Police Discretion with Young Offenders

Methodological Appendix

II.  Statistical data on young offender cases

Custom tabulations of statistical data from the Incident-Based Uniform Crime Reporting (UCR2) Survey were provided by Canadian Centre for Justice Statistics.

For the analyses reported in Chapter II, we used a tabulation of all youth-related incidents reported to the UCR2 for 2001, broken down by province and clearance status of the incident; and a tabulation of all youth-related incidents reported for 1995-2001 by a subset of police services which have been reporting continuously to the UCR2 between 1995 and 2001 (the "Trend Database").

For the analyses reported in Chapter V, we used a tabulation of all young persons apprehended in 2001 who were reported to the UCR2 by a subset of police services. This tabulation was broken down simultaneously by the police disposition (charged vs. processed otherwise) and several independent variables. Because this tabulation incorporated information pertaining to the years 1995-2001 (see below), the sample of police services was restricted to the Trend Database (see above). It was further restricted by omitting one police service (Toronto) which does not report youth who are not charged - since the dependent variable in the analyses was whether or not the youth was charged. The resulting sample included 186 police services in 6 provinces: New Brunswick, Quebec, Ontario, Saskatchewan, Alberta, and British Columbia.

The youth's record of prior criminal activity has been identified as an important determinant of the police disposition, both by previous research and by interviews in the present research. Special programming work was required in order to create this variable, since it is not routinely captured by the UCR2. The work was done for this project by staff of Statistics Canada and the Principal Investigator. The procedure involved searching through all UCR2 records for 1995-2001 for the selected sample of police services, and matching records of apprehensions pertaining to youths apprehended in 2001. Each record (except the latest apprehension in 2001, which was the apprehension whose outcome was being analyzed) constituted one prior apprehension. These were counted and classified.

Matching of records for the same person was not straightforward, since there is no unique person identifier in the UCR2. The person's surname is encoded in a 4-character SOUNDEX code, which is not unique; i.e. many surnames are encoded with the same SOUNDEX. Thus, simply matching on the SOUNDEX would result in many false positive matches; i.e. many records for different people would be erroneously treated as prior apprehensions of a single person. The result would be an underestimate of the number of unique persons and an overestimate of the length of their prior records. This is not necessarily as great a problem in the present research as it might be in other types of research, because we are not concerned here with prior record in itself, but in its correlation with the probability of being charged. In general, errors in measurement of variables (such as overestimates of prior records) result in attenuation of correlations, so the result of such error would be a small underestimate of the impact of prior record on police dispositions, and a small overestimate of the impact of other related variables, such as the youth's age. False positives can be greatly reduced by matching simultaneously on SOUNDEX, birth date, and sex (which are all in the UCR2), but are still a potential problem.

Methodologists at Statistics Canada conducted an exhaustive analysis of the probability of false positive matches by comparing the rate of occurrence of each SOUNDEX in the UCR2 with the rates of occurrence of the corresponding surnames in the populations of the provinces of Canada, using electronic telephone directories. This enabled them to establish, for each SOUNDEX, the expected rate of false positives, when it was used for matching in combination with birth date and sex. SOUNDEXES vary greatly in their vulnerability to false positive matches, since some encode very common surnames and others do not. Assessments of SOUNDEX "match quality" (i.e. non-vulnerability to false positives) were made under the assumption that UCR2 records would be matched only within the police services in a Census Metropolitan Area (CMA), or within the jurisdiction of individual police services outside CMA's (since there was no obvious principle with which to group non-CMA police services). Consideration was also given to the possibility of matching within larger areas, such as an entire province, in order to capture a youth's apprehensions in different jurisdictions. The basic principle here is that the probability of false positives is directly related to the size of the population within which one is matching.

On the basis of this quality analysis, four categories of SOUNDEXes were defined:

"Match efficiency" refers to the absence of false positives; e.g. 99% match efficiency means that 1% of matches are expected to be false positives, and "99% or better" means that 1% or fewer false positives are expected.

Using 95% match efficiency as a criterion of acceptability, we decided to omit all records with SOUNDEXes with a quality code of 2 or 3, except in Montreal. This omission is quite acceptable elsewhere, since most jurisdictions have small enough populations that there are very few or no SOUNDEXes with quality codes of 2 or 3: the only jurisdictions with more than 0.0% of these SOUNDEXes are Montreal (28.4%), Quebec City (2.2%), Calgary (1.3%), Edmonton (3.5%), and Toronto (15.1%), but Toronto was already omitted from our sample because of its non-reporting of youth who are not charged. Due to the large number of records which would be omitted for Montreal if we adopted this criterion, we included records with a SOUNDEX quality code of 2 in that jurisdiction.

The population of areas of New Brunswick reporting to the UCR2 is small enough that matching could be done with all police services treated as one unit, for all SOUNDEXes. For Saskatchewan and Alberta, matching was done with all police services treated as one unit for SOUNDEXes with a quality code of 0, but within individual police services for SOUNDEXes with a quality code of 1. For Ontario, Quebec, and British Columbia, matching was done within CMA or individual non-CMA police service for SOUNDEXes with quality codes of 0 and 1. This resulted in a sample of 38,727 unique young persons apprehended in 2001, with an average of 2.9 apprehensions, including the current one; or 1.9 prior apprehensions. We also examined the results of three other plausible but less conservative sets of matching criteria, which produced very similar results, ranging from 38,369 to 38,411 unique youths, and an average number of apprehensions (in all three cases) of 3.0. Thus, for this study, the results of matching were robust even when less stringent matching criteria were used.

Although the number of prior apprehensions of youths in our sample ranged from 0 to 261, the great majority (96%) had 10 or fewer, and most (90%) had 5 or fewer. In assessing the relationship between the number of prior apprehensions and the police disposition, no information was lost by recoding the number of prior apprehensions as 0, 1, 2, 3-4, and 5 or more.

The police disposition (charged vs. processed otherwise) was cross-tabulated separately with each of the independent variables:

The two latter variables were omitted from further analysis, since they were unrelated to the police disposition. The proportions of youth who were charged, broken down by each of the other variables, are presented in individual tables in Chapter V.

In order to assess the relationship of the independent variables while controlling for related factors, all independent variables were entered simultaneously into a multiple regression analysis with the police disposition (charged vs. processed otherwise) as the dependent variable. Two statistics were calculated: