinion polling and research agency, and the popular Kiwiblog where this article was sourced. He previously worked in the Parliament for eight years, serving two National Party Prime Ministers and three Opposition Leaders.