Section 4: Survey methodology

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Population coverage

The LFS is a monthly household survey of a sample of individuals who are representative of the civilian, non-institutionalized population 15 years of age or older. It is conducted nationwide, in both the provinces and the territories. Excluded from the survey's coverage are: persons living on reserves and other Aboriginal settlements in the provinces; full-time members of the Canadian Forces and the institutionalized population. These groups together represent an exclusion of approximately 2% of the population aged 15 and over.

National Labour Force Survey estimates are derived using the results of the LFS in the provinces. Territorial LFS results are not included in the national estimates, but are published separately.

Why the territories are excluded from the national total

The Labour Force Surveys conducted in the North are extended pilot projects. Difficulties exist with respect to reaching small communities in the territories, and as a result even within the pilot projects there are areas of the territories that are excluded. As well, since the sample design, rotation pattern and reliability criteria are different from those in the ten provinces, estimates for the territories are not included with the provincial totals, but rather they are calculated and reported separately as a part of each of the extended projects.

The LFS has been conducted as a pilot survey since 1991 in the Yukon, and since late 2000 in the Northwest Territories and Nunavut. Given the special difficulties in collecting data in northern areas, and associated data quality issues at the beginning of the pilot projects, data is available since 1992 for the Yukon, 2001 for the Northwest Territories and 2004 for Nunavut.

Details on the survey methodology for the North can be found in the sub-section Differences in survey methodology between the North and the provinces.

Other exclusions from the Labour Force Survey

Indian reserves have historically been excluded from the LFS due to the serious challenges in contacting and interviewing potential respondents, with many of them living in remote locations not easily accessible to LFS interviewers given the short data collection period each month, and the large effort and cost associated with traveling to these locations.

The LFS also excludes residents of institutions (for example, inmates of penal institutions and patients in hospitals or nursing homes who have resided in the institution for more than six months) for conceptual reasons; the LFS is designed to measure the labour force participation in the current labour market. Residents of institutions are for the most part not able to participate in the labour market and are not economically active. There would also be difficulties associated with the practical implications of sampling and interviewing residents of institutions (for example, access within prisons or interviewing people that are very ill) that would make their inclusion operationally problematic.

There would also be practical difficulties associated with sampling and interviewing full-time members of the armed forces, since many of these persons live in locations that are not accessible for the purposes of conducting the LFS, such as naval vessels, military camps and barracks. Many of them may be stationed in other countries. While not included in the Labour Force survey, the monthly numbers of employed full-time members of the Canadian Forces are available from other administrative sources, such as the Department of National Defence.

Sample design and sample size for the provinces

Canada's population lives in various geographic areas such as provinces and regions within provinces. For the purposes of sampling, the population in these areas is further partitioned into strata, in order to maximize the reliability of the estimates while keeping collection costs at a minimum. Dwellings in strata are not selected directly. Small well-defined areas called clusters are mapped across all parts of the 10 provinces. Each cluster contains approximately 200 households according to the 2006 Census. These clusters are used as the unit for stratification, as well as the unit for sample selection within stratum. A sample of clusters is selected in each stratum. All dwellings within selected clusters are listed and a sample of dwellings is chosen from each list. Chosen in this fashion, the sample is representative of the population.

The number of households sampled across the country has varied over the years as a result of varying levels of funding, and improvements in the survey design. The sample size has been approximately 56,000 households. The sample is allocated to provinces and strata within provinces in the way that best meets the need for reliable estimates at various geographic levels. These include national, provincial, census metropolitan areas (large cities), economic regions, and employment insurance regions. The following guidelines were used in sample allocation:

  1. Canada and provinces: estimates of unemployment should not have a CV (standard error relative to the estimate) greater than 2 percent for Canada, and 4 to 7 percent for the provinces. (See section 7 for explanation of sampling error and CVs (Coefficients of variation)).
  2. Census Metropolitan Areas (CMAs)/Employment Insurance Regions: CVs of 15 percent or less for 3 month average estimates of unemployed.
  3. Economic Regions (ERs): CVs of 25 percent or less for 3 month average estimates of unemployed.

Sample rotation for the provinces

The LFS follows a rotating panel sample design, in which households remain in the sample for six consecutive months. The total sample consists of six representative sub-samples or panels, and each month a panel is replaced after completing its six month stay in the survey. Outgoing households are replaced by households in the same or a similar area. This results in a five-sixths month-to-month sample overlap, which makes the design efficient for estimating month-to-month changes. The rotation after six months prevents undue respondent burden for households that are selected for the survey.

Selection of household members for the provinces

Demographic information is obtained for all persons in a household for whom the selected dwelling is the usual place of residence. Labour force information is obtained for all civilian household members 15 years of age or older. Respondent burden is minimized for the elderly (age 70 and over) by carrying forward their responses for the initial interview to the subsequent five months in survey.

For comprehensive information on the LFS methodology see the publication Methodology of the Canadian Labour Force Survey, catalogue no.71-526-X.

Differences in survey methodology between the North and the provinces

With the recent increase in coverage in Nunavut, the sample for all three territories is representative of the working-age population of each territory. Nunavut was initially designed to cover ten of the largest communities in the region, representing about 70% of all Nunavut residents aged 15 years and over. The increase in survey coverage in that territory, effective in the spring of 2009 and retroactive to the winter of 2008, brings it on par with the other two territories (96% in the Northwest Territories and 92% in both Yukon and Nunavut).

While persons living on reserves and other Aboriginal settlements are not included in the sample for the provinces, the northern sample includes both aboriginal and non-aboriginal communities.

In order to get a representative sample of the target population for each territory, the sample for three consecutive months are added. For this reason, estimates for the North are only available as 3 month moving averages.

The following guidelines were used in sample allocation for the territories:

  1. CVs of 25 percent or less for 3 month average estimates of unemployed. (See section 7 for explanation of sampling error and CVs (Coefficients of variation)).

Like the provinces, the northern sample design is also based on a rotating panel but at different intervals. Selected householders in the North are interviewed eight times, once every three months over a two-year period. For example, if a household was first selected for the month of January 2006, household members will be interviewed again every three months (for example, April, July, and October 2006, January, April, July and October 2007). After eight interviews, the household is replaced by another from the same community or from another community in the same stratum. Each quarter, one-eighth of the sampled households are experiencing their first interview.

The same Labour Force Survey questions are asked in the territories as in the provinces, with a few exceptions. The rent questions are not asked in the territories nor are there any supplements to LFS in the North.

Like the provinces, survey operations are conducted by Statistics Canada staff. The first contact is generally in person and most of the other interviews are done over the phone (where possible). Data collection and processing for the North is otherwise the same as for the provinces.

The community boundaries are usually based on 2006 Census geography. The communities included on the frame and eligible to be sampled are:

Yukon– The Census agglomeration of Whitehorse is always in sample; other communities will transition to updated groupings during 2011. April 2011 onward, Dawson and Watson Lake are always in sample; plus one community is selected from Carmacks, Mayo, Haines Junction or Teslin; and one community from Pelly Crossing, Ross River, Carcross or Faro. Watson Lake includes the small neighbouring villages of Upper Liard, Two Mile Village and Two and One-Half Mile Village.

Northwest Territories– Yellowknife, Norman Wells, Hay River and Inuvik are always in sample; one community from Fort Smith or Fort Simpson is selected; plus one community from Behchokò, Fort Liard, Fort Providence or Fort Resolution; one community from Tuktoyaktuk, Fort McPherson or Aklavik; one community from Fort Good Hope, Déline or Tulita; one community from Hay River Dene 1 or Detah; and one community from Wha Ti, Wekweti, Gamèti, Lutselk'e, Tsiigehtchic or Wrigley.

Nunavut– Iqaluit, Rankin Inlet, Cambridge Bay and Kugluktuk are always in sample. One community from each of the following groups are also selected: Baker Lake or Arviat; Igloolik or Pond Inlet; Cape Dorset or Pangnirtung; Taloyoak, Gjoa Haven or Kugaaruk; Coral Harbour or Repulse Bay; Qikiqtarjuaq, Arctic Bay, Hall Beach or Clyde River.

Next | Previous