If you look through the source data you will see that the descriptive race column uses the categories for the Canadian “Visible Minority” framework. I used this categorization because I am Canadian and it’s the one I’m familiar with, since it’s used in the Canadian Census and therefore most other Canadian research. It also happens to line up fairly closely with the categories used by the Institute for Composer Diversity in their database.
However, it’s important to note that it has drawbacks. Any framework that attempts to provide categories for race is going to have flaws, because race is a social and political invention. The Canadian categorization was created in the 1980s for the Employment Equity Act, and has mostly remained the same since (presumably for the purpose of long-term comparability) 1.
To give more context on how quickly these concepts change, the 1971 Canadian census only asked for a single ethnic ancestry on the paternal side. The 1986 census had one question that included categories such as French, English, Scottish, Irish, Jewish, Polish, Ukrainian, and Dutch alongside Chinese (no other Asian ethnicities listed), Black, and a handful of categories for people who were Indigenous. In the current census, Visible Minority (i.e. race), ethnicity, and Indigenous identities are all addresed with distinct questions.
In my notes, the source for the details above was Jain, Harish C. “Racial Minorities and Affirmative Action/Employment Equity Legislation in Canada.” Relations Industrielles / Industrial Relations, vol. 44, no. 3, 1989, pp. 593–614. JSTOR, www.jstor.org/stable/23073601. Accessed 4 Jan. 2021.
One particularly egregious feature of the Canadian framework is that it essentially has built-in assumptions about which mixed-race people with one white parent “pass” for white. This is pretty explicitly outlined in the 2016 Census reference guide:
In the visible minority variable, persons who reported ‘South Asian,’ ‘Chinese,’ ‘Black,’ ‘Filipino,’ ‘Southeast Asian,’ ‘Korean,’ or ‘Japanese,’ in combination with ‘White’ or a write-in response are included in the visible minority count for the specific visible minority group reported. For example, respondents who checked ‘Black’ and ‘White’ are included in the ‘Black’ count. Respondents who checked ‘Black’ and wrote in ‘French’ or ‘Malaysian’ are also included in the ‘Black’ count.
In contrast, in accordance with employment equity definitions, persons who reported ‘Latin American’ and ‘White,’ ‘Arab’ and ‘White,’ or ‘West Asian’ and ‘White’ have been excluded from the visible minority population. Likewise, persons who reported ‘Latin American,’ ‘Arab’ or ‘West Asian’ and who provided a European write-in response such as ‘French’ have been excluded from the visible minority population as well. These persons are included in the ‘Not a visible minority’ category.
There is at least one case among the artists I researched where this official distinction would make a difference. However, I chose to ignore this and counted all composers who described themselves as having one white or European parent and one parent who was a person of colour as composers of colour.
For at least one of the composers this distinction applied to, I was able to find a media interview specifically asking them about whether their cultural heritage influenced their music. So, in addition to being overly-broad and presumptuous, the Canadian government definitions don’t really matter — in the world of classical music, having one non-white parent of any ethnicity is likely going to define your experience in the industry in a way that is different than the experience of a composer with two white parents.