Race Questions and the Census

The CivicAccess list from time to time has some really good discussions.  As of late there have been a couple of great ones.  They are honest and often bring up issues that require some external expertise.

The question of race questions in the former long-form Canadian census was the latest.  It was inspired by comments associated with the NYTimes map – Mapping America: Every City, Every Block posted to the list.  The list threads on this topic can be read here.

Debra Thompson was invited to read the threads and to respond.  Debra wrote The Politics of the Census: Lessons from Abroad in the Journal of Canadian Public Policy as a response to the recent cuts to the Canadian census and won the McMenemy Prize for her paper in the Canadian Journal of Political Science: Is Race Political?

Here is Debra’s Response:

That type of argument (race is dangerous, we shouldn’t be asking a question on it anyway) is actually pretty common – and came up back in 1996 when the question was  first put on the census. Unsurprisingly, it’s more often the white majority that claims race is dangerous, rather than racial minorities who largely understand that race is socially constructed, but carries consequences nonetheless. The basic fact of the matter is that we have a range of policies that depend on accurate census data. Yes, employment equity is one of those policies. Yes, it has its problems – especially in that it can’t account for variation in discrimination within the population we call “visible minority”. Some visible minorities are clearly discriminated against in a variety of socio-economic indicators – housing, employment, services, etc. Most often, these are Aboriginal peoples, Black Canadians and some Asian population groups. Other VM groups don’t necessarily need the policy in order to ensure their labour force representation is equitable. But can you imagine a Black-only or Aboriginal-only employment equity policy? It just wouldn’t fly.

The debates over whether or not a question *should* be in the census is more often than not a debate about the efficiency and equity of affirmative action-type policies. In my opinion, these debates are very important, but should take place elsewhere. I personally think employment equity is a good idea. It means the state has a positive obligation to promote racial equality. We know that the marketplace won’t do this on its own. It also sends important signals about citizenship and social justice as important priorities for the Canadian state. I would also tell critics that our employment equity policy is actually very very weak. VERY weak. It has little by way of actual monitoring, the courts have rarely backed it up, and it doesn’t compel the private sector at all. If our policy was stronger, we would have seen more VMs in the public service by now. Yet, if you look at the data, women have almost achieved representative parity, and VMs are still very much underrepresented – not nearly as badly as persons with disabilities, but still.

No matter the pros and cons of this legislation, it’s the law of the land. And we can’t make this law work properly without accurate data. In the 1980s before there was a “race question”, StatsCan used the ethnic origin question and other proxies to determine which respondents were VMs. But it was highly problematic – think about my father’s family, for example. We came to Canada in the 1860s, via the Underground Railroad. We’ve been here for a very long time. (This is why I find it so frustrating when white Canadians ask me, “no, where are you REALLY from?” I’m from HERE.) In response to the ethnic origin question, what would Dad write? “Black” is a racial group, not an ethnic group. But to put “Canadian,” “American,” or “British,” as Dad might have done, wouldn’t capture the fact that he’s black. StatsCan also had the same types of problems with other groups – Jamaicans and Indians (from India) who put their ethnic origins as “British,” Haitians who put theirs as “French”. If you want to measure race, you need a question on race.

Canada is also not alone in this regard. Over 60% of countries in the world have a question on nationality/ethnicity/race, though they use diverse conceptualizations of what these terms mean. It’s also been proven time and time again by places like the United States and Great Britain that having a question on race in the census is actually helpful if the society’s ultimate goal is racial equality. Canada has had this question on its census since 1996. And we’re not more divided than before, race riots haven’t broken out (though there are some places in Canada where racial minorities are living so far below the poverty line that I wouldn’t be surprised if they did), and our kitten-hugging version of multiculturalism – high on rhetoric, low on actual results in terms of lessing racial disadvantage – is still intact. So, you see, having a question on race isn’t about making Canadian society more divided. It’s about making it more equal. I think that’s a pretty important goal.