Assessing Witness Evidence: How Does the Judge Know Who to Believe?

There are, however, cases where there is a dispute as to what was said or done and when. Consent cases are one obvious example. The patient says she was not warned of a risk. Her doctor says she was. Sometimes birth injury cases turn on timings and there may be different accounts of what was said and done when. In these cases the evidence of witnesses who were there will be important.

What happens when there is a conflict of evidence? Is it just one person's word against another's? How does judge decide who to believe? Is it all down to the impression they make in the witness box?

Some of the reported cases give helpful guidance on how to assess witness evidence. In one recent tax case the judge summarized the right approach. Witness evidence must be weighed with 'objective facts'. Here is an outline of that approach.

Layer 1: Agreed Facts

Assessing a witness's evidence starts with looking at how it fits against the overall background. The first part of that background is the facts which are not in dispute. These should be clear from the statements of case on both side. For instance, in a clinical negligence case, it may be agreed that the patient was suffering from a particular medical condition, was treated in a particular hospital and seen by a named doctor. Witness accounts which fit with this background will be persuasive. Accounts which do not fit will not.

Layer 2: 'Incontrovertible' Evidence The next part of the background is facts which cannot really be doubted whether they are agreed or not. They are 'incontrovertible'. These are matters independent of a witness's memory. There may appear from documents. Or there may be some form of hard evidence. In a traffic accident claim they may be marks on the road or damage to the vehicle suggesting how a collision happened.

In some cases it is not the content of documents but their absence which is important. There are times when if a witness's account were true, one would expect to see a particular document. For instance, there may be a clinical record saying that a patient failed to attend an appointment. Can it be assumed this was the patient's fault? Did the patient know? If the patient had been told of the appointment one might expect to find in the records a letter notifying her. The absence of a letter might suggest that the 'no show' was administration failure rather than patient default.

In one case the Court of Appeal commented, 'what has impressed the judge most in his task of fact-finding was the absence, rather than the presence, of contemporary documentation or other independent oral evidence to confirm the oral evidence of the respondents to the proceedings'.

Documents will be particularly persuasive where they meet 2 tests. The first is that they were contemporaneous – they were made at the time of the events. The second is that they were made without the dispute in mind.

In an important article, the late Lord Bingham commented, “In many cases, letters or minutes written well before there was any breath of dispute between the parties may throw a very clear light on their knowledge and intentions at a particular time”.

In clinical negligence claims the medical records are important documents. Patients have no input into the content of these records and may complain they are wrong. It is possible that not every entry is correct but it is unlikely that many are wrong. Where they are computerized, later changes can be tracked. So despite what some patients fear, it is unlikely that computerized records will have been doctored. My own rule of thumb in assessing what clients tell me is that the more they question the records, the less convincing their account is likely to be.

There is an exception where notes are not made until after a period of crisis and the exact details may be open to dispute. For instance when things suddenly go wrong during labour, midwives and doctors may not be able to make notes until after the baby has been born and is breathing properly. Sequences and timings may be approximations and open to dispute. In these claims, working out exactly when everything happened may be crucial.

I would add 2 more rules of thumb. The first is that the more detailed a record, the more persuasive it is likely to be. A very brief GP record without an adequate note of the history taken or the examination performed will not convey the impression of a thorough GP (although she may have been). The second is that the more people who have taken the same information independently and recorded it, more likely they are to be correct. But it is important that they have taken the information independently. There are instances where an incorrect story is recorded at the outset and repeated by people who have read the incorrect note. Mere repetition of information not derived independently should not increase credibility.

So the first 2 stages in assessing witness evidence at to view it against the background of agreed facts and 'incontrovertible evidence'. Evidence is often 'incontrovertible' because it is established by contemporaneous documents. To be persuasive, witness accounts will as a general rule need to be consistent with those documents.

Assessing the Witness: Demeanor

How a witness comes across in the box is important. We generally learn about people from how they behave. But it may not be quite as important as we think.

Again the Court of Appeal commented, 'This task is not to be carried out merely by reference to the impression that a witness made giving evidence in the witness box. It is not solely a matter of body language or the tone of voice or other factors that might generally be called the 'demeanour' of a witness.'

That is a good thing because forming judgments of people is very subjective. We may warm to one person but not to another. Psychologists refer to the 'halo effect'. If we like a person, we are more inclined to accept what they say. But a person may be likeable without being truthful. An unconscious bias is at work.

As one judge warned that our views of people can be 'distorted through the prism of prejudice' .

Assessing the Witness: Tests of Truthfulness

So how can one test the truthfulness of a witness if demeanor is an unreliable guide? In a 1968 House of Lords case, Lord Pearce pointed out that assessing credibility involved a number of questions :

1. Is this witness is a truthful or untruthful person?

2. Is the witness telling the truth on this issue? (An untruthful witness still tells the truth sometimes.)

3. Has the witness remembered correctly? Or has his recollection been altered by unconscious bias or wishful thinking.

4. Even is the honest witness correct in what he thinks he saw? Or is it more likely that he made a mistake?

These break down the issue of credibility into parts. But they do not help the judge work out whether a witness is honest or not. Lord Bingham suggested 5 tests to assess the honesty of a witness:

1. Consistency of the witness's evidence with what is agreed or clearly shown by the evidence to have occurred (and this is what I have referred to as the first 2 layers);

2. The internal consistency of the witness's evidence;

3. Consistency with what the witness has said on other occasions;

4. The credit of the witness in relation to matters not germane to the litigation; and

5. The witness's demeanour.

There are others relevant factors. One is motive. A witness who has nothing to gain from the evidence may be more compelling than one suspected of an ulterior motive. That does not of course mean that a party giving evidence in support of their case is assumed not to be telling the truth.

As a general rule a simple story is often more persuasive than a complex one. Again that does not mean that the more complex story is not correct. But a simple story will often appear more coherent. And a longstanding principle of philosophy is that when assessing different answers to a problem you should normally favour the simplest ('Occam's razor').

Conclusions

So in assessing witness evidence, how does the judge know who to believe? I have set out some principles considered by the courts in reported cases and suggested some other principles which guide me. The starting point is agreed facts. The next layer is 'incontrovertible evidence', facts for which there is good evidence independent of human memory. With that as a base, there are then some tests of witness credibility which go beyond demeanor.

Put in diagrammatic form, there is a hierarchy of factors which could set out as the pyramid below.

Paul Sankey is a solicitor and partner at Enable Law, specializing in clinical negligence claims for patients. https://www.enablelaw.com/team/paul-sankey/

Many thanks to Paul and Enable law, for permission to reprint this article.