The Impact of the Youth Criminal Justice Act on Police Charging Practices with Young Persons: A Preliminary Statistical Assessment
The objective of this research was to use available statistical data to assess the initial impact of the YCJA on police charging practices with apprehended youth. The only source of national data on police practices is the Uniform Crime Reporting (UCR) Survey, operated by the Canadian Centre for Justice Statistics, a division of Statistics Canada. This Survey is a census of all criminal incidents and alleged offenders dealt with by all police services in Canada. UCR data are collected on a monthly basis, but released once a year, generally about eight or nine months after the end of each calendar year. At the time when we began the research, data for 2003 were the most recent available.
Since 1988, police services in Canada have had a choice between two formats for reporting data to the UCR Survey. The older “aggregate” format – sometimes called the “aggregate UCR Survey” – was designed when many police services did not yet have electronic Records Management Systems, and requires only the reporting of aggregate monthly numbers of youth charged, youth not charged, etc. The “incident-based” reporting format – sometimes called the “incident-based UCR Survey”, or UCR2 Survey – requires much more detailed information, reported at the level of the individual incident, alleged offender, and victim. Although the objective of the UCR Survey is that all police services will eventually use the UCR2 reporting format, adoption by a police service of the UCR2 necessitates modification of its Records Management System, so the new format is being phased in gradually. In 2003, police services accounting for approximately 61% of reported crime in Canada were using the UCR2 reporting format (Canadian Centre for Justice Statistics, 2004: 73). In order to report aggregate crime data for the whole of Canada, the Canadian Centre for Justice Statistics converts data reported by police in the incident-based (UCR2) format to the aggregate format and combines it with the aggregate data supplied by the remaining police services (Canadian Centre for Justice Statistics, 2004: 78).
The UCR2 format was modified as of April 1, 2003 to require reporting of the type of extrajudicial measure used with each chargeable youth who was not charged (Canadian Centre for Justice Statistics, 2003). Therefore, our original research plan included analysis of this data element, for the police services which report to the UCR2 Survey. However, due to some reporting problems, the data reflecting the new requirements for reporting extrajudicial measures will not be available until the release of the 2004 UCR2 data (private communication, Canadian Centre for Justice Statistics, March 8, 2005 ).
The results reported below are based on analyses of data covering all of Canada, supplied for this research in the form of custom tabulations by the Canadian Centre for Justice Statistics. These tabulations are based on the aggregate UCR Survey (i.e. data reported by some police services in the aggregate UCR format, combined with data reported by other police services in the incident-based UCR2 format and converted to the aggregate format).
For the purposes of the present research, the two key data elements of the UCR are the per capita rate  of young persons who were charged or recommended by police to be charged,  and the rate of youth who were not charged. The sum of these two quantities is the total rate of chargeable youth, also known as the police-reported youth crime rate. A “chargeable” person  is defined by the UCR2 Survey as one who
“has been identified by police as being involved in a criminal incident and against whom an information [i.e. charge] could be laid as a result of sufficient evidence/information” (Canadian Centre for Justice Statistics, 2004: 80). Thus, the rate of young persons who were chargeable but not charged is an indicator of the use of extrajudicial measures, since extrajudicial measures in the YCJA are measures other than judicial proceedings (i.e. other than laying a charge) used with a young person whom an officer has reasonable grounds to believe has committed an offence.
There has been some controversy in the criminological literature concerning the accuracy of the UCR data on youth not charged.  The problem is that the UCR Survey is a form of “administrative data” – that is, data which are extracted from records created and maintained by organizations (in this case, police services) primarily for operational purposes, rather than to serve the needs of criminological research. For police operations, it is crucial to maintain complete and accurate records of youth who are charged, but not so important to record every youth who could have been charged but was not. Indeed, the notion of “chargeable” is open to individual interpretation. Individual police officers could be expected to vary considerably in the completeness of their recording of chargeable youth, and research has found that police services vary widely in the numbers of youth which they report as “not charged”, relative to the numbers which they report as charged. Some police services report zero youth not charged, so that their calculated “proportion of chargeable youth who were charged” – an indicator of the level of the (non-)use of police discretion – is 100%. Because of the unreliability of the reporting by some police services of numbers of youth not charged, some criminologists have warned against using this data element in analyses (e.g. Hackler and Don, 1990; Hackler and Paranjape, 1983, 1984; Markwart and Corrado, 1995). Others, including the authors of this report, have argued that although this data element should probably not be used to compare the charging practices of individual police services, the biases in reporting are sufficiently stable over time that it can be used in time series analyses, especially when it is aggregated to the level of the province or territory (Carrington, 1995, 1999; Carrington and Schulenberg, 2004a; Scanlon, 1986).
The coming into force of the YCJA introduced a new consideration concerning this methodological issue. Unlike the Young Offenders Act, the YCJA explicitly addresses the exercise of discretion by police in the decision whether to lay a charge against an alleged young offender, or to deal with him or her by extrajudicial measures.  Police are required by the YCJA to consider using extrajudicial measures before laying a charge. One would expect that, in response to this requirement, individual police officers would begin to record more instances of their use of extrajudicial measures, in order to demonstrate compliance with the Act, and that police services would modify their Records Management Systems and their reporting expectations for their officers, in order to monitor compliance. The new requirement in the UCR2 Survey for more detailed reporting of extrajudicial measures (see above) has necessitated modifications to police Records Management Systems, from which the UCR2 data are extracted. This, in turn, can be expected to increase the reporting by police officers and police services of the use of extrajudicial measures.
These developments suggest the distinct possibility of an increase in 2003 in the number of young persons reported to the UCR Survey as receiving extrajudicial measures (formerly, “youth not charged”),  which in reality reflects nothing more than improved recording and reporting to the UCR Survey by police. The analysis reported in Section 3 takes this possibility into account.
This research uses the interrupted time series design: that is, a series of annual measurements (in this case, of youth charged and youth not charged) is “interrupted” by an intervention whose impact is to be assessed – in this case, the coming into force on April 1, 2003 of the YCJA. Changes in the level and/or trend of the time series following the intervention are interpreted in this design as evidence of the effect of the intervention.  Although many of the changes reported below are based on comparisons of levels in 2002 and 2003, the analyses always refer to the entire time series from 1986 to 2003 as the context for the comparisons, for two reasons: first, to protect against the possibility that 2002 was an anomalous year; and, second, to protect against the possibility of misinterpreting changes from 2002 to 2003 as evidence of the effect of the YCJA, when in fact they might have been merely the continuation of pre-existing upward or downward trends.
Since the UCR data for 2003 include three months (January to March) when the YOA was in force, we also analyzed quarterly data for 2001 to 2003. These data enabled us to distinguish between the levels of the statistical indicators during the first quarter of 2003, under the YOA, from levels during the last three quarters under the YCJA . The quarterly analyses also enabled us to identify changes which began to occur during the second quarter of 2003. One can have more confidence that such changes were the result of the YCJA than changes known only to have occurred at some time in 2003. 
Due to implementation of a new Records Management System, UCR data for the Toronto Police Service were not available for September to December, 2003, and were estimated by the Canadian Center for Justice Statistics, based on data for the same period in 2002 (Canadian Centre for Justice Statistics, 2004: 75). This posed a potential problem for the research, since Toronto accounts for approximately 12% of reported youth crime in Ontario. The period in question includes four of the nine months during which the YCJA was in force, and the substitution of data from 2002 could be expected partially to obscure any impact of the YCJA on police charging practices in Toronto, and therefore to bias downwards any assessment of its impact on police practices in Ontario. Therefore, we obtained two sets of data for Ontario and for Canada - one which included Toronto and one which excluded it – and have reported the results for analyses of both datasets. As it turned out, the inclusion or exclusion of data for Toronto did not materially affect the overall results.
It is possible, using UCR2 data, to construct variables capturing aspects of the prior contacts with police of a chargeable person.  This is done by searching the data for other records pertaining to the same person (which is not a straightforward process, since personal identifiers such as name, address, etc. are not captured by the UCR2 Survey). Such a procedure would have enabled us to address the hypothesis concerning the impact of the provision in Section 4(c) that extrajudicial measures are presumed to be sufficient to deal with a non-violent first offender. However, constructing a prior record variable using UCR2 data is a highly resource-intensive procedure which (like any record linkage project using Statistics Canada data) requires the approval of the Chief Statistician of Canada. We concluded that the time and resources which would have been required to carry out such an analysis were not warranted by its likely contribution to the results of this preliminary assessment.
-  All numbers of young persons were standardized to rates per 100,000 youth population by dividing by the relevant annual youth population estimate, supplied by the Canadian Centre for Justice Statistics.
-  In New Brunswick, Quebec, and British Columbia, the decision whether to lay a charge against a young person is made by the Crown, on receipt from police of a recommendation to charge. UCR data on youth charged for New Brunswick and parts of Quebec report the numbers of youth who were actually charged, as a result of the Crown's decision. UCR data for British Columbia and the remainder of Quebec reflect the numbers of youth who were recommended by police to be charged, not the numbers actually charged (Canadian Centre for Justice Statistics, 2004: 79).
-  The term used by Canadian Centre for Justice Statistics is “charged/suspect-chargeable (accused)”, which we have shortened to “chargeable”. In this report, we use the term “apprehended youth” to refer to the concept of a young person who a police officer has reasonable grounds to believe has committed an offence. We use the term “chargeable youth” to refer to the UCR data element which is used in the analysis as an indicator, or operationalization, of the concept.
-  See, e.g., Carrington, 1995, 1999; Federal-Provincial-Territorial Task Force on Youth Justice, 1996; Hackler and Don, 1990; Hackler and Paranjape, 1983, 1984; Markwart and Corrado, 1995.
-  See Section 1.2 above.
-  And therefore, an increase in the total number of youth reported as chargeable (that is, in the police-reported youth crime rate).
- See , e.g., Campbell and Stanley, 1963: 37-42; Mohr, 1995: Chapter 9.
-  The precise location in time of a change in the time series is a valuable form of protection against a threat, usually referred to as “history”, to the validity of conclusions from an interrupted time series evaluation design. The threat of “history” is that changes in the phenomenon of interest are mistakenly attributed to the intervention when in fact they are due to other events which took place at approximately the same time as the intervention. For example, if there had been a substantial change in 2003 in the underlying level of youth crime, unrelated to the coming into force of the YCJA, it would probably have caused changes in the levels of reported youth charged and youth not charged, which could be misinterpreted as an effect of the YCJA. The more precisely the research can identify the point in time when the change occurred, the more confident one can be in distinguishing between effects of the intervention and effects of other social changes.
-  For examples of analyses of UCR2 data incorporating a prior contacts variable, see Carrington and Schulenberg (2004a; 2004c) and Schulenberg (2004).
- Date modified: