Report and Draft Recommendations: Engagement on Corrections Disaggregated Data and Analysis Strategy

Report and Draft Recommendations: Engagement on Corrections Disaggregated Data and Analysis Strategy (PDF, 866 KB)

Canadian Centre for Justice and Community Safety Statistics (CCJCSS)

Table of Contents

Executive Summary

Over the years, there have been increasing demands for better disaggregated data to shed light on people's diverse experiences with the police and the justice system. Disaggregated data can help to identify and respond to issues of social inequities, discrimination, and systemic racism within Canadian society. Concerns for the disparate treatment of Indigenous and racialized peoples in the Canadian criminal justice system revealed important gaps in the availability of disaggregated data. This situation is especially true for information on the identity of people who encounter police for various reasons, including criminal incidents.

In response to these growing demands, Statistics Canada is in the process of developing a Corrections Disaggregated Data and Analysis Strategy on the Representation of Indigenous and Racialized Groups in Canada's Correctional Systems. This strategy aims to better understand the experiences of Indigenous and racialized groups in terms of their interactions and involvement with correctional systems and with the criminal justice system more generally. Prior to starting development of the strategy, the Canadian Centre for Justice and Community Safety Statistics (CCJCSS) engaged numerous partners of interest, including Indigenous and racialized community groups and organizations, and sought input through the Engagement on Corrections Representation Data & Analysis Strategy. This report provides the background and key results of the Engagement, including seven key recommendations.

Recommendation 1
Statistics Canada should develop population-based indicators and re-contact indicators using disaggregated data to measure representation of sub-populations in correctional systems;

Recommendation 2
Statistics Canada should disaggregate its correctional services data as much as possible;

Recommendation 3
Relationships between socio-economic and mental health issues and over-representation should be analyzed further;

Recommendation 4
Statistics Canada should engage in data quality evaluations of the disaggregated data information collected;

Recommendation 5
Statistics Canada should include appropriate context as part of the data analysis process, and work closely with affected populations when publishing disaggregated data;

Recommendation 6
Statistics Canada should review options for dissemination to enhance data accessibility and use; and

Recommendation 7
Statistics Canada should regularly review its definitions, terminology and categorizations in order to ensure that the language used is appropriate and culturally-sensitive.

Background

Through CCJCSS, Statistics Canada has a long history of publishing data on Indigenous persons in corrections. Most of the reporting to date has focused on presenting the number of admissions (with breakdowns by custody/community and adult/youth) for a given fiscal year by Indigenous identity. Data on Indigenous identity by admission is available back to 1997/1998 for youth and 2000/2001 for adult corrections.

The CCJCSS is currently in the process of developing a Corrections Disaggregated Data and Analysis Strategy: Representation of Indigenous and Racialized Groups in Canada's Correctional Systems (for ease of reference, referred to hereafter as: The Corrections Representation Data & Analysis Strategy).

The Engagement on Corrections Representation Data & Analysis Strategy sought input from partners and data users outside of Statistics Canada to guide the development of its statistical program. Our goal is to provide our partner organizations and the public at large with up-to-date disaggregated data and information pertaining to correctional services in Canada. The Engagementinvolved respondents from a wide and diverse range of perspectives, including: Indigenous and racialized groups and organizations; corrections agencies; academics; and other interested parties at the national and provincial/territorial government levels. Feedback was sought on the following:

  • Current key indicators produced by Statistics Canada related to Indigenous peoples' over-representation in corrections, and ways in which the indicators and analysis can be improved to be more relevant to data users and the public;
  • In addition to the historical focus at Statistics Canada on Indigenous peoples' involvement in correctional systems, explore the importance of expanding the national corrections statistical program to collect, analyze and produce indicators that provide relevant and timely information for Black, Hispanic, South Asian, East and Southeast Asian and other racialized groups;
  • Important contextual information to consider and present when disseminating disaggregated data; and
  • Assessments of data quality.

Feedback was obtained in two ways: (1) from written responses to an engagement document; and (2) from participation in small-group discussions led by Statistics Canada.

Methods

Engagement Document

The written engagement document, both in French and English, was sent to 322 partners and community organizations on November 6, 2021. The original deadline was extended from November 26, 2021 to December 24, 2021. As of December 31, 2021, 33 responses were received and analyzed. Respondents were asked to answer 16 questions in the written document which consisted of a combination of yes/no questions, Likert scales, and open-ended questions.

