Abstract
The recruitment of women and minority group members was intended to move Canadian police forces towards societal representation and to enhance services provided to, and improve relations with, women and racially marginalized groups. This review contemplates progress towards these goals at a time of extraordinary public dissatisfaction with Western policing. A rationale is offered for reconsidering the 50% representation target for women and it is emphasized just how little we yet know about racial bias in policing. The review ends with a call for rigorous, apolitical, research to untangle the complex interactions underscoring the considered questions within.
The social unrest of the 1960s prompted three categories of police reform: the ascendance of community-oriented policing (COP), greater civilian oversight and the creation of a more diverse institution through the recruitment of women and visible minority group members (Sklansky, 2006). The fallout from George Floyd’s death in Minneapolis, including the ‘defund the police’ initiative and a resurgence in Black Lives Matter–focused protests, marked 2020 as an annus horribilis for the policing institution, not just in the US but throughout the Western world (see Bourne, 2020). Now, 50 years after the initial reforms inspired by the civil rights movement, just how much progress has been achieved?
It is the diversification of police personnel, and its impact on the institution, with which this review is concerned. Although focused on Canadian reform, the experiences of other English-speaking Western democracies are drawn upon for context, comparison and the scarcity of Canadian-based studies. Moreover, US findings play a disproportionate role in this review not only because of their availability but because of the impact that the American experience has had on police practice, and perceptions thereof, throughout the Western world (Goff and Kahn, 2012).
Arguments for employment equity (aka positive discrimination and affirmative action) in police hiring fall into two categories: social justice and utilitarian. The justice rationale focuses on redressing historic discrimination by having workplaces better reflect disadvantaged groups in numbers proportionate to their societal composition (Morabito and Shelley, 2015). The practical goals with minority communities were to enhance cultural competency and promote trust and with women to improve the insight and empathy brought to victims. Moreover, women officers might also bring a less aggressive, more calming, style to conflict resolution (Rabe-Hemp, 2008).
Women in Canadian policing
Progress in representation
The history of women in Canadian policing parallels that of the US and the UK where more than a century has passed since women began to move from matron-only functions (Myers, 1993; Schmidt, 2011); it was not until the 1970s that women began to enter roles meant to be equivalent to those of men (Schmidt, 2011). Although initial aspirations were to achieve an equal female-to-male ratio, by 2019 only 22% of Canada’s 68,718 police officers were women (Statistics Canada, 2019). By comparison, by 2013, the UK stood at 27%, Australia 24% and New Zealand at 18% (Prenzler and Sinclair, 2013). The US lagged at 12% with continuing low representation in small municipalities and rural areas but with progress proximal to other western nations in urban areas (Archbold and Schulz, 2012).
After steadily increasing for two decades at about 0.75% per year on average in Canada, growth has slowed over the last decade. So, whereas the period from 1990 to 1999 saw an increase of almost 7% in female representation, cumulative growth over the last decade was just under 3% (Statistics Canada, 2019). This pattern of steady growth followed by a levelling off has also been observed in the US (Cordner and Cordner, 2011), Australia (Prenzler et al., 2010) and elsewhere (Prenzler and Sinclair, 2013). Although the recruitment of women remains a priority, with an even greater emphasis in some jurisdictions (Ward et al., 2020), this plateauing effect seems to be well established (Matusiak and Matusiak, 2018).
The movement of women up the ranks parallels overall growth with a predictable lag. In 1986, when 5.4% of Canadian officers were women, only 0.5% were found in supervisory and just 0.2% in management roles. By 2014, with 22.2% representation overall, women held 17.6% of supervisory and 11% of management positions. By 2019, those percentages had reached 20% and 19%, with promotions outpacing overall growth (Statistics Canada, 2015). Linked to recruitment are historic problems in retention such as seen between 1975 and 1978 when the RCMP lost female over male members at a 2:1 ratio (Seagram and Stark-Ademac, 1992). A 2012 audit suggested retention was improving in the national force; whereas the exit of long-serving female members continued to be disproportionate, the gender gap had disappeared for those with fewer than 20 years of service. Although no more recent statistics are available, Langan et al. (2017) argue that female retention remains a problem in Canadian policing.
Obstacles and competencies
Historically, the greatest obstacle faced by Canadian female officers – as elsewhere – has been their male counterparts, whether this manifests as harassment, hostility or more subtle undermining (Jackson, 1997; Linden, 1983; Schmidt 2011; Seagram and Stark-Ademac, 1992). Experience in the US and UK shows this problem magnified for racial and sexual minority women (Holder et al., 2000; Martin 1994; Miller et al., 2003) and a Canadian researcher has called for greater attention to intersectional considerations (Corsianos, 2009). The most pronounced objection to the entry of women in Canadian policing (Linden, 1983) and elsewhere (Archbold and Schulz, 2012) was that they lacked sufficient physical strength and authoritative presence. Initial fitness standards have since given way to validated protocols for bona fide occupational requirements and the Supreme Court’s 1999 Meiorin decision, involving unjustifiable parameters set for a female firefighter, stimulated Canadian police forces to develop non-discriminatory standards for screening women recruits (Shephard and Bonneau, 2002).
Concern about performance-relevant competencies in women varies upon the perceived magnitude of any difference and the prevailing Zeitgeist as to whether difference is to be valued. While early research downplayed differences, later focus shifted to the distinctive characteristics women brought to policing, such as better conflict resolution skills (Morash and Haarr, 2012; Schulze, 2010). Supporting this view is a relatively recent analysis on the use of lethal force in Canada’s largest cities suggesting an inverse relationship between the proportion of female officers and the number of police killings (Carmichael and Kent, 2015).
Changing perceptions, changing culture?
Kingshott (2013) addresses the gender differences debate by calling for more androgynous officers possessing qualities that optimize service whether traditionally male, female or a blend of the two. COP, launched shortly after women commenced patrol duties, emphasized service relative to crime-fighting (Perrott and Taylor, 1995), presumably a better fit for the female ethic of care and emotional labour (Gilligan, 1982; Rabe-Hemp, 2008). As to the question of critical mass, 22% representation seems a sufficient proportion to affect Canada’s overall policing culture in a move towards greater androgyny (Carmichael and Kent, 2015).
Studies conducted over the last 15 years may be capturing a cultural shift insofar as null (e.g. Paoline and Terrill, 2005), mixed (e.g. Schuck, 2014) and counterintuitive results (e.g. Wentz and Archbold, 2012) are emerging. It may be that differences were never as marked as first argued but also that socialization pressures have begun ‘to erode what has traditionally been an “all-boys club”’ (Archbold and Schulz, 2012: 702). Archbold and Schultz conclude ‘that male and female officers are more similar than they are different’ (697), whether the criterion be arrest decisions, use of force, the drawing of citizen complaints or demonstration of empathy.
In terms of perceived stressors, in a Buffalo-based study, Violanti et al. (2016) found little difference between male and female patrol officers. However, homogeneity ends with perceptions of perceived support. Results showed a 37% higher prevalence of women reporting inadequate support from direct supervisors and these women also reported more problems with colleagues not performing their jobs properly and with low quality equipment.
Whither the recruitment wall?
Why has the growth of women in Canadian policing, and other Western nations, stalled and to what degree does this constitute a failure for the equity goals set more than 40 years ago? A three-level typology is proposed to explain ongoing obstacles:
Overt discrimination and hostility
Some continue to attribute the problem to ongoing hostility, and even misogyny (Franklin, 2005), from within the institution and the men who continue to dominate (Prokos and Padavic, 2002). From this perspective, little has changed. In the only Canadian study situated at this level, Corsianos (2004) used a feminist phenomenological approach in finding all her female detective interviewees to be oppressed.
Growing acceptance, residual resistance and institutional roadblocks
More recently, researchers have focused on subtle forms of resistance and institutional bias where women’s needs are yet to be fully accommodated (Cordner and Cordner, 2011). For example, a study located in the American Midwest showed continuing opposition to women but overall acceptance for those who had proven themselves (Rabe-Hemp, 2008), a finding replicated a decade later in a Canadian sample (Langan et al., 2017).
In a now dated Canadian-based study, Jackson (1997) integrated survey findings with her 8 years of police experience in concluding that although male resistance remained a reality, relations were significantly improved relative to when she began. More recently, Perrott (2018) surveyed officers from a mid-sized Canadian municipal force for perceptions of progress in the quality of services provided to women and in the degree to which the workplace had become inclusive for female members. Although men and women reported steady positive change across specified periods of time, female officers reported less change than did their male counterparts.
The greatest institutional obstacle for female officers is home–work conflict, an important consideration given women’s disproportionate responsibility for childcare (Schulze, 2010). Duxbury and Higgins’ (2012) study of 4000 officers across 25 forces reinforced the importance of such conflict in workplace stress but found, perhaps surprisingly, that this challenge existed equally for male and female officers. This was attributed to growing equalitarianism in the spousal relationships of the officers surveyed. Qualifying this finding, however, was that of those officers with children at home, women spent significantly more time than men with childcare.
Relative to their American counterparts, Canadian female officers are advantaged due to more progressive pregnancy legislation, the provision of paid maternity leave and the option for an additional protected period of unpaid parental leave (Ray et al., 2008). Nonetheless, female officers in Canada continue to be stigmatized and disadvantaged by their pregnancy (Langan et al., 2017; Langan et al., 2019).
Job fit
Cordner and Cordner (2011) allude to an additional reason that women do not pursue policing careers in greater numbers. Perhaps certain action-oriented, assertive and competitive qualities – linked to the traditional male role (Abele, 2003) and bona fide requirements for policing – are not as appealing, on average, to women. One such agentic feature is sensation seeking (Glicksohn and Abula, 1998).
A recent Canadian study showed female firefighter applicants scoring significantly lower than their male counterparts on this trait but higher than female undergraduates (Perrott and Blenkarn, 2015). After considering subtypes of sensation seeking the researchers concluded that the women were not only suited to the challenges of firefighting but were perhaps even better predisposed than were male recruits. However, that these women differed significantly from female undergraduate controls suggested theirs to be an atypical pattern, a finding that may help explain the disproportionate male to female ratio in those applying in the first place.
This job fit hypothesis implies a self-selection bias that may exist independent of de facto discrimination or of easily identified obstacles. A well-known way that gendered vocational interests separate is the ‘things versus people’ dichotomy, a robust finding backed by highly significant effect sizes (Su et al., 2009). Another well supported gender difference, even more salient to self-selection in policing, is communal (or ethic of care) versus agentic orientations (Folberg et al., 2020), a dichotomy that can accommodate the sensation seeking findings in recruit firefighters reviewed above (Perrott and Blenkarn, 2015).
Questioning fit may seem to resurrect discredited arguments of women not being suited to policing. However, this is not a question of ability but rather of choice. Further, this hypothesis is not applicable universally but suggests that the subset of women drawn to policing will continue to be proportionately smaller than the corresponding subset of men. From this perspective, further work remains that can, and should, be done, but slower growth should be expected.
The suggestion here is not to abandon questions of discrimination or of the adequacy of accommodation as barriers to recruitment but rather to consider job fit as an additional hypothesis for consideration. Those police researchers who remain inclined to reject this proposition out-of-hand should be aware that the empirical research is not on their side. It has been a half century since second wave feminism emerged in the West, and men and women across nations remain clearly demarcated across these dimensions of agency and communion (Froehlich et al., 2020). What is especially striking is that countries that are most progressive in laws and policy that support gender equity are those where men and women are most likely to self-select into gender typed occupations (Falk and Hermle, 2018). Cordner and Cordner (2011) argue the job fit view is further circumstantially supported by a drop in the litigation initiated by women for unfair hiring practices in US police forces.
Less focus on polarizing arguments about overt discrimination for clearer analysis at the second and third levels of this typology may allow for the identification of the steps necessary to make policing more woman-friendly. For example, what changes are likely to improve work–family balance and otherwise make policing more accessible? How might we incentivize service provision relative to crime-fighting aspects of police job performance in a way that narrows the agentic versus communal gap? A realistic analysis must include recognition that current trends in policing, including the growth of COMPSTAT, intelligence-led policing and militarization, all reflect movement away from COP and the service provision aspects of policing which women seem to find more meaningful, on balance, than do men (Corsianos, 2009; Perrott, 2014).
Visible minorities in Canadian policing
Although concerns about racial minority recruitment in Canada overlap with those in other English-speaking democracies, the similarities are fewer than is the case with women. Police–race relations, usually focusing on the Indigenous and Black communities (Reasons et al., 2016), are often tense and racially charged incidents frequently precede protests and/or official inquiries (Comack, 2012; Tanovich, 2006). Concerns have also been raised that those from Middle Eastern and/or Muslim backgrounds have been unjustly targeted in the post 9/11 (Odartey-Wellington, 2009). Of special concern of late are the assaults, sexual assaults and murder of Indigenous women and girls; not only have police been accused of providing insensitive and inadequate responses but it has also been alleged that they have been perpetrators in some instances (Palmater, 2016). Still, the Canadian situation is not comparable to that of the US, especially over the last 6 years with racially charged police-linked deaths in Ferguson, Baltimore, New York City, Minneapolis and elsewhere (Bourne, 2020; Mullainathan, 2015).
Discussions of police–minority relations in Canada involve fluid definitions and talking in code. Canadian manners feed a nation-wide reticence to collect race-based data, obscuring knowledge due to the inability to disaggregate groups under the visible minority umbrella (Gabor, 1994; Wortley, 1999). This problem with precision, or with the absence of data altogether, limits understanding especially considering that the policing literature generated out of Canada is already scant for a Western nation of its size (Ricciardelli and Griffiths, 2017).
Progress in representation and promotion
Eight percent of Canadian officers identify as visible minorities and 4% as Indigenous, against national proportions of 22% and 5%, respectively (Statistics Canada, 2019). By comparison, 5% of the constabulary in England and Wales is comprised of minorities (Home Office, 2013) against a societal benchmark of 14% (Office for National Statistics, 2011). About 28% of police officers in the US were members of minority groups in 2016 (Leatherby and Oppel, 2020), unevenly distributed from where they form the majority in larger cities like Washington and Detroit (Sklansky, 2006) to hundreds of smaller municipalities where mostly White officers still police mostly minority communities (Ashkenas and Park, 2014).
Regional sampling shows that the City of Toronto has the highest percentage of minority officers at 24% (CBC News, 2016b) though this is against a population benchmark of 50.5% in Canada’s largest city (Whalen, 2017). As of 2020, the RCMP had reached levels of 11.9% minority and 7.2% Indigenous members nationwide (Royal Canadian Mounted Police, 2020). The only place to have achieved parity, and actually exceed, for both visible minority and Indigenous officers is the Halifax Regional Municipality located on Canada’s east Coast (CBC News, 2016b). Data on progress through the ranks anywhere is scant (Gustafson, 2013) and almost non-existent in Canada. In a dated study of 13 Canadian forces, Jain and Singh (2000) found that of 1506 promotions between 1996 and 1997, 8.6% were to visible minorities and 5.1% to Indigenous persons.
Much lauded when launched about 25 years ago was the First Nations Policing Programs initiative (see Kiedrowski et al. (2017). Meant to promote culturally sensitive and autonomous delivery of services by Indigenous police officers country-wide, governments provided funding to individual bands to set up and manage their own policing systems on reservation land. Although this initiative limps on, it is with significant problems and many of these Indigenous forces have disbanded for a return to state managed police services.
Obstacles and job fit
In a study conducted in Halifax, a municipality reported on in the previous section, Perrott (1999) captured the perceptions of 80 African Canadian applicants of obstacles to, and motivations for, joining a police force. Applicants reported that their primary motivation was to serve the community at large (as compared to the Black community), though the opportunity to be a role model for one’s community was ranked as next most important. Police prejudice, societal racism and the lack of opportunity were cited as the top three obstacles to minority recruitment and a factor analysis showed that problems associated with the policing institution were differentiated from those linked to societal racism. Other responses cited disapproval within the ethnic community as an obstacle, consistent with reports from the US (Campbell, 1980), the UK (Cashmore, 2001) and Canada Jain and Singh (2000) that potential applicants fear being seen as a traitor to their community or because they perceive policing to be a non-prestigious career.
Jain (1987, 1988) previously identified height and weight requirements, a dress code that fails to accommodate all religions, the absence of critical mass from one’s group already serving, and concerns about the adverse impact of selection tests as obstacles to minority recruitment (of these, only the testing issue remains unresolved in the Canadian context). Interestingly, although Perrott’s (1999) participants perceived that racism disadvantaged them in the hiring process they rejected the notion that they should receive preferential treatment.
Jain and Singh (2000) first captured the views of serving Canadian minority officers who reported little blatant racism but did emphasize bias at intake and promotion, stressing that the only way to get promoted was to fit into the ‘Old Boys’ (i.e. White) network. Some offered guarded optimism about the situation improving over time. Twenty years later Rigaux and Cunningham (2020) published another study in this vein with minority officers serving in Western Canada. Expected reported obstacles were a lack of trust in the police and the perception that they would not be welcome in a police force. However, most problems cited by these serving officers were generic grievances having no apparent relationship with race. Much like Perrott’s (1999) applicants, a common theme was to reject the notion that minority applicants be promised or given any preferential treatment; some comments reflected dismay about the ‘political correctness’ in play with efforts to recruit minority officers.
Effects on police culture
Questions arise as to whether minority officers assimilate into the dominant White, hypermasculine culture to survive or, as previously suggested with a possible movement towards androgyny, are they affecting policing culture? There is no Canadian research to address this question but the British experience suggests that minority officers are adding cultural elements insofar as police solidarity seems to be breaking down with evidence of rifts between minority and majority group officers (see Holdaway and O’Neill, 2004, 2006). Sklansky (2006) does not view these divisions as a problem, necessarily, but sees them as evidence of a cultural shift.
Community impact and the question of racial bias
A primary goal of minority recruitment is for officers to have ‘greater understanding of … and greater credibility in minority communities’ (Sklansky, 2006: 1224). Are minority officers more culturally competent when policing minority communities and does this, in turn, have positive effects spilling over to policing as a whole? Although some early research supports this view (e.g. Walker, 1983), Sklansky’s (2006) view on this question is more equivocal. If a more inclusive police force can promote better race relations, there are at least four sources of knowledge from which we might draw evidence: (1) minority community sentiment; (2) racial profiling research; (3) empirical approaches and (4) critical theory (CT). The first three of these sources are considered briefly below and the fourth, that is, research flowing from CT, receives considerable attention and criticism because of (1) its current dominance in the emerging police literature, (2) its detrimental impact on our understanding of the matters discussed in this review and (3) its impact on the more traditional researchers contributing to the policing literature.
Minority community opinion
What little research exists consistently shows approval of the police to be lower in Canada’s minority communities than in the overall public (O’Connor, 2008), replicating findings from other Western nations (Brown and Benedict, 2002). Cao (2011, 2014) found that minority and Indigenous respondents reported lower levels of confidence in the police than did Whites. Sprott and Doob (2014) disaggregated minority groups to show that dissatisfaction was limited to First Nations and Black Canadians; whereas Blacks provided lower ratings on indices of fairness and approachability, but not competence, Indigenous respondents rated the police lower on all dimensions. A Winnipeg-based study supports this pattern showing minority group members to less satisfied than majority group participants and Indigenous respondents more dissatisfied still (Weinrath et al., 2012; see also Comack, 2012).
Studies of racial profiling/carding
The Canadian racial profiling/carding debate is like that in other Western nations. In assessing the meaning of study findings, one research group, led by criminologist Scott Wortley, concludes that race-based profiling, borne of prejudice and stereotyping, is real (Wortley and Owusu-Bempah, 2011; Wortley and Tanner, 2003). Others situate themselves in an opposing camp in arguing that although the police cannot be declared free of discriminatory behaviour, the available evidence does not prove the inverse either (Gabor, 2004; Gold, 2003; Melchers, 2003).
Foster et al. (2016) Ottawa-based project is typical of profiling studies in showing that young male Middle Eastern and African Canadian drivers were disproportionately stopped relative to Whites but with no differences in charge or release outcomes across the three groups. These researchers were careful to note that their data could not address the question of racial/ethnic discrimination. By comparison, just released data reported by Wortley showed a huge overrepresentation of young men of colour being stopped by the Toronto Police (Ontario Human Rights Commission, 2018). Unlike the more cautionary tone taken by Foster et al., Wortley’s report alluded to racial discrimination as the explanatory mechanism.
Profiling investigations cannot prove bias as they do not isolate discrimination as the causal agent. Even those attempting to account for sources of variability fall short and, given the research question, likely always will (Reasons et al., 2016). However, this does not prevent the public, aided by media reports, from viewing such studies as de facto evidence of discrimination and few social scientists are inclined to identify the error (and some, especially from the CT tradition, actively promote it). The problem in taking disproportionate outcomes alone as bona fide evidence of racial discrimination extends to most questions of police bias.
Empirical approaches to assessing racial bias
Archival data
Goff and Kahn (2012) point out that there is a wide-spread public belief that racial bias is pervasive in American policing, a view shared in more tempered form amongst social scientists. However, evidence is complicated by problems accessing data, methodological flaws and various confounds, leaving us knowing much less than is necessary for definitive conclusions. We cannot yet properly formulate the questions and there remains ‘a shocking dearth of scientific certainty about how to assess racial bias in policing’ (177). Goff and Kahn blame both police and academic culture as contributing to this state of ambiguity.
There are no recent Canadian studies that can reliably inform as to racial bias and/or discrimination in the police but two recent ones from the US are sufficiently rigorous to be given serious consideration. Both have methodological limitations that merit critique and the need for measured conclusions. Instead, both were subjected to firestorms of criticism, much of which was politically motivated. This is unfortunate because these represent the best we have on the topic insofar as significant efforts were made to control for alternate explanations.
The first of these, by Harvard economist Fryer (2019), was a large-scale study on police use of force across various major population centres around the US, revealed that African Americans were no more likely than Whites to be killed by the police. Another finding showed that police used significantly more force on Blacks than Whites, but this finding was mostly forgotten in ensuing furore that attended the first. Many scholars took umbrage with purported methodological flaws in the study and, as good scholars might, worked to publish their critiques (e.g. Durlauf and Heckman, 2020) to which, in turn, Dr. Fryer responded (e.g. Fryer, 2020). Added to these academic challenges was a firestorm of outrage reported in the popular media, much of it highly political in nature and some of it involving ad hominem attacks.
The second study proved to be similarly controversial. It demonstrated that White officers were no more likely to shoot minority citizens than were Black or Hispanic officers (initially published as Johnson et al., 2019, but since retracted). Pertinent to this review was the conclusion of these researchers that the recruitment of more minority officers is unlikely to alter the cross-race differences in police shootings. As was the case with the Fryer study, criticism of methodological problems in this study was intense. The popular media exploded with strong endorsements or harsh denunciations primarily, it seems, on the basis of whether the publication held a left versus right editorial stance (e.g. Knox and Mummolo, 2020). For a period, these researchers responded to criticisms with official corrections, but it was not long before they requested that their paper be officially retracted (Johnson et al., 2020). Although it was reported that these researchers initially claimed to have retracted their work for political reasons, they later claimed that they were not under pressure.
It is desirable that research be open to challenge and scholarly debate, but one has to question the severity of attacks to which both of these papers were subjected especially given that they were peer reviewed and published in premiere journals. Moreover, the way in which the paper by Johnson and colleagues was retracted is perhaps unprecedented. Surely, legitimate criticisms might have been published by the journal editor and the original article left to stand.
Laboratory studies on implicit bias
It is frequently argued that racially discriminatory behaviour occurs unconsciously due to stereotyping and implicit associations (Devine, 1989). This premise, coupled with advances in technology developed for simulated use-of-force police training (see Bennell et al., 2007), provides a rationale and means by which to test implicit bias in the well-established ‘shoot/don’t shoot’ paradigm. By manipulating race, the bias hypothesis is assessed by measuring (1) time latencies to fire upon armed subjects and (2) the number of Type 1 errors firing on unarmed ones. Tests of racial bias using this paradigm have only been conducted in the US and, until recently, provided findings showing anti-Black bias. For example, Sadler et al. (2012) found evidence of implicit bias in that their police participants took less time to shoot armed Black as opposed to White suspects.
More recently, a Washington State team utilized this paradigm after ensuring their stimuli more realistic and externally valid (James et al., 2013, 2014). Across three experiments civilian, military and police samples were slower to shoot armed Black as compared to Hispanic or White suspects and more likely to shoot unarmed White than Black suspects (James et al., 2013). In a variation restricted to civilians, the latency in shooting armed Black suspects was accompanied by an enhanced alpha suppression response (James et al., 2014), making race appear salient. James et al. (2013) labelled this as an unconscious ‘counter-bias’ resulting from fear of discipline, civil litigation and negative public sentiment.
In a third study with 80 police officers, James et al. (2016) replicated a ‘reverse racism’ effect, with findings linked to responses on the Implicit Association Test. Although results suggested an association of weapons with Black subjects, this bias did not predict behaviour. The authors contend that this counter-bias response is explained by participant awareness of the state of US race-relations, the media backlash that automatically follows the shooting of a Black suspect and the fear that one will not be supported by one’s department.
Contributions from critical theory
Over the past decade, CT has come to dominate the policing literature, much as it dominates the social sciences and humanities. It is contended here that research from the CT paradigm has distorted what is known about police reform and that the theory makes distortion inevitable. CT’s basic tenets (see George, 2021), combined with its advocacy orientation, require study findings to support pre-determined conclusions. Studies involving race or gender and the police, therefore, are bound to support established orthodoxies.
Duarte et al. (2015) refer to such research as commencing from a position of embedded values. Consider, for example, a recent article extending the oppressor-victim narrative in the Michael Brown shooting case in Ferguson years after the ‘hands up-don’t shoot’ storyline was debunked (Campbell, 2020). According to criminologist and policing journal editor Lori Fridell, ‘they seem to go out of their way to produce the desired conclusion’ (2017: 510).
Those following the critical tradition may not perceive the need to support their claims with evidence. For example, Henry and Tator (2006) assert that they are ‘less interested in determining the numbers involved in racial profiling practices – in what constitutes an actuarial approach – and more concerned with analysing why these kinds of differential behaviours are targeted towards racialized groups’ (37). Khoury (2009), in a similar vein, argues that ‘we depart from mainstream research on racial profiling that attempts to prove or disprove the existence of racial profiling practices. We argue that racial profiling exists extraordinarily and constitutes a social problem, not a police problem’ (56). As Subotnik (1998) points out, traditional scholars have been reticent to challenge this novel practice of publishing conclusions without evidence. Criminologist Phillip Stenning (2011) was, however, an early exception to this pattern of acquiescence in pointing out how the ‘critical’ camp’s contributions have had ‘very negative consequences for rational debate about racial profiling’ (114).
When CT researchers do cite evidence, it is often anecdotal, a case study or from small, nonrepresentative, samples (e.g. Bundy, 2019). Postmodern interpretations of qualitative findings may be used to ensure conclusions flow from preferred narratives. Consider an earlier cited study by Corsianos (2004) with 60 Toronto female detectives. Although 37% reported not facing any sexist behaviour, Corsianos considered their self-reports to be lacking in awareness or evidence that they had been co-opted by a patriarchal, hypermasculine culture. She could then reinterpret what they said in a way to justify her conclusion that all were victims of oppression.
Through means such as just described, CT has developed into a largely non falsifiable paradigm. Those who do not accept its premises may receive signals that dissent will not be well tolerated on matters of police–race relations (e.g. Tobias and Joseph, 2020) or of race more generally (e.g. Ng and Lam, 2020). The rise of CT preceded the recent growth in retracted papers, not due to gross flaws or dishonesty but rather to political pressures associated to cancel culture (see Darbyshire et al., 2020; www.retractionwatch.com). Extraordinary criticism may be directed to unpalatable findings on supposed methodological grounds when they are political ones. Efforts are made to control which findings pass peer review or, if unpalatable findings slip through, to have the work retracted (such as, some argue, was the case with the retraction of Johnson et al., 2019 paper).
Conclusions
This review, meant to examine the degree to which the targeted hiring of women and members of visible minority groups to Canadian police forces has succeeded against original goals, shows that answers to the most important of the questions remain elusive. Perhaps the greatest success has been the recruitment of minority group numbers to closer approximate representative samples from the communities being policed, although representation in most jurisdictions still falls short of targets.
The growth of women, by comparison, is stalled out at about 22%. If gender parity continues as an absolute marker of success, this outcome represents clear failure. However, a more optimistic picture emerges if one considers the job fit argument proposed above to have legitimacy; the degree to which one does so is likely to affect estimates of the proportion of women to be expected were a police force free of gender bias.
There is no way to judge the degree to which targeted recruitment has affected minority group member perceptions of the police or to judge satisfaction with the quality of services now provided to women and girls. There are simply too many factors left unaccounted for that may affect such perceptions that are independent of actual police performance. We do know that policy has changed over the years on many relevant matters (e.g. response to domestic violence and sexual assault); it seems likely that police culture has been affected by these policy changes in conjunction with the influx of women and minority group members. We can conclude, with confidence, that targeted recruitment has been no panacea in addressing minority community dissatisfaction with the police, probably now at an all-time low since the Civil Rights Movement.
Despite the current political orthodoxy, those of us who are police researchers would be wise to remind ourselves just how little is known about racial bias and discrimination in the police (Goff and Kahn, 2012). Much difficult, complex and messy work has yet to be conducted if we are to move towards a better understanding of the dynamics and pathways at play in police–minority relations. The challenge is real, but if we are committed to the importance of evidence supporting sound argument and theory, we need to proceed without undue concerns about political pressure. It is ironic, though not surprising, that the two strongest studies on this topic, both reviewed here, were met with such disproportionate amounts of criticism whereas other, weaker studies, generate overinflated claims and are readily accepted into the literature base. And, just how likely is it that other researchers will attempt to address this question if they fear that their data will point them in an unfavoured, perhaps career threatening, direction?
There are few goals in the social sciences as important as accurately understanding how our police interact with those they serve, especially those who are most vulnerable, in order to formulate the best policy possible going forward. Whereas political action and advocacy will always serve an important function in moving us towards a more just society, even those actions should be informed by the best evidence available. That evidence must come via the scientific method by researchers who approach their work from a position of dispassion and with the assurance that they are free to follow their data where it leads. Continuing our current path, so laden with political considerations and limited by embedded values, serves no one’s interests, least of all those vulnerable persons for whom we are supposedly trying to get it right.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
