Does Racial Bias in Sentencing Contribute to Indigenous Overrepresentation in Prison?

AuthorDon Weatherburn,Lucy Snowball
Published date01 June 2007
Date01 June 2007
Crim40-3 text-final.x Does Racial Bias in Sentencing
Contribute to Indigenous
Overrepresentation in Prison?
Lucy Snowball and Don Weatherburn
NSW Bureau of Crime Statistics and Research, Australia
The possibility of racial bias in the criminal justice system is a recur-
ring concern in Australia, as it is in other countries with high rates of
minority overrepresentation in prison. In this country, however, very
little research has been conducted on racial bias in sentencing. This
article reports the results of a logistic regression analysis designed to
test for racial bias in the sentencing of Indigenous adult offenders. Our
results show that the unadjusted risk of imprisonment is much higher
for Indigenous than for non-Indigenous offenders but this effect progres-
sively diminishes as more sentence-relevant factors are taken into
account. Indigenous status remains a significant predictor of imprison-
ment when all measured legal factors have been taken into account but
in the median case the difference between Indigenous and non-
Indigenous offenders in the risk of imprisonment is less than one
percentage point. Sentencing courts were found to place less weight on
the prior criminal record of an Indigenous offender than on the prior
criminal record of a non-Indigenous offender. The implications of the
study for future research and policy are discussed.
In the 1990s concern about the high number of Indigenous deaths that were occur-
ring in police and prison custody in Australia led to the establishment of a Royal
Commission (Commonwealth of Australia, 1991). The Commission found that
Indigenous Australians were no more likely to die in custody than non-Indigenous
Australians. The high number of deaths in custody turned out to be a direct reflec-
tion of the gross overrepresentation of Indigenous Australians in police and prison
custody. Most Australian States and Territories have undertaken significant crimi-
nal justice reform in an effort to reduce the rate of Indigenous imprisonment
(Cunneen & McDonald, 1996). Instead of declining, however, the rate of
Indigenous imprisonment increased (Australian Bureau of Statistics, 2006; Biles,
1992). The gap between Indigenous and non-Indigenous imprisonment rates in
Australia is now larger than the disparity between African-American and white
imprisonment rates in the United States.1
Address for correspondence: Don Weatherburn, Director, NSW Bureau of Crime Statistics
and Research. E-mail:
VOLUME 40 NUMBER 3 2007 PP. 272–290

The question of why this is so is of fundamental importance to policy. If
Indigenous overrepresentation in prison is partly the result of racial bias in the
response of the criminal justice system to Indigenous defendants, policy makers
need to consider ways of reforming the justice system to reduce or remove the racial
bias. If, on the other hand, Indigenous overrepresentation in prison is not a reflec-
tion of racial bias, or if racial bias exerts only a small impact on the risk of imprison-
ment, then policy makers need to look outside the justice system for ways of
reducing the rate of Indigenous imprisonment. Suggestions of racial bias in the
criminal justice system have been a recurring theme in the Australian literature on
Indigenous imprisonment. The Royal Commission into Aboriginal Deaths in
Custody (Commonwealth of Australia, 1991) placed most of its emphasis on
Indigenous disadvantage and substance abuse as the principal reasons for Indigenous
overrepresentation in prison but also highlighted a number of areas where, in its
opinion, institutional bias in the criminal justice system also played a role. These
areas included bias against Indigenous offenders in the willingness of police to
employ alternatives to arrest (Commonwealth of Australia, 1991, paras:
30.2.2–30.2.14; 14.14.8–14.14.9; 21.1.7; 30.2.2–30.2.7), lack of community-based
alternatives to prison in rural communities (Commonwealth of Australia, 1991,
para: 22.4.11), inadequate funding of Aboriginal legal aid (Commonwealth of
Australia, 1991, paras: 22.4.52–22.4.75) and excessively punitive sentencing
(Commonwealth of Australia, 1991, paras: 14.4.32–14.4.38, 22.3.3).
The idea that the criminal justice system is biased against Aboriginal people has
also received a good deal of attention from Australian criminologists. In the 1980s,
Ronalds, Chapman and Kitchener (1983), Milne (1983) and Cunneen and Robb
(1987, p. 220) all drew attention to the overrepresentation of Aboriginal people
among those arrested for minor public order offences. In the 1990s, Cunneen (1992,
p. 1) argued that there is a ‘widespread and long-held view that judicial racism is a
problem’, while Craigie (1992, p. 1) criticised the Royal Commission for failing to
address what he called ‘the whole question of judicial bias’. In South Australia,
Gale, Bailey-Harris and Wunderitz (1990, p. 7) attributed Indigenous overrepresen-
tation in the South Australian custodial institutions to ‘the compounding effect of
discrimination suffered at earlier steps in the criminal justice process’. Two years
ago, Blagg, Morgan, Cunneen and Ferrante (2005, p. 15) argued that ‘systemic
racism is a factor of established significance’ to Indigenous overrepresentation in the
criminal justice system. Most recently, Cunneen (2006, p. 340) has questioned
whether Indigenous young people are being treated equitably in relation to diver-
sion from the juvenile criminal justice system.
At first blush, the Australian criminal justice system does appear to treat
Indigenous offenders much more harshly than non-Indigenous offenders. Harding,
Broadhurst, Ferrante and Loh (1995), for example, found that Aboriginal people
made up just 2.7% of the Western Australian population in 1993 but accounted for
20% of all arrests. They also found that 44.1% of Aboriginal juvenile offenders
received a detention sentence, compared with just 25.6% of non-Aboriginal
juvenile offenders. In New South Wales (NSW) Indigenous defendants appear in
court on criminal charges at a rate that is 13 times higher than that of their non-
Indigenous counterparts (Snowball & Weatherburn, 2006). A convicted Indigenous
offender with no prior record of imprisonment is 2.5 times more likely to be impris-

