Police Discretion with Young Offenders

III. Environmental Factors Affecting Police Discretion

4.0 The nature of the community

According to Grosman, "...Variations in police policies and priorities depend to a large extent on the nature of the community policed and the subtle pressures placed upon the force and its leadership" (1975: 7). Taking the community into consideration is important when analyzing police decision-making, as officers "are boundary personnel who are utterly immersed in the environment of the districts they patrol" (Klinger, 1997: 287). To understand the police reaction to youth crime, one must take into account the nature of a patrol district, as the maintenance of public order and law enforcement are affected by community-level characteristics and crime patterns (Hale, 1992; Klinger, 1997; Werthman & Piliavin, 1967). Thus, the stance which a police officer adopts is, to some extent, a function of the area policed. However, the police are not passive actors within an environment; rather, they adapt their exercise of authority, based on their perceptions and understandings of the areas they patrol (Meehan, 1993; Sampson, 1986).

A predominant feature of Canadian policing is the variety of environments in which departments operate and the distinct community divisions within each jurisdiction. These attributes can place differential demands on police that vary between divisions in a police force and between police jurisdictions, which may account for a portion of the regional variation in formal and informal processing of young persons. However, there is very little research that explores the social context of discretion and police behaviour across physical space (Klinger, 1997) within a Canadian context.[32]

Most American research indicates that community-level variables are significant predictors of the police arrest decision (Cohen & Felson, 1979; Crank, 1990, 1992; Hale, 1992; Klinger, 1997; McCarthy, 1991; Riksheim & Chermak, 1993). Werthman & Piliavin argue that "residence in a neighbourhood is the most general indicator used by police to select a sample of potential law violators" (1967: 76; emphasis in original). Communities characterized by low socio-economic status tend to have higher crime rates and higher arrest rates (Hale, 1992). As the level of deviance in a neighbourhood increases, a higher proportion of deviant individuals are encountered by police officers (Klinger, 1997; Skolnick, 1967; Stark, 1987), and police become more likely to arrest them (Klinger, 1997; Morash, 1984; Sampson, 1986; Stark, 1987).

Typical ecological research involves an examination of the percentage of suspects charged, the crime rate, the unemployment rate, residential stability (moving rates, home ownership/rental rate), economic variables (unemployment rate, low income family rate), and the population (or population density). However, deriving conclusions concerning police behaviour based only on ecological variables from individual police forces does not allow for the consideration of variation in the composition of areas being policed (Leonard, 1997). Thus, it is important also to incorporate qualitative data to understand police perceptions of the neighbourhoods within their jurisdiction, as the use of an entire police jurisdiction, such as Toronto, as the unit of analysis does not capture the variety of distinct communities (e.g. wealthy and poor neighbourhoods, youth hangouts, and areas with high crime rates) that might elicit different types of police response.

4.1 Degree of urbanization

A great deal of theoretical work and empirical research has been done on the relationships among urbanization, crime, and the police response to crime. Much of this work was motivated by concern about very high rates of crime in the inner-city areas of large American cities.[33] Ecological theories of crime attempt to explain this observed positive relationship between the size of a community and its crime. Urbanization theory characterizes life in big cities as culturally heterogeneous, anonymous, impersonal, and uncaring, in contrast with life in villages and small towns, which have close-knit relations, common values, and high social cohesion. Deviance and crime are constrained in small communities by various forms of informal social control exerted by family, friends, and neighbours; whereas, in the big city, deviance and crime flourish unchecked, and there is more demand for formal social control of the type exemplified by the police, and for the more formal modes of police intervention, such as arrest and charging.

Social disorganization theory elaborates on urbanization theory by specifying the circumstances in which, and the processes by which, urbanization leads to an attenuation of informal social control, and resultant higher crime rates. Urbanization is said to lead to social disorganization - a state in which the residents of neighbourhoods are unable to police themselves through informal social control because neighbourhood relationships have become dis-organized. The networks of relationships which are characteristic of small communities, and through which informal social control is exerted, do not exist, or are much more attenuated, in city neighbourhoods (Lyerly & Skipper, 1981; Gardner & Shoemaker, 1989, Weisheit, et al., 1999). Indicators of social disorganization include any social problems which inhibit the formation or activation of neighbourly social networks: poverty, unemployment, ethnic or racial heterogeneity, non-conventional family structures, and residential instability manifested by high rates of geographic mobility, the presence of transients, and high proportions of rental housing.

The basic premise of the urbanization and social disorganization theories of crime and social control - a positive relationship between the size of the municipality and the crime rate - does not hold in Canada. As Leonard put it:

Many Canadians believe that there is more crime in a large city than in a small city or rural community. However, statistics do not support this perception…..In 1995, 61% of Canadians lived in 24 major metropolitan areas…and 61% of Canada's 2.6 million Criminal Code violations occurred within these metropolitan areas. Thus, a proportionate amount of crime occurred within these big cities …Another commonly held perception is that violent crime, in particular, tends to occur in major metropolitan areas…in fact, in 1995, 58% of violent crime occurred in the 24 biggest cities, which accounted for 61% of the population…of the 18 million Canadians living within a CMA [Census Metropolitan Area, or city of 100,000 or more and its periphery], 80% lived in the nine largest CMAs…[and]…these CMAs accounted for nearly 80% of all crime. Thus, crime occurred in larger and smaller CMAs [i.e. urban areas] in equal proportions. (1997: 2)

Why does Canada differ from the United States in this regard? According to Ouimet, the high crime rates characteristic of big American cities are the result of social conditions which are specific to that nation, namely inner-city slums and ghettos characterized by extremes of poverty and social decay. In Canada, "there are no real ghettos…although some areas of major cities have become more and more disorganized over the past few years" (1999: 402-404). In this respect,

…most important is the fact that the majority of social services…are administered at the provincial or federal [and not the municipal] level…Therefore, there are no pressures toward a centralization of the poor within the limits of a central city…Also, the solution adopted by many American cities has been to group welfare recipients in contiguous areas, often in high-rise buildings. This solution has been disastrous…In Canada, many jurisdictions have instead dispersed welfare apartments across much of the city's territory. (1999: 404-405)

Several researchers have found differences between urban and rural communities in crime rates and policing patterns, particularly in the degree of formal police action employed; however, research results on this topic are mixed (Jackson, 1984; Ouimet, 2000[34]). In relatively homogenous communities or where there is a stable community power structure (as in most rural areas), the police have been found consistently to enforce the law more informally (Cain, 1973; Conly, 1978); whereas, within cities that are more heterogeneous, the police rely on a more formal approach to crime (Cain, 1973). Research has also indicated that an area with a large number of young people seems to have a lower crime rate (Jackson, 1984). This is consistent with the notion that in suburban areas with higher levels of socio-cultural homogeneity and social cohesion, there is a higher probability of using informal means to handle youth crime. In rural areas, increases in the per capita income were associated with increases in arrest rates while the percentage unemployed had no significant relationship (Crank, 1990). In contrast, increases in per capita income and unemployment were positively associated with an increase in the charge rate for urban areas (Crank, 1990). Yet, Riksheim & Chermak (1993) found that decreases in the average per capita income increase the likelihood of arrest[35] and Hartnagel & Lee (1990) found a strong negative relationship between poverty and violent crime rates.[36]

To some extent, differences in these findings may be the result of rural areas having unique characteristics. Features such as geographic isolation, economic factors, and a distinctive social climate (Weisheit, Falcone & Wells, 1999) can influence both the types of crime and the operation of the criminal justice system. Informal social control is facilitated by the "density of acquaintanceship" (Weisheit et al., 1999), which refers to the degree to which people in a geographic area are familiar with one another. Freudenburg (1986) found that communities where there were high levels of density of acquaintanceship reported being victims of crime much less frequently.

The degree of urbanization has also been found to be associated with official modes of response to crime in Canada. Caputo and Kelly (1997: 9) interviewed representatives of 150 police forces, and found that police in larger communities are more likely to say that they use warnings and pre-charge diversion. This could be the result of a larger number of community-based programs available to police for informal dispositions.[37] However, Schulenberg's (2003) ecological analysis of UCR data found that the probability of police laying charges against young persons rather than using informal action increased with population of the municipality. She also found that the probability of laying charges increased with the unemployment rate and the proportion of rental dwellings, both of which increase with municipal population.

However, ecological relationships - whether positive or negative - between formal action by police and municipal population may be spurious. It is possible that rural areas which are close to urban centres might exhibit higher charge rates as a result of youth traveling between jurisdictions (Osgood & Chambers, 2000). Also, suburban areas have lower crime rates than the urban centres and "any differences in the mix of urban/suburban areas policed by various police forces can result in artificial differences" (Leonard, 1997). Since low crime rates have been found in areas of high and low population density, (Neuman & Berger, 1988), it is possible that population is a contextual variable that might condition the effect of other variables.

Variations in style of policing among different types of communities have not been adequately addressed within the Canadian context. Treating everything outside metropolitan areas as rural can distort or conceal important patterns. It has been suggested that the differences in policing conditions among rural communities are larger than most differences between metropolitan and rural communities (Weisheit et al., 1999). This may be particularly true in Canada, because of its unique arrangements for providing police services to rural areas and small towns. Each province is responsible for providing police services to its rural areas (i.e. areas outside municipal boundaries); and, in addition, establishes a threshold municipal population, below which the provincial government undertakes responsibility for municipal policing. This threshold varies between 500 and 50,000, depending on the province (Seagrave, 1997: 32).[38] All provinces except Ontario and Québec contract with the RCMP to fulfill their provincial policing responsibilities.[39] The RCMP also provides both territorial and municipal police services in the three Territories. In addition, the Ontario Provincial Police, and the RCMP, in the eight provinces in which the RCMP provides provincial policing, provide police services under contract with "mid-sized" municipalities which are not entitled to provincial policing because they are over the threshold size. In Québec, many mid-sized municipalities are also policed by the provincial police, the Sécurité du Québec, though as a provincial responsibility, not under municipal contract. In 1995, the RCMP alone provided contract municipal police services to 201 municipalities, or 32% of all municipal police jurisdictions in Canada. Everywhere in Canada except Ontario and Québec, the RCMP provides police services to all but a handful of the largest municipalities, as well as to all rural areas (see, e.g., Dunphy & Shankarraman, 2000: 32-63). The result is that all of the vast expanse of rural Canada,[40] and many small and mid-sized towns, are policed, not by small, local agencies staffed by local residents, but by three very large, modern, professional, and bureaucratic police agencies.

Inevitably, this calls into question the applicability of much of the literature on rural and small town policing, which is based mainly on research done in the U.S.A., where small, local police agencies are the rule. For example, the authors of one text on rural and small town policing in the U.S.A. describe the policing environment thus:

…The major difference between rural and urban [policing] settings is the far greater importance of the county sheriff's office in the administration of rural policing and law enforcement…all unincorporated areas outside of municipal units…are by statute the primary jurisdiction of the county sheriff…even incorporated places in rural areas may depend on the county sheriff for basic policing services…the sheriff is an elected official in all but two states…[this] means that the sheriff is directly subject to the community and to the power of public opinion…but this does not mean that small-town municipal chiefs are free of political influences…those pressures can often be more intense and less predictable…For [small-town] municipal chiefs, the pressure is less from the electorate than from local political and business leaders…(Weisheit et al., 1999: 98-100).

The contrast between the situation of one of these county sheriffs or local municipal police chiefs in the U.S.A., and a detachment commander of the RCMP, OPP, or SQ could not be greater. Both the RCMP and OPP make explicit attempts to insulate members of their detachments from local affiliations or pressures. [41] According to Murphy, detachment police are intended to be "detached" from the community; the RCMP, OPP, and SQ do not allow members to police in their home community, and rotate them regularly from place to place in order to maintain their "detachment" from local conditions (1991: 335-336). Seagrave points out that RCMP detachment commanders see themselves as accountable primarily to their superiors in the RCMP, and, ultimately to Headquarters in Ottawa, rather than to local or provincial residents or officials; and that the official policy with respect to contract policing of the RCMP is that "…control remains with the government of Canada" (1997: 198).

To the authors' knowledge, no systematic research has been done on the conditions and style of rural and small town policing in Canada.

In order to investigate the relationship, if any, between urbanization and the use of discretion by police in dealing with youth, we classified our sample into three types of communities. A metropolitan area refers to either the core of a Census Metropolitan Area (CMA) or the equivalent urban core which has insufficient population (100,000+) to qualify as a CMA (e.g. Charlottetown, P.E.I.). A suburban/exurban area includes all other places in a CMA or the equivalent areas surrounding a non-CMA metropolitan area. A rural/small town area is any other police jurisdiction: that is, an area which is neither an urban centre nor an area peripheral to one.

Several principles motivated this classification. First, we freely acknowledge that there are many "valid" ways of defining the level of urbanization of a community (Weisheit et al., 1999: 179-196). Second, we did not want to categorize the communities in our sample simply by population. To do so would be to exclude cities such as Whitehorse or Charlottetown from being classified as urban centres, since their populations are well below 100,000. Similarly, to classify communities by population density could also be misleading, since some metropolitan areas have low population densities due to their incorporation of large rural fringe areas; and rural policing jurisdictions vary enormously in their geographic size. Third, our classification is oriented toward what we take to be a community's relationship with policing conditions. We presume that if the "degree of urbanization" affects the style of policing, it is due to differing ways of life in communities with different levels of urbanization, and our expectation is that these different ways of life are captured at least as well by this classification into urban centres, their peripheries, and rural areas and small towns which are relatively remote from urban centres, as by other typologies. Finally, although population data are readily available for municipal police jurisdictions, systematic data on the populations of rural jurisdictions are difficult to obtain. The Canadian Centre for Justice Statistics cannot supply them, because CCJS relies on population data from the Census, and is not able to match the boundaries of rural police jurisdictions with Census subdivision boundaries, as it does with municipal jurisdictions. Also, the populations of many rural jurisdictions vary substantially on a seasonal basis, due to tourism.

Before ruling out the use of community population or population density as indicators of urbanization, we calculated correlations between the proportion of apprehended youth who were charged, and municipal population and population density.[42] No relationships were found; thus confirming that simple population or population density may not be the best indicators of urbanization, when analyzing its relationship with police practices.

Figure III.2 shows the distribution of the three types of community in our sample. We tried to sample as evenly as possible between metropolitan, suburban/exurban, and rural/small town police agencies.[43] The actual sample is divided almost evenly between rural and urban/suburban jurisdictions, with 52% of the agencies and detachments located in metropolitan or suburban/exurban areas, and 48% located in rural or small town jurisdictions.

Figure III.2 Types of communities served by police agencies in the sample

Figure III.2 Types of communities served by police agencies inthe sample - If the following image is not accessible to you, please contact Youth Justice Policy at Youth-Jeunes@justice.gc.ca for an alternate format.

Description of Figure III.2

During the interviews, we raised the issue of differences between rural, suburban, and urban policing conditions and styles. The answers were inconsistent: rural officers say that they use either (1) less discretion, or (2) the same amount or more discretion, than urban and suburban police. All of our rural and small town respondents agreed that policing in an urban area would be like "night and day," compared to their experiences. Rural and small town officers suggested that police in metropolitan areas may be able to divert more youth as their use of police discretion is not as visible. In smaller communities, there is a higher density of acquaintanceship (the extent to which members of a community know each other). Rural and small town police officers said that they face a high degree of accountability to their community as they also interact with citizens socially. An officer stationed in the Northwest Territories said "even my private life is scrutinized against their moral ideas of what I should be like". It appears that in metropolitan and suburban areas, a police officer could expect respect by virtue of his or her official position; whereas, respect is earned (or not) by a rural or small town police officer on the basis of his or her character and "record". When a rural or small town officer wishes to use informal action, s/he must anticipate that members of the community will learn about, and remember, his or her handling of the incident. Thus, we found that rural officers tended to be focused on being responsive to the young person and to the anticipated responses of the members of the community, although they are, in many cases, employed by non-local police agencies, such as the RCMP or OPP.

We are not suggesting that our data portray rural and small town officers as using less discretion with young offenders. We found rural and small town officers were more likely to say that they "almost always" use informal action "with minor offences". Middle and upper management in rural and small town jurisdictions hinted at the notion that more arrests and more tickets are indicative of poor policing. It appears that making too many arrests may reflect the junior officer's inability to handle youth-related incidents informally. Respondents in rural and small town police agencies commonly referred to shoplifting and drinking offences as examples of minor offences that should be dealt with informally or sent to alternative measures. This corresponds with our findings concerning the recording of informal action. About one-half of the officers in metropolitan and suburban areas indicated that they always record their informal actions; whereas, only one-third of rural and small town officers said that they usually record their informal dealings with youth. In some of these smaller jurisdictions, officers said that recording informal action turned it into a formal action since a permanent record is created. Thus, when dealing with youth on very minor offences they would only make notations in their notebooks and notify other officers informally. This approach does not appear to inhibit the distribution of intelligence, due to the ease of communication in small police services, and the high level of knowledge these officers had about the youths in the area. However, it does suggest that the amount of informal action reported in rural and small town police agencies is underestimated by the statistics.

Suburban/exurban and rural/small town police agencies do share one characteristic in terms of using informal action with young offenders. In both of these community types, almost half of the officers indicated that they are never able to make referrals to external agencies; whereas, only about one-quarter of the officers in metropolitan areas felt they had the same limitations. This raises the issue of how local economic factors can influence the exercise of police discretion. In some cases, exurban areas can be as isolated as rural or small town jurisdictions. An exurban area does not operate in the same manner as a metropolitan or suburban area, which is more likely to have its own resources and sufficient tax base to pay for them. In many cases, the human and financial resources are minimal or nonexistent. Thus, officers in these jurisdictions must do their job in a different way.

Some writers suggest that "the larger the community the more likely citizens were to believe that police should limit their role to enforcing criminal laws" (Weisheit et al., 1999: 110). Our data support this assertion. Our respondents in smaller communities suggested that traditional law enforcement is by no means the only service their community expects of them. On the contrary, many of these communities do not have the resources to cope with many social problems, and the responsibility for compensating for this deficit falls upon the police service. We found that exurban and rural/small town police agencies showed much more concern for non-crime-related services than agencies in metropolitan areas.

Metropolitan and suburban/exurban areas also differ in how police officers decide to use informal action. These differences originate with how police become aware of youth-related incidents. Officers in suburban/exurban areas were the least likely of the three types of police to find out about youth-related incidents by a call from a parent or guardian. One of the defining criteria for a suburban/exurban jurisdiction is the concept of a "bedroom community". This implies that a significant proportion of the population commutes to a metropolitan area to work for the day. Whereas over one-half of the police officers in both metropolitan and rural/small town areas indicated that parents call them about youth-related incidents, the nature of a suburban/exurban bedroom community appears to change police-community relations. Suburban/exurban officers told us that they commonly have difficulty reaching the parents when they have arrested a young person. They also gave examples of a youth having to wait at the police station all day for one of their parents to finish work to come and get him. For minor offences, these officers may bring the youth back to school and have the principal take care and control of the young person. However, for more serious offences, this is not an option.

Officers in the different types of communities also differ in how often they become aware of youth-related incidents from another system agent. Over three-quarters of police officers in metropolitan areas indicated that they find out about youth crime through other system agents, compared to less than half of the officers in suburban/exurban and rural/small town agencies. This finding suggests differing levels of communication between organizations in the criminal justice system. As respondents generally felt that they have little or no discretion with offences referred to them by other system agents, awareness of youth-related incidents from system agents would tend to play a larger role as a factor in the proportion of youth charged in metropolitan areas. Similarly, suburban/exurban officers were the least likely to receive feedback about alternative measures dispositions.

Among the various methods used to compel appearance, clear distinctions arose between metropolitan and rural/small town police officers. Officers in metropolitan areas were the least likely to use a summons to compel the appearance of a young person. More officers in suburban/exurban and rural/small town agencies considered a summons to be a viable method of compelling appearance for youth-related incidents. They tended not to see a problem in locating the youth for service after the fact. Many metropolitan officers (e.g. in the Toronto Police Service) indicated that they deal with quite a few transient and out-of-town youth. This presents unique problems in serving process (e.g. a summons or a Notice to Parent). These officers tend to release a young person on an appearance notice or PTA whenever possible. In contrast, officers in rural/small town areas were much more likely to use a summons to compel appearance. Just under one-half of rural/small town police officers said that they rarely use appearance notices. Suburban/exurban officers fell in the middle of this continuum. This suggests that the density of acquaintanceship plays a role in the method of compelling the appearance of a young person. Compared to metropolitan areas, "rural life is characterized by greater levels of physical distance among citizens but lower levels of social distance" (Weisheit et al., 1999: 164; emphasis in original). The density of acquaintanceship varies inversely with community population (or population density), as does the feasibility of using a summons to compel the appearance of a young person.

Differences between types of community also arise in the use of conditions with an Officer in Charge Undertaking. Rural and small town officers in our sample are more likely to attach the condition of keep the peace and be of good behaviour. Officers in metropolitan and suburban areas generally told us that this clause was meaningless and had very little impact on the young person. In contrast, officers in rural and small town areas tended to feel that this condition is meaningful and enforceable. An officer in a small town in the Atlantic region gave the example of having lunch and watching a youth on this condition through the window. The young person should have been in school and was hanging around a "questionable establishment". The officer did not charge the youth with a breach but used his discretion. He was pleased to report that the youth did not get into any further trouble with the law. Granted, this is an ambiguous example, but it does shed some light on how officers view the use of this condition in smaller jurisdictions.

Officers in metropolitan areas do not seem to have the same expectation that they will come across a young person who is on release conditions in the course of an ordinary day's activities. In contrast, officers in metropolitan areas are more likely to attach a curfew to an undertaking. They said that a curfew condition is imperative to control youths who commit crimes at night. Rural officers did not feel this condition was as necessary since they tended to see "listening to your parents" (i.e. abiding by a parental curfew) as part of "keeping the peace and being of good behaviour". Again, the likelihood that an urban officer would know the parents of a young offender is much smaller.

When we asked under what circumstances an officer would detain a young person for a judicial interim release hearing, rural and small town police officers were less likely than others to cite "serious" offences, repeat offending, multiple breaches, the youth being before the courts, getting judicial bail conditions, intoxication, or the youth's best interests. Several factors appear to contribute to this finding.

First, the remote location of many rural/small town police agencies makes it difficult to detain young offenders (cf. Canadian Criminal Justice Association, 2000). Many rural and small town detachments and police agencies in our sample were more than three hours' drive away from the nearest juvenile detention facility. Thus, transportation of a juvenile would require at least six hours and two police officers. In some cases, this would constitute the entire complement on duty at that time. Sometimes they have used one officer to transport a youth; however, this occurs infrequently. Coupled with the cost of transporting the youth is the distance which the family will have to travel in order to visit their son or daughter. Officers frequently cited the family as a reason for not detaining a youth in these jurisdictions. In some cases, the family does not own a vehicle and the youth is then completely isolated from his or her family for an unpredictable period of time (as s/he may be held until trial). These officers consider it undesirable to separate a youth from his or her family in this way. As a result, rural and small town officers tend only to detain youth if they feel that they have no choice, given the nature of the offence and the circumstances of the offender.

Second, rural and small town police agencies in our sample were more likely to have written policy and protocols for handling youth (30%) than suburban/exurban (21%) and metropolitan police agencies (17%). Even though the majority of the policies and procedures for youth are not extremely detailed, they do include the sections of the Criminal Code pertaining to arrest, detention, and release of both adults and youth. Junior officers told us that they found these aids helpful in their decision-making. Further, in cases where a youth could be released on a PTA or be held for a judicial interim release hearing, a senior officer or watch supervisor might help with the decision.

In larger forces, the average patrol officer does not usually make the final decision to detain for a JIR hearing. In most cases, an officer (the "Officer in Charge"), usually a sergeant, has been assigned to the cell blocks and reviews all incoming prisoners. It is the cell block sergeant who makes the final determination concerning detention. Thus, the existence of written policy and protocol regarding detention and release is not as pivotal in the decision-making process of suburban and metropolitan patrol officers, as it would be for those in rural and small town agencies.

Officers in rural and small town areas also expressed different views from police in larger communities of the role which the nature of the offence and the characteristics of the offender play in their decision-making. These officers were more likely to say that they consider the presence of a weapon, the extent of harm done, victim/complainant preference, the relation between the offender and the victim, and the age of the offender in deciding how to deal with youth-related incidents. Suburban/exurban and metropolitan officers were more likely to cite the location and time of day of the incident, the demeanour of the accused youth, [44] and the involvement of peer groups and gangs.

In summary, analysis of the interviews suggests that rural and small town police are the most likely of the three types to use informal action to resolve youth-related incidents, and the least likely to use pre-charge diversion, due to the unavailability of programs. They also appear to be the least likely to detain youth for a JIR hearing. Police in suburban and exurban jurisdictions appear to fall somewhere between rural and metropolitan police on these dimensions of police discretion.

These findings from the interview data can be verified by analysis of UCR data on the proportion of apprehended youth who were charged. The problem with the UCR data is that they classify an apprehended youth as either charged or not charged, and do not distinguish between informal action and pre-charge diversion to a program. Thus, the higher use by rural and small town police agencies of informal action and lower use of diversion to a program which are suggested by the interview data may cancel each other out in the UCR statistics. Also, we have some suggestions from rural and small town police that they often do not record the use of informal action, which would cause the UCR to underestimate their use of it. Nevertheless, of the 85 police agencies which reported less than 95% charging of apprehended youths,[45] 40 were in rural areas and small towns, and reported charging an average of 61% of apprehended youth; whereas, the 27 metropolitan police services charged, on average, 66% of apprehended youth. Suburban/exurban police services reported charging the lowest proportion - 57% - perhaps because they are able to combine the more informal style of rural and small town agencies with the access to diversion programs characteristic of big city police agencies.

We also tried to test the applicability of social disorganization theory, which suggests that police respond to higher levels of "social disorganization" in more urban areas by relying more on formal methods of control such as arrests and laying charges. We were able to obtain data at the level of the police jurisdiction (municipality) on indicators of social disorganization for 61 non-rural communities in the sample: the mean household income of residents of the municipality, the proportion of adults with incomes below various poverty cut-off lines, proportion of households living in owned, versus rented, homes, and an index of ethnic heterogeneity. [46] None of these had any relationship with the proportion of apprehended young persons who were charged during 1998-2000 in that jurisdiction. This is not surprising, since social disorganization theory assumes a relationship between urbanization, social disorganization, and the crime rate; and it was shown above that in Canada, larger urban areas are not characterized by higher levels of crime.

The findings from the interview data suggest a different style of policing in rural and small town areas, and also some differences between policing in urban centres and their suburban and exurban fringes. Rural and small town communities have a distinctive social climate that appears also to influence this aspect of police decision-making. With a higher density of acquaintanceship, rural and small town officers feel more accountable to the community. On the other hand, detachment commanders in the RCMP and OPP are accountable to their superiors, and, ultimately, to headquarters in Ottawa or Orillia. American research has found that, in rural areas, characteristics of the community are better predictors of police style than organizational factors (Crank, 1990). As a result, individual officers are held accountable to the community for their actions to a much larger extent than in other types of communities. Rural and small town officers whom we interviewed - whether in independent municipal agencies, or RCMP or OPP detachments - suggested that the communities they police want the police to be tough on youth crime but not to incarcerate their youth. Each jurisdiction handles these pressures differently. However, to some degree it affected all of our respondents in rural and small town areas. This topic has been studied by American criminologists, but has received practically no attention in Canada, although there is enormous variation in the types of communities served by Canadian police agencies. Further research is needed that focuses exclusively on the impact of these differences.

4.2 Socio-demographic characteristics

Research on police suggests that - like most people - they believe that lower socio-economic status individuals commit more crimes (Morash, 1984; Sampson, 1986; Wilson, 1968). This belief may be partly a result of middle class crime being less evident, as it frequently occurs within private space. According to Sampson (1986), the higher vulnerability to arrest within perceived high crime rate areas is independent of the area's actual crime rate. In other words, "ecological contamination" occurs, as the type of community (based on public and police perceptions) where police-juvenile encounters take place will influence the actions taken by police (Bittner, 1970; Sampson, 1986: 884, 887; Werthman & Piliavin, 1967: 75-85). Thornberry (1973) found that lower-class youth were treated more severely regardless of their offence and prior record; however, the class differences were the largest with more serious offences.

Some Canadian research has found that police attitudes toward, or perceptions of, crime prone areas, coupled with citizen complaints, were strong predictors of official delinquency rates. Hagan et al. (1978: 100) argue that "actual class differences in the experience of juvenile crime are amplified by underclass housing conditions and complaint practices, and in turn, even more so by police perceptions", suggesting a community-specific bias in police departments (Bursik, 1988). These findings are based on neighbourhood characteristics, independent of individual level factors such as the sex, race, or demeanour of the accused, the type of crime, etc. Overall, the literature suggests that the probability of police-juvenile encounters ending in arrest declines in neighbourhoods that have a higher perceived socio-economic status.

Variations in the degree of formal police response in any given area may be partly determined by the tolerance of crime by the adults within the community. When a community has high levels of crime and deviance, only the more serious acts may be punished (Klinger, 1997; Stark, 1987). Thus, areas with high levels of crime will see a decrease in the arrest rate (relative to the number of incidents), as only the more serious offences will elicit formal action. The implication is that "variation in levels of deviance across patrol districts means that…officers will have different approaches to the police mandate to regulate deviance and different work to do and that ultimately work group negotiations will occur in different structural contexts" (Klinger, 1997: 287). [47] For example, although police officers were more likely to file official reports in high crime communities, they were less apt to file victimization reports (Smith, 1986; Worden, 1989).

