1. Estimates of population coverage error

1.1 Introduction

The census defines the population to be enumerated as well as the rules by which the population is to be counted (see Section 4). Coverage error occurs when errors are made relative to these definitions and rules. The main sources of coverage error are omission of a dwelling, which results in omission of the dwelling's residents, and failure by the respondent to include all persons who should be included or to exclude persons who should not be included. This section presents estimates of 2011 Census population net undercoverage, undercoverage and overcoverage. Net undercoverage indicates the extent to which the number of enumerations included in census data is higher or lower than complete enumeration. Both undercoverage and overcoverage may produce a bias in official counts and estimates, because the characteristics of persons who are not included may differ from the characteristics of persons who are included, and the characteristics of duplicates may differ from the characteristics of persons who are included only once.

Top of Page

1.2 Net undercoverage

The population net undercoverage rate for the 2011 Census is estimated at 2.22%.^Footnote1 In other words, the difference between the number of persons who were not included in the census but were members of the census target population and the number of duplicate enumerations was estimated to be 2.22% of the census target population. The population undercoverage rate is estimated at 4.07% (1,391,971 persons), while the population overcoverage rate is estimated at 1.85% (632,846 persons). An undercoverage rate of 4.07% means that persons who were not included but were members of the target population make up 4.07% of the census target population. An overcoverage rate of 1.85% means that duplicate enumerations make up 1.85% of the census target population.

The estimated undercoverage rate is lower than in the 2006 Census, and the estimated overcoverage rate is higher. Consequently, net undercoverage, which is the difference between undercoverage and overcoverage, is lower.

This section presents estimates of net undercoverage for a variety of geographic and demographic variables:

• Province or territory of current residence at the time of the census
• Age and Sex
• Legal marital status and Sex
• Marital status and Sex
• Mother tongue
• Census metropolitan area (CMA) of Census Day usual residence

Table 1.2.2 provides the estimated net undercoverage, the standard error associated to the estimate, as well as the corresponding estimated net undercoverage rate and the standard error associated to various characteristics. Negative estimates of net undercoverage indicate that overcoverage was larger than undercoverage. For an explanation of how this can occur, see Section 9.

The standard error provides a measure of the accuracy of estimates based on sampling. The estimates are considered accurate to within plus or minus two standard errors 19 times out of 20. In other words, there are approximately 19 chances in 20 (95%) that the actual population net undercoverage rate for the 2011 Census was between 1.91% and 2.53% (i.e., 2.22% ± two standard errors), or about two chances in three (68%) that the actual rate was between 2.06% and 2.38% (i.e., 2.22% ± one standard error).

Since net undercoverage is a reflection of both undercoverage and overcoverage, the reader should also consult the estimates of undercoverage and overcoverage presented in Table 1.3. A low rate of net undercoverage may indicate low undercoverage or a combination of high undercoverage and high overcoverage.

Population net undercoverage was highest in the territories. The three territories had the three highest net undercoverage rates in Canada, with Nunavut at 6.22%, the Northwest Territories at 4.55% and Yukon at 3.85%. Among the provinces, Alberta had the highest net undercoverage rate, at 3.41%, followed by Ontario at 2.80% and Saskatchewan at 2.78%. New Brunswick had the lowest rate of population net undercoverage, at 0.52%, followed by Quebec at 0.92%. In 2006, Ontario and Alberta also had high rates, while Quebec had the lowest rate. The rates for the territories were also the highest in the country.

Population net undercoverage is generally higher for men, and highest for young adults (Table 1.2.2).The net undercoverage rate for males was slightly more than two and a half times the rate for females, 3.21% compared with 1.25%. In the general population, net undercoverage is highest in the 20-to-34 age group for both males and females. It is 6.49% in the 20-to-24 age group and 7.27% in the 25-to-34 age group. Men aged 25 to 34 had the highest net undercoverage rate, at 8.79%, compared with 6.34% for the 20-to-24 group. For women, the rate reaches 6.64% for the 20-to-24 age group and 5.74% for the 25-to-34 group. Net undercoverage was negative for women aged 15 to 17, as well as, for women aged 45 and over and for men aged 65 and over, indicating that there were more excess enumerations than persons not enumerated. This may also be due to excessive imputation of persons in these age groups when whole household imputation was carried out to compensate for occupied dwellings misclassified as unoccupied and non-response dwellings on the basis of the results of the Dwelling Classification Survey (see Section 6.2.4).

The net undercoverage rate for the population aged 15 and over was higher for never-married persons.If we consider marital status, more than two-thirds of net undercoverage in the 15-and-over population consisted of persons who had never been legally married and were not in a common-law relationship. The net undercoverage rate for this group was 6.04%. Net undercoverage for persons who were separated and not in a common-law relationship is even higher (10.94%), especially for men (14.19%).

The net undercoverage rate was higher for persons whose mother tongue is English than for persons whose mother tongue is French (1.95% compared with -0.04%), which partly explains the lower net undercoverage rate in Quebec. The net undercoverage rate for allophones, persons whose mother tongue is neither English nor French, was higher (4.06%).

Population net undercoverage was slightly more common outside census metropolitan areas. At the Canada level, the net undercoverage rate was 2.50% for persons who should have been enumerated outside of census metropolitan areas (CMAs). This is slightly higher than the net undercoverage rate of 2.09% for persons not living in CMAs. The only jurisdiction where the non-CMA rate was higher than the CMA rate was Ontario. In Quebec, there is little difference between the two rates.

Top of Page

1.3 Undercoverage

Persons counted as undercoverage are generally persons who were not included as usual residents in the questionnaire that was completed for their usual residence, or persons for whom no questionnaire were completed for their usual residence. For example, persons who regard their residence as temporary may not have been included as usual residents elsewhere. Persons who have no usual residence and were therefore not enumerated, e.g., the homeless are also part of undercoverage.

This section presents estimates of undercoverage for a variety of geographic and demographic variables:

• Province or territory of current residence at the time of the census
• Age and Sex
• Legal marital status and Sex
• Marital status and Sex
• Mother tongue
• Census metropolitan area (CMA) of Census Day usual residence

Table 1.3 provides the estimated undercoverage in terms of the number of persons missed, the corresponding estimated undercoverage rate, the standard error associated with the estimate and the related standard errors. In some cases, the estimated undercoverage is negative (e.g., -4,127 for women aged 55 to 64). For an explanation of how this can occur, see Section 9.

There are some demographic trends in undercoverage:

• The undercoverage rate for males is more than one and a half times the rate for females, 5.07% compared with 3.08%.
• For both males and females, undercoverage is highest in the 20-to-34 age group.
• Among young adult males, undercoverage is 9.37% for the 20-to-24 age group and 10.54% for the 25-to-34 age group.
• If we consider marital status, undercoverage is higher for people aged 15 and over who never married and are not in a common-law relationship. The rate for this group is 8.49%. Undercoverage is also higher for separated persons who are not in common-law relationships. The rate for this group is 12.26%. In both cases, the rates are higher for men than for women.

As shown in Table 1.3, the persons most likely to have been missed in the 2011 Census have the following profiles: men aged 20 to 34 who have never married and are not in a common-law relationship, and separated persons. Mother tongue is also important, as undercoverage is lower for persons whose mother tongue is French (2.13%), followed by persons whose mother tongue is English (3.71%). For persons whose mother tongue is neither English nor French, the rate is even higher, at 5.85%.

Top of Page

1.4 Overcoverage

Population overcoverage is the number of excess enumerations in the census counts for persons enumerated more than once (usually twice) or persons who should not have been enumerated. This error produces a bias because these persons should have been enumerated only once. Examples of overcoverage include children whose parents do not live together and who are included in each parent's census questionnaire, persons who live apart from their family for work reasons and are listed in both their family's questionnaire and the questionnaire for the dwelling where they live while working, students who are away at school and are included in both their roommates' census questionnaire and their parents' questionnaire, and foreign visitors who are enumerated.

This section presents estimates of overcoverage for a variety of geographic, demographic and other variables:

• Province or territory of current residence at the time of the census
• Age and Sex
• Legal marital status and Sex
• Marital status and Sex
• Mother tongue
• Census metropolitan area (CMA) of Census Day usual residence

Table 1.3 also contains estimates of the number of excess enumerations and the associated standard error.

The overcoverage estimates in Table 1.3 have lower standard errors than the undercoverage estimates. There are some demographic trends in overcoverage:

• The estimated overcoverage rate varies less than the undercoverage rate for all provinces and territories except Nunavut, which has a lower overcoverage rate (1.17%) than any other territory or province.
• Males and females have similar overcoverage rates, at 1.86% and 1.83% respectively.
• Overcoverage is higher for the 5-to-24 age range. The highest rate is in the 18-to-19 age group (3.37%). In this age group, the rate is higher for women (3.69%) than for men (3.06%). It is the highest rate for any age group.
• If we look at marital status, overcoverage, like undercoverage, is higher for persons who had never married and are not in a common-law relationship (2.45%). This is true for both men and women.

In summary, as shown in Table 1.3, persons who are most likely to be enumerated more than once are just as likely to be male as female and just as likely to be children as young adults, and for persons aged 15 and over, they are more likely to have never married.