Arts & culture data sources: gaps, specificity to culture, and coverage of culture
Have the imperfect, more generalized statistics (e.g., "arts, heritage, and entertainment") met a need for you?
Last week, I offered the briefest of summaries of the 59 posts in year 1 of this new format for Statistical insights on the arts (SIA).
Today, I offer my thoughts on data gaps, plus the specificity and coverage of the data sources over the past year. I decided to have some fun and offer letter grades for each source’s specificity and coverage. Enjoy!
June is “free month”. Thanks to my paid subscribers, all four posts this month will be free. If you’d like to join them — and support vital research into the arts, you can do so here.
Challenges and improvements
Things could certainly be better when it comes to data on the arts. I’ve lived this for over 25 years, including my time in the sector before founding Hill Strategies Research.
I certainly wish that we could have access to a much broader array of statistics on the arts in Canada. Most weeks, I get an email request for information on a topic for which we do not have data.
Here are my thoughts on some gaps:
Measurements of government spending on culture, even at the national level.
Limited statistics on the sector’s environmental impacts, both positive and negative (but this is expanding).
Only occasional reports on arts markets (e.g., who buys visual art?). So far, my SIA posts have tended to cover the supply side of culture much more than the demand side.
My equity related posts (so far) have mainly examined gender equity. This is an important topic, no doubt, but obviously not the only equity related consideration.
You’ll start to see a broader range of equity posts now that I have received custom census data. (Yay! Only 6 months and 1 day — and $3,400 — after I submitted my request.)
However, census data can only tell us about occupational information, not, for example, inequities in arts funding, the opportunities available to different groups of artists, the challenges faced by different groups of cultural workers, or how different cultural organizations are doing in a sort-of-post-COVID environnent.
Once we drill down below the national level, reliable datasets are relatively sparse. This challenge is particularly acute at the local / regional level. Territorial data are incredibly rare.
One of my biggest disappointments this year is that I have not been able to offer my posts in French. This has just not been financially feasible so far. But I’m working on that.
What are your thoughts on culture data gaps? Use the button below to post a comment on the web version of this newsletter. Or you can reply to the email version if you'd prefer to keep the conversation private.
Over my 21 years (plus 5 before that at the Ontario Arts Council), it has always been a challenge to get culture-specific data. I’ve tried to mine some new sources as part of this newsletter. Here are my thoughts on the sources’ specificity to culture and coverage of the sector.
How specific to culture are the datasets analyzed to date? How well do they cover the arts and culture?
To date, I’ve kept a close watch for data sources that might have an arts component, with a particular interest in ones that I’ve never analyzed before. Here are the main sources to date:
Census 2021
Labour Force Survey
Job Vacancy and Wage Survey
National Culture and Sport Indicators (and provincial/territorial indicators)
Canadian Survey on Business Conditions
Canada's Core Public Infrastructure Survey
Rural Canada Non-Profits Database
Other occasional surveys, such as the Arts and Heritage Access and Availability Survey and the Canadian Artists and Content Creators Economic Survey
What culture-specific data are available in each of them? Do they cover the sector well, or are there gaps? Do they provide insights specifically into the arts (as per the name of this newsletter)?
Here are my letter grades on those. (A=excellent, B=very good, C=adequate, D=less than adequate, F=seriously lacking.)
I have not analyzed any datasets with less than a C grade, because who wants less than adequate data?
As always, your feedback is welcome.
Because this is a free post, sharing is encouraged.
Census 2021
Specificity to culture: A
Coverage of culture: B
The census is maybe the second best source of arts specific data that we have. Which makes me sigh.
It provides very relevant information on artists and cultural workers. It covers the range of cultural occupations sort-of well.
It obviously doesn’t cover the myriad other issues in the arts sector (e.g., funding, opportunities, challenges).
The difficulties with the census are less about its specificity to and coverage of culture, and more about its timeliness and the summary nature of the occupation groups.
I have already written a lot about these issues. If you’re interested, here are direct links to two recent posts.
Labour Force Survey
Specificity to culture: B
Coverage of culture: C
I think that I have analyzed the Labour Force Survey more in this year than in the rest of my career combined. It provides some useful insights, especially at a time when the other main option — census earnings data — relates to the pandemic lockdown year of 2020, when artists had very few paid gigs.
The LFS’s coverage is nearly the same as the census (both use the same occupation categories), but LFS data are only reliable at a more abstracted level than the census. And LFS earnings data exclude the self-employed. Hence my grade of C, rather than the B for the census.
