In my days prior to becoming a lawyer I profess that I was a sucker for all things Law & Order. Not the high-minded ‘concerned citizen’ use of that phrase but rather the fictional inner workings of two separate but equally important groups: the police, who investigate crime, and the district attorneys, who prosecute the offenders … in neat 40 minute episodes.

In those tightly written episodes the defendant was often baited into participating in an inculpatory interview with Police with his seemingly mute public defender by his side. In contrast to this legal ineptitude, what I really pined for was the courtroom scenes. Like those Tom Cruise cross examining Jack Nicholson moments (“You can’t handle the truth!” Mixing my references here, indulge me) where the wily prosecutor swayed the jury through devastating cross examination of the cocky defendant.

As I eagerly awaited these mic-drop moments I learned something else which, I am more confident in saying, is reflected more commonly in criminal practice…. Science is King. Whenever the State wheeled out a scientific expert their evidence was afforded significant weight as the jurors looked on nodding and taking fervent notes. Science, it seemed, was the backbone of every good prosecution.

Far more eminent minds than mine have spoken about the ‘Lab Coat’ and ‘CSI’ (a far inferior show, BTW) effects on jurors. Suffice to say, the evidence of an expert witness with all its real and imagined objectivity can cut through the clouds of lawyerly excrement and provide the jury with compelling evidence upon which to base their verdicts.

In the case of DNA juries have become increasingly exposed to the use of probability statistics and, when faced with competing arguments from either side of the bar table, jurors could be excused for falling back on those verifiable scientific figures as a sort of universal truth which trumps the fallibilities of the ‘human’ evidence.

‘Good Science’, therefore, helps both the Prosecution and the Defence. I am not going out on a limb by saying that advances in forensic science have led to a reduction in miscarriages of justice and, equally, helped convict offenders who might otherwise have escaped prosecution.

It was with all of this in mind that I read the recent Court of Appeal decision in R v Lang[1]. In their joint judgment the Court overturned a murder conviction and ordered a retrial.

In Lang the Appellant successfully argued that the jury had received statistics from an expert which were “irrelevant and hence inadmissible”.

The deceased, who was in a romantic relationship with the Applicant at the time of her death, had been found with a knife protruding from her chest. The defence case was that she had committed suicide, the Prosecution alleged that she was murdered by the Applicant. It was common ground at the trial that the deceased had Bi-Polar Disorder.

During the trial statistical evidence of the rates of suicide in women and, further, statistics about the use of sharp implements in such suicides, was adduced from an expert witness. The Court of Appeal explained the problem with such evidence:

[40] A death caused by a knife wound is itself a rare occurrence, whether the wound is self-inflicted or otherwise. This event having occurred, there were two hypotheses which had to be considered by the jury. The prosecution case was that it was very rare for a woman, even with bipolar disorder, to kill herself, particularly by stabbing herself, which, whilst not of itself proving the fact, made it more probable that she had been murdered. The fallacy in that argument is that the evidence signified nothing about the relative likelihood of the alternative hypothesis, namely that she was murdered….[42] In a case where the jury had to consider two alternative hypotheses, evidence which was said to indicate the likelihood of one hypothesis could mean nothing without evidence of the likelihood of the alternative…

Counsel for the Applicant in Lang cited the UK case of R v Clark.[2] When I noticed that name I was reminded of another example of how statistics can prove to be extremely dangerous in the jury room. It is here I must again digress. True to my clichéd middle-aged white male existence, I am now quite interested in podcasts. It was during an episode of They Walk Among Us, a fantastic podcast focusing on UK criminal justice, that I was first exposed to the harrowing story of Sally Clark.

In 1999 Sally Clark was convicted of the murder of her two infant sons. The first had lived for 2 months before being found unresponsive. At the time she was not charged with this child’s death as it was ruled the death was the result of a SIDs related illness. The second child, born the following year, tragically also passed away at a similar age. This time the expert opinion was that the child had died as a result of being shaken. The same expert re-examined the death of the first child and made new findings, ultimately resulting in the Mother being charged and later convicted of the murder of her children.

Vital to those convictions was the use of statistics. For example, at trial one expert suggested the likelihood of two children from the same family passing away due to SIDs was “1 in 73 million’. With such a statistic being front and centre in the minds of the jury it was perhaps unsurprising that Ms Clark was convicted.

It was therefore not without a healthy dose of irony that, after spending more than 2 years in jail, science –which had played a pivotal role in her conviction- came to Ms Clark’s rescue. It turned out that scientific evidence had existed at the time of the original trial which provided a compelling basis to suggest the deaths were as a result of an infection, rather than by an overt act of the Mother. Problem was, this evidence was not disclosed to the defence and their experts.

More importantly for present purposes the Court of Appeal also found that the use of statistics in the conviction of Ms Clark were ‘misleading’, ‘irrelevant’ and dangerous.

The Appeal was upheld and to their credit, the Crown did not seek a retrial. The case would go on to have profound effects on the way SIDs deaths were investigated in the United Kingdom.

The Takeaway

Statistics are powerful.

When either side of the bar table seeks to rely on them great care should be taken in questioning their underlying presumptions, the data upon which they rely and, most importantly, whether they are in fact probative in the case.

Both Lang and Clark demonstrate that even if statistics are accurate, they have the power to distract the jury to the point of appellable error. It would seem then that if there is any doubt about the probative nature of a statistic it is best that it not be put before the jury.

Statistics then, it seems, can be misleading. Much like a Law & Order episode they can be enticing at first blush but, upon closer inspection, may not stand up to scrutiny.


Remy Kurz
24 January 2020

The views of this author are his own and do not necessarily reflect the views of Robertson O’Gorman