Data, denials, and confusion: the racial profiling debate in Toronto.

• Author: Wortley, Scot

On 19 October 2002 the Toronto Star began publication of a series of articles on the controversial topic of race and crime. In addition to reviewing previous Canadian research on this issue, the Star provided its own, original analysis of data derived from the Toronto Police Service's Criminal Information Processing System (CIPS). The CIPS data set, received from the police through a freedom of information request, contains information on over 480,000 incidents in which an individual was either charged with a crime or ticketed for certain types of traffic offences. (1) The final data set is supposed to represent the total population of criminal charges (approximately 800,000) laid by the Toronto Police Service from "late" 1996 until "early 2002" (see Rankin, Quinn, Shephard, Simmie, and Duncanson 2002a).

The Star's analysis revealed that black Torontonians are highly over-represented in certain charge categories--including drug possession. The Star maintains that this pattern of over-representation is consistent with the idea that the Toronto police engage in racial profiling (Rankin et al. 2002b). The Star's analysis also reveals that blacks may be treated more harshly after arrest than their white counterparts. In particular, white offenders are more likely to be released at the scene, while black offenders are more likely to be detained and taken to the station for processing. Furthermore, once at the station, black offenders are much more likely than whites to be held in custody for a bail hearing. The Star maintains that these racial differences remain after other relevant legal factors have been taken into statistical account (Rankin et al. 2002a).

In response to the Star series, the Toronto police vehemently denied all allegations of racial bias. Chief Fantino declared that "[w]e do not do racial profiling ... There is no racism ... We don't look at, nor do we consider race or ethnicity, or any of that, as factors of how we dispose of cases, or individuals, or how we treat individuals" (quoted in "Racism" 2002: A14). Craig Brommel, president of the Police Association, stated in a news release that "[n]o racial profiling has ever been conducted by the Toronto Police Service and we question the Toronto Star's interpretation of its statistical information" (quoted in Porter 2002: A6). These sentiments were echoed by several local politicians. Mayor Lastman, for example, declared that: "I don't believe that the Toronto police engage in racial profiling in any way, shape or form. Quite the opposite, they're very sensitive to our different communities" (quoted in "Analysis" 2002: A9). Even Norm Gardiner, chair of Toronto's civilian--and supposedly non-partisan--police oversight board, claimed that he was confident that the Toronto police did not engage in racial profiling. In fact, he went so far as to suggest that the researchers involved in the Star series just wanted to stir up controversy during a period of excellent police-minority relations: "Some of the people involved, who keep on bringing this stuff up ... they make a living out of social unrest" (quoted in "Analysis" 2002: A9).

Nonetheless, despite all of their denials, police officials provided no systematic critique of the Star's analysis in the four months following the release of the race-crime series. Nor did they provide any new information that might challenge the racial-profiling hypothesis. All that changed, however, at a Police Services Board meeting held on 20 February 2003. On this occasion Chief Fantino produced his own "experts" on racial profiling, who subsequently slammed the Star report as being based on faulty research. Edward Harvey, the University of Toronto sociology professor hired by the police to conduct the re-analysis of the Star data, concluded that his "independent review results do not provide evidence of systemic racial profiling being practiced by the Toronto Police Service" (Harvey 2003; see also Granatstein 2003: 22). Lawyer Alan Gold added that "the Star articles are what we would call junk science" (Moloney 2003: A1). Although newspaper headlines the next day screamed that "allegations of racial profiling are junk science" and "Study finds races treated equally" (Blatchford 2003: A1), Harvey did not actually release his full report to the public until more than a month after his well-publicized presentation. This strategy served to prevent interested academics from commenting on his work in the immediate aftermath of his controversial statements. However, on the day of the presentation, the Police Service Web site did provide an "executive summary" of the Harvey report. This summary provided very few details about Harvey's actual re-analysis of the data, but it did include the following text:

