Findings

Seriousness of the alleged crime
Prior contacts with the police
Presence of a weapon
Injury to a victim
The relationship between the victim and the apprehended youth
Group crime
Age of the youth
Sex of the youth
The youth’s aboriginal status
Summary

Seriousness of the alleged crime

Table 1 shows the percentage of apprehended young persons who were charged, by offence category. Clearly, the type of alleged offence has a large influence on the probability of a charge being laid: a youth apprehended for mischief or arson has a one in three chance of being charged; those apprehended for major offences against the person and offences against the administration of justice are almost sure to be charged. However, the percentages shown in Table 1 suggest that the probability of charging is not related in a simple way to the “seriousness” of the offence, unless one believes that failure to appear in court, provincial traffic violations, etc. are exceeded in seriousness only by murder. Thus, other factors than simply the seriousness of the alleged offence appear to have an impact on the decision.

The second column of percentages are adjusted to remove the confounding effects of related factors, such as the youth’s age and prior contacts. These are the percentages of youth apprehended for each category of offence who “would have been charged if everything about all the alleged offences and offenders were the same, except for the type of offence”. For example, 86% of youth apprehended for robbery were charged, but the adjusted percentage is only 74%. This is because robbery tends to be committed by older youth with more prior contacts, and these factors contribute to making robbers more likely to be charged; but 74% would have been charged if robberies were committed by youth who were of average age and with an average number of prior contacts.

Table 1. Proportion of apprehended youth charged, by type of offence, Canada (parts), 2001

Very little discretion not to charge apprehended youth is exercised by police in recorded incidents involving certain types of offences (Table 1). These include murder and attempt murder, and administrative offences such as failure to appear for court, bail violations, offences under the Young Offenders Act (almost all being failure to comply with a disposition), breach of probation, and escaping custody or being unlawfully at large. They also include drinking-driving offences and provincial traffic and liquor offences. It is possible that police exercise discretion in many such incidents but do not record the incident in the police information system, with the result that it is not reported to the UCR Survey. Cases involving these types of offences, and certain others (dangerous operation of a motor vehicle, aggravated assault and sexual assault) were omitted from further analysis, because little or no discretion is exercised by police in recorded instances of these offences and/or there were too few youths in the “not charged” group for reliable statistical analysis.

Evidently, a considerable amount of the variation in charging rates for different types of offences is due to related factors, since the range of variation is narrowed considerably when the influence of other related factors is statistically controlled. The main related factors are the youth’s age and record of prior contacts. Older youth commit more serious offences (Table 2), and more serious offences tend to be committed by youth who have had previous contacts with the police (Table 3). Thus, part of the reason why some types of offences are charged in relatively high proportions is that they are committed by older youth and/or youth with longer prior records. Robbery and the more serious property offences (e.g. break and enter, possess stolen property, theft over) are examples. When we control statistically for these related factors, the charge rate for these offences is reduced (Table 1). On the other hand, arson and level 1 sexual assault tend to be committed in higher proportions by younger youth, and/or those with fewer prior police contacts, so the charge rate increases when these related factors are statistically controlled.

Table 2. Proportion of each age group apprehended, by offence category, Canada (parts), 2001

Table 3. Proportion of apprehended youth, by type of offence and number of prior police contacts, Canada (parts), 2001

Top of Page

Prior contacts with the police

Prior contacts with the police play an extremely significant role in the decision to charge an apprehended youth. The first column of Table 4 shows the actual percentage charged. Apprehended youth with five or more prior contacts are more than twice as likely as those with no previous contacts to be charged. The second column of percentages are adjusted to remove the confounding effects of related factors, such as the youth's age and the seriousness of the current alleged offence. Even when related factors are controlled, the probability of a charge being laid rises with increasing numbers of prior contacts, from 32% of those with no previous contacts to 66% of youth with five or more prior contacts.

Table 4. Proportion of apprehended youth charged, by the number of prior police contacts, Canada (parts), 2001

The number of prior contacts with police is also related to every other factor affecting police decision-making which was analyzed in this research, and therefore must be statistically controlled when assessing their impact (Tables 5 and 6). The number of prior contacts increases with the age of the apprehended youth: with each year of age, the probability of an apprehended youth having no prior contacts decreases by an average of 5%, and the probability of having 5 or more prior contacts increases by an average of more than 2% - from 2% of 12 year olds to 16% of 17 year olds (Table 5).

Table 5. Number of prior contacts, by the age of the apprehended youth, Canada (parts), 2001

The number of prior police contacts is related to the presence of a firearm, major injury to a victim, the type of relationship, if any, between the youth and a victim in the incident, solo versus group offending, and the apprehended youth's sex and aboriginal status (Table 6). Apprehended youth with 5 or more prior contacts are much more likely to have a firearm (although the probability is still extremely low), and to cause major injury to a victim, to have a stranger as a victim, to act alone, and to be male and aboriginal.

Table 6. Proportion of apprehended youth, by number of prior police contacts and other explanatory variables, Canada (parts), 2001

Top of Page

Presence of a weapon

Table 7 shows the proportion of apprehended youth who were charged, by the presence and type of weapon. The UCR2 records information about weapons only in incidents involving an alleged offence against the person; thus there are only small numbers of youth in this analysis. The presence of a weapon, especially a firearm (which is rare) during the commission of a youth crime greatly increases the probability of charging, even when other relevant factors are controlled. The percentage charged for incidents involving a firearm is substantially reduced when other related factors are controlled, partly because of the relationship with the number of prior police contacts (Table 6), and partly because the presence of a firearm usually results in the classification of the offence as a serious indictable offence; therefore much of the impact of this variable is already accounted for by the variable, Seriousness of the Alleged Crime, which is discussed above.

Table 7. Proportion of apprehended youth charged, by the presence and type of weapon, offences against the person, Canada (parts), 2001

Top of Page

Injury to a victim

Table 8 shows the relationship between injury to a victim and the likelihood of charging. Obviously, injury is a factor only in incidents involving offences against the person. Major injury to a victim is rare, but greatly increases the probability that charges will be laid. The increase is much less when other related factors are controlled, because major injury usually results in the classification of the offence as a serious indictable offence; so much of the impact of this variable is already accounted for by the variable (See Seriousness of the Alleged Crime).

Table 8. Proportion of apprehended youth charged, by the type of injury to a victim, offences against the person, Canada (parts), 2001

Top of Page

The relationship between the victim and the apprehended youth

The relationship between a victim and an apprehended youth plays a significant role in the decision to charge, even when other relevant factors are controlled (Table 9). This variable is coded in the UCR2 only for incidents involving an alleged offence against the person. The probability of a charge is higher if the victim is a parent or close friend, and lower if s/he is another family member or an acquaintance. As with other circumstances of the incident, the percentage differences are much reduced when other factors are controlled. The association between the victim-youth relationship and the number of prior police contacts is shown in Table 6. Also, young persons tend to commit different types of offences against different types of people: robbery and major assault and sexual assault against strangers, and level 1 assault and sexual assault against family members and/or close friends and acquaintances (Table 10).

Table 9. Proportion of apprehended youth charged, by the relationship between a victim and the apprehended youth, offences against the person, Canada (parts), 2001

Table 10. Proportion of apprehended youth, by type of offence and relationship with victim, offences against the person, Canada (parts), 2001

Top of Page

Group crime

A youth who allegedly commits an offence with one or more accomplices is less likely to be charged (Table 11). Group crimes allegedly committed by youth tend to be the least serious, such as theft under $5,000, and to be committed by youth with fewer prior police contacts (Table 6), and by younger youth (Carrington, 2002). Therefore, controlling for related factors reduces the difference in the estimated probabilities of being charged between alleged solo and group offenders.

Table 11. Proportion of apprehended youth charged, by whether accomplices were involved, Canada (parts), 2001

Top of Page

Age of the youth

The age of the youth plays a major role in the decision to charge. An apprehended seventeen year old is more than twice as likely to be charged as a twelve year old (Table 12, first column). Some of the effect of the youth's age is mediated by other factors, especially his or her accumulated record of prior contacts (Table 5), and the increasing seriousness of offences committed (Table 2). However, even when these other factors are held constant, for each additional year of age, the probability of being charged increases by approximately 10% over the previous year, so that a seventeen year old whose offence, prior contacts, etc. are the same as those of a twelve year old, still has a more than 50% higher probability of being charged (Table 12, column 2). Some of this differential might be due to factors not included in the statistical analysis, such as the availability of diversion programs, the demeanour of the youth, or the role of the parents, but it seems unlikely that these could account entirely for the clear relationship shown in the second column of Table 12.

Table 12. Proportion of apprehended youth charged, by the age of the youth, Canada (parts), 2001

Top of Page

Sex of the youth

Although apprehended male youth are more likely than females to be charged (Table 13, column 1), practically all of this difference disappears when other related factors are statistically controlled.

Table 13. Proportion of apprehended youth charged, by the sex of the youth, Canada (parts), 2001

Top of Page

The youth’s aboriginal status²

There is a large difference (19%) between the charge rates for apprehended youth identified as aboriginal and  those who are identified as non-aboriginal or whose aboriginal status is not known or not reported (Table 14). Some of this difference is due to related factors, such as the history of police contacts (Table 6), but when these are controlled, apprehended aboriginal youth are still 12% more likely to be charged. It is not possible to determine from the available data whether this substantial difference is due to other related factors which were not included in the statistical analysis, such as the youth’s demeanour, the role of the parents, the victim’s preference,³ the availability of diversion programs which could serve as alternatives to charging, etc.

Table 14. Proportion of apprehended youth charged, by the aboriginal status of the youth, Canada (parts), 2001

Top of Page

Summary

Table 15 shows the relative importance of the situational factors affecting police decision-making with youth which could be analyzed in this paper. The value of partial eta squared is an indicator of the overall impact of the variable on the entire population of police decisions, excluding the types of offences in which little or no discretion not to charge is exercised (Table 1). A variable may have a large impact on individual incidents (indicated by large adjusted percentage differences in the tables above) while having only a small overall impact (indicated by a small value of partial eta squared in Table 15). For example, major injury to a victim increases the adjusted probability of the accused youth being charged by 12% (Table 8), but injury to a victim has very little overall impact in this population of police decisions (Table 15). The reason is that major injury to a victim was recorded only rarely in the incidents which were analyzed (in 179 of 30,812 incidents), so it has little overall importance. The same is true of the aboriginal status of the accused youth: youths who are coded as aboriginal have a substantially higher probability of being charged, but there are very few of them (Table 14), so this factor has very little overall impact.

Table 15. Overall ranking of situational factors affecting police discretion with apprehended youth

The most influential factor overall is the history of prior contacts with the police. Next in importance comes the seriousness of the offence, as indicated by its Criminal Code classification. The age of the youth, and whether the incident involved a lone alleged offender or a group, also have substantial impacts on police discretion. The other two indicators of the seriousness of the incident – the presence of a weapon and the level of injury to a victim – have only minor overall impacts on police decision-making. However, their minor overall impacts are due to their rarity: when a weapon or major injury is present, the probability of charges being laid is elevated considerably. The remaining variables – the sex and aboriginal status of the youth, and any relationship between the accused youth and a victim – have minimal overall importance when other factors are controlled.

Both hypotheses of this research were strongly confirmed. The number of prior contacts with police has a substantial impact on police decision-making with youth. The probability of charges being laid increases substantially with the number of prior police contacts, even when other factors are controlled, and the overall impact of prior contacts is the greatest of any variable considered in the multivariate analysis of factors affecting police decision-making. The number of prior contacts with police is also correlated – strongly, in most cases - with every other explanatory variable. Thus, it is crucial to control for prior police contacts when assessing the impact of other correlated factors on the police disposition – as is clear from a comparison of the raw and adjusted percentage differences in the tables of police dispositions by such variables as offence seriousness and the age of the youth.

Footnotes

2. Results based on the ‘aboriginal status’ variable in the UCR2 Survey must be interpreted with caution, for two reasons: (i) some police services which report to the UCR2 Survey do not report data for this variable; and (ii) the variable is coded as “unknown” for a large proportion of individuals. Reporting police services use either self-identification or visual identification to identify aboriginal status.
3. Some research on the question of whether the suspect's race affects police decision-making in the USA has concluded that the apparent over-charging of black youth is explained by the expressed preferences of (black) complainants (e.g. Black and Reiss, 1970).