The summary categories that are readily available from the LFS are an awkward fit for artists and cultural workers. I wrote about them here:
National Culture and Sport Indicators (plus similar provincial and territorial indicators)
Specificity to culture: A+
Coverage of culture: A+
You can’t get more culture specific than this. Plus this dataset has detailed breakdowns by culture domain — roughly equivalent to disciplines in the arts, culture, and heritage.
By definition, this dataset covers the whole cultural sector — it uses as its source the Framework for Culture Statistics.
Probably the biggest issue here is the timeliness of the provincial and territorial data.
Like the census, the indicators cover a relatively narrow range of data: jobs and direct economic impacts. No demographic breakdowns of jobs / workers are possible.
The rest of the data sources are ones that I’d never really analyzed before June of 2022.
Job Vacancy and Wage Survey
Specificity to culture: C+
Coverage of culture: C
I believe that some very useful insights have come out of this dataset, but its data only cover payroll employees, not the self-employed.
Like many data sources, the statistics get less and less reliable the more one drills down, and one needs to drill down to access reasonably culture-specific information. This becomes a balancing act, with data reliability on one side and culture specificity on the other.
As I’ve noted in many posts this year, the broad category that I call “arts, heritage, and entertainment” and that Statistics Canada refers to as “arts, entertainment, and recreation” includes performing arts, spectator sports, and related industries (code 711), heritage institutions (code 712), as well as amusement, gambling, and recreation industries (code 713).
In the job vacancy dataset, I have focused solely on the first two categories, ignoring the amusement, gambling, and recreation statistics. However, the performing arts statistics include spectator sports, which is obviously less than ideal for arts research.
Canadian Survey on Business Conditions
Specificity to culture: C
Coverage of culture: C
This series of surveys have provided many insights that were not possible before its inception in 2020, which is why I’ve regularly analyzed its data.
Examples include:
Key obstacles (e.g., inflation)
In general, I’d say that this series of surveys is about the least culture-specific that I’d analyze. The dataset provides no statistics below the summary level of “arts, heritage, and entertainment”. It is therefore not feasible to exclude non-cultural elements, in particular the grouping of “amusement, gambling, and recreation industries”.
This poses a significant problem: there are many more businesses in amusement, gambling, and recreation industries than there are businesses and not-for-profit organizations in the performing arts and heritage.
On a separate note, the cultural industries (typically for-profit businesses) are included in a different category — one that also has many non-cultural businesses.
I should also note that the sample for this survey is restricted to businesses and organizations with at least one employee, which leaves out a number of arts organizations.
Please let me know: Have the more generalized statistics (on the “arts, heritage, and entertainment") met a need for you? Do you trust these numbers? Do you approach them with some caution? (I try to.)
Canada's Core Public Infrastructure Survey
Specificity to culture: A
Coverage of culture: B
This survey covers arts and culture spaces that are owned by municipal, provincial, and federal governments, but it excludes spaces owned by First Nations and other Indigenous governments.
It provides culture-specific data related to libraries, museums and archives, performance / presentation spaces, art galleries, and Indigenous culture facilities (just not ones owned by Indigenous governments).
Based on this dataset, I offered a Canada-wide analysis and a separate one for the provinces and territories. (Yes, data on the territories!)
Rural Canada Non-Profits Database
Specificity to culture: C+
Coverage of culture: C
Not-for-profit organizations in Statistics Canada’s Business Register are included in this dataset, whether or not they have charitable status. The only condition is that they must have reported some revenues, at least one employee, or both. So that seems like pretty good coverage.
However, the dataset provides no statistics below the summary level of “arts, heritage, and entertainment”. Like the Business Conditions survey, it is therefore not possible to exclude amusement, gambling, and recreation industries.
That is a bit less of a problem for this dataset, given that there are not as many not-for-profit organizations as for-profit businesses in the amusement, gambling, and recreation industries.
Again, I provided a Canada-wide analysis and a separate one for the provinces.
Other occasional surveys, such as the Arts and Heritage Access and Availability Survey and the Canadian Artists and Content Creators Economic Survey
Other culture-specific surveys have popped up from time to time, and I have tried to provide value-added analysis of certain elements of them.
Specificity to culture: A+
Coverage of culture: A
There are other arts-specific surveys that I have not analyzed, simply because I didn’t believe that I could offer any value added insights.
Phew. That was long. That’s all for today.