Statistics Canada
www.statcan.gc.ca
Common menu bar links
Projections of the Diversity of the Canadian Population
This statistical program develops projections of the ethnocultural composition of population for Canada, provinces and Census metropolitan areas, based on various assumptions and scenarios on the components of population growth.
Detailed information for 2006 to 2031
Data release – March 9, 2010
Description
This statistical program develops population projections according to some ethnocultural variables (visible minority group, generation status, place of birth, religious denomination and mother tongue) for Canada, provinces and census metropolitan areas from 2006 to 2031, based on various assumptions and scenarios on the components of population growth and differential demographic behaviors.
These projections are useful to decision makers in their planning of programs related to social cohesion, labor market integration, fight against racism and discrimination, multiculturalism, immigration and urban development. The projections are also intended for others organisations, researchers, academics, students as well as any other person interested in the evolution of the ethnocultural diversity in Canada.
Subjects
- Ethnic diversity and immigration
- Population and demography
- Population estimates and projections
- Visible minorities
Data sources and methodology
Data sources
Data are extracted from administrative files and derived from other Statistics Canada surveys and/or other sources.
The base population for the projections is drawn from the database of the 2006 Census (20% microdata file), which was adjusted to take into account census net undercoverage.
The assumptions underlying the population projections were developed from various data sources: 1996, 2001 and 2006 population censuses (20% samples), surveys (General Social Survey, Ethnic Diversity Survey), and administrative data (population estimates, vital statistics, files of Citizenship and Immigration Canada, Longitudinal Administrative Database, 1991 Canadian census mortality follow-up database).
Estimation
The projections were established using a microsimulation model that differs from the traditional cohort component model in that it uses microdata rather than aggregated data, which allows to project a great number of individual characteristics.
The starting point for the projections consists of the microdatabase of the 2006 Census (20% sample), adjusted to take into account census net undercoverage. This database includes more than six million respondents, each with their individual characteristics: age, sex, marital status, place of residence, generation status and year of immigration, place of birth, visible minority group, and mother tongue.
As is also the case for the cohort-component model, the population at time T+1 results from changes undergone by the population in the previous year. But the changes occur at the individual level rather than at the level of population aggregates. In the model, an individual may change marital status, bear a child, change place of residence, leave Canada or die. Also, new individuals are added over time, either through birth or through immigration.
For each person, the model calculates the probability that these events will occur based on the person's own characteristics. Thus, for example, the probability that a woman will give birth to a child differs depending on whether she is married or single, belongs to a visible minority group, etc. Using a Monte Carlo process and based on the probabilities associated with each event, the model determines for each individual which event will occur first and calculates the amount of time that will elapse before it occurs. Each time an event occurs, the probabilities are recalculated to take account of the individual's new characteristics. For example, the probability of changing one's place of residence declines after the birth of a child. The model advances individuals in this manner until 2031, unless they die or leave Canada in the meantime.
To operationalize such a model, it was necessary to use various data sources and methods. The parameters of the model--especially for estimating relative risks (e.g., risk of changing place of residence depending on presence of children in the home, visible minority group, place of birth, age, etc.)--were obtained using different multivariate analysis methods on different data sources. To reduce the variability of the results, vital statistics were used to estimate basic risks for fertility, mortality and emigration.
The projection model also requires that assumptions be developed in advance on each of the components of population growth (including differential behaviours). Thus, this projection exercise is based on three fertility assumptions, three immigration levels, two ways of selecting immigrant characteristics, two assumptions regarding internal migration, and a single assumption regarding mortality and emigration. Five combinations of assumptions were then selected to create projection scenarios that are plausible in light of past trends.
Disclosure control
Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Various confidentiality rules are applied to all data that are released or published to prevent the publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.
Revisions and seasonal adjustment
The model may be subject to revisions to be carried out on a cost recovery basis.
Data accuracy
The accuracy of any projection depends on the quality of the data relating to the base population and the components of population growth, as well as the adequacy of the assumptions on future trends. Projections are not predictions; they instead represent an effort to establish plausible scenarios based on assumptions regarding population growth, which are themselves subject to uncertainty. Accordingly, it cannot be claimed that the values to be observed over the coming years will always remain within the range suggested by the low and high growth scenarios.
