It is possible to point to areas where there have been significant cultural changes in the past 20 years. A concerted effort has been made through various bodies responsible for judicial education to ensure a greater understanding of the complexity of the issues that are coming before the courts. Evidence of the eventual success of this cultural shift in judicial attitudes is perhaps seen in recent research by Dr Samantha Jeffries and Dr Christine Bond at the Queensland University of Technology. It shows that the situation that existed at the time of the commission, where indigenous people would get harsher sentences than non-indigenous offenders for the same crime, seems to have changed in some jurisdictions to sentences being about on par.

There continue to be cases that will highlight particular instances where — like the recent case of the killing of Aboriginal ranger Kwementyaye Ryder in Alice Springs — indigenous families and communities feel aggrieved by the judicial process.

Another improvement within the criminal justice system has been the employment of more indigenous staff in courts and prisons but, perhaps even more profoundly important, has been the exploration of alternative forms of community-based sentencing. Examples of this include courts where elders from indigenous communities sit with magistrates throughout the criminal process, such as the Koori courts in Victoria and circle sentencing models that have been employed in the ACT and parts of NSW.

But these significant areas of gain have failed to fundamentally shift the entrenched problems of over-representation. While indigenous people remain statistically the poorest sector of the community – compounded by the inevitable lower levels of education and employment that are a part of the poverty cycle – poverty-based crime will continue to be an issue. One sobering figure that underscores the broader implications of this statistic is that, according to Australians for Native Title and Reconciliation (ANTaR), the recidivism rate for indigenous offenders is 66 per cent (much higher for juveniles) and the retention rates through high school are 46.5 per cent. Indigenous people return to prison more frequently than indigenous students stay in school.

And the other barrier in the system to reducing over-representation of indigenous people is the inevitable incompatibility between that goal and tough law-and-order campaigns, for instance laws introduced in NSW that toughen the criteria in which bail can be given. A large number of the indigenous people are in jail today because they were deemed ineligible for bail. Requirements about permanent addresses disadvantage homeless and transient populations – and young people whose homes are not safe places for them to be. Another significant factor is the non-payment of fines, a consequence of too-few sentencing options for members of the judiciary, especially outside of the major cities.

But the real frustration for many indigenous people is the continuing number of deaths in custody that occur in circumstances that fuel suspicion within indigenous communities such as the death on Palm Island of Cameron Mulrunji Doomadgee in 2004. This is a matter of justice not just being done but being seen to be done.

The Royal Commission carefully documented the deep history that Aboriginal communities have with police. It acknowledged that this was a complex relationship and that many, particularly new recruits to the police force, were unaware of how deep those tensions ran, particularly in rural communities.

The commission still provides a blueprint for much of what could be done to improve the entrenched problem of over-representation of indigenous people in the criminal justice system. A review of what has been implemented, what hasn't, what is working and what isn't would be a fitting follow-up on this landmark anniversary. It is also time to tackle the issue of investigating deaths in custody with more transparency and less appearance of bias. But serious consideration also needs to be given to forming an independent statutory authority responsible for investigating allegations of assault or death against any person – black or white – in custody. It needs to have investigative powers and consequences for officers who refuse to assist. This is not a new concept and it has worked successfully in other countries.

The Police Ombudsman Northern Ireland has been described as the ''golden standard'' in investigations. The procedure includes being at the scene of a crime within an hour, interviewing all civilian and police witnesses and using independent scientists and medical experts. It is a way of ensuring all deaths in custody are investigated in a way that avoids the perception of conflict of interest. Moreover, it gives the families and communities of people who die in custody some confidence that the process is impartial — and would also be a real force for change.

Professor Larissa Behrendt is the chair of Indigenous Research at the University of Technology, Sydney. Her current research program includes a project with the NSW Attorney-General's Department investigating factors that explain differences in crime rates in different indigenous communities.