Select Page

The case for judging bias through track records

23 September 2022

In June last year, when the Australian Law Reform Commission was deep into its inquiry on judicial impartiality, it received a submission outlining the track record of the judges of the Federal Court when dealing with disputes between unions and employers.
The commission declined to publish that document and issued a statement saying it had taken legal advice that the submission, produced by the Menzies Research Centre, was defamatory.

That must have given heart to those who believe there is nothing to be gained by assembling statistics on particular judges because the outcome of every case depends on unique circumstances.

Joe Public might believe that a judicial track record of favouring one side of particular disputes is enough to give rise to a reasonable apprehension of bias – which is the common law test that requires judges to recuse themselves.

That, however, is not how the law works – at least for now.

Yet the Menzies Centre believed there was still a good case for taking account of the track record of individual judges: it promotes transparency and would help the courts allocate cases in ways that might avoid potential problems.

It is worth recounting what was contained in that suppressed submission.

The Menzies Centre analysed 109 Federal Court disputes between unions and employers and found unions were successful in 52 of those matters – 48 per cent – which is close to half of the matters that came before the court.

But one judge – who was identified in that submission – was an outlier. He found for unions in 10 of the 11 matters that came before him.

While that is a very small sample, it amounts to a 91 per cent success rate for unions, compared to 48 per cent for the court as a whole.

Despite what happened last year, traditionalists are in for a shock: the commission’s final report on judicial impartiality has sided with those – like the Menzies Centre – who believe there is a case for subjecting judges to statistical analysis.

The commission wants the Commonwealth courts to take control of the process to ensure this exercise is meaningful.

Recommendation 13 says: “The Commonwealth courts (individually or jointly) should develop a policy on the creation, development and use of statistical analysis of judicial decision-making.”

The commission’s report says this is not intended to directly affect the law of bias but is intended to “promote reflection within the courts” and enable them to engage effectively when other parties assemble their own data on the track record of judges.

Compare this to the suppressed submission. It had recommended that “court staff compile and report in real time the judgment records of judges in specific areas of law and the parties involved in the case”.

Nick Cater, the centre’s executive director, found it “gratifying the commission has adopted our reform recommendation. But it makes it even more puzzling that they chose to censor our submission.”

The big issue left unresolved by the commission is whether the courts should publish their data – as proposed by the Menzies Centre – or keep it to themselves.

The Menzies Centre has shown that anyone with a computer and enough time can assemble statistics on the track record of judges.

The issue, therefore, is not whether this sort of data should be made public, but whether the courts are happy for others to be the prime source of data about the judiciary.

The reluctance in some quarters to publish statistics on judicial officers is reminiscent of the way in which some judges were horrified by the move to produce regular figures showing the nations’ slowest courts.

That exercise is now the responsibility of the Productivity Commission, which publishes reliable data every year outlining the backlog of cases in the Supreme Courts. The sky has not fallen.



And the Productivity Commission draws on data from the courts.

In a data-rich nation, the growth of judicial analytics is inevitable – unless, of course, the federal government follows the example of France which, according to the commission, has made it a criminal offence to publish some forms of analysis about that country’s judiciary.

So unless the judiciary accepts the commission’s recommendation, it will be vacating the field and leaving it to others to identify those judges with track records of note.

That, however, leaves unaddressed the current law of bias which seems almost immune to the reality of a skewed judicial track record.

One of the most powerful voices for reform in this area is legal academic Matthew Groves, who is a professor of law at Deakin University and takes a realistic view of what an apprehension of bias means.

In a 2020 article for the Australian Institute of Administrative Law, he wrote that statistics on the track record of judges are sometimes “so extreme, so one-sided, that their sheer weight alone might say something even in the absence of detailed analysis of the cases that comprise the statistical set. The difficult question that follows is whether courts can even conceive of that possibility, let alone hear it”.

For the record, the latest figures from the Productivity Commission’s report on government services show that in the year to June 30 last year – the peak of the pandemic – not one of the nation’s Supreme Courts met the benchmark of having no more than 10 per cent of criminal cases still before the courts after 12 months.

The Queensland Supreme Court came closest to hitting that benchmark with just 12.5 per cent of criminal matters still before the courts after a year.

 Published in the Australian Newspaper