The Coalition has decided to drop the privatisation of polygraph, or ‘lie-detector’ tests for sex offenders. But the continued use of this flawed technology within the probation service is misguided and the whole programme should be scrapped.
Since the Offender Management Act was changed in 2007 to allow for the attachment of a polygraph condition to terms of probation, trials of the device for use with post-conviction sex offenders have been taking place in the Midlands. These concluded in 2012, were reported to be a success by the government and are due to be rolled-out nationally.
However, in scientific and legal communities the polygraph’s validity (whether it can detect lies at all) and its reliability (how regularly is makes an error in categorising truths as lies and vice versa) are highly contested and have been so since its early development at the turn of the 20th century.
Above all, it is far from clear whether the use of these measures reduce reoffending rates or improve offender rehabilitation outcomes.
The polygraph machine was first adopted in the UK at the peak of a media-provoked cycle of one-upmanship between Labour and the Conservative Party over who could be toughest on crime.
Tony Blair’s 2005 Labour Party manifesto promised to trial the lie detector for use in the monitoring and treatment of paedophiles post-conviction and thus opened-up the UK to the official use of the device for the first time.
The political justification for their use continues to rely on the idea that parties have to be seen to be being tough, as Justice Minister Jeremy Wright has said, “Introducing lie detector tests, alongside the sex offenders register and close monitoring in the community, will give us one of the toughest approaches in the world to managing this group.”
What about the scientific justification? According to proponents the polygraph works by measuring the concurrence of certain physiological responses (e.g. pulse rate) with deceptive behaviours. The use of the device with sex offenders would be to ensure they are being truthful about their behaviour during probation. However, deception can occur in the absence of these physical responses, and the physical responses can occur in the absence of deception. Moreover, the most extensive US National Academy of Science report on the device concluded that the polygraph’s reliability was flawed as regards its real-world generalizability and that additional basic research was needed.
Scientists involved in the UK trials with sex offenders have conducted research into its efficacy. However, rather than focussing on validity and reliability the research has concentrated on the value of the polygraph as regards the elicitation of ‘clinically significant disclosures’ (CSDs).
Such CSDs are typically used to evaluate the offender’s riskiness, change their treatment strategies or alter their probation conditions.
In this context, the validity of the polygraph is seen to rest not so much whether it can detect lies but whether it can get offenders to make more disclosures about their behaviours. Even if we accept this premise, the question of reliability is still valid: how often does it make mistakes in categorising those disclosures true or false?
The report on the 2012 trial has a worrying feature in this regard. It turns out that the majority of the disclosures are not made during the actual test but in the pre-test interview. As such, the polygraph’s role appears to be less a lie detector and more a threat or method of inducing confessions from offenders.
Reportedly, the use of the test is of value to offender managers because it gives them confidence that offenders are sticking to their probation conditions, discloses risk and allows managers to challenge risk.
Given the focus on numbers of CSDs it seems distinctly unwise to rely on the polygraph as a method for helping to determine risk and of inducing confessions, particularly when many of these disclosures are being made before actually connecting up the device to the offender. As such, a significant risk in adopting the polygraph in this context may be an over-reliance on the veracity of CSDs.
Offenders are likely to be able to manipulate the examination to their own means just as much as managers and examiners are able to use the examination to elicit CSDs.
There is no research at all on strategies the offenders may use to make CSDs in relation to anticipated examinations or during pre- and post-test interviews.
If an offender knows that they have breached the terms of their probation and fear that their examination is going to result in a ‘deception indicated’ result, then they might well offer less significant CSDs in advance of the examination in order to help shape the interpretation of their results.
Furthermore, we don’t have any information on what might happen when an offender has an erroneous result of ‘deception indicated’. When false positives occur, it could be risky for offenders to maintain that they are not lying. If the polygraph is to be believed, this means they are not acting in a trustworthy manner. In line with this suspicion their treatment and probation conditions might be changed. Might offenders provide CSDs that are themselves lies in order to convince the examiner and manager that they are now telling the truth? We have to know a lot more about this technology and how it is used before we trust it to help determine the riskiness of offender behaviours.
Finally, technologies used to manage sexual offenders often leak into other areas and it is possible that polygraphs could in future find use in other contexts of treatment and rehabilitation, particularly if corporations could profit from their adoption.
However, the government’s idea to privatise the probation service has received a poor assessment from its own internal reviews and has been dropped. Instead, the Ministry of Justice plans to go ahead with the programme, but will keep this part of the probation service in-house.
But this doesn’t go far enough. Focussing on how tough the programme is on offenders and on the value of ‘clinically significant disclosures’ papers over the critical question of whether these measures reduce reoffending rates and improve offender’s rehabilitation outcomes.
Ultimately, the safety of children, communities and the offenders themselves will not be improved by increasingly punitive measures if they do not tackle the causes of sexual offending, improve rehabilitation and reduce the rates of sexual offences.
The adoption of technologies that are seen to be tough on offenders should not be an end in themselves.