Small Group Discussions

Due to the impact of the on-going COVID-19 pandemic, small group discussions were held virtually between November and December 2021. Small group discussions for the Canadian Correctional Statistics Survey (CCSS) were conducted in tandem with the Uniform Crime Reporting Survey (UCR) for a total of 41 participants. There were 25 participants who joined the CCSS-specific small group discussions; they were invited to attend a session based on their organization affiliation:

  • Indigenous groups and organizations;
  • Racialized groups and organizations;
  • Correctional services agencies;
  • Academics; and
  • Other interested parties at the national and provincial/territorial government levels.

The discussions were held in both French and English, and were facilitated by a neutral, third-party moderator from Statistics Canada's Internal Engagement team. For each session, the moderator posed the same 12 questions in the same order. These questions were derived directly from the written engagement document and were shortened to be more 'interview-friendly'. Moreover, the moderator encouraged further discussion among the participants and at times would offer prompts such as "How do you see this being useful?" to help move the conversation along and to gain more insight into a particular answer.

Response Rates

Engagement Document

Thirty-three (33) written document responses were received for an overall response rate of 10.3%. The breakdown was as follows:

  • Federal Partners / Stakeholders: 33.3%
  • Criminal Justice: 27.3%
  • Correctional Services Programs: 15.2%
  • Ethno-Cultural Groups / Organizations: 6.1%
  • Indigenous Representative Bodies: 6.1%
  • Other Partners / Stakeholders: 12.2%

Please note that the total sum of the individual percentages above does not add up to 100.0% due to rounding.

The low total response rate can be attributed, at least in part, to response and respondent burden. For instance, some respondents had indicated that they found the engagement document too long and parts of it too complex.

In the coming months, there will be additional opportunities for comments and engagement with partners and stakeholders as Statistics Canada, through CCJCSS, develops its five-year Corrections Representation Data & Analysis Strategy.

Small Group Discussions

Of the 25 individuals participating in the CCSS-specific virtual small group discussions, the majority (32.0%) represented organizations from the Federal Partners / Stakeholders category.

Recommendations

Recommendation 1

Statistics Canada should develop population-based indicators and re-contact indicators using disaggregated data to measure representation of sub-populations in correctional systems

Current Measures of Over-Representation

Recent studies and reports, as well as those going back decades, have highlighted and drawn attention to how the experiences of First Nations peoples, Métis and Inuit and other racialized groups and populations in the Canadian criminal justice system have been marked by over-representation and inequitable treatment. First Nations peoples, Métis and Inuit, for instance, have long and unique social, cultural and political histories in Canada. Most notably, the history of colonialism – including residential schools and culturally insensitive and inaccessible programs – and settler colonial policies continue to seriously impact Indigenous people and communities to this day. As a result, Indigenous peoples have experienced and continue to experience social, economic and institutional marginalization, and also various forms of inter-generational trauma.

Statistics Canada currently measures over-representation by proportion of admissions, where the individual self-identifies as Indigenous. Admissions are counted each time a person begins any new legal status while being supervised in a correctional institution. This means that the same person may be included several times in the admission counts where they move from one correctional program to another (e.g., person initially enters corrections on a remand status then to sentenced custody later in the same involvement – this scenario would count as two admissions, one for remand and one for sentenced custody) or re-enter the system later in the same year. Admissions therefore represent the number of all entries during a fiscal year to each type of legal status (remand, sentenced custody or a community supervision program), and track correctional events more than persons supervised by the system.

Proposed New Indicators

Population-based Measures (Incarceration Rates and Custodial Involvement Rates)

Statistics Canada is exploring developing population-based measures that would present involvement in the correctional system as a rate or percentage of the overall target population as our main indicator of representation. Population-based measures present involvement in the correctional system as a rate or percentage of the overall target population as the main indicator of representation. This includes taking into account incarceration rate and custodial involvement rate indicators.

Incarceration rate measures the proportion of a population in custody on an average day in the year. It is calculated by taking the Average Daily Count (ADC) of the correctional population then dividing it by the general population estimate for that same year. For the CCSS, the rate is expressed as the number of incarcerated persons per 10,000 population. For example, a rate of 100 for Canada means that, on an average day in the year, 1% of the population in Canada was incarcerated. With the CCSS, ADC is now available by Indigenous and racialized groups, allowing Statistics Canada to produce both Indigenous and Non-Indigenous incarceration rate for the population as a whole, or apply intersectionality to the rates (e.g., limit to adult males, adult females, youth between the ages of 12 to 17, etc.). Over-representation (for Indigenous persons for example) would be measured by the relative difference between the Indigenous and Non-Indigenous Incarceration Rates.

Custodial involvement rate measures the proportion of a specific populations experiencing custody over a reference period. The measure identifies the number of unique persons spending at least one day in custody during the reference period for a defined population (Indigenous, Black, young males, etc.), then calculates the percentage of the population experiencing incarceration. Individuals are counted equally, whether or not they spent one night in custody or the whole year.

Over-Representation Index (of Indigenous Populations)

Studies have shown that being young and male are risk factors for involvement in crime. The Over-representation Index would account for differences in age and sex profile of different populations when calculating rates of incarceration. The Over-representation Index recalculates the relative difference between Indigenous and non-Indigenous rates, as if both populations have an age/sex profile identical to the national population distribution. For instance, for a province or territory where the Indigenous population is younger or proportionally more male, there could be a lower score on the Over-representation Index than the actual relative differences between the Indigenous and non-Indigenous Incarceration Rates. Likewise, over time, as population profiles change for a province or territory, the Over-representation Index controls for these changes, reducing the impact demographic shifts may have on the over-representation measurement.

While our initial work on the Over-representation Index has been limited to measuring Over-representation in Indigenous populations, the concepts developed can be applied to other populations, such as Black people in Canada, as well as applying intersectionality.

Re-contact Indicator

Re-contact refers to a measure of return to the justice system after initial release. The re-contact indicator is complicated, with a number of different elements: (1) The reference event is the starting point from which re-contact for an individual is assessed (i.e. when the "clock" starts); (2) Follow-up period: Length of time over which an offender is observed (e.g. 2 years from reference event); (3) Re-contact Event: The re-contact event is defined as the occurrence and start date of any new legal hold status occurring after the reference event and within the follow-up period. There are three metrics associated with the definition of re-contact, listed below:

  • Prevalence (size of the issue): number/proportion of offenders with a re-contact event within the follow-up period?
  • Frequency (how active): how many re-contacts did individuals have during the follow up period (e.g., are most re-contacts due to a small group with a large number of recontacts?)
  • Elapsed time (time to re-contact): how much time passed between the reference event and a re-contact event?

Statistics Canada is currently working with Public Safety on the Pan-Canadian Re-contact Strategy. The strategy aims to continue work on the re-contact indicators, with an initial focus on persons re-contacting the justice system after going through corrections. Statistics Canada is currently reviewing the re-contact definitions and is evaluating adding an escalation metric (indicator of whether the severity of offences associated with re-contacts are increasing or decreasing).

What we heard

Current Measures

Four in ten (42%) of the written document respondents and 52% of the small group discussion participants indicated that observing the rate of admission is useful for measuring the over-representation of incarcerated Indigenous and racialized groups. However, there are concerns surrounding how this method can lead to multiple counts of one individual and therefore, can impact the accuracy and quality of the data. As such, an alternative method was suggested by the majority of respondents: measure over-representation via daily count to avoid the risk of double-counts of re-admission. This would also paint a more accurate picture of the situation inside the correctional facilities.

There is a strong consensus among the Engagement participants that there is an immediate need to develop an official definition for over-representation, and to also implement a national, standardized Over-Representation Index to be used across all correctional systems.

Population-based Measures

Seven in ten (70%) of the written document respondents and 80% of the small group discussion participants have indicated that developing population-based measures would be useful to meet their data needs. There is an overall consensus among all the participants in the Engagement project that population-based measures would help advance our knowledge because it would allow researchers to draw comparisons across provinces and territories. Moreover, this measure would be beneficial to design culturally appropriate programs and to engage with communities in a meaningful way.

Over-Representation Index

Eight in ten (82%) of the written document respondents and 84% of the small group discussion participants have indicated that, in general, the Over-Representation Index would be a very useful measure. More specifically, 70% of the written document respondents also indicated that it would be especially useful to present a national indicator of over-representation by integrating CSC (Correctional Service of Canada) and Provincial/Territorial results (see Figure 1 below).

Percent of written document respondents who found population-based indicators useful for measuring and analyzing over-representation of vulnerable populations
Description for Figure 1: Percentage of respondents finding indicator very useful for measuring and analysing over-representation of vulnerable populations
Percentage of respondents finding indicator very useful for measuring and analysing over-representation of vulnerable populations
  Percentage (%)
Proportion of admissions 48.48
Population/rate based indicators 72.72
Over-representation index 87.88
Recontact indicators 90.91

Re-contact Indicators

Almost nine in ten (88%) of the written document respondents agree that it would be useful to them and their organization to enhance current re-contact indicators. This would include an initial emphasis on correctional data by identifying criminal justice system re-contacts after release from correctional services. The strategy would produce rates of re-contact for the overall correctional population and disaggregate into groups. Similarly, 84% of the small group discussion participants agree that disaggregating rates of re-contact for the overall population would be useful. Furthermore, 32% of the participants have suggested that greater level of detail needs to be included within the re-contact indictor, such as reason for re-contact, quantity of re-contacts, length of stay, and type of offence.

Measures of Over-Representation Used by STC (Statistics Canada) Partners

Almost nine in then (89%) of the written document respondents have indicated that their organization collects and analyzes data on over-representation. 56% of the small group discussion participants collect their own data on over-representation while 44% of participants work with the data collected by Statistics Canada. Furthermore, the majority of participants from both the written document (66%) and the small group discussions (68%) have indicated that they do not have an official, standardized method of measuring and evaluating over-representation. Four in ten (42%) of the written document respondents and 40% of the small group discussion participants measure over-representation by comparing the proportion of the Canadian population to the number of incarcerated individuals. On the other hand, some respondents have indicated that they do have a method of measuring over-representation through the use of several indicators, such as through admission rate, observing the proportion of Indigenous and racialized prisoners, evaluating proportion of dangerous prisoners, and understanding mental health needs.

Recommendation 2

Statistics Canada should disaggregate its correctional services data as much as possible

Overall consensus
The present data on racialized identity is too general and does not accurately represent the cultural diversity that exists within Indigenous peoples and racialized groups

In addition to understanding the incarceration experiences of Indigenous people, Statistics Canada seeks to expand its program to also collect and analyze racialized identity. Respondents were asked several questions on the collection, assessment, and presentation of corrections data for Indigenous and racialized groups. The responses varied when respondents addressed issues facing racialized groups and Indigenous peoples.

Results drawn from current data cannot be used to draw conclusions about the larger population and should only be strictly applied to those in the dataset. Therefore, Statistics Canada needs to develop a strategy to disaggregate the data as much as possible so that the correctional population can be accurately represented.

Disaggregated data for Racialized groups

The findings from the Engagement suggest that there is a strong need to disaggregate race data as much as possible. Racialize identity and ethnicity are often conflated with one another, but they are separate. "Race" is a social construct and does not have any biological basis; it is ascribed to individuals based on their physical characteristics (e.g., skin colour). Ethnicity, on the other hand, encompasses everything from language, to nationality, to religion, and culture. As such, an individual can have multiple/mixed racial identities and ethnicities.

Fifty-two (52%) of the written document respondents indicate race data should be disseminated to include ethnic origin and allow for multiracial identities. It is harmful to collapse identities in one category, as individuals have very different experiences depending on their racial and ethnic identity and should be avoided where possible. Respondents also underscored that it is crucial to allow for the reporting of multiple/mixed racial identities, and it is equally important that such data is disaggregated appropriately. However, no feedback was offered on approaches to analyze and disseminate multi-racial categories, therefore the ideal approach CCJCSS should take regarding this will require further exploration and engagement.

Overall, the respondents would like to see a clear distinction made between race-based data and ethnicity data. Just as how the historical and cultural diversity among Indigenous people is recognized, the same is recommended for racialized groups. For example, "Black" is race-based data because it is socially constructed. However, it does not accurately portray the cultural diversity that exists among Black communities. The experiences of individuals who descend from African-Americans fleeing slavery from the United States is very different than recent immigrants from North Africa or the Caribbean. As such, collecting ethnic-based data would allow for a more accurate portrayal of the incarcerated population, rather than just grouping everyone under the same label as their racialized identity.

Indigenous Groups and Peoples

A frequent misconception is the idea that Indigenous people are homogenous. Thirty percent (30%) of the written document respondents and 16% of the small group discussion participants recognize the diversity and heterogeneity of Indigenous groups and peoples. They agree there is a need to disaggregate the correctional data as much as possible otherwise it becomes difficult to determine the extent to which the aggregated data among Indigenous populations can be generalized to each sub-group. Furthermore, respondents indicated that it would be helpful to examine the differences and similarities between First Nations, Inuit, and Métis people to better understand their experiences with the correctional system and to develop culturally appropriate programs. There is also a need to understand the major regional variation regarding the over-representation of Indigenous people. 18% of the written document respondents and 16% of the small group discussion participants have indicated that the data should be disaggregated by geographical location, such as those living on- or off-reserve. Moreover, 12% of the written document respondents stressed that the data should also be disseminated by regional differences, such as urban or rural dwellings.

Recommendation 3

Statistics Canada should further analyze the relationship between socio-economic and mental health issues and the over-representation of certain groups in correctional systems

Overall consensus
Additional socio-economic and mental health factors should be considered by Statistics Canada when conducting analysis of over-representation

Respondents were specifically asked to identify important contextual information and factors that Statistics Canada should consider when presenting data and writing analytical reports pertaining to Indigenous, racialized and other diverse populations. Understanding additional contextual factors that may contribute to over-representation will ultimately allow for the development of evidence-based approaches to appropriately respond to the phenomenon.

Responses to both the written engagement exercise and small group discussions indicated a rich variety of additional socio-economic and mental health factors that should be considered by Statistics Canada when conducting analysis of over-representation. Some factors were more frequently identified than others (such as education), but the majority of respondents were very clear in asserting that over-representation data cannot be analyzed solely on its own accord. As one respondent explained, "We cannot consider over-representation data in isolation of other data sets such as health and education data."

Additional socio-economic and mental health factors

Respondents to the written engagement noted many socio-economic factors and mental health issues that are associated with over-representation. Education was the area respondents most frequently noted, with 91% indicating that education needs to be considered when conducting data analysis on over-representation. The next factors most frequently noted by respondents were mental health, labour force, and age and gender, with 85% of respondents observing that these areas need to be considered. Other areas to be considered in data analysis on over-representation are noted in Figure 2 below:

Top 10 areas to be considered when doing data analysis on over-representation
Description for Figure 2: Top 10 areas to be considered when doing data analysis on overrepresentation
Top 10 areas to be considered when doing data analysis on overrepresentation
  Percentange (%)
Breach 54.54545
Community Supervision 60.60606
Risk/Need 69.69697
Gang 72.72727
Offences 75.75758
Marginalization 81.81818
Age & Gender 84.84848
Labour Force 84.84848
Mental Health 84.84848
Education 90.90909

Respondents to the written engagement also noted substance use, marital/family stability, housing, sexual orientation, programming availability, length of incarceration, intergenerational trauma, and childhood maltreatment as other factors impacting over-representation.

The small group discussions further identified many factors impacting over-representation, such as homelessness (16%), mental health (16%), experience of violence and/or sexual violence (12%), housing (8%), state of crisis/emergency (8%), cost of living (8%), substance use (8%), and immigration status (8%).

Going forward, it will be very important for Statistics Canada to consider data on over-representation within the context of social issues. The aforementioned factors identified by respondents will guide the focus of Statistics Canada in conducting future analyses of over-representation data, with particular emphasis on those factors that were most frequently identified.

As one respondent commented, "it is important to ensure that over-representation is contextualized with an understanding of the 'social determinants' of interaction with the criminal justice system." Ultimately, as another respondent noted, "understanding the intersection among health, education, poverty, age, gender, etc. and crime informs broader evidence-driven solutions to understanding crime prevention and successful community integration."

Recommendation 4

Statistics Canada should conduct data quality evaluations of the disaggregated information collected

Overall consensus

  • Current method of collecting self-reporting identity is concerning and can inflict or deepen existing trauma, which can impact the quality of the data
  • Trauma-informed data collection methods may help to mitigate the risk of re-traumatization

The following findings were derived from an open-ended question which asked respondents to indicate any concerns they may have about the quality of data and information collected by correctional programs and other sectors of the criminal justice system in regard to Indigenous people and racialized groups. Overall, the majority of Engagement respondents (85% of the written respondents and 64% of the small group discussion participants) expressed concern about the quality of disaggregated data collected by correctional services and other sections of the justice system regarding Indigenous peoples.

The strongest concern that was discussed by the Engagement participants (15% of the written document participants and 12% of the small group discussion participants) revolve around data quality regarding identity. 40% of the written document respondents are concerned about Indigenous identity and 48% of respondents indicated that the current measures of self-reporting identity is problematic. Participants indicated that Indigenous individuals are often more hesitant to disclose their Indigenous identity at intake, for various reasons. Indigenous individuals may feel that declaring their identity will invite further discrimination within the correctional facility from both prison workers and prisoners, especially if the prison worker in question is White. Moreover, 12% of the small group discussion participants are concerned with the lack of consent surrounding identity collection, and 10% are concerned with the potential for re-traumatization due to data collection. Furthermore, 15% of the small group discussion participants have identified that there could be privacy implications among small cohort sizes. It may be easier to identify the individual in question if the admissions of a particular group is small.

In terms of the data collection methods, 30% of the written documents respondents are concerned with how Indigenous identity data is collected. More specifically, 6% of the written document respondents and 12% of the small group discussion participants are concerned that the data collection officer does not have any culturally appropriate training, and may create further trauma for Indigenous peoples at intake. 6% of the written documents respondents have suggested that trauma-informed data collection methods may help to mitigate the risk of re-traumatization. In addition, 36% of the small group discussion participants problematized the lack of standardized identity categories and 29% participants noted the lack of identity categories for Indigenous people, which contributes significantly to the rates of misidentification.

Recommendation 5

Statistics Canada should include appropriate context as part of the data analysis process, and work closely with affected populations when publishing disaggregated data

Overall consensus

  • The historical and social context for marginalized groups needs to be strongly articulated and acknowledged to avoid perpetuating stereotypes
  • Statistics Canada needs to consult and collaborate with marginalized communities in research methods development
  • Racial identity and ethnicity need to be understood and operationalized as separate variables

A key concern noted by respondents was the possibility of adverse impacts on vulnerable populations resulting from Statistics Canada publications of representation and disaggregated data. The majority of respondents (64% from the written engagement raised this concern) noted the importance of ensuring over-representation is understood as a systems-based problem.

Respondents also highlighted the risk of reinforcing stereotypes and racial biases. A further 39% of respondents emphasized the importance of consulting with marginalized communities, and including their voices in the data collection and analysis processes.

Indigenous populations

Discussion group participants echoed the concerns of the written engagement, with many commenting on the need for community-based approaches with Indigenous populations. When working with Indigenous populations, respondents noted that it will be necessary for Statistics Canada to acknowledge historical injustices, the impact of colonialism, and recognizing that Indigenous peoples are not a single, homogenous group.

Many discussion group participants also commented on the need to provide historical and colonial context, traumas, and systemic racism when presenting the data. Thirty-six percent of both the written document respondents and small group participants agree that the histories and present dynamics of systemic racism should be taken into account when dealing with race-based data. Approximately three out of ten respondents (27%) of the written document respondents are concerned with how systemic discrimination against certain groups and communities have impacted the collection and quality of corrections data. Due to the plurality of histories and experiences of Black and other racialized groups, the data should be collected and presented in a culturally appropriate manner. Similarly, 28% of the small group discussion participants suggested that providing additional socio-economic context and highlighting the differences among racialized and ethnic communities would be beneficial.

Going forward, it will be crucial for Statistics Canada and CCJCSS to work with and engage proactively with Indigenous communities to achieve progress in responding to this recommendation. This could be completed through community engagement exercises such as town halls, as well as by collaborating with Indigenous communities and considering the incorporation of Indigenous research methods that promote Indigenous values and knowledge. A statement from one of the respondents who quoted an Indigenous professor illustrates the importance of doing so:

"We have a very long history in this country of being studied and researched and having data collected on us, only to twist that around to blame the victim in a sense. If you want to collect that data, then you do it with us. And you do it for us."

Racialized groups

In the case of racialized groups, respondents noted that Statistics Canada should refrain from conflating racialized identity and ethnicity.

It was noted that racial and ethnic groups are social constructs, and consequently, factors that lead to the over-representation of certain groups are socially-based (e.g. systemic racism), and not due to any perceived differences on the basis of genetic factors.

Further to this, it was noted that the purpose of the data for these populations is very important in determining how it will be used in the future:

"If the intent is to monitor and address inequalities stemming from racism or systemic bias, then race-based data should be collected and disaggregated. However if the intent is to tailor services or interventions in the community to reduce the overrepresentation of certain groups in correctional settings, then ethnicity-based data should be collected and disaggregated so that language, cultural or religious needs of individuals can be better anticipated. If both these purposes are relevant, than both race and ethnicity data should be collected and disaggregated."

Finally, respondents underscored the need for Statistics Canada to recognize that the experiences and needs of individuals vary significantly due to their racial and cultural histories. As one respondent explained, the histories and current dynamics of systemic racism, the plurality of histories and experiences of racialized groups, as well as other factors contributing to challenges in accessing services and supports need to be considered when Statistics publishes disaggregated data.

Recommendation 6

Statistics Canada should review options for dissemination to enhance data accessibility and use

Overall consensus

  • Publication of a specialized Juristat on over-representation would help deepen current understanding of over-representation
  • Accessing micro-data would help researchers in their research endeavours
  • Introducing additional analytical products that allow for a clearer visual representation of the data

Special topic Juristat reports

Specific to this recommendation, 30% of respondents from the written engagement exercise highlighted that a special topic published in Juristat would allow for a greater understanding of the variables that cause over-representation, as well as access to a direct comparison of findings that provide a more transparent view of how Indigenous and racialized groups are represented. As one respondent commented, "…special topics Juristats are helpful and focusing on characteristics of a specific population for a report can be informative."

Accessibility of micro-data

An additional 12% of respondents from the written engagement exercise proposed that increased accessibility to micro-data will allow researchers to examine a number of questions that will contribute to the advancement of knowledge on over-representation. As one respondent mentioned: "…having this data available as microdata in the Research Data Centres will allow for examination of some of the precursors to over-representation via integrated data projects.

Additional products

Search parameters were also suggested by 9% of respondents from the written engagement exercise. Allowing individuals to help filter data by jurisdiction, types of offence, racialized identity, and other parameters would make it more accessible to users.

Another 6% of respondents suggested that the inclusion of data tables and infographics will allow for a clearer visual representation of the data.

Finally, some respondents from the discussion groups suggested communicating data using familiar platforms, organizations, and resources. For instance, the Aboriginal Peoples Television Network (APTN) was suggested as a platform that could be used to share data.

Recommendation 7

Statistics Canada should regularly review its definitions, terminology and categorizations in order to ensure that the language used is appropriate and culturally-sensitive

Overall consensus

  • Several terminologies need to be replaced and/or redefined
  • Current uses of the words 'over-representation', 'offender', and 'gang affiliation' can be harmful and stigmatizing

Respondents highlighted the need for Statistics Canada to thoughtfully consider its terminology. As one respondent explained:

"The language and terminology used in collection practices, and the framework used to analyze data, should be determined in collaboration with communities that would be impacted by the data…The terms and language should then be shared widely through robust training."

Respondents most often referred to 'over-representation', 'offender', and 'gang affiliation' as terminology that should be reviewed by CCJCSS.

Defining 'over-representation'

The majority of discussion group respondents (56%) stressed the need for an official definition of over-representation, or a standard means by which it can be measured. Academics recommended using the term disproportionate in place of over-represented when commenting on imprisonment data.

Use of 'offender' and 'gang affiliation'

Respondents pointed to the term 'offender' as something that categorizes someone within the correctional system as an 'other'. Categorizing someone as an 'other' minimizes persons to a single characteristic, namely that they are housed within a correctional facility. Further to this, it is an inaccurate descriptor given that people held on remand may be later deemed innocent of their offence.

CCJCSS was also asked to exercise caution with use of the term 'gang affiliation'. This term has been disproportionately applied to Indigenous and racialized groups in the past, and has been applied to individuals who may have associated with individuals in gangs for however brief a period. The label can also be assigned to someone in a correctional facility without the requirement of police or court documentation. Once an individual in a correctional facility is assigned this label, removing it or correcting is "extremely difficult".

Next Steps

Based on these seven recommendations, Statistics Canada, through its CCJCSS, will develop a strategic five-year plan for developing its disaggregated data program with correctional services data. The plan will be comprehensive, addressing survey development, respondent and partner relations, analytical plans, indicator development and data quality evaluations, along with other components of its statistical program.

In the short-term, for fiscal year 2022/2023, CCJCSS will produce a special topic Juristat on Indigenous representation in adult custody (for provinces reporting to CCSS), and include a table on visible minority admissions on its website as part of its annual analysis. Within this context, it is important to note that CCJCSS has not yet developed standards for categorizing racialized groups as it awaits guidance from Statistics Canada on new classifications pertaining to racialized identity and ethnicity.