9. Data Challenges and Limitations
This chapter describes the various data challenges encountered in the production of this labour market information report and identifies the associated limitations. The first section discusses the specific data challenges faced in the production of this report. The second section discusses the associated limitations to the findings in this report. Where possible, the next chapter (Chapter 10: Recommendations) discusses possible solutions to remedy or minimize the challenges associated with the data.
9.1 Data Challenges
While Statistics Canada’s census provides the most detailed data with respect to cultural workers’ demographics and labour-related information (work activity, labour force status, employment income, etc.), the implementation of the census dictates that it can capture only a snapshot of the cultural labour force during the week in which the 2016 Census was conducted (Sunday, May 1, to Saturday, May 7, 2016). Indeed, given the high rate of seasonal workers in the cultural sector, there is a high likelihood that such a short survey period is insufficient to gain a full grasp of the cultural labour force. Specifically, a seasonal cultural worker’s involvement in the cultural sector would not be captured by the census if the worker was engaging in other non-cultural occupation during the survey week.
In addition, since the census probes only an individual’s primary occupation, two possible scenarios could arise and contribute to the underestimation of an individual’s involvement in the cultural sector. The first scenario is if a cultural worker spends more hours working in other sectors, as is common among individuals in the cultural sector. Because the census records only the occupation in which a worker spends the most time, this segment of cultural workers will be left out. The second scenario centres on the fact that it is common for a cultural worker to take on multiple cultural roles/occupations. While the census evaluates and assigns a person’s occupation according to both job title and work activity, the chance of mis-assigning that occupation would increase if a cultural worker were involved in a series of heterogeneous work activities.
To estimate the degree to which an individual’s involvement in the cultural sector was underestimated, the Conference Board’s labour market information survey of individuals in the cultural sector asked respondents to select multiple occupations that most accurately reflected their involvement in culture. As presented in Table 9.1, every cultural worker holds on average two occupations, suggesting the likelihood of a person’s involvement in culture not being fully captured by the census could run high. In fact, in the sound recording domain, each individual has 3.2 occupations on average.
Table 9.1.1: Average Number of Occupations Reported, by Domain
|
Domain |
Average number of occupations reported |
|
Heritage and libraries |
1.5 |
|
Live performance |
1.9 |
|
Visual and applied arts |
2.5 |
|
Written and published works |
2.0 |
|
Audio-visual and interactive media |
1.5 |
|
Sound recording |
3.2 |
|
Total |
2.0 |
Source: The Conference Board of Canada, Cultural Sector Labour Market Information Survey (Individual), 2018–19.
The survey of individuals in the cultural sector also probed respondents’ working hours outside the cultural sector, as well as their nature of employment (whether they had worked on a full-time, part-time, or seasonal basis over the past year). The results suggest that, in aggregate, nearly 8 per cent of respondents worked more hours outside the cultural sector and thus would likely not be captured by the census. Meanwhile, close to 16 per cent of respondents working in the cultural sector worked on a seasonal basis. While the chance of these individuals not working in culture during the census week varied by occupation and by domain, it certainly added a layer of downward bias to the estimation of cultural labour force.
A more mixed picture was found at the domain level. Only a small share of respondents in heritage and libraries domain indicated they spent more hours working outside culture. But in the audio-visual and interactive media domain, 20 per cent of camera operators, scriptwriters, and actors said they did. A similar pattern was found in live performance and sound recording, with circus artists and music arrangers having the highest likelihood of working more hours outside culture. A higher percentage of individuals working outside culture tends to equate to a higher likelihood of underestimating the size of cultural labour force.
Table 9.1.2: Likelihood of Underreporting—Heritage and Libraries
|
Occupation |
Working more hours outside the cultural sector |
Seasonal worker |
|
Archivist |
3% |
0% |
|
Exhibitions designer or planner |
3% |
3% |
|
Conservator/curator |
2% |
2% |
|
Library worker |
1% |
3% |
|
Librarian (master’s degree or equivalent) |
1% |
1% |
|
Registrar |
0% |
7% |
|
Library and public archive technician |
0% |
4% |
Source: The Conference Board of Canada, Cultural Sector Labour Market Information Survey (Individual), 2018–19.
Note: occupations in which workers did not work more hours outside the cultural sector or did not work on a seasonal basis are not shown.
Table 9.1.3: Likelihood of Underreporting—Live Performance
|
Occupation |
Working more hours outside the cultural sector |
Seasonal worker |
|
Circus artist |
25% |
0% |
|
Choreographer |
11% |
17% |
|
Actor |
10% |
33% |
|
Playwright |
10% |
19% |
|
Comedian |
9% |
18% |
|
Musician/singer |
8% |
27% |
|
Wardrobe worker |
8% |
54% |
|
Dancer |
5% |
18% |
|
Stagehand |
4% |
37% |
|
Presenter |
3% |
37% |
|
Manager (administration, HR, communications, marketing, theatre, production, etc.) |
2% |
12% |
|
Technician (automation, lighting, rigging, sound, etc.) |
1% |
12% |
|
Props master or worker |
0% |
31% |
|
Conductor/composer/arranger |
0% |
50% |
Source: The Conference Board of Canada, Cultural Sector Labour Market Information Survey (Individual), 2018–19.
Note: occupations in which workers did not work more hours outside the cultural sector or did not work on a seasonal basis are not shown.
Table 9.1.4: Likelihood of Underreporting—Visual and Applied Arts
|
Occupation |
Working more hours outside the cultural sector |
Seasonal worker |
|
Metalsmith |
14% |
19% |
|
Painter |
9% |
14% |
|
Illustrator |
9% |
17% |
|
Sculptor |
6% |
15% |
|
Artisans/craftsperson/patternmaker |
6% |
13% |
|
Visual Artist |
6% |
12% |
|
Manager (art gallery management, administration, HR, communications, marketing, etc.) |
5% |
5% |
|
Designer |
4% |
15% |
|
Graphic designer |
3% |
16% |
|
Theatre, fashion, exhibit and another creative designer |
0% |
16% |
|
Photographer |
0% |
25% |
|
Art dealer |
0% |
18% |
|
Technician (graphic arts, etc.) |
0% |
13% |
|
Printmaker |
0% |
11% |
|
Curator |
0% |
10% |
|
Potter |
0% |
10% |
Source: The Conference Board of Canada, Cultural Sector Labour Market Information Survey (Individual), 2018–19.
Note: occupations in which workers did not work more hours outside the cultural sector or did not work on a seasonal basis are not shown
Table 9.1.5: Likelihood of Underreporting—Written and Published Works
|
Occupation |
Working more hours outside the cultural sector |
Seasonal worker |
|
Journalist |
13% |
25% |
|
Author/writer |
12% |
15% |
|
Operations/logistics |
8% |
0% |
|
Production/design |
5% |
5% |
|
Publisher/editor |
5% |
5% |
|
Manager (administration, HR, communications, etc.) |
5% |
0% |
|
Marketing/advertising associate |
4% |
4% |
Source: The Conference Board of Canada, Cultural Sector Labour Market Information Survey (Individual), 2018—19.
Note: occupations in which workers did not work more hours outside the cultural sector or did not work on a seasonal basis are not shown
Table 9.1.6: Likelihood of Underreporting—Audio-Visual and Interactive Media
|
Occupation |
Working more hours outside the cultural sector |
Seasonal worker |
|
Camera operator |
20% |
0% |
|
Scriptwriter |
20% |
27% |
|
Actor |
20% |
35% |
|
Producer |
15% |
19% |
|
Announcer/ another broadcaster |
14% |
14% |
|
Director |
13% |
22% |
|
Editor |
7% |
18% |
|
Technician (lighting, sound, broadcast, audio and video recording, etc.) |
4% |
18% |
|
Manager (studio, HR, communications, administration, marketing, production, etc.) |
4% |
15% |
|
Engineer (sound, etc.) |
0% |
33% |
Source: The Conference Board of Canada, Cultural Sector Labour Market Information Survey (Individual), 2018–19.
Note: occupations in which workers did not work more hours outside the cultural sector or did not work on a seasonal basis are not shown.
Table 9.1.7: Likelihood of Underreporting—Sound Recording
|
Occupation |
Working more hours outside the cultural sector |
Seasonal worker |
|
Conductor/music director |
100%* |
0% |
|
Music arranger |
29% |
14% |
|
Manager (music artist management, administration, HR, communications, marketing, etc.) |
20% |
11% |
|
Musician |
18% |
10% |
|
Singer (opera, pop, jazz, etc.) |
14% |
14% |
|
Composer |
13% |
0% |
|
Songwriter |
8% |
13% |
|
Music publisher |
0% |
15% |
|
Technician (instrument, recording, sound effects, etc.) |
0% |
67% |
|
Producer |
0% |
21% |
Source: The Conference Board of Canada, Cultural Sector Labour Market Information Survey (Individual), 2018–19.
Note: occupations in which workers did not work more hours outside the cultural sector or did not work on a seasonal basis are not shown.
* Although 100 per cent of conductor/music directors indicated more hours working outside the cultural sector, the percentage was not representative, as only one conductor/music director responded to the labour market information individual survey.
9.2 Limitations
While this report presents The Conference Board of Canada’s best effort to provide a comprehensive profile of the cultural sector, readers should be aware of limitations.
The first set of limitation stem from the use and reliance of census data. In particular, this study relies heavily on Statistics Canada’s census to inform readers on cultural workers’ demographics, employment status, income, and many other aspects associated with working in the sector. Unfortunately, the design of the census questionnaire and its implementation in the field needs to strike a careful balance between the level of analysis possible and the level of effort required from respondents to complete the survey. As a consequence, much of the information about the distinct employment arrangement of many cultural workers—including the high incidence of holding multiple jobs and working on a seasonal or as-needed (gig) basis—is not captured in the census.
Another limitation of the census data is that it fails to accurately account for the important non-paid time spent by many cultural workers on practice and preparation. This missing component likely not only contributes to an underestimation of the size of the cultural workforce but also leads to a potential misclassification of individuals’ work activity in the cultural sector. In addition, many occupation titles in the current National Occupational Classification system are outdated, making it challenging to accurately measure involvement in the cultural sector.
The second limitation stems from the use of Statistics Canada’s Business Register data to compile the employer profile. While the Business Register provides a detailed snapshot of all Canadian establishments, a business needs to meet one of the three criteria (has employees, is incorporated, is a GST/HST registrant) to be included in the Business Register. Small cultural establishments, particularly sole proprietorships, often do not meet any of these criteria and thus are not captured by the Business Register.
Another layer of complication is attributed to the diverse business activities in which many cultural establishments engage. Indeed, large cultural enterprises often tap into multiple lines of culture-related businesses. Many internet streaming services, for instance, both produce and disseminate video content. Assigning a multi-faceted cultural business to a specific industry, therefore, becomes difficult under the current industry classification system and is up to Statistics Canada’s discretion. As a result, the employer profile in this study may be skewed due to particular singular domain classifications.
A third limitation results from the lack of official data about some domains and business activities. Specifically, statistics related to the heritage and libraries domain in the current Culture Satellite Account (and thus in this report) represent only private institutions, a small portion of all heritage and library institutions. The exclusion of public institutions by Statistics Canada is due to the inherent difficulties of collecting and segregating public expenditure data into the appropriate heritage and libraries subdomains.
Moreover, the rise of digitalization within the cultural sector poses significant challenges to measuring the digital presence of cultural activities and its contribution to the Canadian economy. As Statistics Canada is still in an early stage of estimating the digital economic activity, a considerable portion of the digital goods and services produced in Canada may be missing in the findings of this report. Another by-product of digitalization is that many cultural products consumed by Canadians today are provided from outside Canada, most notably, by foreign audio and video streaming services. The jobs these companies create in Canada and the revenues they generate from Canada are difficult to measure given their foreign ownership. As a result, it is likely that much of this activity is not included in this report.
Lastly, despite the best efforts by the Conference Board and its partners to disseminate the cultural sector labour market information surveys, only a relatively small number of individuals (n = 34 out of 1,867) and a small number of employers (n = 4 out of 311) reported being involved in the sound recording domain. As a result, the findings from this domain should be used with caution.