The Star's allegations of systemic police bigotry, based upon claims of statistical disparity in the particular examples it has publicized, are scientifically unsound and an unfair selection of the available data. in a word they are bogus, bogus and bogus.... The truth is that the Star's conclusions are simply false based upon the data that they had and the Star's mistakes in considering the matter are fundamental, basic and simply embarrassing. Their whole project is fundamentally flawed. Their articles will enter the junk science hall of fame. ("Text" 2002: A 19) It is the purpose of this article to review--in detail--Edward Harvey's re-examination of the Star data and his conclusion that racial profiling does not exist in the Toronto area. It is our opinion that, in his report, Professor Harvey (1) does not properly define racial profiling and totally ignores the published criminological literature on this topic; (2) makes several incorrect and/or misleading statements designed to discredit the Star's analysis; (3) engages in questionable "data cleaning" procedures that may dramatically reduce racial disparities in the arrest statistics; (4) provides neither a transparent nor a complete re-analysis of the Star's major findings; and (5) provides no concrete evidence that can disprove the Star's allegations of racial bias. Indeed, many of the findings produced by Harvey are completely consistent with the racial-profiling argument. We conclude with a brief discussion of what this racial-profiling debate--a debate fought largely in the media rather than academic journals--could mean for the reputation of criminology in Canada.

Research on racial profiling

In his report, Professor Harvey claims to provide evidence that systematic racial profiling does not exist in the Toronto area (Harvey 2003: 39). In fact, Harvey does not actually examine racial-profiling data at all. Indeed, his entire report is based on an examination of the Toronto police arrest data set (CIPS). It is our contention that such arrest data are only produced after racial profiling has already taken place. In the criminological literature, racial profiling is said to exist when the members of certain racial or ethnic groups become subject to greater levels of criminal justice surveillance than others. Racial profiling, therefore, is typically defined as a racial disparity in police stop and search practices, racial differences in customs searches at airports and border-crossings, increased police patrols in racial minority neighbourhoods and undercover activities, or sting operations that selectively target particular ethnic groups (see Weitzer and Tuch 2002; Meehan and Ponder 2002; Engel, Calnon, and Bernard 2002; Harris 1999). Racial profiling, therefore, is associated with racial bias in police investigation--not racial bias in arrest decisions or racial bias in police treatment after arrest. This is not to say that arrest statistics do not reflect profiling. For example, the over-representation of blacks in Toronto arrest statistics could mean that blacks are, indeed, subject to greater police surveillance. However, it could also mean that blacks are simply more involved in certain criminal activities. Thus, the racial-profiling hypothesis cannot truly be proven or rejected unless we first examine information on police surveillance activities. Harvey, unfortunately, makes no attempt to look at such data.

Do black people come under greater criminal justice surveillance than people from other racial backgrounds? Are black people more likely to be stopped, questioned, and searched by the police? Police data from both England (Bunyan 1999) and the United States (see Engel et al. 2002; Harris 1997) suggest that they are. In England, for example, the passage of the Police and Criminal Evidence Act (PACE) gave the police the authority to stop and search persons or vehicles on the reasonable suspicion that they would find drugs, stolen goods, or other prohibited items. However, the PACE legislation also mandated that the police make a written record of the racial background of all people who were subjected to police stops and searches. Police statistics from 1997-1998 reveal that black people were stopped and searched at a rate of 142 per 1,000, compared to 45 per 1,000 for Asians and 19 per 1,000 for whites. Overall, the English data suggest that blacks are approximately eight times more likely to be stopped and searched by the police than whites (Bunyan 1999; Brown 1997).

Unfortunately, unlike England and the United States, the police in Canada are not required to record the race of the people they stop and/or search. Thus, official police statistics cannot be used to investigate the presence or absence of racial profiling in this country. However, a number of field studies have uncovered evidence that racial profiling may exist. For example, James (1998) conducted intensive interviews with over 50 black youth from southern Ontario. Many of these youths reported that being stopped by the police was a common occurrence for them. James concludes that the adversarial nature of these police stops contributes strongly to black youths' hostility...