Introduction
Immigrant integration hinges on newcomers actively engaging with their host. This involvement includes expressing opinions, understanding societal dynamics, and participating in decisions that shape the future of their new nation.1 Identifying and understanding specific social behaviour indicators and outcomes is for defining immigrant integration through the lens of civic engagement.
Integration is a dynamic, interactive process often viewed as a “two-way street” between immigrants and non-immigrants. Achieving successful integration requires that there be relative parity between immigrants and non-immigrants in several key areas. The Canadian Index for Measuring Integration (CIMI)2 is a comprehensive evaluation tool focusing on four dimensions: Economic, Social, Civic & Democratic Participation, and Health. By accounting for sociodemographic differences, the CIMI ensures fair comparisons between immigrants and Canadian-born individuals across geographies and over time.
Recently, discussions around immigrant integration emphasize the importance of “inclusion.” Social inclusion spans multiple dimensions, enabling groups and individuals to fully participate in economic, social, cultural, and political systems.34 The CIMI defines inclusion as equal access to opportunities and resources necessary for full participation in all aspects of life. In our study, we explore voting in elections as an indicator of citizens’ access to political systems and their exercise of civic rights. We also explore how indicators related to social inclusivity, such as having a network of close friends and relatives, participating in groups and organizations, etc., affect these outcomes.
Background
An analysis of historical voting trends in Canadian federal elections shows that that naturalized citizens (immigrants) generally have lower turnout rates compared to citizens by birth.5 New Canadians defined as recent immigrants who arrived within the past decade, differ from established immigrants who have lived in Canada for over 10 years. In the last three federal elections (2011, 2015 and 2019), established immigrants had a similar voter turnout rate to Canadian born citizens. However, voting trends among recent immigrants has been relatively low – almost 15% lower than both established immigrants and citizens by birth. Notably, voter turnout among recent immigrants has gradually increased over the past two election cycles in 2015 and 2019 (Figure 1).6
Compared to other indicators of immigrant integration, the increase in voter turnout among immigrants is notably slower. For instance, within less than a decade of arriving in Canada, immigrants’ sense of belonging,7 interpersonal trust,8 and attachment to the Canadian political community9 reach levels similar to those of native-born citizens. Additionally, Canadian immigrants adopt citizenship at a high rate – the 2021 Canadian census reports that 58% acquire citizenship within 6–10 years, and 79% become naturalized citizens within 10–15 years.10
Several factors influence citizens’ active participation in the electoral process. Evidence from various countries suggests that the longer immigrants reside in their settlement country, the more likely they are to vote.111213 However, White14 also argues that the timing of immigrants’ arrival, rather than their length of residence, is a critical determinant of their electoral participation in Canada. Regardless, both immigrants and non-immigrants exhibit differences in voter turnout based on demographics, sociocultural factors, and political characteristics. Voting is a social act influenced by our interactions with others, especially within our immediate social network (friends and family). Political interest and knowledge play essential roles in civic participation and can be nurtured by social surroundings from early life to adulthood.15
Grouped bar chart. The vertical axis runs from 0 to 100 in steps of 20. The horizontal axis shows three groups: 2011 Federal Election, 2015 Federal Election and 2019 Federal Election, each with three bars: Citizen by naturalization – Recent (10 years or less since immigration); Citizen by naturalization – Established (more than 10 years since immigration); and Citizen by birth. Printed values: 2011 Federal Election, recent 56%, established 71%, citizen by birth 70%; 2015 Federal Election, recent 70%, established 76%, citizen by birth 78%; 2019 Federal Election, recent 72%, established 75%, citizen by birth 78%.
Voir la page 34 du PDF (s’ouvre dans un nouvel onglet)
Methodology
This study examines the impact of social inclusion on civic engagement, using several indicators that are included in the CIMI. Key indicators include the size and depth of an individual’s social network, their involvement in community (through group memberships, volunteering), and experiences of discrimination., These elements of social inclusion, this study will examine two principal research questions:
- Does social inclusion affect civic engagement and participation of citizens?
- How does civic engagement differ between Canadian born and naturalized or immigrant citizens, especially when social inclusion factors are considered?
To address these questions, we utilize data from the General Social Survey (GSS) a periodic, cross-sectional survey by Statistics Canada with each cycle covering a specific topic in-depth. The latest GSS Social Identity (cycle 35) in 2020 provides an overall picture of Canadians’ identification, attachment, sense of belonging and pride in their social and cultural environment. Two of the key components included in the survey were social networks and civic participation and engagement, allowing us to examine several relevant indicators in our modelling study.
We used binary logistic regression models to predict the binary outcome (yes or no) of ‘Voted in the last federal election’ (Y). The first model (as shown in Figure 2) looked at immigrant type (established and recent immigrants) with Canadian-born citizens as the reference category and socio-demographic variables (gender, age groups, visible minority status, income and education levels) used as controls.
Diagram of three model equations, one per row. Model 1: Y = Immigrant type + Sociodemographic Controls. Model 2: Y = Immigrant type + Sociodemographic Controls + Social Inclusion Variables. Model 3: Y = Immigrant type + Sociodemographic Controls + Social Inclusion Variables + Civic Inclination Variables.
Voir la page 35 du PDF (s’ouvre dans un nouvel onglet)
In the second model, we introduced the following indicators of social inclusivity as predictor variables to those from Model 1:
- Number of close friends
- Number of close relatives
- Membership in a group or organization
- Experienced discrimination in the past 5 years*
- Volunteered in the past 12 months
- Level of trust in people*
- Religious affiliation
*Note: Variables were derived from individual indicators. Detailed description is provided in the appendix.
Finally, in the third model, we added another set of controls under the category ‘Civic inclination’ which included variables such as confidence in institutions, interest in politics and political activity.
Results and discussion
In the GSS 2020 sample, the voter turnout rate in the last federal election (2019) among Canadian born, established immigrants and recent immigrants was 86.4%, 80.6% and 68.5%, respectively. These estimates are slightly different from those reported in the Labour Force Survey,16 which is a common occurrence given that the survey sample and frequency of the two surveys vary. Table 1 shows the model effects of immigrant status and type, socio-demographic controls, social inclusivity variables, and political inclination factors as ‘odds ratio’ – values greater than 1 denote positive probability of the outcome per unit change in the independent variables, and values less than 1 denote negative odds. The results can also be viewed in this interactive Tableau dashboard.
We conducted a series of models, initially examining immigrant status (established and recent immigrants compared to Canadian born individuals) along with basic demographic controls: age, gender, visible minority status, education level, and income level. In Model 2, we added social inclusion variables to the factors in Model 1, and in Model 3, we added civic inclination-related variables to those in Model 2. The pseudo (Nagelkerke) R² improved significantly as we incorporated social inclusion and civic inclination variables in Models 2 and 3, respectively.
All three models consistently showed that being an immigrant was associated with a significantly lower likelihood of voting in federal elections compared to Canadian-born respondents. Although the addition of social inclusion and civic inclination variables in Models 2 and 3 slightly increased voting probability among immigrants, it remained significantly lower than that of Canadian-born respondents. Notably, in the final Model 3, the odds of voting among established and recent immigrants were, respectively, 30 percent and 60 percent lower than those of Canadian-born respondents.
Gender was not a significant predictor of voting behaviour in the first two models; however, when controlling for civic inclination behaviours (such as interest in politics and political activity) in Model 3, female respondents had significantly higher odds of voting compared to males. Additionally, higher age, income, and education levels were consistently associated with greater odds of voting across all models. The impact of social inclusion and civic inclination variables was particularly pronounced for each age group in Models 2 and 3. Similar trends were observed in the Labour Force Survey reporting of the federal elections of 2015 and 2019 – showing a gradual increase in voter turnout with increasing age, from 67–68 percent for the 18–24 age group to 85–86 percent for the 65–74 age group.17
| Variable Category | Indicator | Odds Ratio: Model 1 | Odds Ratio: Model 2 | Odds Ratio: Model 3 |
|---|---|---|---|---|
| Immigrant Type | Established immigrant | 0.659 (***) | 0.682 (***) | 0.692 (***) |
| Immigrant Type | Recent immigrant | 0.390 (***) | 0.392 (***) | 0.392 (***) |
| Demographics | Female | 1.055 | 1.002 | 1.176 (***) |
| Demographics | Age: 25-34 | 1.530 (***) | 1.702 (***) | 2.062 (***) |
| Demographics | Age: 35-44 | 1.811 (***) | 1.972 (***) | 2.450 (***) |
| Demographics | Age: 45-54 | 2.174 (***) | 2.356 (***) | 2.883 (***) |
| Demographics | Age: 55 and over | 3.407 (***) | 3.538 (***) | 4.221 (***) |
| Demographics | Visible minority | 0.768 (***) | 0.794 (***) | 0.933 |
| Demographics | Income | 1.168 (***) | 1.133 (***) | 1.127 (***) |
| Demographics | Education | 1.467 (***) | 1.360 (***) | 1.188 (***) |
| Social Inclusion | Number of close friends | 1.101 (***) | 1.050 (*) | |
| Social Inclusion | Number of close relatives | 1.148 (***) | 1.133 (***) | |
| Social Inclusion | Experienced discrimination in the past 5 years | 1.104 (*) | 0.900 (*) | |
| Social Inclusion | Group membership | 1.419 (***) | 1.222 (***) | |
| Social Inclusion | Volunteered | 1.104 | 1.027 | |
| Social Inclusion | Trust in people | 1.241 (***) | 1.116 (***) | |
| Social Inclusion | Religious affiliation | 1.274 (***) | 1.326 (***) | |
| Civic Inclination | Confidence in Institutions | 1.052 | ||
| Civic Inclination | Interested in Politics | 3.217 (***) | ||
| Civic Inclination | Political Activity | 1.230 (***) | ||
| Constant | 0.960 | 0.237 (***) | 0.134 (***) | |
| Sample size (N) | 30295 | |||
| Pseudo (Nagelkerke) R² | 0.100 | 0.126 | 0.219 |
Note: Significance levels are denoted by asterisks (*). One * means p < .05; Two ** means P < 0.01; Three *** means P < 0.001; No asterisk means ‘not significant’.
For the Age variable, 18-24 is the reference category in the models.
Having a robust social network (i.e., many close friends and relatives) was linked to higher odds of voting in both Models 2 and 3. However, the effect size and significance of the number of close friends diminished in Model 3 after controlling for civic inclination variables. Interestingly, respondents who reported experiencing discrimination in the past five years were slightly (+10 percent) more likely to vote according to Model 2, but this trend reversed after introducing civic inclination variables in Model 3. Being a registered member of any group or organization within the last 12 months significantly increased the odds of voting in both Models 2 and 3. However, volunteering was not a significant predictor in either model.
Trust in others and religious affiliation emerged as highly significant predictors of voting behaviour in both Models 2 and 3. Finally, the civic inclination variables of ‘having interest in politics’ and ‘being politically active’ were, as expected, associated with significantly higher odds of voting, as observed in Model 3.
Conclusion
The results from all three models consistently demonstrate that the odds of voting among immigrants are significantly lower compared to Canadian-born individuals, even when accounting for various demographic groups, diverse social behaviors and experiences, and civic inclinations. Among the sociodemographic variables, belonging to the higher age group (55+) emerges as having the most substantial positive effect on the model outcome. While most indicators in the social inclusion category positively impact voting, being a registered group or organizational member, affiliated with a religious group, and having higher trust in people appear to exert a more significant influence. These findings suggest that strategies to promote civic engagement, especially among new Canadians, should focus on strengthening social networks, addressing social isolation, and fostering community participation – essential components for building an inclusive society.