We asked respondents to describe their communities in terms of level of wealth, age profile, ethnic or racial diversity or homogeneity of the residents, and whether they considered the population to be predominantly stable or to include a substantial number of transients. We then attempted to relate their characterizations of their communities with the ways in which they exercise their discretion with youth. Although interviewees' descriptions of their communities might be considered "subjective", they are, in our view, at least as relevant as "objective" indicators of community characteristics, because it is surely officers' perceptions of their environment, regardless of their accuracy, which condition their attitudes and decision-making. However, we also did statistical analysis of the correlations between the proportion of apprehended youth who were charged by a police service during 1998-2000, and indicators of socio-demographic characteristics of the community.

4.2.1 Level of wealth

Because our respondents police entire towns, cities, and even regional municipalities, our queries about the overall "level of wealth" and income in the community elicited rather vague answers. How can one characterize the level of wealth of a heterogeneous area like Toronto, or even Kelowna? However, we also asked whether there is an area in their jurisdiction that presents a policing problem due to low income or poverty. Almost one-quarter (23%) of the sample answered in the affirmative. This was much more common in metropolitan areas: 50% of metropolitan police services identified a poverty problem area, compared with only 16% of suburban/exurban services, and 9% of rural and small town services. Poverty problem areas were identified most often by police in the Prairies and least often by police in the Territories and Quebec (Figure III.3).

Figure III.3 Police services in communities with poverty problem areas, by region

Figure III.3 Police services in communities with poverty problem areas, by region - If the following image is not accessible to you, please contact Youth Justice Policy at Youth-Jeunes@justice.gc.ca for an alternate format.

Description of Figure III.3

Analysis of UCR data on the proportion of apprehended young persons who were charged showed that police services which identified a poverty problem area were slightly more likely to charge: on average, 65% of apprehended young persons were charged in these areas, versus 61% in areas with no identified poverty problem.

Officers in agencies and detachments which police poverty problem areas were slightly more likely (100%) than others (90%) to say that they use informal warnings. Similarly, they are more likely (45%) to use formal warnings than agencies with no identified poverty problem area (27%). This suggests, contrary to the findings from analysis of UCR data (above), that formal action (laying charges) is used less by police agencies which serve "problematic" areas, corresponding with the theory discussed above that in communities with high levels of crime and deviance, only the more serious acts are punished. However, this does not necessarily contradict the opposing theory that the use of arrest in police-juvenile encounters is lower in neighbourhoods that have a higher perceived socio-economic status, since our data on formal and informal action pertain to entire towns and cities.

Police in agencies with an identified problem poverty area in their jurisdiction were less likely to say that they would resolve an incident by questioning a youth at home or at the police station (18%) than police in other jurisdictions (30%). We can only speculate as to the reasoning behind this trend. It appears that officers policing these types of areas are more likely either to deal with a young person informally at the scene, or to charge him or her. They appear to be less inclined to bring a youth back to the station unless they are going to process a charge.

We found no significant differences between police services which do or do not have poverty problem areas in relation to their use of pre- or post-charge alternative measures.

Police services with poverty problem areas also tend to differ from the others concerning the method of compelling appearance. They are less likely to use a summons (45%) than other police services (19%), and more likely to say that they use an appearance notice for "minor offences" (55%) than other police services (23%). This may be due to the mobility of youth and the difficulty of serving process in metropolitan areas.

Agencies policing poverty problem areas are also more likely to attach certain conditions to OIC undertakings. Figure III.4 (below) shows the frequency of use of various conditions with young persons by police services with and without poverty problem areas.

Police services that police low-income and problem areas are more likely to attach the condition of no go, no association, keep the peace and be of good behaviour, no alcohol or drugs, no weapons, and attend school.

Similarly, police agencies that police these areas will tend to detain young persons more often for specific reasons. Figure III.5 (below) shows the frequency of reasons given to detain young offenders.

Police agencies that police poverty problem areas are more likely to detain for repeat offenders, if they are before the courts, multiple breaches, intoxication (alcohol or drugs), bail conditions, in the best interests of the young person and to say that they "follow the law" in detention and release decisions. Similarly, these police agencies are also more likely to "almost always" detain a young person for a serious offence (77%), repeat offenders (50%), and to get judicial release conditions (45%), compared to agencies that did not identify a poverty problem area (55%, 29%, and 11% respectively).

Figure III.4 Conditions of OIC Undertakings by low-income/poverty area

Figure III.4 - Conditions of OIC Undertakings by low-income/property area - If the following image is not accessible to you, please contact Youth Justice Policy at Youth-Jeunes@justice.gc.ca for an alternate format.

Description of Figure III.4

Figure III.5 Reasons to detain for a JIR hearing, by poverty problem area

Figure III.5 Reasons to detain for a JIR hearing, by poverty problem area - If the following image is not accessible to you, please contact Youth Justice Policy at Youth-Jeunes@justice.gc.ca for an alternate format.

Description of Figure III.5

We also calculated correlations between the proportion of apprehended youth charged during 1998-2000 in each jurisdiction, and indicators of its level of wealth: the mean income of adults in that municipality, the proportion of adults living below a certain level of income (using several different cut-off points), and the proportion of households living in owned, versus rented, homes. None of these had any relationship with the proportion of apprehended youth who were charged.[48]

Overall, the existence of an identified area of poverty appears to affect police decision-making to a certain degree, in relation to the use of informal action, laying charges, and compelling appearance. We cannot be sure, however, because most of these police services are located in metropolitan areas, so the differences could be due more to the conditions of metropolitan policing than to the existence of the poverty problem area per se.

4.2.2 Residential instability

Since the literature suggests that police are more likely to use formal action in areas characterized by "residential instability", we calculated correlations between the proportion of apprehended youth who were charged during 1998-2000 in each jurisdiction and two indicators of instability: the proportion of rented homes in the municipality, and the proportion of the population who had moved in the past 5 years.[49] No relationships were found.

We also asked officers whether the population which they served tended to be stable or to experience significant geographic mobility. Officers in 28% of our sample answered this question by identifying a problem with young persons who were "transients".

The distribution across Canada of police services which identified a significant population of transients is shown in Figure III.6. Police services in metropolitan areas were much more likely (60%) to identify a problem with transient youth than rural and small town (18%) or suburban/exurban agencies (5%). One would expect, then, that it would be mainly independent municipal agencies that deal with transients, but, in fact, 42% of the OPP detachments in the sample mentioned transient youth, compared with 32% of independent municipal agencies, 19% of RCMP detachments, and none of the three First Nations police services. Many of the transients identified by these OPP officers in our sample, who were stationed predominantly in small Northern Ontario municipalities, are "passing through" areas which straddle major highways, such as the Trans-Canada Highway; and do not fit the stereotype of the urban, homeless, skid-row transient.

Figure III.6 Police services in communities with significant transient youth, by region

Figure III.6 Police services in communities with significant transient youth, by region - If the following image is not accessible to you, please contact Youth Justice Policy at Youth-Jeunes@justice.gc.ca for an alternate format.

Description of Figure III.6

Analysis of UCR data showed that police services which identified transients were no more or less likely to lay charges against apprehended youth.

Officers in police services that have a significant transient population were more likely (89%) than those which do not (71%) to say that they use informal action with young persons. They were also more likely to say that they use each individual type of informal action (informal warning, formal warning, parental involvement, taking the youth home or to the police station, questioning the youth at home or at the police station, or referring a youth to an internal or external program). Thus, it appears that police officers that have transients within their jurisdictions are on the whole more likely to use informal action in all types of cases. There were no significant differences in the use of pre- and post-charge alternative measures between police agencies which did and did not mention significant transient populations.

Differences also emerge in the methods that are used by police services who police transients to compel appearance in court. Over one-half (52%) of police agencies that police transients said that they will use an appearance notice when no other options are available compared to 39% of the agencies that do not police transients. This is not surprising since over one-half of these police agencies (56%) said that they do not use a summons to compel attendance, compared to 29% of the agencies which do not police transients. Another 37% of agencies policing transients "rarely" use a summons.

If the youth is arrested and brought back to the police station, police agencies that police transients are slightly more likely to use a Promise to Appear with an attached OIC undertaking (70%) than other agencies (56%).

The various reasons which were given for detaining a youth for a JIR hearing were all offered more frequently by officers in police services with significant transient populations. They are twice as likely to say that they detain "in the best interests of the youth" (30%) as agencies that do not police transients (16%). They are considerably more likely (41%) than agencies which do not police transients (24%) to say that they will detain a youth in order to get bail conditions. They are twice as likely (41% versus 18%) to detain a youth because s/he is under the influence of alcohol or drugs. More than half (59%) of the police agencies with transients in their community will detain a youth for multiple breaches compared to only just over one-quarter (26%) of the agencies without significant transient populations. Almost three-quarters (70%) of police agencies with a transient population will detain a repeat young offender compared to just over one-third (37%) of the agencies without transients. Finally, all of the responding agencies that indicated they would detain a youth in order to get them admitted to a program have a significant proportion of transients. Thus, it is not surprising that agencies that police a transient population are more likely to say that they "almost always" detain young offenders for serious offences, repeat young offenders, due to departmental policy, to get release conditions, and for offences committed under the influence of alcohol or drugs. It is striking that 81% of these police agencies will "almost always" detain for serious offences, compared to just over half (51%) of those agencies who do not police a transient population.

Thus, police agencies which police a significant transient population have different patterns in their use of informal action and methods of compelling appearance. Our data suggest that transient youth present a different type of problem to the police, which they see as requiring a different type of response. Officers in all types of communities agreed that transient youth are usually youth in crisis as a result of alcohol and/or drug addictions, abusive home situations, or prostitution. Respondents emphasized that these youth are the ones that require the most intervention and are at the highest risk of re-offending.