Ethnicity, Language and Immigration Thematic Series
Choosing the reference population for census statistics on language of work
Correction Notice
The following correction has been made to this document: in table 3, the last two percentages of the 4th column (Quebec, Non-employed persons with recent work experience) are 1.9% (“Outside Canada”) and 17.3% (“No usual place of work”).
Skip to text
Text begins
Since 2001, the Census of Population long-form questionnaire has contained a question about languages used at work. This question is included in a section containing several questions about employment status and labour, asked to people aged 15 and older.
As is the case for other labour-related questions, the question about languages used at work can refer to a job held by the respondent during two separate periods. Respondents are first asked to answer the question in reference to the job they spent the most hours at during the census reference week—for the 2021 Census, this was the week of May 2 to 8, 2021. If respondents did not hold a job during that week, they were asked to answer in reference to the job they held the longest since January 1 of the previous year (which for the 2021 Census was January 1, 2020), if applicable.
There are therefore two distinct populations for which we have information on languages of work: people employed during the census reference week (17,321,700 people in Canada in the 2021 Census) and people with recent work experience (between January 1 of the previous year and the census reference week) but who did not hold a job during the reference week (3,308,820 people). Among the latter, the majority (55%) were no longer part of the active labour force during the census reference week,Note and the others were unemployed.
This technical note aims to discuss the choice of the reference population for census statistics on language of work, and to present the approach on the matter adopted by Statistics Canada when disseminating data from the 2021 Census.
Choosing a reference population to facilitate the use and interpretation of data on languages of work
In Statistics Canada’s products from census cycles from 2001 to 2016, data on languages of work were usually presented by including both people employed during the census reference week and those with recent work experience but who were no longer employed during the census reference week.
For the dissemination of 2021 Census data, there was a change of approach. For most analytical and data visualization products, the default reference population includes only those who held a job during the census reference week. That said, most census data tables allow users to select a reference population based on labour force status by themselves, according to their preference.
This change in the default reference population is driven by considerations related to the availability of complementary information about labour, as well as ease of data interpretation.
First, considering only employed workers for language-of-work products ensures consistency with other census information about labour (occupation, industrial sector, etc.) and the analytical and data products on these topics. In those products, the reference population usually only includes persons employed during the census reference week. In fact, some key information about labour, including place of work, is available only for people who were employed during the census reference week.
Using this narrower reference population also facilitates the interpretation of data. Interpreting information about people who are not employed, but who have had recent work experience, can be complex. This is in part due to the heterogeneity of this population, which includes unemployed people, retirees, students who had a summer job the previous year, recent immigrants whose last job was outside Canada, etc. In all cases, the information pertains to a job that is no longer being held by the person at the time of the census. The position that was held by that person may have been assigned to someone else (in which case it is a form of double counting), or may simply no longer exist.
In some cases, there are also measurement differences depending on people’s work activity, for instance in cases where people are holding more than one job at once.Note If the job of interest for the employed population is the one at which the worker spent the most hours during the census reference week, the job of interest for the non-employed population is that held the longest by the person during the 16-month period prior to the census. These two definitions can refer to different jobs.
Finally, for people who were not employed but had recent work experience, the exact period during which a job was held since January 1 of the previous year is unknown. In all cases, there is a lag between the time that job was held and the time of the census. Therefore, using this population to analyze language dynamics at work can lead to inconsistencies and erroneous conclusions, especially when cross-tabulating these data with other census data that have the time of the census as a reference period (place of residence, languages known or spoken at home, etc.). Furthermore, if the census does provide a complete portrait of people employed at the time of the census, the same cannot be said for non-employed people who had held a job since January 1 of the previous year, as some of these people will in the meantime have left Canada (a frequent occurrence for certain sub-populations such as non-permanent residents), joined a collective household,Note or died.
In short, interpreting data regarding people who are not employed but who have recent work experience is complex, and above all it is different from interpreting data about people who were employed during the census reference week.
How the choice of a reference population affects language of work statistics
Choosing a reference population, that is, excluding or including people who were not employed during the census reference week but who had recent work experience, has a significant impact on the number of people in the population, but a limited impact on the proportions of people with a given language-use profile. However, this impact can vary based on the sub-population of interest.
Here, we examine the effect of the choice of reference population on counts of workers, rates of use of languages at work, characteristics of workers and trends over time. Data are presented for Canada as a whole, as well as for Quebec and New Brunswick, as these provinces stand out when it comes to languages of work.
Description for Figure 1
This figure shows the distribution of the population by employment situation at the time of the 2021 Census.
In total, there were 30.3 million people aged 15 or more in Canada.
- 17.3 million people, or 57% of the population, were employed during the census reference week.
- 3.3 million people, or 11% of the population, were not employed, but had worked at some point since January 1, 2020.
- 9.7 million people, or 32% of the population, had not worked since January 1, 2020.
Impact on worker counts
The main impact of this choice involves the total number of workers in the reference population (Table 1). Considering only people who were employed during the census reference week reduces the total size of the reference population by 16.0% in Canada (from 20.6 million people employed or having held a job since January 1, 2020, to 17.3 million employed people).
This proportion can vary from one region to another. For example, the reduction in the number of workers was less pronounced in Quebec (-13.6%) and more pronounced in New Brunswick (-17.5%). These differences can be explained by a set of factors, including variations in the labour market conditions or in the characteristics of the working population (age structure, for instance).
| Canada | Quebec | New Brunswick | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Persons employed or with recent work experience | Employed persons only | Non-employed persons with recent work experience | Persons employed or with recent work experience | Employed persons only | Non-employed persons with recent work experience | Persons employed or with recent work experience | Employed persons only | Non-employed persons with recent work experience | |
| number | |||||||||
| Language used most often at work | |||||||||
| Total | 20,630,525 | 17,321,700 | 3,308,820 | 4,747,110 | 4,100,445 | 646,660 | 423,255 | 349,210 | 74,040 |
| English | 15,879,855 | 13,260,375 | 2,619,475 | 668,485 | 570,385 | 98,100 | 317,535 | 264,635 | 52,900 |
| French | 3,986,330 | 3,446,695 | 539,630 | 3,773,015 | 3,270,095 | 502,920 | 88,705 | 70,110 | 18,600 |
| Indigenous language | 19,435 | 15,105 | 4,335 | 10,215 | 8,160 | 2,060 | 150 | 100 | 50 |
| Other non-official language | 291,020 | 211,030 | 79,985 | 26,030 | 18,480 | 7,550 | 615 | 400 | 220 |
| English and French | 321,975 | 280,000 | 41,975 | 249,940 | 217,520 | 32,420 | 15,610 | 13,465 | 2,145 |
| English and non-official language(s) | 115,955 | 95,500 | 20,450 | 7,155 | 5,725 | 1,430 | 595 | 480 | 115 |
| French and non-official language(s) | 6,285 | 5,065 | 1,215 | 6,010 | 4,855 | 1,155 | 10 | 0 | 0 |
| English, French and non-official language(s) | 7,400 | 6,205 | 1,195 | 6,170 | 5,165 | 1,000 | 40 | 25 | 10 |
| Multiple non-official languages | 2,270 | 1,715 | 555 | 90 | 60 | 30 | 0 | 0 | 0 |
| Languages used at least regularly at work | |||||||||
| Total | 20,630,525 | 17,321,700 | 3,308,820 | 4,747,110 | 4,100,445 | 646,660 | 423,255 | 349,210 | 74,040 |
| English | 14,942,170 | 12,452,045 | 2,490,120 | 350,525 | 290,920 | 59,605 | 284,045 | 235,385 | 48,660 |
| French | 3,164,890 | 2,724,635 | 440,255 | 3,032,285 | 2,617,765 | 414,520 | 63,250 | 48,735 | 14,515 |
| Indigenous language | 6,935 | 5,245 | 1,685 | 4,170 | 3,275 | 900 | 50 | 35 | 20 |
| Other non-official language | 162,265 | 109,865 | 52,405 | 14,650 | 9,835 | 4,810 | 340 | 225 | 120 |
| English and French | 1,682,510 | 1,475,875 | 206,630 | 1,245,530 | 1,096,060 | 149,475 | 72,790 | 62,660 | 10,130 |
| English and non-official language(s) | 570,220 | 470,660 | 99,560 | 25,880 | 20,245 | 5,630 | 2,330 | 1,860 | 470 |
| French and non-official language(s) | 20,935 | 17,235 | 3,700 | 20,370 | 16,835 | 3,535 | 20 | 0 | 20 |
| English, French and non-official language(s) | 69,335 | 58,300 | 11,040 | 52,965 | 45,050 | 7,910 | 415 | 320 | 100 |
| Multiple non-official languages | 11,260 | 7,835 | 3,430 | 730 | 455 | 270 | 15 | 0 | 15 |
| Source: Statistics Canada, 2021 Census. | |||||||||
Impact on the rates of language use at work
The choice of the reference population also has an impact, though relatively minor, on the rates of language use at work (Table 2). Nationally, when people who were not employed during the census reference week are excluded, we see slightly lower rates of use of English, slightly higher rates of use of French, and lower rates of use of other, non-official languages, particularly non-Indigenous languages.
Some of these differences at the national level, particularly regarding English and French, are caused by a compositional effect. As mentioned above, retaining only employed people has a smaller impact on the total number of workers in the reference population in Quebec than in Canada as a whole. Because the majority of workers using French are located in this province, this has the effect of increasing the relative share of French at work in Canada.
For Quebec, the impact of the reference population choice is similar to that for Canada as a whole (namely, higher rates of use of French when considering only employed people), but this impact is much more limited. For New Brunswick, the impact is the opposite: retaining only employed people leads to slightly lower rates of use of French and higher rates of use of English.
Proportionally, the most significant impact of changing the reference population involves the exclusive use of non-official languages other than Indigenous languages. Considering only employed people results in a 32% decrease in the number of workers who reported using neither English nor French on a regular basis at work. This is partly because the population without a job but with recent work experience more often includes people whose last job was outside Canada. In 2016,Note 1.8% of people with recent work experience answered the census questions about labour by referring to a job held outside Canada, compared with 0.5% of those employed during the census reference week. Jobs held outside Canada more often involved the main or exclusive use of languages other than English or French.
| Language | Canada | Quebec | New Brunswick | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Persons employed or with recent work experience | Employed persons only | Non-employed persons with recent work experience | Persons employed or with recent work experience | Employed persons only | Non-employed persons with recent work experience | Persons employed or with recent work experience | Employed persons only | Non-employed persons with recent work experience | |
| percent | |||||||||
| Language used most often at work | |||||||||
| Total | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 |
| English | 77.0 | 76.6 | 79.2 | 14.1 | 13.9 | 15.2 | 75.0 | 75.8 | 71.4 |
| French | 19.3 | 19.9 | 16.3 | 79.5 | 79.7 | 77.8 | 21.0 | 20.1 | 25.1 |
| Indigenous language | 0.1 | 0.1 | 0.1 | 0.2 | 0.2 | 0.3 | 0.0 | 0.0 | 0.1 |
| Other non-official language | 1.4 | 1.2 | 2.4 | 0.5 | 0.5 | 1.2 | 0.1 | 0.1 | 0.3 |
| English and French | 1.6 | 1.6 | 1.3 | 5.3 | 5.3 | 5.0 | 3.7 | 3.9 | 2.9 |
| English and non-official language(s) | 0.6 | 0.6 | 0.6 | 0.2 | 0.1 | 0.2 | 0.1 | 0.1 | 0.2 |
| French and non-official language(s) | 0.0 | 0.0 | 0.0 | 0.1 | 0.1 | 0.2 | 0.0 | 0.0 | 0.0 |
| English, French and non-official language(s) | 0.0 | 0.0 | 0.0 | 0.1 | 0.1 | 0.2 | 0.0 | 0.0 | 0.0 |
| Multiple non-official languages | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| Languages used at least regularly at work | |||||||||
| Total | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 |
| English | 72.4 | 71.9 | 75.3 | 7.4 | 7.1 | 9.2 | 67.1 | 67.4 | 65.7 |
| French | 15.3 | 15.7 | 13.3 | 63.9 | 63.8 | 64.1 | 14.9 | 14.0 | 19.6 |
| Indigenous language | 0.0 | 0.0 | 0.1 | 0.1 | 0.1 | 0.1 | 0.0 | 0.0 | 0.0 |
| Other non-official language | 0.8 | 0.6 | 1.6 | 0.3 | 0.2 | 0.7 | 0.1 | 0.1 | 0.2 |
| English and French | 8.2 | 8.5 | 6.2 | 26.2 | 26.7 | 23.1 | 17.2 | 17.9 | 13.7 |
| English and non-official language(s) | 2.8 | 2.7 | 3.0 | 0.5 | 0.5 | 0.9 | 0.6 | 0.5 | 0.6 |
| French and non-official language(s) | 0.1 | 0.1 | 0.1 | 0.4 | 0.4 | 0.5 | 0.0 | 0.0 | 0.0 |
| English, French and non-official language(s) | 0.3 | 0.3 | 0.3 | 1.1 | 1.1 | 1.2 | 0.1 | 0.1 | 0.1 |
| Multiple non-official languages | 0.1 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| Source: Statistics Canada, 2021 Census. | |||||||||
Impact on worker characteristics
In addition to place of work, the characteristics of people employed during the census reference week and those of people not employed but who had recent work experience differ on a number of other points (Table 3).
First, among those not employed but with recent work experience, there is significant overrepresentation of people aged 15 to 24 (an age group that includes many students working only part of the year) and of those aged 60 and older (an age group that includes many people who are retiring).
People not employed at the time of the census were also overrepresented in some industrial sectors, particularly those affected by strong seasonality (agriculture, forestry, fishing and hunting), as well as those most disrupted by the COVID-19 pandemic (arts, entertainment and recreation; accommodation and food services).
In parallel to differences pertaining to languages of work, there are also differences in languages spoken at home. In Quebec, those not employed but with recent work experience were slightly more likely to speak English most often at home, and slightly less likely to speak French most often. In New Brunswick, the opposite was seen.
| Characteristic | Canada | Quebec | New Brunswick | |||
|---|---|---|---|---|---|---|
| Employed persons only | Non-employed persons with recent work experience | Employed persons only | Non-employed persons with recent work experience | Employed persons only | Non-employed persons with recent work experience | |
| percent | ||||||
| Labour force activity (census reference week) | ||||||
| Employed | 100.0 | Note ...: not applicable | 100.0 | Note ...: not applicable | 100.0 | Note ...: not applicable |
| Unemployed | Note ...: not applicable | 45.4 | Note ...: not applicable | 37.8 | Note ...: not applicable | 43.6 |
| Inactive | Note ...: not applicable | 54.6 | Note ...: not applicable | 62.2 | Note ...: not applicable | 56.4 |
| Age | ||||||
| 15 to 24 years | 11.7 | 25.2 | 12.4 | 26.9 | 12.3 | 24.0 |
| 25 to 59 years | 75.7 | 51.1 | 75.7 | 46.1 | 74.4 | 47.3 |
| 60 years and over | 12.6 | 23.7 | 11.9 | 26.9 | 13.3 | 28.7 |
| Immigration status | ||||||
| Non-immigrants | 71.3 | 71.7 | 80.4 | 78.9 | 91.3 | 93.5 |
| Immigrants | 25.5 | 25.3 | 16.6 | 17.5 | 6.6 | 5.1 |
| Non-permanent residents | 3.2 | 3.1 | 3.0 | 3.6 | 2.1 | 1.4 |
| Industrial sector | ||||||
| Agriculture, forestry, fishing and hunting | 2.3 | 2.6 | 1.8 | 2.4 | 3.4 | 8.0 |
| Mining, quarrying, and oil and gas extraction | 1.3 | 1.2 | 0.6 | 0.5 | 0.8 | 1.1 |
| Utilities | 0.8 | 0.4 | 0.7 | 0.3 | 1.1 | 0.5 |
| Construction | 7.7 | 8.0 | 6.8 | 6.6 | 6.5 | 11.7 |
| Manufacturing | 8.3 | 6.5 | 10.5 | 8.1 | 7.9 | 8.8 |
| Wholesale trade | 3.3 | 2.2 | 3.4 | 2.5 | 2.5 | 1.6 |
| Retail trade | 11.2 | 13.9 | 11.7 | 14.4 | 12.1 | 12.7 |
| Transportation and warehousing | 5.2 | 5.2 | 4.8 | 4.9 | 4.8 | 4.6 |
| Information and cultural industries | 2.1 | 1.7 | 2.2 | 1.5 | 1.8 | 1.2 |
| Finance and insurance | 4.6 | 2.0 | 4.1 | 1.8 | 3.7 | 1.4 |
| Real estate and rental and leasing | 1.9 | 1.6 | 1.5 | 1.4 | 1.1 | 1.1 |
| Professional, scientific and technical services | 8.6 | 6.1 | 8.1 | 5.9 | 5.2 | 3.7 |
| Management of companies and enterprises | 0.3 | 0.1 | 0.1 | 0.1 | 0.2 | 0.1 |
| Administrative and support, waste management and remediation services | 4.0 | 5.9 | 3.8 | 5.8 | 4.2 | 5.7 |
| Educational services | 7.7 | 7.1 | 8.1 | 6.5 | 7.9 | 5.1 |
| Health care and social assistance | 13.6 | 8.6 | 14.7 | 9.3 | 15.5 | 10.1 |
| Arts, entertainment and recreation | 1.5 | 5.4 | 1.4 | 5.8 | 1.4 | 3.3 |
| Accommodation and food services | 5.0 | 12.4 | 4.6 | 13.4 | 5.8 | 9.2 |
| Other services (except public administration) | 4.0 | 5.9 | 4.3 | 5.1 | 4.5 | 4.4 |
| Public administration | 6.7 | 3.5 | 6.9 | 3.7 | 9.7 | 5.7 |
| Language spoken most often at home | ||||||
| English | 63.6 | 66.6 | 10.7 | 12.6 | 69.2 | 67.2 |
| French | 19.8 | 15.8 | 77.7 | 73.6 | 25.7 | 29.0 |
| Indigenous language | 0.1 | 0.2 | 0.3 | 0.4 | 0.0 | 0.1 |
| Other non-official language | 12.1 | 13.0 | 7.0 | 8.6 | 2.8 | 1.9 |
| English and French | 0.6 | 0.6 | 1.7 | 1.9 | 1.3 | 1.2 |
| English and non-official language(s) | 3.2 | 3.3 | 0.7 | 1.0 | 0.8 | 0.6 |
| French and non-official language(s) | 0.4 | 0.3 | 1.4 | 1.4 | 0.1 | 0.0 |
| English, French and non-official language(s) | 0.1 | 0.1 | 0.5 | 0.5 | 0.0 | 0.0 |
| Multiple non-official languages | 0.1 | 0.1 | 0.0 | 0.0 | 0.0 | 0.0 |
| Usual place of work (2016Table 3 Note 1) | ||||||
| In the province of residence | 87.1 | 76.0 | 89.2 | 79.0 | 86.8 | 69.3 |
| In another province | 0.9 | 2.1 | 1.7 | 1.8 | 1.6 | 4.5 |
| Outside Canada | 0.5 | 1.8 | 0.3 | 1.9 | 0.4 | 1.0 |
| No usual place of work | 11.5 | 20.0 | 8.9 | 17.3 | 11.2 | 25.2 |
... not applicable
|
||||||
Impact on trends over time
Regarding comparisons over time, the choice of the reference population has little effect on trends, particularly for specific geographic areas (Chart 1).Note
Nationally, excluding people not employed during the census reference week transforms a slight increase in the percentage of workers using mainly English at work between 2016 and 2021 into a slight decrease, and slightly inflates an upward trend for French. However, this essentially reflects a compositional effect, due to the fact, discussed above, that the change in the reference population does not have the same impact on the counts of workers in all provinces. More specifically, only keeping employed people has a smaller impact on the counts of workers in Quebec—where the large majority (95% in 2021) of workers using mainly French are concentrated—than in the rest of Canada. The proportion of people residing in Quebec among employed people rose from 22.9% in 2016 to 23.7% in 2021, while it only rose from 22.7% to 23.0% when including people with recent work experience.
When only Quebec or New Brunswick is considered, excluding people not employed during the census reference week does not significantly affect the trends observed between 2016 and 2021.Note
Data table for Chart 1
| Canada | Quebec | New Brunswick | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Persons employed or with recent work experience | Employed persons only | Persons employed or with recent work experience | Employed persons only | Persons employed or with recent work experience | Employed persons only | |||||||
| 2016 | 2021 | 2016 | 2021 | 2016 | 2021 | 2016 | 2021 | 2016 | 2021 | 2016 | 2021 | |
| percent | ||||||||||||
| French mainly used | 19.2 | 19.4 | 19.4 | 19.9 | 79.9 | 79.6 | 80.0 | 79.9 | 21.6 | 20.9 | 20.9 | 20.1 |
| French and English used equally | 2.1 | 1.6 | 2.2 | 1.7 | 7.2 | 5.4 | 7.3 | 5.4 | 4.4 | 3.7 | 4.4 | 3.9 |
| English mainly used | 77.3 | 77.5 | 77.2 | 77.1 | 12.1 | 14.2 | 12.0 | 14.0 | 73.8 | 75.2 | 74.5 | 75.9 |
|
Note: “French mainly used” and “English mainly used” include people who declared using only one of these languages most often at work, including those who used one of these languages equally with a non-official language. Sources: Statistics Canada, 2016 and 2021 Censuses. |
||||||||||||
Conclusion
When disseminating 2021 Census data on languages of work, the reference population used by default by Statistics Canada in analytical and data visualization products is that of people who were employed during the census reference week. The rationale for this choice was a desire to harmonize the release of these data with that of other labour-related census data and to make it easier for users to interpret the data. Although this way of presenting the data has relatively minor impacts on language of work statistics, it presents several advantages in terms of coherence and interpretation of data.
Compared with people employed during the census reference week, non-employed people with recent work experience have a particular set of characteristics (overrepresentation among certain age groups and sectors, etc.). Depending on their needs, some users may find it appropriate to focus on these non-employed people, or to include them in their population of interest along with people employed during the census reference week. Several 2021 Census data tables allow users to select the population of reference they prefer, providing them with more flexibility. However, those users must bear in mind that there are differences in the interpretation of data on languages of work depending on the reference population.
- Date modified: