@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Arts, Faculty of"@en, "Sociology, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Hwang, Monica, Mi Hee Clara"@en ; dcterms:issued "2013-09-12T14:29:58Z"@en, "2013"@en ; vivo:relatedDegree "Doctor of Philosophy - PhD"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """This thesis considers ethno-racial differences in social and political trust, which leading scholars see as the two key dimensions of social cohesion in Canada. Although not directly addressing problems of prejudice and intolerance, the analysis relates to this same research tradition. I compare trust among eight ethno-racial groupings: British, French, “Canadians,” other Europeans, Aboriginal Peoples, visible minorities, mixed-origins respondents, and all others. Building from the concepts of “social distance” and “social boundaries,” I test three sets of factors for explaining ethno-racial differences in trust: (1) three ethno-cultural “markers” – religion, language, and immigration status; (2) two socioeconomic influences –education and income; and (3) two social engagement indicators – voluntary association activity and ethnic diversity of friendships. Models also include controls for region, age, and gender. Based on the 2008 General Social Survey Public Use Microdata file, findings indicate that a perspective employing concepts of social distance and social boundaries helps in understanding many, though not all, of the differences in trust across ethno-racial communities. The results show that, compared to more established groups like the British, the most culturally distinctive minorities – visible minorities, French, and Aboriginal Peoples – express less social trust. This is consistent with the interpretation that groups subjected to more social distance/social boundaries experiences are less likely to develop social trust. Nevertheless, these same groups, except for Aboriginal Peoples, exhibit relatively high political trust. The latter finding suggests that some minorities, when treated or perceived by others as different or distant from the “mainstream,” may see government agencies as defending their minority rights and interests against discrimination. Aboriginal Peoples are an exception in being the only minority grouping to express lower levels of both social and political trust. This underscores their unique position in Canada. Despite being the country’s original inhabitants, they have long endured processes of discrimination, exclusion, and racism that understandably contribute to lower trust in other people. At the same time, historical and present-day governments have ignored, exacerbated, or created many of these injustices, giving Aboriginal Peoples far less reason than other groups to trust Canadian political institutions."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/45063?expand=metadata"@en ; skos:note " IN CANADA WE TRUST? UNDERSTANDING ETHNO-RACIAL VARIATIONS IN SOCIAL AND POLITICAL TRUST by MONICA MI HEE CLARA HWANG A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Sociology) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) September 2013 ? Monica Mi Hee Clara Hwang, 2013 ii ABSTRACT This thesis considers ethno-racial differences in social and political trust, which leading scholars see as the two key dimensions of social cohesion in Canada. Although not directly addressing problems of prejudice and intolerance, the analysis relates to this same research tradition. I compare trust among eight ethno-racial groupings: British, French, ?Canadians,? other Europeans, Aboriginal Peoples, visible minorities, mixed-origins respondents, and all others. Building from the concepts of ?social distance? and ?social boundaries,? I test three sets of factors for explaining ethno-racial differences in trust: (1) three ethno-cultural ?markers? ? religion, language, and immigration status; (2) two socioeconomic influences ?education and income; and (3) two social engagement indicators ? voluntary association activity and ethnic diversity of friendships. Models also include controls for region, age, and gender. Based on the 2008 General Social Survey Public Use Microdata file, findings indicate that a perspective employing concepts of social distance and social boundaries helps in understanding many, though not all, of the differences in trust across ethno-racial communities. The results show that, compared to more established groups like the British, the most culturally distinctive minorities ? visible minorities, French, and Aboriginal Peoples ? express less social trust. This is consistent with the interpretation that groups subjected to more social distance/social boundaries experiences are less likely to develop social trust. Nevertheless, these same groups, except for Aboriginal Peoples, exhibit relatively high political trust. The latter finding suggests that some minorities, when treated or perceived by others as different or distant from the ?mainstream,? may see government agencies as iii defending their minority rights and interests against discrimination. Aboriginal Peoples are an exception in being the only minority grouping to express lower levels of both social and political trust. This underscores their unique position in Canada. Despite being the country?s original inhabitants, they have long endured processes of discrimination, exclusion, and racism that understandably contribute to lower trust in other people. At the same time, historical and present-day governments have ignored, exacerbated, or created many of these injustices, giving Aboriginal Peoples far less reason than other groups to trust Canadian political institutions. iv PREFACE This dissertation is original, unpublished, independent work by the author, M. Hwang. v TABLE OF CONTENTS ABSTRACT ............................................................................................................................. ii PREFACE ................................................................................................................................ iv TABLE OF CONTENTS ......................................................................................................... v LIST OF TABLES ................................................................................................................... ix CHAPTER ONE: LITERATURE REVIEW AND THEORETICAL ISSUES ....................... 1 INTRODUCTION ................................................................................................................ 1 ETHNIC DIVERSITY AND TRUST: INTERNATIONAL PATTERNS ........................ 12 Putnam on Ethnic Diversity and Trust in the United States ........................................... 12 Ethnic Diversity and Trust in European Nations ............................................................ 17 ETHNIC DIVERSITY AND TRUST IN CANADA ......................................................... 19 Canada?s Multiculturalism Policy .................................................................................. 20 Immigration Policy and the Points Selection System ..................................................... 21 Are Canadians Concerned about Ethnic Diversity and Multiculturalism? ..................... 23 The Unique Case of Aboriginal Peoples ........................................................................ 26 THEORETICAL ISSUES: SOCIAL DISTANCE AND SOCIAL BOUNDARIES ........ 28 Some Personal Observations .......................................................................................... 28 Religion, Language, and Immigration Status as Ethno-cultural Markers ...................... 30 Social Distance and Cultural Distinctiveness ................................................................. 32 Social Boundaries and Cultural Distinctiveness ............................................................. 36 Research on Ethnicity and Trust ..................................................................................... 41 Breton?s Research ....................................................................................................... 44 The Findings of Soroka and Colleagues ..................................................................... 46 Reitz and Banerjee?s Research ................................................................................... 48 Ethnicity and Political Trust in Canada and Other Societies .......................................... 53 vi KEY EXPLANATORY VARIABLES .............................................................................. 58 Ethnocultural Markers: Immigration Status, Religion, and Language ........................... 58 Immigration Status: Nativity and Time in Canada ..................................................... 58 Religion....................................................................................................................... 60 The Effects of Language ............................................................................................. 63 Socioeconomic Factors: Income and Education ............................................................. 64 Social Engagement Factors: Voluntary Association Membership and Diversity of Friendship Networks ....................................................................................................... 67 Ethnicity and Voluntary Association Membership .................................................... 68 Ethnic Diversity of Friendship Networks ................................................................... 70 ADDITIONAL CONTROL VARIABLES: REGION, AGE, AND GENDER ................. 72 Region ............................................................................................................................. 72 Age .................................................................................................................................. 74 Gender............................................................................................................................. 75 MAIN RESEARCH QUESTIONS..................................................................................... 75 SUMMARY ........................................................................................................................ 77 CHAPTER TWO: DATA AND METHODS ......................................................................... 79 DATA SOURCE ................................................................................................................ 79 VARIABLES ...................................................................................................................... 81 The Dependent Variables: Measuring Social Trust and Political Trust ......................... 81 Social Trust ................................................................................................................. 82 Political Trust.............................................................................................................. 84 The Independent Variable: Ethno-racial Background .................................................... 86 Explanatory Control Variables ....................................................................................... 87 Ethno-cultural Markers: Immigration Status, Religion, and Language ...................... 87 Socioeconomic Factors: Education and Income ......................................................... 89 Social Engagement Factors: Voluntary Association Membership and Ethnic Diversity of Friendship Networks .............................................................................................. 91 vii Other Control Variables: Region, Age, and Gender ....................................................... 94 CHAPTER THREE: BIVARIATE RESULTS ...................................................................... 95 Relationships Between Ethno-racial Background and the Control Variables ................ 95 The Association Between Ethno-racial Background and Social Trust ......................... 100 The Association Between Ethno-racial Background and Political Trust ..................... 103 The Relationships Between the Control Variables and Social Trust ............................ 107 The Relationships Between the Control Variables and Political Trust ........................ 110 SUMMARY OF THE BIVARIATE FINDINGS ............................................................ 113 CHAPTER FOUR: MULTIVARIATE ANALYSIS ........................................................... 115 PART I: MODELLING SOCIAL TRUST ....................................................................... 116 Ethno-racial Background, Social Trust, and the Three Ethno-cultural Markers .......... 116 Model 1 ..................................................................................................................... 116 Model 2 ..................................................................................................................... 118 Model 3 ..................................................................................................................... 119 Model 4 ..................................................................................................................... 120 Ethno-racial Background, Social Trust, and the Effects of Socio-economic Status .... 121 Model 5 and Model 6 ................................................................................................ 121 Ethno-racial Background, Social Trust, and Social Engagement Influences ............... 124 Model 7 and Model 8 ................................................................................................ 124 The Full Explanatory Model of Social Trust: Model 9 and Model 10 ......................... 127 Model 9 ..................................................................................................................... 129 Model 10 ................................................................................................................... 131 PART II: MODELLING POLITICAL TRUST ............................................................... 132 Ethno-racial Background, Political Trust, and the Three Ethno-cultural Markers ....... 132 Model 1 ..................................................................................................................... 134 Model 2 ..................................................................................................................... 134 Model 3 ..................................................................................................................... 135 viii Model 4 ..................................................................................................................... 135 Ethno-racial Background, Political Trust, and the Effects of Socio-economic Status . 137 Model 5 and Model 6 ................................................................................................ 137 Ethno-racial Background, Political Trust, and Social Engagement Influences ............ 139 Model 7 and Model 8 ................................................................................................ 139 The Full Explanatory Model of Political Trust: Model 9 and Model 10 ...................... 141 Model 9 ..................................................................................................................... 144 Model 10 ................................................................................................................... 144 SUMMARY OF THE MULTIVARIATE FINDINGS .................................................... 145 CHAPTER FIVE: CONCLUSIONS .................................................................................... 150 VISIBLE MINORITIES ................................................................................................... 152 Interpreting the Findings for Social Trust Among Visible Minorities ......................... 153 Interpreting the Findings for Political Trust Among Visible Minorities ...................... 155 THE FRENCH AND THE ?CANADIANS? ................................................................... 161 Interpreting the Findings for Social and Political Trust Among the French and Canadians ...................................................................................................................... 162 ABORIGINAL PEOPLES ............................................................................................... 167 Interpreting the Findings for Social and Political Trust Among Aboriginal Peoples ... 168 THE BRITISH, OTHER EUROPEANS, AND THE ?MIXED-ORIGINS? CATEGORY .......................................................................................................................................... 172 CONCLUDING REMARKS ........................................................................................... 175 REFERENCES ..................................................................................................................... 177 APPENDIX .......................................................................................................................... 196 A Note on Regional Differences in Social and Political Trust ..................................... 196 ix LIST OF TABLES TABLE 3.1 0.1 Ethno-racial Background and Immigration Status........................................ 96 TABLE 3.2 0.2Ethno-racial Background and Religion......................................................... 96 TABLE 3.3 0.3Ethno-racial Background and Language....................................................... 97 TABLE 3.4 0.4Ethno-racial Background and Region ........................................................... 98 TABLE 3.5 0.5 Ethno-racial Background and Education (in Years), Annual Household Income (in Dollars), Number of Voluntary Memberships, Ethnic Diversity of Friendships (Mean Score on 5-Point Scale), Age (Average in Years), and Percent Female........................................................................................................................... 100 TABLE 3.6 0.6 Social Trust Measures by Ethno-racial Background .................................. 101 TABLE 3.7 0.7Political Trust Measures by Ethno-racial Background ............................... 104 TABLE 3.8 0.8 Social Trust Scale Scores by All the Control Variables ............................. 108 TABLE 3.9 0.9 Political Trust Scale Scores by All the Control Variables .......................... 111 TABLE 4.1 0.1 Multiple Regression Analysis (Unstandardized) Showing the Effects of Ethno-racial Background, Immigration Status, Religion, and Language on Social Trust ...................................................................................................................................... 117 TABLE 4.2 0.2Multiple Regression Analysis (Unstandardized) Showing the Effects of Ethno-racial Background, Education, and Income on Social Trust.............................. 122 TABLE 4.3 0.3 Multiple Regression Analysis (Unstandardized) Showing the Effects of Ethno-racial Background, Number of Voluntary Memberships, and Ethnic Diversity of Friendships on Social Trust .......................................................................................... 125 TABLE 4.4 0.4Multiple Regression Analysis (Unstandardized) Showing the Effects of Ethno-racial Background, Immigration Status, Religion, Language, Education, Income, Voluntary Membership, Ethnic Diversity of Friendships, Region, Age, and Gender on Social Trust ................................................................................................................... 128 TABLE 4.5 0.5 Multiple Regression Analysis (Unstandardized) Showing the Effects of Ethno-racial Background, Immigration Status, Religion, and Language on Political Trust .............................................................................................................................. 133 TABLE 4.6 0.6 Multiple Regression Analysis (Unstandardized) Showing the Effects of Ethno-racial Background, Education, and Income on Political Trust .......................... 138 TABLE 4.7 0.7 Multiple Regression Analysis (Unstandardized) Showing the Effects of Ethno-racial Background, Number of Voluntary Memberships, and Ethnic Diversity of Friendships on Political Trust ....................................................................................... 140 x TABLE 4.8 0.8Multiple Regression Analysis (Unstandardized) Showing the Effects of Ethno-racial Background, Immigration Status, Religion, Language, Education, Income, Voluntary Membership, Ethnic Diversity of Friendships, Region, Age, and Gender on Political Trust................................................................................................................ 142 TABLE A1 0.1 Social Trust and Political Trust Scores of the Eight Ethno-racial Groups, for Canada and the Five Regions ....................................................................................... 197 TABLE A2 19 Descriptive Statistics for Independent and Control Variables .................... 198 1 CHAPTER ONE: LITERATURE REVIEW AND THEORETICAL ISSUES INTRODUCTION Canada is widely recognized as an ethnically and culturally diverse society. This diversity is evident in the high percentage of Canadians who are foreign-born, which stood at almost 20 percent as of the 2006 Census, and surpassed 20 percent in 2011 (Proussalidis 2013). It is also reflected in the linguistic duality of the nation, in the sizeable Aboriginal Peoples population, and in the proportion of Canadians who are racial minorities, as judged by their membership in one of Statistics Canada?s ?visible minority?1 categories. This proportion was 1 This term is employed advisedly in the thesis, because of its standard usage in Canadian government research and data sources, including those that are accessed in my analysis. Canada?s Employment Equity Act identifies people who are non-white in colour and non-Caucasian in race as ?visible minority.? Individuals who are Chinese, South Asian (e.g., East Indian, Pakistani, Punjabi, Sri Lankan); Black (e.g., African, Haitian, Jamaican, Somali); Arab/West Asian (e.g., Armenian, Egyptian, Iranian, Lebanese, Moroccan); Filipino; South East Asian (e.g., Cambodian, Indonesian, Laotian, Vietnamese); Latin American; Japanese; Korean; and Other fall under this classification. This term excludes Aboriginal peoples. (See, e.g., Boyd and Vickers 2009). While the visible minority term is widely used in Canada and informs policy decisions, not all researchers accept it universally. Historian Henry Yu (2009) has questioned whether the visible minority concept is applicable to the demographic composition in metropolitan areas, where a majority of nonwhite Canadians reside (96%). In this instance, Yu (2009) argues that the term may become a ?demographic oxymoron,? because certain visible minority groups are on the verge of becoming the majority group in areas such as Toronto (43%) and Vancouver (42%) as of the 2006 Census. While Yu makes an interesting point, it is worthwhile to mention that the visible minority designation does not refer to numerical majorities or minorities. As Yu, himself, points out, ?[t]he term has always been about visibility in a society dominated by people who are invisible because they are white? (Ward 2008). Hence, for Yu, this term is more than a statistical indicator, but a reference to a history of discrimination against nonwhite Canadians. Notwithstanding some 2 16 percent in 2006 and rose to more than 19 percent in 2011(Statistics Canada 2008a, 2008b; Proussalidis 2013). The comparable proportions are even larger in metropolitan areas like Toronto and Vancouver, which are quickly becoming ?majority-minority? cities, i.e., cities in which most of the population have visible minority backgrounds (Kalbach, Verma, George, and Dai 1993; Statistics Canada 2005; Yu 2009). Such figures clearly place Canada among the most diverse societies in the world (see also Li 2003: 4). Cultural diversity extends beyond the visible minority population to the Aboriginal Peoples2 population as well. There concerns about the term, Yu maintains that the visible minority label remains a meaningful description of certain non-white Canadians who continue to be susceptible to racial discrimination, such as Muslim, Filipino, South Asian, and Caribbean-Canadian immigrants. There is much internal diversity within the visible minority grouping, with an array of linguistic and social variations reflecting diverse origins. The data used for my thesis did not permit a further breakdown of the visible minority category. Despite this, my thesis still makes a significant addition to our understanding of ethno-racial differences in trust. Furthermore, the term itself continues to have important meaning in Canada, because it works effectively as a legal concept in employment law and in developing initiatives related to equal opportunities for all. Furthermore, social geographer Daniel Hiebert notes that the concept is demographically relevant as nonwhites still account for only 16% of the Canadian population (Ward 2008). 2 Aboriginal Peoples are a distinctive group in Canada. Due to the nature of the dataset, a further breakdown of this grouping could not be done. I follow research conventions, by treating them as a single category to be compared with the other groups in the study (See e.g., Guimond 2003; Pendakur 2005). I also employ the term Aboriginal Peoples to represent the heterogeneous grouping, which includes M?tis, Inuit, and culturally diverse grouping of North American Indian. I use the term Aboriginal Peoples instead of the widely-used ?Aboriginal? as a single referent for Indigenous Peoples in Canada. While many Indigenous Peoples self-identify as Aboriginal, Alfred and Corntassell (2005) contend that this identity is highly problematic and inherently unjust. They assert that the Aboriginal identity was created by the Canadian government, as a means of advancing the state?s own legal, cultural, and political interests. According to Alfred and Corntassel (2005: 598-599), Aboriginalism reflects an ongoing colonial process whereby Aboriginal Peoples are forced to cooperate individually and collectively with the Canadian government, out of sheer necessity. Consequently, Aboriginal Peoples strictly see themselves in terms of their political-legal identity, rather than in terms of the social and cultural ties to their Indigenous communities or cultures. Moreover, while it is important to articulate an Indigenous identity that is dynamic and pliant across time and space, Corntassel believes that there is value in establishing a widely-accepted and consistent definition of a collective group identity, as the practice of 3 is often a misconception that Canada?s Aboriginal Peoples are a homogeneous group. However, this is far from the case. Aboriginal Peoples have distinctive legal, social, and cultural differences (Corntassel 2003; Guimond 2003). Academics and various international organisations have developed their own respective definitional frameworks for Aboriginal Peoples, in an attempt to provide a term that is both inclusive of and sensitive to the complex history and unique ancestry of Aboriginal Peoples. However, as Corntassel observes, consensus on the question of ?who is indigenous? (2003: 75) remains elusive, and existing definitions that seek to capture the elements of an authentic indigenous identity are incomplete and highly politicised endeavours (2003: 78). In light of this significant demographic reality, Canadian social scientists have been interested for some time in exploring the implications of ethno-racial3 diversity for explaining a number of important social processes. Among these crucial questions are the extent to which ethnic groups differ in their involvement and inclusion within the major institutions of Canadian society. Some of the central indicators of ethnic incorporation that have been considered in previous research are socioeconomic status attainment, social capital mobilization, civic or voluntary activity, and expressed feelings of social cohesion, unlimited self-identification for Aboriginal Peoples can be problematic (for further discussion see Corntassel 2003). 3 I use the term ?ethno-racial? as a single broad referent to denote both ethnic and racial background. This is a common practice in contemporary ethnic and racial discourse (Richomme 2009: 1-3; Yanow 2003: viii). However, it is important to note that race and ethnicity are multifaceted concepts and that each concept has its own attendant meanings. This term also encompasses the ancestral heritage of Aboriginal Peoples. The usage of ethno-racial is not meant to gloss over important political claims and rights of various groups, including Aboriginal Peoples. Since my focus is on culturally distinct groups and not on issues of political identity, I include Aboriginal Peoples, as with all other distinctive groups, under this ethno-racial concept. 4 connectedness, or inclusion within the larger society (see, e.g., Aizlewood and Pendakur 2006; Bevelander and Pendakur 2007; Breton 1964, 1990, 2005; Breton, Hartmann, Lennards, and Reed 2004; Breton, Isajiw, Kalbach and Reitz 1990; Fong and Ooka 2006; Hou and Wu 2009; Kazemipur 2006; Mata and Pendakur 2010; Roth, Seidel, Ma, and Lo 2012; Reitz 1980a, 1980b; Reitz, Breton, Dion, and Dion 2009; Wu, Hou, and Schimmle 2011). The purpose of this doctoral thesis is to extend our knowledge and understanding of social cohesion and inclusion among Canada?s ethnic and racial groups, by assessing ethno-racial differences in two key indicators of cohesion or connectedness. The first of these is trust in other people, or what I refer to in the thesis as ?social trust?; the second is level of confidence in the society?s major political institutions, or what is referred to in the thesis as ?political trust.? For some time now, eminent scholars in the field of ethnic relations have concluded that these two forms of trust represent the two most important ?facets? or ?dimensions? of social cohesion in Canadian society (Breton, Reitz, and Valentine 1980: 3; Soroka, Johnston, and Banting 2007: 282). Social cohesion refers to the willingness of members of a society to cooperate with each other, in order to ensure the welfare of all its members. The thinking of these writers appears to be that a society coheres or holds together in large part because people are able to trust each other and are also able to trust in the major organizations that shape or structure their lives (for discussion see, e.g., Jenson 1998; Jeannotte 2000). The principal goal is to determine the extent to which three related but distinct sets of explanatory variables help us to understand ethno-racial differences in both social and 5 political trust in Canada. The three types of factors subsume: (1) a set of three ethno-cultural4 ?markers? or correlates of ethno-racial background, including religion, language, and immigration status; (2) two important socioeconomic or class-related influences, namely, education and income; and (3) two significant indicators of social engagement or social connection, i.e., voluntary association activity and social network composition. My approach is to treat the three categories of factors as competing but complementary explanations for why some ethno-racial groups in Canada express more or less trust than others. My central interest is in determining which of these sets of explanations is most effective in accounting for differences in social and political trust across a range of ethno-racial groups, while also assessing the main effects of each explanatory factor on the two types of trust. The full models in the analysis also include region, age, and gender as additional control variables, for reasons that are outlined later in Chapter 1. The main research questions to be addressed in the thesis are as follows: 1. Are there significant ethno-racial differences in social and political trust? In particular, how do visible minorities, French Canadians, and Aboriginal Peoples compare with others in their levels of trust? 2. Are ethno-racial differences in social and political trust mainly or partly attributable to the three significant markers of cultural distinctiveness that I identify in this chapter: 4 The term ?ethno-cultural? is used in the thesis to distinguish these three factors from ethno-racial background per se, and to signify that these three factors amount to cultural variables that also have clear empirical or conceptual overlaps with ethnicity. While not a relevant explanatory factor for groups like the French and Aboriginal Peoples, who are overwhelmingly Canadian-born, it proves to be quite relevant for understanding trust patterns among other groups, especially visible minorities. 6 immigration status5, religion, and language? That is, do those who are the most distinctive from the British and other established groups on these three ethno-cultural dimensions express lower levels of social and/or political trust than other Canadians? Here I allow for the possibility that the patterns could be different for the two types of trust, with culturally distinctive groups being lower on social trust but higher on political trust than other Canadians. 3. Are ethno-racial differences in trust mainly or partly attributable to differences in the socioeconomic status attainment of these groups, as measured by education and income? Is it the case, in particular, that those groups who have higher education and higher incomes are more likely than those with lower education and lower incomes to express more social and political trust, because of the greater potential for social integration provided by these advantages, resources, and successes? 4. Are ethno-racial differences in trust largely or partly due to differences in voluntary association activity? More specifically, are some ethno-racial groups more likely than others to belong to formal voluntary organizations, which could aid in breaking down social boundaries and undermining feelings of social distance, thereby increasing levels of social trust and political trust in these groups? The underlying assumption here is that experience with these organizations on average enhances people?s feelings of being accepted within the larger Canadian community, making them more trusting of others and of the society?s major public institutions. 5 Immigration status is used as a relevant explanatory control for groups other than the French and Aboriginal Peoples, the large majority of whom are Canadian-born. This factor proves especially useful, for understanding patterns of trust for visible minorities. 7 5. Are ethno-racial differences in trust mainly or partly due to differences in social network composition? More specifically, are those groups whose friendship networks include people who are ethnically different from themselves more likely to express feelings of social and political trust than are other groups? The rationale is that having a diversity of friendships is more likely to establish bridging connections that span across ethnic communities, leading to a greater sense of trust in others, and ultimately trust in public institutions as well. 6. Are the patterns of ethno-racial differences in social and political trust affected by controls for region, age, and gender? My contributions to new knowledge come in several ways. To my knowledge, my analysis is the first to consider the simultaneous influences of all these factors in a multivariate design, with the express goal of comparing the explanatory value of the three different kinds of explanations derivable from the theoretical and research literature. My research also provides more insight into how ethno-racial groups vary in their confidence in public institutions, which has not been fully explored in the current sociological literature. My thesis also shows empirically the uniqueness of Aboriginal Peoples in Canada. In contrast to visible minorities and French/Canadians, a different set of social distance/social boundary processes seems to be at work for Aboriginal Peoples. The trust ?divide? between Aboriginal Peoples and other Canadians appears to be rooted in long-standing problems, including colonialism, paternalism, and racism. Lastly, my thesis is also one of the only analyses yet to be conducted in Canada, in which these two quite distinct types of trust are considered side by side. As demonstrated below, the joint investigations of social trust and political trust reveal quite different outcomes, especially in regard to which ethno-racial groups exhibit more or less trust in Canadian society. The issue of interpersonal trust is 8 important to study for, indeed, leading writers like Giddens, Putnam, and even Durkheim, assert that human interactions, especially positive and constructive interactions that are fundamental to and constitutive of society itself, are not possible without a moderately high degree of interpersonal trust. Before proceeding, I should also comment on what I believe are important linkages between the issues of trust that I explore within Canada?s population and the problem of ethnic or racial prejudice, intolerance, or discrimination. Although my thesis project cannot deal directly with the question of prejudice and discrimination, my analysis is in the same general tradition as the existing literature on this important topic. Furthermore, I believe that my findings are relevant for understanding some of the causes and consequences of prejudice and discrimination, especially toward minority groups. In particular, I suggest that one of the main underlying factors for understanding ethno-racial differences in trust is the influence that experiences with prejudice and discrimination can have on the extent to which minority groups come to trust other people and to rely on Canada?s major public institutions. As I discuss later in this chapter, Canada has a reputation in some circles as an exemplary multicultural society, where all people are encouraged to retain their cultural heritage and identity if they choose to do so, and where all of us are expected to accept, and indeed celebrate, the diversity of our population. However, previous research on whether this reputation is deserved has produced quite mixed results. In a 2010 joint CBC-Environics Research Group poll of 2000 people, 28% of survey participants identified Aboriginal Peoples and Muslims as groups that were the most frequent targets for discrimination, while 20% of those surveyed believed that blacks faced discrimination on a recurring basis. Although there is recognition that minority groups encounter prejudice, people?s perception 9 of its prevalence in Canada is quite restrained. In fact, the poll found a dramatic decline in perceptions of pervasive discrimination against blacks (12%), Aboriginal Peoples (9%), and Muslims (9%) since a similar poll done in 2006. Various advocacy groups believe that the decrease in perceived discrimination does not correspond to the actual reality of minorities? experiences with racism. Instead, this may reflect how our pride in multiculturalism shapes our understanding of racism in Canadian life. As Reitz explains: ?Canadians have as part of their self-image the belief in being inclusive, open, multicultural... [Hence], it would be inconsistent to then believe that there's pervasive discrimination? (quoted in Hildebrandt 2010). In one of the most complete analyses of this subject to date, Reitz and Breton (1994) determined that, in many respects, the claim that Canadians are especially progressive or enlightened about ethnic and race relations is overstated; they conclude, in particular, that the perception of Canadians as more ethnically tolerant than Americans is largely an ?illusion of difference? (Reitz and Breton 1994). For example, Reitz and Breton found that, while Canadians are more supportive than Americans of inter-racial marriage, and are also somewhat more supportive of immigration, Canadians are also more likely than Americans to expect immigrants to assimilate to the county?s dominant values, with the two peoples being very much alike on a range of other issues (Reitz and Breton 1994: 28, 81). Similarly, more recent analyses indicate that Canadians are somewhat more tolerant of minority group rights than are Americans in some areas, e.g., in their willingness to accept immigrants or racial minorities as neighbours (Andersen and Milligan 2009: 398; Grabb and Curtis 2005: 212); at the same time, however, studies indicate that the level of racial discrimination in the 10 job market is substantial in Canada, and is almost identical to that found in comparable studies in the United States (Henry and Ginsberg 1985; Henry 2004). The treatment of Aboriginal Peoples also appears to be much the same in both countries, with the United States actually being superior on such questions as legal rights and sovereignty for Aboriginal Peoples (Grabb and Curtis 2005: 203-205). Regarding Canada?s Aboriginal Peoples, Satzewich reviews research showing that, despite expressing some sympathy toward the plight of Aboriginal Peoples, Canadians have little knowledge of the issues involved, do not put a high priority on these issues, and are strongly opposed to the protest movements associated with Aboriginal Peoples (Satzewich 2011: 53; also Ponting and Kiely 1997). Given these kinds of findings, it seems clear that problems of prejudice, discrimination, and intolerance, especially toward ethnic or racial minorities, continue to be a part of the Canadian reality (see especially Satzewich 2011). I see my research on ethno-racial differences in social and political trust as adding further insights, albeit indirectly, into this crucial aspect of ethnic and race relations in Canada. That is, to the extent that my research shows substantial differences in trust between ethno-racial minorities, on the one hand, and more established or dominant groupings, on the other hand, my analysis should also enhance our understanding of the processes and mechanisms through which minorities come to be perceived and treated differently by other Canadians. I expect that the ethno-racial differences in trust that I identify will tend to parallel and partly reflect that patterns of discrimination and prejudice experienced by minority groups. This chapter begins with a brief review of studies that have considered the linkages between ethnic diversity and trust in other societies, especially the United States and Europe. 11 While it is not the purpose of my thesis research to test directly the contextual effects of ethnic heterogeneity on trust, this review provides a useful backdrop for understanding individual-level patterns of trust, which is the main focus of the present analysis. I then consider the Canadian experience with ethnic diversity, beginning with a short assessment of the country?s multiculturalism and immigration policies, and a discussion of the concerns that some observers have recently expressed about multiculturalism in Canada. Here I also point out the unique situation that Aboriginal Peoples occupy in the larger discussion of ethnic diversity in Canada. In the next section of Chapter 1, I outline important theoretical issues that inform the analysis in the rest of the thesis. Of special significance here are the concepts of ?social distance? and ?social boundaries.? I suggest that the application of these concepts leads to a set of predictions about which ethnic or racial minorities face the greatest challenges in developing a sense of trust and connectedness, both to their fellow citizens and to their nation?s major public institutions. With this theoretical backdrop, I then consider the results of leading research on the relationship between ethno-racial background and trust in Canada and elsewhere. The initial review of research focuses more specifically on social trust, which has received far greater attention than political trust in the existing literature on ethnic differences. Previous studies of ethnic variations in political trust are then considered in the subsequent section. In the process of reviewing these studies, I also indicate instances in which earlier research has touched on the three key sets of explanatory factors noted above, i.e., ethno-cultural factors --- religion, language, and immigration status ---, socioeconomic or class factors, and social engagement or social network influences. As a general comment, I should point out that most 12 previous analyses have not taken into account these three sets of factors, and the remaining studies have included controls for some of the variables but not others. In the final section of the chapter, I discuss the measures used to assess the effects of the three groups of explanatory variables, as well as important additional control variables that will be incorporated into the analysis. Chapter 1 concludes by delineating the main research questions to be addressed in the thesis. ETHNIC DIVERSITY AND TRUST: INTERNATIONAL PATTERNS Putnam on Ethnic Diversity and Trust in the United States While it is typical of many scholars to emphasize the virtues of diversity for modern societies, Robert Putnam, who is probably the leading writer on social cohesion and social engagement in the United States, has suggested that ethnic and racial heterogeneity can have potentially negative effects, both for social involvement and for various dimensions of social capital. In particular, Putnam has raised some troubling questions about increased ethnic and racial diversity for a vibrant civil society (Putnam 2007: 137, 149, 150,157). It was in Bowling Alone (2000) that Putnam first examined in detail the possible factors contributing to what he saw as a precipitous decline in the levels of civic activity in the United States. Specifically, Putnam discusses the importance of social capital as the marker of more socially cohesive societies (1995: 67). He conceives of social capital as primarily related to value and belief orientations regarding trust (Putnam 1995: 67; see also Johnston and Soroka 13 2001: 31), and defines social capital as ?social networks and norms of reciprocity and trustworthiness? (2007: 42; see also 1995: 67). The acquisition of social capital, in its different forms, is said to enable people to work collectively in the pursuit of common goals. Putnam identifies the propensity and ability to form and join associations as central to the creation of social capital, as judged by levels of interpersonal trust, understanding, cooperation, and solidarity. More recently, in his first systematic attempt to address the issue of race and ethnicity, Putnam (2007) has considered how increasing multiculturalism and diversity within a sample of American communities may affect the formation of social capital as he defines it. Using data from the Social Capital Community Benchmark Survey, which included 41 different communities across the United States, he advances and tests three competing hypotheses concerning diversity?s effects on social trust (2007: 141). First, according to the contact hypothesis, diversity fosters inter-ethnic tolerance and social solidarity. This argument is based on the premise that increased contact among different ethnic groups allows individuals to overcome the initial ignorance, wariness, or reluctance that may exist regarding those who are different from themselves. This increases people?s trust in others, leading to a noticeable ?reduction of ethnocentric attitudes and development of out-group trust and solidarity? (2007: 142). Thus, diversity erodes inter-group differences, producing ?out-group solidarity or bridging forms of social capital? (2007: 147). In contrast, Putnam poses a second explanation, the conflict hypothesis. This prediction is that ethnic or racial diversity tends to exacerbate inter-group distinctions. Inter-group differences can strengthen in-group solidarity and ?bonding? forms of social capital, which 14 tie people of the same ethnic group together in a common identity (Putnam 2007: 145). However, there is a danger that such bonding can also promote the building of barriers rather than bridges between different ethnic or racial communities, leading to ethnocentrism. As a result, people?s interactions are limited mainly to other people who are similar to themselves, which may actually promote distrust between groups, increase social exclusion, and thereby hamper public-minded action (Putnam 2007: 145; see also Ramakrishnan and Bloemraad 2008). A third hypothesis, which Putnam tentatively labels ?constrict theory? (Putnam 2007: 149), suggests that ethnic heterogeneity may have a joint effect, reducing not only the bridging connections across different groups but also the bonding connections and levels of trust within individual groups. It is this possibility that is the most worrisome in Putnam?s view, since it means that residents of ethnically diverse communities are not engaging in activities that either bridge or bond. Involvement in bridging networks of voluntary associations promotes integration into the wider society, while bonding networks of association activity reinforce social boundaries and preserve distinct ethnic or racial identities (Breton et al. 2004: 189, 195; see also Dekker and Uslaner 2001: 180; Onyx and Bullen 2001: 46; Putnam 2007: 143). Putnam presents evidence that is primarily from the United States, and that appears to support the second and third hypotheses. On one question about inter-racial trust, which asked respondents how much they trusted people of a different racial background, he finds a strong negative relationship between inter-racial trust and ethnic heterogeneity, which seems consistent with the conflict argument. On a second measure of social capital, assessing the trust that individuals feel toward the people who live in their neighbourhoods, Putnam 15 likewise finds that, in more diverse neighbourhoods, people trust their neighbours less. Finally, on a measure of intra-racial trust, which considered respondents? trust in people of their own race, he showed that such trust is lower in more ethnically diverse settings, which appears to be consistent with the constrict hypothesis (2007: 149). Overall, Putnam believes that his findings indicate there is less interpersonal trust both within and across ethnically and racially heterogeneous communities, with the result that people willfully ?hunker down? and retreat from civic activity in such communities (Putnam 2007: 150, 157). This means decreased social cohesion, not only among out-group members or people who are different, but also within homophilous groups (2007: 149). According to Putnam, people who live in more diverse areas have less trust in their political figures and neighbours, lower expectations of receiving help from others in resolving collective problems, decreased levels of voter registration, and a lower propensity to work together on community endeavours (2007: 149-51). He also cites other studies from the United States and elsewhere which show that ethnic diversity is associated with lower social trust and cohesion (Alesina and La Ferrara 2000, 2002; Anderson and Paskeviciute 2006; Costa and Kahn 2003; Hooghe, Reeskens, Stolle, and Trappers 2006). However, some researchers have raised questions about this finding, including a recent review of American research, which concluded that the negative effect of ethnic heterogeneity on trust is ?attenuated by more frequent and substantive cross-racial/ethnic interactions,? or what Putnam might refer to as ?bridging? social ties (Smith 2010: 470). A similar conclusion is drawn by Stolle and colleagues in an analysis that included samples from both Canada and the United States (Stolle, Soroka, and Johnston 2008). Most recently, another leading researcher, in a comparison of the United States and the United Kingdom, 16 has concluded that the supposed negative effect of ethnic diversity on trust within communities is largely explicable by residential segregation, and not heterogeneity itself; he argues further that the establishment of ?close personal ties across ethnic groups in integrated diverse communities builds trust? rather than undermining it (Uslaner 2011: 221; for further debate and discussion, see Delhey and Newton 2005; Fieldhouse and Cutts 2010; Grabb 1980; Hooghe, Powers and Ellison 1995; Kesler and Bloemraad 2010; Reeskens, Stolle, and Trappers 2009; Rotolo 2000; Wu et al. 2011). It may be because of these kinds of considerations that Putnam believes the negative effects on social solidarity of increased ethnic diversity tend to be short-term or medium-term, and that ethnic diversity can indeed provide long-term cultural and economic benefits. Ultimately, he suggests that diversity through immigration can be desirable, provided that ?successful immigrant societies create new forms of solidarity that dampen the negative effects of diversity by creating new encompassing identities? (2007: 137-138). My reading of Putnam?s perspective is that he believes diversity undermines social capital formation in a way that is compositional, or what Johnston and Soroka would refer to as ?congenital? (Johnston and Soroka 2001: 32). That is, Putnam sees the development of interpersonal trust as contingent upon people?s own civic actions, suggesting that it is individuals themselves who are most responsible for redressing problems of low social capital in their communities (see also de Hart 2001: 98). We can infer that Putnam sees increased civic education as the best means for ethnic or racial groups that are deficient in social capital to attain it. He argues that immigrants, in particular, can increase their levels of interpersonal trust through civic involvement. Putnam also calls upon other agents, including local churches and community organizations, to do their part in integrating new immigrants: 17 ?tolerance for difference is but a first step. To strengthen shared identities we need more opportunities for meaningful interaction across ethnic lines ... enabling us all to become comfortable with diversity? (Putnam 2007: 164). The implicit assumption in Putnam?s research is that if people want more trust in others and cooperation in their lives, they should join associations, because association members tend to develop more social trust than do non-members. To some critics, Putnam?s proposed solutions for enhancing trust may seem rather simplistic, to the extent that they may overlook some of the significant structural challenges that disadvantaged ethnic and racial groups face in seeking to attain social capital. For example, Portes notes that ?the alleged decline in social capital is put squarely on [such individual factors as] the leisure behavior of the masses, rather than on the economic and political changes wrought by the corporate and governmental establishment? (Portes 1998: 19; see also Hero 2003; Portney and Berry 1997). At a minimum, then, it would seem necessary to assess the potential positive effects of civic engagement on trust as part of a larger analysis that includes an assessment of how socioeconomic factors, as well as ethno-racial correlates such as religion, immigration status, and language, influence patterns of trust among different minority groups. Notwithstanding these concerns, however, Putnam?s postulations about the role of civic education and social networks raise interesting questions about the possible links between ethnic diversity and trust. Ethnic Diversity and Trust in European Nations While Putnam`s work has focused attention on the issue of diversity and trust in the United States, similar questions about heterogeneity and the growth of multiculturalism are being 18 posed in many European states as well. In several of these societies, increased diversity has appeared to intensify inter-ethnic tensions. Public discourse in Europe suggests considerable uneasiness about diversity. For example, in the 2003 European Social Survey, 58 percent of people in the fifteen European Union (EU) member states agreed with the idea that minorities posed a collective (economic and cultural) threat (Coenders, Lubbers, and Scheepers 2005: 7, 10, 15). The same study revealed that 55 percent of respondents in EU countries harboured a ?resistance to diversity,? or had a clear preference for ?a monocultural society in which people share the same customs and traditions? (Coenders, Lubbers, and Scheepers 2005: 12). Consistent with these findings are the results from the 2006 Global Pew Attitudinal Survey on Muslims in Europe. The Pew Survey indicates that, while a serious backlash against Muslims has not yet occurred in European countries, most people do express doubts about the desire of Muslims to integrate into mainstream society, and feel that most Muslims want to remain distinct from the larger population (Kohut, Allen, and Wilke 2006: 8; see also Bleich 2009). Given the generally cautious attitudes of many Europeans toward the question of minority integration, it may not be surprising that the political agenda in many European nations is pulling away from endorsing multiculturalism, and toward greater support for the civic and cultural assimilation of minorities (Foner 2008: 407, 408; Hansen 2007; Joppke 2007; Koopmans 2010: 1, 20; Michalowksi 2003: 3). Several European member states have established integration programs since the 1990s, in an attempt to address the problems faced by ethnic ghettos in urban centres. These problems include lower educational achievement for second-generation immigrants, poor language skills among both recent arrivals and migrants who have lived in the host country for a longer time, and high joblessness and state 19 dependency among immigrants (Michalowski 2003: 67-68). The general indication is that many European states are now focusing on more assimilative rather than multicultural policies in dealing with such problems (Brubaker 2003; Wilkes, Guppy, and Farris 2008; Van Oers 2008; see also Soroka, Johnston, and Banting 2006: 2-3). ETHNIC DIVERSITY AND TRUST IN CANADA Compared with the United States and Europe, the Canadian experience with ethnic diversity and multiculturalism has generally been portrayed as much more positive. It has been argued that Canada?s policies have proven to be more effective in avoiding the problems associated with minority integration. Canada, in fact, is often held to be an especially successful example to other nations of how multiculturalism should work. Policymakers assert that, for other countries that are dealing with issues of increasing diversity, the ?Canadian model? of immigration and integration is a good blueprint to follow. Canada is said to surpass other countries that traditionally have had high levels of immigration, such as Australia and the United States, as well as newer countries of reception, like the Netherlands and Sweden, in actively facilitating the incorporation of minorities (Kesler and Bloemraad 2010; see also Reitz 1998; Wilkes, Guppy, and Farris 2008). For example, Bloemraad (2006) cites Canada?s extensive settlement programs, including employment services and language classes, as contributing factors accounting for the higher and quicker rates of naturalization among immigrants in Canada relative to the US. Because the practices of settlement and incorporation are under the purview of the Canadian state, minority groups have access to an array of government interventions, from basic language training to a number of equality and 20 human rights protections (Bloemraad 2006: 680; Reitz 2006: 5). Two of the most important components of the Canadian approach, which are often cited as reasons behind Canada`s allegedly smoother incorporation of immigrants, are the official policy of multiculturalism and the policies governing immigrant selection. Canada?s Multiculturalism Policy Canada?s multiculturalism policy, as defined in the Multiculturalism Act of 1988, is perhaps the most well known of Canada?s ethnic incorporation programs. Supporters of this policy argue that its general approach is different from more assimilative integration policies, in that immigrants to Canada are encouraged to maintain different aspects of their ethnic and racial heritage, including their religious and cultural practices, dress, food, and ethno-racial affiliations (Kymlicka and Cohen-Almagor 2000: 92). Although Canadian multiculturalism primarily focuses on the social and cultural integration of immigrant groups, the policy is also supposed to help the descendents of immigrants and other cultural groups. The Multiculturalism Act, which was established in 1988, specified the following four ?supports? ??1) promoting contribution to Canada; 2) full participation in Canadian institutions; 3) interchange between groups in the interest of national unity; and 4) acquisition of an official language? (Reitz 2006: 8-9). In order for immigrants to achieve these goals, the Multiculturalism Act provides some government funds and support for ethnic or immigrant associations, accommodations for cultural and religious minorities, and public recognition of minority groups (Bloemraad 2006). Collectively, these institutional supports are intended to cultivate and bolster the connection of minority groups to the wider society. According to Kymlicka (1995, 2001), such state support is helpful in fostering a strong 21 attachment to Canada among immigrant groups, while at the same time celebrating their cultural distinctiveness. Even so, it remains an open question whether the multiculturalism program has fulfilled its stated objectives regarding immigrant inclusion or the accommodation of immigrant distinctiveness within the larger society. There is some reason to doubt, for example, that Canada`s approach has produced results that are markedly different from those in other countries, such as the United States. One of the most comprehensive comparative studies of the US and Canada found that, although Canadians are generally more supportive of immigration than Americans, the two populations were actually similar in the degree to which new immigrants were encouraged to maintain distinct ethnic communities and cultural practices (Reitz and Breton 1994; see also Reitz et al. 2009; Grabb and Curtis 2005). Because the United States has no formal government policy fostering immigrant diversity of any kind (Bloemraad 2006: 680), it is surprising that there apparently is not a larger difference between the two nations in this area. Perhaps, as Reitz (2006) has argued, the power of Canadian multiculturalism lies more in its symbolic importance than in its policy outcomes, because funding for multiculturalism has at times been inconsistent, and not sufficient to improve substantially the social standing of ethnic and racial groups. Furthermore, the multiculturalism policy does little to provide ethnic and racial groups with the weapons necessary to address racial discrimination (Bloemraad 2006: 687). Immigration Policy and the Points Selection System The structure of the Canadian immigration system might also explain why, according to some observers, Canadian immigrants experience an easier social transition than their 22 counterparts elsewhere. In 2011, Canada planned to accept between 240,000 and 265,000 new permanent residents, a figure that is comparable to those for most other years in recent decades; this helps explain why, as reported in the 2011 Census, two-thirds of the people added to the Canadian population since 2006 have been immigrants (Johnson 2012). Canada?s immigrants fall into three major categories: 1) economic, 2) family reunification, and 3) refugees. These categories grew out of policy changes made to the original Immigration Act of 1952, which was modified in 1978 and then followed by another Act, the Immigration and Refugee Protection Act of 2002 (see Boyd and Vickers 2009; Li 2003;). While Canada?s immigrants can enter as refugees, or as part of the government?s family reunification program, the majority of immigrants in most years are chosen using the points selection system, which was first created in 1967. Under this system, aspiring immigrants are awarded points based mainly on economically relevant human capital criteria, including their educational credentials, work experience, and knowledge of one of Canada?s official languages. Points are also given to immigrants who fill the needs of various occupational sectors in health, skilled trades, finance, and resource extraction (Citizenship and Immigration Canada 2009). Prior to the creation of the points system, immigrants to Canada were much more likely to comprise individuals with lower education and fewer skill qualifications (Reitz 2002). Applying the points selection criteria is meant to ensure that immigrants are more likely to integrate well into the Canadian labour market and, hence, into society at large. In contrast to earlier times, when immigration served the purposes of nation-building and regional settlement, in recent years, the economic benefits of immigration, including increasing the nation?s productivity and labour supply, have driven many of 23 Canada?s policies and programs (Reitz 2006: 7; see also Li 2003; Boyd and Vickers 2009). Are Canadians Concerned about Ethnic Diversity and Multiculturalism? For many years now, Canadians have generally expressed considerable popular support for immigration and the general principle of multiculturalism (see, e.g., Grabb and Curtis 2005: 205-206; Reitz 2006: 7). Nevertheless, Canadian attitudes, toward both immigration and multiculturalism, are also marked by some fluctuations and inconsistencies, depending on economic conditions. For example, the opinion polling agency, Environics, has reported data for the period 1972 to 2002, which showed lower public support for multiculturalism and bilingualism programs during periods of recession and high unemployment (Dasko 2005; see also Reitz 2006: 4; Wilkes, Guppy, and Farris 2008: 304, 322). There is also some concern in the Canadian population that multiculturalism may foster ethnic social isolation, and thereby undermine rather than encourage social cohesiveness (Dasko 2005). Quebecers, in particular, have expressed relatively strong opposition to the federal multiculturalism policy since its formation, and the Quebec government generally does not embrace the multiculturalism label (Reitz and Banerjee 2009: 287). This opposition may stem partly from a fear that multiculturalism might weaken Quebec?s distinct status and special partnership with English Canada (Dewing and Leman 2006: 10). Some Quebecers even view the federal multiculturalism policy as a political tactic to ?downgrade? French Canadian cultural rights to the same level as ethnic minorities (Dewing and Leman 2006: 10). The apparent difference between Quebec and the rest of the country on this issue is one indication of the important regional differences that exist, and that need to be addressed when assessing patterns of trust and connectedness among Canada?s ethnic communities. 24 While many commentators praise the multiculturalism policy for its emphasis on the recognition of cultural minorities, others have raised concerns about the direction that Canada?s pursuit of multiculturalism may take us. Jason Kenney, the current federal Minister of Citizenship, Immigration, and Multiculturalism, has suggested that ?pluralism? rather than multiculturalism would be a better approach to dealing with ethno-racial diversity in Canada, since the latter term emphasizes a distinctive ?them? rather than a cohesive ?we? mentality (Delacourt 2009). Kenney has also expressed the hope that the new term ?pluralism? will translate into public policies that help newcomers to integrate more quickly into the Canadian mainstream. In particular, he seems to favour a greater emphasis on what Putnam and others have called bridging connections, involving outward-looking networks that bring together different kinds of people, with relatively less emphasis on bonding connections, which entail inward-looking networks that bring together similar kinds of people. Other critics have voiced more serious objections about the potential negative effects of Canadian immigration and multiculturalism policies. Included here is the claim that multiculturalism promotes a particular brand of identity politics, which focuses on self-entitlement and a preoccupation with perceived racism. Along these lines, the Canadian journalist, Barbara Kay, wrote in The National Post (April 8, 2008) that multiculturalism is Canada?s ?biggest mistake.? She asserts that some aspects of multicultural policy foster a weak commitment to Canadian values of egalitarianism and freedom of speech, and instead encourage a kind of ?postmodern, one size fits all sense of identity,? as well as guilt or even ?self-loathing? among so-called ?heritage Canadians.? Writing in Walrus magazine, Allan Gregg (2006) expressed somewhat similar concerns about the increasing ?atomization of our society along ethnic lines? (cited in Soroka et al. 2006). Such concerns are also evident in 25 earlier work by another Canadian journalist, Richard Gwyn (1995). While Gwyn is generally supportive of multiculturalism, he worries about the danger of having too many policies that could build ?walls? separating individual ethnic communities from each other, with the result that citizens could lose their sense of a common identity and common purpose as Canadians. A final and potentially more divisive set of concerns can be found in comments by observers at the conservative think-tank, the Fraser Institute (e.g., Grubel 2009). In the view of these commentators, Canada?s current immigration program is seriously flawed and must be altered to reduce the number of newcomers to the country. The authors perceive several heavy burdens stemming from Canada?s high immigration levels, including possible threats to national security, high state expenditures on immigrant-related programs, and even increased traffic congestion. These kinds of allegations about the possible negative effects of multiculturalism, ethnic diversity, and immigration in Canada are consistent with the views of some commentators in other countries, including those who have concluded that higher immigration and the recognition of diverse ethnic groups may be contributing to an ?integration crisis? in countries like the Netherlands, France, and Germany (for discussion, see Joppke 2007). Is it possible that a similar integration crisis is now occurring, or is about to take place, in contemporary Canadian society? One way to address this question is to study the opinions and attitudes about multiculturalism and minority immigration that have been posed by the dominant members of the population and by prominent observers. An equally important task, however, and the central approach taken in this thesis, is to assess and understand the views and perceptions of minority groups themselves. I am especially interested to know whether, despite the 26 supposed Canadian emphasis on multicultural acceptance, there are still significant differences between ethno-racial minorities and other Canadians when it comes to levels of trust in other people and in Canada?s major political institutions. The discussion turns now to a consideration of some important ideas in the theoretical literature, which help provide a better understanding of how cultural distinctiveness may act as a key driver of variations in trust across Canada?s ethnic groups. I focus in particular on the two related concepts of ?social distance? and ?social boundaries.? Prior to discussing these ideas, I present in the next section some personal observations about the significance that members of society attach to observable cultural distinctions, and how these often serve as ethno-cultural ?markers? that shape the perceptions and interactions of both minorities and non-minorities in Canada and elsewhere. The Unique Case of Aboriginal Peoples When contemporary observers have commented on the state of multiculturalism and ethnic diversity in Canada, much of their interest has centred on the important role of visible minorities, including the many people in this group who were born outside the country. Considerable attention has also been devoted to the special situation occupied by French Canadians, as a distinct community that has long enriched the ethnic and linguistic makeup of the country. Aboriginal Peoples, however, who can be arrayed alongside visible minorities and the French as one of the three most culturally distinct minorities in Canada, have sometimes been overlooked, or mentioned only in passing, when discussing the fabric of ethnic or racial distinctiveness in our society. This pattern is evident, as well, in major policy 27 developments in this area. For example, the 1969 Official Languages Act, the 1978 Immigration Act, the 1982 Charter of Rights and Freedoms, and the 1988 Multiculturalism Act have been geared largely toward visible minority or French Canadian concerns, and have typically incorporated little or no mention of Aboriginal issues. In other words, then, both social commentators and policy-makers often fail to recognize that, in many respects, Aboriginal Peoples belong in the same conversations about cultural diversity, tolerance, and inequality as do the other groups. Of course, part of the reason why Aboriginal Peoples may not always be discussed in the same context as these other minorities is that they are, in fact, a unique ethno-racial community, unlike any other in the society. First of all, Aboriginal Peoples are the one group whose ancestors comprise the original inhabitants of the country. This means, moreover, that they are almost entirely Canadian-born, and in this way are quite different from visible minorities, for example. At the same time, even though Aboriginal Peoples are similar to French Canadians in being mainly native-born, they are clearly different from the French (though similar to visible minorities), in that their physical characteristics often identify them as a distinct ethno-racial grouping in the society. Unfortunately, another unique factor distinguishing Aboriginal Peoples is that they have endured problems of discrimination, injustice, and ill treatment that are arguably more deep-seated and longstanding than those faced by any other ethno-racial minority in Canada. Many of these problems stem directly from a deplorable historical legacy that began with early French and British colonial policies, and that continued with the actions, or the inactions, of past and present Canadian governments. The forced removal of Aboriginal Peoples from their lands and their resettlement on reserves, the compulsory relocation of Aboriginal 28 children to the now infamous residential schools, and the failure of many political leaders to honour treaty agreements or settle land claims are all part of the sadly unique experiences that Aboriginal Peoples have had with government (e.g., Frideres and Gadacz 2011; Haig-Brown 1988; Menzies 2009). At the societal level, as well, there is no question that, over the years, members of the Canadian population have also played an active part in the process of excluding Aboriginal Peoples. Many individuals have harboured intolerant attitudes and engaged in discriminatory practices when dealing with Aboriginal Peoples in their daily lives, and many have also ignored their serious socioeconomic disadvantages relative to other Canadians (e.g., Satzewich 2011; White, Beavon, and Spence 2007) . As will be shown in the present analysis, this combination of political and social injustice has important implications for explaining or understanding the levels of trust, both in government and in other people, which exist within the Aboriginal Peoples community. THEORETICAL ISSUES: SOCIAL DISTANCE AND SOCIAL BOUNDARIES Some Personal Observations In one of the most eloquent lines from Martin Luther King's famous ?I have a dream? speech, he spoke of his hope that, in a future America, his children would ?not be judged by the color of their skin, but by the content of their character.? This hope, of course, is not unique to minority peoples living in the United States. Despite Canada's popular image as an unusually tolerant and multicultural society, it is clear that here, too, the simple visible facts 29 of having a different skin colour from the majority, of wearing clothes that signify a different religion from the majority, of speaking a language in public that is different from that of the majority (or speaking the majority language with a discernible accent), can and do have effects on the manner in which people relate to and are treated by others. I can describe numerous experiences of my own, and those of family members and friends, in which identifiable ethno-cultural attributes traceable to race, religion, language, or immigration status led to the feeling of not being fully accepted or included by other citizens of Canada, or by officials employed by major public institutions. For example, it is still a fairly common occurrence for me to be asked by people, ?Where are you from?? or ?What is your citizenship?? and to find that, when I say I am Canadian, to be asked a follow-up question: ?But where are you really from?? These experiences are what motivated me to start this project. I can mention another story that occurred outside Canada, when I was studying in France. One day I introduced myself as a Canadian to a group of international students, only to have a fellow (white) Canadian in attendance add, ?But most Canadians don't really look like her.? Similar experiences clearly occur for many, and probably most, members of racial, religious, linguistic, and immigrant minorities in Canada. Indeed, some people can have such experiences for all of these reasons simultaneously, as could occur, for example, in the case of a Muslim immigrant born in Somalia, who has only recently arrived in Canada and speaks little or no English. Individuals with these experiences are liable to have a deep sense of their distinctiveness from, or lack of close connection to, Canada and its wider population. Such individuals are also likely at times to be treated as ?different,? both by other citizens and by agents of the 30 nation's public organizations. In the latter instance, for example, encounters with welfare workers, police officers, politicians, and other civil servants can play an important part in the extent to which minorities feel that they can trust their government and its representatives to act in their interests. In other words, then, the ?being different? experience, which is common for people who are visibly distinct from the presumed ?typical? Canadian, cannot be overestimated for its power to shape the way that members of ethno-racial minorities come to feel about how they ?fit? inside the larger Canada picture, about their relations with other people living in the society, and about the public institutions that lend structure to their everyday lives. Religion, Language, and Immigration Status as Ethno-cultural Markers As was briefly noted at the beginning of this chapter, it is well-known that, for many years now, and especially in the last several decades, the ethnic landscape of Canada has changed dramatically, with a large and growing influx of people from countries outside Europe, especially from nations in Asia, Africa, the Middle East, the Caribbean, and Latin/Central/South America. People with these non-European backgrounds bring to the wider society a rich and varied set of cultural and related differences. A central issue to consider is whether such differences have important impacts on the way that individuals originating or descending from these nations come to feel about their position in Canada. Everything from distinctive diets, clothing, and hairstyles, to different religious practices and languages are likely to play a role in the degree to which individuals from such backgrounds are able to, or are encouraged to, acculturate, integrate, and be accepted by the host society. 31 Therefore, while my main focus is on how ethno-racial minorities differ in their levels of social and political trust, the concept of ?ethno-racial minority? is itself intimately intertwined with other closely related forms or types of minority status. In the present analysis, I am especially concerned with the statuses that are defined by religion, language, and immigration. Although these statuses are not identical to ethnicity or race, each of these characteristics tends to be correlated with ethnic or racial background, and each of them commonly operates as an observable or otherwise identifiable ?marker? of ethno-cultural distinctiveness. In contemporary Canada, people who belong to religious minorities, linguistic minorities, and immigrant minorities all tend to have substantial overlaps with each other and with those who are identifiable as ethnic or racial minorities. These overlaps mean that factors like religion, language, and immigration status are, in a sense, more than just additional control variables to be considered when doing research on ethnicity in this country. Instead, these factors can and should be seen as component elements in a more general theoretical approach that seeks to uncover key social processes or social mechanisms that may underlie ethno-racial variations on a host of important issues, including the question of trust or connectedness. Empirically, this also means that it is important, where possible, not only to control for the effects of these ethno-cultural markers, but also to determine whether some are more important than others in accounting for why some ethno-racial groupings express less social or political trust than others (for some discussion, see Chai 2005). In particular, if we find lower levels of trust among visible minorities, is this pattern attributable more to their religious distinctiveness, their linguistic background, or their immigration status? Moreover, in the case of other distinctive groups that are predominantly Canadian-born, like 32 the French and Aboriginal Peoples, does either religion or linguistic affiliation play a more important part in accounting for their levels of trust? Hence, a central theoretical and empirical question that is addressed in this thesis is: which of the overlapping and interconnected ethno-cultural markers is more or less significant in accounting for ethno-racial differences in trust and connectedness? In addition, as discussed below, how do the effects of these ethno-cultural influences play out when juxtaposed with other explanatory factors that are alternatives to them but also complementary with them. The most important additional factors considered in the present analysis are socioeconomic or class effects, as well as social engagement and social network influences. These are crucial questions that have only been partially explored, if at all, in the existing literature on ethnic relations and ethnic inequality in Canada. Social Distance and Cultural Distinctiveness My main working hypothesis is that ethno-racial groups that are culturally the most distinct, as indicated by the three ethno-cultural ?markers, will have the greatest challenges in being accepted by the host society, and that this in turn will lead to lower levels of trust among the most distinctive groups. Some of the basis for this hypothesis can be found in the classical work of Emory Bogardus on ?social distance,? which he defines as ?the grades and degrees of understanding and intimacy which characterize pre-social and social relations generally? (1925: 216). Bogardus was the first to operationalize and measure social distance, a concept that has its origins in Simmel's early work on The Stranger (1908), and that was also employed by such early sociological theorists as Robert Park (1924) (for some discussion, see Williams 2007; 33 Karakali 2009). In its simplest form, Bogardus?s theoretical approach suggests that the more distinct that certain groups are from the established or dominant population, the more likely it is that the host population will maintain separation or distance from these groups, sometimes to the point of total exclusion from any form of contact. This hypothesis predicts that, the greater the social distance between groups, the less likely they are to influence each other, or to be connected to each other in a socially meaningful way. It is plausible to argue that this situation is in turn conducive to less trust among those groups who are at the greatest social distance from the dominant population. In his research in the United States, Bogardus found evidence that clearly supported his main hypothesis. In one of his earliest studies, Bogardus (1925) asked a sample of 248 subjects, the vast majority of whom were identified as having northern or western European backgrounds (i.e., ?English,? ?Scotch,? ?Irish,? ?German,? ?French,? or ?Dutch?), to rate 36 different ?races? in terms of how much ?antipathy? they felt towards each of the 36 categories. Bogardus developed a scale in which the acceptability of each group was rated along a spectrum of social distance situations. These situations ranged from close or small distances (e.g., accepting kinship through marriage with each group), to intermediate distances (e.g., acceptance of each group as members of the same club, or as neighbours), to large distances (e.g., whether such groups should be allowed to be citizens or even to visit the country). The results of his early studies showed overwhelmingly that the greatest ?anitpathy,? or lowest acceptability, was directed toward those who were members of the most distinct ethno-racial or ethno-religious minority groups, and hence were the most different from the subjects themselves. Examples of these distinct minorities included those who were ?Turk,? 34 ?Negro,? ?Mulatto,? ?Japanese,? ?Hindu,? ?Mexican,? ?Chinese,? and ?Jewish?. Somewhat more acceptable were eastern and southern European groups (e.g., ?Greek,? ?Hungarian,? ?Servian (sic),? ?Russian,? ?Italian?), who shared with the subjects some cultural characteristics (including being European and mainly Christian, for example), and so were less different from the subjects than the other minorities. Finally, the groups that were considered the most acceptable were those who were most similar to the subjects themselves, that is, those who were northern or western European (e.g., ?Scotch,? ?Norwegian,? ?Dane,? ?English,? ?Swedish,? French,? as well as ?Canadian?) (see Bogardus 1925; 1928). This evidence seemed to show rather convincingly that, at least back in the 1920s, people belonging to dominant groups generally preferred to maintain social distance --- to have limited or in some cases no contact --- with minorities who were cultural different, and especially from those who were the most different. Subsequent research by Bogardus confirmed the same general pattern right up to the 1960s, although the overall social distance between groups did decline somewhat over time (Bogardus 1967). In a more recent American analysis that applied the Bogardus scale for the period 1925 to 1993, Kleg and Yamamoto (1998) found, like Bogardus himself, that social distance did become smaller with time, suggesting a greater acceptance of ethnic and racial minorities compared to the past; nevertheless, these researchers also found that ?the rankings based upon the distance values remained remarkably similar? over time, with results still indicating the same ranking of groups. Hence, those who were the most welcome or accepted by Americans were ?western and northern Europeans? and those who were the least accepted were ?middle Easterners, Orientals, and African Americans,? with eastern and southern Europeans generally in between (Kleg and Yamamoto 1998: 183; see also Parillo and Donoghue 2007). 35 Similar ethnic ?hierarchies? based on social distance have been identified in European nations as well (see, e.g., Bleich 2009; Hagendoorn 1993). A number of studies in Canada tend to show the same patterns and rankings of individuals. Although these studies do not employ the Bogardus social distance scale per se, they find essentially the same ethnic rankings using similar kinds of measures. For example, one analysis based on a 1974 national survey showed that Canadians were more likely to attribute such positive characteristics as being ?hard-working,? ?important,? or genuinely ?Canadian? to people of European origin than to people who were labelled ?Chinese,? ?Canadian Indian,? ?Negro,? or ?East Indian? (Berry, Kalin, and Taylor 1976: 106; see also Li 2008: 24). Another study from the 1970s used national survey data to compare Canadians' subjective rankings of the ?social standing? of different ethnic and racial groups (Pineo, 1977: 154). The subjective rankings revealed that respondents in the survey placed the British at the top, with the French and other Europeans generally in the middle, and ?Japanese,? ?Chinese,? ?Canadian Indians,? and ?Negroes? at the bottom (see also Reitz and Breton 1994: 76). A more recent study based on 1991 national survey data, which was conducted by the Angus Reid opinion poll agency for the Ministry of Multiculturalism and Citizenship, found highly similar rankings, this time using Canadians' subjective assessments of how ?comfortable? they felt about the presence of different ethnic groups in the country. The British again ranked highest on this subjective evaluation, with the French and other Europeans in the middle, and the following groups ranked at the bottom: ?Native Canadians,? ?Chinese,? ?West Indians/Blacks,? ?Arabs,? ?Muslims,? ?Indo-Pakistanis,? and ?Sikhs? (see Li 2003: 26; Satzewich 2011: 12-13). It is also notable in these Canadian studies that, in contrast to most findings from the United States, the French tend to be ranked lower 36 than the British and other northern or western Europeans, a result that suggests the distinctive experience of French Canadians. What this review of research pertaining to social distance does not directly address, of course, are the attitudes and perceptions of the groups who experience lower levels of acceptance, or higher levels of antipathy and exclusion, from the majority population. Do these same groups, in consequence, develop lower levels of trust and connectedness? If so, then Bogardus?s social distance perspective would seem to work well as an explanation for why ethnic groups rank as they do on measures of trust and social cohesion. Later in this chapter, I review research from Canada and other countries that bears on this question. Before doing so, however, I consider the concept of social boundaries and how this idea may also help us to understand or account for ethnic differences in trust and connectedness. Social Boundaries and Cultural Distinctiveness A review of the work done on social boundaries provides further insights concerning how ethnic and racial groups can either narrow or widen the gaps that may separate them. Leading writers in this area have pointed out that the cultural practices, beliefs, and institutions that set groups apart from one another are fundamental features of social life. Lamont and Fournier (1992: 1-2), for example, cite Marcel Mauss?s early observation that ?the domain of social life is essentially a domain of differences.? They note, as well, Mauss?s research with Durkheim, which led these two classical social scientists to conclude that it was because people ?were grouped and thought of themselves in the form of groups, that in their ideas they grouped things together? (Durkheim and Mauss 1903: 82). 37 The study of human societies suggests that there is a near-universal tendency for people to classify and categorize the world, based on ?natural? boundaries that are defined by observable characteristics such as biological sex, for example. Moreover, it appears that the identification or recognition of supposed natural boundaries can give rise to symbolic boundaries that are ?in people?s heads,? and also to socially constructed or socially defined boundaries that are often ?the product of sociopolitical forces? (Lamont and Fournier 1992: 1). Much of the research on the processes that lead to social boundaries has examined how seemingly inconsequential distinctions between people, such as skin colour, for example, can have highly significant implications for shaping patterns of institutionalized social differences and entrenched inequalities (Pachucki, Pendergrass, and Lamont 2007: 331; see also Grabb 2007: 1-2). Hence, work on social boundaries continues to inform our understanding of such crucial questions as cultural group membership, social exclusion, immigration, religion, and identity politics (see, e.g., Ashmore, Deaux, and McLaughlin-Volpe 2004; Bail 2008; Barth 1969; Fleming, Lamont, and Welburn 2012; Fuller 2003; Gans 2007; Lamont and Molnar 2002). Bourdieu?s work on tastes and cultural practices likewise added depth and texture to the broader analysis of symbolic and social boundaries (e.g., Bourdieu 1977; 1984; Bourdieu and Passeron 1972; see also Bryson 1996; Erickson 1996; Hall 1992; Halle 1993; Lamont 1992; Lamont and Lareau 1988; Peterson and Kern 1996; Waters 1999). Some argue that symbolic boundaries can be seen as a kind of social ?tool kit? that groups use to draw lines of identification for the inclusion of some and the exclusion of others (Epstein 1992: 232; see also Abbott 1995; Bail 2008: 38; Barth 1969; Jenkins 1996; Lamont 38 and Molnar 2002: 168). These markers can be articulated in myriad ways that include cultural practices and beliefs, binary classification schemas (e.g., Muslim/non-Muslim, black/white, English-speaker/non-English speaker, etc.), and normative proscriptions or cultural taboos (see Alexander 1992). By comparison, social boundaries have been conceptualized as ?objectified forms of social differences manifested in unequal access to an unequal distribution of resources...and social opportunities? (Lamont and Molnar 2002: 168). There is no question, of course, that symbolic and social boundaries are closely connected. However, whereas social boundaries eventually become institutionalized guides for social interaction, symbolic boundaries can often shift as a result of classification struggles, in which dominant groups seek to preserve the rewards and advantages associated with their group position (Bail 2008: 39). This means that how a group defines itself is often a dialectical process. On the one hand, a group distinguishes itself from others by drawing on a distinctive set of criteria that provide a shared sense of belonging and community. On the other hand, the internal identification of group members with themselves must also be recognized and legitimated by outsiders of the group, in order for an objectified group identity to develop and form (Jenkins 1996). It is then that symbolic boundaries become social boundaries, for then they can translate into identifiable patterns of social exclusion involving racial segregation, class distinctions, and so on (see, e.g., Logan et al. 1996; Massey and Denton 1993; Stinchcombe 1995). The links between this conception of social boundaries and the previous discussion of social distance are apparent. That is, in both conceptions, there is a kind of reciprocal process, in which both the dominant groups, who seek to establish social boundaries and maintain social distance, and the subordinate groups, who are excluded by boundaries and 39 are kept at a distance, come to observe their distinctiveness, and to some degree even accept it as a social reality. It should also be noted that similar arguments can be found in social psychology, where researchers have developed what has come to be known as ?social identity theory.? These writers have likewise attended to the segmentation between ?us? and ?them,? and how the ?pressures to evaluate one?s own group positively through in-group/out-group comparison lead social groups to attempt to differentiate themselves from each other? (Tajfel and Turner 1985: 16?17; see also Brewer and Brown 1998; Hogg and Abrams 1988). The concept of boundaries has been central to the study of ethnic and racial inequality as an alternative to more static cultural or even biological theories of ethnic and racial differences. Particularly important here was the work of Barth (1969), who rejected a view of ethnicity that stressed shared culture, in favour of a more relational approach emphasizing feelings of communality that are defined in opposition to the perceived identity of other racial and ethnic groups (see also Hechter 1975, Horowitz 1985). Among several subsequent contributions inspired by this work is the research of Verdery (1994), who analyzed how the nation-state can act as both a producer of differences between societies and as a homogenizer of populations within societies (see also Starr 1992). Following research by Davis (1991) and others, the study of the production of racial and ethnic classification by the nation-state in the United States has provided a particularly fruitful terrain for studying shifts in the definition of social boundaries. In the past, racial and ethnic classifications forced people to be included in only one category, with the assumption that the groups were always mutually exclusive (Lee 1993). Such assumptions often can have polarizing effects. For example, Shanahan and Olzak (1999) and Gans (1999) have analyzed how polarization arises between whites and nonwhites. These authors note that some individuals will identify strictly with the 40 white population as a means to attain or defend their privileged market position or status, which in turn can lead to tension and even violence toward non-whites. This research, along with previous work by Rieder (1987), points to self-interest as a major source of ethnic conflict, and shows how such conflicts are tied to closure and the protection of acquired privileges. Such an approach, of course, is consistent with the earlier theoretical insights offered by Parkin (1979), who points to ethnicity, class, and various other social factors (e.g., gender, religion) as integral to forms of social closure in systems of inequality. The dynamics of boundary creation can also be seen in studies of racial and ethnic identity construction, which consider how these identities result from a process of self-definition, involving the creation of symbolic boundaries and the assignment of collective identities by others (Cornell and Hartmann 1997, Ch. 4; also Portes and Rumbaut 2001). Sometimes ethno-racial boundaries can become more pronounced or more blurred, depending on the extent to which other kinds of boundaries arise that either are congruent with them or cut across them. These other boundaries can arise around religious differences, linguistic differences, and so on, but a particularly important instance in the literature concerns class boundaries. In one study, for example, Lamont (2000) considered the tendency for white workers to associate blacks with being poor and lacking a work ethic, and for black workers to associate whites with a different set of middle or upper class values emphasizing material self-interest and egotism. Such findings from the United States suggest that class and race may be highly correlated, both symbolically in the minds and perceptions of whites and blacks, and socially in the form of real socioeconomic gaps separating the two racial groupings. On the other hand, where class and race differences do not overlap so closely, as in communities or societies in which there are substantial numbers of whites in 41 the working class and non-whites in the middle or upper class, then social boundaries based on race or ethnicity may be weakened or undermined (for some discussion, see Parkin 1979). There is some evidence to suggest that this can occur, although sometimes in only limited ways (e.g., Anderson 1999; Pattillo-McCoy 1999). Nevertheless, to the extent that class boundaries and ethno-racial boundaries are not congruent with each other, it is plausible to expect that the levels of distrust or exclusion experienced by ethno-racial minorities should be reduced. As discussed further below, these considerations point to the need to take into account socioeconomic differences across ethnic groups when assessing variations in trust and connectedness. I turn now to a review of previous empirical analyses of the relationship between ethno-racial background and trust. This section begins with a review of studies that have addressed this topic in societies other than Canada, followed by a more detailed assessment of leading Canadian research on ethno-racial differences in trust. Research on Ethnicity and Trust Giddens describes trust as \"confidence in the reliability of a person or system, regarding a given set of outcomes or events, where that confidence expresses a faith in the probity or love of another, or in the correctness of abstract principles\"(1990:34). It entails confidence and faith in certain traits of another individual, system or principle. Trust has been associated with values of honesty and fairness (Tonkiss and Passey 1999), and can rest on the degree of faith, predictability, or cooperation existing in a given situation, on expectations that the other will act in ways that are competent and caring, and on perceptions of commitment to a goal and the fulfilment of obligations (Kasperson, Golding, and Tuler 1992). To Simmel, 42 trust is belief in someone, a belief that they possess some consistency upon which one can rely, based on reasons but not explained by them. Simmel believed that ?without the general trust that people have in each other, society itself would disintegrate? (1990: 178) and asserted that trust is ?one of the most important synthetic forces within society (1950: 318). Given the significance of trust in society, very few studies exist in Canada that explain and account for the nature and the scope of ethno-racial differences in trust. Assessing the level of trust among minority groups should presumably be a crucial concern for those who are interested in fully understanding the patterns and problems of ethnic and race relations in contemporary societies. In fact, one of the leading scholars in this area has argued that ?race is the life experience that has the biggest impact on trust? [emphasis in the original] (Uslaner 2002: 91). Smith (2010: 47) has commented similarly that ?race is the most important determinant of trust? in the United States. It is somewhat surprising to discover, then, that relatively few studies have been conducted on the question of whether, at the individual level, ethnic or racial minorities evince higher or lower trust than do non-minorities. This is in contrast to the many studies, which were discussed earlier in this chapter, that have looked at the effects on trust of ethnic diversity or heterogeneity at the contextual level (recall, e.g., Anderson and Paskeviciute 2006; Alesina and La Ferrara 2000; Costa and Kahn 2003; Fieldhouse and Cutts 2010; Hooghe et al. 2006, 2009; Kesler and Bloemraad 2010; Putnam 2007). The two societies in which the topic appears to have received the greatest attention are the United States and Canada. However, the American analyses have tended to be somewhat limited, in that they have focused mainly on comparing whites with blacks, and rarely consider a wider array of minority groups. The American research has shown consistently 43 that blacks express significantly less interpersonal trust than do whites, and that this pattern has existed for many decades (see, e.g., Alesina and La Ferrara 2002; Smith 2010; Uslaner 2002; Wilkes 2011). One other study by Uslaner (2011) has found this same result both in the United States and, to a lesser extent, in the United Kingdom as well. In her extensive review of American research on ethnicity and trust, Smith was able to identify only a few studies that looked beyond black-white differences, noting that the relative lack of broader comparisons may be partly because it is between blacks and whites that the starkest differences in trust are found (Smith 2010: 454). The American analyses that consider a more extensive array of minority groups suggest that blacks have the lowest trust, whites have the highest, and Latinos and Asians tend to fall in between the two extremes (see Patterson 1999; Putnam 2000). The most complete comparison is found in research by Smith (1997), who looked at levels of distrust or ?misanthropy? across 40 different ethnic categories. His results, with controls for other factors, showed that distrust was highest among blacks, followed by people from southern or eastern Europe, Hispanics, Amerindians, people from ?middle? Europe (e.g., France, Germany, Ireland, as well as French Canadians), Asians, and finally people from Britain and northern Europe (i.e., Scandinavia) (Smith 1997: 190). If we recall the rankings of ethnic groups on the Bogardus social distance scale discussed earlier in this chapter, we can see an obvious correspondence in the rankings for trust that were reported by Smith. Smith?s analysis is also one of the few American studies to consider some of the same controls that form a central component of my thesis research. With respect to religion, for example, Smith showed that the highest trust levels were among Jews and the two major Christian categories (Roman Catholic and Protestant), with the lowest trust among those with 44 ?Other? religions and those who espoused no religion (1997: 188). He also reported that socioeconomic factors, specifically income and education, were positively related to trust (Smith 1997: 184,189), which is consistent with most research in this area (see Smith 2010). Smith (1997: 188) determined, as well, that people who were members of voluntary associations expressed more trust in others than did people who had no voluntary association involvements. It appears, however, that it is the Canadian population that has received the closest scrutiny when it comes to assessing differences in trust and related variables across a wider range of ethno-racial groups. I turn now to a more detailed review of the leading Canadian studies that have explored this important question. Breton?s Research Raymond Breton stands as one of the first and most prominent sociologists to consider the question of ethnicity and social cohesion in Canada. In his classic discussion of ?institutional completeness? within ethno-racial communities (Breton 1964), Breton argued that ethnic and immigrant communities often establish their own set of parallel institutions and organizations that operate relatively autonomously from those in the larger society. These entities, which can include both informal networks or groups and more formal voluntary organizations, can have different influences when it comes to integrating or separating ethnic groups from the receiving society. Breton suggests that these influences vary across the minorities themselves, with those groups who are ?more different? from the host population --- on such factors as language, religion, and race --- being more likely than others to develop separate institutions (Breton 1964: 204). He confirmed this pattern for language in some of his early 45 research, in which he compared a sample of Anglophones and Francophones in Montreal from the late 1950s. In this analysis, he also found a length-of-stay effect on ethnic integration for immigrant respondents, whose bridging connections with the host community and its institutions tended to increase over time (Breton 1964: 196). In the latter findings, we can see one of the first indications in Canadian research of the role that immigration status can play in patterns of social cohesion and connectedness. The tendency in Breton?s early work for the ?more different? ethnic groups to establish their own distinct institutions, informal groups, and formal organizations is consistent with the social distance and social boundaries arguments. So, too, is some of the research that he later conducted using a Toronto sample from the 1970s (Breton 1990). Breton found that two ethno-racial minorities (West Indians and Chinese) and one ethno-religious minority (Jews) were the least likely to feel that they were accepted by the established population, as measured by their impressions of neighbours? attitudes (Breton 1990: 199-201). He also showed that two of these same groups, the West Indians and Chinese, were low in ?ethnic-community organization? memberships, but that Jews, in contrast, were very high on such memberships (see also Fong and Ooka 2006). These mixed results leave open the question of whether ethnic groups who are culturally distinct from the dominant population, and who consequently may be less accepted than others, will necessarily be more likely to establish in-group bonding through ethnically based social network involvement or organizational activity. In a more recent study based on national survey data from 1997, Breton and his colleagues also examined the issue of ethnic acceptance or connectedness, by considering what these researchers refer to as Canada?s ?social fabric? (Breton et al. 2004). The results 46 of this study, which did not include systematic multivariate controls, show a generally strong commitment by most Canadians to the four core values of trust in social interactions, fairness, recognition for societal contribution, and sense of belonging. At the same time, however, the findings indicate potential sources of social disintegration in Canada?s social fabric, including some fragmentation along ethno-racial lines. The authors detected no substantial differences in trust across ethnic groups (Breton et al. 2004: 130), but they did report that visible minorities, and also French Canadians, were the least likely to feel that Canada treats them fairly (Breton et al. 2004: 117). In addition, visible minorities and French respondents were the least likely to feel ?very satisfied? that their contributions to the society were recognized (Breton et al. 2004:120). On the fourth criterion, regarding feelings of belonging, the results again showed that visible minorities, and also French-speaking Canadians, were the least positive, expressing the weakest sense of belonging to society as a whole and to their own local communities (Breton et al. 2004: 130-131, 127; see also Nakhaie 2006). The Findings of Soroka and Colleagues We can discern many of the same patterns, along with some complicating differences, in subsequent research that Soroka and colleagues have conducted, using data from the 2002 Economy, Security, and Community Survey (ESCS) and also from the 2004 Canadian Election Survey (CES) (Soroka et al. 2006; see also Johnston and Soroka 2001; Soroka et al. 2007). Their analyses used some measures that were similar to those employed by Breton and his colleagues, including questions about trust and sense of belonging. The researchers were also able to compare the answers to these questions across a range of ethno-racial 47 minorities, although sample size limitations forced them to combine several groups into a smaller number of categories. In contrast to Breton et al.?s (2004) finding of little variation in levels of trust across the ethnic groups that they were able to compare, Soroka et al. (2006: 16) determined that all of the minority groups in their sample expressed less trust than the ?British/northern European? grouping, which they used as their reference category. The lowest levels of trust occurred among Quebec Francophones, Aboriginal Peoples, the ?Caribbean/African? category, and the ?South Asian/Middle Eastern? category (2006: 16). Johnston and Soroka (2001: 38) found similar results using data for a more limited set of ethnic categories in the 1991 World Values Survey. Soroka and colleagues also determined that, even though there was a fairly high sense of belonging to Canada among all the ethnic groups, the British/northern European reference category still ranked highest, with Francophone Quebecers ranking lowest and most racial minority categories, including the Caribbean/African, East Asian, South Asian/Middle Eastern, and Aboriginal respondents, also ranking lower than the reference category (Soroka et al. 2006: 14). These patterns are consistent with those found by Breton et al. (2004) and again show a tendency for ethno-racial minorities to have a comparatively low sense of connection to their society relative to more established groups such as the British and northern Europeans. Finally, the 2006 study by Soroka et al. also included an item on national pride, which could be seen as another possible indicator of people?s connectedness or commitment to their society. For this measure, the lowest levels were again reported by Francophone Quebecers, followed by Aboriginal Peoples (Soroka et al. 2006: 13). In this case, however, one of the 48 minority groupings --- the South Asian/Middle Eastern category --- actually ranked highest, with the other minority categories not much different in national pride from the British/northern European reference category (see also Johnston and Soroka 2001: 40). The latter differences were especially small in analyses that included controls for immigration status, age, and religion. Like some of Breton?s research, then, the controlled analyses by Soroka and colleagues also underscore the importance of taking into account immigration status, as well as religion and age, when trying to understand ethnic variations in cohesion and connectedness among Canada?s ethnic groups (see also Soroka et al. 2007: 187). While not including all of the other potentially important explanatory factors that I consider in the present analysis, the research by Soroka and his co-researchers does point to several of them. Moreover, the findings of these important studies are consistent with the argument that cultural distinctiveness, stemming from key mechanisms associated with social distance and social boundaries, may lie at the root of ethno-racial differences in trust. Reitz and Banerjee?s Research Another prominent body of work on the relationship between ethnicity and sense of trust or connectedness in Canada can be found in research by Reitz and his colleagues. A particularly relevant and significant analysis for my thesis project is Reitz and Banerjee?s (2009) research based on the Ethnic Diversity Survey (EDS), a large nationally representative sample survey of Canadian respondents that was conducted in 2002 (see also Reitz 1980a, 1980b, 1998; Breton and Reitz 1994; Reitz and Banerjee 2007). The EDS incorporates several indicators that parallel the measures used by the other key researchers that we have already discussed. Regarding the issue of social trust, or whether 49 ?most people can be trusted,? Reitz and Banerjee?s findings are close to those of Breton et al. (2004) in showing little difference between visible minorities (48 percent) and non-minorities (50 percent) as a whole (Reitz and Banerjee 2009: 286; see also Nakhaie 2006). Nevertheless, within the broad visible minority category, more detailed comparisons show considerable variation, with the level of trust being relatively high for the Chinese (60 percent), about average for South Asians (49 percent) and well below average for Blacks (just 31 percent). On a second measure, which asked respondents how often they felt ?uncomfortable or out of place in Canada? due to their race or cultural background, Reitz and Banerjee?s findings are consistent with those reported by both Breton et al. (2004) and Soroka et al. (2006). That is, their results show that visible minority groups as a whole (at 37 percent) are more than twice as likely as non-minorities (at 16 percent) to say that they sometimes feel uncomfortable or out of place in Canada; the highest levels are reported by Koreans (49 percent), Filipinos (49 percent), and Blacks (43 percent) (Reitz and Banerjee 2009: 276). Another important result reported by Reitz and Banerjee is that this feeling of being out of place is significantly affected by both nativity and length of time in Canada, which are among the key explanatory factors that I emphasize in my thesis project. Their findings indicate that recent minority immigrants (at 42 percent) are the most likely to say that they feel this way at least some times, compared to a much lower 27 percent among Canadian-born minority respondents (see Reitz and Banerjee 2007). Once again we see the significance that immigration status is likely to have in accounting for ethno-racial variations in social cohesion. 50 Reitz and Banerjee also considered strength of Canadian identity as a potential indicator of social connectedness. Here, as well, the results suggest that visible minorities may feel comparatively less connected to the wider society or polity than non-minorities. While 64 percent of non-minorities expressed a strong sense of Canadian identity, only 34 percent of visible minorities did so, and the level was low across all the individual minority categories that they considered (Reitz and Banerjee 2009: 286). Another measure in their analysis, which may be seen as an indirect indicator, cause, or correlate of trust and connectedness, is perceived discrimination. Reitz and Banerjee found that, when respondents were asked about whether they had previously experienced at least one instance of discrimination or unfair treatment over the previous five years (or the period since their arrival in Canada, if they were new immigrants), about 50 percent of Black respondents, 43 percent of Japanese respondents, and 41 percent of Korean respondents said that they had (Reitz and Banerjee 2009: 276, 278). The figure for all visible minority groups combined was 36 percent, compared to only 11 percent for all non-minority respondents. In other words, there is a clear tendency for those who some have called ?people of colour? to be the most likely to experience or sense discrimination from other Canadians. This pattern is consistent with other findings reported by Reitz and Banerjee from a 2003 Globe and Mail survey, concerning perceptions of prejudice regarding employment opportunities. The latter results showed that ?42 percent of visible minorities think that prejudice affects opportunities, compared with 30 percent of whites? (Reitz and Banerjee 2009: 280; see also Breton et al. 2004: 118; Reitz 1980b: 373). The results for perceived discrimination also revealed the importance of immigration status. In this case, the findings showed that the feeling of being discriminated against was 51 actually more prevalent among Canadian-born visible minority respondents than among those who were either recent immigrants or earlier immigrants. As an example, the proportion who reported that they experienced discrimination was 61 percent for Canadian-born Blacks, compared to less than 50 percent for other Blacks (Reitz and Banerjee 2009: 280). These findings again suggest the need to take into account the explanatory impact of both nativity and time in Canada when assessing ethnic differences in social cohesion and connectedness (recall also Breton 1964; Soroka et al. 2006). Finally, Reitz and Banerjee performed one other set of comparisons from the EDS, focusing on sense of belonging to Canada. In this instance, however, the findings provided no support for the view that minority groups have a greater sense of exclusion or disconnection than other Canadians. Instead, about the same proportion of visible minority respondents (59 percent) and non-minority respondents (55 percent) expressed a sense of belonging. This feeling was especially high among South Asians (65 percent) and Blacks (61 percent). In general, however, and despite some mixed outcomes, the findings of Reitz and Banerjee, along with those of Breton, Soroka, and others, indicate a relatively consistent pattern. The pattern is that those ethno-racial groupings who, following Breton, are ?more different? from the established groups in the host population, often seem to face greater challenges or difficulties when it comes to developing trust in others or a sense of connectedness to the larger society. Many of the groups that face these challenges include those who are classified among Statistics Canada?s ?visible minorities.? The latter term of reference, while it may have some conceptual limitations, does capture well the sense that some people are more likely than 52 others to be readily identifiable as distinctive in regard to certain observable cultural attributes (e.g., physical appearance, manner of dress or speech, the wearing of religious symbols) and that, as a result, are more likely to be treated differently or made to feel separate from others in society. Before proceeding, it is important to remark on the two other groups who, in previous Canadian studies, also tend to express lower levels of trust and connectedness. Neither of these groups is understood as a ?visible minority,? according to the conventional Statistics Canada usage. One is the French, who are a long-established and largely non-immigrant grouping that is typically ?non-visible? in terms of observable criteria such as physical appearance. The second is Aboriginal Peoples, who, of course, represent the original inhabitants of Canada, and also are not a ?visible? minority by the standard Statistics Canada definition, but who are often distinguishable in their physical or cultural attributes from non-minority or ?white? Canadians. Anyone familiar with Canada?s social history will be aware of the unique circumstances that help to explain why each of these two groups, although they are not defined as visible minorities by government agencies, do represent culturally distinctive communities within Canadian society. It is not the purpose of this thesis to outline in detail the various factors that lie behind this distinctiveness, which include historical problems of economic inequality, political subordination, and cultural isolation that have been experienced by both the French and Aboriginal Peoples. Such factors will, of course, are at different points in the analysis and conclusions, and are considered in more depth and detail when seeking to explain why these two groups seem to express comparatively low levels of trust and social connectedness in Canada. What is clear for purposes of the present analysis is that French and Aboriginal 53 Peoples share a fundamental commonality with the nation?s so-called visible minorities; that is, they too have linguistic, religious, or other cultural characteristics that show them to be different in observable ways from the British and/or northern European groups that have predominated in much of the society?s history. It is this common denominator of cultural distinctiveness, broadly understood, that I argue is a key mechanism at work in accounting for the lower levels of trust and connectedness reported in previous studies, both among many visible-minorities and among many French and Aboriginal Canadians. In summary, then, the findings from Canadian and other studies are generally consistent with the view that mechanisms associated with social distance and social boundaries underlie significant ethno-racial differences in social trust. Moreover, although this previous research only partially takes into account the key explanatory factors that are emphasized in my thesis research, those studies that include at least some of these factors generally suggest that the influences of variables like religion, language, immigration status, socioeconomic status, and social engagement could provide part of the answer as to why certain ethno-racial groups express more social trust than others. Ethnicity and Political Trust in Canada and Other Societies This brings us to the discussion of political trust, and the extent to which there are significant ethnic or racial differences on this second important dimension of social cohesion. As pointed out by Grabb et al. (2009b: 380), many social scientists have studied political trust. One of the recurring findings in this literature is that people?s feelings of trust, confidence, or deference regarding their political leaders and institutions have generally declined over the last several decades, and across a wide range of countries that includes Canada, the United 54 States, and many other societies in Europe and elsewhere (see, e.g., Adams 2003; Brooks and Cheng 2001; Catterberg and Moreno 2005; Dalton 2004, 2005; Inglehart 1997; Levi and Stoker 2001; Miller and Listhaug 1990; Nevitte 1996; Norris 1999; Perlin 1997; Putnam 2000;). Nevertheless, in spite of this large body of research, scholars in most countries seem to have devoted little attention to studying ethnic or racial variations in political trust, even less, in fact, than they have to ethno-racial differences in social trust. The main exceptions to this assertion appear, once more, to be found in American and Canadian studies. As was the case for social trust, the American research has focused almost exclusively on black versus white differences in political trust, and has not typically considered a broader range of ethnic or racial categories for comparison. As we might expect, the findings in the United States tend to show that blacks are less trusting of their political leaders and institutions than are whites, perhaps because blacks have less power in the society and are often treated less favourably by government; even so, most of this research shows racial differences that are weak, and considerably weaker than the racial differences for social trust (see, e.g., Abrahamson 1983; Howell and Fagan 1988; Cole 1973; Lewis and Stoker 2001; Nelson 1979: Putnam 2000; Rahn and Rudolph 2005). Some studies actually find no racial differences, or indicate that blacks at times are more trusting or positive about government than are whites (Cantril and Davis Cantril 1999; Emig, Hesse, and Fisher 1996;). One recent analysis, which also included Latinos, likewise found no significant ethno-racial differences in attitudes about the federal electoral system (Nunnally 2011). These findings suggest, then, that racial minorities in the United States, including blacks, are not very different from other 55 Americans in regard to their views about their nation?s governmental agencies and political institutions. There appears, however, to be one exception to this assessment. The most consistent finding indicating lower political trust among American blacks occurs when respondents are asked about one area of government activity: the police and criminal justice. Here the results indicate that blacks report having more negative experiences with law enforcement than other groups and often feel that they are treated differently by the police and the justice system (Carr, Napolitano, and Keating 2007; Tuch and Weitzer; Weitzer and Tuch 2004;). There is some research that has compared a larger range of ethno-racial groups, and it, too, suggests that blacks have the most negative views about criminal justice, followed by Latinos and Native Americans, with Asians and whites being the least negative (Hagan, Shedd, and Payne 2005: 391-394). Canadian analyses of ethnic differences in political trust have also been rather limited to date, with much of the focus being on the ethno-linguistic dichotomy between French and English, or the related regional division between Quebec and ?ROC? (the rest of Canada). Some of the early research on this question found that French Canadians, especially in Quebec, were more likely than other Canadians to express dissatisfaction with their treatment by the federal political system and its leaders, and to feel isolated from federal political institutions (e.g., Roseborough and Breton 1968; Simeon and Elkins 1974; Grabb 1979). However, most recent research has found that French respondents, including those in Quebec, appear if anything to express slightly higher levels of political satisfaction, as measured by such indicators as political efficacy (Henderson 2004: 600) and trust in government (Soroka et al. 2007: 188; also Grabb and Curtis 2005: 279). 56 One study that used a somewhat more detailed set of ethnic categories also showed weak and mixed patterns. That analysis was based on data from the 1991 World Values Survey, and compared political trust across four broad groupings that included ?white non-ethnic,? ?white ethnic,? ?non-white,? and French respondents (Soroka and Johnston 2001). The white non-ethnic category, which was largely composed of people who identified themselves as English Canadians, expressed slightly higher political trust than the others, but the differences across groups were very small (Soroka and Johnston 2001: 39). The most recent Canadian analysis to address the question of ethnicity and political trust appears to be a study by Grabb and colleagues, which used data from the Canadian and American samples of the 2001 World Values Survey (Grabb et al. 2009b). This study relied on indicators of political trust that are similar to those used in my thesis project, based on questions about people?s level of confidence in four major political institutions (i.e., the federal government, the civil service, political parties, and the police). Because of sample size limitations, the ethno-racial comparison in this analysis was again restricted to a simple dichotomy between ?white? and ?non-white? respondents. Like most studies in the US, the findings indicated a lower level of confidence in the police among American non-whites, but only in areas outside the South; in addition, and in contrast to several Canadian analyses (e.g., Hagan, and Macmillan 1997; Roberts and Doob 1997; Wortley 1996; Wortley, Wortley and Tanner 2003;), the Canadian sample considered in this study revealed no differences between whites and non-whites in regard to their confidence in the police. The Canadian findings showed, moreover, that non-whites were significantly more likely than whites to express confidence in the other three political institutions that were considered (Grabb et al. 2009b: 386-387). The authors tentatively concluded that the generally higher political trust 57 among non-white Canadians could be partly due to their being majority foreign-born people, who typically ?have actively and willingly chosen to come to Canada,? and so might be more likely to see the government in a positive light (Grabb et al. 2009b: 392). Overall, the Canadian evidence for ethno-racial differences in political trust, like the American evidence, tends to show rather weak and sometimes inconsistent patterns. Especially in recent research, the most consistent finding is that, with the possible exception of their views about the police, Canada?s French and non-white minority groups seem, if anything, to have slightly more trust and confidence in the country?s major political institutions. This pattern stands in clear contrast to the evidence on social trust, which, as we have seen, generally suggests that French, Aboriginal, and visible minority Canadians all have lower levels of trust than do other groups. One possible explanation for this apparent discrepancy in the patterns for social trust and political trust in Canada could relate to the existence of social distance and social boundary mechanisms, and their tendency to distinguish and separate ethnic groups from one another. To the extent that these mechanisms operate, they could contribute to comparatively less trust in other people among minority members, while at the same time leading minority groups to perceive the government and most of its agencies as important buttresses for protecting their rights and interests. The latter process could be especially important for immigrants, many of whom might see the Canadian government as generally sympathetic and responsive to their needs, particularly when compared to the government agencies in their countries of origin. Such possibilities will be considered again in greater detail later in the present analysis. 58 KEY EXPLANATORY VARIABLES As discussed previously at various points in Chapter 1, there are three groups of explanatory controls that I wish to assess for their relative roles in accounting for ethno-racial differences in social and political trust. In this section my purpose is to comment briefly on each of these sets of factors and outline my expectations for how they relate to ethno-racial variations in trust. Ethnocultural Markers: Immigration Status, Religion, and Language Immigration Status: Nativity and Time in Canada Our assessment of earlier research indicates that nativity, i.e., being born in Canada or elsewhere, can have a substantial influence on the social trust and sense of social cohesion of the population, with the foreign-born expressing lower levels on several, though not all, measures (recall, e.g., Breton 1964; Soroka et al. 2006, 2007; Reitz and Banerjee 2009). Given that certain ethnic groups in Canada, especially visible minorities, are more likely to be foreign-born than are other ethnic groups, it is necessary to determine if nativity is an especially important factor for understanding why some ethno-racial groups are less trusting than others when it comes to their views about their fellow Canadians. However, the expectation based on previous studies is that respondents born outside Canada will not be less trusting or confident regarding major government organizations, with the possible exception of the police. 59 In addition, it is essential to sub-divide the foreign-born respondents in my investigation, by taking into account the length of time that they have resided in Canada. That is, we need to consider whether the length of stay of immigrant Canadians has an additional impact on the trust levels of different ethnic groups. Soroka and colleagues found that, in assessing ethnic differences in social integration, ?much of the apparent difference is driven by the fact that many members of ethnic minorities are first generation immigrants,? and that ?the differences between immigrant and native-born Canadians narrow considerably the longer the newcomers are in Canada? (Soroka et al. 2006: 4). It is also notable in their research that two native-born, rather than foreign-born, groupings, i.e., French Canadians and Aboriginal Peoples, generally express the lowest levels of integration in Canadian society (Soroka et al. 2006: 5, 24; Johnston and Soroka 2001: 40). As touched on previously, Breton?s early work (1964) also found that immigrants were more likely to be integrated into the larger society with increased time in Canada. Regarding economic integration, as judged by economic indicators like income, Hou, Balakrishnan, and Jurdi (2009: 268) have shown, as well, that the longer immigrants are in Canada, the more likely they are to catch up with, and at times surpass, native-born Canadians (see also Hou and Balakrishnan 1996; Reitz and Banerjee 2009: 280). Reitz and Banerjee?s (2009) findings also suggest that the level of perceived vulnerability expressed by both visible minorities and non-minorities generally declines as we move from recent immigrants, to immigrants who have been in Canada longer, and finally to the Canadian-born, and this pattern could be congruent with changing levels of trust. On the other hand, Smith (1997: 188-189) did not find clear evidence of this same period-of-immigration effect when he looked at trust in the US, (1997: 188-189). This suggests that 60 the integration of different ethnic groups need not always follow a simple linear path, but instead can entail more intricate patterns, depending on the measure of social cohesion that is being considered. This has led some to question ?straight-line? assimilation models of integration, which view immigrant adaptation as occurring in a positive ?linear? fashion (Gans 1992: 174; 1997). These classic models, which are drawn mainly from American research, implicitly assume that, with time, subsequent generations become virtually indistinguishable from the receiving society, shedding their distinct cultural practices, identities, values, and languages (e.g., Gordon 1964; Park 1930; Warner and Srole 1945). American scholars have recently raised doubts about linear depictions of immigrant integration (Portes 1995; Portes and Zhou 1993; Ramakrishnan and Espenshade 2001; Waters 1994;). Therefore, while there is some indication in Canadian studies that longer-term immigrants will have higher social trust than more recent immigrants, it is also possible that I will find a different result, given the doubts raised by American researchers on this topic. Religion It is clear, especially in the present day, that a thorough understanding of ethnic relations in Canada requires greater attention to the increasing overlaps between ethnicity and religion. Until recently, there has been little systematic research on whether and how religious affiliation and religious commitment influence the level of social cohesion among Canada?s ethnic and racial groups (see Breton 2012: vii, 4, 18). While there was a time when Canadian society was predominantly Christian in religious composition, the proportion of the population who belong to non-Christian faiths has risen appreciably in the last few decades, 61 especially among visible minorities (see, e.g., Bibby 1993, 2002; Breton 2012; Statistics Canada 2005b;). The shifting religious landscape has been due in large part to significant changes in immigration policy, which have increased substantially the number of newcomers from non-European countries. Using evidence from the 2002 Ethnic Diversity Survey, Breton (2012) has shown that the vast majority of Buddhists in Canada come from China and Southeast Asia, while the large majority of Muslims come from Arab/West Asian and South Asian societies. Hence, there is now a relatively close intertwining of ethno-racial background and religious affiliation for a significant and growing segment of the Canadian population. The same linkage of ethnicity and religion is evident for other groups as well. For example, many followers of Judaism, when asked about their ethnicity on Canadian Censuses, are apt to declare their ethnic background as Jewish. The characterization of Jews as an ?ethno-religious? group is reflective of this interconnection, and is consistent with the concept of ?secular? Judaism, an idea that some writers have developed to capture the reality of those people who see their Jewish background as more of a cultural than a religious identity (e.g., Dershowitz 1997). A similar overlap also occurs to some extent in the case of French Canadians, for whom Roman Catholicism is a central component of their cultural identity, even though only a small proportion of French Canadians actually attend religious services (e.g., Grabb and Curtis 2005: 143-145; see also Gagnon 2003). Another issue to keep in mind is that religion can provide one of the primary contexts in which ethnic minorities define and affirm their social identities. This may be especially true for those minorities who are also recent immigrants to Canada, because religious institutions can serve as an immediate, familiar, and inviting source of internal cohesion or in-group 62 bonding for people who have yet to integrate into the larger society. Scholarship on immigrant incorporation has shown that religion provides ?symbolic resources? to immigrant groups as they adapt to new social and cultural conditions in the host society (Breton 2012; Gordon 1964; Min and Kim 2001). These symbolic resources consist of shared symbols, rituals, and traditions (Swidler 1986: 277-78), from which members can reconstruct or maintain social boundaries (see Hirschman 2004). Hence, for purposes of the present analysis, it is important to include religious affiliation as a control variable when assessing ethnic variations in social and political trust, because of the overlap between ethno-racial background and religion, and the possibility that social distance and social boundaries processes arising from religious differences will affect levels of trust. As to which religious groups are more likely to exhibit social or political trust, findings have varied across different times and societies. However, most research seems to show that Protestant denominations are typically associated with higher levels of social trust, but lower levels of political trust, than are Catholics or members of other religions (e.g., Soroka et al. 2007; Paxton 2007; Stolle et al. 2008; Grabb et al. 2009b; You 2012); one partial exception is a study by Veenstra (2002), who found Protestants were higher on both types of trust. The reasons for the prevailing pattern may relate to the less hierarchical structure of authority in Protestantism compared to other faiths, which can promote more trust in other people but less trust or deference regarding political authority (e.g., Uslaner 2004; see also Uslaner 2002; Wood 1992; Grabb and Curtis 2005). 