Social capital and voting participation of immigrants and minorities in Canada

Abstract

Using the social capital literature as a base, we explore the impact of interaction with others on voter participation with particular emphasis on exploring the differences between Canadian-born majority and minority residents.

We use the 2002 wave of the Equality Security Community survey to explore the relationship between voting and personal characteristics, work characteristics, social capital attributes and ethnic characteristics. We find that the odds of voting are largely a product of socio-demographic and social capital attributes. The impact of immigration and ethnicity is largely overridden. This suggests that it is not the minority attribute that impacts voting. Rather it is age, level of schooling and level of civic engagement which affects the probability of voting, both federal and provincial.

Keywords:

• Electoral behaviour
• ethnic minorities
• immigrants
• social capital
• political participation
• voting participation

Acknowledgements

We wish to thank Uttara Chauhan, Patrick Fafard, John Solomos and two anonymous reviewers for their comments and suggestions.

Notes

1. While some countries such as the Netherlands, Denmark and Sweden allow legally resident non-citizens to vote in local and regional elections this right is restricted to citizens in Canada.

2. Voter turnout for Canadian elections can be found at the following website: www.sfu.ca/~aheard/elections/historical-turnout.html

3. Detailed information on the survey can be found at the Institute for Social Research website at York University: www.yorku.ca/isr/download/esc.html

4. There are about fifty respondents claiming Aboriginal origins on the dataset. These respondents were also dropped because there are too few people to get any significant results. Also, to do justice to an analysis of Aboriginal voting propensities we would have to interact Aboriginal status with the education variables which would further complicate the model.

5. We note that our trust variable is different from that used by Putnam (2007) who uses a more general ‘trust in others’ variable. However we argue that for the purposes of a voter participation study, trust in government is a stronger indicator.

6. The question used to derive ethnic origin asks about the origins of the respondent's parents and grandparents. Respondents can provide up to four answers. As can be seen in Appendix Table A, the largest group is British followed by French and then Canadian. For the purposes of our research, we roll British, French and Canadian to create the majority category.

7. Wave II of the ESC does not include a weighting variable; however we have added information about the sample design (province, and the oversample information for Montreal, Toronto and Vancouver) to the model using Stata's survey set criteria.