Improving Eyewitness Testimony
by Laura Mickes and Travis Seale-Carlisle
Crime rates across the UK are on the rise with knife and acid attacks featured prominently and regularly in the news. Eyewitnesses can provide valuable evidence, but the way that evidence is collected and used needs much improvement. Psychology has had a lot to say about memory in general and the pitfalls of memory as an accurate recording of past experiences. Fortunately, the solutions are simple and inexpensive.
Eyewitnesses to crimes are often asked to try to identify the perpetrator out of an identity parade. They do not have to pick anyone, but if they did, the person is either a stooge (known innocents) or the suspect. If the suspect is picked, then that provides evidence against that person.
It’s good news if that person is guilty. But it’s bad news if that person is innocent.
Eyewitness errors are costly
Consider a hypothetical case of an innocent suspect, we’ll call James, who is identified from a parade by an eyewitness. Based on that evidence, James is further investigated, charged, and found guilty in a court of law (largely based on the eyewitness testimony). He then is imprisoned.
What is the cost of this error? That’s not a number at our fingertips because it includes personal, societal, reputational, and financial costs. We can try to estimate it.
Investigation costs. An attending officer, crime scene investigators, an investigating officer, an ID officer, and their supervisors are all involved, which takes on average 55 days from the commission of a crime to the charge of a suspect. The average salary for a police officer is about £31,000 per year. Salary costs and resources substantially add up during this process.
Court costs. According to the Crown Prosecution Service, the cost of one committal for trial at the Crown Court is £3,500.
Prison costs. The yearly cost of housing a prisoner is £35,000. And let’s say James is imprisoned for 15 years (the average time those who have been exonerated are imprisoned). That’s a total of £525,000.
If James is exonerated, and if he can prove his innocence, he may receive compensation. The maximum compensation amount is £1 million for those imprisoned for over nine years.
Once exonerated, the police investigators have to re-open the case.
Meanwhile, the real perpetrator has been free to victimise more people.
Then there’s the emotional fallout. James is at risk of poverty due, in part, to reputational damage and years of lost wages. Chances are high that his relationships have dissolved. He will likely have mental health issues (depression, anxiety, drug dependence, etc). It doesn’t stop there. All involved are affected to some degree. How do the players (eyewitness, police investigators, judges and jurors) feel about the role they played in sending an innocent man to prison? How do society members feel about the integrity of the criminal justice system? These costs are immeasurable.
A grim scenario. But it doesn’t have to be. Research suggests that we can improve the value of this evidence.
Let’s go back in time and allow the eyewitness express confidence in their identification of James. It would have been expressed with low confidence. This potentially could have saved James 15 plus years of turmoil and the taxpayers well over £500,000 for just this one case (not including the £1 million compensation).
When we express low confidence in a memory, we are effectively saying that there’s a good chance we’re not right. By ignoring this information, there’s an increased risk of sending innocent people to prison. The police can continue investigating those suspects, but should keep pursuing others.
When we express high confidence in a memory, the accuracy of that memory is usually high. If an eyewitness expresses high confidence in an identification, the police can have more certainty that their suspect is guilty and continue to find more evidence to support that case.
The value of confidence is one conclusion drawn by the ESRC-funded project ‘Investigating New Ways to Improve Eyewitness Identifications Using Receiver Operating Characteristic Analysis’. Importantly, only during the initial identification procedure does confidence tells us anything about accuracy. It loses predictive value in the court because the forces that contort memory have been at play in the months between the time of initial identification and the trial.
The easy, inexpensive solution is for the police to collect expressions of confidence at the initial identification procedure and for the courts to use it. This change negligible in cost, and has great potential to drastically reduce costs of a false identification.