Methodological details regarding the National Household Survey and the Labour Force Survey
Baseline data on artists are used for multiple purposes, including as evidence for policy analysis, trend analysis, benchmarking in reporting on outcomes, analysis of particular issues such as regional or cultural equity, and the assessment of long-term impacts of arts funding. This report assesses the reliability and usefulness of potential data sources on the working lives of artists, including a close examination of the National Household Survey (NHS) and the Labour Force Survey (LFS).
In order to prepare this methodological report, Hill Strategies Research has:
- Reviewed published methodology relating to the National Household Survey labour force data.
- Consulted with a limited number of experts on the NHS.
- Assessed the likely impact on baseline reporting of data on artists.
- Assessed the ability to compare data from the 2011 NHS with data from the 2006 census.
- Identified other sources of data on artists/arts occupations to supplement the NHS.
The main findings of the analysis of methodologies and data quality are that:
- Neither the National Household Survey (NHS) nor the Labour Force Survey (LFS) are an ideal source of data on artists. However, both may provide some useful information.
- While the NHS is less reliable than the previous long-form census, there is still valuable information in the survey that can be used to examine the working lives of artists.
- Careful attention should be paid to the reliability of statistics presented from either the LFS or the NHS. Where possible, both sources should be used (and checked against each other) in order to enhance confidence in the data.
- The NHS has a much larger sample size but a much lower response rate than the LFS.
- There are large changes in many estimates from the 2011 NHS compared with the 2006 long-form census (e.g., specific arts occupations, most provinces, territories, and Census Metropolitan Areas). It is highly improbable that these changes are all “real” differences in the amounts. As such, the two sets of estimates should not generally be compared.
- There are differences between data on artists from the NHS and the LFS. However, in Canada and all 10 provinces, the NHS estimates of artists fall within the margin of error of the LFS data.
- The small sample size of the LFS leads to limited reliability of breakdowns of the number of artists. The LFS does not publish any amount below 1,500 in certain jurisdictions (and below 500 in smaller jurisdictions).
- The LFS provides the best estimate of trends in the overall number of artists in Canada. The annual averages from the LFS are also timelier than the five-year census or NHS.
- Data on artists in less populated areas and for smaller demographic breakdowns will not be available from either the LFS or the NHS.
- In the 2006 census, the minimum number for reliable estimates was 40 artists. A general rule for the NHS might be to examine estimates of 500 artists or more (possibly even 1,000). Below population levels of 1,000 people, and especially below 500, the NHS population estimate tends to diverge from the census population figure (which, by definition, will be more accurate). Transposing this analysis onto artists, we might be able to consider any jurisdiction with at least 500 to 1,000 artists as being reliable.
The report provides detailed comparisons of the methodologies of the National Household Survey and the Labour Force Survey, data comparisons between the 2011 NHS and the 2011 LFS, detailed comparisons of 2006 census and 2011 NHS data, and a comparison of data trends from historical LFS data and census / NHS data from 2001, 2006, and 2011. In addition, the report compares overall population estimates at the municipal level between the 2011 NHS and 2011 census. Based on these comparisons, an assessment of the potential impact on reports related to artists is provided. Before concluding, the report provides an overview of the general methodology of the 2011 NHS. Each section of the report begins with key findings and continues with details of the analysis.