A "first of its kind" program in Australia that collects data from court decisions on refugee status to examine judicial bias has revealed dramatically different outcomes depending on the tribunal member who oversees the case, according to UNSW's Kaldor Research Centre for International Refugee Law.

The aim of the centre's Data Lab is to use the data to ensure fairer outcomes for refugees and people seeking asylum and to inform policy reforms and improve processing at courts and tribunals.

The lab's data analysis tracks trends in court rulings and can point to potential biases held by individual judges. The centre's director Dr Daniel Ghezelbash said using data to analyse decisions made by individual judges is the “most controversial part of our work”. A similar project in France using statistical analysis to name judges and their decisions led the French government to pass criminal prohibitions. “If I was doing this work in France, I could potentially be sent to jail,” he said.

Data analysis is a new and not yet common form of legal research in refugee law. This is in part due the subjectivity of every case and a misunderstanding of what statistical analysis is.

“I think it makes certain corners of the legal profession and legal establishment very uncomfortable,” Ghezelbash said. “The common response is that every case is unique and should be examined on its merits and this sort of quantitative comparative analysis is not meaningful.”

The data is collected at judicial review and can be used to inform the litigation strategies of refugee lawyers. This means that a refugee lawyer using the data could determine the success rates and decision strategies of an individual judge, and then tailor their litigation argument to be favourable to that judge. In an area of law where the number of cases is huge, being able to pick out relevant decisions is important.

'I think it makes certain corners of the legal profession and legal establishment very uncomfortable.'

Barrister Chris Honnery, who works in refugee law, described using the Kaldor Centre’s data findings as helpful to identify relevant judicial decisions where the volume of cases was a challenge to practitioners. “I want to understand the process judges have adopted in analogous cases to the one I am preparing for,” he said in a statement sourced through the Kaldor Centre's director Ghezelbash. “The data lab provides access to resources that help identify relevant decisions of judges and is a valuable resource for traversing and filtering decisions.”

The Kaldor Centre’s data analysis found different success rates for refugee protection visas dependent on the tribunal member who reviewed the case. The research showed one member had an approval rate of less than 5 per cent while another approved 86 per cent of visa cases. In an article for The Conversation in 2020, Ghezelbash said it was difficult to draw conclusions from the variations in data due to individual preference or the allocation of members which is influenced by expertise on specific countries.

Sydney barrister Geoffrey Watson SC said arguing a case in favour of a particular judge was not uncommon during court proceedings and it was “crucial” and the responsibility of a barrister to present a case in a way that was “most palatable to the individual judicial officer”.

“As a barrister, I’ve done these sorts of things,” said Watson, a former counsel to the NSW ICAC, a director of The Centre for Public Integrity, who is active in pro bono work for asylum seekers. “I haven’t changed the evidence, but I’ve certainly played up the fact before one particular judge.”

Watson thinks if there is a bias, it’s “person to person” and not systemic.

“The point is that until you have the evidence, the data, you just don’t know which way the bias is going,” he said. “I could give instances where I think courts, tribunals have been very, very biased in favour of refugees.”

Although Watson said collecting data on the decisions made by individual judges was useful, he was not sure it was the best use of the statistics which is more for “micromanaging figures to micromanage problems”.

“This big data set could prove other bigger issues,” he said.

Ghezelbash said collecting evidence on Australia’s refugee and asylum seeker review process allowed researchers to identify why previous government efforts to reform the system to fast track the process had failed.

“Often we talk about the fact that we don’t have a Bill of Rights … but what we do have is the constitutionally entrenched right to seek judicial review for decisions made by government officials,” he said.

Part 2: Zahra's mission to help refugee community is personal