Discrimination and Multiculturalism in Canada: Exceptional or Incoherent Public Attitudes?

ABSTRACT

In this article, I argue that satisfaction with multiculturalism and support for discrimination in the immigration system are conceptually linked but distinct in the Canadian public’s mind. Following a large literature, I make the case that despite a normative assumption of nondiscrimination in the intellectual framework and policy rhetoric of multiculturalism, the public can support discrimination while also supporting multiculturalism. To support this, I present the results of a 2017 survey of Canadians. I show that slightly less than half of respondents are willing to explicitly support discrimination. Next, I show that, when faced with a more complex decision that offers the chance to discriminate, many do so. Finally, I compare experimental results to a nearly identical experiment in the United States, which reveals that Canadian and American respondents discriminate at similar rates.

KEYWORDS:

• Canadian exceptionalism
• immigration
• multiculturalism
• discrimination
• public opinion

Acknowledgments

I wish to thank Andrew Potter and the McGill Institute for the Study of Canada for funding the data collection. For helpful comments I wish to thank Keith Banting, Randy Besco, Eve Bourgeois, Sophie Borwein, Nicholas Fraser, Joe Heath, Peter Loewen, Erin Tolley, and Phil Triadafilopoulos. Ben Allen-Stevens provided able research assistance.

Disclosure Statement

No potential conflict of interest was reported by the author.

Notes

1. This is derived from a five-point strongly oppose—strongly support scale.

2. There are, of course, anti-immigration voices within each party, and particularly within the Conservative Party. These voices, though, have not set the agenda and do not control policy or campaign messaging.

3. This timing is important for a number of reasons, most notably because it predates the resurgence of the Bloc Québécois in the polls, and so all inferences about Bloc supporters below rely on a relatively small sample.

4. The sample was purchased from Ipsos Observer and consists of 1,522 respondents. The survey was hosted on Qualtrics. The survey included a variety of questions about immigration and identity. As is the case with many online surveys, this was an opt-in panel in which respondents participated in order to earn a variety of rewards. More details, including descriptive statistics, are available in Donnelly (2017).

5. It cannot be separated in a salience or priming sense, because this entire survey focused on immigration questions, but it does not ask whether the immigration piece of multiculturalism is working.

6. That this was a preference question was emphasized in order to distinguish this question from a previous knowledge question about government policy.

7. The order of the characteristics was randomized.

8. For Canada, the countries were China, Germany, India, the Philippines, Poland, Sudan, and Syria. For the US, the countries were China, France, Germany, India, Iraq, Mexico, the Philippines, Poland, Sudan, and Syria.

9. The wording of their question was “Now I’d like to ask you a few questions about President Trump’s executive order on immigration. Do you support or oppose suspending all travel by citizens of Iraq, Syria, Iran, Libya, Somalia, Sudan and Yemen to the U.S. for 90 days?”

10. Partisanship here is measured using the question “If a national election were held today, which party would you be most likely to vote for?”

11. As noted above, the number of respondents supporting both the Bloc and the Green Party is relatively small, so the confidence intervals on these estimates are correspondingly wide.

12. Whether this is a glass-half-empty-or-half-full story requires some comparison by which to judge party unity. Returning to the previously mentioned Quinnipiac (2017) poll, we can see that 88% of Republicans supported Trump’s suspension of immigration from Muslim countries, while 88% of Democrats opposed it.

13. In other words, “Ontario” in the figures can be read as “Non-GTA Ontario.” Note also that the “West” includes one respondent from Yukon.

14. While it is true that the religions among both Sudanese and Syrians arriving in Canada are mixed, it is probably the case that most respondents assume that these profiles are nonwhite. It seems likely that people are aware that Syrians are largely Muslim, though knowledge of the typical religion of Sudanese migrants is probably lower. The results for Sudan, then, may be impacted more by a preference for racial than religious discrimination.

15. Also note, as mentioned above, that the sample for Bloc respondents is quite small, and given the power constraints of the conjoint design, the standard errors are unsurprisingly large on their estimates.

16. Because the list of countries included in the two experiments differ slightly, doing this requires subsetting both data sets to only those pairs of profiles that contain two immigrants from countries that are included in both lists. This reduces the statistical power of the analysis, but, under reasonable assumptions, does not alter the validity of the design.

17. It should be noted that there is some evidence against this. Hopkins (2015) shows in a video experiment that the skin tone of Hispanic migrants does not affect American attitudes toward immigration. Creighton and Jamal (2015) show that while there is a big difference in the share of Americans who express explicit opposition to the naturalization of a Muslim versus a Christian migrant, there is a much smaller difference when opposition was elicited in a list experiment.

18. Of course, it is not possible from the data in hand to be confident that this correlation is causal or that causation runs from policy to popular attitudes. Instead, it could be that less discriminatory respondents are more open to multicultural policy.

19. The taboo against anti-Muslim discrimination is probably much weaker, but it is hard to argue against the existence of a strong and widespread norm against expressing support for racial discrimination.

20. The PPC’s 2019 manifesto called for ending multiculturalism, slashing annual levels of immigration by more than half, and abolishing the family reunification policy for parents and grandparents (PPC 2019).

Additional information

Funding

The author thanks the McGill Institute for the Study of Canada, which funded the data collection as part of its 2017 Annual Conference.

Notes on contributors

Michael J. Donnelly

Michael J. Donnelly is an Assistant Professor of Political Science and Public Policy at the University of Toronto. His teaching and research focuses on comparative politics, with an emphasis on the ways in which social diversity and public policy are related.