Black people in criminal courts in Canada: An exploration using the relative rate index

Method

Procedures and measures

This study uses the same methodology and datasets as a separate recent study that focused on Indigenous people in criminal courts (Saghbini, Bressan, & Paquin-Marseille, 2021). This project was a collaborative effort between the Research and Statistics Division at the Department of Justice Canada and the Canadian Centre for Justice and Community Safety Statistics at Statistics Canada. The data presented in this report were obtained through a data linkage: records from Statistics Canada’s 2016 Census of Population long-form (Census)^Footnote5 and the Integrated Criminal Court Survey (ICCS)^Footnote6 were assessed and linked based on the probability that they belonged to the same person. Both the ICCS appearance and charge files were required to complete the linkage since each file contained specific information on the court outcomes examined in this study. More specifically, data from the ICCS appearance file with personal identifiers were first sent to the Social Data Linkage Environment (SDLE)^Footnote7 to identify unique individuals. These individuals were then linked back to the ICCS charge file, which is composed of charges for completed court cases (i.e., cases where all charges received a final decision).^Footnote8 To obtain the Black identity of ICCS accused, individuals from the ICCS charge file were linked to the long form Census, which was administered to 25% of the Canadian population. The ICCS cohort had a linkage rate of 13% (or 52% of the possible 25%). In order to make inferences to the entire population of ICCS individuals, Census weights were adjusted to match the complete ICCS cohort.^Footnote9 Within the ICCS cohort, only individuals who had a completed court case and were linked to the Census were retained for this study, along with their weights and Black identity.

The data obtained through the data linkage procedure are national in scope and include information from 11 provinces/territories. Data from Quebec and Alberta were excluded since the personal identifiers required for linkage were not available in the ICCS.^Footnote10

The “Visible minority group” variable in the Census was used to create two study groups: Black accused and White accused. The term “Black accused” represents those who identified as Black on the 2016 Census.^Footnote11 For the purpose of this study, the term “White accused” includes individuals identified as neither Indigenous, nor as a visible minority^Footnote12 (hereinafter “racialized group”) in the Census.

The linked data for Black and White groups were used to generate two types of metrics: 1) proportions of Black and White accused in criminal courts; and 2) Relative Rate Index.

Proportions of Black and White accused in criminal courts

To obtain the representation of Black people as accused in criminal courts, proportions of both Black and White accused were calculated based on the total number of accused in the ICCS. These were contrasted with the proportions of Black and White people in the Canadian population, which were calculated using the total Census population. For comparison purposes, individuals under 12 years of age were excluded from the Census population counts, as individuals within this age group are excluded in the ICCS.^Footnote13 Proportion data (both ICCS- and Census-based) were generated for the three most recent Census years for which there were exact population counts: 2006, 2011 and 2016.^Footnote14

Relative Rate Index

The RRI method measures the likelihood of a selected group (Black accused) encountering an outcome (e.g., guilty finding, custodial sentence) relative to a reference group (White accused) encountering the same outcome. In this study, RRIs were calculated by dividing the rate of Black accused experiencing a court outcome by the rate of White accused experiencing the same outcome (see Annex 1). These rates are based on the number of Black and White accused experiencing a court outcome out of the total number of Black and White accused, respectively, “at risk” of experiencing the outcome. The term “at risk” refers to the different stages of the criminal court process; only those accused present in the court system at the previous stage are “at risk” of moving through to the next stage. For example, only those convicted—as opposed to all accused—are “at risk” of receiving a custodial sentence.^Footnote15 Consequently, RRIs represent the level of disproportionality at key stages/decision points of the criminal court process, independent of any disproportionality that may have occurred at an earlier stage in the court process.

This RRI study examines the representation of Black accused relative to White accused at three key stages/decision points in the criminal court process:

court decisions;^Footnote16
sentencing;^Footnote17 and
length of custodial sentences.^Footnote18 In addition to the RRIs, for information purposes, data on the length of custodial sentences are also presented using the median length of custodial sentences in days as a measure (see Annex 2, Table 8 and Table 10).^Footnote19

RRIs were generated for each of the years from 2005-06 to 2015-16.^Footnote20 To limit the scope of the report and facilitate reporting of results for Black and White accused, RRIs are primarily presented in the text of the report as a single average across the 11-year period (i.e., rather than as 11 separate data points). Unless otherwise stated, RRIs are reported only if the data were available for each year of the 11-year period. This allowed for consistency in the reporting timeframe across court outcomes (e.g., court decisions and sentence types), as well as sub-outcomes (e.g., guilty finding and stay of proceedings or custodial sentence and probation). Since the average RRI may hide important variations from year to year, charts presenting the yearly RRIs are included for each court outcome examined, and notable trends are discussed.

The key RRI data presented in the report capture the total ICCS population. The RRI data were broken down by age groups (adults and youth), by sex (male and female), and by type of offence (violent and non-violent). RRIs for these groups are presented in the text where the data show a different trend than that of the Black population as a whole or where disproportionality is more pronounced at a specific juncture of the criminal court process when additional variables are taken into account. These breakdowns are also reported as an average over the 11-year period. In some cases, these reported breakdowns may hide important variations over time. It should be noted that certain breakdowns, such as for youth and female accused, resulted in low sample sizes; in these cases, certain data points had to be suppressed to ensure data quality and the confidentiality of individuals.^Footnote21 These breakdowns are therefore unavailable for certain years. In these various instances, a note was made in the report. This issue was particularly pronounced with jurisdiction-specific data where breakdowns for most jurisdictions were not available. Therefore, jurisdictional data are not presented in the text of the report. Finally, data reported in the text of the report focus on the most notable results; complete data tables, including all available breakdowns, are presented in Annex 2.

Interpretation of RRI results

For each court outcome and breakdown, the average RRI of Black accused is established in comparison with that of their White counterparts, which constitute the reference group. For example, the likelihood of Black female accused being found guilty is established in comparison to White female accused being found guilty. A RRI of 1.00 means there is no disproportionality compared to the reference group. A RRI over 1.00 means that Black accused are more likely to encounter a court outcome than the reference group. A RRI lower than 1.00 means that Black accused are less likely to encounter a court outcome than the reference group.

For the purpose of this report, these thresholds were slightly adjusted. RRI values that were within four percent of the reference category (i.e., 1.00 +/- 0.04) were considered to present no disproportionality. For instance, Black accused and White accused are considered equally likely to encounter a court outcome when the RRI value is situated between 0.96 and 1.04. A RRI of 0.95 (or -5%) or less would indicate that Black accused are less likely than White accused to encounter a court outcome. A RRI of 1.05 (or +5%) or more would indicate that Black accused are more likely than White accused to encounter a court outcome.

Table 1: Reporting and interpretation of RRI results
RRI value	Data reporting	Data interpretation
1.05 or more	+5% or more	Black accused are more likely than White accused to encounter an outcome
0.96 to 1.04 (1.00 = Reference)	-4% to +4%	Black and White accused are equally likely to encounter an outcome
0.95 or less	-5% or less	Black accused are less likely than White accused to encounter an outcome

In this report, RRI values are presented as percentages (see Table 1). For example, a RRI value of 1.20 would be reported as 20% more likely or +20%. The same applies to RRI values that are less than 1.00. For example, a RRI value of 0.85 would be reported as 15% less likely or -15%. RRI values of 2.00 or greater can also be reported in multiples. For example, a RRI of 2.00 (or +100%) would be reported as twice as likely. However, charts plotting the RRI data trends over an 11-year period (presented in the findings) use the RRI values, rather than percentages.

Footnotes

Footnote 5

The Census (long-form) is a sample-based and mandatory survey, conducted every five years, that provides demographic, social and economic information on Canada’s population. The Census excludes Canadian citizens living temporarily in other countries, full-time members of the Canadian Forces stationed outside Canada, persons living in institutional collective dwellings such as hospitals, nursing homes and penitentiaries, and persons living in non-institutional collective dwellings such as work camps, hotels and motels, and student residences.

Return to footnote 5

Footnote 6

The ICCS collects data from administrative court records and maintains a national database of statistical information on appearances, charges and cases in both youth and adult criminal courts. The Survey covers only provincial and superior courts and excludes appeal courts, federal courts and the Supreme Court of Canada.

Return to footnote 6

Footnote 7

The SDLE at Statistics Canada promotes the use of existing administrative and survey data to address important research questions and inform socioeconomic policy through record linkage. The SDLE expands the potential of data integration across multiple domains, such as health, justice, education, and income through the creation of linked analytical data files without the need to collect additional data from Canadians.

Return to footnote 7

Footnote 8

A final decision consists of a finding of guilt (including guilty pleas), an acquittal, a stay of proceedings, charges being withdrawn by the prosecution, a case being dismissed or an accused being discharged, a verdict of not criminally responsible, and other decisions, such as mistrials, special pleas (e.g., autrefois acquit) and being unfit to stand trial.

Return to footnote 8

Footnote 9

The maximum linkage rate possible was 25% since only 25% of the Canadian population received the Census long-form questionnaire. Census weights were used to represent the entire population based on the information gathered from the sample, such as sociodemographic information. The ICCS cohort had a 13% linkage rate to the records of the Census. Census weights were re-adjusted to represent the remaining 87% of the cohort that did not link using information available in the ICCS dataset. This means that each person from the cohort who linked to the Census represents a certain number of persons in the total cohort based on specific characteristics, which include the age, the sex, the region, and the common offence classification group.

Return to footnote 9

Footnote 10

As of 2005-06, all provinces and territories report provincial/territorial court data to the ICCS. Most provinces and territories also report superior court data, with the exception of Ontario, Manitoba, Saskatchewan, and until 2018-19, Prince Edward Island.

Return to footnote 10

Footnote 11

The data presented in this study are part of Statistics Canada’s “accused file.” Therefore, the study unit is an accused. As an accused progresses through the CJS, they eventually become an offender based on their court outcomes (i.e., when found guilty). In this report, to ensure consistent terminology and avoid confusion, the term “accused” is used throughout to refer to both accused and offenders.

Return to footnote 11

Footnote 12

Visible minority is defined by the Employment Equity Act as “persons, other than [Indigenous] peoples, who are non-Caucasian in race or non-white in colour.” The visible minority population consists mainly of the following groups: South Asian; Chinese; Black; Filipino; Latin America; Arab; Southeast Asian; West Asian; Korean, and Japanese. The “not a visible minority” category predominantly consists of White people, but also includes individuals that identified themselves as both White and a small subset of visible minority groups (i.e., White and Latin American or White and Arab). These individuals represent less than 1% of the “not a visible minority” category in the Census.

Return to footnote 12

Footnote 13

Children under 12 years of age are not subject to criminal prosecution in Canada.

Return to footnote 13

Footnote 14

For 2006 counts, all individuals (ICCS and Census population) with an invalid value for the Black identity indicator were excluded. The 2016 Census counts were compared to 2015-16 ICCS data.

Return to footnote 14

Footnote 15

For certain key stages (i.e., preliminary inquiry and trial), RRIs were calculated based on all accused at previous stages, since the data did not allow to identify the “at risk” group. Rates were only calculated where 30 or more individuals were “at risk” of experiencing the event or the outcome or where 10 or more individuals experienced a specific court stage/decision point. In cases where the White group had zero individuals, the RRI could not be calculated.

Return to footnote 15

Footnote 16

Court decisions include: guilty; acquitted; stay of proceedings; withdrawn, dismissed and discharged; and, other (e.g., not criminally responsible, unfit to stand trial, special pleas, and waived out of province or territory).

Return to footnote 16

Footnote 17

Sentence types include: custody; conditional sentence; probation; fine; and, other (e.g., absolute and conditional discharge, suspended sentence, community service order, and prohibition order).

Return to footnote 17

Footnote 18

Other relevant court outcomes could not be included in this study due to data unavailability. These include bail hearings and decisions; pre-trial hearings, such as Charter challenges and preliminary inquiries; and trials.

Return to footnote 18

Footnote 19

The median represents the point at which half of all cases had longer custodial sentence lengths and half had shorter custodial sentence lengths.

Return to footnote 19

Footnote 20

A case year refers to a twelve-month duration beginning April 1 and ending March 31, and represents the year in which all charges in the case reached a final decision. Individuals can appear across multiple years for different cases.

Return to footnote 20

Footnote 21

Census suppression rules are applied to prevent direct or residual disclosure of any information deemed confidential that could identify respondents. Consequently, jurisdictions with a population below a certain threshold are not disclosed.

Return to footnote 21

Date modified:: 2023-03-15

Language selection

Search

Phone Scam Warning

Black people in criminal courts in Canada: An exploration using the relative rate index

Method

Procedures and measures