Expert evidence in court

Let me start by saying that this piece has nothing whatever to do with accountancy or tax.  So if that is what you are looking for stop reading now!

But it does have to do with how courts use evidence from expert witnesses - and accountants (such as myself) sometimes give evidence as expert witnesses in court.  In the last few months there have been three interesting Court of Appeal cases concerning expert witness evidence.

A 'normal' witness in court is required to restrict his evidence, generally speaking, to what he saw, or heard, or did.  He is not allowed to give his opinions.

But an expert witness is allowed to give his opinions in evidence where he has skills / knowledge not available to the court because they are outside the range of skills / knowledge which is common in the community.  So, for example, the court would not allow expert evidence as to whether a defendant could properly be described as 'tall' because a judge (in a civil case) or a jury (in a criminal case) can perfectly well decide that for themselves when they see the defendant.  But an expert could give evidence as to whether a fingerprint found on a knife used in a stabbing was the defendant's fingerprint.

Rather than say simply (and certainly) that the fingerprint was that of the defendant the expert is likely to express his opinion as to how strong the evidence is to suggest that the fingerprint in question is that of the defendant.  In attempting to describe the level of support which the evidence gives to the proposition that the fingerprint was that of the defendant he might select one of the following expressions, "weak or limited support, moderate support, moderately strong support, strong support, very strong support, extremely strong support".  But he is unlikely to say, for example, that there "is an 87% chance that the fingerprint on the knife is that of the defendant" because there would be no scientific basis on which he could come up with that figure.

Experts have, for understandable reasons, attempted to create some logical basis on which they can put some figures to the strength of their opinion and then relate those figures back to the words they use in evidence.  This is sometimes referred to as the Bayesian approach.  The courts have ruled however that (except in the case of DNA samples - where there is a large database of underlying data) it would be wrong for an expert witness to try to 'calculate' the strength of his opinion in that way.  This is because there is not sufficient underlying data to allow this calculation to be done reliably.

So, for example, an examiner of footprints was recently criticised for using a mathematical method to calculate how sure he was that a footprint had been made by a trainer belonging to the defendant in a murder case (R v T [2010] EWCA Crim 2439).  He did this by considering the sole pattern (apparently Nike trainers have 1,200 different sole patterns), the size of trainer, the wear in the pattern, and any damage to the tread pattern in the footprint and on the defendant's trainer.  He then put a numerical probability on a match for each of these four factors and multiplied the four figures together.  The Court of Appeal was not happy with his methodology, and also not happy that he had failed to disclose in his evidence that he had used a mathematical approach in forming his opinion.  The result was that the defendant's conviction was quashed.

In a another recent appeal, again in a murder case, there was conflicting evidence concerning whether the fibres found at the crime scene and on the victim were from the same source as fibres found in the defendant's wardrobe and in his car.  No garment was found which appeared to be the source of any of the fibres.  There were a range of similar fibres found at the crime scene and another range of similar fibres found in the wardrobe.  Clearly some fibres from one sample did not match some fibres from the other - but some fibres from the two samples did match.  Did both samples come from the same garment?  That proved to be a question on which the experts disagreed.  Ultimately the court found one of the experts most credible (and not the one who most fiercely stuck to his original view under cross-examination).  They preferred the approach of the expert who seemed willing to take on board different views but ultimately came to a view as to which conclusion should be drawn (R v Hall [2011] EWCA Crim 4).  In this case the Court of Appeal found that the evidence did link the defendant to the crime scene.

In a third recent case the court had to consider whether the 'expert witness' was recognisably an expert in anything at all.  The expert (who described himself as "a podiatrist and specialist in lower limb gait, pathomechanics and biomechanics") gave an opinion that the 'gait' of the defendant seen walking on one piece of CCTV footage (filling his car with petrol at a garage) was the same as the 'gait' of a person seen on another piece of CCTV footage (showing a gunman shooting dead his victim at a crime scene, with an accomplice) - meaning that the defendant was likely to be the accomplice visible on the crime scene video.  (The faces on the videos were indistinct.) The prosecution relied upon the expert evidence concerning the way in which the 'two people' seen in the videos had a similar erect posture with head slightly forward, both kept their feet close together when walking (as opposed to splaying their feet), etc.  The defence argued that this expert evidence was nothing more than subjective opinion based on the expert's personal view of the comparability of the walking movements on the two videos.  The court accepted that the expert evidence was properly put before the jury and the conviction could stand (R v Otway [2011] EWCA Crim 3).  The convicted man is serving a life sentence with a minimum recommended term of 27 years.

A while ago I was an expert witness instructed by the defence in a case involving alleged theft of cash from shop takings.  Part of the prosecution evidence was that recorded weekly takings of the shop apparently fell after the defendant commenced his employment there, and rose again after he left.  The suggestion was that this was evidence of theft of cash by him while he worked there.  During the evidence myself and the expert instructed by the other side got into considerable detail of recorded takings week by week (and on average) over the 18 months or so that the defendant was employed, and recorded sales of corresponding weeks in the year before and the year after.  At one point a jury member asked if a chi-squared test had been undertaken.  In my view, if the key evidence is that tenuous then the Crown should not have pursued the prosecution - but that's another story.  The defendant was not convicted of the theft of cash.

Ultimately a jury is by no means bound to accept the expert's opinion as being correct.  It is for the jury to decide if they find the expert's evidence persuasive.

However if you ever find yourself in court as a member of a jury hearing evidence from an expert witness you will now be aware of at least some of the legal context in which that evidence is being given.

David

www.AccountingEvidence.com

Comments
b.clarke's picture

Footprint examiner knows numbers?

b.clarke | | Permalink

"... an examiner of footprints was recently criticised for using a mathematical method to calculate how sure he was that a footprint had been made by a trainer belonging to the defendant in a murder case (R v T [2010] EWCA Crim 2439).  He did this by considering the sole pattern (apparently Nike trainers have 1,200 different sole patterns), the size of trainer, the wear in the pattern, and any damage to the tread pattern in the footprint and on the defendant's trainer.  He then put a numerical probability on a match for each of these four factors and multiplied the four figures together.  The Court of Appeal was not happy with his methodology,... "

A very interesting report. May I summarise para 80 of the judgement as "Garbage in, garbage out"?

I think there is in fact a relevance to accountancy here; be very, very, VERY careful if you use statistics in a report.

 

 

Add comment
Log in or register to post comments
Group: Money laundering and crime
A group for discussing issues relating to suspected money laundering and other crime