Black Lives Matter: Why are Indigenous people more likely to be in custody in Australia?

This week, I’ve been looking for evidence that the Australian justice system is racist against Aboriginal people. Given Indigenous people are locked up at 11 times the rate of non Indigenous people, racism seems likely to be at least some of the reason. Where statistics exist (and I’ve found some), they often show racism. But they are hard to find.

After my last post on Australian statistics on Indigenous deaths in custody,  I’ve been asked whether we do have a racist justice system, or maybe its just that Indigenous people commit more crimes? How much really is bias in the system?

This line of argument is best encapsulated by Tony Abbott’s comments last weekend (described as incorrect in this article):

“Obviously the Indigenous incarceration rate is much higher than the general incarceration rate. That shouldn’t be so unless there’s evidence that courts are more likely to imprison Indigenous offenders than non-Indigenous offenders, and there is none,[my emphasis]” he told The Sun-Herald in his first substantive comments on the protest movement sweeping the globe.

“The higher Indigenous incarceration rate is a function of the higher Indigenous offending rate. It is absolutely tragic. But we know that when it comes to domestic violence and a lot else, this is much worse in Indigenous communities than most parts of the country.”

Indigenous incarceration rates that are more than ten times higher than population incarceration rates don’t, of themselves, prove bias in the system. They are very strongly suggestive, but they aren’t proof.

So this week I’ve been looking for proof. This article in the Guardian describes the data problems much better than I could, but despite a number of the Royal Commission recommendations being about better data, it is still very hard to find.

Despite the ramifications of over-policing, there remains a paucity of data, publicly at least. Most police forces do not release detailed breakdowns of their interactions with Indigenous Australians, NSW being a notable exception.

In Victoria, despite a commitment to begin mandatory recording of ethnic appearance descriptors in police interactions with the public, the force has refused to make the data public. In a statement this week, Victoria police told the Guardian there was “considerable community concern and risk associated with the reporting of ethnicity data”.

It is indisputable that an Indigenous person is worse off economically, educationally, and health wise than a non Indigenous person in Australia. That’s why we have Closing the Gap targets, and some of those Gaps are large, with life expectancy implicitly being a summary of many of the health and economic gaps:

In 2015–2017, life expectancy at birth was 71.6 years for Indigenous males (8.6 years less than non-Indigenous males) and 75.6 years for Indigenous females (7.8 years less than non-Indigenous females).

So how much of the difference in statistics for Incarceration and deaths in policy custody come from the underlying disadvantage of Indigenous people (which is probably also partly due to systemic racism), and how much comes from further racial prejudice in the judicial system?

I’ve been looking for evidence this week. It is hard to find. And not just because the judicial system appears reluctant to release statistics. The ideal statistical evidence would be a measure of the actual crimes committed by all Australians (of all racial groups) and then a comparative look at the progress of each group through the judicial system – from initial attention from the police, to being charged with a crime by police, to being convicted of a crime, and then the level of sentencing. Those first two steps of initial attention and being charged, are by their nature, almost impossible to find an underlying population level of statistics.

• Initial attention from the police – if we are looking (for example) at public drunkenness (which is a crime in Queensland) – without a measure of all the people who are drunk in public, it is impossible to be definitive about whether Indigenous people are more likely to be charged with it. Given how few non Indigenous people are arrested at sporting events for public drunkenness, I find it easy to believe that Indigenous people are disproportionately targeted for this offence. But nobody is sitting in public places counting the people who are drunk in public to work out the proportion of Indigenous and non Indigenous drunk people who are arrested.
• Being charged with a crime after coming to police attention – similarly unless there is public data tracking all the interactions the police have, by Indigenous status, it is very difficult to track whether someone who has a conversation with police is more likely to be charged if they are Indigenous, rather than not Indigenous.

Here’s the limited range of evidence I’ve found so far:

• NSW police let nearly four times as many non Indigenous as Indigenous offenders off with cautions after being caught with small amounts of cannabis
  

  The data shows police were four times more likely to issue cautions to non-Indigenous people. In the five years to 2017, only 11.41% of Indigenous Australians caught by police with small amounts of cannabis were issued cautions, compared with 40.03% of the non-Indigenous population.

• While non-Indigenous people had a similar conviction rate for drug offences, evidence suggests Indigenous people receive harsher sentences, for example this analysis of data in NSW for women convicted of drug offences:
  

  A global report prepared for Penal Reform International used Bocsar data from 2013 to 2017 to find Indigenous women were more likely to receive harsher prison sentences for drug possession offences, “including triple the rate of prison sentences”.

• This article looks at strip searches in NSW and concludes that they are disproportionately of Indigenous people:
  

  Twelve per cent of strip-searches in the two year period were Indigenous people, even though they only make up a little more than three per cent of the state population. It included one 10-year-old and two 11-year-olds.

And this article from the Economist shows a similar example for the US (where there is much richer data available):

You don’t need a degree in statistics to believe that racial disparities plague American law enforcement. Of every 100,000 black adults, 2,300 are incarcerated—five times the rate for whites. This gap is not proof of discrimination: blacks could be five times as likely to break the law.

The graphs in this article show one study that The Economist quotes on how prosecutors use cut offs in the legal sentencing rules disproportionately to drive higher sentences for Black and Hispanic drug convictions.

Mr Tuttle finds that only a minority of prosecutors (around 20-30%) display this bias. These officials tend to work in states with above-median rates of searches on Google for racial slurs, suggesting racism is more common in their regions. In other states, “bunching” at 280-290g did occur, but affected blacks and whites equally.

An incarceration rate for Indigenous people 10 times higher than non Indigenous people seems very likely to indicate racism in the system. And it does exist. Several specific examples show racism does exist for particular police interactions and crimes. But statistics aren’t well kept, or available, and even if they were, not all aspects of the stages of the justice system are amenable to statistical analysis.