Aboriginal Peoples Technical Report, Census of Population, 2016
5. Comparability

5.1 Historical comparability

The counts for Aboriginal groups, the Registered or Treaty Indian status group, and First Nation or Indian band membership may change over time for a number of reasons. The change (increase or decrease) in the number of people reporting Aboriginal group, Registered or Treaty Indian status, or First Nation or Indian band membership is partly caused by demographic growth, and partly caused by changes in reporting patterns between Aboriginal groups and between Aboriginal and non-Aboriginal populations.

Changes in Aboriginal counts over time can also be affected by differences in the wording and format of Aboriginal questions, legislative changes, differences in the set of incompletely enumerated reserves, and changes made to the definition of reserves.

The counts for the Aboriginal variables may also change as a result of methodology differences and changes between the 2016 Census, the 2011 National Household Survey (NHS), and the 2006 and 2001 censuses (changes related to weighting and calibration, changes to the universe, etc.).

5.1.1 Differences in wording and format of Aboriginal questions

There are various ways to define the Aboriginal population based on the four Aboriginal questions asked in the census (Aboriginal ancestry [Question 17], Aboriginal group [Question 18], Registered or Treaty Indian status [Question 20], and membership in a First Nation or Indian band [Question 21]).

Although they measure the same concepts, the four Aboriginal questions in the 2016 Census differ from the Aboriginal questions in the 2006 Census. The question wording was modified to reflect current terminology and to ensure ongoing accuracy in Aboriginal population measurement. The question wording remained consistent between the 2011 NHS and the 2016 Census.

To see how these questions have changed between 2006 and 2016, please refer to the 2016 Census 2A-L and 2A-R questionnaires, and to the 2006 Census 2B and 2D questionnaires.

5.1.2 Legislative changes

Legislative changes, such as the amendments to the Indian Act of Canada (e.g., Bill C-31 in 1985 and Bill C-3 in 2011^{Note 1}), may affect concepts such as Aboriginal identity and Registered or Treaty Indian status. Legislative changes may have affected how individuals responded to the Aboriginal identity and Registered or Treaty Indian status questions in the 2016 Census. The net effect of these changes cannot be measured. Users should be careful when interpreting the results.

In addition, the Government of Canada, by order in council, recognized the Qalipu Mi’kmaq First Nation in Newfoundland and Labrador under the Indian Act in September 2011. The number of Registered or Treaty Indians living in Newfoundland and Labrador grew from 6,600 in the 2006 Census to 21,155 in the 2016 Census.

5.1.3 Response mobility

In addition to the factors listed above, some people report their Aboriginal identity or ancestry differently from one data collection period to another, for a variety of reasons. Response mobility is the movement of responses from one category to another, such as “No, not an Aboriginal person,” to “Yes, Métis.” Response mobility has a greater impact on estimates for First Nations and Métis populations than on estimates for the Inuit population. Changing attitudes about Aboriginal identity, judicial decisions or anticipated legal changes, the social climate, and other factors may influence how people identify themselves.

Further research from the Social and Aboriginal Statistics Division will use the term “response mobility” to better reflect that the analysis focuses on changes to individuals’ responses to a questionnaire, not on shifts in ethnic or cultural identity.

5.1.4 Other factors that affect responses to the Aboriginal questions

In summary, population estimates for concepts such as Aboriginal identity and Registered or Treaty Indian status are influenced by numerous factors. Users should be aware that point estimates and changes over time are influenced by a combination of natural growth, changes to coverage and measurement, and other factors that affect how people self-identify. It is not possible to quantify the impact these changes have in isolation from each other.

5.1.5 Differences in the list of incompletely enumerated reserves

In 2016, 14 Indian reserves and Indian settlements were incompletely enumerated in the census. Incompletely enumerated reserves have an impact on census estimates for the First Nations population living on reserve, and for the Registered or Treaty Indian population living on reserve. Estimates for other populations could also be affected, including the total Aboriginal identity population, the First Nations (North American Indian) identity population, the total Aboriginal ancestry population, the First Nations (North American Indian) ancestry population, the Registered or Treaty Indian population, the population who reported membership in a First Nation or Indian band, and the population living on Indian reserves and Indian settlements.

Estimates associated with other variables related to First Nations, such as language and band housing, may also be affected by the incomplete enumeration of certain Indian reserves and Indian settlements in the census. This issue does not have an impact on the estimates for the Inuit or Métis populations since those populations are much less likely to live on reserves.

The extent of the impact of the exclusion of the incompletely enumerated reserves depends on the geographic area under study. The impact is much lower for higher geographic areas, such as Canada, provinces and territories, census metropolitan areas, and census agglomerations. The impact may be more significant for lower geographic areas, such as census subdivisions.

5.1.6 Changes made to the census subdivision types associated with the “residence on reserve” population

Statistics Canada uses the definition of “residence on or off reserve” provided by Crown-Indigenous Relations and Northern Affairs Canada (CIRNAC). From time to time, changes are made to the geographies that define Indian reserves and Indian settlements (see the census subdivision (CSD) definition in the Dictionary, Census of Population, 2016, Catalogue no. 98-301-X). In 2006, there were 1,174 on-reserve CSDs; in 2011, there were 997; and, in 2016, there were 984. These changes could be the result of additions, deletions, boundary changes or amalgamations. Additionally, only reserves that contain population are recognized as individual CSDs. Data are adjusted for these changes to compare populations of the affected geographies (e.g., to compare the on- and off-reserve populations). These comparisons can be requested through custom tabulations.

For more information on Indian reserves, consult the CIRNAC website. Users need to request special tabulations for data that have been adjusted for these changes to compare populations of the affected geographies (e.g., comparing on- and off-reserve population).

5.1.7 Comparability of population and dwelling counts over time

In 2016 and 2006, dwellings on reserves for which the occupancy status could not be verified or determined during collection were mostly assumed to be unoccupied. In 2011, dwellings for which the occupancy status could not be verified or determined were mostly assumed to be occupied. Actual occupancy for these dwellings was unknown, and the assumptions made during collection may reflect the actual situation across reserves with different degrees of accuracy.

This difference in methodology does not affect the comparability of counts over different censuses for all reserves. It might have an impact on comparisons of counts for reserves that had a significant number of dwellings with unknown occupancy status at a given point.

Prior to the release of the population and dwelling counts from the 2016 Census, six reserve CSDs were identified as being potentially underestimated. For these CSDs, Statistics Canada initiated a formal review of the population and dwelling counts. For any changes made as a result of the formal review, refer to the Population and dwelling count amendments, 2016 Census. For more information on formal reviews, refer to Statistics Canada’s Policy on response to formal review requests of 2016 Census population and dwelling counts.

When changes in population and dwelling counts are analyzed over time, it is recommended that multiple reference years be included, especially for smaller geographic areas. Smaller population counts may be more affected by non-response. The methodology for the 2016 Census is comparable with that employed for the 2006 Census. Therefore, comparisons between the 2006 Census and the 2016 Census are not affected by this difference in methodology.

5.1.8 Comparability with the Indian Register

The 2016 Census, previous censuses and the Indian Register maintained by CIRNAC are the principal sources of demographic data on the Registered or Status Indian population.

The censuses are snapshots of the population that self-identifies as Registered or Treaty Indians at a particular time, while the Indian Register is a continuous statutory administrative file based on the registration of individuals who meet specific criteria, as defined by the Indian Act.

The census estimate is affected by the undercounting of Registered Indians living on incompletely enumerated reserves where enumeration was not permitted or was interrupted before completion, by the undercoverage of Registered Indians living on participating reserves and settlements, and by the undercoverage of Registered Indians living off reserve. Furthermore, because the short-form questionnaires were used, the census did not collect data that identified Registered Indians living in institutions (hospitals, senior citizen homes, jails, shelters, etc.), or living outside the country.

The Indian Register, conversely, forms one cluster of CIRNAC’s entire Indian Registration System. Although it is considered the authoritative source for the number of Registered Indians as per criteria established in the Indian Act, the Indian Register also comes with its own set of limitations. The main purpose of the Indian Register is to record individual names and a range of non-statutory data in accordance with specific subsections of the Indian Act. It was not designed for statistical analysis. Therefore, data from the application may not fully meet the requirements of some statistical activities (e.g., demographic projections, migration patterns).

There is also a reporting lag between the occurrence of a given life event and its recording in the Indian Register. This means that events that occur in a given year may not be reflected in the Indian Register during that year. The following examples taken from the Registered Indian Population by Sex and Residence 2016 report, the most recent year for which the report is available, illustrate this concept:

1. Infants entitled to be registered at birth may not be registered by the December 31 reporting date. Parents often do not report the birth until a later year. It is estimated that, of all births reported in 2016, approximately 79% actually occurred prior to 2016.
2. Individuals can remain on the Indian Register for some time after they are deceased. A certificate of death or a confirmation of presumed death is normally required to remove a name from the Indian Register.
3. Residency codes are typically updated when a life event is reported, although some bands update them more frequently. Residency remains a voluntary field and is not mandatory to collect.

For additional information on the Indian Register, refer to How to update the Indian Register?

Furthermore, Registered Indian status does not guarantee Canadian residency. There are a number of registrants on the Indian Register who live in other countries. For example, at the end of 2016, there were nearly 23,000 registrants whose province of residence was listed as “Outside of Canada.”

The different purposes, and the methodological and conceptual differences, of the Indian Register and the census result in estimates that are not directly comparable. The estimate of Registered Indians from the census on May 10, 2016, was 820,120. This was 15.5% lower than the estimate from the Indian Register as of December 31, 2016, which was 970,562.

5.1.9 Comparability with Demosim projections

Demosim is a demographic microsimulation projection model developed and maintained at Statistics Canada. Demosim produces dynamic population projections at the levels of provinces, territories, census metropolitan areas and selected smaller geographies, based on a number of characteristics (age, sex, place of birth, Aboriginal identity, Registered Indian status, visible minority group, generation status, languages, highest level of educational attainment and labour force participation, etc.). It produces projections by simulating events such as births, deaths, migrations and changes in level of education according to various population growth scenarios.^{Note 2}

Custom projections^{Note 3} from the Demosim model based on the 2011 NHS were prepared specifically for the purposes of comparing the 2016 Census of Population data with projection results as of May 10, 2016, for a number of variables, including the Aboriginal identity variable.

For the Aboriginal identity variable, two projection scenarios were prepared with different assumptions related to intra-generational response mobility.^{Note 4} One scenario included response mobility according to the average of three periods (from 1996 to 2001, from 2001 to 2006 and from 2006 to 2011), and the other scenario was prepared with an assumption that there was no response mobility as of 2011.

At the Canada level, the 2016 Census estimated the Aboriginal identity population at 1,673,785, a number relatively close to the 1,654,000 Aboriginal people projected in the Demosim scenario with response mobility. The difference was 19,785, or 1.2%. However, the difference increased to 126,785 (7.6%) when the census estimate was compared with the projected number based on the scenario without response mobility. Furthermore, the difference between the two sources was more noticeable when specific Aboriginal groups, especially the Métis and First Nations people, were examined.

Consequently, the overall differences in results between the two sources are smaller between the census and the Demosim projection scenario with response mobility, compared with the results between the census and the Demosim scenario without response mobility.

Notes

• Previous
• Next