Skip to main content

Making waves in his first speech after taking office in 2009, U.S. Attorney-General Eric Holder described his country as a "nation of cowards," afraid to confront racial issues. If fear of confronting race is cowardice, what does one call fear of even asking about it? Because that's where Canada is at the moment.

By largely avoiding collecting race-based statistics, Canada is failing to take into account how racial discrimination affects important social and economic outcomes – in the process starving researchers and policy makers of the information necessary to assess and address the issue. By omission, it is continuing to ignore the issue, even as the demographic reality dictates a pressing need to acknowledge it.

Last month, the U.S. Federal Reserve Board released results of its 2013 Survey of Consumer Finances (SCF). Piggybacking on the Fed release, the following week Canada's Broadbent Institute released custom tabulations using Statistics Canada's 2012 Survey of Financial Security (SFS). While Statscan released SFS results back in February, the Broadbent Institute thought it novel to publish the data by deciles, i.e. an equal 10-way split of families, as the Fed did with SCF. (Statscan's SFS release used quintiles, i.e. an equal five-way split.)

The Fed's SCF collects information about family incomes, net worth, balance sheet components, credit use and other financial outcomes, much like Statscan's SFS does. There's one glaring difference between the two surveys: The Fed's SCF inquires about race. Statscan's SFS does not.

The more complete SCF demographic profile enables the Fed, as well as U.S. academic and public policy researchers, to provide more timely and insightful analysis, including on racialized families' financial security.

Given the limited Canadian SFS demographic profile, neither Statscan nor the Broadbent Institute could provide any insight on racialized families. That the purportedly progressive Canadian think tank did not see fit to mention the glaring omission, especially given the timing and context of its release, is disappointing.

The U.S. Bureau of Labor Statistics' monthly labour force reports, based on the Census Bureau's Current Population Survey (CPS), likewise provides analysis of labour market outcomes by race. Statscan's monthly Labour Force Survey (LFS) does not.

Before being discontinued in 2012, Statscan's related (albeit voluntary) Survey of Labour and Income Dynamics (SLID) used to inquire about race. The limited public information available suggests the Canadian Income Survey replacing it does not. (To date, Statscan declined requests to disclose the 2012 CIS questionnaire contents.)

The long-form census, the only mandatory Canadian household survey that collected information on race, became a voluntary survey in 2011. As expected, the change resulted in a significantly less reliable 2011 National Household Survey (NHS). One recent paper found the voluntary 2011 NHS and SLID surveys showed different results for racial minority labour market outcomes. It's impossible to tell if either result is meaningful.

What's curious is that a number of Canadian surveys that exclude questions on race – such as the SFS and LFS – do include questions on immigration status.

A Statscan analyst advised that race had been among the variables considered for inclusion during past LFS redesigns. For rather convoluted reasons (fit, general interest, operational considerations, questionnaire length, cost, respondent confidence), Statscan decided to choose between either race or immigration status. (The U.S. and U.K. labour force surveys manage to include both.)

An alternative view is that poorer socio-economic outcomes for immigrants are easier to explain away as poorer assimilation rather than as racism. Including both race and immigration status would allow researchers to factor for assimilation effects, and focus on the impact of racial discrimination. One such research paper found that "for most visible minority groups, earnings gaps are identified among third- and higher-generation Canadians." That paper relied on the long-form census; it was published in August, 2010, about a month after the current federal government decided to cancel the long-form census.

About one in five Canadians identified as a racial minority in the 2006 census. Demographic projections published in 2010, largely based on the long-form census, suggested a rise to nearly one in three by 2031. While Canada's population projections were recently updated, its demographic projections have yet to be updated. If/when the 2011 NHS is used to update them, the projections will likewise become significantly less reliable.

The dearth of statistics on race is effectively a compound problem: It not only denies Canadians information necessary to assess the prevalence, persistence and effect of racial discrimination, it makes it difficult to even estimate the share of the population potentially affected.

Shortly before the Broadbent Institute release, a fellow by the name of Tim Uppal took to Twitter to recount a racist episode he and his wife had just experienced. For readers not familiar with the name, Mr. Uppal is Canada's recently appointed Minister of State for Multiculturalism. Hopefully he can sum up the courage Mr. Holder had when discussing the issue of race. Just asking why Canada avoids the question would be a good start.

Sam Boshra is an independent Montreal-based economist and editor for

Interact with The Globe