A tool designed to predict future crime in terrorist offenders considered them at greater risk of offending if they were autistic despite having no empirical basis to do so, an independent report has found.
The report into the Vera-2R tool, which was released to Guardian Australia and others under freedom of information laws, found a lack of evidence underpinning the instrument had “potentially serious implications for [its] validity and reliability” and found it was “extremely poor” at predicting risk.
It also found that autism spectrum disorders, as well as non-compliance with conditions or supervision, were included as risk factors in the tool, despite a lack of empirical evidence.
The most comprehensive systematic review that has been conducted on the drivers of radicalisation and terrorist behaviour and violence found not a single piece of empirical evidence that supported the inclusion of those two factors, the report found.
The release of the critical report came after a series of Guardian Australia stories which exposed the federal government’s continued use of the tool despite criticisms. The federal government did not disclose the report to the lawyers of convicted offenders, its own experts, or the NSW government, which uses the tool to justify harsh post-sentence orders for offenders, including ongoing detention.
The report, titled “Testing the Reliability, Validity and Equity of Terrorism Risk Assessment Instruments”, was completed by Australian National University academics Dr Emily Corner and Dr Helen Taylor for the Department of Home Affairs and runs to more than 270 pages.
The government received the report, known as the Corner report, in May 2020.
It was released under FoI with minimal redactions, despite the federal government previously claiming it could not be released because of national security.
“The lack of evidence underpinning both instruments has potentially serious implications for their validity and reliability,” the report found.
“Without a strong theoretical and empirical basis for factor inclusion, it is not reasonable to anticipate that the instruments are able to predict their specified risk with anything other than chance.
“If an instrument with a weak evidence base is employed as a predictive instrument by practitioners, it is not possible to determine if individuals who pass through assessment processes would ever be suitable for the management plan as determined by the risk decision outcome made on the instrument.”
Guardian Australia has previously reported the tool was used 14 times by the federal government after they received the report, and the NSW government also continued to use it to assess offenders, including some without convictions for terrorism.
Corner and Taylor also assessed the Radar tool, a classified instrument which is used less frequently to justify post-sentence orders than Vera-2R, and said their report was the first piece of research to be performed on these instruments.
The report made four recommendations, including that another evaluation is conducted, that the authors of both instruments “enact far more thorough evaluations of the wider theoretical and empirical literature to help develop risk factors that accurately reflect behavioural trajectories towards radicalisation and terrorist violence”, and that the risk specification of each instrument should be refined.
It found that the correct instrument, implemented correctly, could allow risk assessment that ensured those at highest risk of offending received the most intensive interventions, while those presenting as lower-risk were protected from too much intervention and therefore avoided “the disruption of circumstances that were making them low risk in the first place”.
But it found there was little basis for the claims made by the developers of the Vera-2R of its strong reliability and validity.
It said it was “concerning” that almost 60% of the cited evidence base for factor development of the tool was not empirical, and less than half of the works cited in the Vera-2R accurately reflected what was in the recorded texts.
The test can only be used by people who complete training accredited to the author, with the report noting that “there is no information within any documentation as to the time or costs of this training program, but it is believed to be a multiple day program with substantial costs attached”.
The federal and NSW governments both say they are considering the Corner report findings, but this week’s federal budget shows the Labor government is committed to continuing to fund its regime of post-sentence orders.
The budget gave $130.1m over two years to the high-risk terrorist offenders scheme.
A NSW communities and justice department spokesperson said the Vera-2R tool was just one of a number of tools used to assess risk.
“The Department of Communities and Justice is currently considering the Corner report in collaboration with other stakeholders,” the spokesperson said.
