ANNEX II: International perspective from Canada

Canada’s multi-pronged approach to inclusive data

Canada’s Disaggregated Data Action Plan (DDAP) is a whole-of-government approach led by Statistics Canada to strengthen the government‘s efforts to address equity, diversity and inclusion. The DDAP fills data and knowledge gaps, supports more representative data collection methods, enhances statistics on diverse populations to allow for intersectional analyses, and supports government and societal efforts to address known inequalities and promote fair and inclusive decision-making. Concretely, the DDAP prioritizes enhanced engagement and communication, expanded disaggregated data holdings, increased access to disaggregated data, increased analytical insights on diverse groups of people, and the promotion of national statistical standards, explained further in the Statistics Canada DDAP infographic.

Disaggregated data and the examination of data from an intersectional perspective are important because, unlike aggregated data, they may reveal important insights between and among different groups.

While diversity is a key characteristic of Canadian society and producing statistics and insights is part of Statistics Canada’s mandate, there have been government initiatives to help the agency set priorities. These enablers include, but are not limited to, the Gender-based Analysis Plus, and Canada’s Anti-Racism Strategy.

Canada also disseminates data and monitors progress in relation to several key Government of Canada commitments. This includes the Canadian Indicator Framework for Sustainable Development Goals, which includes data disaggregation by identity factors, subject to data availability. The Gender Results Framework, which is aligned with the Government of Canada’s policy of Gender-based Analysis Plus, ensuring that gender is considered in addition and in relation to other intersecting identity factors, including age, disability, education, ethnicity, race, geography, sex, religion, economic status, and language. Two other frameworks, curated for different purposes, are worth mentioning: the Quality of Life Framework and the Social Inclusion Framework, which was developed to address racism, hence disaggregation by racialized groups. The indicators included in these two frameworks are similar and for both, the objective is to be able to produce the indicators for different population groups (i.e. to further data disaggregation).

Statistics Canada regularly produces data tables, analytical insights and in-depth research, which shed light on inequality issues experienced by different population groups. It is possible to find this information by searching by topics on the Statistics Canada website. Statistics Canada also has a Hub that is dedicated to statistics concerning gender, diversity and inclusion in Canada, as well as relevant Government of Canada initiatives. Within this Hub, there are different portals that further focus on specific population characteristics (e.g., Accessibility statistics, Immigration and ethnocultural diversity statistics; Social inclusion statistics).

Inclusive data and the Census in Canada

From Canada’s perspective, the Census of Population, conducted every 5 years, best facilitates the collection of identity factors in terms of coverage, especially for vulnerable population groups.

At present, Canada’s population estimation relies on Census data for benchmarking and re-basing, and Canada’s population projections on racialized groups, religions, etc. use the Census long-form (where ethnocultural characteristics are collected).

Statistics Canada recognizes that there are other data quality concerns for using solely administrative data sources. For example, Census is stock data; it captures the portrait of the nation’s population at a given time (Census Day). Administrative data, especially population characteristics, are usually collected at the beginning of program administration, hence could be static. Some identity factors could change over time (e.g., gender identity, ethnic identity, religion, disability status, etc.). Further, unless multiple data sources are linked, the content of administrative data could be limited, hence the capacity to explain a given estimate (e.g., income) could be limited. As such, administrative data could produce good indicators, but not the necessary context or explanations.