EVIDENCE (IDENTIFICATION) AMENDMENT BILL

Second Reading

Adjourned debate on second reading.

(Continued from 18 May 2011.)

The Hon. S.G. WADE (12:49): This bill highlights a clash between the law and the science, but let us not miss the wood for the trees. The law and the science agree on one point, the key point, that identification testimony is inherently unreliable. From the perspective of law, the innocence project in the United States shows that about three-quarters of wrongly convicted people were wrongly convicted on the basis of identification evidence. On average, these people served between 12 and 15 years of prison time when they were innocent. Juries find identification testimony very persuasive—perhaps more persuasive than the science would suggest it should be.

From the perspective of science, studies suggest that, whatever method you use, misidentification rates are significant. Let me highlight the misidentification rates in four scenarios. The first scenario is where the offender is present and the line-up is rejected; that is, a witness says they are not there, or cannot pick them. That is incorrect on 20 to 50 per cent of occasions. Scenario two is where the offender is present and a known innocent person is selected. That occurs about 20 per cent of the time, and for the sake of clarity we might call that the 'Kevin Foley scenario'.

Scenario three is where the offender is not present and a known innocent person is selected. That occurs in around 20 to 50 per cent of cases. Scenario four is where the offender is present and the offender is selected. That could be anywhere from 20 per cent of the time, if conditions were such that on average witness memory is poor, to 90 per cent of the time, if conditions led to good witness memory.

Given that this council is looking at the need for a criminal cases review commission to provide appeals for miscarriages of justice, it would be bizarre for us to treat lightly an area of law which is like a bag full of seeds for potential miscarriages of justice. So, what does the criminal justice system do to reduce this risk? Firstly, the courts have a well-established preference for live line-ups. The law regards live line-ups as more reliable than other forms of identification.

In Alexander v The Queen (1981), the High Court established the practice that a live line-up should be used whenever possible and that judges have discretion to exclude evidence such as photo board line-up evidence if a live line-up is possible. The law also recognises that putting aside the quality variables, live line-ups have other significant benefits over photo boards.

Firstly, to be able to do a photo board identification you must have a photograph of the suspect, and juries could be led to assume that if you have a photograph of a suspect they must already have a criminal record and therefore are more likely to be guilty. With a photo board there is no safeguard against police suggestion. At least the accused and often a lawyer or other independent person is present at a live line-up. That may not be the case with a photo board identification.

Thirdly, a live line-up provides a wider range of identifying characteristics to be used by the witness in making the identification, for example, a person's body shape. I received a copy of a letter dated 18 March from the Law Society to the Attorney-General which highlights these points:

We note that the existing law in this State, which is in line with High Court authority, is that parade identification is the best identification...We accept that this is so. We are in general disagreement with the comments accompanying the Bill to the extent that they advocate for photoboard identification to the exclusion of parade identification.

If the Bill is passed in this form, we believe that it will benefit the accused and, we accept, the police by way of providing a saving of resources. The consensus of the Criminal Law Committee (by experience) is that parade identification is much more reliable. The witness is able to view the suspect in three-dimensional form and assess factors such as height, build etc.

At the same time we acknowledge that police resources are not finite—

I think they meant 'not infinite'—

Nevertheless, an effective change from parade identification should not be hijacked by the resource issue if it otherwise may lead to less identifications and, therefore, less convictions.

A key goal of this bill is to negate the judicial preference for live line-ups. In doing so, the government has tried to use science to discredit live line-ups. As I will show later, the science in fact discredits all line-ups done badly.

Secondly, the law seeks to control jury warnings. In relation to jury warnings, there is a standing obligation on a trial judge to adequately direct any jury about the possible unreliability of identification testimony, whatever method is used. In the text Australian Evidence, a leading Australian resource on this area of the law, our own South Australian former academic, Andrew Ligertwood, expresses the law in the following terms:

Consequently, courts have been conscious that juries should be made fully aware, by clear and careful warning, of the risks inherent in both observation and recognition, and that police should employ recognition procedures which eliminate the most obvious risks. But this consciousness has so far fallen short of the imposition of either mandatory guidelines for adequate warning or mandatory recognition procedures to be followed by police, despite appellate courts being given opportunities to tighten up their approach. This failure has led to recommendations for law reform in this area.

In that passage, Mr Ligertwood suggests that to deal with unreliability of identification evidence, the laws of evidence in relation to the identification testimony could be changed to do two things: firstly, enhance judicial warnings to juries and, secondly, provide for mandatory recognition procedures to be followed by police. This bill focuses on the first and does nothing about the second.

Both law and science suggest that identification procedures do matter. In fact, the government's preferred authority in this area, Professor Neil Brewer of Flinders University, suggests that how an identification method is implemented matters more than the identification method used. According to science, there is a negligible difference between the reliability of identification using live line-ups and identification using photoboard line-ups. In a recent police journal article, Professor Brewer wrote:

Across a number of controlled studies, the evidence does not indicate that either method produces superior performance in terms of preventing identification of innocent suspects or maximising the identification of offenders.

This is not to say that, at some stage, evidence will come to light that demonstrates the superiority of one or other method under particular conditions. But, at present, there is no compelling evidence for the superiority of either method.

I suspect that this might reflect the fact that witnesses generally simply do not store the rich details in memory that would allow them to take advantage of the additional memorial cues that the live line-up would appear to offer.

What makes a huge difference to the reliability of a line-up is how the line-up is constructed and how it is delivered. My understanding is that Professor Brewer supports the use of photoboard line-ups in preference to live line-ups because a photoboard line-up gives you a greater capacity to change the quality variables. It gives you a better opportunity to promote quality, but only an opportunity. As Professor Brewer put it at a recent public lecture, 'A photoboard done badly is just bad as any other method.'

The opposition supports ensuring that our law allows the justice system to rely on the best available evidence and the best available science. We support the removal of the judicial preference for line-ups, but we do not support doing so without firstly ensuring that scientifically verified quality identification procedures are in place for all modes of identification and, secondly, ensuring other steps are taken to maximise the quality of identification evidence.

Last year, Professor Brewer and a colleague, Matthew A. Palmer, published an article in Legal and Criminological Psychology, a journal of the British Psychological Society called 'Eyewitness identification tests'. The article concludes with a section called 'Summarising what works', which states:

In sum, what should line-up practitioners focus on given the current state of knowledge?

What are the procedures most likely to ensure maximum diagnosticity for eyewitness identification test evidence?

The following points capture the main conclusions to emerge from the preceding discussion:

(1) Line-ups should contain only one suspect. Foils should capture the key features of the witness's description and be plausible matches for, but not too similar in appearance to, the suspect.

(2) Do not be preoccupied with the line-up presentation medium as there are many other factors that seem to be far more important.

(3) Double-blind line-up administration should be preferred.

(4) Witnesses who have previously seen a suspect's face in a mugshot search or showup should not subsequently view a line-up for that suspect.

(5) Ensure unbiased instructions providing clear warnings regarding possible culprit absence.

(6) Immediately after the identification decision, ensure there is an (independent) record of the witness's exact identification response, the confidence in that decision, the witness's decision latency, and other perceptions that witnesses had of the encoding and identification experiences.

(7) Avoid giving disconfirming feedback to witnesses following an initial identification decision if there is a chance that the witness may subsequently be asked to view another line-up.

(8) Discount courtroom expressions of identification confidence and other witness perceptions of encoding and the line-up.

At this point it is important to reiterate a point we made at the start. The available evidence clearly suggests that, even if all of the suggested guidelines are diligently observed, eyewitnesses will still make inaccurate decisions when examining line-ups, including the selection of innocent suspects and line-up foils and the rejection of line-ups containing the culprit. Moreover, while we are hopeful that future work will gradually help us refine these guidelines, we believe that the traditional identification test formats that we have been discussing in this chapter will remain limited in their capacity to discriminate the guilty from the innocent.

Debate adjourned on motion of Hon. G.E. Gago.

[Sitting suspended from 13:01 to 14:17]