Crime and Race Statistics: Toward a Canadian Solution
Roberts, Julian V., Canadian Journal of Criminology
In this comment, I argue that routinely collecting and publishing crime-race statistics has more disadvantages than advantages. I believe that ethnicity data should be gathered only on a periodic basis, as part of the "special study" initiative of the Canadian Centre for Justice Statistics. I begin by looking at the U.S. experience, for that is the model advocated by several politicians in Canada, and in my view best illustrates the dangers of routinely collecting and distributing race-crime data.
The widely-distributed, annual publication, The Sourcebook of Criminal Justice Statistics, provides information on several aspects of criminal justice in the U.S., including a breakdown of the percentage of arrests in which the suspect was black, white or Amerindian. A recent edition of this publication (U.S. Department of Justice 1989) notes that blacks accounted for 30% of arrests for a list of offences (Table 4.9). On the same page, census data are provided showing that 12% of the U.S. population are black. Anyone who can subtract 12 from 30 can arrive at the "over-representation" of black Americans in arrest statistics.
The racial statistics sit as comfortably on the page as a statistic such as the percentage of arrests that involve individuals with a criminal record. Taken at face value then, the colour of a person's skin (an "ascribed" characteristic(2) is as important to the system as whether he is a recidivist (an "achieved" characteristic). The fact that this kind of presentation passes without comment in the U.S. reflects how inured Americans have become to the concept of race, particularly the distinction between black and white. Skin-colour (since almost all arrests involve native-born Americans rather than naturalized residents, country of origin is of little relevance) is a central defining personal characteristic in American society. It therefore does not seem unnatural to present white/non-white arrest statistics.
Canadians are, in my opinion, less inclined, on two principal grounds, to define themselves in this way. First, because black (and white) Canadians come from a variety of countries and cultures, and, second, because Canadians do not have a history of classifying people in terms of skin colour.
1. What use are race-crime data?
Several possible uses of crime-race data have been identified; not many stand up to scrutiny. Some law enforcement officials have argued that these data are useful to the criminal justice system because they permit the police to better allocate resources. This presumably means identifying neighbourhoods that have high crime rates. This can be done, indeed is done without the necessity of first establishing the racial composition of the area. As well, it is of little use in neighbourhoods that are racially mixed, as is the case in several Canadian cities. A second argument made by enforcement officials is that ethnicity data facilitate the investigation of a case. Clearly the police need as much information as possible to help identify, locate, and arrest a suspect. But this is independent of whether we need this kind of information collected and distributed throughout the system, and to the general public. Publishing racial statistics in the annual "Canadian Crime Statistics" is hardly going to facilitate the investigation in any individual case.
Some people have argued that knowing about the Canadian crime rates of different ethnic groups somehow helps immigration officers screen new applicants from around the world. This is a slippery slope: it implies a priori categorizations and differential admission (or investigative) thresholds. Do we really want immigration authorities to conduct more rigorous investigations of people applying for entry to Canada from different parts of the world? It seems to me that principles of equity are violated if this policy is followed. As well, curious screening strategies might emerge. Would this not suggest we pay more careful attention to applicants from Palermo rather than from Florence, on the reasoning that the former are more likely to have links to organized crime? …