2.1.1 Comment

Ethnic disparities in apprehension figures are of course consistent with both explanatory approaches discussed in this paper. However, the possibility that certain sub-groups of the population are more susceptible to Police stopping and checking is a reasonably well-researched issue. A considerable number of arrests follow from Police stopping and questioning individuals in public places. From the perspective of “amplification”, differences in apprehension could therefore result from some bias in policing practice.

Younger age, certain styles of clothing and headgear, hairstyles, type of car driven, and incongruity of person with time or location are all known to trigger Police suspicion. Research conducted in the UK indicated that ethnicity also influenced Police officers’ decisions to stop and question potential offenders 1. One US study 2 found that African American male drivers were more likely than other groups to be stopped by the Police. Similar findings emerged in some UK studies 3, although one 4 found no such effect. Its authors commented that this may not be surprising, given that many stops are made at night when the appearance of the driver might not be necessarily be visible to Police.

Interestingly, traffic offending (excess breath/blood alcohol, disqualified driving, dangerous driving, etc) here in New Zealand, as well as drug offending, are classes of offending in which the vast majority of offenders are apprehended directly as a result of Police activity, rather than by way of crimes being reported by the public. Were Police bias an important influence affecting number of offenders arrested, then arguably Māori disproportionality should be more pronounced in traffic and drug offence figures. However, according to Ministry of Justice data, while still making up a significant number of those arrested, the actual proportions made up by Māori of all offenders convicted for both traffic (37%) 5 and drugs (40%) offences are in fact the lowest across the main offence classes (e.g., comparable figures for violence are 47%, and dishonesty offences 48%). This does still beg the question as to what the relative detection rates of different types of underlying crime are, however. Ferguson’s 2003 study cited above suggests that Māori are more likely to be apprehended for cannabis use than non-Māori, but we are not aware of comparable work on other types of offences.

Responses of suspects stopped by the Police similarly may cue officers to search suspects and/or their vehicle. Again, research has shown that certain behaviours – facial expression, gaze aversion, speech disruptions and hand gestures – are liable to be interpreted by Police officers as “suspicious” 6. Some behaviours of this type may however be culturally influenced. In a naturalistic study in the USA, African American and Hispanic people showed significantly higher levels than Europeans of behaviours that were interpreted as suspicious 7. Cultural differences in degree of eye contact have been found in New Zealand 8 .

The extent to which the detained person cooperates with the Police officer may also be important. A 1998 study suggested that Māori held relatively negative attitudes towards the Police generally: perceptions of bias were common amongst both Māori and Police officers themselves 9. It is conceivable that this negativity may motivate hostile and uncooperative responses when in direct contact with the Police. Such behaviour could in turn increase the likelihood that the Police took the matter further 10. Family experience and circumstances can also bear on this, underlining the complexity of interactions between official agencies and suspected offenders. Work by Canadian criminologists based on research with youth living on the streets of Toronto and Vancouver, suggested that, for youth who had suffered from abuse within their family, subsequent police sanctions “more often result in criminal acts of defiance than in deterrence”11 .

Further, Police engage in offender “profiling”, both explicitly and consciously as an aid to crime resolution, and implicitly (even unconsciously) as they individually go about their work. Given the salience of ethnicity as a personal characteristic, successes achieved (at least in part) as a result of ethnicity-based profiling inevitably reinforce the tendency to pay closer attention to that characteristic in future. Similarly, the attention of individual officers inevitably is drawn towards specific individuals already known to them, from which the same circularity of cause and effect can ensue: Police become acquainted with significant numbers of Māori offenders through their arrest and prosecution, which inevitably influences subsequent Police attentiveness to those same individuals. This may partly explain a consistent finding from the Department’s annual recidivism analyses, which reveal rates of reconviction for Māori that are around 20% higher than those of NZ Europeans 12  .

Apprehension rates will also be affected by the extent to which crime is reported – and those reporting crime may be strangers, passers-by, teachers, social service agency officials, victims or family and friends. A subsequent exercise of discretion may occur as to whether a report proceeds to be recorded as an official complaint, and then onto subsequent arrest and charges. A large number of factors can affect the discretion being exercised at each point by the various parties involved.

One obvious perspective that has not yet been explored - which would shed light on the key perspectives examined in this report - is the victim survey, with respect to victims’ report of offender ethnicity. For example, if court data indicated that 50% of all persons convicted of aggravated robbery were Māori, but that just 30% of those victimised in this manner reported that the offender was Māori, then it might reasonably be inferred that some form of bias was intruding at the arrest or prosecution stages. Of course, such an analysis is problematic, given the uncertainty which surrounds crime victims’ ability to identify the ethnicity of someone with whom they interact only fleetingly, who may be masked, and in a situation involving extremes of emotion. Further, given the frequency of media reports of crimes where the offender was described by witnesses simply as “Polynesian”, the problems are compounded. The difficulties outlined in section 1.1.3 above in having a consistent definition of ethnicity would also be significant.

In any case, such data is not yet available in New Zealand. However, studies of this nature have been undertaken in the United States (US), where similar problems of over-representation apply to black Americans. In the US, African Americans constitute about 12.8% of the population, but in year 2000 accounted for 38% of the arrests for violent crimes, and 31% of the arrests for property crimes 13. Two reasonably comprehensive studies, one in 1978, and a more recent one in 2003, have addressed the question of difference in ethnic involvement between actual crime, and official crime statistics 14,15, . The first study used official crime data (Uniform Crime Rates 16 ) which were analysed in conjunction with annual crime victim survey data. The second used the more modern National Incident-Based Reporting System (NIBRS), which gathers information from individual crime reports recorded by Police officers at the time of the crime incident. A wide range of different criminal offenses are reported on, with information collected on victim and offender demographics, victim/offender relationship, time and place of occurrence, weapon use, and victim injuries.

Both of these studies set out to determine whether the over-representation of African Americans in criminal arrest data was a consequence of their differential involvement in crime, or whether it reflected racially biased reporting of crime, or Police practice. Significantly, both studies reached the same conclusion - that there was no basis to conclude that the disproportionate representation of Africa Americans in crime statistics was a consequence of anything other than their unusually high rates of participation in crime.

Finally, a recent Australian study is also worth mentioning 17 . The researchers addressed a very similar question to that which is part of the focus of the current paper - is over-representation of Aboriginal people in prison the result of systemic bias in policing, the law, and/or the operation of the criminal justice system, or is it the result of high rates of Aboriginal involvement in crime? The report traverses data similar to the type covered in the first half of this report, such as disproportionality in arrests, convictions, etc. They also present data which reveals that, on average, Aboriginal offenders tend to have longer criminal histories. However, the authors of this report had access to a unique form of information, data jointly gathered by the NSW Bureau of Crime and NSW Health Department, which are derived from periodic representative sample surveys of secondary school students. These surveys seek to ascertain self-reported involvement in crime. The survey data clearly indicate that the prevalence of self-reported involvement is considerably higher amongst Aboriginal juveniles than non-Aboriginals. There were particularly striking differences in the self-reported frequency of crime - across most crime categories, the percentage of Aboriginal respondents who report having committed more than five offences in the previous 12 months is more than double that of the non-Aboriginal respondents. The authors conclude therefore that arrest and conviction rate differences are simply reflective of real differences in rates and patterns of involvement in crime.

It should be noted that there have been important critiques made of aspects of this study, particularly in terms of the policy claims and prescriptions made by Weatherburn et al, and the risk of adopting a simplistic view that over-representation must be a result of either justice system bias or actual offending levels 18 .

In summary, this brief review of apprehensions figures, and relevant research, offers a mixed picture. Evidence does exist to indicate that apprehension rates do not simply reflect actual offending behaviour of persons in the community. Instead there is some support for the notion that the interactions between Police and Māori on the “front-line”, as well as social interactions within families or communities, lead to an increased probability of Māori offenders being subject to criminal apprehension, independently of rates of actual offending. Some form of bias appears to be occurring – the precise magnitude of the effect, however, is less clear.