63 The Effects of Language Language stands as another significant ethno-cultural marker, a factor that tends to be closely correlated with ethno-racial background, that is often readily observable in everyday social interactions, and that, consequently, can have a notable impact on how we relate to others and to the society as a whole. Several of the previous studies discussed in this chapter indicate that both social and political trust seem to vary by language affiliation, especially when comparing Francophones and Anglophones. Our review of previous research indicates that French-speaking Canadians, particularly in Quebec, have a comparatively lower sense of social trust than other Canadians, but that, at least in recent times, Francophone Canadians may have a somewhat higher level of trust or confidence in the country?s political institutions than do other Canadians (recall, e.g., Breton et al. 2004; Soroka et al. 2006, 2007). As cited previously, some early studies that used various indicators of integration or inclusion within the larger society found that the main linguistic cleavage was between French Canadians, on the one hand, and English and ?Other? Canadians, on the other hand (recall Roseborough and Breton 1968; Simeon and Elkins 1974; Grabb 1979; see also Breton and Stasiulis 1980). Nevertheless, significant changes in immigrant composition in recent decades have also transformed Canada?s linguistic composition. This is especially true for ?Other? or Allophone Canadians, who are much more likely than in years past to speaking non-European languages. This increase in the cultural distinctiveness of languages spoken by Allophones could mean that Allophones could also have different levels of social and political trust than Anglophone majority. More specifically, it could be that the Allophone minority, like the Francophone minority, expresses lower social trust but higher political trust compared to Anglophone Canadians. 64 Socioeconomic Factors: Income and Education As has been noted previously, there is a basis for arguing that one of the key reasons why we might expect lower levels of trust among certain ethno-racial minorities is socioeconomic disadvantage, and the related process of discrimination. With few exceptions, research in a number of societies has shown a clear tendency for ethno-racial minorities to experience relatively low socioeconomic status attainment, with below-average levels of income, educational attainment, and occupational status (Duncan and Murnane 2011; Farrell 2012; Kochhar, Fry, and Taylor 2011; McBrier and Wilson 2004; Pendakur and Pendakur 2011; Weller, Ajinkya, and Wilson 2012;). The pattern is not entirely consistent, since some minority groups, including Asians in the United States, for example, tend to have above-average socioeconomic status (Kao and Thompson 2003; Pew Research Center 2012; Sakamoto, Goyette, and Kim 2009). In Canada, as well, there are exceptional cases, especially if we focus on measures of educational attainment, where a number of minority groups have been found to be among the most highly educated Canadians (e.g., Hou et al. 2009). Even so, as discussed below, most evidence points to continuing problems of socioeconomic disadvantage among Canada?s minority groups. Such disadvantages stand as an important intervening explanation in accounting for why ethno-racial minorities may have lower levels of trust and connectedness in Canada. Many years ago, John Porter (1965) posed the so-called ?blocked mobility? thesis when discussing the obstacles that ethnic groups are likely to face in their pursuit of socioeconomic success in Canadian society. Evidence compiled over recent decades has demonstrated that there are real problems of socioeconomic disadvantage for many of Canada?s minorities, 65 although the patterns differ substantially, depending on the socioeconomic indicator that is employed. First, with regard to education, except for a few groups, including Blacks and Native peoples, ethnic and racial minorities have generally been found to have higher levels of education than non-minority Canadians. Educational attainment is especially high among such visible minority groups as the Chinese, South Asians, and Filipinos, and also for some ethno-religious minorities, especially Jews (e.g., Davies and Guppy 2010; Hou et al. 2009: 257, 259; Li 2003). This pattern holds true, whether or not the minorities in question are Canadian-born or foreign-born; indeed, partly because of the points selection policy, immigrant Canadians on average have higher levels of education than non-immigrants (see also Hou and Balakrishnan 1996; Li 2003; Pendakur and Pendakur 1998, 2002; Reitz and Banerjee 2009; Grabb and Curtis 2005). Nevertheless, these same studies also reveal that minority groups, especially if they are immigrants, usually do not receive the same economic ?returns? on their education that non-minority Canadians receive. The result is that ethno-racial minorities are more likely than non-minorities to hold lower-level occupations, to be underemployed, and to earn below-average incomes (e.g., Hou et al. 2009; Lian and Matthews 1998; Reitz 2001a, 2001b; Reitz and Banerjee 2009; Soroka et al. 2006). It is understandable, then, that ethnic relations and attitudes in Canada are at times complicated by the more frequent and at times inconsistent experiences of socioeconomic disadvantage faced by ethno-racial minority groups. On the one hand, we might expect that the relatively high levels of educational attainment among many minority groups should be conducive to their greater sense of integration within the Canadian institutional structure, which in turn should promote a greater sense of trust and connectedness, both toward other 66 Canadians and toward the society?s major political institutions. On the other hand, the fact that minority Canadians tend to have lower average incomes, and do so despite their comparatively high education, could reduce their sense of connectedness, and exacerbate any feelings of distrust that they might harbour. A few previous studies of ethnic variations in social cohesion in Canada have addressed the relative effects of education and income on trust, and have found inconsistent results. Soroka and colleagues (Soroka et al. 2007) found weak positive effects for education and income on trust in other people, along with a weak positive effect and a weak negative effect of income on trust in government. Reitz and Banerjee (2009: 287) looked at income and found that it had a relatively weak effect in accounting for ethnic differences in various measures of social integration (see also Reitz and Banerjee 2007). Grabb et al. (2009b: 387) reported inconsistent effects for both education and income on four individual measures of political confidence, while Veesntra (2002) reported no relationships between either education or income and trust. Therefore, a goal of the present analysis is to explore further and understand better the complex and perhaps contradictory influences that education and income seem to have on the trust levels among Canada?s ethno-racial minorities, while simultaneously controlling for other factors. Reitz and Banerjee (2009) have suggested that it is really perceived or actual discrimination against minority groups that lies behind these socioeconomic influences, and that may be the key factor affecting how minority groups feel about their connection to the larger society and their trust in other people. We have seen evidence of perceived discrimination among Canada?s minorities, but some might question whether such measures 67 reflect actual ethnic or racial discrimination, since they are based on respondents? personal interpretations of their experiences. However, evidence from previous Canadian studies of actual employment discrimination against racial minorities suggests that the perceptions among Canada?s minorities have a clear basis in reality. Research based on Toronto field experiments, in which equally qualified white and non-white subjects applied for the same jobs, revealed that non-whites were subject to discriminatory hiring practices in about one-quarter of the cases (Henry and Ginsberg 1985; see also Henry 2004; Oreopoulos 2009). These results were strikingly similar to findings from studies conducted in major cities in the United States and England. Other early research has shown that Canadians in the past have tended to rank racial minorities as lower in ?social standing? compared to other groups (see Reitz and Breton 1994). On the other hand, some studies suggest that, on certain measures of ?social distance,? including opposition to interracial marriage or to having racial minorities as neighbours, Canadians are relatively more accepting of interracial contact than are most other populations (see Andersen and Milligan 2009: 398; Grabb and Curtis 2005: 206-212; Satzewich 2011) Still, problems of ethnic prejudice and racial marginalization continue to exist in Canadian society (see Henry and Tator 2005; Satzewich 1998; Satzewich and Liodakis 2007). These problems could underlie the possible effects of education and income on social and political trust in the present analysis. Social Engagement Factors: Voluntary Association Membership and Diversity of Friendship Networks The final set of explanatory controls that I consider in the thesis concerns two factors that fall under the broad rubric of social engagement. Two important means for removing the social 68 boundaries and reducing the social distance between ethnic groups is through involvement in formal community organizations and informal social networks. Both of these forms of activity can bring people from different backgrounds into direct contact with each other, thereby enhancing trust in others and in the society more generally. I first discuss the possible role of voluntary organization membership, and then consider the possible effects of the ethno-racial composition of people?s social networks. Ethnicity and Voluntary Association Membership It is a plausible hypothesis that people who are members and participants in formal voluntary organizations will be more likely than other people to possess a stronger sense of connection to and trust in their fellow citizens and their society?s institutions (see, e.g., Uslaner 1998; Stolle 1998; Veenstra 2002). This may be especially likely if these activities involve more general types of organizations, the kind that can provide bridges linking different ethnic communities with each other. A key issue for the present study is the extent to which ethnic groups vary in their levels of voluntary association membership. Only a few studies have been conducted to date on this topic, and even fewer have used national survey data. Some analyses have looked at the related topic of linguistic and regional differences, specifically between French and English Canadians, and between Quebecers and people living elsewhere in Canada (e.g., Caldwell and Reed 2000; Curtis and Grabb 2002; Curtis, Lambert, Brown, and Kay 1989; Grabb and Curtis 1992). These studies, which are typically based on representative national surveys, generally indicate that Francophones and Quebecers are less likely than Anglophones and non-Quebecers to be involved in voluntary association activity. However, in one study that 69 looked at the joint effects of language and region, researchers used data from the 2002 Economy, Security, and Community Survey (ESCS) to show that it is mainly Francophones within Quebec who are low in such activity (Hwang, Andersen, and Grabb 2007). This suggests the need to look at both language group and region when considering ethno-racial differences in voluntary activity. Another study, based on Canada?s 2003 General Social Survey, compared a wider range of ethnic groups, although the findings were restricted to immigrant respondents only (Baer 2008). Results showed that levels of voluntary involvement were comparatively low for individuals born in India, the Philippines, Vietnam, Poland, and China/Hong Kong (Baer 2008). This outcome suggests that, at least for foreign-born Canadians, certain racial minorities may be comparatively less likely to engage in civic activity. Such results are largely consistent with an earlier study based on a 1969-1970 Toronto sample, which looked at the ?political organization memberships? of immigrants. Results showed that some ethno-racial minority groups, namely ?Asians? and ?Blacks,? had relatively low membership levels, although certain European groups (Italians, Greeks, and Portuguese) also had low levels, and one ethno-religious minority, Jews, had the highest level (Richmond and Goldlust 1977: 52; see also Chui, Curtis, and Lambert 1991). Some studies that have looked at both immigrant and native-born respondents have found basically similar patterns, including research using a Toronto sample for the late 1970s. This study determined that two minority groups, West Indians and Chinese, had low levels of membership in ?ethnic-community organizations,? although here too there were some European groups (Italians, Portuguese, and Germans) who also had low levels, and Jewish respondents once again ranked highest (Breton 1990: 228-229). Another analysis, which 70 drew on national survey data from 1997, but which had a limited set of ethnic categories, reported that ?visible minorities? ranked lower on a composite measure of ?communal activity? than did several other groups, including ?Western Europeans,? ?Eastern Europeans,? and those of ?British? background (Breton et al. 2004: 129; see also Breton 2005: 235, 252-255; Johnston and Soroka 2001: 38). To my knowledge, there exists only one Canadian study based on national survey data that considers a range of voluntary association involvements, and that includes a relatively detailed set of ethnic and racial groups (Grabb, Andersen, and Hwang 2009a). That analysis used the 2002 ESCS data to show that some visible minorities, especially those of Latin American and East Asian heritage, had below-average voluntary membership levels, but that other ethno-racial groups, including South Asians and Native peoples, were at or above the average. All of these studies, taken together, suggest that there will be significant ethnic variations in voluntary association membership and that they could, in turn, mediate ethno-racial differences in trust. While it is not entirely clear from previous findings, I expect that those ethno-racial groupings who experience the greatest social distance and social boundaries influences will exhibit lower levels of both social and political trust, and that lower voluntary association activity among these groups will be one of the reasons for this outcome. Ethnic Diversity of Friendship Networks In much the same way that voluntary association activity could have important implications for the levels of trust among different ethnic groups, so too could the more informal patterns of social networking that operate in different ethnic communities. The informal personal 71 connections deriving from social networks can provide individuals with many resources, or what Bourdieu (1985) has referred to as ?profits? (see also Musick and Wilson 2008: 120). Research in the United States and other countries has shown that network connections can enhance people's general welfare (Lin and Erickson 2008: 1, 4); their chances for long-term success in the educational system or labour market (Brady, Schlozman, and Verba 1999; Coleman 1988; Granovetter 1973, 1995; Smith 2008; Wilson and Musick 1998); and their opportunities to engage in politics or voluntary activity (Bekkers, V?lker, van der Gaag, and Flap 2008; Flap 1999; Putnam 1993; Brady, Verba, and Schlozman 1995; ; Lin 2000; Tindall and Cormier 2008; Wilson and Musick 1997). Partly for these reasons, as discussed at the beginning of this chapter, Putnam has explored how informal networks and contacts affect social trust and cohesion (Putnam 1995, 2000, 2007). A key question for the present analysis is how the social networks of ethnic minorities in Canada influence their potential levels of both social and political trust. For purposes of this study, the specific focus is on friendship networks and the extent to which the various ethno-racial groups in the analysis have friends with ethnic backgrounds that are similar to or different from their own. This factor has received virtually no attention in previous Canadian research on trust, but could have significant implications for why some groups have greater levels of social or political trust than others. Research in the United States indicates that social networks are crucial for the development of generalized trust (Glanville and Paxton 2007; Stolle 2001). As previously discussed, American evidence also reveals significant ethno-racial differences in trust, which is generally lower among racial minorities, and especially low for blacks compared to whites (recall Smith 1997; 2010; Uslaner 2002; Wilkes 2011). But does diversity in the ethnic 72 composition of people?s social networks produce greater trust in others at the individual level, as some have suggested (recall Uslaner 2011), or do heterogeneous networks have the opposite effect, as some aggregate level studies have implied (recall Putnam 2007)? The virtual absence of any research on this question in Canada makes it essential that I explore the impact of ethnic friendship diversity on ethno-racial differences in social and political trust in the present analysis. My initial expectation is that friendship diversity should serve to bridge the gaps in social distance and the social boundaries separating groups from each other, leading to relatively greater social trust and trust in public institutions in Canada. ADDITIONAL CONTROL VARIABLES: REGION, AGE, AND GENDER Apart from the three main sets of explanatory controls that are the focus of the thesis analysis, I include three other factors that should be taken into account when examining ethno-racial variations in trust. These other controls are region, age, and gender. Region It is well-known that, in Canadian research, the effects of region often closely overlap with the three ethno-cultural markers that are among the central explanatory influences included in the present analysis. A prime example of this overlap concerns the division between Quebec and the rest of Canada. For example, as Breton has pointed out, it is largely because of the predominant use of French in Quebec that language and region, though conceptually separate, are so closely related empirically in this country (Breton 2005: 145; also Breton, 73 Reitz, and Valentine 1980). On other ethno-cultural dimensions, as well, we see a significant overlap, including the over-representation of Roman Catholics in Quebec, coupled with the considerable under-representation of both visible minorities and immigrants in that province (see, e.g., Grabb and Curtis 2005). More generally, however, the diversity of Canada?s regions is widely recognized as one of the distinguishing features of the nation, possibly rivalling ethnic diversity in its potential significance for understanding the social, political, economic, and cultural composition of the society (e.g., Gibbins 1982; Henderson 2004; Corrigall-Brown and Wien 2009; Schwartz 1974; Simeon and Elkins 1974). Research has shown that, along with French Quebecers, Atlantic residents have generally been the most likely to develop and maintain a regional identity that matches, and often surpasses, their sense of connection with Canada as a whole. As in Quebec, the evidence also indicates that ethno-racial minorities form a comparatively smaller proportion of the Atlantic population than is the case elsewhere. These kinds of differences in ethnic composition and related factors across the regions make it necessary to control for region when exploring ethno-racial variations in social and political trust in this thesis. My central expectation is that Quebec will stand out as the most distinctive region in regard to patterns of trust, reflecting its status as a so-called ?region apart? from all the others on a wide range of issues (Grabb and Curtis 2005; see also, e.g., Adams 2003; Baer, Grabb, and Johnston 1993). More specifically, I expect that Quebecers will have below-average social trust but above-average political trust, in part because of the disproportionate number of residents who are Francophones and of French ethnic origin, and the levels of trust that people with these backgrounds have displayed in previous research. 74 Age Age is another potentially important control variable to be included in the present analysis. Age appears from previous research to have an effect on trust, although the relationship has not always been consistent across different studies. Soroka et al. (2006, 2007) found that age had a positive effect on generalized trust in other people, but a negative effect on trust in government. Veenstra?s (2002) research showed a positive effect of age on both types of trust. Grabb et al. (2009b) found virtually no age effect on political trust in a more recent Canadian analysis. American research has also shown inconsistent patterns, with early studies indicating a negative relationship between age and political trust, but more recent analyses indicating a positive relationship between these two variables (Dalton 2004). In the case of social trust, some American studies suggest that age has a positive effect (Smith 1997; Uslaner 1998), while others have found a generally positive though non-monotonic effect, with trust in other people being higher among the middle-aged than among younger and older groups (e.g., Robinson and Jackson 2001). In the present analysis, age is expected to have a significant effect on trust in its own right. In addition, however, age is also correlated with several of the key explanatory controls in the present study, and so may affect the patterns of results for the ethno-racial groupings themselves. For example, evidence shows that, on average, visible minorities, Aboriginal Peoples, and recent immigrants all tend to be younger than other Canadians (Samuel and Basavarajappa 2006; Statistics Canada 2006a; Statistics Canada 2006b). It is important, then, that we control for the effects of age in the present study. 75 Gender Finally, the thesis research will also include a control for gender. Some previous research in Canada and the United States has found that females have somewhat greater trust in other people than do males (Smith 1997; Soroka et al. 2007) and also slightly greater trust in political institutions, especially the police (Grabb et al. 2009b). This control variable is included both to determine if there are appreciable gender differences in social and political trust in Canada, and also to take into account any gender compositional differences that might influence the pattern of ethno-racial differences in trust. MAIN RESEARCH QUESTIONS The main research questions to be addressed in the thesis are as follows: 1. Are there significant ethno-racial differences in social and political trust? In particular, how do visible minorities, French Canadians, and Aboriginal Peoples compare with others in their levels of trust? 76 2. Are ethno-racial differences in social and political trust mainly or partly attributable to the three significant markers of cultural distinctiveness that I have identified in this chapter: immigration status, religion, and language? That is, do those who are the most distinctive from the British and other established groups on these three ethno-cultural dimensions express lower levels of social and/or political trust than other Canadians? Here I allow for the possibility that the patterns could be different for the two types of trust, with culturally distinctive groups being lower on social trust but higher on political trust than other Canadians. 3. Are ethno-racial differences in trust mainly or partly attributable to differences in the socioeconomic status attainment of these groups, as measured by education and income? Is it the case, in particular, that those groups who have higher education and higher incomes are more likely than those with lower education and lower incomes to express more social and political trust, because of the greater potential for social integration provided by these advantages, resources, and successes? 4. Are ethno-racial differences in trust largely or partly due to differences in voluntary association activity? More specifically, are some ethno-racial groups more likely than others to belong to formal voluntary organizations, which could aid in breaking down social boundaries and undermining feelings of social distance, thereby increasing levels of social trust and political trust in these groups? The underlying assumption here is that experience with these organizations on average enhances people?s feelings of being accepted within the larger Canadian community, making them more trusting of others and of the society?s major public institutions. 77 5. Are ethno-racial differences in trust mainly or partly due to differences in social network composition? More specifically, are those groups whose friendship networks include people who are ethnically different from themselves more likely to express feelings of social and political trust than are other groups? The rationale is that having a diversity of friendships is more likely to establish bridging connections that span across ethnic communities, leading to a greater sense of trust in others, and ultimately trust in public institutions as well. 6. Are the patterns of ethno-racial differences in social and political trust affected by controls for region, age, and gender? SUMMARY In this chapter I have reviewed the previous literature on ethno-racial variations in social cohesion and connectedness, with a specific focus on two key dimension of cohesion: social trust and political trust. I also have outlined three sets of key explanatory factors, including: (1) three ethno-cultural correlates or markers of ethno-racial background: immigration status, 78 religion, and language; (2) two crucial indicators of socioeconomic status attainment: income and education; and (3) two indicators of social engagement: voluntary association activity and friendship network composition. In addition, I have discussed a number of other control variables that will be included in the thesis and have delineated the major research questions that will be addressed in the analysis. In the next chapter, I describe the data source and list the measures that will be used to test the main hypotheses and research questions. 79 CHAPTER TWO: DATA AND METHODS DATA SOURCE The data come from the Public Use Micro-data File (PUMF) for Cycle 22 of the Canadian General Social Survey (GSS). The survey, which focuses mainly on social engagement in Canada, was conducted in 2008 by the Social and Aboriginal Statistics Division of Statistics Canada. Data were collected in 5 waves, between February 1 and November 30 of 2008.6 The survey involved computer assisted telephone interviewing (CATI), and employed random digit dialing (RDD). This approach produced a final sample of 20,401 respondents, representing a response rate of 57.3 percent. The target population included all persons 15 years of age and older in Canada, excluding full-time residents of institutions and people living in the Yukon, Northwest Territories, and Nunavut7. Because my principal interest is in 6 Another approach to comparing trust levels among Canada?s ethno-racial groups would be to employ mixed methods, including qualitative interviewing, for example. I believe that this would be a useful approach to pursue in future studies of trust in Canada. In the present analysis, I have concentrated on the large scholarly task of providing a national profile, using representative sample survey data, of factors that explain how and why eight distinct ethno-racial groupings differ in both social and political trust. The findings reported here should serve as an important benchmark, to which future studies, including those involving qualitative analysis, can be compared. 7 One possible limitation of the present study concerns the exclusion from the sample of respondents from the northern territories, which conceivably could affect the representation of certain groups, especially Aboriginal Peoples. It is true that, according to the 2006 Census, Aboriginal Peoples accounted for a large proportion of the population in Nunavut (85 percent), the Northwest Territories (50 percent), and the Yukon (25 percent) (Statistics Canada 2010). At the same time, however, the Census data show that only about 4.5 percent of the total population of Aboriginal Peoples lives in these three northern regions, with about 80 percent living in Ontario and the four western provinces, and most of these individuals also living in urban areas. Therefore, the exclusion of the northern territories from the GSS 80 adult individuals who are well established in their lives (e.g., have completed their education), I selected only those respondents who were older than age 24. This limited the sample to 18,457 cases. Non-responses and other considerations meant that the working sample was 17,374 cases. Households without regular telephones were excluded from the survey, as well as households that only had cellular telephones.8 Research on survey design suggests that households without telephones represent less than 0.9% of the target population (Statistics Canada 2007).9 All telephone interviews were conducted in the preferred official language of the respondents. The sample was drawn using a stratification procedure based on geographic location10. There were 27 strata in all. Fourteen of the strata were Census Metropolitan Areas (CMAs) cellpthat included: St. John's, Halifax, Saint John, Montreal, Quebec City, Toronto, Ottawa, sampling frame seems likely to have little effect on the representativeness of the Aboriginal Peoples subsample in the present analysis. 8 People without telephones are excluded from the sampling frame (0.9% of the population) as were individuals who rely exclusively on cellphones (6.4%). 9 The exclusion of people without regular landline telephones may place some limitations on the data used in this study, given that more Canadians are switching to cell phones and abandoning landlines. However, Statistics Canada has shown that, as recently as 2010, a large majority (87 percent) of Canadian households still had landlines (Statistics Canada 2011). Therefore, it is likely that the sample used in present study, with data collected in 2008, was not unduly affected by being restricted to households with landlines. Even so, the increasing use of cell phones and declining use of landlines in Canada probably make it necessary for future studies to include cell phone users in their sampling frames. 10 One of the most intriguing topics to include in future studies of trust includes spatial or geographic factors. This seems especially interesting because some of the measures used for social trust in the present analysis have a spatial component, asking about how Canadians feel about people in their neighbourhoods, for example. Where appropriate data are available, subsequent analyses of trust would benefit from the inclusion of spatial variables, especially contextual variables. Examples include the degree of ethno-racial diversity within Census tracts, the average household income within Census tracts, etc. 81 Hamilton, Winnipeg, Regina, Saskatoon, Calgary, Edmonton, and Vancouver. Three additional strata were formed by combining the remaining CMAs in each of Quebec, Ontario, and British Columbia. The other ten strata were based on combining the non-CMA areas in each of the ten provinces. Survey estimates were weighted to represent all persons in the target population. (For further details see Statistics Canada 2010: 13-16.) Survey estimates were weighted to represent all persons in the target population. Applying the weight variable to the actual sample produces a very large weighted sample, one that is too large for conducting meaningful tests of statistical significance. Hence, the weight used in the analysis was adjusted by dividing the sample weight variable by its average weight. The weight used throughout the analysis is WGHT_PER/1235.315334. (For further details on the original sampling and weighting procedures, see Statistics Canada 2010: 13-16.) VARIABLES The Dependent Variables: Measuring Social Trust and Political Trust As noted at the beginning of Chapter 1, leading scholars in Canada and elsewhere point to social trust and political trust as the two most important dimensions or gauges of social cohesion in society (recall Breton, Reitz, and Valentine 1980; Soroka, Johnston and Banting 2007; Putnam 2000). Below I describe how both of these dimensions are measured in the present analysis. 82 Social Trust The first of my two dependent variables concerns social trust, or the extent to which people feel that they can trust other individuals in society. There are a number of items in the GSS that touch on different aspects of trust in others, and that were considered for possible inclusion in the thesis research. Initial analyses ultimately led to the selection of six items that were found to inter-correlate well with each other, and that could be combined to provide a reliable scale of social trust. Factor analyses confirmed that the six items produced a clear single-factor solution. Moreover, a check for the internal consistency reliability of the six-item composite scale produced a strong Cronbach?s alpha of .758. Hence, both the factor analysis and the test for reliability indicate that the items essentially measure the same underlying construct. Consequently, although I report some initial findings for the individual component items, most of the analyses of social trust in the thesis focus on the composite scale. All six social trust items were coded so that responses that indicated the highest level of trust in other people were assigned the highest scores. Mean values were substituted for the small number of missing cases for each question. Preliminary analyses revealed that essentially the same results are found if these missing cases are excluded from the findings; therefore, to retain a larger sample size throughout the analysis, the results using the mean-substitution procedure are reported. Because the six items were not all measured on the same metrics, with some having five response categories and others having less than five, I followed conventional standardization techniques (see, e.g., Scavo 2007), in which each component variable was standardized using a Z-score transformation, by subtracting the mean and dividing by the standard deviation for 83 each item . Each item was also weighted using the component score coefficients generated from the factor analysis, so as to take into account the somewhat different weights that the individual measures contributed to the underlying construct that was derived from the factor analysis. The six items were then summed to form the composite scale of social trust. The six questions are listed below, along with their component score coefficients shown in parentheses: 1. ?Generally speaking, would you say that most people can be trusted [coded 2] or that you cannot be too careful in dealing with people [coded 1]?? (.236) 2. ?Using a scale of 1 to 5, where 1 means ?Cannot be trusted at all? and 5 means ?Can be trusted a lot?, how much do you trust people in your neighbourhood? (.277) 3. ?Using a scale of 1 to 5, where 1 means ?Cannot be trusted at all? and 5 means ?Can be trusted a lot?, how much do you trust strangers?? (.260) 4. ?Would you say that you trust most of the people in your neighbourhood [coded 4], many of the people in your neighbourhood [coded 3], a few of the people in your neighbourhood [coded 2], or nobody else in your neighbourhood [coded 1])? (.276) 5. ?If you lost a wallet or purse that contained two hundred dollars, how likely is it to be returned with the money in it if it was found by someone who lives close by? Would it be very likely [coded 3], somewhat likely [coded 2], or not at all likely [coded 1]?? (.229) 6. ?If you lost a wallet or purse that contained two hundred dollars, how likely is it to be returned with the money in it if it was found by a complete stranger? Would it be very likely [coded3], somewhat likely [coded 2], or not at all likely [coded 1]?? (.189) 84 The composite scale of social trust, therefore, is as follows: Social Trust = .236 x (general trust - 1.477)/.491 + .277 x (trust neighbours - 3.446)/1.085 + .260 x (trust strangers - 2.045) /.998 + .276 x (trust local people -2.945)/.990 + .229 x (wallet neighbour - 2.305)/.692 + .189 x (wallet stranger - 1.493)/.573 As a standardized variable, the social trust scale has a mean of 0, and a standard deviation of 1.0. It is quasi-normal with a slight negative skew (skewness = -.076) Political Trust The second dependent variable considered in the thesis is a composite scale of political trust, which includes multiple indicators of respondents? level of confidence in a range of public or political institutions. The GSS includes a battery of items that ask about the extent to which individuals feel they can rely on various organizations. Six of these items deal specifically with government institutions or public agencies. Respondents were asked ?how much confidence? they had in ?the police,? ?the justice system and courts,? ?the health care system,? ?the school system,? ?the welfare system,? and ?federal parliament.? All six items were coded so that responses indicating the highest level of confidence in the institution were assigned the highest scores. The possible responses were ?a great deal of confidence [coded 4],? ?quite a lot of confidence [coded 3],? ?not very much confidence [coded 2],? and ?no confidence at all [coded 1].? Mean values were substituted for the small number of missing cases for each question11. 11 The data set included two other items that asked about confidence in non-governmental organizations or institutions (i.e., ?banks? and ?major corporations?). However, because my interest was only in how ethno-racial groups might differ in their orientation toward political 85 A factor analysis of the six items produced a clear single-factor solution, and a test for internal consistency reliability revealed a strong Cronbach?s alpha of .759, which coincidentally is almost identical to the alpha for the social trust scale. These results again indicate that the items all measure the same underlying construct. As was the case for social trust, I report some initial findings for the individual component items of political trust, but concentrate most of my analyses on the composite political trust scale. For the political trust scale, each component variable was standardized using a Z-score transformation. Before being summed to form the composite scale, each item was also weighted, using the component score coefficients generated from the factor analysis. This procedure again takes into account the different weights that individual items contribute to the underlying construct derived from the factor analysis. The component score coefficients were as follows: the police (.225), the justice system and courts (.268), the health care system (.248), the school system (.249), the welfare system (.247), and federal parliament (.248). The composite scale of political trust is: Political Trust = .225 x (police - 3.212)/.712 + .268 x (justice - 2.700)/.814 + .248 x (healthcare - 2.921) /.759 + .249 x (schools - 2.946)/.705 + .247 x (welfare - 2.567)/.822 + .248 x (federal parliament - 2.411)/.797 As a standardized variable, the political trust scale has a mean of 0, and a standard deviation of 1.0. It is quasi-normal with a slight negative skew (skewness = -.087). Before or governmental organizations and agencies, I focused solely on the six political organizations included in the data set. It is possible that there are ethno-racial differences in trust regarding other types of organizations or institutions, but this topic was beyond the scope of the present study. 86 proceeding, I can also note that the two scales of trust appear to operate as relatively distinct dimensions of trust, for they are significantly but only weakly correlated with each other (r = +.196, p < .000). The Independent Variable: Ethno-racial Background The central independent variable in the thesis is ethno-racial background. Preliminary investigation showed that the optimal procedure for measuring this variable was to combine responses from 3 items in the PUMF of the GSS. The first of these items, ?ethnic7,? grouped respondents into seven ethnic categories, based on their answers to a battery of questions about ?the ethnic or cultural origins of your ancestors.? The original survey item allowed for multiple choices, but the large majority of respondents offered just one response. Those respondents who identified with multiple origins (e.g., ?Canadian and other,? ?British Isles origins and other,? ?French and other?) were grouped together in a category labelled ?Mixed-Origins.? The results for the latter category pose some difficulties in interpretation, because it cannot be determined from their responses what the respondent?s predominant ethnic origin or ethnic identity might be. However, so as to avoid sample size reduction, they are included. Their inclusion also allowed for a speculative test as to whether those respondents who stressed their multiple backgrounds might also have higher scores on the two trust measures, based on the assumption that their plural identities were a sign that they themselves had ?bridged? between ethnic groups. The second item for measuring ethno-racial background is the ?vismin? variable, which simply categorized respondents into ?visible minority? or ?non-visible minority,? based on the standard Statistics Canada classification. The third item, ?air_110,? divided the sample 87 into Aboriginal Peoples and non-Aboriginal Peoples, as determined from the individual?s ?Yes? or ?No? response to the question: ?Are you an Aboriginal person, that is, North American Indian, M?tis or Inuit?? Cross-referencing these 3 items produces a measure of ethno-racial background that includes the following eight categories: (1) British, which is used as the reference category in the analysis (N=4724); (2) French (N=1848); (3) ?Canadian? (N=1382, about two-thirds of whom [896/1382] are Francophone Quebecers, making them very similar to the French category; in fact, the majority of this category are perhaps more aptly labelled as ?Canadiens?); (4) Other European (N=2344); (5) Aboriginal (N=561); (6) Visible Minority (N=2160); (7) Other (N=1831); and (8) Mixed-Origins (N=2524). Explanatory Control Variables Ethno-cultural Markers: Immigration Status, Religion, and Language The first set of explanatory control variables, which includes what are referred to in the thesis as ethno-cultural markers, subsumes immigration status, religion, and language. 1. Immigration Status Immigration status12 is constructed by combining two items, one that asks about nativity, i.e., whether the respondent is born in Canada or outside Canada, and a second that asks, for 12 As noted previously, immigration status is not relevant for Aboriginal Peoples or French respondents, almost all of whom are Canadian-born. As Table 3.1 below shows, 96% of 88 those who were born outside Canada, the period in which they first came to the country. The responses were categorized as follows: (1) Canadian-born, which is used as the reference category in the analysis; (2) immigrated in 2000 or later; (3) immigrated in 1990-1999; (4) immigrated in 1980-1989; (5) immigrated in 1970-1979; and (6) immigrated before 1970. 2. Religion The measure for religion is collapsed in the PUMF data, and includes the following five categories: (1) no religion; (2) Roman Catholic; (3) Protestant, which is used as the reference category in the analysis; (4) Non-Christian; and (5) Other. 3. Language The third ethno-cultural marker that I consider in the thesis is language affiliation, which is measured with a question asking about the respondent?s ?household language,? and is grouped into four categories: (1) Anglophone, which is used as the reference category in the analysis; (2) Francophone; (3) Allophone; and (4) no answer. The findings for the latter category are not interpreted, but this category is retained in the analysis so as to avoid reducing the size of the working sample. Aboriginal Peoples in the sample are Canadian-born, while 4% of Aboriginal Peoples are foreign-born. 89 Socioeconomic Factors: Education and Income The second set of explanatory control variables used in the analysis involves measures of socioeconomic status or class location. The two variables chosen are educational attainment and income level.13 1. Education The measure of education is based on a 10-category item that asked the respondents to indicate the highest level of education that they attained. This measure was coded so as to reflect the average number of years of formal schooling that would typically be represented by each category, as follows: (1) post-bachelor?s degree or higher, which was coded as 18; (2) bachelor?s degree, coded as 16; (3) some university or completed community college diploma, coded as 14; (4) some community college or trade school, coded as 13; (5) completed high school, coded as 12; (6) some high school, coded as 10; and (7) elementary school or less, coded as 4. In addition, however, I also treat education as a multi-category dummy variable for some analyses, and use the highest education grouping as the reference category in these instances. I do this specifically for the multivariate analysis of political trust, because, as is discussed 13 Occupation was considered in preliminary research but sample size attenuation, which was due to the substantial number of respondents who were not in the labour force, led to the exclusion of this measure (for a similar approach, see, e.g., Smith 1997; Soroka et al. 2007; Reitz and Banerjee 2009). 90 below, education has a non-linear association with political trust that cannot be captured by treating education as a ratio-level or continuous variable. 2. Income For income, I chose a question that asked the respondents about their total annual household income from all sources. This choice is likely to reflect more accurately the overall socioeconomic ranking of respondents than does personal income, which can be especially misleading as a socioeconomic indicator for people who are not in the labour force. The item asks: ?What is your best estimate of the total income, before deductions, of all household members from all sources during the past 12 months?? This variable was categorized into 12 ranked categories in the PUMF data set. For purposes of the present analysis, I assigned a value, in thousands of dollars, to each of these categories to reflect the mid-point of each category?s income range, as follows: (1) no income, which was coded as 0; (2) less than $5,000, coded as 2.5; (3) $5,000 to $9,999, coded as 7.5; (4) $10,000 to $14,999, coded as 12.5; (5) $15,000 to $19,999, coded as 17.5; (6) $20,000 to $29,999, coded as 25.0; (7) $30,000 to $39,999, coded as 35.0; (8) $40,000 to $49,999, coded as 45.0; (9) $50,000 to $59,999, coded as 55.0; (10) $60,000 to $79,999, coded as 70.0; (11) $80,000 to $99,999, coded as 90.0; and (12) $100,000 or more, coded as 110. As I did for education, I also treated income as a multi-category dummy variable for some analyses, and used the highest income grouping as the reference category in these cases. I do this specifically for the multivariate analysis of political trust, because, as I describe below, 91 income has a non-linear association with political trust that cannot be captured by treating income as a ratio-level or continuous variable. Social Engagement Factors: Voluntary Association Membership and Ethnic Diversity of Friendship Networks 1. Voluntary Association Membership As discussed in Chapter 1, I wish to explore whether different forms of social engagement, including voluntary association involvement and social network composition, have a positive impact in breaking down social boundaries, reducing social distance, and consequently increasing the levels of social and political trust that different ethnic groups experience in Canadian society. Unfortunately, the GSS survey does not include a large array of items with which to test for these kinds of effects or patterns. For example, in contrast to some survey data used in earlier studies (e.g., Grabb et al. 2009a), the GSS does not provide a measure for people?s involvement in ethnically-based formal organizations (e.g., ethnic sports clubs or ethnic cultural organizations). Nevertheless, there is a good measure of general involvement in voluntary associations, which permits me to assess whether the overall level of voluntary activity of ethno-racial groups has a positive effect on their levels of trust. Such a finding would be consistent with the hypothesis that voluntary membership helps to bridge the social distance gaps separating ethnic communities from one another. Therefore, to measure general voluntary organization involvement, I use a count of the number of association memberships that respondents report, based on a question that asked 92 about eight ?groups, organizations, networks or associations? to which the respondent belonged in the previous 12 months. The question also noted that ?these could be formally organized groups or just groups of people who get together regularly to do an activity or talk about things.? Respondents read the list of eight types of organizations or groups and replied ?yes? or ?no? to being a member in each case. The organizations were: (1) ?a union or professional association?; (2) ?a political party or group?; (3) ?a sports or recreational organization (such as a hockey league, health club or golf club)?; (4) ?a cultural, educational, or hobby organization (such as a theatre group, book club or bridge club)?; (5) ?a religious-affiliated group (such as a church youth group or choir),? excluding the respondent?s church or religion itself; (6) ?a school group, neighbourhood, civic or community association (such as PTA, alumni, block parents or neighbourhood watch)?; (7) ?a service club or fraternal organization (such as Kiwanis, Knights of Columbus or the Legion)?; and (8) ?any other type of organization that you have not mentioned.? Because the distribution of this variable is non-normal, with the large majority of respondents having either no memberships or one membership, it was converted into a dummy variable in the analysis, as follows: no memberships, which is the reference category; one membership; two memberships, three or more memberships. 2. Ethnic Diversity of Friendship Networks One way of assessing the possible ?ethnic? or cultural nature of social engagement is to look outside formal voluntary associations and consider more informal types of social 93 engagement, including friendship networks. There is one measure in the GSS that allows me to assess the possible effect on social and political trust of having a more ethnically homogeneous or more ethnically diverse friendship network. That question is worded as follows: ?Think of all the friends you had contact with in the past month, whether the contact was in person, by telephone, or by e-mail. Of all these people, how many come from an ethnic group that is visibly different from yours?? Respondents had five possible answer choices: ?all,? which was coded as 5; ?most,? coded as 4; ?about half,? coded as 3; ?a few,? coded as 2; and ?none,? coded as 1. With this coding scheme, the variable can be interpreted as a measure of ethnic diversity of friendships.14 The expectation is that respondents who have more diverse friendship networks will express higher levels of social and political trust than will respondents who have less diverse friendship networks. As discussed below, I also conducted analyses in which this variable was converted into dummy variables, with ?none? as the reference category.15 14 This item may not be ideal as a measure of diversity for some respondents, as in the case of a person with friends that are all ethnically different from her or him, but who themselves have homogeneous backgrounds, for example, an Aboriginal Person who has all British friends. Nevertheless, this measure is the best available to test the hypothesis, and so is included partly for exploratory purposes. 15 In preliminary analyses, I considered another social engagement measure, which asked respondents to think of the voluntary association that they were most active in, if any, and then asked: ?thinking of all the people you met through this organization ? how many come from an ethnic group that is visibly different from yours?? Results, however, revealed that this variable had no significant effect on either social trust or political trust, and so it was excluded from the analysis for reasons of parsimony and to trim the models. 94 Other Control Variables: Region, Age, and Gender The remaining controls included in the thesis are region of residence, age, and gender. Region Region is measured using province of residence, which is grouped into five categories: the Atlantic region; Quebec; Ontario, which is used as the reference category; the Prairie region; and British Columbia. Age is measured as a categorical variable in the PUMF, with 12 ranked categories that are grouped into 5-year ranges, between age 25 and age 80 or more. So as to approximate a continuous variable, these categories were assigned values that reflected the mid-point of each category?s age range, as follows: age 25-29 coded as 27, 30-34 coded 32, 35-39 coded 37, 40-44 coded 42, 45-49 coded 47, 50-54 coded 52, 55-59 coded 57, 60-64 coded 62, 65-69 coded 67, 70-74 coded 72, 75-79 coded 77, and 80 coded as 80. Finally, gender is a dichotomous dummy variable with males used as the reference category and coded as 0, and females coded as 1. This completes the description of the data source, variables, methods, and measurement approach employed in the thesis. In Chapter 3, I present the bivariate results, followed by the multivariate analyses in Chapter 4. 95 CHAPTER THREE: BIVARIATE RESULTS The initial step in the empirical analysis is to examine and describe the various bivariate findings in the thesis. The main concern is to describe the relationship between ethno-racial background and both social trust and political trust. In addition, however, it is useful first to consider some of the other bivariate associations. These include the relationship between each of the control variables and the two trust measures, as well as the relationships linking ethno-racial background to each of the other independent variables. An awareness of these other empirical linkages will provide some context for the multivariate analyses in Chapter 4, and will make it easier, in Chapter 4, to demonstrate and interpret the effects that the control variables have on the relationship between ethno-racial background and the two types of trust. I begin with a brief description of the relationship between ethno-racial background and the other independent variables, focusing on the most salient patterns only. Relationships Between Ethno-racial Background and the Control Variables The bivariate associations between ethno-racial grouping and the control variables are shown in Tables 3.1 to 3.5. The main patterns in these tables are as follows. First, certain ethno-racial groupings are quite distinct from the others in regard to their distributions on the three ethno-cultural ?markers?: immigration status, religion, and language (Tables 3.1, 3.2, and 3.3). 96 TABLE 3.1 0.1 Ethno-racial Background and Immigration Status Canadian-born (%) Immigrated since 2000 (%) Immigrated 1990-1999 (%) Immigrated 1989-1990 (%) Immigrated 1979-1979 (%) Immigrated before 1970 (%) British 88 1 1 1 2 7 French 97 1 1 <1 <1 1 Canadian 99 <1 <1 <1 <1 1 Other European 70 3 2 4 3 18 Aboriginal Peoples 96 1 <1 1 1 1 Visible Minority 10 26 26 18 15 5 Other 73 5 6 4 3 9 Mixed Origins 96 <1 <1 1 1 2 TOTAL 78 4 5 4 3 6 TABLE 3.2 0.2 Ethno-racial Background and Religion No Religion (%) Roman Catholic (%) Protestant (%) Non-Christian (%) Other (%) British 25 19 54 1 1 French 12 82 5 1 <1 Canadian 15 69 15 <1 1 Other European 19 42 30 9 1 Aboriginal Peoples 23 43 28 4 2 Visible Minority 24 22 20 33 2 Other 22 33 32 12 1 Mixed Origins 24 43 30 2 1 TOTAL 21 39 32 7 1 97 TABLE 3.3 0.3 Ethno-racial Background and Language Anglophone (%) Francophone (%) Allophone (%) British 96 3 1 French 23 75 2 Canadian 31 68 1 Other European 83 4 14 Aboriginal Peoples 82 9 9 Visible Minority 38 4 57 Other 73 13 13 Mixed Origins 75 24 1 TOTAL 68 21 11 With respect to immigration status, we see that visible minorities are clearly distinct, with only 10 percent of this group being Canadian-born, whereas between 73 percent and 99 percent of the other seven ethno-racial groupings are Canadian-born. On religion, the most prominent differences are that: visible minorities (at 33 percent) are more likely than any other grouping to be non-Christian; the French and Canadian categories (at 82 percent and 69 percent) are more likely than any other grouping to be Roman Catholic; and the British (at 54 percent) are more likely than any other category to be Protestant. As for language, all but three of the groups are majority Anglophone. The exceptions are the French and the Canadians, who are the only categories that are majority Francophone (at 75 percent and 68 percent) and the visible minorities, who are the only category that is majority Allophone (at 57 percent). 98 TABLE 3.4 0.4 Ethno-racial Background and Region Atlantic (%) Quebec (%) Ontario (%) Prairies (%) British Columbia (%) British 13 5 48 17 17 French 7 70 15 4 4 Canadian 7 67 14 8 5 Other European 2 11 45 28 14 Aboriginal Peoples 9 11 31 34 15 Visible Minority 1 11 52 14 22 Other 5 20 40 19 17 Mixed Origins 7 25 38 18 12 TOTAL 7 23 39 17 14 The findings in Table 3.4 show the relationship between ethno-racial background and region of residence. The key results are that, not surprisingly, a large majority of both the French and the Canadian respondents live in Quebec. Visible minorities are disproportionately to be found in Ontario and British Columbia, but are underrepresented in Quebec and the Atlantic region. Aboriginal Peoples are overrepresented in the Prairies but underrepresented in Quebec, while the British tend to be overrepresented in Ontario and the Atlantic, but underrepresented in Quebec. Table 3.5 shows how the eight ethno-racial groupings compare on the other independent variables. For the two socioeconomic variables, it is notable that Aboriginal Peoples rank the lowest on both education and income. Visible minorities rank the highest on education but 99 are only sixth out of the eight groups on income; this finding, which suggests that visible minorities tend not to get the same economic returns from their educational credentials as do other groups, has, of course, been found in a number of Canadian studies over the years (e.g., Davies and Guppy 2010; Reitz 1998; Reitz and Banerjeee 2009; Grabb and Curtis 2005). It is also the case that the British rank highest on income and rank third among the eight groups on education, while the French and the Canadians fall below the middle rank on both education and income. Regarding the two social engagement controls, we see in Table 3.5 that the highest level of voluntary association membership is found among the British respondents, with Other Europeans and the Mixed-Origins category also having above-average membership levels; the groups with the lowest membership levels are the Canadians, the visible minorities, and the French. For the friendship networks variable, visible minorities, along with Aboriginal Peoples, have the most ethnically diverse sets of friends, while the French have the least diverse or most homogeneous friendship networks. Finally, in Table 3.5, there are the two remaining control variables, age and gender. With regard to age, visible minorities and Aboriginal Peoples stand out as being much younger on average than all the other ethno-racial groupings. The British, French, Canadians, and Other Europeans are the oldest groupings. Lastly, there are no important differences in the gender composition of the eight ethno-racial categories, although Canadians are somewhat higher in female proportion (55 percent) and Aboriginal Peoples are somewhat lower (47 percent). 100 TABLE 3.5 0.5 Ethno-racial Background and Education (in Years), Annual Household Income (in Dollars), Number of Voluntary Association Memberships, Ethnic Diversity of Friendships (Mean Score on 5-Point Scale), Age (Average in Years), and Percent Female Education (Years) Income (Dollars) Voluntary Memberships Friendship Diversity Age Female (%) British 13.7 72,036 1.57 1.98 52.7 51 French 13.5 68,198 1.24 1.75 51.3 49 Canadian 12.5 63,107 1.02 2.01 50.6 55 Other European 13.5 72,023 1.46 2.20 50.3 50 Aboriginal Peoples 12.4 57,431 1.31 2.48 45.4 47 Visible Minority 14.4 66,995 1.05 3.09 44.2 50 Other 13.9 70,186 1.38 2.12 48.4 52 Mixed Origins 13.7 71,977 1.45 1.95 48.0 52 TOTAL 13.6 69,614 1.36 2.15 49.6 51 The Association Between Ethno-racial Background and Social Trust In this section, I move to the first research question for the thesis, concerning whether and to what extent ethno-racial minorities differ from other Canadians in their levels of social and political trust. In Table 3.6, the basic relationship between ethno-racial background and social trust is displayed for each of the six individual items, and also for the standardized overall scale of social trust. One of the first points to note in Table 3.6 is that all of the results for the social trust measures are highly significant statistically, with Chi-square values for each of the individual items that are significant beyond the .000 level. 101 TABLE 3.6 0.6 Social Trust Measures by Ethno-racial Background British (%) French (%) ?Canadian? (%) Other European (%) Aboriginal Peoples (%) Visible Minority (%) Other (%) Mixed Origins (%) TOTAL (%) 1. Generally speaking, would you say that most people can be trusted or that you cannot be too careful in dealing with people? Can trust 56 39 35 52 40 44 47 50 48 *** Be careful 44 61 65 48 60 56 53 50 52 2. How much do you trust people in your neighbourhood? On a 1-to-5 scale Category 5 26 18 19 18 17 12 18 19 20 *** Category 4 32 29 27 32 19 23 29 32 29 3. ?How much do you trust strangers?? On a 1-to-5 scale? Category 5 2 1 1 1 1 1 1 1 1 *** Category 4 9 4 4 6 6 3 6 8 6 4. ?Do you trust most, many, a few or none of the people in your neighbourhood?? Trust most 55 31 33 48 33 23 43 45 42 *** 5. How likely is it that someone living close by would return a lost wallet or purse with the money in it? Very likely 51 46 45 45 37 32 45 45 45 *** Somewhat 38 41 38 41 37 46 41 43 41 6. How likely is it that a complete stranger would return a lost wallet or purse with the money in it? Very likely 5 3 3 5 5 7 6 4 5 *** Somewhat 50 27 25 40 38 35 42 43 40 Score on the overall social trust scale +.293 -.142 -.225 +.074 -.288 -.377 +.011 +.101 0.0^ N = (4724) (1848) (1382) (2344) (561) (2160) (1831) (2524) (17374) *** Chi-square values statistically significant at .000 ^ ANOVA between groups statistically significant at .000 Trust measures are highly significant statistically, with Chi-square values for each of the individual items that are significant beyond the .000 level.16 16 Given the large sample size in the present study, there is an issue of reporting results that, though statistically significant at conventional levels (i.e., p < .05, p < .01, and even p < .001), represent very minor substantive or practical effects. Therefore, I chose a more conservative alpha level than is conventionally used, and concentrate almost exclusively on discussing those findings that are significant at .000 or better. Other findings, for the most part, are not discussed or interpreted. 102 We also see, as expected from the previous literature, that it is the two most culturally distinct ethno-racial minority categories --- that is, the visible minority grouping and Aboriginal Peoples --- who generally rank the lowest on the individual measures of social trust, and also on the overall scale. As an illustration, only about 35 percent of visible minority respondents and 36 percent of Aboriginal Peoples choose category 4 or 5 on the scale measuring trust of other people in their neighbourhood. This compares with 58 percent among British respondents. A similar pattern occurs for item 4, where 55 percent of the British say that they trust most of the people in their neighbourhood, compared to 33 percent of Aboriginal Peoples and just 23 percent of visible minorities. The findings for visible minorities and Aboriginal Peoples are consistent with the hypothesis that groups with the greatest social distance from the dominant population, and those people facing the greatest social boundaries, will be among the least likely to develop trust in other people. Two other categories that tend to be below-average on all or most of the social trust measures in Table 3.6 are the French and the Canadians (the large majority of whom, as noted earlier, are Francophone Quebecers and hence mainly ?Canadiens?). This pattern is especially pronounced for the first item in Table 3.6, where 61 percent of the French and 65 percent of the Canadian category say that ?you cannot be too careful in dealing with people.? These percentages are considerably higher than the 44 percent of British respondents who hold this view. As is already evident from these comparisons, the respondents who consistently express the highest social trust are the British, with Other Europeans and the Mixed-Origins category 103 also above the average on all six items and the overall scale. The findings for the British and Other Europeans are again to be expected, given previous studies showing that these groups are relatively high on trust (recall, e.g., Soroka et al. 2006, 2007). Both of these groups presumably have comparatively little experience in Canada with social distance or social boundaries mechanisms that would negatively affect their sense of trust in other people. Although it is somewhat speculative, the finding of above-average social trust for the Mixed-Origins grouping also appears to support a social distance/social boundaries interpretation, since individuals who self-identify that they have mixed or multiple backgrounds are presumably among those Canadians who have been the most able or the most willing to lower the social barriers between ethnic groups, leading to higher levels of trust in others. The Association Between Ethno-racial Background and Political Trust In Table 3.7, the basic relationship between ethno-racial background and political trust is shown, both for the six component items and for the overall scale. As was the case for the social trust measures, the Chi-square tests once again show highly significant results, with p-values that are beyond .000 for each individual item. Even so, the results are different from those for social trust in some interesting and important ways. 104 TABLE 3.7 0.7 Political Trust Measures by Ethno-racial Background British (%) French (%) ?Canadian? (%) Other European (%) Aboriginal Peoples (%) Visible Minority (%) Other (%) Mixed Origins (%) TOTAL (%) Do you have a great deal, quite a lot, not very much, or no confidence at all in ... 1. the police? Great deal 36 42 40 34 30 34 35 36 36 *** Quite a lot 53 49 50 51 45 48 52 52 51 2. the justice system and courts? Great deal 12 17 16 12 14 26 15 12 22 *** Quite a lot 47 55 54 45 35 47 44 48 48 3. the health care system? Great deal 18 28 27 19 20 28 19 19 22 *** Quite a lot 55 54 56 52 46 46 52 54 53 4. the school system? Great deal 16 23 28 16 22 25 18 18 19 *** Quite a lot 60 61 56 59 50 55 55 59 58 5. the welfare system? Great deal 7 17 18 9 13 17 10 10 12 *** Quite a lot 38 56 53 36 33 45 40 43 43 6. federal parliament? Great deal 6 8 10 7 8 15 6 6 8 *** Quite a lot 34 46 40 37 27 41 37 37 37 Score on the overall political trust scale -.100 +.257 +.227 -.127 -.264 +.189 -.109 -.053 0.0^ N = (4724) (1848) (1382) (2344) (561) (2160) (1831) (2524) (17374) *** Chi-square values statistically significant at .000 ^ ANOVA between groups statistically significant at .000 105 First of all, whereas British respondents ranked the highest of all the ethno-racial groupings on each of the measures of social trust, the British do not rank the highest on any of the political trust items, and in fact fall below the mean on the overall political trust scale. Other Europeans and the Mixed-Origins category also generally rank in the bottom half of the eight ethno-racial categories on political trust, which contrasts with their above-average rankings on social trust, but again reflects their similarity to the British. The second important result is that visible minorities, who were lowest on social trust among all the ethno-racial groups, actually are among the most trusting on 5 of the 6 individual political trust items, and also rank among the highest groups on the overall political trust scale. It is only for the item about the police that their level of trust is low in comparison to other ethno-racial groupings, a finding that is consistent with previous Canadian and American research cited in Chapter 1. Even for the police, however, fully 82 percent of visible-minority respondents indicate that that have ?a great deal? or ?quite a lot? of confidence. Hence, we find initial support for the view that visible minorities may see the country?s major political institutions as significant and relatively effective organizations for furthering their interests and protecting their rights within Canada. A third finding of note, and one that tends to parallel the pattern for visible minorities, is that French and Canadian respondents, despite being among the lowest-ranked groupings on the questions about social trust shown in Table 3.7, rank as the most trusting or confident of all the ethno-racial categories when it comes to the country?s political institutions. This is a result that is consistent with some previous studies reviewed earlier in the present analysis, and so may not be entirely surprising (recall Henderson 2004; Soroka et al. 2007). Nevertheless, the difference in the way that the French and Canadian respondents rank on 106 political trust versus social trust is quite striking in these data, and underscores the importance of distinguishing between these two dimensions of trust when drawing conclusions about patterns of social cohesion in Canada, and when seeking to understand the underlying processes that are at work. It should be pointed out, of course, that several of the political institutions included in the set of political trust items, such as those dealing with health care and education, for example, are not officially areas of federal responsibility. Hence, French and Canadian respondents, most of whom live in Quebec, may well be thinking of their relatively autonomous and distinct set of provincial institutions when indicating their comparatively high level of trust in these public agencies. Even so, we also see that, on the item asking specifically about the federal parliament, the French and Canadian respondents again rank among the most trusting of all groups, with only the visible minority category matching their trust in this federal institution. A comparison of the results in Tables 3.6 and 3.7 reveals that the only ethno-racial grouping that ranks consistently low on both types of trust is the Aboriginal category. In other words, just as Table 3.6 showed that Aboriginal Peoples are among those expressing the lowest levels of social trust in Canadian society, Table 3.7 shows that, in relative terms at least, Aboriginal Peoples are also the least trusting of all the ethno-racial groupings when they are asked about political institutions. It is worth stressing that this does not mean Aboriginal Peoples are exceptionally distrustful in an absolute sense. For example, about three-quarters of Aboriginal respondents say they have a great deal or quite a lot of confidence in both the police and the schools, and about two-thirds say the same for the health care system. However, when compared to the 107 other major ethno-racial categories, Aboriginal Peoples are the one group that is lowest overall in both social and in political trust. The Relationships Between the Control Variables and Social Trust In Table 3.8, findings are presented that show the bivariate associations between each of the control variables and the overall scale of social trust. Among the more notable patterns are those for the three ethno-cultural controls. First, for immigration status, we see that the highest levels of social trust occur among long-term immigrants and, less so, among the Canadian-born, with recent and medium-term immigrants having the lowest social trust levels. This result hints at the possibility that there is an acculturation or assimilation effect at work, with immigrants becoming increasingly similar in social trust to native-born Canadians, the longer that they live in the country. Of course, it is also possible that this is a cohort effect, rather than a life-course effect. That is, perhaps Canada?s immigration system admitted immigrants who were generally more trusting in the past than the immigrants who have been admitted in recent decades, but this seems a less plausible interpretation. Another possibility is that more recent immigrants are, on average, younger than other Canadians, and younger respondents, as we can see in Table 3.8, are less trusting than other people; this possible effect of age is taken into account in the multivariate analysis in Chapter 4. 108 TABLE 3.8 0.8 Social Trust Scale Scores by All the Control Variables Immigration Status:^ Canadian born +.045 Immigrant since 2000 -.330 Immigrated 1990-1999 -.346 Immigrated 1980-1989 -.409 Immigrated 1970-1979 -.067 Immigrated before 1970 +.192 Religion:^ No religion +.014 Roman Catholic -.163 Protestant +.259 Non-Christian -.217 Other -.144 Language:^ Anglophone +.167 Francophone -.346 Allophone -.332 Education Level:^ Elem. or less -.299 Some h.s. -.227 h.s grad -.069 some postsec +.023 post sec dipl -.038 bachelor deg +.145 post-bach deg +.345 r = +.145, p = .000 Annual Income:^ < $20,000 -.388 $20-29,999 -.189 $30-39,999 -.148 $40-49,999 -.147 $50-59,999 -.041 $60-79,999 +.017 $80-99,999 +.130 $100,000+ +.270 r = +.179, p = .000 Voluntary Memberships:^ none -.234 one -.061 two +.170 three+ +.349 r = +.222, p = .000 Diversity of Friendships:^ none -.005 a few +.105 half -.062 most -.104 all -.272 r = -.092, p = .000 Region:^ Atlantic +.257 Quebec -.363 Ontario +.080 Prairies +.116 Br. Col. +.131 Age:^ 25-29 -.382 30-34 -.209 35-39 -.074 40-44 -.034 45-49 +.002 50-54 +.128 55-59 +.080 60-64 +.158 65-69 +.137 70-74 +.225 75-79 +.202 80+ +.230 r = +.162, p = .000 Gender: Males +.017 Females -.017 ^ ANOVA between groups statistically significant at .000 109 Second, in the case of religious affiliation, we see that Protestants are the most trusting, while non-Christians and Roman Catholics tend to have the lowest levels of social trust. The findings for the latter two affiliations may relate to the fact that visible minorities, who as we have seen are lower in social trust, are the most likely of all the ethno-racial groups to be non-Christians, while French and Canadian respondents, who are also below-average in social trust, are the most likely of all the groups to be Roman Catholic. (For evidence of these associations, see Table 3.2 above.) A parallel pattern seems also to be at work for the third ethno-cultural control variable, language. Here the results show that Francophones, who are mainly French or Canadian in ethno-racial background, and Allophones, who have a large visible-minority component (see Table 3.3), have lower levels of social trust than do Anglophones. The results for the two socioeconomic controls are generally consistent with expectations, although the effects are not strong. Using education, we find that the more highly educated respondents tend to express higher levels of social trust, and the correlation when education is measured in years is positive, but low (r = +.145, p = .000). For income, it is evident that those respondents living in households with higher incomes tend to have higher social trust, and that the correlation, when income is measured in thousands of dollars, is positive, but low (r = +.179, p = .000). Next are the two measures of social engagement. First, as expected, the number of voluntary associations that individuals belong to is positively related to social trust (r = +.222, p = .000). However, ethnic diversity of friendship networks is, contrary to expectations, negatively and weakly associated with social trust (r = -.092, p = .000). 110 As for the remaining control variables, Quebec is the only region in the country with below average social trust levels, which may be due to the high proportion of French and Canadian respondents in that region (see Table 3.4). For age, it is evident, as expected, that older respondents are more trusting of other people than are young people (r = +.162, p = .000), although the relationship is not strong. Finally, there is no real gender difference in social trust. The Relationships Between the Control Variables and Political Trust Table 3.9 repeats the analyses shown in Table 3.8, this time using the political trust scale as the dependent measure. The bivariate results in this case are somewhat more complex. First, the results for immigration status again suggest a possible acculturation or assimilation effect for political trust, although in a sense it reverses the pattern for social trust shown in Table 3.8. That is, we see here that, when asked about political institutions, the lowest level of trust is among long-term immigrants, who basically match the Canadian-born in this regard. The highest levels of political trust occur for the most recent immigrants. This outcome is consistent with the view discussed in Chapter 1, which is that new immigrants have a relatively high level of confidence in the country?s political institutions, perhaps because they see them as mechanisms for protecting their rights and interests. However, after immigrants have been in the country for a longer time, they may become more and more like their Canadian-born counterparts and develop a comparatively less positive impression of the nation?s public agencies and organizations. 111 TABLE 3.9 0.9 Political Trust Scale Scores by All the Control Variables Immigration Status:^ Canadian born -.018 Immigrant since 2000 +.350 Immigrated 1990-1999 +.091 Immigrated 1980-1989 +.031 Immigrated 1970-1979 +.016 Immigrated before 1970 -.117 Religion:^ No religion -.210 Roman Catholic +.138 Protestant -.063 Non-Christian +.173 Other -.058 Language:^ Anglophone -.134 Francophone +.327 Allophone +.171 Education Level:^ Elem. or less +.073 Some h.s. +.013 h.s grad -.026 some postsec -.085 post sec dipl -.076 bachelor deg +.120 post-bach deg +.089 r = +.024, p = .001 Annual Income:^ < $20,000 +.011 $20-29,999 +.079 $30-39,999 -.010 $40-49,999 -.015 $50-59,999 -.006 $60-79,999 -.016 $80-99,999 +.029 $100,000+ +.002 r = -.004, p = .622 Voluntary Memberships:^ none +.019 one -.017 two -.011 three+ -.001 r = -.008, p = .273 Diversity of Friendships:^ none +.055 a few -.046 half +.023 most -.082 all -.180 r = -.031, p = .000 Region:^ Atlantic +.018 Quebec +.277 Ontario -.021 Prairies -.165 Br. Col. -.234 Age:^ 25-29 -.011 30-34 +.057 35-39 +.005 40-44 -.049 45-49 -.049 50-54 -.074 55-59 -.062 60-64 +.007 65-69 +.057 70-74 +.080 75-79 +.098 80+ +.198 r = +.030, p = .000 Gender: Males -.016 Females +.016 ^ ANOVA between groups statistically significant at .000 112 A similar reversal of the social trust results shown in Table 3.8 seems to occur in Table 3.9 for the relationships between political trust and both religion and language. We find that it is Catholics and non-Christians, rather than Protestants, who express the most political trust, as do Francophones and Allophones, when compared to Anglophones. These findings are the opposite of those shown for social trust in Table 3.8, and again suggest that French, Canadian, and visible-minority respondents, who have higher political trust levels than the other ethno-racial groups, are probably driving these patterns. The results for the two socioeconomic control variables do not show a clear positive relationship with political trust. The linear association between education and political trust is marginally significant and very weak (r = +.024, p = .001); in addition, an inspection of the mean political trust scores for each of the seven ranked categories of education reveals a curvilinear pattern, with the lowest political trust being found among those people in the middle of the education range and the highest trust levels being expressed by the those respondents with the most and the least education. For this reason, I convert educational attainment to a set of dummy variables in the multivariate analysis of political trust, so as to be able to display and interpret this apparent curvilinear influence. In the case of income, a very similar pattern occurs, with no evidence of a linear effect on political trust (r = -.004, p = .622), and instead a curvilinear association in which people with the lowest and the highest income levels indicate more trust in political institutions than do middle-income earners. Here as well, then, income will be converted into a set of dummy variables for the multivariate analysis of political trust. The two social engagement measures do not behave as expected in their relationships with political trust. There is no relationship between voluntary association membership and 113 political trust (r = -.008, p = .273). Having an ethnically diverse friendship network appears to have a weak linear correlation with political trust, but, contrary to expectations, it is negative (r = -. 031, p = .000); moreover, an inspection of the political trust scores across the five categories of this control variable reveals that it actually has a complex non-linear association. Given these results, I create a set of dummy variables for this predictor in the multivariate analysis of political trust. The remaining control variables show the following bivariate relationships with political trust. For region, we again see a reversal of the pattern for social trust reported in Table 3.8, with Quebecers expressing the highest political trust of all the regional sub-groups; this is probably attributable to the large proportion of French and Canadien respondents living in Quebec (see Table 3.4). Age is positively but very weakly related to political trust (r = +. 030, p = .000). Finally, there is no appreciable gender difference in political trust. SUMMARY OF THE BIVARIATE FINDINGS This completes the description of the bivariate associations linking ethno-racial background to both social and political trust, and also the bivariate associations between all the control variables and the two trust measures. The general pattern is for minority groups, specifically visible minorities, Aboriginal Peoples, and the French and Canadian (mainly Canadien) respondents, to be lower in social trust than the British and Other Europeans. This finding is consistent with expectations regarding the first of my main research questions. In contrast, several of the minority categories, specifically visible minorities, the French, and 114 the Canadians, but not Aboriginal Peoples, have relatively high levels of political trust, while the British and Other Europeans are comparatively low on this measure. In Chapter 4, I undertake a series of detailed multivariate analyses that will enable me to determine which of the control variables is most important for explaining these ethno-racial variations in social and political trust. 115 CHAPTER FOUR: MULTIVARIATE ANALYSIS In Chapter 3, we established that there are some important differences in the levels of social trust and political trust exhibited by the eight major ethno-racial categories under consideration in my thesis. While members of what might be termed the traditionally dominant population --- including the British and Other Europeans, in particular --- show relatively high levels of social trust, those individuals who belong to groups that are the most culturally distinct from this population, especially visible minorities and Aboriginal Peoples, as well as the French and Canadian categories, reveal comparatively low levels of trust in other people. With respect to political trust, however, we saw in Chapter 3 that the pattern is close to the reverse; that is, these same distinctive ethno-racial groups, for the most part, have higher levels of trust or confidence in Canada?s public institutions than do the British and Other European respondents. The one key exception to the latter assertion is Aboriginal Peoples, who, at least in relative terms, express lower levels of political trust than do all the other groups. Having established these base-line results, I now address the other central research questions that were outlined in Chapter 1. That is, I determine the extent to which the three sets of explanatory factors discussed in Chapter 1 help to account for the ethno-racial differences in social and political trust that are evident in the analysis so far. In addition, I assess the potential predictive effects on social and political trust of three other control variables: region, age, and gender. The multivariate analyses begin with a set of models that focus on social trust, with the results presented in Tables 4.1 to 4.4. Then, in Tables 4.5 to 4.8, I test a set of models for predicting political trust. 116 PART I: MODELLING SOCIAL TRUST Ethno-racial Background, Social Trust, and the Three Ethno-cultural Markers In Table 4.1, the results of a series of unstandardized multiple regression analyses are provided, showing four different models. Model 1 is the base-line bivariate model, in which the main independent variable, ethno-racial background, is used to predict social trust. The findings in Model 1 essentially correspond to the results for the overall scale of social trust in Table 3.6; in this case, however, the social trust scores are relative to the British reference group. The remaining three models indicate how the effects of ethno-racial background vary with the introduction of the three ethno-cultural controls, beginning with immigration status, then religion, and finally language. Model 1 Using the British as the reference category, Model 1 shows once again that, without controls for other factors, the lowest levels of social trust are to be found, first, among visible minorities (b = -.567, p = .000), followed by Aboriginal Peoples (b = -.478, p = .000), the Canadian category (b = -.415, p = .000), and the French (b = -.332, p = .000). In fact, all seven ethno-racial groupings shown in Model 1 have significantly lower social trust than do the British, although the differences for people in the Mixed Origins, Other European, and Other categories are relatively small. 117 TABLE 4.1 0.1 118 Model 2 When immigration status is introduced as a control in Model 2, the only real change of note is that the visible minority score weakens (from b = -.567 to b = -.406), although it still remains significant beyond the .000 level. This indicates that part of the reason why visible minorities express lower trust in people than do other groups is that they are the most likely of all the groups to be foreign-born. In truth, visible minorities are the only group among the eight considered in the analysis that is majority foreign-born (see Table 3.1); this also explains why controlling for immigration status has little or no effect in changing the social trust scores for the other ethno-racial groups. The second point of note in Model 2 concerns the effect of immigration status itself. The coefficients show that immigrants who arrived in the country during the period from 1980 to the present are significantly lower in social trust than the Canadian-born; however, immigrants who arrived in the 1970-1979 period are no less trusting of other people than are the Canadian-born, while those who arrived before 1970 are actually somewhat more trusting than the Canadian-born (b = +.123, p = .000). These findings are consistent with the hypothesis suggested earlier that immigrants who have lived in Canada longer may tend to become more similar to native-born Canadians in their level of social trust. One cautionary note is that this pattern is not linear or monotonic, since the biggest gap in social trust is not for the most recent cohort of immigrants, who arrived since the year 2000 (b = -.202, p = .000), but instead for the cohort that arrived in the 1980-1989 period (b = -.322, p = .000). Even so, these results support the conclusion that being an immigrant is an important marker of ethno-cultural distinctiveness that can affect social trust, and that the effects on social trust of being an immigrant may indeed fade with time. 119 Model 3 Model 3 shows how the differences in trust across the different ethno-racial groupings are altered when religion is also included in the analysis. Again we see a decline in the gap in social trust for the visible minority grouping (from b = -.406, p = .000, to b = -.373, p = .000). This change fits with the interpretation that visible minorities show lower levels of social trust in part because they are disproportionately non-Christian (see Table 3.2) and, as Model 3 also shows, non-Christians are less trusting of other people than are the Protestant reference group (b = -.201, p = .000). This supports the argument that being identifiably different from the Christian religious ?mainstream? may be one reason behind the lower levels of social trust among some minorities. Respondents with no religious affiliation, who likewise could be considered outside the religious mainstream, also express lower levels of social trust (b = -.196, p = .000). The most substantial changes in trust levels that occur with the religion control in Model 3 concern the patterns for the French and Canadian categories. The differences in trust between both of these groupings and the reference category drop substantially when religion is taken into account. In the case of the French the score is cut in half (from b = -.332, p =.000, to b = -.165, p = .000). The reason for this change seems clearly to be that Roman Catholics have significantly lower levels of social trust than the Protestant reference group (b = -.363, p = .000), and a large majority of the French (as well as the Canadian category) are Roman Catholic (see Table 3.2). Here again we can see that the ethno-cultural marker of religion has implications for the amount of social trust that some Canadians feel relative to others. 120 Model 4 The last model to consider in Table 4.1 is Model 4, which adds language affiliation to the regression equation. In this model, all three of the ethno-cultural explanatory controls are included in the analysis. The results again indicate a slight decline in the difference between visible minorities and the reference group, which is partially attributable to the fact that visible minorities are majority Allophone (see Table 3.3), and Allophones, as shown in Model 4, are less trusting of people than the Anglophone reference group (b = -.164, p = .000). What may be even more notable in Model 4, however, is that when language is taken into account, both the French coefficient (b = +.065, p = .03) and the Canadian coefficient (b = -.077, p = .02) reduce to very small values with marginal significance levels. The interpretation here is that one of the key reasons why French and Canadian respondents were found in Model 1 to have lower social trust is that the vast majority of these two groups are Francophone (see Table 3.3), and Francophones are considerably less trusting than their Anglophone counterparts (b = -.477, p = .000). In other words, then, it is not French ethnic background per se that explains the relatively low levels of trust in this group; instead, it is the combination of being distinctly Francophone and predominantly Roman Catholic that lies at the root of this tendency. This is an argument that also applies to the Canadian category. I explore the theoretical implications of this particular finding later in the thesis, and also incorporate into the analysis the related factor of region of residence, which takes into account the fact that most Francophones live in Quebec. As for the other ethno-racial categories analysed in Models 1 through 4, probably the key 121 finding concerns Aboriginal Peoples, who remain considerably less trusting than other Canadians, with or without controls for the three ethno-cultural factors of immigration, religion, and language. Aboriginal Peoples have one of the lowest overall social trust scores in Model 1 (b = -.478, p = .000) and this score is almost unchanged by controlling for these three factors in Model 4 (b = -. 434, p = .000). Therefore, in the specific case of Aboriginal Peoples, I do not find support for my hypothesis that the three ethno-cultural markers, and the social distance/social boundaries processes that they may promote, lie behind the lower levels of social trust of ethno-racial minority groups. For several of the other minorities, however, the incorporation of these ethno-cultural factors does clearly help us to understand why some groups have lower social trust than others. Ethno-racial Background, Social Trust, and the Effects of Socio-economic Status Model 5 and Model 6 I now turn to my third research question, which asks whether socioeconomic or class-related influences are important in explaining why some ethno-racial groups have more social trust than others. The expectation is that both of the two socioeconomic factors, education and income, will have a positive effect on social trust, and that controlling for their effects will weaken the association between ethno-racial background and trust in other people. With these hypotheses in mind, a second multiple regression analysis was conducted, as shown in Table 4.2. In this table, the base-line model, Model 1, is displayed and compared to two other models, in which education (Model 5) and then income (Model 6) are brought into the equation. 122 TABLE 4.2 0.2 The results show, as expected, that both education (b = +.040, p = .000) and income (b = +.004, p = .000) are positively related to social trust. These findings indicate that, for each one-year increment in education, respondents? social trust rises, on average, by .04 units on the overall social trust scale, and that, for each one-thousand dollar rise in annual household income, respondents? social trust rises by .004 units on the overall social trust scale. Nevertheless, while the main effects of education and income are as predicted, controlling for their influences has only a limited impact in accounting for any ethno-racial variations in 123 social trust. When education is added in Model 6, there is a reduction in the trust gap between the reference group and two of the other categories, i.e., Aboriginal Peoples and Canadians; the reductions are slight but lend minor support to the hypothesis that these two groups have less trust in other people because of having below-average educational attainment (see Table 3.5). For some ethno-racial groupings, however, especially the visible minority category, the gap actually increases slightly (from b = -.567 in Model 1 to b = -.616 in Model 5); this finding indicates that, if it were not for their above-average education levels (see Table 3.5), visible minorities would express even less social trust than they do. To some extent, the same pattern occurs for the Other and Mixed Origins groups as well. When income is added in Model 6, there are some further minor changes in the social trust coefficients for the ethno-racial groupings, with slight declines for most of the groups, including Aboriginal Peoples, visible minorities, the French, and the Canadians. These small shifts do not demonstrate any strong income effect on the relationship between ethno-racial background and trust, but they do indicate that some of the gap in trust between these groups and the reference category is attributable to these groups having below-average incomes. Overall, then, the results in Table 4.2 lend only minor support to the hypothesis that socioeconomic factors mediate the relationship between ethno-racial background and trust in other people in Canada. 124 Ethno-racial Background, Social Trust, and Social Engagement Influences Model 7 and Model 8 The third set of explanatory controls to consider involves the two social engagement factors. In this set of models, two related research questions are addressed. First, is it true that individuals who belong to voluntary associations will be more likely than other individuals to develop a sense of trust in other people, and does this influence account for some of the ethno-racial differences in social trust? Second, are respondents more likely to develop a sense of trust in others if their friendship networks include people whose ethnic backgrounds are different from their own, and does this influence partly account for ethno-racial variations in social trust? In Table 4.3, I test these hypotheses, by comparing the base-line Model 1 with Model 7, in which voluntary association membership is taken into account, and then with Model 8, in which the diversity of friendships variable is also added to the equation. The results show, as expected, that belonging to voluntary associations is positively associated with social trust and that this association is monotonic, with the lowest trust level found among the respondents with no memberships (the reference category) and the coefficient increasing steadily as we move to one, to two, and then to three or more memberships. This pattern is evident in Model 7, and also in Model 8, when the diversity of friendship control is added. 125 TABLE 4.3 0.3 126 As for diversity of friendship, it has a significant overall effect, given that the block F-test for this variable has a p-value of .000. However, the pattern is far from monotonic or linear, as shown in Model 8. More specifically, it appears that respondents who have a few ethnically different friends are marginally higher in social trust than those who have no such friends (b = +.047, p = .003), but respondents whose friends are about half or mostly from different ethnic backgrounds are not significantly different from respondents who have no friendship diversity; moreover, there is some evidence that respondents whose friends are all ethnically different from themselves actually have slightly lower social trust than people with no friendship diversity at all (b = -.163, p =.000). This pattern is difficult to interpret, although the effects are rather small in any event. I speculate on their possible theoretical implications later in the thesis. The other findings shown in Models 7 and 8 suggest that the two social engagement controls have only minor effects on the pattern of ethno-racial differences in social trust. The diversity of friendship control variable is especially weak in this regard. It does appear that the lower levels of social trust revealed by the visible minority category, as well as the French and the Canadian groups, are partly related to these groups having lower voluntary activity than the British reference category, which ranks first of all eight groups on voluntary membership (see Table 3.5). As an illustration, we see that the gap in social trust for visible minorities drops from b = -.567 (p = .000) to b = -.485 (p = .000), when voluntary membership as added in Model 7. Still, the effects of the social engagement factors in accounting for ethno-racial differences in social trust are relatively minor. 127 The Full Explanatory Model of Social Trust: Model 9 and Model 10 We now have a sense of how the three sets of explanatory controls, when considered separately, affect the social trust levels of different ethnic groups in Canada. To this point, it appears that the three ethno-cultural factors have somewhat more influence in accounting for ethno-racial differences in trust than either the socioeconomic or the social engagement variables, although the R-square values for all three sets of models in Tables 4.1 to 4.3 are quite small. Still, we have yet to consider the combined influences of all these controls operating together, which may alter the overall pattern of relationships. The results of this combined analysis (Model 9) are shown in Table 4.4, along with a final set of results (Model 10) that includes the three other control variables included in the thesis: region, age, and gender. 128 TABLE 4.4 0.4 129 Model 9 The first finding that stands out from Table 4.4 is that the differences in social trust across the eight ethno-racial categories are affected substantially when all the ethno-cultural, socio-economic, and social engagement control variables are included in Model 9. The biggest changes occur for the French and Canadian groupings, which are shown to have the same levels of social trust as the British when all of these factors are taken into account. This 130 outcome, which also applies to the Mixed Origins category, indicates that the three sets of explanations for ethno-racial variations in social trust apply very well to these particular groups, and help us to understand that ethno-cultural influences, socioeconomic factors, and social engagement differences are primarily behind the lower levels of social trust among the French and Canadian (or Canadien) members of the national community. The combination of these controls also reduces the trust gap for the two most culturally distinct categories: the visible minorities and Aboriginal Peoples. The reduction from Model 1 to Model 9 is more substantial for visible minorities than for Aboriginal Peoples, and both groups still express relatively low levels of trust in other people, even when all three sets of explanatory factors are taken into account. Therefore, further research is necessary if we are to understand fully the reasons for the lower levels of social trust exhibited by these two groups. As for the other predictors of social trust displayed in Model 9, I can briefly note their effects, which are little different from those shown in previous models. First, there is the same tendency for recent immigrants to be less trusting than either long-term immigrants or the Canadian-born. As before, we find again that all the other religious groups are less trusting than are Protestants. We see once more that Francophones are less trusting than Anglophones, although the tendency in earlier models for Allophones also to have less trust is far less evident in the full model, and of marginal statistical significance (b = -.070, p = .020). Education, income, and voluntary association membership still show the same positive effects on social trust as in previous models. Finally, the tendency for ethnic diversity of friendships to have the unexpected, though weak, negative relationship to social trust, holds when all the main explanatory controls are incorporated in Model 9. 131 Model 10 The final model used to predict social trust is Model 10, which adds region, age, and gender to the variables used in Model 9. Two of these three variables, region and age, have significant main effects, but gender has no effect whatever and so is not discussed further in this section. It is not necessary to examine in detail the Model 10 results, most of which do not change substantially from Model 9. Probably the most notable finding is that the trust gaps for visible minorities (b = -.286, p = .000) and Aboriginal Peoples (b = -.246, p = .000) reduce further, although they remain significant. The main reason for this further reduction is the age control. These two groups are the youngest ethno-racial categories considered here (see Table 3.5) and, as Model 10 shows, younger people have less social trust than older people (b = +.014, p = .000). Controlling for age also substantially weakens the tendency for recent immigrants to be less trusting than others, reflecting that recent immigrants are considerably younger than long-term immigrants or the Canadian-born.17 Another important change from Model 9 to Model 10 concerns the social trust coefficient for Francophones, which drops by more than half, from b = -.403 (p = .000) in Model 9 to b = -.156 (p = .000) in Model 10. This clearly occurs because of the control for region in Model 10, and stems from the fact that the vast majority of Francophones live in Quebec, and that, as shown in Model 10, Quebec residents are the most likely of all respondents to exhibit 17 The relationship between age and immigration is not shown in tables, but the results indicate, for example, that immigrants who came to Canada after 2000 had an average age of 37.7 years, compared to 65.2 for immigrants who came before 1970, and 49.4 for the Canadian-born. 132 low social trust. Residents of the Prairies and British Columbia match the Ontario reference group in social trust, while Atlantic residents are somewhat more trusting than Ontarians. PART II: MODELLING POLITICAL TRUST Ethno-racial Background, Political Trust, and the Three Ethno-cultural Markers In Part II of the multivariate analysis, our attention turns to political trust. I follow a similar strategy to that used in Part I of Chapter 4, where the focus was on social trust, and now test the effects for each of the three sets of explanatory factors on ethno-racial differences in political trust. I then consider a full model that includes all of these factors together, and finally add controls for region, age, and gender to complete the analysis. Model 1 in Table 4.5 displays the differences in political trust, before controls, across the eight ethno-racial categories. Models 2, 3, and 4 show how the Model 1 patterns change as each of the three ethno-cultural control variables is added to the analysis. In Model 1, we find results that parallel those found in Table 3.7 in Chapter 3; in this case, however, the political trust scores are relative to the British reference group. 133 TABLE 4.5 0.5 134 Model 1 The findings indicate again that the French (b = +.329, p = .000), Canadian (b = +.298, p = .000), and visible minority (b = +.260, p = .000) respondents are all significantly more trusting of the country?s political institutions than are the British reference category. The Mixed Origins and Other groupings are no different from the British in political trust, while the Other Europeans are similar to the British, though perhaps marginally less trusting. As before, the one ethno-racial group that stands out as the least trusting of government institutions is the Aboriginal Peoples (b = -.193, p = .000). Model 2 When immigration status is added to the equation in Model 2, the most notable change is that the tendency for visible minorities to exhibit more political trust than the reference group decreases somewhat. This occurs because visible minority respondents are much more likely to be immigrants, especially recent immigrants, than any of the other ethno-racial groupings and, as shown in Model 2, recent immigrants are the most trusting of all when it comes to government agencies. We see, as well, that immigrants who have been in Canada for a longer period are no more trusting of their political institutions than are the Canadian-born. This outcome fits once again with an assimilation or acculturation interpretation, although it is the reverse of the pattern for the social trust dependent variable discussed earlier in this chapter; that is, we found previously that recent immigrants have less social trust than the Canadian-born, but that longer-term immigrants seem to increase in social trust, and become much like the Canadian-born in this regard. For political trust, on the other hand, we see that recent 135 immigrants trust the government more than do native-born Canadians, but that longer term immigrants do not, suggesting that immigrants gradually come to resemble their Canadian-born compatriots in being relatively less trusting of government. Model 3 Adding religion to the analysis in Model 3 changes a number of the political trust coefficients for the ethno-racial categories. There is a reduced tendency for the French and the Canadians to be more trusting of their political institutions than the reference group, which is due to these two groups being predominantly Roman Catholic, and Roman Catholics having more political trust than Protestants (b = +.139, p = .000). Similarly, the visible minority coefficient also drops somewhat, reflecting that this group is disproportionately non-Christian and non-Christians have more political trust than Protestants (b = +.126, p = .000). As for those respondents who belong to no religion, they exhibit significantly less political trust than the reference group (b = -.173, p = .000). Hence, respondents with no religious affiliation are the one religious category that is low on both social trust and political trust in this analysis. Model 4 The inclusion of the third ethno-cultural control variable, language, clarifies further the underlying reasons why some ethno-racial groupings have more political trust than others. It is evident, first of all, that the French and the Canadians, who express the highest political trust of the eight groups without controlling for other influences, are no longer significantly 136 different from the reference group when the three ethno-cultural controls (especially religion and language) are incorporated in Model 4; the score for the French is now b = +.044 (p >.05), and for the Canadians the score is b = +.051 (p > .05). This outcome suggests that the French and Canadians retain more political trust than other Canadians precisely because they are predominantly Roman Catholic and Francophone. One plausible interpretation is that, being largely composed of members who are both religious and linguistic minorities, the French and Canadians see the government as an effective bulwark for protecting their minority rights, perhaps because of the establishment of such mechanisms as the Charter of Rights and Freedoms (including Section 33, the so-called ?notwithstanding clause,? which allows provincial governments to override the Charter in some circumstances). A somewhat similar argument can be applied to the visible minority category, most of whom are immigrants and many of whom are non-Christian (or non-Protestant) and Allophone. Taking into account their minority status on all three of these ethno-cultural dimensions appears partly to explain their positive views about government, since their higher levels of political trust relative to the reference category reduce from Model 1 (b = +.260, p = .000) to Model 4 (b = +.117, p = .000). Still, visible minorities exhibit above-average political trust, even when immigration status, religion, and language are controlled. In regard to the other ethno-racial groupings, the one category that stands out from the rest is again the Aboriginal Peoples. Aboriginal Peoples respondents have the lowest political trust coefficient in Model 1 (b = -.193, p = .000) and this result is not reduced (and actually increases slightly) with the addition of the controls in Model 4 (b = -.205, p = .000). As also occurred with the social trust dependent variable, then, the Aboriginal Peoples express the lowest level of political trust among all the ethno-racial groups and this outcome cannot be explained by 137 influences that are associated with the ethno-cultural markers of immigration status, religion, and language. Ethno-racial Background, Political Trust, and the Effects of Socio-economic Status Model 5 and Model 6 Table 4.6 shows the effects of controlling for education (Model 5) and then income (Model 6) in the analysis of ethno-racial differences in political trust. Because of the evidence in Chapter 3 that both education and income had non-linear effects on political trust, these two variables are converted to categorical measures in Table 4.6, with post-bachelor degree or higher used as the reference category for education, and a level of $100,000 or higher used as the reference category for income. The multivariate analysis in Model 5 reveals a significant non-linear effect for education, in which respondents who are either at the lowest or the highest end of the education spectrum are more trusting of government than those in the middle range. This pattern holds in Model 6, when income is introduced into the analysis, but income itself is found to have no appreciable effect, non-linear or otherwise, on political trust, with a block F-test that is not significant even at the .05 level. The low R-square values for Models 5 and 6 indicate that neither of the socioeconomic factors is especially important in accounting for variations in political trust in the present study, and controlling for these two factors has little impact on the political trust levels of the eight ethno-racial groups. 138 TABLE 4.6 0.6 139 Ethno-racial Background, Political Trust, and Social Engagement Influences Model 7 and Model 8 In Table 4.7, in Models 7 and 8, we consider the effects on political trust of the two social engagement influences: voluntary association membership and ethnic diversity of friendships. The results in these models lend little or no support to the hypothesis that differences in social engagement factors, at least as measured in the present analysis, have an important impact in explaining ethno-racial differences in political trust. Comparing Model 1 to both Model 7 and Model 8 reveals that the political trust coefficients for the eight ethno-racial categories are virtually unchanged by the two control variables. Voluntary membership actually has no effect on political trust, as shown by the non-significant block F-test for this predictor. Ethnic diversity of friendships does have a significant effect, although it is both weak and again contrary to expectations. The findings show that, if anything, the more ethnically diverse the respondent?s circle of friends, the lower is her or his level of political trust. It will be recalled that this result is similar to the effect of friendship diversity on social trust reported earlier. The R-square values for Models 7 and 8 are very low, at less than 3 percent. Such findings make it clear that the reasons why some respondents have more trust in government than others has very little to do with their voluntary membership activity or the ethnic diversity of their friendship networks. 140 TABLE 4.7 0.7 141 The Full Explanatory Model of Political Trust: Model 9 and Model 10 In Part II of Chapter 4 we have found so far that, as was the case for the social trust dependent variable considered in Part I, the three ethno-cultural factors of immigration status, religion, and language appear more important for explaining ethno-racial differences in political trust than do the socioeconomic factors or social engagement influences being considered in the thesis. It still remains, however, to assess the patterns when all three sets of explanatory controls are included simultaneously. Table 4.8 displays the results of this analysis in Model 9, while Model 10 shows the results with region, age, and gender added to the equation. These models have been trimmed and simplified somewhat, by excluding both income and voluntary membership from the analysis; these exclusions were made because the block F-tests for these two variables in the full model were found to be non-significant. 142 TABLE 4.8 0.8 Continued on next page 143 144 Model 9 With the three sets of explanatory controls included together in Model 9, there are few major changes of note from the findings discussed previously. Once again, the French and Canadian respondents have trust levels that are not significantly higher than those for the British and most other groups, when all the controls are incorporated. Once again, as well, Aboriginal Peoples continue to rank the lowest on political trust, whether the control variables are taken into account or not. The overall R-square of just 6.6% underscores the weak predictive value of adding the three sets of explanatory controls. The R-square in this model is less than half the size of the 17.8% R-square attained for the social trust measure in Table 4.4, suggesting that the explanatory variables used in the thesis do a much better job of modelling social trust than they do of predicting political trust. Model 10 The final model, Model 10, provides only a small amount of new information for understanding why some Canadians have more political trust than others in the present analysis. The differences in political trust across the eight ethno-racial groupings are essentially unchanged by controlling for gender (which again has a non-significant effect), region, and age. Age has the expected positive effect on political trust (b = +.003, p = .000). The regional pattern is significant, but is different from that for social trust. While Quebecers exhibited the lowest social trust in the country, even with controls on all other factors, it is the residents of the two western regions --- the Prairies and British Columbia --- who have 145 the lowest political trust scores when the full set of predictors is incorporated into the analysis. SUMMARY OF THE MULTIVARIATE FINDINGS There are a number of important and interesting findings that are revealed in the multivariate analysis. Here I briefly summarize the key findings, before moving to the conclusion, where I discuss their theoretical significance and substantive importance for understanding ethno-racial differences in social and political trust in Canada. One of the most consistent findings in the analysis has been the clear tendency for the most culturally distinctive ethno-racial groups --- that is, visible minorities, Aboriginal Peoples, the French, and the Canadian (or Canadien) category --- to differ considerably from the one group that traditionally has been the most established or dominant in the population, namely, the British. This finding is true, moreover, for both dimensions of trust; however, an intriguing outcome is that the rankings of the ethnic groups are almost the reverse, depending on which trust dimension is compared. That is, visible minorities, Aboriginal Peoples, and French/Canadian respondents generally rank the lowest in social trust of all the groups, with the British highest and several of the remaining groupings, especially Other Europeans and the Mixed Origins category, ranking relatively close to the British. In contrast, the same culturally distinctive ethno-racial groups, with the important exception of Aboriginal Peoples, generally rank the highest in political trust, with the British, Other Europeans, and Mixed Origins respondents toward the bottom on this trust dimension. That Aboriginal Peoples rank at the bottom on both types of trust is a key finding that will receive further 146 attention and discussion in the concluding chapter. Another result, which is also consistent for both social trust and political trust, is that the three ethno-cultural explanatory factors --- immigration status, religion, and language --- proved to be more effective than either of the two socioeconomic factors --- education and income --- or the two social engagement factors --- voluntary association activity and diversity of friendship networks --- in accounting for ethno-racial differences in trust. This outcome is apparent from comparing the reductions in the trust gaps across the eight ethno-racial categories when the three sets of explanatory factors are considered in separate models. The greater impact of the ethno-cultural factors is especially evident for the French, Canadians, and visible minorities. It is far less apparent in the case of Aboriginal Peoples, for whom the socioeconomic factors are relatively more important in accounting for their gap in trust with the British reference group. These results answer one of the key research questions in the thesis, concerning whether the ethno-cultural explanations for differences in trust are more powerful than the other explanations. As for the main findings involving the three ethno-cultural factors, they are as follows: the traditionally more ?established? or dominant groups on each of these dimensions, i.e., the Canadian-born, Protestants, and Anglophones, are in general relatively high on social trust and low on political trust, while, in contrast, the less established or traditionally less dominant groups, i.e., immigrants (especially recent immigrants), both non-Christians and Catholics, and both Allophones and Francophones, are relatively low on social trust, and are either higher on political trust than the dominant groups or at the same level. The fact that these more or less reverse effects occur for social and political trust, even with controls for other variables, is one of the more interesting patterns of findings in the thesis, and 147 something that I will discuss further in the concluding chapter. One other notable example of a reverse pattern concerns the effect of immigration status on social and political trust. Although this result is not completely monotonic, the findings generally show that, the longer immigrants are in Canada, the more likely they are to resemble the Canadian-born in both types of trust; that is, they come to be more trusting of other people, much like the Canadian-born, but less trusting of political institutions, which is again much like the Canadian-born. The effects of the two socioeconomic factors on social trust are consistent with the expectations that were set out in the research questions in Chapter 1. That is, both education and income have positive effects on social trust in all the multivariate models, although their effects are weaker than was originally anticipated. For political trust, however, the effects appear to be even weaker, especially for income, which has no effect on political trust in the multivariate analysis. Education has a weak curvilinear relationship to political trust with multivariate controls, with people in the middle of the education range expressing somewhat less trust than the lowest and the highest educated individuals. The multivariate results for the two social engagement factors are quite mixed, with voluntary association membership having a consistent positive relationship to social trust, as expected, but showing no relationship whatever to political trust, which is contrary to expectations. Even more unexpected are the findings involving ethnic diversity of friendships, which has a weak but negative effect on both social trust and political trust in the multivariate models. The multivariate findings, especially those for the full models, also show that the three sets of explanatory factors do a good job in accounting for the trust gaps involving several of 148 the ethno-racial groups. This is quite apparent for the French and Canadian categories, which are no different from the British reference group on either social trust or political trust when all the explanatory controls are included in the models. To a lesser extent, the same pattern occurs for visible minorities, whose gaps in both social trust and political trust compared to the British are still significant with controls, but are reduced by about half when all the factors are taken into account. Aboriginal Peoples, however, remain lowest on both social and political trust, with or without controls, and the trust gap between them and the reference category changes only modestly when the three sets of explanatory factors are taken into consideration. Aboriginal Peoples, therefore, stand out as the one ethno-racial minority in the analysis who fit least well with the social distance/social boundaries hypothesis, at least as laid out and tested here. That is to say, while it still seems likely that social distance and social boundaries processes are at work in the case of Aboriginal Peoples, the causes for their relatively low levels of trust, both in other people and in the society?s public institutions, go beyond the ethno-cultural, socioeconomic, and social engagement explanations that I have tested in the present analysis. Regarding the three additional control variables and their effects, there are virtually no gender differences on either type of trust. Age, however, has a positive effect on both social trust and political trust and, in the final models, age is also seen to account for at least some of the lower social trust among visible minorities and some of the lower social trust and lower political trust among Aboriginal Peoples; both visible minorities and Aboriginal Peoples are significantly younger than the rest of the population. Controlling for region in the final models shows that it has a notable bearing on both social trust and political trust: 149 Quebec residents are clearly the least likely to express trust in other people, even with all other factors statistically controlled, and residents of the Prairies and British Columbia exhibit the lowest trust in government agencies. Finally, it should be noted that the overall model (Model 10) for social trust has better explanatory power than the comparable model for political trust, with R-square values of about 18 percent versus 7 percent, respectively. 150 CHAPTER FIVE: CONCLUSIONS In this thesis, I set out to add to our understanding of the nature of social cohesion in Canadian society, by addressing the question of whether ?in Canada we trust?? More specifically, I wished to determine if members of some ethno-racial communities are more or less likely than others to have high levels of trust, both in their fellow Canadians and in their society?s central public or political institutions. By focusing on patterns of social and political trust, I have examined what leading scholars see as the two most important indicators or dimensions of social cohesion in Canadian society (recall Breton, Reitz, and Valentine 1980; Soroka, Johnston, and Banting 2007). Drawing on theoretical insights gleaned from the literature on social distance and social boundaries formation, I derived my main working hypothesis. That hypothesis was that those ethno-racial groups who were the most culturally distinct from the traditionally dominant or ?mainstream? members of the population, especially British Canadians, would express different levels of both social and political trust. That is, I expected that the most culturally distinctive groups, which include visible minorities, Aboriginal Peoples, and French Canadians, would experience greater social distance and more serious social boundaries in relation to the rest of the population. This, in turn, would probably lead them to express significantly lower levels of social trust, but higher levels of political trust, than other groups. Briefly stated, I expected that their comparatively low social trust would reflect their lower level of social connectedness with the dominant population, while their comparatively high political trust would reflect their greater faith in the country?s government agencies to act on their behalf, as protectors of their minority rights and interests. 151 In my analysis of the trust levels across the country?s major ethno-racial groupings, I outlined and tested three related but distinct sets of explanations for why some ethno-racial groups are likely to exhibit higher degrees of social and political trust than others. The first set of explanations centred on what I have referred to in the thesis as key ethno-cultural ?markers,? the observable or identifiable differences among ethnic groups that arise because of the rich diversity in immigration status, religious affiliation, and linguistic background that exists within the Canadian population. I expected that the differences in trust across ethno-racial groups would occur in part because some groups are more visible or otherwise discernible as culturally distinct communities, due to the three ethno-cultural markers. In addition, I suggested a second set of explanations, in which I expected that socioeconomic differences across groups, specifically in regard to education and income levels, would partly explain the differences in trust levels. Finally, I also expected that two social engagement influences --- voluntary association activity and ethnic diversity of friendship networks --- would help in explicating the patterns of differences in social and political trust among the major ethno-racial groupings considered in the thesis. As was already outlined in Chapters 3 and 4, several, but not all, of these expected outcomes did in fact emerge from the data analysis. I will now comment on and interpret the major findings for the ethno-racial groupings, and also discuss the degree to which the results are consistent with the social distance/social boundaries arguments that guided my analysis. I concentrate, in particular, on the findings for the most distinct groups, i.e., visible minorities, French/ ?Canadian? respondents, and Aboriginal Peoples. 152 VISIBLE MINORITIES My initial interest in this thesis project grew out of my concerns about the possibility of an impending crisis in the level of social cohesion and social connectedness experienced by Canada?s ethno-racial minorities, particularly by those people that government agencies include under the rubric ?visible minorities.? I wondered, especially, about how recent claims by Putnam and others, about the negative effects of immigration and ethnic diversity on trust in the United States and Europe, might play out among visible minorities in Canada, notwithstanding our reputation in some circles as perhaps the quintessential multicultural society. As discussed previously, although the label, ?visible minority,? raises a number of conceptual issues, one common denominator shared by all the people subsumed under this term is that these Canadians are more likely than others to be physically observable or identifiable as minority members within the larger society. Moreover, apart from any physical distinctiveness, people within this group are also more likely to differ from Canada?s non-minority population on one or more of three important ethno-cultural markers, which relate to their immigration status, religion, and language. One of my central goals in the thesis was to determine whether and to what extent members of this visible minority grouping would express different levels of social and political trust than other Canadians. 153 Interpreting the Findings for Social Trust Among Visible Minorities As is evident from the results presented in Chapters 3 and 4, my expectation that visible minority Canadians would have relatively low social trust was borne out by the findings. So, too, was the hypothesis that some of the main reasons for this lower social trust could be traced to visible minorities? being disproportionately immigrants (especially recent immigrants), disproportionately Allophone, and disproportionately non-Christian (although the latter effect weakens substantially with the addition of controls for other factors in the models). Overall, more than one-third of the gap in social trust between visible minorities and the British reference group is attributable to the three ethno-cultural factors. By comparison, the socioeconomic and social engagement influences that I had posed as important explanatory factors played a far less noticeable role, although a small portion of the gap in social trust is accounted for by the somewhat lower incomes of visible minorities (especially when contrasted with their high average education levels), by their lower levels of voluntary association activity, and by their high diversity of friendships, which unexpectedly was associated with lower rather than higher social trust. Being younger on average than other Canadians also contributed to relatively lower social trust among the visible minorities in my analysis. In general, the most plausible interpretation of these findings is one that fits with the social distance/social boundaries arguments that have informed my thesis. That is, visible minorities do express lower levels of trust in other people and this seems to occur in part because they are perceived by others, and are treated by others, as culturally different, distant, or separate from the Canadian ?mainstream.? Clearly, this is not the only social 154 process or social mechanism at work, since the gap in social trust between visible minorities and others is still apparent, even with controls for the ethno-cultural factors, as well as for all the other predictors that I included in the analysis. Even so, the simple fact of identifiable cultural distinctiveness seems to have a substantial influence on social trust and, hence, on social cohesion and social connectedness, among Canada?s visible minorities. As I comment on further below, processes and experiences of discrimination and prejudice in the lives of visible minorities seems almost certain to lie behind much of this pattern. Another finding of note is that, because most people in the visible minority category are recent immigrants, and because it is mainly recent immigrants, and not longer-term immigrants, who harbour lower trust in other people, there is reason to conclude that the gap in social trust between visible minorities and other Canadians tends to close with time. This could be seen as an overly optimistic interpretation of the immigration effect, since it is based on cross-sectional findings rather than panel data. Nonetheless, the results do point to the possibility that visible minorities eventually go through a process of acculturation or integration, in which they increasingly accept and are accepted by the established population, and come to adopt much the same level of social trust as their native-born counterparts. As for the effects on social trust of the other explanatory factors included in my models, the analysis confirmed my predictions that variables such as education, income, and voluntary activity are positively associated with social trust, even though these factors do not account for much of the difference in social trust between visible minorities and others. Perhaps the most surprising result in my analysis of social trust is the role played by ethnic diversity of friendships. Visible minorities, as shown in Table 3.5 in Chapter 3, have the highest average friendship diversity of all eight ethno-racial groups in the analysis. I had 155 expected that high friendship diversity would promote higher social trust; instead, however, my results suggest a negative, albeit weak, association with social trust. Therefore, visible minorities may have lower social trust in part because they have friendship networks that are more, and not less, ethnically diverse. My interpretation of this unexpected outcome is a methodological one, having to do with the nature of this measure and the fact that it is really only those respondents with friends that are ?all? in a different ethnic group from their own who have lower social trust. It is possible that this extreme category is capturing individuals who, either in reality or in their own perceptions, are so isolated from their own ethnic communities that they have difficulty developing trust in people generally. There is also a question as to whether individuals whose friends are entirely from a different ethnic group from their own actually have diverse friendships. It could be that such people have friends who are all or predominantly British, for example, and so their friendships do not represent a diverse mix of ethnic backgrounds at all. In any event, I suggest that additional research, using more and better indicators of network diversity, is required before drawing strong conclusions about this outcome, especially given the weakness of the finding. Interpreting the Findings for Political Trust Among Visible Minorities Regarding political trust, I have determined, again as predicted, that visible minorities rank higher than other Canadians, especially when compared to traditionally established groups such as the British and Other Europeans. Here, as well, the most likely interpretation seems to be the one that was raised in the initial chapter of the thesis. That is, visible minorities 156 probably develop this comparatively positive view about Canada?s public institutions for a combination of reasons, many of which stem from the fact that this group is, indeed, a minority, and also a predominantly foreign-born minority, with about 90 percent being born outside Canada in the present analysis.18 One reason for their high political trust could be that immigrant minority members typically came to Canada voluntarily, which suggests that they had a positive predisposition toward the country and its institutions from the outset. In addition, for the many minority individuals who arrived in this country from less prosperous and less democratic parts of the world, Canadian political institutions may seem relatively more just, generous, efficient, and trustworthy, at least in comparison to institutions in their countries of origin. Third, members of the visible minority community could simply be more likely than other Canadians to believe that the government has their rights and interests at heart; this impression could arise from their awareness of such initiatives as the Canadian Charter of Rights and Freedoms, which explicitly asserts that minority rights are to be protected, and also the Multiculturalism Act, which made Canada the first country in the world to establish official legal statutes as safeguards for distinct ethno-racial communities (see, e.g., Merelman 1991; Breton 1998; Fleras and Elliott 2002; Grabb and Curtis 2005). Other considerations that could encourage a positive view of the Canadian government among visible minorities are that, compared to the immigration systems in most other nations, the Canadian system has been among the most 18 For respondents aged 25 or older, the 90 percent figure for the proportion of foreign-born visible minorities in the 2008 GSS sample is very close to the comparable proportion from the 2011 National Household Survey (which now replaces the long-form Census), which is 89 percent (Statistics Canada 2013). 157 accepting of new arrivals (Gwyn 1996; Li 2003), among the quickest to grant citizenship (Bloemraad 2002), and among the most likely to take in refugees, with annual resettlements of about ten percent of the world?s refugees (Citizenship and Immigration Canada 2011)19. This is not to say that the government and all its agencies are above reproach in the eyes of visible minorities. For example, the findings in Chapter 3 showed that some institutions, (e.g., the federal parliament and the welfare system) receive a substantially weaker vote of confidence than others (e.g., the police and the health care system). There is also some reason for concern about the future, and the effects of possible changes in government policy. As noted in Chapter 1, the current Conservative administration is considering immigration rules that may be more restrictive, and that therefore could lower the level of political trust among visible minorities, especially at the federal level. Another important finding in the multivariate analysis is that the tendency for visible minorities to be relatively more trustful of government occurs mainly among recent immigrants, and not among longer-term immigrants, which suggests that minority members may become more disillusioned with or critical of government over time; this is again consistent with the conclusion that minority immigrants eventually acculturate to the Canadian ?norm,? so to speak, with about 19 In analyses not reported in tables, I assessed the possibility that the trust patterns for Canadian-born visible minorities might be different from those for foreign-born visible minorities. However, results showed that, for both types of trust, the same basic patterns occur as those found in the main analysis. For social trust, longer-term visible minority immigrants resemble Canadian-born visible minorities in having higher social trust, with recent visible minority immigrants expressing relatively low social trust. For political trust, visible minority immigrants generally have higher political trust levels than Canadian-born visible minorities, with their political trust levels declining the longer that visible minority immigrants have been in Canada; this suggests that they become increasingly similar to Canadian-born visible minorities over time. These are the same basic patterns as those that were reported for the full sample of respondents. 158 the same level of political trust or distrust as that found among native-born Canadians. As for which set of explanatory factors is more important in accounting for the gap in political trust between visible minorities and the more established groups, my findings have shown that the three ethno-cultural markers of immigration status, religion, and language are once again more influential than are the socioeconomic or social engagement factors that I have considered. Income and voluntary activity appear to have no appreciable relationship to political trust, while education and friendship diversity have quite minor effects. None of the explanatory factors considered in the thesis accounts for much of the variance in political trust in my models, especially in comparison to the results for social trust20. The full multivariate model also reveals that, even with all the controls in place, visible minorities are more trusting of political institutions than any of the other ethno-racial groups. These outcomes indicate that additional explanations should be considered to account completely for the higher political trust exhibited by Canada?s visible minorities. One methodological influence that might be at work is a possible social desirability effect. For example, it could be that visible minorities, especially if they are recent immigrants, are more 20 Two possible reasons for why the models account for more variance in social trust than in political trust are as follows. First, it could simply be that the explanatory factors included in the analysis just have a greater impact on how Canadians think about their fellow citizens than they do on how Canadians think about their government. Second, it could be that Canadians can more readily think of concrete examples of their everyday dealings with other people, resulting in sharper differences in social trust responses across groups of respondents; in contrast, Canadians could have a more distant and vague sense of their dealings with government institutions; for example, it is probable that many respondents pay little attention to Parliamentary activities and most have never been to court. Their less immediate and less frequent experiences with political institutions could lead to weaker distinctions in the political trust levels reported by respondents from different groups, which in turn would lead to less explained variance overall. 159 hesitant than other Canadians to state any negative views they have about government, out of a concern that they might seem ungrateful if they speak too frankly, or even out of a fear of reprisal by officials in some cases. While this pattern might seem unlikely to the average Canadian, it should not be dismissed out of hand. There is also some potential for response set bias, since all of the political trust items are worded using the same sentence construction. A response set might be more likely among visible minorities than the other ethno-racial groups, if we consider that most of the visible minority category includes immigrants, a number of whom are either Allophone or use English or French as their second language. While it seems improbable that such a bias could account completely for the higher political trust among visible minorities in my analysis, it is conceivable that it played some role. Therefore, where possible, future studies would be improved by using batteries of items that are reverse-coded, which would reduce the risk of a response set problem. Before moving on, it is important to stress again another important factor that is implicit in, and almost certainly contributes to, social distance and social boundaries processes. That factor is the underlying question of ethno-racial prejudice or discrimination. In other words, when considering how and why some minority group members develop greater feelings of social distance or separateness from other groups, leading in turn to different levels of social and political trust, it is quite plausible that minority experiences with discrimination and prejudice lie at the root of many of these feelings. In fact, discrimination or prejudice and the processes of social distance and social boundaries would seem clearly to go hand in hand. While the data set I employed lacks direct measures of discrimination and prejudice, this must be considered a likely explanation for why visible minorities (as well as others like the French and Aboriginal Peoples) exhibit different levels of trust when compared with more 160 established groups like the British and many other Europeans. Unfortunately, there are no direct measures of discrimination and prejudice in the data set that I have employed, so that this must remain a likely, yet unproven, additional explanation for why visible minorities (as well as others like the French and Aboriginal Peoples) exhibit different levels of trust when compared with more established groups like the British and many other Europeans. It is both disappointing and surprising that Canadian research on prejudice and discrimination is rather sparse, especially given the supposed importance attached to understanding ethnic pluralism and multiculturalism in this country. The existing research indicates that, while Canadians are relatively more supportive of multiculturalism policies and ethnic diversity than they once were, and also appear to be more accepting of ethno-racial minorities than are most people living elsewhere in the world, Canadian society continues to face real problems of prejudice and discriminatory practices that are unresolved (see, e.g., Reitz and Breton 1994; Grabb and Curtis 2005; Reitz and Banerjee 2009; Andersen and Milligan 2009). Hence, if and when the appropriate data become available, a crucial issue to consider in future studies of social cohesion in Canada is the important role that patterns of prejudice and discrimination probably play in accounting for differences in social and political trust across the nation?s ethnic groups. Another fruitful avenue for future research would be to examine the internal variations within the visible minority category. I emphasize, nevertheless, that my thesis makes an important contribution in establishing our understanding of the overall pattern for visible minorities, while also comparing them to two other distinctive minorities, the French and Aboriginal Peoples. 161 THE FRENCH AND THE ?CANADIANS? Taken together, the French and the highly similar category of ?Canadian? (or ?Canadien?) respondents represent a second culturally distinctive grouping, one that is different in important ways from ?establishment? groups like the British. Of course, in one key respect, it could be argued that the French/Canadian categories are also establishment groups, since many of their number come from long-standing historical communities. Along with the British, the French population is seen by most observers as one of the two ?charter groups? of Canadian Confederation. As well, the French language, in conjunction with the English language, is integral to Canada?s founding image as a bilingual and bicultural society, going back to 1867 and the British North America Act. In fact, French colonial influence in Canada actually pre-dates the British influence. For this reason, some French/Canadian respondents, especially those descending from the so-called ?pure laine? or ?ancienne souche? Quebec communities that existed before the British conquest in the mid-1700s, could probably claim a longer lineage and deeper roots than the British themselves (see, e.g., Careless 1971; Harris and Warkentin 1974). Nevertheless, at least until recent decades, and especially outside Quebec, the French have not formed an ?established? community like the British, in the sense of being economically, politically, and socially dominant within Canada. Therefore, in the present analysis, it seems appropriate to treat the French/Canadian categories as a distinct minority grouping that, in some respects, shares an affinity with the visible minorities. Some evidence of this affinity is shown in the thesis findings, which reveal that the French/Canadian respondents are much like the visible minority category in having substantially lower social trust, but demonstrably 162 higher political trust, than do the British. Below I discuss both of these outcomes. Interpreting the Findings for Social and Political Trust Among the French and Canadians As an initial comment, I should note that there is at least one obvious difference between the French/Canadians and the visible minorities in the present analysis. The difference is that the former are almost 100 percent native-born, while the latter are about 90 percent foreign-born. Therefore, it is not surprising that, in contrast to the results for visible minorities, immigration status has no impact in accounting for the gap in trust levels between the French/Canadians and the British. At the same time, however, I have found that the other two ethno-cultural markers --- religion and language --- are, as they were for visible minorities, very important for understanding the differences in both social trust and political trust between the French/Canadian respondents and other groups. In fact, the influences of religion and language are particularly striking in the case of the French/Canadian groupings, with my findings showing that it is almost entirely because the vast majority of these respondents are Roman Catholic and Francophone that they express lower social trust than the British, and being Francophone is the key reason why they express higher political trust, than the British. In the case of the French/Canadian respondents, the results for both measures of trust indicate that the socioeconomic and social engagement explanations play a minimal role at best. The findings for religion and language raise an important question: what is it about the Roman Catholic religion and the French language that leads to these outcomes? First, with respect to religion, I have previously discussed work by writers such as Uslaner (2004) 163 suggesting that, when compared to Protestantism in particular, Roman Catholicism may foster in its adherents a more authoritarian and collectivist world view that is conducive to a greater distrust of other people, combined with more trust and acceptance of political and other forms of authority. This religion effect may seem less plausible for the French/Canadian respondents, since French Canadians in the present day are among the least likely in the Canadian population to attend religious services, and so could be only nominally Roman Catholic (Bibby 2002; Clark 2003). However, as other writers have noted (e.g., Gagnon 2003; Grabb and Curtis 2005), most French Canadians do retain a strong sense of their Roman Catholic heritage, and still wish to preserve it as one central component of their ethnic or cultural identity, notwithstanding their low levels of religious practice. Therefore, it is possible that the French and Canadian respondents in my analysis express comparatively low social trust and high political trust in part because of the same Roman Catholic influences noted by Uslaner. A somewhat parallel argument can be made for the effects of language on the trust levels of the French/Canadian respondents. The use of the French language is clearly another unifying component of French Canadian identity, and symbolizes another fundamental difference between the French/Canadian grouping and the traditionally dominant British, the large majority of whom are Anglophones. As Grabb and Curtis (2005: 38) have noted, Tocqueville was one of the first to emphasize the ?tie of language? as ?perhaps the strongest and most durable? force for uniting an ethnic or cultural group (Tocqueville 1835: 29). At the same time, however, language can also be a powerful force for separating one group from another, as Canada?s history of French-English relations has often shown (Posgate and McRoberts 1976; Young 1995). 164 Therefore, being immersed in a minority culture, which for centuries has literally been written and spoken using a language that is distinct from that of the British majority, may have helped to instill a quite different perspective in French Canadians about a range of issues and concerns, including their orientations to social and political trust. Especially when combined with the Roman Catholic influence discussed previously, the shared influence of a common language may have contributed to processes of social distance and social boundaries formation between the French/Canadian respondents and other groups, which in turn could have contributed to a distinct French/Canadian viewpoint about trust in other people and trust in political authority. In the case of political trust, another impetus could be the sense among many Francophones that government agencies and institutions (e.g., the education system and the justice system) have been constituted with built-in safeguards for their language rights, and so can be trusted to play their part in preserving the French language in Canadian society. As mentioned in Chapter 3, one other point to consider when interpreting the French/Canadian levels of political trust is that several of the political institutions included in the political trust scale are provincial jurisdictions, which might mean that the higher political trust among French/Canadian respondents, most of whom are Quebecers, pertains more to their views about Quebec?s political institutions than to their trust in the Canadian government. Nevertheless, it does appear from the findings in Chapter 3 that French/Canadians are also more trusting of federal institutions (e.g., the federal parliament) than are the British and Other Europeans. This is a pattern that is different from that shown in early studies (recall Roseborough and Breton 1968; Simeon and Elkins 1974), and may signal that French/Canadians now see both levels of government as comparatively effective in 165 protecting their rights and interests. This leads to one other salient factor to keep in mind when assessing French/Canadian patterns of social and political trust in my analysis. That factor is region, and the large concentration of French Canadians in Quebec. The existence of a long-standing and separate regional society in Quebec is well-known and widely acknowledged by those who are familiar with Canada. Even Prime Minister Harper, who in the past resisted the concept of a distinct Quebec society, now appears to take this view, for in 2006 he became the first federal leader to grant Quebec official status as its own ?nation? within the larger nation of Canada. In addition to this acknowledgement of Quebec?s unique role in the country, it is important to be aware that living in Quebec, versus Canada?s other regions, has a notable influence on the social and political trust expressed by the respondents in my analysis. The findings in Chapter 4 show that, even with controls for all other factors, there is a significant region effect: Quebecers express the lowest social trust of all respondents, and Quebecers are also among those with the highest political trust (although they are matched on political trust by Ontarians and Atlantic residents in the full multivariate analysis). Because of Quebec?s special position within the country, and given that the large majority of French/Canadian respondents also reside in Quebec, I explored, in preliminary analyses, the possibility of an interaction effect involving region and ethno-racial background. That is, I looked at whether the lower social trust and higher political trust exhibited by French/Canadians might be especially evident for French/Canadian respondents living in Quebec, and less evident for the other ethno-racial groupings living in Quebec. I found no such pattern, however, and instead determined that any ?Quebec effect? on social and 166 political trust also applies in more or less the same way to the other ethno-racial groupings in my study, and not just the French/Canadian categories (for the basic patterns, see Table A1 in the Appendix). In a comparison of ethno-racial variations in social and political trust across five major areas that included the Atlantic region, Ontario, the Prairies, British Columbia, and Quebec, Quebec stands out as the one region of Canada where people from all eight ethno-racial groups express their lowest average levels of social trust. Quebec is also where people from all but two of the eight ethno-racial groups express their highest average levels of political trust. The exceptions are Aboriginal Peoples and the French themselves, both of which have their second highest political trust scores in Quebec, but their highest in the Atlantic region. Overall, then, there appears to be something about the regional culture and institutions of Quebec that leads its residents to be somewhat less trusting of people and somewhat more trusting of government than other Canadians, and this pattern holds almost regardless of ethno-racial background. For political trust, one contributing factor could be that, for some time now, Quebec has typically been the most ?statist? region in Canada, as measured by such indicators as the level of spending on government services; this could have promoted a more positive or accepting view about political institutions among Quebecers (see, e.g., Grabb and Curtis 2005). Grabb and Curtis (2005: 172-174) review research indicating that, for many decades now, Quebec has generally been the most ?statist? or ?social-democratic? region of the country, with higher government spending per capita, higher government debt per capita, and higher overall taxation levels than virtually all other Canadian provinces (see also Bertrand 2013; Finlayson 2012; Palacios, Veldhuis, and Walker 2008). Another speculation is that the social distance and social boundaries mechanisms at 167 work in Canada are even more pronounced in Quebec, where the degree of ethno-racial diversity is the lowest of all the major regions of the country. For example, in the GSS sample used for my analysis (these findings are not shown in tables), the Quebec sub-sample has the lowest representation of Aboriginal Peoples among all the five major regions, with less than 2 percent of Quebecers indicating membership in this group. Quebec also has the lowest representation of British respondents (6 percent) and the second lowest representation of visible minorities (6 percent), with only the Atlantic region (at 2 percent) having a lower proportion of visible minorities. Therefore, Quebec?s status as the most ethnically and culturally homogeneous region of Canada could be influencing the findings, generating lower trust in other people, but higher trust or acceptance of political institutions, than occurs elsewhere. While further analysis of this pattern is beyond the scope of my thesis objectives, it is an intriguing outcome that should be explored in future research. ABORIGINAL PEOPLES Along with visible minorities and the French/Canadian respondents, Aboriginal Peoples represent one of the three most culturally distinct ethno-racial groups in my analysis. Aboriginal Peoples have some obvious similarities, but also some obvious differences, when they are compared to the other two groups. For example, like visible minorities but unlike French/Canadians, Aboriginal Peoples tend to be more easily identifiable by their physical attributes as a distinct ethno-racial grouping. In addition, like French/Canadians but unlike visible minorities, Aboriginal Peoples are overwhelmingly Canadian-born or non-immigrant. Perhaps this unique mix of similarities and differences relative to visible minorities and 168 French/Canadians is part of the reason why my findings for Aboriginal Peoples are unlike those for other groups. I have found that Aboriginal Peoples are the one ethno-racial group whose levels of social trust and political trust are both substantially lower than those of the average Canadian. In the next section, I offer some interpretations and speculations about this exceptional outcome for Aboriginal Peoples. Interpreting the Findings for Social and Political Trust Among Aboriginal Peoples While each of the ethno-racial groupings that I have compared in the thesis can be thought of as unique in its own way, Aboriginal Peoples clearly stand apart from the others in a number of critical respects. First of all, Aboriginal Peoples are the only group that can make the claim to being the descendents, more or less, of the original inhabitants of the country. In addition, however, and despite their primacy of place as the original population, research has shown consistently over time that Aboriginal Peoples have tended to gain the fewest benefits and opportunities from living in this country. Aboriginal Peoples are among the groups with the lowest income and education levels, as well as the highest poverty, mortality, and morbidity rates (e.g., Frideres and Gadacz 2011; Satzewich 2011:59-68; White, Beavon, and Spence 2007; Menzies 2009). In general, they have the worst life chances of all the ethno-racial communities in Canada. The reasons for the often deplorable living conditions for Aboriginal Peoples in Canada are many, but the most obvious and most telling can be linked to the history of colonialism and betrayal of Aboriginal rights and interests by a series of French, British, and later Canadian governments. Other important factors include the racist actions and attitudes of 169 political decision-makers and many members of the general population over the years. Moreover, almost from the first encounters with European settlers, there has been a fundamental clash of values, cultures, and world views that has made co-existence and acculturation extremely difficult for Aboriginal Peoples (Fleras and Elliott 1992; Frideres and Gadacz 2011; Maaka and Fleras 2005; Menzies 2009; Satzewich 2011). These difficulties have been made worse because, unlike French Canadians, for example, Aboriginal Peoples comprise a diverse set of tribes and communities that do not always share the same traditional languages and religions, and that are further divided by their residence in what are often sparsely populated, scattered, and remote areas. In addition to these long-standing problems, of course, are the numerous examples of injustice and discrimination faced by Aboriginal Peoples in more recent times, including the scandalous treatment of Aboriginal Peoples in the residential school system (Haig-Brown 1988; Furniss 2002) and the prolonged and ongoing resistance by government to the recognition of Aboriginal Peoples? treaty rights and the settlement of land claims (Wilkes 2006; Corrigall-Brown and Wilkes 2012; Frideres and Gadacz 2011). Given this background of historical experiences and contemporary circumstances, it is not surprising that, when compared to the other ethno-racial groupings in my analysis, Aboriginal Peoples are among the least trusting toward other people and the least trusting toward government agencies and institutions. In fact, the surprise may be that, in spite of all this, Aboriginal Peoples do not express even more distrust or disillusionment than they do. Hence, while Aboriginal Peoples are less trusting of other people than are most groups in my analysis, they are not lower in social trust than are visible minorities, both before and after controls. Aboriginal Peoples do rank the lowest of all groups on political trust, both before 170 and after controls; however, as discussed in Chapter 3, their absolute levels of political trust are not uniformly low, with more than two-thirds of Aboriginal Peoples reporting that they have a great deal or quite a lot of confidence in the police, the justice system and courts, and the school system. Nevertheless, it is undeniable that, relative to other Canadians, Aboriginal Peoples have particularly low levels of both social and political trust. Another outcome that is exceptional in the case of Aboriginal Peoples is that, in contrast to the findings for visible minorities and French/Canadians, the three sets of explanations that I have posed generally do not work very well in accounting for the patterns of trust among Aboriginal Peoples. First of all, the two social engagement factors have almost no impact. Second, and more important perhaps, the three ethno-cultural markers play very little role in explaining the differences in either social trust or political trust between Aboriginal Peoples and the British reference group. I still would suggest that social distance/social boundary processes are at work, but clearly these processes seem not to be tied to such factors as immigration, religion, and language. The basis for the trust ?divide? between Aboriginal Peoples and other Canadians appears instead to stem from the long-standing problems noted earlier, including colonialism, paternalism, racism, and a widespread indifference to Aboriginal Peoples? needs and aspirations. These factors have been so deeply ingrained in Canada?s social fabric that their effects on social and political trust probably cannot easily be addressed or reduced, at least in the near term. This brings me to the third set of explanations, involving the two socioeconomic factors: education and income. I have found that socioeconomic influences have little or no effect in accounting for the lower political trust among Aboriginal Peoples. However, for social trust, 171 the models shown in Chapter 4 reveal that education and income differences do appear to explain some, though certainly not all, of the gap between Aboriginal Peoples and other Canadians. To understand this effect further, in analyses not reported in the tables, I looked at the relationship between the two socioeconomic factors and social trust within the Aboriginal Peoples? category, and found positive correlations for both years of education (r = +.253, p = .000) and income (r = +.305, p = .000). Using the categorical measures of education and income, I also found that, at least for the small number of Aboriginal Peoples in the two highest education categories (N=65) and the highest income grouping (N=77), their social trust scores were well-above the average for all respondents. There is little basis for optimism in these outcomes, given the small number of Aboriginal Peoples who are able to achieve high socioeconomic status in this country. Nevertheless, these results at least hint at the possibility that, if there were fewer structural and cultural barriers limiting the opportunities for Aboriginal Peoples to attain socioeconomic success, there might be reason to hope that Aboriginal Peoples? feelings of social cohesion and connectedness would be substantially improved. Enhancing Aboriginal Peoples? access to higher education is one mechanism that would probably help the situation. Also, more could be done to encourage greater Aboriginal Peoples? economic independence, for example, by establishing their own successful business enterprises, as has occurred for some Aboriginal Peoples in British Columbia and elsewhere. (For some discussion and analysis, see Canadian Council for Aboriginal Business 2011.) Then Aboriginal Peoples would have greater reason to believe that they have a real stake in the nation, and another basis for developing trust in their fellow citizens. A related factor to consider when assessing the trust levels of Aboriginal Peoples is age. 172 The findings for the full models in Chapter 4 indicate that one of the main reasons why Aboriginal Peoples rank low on both social trust and political trust is that, on average, they are one of the youngest ethno-racial groups, and younger people are less trusting than are older people in my analysis. When age is controlled in these models, the social trust gap for Aboriginal Peoples drops to about half of the gap in Model 1, and even the political trust gap diminishes somewhat. A number of studies have shown that young Aboriginal Peoples are particularly disillusioned with the political process in Canada and are less likely to participate politically, because they distrust the government and its agencies (Taiaiake, Pitawanakwat, and Price 2007; Kinnear 2003; Elections Canada 2004). Clearly, it is necessary to find the means to improve both the prospects and the realities of young Aboriginal Peoples, if we are to make any progress toward giving this key group a reason to feel that they are integral to the nation as a whole and have their own stake in making the country a more unified, coherent, and beneficent place for all Canadians to live. THE BRITISH, OTHER EUROPEANS, AND THE ?MIXED-ORIGINS? CATEGORY Before concluding, I should comment briefly on the results for the other groups that I have discussed in the thesis, which include the British, Other Europeans, and those in the ?Mixed-Origins? category. The bivariate findings show that the British tend to have the highest social trust among all the groups that I have considered, with the Other Europeans and Mixed-Origins groupings being very similar to the British on this measure. The bivariate findings 173 show, as well, that these same three groups generally rank as the least trusting when it comes to the society?s political institutions. The full multivariate models that I have tested offer some insight into these bivariate findings, for they suggest that the gaps in social and political trust between these three categories, on the one hand, and the more culturally distinctive minority groups, on the other hand, occur in part because the British, Other European, and Mixed-Origins respondents tend not to share the distinctive ethno-cultural markers of the minority categories; in particular, people from British, Other European, and Mixed-Origins backgrounds are disproportionately Canadian-born and Anglophone, and for these reasons may be less subject to the social distance/social boundaries processes that shape the social and political trust levels of the other groups. There is also some evidence in the full models that the socioeconomic and social engagement resources of the British, Other European, and Mixed-Origins respondents, especially their comparatively high incomes and high levels of voluntary association membership (see Table 3.5 in Chapter 3), might play some role, at least in explaining their higher levels of social trust compared to other groups. Apart from these outcomes, however, my findings do not offer a complete understanding of the reasons why these three groups rank as they do when it comes to social and political trust. A possible explanation is that groups which have traditionally predominated in the society, like the British and many Other Europeans, are better positioned to develop trust in other people, since they have social, economic, and political advantages or resources that generally make their social interactions and experiences more successful, predictable, and enjoyable than occurs for less established and less privileged ethno-racial groups. At the same time, the relatively lower trust in political institutions of the British and Other 174 Europeans could occur because, compared to members of ethno-racial minority communities, they have greater knowledge of what they see as deficiencies in the nation?s public agencies, higher expectations about what these agencies should provide to citizens, or perhaps even some resentment that they do not receive the same ?special? status or treatment that they may believe the government offers to minority groups in Canada. Finally, there is the Mixed-Origins category. It will be recalled that this grouping includes all of those in the survey who gave more than one response to the question about their ethnic background. The initial reason for including this category in my analysis was to allow for a speculative test of whether respondents who emphasized multiple ethnic backgrounds might have high levels of trust, based on the assumption that their plural identities were a good sign that they themselves had ?bridged? between ethnic groups and therefore would be more trusting of others. As already noted, my findings are consistent with this speculation in regard to social trust, with the results indicating that Mixed-Origins respondents rank with the British and Other Europeans as the most trusting of all the groups, both before and after controls. On the other hand, Mixed-Origins respondents rank among the least trusting when it comes to their attitudes about Canada?s political institutions, and in this way are again strikingly similar to the British and Other European categories. This similarity with the more established groups on both trust measures could suggest that the Mixed-Origins category includes individuals who are simply more acculturated to the dominant ideology, or more attuned to prevailing views about trust. It must be admitted, however, that these interpretations are highly speculative, because it is not possible to know with any precision the composition of this particular group in the analysis. Therefore, a satisfactory test of my 175 hypothesis about the pluralist nature of ?mixed? Canadians and how this might affect such people?s levels of trust must wait for additional research involving better measures. CONCLUDING REMARKS This concludes my analysis of ethno-racial variations in social and political trust in Canada. In asking whether in Canada we trust, I have found that the answer is ?yes,? but that the amount of trust can differ substantially, depending on which ethno-racial group is asked the question, and which type of trust is being considered. The analysis shows that a theoretical perspective drawing on the concepts of social distance and social boundaries is useful for understanding many, though not all, of the differences in social and political trust that exist across Canada?s major ethno-racial communities. My results show that, in contrast to more established groups like the British and Other Europeans, those groups that are more culturally distinctive, including visible minorities, French Canadians, and Aboriginal Peoples, have comparatively less trust in other people. This outcome is consistent with the interpretation that ethno-racial communities who experience more social distance from established groups, or are confronted with relatively greater social boundaries, are less able to develop social trust. In addition, these same groups, with the exception of Aboriginal Peoples, exhibit comparatively high trust in political institutions. The latter finding is consistent with the interpretation that culturally distinct minorities, in response to the impact of social distance/social boundaries processes, are more likely than other groups to see the government and its agencies as defenders of their minority rights and interests within Canada. That Aboriginal Peoples are an exception to this pattern, and express lower levels of 176 both social trust and political trust, could signify that this group actually has the same response to social distance/social boundaries influences as do other minorities in regard to social trust. However, because of a long history of problems with Canada?s political institutions and agencies, Aboriginal Peoples have far less reason than either visible minorities or French Canadians to see the government as its protector and defender. Unfortunately, without important changes in the ways that Aboriginal Peoples? needs and aspirations are addressed in this country, it seems unlikely that this result will change any time soon. The findings of my analysis lead to conclusions that are generally consistent with those of the eminent scholar, Raymond Breton, who along with his colleagues (2004: 189) has asserted that one of the crucial ways that Canadians can protect and strengthen our ?fragile social fabric? is by ?bridging the social boundaries? that separate ethno-racial communities from each other. Likewise, one of the best things that we can do is to find more ways to reduce the social distance between us, especially between our distinct ethno-racial minorities, on the one hand, and our more established ethno-racial groupings, on the other hand. It seems crucial, in particular, to do whatever is required to reduce existing problems of discrimination and prejudice against minority Canadians. Through the concerted efforts and sustained initiatives of the Canadian people and Canadian institutions, we may yet achieve a fuller appreciation of the diversity that enriches our society, and a better understanding of the many similarities that serve to bind us all together. 177 REFERENCES Abada, Teresa, Feng Hou, and Bali Ram. 2009. \"Ethnic differences in educational attainment among the children of Canadian immigrants.? Canadian Journal of Sociology 34(1): 1-28. Abrahamson, Paul. 1983. Political Attitudes in America. San Francisco: Freeman and Co. Adams, Michael. 2003. Fire and Ice: The Myth of Value Convergence in Canada and the United States. Toronto: Penguin. Aizlewood, Amanda and Ravi Pendakur. 2005. Ethnicity and social capital in Canada. Canadian Ethnic Studies 37(2): 77?102. Alesina, Alberto and Eliana La Ferrara. 2000. ?Participation in heterogeneous communities.? The Quarterly Journal of Economics 115(3): 847-904. Alesina, Alberto and Eliana La Ferrara. 2002. ?Who trusts others?? Journal of Public Economics 85, 2: 207-234. Alexander, Jeffrey C. 1992. ?Citizens and enemy as symbolic classification: on the polarizing discourse of civil society.? In M. Lamont and M. Fournier eds., Cultivating Differences: Symbolic Boundaries and the Making of Inequality, pp. 289-308. Chicago: University of Chicago Press. Andersen, Robert and Scott Milligan. 2009. ?Inequality and intolerance: Canada in cross-national perspective.? Pp.409-428 in Social Inequality in Canada: Patterns, Problems, and Policies, edited by Edward Grabb and Neil Guppy. 5th ed, Toronto: Pearson Prentice Hall. Andersen, Robert, James Curtis, and Edward Grabb. 2006. ?Trends in civic association activity in four democracies: The special case of women in the United States.? American Sociological Review 71: 376-400. Anderson, Christopher J. and Aida Paskeviciute. 2006. ?How ethnic and linguistic heterogeneity influence the prospects for civil society: A comparative study of citizenship behavior.? The Journal of Politics 68(4): 783-802 Baer, Douglas. 2008. ?Community context and civic participation in immigrant communities: a multi-level study of 137 Canadian communities.? Metropolis British Columbia Working Paper Series No.08-03, February. Baer, Douglas, Edward Grabb, and William Johnston. 1993. ?National character, regional culture, and the values of Canadians and Americans.? Canadian Review of Sociology and Anthropology 30, 1 (February): 13-36. Bekkers, Ren?, Beate V?lker, Martin van der Gaag, and Henk Flap. 2008. ?Social networks of participants in voluntary associations.? Pp. 185-205 in Social Capital: An International Research Program, edited by Nan Lin and Bonnie Erickson. New York: Oxford University Press. 178 Bibby, Reginald. 1993. Unknown Gods. Toronto: Stoddart. Bibby, Reginald. 2002. Restless Gods. Toronto: Stoddart. Bleich, Erik. 2009: ?Where do Muslims stand on ethno-racial hierarchies in Britain and France? Public opinion surveys, 1988-2008.? Patterns of Prejudice 43, 3-4, July/September 2009: 379-400. Bloemraad, Irene. 2002. ?The North American naturalization gap: an institutional approach to citizenship acquisition in the United States and Canada.? International Migration Review 36, 1 (Spring): 193-228. Bogardus, Emory S. 1925. ?Social distance and its origins.? Journal of Applied Sociology 9: 216-226 Bogardus, Emory S. 1928. Immigration and Race Attitudes. Boston: D.C. Heath. Bogardus, Emory S. 1947. ?Measurement of personal-group relations.?? Sociometry 10, 4: 306-311. Bogardus, Emory S. 1967. A Forty Year Racial Distance Study. Los Angeles: University of Southern California Press. Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press. Bourdieu, Pierre. 1985. ?Forms of capital.? In Handbook of Theory and Research for the Sociology of Education, edited by John G. Richardson. Westport, Conn.: Greenwood Press. Boyd, Monica. 2002. ?Educational attainments of immigrant offspring: success or segmented assimilation? International Migration Review 36 (Winter): 1037-1060. Boyd, Monica. 2008. ?A socioeconomic scale for Canada: measuring occupational status from the Census.? Canadian Review of Sociology 45, 1 (February): 51-91. Boyd, Monica and Michael Vickers. 2009. ?The ebb and flow of immigration in Canada.? Pp. 237-252 in Social Inequality in Canada: Patterns, Problems, and Policies, edited by Edward Grabb and Neil Guppy. 5th ed, Toronto: Pearson Prentice Hall. Brady, Henry, Sidney Verba, and Kay Lehman Schlozman.1995. ?Beyond SES: a resource model of political participation.? American Political Science Review 89: 271-94. Brady, Henry, Schlozman, Kay Lehman and Verba, Sidney. 1999. ?Prospecting for participants: Rational expectations and the recruitment of political activists.? American Political Science Review 93: 153-69. Breton, Raymond. 1964. ?Institutional completeness of ethnic communities and the personal relations of immigrants.? American Journal of Sociology 70(2): 193-205 179 Breton, Raymond. 1990. ?The ethnic group as a political resource in relation to problems of incorporation: perceptions and attitudes.? Pp. 196-255 in Breton, Raymond, Wsevolod Isajiw, Warren Kalbach, and Jeffry Reitz. 1990. Ethnic Identity and Equality: Varieties of Experience in a Canadian City. Toronto: University of Toronto Press. Breton, Raymond. 1998. ?Ethnicity and race in social organization: recent developments in Canadian society.? Pp. 60-115 in Rick Helmes-Hayes and james Curtis, eds., The Vertical Mosaic Revisited. Toronto: University of Toronto Press. Breton, Raymond. 2012. Different Gods: Integrating Non-Christian Minorities into a Primarily Christian Society. Montreal and Kingston: McGill-Queen?s University Press. Breton, Raymond, Jeffrey Reitz, and Victor Valentine (eds.). 1980. Cultural Boundaries and the Cohesion of Canada. Montreal: Institute for Research on Public Policy. Breton, Raymond and Daiva Stasiulis. 1980. ?Linguistic boundaries and the cohesion of Canada.? Pp. 137-309 in R. Breton, J. Reitz, and V. Valentine (eds.), Cultural Boundaries and the Cohesion of Canada. Montreal: Institute for Research on Public Policy. Breton, Raymond, Wsevolod Isajiw, Warren Kalbach, and Jeffry Reitz. 1990. Ethnic Identity and Equality: Varieties of Experience in a Canadian City. Toronto: University of Toronto Press. Breton, Raymond, Norbert Hartmann, Jos Lennards and Paul Reed. 2004. A Fragile Social Fabric? Fairness, Trust, and Commitment in Canada. Montreal and Kingston: McGill-Queen?s University Press. Brooks, Clem, and Simon Cheng. 2001. ?Declining government confidence and policy preferences in the U.S.: devolution, regime effects, or symbolic change?? Social Forces 79: 1343-75. Caldwell, Gary and Paul Reed. 2000. ?Civic participation in Canada: is Quebec different?? Inroads 8: 215-222. Canadian Council for Aboriginal Business. 2011. Promise and Prosperity: The Aboriginal Business Survey. Ottawa: Indian and Northern Affairs Canada, April. Cantril, Albert, and Susan Davis Cantril. 1999. Reading Mixed Signals: Ambivalence in American Public Opinion about Government. Washington, DC: Woodrow Wilson Center Press. Careless, J.M.S. (ed.) 1971. Colonists and Canadiens. Toronto: Macmillan Carr, Patrick, Laura Napolitano, and Jessica Keating. 2007. ?We never call the cops and here is why: a qualitative examination of legal cynicism in three Philadelphia neighborhoods.? Criminology 45: 445-80. 180 Catterberg, Gabriela, and Alejandro Moreno. 2005. ?The individual bases of trust: trends in new and established democracies.? International Journal of Public Opinion Research 18: 31-48. Chai, Sun-Ki. 2005. ?Predicting ethnic boundaries.? European Sociological Review 24, 1 (September): 375-391. Chui, Tina, James Curtis, and Ronald Lambert. 1991. ?Immigrant background and political participaton: examining generational patterns.? Canadian Journal of Sociology 16, 4 (Fall): 375-396. Citizenship and Immigration Canada. 2011. The refugee system in Canada. http://www.cic.gc.ca/english/refugees/canada.asp . Last modified June 28, 2011. Clark, Warren. 2003. ?Pockets of belief: religious attendance patterns in Canada.? Canadian Social Trends 68 (Spring): 2-5. Coenders, Marcel, Marcel Lubbers, and Peer Scheepers. 2005. ?Majority populations' attitudes towards migrants and minorities.? Report for the European Monitoring Centre on Racism and Xenophobia. Report 4: Majorities' attitudes towards minorities in European societies: results from the European Social Survey 2002-2003, Commissioned by European Monitoring Centre on Racism and Xenophobia (EUMC), Vienna. Cole, Richard L. 1973. ?Toward a model of political trust: a causal analysis.? American Journal of Political Science 17: 809-17. Coleman, James. 1988. ?Social capital in the creation of human capital.? American Journal of Sociology 94: S95-S120. Corntassel, Jeff. 2003. ?Who is Indigenous? ?Peoplehood? and ethnonationalist approaches to rearticulating Indigenous identity.? Nationalism & Ethnic Politics 9: 75?100 Corrigall-Brown, Catherine and Fred Wien. 2009. ?Regional inequality: causes and consequences.? Pp. 324-344 in Social Inequality in Canada: Patterns, Problems, and Policies, edited by Edward Grabb and Neil Guppy. 5th ed, Toronto: Pearson Prentice Hall. Corrigall-Brown, Catherine and Rima Wilkes. 2012. ?Picturing protest: The visual framing of collective action by First Nations in Canada.? American Behavioral Scientist 56: 223-243. Costa, Dora and Matthew Kahn. 2003. ?Civic engagement and community heterogeneity: an economist?s perspective.? Perspectives on Politics 1, 1: 103-111. Curtis, James and Edward Grabb. 2002. ?involvement in the organizational base of new social movements in English Canada and French Canada.? Pp. 164-181 in Douglas Baer, ed., Political Sociology: Canadian Perspectives. Toronto: Oxford University Press. 181 Curtis, James, Ronald Lambert, Steve Brown, and Barry Kay. 1989. ? Affiliating with voluntary associations: Canadian-American comparisons.? Canadian Journal of Sociology 14(2): 143-162. Dalton, Russell. 2004. Democratic Challenges, Democratic Choices: The Erosion of Political Support in Advanced Industrial Democracies. Oxford: Oxford University Press. Dalton, Russell. 2005. ?The social transformation of trust in government.? International Review of Sociology 15: 133-54. Dasko, Donna. 2005. ?Public attitudes towards bilingualism and multiculturalism in Canada.? Pp. 119-125 in Margaret Adsett, ed., Canadian and French Perspectives on Diversity: Conference Proceedings. Ottawa: Department of Canadian Heritage. Davies, Scott and Neil Guppy. 2010. The Schooled Society: An Introduction to the Sociology of Education 2nd edition, Toronto: Oxford University Press. Delhey, Jan and Kenneth Newton. 2005. ?Predicting cross-national levels of social trust: global pattern or nordic exceptionalism?? European Sociological Review 21, 4: 311-327. Dekker, Paul and Eric M. Uslaner. 2001. eds. Social Capital and Participation in Everyday Life. London: Routledge. Dershowitz, Alan M. 1997. The Vanishing Jew: In Search of Jewish Identity for the Next Century. New York: Simon and Schuster. Durkheim, Emile and Marcel Mauss. 1903 [1963]. Primitive Civilization. Chicago: University of Chicago Press. Elections Canada. 2004. ?Aboriginal and youth participation in federal elections.? Electoral Insight. http://www.elections.ca/vot/abo/EC91007_e.pdf Emig, Arthur, Michael Hesse, and Samuel Fisher III. 1996. ?Black-white differences in political efficacy, trust, and sociopolitical participation: a critique of the empowerment hypothesis.? Urban Affairs Review 32, 2 (November): 264-276. Fieldhouse, Edward and David Cutts. 2010. ?Does diversity damage social capital? A comparative study of neighbourhood diversity and social capital in the US and Britain.? Canadian Journal of Political Science 43, 2 (June): 289-318. Finlayson, Jock. 2012. ?Numbers show that government in Canada is growing too fast.? The Vancouver Province, May 28. Flap, Henk. 1999. ?Creation and returns of social capital: a new research program.? La Revue Tocqueville 20: 5-26. 182 Fleras, Augie and Jean Leonard Elliott. 1992. The Nations Within: Aboriginal-State Relations in Canada, the United States, and New Zealand. Toronto: Oxford University Press. Fleras, Augie and Jean Leonard Elliott. 2002. Engaging Diversity: Multiculturalism in Canada, 2nd edition. Toronto: Nelson Thomson Learning. Foner, Nancy. 2008. ?The Dutch dilemma: Religion, integration, and the crisis of tolerance. Sociological Forum 23(2): 407-413. Fong, Eric and Emi Ooka. 2006. ?Patterns of participation in informal social activities among Chinese immigrants in Toronto.? International Migration Review 40, 2 (June): 348-374. Frideres, James and Rene Gadacz. 2011. Aboriginal Peoples in Canada: Contemporary Conflicts (9th edition). Toronto: Pearson Prentice Hall. Furniss, Elizabeth. 2002. Victims of Benevolence. Vancouver: Arsenal Pulp Press. Gagnon, Lysiane. 2003. ?When rituals soothe.? The Globe and Mail, 9 June, p.A11. Gans, Herbert. 1992. ?Second generation decline: scenarios for the economic and ethnic futures of the post-1965 American immigrants.? Ethnic and Racial Studies 15 (April): 173-191. Gans, Herbert. 1997. ?Toward a reconsideration of \"assimilation \" and pluralism\": The interplay of acculturation and ethnic retention.? International Migration Review 31 (Winter):875-892. Gibbins, Roger. 1982. Regionalism. Territorial Politics in Canada and the United States. Toronto: Butterworths. Giddens, Anthony. 1990. The Consequences of Modernity. Stanford: Stanford University Press. Glanville, Jennifer and Pamela Paxton. 2007. ?How do we learn to trust? A confirmatory tetrad analysis of the sources of generalized trust.? Social Psychology Quarterly 70, 3: 230-242. Gordon, Milton M. 1964. Assimilation in American Life: The Role of Race, Religion and National Origins. New York: Oxford University Press. Grabb, Edward. 1979. ?Relative centrality and political isolation: Canadian dimensions.? Canadian Review of Sociology and Anthropology 16:343-355. Grabb, Edward. 1980. \"Social class, authoritarianism, and racial contact: recent trends.\" Sociology and Social Research 64(2): 208-220. 183 Grabb, Edward. 2007. Theories of Social Inequality (fifth edition). Toronto: Thomson Nelson. Grabb, Edward, Robert Andersen, and Monica Hwang. 2009a. ?Bridging and bonding: ethnic background and voluntary association activity in Canada. Canadian Ethnic Studies 41, 1-2: 47-67. Grabb, Edward, Robert Andersen, Monica Hwang, and Scott Milligan. 2009b. ?Confidence in political institutions in Canada and the United States: assessing the interactive role of region and race.? American Review of Canadian Studies 39, 4 (December): 379-397. Grabb, Edward and James Curtis 1992. \"Voluntary association activity in English Canada, French Canada, and the United States: a multivariate analysis.\" Canadian Journal of Sociology 17, 4: 371-388. Grabb, Edward and James Curtis. 2005. Regions Apart: The Four Societies of Canada and the United States. Toronto: Oxford University Press. Granovetter, Mark. 1973. ?The strength of weak ties.? American Journal of Sociology 78, 6: 1360?1380. Granovetter, Mark. 1995. Getting a Job: A Study of Contacts and Careers. 2nd ed.. Chicago: University of Chicago Press. Gregg, Allan. 2006. ?Identity crisis: a twentieth-century dream becomes a twenty-first century conundrum.? Walrus 3, 2: 38-47. Grubel, Herbert (ed.). 2009. The Effects of Mass Immigration on Canadian Living Standrads and Society. Vancouver: The Fraser Institute. Guimond, Eric. 2003. ?Fuzzy definitions and population explosion: Changing identities of Aboriginal groups in Canada.? Pp. 35-50 in Not Strangers in these Parts, edited by David Newhouse and Evelyn Peters. Ottawa: Policy Research Initiative. Gwyn, Richard. 1995. Nationalism without Walls: The Unbearable Lightness of Being Canadian. Toronto: McClelland and Stewart. Hagan, John, Carla Shedd, and Monique Payne. 2005. ?Race, ethnicity, and youth perceptions of criminal justice.? American Sociological Review 70, 3: 381-407. Hagendoorn, Louk. 1993. ?Ethnic categorization and outgroup exclusion: cultural values and social stereotypes in the construction of ethnic hierarchies.? Ethnic and Racial Studies 16, 1: 26-51. Haig-Brown. Celia. 1988. Resistance and Renewal: Surviving the Indian Residential School. Vancouver: Arsenal Pulp Press. 184 Hansen, Randall. 2007. ?Diversity, integration and the turn from multiculturalism in the United Kingdom.? Pp. 351-386 in Keith Banting, Thomas J. Courchene, and F. Leslie Seidle, eds., Belonging? Diversity, Recognition, and Shared Citizenship in Canada. Montreal: Institute for Research on Public Policy. Harris, R. Cole and John Warkentin. 1974. Canada Before Confederation. New York: Oxford University Press. Henderson, Ailsa. 2004. ?Regional political cultures in Canada.? Canadian Journal of Political Science 37(3): 595-615. Henry, Frances. 2004. ?Two studies of racial discrimination in employment.? Pp. 285-294 in Social Inequality in Canada: Patterns, Problems, and Policies, edited by James Curtis, Edward Grabb, and Neil Guppy. 4th ed, Toronto: Pearson Prentice Hall. Henry, Frances and Effie Ginsberg. 1985. Who gets the work? A test of racial discrimination in employment. Toronto: Social Planning Council of Metro Toronto and the Urban Alliance on Race Relations. Henry, Frances and Carol Tator (eds.). 2005. The Colour of Democracy: Racism in Canadian Society, 3rd ed. Toronto: Nelson Thomson. Hero, Rodney E. Hero. 2003. ?Social capital and racial inequality in America,? Perspectives on Politics 1, 1: 113-122. Hildebrandt, Amber. 2010. ?Aboriginal Peoples, Muslims face discrimination most: Poll.? CBC News, March 15. Hooghe, Marc, Tim Reeskens, Dietlind Stolle, and Ann Trappers. 2006. ?Ethnic diversity, trust and ethnocentrism in Europe. A multilevel analysis of 21 European countries.? Paper presented at the 102nd Annual Meeting of the American Political Science Association, Philadelphia, August 31 to September 3, 2006. Hooghe, Marc, Tim Reeskens, Dietlind Stolle, and Ann Trappers. 2009. ?Ethnic diversity and generalized trust in Europe: a cross-national multilevel study.? Comparative Political Studies 42, 2 (February): 198-223. Hou, Feng, T.R. Balakrishnan, and Rozzet Jurdi. 2009. \"The economic integration of visible minorities in contemporary Canadian society: Revisited.\" Pp. 253-272 in Social Inequality in Canada: Patterns, Problems, and Policies, edited by Edward Grabb and Neil Guppy. 5th ed., Toronto: Pearson Prentice Hall. Hou, Feng and Simon Coulombe. 2010. \"Earnings gap for Canadian-born visible minority groups in the public and private sectors\" Canadian Public Policy 36(1): 29-43. Hou, Feng and Zheng Wu. 2009. \"Racial diversity, minority concentration and trust in Canadian urban neighbourhoods\" Social Science Research 38, 3: 693-716. 185 Howell, Susan, and Deborah Fagan. 1988. ?Race and trust in government: testing the political reality model.? Public Opinion Quarterly 52: 343-50. Hwang, Monica, Robert Andersen, and Edward Grabb. 2007. ?Voluntary association activity in Quebec and English Canada: assessing the effects of region and language.? Canadian Journal of Political Science 40, 1 (March): 209-232. Inglehart, Ronald. 1997. Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Democracies. Princeton: Princeton University Press. Jeannotte, Sharon. 1997. Social cohesion Research Workplan. SRA-266. Strategic Research and Analysis. Department of Canadian Heritage, Ottawa. Jeannotte, Sharon. 2000. ?Social cohesion around the world: An international comparison of definitions and issues.? Prepared for the Strategic Research and Analysis Directorate, Department of Canadian Heritage. Jenson, Jane. 1998. Mapping Social Cohesion: The State of Canadian Research. Ottawa: Canadian Policy Research Networks. Johnson, Richard. 2012. ?The homelands and homes of new Canadians.? The National Post, February 10. Johnston, Richard and Stuart Soroka. 2001. Social Capital in a Multicultural Society. In Social Capital and Participation in Everyday Life, ed. P. Dekker and E. Uslaner, 30-44. London: Routledge. Joppke, Christian. 2007. ?Transformation of immigrant integration in western Europe: civic integration and anti-discrimination policies in the Netherlands, France, and Germany.? World Politics 59, 2 (January): 243-273. Kalbach, Warren E., Ravi Verma, M.V. George, and S.Y. Dai. 1993. Population Projections of Visible Minority Groups, Canada, Provinces and Regions, 1991-2016. Ottawa: Statistics Canada, Working Group on Employment Equity Data, Population Projections Section, Demography Division. Kao, Grace and Lindsay Taggart-Rutherford. 2007. \"Does social capital still matter? Immigrant minority disadvantage in school-specific social capital and its effects on academic achievement.\" Sociological Perspectives 50(1): 27?52. Karakayali, Nedim. 2009. ?Social distance and affective orientations.? Sociological Forum 24,3: 538-562. Kasperson, Roger, Dominic Golding, and Seth Tuler. 1992. ?Social distance as a factor in siting hazardous factors and communicating risks.? Journal of Social Issues 48(4): 161-187. Kay, Barbara. 2008. ?Multiculturalism was Canada's biggest mistake.? The National Post, April 8. 186 Kazemipur, Abdolmohammad. 2006. ?A Canadian exceptionalism? Trust and diversity in Canadian cities.? Journal of International Migration and Integration 7, 2: 219-240. Keown, Leslie-Anne and Georgia Roberts. 2009. ?Using STATA with Statistics Canada data: incorporating complex survey design into analysis.? Presentation of the Canadian Users STATA Group, Ottawa. Kesler Christel and Irene Bloemraad. 2010. ?Does immigration erode social capital? The conditional effects of immigration-generate diversity on trust, membership, and participation across 19 countries, 1981-2000.? Canadian Journal of Political Science 43, 2 (June): 319-347. Kinnear, Michael. 2003. ?The effects of expansion of the franchise on turnout.? Electoral Insight 5, 3. Kleg, Milton and Kaoru Yamamoto. 1998. ?As the world turns: ethno-racial distances after 70 years.? The Social Science Journal 35, 2: 183-190. Kohut, Andrew, Jodie Allen, and Richard Wilke. ?Few signs of backlash from western Europeans: Muslims in Europe: economic worries top concerns about religious and cultural identity.? Pew Global Attitudes Project. Koopmans, Ruud. 2010. ?Trade-offs between equality and difference: immigrant integration, multiculturalism, and the welfare state in cross-national perspective.? Journal of Ethnic and Migration Studies. 36(1): 1-26. Kymlicka, Will. 2001. Politics in the Vernacular: Nationalism, Multiculturalism, and Citizenship. Oxford: Oxford University Press. Kymlicka, Will and Raphael Cohen-Almagor. 2000. ?Democracy and multiculturalism,? pp. 89-118 in Challenges to Democracy: Essays in Honour and Memory of Isaiah Berlin edited by R. Cohen-Almagor. London: Ashgate Publishing Ltd. Lamont, Michele and Marcel Fournier. 1992. Introduction, in M. Lamont and M. Fournier eds., Cultivating Differences: Symbolic Boundaries and the Making of Inequality, pp. 1-18. Chicago: University of Chicago Press. Lamont, Michele and Virag Molnar. 2002. ?The study of boundaries in the social sciences.? Annual review of Sociology 28: 167-195. Levi, Margaret, and Laura Stoker. 2000. ?Political trust and trustworthiness.? Annual Review of Political Science 3: 475-507. Li, Peter. 2003. Destination Canada: Immigration Debates and Issues. Toronto: Oxford University Press. 187 Lian, Jason and D. Ralph Matthews. 1998. ?Does the vertical mosaic still exist? Ethnicity and income in Canada, 1991.? Canadian Review of Sociology and Anthropology 35, 4: 461-481. Lin, Nan. 2000. ?Social capital: social networks, civic engagement, or trust?? Paper prepared for the Workshop on Social Capital, University of Trento, October 2000. Lin, Nan and Bonnie Erickson. 2008. ?Theory, measurement, and the research enterprise on social capital.? Pp. 1-26 in Social Capital: An International Research Program, edited by Nan Lin and Bonnie Erickson. New York: Oxford University Press. Marotte, Bertrand. 2013. ?Quebec?s ?welfare state? on the decline? On the contrary, says new study.? The Globe and Mail (Montreal edition), April 3, 2013. Maxwell, Judith. 1996. ?Social Dimensions of Economic Growth.? Department of Economics, University of Alberta: The Eric John Hanson Memorial Lecture Series, Volume VIII, January 25, 1996. Merelman, Richard. 1991. Partial Visions: Culture and Politics in Britain, Canada, and the United States. Madison: University of Wisconsin Press. Maaka, Roger and Augie Fleras. 2005. The Politics of Indigeneity. Dunedin, NZ: University of Otaga Press. Mata, Fernando and Ravi Pendakur. 2010. ?Social capital formation and diversity in Canadian cities: impacts of individual and place-related characteristics? Working paper series No. 10-02 January 2010 Metropolis British Columbia Centre of Excellence for Research on Immigration and Diversity. Menzies, Charles. 2009. ?First Nations, inequality, and the legacy of colonialism.? Pp. 295-304 in E. Grabb and N. Guppy (eds.), Social Inequality in Canada: Patterns, Problems, and Policies. Toronto: Pearson Prentice Hall. Michalowski, Ines. 2003. ?What is the Dutch integration model and has it failed?? Policy Brief No. 1. Focus Migration. Hamburgiches WeltWirtschafts Institut Migration Research Group. Miller, Arthur H., and Ola Listhaug. 1990. ?Political parties and confidence in government: a comparison of Norway, Sweden, and the United States.? British Journal of Political Science 20: 357-86. Musick, Marc and John Wilson. 2008. Volunteers: A Social Profile. Bloomington, IN: Indiana University Press. Nakhaie, M. Reza. 2006. ?Contemporary realities and future visions: Enhancing multiculturalism in Canada.? Canadian Ethnic Studies 38, 1: 149-158. 188 Nelson, Dale C. 1979. ?Ethnicity and socioeconomic status as sources of participation: The case of ethnic political culture.? American Political Science Review 73: 1024-38. Nevitte, Neil. 1996. The Decline of Deference. Peterborough, ON: Broadview Press. Norris, Pippa, ed. 1999. Critical Citizens: Global Support for Democratic Government. Oxford: Oxford University Press. Nunnally, Shayla. 2011. ?(Dis)Counting on democracy to work: perceptions of electoral fairness in the 2008 presidential election.? Journal of Black Studies 42, 6: 923-948. Onyx, Jennifer and Paul Bullen. 2001. ?The different faces of social capital in NSW Australia.? Pp. 45 - 58 in Social Capital and Participation in Everyday Life, edited by Eric M. Uslaner. London: Routledge. Oreopoulos, Philip. 2009. ?Why do skilled immigrants struggle in the labor market? A field experiment with six thousand r?sum?s.? (May) Unpublished manuscript. Palacios, Milagros, Niels Veldhuis, and Michael Walker. 2008. Tax Facts 15. Vancouver: The Fraser Institute. Parillo,Vincent and Christopher Donoghue. 2007. ?Updating the Bogardus social distance studies: a new national survey.? The Social Science Journal 42,2: 257-271. Park, Robert E. 1924. \"The concept of social distance as applied to the study of racial attitudes and racial relations.\" Journal of Applied Sociology 8: 339-344. Park, Robert E. 1930. ?Assimilation, social.? In Encyclopedia of the Social Sciences, edited by E.R.A. Seligman and A. Johnson. New York: Macmillan. Patterson, Orlando. 1999. ?Liberty against the democratice state: on the historical and contemporary sources of American distrust.? In Democracy and Trust, ed. Mark Warren, pp. 151-207. Cambridge: Cambridge University Press. Pattillo-McCoy, Mary. 1999. Black Picket Fences: Privilege and Peril among the Black Middle Class. Chicago: University of Chicago Press. Paxton, Pamela. 2007. ?Association membership and generalized trust.? Social Forces 86,1: 47-76. Pendakur, Krishna. 2005. ?Visible minorities and Aboriginals in Vancouver?s labour market.? Vancouver Centre of Excellence. Research on Immigration and Integration in the Metropolis. Working Paper Series No-05-12. Pendakur, Krishna and Ravi Pendakur. 1998. ?The colour of money: earnings differentials among ethnic groups in Canada.? Canadian Journal of Economics 31, 3: 518-548. 189 Pendakur, Krishna and Ravi Pendakur. 2002. ?Colour my world: have earnings gaps for Canadian-born ethnic minorities changed over time?? Canadian Public Policy 28, 4: 489-512. Perlin, George. 1997. ?The constraints of public opinion: diverging or converging paths?? In Degrees of Freedom: Canada and the United States in a Changing World, ed. Keith Banting, George Hoberg, and Richard Simeon, 71-149. Montreal and Kingston: McGill-Queen?s University Press. Portes, Alejandro. 1995. ?Children of immigrants: Segmented assimilation and its determinants.\" Pp. 248-280 in Alejandro Portes (ed.) The Economic Sociology of Immigration. New York: Russell Sage. Portes, Alejandro. 1998. ?Social capital: Its origins and applications in modern sociology.? Annual Review of Sociology 24:1-24. Portes, Alejandro and Min Zhou. 1993 ?The new second generation: segmented assimilation and Its Variants.? Annals of the America Academy of Political and Social Science 530:74-96 Portney, Kent E. and Jeffrey M. Berry. 1997. ?Mobilizing minority communities: Social capital and participation in urban neighbourhoods.? American Behavioral Scientist 40(5): 632- 644. Powers, Daniel A and Christopher G. Ellison. 1995. ?Interracial contact and black racial attitudes: The contact hypothesis and selectivity bias.? Social Forces 74:205-27. Porter, John. 1965. The Vertical Mosaic: An Analysis of Social Class and Power in Canada. Toronto: University of Toronto Press. Posgate, Dale and Kenneth McRoberts. 1976. Quebec: Social Change and Political Crisis. Toronto: McClelland and Stewart. Proussalidis, Daniel . 2013. ?2011 Census: nearly one in five Canadians are visible minorities, StatsCan reports.? Toronto Sun, May 8. Putnam, Robert. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. Putnam, Robert D. 1995. ?Bowling alone: America?s declining social capital.? Journal of Democracy 6: 65-78. Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon and Schuster. Putnam, Robert D. 2007. ?E pluribus unum: Diversity and community in the twenty-first century.? The 2006 Johan Skytte Prize Lecture. Scandinavian Political Studies 30, 2: 137-74 190 Ramakrishnan, Karthick and Thomas Espenshade. 2001. ?Immigrant incorporation and political participation in the United States.? International Migration Review 35(3): 870-909. Rahn, Wendy and Thomas Rudolph. 2005. ?A tale of political trust in American cities.? Public Opinion Quarterly 69, 4 (Winter): 530-560. Ramakrishnan, Karthick and Irene Bloemraad (eds.). 2008. Civic Hopes and Political Realities: Immigrants, Community Organizations, and Political Engagement. New York: Russell Sage Foundation Press. Reitz, Jeffrey G. 1980a. The Survival of Ethnic Groups. Toronto: McGraw-Hill Ryerson. Reitz, Jeffrey G. 1980b. ?Immigrants, their descendents, and the cohesion of Canada.? Pp. 329-417 in R. Breton, J. Reitz, and V. Valentine (eds.), Cultural Boundaries and the Cohesion of Canada. Montreal: Institute for Research on Public Policy. Reitz, Jeffrey G. 1998. Warmth of the Welcome: The Social Causes of Economic Success for Immigrants in Different Nations and Cities. Boulder, CO: Westview Press. Reitz, Jeffrey G. 2001a. ?Immigrant skill utilization in the Canadian labour market: implications of human capital research.? Journal of International Migration and Integration 2, 3: 347-378. Reitz, Jeffrey G. 2001b. ?Immigrant success in the knowledge economy: institutional change and the immigrant experience in Canada.? Journal of Social Issues 57, 3: 577-611. Reitz, Jeffrey G. 2002 ?Host societies and the reception of immigrants: research themes, emerging theories, and methodological issues, ? International Migration Review 36,4 (Winter): 1005-1019. Reitz, Jeffrey G. 2006. ?Does Canadian experience in immigrant incorporation have lessons for Europe?? Based on a keynote address at the Third Annual Conference of IMISCOE (International Migration, Integration, and Social Cohesion) network of excellence in European migration research, Vienna, September 5. Reitz, Jeffrey G. and Rupa Banerjee. 2007. ?Racial inequality, social cohesion, and policy issues in Canada.? Pp. 489-545 in Keith Banting, Thomas J. Courchene, and F. Leslie Seidle, eds., Belonging? Diversity, Recognition, and Shared Citizenship in Canada. Montreal: Institute for Research on Public Policy. Reitz, Jeffrey G. and Raymond Breton. 1994. The Illusion of Difference: Realities of Ethnicity in Canada and the United States. Toronto: C.D. Howe Institute. Reitz, Jeffrey G. and Rupa Banerjee. 2009. ?Racial inequality, social cohesion, and policy issues in Canada.? Pp. 273-294 in Social Inequality in Canada: Patterns, Problems, and Policies, edited by Edward Grabb and Neil Guppy. 5th ed, Toronto: Pearson Prentice Hall. 191 Richmond, Anthony and John Goldlust. 1977. Family and Social Integration of Immigrants in Toronto. Toronto: Ethnic Research Program, York University. Richomme, Olivier. 2009. ?The role of ?ethno-racial? classification in the Americanization process? Cercles 19: 1-17. Roberts, Julian, and Anthony Doob. 1997. ?Race, ethnicity, and criminal justice in Canada.? Crime and Justice 21: 469-522. Robinson, Robert B. and Elton F. Jackson. 2001. ?Is trust in others declining in America? An age?period?cohort analysis.? Social Science Research 30, 1 (March): 117-145. Roseborough, Howard and Raymond Breton. 1968. ?Perceptions of the relative economic and political advantages of ethnic groups in Canada.? Pp. 604-628 in Canadian Society: Sociological Perspectives, 3rd ed., edited by Bernard Blishen, Frank Jones, Kaspar Naegele, and John Porter. Toronto: Macmillan of Canada, Roth, Wendy. 2009. ??Latino before the world?: the transnational extension of panethnicity.? Ethnic and Racial Studies 32(6): 927-947. Roth, Wendy, Marc-David Seidel, Dennis Ma, and Eiston Lo. 2012. ?In and out of the ethnic economy: A longitudinal analysis of ethnic networks and pathways to economic success across immigrant categories.? International Migration Review 42(2): 310-360. Rotolo, Thomas. 2000. ?Town heterogeneity and affiliation: A multilevel analysis of voluntary association membership.? Sociological Perspectives 43: 271-289. Samuel, John and Kogular Basavarajappa. 2006. ?The visible minority population in Canada: a review of numbers, growth, and labour force issues.? Canadian Studies in Population 33, 2: 241-269. Satzewich, Vic (ed.). 1998. Racism and Social Inequality in Canada: Concepts, Controversies, and Strategies of Resistance. Toronto: Thompson Educational Publishing. Satzewich, Vic. 2011. Racism in Canada. Toronto: Oxford University Press. Satzewich, Vic and Nikolaos Liodakis. 2007. ?Race? and Ethnicity in Canada: A Critical Introduction. Toronto: Oxford University Press. Scavo, Carmine. 2007. ?Constructing data sets and manipulating data.? Pp. 303-322 in K. Yang and G. Miller (eds.), Handbook of Research Methods in Public Administration 2nd ed. Boca Raton, FL: CRC Press. Schwartz, Mildred. 1974. Politics and Territory: The Sociology of Regional Persistence in Canada. Montreal: McGill-Queen?s University Press. Simeon, Richard and David Elkins. 1974. ?Regional political cultures in Canada.? Canadian Journal of Political Science 7: 397-437. 192 Simmel, Georg. 1908. Soziologie: Untersuchungen ?ber die Formen der Vergesellschaftung. Leipzig: Duncker & Humblot. Simmel, Georg. 1950. The Sociology of Georg Simmel. Translated edition. New York: Free Press. [German original, 1908]. Simmel, Georg. 1990. The Philosophy of Money. London: Routledge. [German original, 1900]. Smith, Sandra Susan. 2005. ?'Don?t put my name on it': social capital activation and job-finding assistance among the Black urban poor.? American Journal of Sociology 111, 1: 1-57. Smith, Sandra Susan. 2007. Lone Pursuit. Distrust and Defensive Individualism Among the Black Poor. New York: Russell Sage Foundation. Smith, Sandra Susan. 2008. ?A question of access or mobilization? Understanding inefficacious job referral networks among the black poor.? Pp. 157-184 in Social Capital: An International Research Program, edited by Nan Lin and Bonnie Erickson. New York: Oxford University Press. Smith, Sandra Susan. 2010. ?Race and trust.? Annual Review of Sociology 36: 453-475. Smith, Tom W. 1997. ?Factors relating to misanthropy in contemporary American society.? Social Science Research 26, 2 (June): 170-196. Soroka, Stuart N., Richard Johnston, and Keith Banting. 2006. ?Ties that bind? Social cohesion and diversity in Canada.? Pp. 561-600 in Keith Banting, Thomas J. Courchene, and F. Leslie Seidle, eds., Belonging? Diversity, Recognition, and Shared Citizenship in Canada. Montreal: Institute for Research on Public Policy. Soroka, Stuart, Richard Johnston, and Keith Banting. 2007. ?Ethnicity, trust, and the welfare state.? In Social Capital, Diversity, and the Welfare State, ed. Fiona Kay and Richard Johnston, 279-303. Vancouver: University of British Columbia Press. Statistics Canada. 2003. Ethnic Diversity Survey: A Portrait of a Multicultural Society. Catalogue no. 89-593-XIE. Statistics Canada. 2005a. ?Fears, perceptions, and attitudes toward the justice system.? Canadian Centre for Justice Statistics Profile Series, Number 11, Catalogue no. 85F0033MWE. Statistics Canada. 2005b. Population Projections of Visible Minority Groups, Canada, Provinces and Regions, 2001 to 2017. Catalogue no. 91-541-XIE. Ottawa: Statistics Canada. Statistics Canada. 2006a. Aboriginal Peoples in Canada in 2006: Inuit, M?tis and First Nations, 2006 Census. Catalogue no. 97-558-XIE. 193 Statistics Canada. 2006b. Immigration in Canada: A Portrait of the Foreign-born Population, 2006 Census. Catalogue no. 97-557-XIE 2006001. Statistics Canada. 2007. Residential Telephone Services Survey, December. Catalogue no. 56M0001XCB. Statistics Canada. 2008a. ?Aboriginal peoples in Canada in 2006: Inuit, M?tis, and First Nations, 2006 Census.? The Daily. Tuesday, January 15. http://www.statscan.gc.ca/daily-quotidien/080115a-eng.htm Statistics Canada. 2008b. ?Census: ethnic origin, visible minorities, place of work and mode of transportation.? The Daily. Wednesday, April 2. http://www.statscan.gc.ca/daily-quotidien/08402/dq080402a-eng.htm Statistics Canada. 2009. ?Perceptions of the criminal justice system, by Aboriginal identity, Canada's ten provinces, Table 13.? statcan.gc.ca/pub/85-002-x/2011001/article/11415/tbl/tbl13-eng.htm Statistics Canada. 2010. Aboriginal Statistics at a Glance: Provincial/territorial Distribution of Aboriginal Peoples. http://www.statcan.gc.ca/pub/89-645-x/2010001/territorial-territoire-eng.htm Statistics Canada. 2011. ?Residential telephone survey.? The Daily, Tuesday, April 5. http://www.statcan.gc.ca/daily-quotidien/110405/dq110405a-eng.htm Statistics Canada. 2013. ?Visible minorities, immigration status and period of immigration, age groups, and sex, for the population in private households of Canada, provinces, territories, Census Metropolitan Areas, and Census Agglomerations, National Household Survey.? Catalogue no. 99-010-X2011030. Stolle, Dietlind. 1998. ?Bowling together, bowling alone: the development of generalized trust in voluntary associations.? Political Psychology 19: 497-525. Stolle, Dietlind. 2001. ?Clubs and congregations: the benefits of joining an association.? Pp. 202-244 in Trust in Society, edited by Karen S. Cook. New York: Russell Sage Foundation. Stolle, Dietlind, Stuart Soroka, and Richard Johnston. 2008. ?When does diversity erode trust? Neighborhood diversity in interpersonal trust and the mediating effect of social interaction.? Political Studies 56: 57-75. Taiaiake, Alfred, Brock Pitawanakwat, and Jackie Price 2007. ?The meaning of political participation for indigenous youth: charting the course for youth civic and political participation.? Indigenous Governance Programs, University of Victoria. Canadian Policy Research Networks. Taiaiake, Alfred and Jeff Corntassel. 2005. ?Being Indigenous: resurgences against contemporary colonialism.? Government and Opposition 9: 597?614 194 Tigges, Leann M., Irene Browne, and Gary P. Green. 1998. ?Social isolation of the urban poor: Race, class, and neighborhood effects on social resources.? The Sociological Quarterly 39, 1: 53?77. Tindall, David and Jeffrey Cormier. 2008. ?Gender, network capital, social capital, and political capital: the consequences of personal network division for environmentalists in British Columbia.? Pp. 282-307 in Social Capital: An International Research Program, edited by Nan Lin and Bonnie Erickson. New York: Oxford University Press. Tonkiss, Fran and Andrew Passey. 1999. ?Trust, confidence, and voluntary organizations: Between values and institutions.? Sociology 33(2): 257-274. Tuch, Steven, and Ronald Weitzer. 1997. ?Racial differences in attitudes toward the police.? Public Opinion Quarterly 61: 642-63. Uslaner, Eric. 1998. ?Social capital, television, and the ?mean world?: trust, optimism, and civic participation.? Political Psychology 19: 441?467. Uslaner, Eric. 2002. The Moral Foundations of Trust. Cambridge, UK: Cambridge University Press. Uslaner, Eric. 2004. ?Trust and corruption.? Pp. 76-92 in J. Graf, L. Taube, and M. Schramm, eds., Corruption and the New Institutional Economics. London: Routledge. Uslaner, Eric. 2011. ?Trust, diversity, and segregation in the United States and the United Kingdom.? Comparative Sociology 10: 221-247. Ward, Doug. 2008. Is visible minority still relevant? www.vancouversun.com, April 8. Retrieved March 1, 2013 (http://www.canada.com/vancouversun/news/story.html?id=8d16d9cd-5342-451f-8aef-33491a35526e) Yu, Henry. 2009. ?Global migrants and the new pacific Canada.?International Journal (August): 147-62. Veenstra, Gerry. 2002. ?Explicating social capital: trust and participation in the civil space.? Canadian Journal of Sociology 27, 4: 547-572. Warner, Lloyd, and Leo Srole. 1945. The Social Systems of American Ethnic Groups. New Haven, Connecticut: Yale University Press. Waters, Mary. 1994. ?Ethnic and racial identities of second-generation black immigrants in New York City.? International Migration Review 28, 4: 795-820. Weitzer, Ronald, and Steven Tuch. 2004. ?Race and perceptions of police misconduct.? Social Problems 51: 305-325. 195 White, Jerry, Dan Beavon, and Nicholas Spence (eds.). 2007. Aboriginal Well-Being in Canada. Toronto: Thompson Educational Publishing. Wilkes, Rima. 2006. ?The protest actions of indigenous peoples: A Canadian-U.S. comparison of social movement emergence.? American Behavioral Scientist 50, 4: 510-525. Wilkes, Rima. 2011. ?Rethinking the decline in trust: a comparison of black and white Americans.? Social Science Research 40, 6 (November): 1596-1610. Wilkes, Rima, Neil Guppy, and Lily Farris. 2007. ?Right-wing parties and anti-foreigner sentiment in Europe.? American Sociological Review 72 (October): 831-840. Wilkes, Rima, Neil Guppy, and Lily Farris. 2008. ??No thanks, we?re full?: individual characteristics, national context, and changing attitudes toward immigration.? International Migration Review 42(2) Summer: 302-329. Williams, Joyce E. 2007. The Blackwell Encyclopedia of Sociology (Online). Edited by George Ritzer. London: Basil Blackwell. Wilson, John and Marc Musick. 1997. ?Who cares? Toward an integrated theory of volunteer work.? American Sociological Review 62: 694-713. Wilson, John and Marc Musick. 1998. ?Social resources and volunteering.? Social Science Quarterly 79: 799-814. Wood, Gordon S. 1992. The Radicalism of the American Revolution. New York: Knopf. Wortley, Scot. 1996.?Justice for all? Race and perceptions of bias in the Ontario criminal justice system: a Toronto survey.? Canadian Journal of Criminology 38: 439-67. Wortley, Scot, John Hagan, and Ross Macmillan. 1997. ?Just des(s)erts: The racial polarization of perceptions of injustice.? Law and Society Review 31: 637-676. Wortley, Scot, and Julian Tanner. 2003. ?Data, denials and confusion: the racial profiling debate in Toronto.? Canadian Journal of Criminology and Criminal Justice 45: 367-89. Wu, Zheng, Feng Hou, and Christoph Schimmele. 2011. ?Racial diversity and sense of belonging in urban neighborhoods.? City and Community 10, 4 (December): 373-392. Yanow, Dvora. 2003. Constructing Race and Ethnicity in America: Category-Making in Public Policy and Adminstration. New York: M.E. Sharpe, Inc. You, Jon-Sung. 2012. ?Social trust: fairness matters more than homogeneity.? Political Psychology 33 (first published online, June). Young, Robert. 1995. The Secession of Quebec and the Future of Canada. Montreal and Kingston: McGill-Queen?s University Press. 196 APPENDIX A Note on Regional Differences in Social and Political Trust Table A1 in the Appendix illustrates some of the ways in which both the social climate and the political culture of Quebec appear different from that found in other regions of the country. The data show the mean scores on the overall social trust scale and the overall political trust scale for the eight ethno-racial groupings, both for the full Canadian sample and for each of the five regional samples. The results suggest that there may be a Quebec regional culture phenomenon. That is, first, we see that, for every one of the eight ethno-racial groupings in the analysis, their lowest social trust scores are to be found in Quebec. Second, we see that, except for the Aboriginal Peoples and French categories, all of the ethno-racial groups have their highest political trust score if they reside in Quebec. Therefore, it would appear that Quebec provides a regional culture in which, in relative terms, trust in other people is exceptionally low, but trust in political institutions is exceptionally high, and this occurs almost regardless of which ethno-racial grouping is being considered. 197 TABLE A1 0.1 198 TABLE A2 2 Descriptive Statistics for Independent and Control Variables Percent N Ethno-racial Group: British 27% (4724) French 11 (1848) Canadian 8 (1382) Other European 14 (2344) Aboriginal Peoples 3 (561) Visible Minority 12 (2160) Other Mixed Origins Immigration Status: Canadian-born 78% (14443) Immigrated since 2000 5% (792) Immigrated 1990-1999 5% (786) Immigrated 1980-1989 4% (655) Immigrated 1970-1979 3% (616) Immigrated before 1970 6% (1099) Religion: No Religion 21% (3856) Protestant 32 (5660) Roman Catholic 39 (7124) Non-Christian 7 (1266) Language: Anglophone 68% (12260) Francophone 21 (3889) Allophone 11 (2004) Education (mean in years): 13.6 Education (categories): Elementary or less 3% (599) Some high school 13 (2275) High school graduate 14 (2518) Some post-secondary 13 (2299) Post-secondary diploma 29 (5346) Bachelor degree 20 (3671) Post-bachelor degree 8 (1540) Income (mean in dollars): 69,614 Income (categories): Less than $20,000 7% (1044) $20,000-29,999 8 (1183) $30,000-39,999 9 (1355) $40,000-49,999 9 (1382) $50,000-59,999 11 (1592) $60,000-79,999 16 (2426) $80,000-99,999 13 (1877) $100,000 or more 28 (4211) 199 TABLE A2 continued Descriptive Statistics for Independent and Control Variables Percent N Vol. Memberships: None 36% (6559) One 26 (4788) Two 19 (3472) Three or More 20 (3587) Diversity of Friendships: No ethnically different friends 48% (8373) A few 39 (6697) About half 6 (1108) Most 4 (733) All 3 (502) Region: Atlantic 7% (1332) Quebec 23 (4408) Ontario 39 (7149) Prairies 17 (3045) British Columbia 14 (2523) Age (mean in years): 49.6 Age (categories): 25-29 10% (1853) 30-34 10 (1817) 35-39 10 (1885) 40-44 11 (2053) 45-49 12 (2209) 50-54 11 (2017) 55-59 9 (1725) 60-64 8 (1441) 65-69 6 (1065) 70-74 5 (835) 75-79 4 (695) 80 or older 5 (862) Gender: Female 51% (9413) Male 49 (9044) Descriptive Statistics for Dependent Variables Social Trust: Mean = 0.0, Standard Deviation = 1.0, Range from -2.26 to + 2.44 Political Trust: Mean = 0.0, Standard Deviation = 1.0, Range from -3.48 to + 2.33 "@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2013-11"@en ; edm:isShownAt "10.14288/1.0042039"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Sociology"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "Attribution 2.5 Canada"@en ; ns0:rightsURI "http://creativecommons.org/licenses/by/2.5/ca/"@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "In Canada we trust? understanding ethno-racial variations in social and political trust"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/45063"@en .