Economic and Social Reports
Unity in Canada: Experimental measures of feelings towards people with similar or different views
DOI: https://doi.org/10.25318/36280001202500500004-eng
Text begins
Abstract
Unity and social connectedness are key to forming cohesive societies, which typically experience higher well-being, growth and resilience. Using the Survey Series on People and their Communities (wave 5), this study focuses on six new experimental questions about respondents’ feelings towards people based on whether they share similar or differing views on politics, racism and gender identity. These questions were developed in response to recommendations from previous international work by the United Nations Economic Commission for Europe’s Task Team on Social Cohesion, which highlighted the lack of measures regarding affective ties between social groups. Overall, around one-third of survey respondents reported feeling the same towards people with similar views as those with different views on racism, while over 40% reported the same scores based on people’s political views and their views on gender identity. Findings also show that a small percentage of respondents (approximately 2%) who had “very warm” feelings towards people with similar political views also had “very cold” feelings towards those with differing views. In comparison, 24% of people had these feelings towards others based on their views on racism, and 15% had them based on their views on gender identity. The study also examined the sociodemographic characteristics associated with expressing such varying affective distances, which includes higher affective distance observed among respondents who were highly educated, younger, and those born in Canada.
Keywords: Affective distance, social cohesion, between-group relationships, in-group ties, out-group ties
Authors
Samuel MacIsaac and Grant Schellenberg (now retired) are with the Social Analysis and Modelling Division, Analytical Studies and Modelling Branch, at Statistics Canada.
Background
Unity, social connectedness, trust and belonging among individuals and social groups are key to forming cohesive societies (Jenson, 1998; Jeannotte et al., 2002; Fonseca et al., 2019; MacIsaac et al., 2023a; Aruqaj, 2023). Socially cohesive societies are typically healthier, more resilient to external shocks and crises, and experience higher economic growth and well-being (Organisation for Economic Co-operation and Development, 2011; Schiefer and Van der Noll, 2017; Aruqaj, 2023). Conversely, divisions between individuals and social groups can adversely affect social cohesion—generally understood as the bonds or social “glue” between societal members. While differences in views and opinions are a fundamental part of any democratic society, increased affective distanceNote between groups can impede functional discourse and dialogue (Owen et al., 2019). Moreover, affective rifts based on one’s “in-group” (i.e., shared values) versus “out-group” (i.e., differing values) could erode social cohesion (Jeannotte et al., 2002; MacIsaac et al., 2023b).
In 2022, based on widespread interest in questions about potentially widening rifts within society, the United Nations Economic Commission for Europe (UNECE) Conference of European Statisticians (CES) Bureau selected the concept of social cohesion for an in-depth review (MacIsaac et al., 2023a). This conceptual review paved the way to further explore the measurement of social cohesion, leading to the launch of an international Task Team to collect, review and analyze gaps and identify challenges to measurement practices.
Despite the existence of several survey questions on relationships with individuals in one’s in-group (e.g., family, friends, coworkers, neighbours), there are fewer measures of relationships with out-group members (e.g., groups with different identities, views, values or customs). The Task Team’s work focused on these “between-group” measures of social cohesion (MacIsaac et al., 2024). Putting together an inventory of international survey content, the Task Team identified several measures of affect towards out-group members, perceived degrees of tension or prejudice within society, and differences in values across social groups—most of which were based on affect towards people based on their sociodemographic characteristics (e.g., immigration status, ethnicity, gender, age). Markedly, the report highlighted a dearth in affective measures across social groups based on similarities and differences in views or values (MacIsaac et al., 2024).
This study responds to the Task Team’s recommendations to develop new survey content that addresses the dearth of measures of between-group affect and conduct research to better understand Canadians’ feelings towards those who share or do not share their views or values. Using new experimental “feeling thermometer” questions from wave 5 of the Survey Series on People and their Communities (SSPC), the study explores affect towards others based on whether respondents shared similar or differing views on politics, racism and gender identity. This study aims to provide insights for Canadians on affective ties between groups, with a specific focus on individuals who feel “very warm” towards their in-group and “very cold” towards their out-group.
The study is divided into several sections. Following a brief overview of the data and new experimental measures, the study presents summary statistics on respondents’ feelings towards others based on their similar or different views. The next sections delve deeper in the in-group versus out-group affective distance and its relation to respondents’ sociodemographic characteristics. The study concludes with a brief discussion of its findings and implications.
Data
Data collection for wave 5 of the SSPC – Social Cohesion and Experiences of Discrimination took place from April 2 to 22, 2024, with 8,855 respondents surveyed. Using the 2021 Census survey frame, the sample for the SSPC includes an oversampling of immigrants and people belonging to racialized groups. All reported numbers in this study use survey weights to better reflect the broader Canadian population and a set of 1,000 replicate bootstrap weights for the calculation of statistical variance.
New experimental feeling thermometer questions, as part of the content of the SSPC (wave 5), ask respondents about their feelings towards people with similar or differing views on politics, racism and gender identity. These questions are inspired by an extensive body of research on affective polarization (Iyengar et al., 2012; Iyengar et al., 2019; Johnston, 2023) and measure feelings of warmth or coldness towards selected groups across experimental topics that may unite or divide various groups. This temperature scale serves as a gauge of positive or negative feelings towards people, ranging from very warm (i.e., very positive) to very cold (i.e., very negative). The SSPC wave 5 thermometer questions use a scale from 0 to 10, where 0 represents very cold feelings, 10 represents very warm feelings, and 5 indicates neutral feelings towards a specific group.Note
As stated in the SSPC wave 5 microdata user guide, “it is important to note that the thermometer questions do not measure respondents’ personal views on politics, racism or gender identity” (Statistics Canada, 2024, p. 25). Because respondents’ personal views are unknown,Note those who report identical scores for their feelings towards people with similar and different views may still hold substantially different personal views. Therefore, one should assume heterogeneity in views among respondents, even among those with similar thermometer scores. No valid inferences can be made about the similarities or differences in respondents’ views.
As a final methodological note, these survey questions measure complex information, and respondents may have interpreted them differently. Beyond the effects of the questionnaire design on responses, survey results are also subject to individuals’ numerous subjective evaluations (Tourangeau et al., 2000), including their assessments of their own views, those who share their views, their feelings towards this group, and their feelings towards people with a potentially wide range of differing views. As such, respondents’ sociodemographic characteristics, ideological inclinations, perceptions of which groups they see as having similar or different views, and several other factors may shape their responses.
Descriptive feeling thermometer results
Table 1 provides summary measures of the feelings SSPC respondents expressed towards people with views that are similar and different from their own. Respondents expressed comparable feelings of warmth towards Canadians in general and towards people with similar views. When asked about their feelings towards Canadians in general, respondents reported an average feeling thermometer score of 7.2 on the 0 to 10 scale, with a median score of 7. Though not a measure of affect between different groups based on their views, feelings towards Canadians in general provide a useful baseline for comparison. Average feelings towards people with similar views on politics (6.7), racism (7.1) and gender identity (6.9) were similarly warm. In contrast, average feelings towards people with different views on racism (3.6) and gender identity (4.3) were markedly colder than average feelings towards people with different views on politics (5.2). Overall, respondents expressed warmer feelings towards people with similar views than towards those with different views, with a difference of 3.5 points for feelings towards people based on their views on racism, 2.6 points for views on gender identity, and 1.5 points for views on politics.
| On the 0 to 10 scale, how do you feel towards … | Average | 25th percentile | 50th percentile (median) | 75th percentile |
|---|---|---|---|---|
| number | ||||
| Source: Statistics Canada, Survey Series on People and their Communities, 2024. | ||||
| Canadians in general | 7.2 | 5 | 7 | 9 |
| People with similar political views | 6.7 | 5 | 7 | 8 |
| People with different political views | 5.2 | 5 | 5 | 6 |
| People with similar views on racism | 7.1 | 5 | 8 | 9 |
| People with different views on racism | 3.6 | 2 | 4 | 5 |
| People with similar views on gender identity | 6.9 | 5 | 7 | 9 |
| People with different views on gender identity | 4.3 | 3 | 5 | 5 |
Comparing responses at the higher and lower ends of the thermometer scale provides further perspective on feelings towards in-groups and out-groups. The bottom 25% (i.e., 25th percentile or lower) felt very cold (score of 2 or lower) towards people with different views on racism, and the top 25% of respondents (i.e., 75th percentile or higher) felt very warm (score of 9 or higher) towards people with similar views on racism (see Table 1). The seven-point gap between them, while drawn between selected points along two response distributions, highlights the potential for positive in-group and negative out-group affect. Similar results are observed for feelings towards people with similar and different views on gender identity, but not for feelings towards people with similar and different political views.
Moving on from the summary statistics above, survey respondents were grouped into mutually exclusive categories based on the feelings they expressed towards people with similar views versus those with different views. These categories, and the percentages of respondents in each, are shown in Tables 2 to 4.
Table 2 shows respondents’ feelings towards people based on their political views. The largest share of respondents (32%) expressed neutral feelings (score of 5) towards people with similar views and those with different views.Note Another 27% of respondents expressed warm feelings (score of 6 to 10) towards people with similar and those with different views on politics. People in this group, expressed positive affect towards both in-group and out-group members. Overall, 15% of respondents expressed warm feelings (score of 6 to 10) towards people with similar political views, but cold feelings (score of 0 to 4) towards those with differing views. This included the small percentage of respondents (2%) in the bottom-left corner of the cross-tabulation with more polarized scores, expressing very warm feelings (score of 8 to 10) towards people with similar views and very cold feelings (score of 0 to 2) towards those with differing views.
Despite the prevalence of more positive affect towards one’s in-group than out-group, the results above show little evidence of large affective gaps. A small share of respondents (2%) expressed diverging feelings towards people with similar and different political views. This is inconsistent with results from studies using the Canadian Election Study that show greater affective distance across political party and ideological lines in Canada (e.g., Johnston [2023]). This may reflect various factors, which are discussed later in the paper. Such feelings were more prominent towards people based on their views on racism (Table 3) and gender identity (Table 4) than on their political views.
| Feelings towards people with different political views | |||||
|---|---|---|---|---|---|
| 0 to 2 (very cold) | 3 to 4 | 5 (neutral) | 6 to 7 | 8 to 10 (very warm) | |
| percent | |||||
| Source: Statistics Canada, Survey Series on People and their Communities, 2024. | |||||
| Feelings towards people with similar political views | |||||
| 0 to 2 (very cold) | < 1 | < 1 | < 1 | < 1 | < 1 |
| 3 to 4 | < 1 | < 1 | < 1 | < 1 | < 1 |
| 5 (neutral) | 1.2 | 2.9 | 31.5 | < 1 | < 1 |
| 6 to 7 | 1.3 | 5.3 | 6.8 | 6.2 | < 1 |
| 8 to 10 (very warm) | 2.4 | 6.3 | 10.3 | 9.0 | 11.7 |
| Feelings towards people with different views on racism | |||||
|---|---|---|---|---|---|
| 0 to 2 (very cold) | 3 to 4 | 5 (neutral) | 6 to 7 | 8 to 10 (very warm) | |
| percent | |||||
| Source: Statistics Canada, Survey Series on People and their Communities, 2024. | |||||
| Feelings towards people with similar views on racism | |||||
| 0 to 2 (very cold) | 1.9 | < 1 | < 1 | < 1 | < 1 |
| 3 to 4 | < 1 | 1.6 | < 1 | < 1 | < 1 |
| 5 (neutral) | 3.9 | 3.2 | 23.2 | < 1 | < 1 |
| 6 to 7 | 2.4 | 4.2 | 3.7 | 2.9 | < 1 |
| 8 to 10 (very warm) | 23.8 | 11.1 | 7.6 | 3.2 | 4.7 |
| Feelings towards people with different views on gender identity | |||||
|---|---|---|---|---|---|
| 0 to 2 (very cold) | 3 to 4 | 5 (neutral) | 6 to 7 | 8 to 10 (very warm) | |
| percent | |||||
| Source: Statistics Canada, Survey Series on People and their Communities, 2024. | |||||
| Feelings towards people with similar views on gender identity | |||||
| 0 to 2 (very cold) | 1.4 | < 1 | < 1 | < 1 | < 1 |
| 3 to 4 | < 1 | 1.4 | < 1 | < 1 | < 1 |
| 5 (neutral) | 2.3 | 2.9 | 31.8 | < 1 | < 1 |
| 6 to 7 | 2.5 | 4.4 | 3.8 | 3.7 | < 1 |
| 8 to 10 (very warm) | 14.7 | 9.4 | 9.1 | 3.4 | 6.9 |
When SSPC respondents were asked about their feeling towards people with similar and different views on racism and gender identity, colder feelings towards out-group members were more prevalent. As summarized in Chart 1, 42% of respondents expressed warm feelings towards people with similar views and cold feelings towards people with different views on racism, with 24% of respondents expressing very warm in-group and very cold out-group feelings.
Similar results were observed for people based on their views on gender identity. Chart 1 shows that 31% of respondents expressed warm feelings towards people with similar views on gender identity and cold feelings towards those with different views, with 15% of respondents expressing very warm in-group and very cold out-group feelings.
Data table for Chart 1
| Political views | Views on racism | Views on gender identity | |
|---|---|---|---|
| percent | |||
| Source: Statistics Canada, Survey Series on People and their Communities, 2024. | |||
| Very warm towards people with similar views and those with different views | 11.7 | 4.7 | 6.9 |
| Warm towards people with similar views and those with different views | 27.4 | 11.1 | 14.3 |
| Neutral towards people with similar views and those with different views | 31.5 | 23.2 | 31.8 |
| Warm towards people with similar views and cold towards those with different views | 15.2 | 41.6 | 31.1 |
| Very warm towards people with similar views and very cold towards those with different views | 2.4 | 23.8 | 14.7 |
Overall, a substantial share of respondents reported strong positive affect towards in-group members and strong negative affect towards out-group members, particularly in terms of their views on racism and gender identity. That said, the results also highlight the extent to which respondents expressed neutral and positive feelings towards others, regardless of whether their views were similar or different
A closer look at respondents with polarized feeling thermometer scores
Neutral and warm feelings towards others capture a positive aspect of social cohesion. Warmth towards Canadians in general is linked to more positive outlooks, including on the unity among Canadians, the functioning of democracy and economic opportunities in Canada (MacIsaac and Arim, 2025). Yet, the prevalence of cold feelings (or negative affect), particularly towards out-group members, is also of interest. The intersection of the two, reflecting positive in-group and negative between-group ties, may be of concern if this reflects “us and them” perspectives and a weakened social fabric. To learn more about this, the focus is on respondents in the bottom-left corners of each cross-tabulation in Tables 2 to 4—those who, respectively, expressed very warm feelings towards in-group members and very cold feelings towards out-group members. Table 5 shows the shares of respondents expressing these views, disaggregated by characteristics, such as age, education and immigration status.
Larger shares of younger respondents than older respondents expressed very warm feelings towards in-group members and very cold feelings towards out-group members, with 29% of individuals aged 15 to 24 years and 20% of seniors doing so in terms of views on racism. The same pattern is observed in terms of views on gender identity. The shares of women expressing such views were slightly larger than the shares of men. Larger shares of those who identified as 2SLGBTQ+ people expressed very warm in-group and very cold out-group feelings, compared with those who did not identify as 2SLGBTQ+ people. This was the case in terms of views on gender identity and racism.
The prevalence of these feelings did not vary systematically across household composition. In terms of religiosity, respondents who identified as having no religion or a secular perspective were more likely to express these contrasting feelings. Across educational attainment, respondents with a university degree were most likely to express them, particularly based on people’s views on racism.
The characteristics associated with being in the group of interest include younger age, higher educational attainment, a secular perspective and 2SLGBTQ+ identity. Results in Table 5 suggest that place of residence also matters, although the pattern becomes relatively small and inconsistent in the multivariate models below.
Feelings towards in-group and out-group members varied considerably by immigration status and whether respondents belonged to a racialized group. Based on people’s views on racism, larger shares of Canadian-born respondents (28%) expressed very warm feelings towards in-group members and very cold feelings towards out-group members, compared with immigrant respondents (14%). Similarly, a larger share of respondents who did not identify as belonging to a racialized group expressed very warm in-group and very cold out-group feelings in terms of views on racism (28%), compared with those who identified with a racialized group (12% to 21%). These relationships are further examined in the multivariate models below.
| Individuals expressing very warm feelings towards people with similar views and very cold feelings towards people with different views | |||
|---|---|---|---|
| Political views | Views on racism | Views on gender identity | |
| percent | |||
| Notes: N.i.e. = not included elsewhere. The term “men+” includes men and some non-binary people, and the term “women+” includes women and some non-binary people.
Source: Statistics Canada, Survey Series on People and their Communities, 2024. |
|||
| Total | 2 | 24 | 15 |
| Age group | |||
| 15 to 24 years | 2 | 29 | 24 |
| 25 to 34 years | 2 | 29 | 16 |
| 35 to 44 years | 3 | 27 | 14 |
| 45 to 54 years | 4 | 20 | 13 |
| 55 years and older | 1 | 20 | 12 |
| Gender | |||
| Men+ | 3 | 22 | 14 |
| Women+ | 2 | 25 | 16 |
| 2SLGBTQ+ people | |||
| Yes | 4 | 38 | 30 |
| No | 2 | 23 | 13 |
| Highest level of education completed | |||
| High school or less | 1 | 19 | 14 |
| College, CEGEP or trades certificate or diploma | 3 | 26 | 15 |
| University degree (bachelor's or above | 3 | 28 | 16 |
| Household composition | |||
| Living alone | 3 | 25 | 15 |
| Multiple person household, no children | 3 | 25 | 14 |
| Multiple person household, with child(ren) | 2 | 21 | 16 |
| Reside in a census metropolitan area | |||
| Yes | 2 | 22 | 14 |
| No | 2 | 29 | 16 |
| Religion | |||
| Identified a religion or spiritual tradition | 2 | 20 | 12 |
| No religion or secular perspectives | 4 | 32 | 19 |
| Immigration status | |||
| Canadian-born individual | 3 | 28 | 17 |
| Immigrant | 2 | 14 | 9 |
| Non-permanent resident | 1 | 13 | 8 |
| Racialized group | |||
| South Asian | 3 | 12 | 8 |
| Chinese | 2 | 12 | 8 |
| Black | 2 | 16 | 7 |
| Filipino | 2 | 12 | 6 |
| Arab | 2 | 16 | 13 |
| Latin American | 2 | 21 | 12 |
| Southeast Asian | 1 | 12 | 8 |
| West Asian | 2 | 16 | 14 |
| Korean or Japanese | 1 | 12 | 9 |
| Racialized group, n.i.e. or multiple racialized groups | 1 | 18 | 9 |
| Does not belong to a racialized group | 3 | 28 | 17 |
Multivariate analysis of affective distance between groups
The above profile is based on responses at the top and bottom of the thermometer scales. While this provides perspective on a specific group of respondents, it does not capture the full range of affective distances expressed by SSPC respondents. For example, a respondent who reported a score of seven towards people with similar views and a score of three towards people with differing views would not be included in the group profiled above, despite a four-point difference in their feelings towards in-group and out-group members.
To capture the full range of this affective distance, a derived value calculated by subtracting each respondent’s feelings score towards people with different views from their score towards people with similar views was used. Differences in feeling thermometer scores ranging from 1 to 10 quantify how much warmer the respondent felt towards in-group versus out-group members, while a difference of 0 indicates that the respondent expressed the same thermometer feeling towards members of both groups. Scores ranging from -1 to -10 were reported among 2% to 3% of individuals who expressed warmer feelings towards out-group than in-group members. The absolute value of this variable was taken and identified as between-group affective distance, or simply, affective distance. This derived variable, ranging from 0 to 10, was included as the dependent variable in a linear regression model and regressed against a set of sociodemographic characteristics.
Results from the regression models are presented in Table 6. A positive coefficient indicates how much of an increase in affective distance between in-group and out-group members is associated with a given characteristic relative to a reference group. Conversely, a negative coefficient indicates how much of a decrease in affective distance between in-group and out-group members is associated with a given characteristics. The constant for each model, at the bottom of the table, shows the between-group affective distance when all the independent variables are set to their reference group category. The discussion primarily highlights coefficients from models 2 and 3, as those from Model 1 are generally consistent but smaller in magnitude.
The regression results from the multivariate models largely mirror the descriptive results above. Again, affective distance expressed towards in-group and out-group members was larger among younger respondents, compared with older respondents. In terms of feelings towards people based on their views on racism and gender identity, between-group affective distances were 1.1 to 1.3 points larger among respondents aged 15 to 24 than among the reference group, respondents aged 55 and over. Still, the affective distances expressed by those aged 25 to 44 suggest a cautious interpretation is warranted.
Women and respondents identifying as 2SLGBTQ+ people were tied to larger affective distances. Women expressed larger affective distances than men (0.4 to 0.5 points) when asked about their feelings towards people based on their views on racism and gender identity. Meanwhile, with regard to views on racism and gender identity, affective distances were 0.8 to 1.5 points larger among those who identified as 2SLGBTQ+ than among those who did not.
Higher educational attainment and secularity were consistently correlated with larger between-group affective distances. Respondents with a trades or college certificate, or a university certificate below a bachelor’s degree, had affective distances 0.4 to 0.8 points larger than the reference group (those with high school or less), while respondents with a university degree had affective distances 0.6 to 1.4 points larger. Secularity remains a significant correlate in the multivariate models, with between-group affective distances 0.5 to 0.8 points larger among respondents reporting no religion or a secular perspective. Results across household composition and place of residence varied in direction across models, and coefficients were generally small.
Consistent with the descriptive results above, respondents born in Canada expressed larger affective distances than immigrants based on people’s views on racism and gender identity. This difference was also narrower (0.6 points, or no statistically significant difference) for those who immigrated to Canada before 1980, compared with their Canadian-born counterparts, than the 1.1 to 1.5 points more for non-permanent residents and immigrants who arrived from 2016 to 2021. These results suggest that more time spent in Canada since immigration is tied to larger affective distances, entailing a smaller affective distance gap between more established immigrants and individuals born in Canada.
Affective distance also varied depending on whether respondents identified as belonging to various racialized groups. Apart from Black and Latin American respondents’ similarly large between-group affective distances based on people’s views on racism, respondents belonging to a racialized group typically reported lower affective distance than White respondents. Meanwhile, Chinese, Filipino, Korean, Japanese, South Asian and Southeast Asian respondents, and those belonging to multiple racialized groups, tended to report lower affective distance, compared with other groups.
| Ordinary least squares regression on difference in feelings towards people | |||
|---|---|---|---|
| (1) | (2) | (3) | |
| Political views | Views on racism | Views on gender identity | |
Source: Statistics Canada, Survey Series on People and their Communities, 2024. |
|||
| Age (reference: 55 years and older) | |||
| 15 to 24 years | 0.2 Table 6 Note ** | 1.3 Table 6 Note ** | 1.1 Table 6 Note ** |
| 25 to 34 years | 0.1 Table 6 Note * | 0.4 Table 6 Note ** | 0.1 |
| 35 to 44 years | 0.0 | 0.9 Table 6 Note ** | 0.0 |
| 45 to 54 years | 0.0 | -0.1 Table 6 Note † | -0.2 Table 6 Note ** |
| Gender (reference: men+) | |||
| Women+ | 0.1 Table 6 Note † | 0.5 Table 6 Note ** | 0.4 Table 6 Note ** |
| 2SLGBTQ+ (reference: no) | |||
| Yes | 0.5 Table 6 Note ** | 0.8 Table 6 Note ** | 1.5 Table 6 Note ** |
| Highest level of education completed (reference: high school or less) | |||
| College, CEGEP or trades certificate or diploma | 0.4 Table 6 Note ** | 0.8 Table 6 Note ** | 0.5 Table 6 Note ** |
| University degree (bachelor’s or above) | 0.6 Table 6 Note ** | 1.5 Table 6 Note ** | 1.0 Table 6 Note ** |
| Household composition (reference: living alone) | |||
| Multiple person household, no children | -0.1 Table 6 Note * | 0.3 Table 6 Note ** | -0.2 Table 6 Note * |
| Multiple person household, with child(ren) | -0.3 Table 6 Note ** | -0.2 Table 6 Note * | 0.0 |
| Place of residence (reference: resides in a census metropolitan area) | |||
| Did not reside in a census metropolitan area | 0.1 Table 6 Note * | 0.2 Table 6 Note ** | -0.1 Table 6 Note * |
| Religion (reference: identified a religion or spiritual tradition) | |||
| No religion or secular perspectives | 0.5 Table 6 Note ** | 0.8 Table 6 Note ** | 0.5 Table 6 Note ** |
| Year of immigration (reference: Canadian-born individual) | |||
| Before 1980 | -0.1 | -0.6 Table 6 Note ** | -0.1 |
| 1980 to 1990 | 0.1 | -0.6 Table 6 Note ** | -0.4 Table 6 Note ** |
| 1991 to 2000 | 0.0 | -1.0 Table 6 Note ** | -0.3 Table 6 Note ** |
| 2001 to 2010 | -0.2 Table 6 Note ** | -1.0 Table 6 Note ** | -0.7 Table 6 Note ** |
| 2011 to 2015 | -0.4 Table 6 Note ** | -1.3 Table 6 Note ** | -1.0 Table 6 Note ** |
| 2016 to 2021 | -0.5 Table 6 Note ** | -1.5 Table 6 Note ** | -1.1 Table 6 Note ** |
| Non-permanent resident | -0.6 Table 6 Note ** | -1.5 Table 6 Note ** | -1.1 Table 6 Note ** |
| Racialized group (reference: does not belong to a racialized group) | |||
| South Asian | 0.0 | -0.9 Table 6 Note ** | -0.7 Table 6 Note ** |
| Chinese | -0.5 Table 6 Note ** | -1.6 Table 6 Note ** | -1.2 Table 6 Note ** |
| Black | 0.1 Table 6 Note † | 0.1 Table 6 Note * | -0.6 Table 6 Note ** |
| Filipino | -0.2 Table 6 Note ** | -0.8 Table 6 Note ** | -0.8 Table 6 Note ** |
| Arab | 0.1 | -0.2 Table 6 Note ** | -0.4 Table 6 Note ** |
| Latin American | 0.0 | 0.1 | -0.1 |
| Southeast Asian | -0.3 Table 6 Note ** | -1.2 Table 6 Note ** | -1.2 Table 6 Note ** |
| West Asian | 0.0 | -0.4 Table 6 Note ** | -0.2 Table 6 Note * |
| Korean | -0.8 Table 6 Note ** | -1.4 Table 6 Note ** | -0.9 Table 6 Note ** |
| Japanese | -0.2 | -1.0 Table 6 Note ** | -0.9 Table 6 Note ** |
| Racialized group, n.i.e. | 0.1 | -0.3 Table 6 Note * | -0.3 Table 6 Note † |
| Multiple racialized groups | -0.1 Table 6 Note * | -0.8 Table 6 Note ** | -1.1 Table 6 Note ** |
| Constant | 1.3 Table 6 Note ** | 2.3 Table 6 Note ** | 2.1 Table 6 Note ** |
Discussion
Similar or compatible social values are known to favour affective ties and social cohesion (Jenson, 1998; Jeannotte et al., 2002; Fonseca et al., 2019; MacIsaac et al., 2023a; MacIsaac et al., 2023b; Aruqaj, 2023). As such, the primary objective of this study was to develop and analyze new survey content that captures affective distance between groups. This distance has been identified as an important yet under-measured aspect of social cohesion. To measure affective distance, this study uses new experimental survey questions on self-reported feelings towards people based on similarities and differences in their views. The definition of in-group and out-group members in terms of similar or different views on politics, racism and gender identity was intended to complement existing measures of out-group affect across other characteristics, such as trust in other social groups. The analyses of respondents’ affective ties across sociodemographic characteristics suggest these new feeling thermometer questions and the derived affective distance metric are valid measures of between-group affect and, more broadly, social unity.
Affective distances based on people’s views on racism and gender identity were larger than those based on politics. One reason may be that views on racism and gender identity are far more specific than political views, making it easier for respondents to identify people with similar or different views. Moreover, people’s views on racism and gender identity could also be more relevant to respondents’ personal experiences,Note identities and sense of morality than people’s political views, thereby reinforcing ties within groups and weakening those between groups. The strong correlation between 2SLGBTQ+ identity and affective distance based on views on gender identity are consistent with this. Ultimately, questions that relate to views on racism and gender identity align closely with the survey’s theme on social cohesion and experiences with discrimination and the focus on discrimination in previous waves.Note
In addition to results based on respondents’ personal identities and experiences, other sociodemographic variables are also associated with in-group versus out-group affective distance. Larger affective distances among younger and more highly educated respondents may reflect this group’s social activism, particularly in relation to racism and gender identity.
Between-group affective distances based on respondents’ immigration status and whether they belonged to a racialized group also stand out. Though the outlooks underlying this affective distance remain unobserved, some may ascribe it to the range of views on racism articulated through ongoing public discussions, activism and debates. Affective distance based on people’s views on racism was highest among Black, Latin American and White respondents. Relatedly, respondents who were born in Canada reported higher affective distances.
In-group versus out-group affective distances may reflect longstanding differences across social groups or more recent societal shifts in values. In an age of increasingly digital interactions that can influence the visibility of similar versus dissenting views (Owen et al., 2019), recognizing how diverging views shape how people feel towards each other could enhance insights into societal unity in Canada. Moving forward, more longitudinal data and research are needed to determine changes in relationships over time and across social groups.
References
Aruqaj, B. (2023). An integrated approach to the conceptualisation and measurement of social cohesion. Social Indicators Research, 168(1): 227-263.
Fonseca, X., Lukosch, S., & Brazier, F. (2019). Social cohesion revisited: A new definition and how to characterize it. Innovation: The European Journal of Social Science Research, 32(2): 231-253.
Iyengar, S., Lelkes, Y., Levendusky, M., Malhotra, N., & Westwood, S.J. (2019). The origins and consequences of affective polarization in the United States. Annual Review of Political Science, 22: 129-146.
Iyengar, S., Sood, G., & Lelkes, Y. (2012). Affect, not ideology: A social identity perspective on polarization. Public Opinion Quarterly, 76(3): 405-431.
Jeannotte, S.M., Stanley, D., Pendakur, R., Jamieson, B., Williams, M., & Aizlewood, A. (2002). Buying In or Dropping Out: The Public Policy Implications of Social Cohesion. Ottawa: Department of Canadian Heritage.
Jenson J. (1998). Mapping Social Cohesion: The State of Canadian Research. Ottawa: Canadian Policy Research Network.
Johnston R. (2023). Affective polarization in the Canadian party system, 1988–2021. Canadian Journal of Political Science, 56(2): 372-395.
MacIsaac, S., Schellenberg, G., & Wavrock, D. (2023a). Social Cohesion: Concept and Measurement. Geneva: United Nations Economic Commission for Europe.
MacIsaac, S., Wavrock, D., & Schellenberg, G. (2023b). What holds us together? Measuring dimensions of social cohesion in Canada. Statistical Journal of the IAOS, 39(3): 589-604.
MacIsaac, S., Schellenberg, G., Arim, R., & Wavrock, D. (2024). Measuring Social Cohesion. Geneva: United Nations Economic Commission for Europe.
MacIsaac, S., & Arim, R. (2025). Hopefulness about Canadian society. Economic and Social Reports 5(5), 1-5.
Organisation for Economic Co-operation and Development. (2011). Perspectives on Global Development 2012: Social Cohesion in a Shifting World. Paris: OECD Publishing.
Owen, T., Loewen, P., Ruths, D., Bridgman, A., Gorwa, R., MacLellan, S., Merkley, E., Potter, A., Skazinetsky, B., & Zhilin, O. (2019). Digital democracy project: Polarization and its discontents. Public Policy Forum.
Schiefer, D., & Van der Noll, J. (2017). The essentials of social cohesion: A literature review. Social Indicators Research, 132(2), 579-603.
Statistics Canada. (2024). Microdata User Guide: Survey Series on People and Their Communities - Social Cohesion and Experiences with Discrimination. Ottawa: Statistics Canada.
Tourangeau, R., Rips, L. J., & Rasinski, K. A. (2000). The Psychology of Survey Response. Cambridge: Cambridge University Press.
- Date modified: