Review of the Roots of Youth Violence: Literature Reviews

Volume 5

Aggressive Policing Strategies2

“Get tough” policing strategies are often controversial. This is because, even if they are effective in reducing crime (and many are not), they raise questions about police powers and due process. This chapter seeks to review and assess some of the “get tough” policing strategies that have become popular or widespread in North America and the United Kingdom.

As with most issues in the social sciences, there seem to be mixed findings for most types of interventions. However, there is some consensus around programs that are either clearly ineffective and those seem promising.

What is “get tough” policing? For the purpose of this chapter, it is the “weed” portion of the so-called “weed and seed” approach to crime. In other words, getting tough is any strategy that involves more police numbers, more police powers, and/or more aggressive policing. Under the umbrella term “get tough,” the following policing strategies are reviewed in this chapter: increasing police numbers; rapid response; broken windows theory; zero tolerance policing; hot spots policing; and “guns and gangs” policing.

There is substantial overlap between these policing strategies; for example, a zero tolerance approach to policing is likely to subscribe to the broken windows theory of crime. However, each strategy is given separate treatment in the chapter because each topic has specific programs and philosophies associated with it. In all cases, the literature is reviewed with a view to understanding the strengths and weakness of each approach, and to identify the existing social science evidence relating to the programs’ efficacy.

Increased Police Numbers

One of the most often cited truisms about policing is: the larger the number of police, the less crime there is (Bayley, 1994; Sherman and Eck, 2002). Governments point to increased hiring of police officers as evidence that they are “getting tough” on crime and taking community safety seriously (cf. Government of Ontario, 2008). However, this conventional wisdom concerning the relationship between police numbers and crime has been seriously questioned by a number of police scholars. Bayley (1994:3), for example, is one prominent criminologist who has long argued that increasing the number of police on the street has little impact on crime rates. According to Bayley, research confirms that urban crime rates are unrelated to the relative size of the local police force. Indeed, many cities with high numbers of police officers per citizen have higher crime rates than cities with relatively small police forces (Bayley, 1994:3). Bayley argues that the lack of a statistically significant relationship between police numbers and crime can be explained by the fact that the root causes of crime have nothing to do with policing. In other words, efforts to increase police numbers without addressing the social conditions that give rise to crime will ultimately have little impact on the amount of crime in a given society.

Sherman and Eck (2002) also question the conventional wisdom surrounding numbers of police. They warn that police may well not be able to prevent crime if they are not focused on “specific objectives, task, places, times, and people” (295), and many of the police’s crime prevention strategies are not.

Sherman and Eck (2002) propose the following explanation for the popularity of the “increased police numbers” hypothesis. According to them, levels of violence in the 19th century declined while police numbers in the UK, U.S., and other western nations increased. Thus, the argument goes, observers attributed the decrease in crime to strong police presence (Sherman and Eck, 2002: 296).

However, most social research tends to debunk the conventional wisdom that police numbers have a direct impact on crime rates. One of the most widely cited studies is the Kansas City Preventative Patrol Experiment (Kelling, 1974). The study design comprised three areas that were given different “treatments.” One area, the control area, experienced no change in patrol. The second area was assigned extra patrols, and the third area had all routine preventative patrols removed (officers only entered the area to respond to calls for service). The study found that “few differences and no consistent patterns of differences in rates of reporting crime to the police occurred across experimental conditions” (Kelling, 1974). This study has been used to demonstrate that increases in police numbers have not been proven to reduce crime.

Sherman and Eck (2002) examine research on the strategy of increasing police numbers in order to reduce crime in a given area. They do not find evidence supporting the increased numbers hypothesis. While multiple studies examining the effect of no police – for example, police strikes (Andenaes, 1974; Clark, 1969; Russell, 1975; Sellwood, 1978; cited in Sherman and Eck, 2002) – show that crime will “skyrocket” in these rare circumstances, studies that examine increases in already-existing police forces do not find significant decreases in crime. Two reviews of police-crime studies found little evidence that numbers of police influence crime (Moody and Marvell, 1996; Eck and Maguire, 2000). However, Moody and Marvell (1996) also analyzed data from 49 states and 56 U.S. cities and found significant negative effects of police numbers on crime in the big cities, but not at the state level. Sherman and Eck (2002) infer the following from the Moody and Marvell study: “The evidence for the marginal effectiveness of adding police officers willy-nilly to police agencies is weak and inconsistent, in the absence of clear indicators of what they do and where they do it.” (305). In addition, the methods used in all the police-crime studies above are rated as poor by Sherman and Eck (2002).

Maguire et al. (2002) also conclude that the evidence supporting the underlying hypothesis of increasing police numbers is inconclusive. Maguire et al. cite the Kansas City patrol study, described above, as well as a review by the Home Office in the UK (Clarke and Hough, 1984). Clarke and Hough (1984) conclude that the evidence on conventional deterrent policing (foot and car patrol and detective work), suggests that crime will not be significantly reduced simply by devoting more manpower to these tasks.

Finally, David Bayley points out that any changes in police numbers that occur “as a result of normal political and budgetary pressures” will be too small to make any difference in rates of crime (1994: 5). Bayley concludes that it is “unlikely that crime will be reduced if we try to spend our way to safety by adding police officers. Changes in the number of police within any practicable range will have no effect on crime” (ibid.).

Rapid Response

The “rapid response” hypothesis – the faster the police respond to calls, the greater the reduction in crime – is a close relative of the “increased police numbers” hypothesis. Conventional wisdom asserts that more police, combined with an emergency call service (such as 911), will result in more rapid police responses. According to proponents of rapid response, there will be less crime because rapid response a) operates as a deterrent to would-be burglars (as an example), b) reduces harm to victims when crime does occur and c) results in more arrests (more criminals caught “red handed”) and therefore increases incapacitation (Sherman and Eck, 2002: 304).

Coupe and Griffiths (1996, 2000) conducted a study of residential burglaries on behalf of the Home Office, in a city located in the West Midlands, UK. The authors identified that the most important factor in detecting burglaries was catching the criminal in the act. The authors also identified that a quick response was lacking in a large portion of the cases examined in the study, and they postulate that a quicker response may have increased detection rates and victim satisfaction. Thus, “catching more offenders in the act, by responding to alerts more quickly and in greater numbers, particularly during weekday afternoons when many of the least successful responses to ‘in progress’ burglaries were made ... offers the greatest opportunity for boosting detections” (Coupe and Griffiths, 1996: vi).

While the Coupe and Griffiths hypothesis is in line with common-sense notions of how police are effective, it is directly contradicted by three early studies on emergency response. According to Bayley (1998), studies on police response time by Bieck and Kessler (1977), Percy (1980), and Spelman and Brown (1981) found that the speed of police response to calls for service “does not affect arrest rates” and “rarely prevents further injury or damage” (Bayley, 1998: 52).

The above discussion on police numbers and rapid response has shown that conventional wisdom about policing is rarely confirmed by social science knowledge about policing. The next section also examines some popular aggressive policing tactics that are also in line with conventional wisdom or “folk knowledge” about policing. Unlike the research on police numbers and rapid response, there are some studies that show support for some of the following “get tough” policing strategies.

Broken Windows, Zero Tolerance and Hot Spots Policing

This section will describe the broken windows theory of crime, as well as two separate yet related policing strategies: zero tolerance and hot spots policing. This section will identify some police crime reduction programs that have arisen as a result of the broken windows approach, review the social science evaluations (if any) of the effectiveness of the broken window programs, and identify strength and weakness of the methodologies (of both the programs and the evaluations of the programs). This section will conclude with a discussion of the implications of the main findings regarding broken windows, including political, community trust and cost-effectiveness.

The Broken Windows Theory of Crime

In an influential magazine article, criminologists Wilson and Kelling proposed the broken windows theory of crime (Wilson and Kelling, 1982). The theory suggests that signs of disorder in a neighbourhood – such as broken windows, dilapidated buildings, litter, etc. – can lead to more serious crime in that same neighbourhood. The hypothesis underlying the theory is that signs of disorder in a neighbourhood can undermine the ways in which residents can exert social control. The lack of social control makes the neighbourhood attractive to other social disorder activities such as public drinking and prostitution. In turn, this degraded environment is attractive to semi-commercial criminal enterprises such as drug dealing and elevates further the level of serious crime in the neighbourhood (Maguire, Morgan and Reiner, 2002; Sherman and Eck, 2002; Skogan, 1990).

The key implication of the broken windows theory for policing is a radical departure from traditional policing strategies: broken windows policing means policing minor offences in order to affect a reduction in serious crimes (Maguire et al., 2002; Sherman and Eck, 2002; Skogan, 1990).

Skogan (1990) studied the relationship between crime and disorder using data from multi-site, cross-sectional surveys of inner cities in the United States. He found some support for the Wilson and Kellings broken windows hypothesis:

Perceived crime problems, fear of crime, and actual victimization all were linked to social and physical disorder in the area. Those relationships were strong even when other important determinants of area crime were taken into account, suggesting that disorder could play a role in stimulating neighbourhood crime. (Skogan, 1990: 84.)

The broken windows theory of crime has been tested in several other studies, with generally supportive results (although with some methodological concerns, according to Sherman and Eck, 2002). Wilson and Boland (1978) and Sampson and Cohen (1988) used national samples of cities to examine the effects of police arrest rates for minor offences on the more serious crime of robbery (cited in Sherman and Eck, 2002: 312). Both studies found support for the hypothesis that aggressive, proactive arrests for minor offences would reduce rates of robbery. However, these studies have been questioned because they use arrests as a measure of policing aggressiveness, which are largely a result of traffic stops, whereas robbery is most likely to occur in pedestrian-heavy areas (Sherman and Eck, 2002).

Further support for the broken windows/ proactive policing hypothesis is found in a quasi-experimental study of the Indianapolis Police Department. McGarrell et al. (2001) found that increases in traffic enforcement resulted in reductions in violent gun-related crime in at least one study area. Weak support for the broken windows hypothesis is found in a longitudinal study of the policing of “incivilities” (disorder) in Baltimore. Taylor (2001) found that while disorder policing did have some effects on serious crime, it did not have an effect on as wide a range of crimes as would be suggested from the original theory (Taylor, 2001).

In a recent commentary on the state of broken windows policing, Skogan reflects on how widely the theory and its hypotheses have been embraced:

As a result of adopting the broken windows theory of neighborhood decline wholeheartedly (see Skogan, 2006), Chicago police now find themselves involved in a host of new activities. They orchestrate neighborhood cleanups and graffiti paint-outs by volunteers and city workers, distribute bracelets that will identify senior citizens if they fall down, and take note of burned out street lights and trees that need trimming. Officers drop into stores to ask merchants to display signs requesting that patrons refrain from giving money to panhandlers. The public steers police toward problems created by the sale of loose cigarettes and individual cans of beer at convenience and grocery stores because they encourage loitering and public drinking. Those sales are illegal but truly soft crimes (Skogan, 2008: 199).

Unfortunately, this rather romantic view of broken windows policing completely ignores the aggressive aspects of this type of policing, and focuses more on the “soft” side of broken windows. The rosy picture that Skogan depicts in the excerpt illustrates the allure of the broken windows approach: police can still align broken windows policing to the wider goals of community policing. This may not hold true for zero tolerance policing, as we shall see.

In sum, Sherman and Eck (2002) generally find supporting evidence for proactive (broken windows) arrests. For example, crackdowns on drunk driving and traffic arrests led to decreases in crime rates in several studies (314). However, drug crackdowns showed modest short-term reductions in crime in the areas where police raids took place, and showed a return to pre-raid crime rates within seven days. The drug raids were deemed to be labour-intensive and not cost-effective (Sherman and Eck, 2002: 314). As with zero tolerance policing (discussed below), the authors express the concern that broken windows policies may have positive short-term effects but may increase crime in the long term (ibid.).

Zero Tolerance

What is zero tolerance policing? It is difficult to find a definition of zero tolerance policing. It often associated with the “get tough” attitudes and policies of Mayor Giuliani and Police Chief Bratten in New York City in the 1990s. Crackdowns on panhandlers, “squeegee” kids, and other “quality of life” policing come to mind. Some of the characteristics of zero tolerance policing include a focus on disorder; proactive stops, searches and arrests; the use of civilian sanctions; low police discretion; and eschewing of community involvement (Goff, 2004).

In an article that traces the genealogy of zero tolerance, Newburn and Jones propose a savvy definition:

It is difficult to specify a particular set of policy interventions that characterize Zero Tolerance anti-drug policies. The term has been used primarily as a rhetorical device, used to signal uncompromising and authoritative action by the State and its agencies, against an external and internal enemy. It has denoted an unambiguous faith in a criminal justice response to the problem of drugs, diversion of resources to enforcement from other response (such as treatment), harsher punishments and a weakening of ‘due process’ considerations in favour of those of ‘crime control’ (Newburn and Jones, 2007: 223).

According to Newburn and Jones, the following activities were associated with NYC policing in the 1990s, and may have been mislabeled as “zero tolerance”:

Several studies of zero tolerance policing in NYC have found tentative support for the hypothesis that changes in the policing of misdemeanours contributed to the dramatic declines in homicides in the 1990s (Kelling and Bratton, 1998). However, some academics have been skeptical of the zero tolerance approach to policing. Bowling (1999) argues that the zero tolerance policing in NYC was oversold to the public through the media and police agencies, and that zero tolerance may be an inappropriate term to apply to policing strategies.

Bowling questions the conventional wisdom around the New York City crime story – that the decline in the city’s murder rate from 1991 to 1997 can be attributed to the mayor’s so-called zero tolerance approach. Bowling suggests that the crack cocaine drug market spiked just before the implementation of intensified policing in New York City, and therefore the decline in homicide may be attributable to a drug decline already under way (Bowling, 1999). Further, others have suggested that the level of public complaints and dissatisfaction resulting from the aggressive policing in New York City was too high a cost regardless (Greene, 1999).

Sherman and Eck (2002) catalogue a number of studies that arguably test the effectiveness of zero tolerance policing. A controlled study of aggressive field interrogations in San Diego (Boydstun, 1975) found that the study area for which all field interrogations were suspended experienced an increase in robbery, burglary, grand theft, petty theft, auto theft, assault, and sex crimes (classified as “suppressible crimes” by the study authors, and as “outdoor crime” by Sherman and Eck). However, Boydstun also found that the area that received an increase in field interrogations (generally considered an important element of zero tolerance policing) did not experience a discernible decrease in these “suppressible” crimes (1975).

The list of zero tolerance studies with mixed results is long. Sherman (1990), in a review of 15 case studies of police crackdowns, found that a) increased field interrogations did lead to fewer “outdoor” crimes such as robbery, larceny and auto theft; however, b) there were no reductions in robbery as a result of police crackdowns on disorder (Sherman, 1990).

Kelling and Coles (1996, cited in Sherman and Eck, 2002) conducted case studies of broken windows policing in several U.S.US cities. They found some evidence in support of police “crackdowns” on minor offences translating into reductions in more serious crime: in their assessment of a police crackdown on fare-dodging in the New York City subway, they found the crackdown did reduce the number of robberies on the subway.

Novak et al. (1999) evaluated a one-month program of aggressive policing on disorder crime – such as public intoxication and loitering – in a Midwestern United States city. The evaluators found no reduction in burglary or robbery as a result of the proactive disorder enforcement (Novak, Harman, and Holsinger, 1999).

Despite these findings, Sherman and Eck (2002) postulate that zero tolerance activities, whether or not effective in the short term, may have unexamined long-term consequences. There is also some consensus, as with broken windows in general, that zero tolerance policies result in increased public complaints against police, and dissatisfaction – especially amongst minority communities.

Hot Spots Policing

The hot spots policing hypothesis is the idea that “the more precisely patrol presence is concentrated at the ‘hot spots’ and ‘hot times’ of criminal activity, the less crime there will be in those places and times” (Sherman and Eck, 2002: 308). Crime mapping technology has become the most well-known innovation associated with hot spots policing; indeed, a recent report found that “7 in 10 departments with more than 100 sworn officers reported using crime mapping to identify crime hot spots,” making hot spots policing one of the most popular crime reduction strategies in the United States (Braga, 2005: 317). The policing activities associated with hot spots policing are directed patrol, heightened traffic enforcement, and aggressive disorder enforcement (Braga, 2005). According to Braga, one-off police crackdowns should not be considered hot spot policing, but crackdowns that are conducted on an ongoing basis with frequent follow-up can be classified as hot spots policing (Braga, 2005). Thus, we can see that although hot spots policing is not as extreme as zero tolerance policing, it does qualify as a “get tough” policing strategy.

There does not appear to be a body of academic literature addressing the relationship, if any, between hot spots policing and broken windows theory; however, it could be argued that the two approaches have much in common. For example, both strategies stem from theories such as rational choice, routine activities, and environmental criminology (Braga, 2005). Basic assumptions that provide the basis for the broken windows theory of crime seem important to the conceptual framework of hot spots policing. For example, hot spots researchers have identified that facilities and site features of particular places (like street corners with abandoned buildings) make them attractive to criminals (ibid.: 319). It can be argued, then, that broken windows-style disorder in neighbourhoods needs to be identified in order for hot spots policing to be implemented in a given neighbourhood.

The research evidence regarding hot spots policing is more supportive over all than the evidence for zero tolerance policing. Sherman and Eck (2002) surmise that the well-funded U.S. research on directed patrol (hot spots policing) has generally had supportive results. Two studies using data from the Minneapolis Police Department patrol project found supporting evidence for the hot spots hypothesis in policing (Sherman and Eck, 2002: 308-309). Sherman and Weisburd (1995, cited in Sherman and Eck, 2002) found that areas in the experiment that received extra patrols (100 per cent more) had a reduction of crime of up to 50 per cent. In a study using the same data from Minneapolis, Koper (1995, cited in Sherman and Eck, 2002) analyzed the impact of the length of the patrol visit on crime reduction. Koper discovered that longer patrol visits (up to 15 minutes long) paid off with longer crime-free periods following the visit.

However, Sherman and Eck (2002) raise two concerns about the evidence for hot spots policing. Firstly, there is concern that many studies do not account for displacement of crime to other areas for the duration of the intervention studied. Secondly, the evidence for hot spots policing is mixed for the policing of drug markets. For example, May et al. (2000) found that low-level enforcement in two UK drug markets did not appear to have an impact on above-street-level distribution. The authors conclude that police enforcement is largely ineffective (May, Harocopos and Turnbull, 2000).

Similar to the UK findings, U.S. studies that test whether sustained drug “crackdowns” have an effect on levels of violence have been inconclusive. In a controlled experiment, Sherman and Rogan (1995) found that crackdowns and raids on crack houses in Kansas City, Missouri reduced crime in the hot spots for only two weeks. The authors therefore conclude that “crackdowns” in the form of raids on crack houses, which are resource intensive, may not be cost-effective (Sherman and Rogan, 1995).

A randomized experimental evaluation of hot spot policing in New Jersey, which included police crackdowns on drug hot spots showed significant impact on disorder offences (Weisburd and Green, 1995). Crackdowns consisted of intensive policing of a particular street corner in order to “shut down” the drug market operating at that location. Crackdowns happen at a specific point in time, with follow-up visits in the next few days if necessary. However, the authors found that the police crackdowns did not influence violent or property crime (Weisburd and Green, 1995:723).

Finally, Braga (2005), in a recent systematic review and meta-analysis of five hot spots policing randomized controlled trials, concluded “four of five experimental evaluations reported noteworthy crime and disorder reductions” (336). In addition, “a meta-analysis of key reported outcome measures revealed a medium statistically significant mean effect size favoring the effects of hot spots policing in reducing citizen calls for service in treatment places relative to control places” (ibid.). The evidence, then, is generally supportive of hot spots policing, but it should be taken into consideration that most of the studies measured neither displacement of crime nor the long-term effects of hot spots policing.

Given these research limitations, a cautionary note from an early study on hot spots policing is still relevant today. Chaiken (1977), reviewing the deterrent effect of various police activities in a RAND corporation report, notes: “most studies are consistent with the view that a substantial increase in police activity will reduce crime for a time, but, in the real world, increases in police manpower tend to follow increases in crime. The magnitude and duration of deterrence effects are essentially unknown.”

Guns & Gangs

This section will review gang interventions that involve aggressive policing of guns and gangs. The earliest programs were “street worker” based, and relied on an individual street worker’s relationship with gang members; this model is dubbed the “gang transformation model.” With the more conservative political movements of the 1980’s, a deterrence model of gang suppression overtook the street worker model. The police took a leading role in deterrence-based suppression measures (Klein and Maxson, 2006).

Before a discussion of aggressive policing of gangs, it is prudent to acknowledge the myriad of other ways police try to solve the “gang problem” through strategies of prevention and intervention. While many of these programs have intentions that are commendable and are popular with the public – parents and teachers especially – almost all gang prevention programs are ineffective (Esbensen, Osgood, Taylor, Peterson and Freng, 2001; J. Greene and Pranis, 2007; Katz and Webb, 2006; Klein, 2006; Stinchcomb, 2002; Weisel and Painter, 1997). According to Klein and Maxson (2006), most gang-related programs suffer from relying on conventional wisdom about gangs, rather than on social science knowledge about gangs. As a result, according to Klein and Maxson, “For the most part, these programs, with their conventional wisdoms, became conventional failures” (91).

Police crackdowns on guns and gangs have been a feature of police responses to gangs since the 1980s. Sherman (1990) defines a police crackdown as “a sharp increase in law enforcement resources applied to the previously under-enforced laws, with a clear goal of enhancing general deterrence of the misconduct” (129). Scholars speculate that the conservatism of the Reagan years in the 1980s, combined with the perception of failed social interventionist strategies of the 1960s and 1970s, paved the way for more “get tough” policing in the 1980s and 1990s (Greene and Pranis, 2007; Katz and Webb, 2006; Klein and Maxson, 2006).

Crackdowns in the context of gang members and gang-related crime have not been systematically evaluated (Fritsch, Caeti and Taylor, 1999; Klein, 2006). However, the negative consequences of police crackdowns are embodied in the experience of Operation Hammer, which was instigated by the LAPD in 1987. It was a “preannounced, media-covered gang sweep” of Black gang-involved neighbourhoods involving over 1,000 officers (Fritsch et al., 1999: 126). Operation Hammer resulted in “1,453 juveniles being arrested; however, 1,350 were released without formal charges being filed” (ibid.). Klein and Maxson note that Operation Hammer resulted in few convictions, but that “community resentment flourished” (Klein and Maxson, 2006: 241).

Aggressive policing of truancy and curfew (crackdowns), as part of an anti-gang initiative in Dallas, was assessed by Fritsch et al. (1999). The authors found that saturation patrols, including intensive stops and frisks of suspected gang members, did not have an effect. However, they found the enforcement crackdowns on curfews and truancy did appear to reduce gang-related violence, although overall gang-related offences reported to the police did not change (Fritsch et al., 1999).

According to Greene and Pranis (2007), an example of a “balanced” model of gang suppression is found in Boston’s Operation Ceasefire. This operation was based on deterrence of gun violence through aggressive policing tactics, but according to Braga et al., the actors involved traded zero tolerance fantasies of the elimination of all gang crime for more realistic reductions in gun violence by leveraging knowledge about gang members’ activities. The philosophy of policing in Operation Ceasefire is: “We’re here because of the shooting.... We’re not going to leave until it stops. And until it does, nobody is going to so much as jaywalk, nor make any money, nor have any fun” (Greene and Pranis, 2007: 85). Time-series analysis evaluation of Operation Ceasefire concluded that the program was “associated with statistically significant reductions in youth and gun violence” (ibid.).

The most popular police response to gangs has been the establishment of specialized police gang units (Katz and Webb, 2006: 3). For example, over 56 per cent of municipal police departments in the U.S. have established gang units (ibid.: 9). On one hand, specialized gang units are a way for the police to show the public they are serious about gang-related crime. On the other hand, overly aggressive gang units themselves have attracted their fair share of negative publicity for the police. For example, Rampart CRASH unit officers in Los Angeles were found to be engaging in hard-core criminal activity. Officers admitted to attacking known gang members and falsely accusing them of crimes they had not committed. The ensuing investigation revealed that officers were routinely choking and punching gang members for the sole purpose of intimidation. In one case, officers had used a gang member as a human battering ram, forcefully thrusting his face repeatedly against a wall. Similarly, in Las Vegas, gang unit officers were found guilty of participating in a drive-by shooting. Two officers, one driving and the other hanging outside a van, had driven around a well-known neighbourhood until they found a group of gang members loitering on a street corner. The officer hanging outside the van shot six times into the crowd, killing a 21-year-old male (Katz and Webb, 2006: 3).

Katz and Webb (2006) report on the results of their study of four police gang units in the Southwestern United States. The methodologies used in the study were field observation, interviews with gang unit officers (n=76) and other stakeholders (n=69), review of documents generated by the gang units and police departments (n=275), and newspaper articles (n=285). Katz and Webb find, contrary to some gang unit research, that the formation of gang units in the cities studied was not the result of knee-jerk reactions to moral panics about gang violence (267), but rather the police responding to a real or objective threat. However, they point out that the response of police to the gang problem – the formation of gang units – is an indirect response to this objective problem. The gang units, then, were responses to political pressure and public concern about high-profile incidents of gang violence, rather than responses to actual crime problems.

Katz and Webb draw some conclusions about the functioning and consequences of the four police gang units studied. Firstly, they find that the gang units and police officers studied lacked accountability to either the community or even their own police departments and supervisors:

We concluded that the gang unit officers . . . were free to undertake any activity that interested them, had few expectations to meet, and had virtually no policies or training to guide their decision making. Gang unit officers were also rarely under the control or supervision of police management. They were physically and operationally isolated from the rest of the police department, and typically had little contact with “regular” police officers, criminal justice officials, the public, or community groups (Katz and Webb, 2006: 268).

The authors hypothesize, drawing from organizational theory, that the lack of accountability stems from the strategic and structural “decoupling” of gang units from their host departments. For example, three of the gang units studied were located off site in “secret” locations:

Nearly all other police officers and criminal justice stakeholder were kept in the dark about their locations. Even those select few who may have been told where to find them could not enter unescorted. . . . In a few instances, we thought that the espoused need for secrecy had become cloaked with a cold war, spy-like quality, some gang officers asserting that their regular precinct stations or police headquarters had become subject to penetration by gang members, rendering intelligence files vulnerable to destruction or manipulation (Katz and Webb, 2006: 276).

Secondly, the researchers found that each gang unit, as a consequence of its isolation, had developed a subculture, which was largely responsible for guiding the activities deemed important for each unit. Thus, the work-group subculture replaced guidance from supervisors or departmental policies. This further encouraged what Katz and Webb dubbed the “buffet-style” policing that characterized the gang united studied. The development of gang unit subculture can also have more sinister undertones: “A still greater problem ... is the potential for [gang units] to develop unique internal subcultures that can become at odds with the mission of the parent department, or even with the law” (278). The authors cite the investigation into corruption of the LAPD’s Rampart CRASH unit (see above).

Thirdly, Katz and Webb reveal that the relationships of the gang units studied with the communities they served and the goals and tenets of community policing was problematic. Given the units’ isolation and self-directed activities, the researchers found that gang unit officers had almost no contact with the community, including very little contact with gang members themselves. They also found that the gang units operated with “no input from the community.” Over all, Katz and Webb conclude that the gang units studied were antithetical to the stated goals of community policing of the wider department, media and public. This finding is echoed in Decker’s recent call for police gang units to be expanded, but only within the framework of community policing (Decker, 2007).

Finally, Katz and Webb believe that the gang units that emphasized intelligence-gathering, both internally and externally, achieved a greater amount of legitimacy from providing a service to other organizations (such as schools). The intelligence function also allowed greater interaction with the community than the units that emphasized enforcement. The authors emphasize that the gang unit that leaned most towards the goal of intelligence-gathering was able to position itself as more legitimate and amenable to community policing goals than the enforcement-oriented gang units that struggled to earn legitimacy and were often at odds with community policing. Importantly, Katz and Webb also conclude that intelligence-gathering units represent a more cost-effective policing strategy than do the enforcement-oriented units.

In light of Katz and Webb’s study of U.S. gang units, the following description of the Toronto Police gang unit is not encouraging:

The logo of Canada’s largest gang-suppression unit, the Toronto Police Guns and Gangs Task force, is perhaps emblematic of this ethic of confident state-sanctioned aggression. Adorning golf shirts and other popular swag sold at Task Force presentations to other gang cops and police service members, the striking logo features a muscular cartoon-like bulldog, its teeth clenched and its eyes shaded by wraparound sunglasses, holding a massive machine gun. Wearing a muscle shirt with the Toronto police logo, the dog has massive arms that feature a red bandana around the left biceps and a blue one around the right – perhaps a ‘taking of the flag’ sign of victory against some indeterminate Blood or Crip gang set (Chettleburgh, 2007: 147).

The Toronto unit would probably be classified as an enforcement-oriented unit and might suffer, according to Katz and Webb, from the problems of isolation and lack of legitimacy and accountability, and may be a resource-intensive policing style that is outside the purview of community policing goals. Unfortunately, there has been no research to date that has been able to measure the impact of Toronto’s gang unit on either crime or community attitudes towards the police.


This review of “get tough” policing strategies illustrates that many aggressive policing strategies have been found to be clearly ineffective, or evidence regarding their effectiveness is inconclusive (Bayley, 1994, 1998; Greene and Pranis, 2007, Klein and Maxson, 2006; Katz and Webb, 2006; Maguire et al., 2002; Sherman and Eck, 2002). The most promising strategies seem to be the less-aggressive hot spots policing (though perhaps not for drug markets), and unique programs such as Operation Ceasefire that are sensitive to social processes associated with gangs and crime (Braga, 2005; Greene and Pranis, 2007; Sherman and Eck, 2002). The table below summarizes the findings of this chapter on “get tough” policing.

Table 1: Get Tough Policing and Crime Reduction
Get tough strategy Underlying crime reduction hypothesis Effectiveness Evidence
Increase Police Numbers More police = less crime Inconclusive and/or not effective Clarke & Hough (1984); Kelling et al. (1974); Marvel & Moody (1996); Eck and Maguire (2000)
Rapid Response The faster the response to incidents, the less crime Inconclusive Bayley (1994,1998); Bieck & Kessler (1977); Spelman & Brown (1981)
Broken Windows Signs of neighbourhood disorder erode community social controls and “invites” criminal activity to the area Weak support for hypothesis Skogan (1990); Wilson & Boland (1978); Sampson & Cohen (1988)
Zero Tolerance Crackdowns on minor offences lead to a reduction on more serious offences Inconclusive – can result in increased public complaints about police aggressiveness Bowling (1999); Greene (1999); Kelling & Bratton (1998); Kelling & Coles (1996); Sherman (1990); Novak (1999)
Hot Spots (Directed Patrol) Focus patrol on “hot spots” and “hot times” = less crime May be effective in some contexts; not effective in hot spot policing of drug markets Braga (2005); Sherman & Weisburd (1995); May et al. (2000)
Guns&Gangs: aggressive policing and specialized Police Gang Units Marshal resources within a police organization to “fight back” against gangs and to show the community the police’s commitment to solving the gang problem Ineffective, with possible exception of Boston’s Operation Ceasefire Fritsch et al., (1999); Greene & Pranis (2007); Katz & Webb (2006); Klein & Maxson (2006)
Adapted from: Maguire et al., 2002; Sherman and Eck, 2002.

Some final thoughts on the popularity of “get tough” policing seem appropriate. The emphasis on the “get tough” approach to crime is perhaps not surprising given the political contexts of crime prevention policies in the United States. For example, the U.S. Department of Justice Weed and Seed program (which provided funding for some of the enforcement strategies discussed in this chapter) was designed to improve public safety and fear of crime by reducing or eliminating “violent crime, drug trafficking, and drug-related crime from targeted high crime neighbourhoods” (U.S. Department of Justice, 1998). Weed and seed programs stem from the broken windows style policing we have discussed throughout this chapter. The “weed” component is the “iron fist” of aggressive enforcement, while the “seed” component is the “velvet glove” of community engagement (Bridenball and Jesilow, 2005; Kraska and Paulsen, 1997).

Not surprisingly, scholars have noted that the “weed” component of the weed and seed program has overshadowed the “seed” component in political will, enthusiasm from law enforcement agencies and, most importantly, funding from federal agencies (Bridenhall and Jesilow, 2005; Miller, 2001). The most telling evidence of “weeding” outstripping “seeding” is the funding choices made by the Bush administration:

The future of Weed and Seed programs, however, is unclear. George W. Bush’s budget for 2003-2004 drastically cut federal funds for the programs, primarily as a means to finance antiterrorist activities. Money that was left in the budget was targeted for law enforcement activities. Seed efforts, already given secondary consideration by policy makers, all but vanished from the budget (Bridenhall and Jesilow, 2005: 66).

In Canada, the situation appears even more perplexing than in the United States. For example, while there have been at least some attempts at evaluating “get tough” policing initiatives in the United States, there is currently no publicly available research assessing the effectiveness or costs of “get tough” policing in Canada (Chettleburgh, 2007). This, despite the fact that the amount being spent in Ontario alone, following the Provincial government’s expansion of the Toronto Anti-Violence Intervention Strategy (TAVIS) and its ongoing Guns and Gangs program, totals approximately $61 million.

Finally, new research on negative police-public interactions and their effect on public perceptions of police (and police legitimacy in general) raise questions about the unwanted side effects of aggressive policing. According to Skogan (2007), when it comes to encounters with police, “the impact of having a bad experience is four to fourteen times as great as that of having a positive experience” (99). Thus, the impact of negative encounters of the general public with police as a result of “get tough” policies (whether effective for crime reduction or not) can have deleterious and irreversible effects on policing legitimacy and ultimately impair their main functions. Indeed, Sherman and Eck (2002) highlight the importance of process, rather than outcome, for the victims of crime and others who come into contact with the police:

One of the most striking recent findings is the extent to which the police themselves create a risk factor for crime simply by using bad manners. Modest but consistent scientific evidence supports the hypothesis that the less respectful police are towards suspects and citizens generally, the less people will comply with the law. Changing police ‘style’ may thus be as important as focusing on police ‘substance’ (293).

The effect of aggressive policing tactics on the lives of ordinary people is unknown. Yet there have been instances where police raids have left ordinary people wondering what the benefits of “get tough” policing are. A recent media report on a raid by the Toronto Police Guns and Gangs Task Force on the home of a Toronto musician raised such questions. Kevin Clarke reported to the CBC that “a massive operation, over 50 police involved” broke down the door to his home, heavily armed, set off smoke bombs, cordoned off the neighbourhood and executed a search warrant (CBC news, 2008a). According to Clarke and his lawyer, the police found nothing and have not told him or his lawyer what they were looking for. Clarke feels both shaken and distrustful as a result of his experience. In a similar incident, heavily armed officers searched the home of Brian Henry, a prominent Black youth worker (ibid.). What impact do such raids have on the community’s perception of the police and their trust in the criminal justice system? Unfortunately, due to a lack of evaluation research, we can’t answer these questions at this time.

The lack of evaluation of aggressive policing in Canada, then, has two major implications. The first is that the effectiveness of these programs — with respect to reducing overall levels of crime and violence — is simply unknown. The second implication is that little is known about how these aggressive policing strategies impact the general population — especially the members of ethnic minority groups who often reside in targeted communities. As discussed above, we must consider the possibility that aggressive policing policies may have a negative impact on community perceptions of the justice system — a result that may temper the benefits of any reductions in local crime rates. In sum, the lack of evaluation means that members of the general public really have no ability to conduct a basic cost-benefit analysis that will help them determine whether their tax dollars are being spent in the most productive manner.

In conclusion, it should be stressed that we are not at this time arguing that aggressive police strategies in Ontario — including TAVIS and the Guns and Gangs Task Force — are proven to be ineffective. Indeed, they may have many benefits. They may, for example, bring dangerous criminals to justice, deter crime, reduce fear of crime in specific communities and increase public trust in the police. However, as the above discussion denotes, previous research suggests that these aggressive policing programs may also be expensive, may not reduce crime rates in the long term and could have a very negative impact on community perceptions of the justice system. The only way to truly determine the long-term benefits – and consequences – of such programs would be to conduct rigorous, transparent evaluations.


Bayley, D. H. (1994). Police for the Future. New York: Oxford University Press.

Bieck, W. and D. A. Kessler. (1977). Response Time Analysis. Kansas City: MO: Board of Police Commissioners.

Bowling, B. (1999). The rise and fall of New York murder: zero tolerance or crack’s decline? British Journal of Criminology, 39(4), 531–554.

Boydstun, J. (1975). San Diego Field Interrogations: Final Report. Washington: Police Foundation.

Braga, A. A. (2005). Hot spots policing and crime prevention: A systematic review of randomized controlled trials. Journal of Experimental Criminology, 1(3), 317–342.

Bridenball, B., and P. Jesilow. (2005). Weeding criminals or planting fear: An evaluation of a Weed and Seed project. Criminal Justice Review, 30(1), 64–89.

CBC News. (2008a). Highly publicized 2005 gang sweep nets few convictions. Retrieved July 16, 2008 from

CBC News. (2008b). Toronto artist seeks explanation for police raid. Retrieved July 14, 2008 from

Chaiken, J. M. (1977). What’s Known about Deterrent Effects of Police Activities. Santa Monica, CA: The Rand Corporation.

Chettleburgh, M. C. (2007). Young Thugs: Inside the Dangerous World of Canadian Street Gangs. Toronto: Harper Collins.

Clarke, R. V. and M. Hough. (1984). Crime and Police Effectiveness Home Office Research Study No. 79. London: Her Majesty’s Stationery Office,

Coupe, T. and M. Griffiths. (1996). Solving Residential Burglary. London: Police Research Group.

Coupe, T. and M. Griffiths. (2000). Catching offenders in the act: An empirical study of police effectiveness in handling ‘immediate response’ residential burglary. International Journal of the Sociology of Law, 28(2), 163–176.

Decker, S. (2007). Expand the use of police gang units. Criminology and Public Policy, 6(4), 729.

Esbensen, F., D. W. Osgood, T. J. Taylor, D. Peterson and A. Freng. (2001). How great is G.R.E.A.T.? Results from a longitudinal Quasi-experimental design. Criminology and Public Policy, 1(1), 87–118.

Fritsch, E. J., T. J. Caeti and R. W. Taylor. (1999). Gang suppression through saturation patrol, aggressive curfew, and truancy enforcement: a quasi-experimental test of Dallas Anti-Gang Initiative. Crime and Delinquency, 45(1), 122–139.

Goff, C. H. (2004). Criminal Justice in Canada (3rd ed.). Scarborough: Thomson Nelson.

Government of Ontario (2008, June 24). Ontario Tackles Guns And Gangs in Toronto. Retrieved 3 July, 2008, from

Greene, J. A. (1999). Zero tolerance: a case study of police policies and practices in New York City. Crime and Delinquency, 45(2), 171–187.

Greene, J. and K. Pranis. (2007). Gang Wars: The Failure of Enforcement Tactics and the Need for Effective Public Safety Strategies. Washington: Justice Policy Institute.

Katz, C. M. and V. J. Webb. (2006). Policing Gangs in America. New York: Cambridge University Press.

Kelling, G. L. (1974). The Kansas City Preventive Patrol Experiment: A Summary Report. Washington: Police Foundation.

Kelling, G. L. and W. J. Bratton. (1998). Declining crime rates: insiders’ views of the New York City story. Journal of Criminal Law and Criminology, 88(4), 1217–1231.

Kelling, G. L. and C. M. Coles. (1996). Fixing Broken Windows: Restoring Order and Reducing Crime in our Communities. New York: Free Press.

Klein, M. W. and C. L. Maxson. (2006). Street Gang Patterns and Policies. New York: Oxford University Press.

Kraska, P. B. and D. J. Paulsen. (1997). Grounded research into U.S. paramilitary policing: Forging the iron fist inside the velvet glove. Policing and Society, 7(4), 253–270.

Maguire, M., R. Morgan and R. Reiner. (Eds). (2002). The Oxford Handbook of Criminology (3rd ed.). Oxford: Oxford University Press.

Marvell, T. B. and C. E. Moody. (1996). Specification problems, police levels, and crime rates. Criminology, 34(3), 609–646.

May, T., A. Harocopos and P. J. Turnbull. (2000). Serving Up: The Impact of Low-Level Police Enforcement on Drug Markets. London: Research.

McGarrell, E. F., S. Chermak and A. Weiss. (2001). Reducing firearms violence through directed police patrol. Criminology and Public Policy, 1(1), 119–148.

Newburn, T. and T. Jones. (2007). Symbolizing crime control: Reflections on Zero Tolerance. Theoretical Criminology, 11(2), 221–243.

Novak, K. J., J. L. Harman and A. M. Holsinger. (1999). The effects of aggressive policing of disorder on serious crime. Policing an International Journal of Police Strategies and Management, 22(2), 171–190.

Sherman, L. W. (1990). Police crackdowns: Initial and residual deterrence. In M. Tonry and N. Morris (Eds.), Crime and Justice: A Review of Research, 12 (pp. 1–48). Chicago and London: University of Chicago Press.

Sherman, L. W. and J. E. Eck. (2002). Policing for crime prevention. In L. W. Sherman, D. P. Farrington, B. C. Welsh and D. L. MacKenzie (Eds.), Evidence-Based Crime Prevention (Revised Ed.), (pp. 295–329). New York: Routledge.

Sherman, L. W. and D. P. Rogan. (1995). Deterrent effects of police raids on crack houses: a randomized controlled experiment. Justice Quarterly, 12(4), 755–781.

Skogan, W. G. (1990). Disorder and Decline. Berkeley: University of California Press.

Skogan, W. G. (2006). Asymmetry in the impact of encounters with police. Policing and Society, 16(2), 99–126.

Skogan, W. G. (2008). Broken windows: Why - and how - we should take them seriously. Criminology and Public Policy, 7(2), 195–201.

Spelman, W. and D. K. Brown. (1981). Calling the Police: Citizen Reporting of Serious Crime. Washington: Police Executive Research Forum.

Stinchcomb, J. B. (2002). Promising (and not-so-promising) gang prevention and intervention strategies: A comprehensive literature review. Journal of Gang Research, 10(1), 27–46.

Taylor, R. B. (2001). Breaking Away from Broken Windows: Baltimore Neighborhoods and the Nationwide Fight against Crime, Grime, Fear, and Decline. Boulder, CO: Westview Press.

U.S. Department of Justice. (1998). Operation Weed and Seed implementation manual. Washington: Government Printing Office.

Weisburd, D. L. and L. Green. (1995). Policing drug hot spots: the Jersey City Drug Market Analysis experiment. Justice Quarterly, 12(4), 711–735.

Weisel, D. L. and E. Painter. (1997). The Police Response to Gangs: Case Studies of Five Cities. Washington: Police Executive Research Forum.

Wilson, J. Q. and G. L. Kelling. (1982, March). Broken Windows. Atlantic Monthly, 249(3), 29–38.

2 This chapter was written with the assistance of Lysandra Marshall, Ph.D candidate, Centre of Criminology, University of Toronto.


Volume 1. Findings, Analysis and Conclusions

Volume 2. Executive Summary

Volume 3. Community Perspectives Report

Volume 4. Research Papers

Volume 5. Literature Reviews