oned in NSW than a convicted non-Indigenous offender with no prior record of
imprisonment. Indigenous offenders in NSW who have been previously imprisoned
are three times more likely to be imprisoned than non-Indigenous offenders who
have been previously imprisoned (Snowball & Weatherburn, 2006).
Sentence disparity by itself, however, does not necessarily indicate that
Indigenous defendants are being subjected to discriminatory treatment by the
courts. To reach this conclusion it is necessary to show that differences in the treat-
ment of Indigenous and non-Indigenous offenders persist after taking into account
the factors that courts are allowed or required to consider when exercising their
discretion. Surprisingly, despite all the attention given to the possibility of racial
bias in the criminal justice system, no-one in Australia to date has examined the
question of whether racial differences in the risk of adult imprisonment persist after
controlling for relevant legal factors. The present study examines this question. The
structure of the article is as follows. In the next section we review past overseas and
Australian research on the issue of racial bias in sentencing. In the third section we
present the results of our analysis. In the fourth and final section we discuss our
findings and highlight their implications for public policy.
Past Research
Studies of racial bias in sentencing in the United States from the 1930s to the 1960s
showed significant disparities in sentence outcomes for white and minority defen-
dants, a result which, at the time, was widely viewed as evidence of racial discrimi-
nation in sentencing (Spohn, 2000). Concern about racial discrimination in
sentencing played a major role in shaping reforms during the 1970s and 80s
designed to restrict judicial discretion in sentencing (e.g. determinate sentencing,
sentencing guidelines, ‘three-strikes’ sentencing legislation). In the 1970s and 80s,
however, it became clear (Hagan, 1974; Hagan & Bumiller, 1983; Kleck, 1981) that
much of the early research on racial bias in sentencing was methodologically
flawed, often lacking even basic controls for relevant legal factors, such as offence
seriousness and prior criminal record. Well conducted studies generally uncovered
little evidence of direct racial discrimination in sentencing once legal factors were
taken into account (Spohn, 2000). This led the National Research Council Panel
on Sentencing to conclude that the sentencing process in the United States is not
characterized by widespread discrimination (Blumstein et al., 1983, p. 93).
It now appears that this conclusion was premature and that racial bias is a
feature of the sentencing process in the United States, even if its manifestations are
subtler than previously assumed. Up until the late 1990s, researchers carrying out
multivariate analyses of the effect of race on sentencing tended to look only for
main effects of race. Steffensmeier, Ulmer and Kramer (1998) criticised this
approach. They argued that race is most likely to influence the sentencing process
when judges lack accurate...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT