Date: April 2022
Program manager: Director, Canadian Centre for Justice and Community Safety Statistics
Director General, Census Subject Matter, Social Insights, Integration and Innovation
Reference to Personal Information Bank (PIB)
Personal information collected through the Uniform Crime Reporting Survey (UCR) is described in Statistics Canada's "Crime Statistics" Personal Information Bank (Bank number: StatCan PPU 029). The UCR collects personal information to produce statistical information relating to the nature and extent of crime in Canada and to describe the circumstances of criminal incidents, as well as the characteristics of victims and accused persons. Personal information is collected pursuant to the Statistics Act (Sections 3, 7, 13).
The Crime Statistics PIB is published on the Statistics Canada website under the latest Information about Programs and Information Holdings chapter.
Description of statistical activity
Statistics Canada conducts the Uniform Crime Reporting Survey (UCR) under the authority of the Statistics ActFootnote 1. The UCR collects data on criminal incidents that come to the attention of, and have been substantiated by, police. Information collected from police services includes personal information such as name, date of birth, gender, and characteristics of victims and accused persons involved in criminal incidents, as well as information relating to the nature and extent of the incidents.
The purpose of the survey is to measure the incidence of crime in Canadian society and its characteristics. The information is used by federal and provincial policy makers as well as public and private researchers.
More than 1,000 separate police services and detachments respond to the survey, comprising of over 190 different police forces. Starting in 2022, the Canadian Forces Military Police will become an active respondent to the UCR survey and will be required to provide the same information already provided by other police services.
Information on the UCR is available on Statistics Canada's website Uniform Crime Reporting Survey (UCR)
Reason for supplement
While the Generic Privacy Impact Assessment (PIA) addresses the privacy and security risks related to statistical activities conducted by Statistics Canada including the UCR, due to the sensitivity of the information being collected and in the context of the current Canadian Armed Forces (CAF)-Department of National Defence (DND) Sexual Misconduct dossier, this supplement describes the measures in place to protect the privacy and confidentiality of the information, and informs Canadians that the Canadian Forces Military Police will now be an active respondent to the Uniform Crime Reporting Survey (UCR) and will be reporting data on Criminal Code and Federal Statute incidents under their jurisdiction. As is the case with all PIAs, Statistics Canada's privacy framework ensures that elements of privacy protection and privacy controls are documented and applied.
Necessity and Proportionality
The collection of personal information for the Uniform Crime Reporting Survey (UCR) by the Canadian Forces Military Police can be justified against Statistics Canada's Necessity and Proportionality Framework:
- Necessity:
The general benefits of the UCR data for Canadians have been well defined in terms of:- providing a comprehensive picture of crime in Canada
- supporting police services and other justice stakeholders in their resource planning and program development for the policing community
- supporting decisions about the distribution of police resources, definitions of provincial standards and for comparisons with other departments and provinces or territories
- providing information for policy and legislative development, evaluation of new legislative initiatives, and international comparisons, and
- helping media, academics and researchers use these data to examine specific issues about crime.
To provide the most accurate picture of crime in Canada, it is important that all police services respond to the UCR. In addition, the CFMP has expressed its interest to be included in the UCR in order to be transparent regarding crimes on their military bases across the country. The Canadian Centre for Justice and Community Safety Statistics (CCJCSS) has therefore been working with the CFMP in including them in the UCR. In compliance with Statistics Canada's Policy on Informing Users of Data Quality and MethodologyFootnote 2, users of the UCR will be informed of the addition of the CFMP as a respondent to the UCR and of the relevant data quality and methodology through the Agency's Integrated Metadata Base (IMDB) UCR web pageFootnote 3.
The CFMP will be reporting the same information as the current UCR respondents, including personal information on victims and accused involved in criminal incidents in Canada. The survey data file, without direct identifiers other than postal code and date of birth, will be made available to researchers in the Research Data Centres (RDC)Footnote 4 upon approval of requests to access the data for statistical research. Statistics Canada's directives and policies on data publication will be followed to ensure the confidentiality of any data released from the RDCs. Only aggregate results, which are fully anonymized and non-confidential, without direct identifiers, which precludes the possibility of re-identifying individuals, can be released from the RDCs. Statistics Canada will retain this data as long as required for statistical purposes, in order to conduct analysis of long‐term impacts.
Data from the Uniform Crime Reporting Survey (UCR) can be linked to other data sets for both specialized data production activities such as the update of the Crime Severity Index (CSI), as well as for cost-recovery or other specialized projects that address specific justice-related research questions and/or data gaps. Statistics Canada's microdata linkage and related statistical activities were assessed in Statistics Canada's Generic Privacy Impact Assessment.Footnote 5 All data linkage activities are subject to established governanceFootnote 6, and are assessed against the privacy principles of necessity and proportionalityFootnote 7. All approved linkages are published on Statistics Canada's websiteFootnote 8. - Effectiveness - Working assumptions:
Statistics Canada's Uniform Crime Reporting Survey (UCR) is an administrative data source that extracts police service files from their Record Management Systems on a monthly basis. It is the only source of data that collects information on all criminal incidents reported by Canadian police services. It is Canada's primary source of information to monitor the nature and extent of police-reported crimes across the country. Accordingly, including the Canadian Forces Military Police (CFMP) will allow for a more complete picture of crime in Canada.
All Statistics Canada directives and policies for obtaining administrative data under the Statistics Act and dissemination of the data are followed, and information released does not disclose any information identifying victims or accused involved in a criminal incident reported by the police. Further, the CFMP is engaged in ongoing discussions with the CCJCSS to determine if additional vetting rules specific to their data are required to ensure the confidentiality of victims or accused.
The data from the UCR survey is representative of crime in Canada and is disaggregated at the provincial and Census metropolitan Area (CMA)Footnote 9 geographical level and at the incident level. Information on victims and accused including age and gender, are grouped in order to ensure anonymity. More specifically, CCJCSS does not release any tables or cross-tabulations that may identify directly or indirectly a particular victim or suspect by the specifics of an offence. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.
The CFMP, in collaboration with CCJCSS, has established that the CFMP data will be systematically disseminated at a higher level of geography where their individual military bases' data will be published together as one respondent and included in the Canada level rollup only (no province breakout for confidentiality reasons). In addition, there will be ongoing discussions between the CFMP and CCJCSS to establish if additional residual disclosure rules for their data will be required. - Proportionality:
Personal information such as name, date of birth, gender, and characteristics of victims and accused persons involved in criminal incidents, as well as information relating to the nature and extent of the incident is collected from more than 1,000 separate police services and detachments, comprising of over 190 different police forces which will now include the Canadian Forces Military Police. To provide the most accurate picture of crime in Canada, it is important that all police services respond to the UCR.
The data collected through the UCR survey provide key information for crime analysis, resource planning and program development for the policing community.
Municipal and provincial governments use the data to inform decisions about the distribution of police resources, to develop definitions of provincial standards, and for comparisons with other departments or jurisdictions. For the Federal Government, the UCR survey provides information for policy and legislative development, for the evaluation of new legislative initiatives, and for international comparisons.
The UCR survey also provides information to the general public on the nature and extent of police-reported crime and crime trends in Canada. As well, media, academics and researchers use these data to examine specific issues about crime.
All data extracted from police services administrative files are carefully considered to address specific data needs and gaps, and consultations with the Police and Information Statistics Committee (POLIS) from the Canadian Association of Chiefs of Police are conducted on an ongoing basis.
Proportionality has also been considered based on ethics:
Data sensitivity: The data collected from the CMPF could be of a sensitive nature due to the current CAF-DND Sexual Misconduct dossier. The risk of sensitive information being disclosed is minimal since their data will be processed according to Statistics Canada's and the UCR-specific current practices. In particular, personal identifier variables (e.g. name and date of birth of victims and accused) are suppressed from the data and accessible to only a limited number of employees on a need-to-know basis. They are never disclosed.
Approved researchers accessing the UCR microdata files in the RDCs must apply well defined vetting rules in order to use the data for research purposes. It has additionally been agreed on that the CFMP, in collaboration with CCJCSS, could develop additional vetting rules that may be applied, if required, to their specific data and those rules will be consistent across military bases for which information will be collected. - Alternatives:
The Canadian Forces Military Police (CFMP) is not currently responding to the UCR Survey, therefore incidents reported under the Canadian Criminal Code by this force are not accounted for in the Canadian crime statistics released yearly by Statistics Canada. Including the CFMP as an active respondent will address this data gap.
Mitigation factors
Some of the variables contained in the Uniform Crime Reporting Survey (UCR) such as names, date of birth and gender are considered sensitive as they directly identify individuals coming into contact with the police in the context of a criminal incident. The overall risk of harm to the survey respondents has been deemed manageable with existing Statistics Canada safeguards that are described in Statistics Canada's Generic Privacy Impact Assessment, as well as with some additional measures implemented specifically for the UCR. Data from the Canadian Forces Military Police (CFMP) could be perceived as of a higher sensitivity due to the CAF-DND Sexual Misconduct dossier. Additional safeguard measures will therefore be applied on those specific cases reported by the CFMP.
Measures for all StatCan surveys
The following describes some of the measures that apply to all surveys but that demonstrate safeguards of particular importance for the UCR.
The Uniform Crime Reporting Survey (UCR) has collected aggregate police-reported data since 1962 and incident-based data on the characteristics of all criminal incidents, victims and accused persons since 1988. The ongoing success of the UCR incident-based survey is dependent on the cooperation and participation of the police services across Canada. Statistics Canada takes great care to respect the trust that respondent police services place in Statistics Canada to safeguard the information that they have supplied. The confidentiality provisions stipulated by Paragraph 17(1)(b) of the Statistics Act require that the characteristics of individual incidents, accused or victims not be disseminated to the public.Footnote 10
In addition to legislative requirements outlined by the Statistics Act, the Government of Canada and Statistics Canada have established a number of policies around collecting and storing sensitive statistical information, such as data collected through the UCR; these include the Federal Policy on Government SecurityFootnote 11, and Statistics Canada's IT Security Policy and Policy on Privacy and Confidentiality.
To ensure compliance with these security requirements, physical security measures are put in place whereby data, including that of the UCR, are stored in a secure location where access is controlled and limited only to authorized persons on a need-to-know basis. More specifically, access to the data is controlled by file permissions and these permissions are only granted in a way to permit the minimum access necessary to program service delivery. All Statistics Canada employees, including deemed employeesFootnote 12, involved in the production of statistics are aware of their obligation to protect confidentiality and of the legal penalties for wrongful disclosure.
Furthermore, all direct identifiers (e.g. for the UCR, names, date of birth, and address of the victims or accused persons) are removed from the data files during UCR data processing steps and the files that contain these variables are stored in a separate secure area (server). Secure practices and processes are used in the production of statistics and access to any protected information is based strictly on the need-to-know principle and is limited to only those persons whose current work-related responsibilities require access to the information.
Measures for the UCR
The following describes some of the additional measures that apply to the UCR:
Statistics Canada's Canadian Centre for Justice and Community Safety Statistics (CCJCSS) also takes specific steps to preserve data confidentiality and where data users have access to microdata, for example, via Statistics Canada's Research Data Centres (RDCs), confidentiality vetting guidelines specific to the UCR survey have been developed in order to prevent the release of potentially sensitive information that pertains to the characteristics of a particular individual or incident.
Generally speaking, these measures include the UCR data made available in the RDCs for statistical research and analysis, and the results (outputs) of this analysis that can exported out of the RDCs.
UCR data available in RDCs
- A review of all the UCR variables was conducted to determine which variables would be included for access in the RDCs and which would not. Variables were marked for exclusion for one of two main reasons:
- Direct identifiers – The variables that identify specific individuals or incidents (e.g., name of victim, name of accused, FPS (Finger Print Information System) number, police report number (incident file number)). These variables are not available in the RDCs under any circumstances.
- Data quality issues – The data received from police services that are of poor or unknown quality.
- Consideration was also given to possible confidentiality risks that might be associated with the response categories for certain UCR variables, as these responses could potentially reveal sensitive information about the incident, the accused or the victim. For example, the precise time and date of particular incidents may be particularly revealing and as such are not releasable. When presenting data by these variables (e.g., cross-tabulating type of violation by time of incident), the researcher is required to collapse the time into 6-hour intervals. Therefore, where necessary, sensitive response categories are aggregated or collapsed into a more generic category to reduce any risk of disclosure.
- Similarly, individual violation codes for criminal offences captured on the UCR were examined to discern if there were any potential confidentiality issues that might result from releasing information for particular violation codes. Sensitive violation codes such as those related to the sexual violation of children, are grouped into a single category as 'Other Sexual Violations', therefore aggregated or collapsed under a more generic code to reduce any risk of disclosure.
- Finally, the retained variables were examined to identify any potentially sensitive variables (i.e., indirect identifiers or sensitive variables), that could, either alone or when appearing in conjunction with other variables, disclose confidential information. Once the potentially sensitive variables were identified, disclosure rules were then developed to manage the confidentiality concerns related to these variables.
UCR research and analysis outputs that can leave the RDCs
In order for statistical outputs resulting from analysis of the UCR data to receive approval for removal from the RDCs, the following confidentiality vetting rules must be satisfied.
First and foremost, statistical outputs may not be removed from the RDC if they can lead to the possible identification of a particular victim, accused person or incident.
For frequency tables (i.e., counts, rates, totals), a scoring method is used by the RDC analyst to determine the disclosure risk score for each table. The score is based on the nature and level of detail of the variables that make up that table, and increases as more sensitive variables are included. Variables that are potentially more sensitive (i.e., those associated with a higher risk of identification or revealing a characteristic that can be attributed to a single individual or incident) are given higher scores. The lowest score is a 0 and is awarded to variables associated with an extremely low risk of disclosure, whereas a score of 8 is given when the risk of disclosure is extremely high. Researchers will score their tables themselves. Then, after tallying the total disclosure risk score for a table, the researcher, in consultation with the RDC analyst, then compares the table score to the established threshold to determine whether or not it can be released.
For the UCR, the threshold for disclosure risk scores is 7. Frequency tables with scores at or below the threshold may be released as is – regardless of the cell count. Tables with disclosure risk scores above the threshold will not be releasable. In cases where tables score above the threshold, researchers may want to consider removing one or more sensitive variables or try using an aggregated variable to lower the table's overall score.
Alternately, with a few exceptions, researchers can choose to apply a controlled rounding process (rounding to base 5) to their tables. The RDC analyst will verify that the rounding program has been applied correctly.
Measures for UCR CFMP data
In collaboration with the Canadian Centre for Justice and Community Safety Statistics (CCJCSS), it has been determined that the Canadian Forces Military Police (CFMP) would like all of their bases' data published together as one respondent and included in the Canada level rollup only (no province breakout for confidentiality reasons). Additionally, measures such as rolling up certain offence types to avoid the release of disaggregated data that could potentially identify individuals and/or specific vetting rules related to the CFMP data may be developed by the CFMP and Statistics Canada, if required, in order to heighten confidentiality protection.
Conclusion
This assessment concludes that, with the existing Statistics Canada safeguards and additional mitigation factors listed above, any remaining risks are such that Statistics Canada is prepared to accept and manage the risk.
Formal approval
This Supplementary Privacy Impact Assessment has been reviewed and recommended for approval by Statistics Canada's Chief Privacy Officer, Director General for Modern Statistical Methods and Data Science, and Assistant Chief Statistician for Social, Health and Labour Statistics.
The Chief Statistician of Canada has the authority for section 10 of the Privacy Act for Statistics Canada, and is responsible for the Agency's operations, including the program area mentioned in this Supplementary Privacy Impact Assessment.
This Privacy Impact Assessment has been approved by the Chief Statistician of Canada.