Concepts of race don’t fit into one or two neat little boxes for our survey collection convenience. Racial differences rarely have any ‘objective’ genetic basis; most people agree that race and ethnicity are mostly social concepts, not physical laws of who we are, how we live and what we’re like. Changes to institutions like the US Census to make room for multiracial self-reporting, as covered in this NPR article, reflect the reality – race is easy to simplify and hard to quantify.*
(*Not-so-fun fact: The Canadian Census dropped its long form mandatory section, which means we simply don’t ask. The National Household Survey does at least ask about race and visible minority and allows the respondent to “check all that apply.”)
Increasing awareness of ‘race’ as a subjective term means that people today are way less likely to identify as one race or ethnicity (explore this through The Race Card Project). This is especially true in countries with high immigration: Datassist conducted a recent survey of teenagers that listed many ethnic identities and asked them to check all that applied to them. More than half checked four or more boxes. Any measure attempting to record totally distinct races can end up acting like a blunt instrument applied to a subtle question – that is, it can go poorly.
Defining race is dynamic and interactive, and survey data needs to keep pace. The easy solution is to just pose the question and let respondents write anything…but people tend to skip open-ended questions. (They’re scary! And hard!) We have to get ahead of the curve by listing lots of possible races and ethnicities plus an inclusive “other” category, and letting people check all the boxes they want.
Most of all, we have to make sure the complexities of race and ethnic identification is recorded accurately and used well. We need to analyze race in careful, thoughtful ways that do justice to the careful, thoughtful ways people define and express their identity. Over the coming weeks, we’ll look at some practical ways to take people out of little boxes and help our data reflect a truer sense of the colours on display.