Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The attitudinal mosaic : forming attitudes about multiculturalism, immigration, and ethnic diversity… Penner, Erin Margaret 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2013_fall_penner_erin.pdf [ 1.24MB ]
JSON: 24-1.0103355.json
JSON-LD: 24-1.0103355-ld.json
RDF/XML (Pretty): 24-1.0103355-rdf.xml
RDF/JSON: 24-1.0103355-rdf.json
Turtle: 24-1.0103355-turtle.txt
N-Triples: 24-1.0103355-rdf-ntriples.txt
Original Record: 24-1.0103355-source.json
Full Text

Full Text

   The Attitudinal Mosaic:  Forming Attitudes about Multiculturalism, Immigration,  and Ethnic Diversity in Canada   by  Erin Margaret Penner    B.A., University of Manitoba, 2002 (Religion) B.A., McGill University, 2004 (Political Science) M.A., McGill University, 2006 (Political Science)     A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Political Science)      The University of British Columbia (Vancouver)    October 2013    ? Erin Margaret Penner, 2013  ii ABSTRACT  Western liberal democracies are grappling with the issue of ethnic diversity.  Canada, with its shifting immigration patterns from ?traditional? to ?non-traditional? sender countries, is no exception.  Since 1971, Canada?s federal government has pursued a policy of multiculturalism in its approach to interethnic relations.  The normative assumption behind the approach is that ethnic diversity is good and that affirming ethnic minority cultures is the best way to achieve a strong, cohesive society for all Canadians.  This dissertation investigates the behavioural aspect of the multicultural assumption.  That is, how do people actually behave in a multicultural society?  What factors determine the shape of the attitudinal terrain?  How do citizens really perceive their neighbors and what are the consequences?   This dissertation is interested not just in what people say about ethnic diversity, but why people say it.  In a series of four articles, it looks at a range of issues, including the attitudinal determinants of comfort with intergroup contact, the marginal effects of individual and contextual factors of attitudes toward immigrant integration for whites and non-whites, and the interaction between political values and personality in overcoming negative reactions to ethnic stereotypes.  I draw on political science theories of behaviour, as well as social psychological and personality psychological theories of attitude and identity formation.  I employ recent public opinion data from various sources to test hypotheses with descriptive statistics and regression analysis.   Given the dissertation?s article structure, there are numerous conclusions to be drawn.  For instance, I highlight the relative importance of prejudice, rooted in an individual?s social identity, in shaping attitudes about ethnic outgroups.  I also show the analytic dangers of conflating value differences with physical differences when assessing the perceptions of  iii minority groups.  Along the same lines, I show the importance of investigating minority opinion separately from majority opinion: key attitudinal determinants impact these groups quite differently.  I illustrate the power of political values in overriding potent negative stereotypes.  Finally, I uncover evidence that supports the assumptions behind Canada?s multiculturalism policy, namely, that ethnic minorities confident in their own culture will be more willing to contribute to a strong Canadian society.    iv PREFACE  This dissertation is the original, unpublished, independent work by the author, Erin Margaret Penner.    v TABLE OF CONTENTS Abstract ....................................................................................................................................... ii Preface ....................................................................................................................................... iv Table of Contents ........................................................................................................................ v List of Tables ............................................................................................................................ vii List of Figures .......................................................................................................................... viii Acknowledgements ................................................................................................................... ix CHAPTER ONE INTRODUCTION .................................................................................. 1 CHAPTER TWO RANKING MINORITIES IN CANADA: THE ROLE OF PERCEIVED ETHNO-RACIAL AND RELIGIOUS DIFFERENCES  ................................. 14 The Categorization of Outgroups ..................................................................................... 19 The Intimacy of the Contact Setting ................................................................................. 22 Data and Methods ............................................................................................................. 24 Dependent variable ....................................................................................................... 24 Independent variables ................................................................................................... 33 Results .............................................................................................................................. 35 Discussion ......................................................................................................................... 40 CHAPTER THREE DETERMINANTS OF WHITE AND NON-WHITE ATTITUDES ABOUT ETHNIC DIVERSITY IN CANADA: THE ROLE OF ETHNIC PREJUDICE, MATERIAL SELF-INTEREST, AND POLITICAL VALUES  ............................................. 50 Determinants of Ethnic Attitudes ..................................................................................... 52 Majority and Minority Attitudes ....................................................................................... 61 Data and Methods ............................................................................................................. 63 Dependent variables ..................................................................................................... 64 Individual-level independent variables ......................................................................... 66 Results .............................................................................................................................. 74 Discussion ......................................................................................................................... 81  CHAPTER FOUR DETERMINANTS OF WHITE AND NON-WHITE ATTITUDES ABOUT ETHNIC DIVERSITY IN CANADA: THE ROLE OF CONTEXT ........................ 88 Contextual Level Influences on Attitudes about Ethnic Diversity ................................... 90 Ingroup Size and Subjective Uncertainty ......................................................................... 94 Data and Methods ........................................................................................................... 100     Contextual-level independent variables ...................................................................... 101 Results ............................................................................................................................ 104 Discussion ....................................................................................................................... 108  CHAPTER FIVE PERSONALITY, POLITICAL VALUES, AND ATTITUDES ABOUT ETHNIC MINORITIES IN CANADA .................................................................................. 115 Psychology and Prejudice ............................................................................................... 118 Personality and prejudice ............................................................................................ 118 Social psychology and prejudice ................................................................................ 120  vi Political Values and Prejudice  ....................................................................................... 122 Personality, Stereotypes, Political Values, and Prejudice against Muslims  .................. 127 Data and Methods ........................................................................................................... 130 The first analysis: Personality in the ?Fatma? experiment ......................................... 130 The second analysis: The expanded model ................................................................ 134 Discussion ....................................................................................................................... 138 CHAPTER SIX CONCLUSION .................................................................................. 145 Summary of Main Findings ............................................................................................ 147 Limitations ...................................................................................................................... 152 Implications .................................................................................................................... 153 BIBLIOGRAPHY .................................................................................................................. 155 APPENDIX ONE ................................................................................................................... 175 APPENDIX TWO .................................................................................................................. 179 APPENDIX THREE ............................................................................................................... 181 APPENDIX FOUR ................................................................................................................. 182        vii LIST OF TABLES 2.1 Determinants of Comfort Ratings for Ethno-Racially and Religiously Defined Outgroups .................................................................................................................................................. 47 3.1 Individual-Level Determinants of Disagreeing Immigration Has Been a Good Thing for Canada and Preferring a Melting Pot to a Mosaic .................................................................... 86 3.2 Marginal Changes for Predicted Strong Disagreement with Immigration and Preferences for a Melting Pot ....................................................................................................................... 86 4.1 Contextual-Level Determinants of Disagreeing Immigration Has Been a Good Thing for Canada, Immigrants are Obligated to Learn Canadian Ways, and Preferring a Melting Pot to a Mosaic .................................................................................................................................... 113 5.1  Ten Item Personality Index .............................................................................................. 142 5.2  Personality Traits and Attitudes toward Fatma?s Multicultural Grant Application, Both Conditions ............................................................................................................................... 142 5.3  Predictors of Opposition to Fatma?s Multiculturalism Grant Application for Openness to Experience and Agreeableness Personalities, Combined Conditions .................................... 143    viii LIST OF FIGURES 2.1 Predicted ?Very Comfortable? Response for Outgroups in Three Settings ........................ 46 2.2a-j  Predicted ?Very Comfortable? Response for Outgroups as In-Laws ............................ 48 2.3a-j  Differences in Predicted ?Very Comfortable? Response for Outgroups as In-Laws ..... 49 3.1ai-bii  Predicted Probabilities after Extreme Changes in Prejudice, Material Self-Interest, and Political Values for Whites and Non-Whites ..................................................................... 87 4.1a-c  Predicted Probabilities after Changes in Local Visible Minority Presence for Whites and Non-Whites ...................................................................................................................... 114 5.1  High in Openness Personalities and Opposition to Fatma?s Multicultural Grant Application at Different Levels of Liberal Egalitarian Values ............................................... 144        ix ACKNOWLEDGEMENTS   This dissertation could not have been written without the assistance, patience, and good humour of many teachers, advisors, colleagues, family members, and friends.  I would first like to thank my doctoral advisor, Fred Cutler, who has helped me in innumerable ways.  The members of my committee, Richard Johnston and Antje Ellerman, also deserve praise for their meticulous comments and criticism that has shaped this dissertation for the better.  I am very grateful to Reginald Bibby from the University of Lethbridge for graciously providing me with his Project Canada data.  I?d also like to thank numerous faculty and friends at UBC and McGill for teaching me, both formally and informally, about politics and the art of political science.  Of specific note is Dietlind Stolle for invaluable comments on earlier drafts, and Stuart Soroka, without whom I would have not pursued this degree in the first place.  And, the most heartfelt thanks goes to my family, who has put up with the highs and lows of a more-than-a-decade-long academic career: Edith and Steve, Lloyd and Emma, Ainsley and Elora, Steven and Jenny, Ronan, Rhedyn, Fin, and Julie, Jason, and Sal.  The biggest thanks, of course, goes to Jeremy and Isla.     1 CHAPTER ONE INTRODUCTION  The issue of ethnic diversity has become an urgent political question in many Western societies.  Increased ethnic diversification due to shifting immigration patterns over the past 40 years has forced these societies to rethink how they integrate new ethnic communities.  In Canada, the demographic impact of immigration from ?non-traditional? countries has been considerable: the 2006 Canadian census found that almost 20 per cent of Canadians were born outside Canada and about 85 per cent of these immigrants were from non-European countries.  When compared to 35 years ago, the change is dramatic: in 1971, 15 per cent of Canadians were foreign born, and only 24 per cent were from non-European countries (Chui et al. 2007).  Moreover, the 2011 National Household Survey, which is completed by 20 percent of Canadians, estimates that about one in five Canadians can be classified as a visible minority.1  While many Western European countries have seen growing xenophobia and a retreat from multiculturalism policies, Canada still enjoys considerable popular support for its policy approach to ethnic integration (Reitz et al. 2009).  In fact, Canada tends to be a positive outlier when citizens? attitudes are compared with citizens from other democracies (Hiebert 2006).  An all-party consensus on the value of immigration and ethnic diversity also exists.  Disagreements between parties focus only on the mechanics of policies, not on the existence of the policies themselves.  But how stable is this widespread support?  Are Canadians truly exceptional in their progressive acceptance of ethnic diversity?  Or does the calm on the surface obscure roiling waters underneath?                                                   1 Statistics Canada. 2011. National Household Survey.  Figure 2.  Ottawa. May 1, 2013.  2 The stakes are high for understanding how Canadians think about ethnic diversity and the related issues of multiculturalism and immigration.  Canada has largely escaped the violent interethnic conflicts that have erupted in other Western democracies, such as England, France, the Netherlands, and Australia.  Still, there have been controversies that reveal tensions between the majority and minority ethnic groups.  The recently foiled plot to derail Via Rail trains by two al-Qaeda-backed men is just one example of interethnic tensions.  Others include the 2006 arrest of the so-called Toronto 18, who were involved in al-Qaeda training camps outside Toronto, as well as the taut conditions between majority and minority Quebeckers that led to the Bouchard-Taylor Commission on Reasonable Accommodation in 2008. Beyond the stakes on the ground, there are normative concerns for the quality of Canadian democracy.  It is vital to understand why some groups are systematically disadvantaged in society if the pluralistic principle of democratic politics is to be realized.  A political system that hinges on the fair competition of disparate interests is hampered when some groups, such as ethnic minorities, must navigate additional hurdles in order to participate.  Recognizing that some groups are systemically advantaged over others is not new.   Pratto, Sidanius, and colleagues? Social Dominance Theory captures how all societies are arranged in group-based hierarchies (e.g., Pratto et al. 2006).  Moreover, in Canada, the unequal power relations between ethnic groups was outlined as early as John Porter?s The Vertical Mosaic (1965).  Examining how individual attitudes contribute to this social and political asymmetry can provide insight into how unequal power relations can be remedied and democratic pluralism achieved.  3 Unfortunately, little is known about how Canadians formulate their attitudes about ethnic diversity, immigration, and multiculturalism.  Commercial opinion polls regularly find Canadians hold positive attitudes about ethnic diversity.  These polls, however, inquire about attitudes at a superficial level and, thus, are limited for those interested in a deeper understanding.  Importantly, there is a new and growing literature on the consequences of ethnic diversity, such as its effects on interpersonal trust (e.g., Stolle et al. 2008), social cohesion (e.g., Soroka et al. 2006), voting behaviour (e.g., Bilodeau et al. 2010), and redistributive policy preferences (e.g., Johnston et al. 2010).  This literature rarely goes beyond basic socio-demographic predictors of attitudes (important Canadian exceptions include Citrin et al. 2012; Harell et al. 2012).  To be sure, it is insightful to examine attitudinal divisions between white and non-whites or domestic- and foreign-born citizens.  In fact, the present dissertation is interested these divisions as well.  I endeavor to go further, however, by digging beneath the surface to reveal the terrain of attitudes that eventually produce the divisions identified in previous work.  Only by scrutinizing the attitudinal terrain underpinning intergroup divisions can we hope to fully explain why these divisions occur.  When we understand why, we improve our ability to predict attitudinal change.  For example, knowing that whites think there is less racial discrimination in Canada than non-whites is a helpful insight (e.g., Reitz and Banerjee 2007).  But, it does not tell us that the more serious trigger of intergroup hostility for whites is not physical differences, but the religious differences often implied in the concept of race and ethnicity (see Chapter Two here).   My goal is not to predict attitudinal changes here, but my results can inform others who assume that task.  I hope to offer a point of departure for future research into the consequences of ethnic diversity in Canada.  The Canadian literature is sorely in need of in- 4 depth analysis of the attitudes around this very salient political issue.  Indeed, Thompson (2008) warns that Canadian political science?s neglect of ethnic issues risks the discipline becoming out-of-touch with the society it purports to analyze.   The present dissertation endeavors to address this gap.  In a series of four papers, I look at various aspects of Canada?s ethnic diversity debate in an attempt to explain some of the fundamental determinants of citizens? attitudes.  As such, the dissertation focuses on the attitudinal aspect of the ethnic diversity issue, rather than normative goals or policy results.  It investigates how Canadian citizens function in their everyday lives, thus seeking to glimpse the fundamental terrain underpinning ethnic attitudes.  The four papers range from addressing the attitudinal determinants of comfort with intergroup contact, the marginal effects of individual and contextual factors of attitudes toward immigrant integration for whites and non-whites, and the interaction between political values and personality in overcoming negative reactions to ethnic stereotypes.  The articles draw on political science theories of behaviour, as well as social psychological and personality psychological theories of attitude and identity formation.  I employ recent public opinion data from various sources to test hypotheses with descriptive statistics and regression analysis.  I do not claim to fully explain the dense intertwining of factors that produce Canadians? attitudes about their fellow citizens.  I do, however, identify some of the essential building blocks of attitudes, taking snapshots of people?s opinions on issues central to the ethnic diversity debate as it plays out in Canada. Importantly, I am necessarily selective in the ethnic minority groups I study.  I focus on immigrant minority groups, or groups that have left another country for Canada at some point in their collective history.  I do not include Canada?s indigenous peoples in the analysis.  Investigating attitudes about indigenous peoples is a worthy line of inquiry, and a few studies  5 exist (e.g., Donakowski and Esses 1996; Wells and Berry 1992).  However, indigenous peoples have a considerably different political history, including different policy arenas, from Canada?s immigrant minorities.  As such, I leave the assessment of attitudes toward them for another time. I also omit any deep investigation into the attitudinal peculiarities of the Quebecois.  Attitudinal studies in Canada have regularly shown that French-speaking Quebeckers tend to hold different attitudes from the rest of the country about a host of political issues.  Researchers have attributed this difference to cultural insecurity.  In Quebec, this insecurity translates into Quebecois individuals and opinion leaders lashing out at other cultural groups because of a perceived existential precariousness of the Quebecois identity (e.g., Gidengil et al. 2004).  Cultural insecurity is related to the social psychological theory of social identity, which is a common theoretical framework in the dissertation.  In fact, it can apply to any cultural group that perceives others as a threat to their cultural identity.  To be sure, cultural insecurity does not necessarily entail the intense hostility sometimes associated with prejudice.  It can manifest in subtle ways, such as a desire for cultural continuity.  This desire, though, is related to the process identified in social identity theory that results in an individual categorizing herself and others into reductive groups.  The Quebec case entails another layer of complexity because the struggle for cultural protection has become highly politicized in Canada, including sovereigntist political parties successfully competing in Quebec elections and pushing cultural demands onto the political agenda at the federal level.  I acknowledge the unique dynamic in Quebec both theoretically and empirically in the following articles.  But, because of their complexity and practical space limits, I treat them as an analytic control, rather than as an additional attitudinal prong to investigate.  Moreover, any influence of  6 generalized cultural insecurity, for example, as felt by Anglophones outside of Quebec, will be captured under the analyses? various indicators of the social psychological conception of social identity. Each chapter in the dissertation examines, to some extent, the role of the ?primary ingredients? of attitudes (Kinder and Sanders 1996).  These primary ingredients ? social identity, material self-interest, and political values ? have been identified as the main components of attitudes toward race-related issues in the United States, and I adopt it as a useful framework here.  In addition, each chapter implements a distinct analytic approach.  These approaches, a sort-of ?value added?, are intended to enrich the analysis.  They also avoid a repetitive chapter layout, hopefully making for more interesting reading.  Chapter Two looks at the dynamics of social acceptance of minority groups.  It takes a step back from the ?contact vs. conflict? debate (e.g., Putnam 2007) and considers one of the ways people assess their willingness to engage in intergroup contact.  Using statements of comfort, the analysis compares white, mainline Christian Canadians? attitudes to six identifiable groups that are defined as either ethno-racially different, religiously different, or both.  It also compares how these statements of comfort change over different hypothetical settings with different levels of ?intimacy?, or social closeness.  The analysis goes beyond previous academic work that tends to focus on either one type of identifiable group, or one contact setting, at a time (important exceptions include Sniderman et al. 2000; Hainmueller and Hangartner 2013; Harell et al. 2012).  Contrasting statements of comfort with six groups across three settings provides a much more nuanced understanding of how Canadians perceive their neighbors.  The comparison between ethno-racial (or the conceptual leftovers of ?ethnicity? once the religious component has been omitted) and religious identifiable groups is  7 also novel.  It isolates the role of religious difference from ethno-racial difference in people?s assessment of comfort, rather than conflating these two overlapping, yet distinct, types of group difference.   The value-added aspect of the chapter is the inductive demonstration of outgroup rankings.  The analysis finds that people rank outgroups in a hierarchy of social acceptance, where religiously different groups have significantly lower social acceptance than ethno-racially different groups.  Cross-categorized groups, or groups that can be defined as both ethno-racially and religiously different, fall between the two extremes.  This snapshot of how Canadians fundamentally rank groups has not been done since the mid-1990s and, to my knowledge, has never included distinctly religiously-defined groups (e.g., Berry and Kalin 1995).2  I also find that respondents are less comfortable with all groups as the contact setting becomes more intimate.  They are the least comfortable with both religious and ethno-racial groups when the scenario relates to someone marrying into the family.   Across the three contact settings, the order of identifiable groups was found to be surprisingly stable: ethno-racial groups at the top, religious groups at the bottom, and cross-categorized groups in the middle.  This result spurred a second analysis to test the resilience of the hierarchy.  Namely, I wanted to see who was driving this hierarchy, who ranked groups differently, or who eschewed ranking at all.  I tested numerous determinants that commonly differentiate individuals in their ethnic attitudes.  Ethnic and religious self-identities were tested, as was material self-interest, education levels, gender, age, regional residency, and so on.  The remarkable finding was that the initial ranking was maintained for largely all of the comparisons.  The over-all levels of comfort changed, but the ethno-racially-defined groups                                                 2 Similar to my analysis, Berry and Kalin (1995) include Muslims and Jews in their analysis.  They also include Sikhs.  While they acknowledge that the perceptions of these groups likely have a religious component, they fail to theorize on the possible implications this distinction might have for people?s perceptions of them.  8 were consistently more accepted than the religiously-defined groups, with cross-categorized groups in the middle.  This result held true even for respondents who might be thought as more socially progressive, such as respondents with low religious or ethnic identities, high education, or young people.   The stability of the outgroup hierarchy across sub-groups and settings hints that there are specific categories of minorities in Canadian society that are far from gaining general acceptance from the majority.  These groups, who represent different religious values from the majority, are not typically addressed by the extensive multiculturalism policy framework that exists in Canada.  Rather, Canada?s multiculturalism policies, as well as the popular conception of the Canadian ethnic mosaic, tend to focus on groups that ?look different? rather than groups that ?believe different?.  The troubling policy implications of this finding are discussed in the chapter?s conclusion.  The third chapter investigates the influence of the primary ingredients ? social identity, material self-interest, and political values ? on individual attitudes about immigration and multiculturalism.  The ?value added? component of the analysis is my comparison of the effect of the primary ingredients on white and non-white attitudes.  The chapter, thus, attempts to draw together multiple research agendas into a single analysis as a way of assessing the relative influence of commonly identified determinants of ethnic attitudes.  Past work tends to focus on one or two determinants, which does not provide a clear picture of the relative role these determinants play in shaping the attitudinal terrain (exceptions include Dustmann and Preston 2007; Fetzer 2000; Kinder and Sanders 1996).  Moreover, few studies look at the attitudes of individuals from ethnic minority groups, and even fewer compare the attitudes of  9 these individuals with the attitudes of the ethnic majority (exceptions include Citrin et al. 2001; Oliver and Wong 2003).  The analysis finds that the influence of the primary ingredients greatly depends on whether the respondent identifies as white or non-white.  For whites, measures of outgroup resentment (especially concerns about outgroup societal power) have the strongest association with attitudes about immigration and multiculturalism.  This finding supports other work attesting to the strength of prejudice in underpinning ethnic and racial attitudes (e.g., Citrin et al. 2001; Kinder and Sanders 1996).  I also find that objective measures of material stakes have no association with the attitudes under scrutiny, which is in line with previous work downplaying material predictors (e.g., Sniderman et al. 2004). For non-whites, outgroup resentment matters, but in a different way.  They are more concerned about interethnic marriage than the societal power of whites.  Material concerns, whether objective or subjective, matters even less for non-whites than whites.  Education, on the other hand, results in supportive attitudes about immigration and multiculturalism for both ethnic groups.  The lesson here is that no assumptions should be made about the applicability of determinants: determinants that strongly influence one group do not necessarily influence the other group.  The analysis empirically demonstrates how minorities have different attitudinal contours compared to the majority.  Often, this difference is acknowledged in passing in attitudinal studies.  But, undoubtedly, there is still a considerable gap in empirical investigations of this difference.  My hope is that this chapter will bridge some of the gap, at least for attitudes in Canada.   Chapter Four is an extension of Chapter Three.  It takes the individual-level analysis of white and non-white attitudes about immigration and multiculturalism and adds contextual  10 measures.  As such, it directly takes on the ?contact vs. conflict? debate in social science that seeks to establish if interethnic contact increases or decreases prejudice (e.g., Pettigrew and Tropp 2006; Putnam 2007).  There is good reason why majority and minority group members will interact with their lived contexts in substantively different ways.  The chapter outlines a social psychological theory of how the formation of identity can be shaped by one?s environment.  The theory centres on concepts of subjective uncertainty, depersonalization, and self-affirmation.  It posits how the ethnic diversity in an ethnic minority?s locale can help increase positive individual identity while helping decrease hostility of others under specific conditions (e.g., Grieve and Hogg 1999).   For Canada, the process described by social psychology would lend much-needed empirical support for the political assumptions around multiculturalism.  Canadian multiculturalism policy, established in 1971 by Prime Minister Pierre Elliot Trudeau, is premised on the assumption that affirming minority cultural identity will result in confident ethnic minorities who are more willing to integrate into the wider Canadian experience.  The belief is that recognizing the ethnic minority ?part? will strengthen the ethnic majority ?whole?. My analysis suggests that the process described above exists in diverse Canadian neighborhoods.  I caution that the evidence is slight.  But, it does appear that, in more ethnically diverse areas, non-whites have higher group confidence (i.e., ?ingroup bias?) and lower wariness of the majority (i.e., ?outgroup resentment?).  This dynamic is precisely what multiculturalism advocates would want to see in the real world.  For whites, however, I find virtually no effect of context on attitudes about immigration and multiculturalism.  The story  11 about attitude formation for whites is really at the individual level, at least how concepts are measured here.  I discuss possible reasons for this in the chapter?s conclusion. The fifth chapter seeks to untangle the influence of political values, personality types, and stereotypes on people?s attitudes about ethnic minorities.  The chapter analyzes an experiment embedded in the 2011 Canadian Election Survey about ethnic difference and policy preferences.  The question motivating the chapter is about attitudes held by individuals who have personalities predisposed to ethnic tolerance.  Numerous psychological experiments find that personalities high in the Openness to Experience or Agreeableness dimensions are consistently more tolerant of ethnic difference than others.  My analysis, however, reveals that this insight can only partially explain attitudes toward Muslims identified in a survey experiment.  When a powerful symbol of Islamic difference (i.e., the hijab) is primed, the typical gap between high and low scores on these personality dimensions vanishes as these people with typically tolerant personalities increase their exclusionary attitudes.  The analysis shows how exposure to the hijab can drive these people to hold attitudes incongruent with their personalities. This result prompted a second investigation into the possible role political values have in overcoming exclusionary reactions to negative stereotypes.  Political values, as aspirational goals, are proposed as a likely source for individuals with tolerant personalities to control intolerant behaviour.  I look at liberal egalitarian values as this moderating mechanism, as its emphasis on equality should result in tolerance of individuals perceived as different.  My analysis shows that these values work: individuals with personalities high in Openness use liberal egalitarian values to override their negative reactions to the Muslim hijab.  The result demonstrates that Canadians, at least Canadians high in Openness, are not simply paying lip  12 service to these values.  Liberal egalitarianism actually moderates negative reactions to potent ethnic stereotypes.  My results have valuable political implications, since the promotion of political values is something of which policymakers are capable.  The result for individuals high in Agreeableness is different.  Liberal egalitarianism has a direct effect on their attitudes toward Muslims.  But, any initial influence of this personality dimension is washed out in the expanded model.  Possible reasons are discussed in the conclusion. The ?value added? component of the chapter is its demonstration of the psychological theory of the ?dual source? of attitudes (e.g., Devine 1989).  Research on the dual source posits that attitudes have an automatic component and a controlled component.  In my analysis, the automatic component is the respondent?s personality and the controlled component is her political values.  My analysis demonstrates how these two components work together to define multiple paths to ethnic tolerance.  When typically tolerant personalities fail to produce tolerant attitudes, liberal egalitarian values can correct, or ?control?, negative reactions.  Importantly, my results demonstrate the dual source theory in the real world context of a public opinion survey, which is an important methodological expansion on previous laboratory-based psychological experiments.  The present dissertation is not intended to provide a single definitive answer to the question of how Canadians arrive at their attitudes about ethnic diversity.  The chapters demonstrate clearly that there is no single answer.  Previous theoretical and empirical studies only offer shallow fragments of the large mosaic that is Canadian public opinion.  The chapters here endeavor to integrate these various fragments to some extent and to provide a richer analysis of attitudes than what has come before.  I hope to provide clearer snapshots of the Canadian ?attitudinal mosaic?, identifying sub-populations of citizens whose attitudes  13 about ethnic diversity, immigration, and multiculturalism are shaped by a network of factors.  In light of the chapters? findings, the concluding chapter will revisit the issue of the attitudinal mosaic and identify practical lessons that can be drawn from the preceding analysis.  Namely, I will address the question of what Canadians? think about ethnic diversity, and, more importantly, why they think it?      14 CHAPTER TWO RANKING MINORITIES IN CANADA:  THE ROLE OF PERCEIVED ETHNO-RACIAL AND RELIGIOUS DIFFERENCES  In recent social science research, considerable attention has been devoted to the effects of intergroup contact (e.g., Berry and Kalin 1982; Bobo 1999; Brewer and Brown 1998; Esses et al. 2001; Hewstone et al. 2002; Oliver and Wong 2003; Pettigrew and Tropp 2006; Putnam 2007; Stolle et al. 2008; Uslaner 2010).  The goal of this research agenda is to determine whether intergroup contact exacerbates or ameliorates prejudice and associated behaviour.  Results have been decidedly mixed, however.  The mixed results may be because little is known about what motivates people to initiate contact in the first place.  Intergroup contact is not randomly assigned; much of it comes from individuals? behavioural choices to make, or avoid, contact with groups perceived as different.  Numerous motivators have been suggested, most commonly in the social psychological research agenda.  Allport?s (1954) seminal work on prejudice, for example, notes that the outcomes of contact might vary according to contact frequency, duration, status, the number of individuals involved, whether the relationship was competitive or cooperative, the personality of the individuals, and so on.   This paper, accordingly, takes a step back from the contact research agenda to look at factors that may determine intergroup contact in the first place.  It finds that the perception of ethno-racial difference and religious difference matter and, importantly, matter at different magnitudes.  Specifically, people are considerably more wary of perceived religious difference (i.e., difference in belief, sacred texts, houses of worship, etc.) than perceived ethno-religious difference (i.e., difference in racialized physical attributes, collective histories,  15 etc.).  The analysis also finds that the setting of contact matters.  People do not form impressions of others in a vacuum.  The setting, specifically the intimacy, or social closeness, of the setting, influences people?s assessments of difference.   These findings are gleaned from people?s own statements of comfort with a number of identifiable minority groups in a number of different settings.  Statements of comfort do not directly lead to outgroup contact, positive or negative.  Moreover, they may be shaped by previous contact with outgroups.  But, statements of comfort do point to a greater social acceptance of the outgroup in question.  Whatever its origins, comfort with groups perceived as different is a critical signal of whether eventual intergroup contact will produce positive or negative attitudinal outcomes.  One of Allport?s four conditions of positive intergroup contact is that the groups must have equal status (ibid.).  Social acceptance, as revealed by feelings of comfort, improves the perceived status of the outgroup in question, which in turn helps produce positive intergroup contact if and when it occurs.3   Beyond the ?contact vs. conflict? debate, investigating how group categorization impacts social acceptance, as indicated with comfort statements, is important if we are to understand why some ethnic minority groups face subtle political or social obstacles, and even outright public hostility, more than others.  Clearly, understanding how social acceptance varies across groups would have consequences for policymakers interested in harmonious interethnic relations.  Being aware that certain types of groups face less social acceptance than others would help the effectiveness of policies meant to improve interethnic relations, such as ethnic outreach or anti-racism programs.  To this end, the paper broadly asks: with whom are                                                 3 The idea of social acceptance is different than the idea of social tolerance.  Tolerance implies putting up with difference, as opposed to accepting difference.  Given the empirical indicators used here are comfort ratings, ?social acceptance? is more appropriate.  16 Canadians comfortable?  In what settings are they the most comfortable? And, lastly, who is comfortable? The argument focuses on two factors that influence whether an individual is comfortable with others perceived as different: 1) the category of target group, that is, whether the target is perceived as ethno-racially or religiously different, and 2) the intimacy of contact, or the social closeness of the potential contact setting.  The first factor is part of a social psychological process where individuals subjectively aggregate themselves and others into ingroups and outgroups (e.g., Tajfel and Turner 1979).  Ethno-racial and religious differences are two typical outgroup categorizations. They often overlap, but are considered independent parameters of intergroup differentiation.  Moreover, they are central to the current political debate around ethnic diversity.  Concerns about, say, language acquisition, religious accommodations, or simply wariness of people who ?look different? fuel discussions about ethnic integration.  Ethno-racial and religious differences, which are constituent parts of diverse societies, likely influence individuals' willingness to engage in intergroup contact.    The second consideration ? intimacy of the contact setting ? is also expected to affect the willingness of individuals to engage in intergroup contact.  The setting where intergroup contact occurs is not neutral and should shape whether an individual tolerates interaction with an outgroup.  The potential for intergroup contact can occur in numerous different settings even in a single day.  A person?s willingness to have contact with someone perceived as part of an outgroup may change considerably over the course of a day, simply because the setting has changed.  Identifying the influence of a particular setting?s intimacy on an individual?s comfort with intergroup contact is an important part of understanding the formation of ethnic attitudes specifically, and intergroup attitudes more broadly.  17 The role of these two factors in shaping individual comfort with intergroup contact is investigated here in the Canadian context.  The factors are critical to the debate about ethnic diversity in most Western democracies.  However, being the first country to have an official policy of multiculturalism (1971), successive Canadian governments at the federal and provincial levels have invested substantial public resources toward the goal of managing the relations between disparate ethnic groups.   As such, understanding not only the effects of contact, but the determinants of contact, is critical in the Canadian setting.  The categorization of the outgroup and the intimacy of potential contact reflect fundamental questions that societies must answer when grappling with immigrant integration.  As Berry (2001) notes, these questions are: how much intergroup assimilation, and how much intergroup contact, is desirable?4  Canada?s multicultural policies advance a program of low assimilation and high contact.  The federal government?s stated purpose for multiculturalism is to ?[ensure] that all citizens can keep their identities, can take pride in their ancestry and have a sense of belonging?, as well as encourage citizens to ?integrate into their society and take an active part in its social, cultural, economic and political affairs.?5  As such, Canada's multicultural rhetoric and policy framework encourages the retention of ethnic difference and encourages contact between different ethnic groups.  The core motivation of Canadian multiculturalism policy, then, directly relates to the categorization of outgroups and the intimacy of the contact setting. Using a 2003 Centre for Research and Information on Canada (CRIC) public opinion survey, I use multivariate regression analysis to investigate the role of different diversities and                                                 4 Berry uses the term ?penetration? instead of ?contact?.  The term contact is used here because it fits with the political science research agenda on intergroup contact. 5 From the Citizenship and Immigration Canada departmental website.  Accessed February 20, 2013.   18 different settings when individuals assess their comfort with intergroup contact.  Similar to previous work on ethno-racial outgroups in Canada and elsewhere (Berry and Kalin 1995; Hagendoorn 1995; Kalin and Berry 1996; Parrillo and Donoghue 2005), the present analysis finds that there is a hierarchical ranking of outgroups.  An inductive assessment of the hierarchy suggests that outgroups best described as ?ethno-racial? are ranked high and outgroups best described as ?religious? are ranked low.  Groups best described as ?cross-categorized? are ranked in the middle.  Moreover, it discovers that attitudes about these groups improve as the intimacy of the setting decreases.   The analysis also finds that, in the aggregate, the outgroup hierarchy remains remarkably stable over the three contact settings.  The second part of the analysis, thus, tests the resilience of the outgroup hierarchy by looking more closely at common attitudinal motivators.  Do motivators linked to ethnic attitudes in the past have an impact on the outgroup hierarchy?  Do relevant self-identities, for example, change the order of outgroups?  Does material self-interest?  What about other common attitudinal motivators, such as age, gender, education, and geographic region?  Again, the analysis finds stability in the outgroup hierarchy: across groups of people that are often differentiated in their ethnic attitudes (i.e., high and low education, high and low ethnic or religious identity, young and old, etc.), the ordering of outgroups is largely the same.  This finding shows that there is a broad consensus, in Canada at least, about the acceptance of certain groups in terms of being comfortable with engagement.  The implication for the contact literature is that not only does an outgroup hierarchy exist, people from different walks of life tend to have the same rankings.  This attitudinal consensus is hidden in studies that look at only one or two outgroups, or one or two  19 settings, at a time.  Only in a study like the present one can the broad consensus about outgroup rankings be detected. The paper begins with a brief outline of the debate around intergroup contact, focusing on previous work identifying drivers of contact, of which social acceptance is one.  After this, it analyzes the relationship between the probability of contact, the type of target group, and contact intimacy.  Finally, it investigates the determinants of this relationship with multivariate regression analysis.  It concludes with discussion on contact and tolerance in Western societies. The Categorization of Outgroups The first factor that likely shapes an individual?s attitudinal willingness to engage in intergroup contact is how the outgroup is categorized.  The concept of outgroups is rooted in social psychology.  The discipline has long established that individuals categorize others into ingroups and outgroups under the thinnest of circumstances (e.g., Tajfel 1982; Tajfel and Turner 1979).  The divisions are based on stereotypes created by the perceiver about the members of her ingroup and members of outgroups (Hagendoorn 1995).  Intergroup competition is triggered when the perceiver favours her ingroup and endeavors to bolster its status at the expense of other groups.  This dynamic not only produces ingroup favouritism, but also outgroup wariness.  Past work has found that outgroups are not perceived with equal wariness, however.  Some groups are considered more different from the perceiver than others.  As such, the perceiver ranks outgroups hierarchically according to various criteria, including ethno-racial difference and religious difference (Pineo 1977; Berry et al. 1977; Berry and Kalin 1995; Hagendoorn 1995; Triandis and Davis 1965).   As will be demonstrated, perceived ethno-racial and religious differences are used by  20 Canadians to assess levels of comfort around outgroups.  Ethno-racial difference refers to immutable differences that a person would be able to detect during a brief interaction, and on which there is a social consensus (Jackson et al. 1995).  This can be, for example, differences in skin colour or language.  Religious differences, on the other hand, are more mutable as beliefs about right and wrong can change.  Moreover, they cannot be detected by simply meeting a person.  More information is needed about the target for religious differences to be assessed, even if this information is a label that has been stereotyped by the perceiver (e.g., the target is known to be a Buddhist).   How might potential differences in treatment of ethno-racial and religious groups be identified?  The overlap between the two outgroup parameters, both theoretically and in popular conception, complicates the study of this question.  It has been empirically established that ethnicity, in terms of physical difference or ancestry, and religiosity are two distinct categories people use when they form impressions of others (e.g., Hagendoorn 1995).  Much of the research in this area of outgroup categorization, however, conflates the two.  Indeed, Horowitz?s classic study of ethnic conflict views ethnic groups as having the myth of a common ancestry and ?differentiated by color, language, and religion: it covers ?tribes,? ?races,? ?nationalities,? and castes.? (1985, 53).  This broad definition that leaves ancestry and religiosity undifferentiated is not improved through operationalization.  Burton et al. (2010) note that past measures of ethnicity have used no less than eight conceptual dimensions, including race, national identity, nationality, ancestry, language, country of birth, religion, and culture.  The typical approach to conceptualizing ethnicity, then, does not suit my purposes and so I diverge from it.  Reflecting the position that they are conceptually and practically different, I separate out the religious dimension of perceived difference from the broader  21 dimension of ethnicity.  This enables a comparison between the impact of ethno-racial cues (e.g., a target who is Asian) and religious cues (e.g., a target who is fundamentalist Christian) on an individual?s comfort with the idea of contact with an outgroup. It is unclear how these two cues might separately impact the probability of intergroup contact.  Past research has typically considered only one of these factors (usually ethnic or national differences), which does not allow a comparison (e.g., Hainmueller and Hangartner 2013; Oliver and Wong 2003; Pettigrew and Tropp 2006; Putnam 2007; Sigelman and Welch 1993; Sniderman et al. 2000; Wainryb et al. 2001).  One study that does compare religious and ethno-racial differences on an American college campus finds that students seem to be more tolerant of the latter, rather than the former (Haidt et al. 2003).  Another study finds that, in the workplace, ?value diversity? is correlated with more conflict amongst coworkers than ?social category diversity? (Jehn et al. 1999).  Empirically, then, there is evidence that religiosity, compared to racialized ethnicy, has a stronger impact on perceptions of outgroup difference.  Theoretically, this may be because religion has more significance for an individual?s sense of identity.  Religious beliefs strongly motivate behaviour, since they are idealized prescriptions of what ought to happen and closely inform people?s fundamental perception of themselves as good (e.g., Rokeach 1979; Skitka and Mullen 2002).  Racialized ethnic identity (or, what conceptually remains of ethnicity once religion is omitted) is certainly important to many people, and ethno-racial difference has been shown to motivate a range of behaviour and attitudes (e.g., Arends-T?th and Van De Vijver 2003; Citrin et al. 1990; Kalkan et al. 2009; Kinder and Sanders 1996; Kinder and Sears 1981).  The social psychological theory of social identity reveals how racialized ethnic categories can be an integral part of an individual?s perception of herself and her place in the world (e.g.,  22 Verkuyten 2005).  So, ethno-racial classification is assumed here to shape individuals? attitudes about intergroup contact.  But, does it shape attitudes more than religious classification?  At the end of their study of ethnic prejudice in Italy, Sniderman et al. speculate that religion may be an even deeper divide than the forms of difference they examined: religion stand ?as a marker not merely of different beliefs but of opposing ways of life? (2000, p. 140).  So, I anticipate that religious difference, compared to ethno-racial difference, will result in greater reluctance to engage in intergroup contact.  The Intimacy of the Contact Setting The second factor that shapes an individual's willingness to engage in contact with an outgroup is the social closeness, or intimacy, of the contact setting.  Willingness to interact with someone perceived as part of an outgroup is likely to differ if this interaction takes place at the grocery store as opposed to the home, workplace, or other setting (e.g., Mutz and Mondak 2006; Parrillo and Donoghue 2005).  Preferences for outgroup contact have been shown to vary across settings, though many studies of the political effects of contact focus on relatively large geographic areas, such as neighborhoods, electoral wards, cities, and states (e.g., Dustmann and Preston 2001; Ha 2010; Hood and Morris 1998; Patchen et al. 1980; Putnam 2007; Stolle et al. 2008).  One study of smaller settings, where intimacy is more clearly a factor, showed that anti-black attitudes were higher when subjects saw depictions of Blacks in a gang-related scenario (versus a family barbecue) and on a street corner (versus in a church) (Wittenbrink et al. 2001).  Another showed similar negative outgroup attitudes when a prison cell (versus a church) and a foggy street (versus a sunny garden) were cued (Maddux et al. 2005).   These experimental studies support the idea that settings that cue negative stereotypes  23 (e.g., street corners, prisons) or feelings of fear (e.g., foggy street) shape negative attitudes about ethno-racial and religious outgroups.  Further, they suggest that a more intimate setting will trigger negativity and fear, hindering positive intergroup contact and even eliminating intergroup contact all together.  Other studies, however, have found how the intimacy of friendships with outgroups contribute to more positive contact (e.g., Ellison and Powers 1994).  Oliner and Oliner (1988), for example, find that many non-Jews who rescued Jews during World War II cited prewar friendships with Jews for their motivation.  The anti-Semitic propaganda in the years just prior to the War increased the perception that Jews were different and, consequently, heightened wariness and hostility toward them.  Still, many non-Jews with pre-war Jewish friends risked their lives to rescue Jewish strangers.  This action illustrates how positive contact can mitigate negative stereotyping.  It also illustrates the importance of the voluntary nature of the contact, identified by Allport (1954).  The voluntary nature of contact gives the individual a sense of control, contributing to positive attitudinal outcomes.  As such, highly intimate settings need not produce negativity and fear if the individual is engaging willingly with the outgroup member. Other studies have taken this idea of setting intimacy further by comparing reactions to different outgroups in different settings.  One study of college students finds that they tend to be more tolerant of outgroups in less intimate situations (e.g., dorm room versus campus), regardless of outgroup difference type (Haidt et al. 2003).  Another finds that subjects? race-related concerns become increasingly important, and belief-related concerns become decreasingly important, as the intimacy of the contact setting increases (Triandis and Davis 1965).  And yet another showed no difference in ethnic or belief intolerance levels for public or private situations (Rokeach and Mezei 1966).  So, it is reasonable to expect the intimacy of  24 the setting to impact an individual?s willingness to engage in intergroup contact.  However, the pattern of this impact is far from well-understood. Data and Methods To test the effects of 1) independent parameters of outgroup difference and 2) setting intimacy on Canadians' comfort with intergroup contact, I draw on a 2003 survey conducted by the Centre for Research and Information on Canada (CRIC).  The survey was commissioned for publication in a series entitled ?The New Canada? in the Canadian newspaper, the Globe and Mail.6   Dependent variable The dependent variable ? comfort ? is from a survey question that asks how comfortable the respondent would be with members from particular outgroups in three different scenarios.  As such, it taps both the effect of different outgroup categorizations and the effect of different setting intimacies.  The use of comfort ratings has been used as an indicator of intergroup tolerance in previous work in the Canadian setting (Berry and Kalin 1995; Kalin and Berry 1996).  The groups used in the current analysis are Muslims, Asians, Blacks, Jews, fundamentalist Christians and Atheists.7  These outgroups were randomized in the survey, so any patterns observed are not due to respondent fatigue where the respondent becomes less attentive for later questions.  The three roles used to test for setting intimacy are a boss at the workplace, a teacher at a local school, and a person marrying a close relative.  (See Appendix One for the question wording and descriptive statistics of all items included in the analysis.)                                                  6 The survey is a national sample of 2000 randomly selected Canadians, interviewed by telephone between April 21 and May 4, 2003.  Interviews were conducted in English and French, which may bias the immigrant sample to respondents who are linguistically integrated into Canadian society.  This potential bias is not a serious problem, however, since the immigrant sub-sample is small and none of the analysis relies on immigrant respondents. 7 The question also included Aboriginal, white supremacist, and French or English Canadian categories (French and English Canadian were rotated depending on the language of the interview).  They have been excluded from the current analysis.    25 The survey sample was split into thirds, with each third being asked about all of the groups, but for only one of the roles.  Thus, one respondent was asked how comfortable she would be if her boss was a Muslim, an Asian, and so on.  A different respondent was asked about her comfort if a Muslim, an Asian, and so on, were teaching at a local school.  And the third was asked about a Muslim, an Asian, and so on, marrying into the family.  Because the survey sample was split into thirds, the present analysis is essentially using three different samples.  This should strengthen any conclusions drawn from the findings. The survey question has a double function.  The identifiable groups tap respondents? attitudes about outgroups, and the three roles tap respondents? attitudes about outgroup contact in various settings of intimacy.  Therefore, differences in estimates in the regression model will help identify how different outgroup categories and different contact settings shape people?s willingness to engage in intergroup contact.   Before delving into the determinants of outgroup perceptions and why these different perceptions may exist, I first test to see if outgroup ranking actually occurs.  As mentioned, outgroup ranking has been identified historically in Canada (Berry et al. 1977; Berry and Kalin 1995) and elsewhere (Hagendoorn 1995; Verkuyten and Kinket  2000).  The question here is whether it still occurs, and, if it does, what might it look like?  To address this initial question, I estimate the direct relationship between outgroups, settings, and the level of comfort with ordered logistic regression (not shown).  No other variables are added at this time.  The sample has been restricted to white, mainline Christian Canadians as a way to isolate perceptions of outgroup difference.  That is, respondents implicated in the six outgroups at issue here have been dropped from the analysis.  So, visible minorities and Jews are excluded, using a self-reported ethnicity variable.  Also, Atheists, Agnostics, and  26 fundamentalist Christians are excluded, using a self-reported religious affiliation variable.8   The survey does not specifically identify Muslim respondents, but the ?other? categories in the ethnic identity and religious affiliation indicators should capture many of them.  So, these categories have been dropped as well.  Consequently, about half the original sample is lost, but considerable analytic clarity is gained.  To be sure, there is great value in analyzing the attitudes of minority ethno-racial or religious groups.  However, the survey questions in the Project Canada data are biased toward the majority group experience.  For example, respondents are not asked if they are comfortable with a white person as a boss, teacher, or in-law.  Also, like many other surveys, the number of cases representing minority respondents is unacceptably small for any meaningful comparative analysis.  So, it is a practical necessity to restrict the sample to white, mainline Christians. Figure 2.1a-c illustrates the ranking of outgroups from simulations predicting ?very comfortable? (comfort=4) responses with the particular outgroup across the three hypothetical scenarios.  The point estimates have been arranged according to the level of comfort for each scenario, rather than any a priori expectation of outgroup rank.  The analysis, then, is inductive.  Indeed, arranging the outgroups in order of comfort levels clearly highlights a pattern.  With the exception of Muslims and Atheists in the in-law scenario (Figure 2.1c), the order of the outgroups places Blacks and Asians at the top, Atheists and fundamentalist Christians at the bottom, and Jews and Muslims in the middle.  This order is observed across the three settings, with the previously noted exception with outgroup in-laws.  Recall that the ordering of groups is randomized in the survey and that the respondents are split into thirds                                                 8 Identifying fundamentalist Christians is tricky, since there is no specific category for them.  Moreover, they are not confined to any one denomination.  So, denominations that can be both fundamentalist and mainline, such as Baptist or Presbyterian, are retained in the sample as mainline Christians.  Instead, I follow the observation that fundamentalists often eschew specific denominational labels and, so, drop respondents who identify as ?Christian?, ?Other Christian?, or ?Protestant? (Campbell 2006).  27 and asked about only one scenario.  Discovering that the outgroup order holds despite these constraints is suggestive that the ranking is not an artifact of measurement and, instead, exists in the perceptions of white, mainstream Christian Canadians.  Moreover, the survey was conducted by telephone, which can trigger social desirability bias (e.g., Nederhof 1985).  However, given the significant differences in comfort levels ? that is, respondents were willing to admit discomfort with some groups ? social desirability bias is not a concern here. The important question to address now is what can account for this ranking?  What motivates respondents to place Blacks and Asians at the top, Atheists and fundamentalist Christians at the bottom, and Jews and Muslims in the middle?  Here, I reintroduce the idea of independent parameters of outgroup difference.  As mentioned, past research on outgroup hierarchies have established that rankings are generated from stereotypes held by the perceiver (e.g. Hagendoorn 1995).  The more dissimilar the outgroup stereotype is from the perceiver?s own identity, the lower the outgroup is placed on the hierarchy.  It has also been established that individuals use different parameters to categorize groups (Berry and Kalin 1995).  Of particular interest to me are the parameters of ethno-racial difference and religious difference.  These two parameters are typically conflated in the literature, but they are central to Canada?s debate around ethnic minority integration.   Observing the repeated order of outgroups in Figure 2.1a-c suggests these two  parameters are at play when white, mainline Christian Canadians are contemplating their comfort levels.  Blacks and Asians ranked together (yet separate from Atheists and fundamentalist Christians) invokes the idea of an ethno-racial parameter.  And, Atheists and fundamentalist Christians ranked together (yet separate from Blacks and Asians) invokes the idea of a religious parameter.  These are natural categories to explain the stable ranking of  28 outgroups in Figure 2.1, especially given the randomization of outgroups and the three separate conditions in the survey.  What explains the mid-ranking of Jews and Muslims?  Jews and Muslims are likely what are called ?cross-categorized? outgroups, or outgroups that are defined by more than one salient parameter of outgroup difference (e.g., Crisp and Hewstone 2007).  For instance, even though many Muslims are ?white?, recent international tensions, such as the 9/11 terrorist attacks in the US and the subsequent war in Iraq, have largely focused on Muslims in the Middle East.  As such, Muslims are likely stereotyped as ethno-racially distinct (e.g., differences in physical appearance, dress, language, collective history, etc.).  However, Muslims are also likely stereotyped as religiously distinct.  In Canada, religious differences between Muslims and non-Muslims have become salient with several recent high-profile public controversies in Canada.9  The international tensions between the West and the Muslim world would have also contributed to a perception of considerable religious distance. The Jewish community is also plausibly cross-categorized as both ethno-racially and religiously distinct by white, mainline Christian Canadians.  The Jewish community is defined by its religion, which would inform any sense of religious distance for non-Jews.  The group may also trigger perceptions of ethno-racial difference, given certain physical stereotypes and a distinct and well-known collective history (e.g., Hoedl 2000).   It is clear, though, that Canadians are more comfortable with Jews than Muslims.  So, there is a ranking within the cross-categorized group itself.  Two processes are likely at work here.  First, negative Jewish stereotypes may be somewhat tempered in the current social and                                                 9 Examples include the 2005 debate over Sharia arbitration in Ontario and the 2008 Bouchard-Taylor Commission on Reasonable Accommodation in Quebec.  See Natalie Alcoba, ?Ontario rejects sharia law,? National Post. September 12 2005, A1. Bill Curry. ?Forum dredges up Quebeckers' ire toward minorities.? Globe and Mail. Retrieved March 1, 2012.   29 political environment, since they have been targeted in campaigns fighting anti-Semitism as exaggerations of the Jewish community's distinctiveness.  Second, Muslims stereotypes are particularly negative in the current political climate with the terrorist attacks on 9/11 and the subsequent ?War on Terror?.  These two processes have likely resulted in upward pressure on the Jewish ranking and downward pressure on the Muslim ranking. Taken together, it is clear that white, mainline Christian Canadian respondents are more comfortable with ethno-racial outgroups than religious outgroups.  They state that differences in religious belief, or at least perceived differences in religious belief, are quite worrisome.  The mid-ranking of Jews and Muslims, as well as the lower ranking of Muslims compared to Jews, reinforces this contention: the presence of perceived religious difference drags down outgroup rankings.  The tight clustering of point estimates around Atheists? and fundamentalist Christians? vertical space is additional evidence that religious difference triggers disproportionate prejudice.  When it comes to Atheists and fundamentalist Christians, tt matters little where the hypothetical intergroup contact is made.  Religious difference, when it is clearly the most salient parameter of difference, is simply off-putting.   The ranking of Jews and Muslims also fits with previous theories of outgroup difference that argues parameters of difference are interactive (e.g., Macrae et al. 1995).  That is, the weighting of one parameter depends on the weighting of another.  Here, for cross-categorized groups, the religious parameter seems to exert a disproportionately negative pressure on the ranking of the outgroup ? hence, the low ranking of Jews and Muslims compared with Blacks and Asians.  The relatively low ranking of Muslims compared to Jews also points to this idea, if the preponderance of negative Muslim stereotypes compared to negative Jewish stereotypes in popular culture has embedded itself into Canadians? comfort  30 levels.  The parameter of ethno-racial difference, however, seems to compensate for the parameter of religious difference, resulting in the mid-level ranking of Jews and Muslims.  This is in contrast with some literature that argues parameters of outgroup difference are additive ? that the perceiver adds up salient parameters of outgroup difference to determine rankings (e.g., Crisp and Hewstone 2007; van Oudenhoven et al 2000; Urban and Miller 1998).  If the additive model is at play here, Muslims and Jews would be ranked at the bottom of the hierarchy.  They are not, however.   What might explain the compensating mechanism that ranks Jews and Muslims in the middle rather than at the bottom?  It might be due to Canada?s political context that emphasizes the normative good of ethno-racial difference, or, at least, tolerance of that difference.  The fundamental premise and enduring message of Canada?s multiculturalism policy has been that ethno-racial difference is a boon to the country?s economics, politics, and most importantly, cultural identity.  Indeed, there is a broad consensus about the desirability of ethno-racial diversity amongst the public and the political elite.  Conversely, religious difference is rarely explicitly acknowledged as a legitimate part Canada?s official cultural mosaic.  Two former officials in charge of Canada?s federal multiculturalism program mused that religion is the diversity that ?dares not speak its name? (Biles and Ibrahim 2005), hinting that policymakers are reticent to directly deal with religious difference.  These two dynamics, that ethno-racial difference has been celebrated and religious difference has been relatively ignored, explains why ethno-racial difference compensates for religious difference.  The rankings in Figure 2.1a-c might change in countries where the government-driven political messaging about ethnic diversity is different.   31 What about comfort differences across the three scenarios?  How is comfort influenced by the idea of an outgroup member being a boss, a teacher, or an in-law?  Intimacy of the setting is thought to influence people?s social acceptance of the outgroup.  Though, as mentioned, the precise influence of intimacy is debated (e.g., Haidt et al. 2003).  It is clear here that the in-law scenario triggers the least tolerance difference, regardless of the group categorization.  For example, the gap between predicted comfort levels for blacks is about .15 (.46 for the teacher scenario and .31 in the in-law scenario).  And, in the in-law scenario, the ordering of Muslims and Atheists flip, with Atheists being preferred to Muslims.  I have no explanation for this, except to speculate that the perceived religious difference of Muslims is more threatening in a family setting than the perceived religious difference of Atheists (or the threat of non-belief).  Regardless, the aversion to difference in the in-law scenario makes sense, given that interethnic marriage has been called the ?litmus test? of assimilation and the ?last taboo? (Alba and Nee, 2005: 90; Qian 2005).  Moreover, an individual has little control over their in-laws.  Thus, the observation that voluntary contact can produce positive attitudes in intimate settings likely does not apply (e.g., Oliner and Oliner 1988).   It appears that the ?teacher at a local school? scenario triggers the most social acceptance of outgroup difference.  For the most part, respondents are relatively comfortable with outgroups as teachers at local schools.  This may be surprising given the perceived vulnerability of children to influence.  Indeed, the recent ban on Muslim headscarves in French schools for students and teachers exemplifies this concern.  The influence of intimacy may change if the respondent has school-aged children, something that cannot be tested here.   Still, the idea of a teacher belonging to an outgroup at a local school may still be quite abstract, even if the respondent has children.  There is no guarantee that the respondent?s  32 children go to a local school, nor is there a guarantee than children who do go to a local school will be taught by the hypothetical teacher.  Moreover, the survey question on which comfort is based asks about ?your boss?, which cues the respondent?s personal experience more than the phrase ?teacher at a local school.?   The previous investigation clearly illustrates that white, mainline Christian Canadians rank outgroups in a way that exposes the analytic importance of considering the difference between types of outgroups and types of scenarios.  But, what are the individual-level factors that motivate people to rank outgroups in this order, and what might change the order?  Who might eschew the rankings all together?  The second investigation adds common attitudinal determinants that have been found to influence political attitudes, particularly around the issue of ethnic diversity.  Here, the motivation of the analysis is to see how stable the outgroup ranking in Figure 2.1a-c is.  That is, how many groups of people adhere to ranking Blacks and Asians at the top, Jews and Muslims in the middle, and Atheists and fundamentalist Christians at the bottom?   I have stacked the original CRIC dataset so that the dependent variable, comfort, indicates the respondent?s comfort rating for a particular outgroup in a particular setting.  The unit of analysis in the original data is the respondent.  After stacking the data, the unit of analysis is a single answer to the comfort survey question.  Since respondents are asked about all six outgroups for only one setting, the respondent appears in the data six separate times.  Recalling that the sample used here is restricted to white, mainline Christians, the number of cases used in the analysis is 5880, though the effective sample size is smaller since standard errors are corrected for ?clustering? on the individual.  This technique is used to recognize that, since respondents are in the dataset six times, the observations (i.e., their answers) are not  33 independent of each other.  The six outgroups and the three contact settings are recoded as dummies, interacted, and added as controls to the model.  Blacks serve as the reference group for the outgroups and the in-law scenario is the reference group for the contact settings.  Independent variables The measures used here to investigate the determinants of comfort levels are individual-level factors commonly identified as attitudinal drivers.  The first two independent variables relate to the social psychological concept of social identity.  The previous discussion of how individuals categorize others into ingroups and outgroups make it clear that quantitative measures of an individual?s self-identity are likely central to any explanation of outgroup hierarchies (Turner 1999).  How an individual perceives herself greatly determines how she organizes others as more or less different from herself.  The CRIC data has two measures that capture the relative importance of ethnic and religious identity to the respondent?s sense of self.  These questions should reflect concerns about outgroups that are ethno-racially and religiously different, which in turn, will shape the respondent?s stated comfort with engaging with these outgroups.  The measures come from a survey question asking about factors that can contribute to ?one's personal feeling of identity?, including ?ethnicity and race? and ?religion?.  The respondent was asked separately if these factors were important or not to their own identity.  The resulting indicators ? ethnic id and religious id ? are ordinal with four possible outcomes that run from ?very important? to ?not at all important?.  As observed above, individuals weigh perceived religious difference more than perceived ethno-racial difference.  So, I anticipate that the religious identity variable will have a greater impact than the ethnic identity variable on comfort ratings.  34 Material self-interest is also a common determinant of individual attitudes.  In terms of intergroup attitudes, material self-interest has been associated with attitudes that, on the face of it, appear prejudiced.  Attitudes appearing to negatively target a group defined by particular characteristic, like ethnicity, can actually be driven by material uncertainty and the perception of intergroup competition in the economic realm (Blake 2003; Bobo 1983; Kluegel and Smith 1983; Mayda 2006; Palmer 1996; Schissel et al. 1989).  A white person?s negative attitudes against Blacks, for example, may be driven by a concern that civil rights pose a challenge to her material advantage in society.  Or, anti-immigrant attitudes may be fueled by a perception that newcomers increase job competition or strain the welfare system.  To capture this idea, I add an objective measure of material uncertainty: the respondent?s income (income).  If material uncertainty is related to how individuals perceive ethno-racially and religiously different outgroups, a lower household income should contribute to lower comfort ratings of outgroups in general. The respondent?s age, sex, and education are also added.  Younger people, females, and the better-educated tend to be more liberal in their political and social views (Blake 2003; Gidengil et al. 2005; Whitley 1999).  So, they are likely going to express more comfort with the outgroups identified here.  Whether the respondent lives in Quebec and whether they live in an urban area are included in the model.  Living in Quebec should influence how one views outgroups.  Previous studies have identified that living in Quebec is associated with more negative outgroup attitudes, due to province?s fragile cultural situation (Berry and Kalin 1995) or its higher levels of anti-Semitism (Sniderman et al. 1993).  Living in an urban rather than a rural area is thought to have an impact on outgroup attitudes, though the precise impact is debated.  The impact depends on whether intergroup contact is thought to produce positive or  35 negative intergroup relations, a question reflected in the ?contact vs conflict? debate (e.g., Putnam 2007).  A recent meta-analysis, however, finds that the majority of evidence point to intergroup contact producing positive relations, though scenarios do exist where it has the opposite effect (Pettigrew and Tropp 2006).  As such, living in an urban centre where intergroup contact is more likely should be associated with more intergroup comfort. Results Table 2.1 displays the ordered logit regression coefficient testing the possible determinants of comfort with intergroup contact.  All the independent variables have been interacted with outgroup type, since I expect that there will be variation depending on the outgroup.  The independent variables in the model can illustrate how variance along these determinants influences the outgroup hierarchy.  The multiple interactions in Table 2.1 render it difficult to interpret.  There are some direct effects that suggest at least some of the factors thought to influence comfort of outgroups are associated with the comfort ratings measured by the dependent variable.  For example, the respondent?s education and age are consistently associated with comfort ratings: higher education and relative youth is associated with higher odds of being comfortable with outgroups.  Factors like living in an urban area and living outside of Quebec are also associated with comfort ratings, but with less consistency.  The associations of other factors, like religious identification, ethnic identification, and income are not significant in the table.  But, predicted probabilities can help identify any subtle associations. Figures 2.2a-2.2h show the predicted effect of low and high levels of the individual-level motivators in the model.  The focus here is on the gap between these groups of respondents.  The probabilities again predict ?very comfortable? responses about the ethno- 36 racial and religious outgroups as in-laws.  As hinted at in the table, some of the motivators fail to distinguish the comfort ratings between the respondents at the low and high levels.  Not surprisingly, for example, respondents with low and high religious self-identity are statistically indistinguishable for their comfort ratings of Blacks and Asians (see Figure 2.2a).  The extent of their religious identity should not figure into their perceptions of ethno-racial outgroups, at least how these groups are defined here.  It is a surprise, however, how low and high religious self-identifiers rank cross-categorized outgroups.  Namely, low and high religious self-identifiers are indistinguishable when it comes to Jews, but not Muslims.  High religious identifiers rank Muslims lower than low religious identifiers, but these two groups of respondents rank Jews the same.  I did expect that respondents, in general, are more comfortable with Jews compared to Muslims.  But I did not expect that high religious identifiers would be as comfortable with Jews as low religious identifiers, especially since their high religious identity figures into the ranking of Muslims (and Atheists and fundamentalist Christians, for that matter).   Another unexpected result is that respondents with low and high ethnic self-identities are indistinguishable across the board, even for groups who are ethno-racially defined (Figure 2.2b).  This might be due to the conceptual softness of ethnicity as compared to religion.  People may have a more vague idea of their ethnic self-identity and so any association between it and reported comfort levels is more tenuous.  Still, to see no gap between low and high ethnic identifiers is a surprise.  Similar to my expectations for religious identity, social psychological theories of self-identity would suggest that strong ethnic identity would influence this type of intergroup attitude.  37 Education operates as expected, with a large gap between high and low educated respondents (Figure 2.3c).  The largest gap is for comfort with Jews, and the smallest is for comfort with fundamentalist Christians.  Similarly, urban dwellers and younger respondents tend to have higher comfort ratings than rural dwellers and older respondents (Figures 2.3e and 2.3f, respectively).  The gaps between these groups are not as large as the gaps observed with education.  But, there is still considerable divergence in the comfort ratings for Jews in the urban/rural predictions and Blacks, Muslims, and Atheists in the young/old predictions.   Having a higher income tends to increase predicted comfort ratings, but only for some groups ? Blacks, Muslims, and Atheists, in particular (Figure 2.3d).  So, material self-interest as measured here influences perceptions of difference, but it is hardly a dominant force.  Living in Quebec is associated with both higher and lower comfort ratings, depending on the group.  Compared to respondents in the rest of Canada, respondents in Quebec are less likely to be comfortable with Jewish, Muslim, and fundamentalist Christian in-laws.  Conversely, they are more likely to be comfortable with Atheists.  As mentioned, studies have found that the Quebecois tend to have more negative attitudes about other minorities because of the province?s distinct status claim (Berry and Kalin 1995) and the high value it places on cultural conformity (Sniderman et al. 1993).  Figure 2.3h only partially reflects this observation, since respondents from Quebec rank Blacks and Asians the same as respondents living outside Quebec.  Still, the relative discomfort with Jews, Muslims, and fundamentalist Christians supports the idea of higher hostility toward outgroups in Quebec.  It also perhaps reflects the provinces tumultuous history with organized religion (Meadwell 1993).  The greater comfort with Atheists reinforces this suggestion.  Finally, the sex of the respondent has little impact on  38 predictions (Figure 2.3g).  Any argument that women tend to be more progressive in their views about ethnic minorities gains no support here (e.g., Whitley 1999).  While the table and figures show how the individual determinants influence comfort ratings, the more illuminating lesson here is the stability of the outgroup hierarchy.  Here, I am focusing more on differences between outgroup categories rather than the differences between low and high values of determinants.  Regardless of the individual-level motivator, whether it drives individuals to be more or less comfortable, the ordering of the outgroups largely remains the same.  This is the case even for groups of respondents who are expected to be more tolerant of ethno-racial and religious difference.   To delve more deeply into this observation, Figures 2.3a-2.3h show the statistical differences between the point estimates for the sub-populations displayed in Figures 2.2a-2.2h.  That is, the point estimates in the current figures (2.3) are showing whether the outgroup rankings that appear for, say, high ethnic identifiers or urban dwellers in Figures 2.2 are a statistical reality.  Each estimate in Figure 2.3 is the statistical difference between two paired outgroups.  The groups being compared are listed on the x-axis and their difference (or lack of difference) is on the y-axis.  So, for example, the point estimate in the upper left corner of the religious identification box (Figure 2.3a) compares the predicted probability that low religious self-identifiers feel very comfortable with Blacks with the probability they feel very comfortable with fundamentalist Christians.  The difference is large and statistically significant (.3, p<0.000), reflecting the clear preference for Blacks over fundamentalist Christians held by low religious identifiers shown in Figure 2.2a.  Next, the estimate for low religious self-identifiers compares the different comfort ratings for Asians and Blacks (-.02, not significant), and then for Jews and Asians (-.08, p<0.05), and so on.  An outgroup  39 hierarchy, then, can be detected when the comfort ratings are statistically different between paired groups (or, different from the zero on the y-axis).  If the paired groups are statistically indistinguishable, the respondent ranks the outgroups the same and the point estimate sits at zerio.   For the most part, the outgroup hierarchies of the sub-populations are the same as the aggregate hierarchy illustrated in Figure 2.1: respondents at low and high ends of the independent variables tend to be most comfortable with Blacks as in-laws, then Asians, Jews, Muslims, Atheists, and fundamentalist Christian.  This observation supports previous Canadian work finding that outgroup hierarchies are ?consensual?, or similar across different groups (e.g., Berry and Kalin 1996).  Again, the replication of the hierarchy is surprising because it appears even amongst sub-populations that should be more tolerant of outgroup difference.  There is no sample of ostensibly progressive individuals who has a flat ranking, that is, where all six outgroups are rated equally.  Perhaps the closest are highly educated respondents, who rank Blacks, Asians, and Jews the same (Figure 2.3c).  Still, these individuals, who are typically more tolerant (e.g, Hainmueller and Hiscox 2007), place Muslims, Atheists, and fundamentalist Christians at a lower level.   Another notable result is that people who claim to have low religious or low ethnic identities still rank outgroups largely in a manner that reflect the outgroup?s religious and ethno-racial differences (Figure 2.3a-b).  So, while they explicitly state that their ethnicity and religion are not important, their comfort ratings hint at an implicit influence of these categories of outgroup difference on their perceptions of outgroups in general.   There are a few deviations from the general hierarchical pattern of outgroup rankings that deserve special note.  For example, respondents with low and high religious identities  40 reverse their ranking of Atheists and fundamentalist Christians (Figure 2.3a).  So, while their rankings of Jews and Muslims are perplexing, the rankings of overtly religiously-defined groups are clearly influenced by the strength of their religious identities.  As mentioned, respondents living in Quebec also are much more comfortable with Atheists than Muslims and fundamentalist Christians (Figure 2.3h).  Another example is that the oldest respondents rank Jews the highest in the hierarchy, then Asians and Blacks (Figure 2.3f).  This is likely a generational effect, reflecting the radically different political and social environments that older respondents grew up in.  In general, though, the ranking of ethno-racial and religious outgroups illustrated in Figure 2.1 is retained, with ethno-racial outgroups having higher comfort ratings than religious outgroups and crossed-categorized groups in the middle.   Discussion The main goal of this paper was to examine the effects of perceptions of ethno-racial and religious difference and the intimacy of the contact setting on an individual's comfort with intergroup contact.  In the Canadian literature, there has been very little attention devoted of the fundamental way citizens perceive and differentiate between outgroups.  There has been even less attention to the consequences of citizens categorizing outgroups as either ethno-racially or religiously different.10  By looking at who is comfortable with whom and where, my analysis identifies one factor that people likely use when assessing the possibility of making actual contact with outgroups.  These factors are often overlooked in the ?contact vs. conflict? literature.  To be sure, being comfortable with outgroups does not guarantee actual contact                                                 10 Berry and Kalin (1995) include Muslims, Jews, and Sikhs in their analysis of outgroup hierarchies.  They acknowledge that there is a religious component to the popular perceptions of these groups.  However, the authors fail to theorize how this component may affect people?s perceptions.  In fact, Muslims and Sikhs are at the bottom of the ethnic hierarchy in their study.  Jews are ranked in the middle.  This lends support to my findings that cross-categorized groups are ranked between ethno-racially-defined groups and religiously-defined groups.  41 with these groups.  But it does hint at a greater social acceptance of outgroups, which relates to the ?equal status? condition of positive intergroup contact (Allport 1954).  As will be discussed more below, the analysis shows that identifying nuance in outgroup categorization is critical: outgroups categorized as religiously different have less social acceptance than groups categorized as solely ethno-racial.  A secondary goal of the study was to identify the individual-level factors that shape comfort with outgroups, in an attempt to test the resilience of the hierarchical rankings found in the initial analysis.  The hope was to discover which people held the hierarchy and which deviated from the trend. In general, the theoretical expectations were supported.  First, white, mainstream Christian Canadians tend to be more comfortable with outgroups that are ethno-racially different as compared to outgroups that are religiously different.  This finding is supported here by consistent results across three different survey subsamples.  Ethno-racial difference can spur wariness of contact, and the different rankings of Blacks and Asians for different sub-samples of respondents (e.g., younger vs. older respondents) illustrates this point.  But, the consistently low ranking of Atheists and fundamentalist Christians highlights the role of religious difference in an individual?s assessments of different outgroups.  The lower rankings of Jews and Muslims compared to Blacks and Asians also highlights the disproportionately negative influence of religious difference.   The waning influence of setting intimacy for Atheists and fundamentalist Christians is yet more evidence.  The occurrence and the results of intergroup contact in the real world, then, will likely be affected by the type of outgroup in question, particularly if the outgroup is defined by religious difference. More work needs to be done to identify precisely why the majority group Canadians analyzed in this study are more comfortable with ethno-racial outgroups as compared to  42 religious outgroups.  A compelling explanation is that an individual's religious outlook is a fundamental component of her identity and, thus, religious differences are perceived as more threatening than racialized ethnic differences.  Past research has pointed in this direction (e.g., Haidt et al. 2003; Jehn et al. 1999).  On the other hand, the dramatic change in social awareness that accompanied such events as the 1960s civil rights movement or the debates around Canadian identity in the late 1970s and early 1980s may have influenced individuals' awareness of social inequalities differently.  White, mainstream Christian Canadians might be more accepting of (or at least publicly say they are more accepting of) ethno-racial, than religious, differences in part because the rights debates were primarily focused on race and ethnicity, rather than on belief.  Canada's complementary immigration and multicultural regimes hints at this asymmetry: policy restrictions do not apply to racial or ethnic origins of minority groups, but do apply to some religious practices of these groups (for example, polygamy).  Perceiving religious and ethno-racial difference in a single outgroup (i.e., Muslims or Jews) might temper the perceiver's attitudes as she partially overcomes wariness of religious difference in response to her relative comfort with racialized ethnic difference.  When religious difference is uncoupled from ethno-racial difference (i.e., Atheists or fundamentalist Christians), white, mainstream Christians? wariness is thrown in sharp relief since the lessons of the civil rights movements and identity debates no longer apply.   The current analysis, too, supports the idea of different reactions to hypothetical intergroup contact at different levels of intimacy.  The respondents in the study appear to be the most wary of both ethno-racial and religious difference in the highly intimate in-law scenario.  This conforms to the expectation that harmonious family relations are highly valued and any possible disruption will be generally eschewed.   The respondents are least wary of  43 outgroups if the outgroup member is a teacher at a local school.  The abstract concept of this setting, due to the imprecise wording of the survey question, may have resulted in respondents being the most comfortable with outgroups in this setting.  If the survey question had specifically referenced the respondent?s children, or had otherwise been more applicable to the everyday life of the respondent, the result might have been different.  Comfort with outgroups as bosses falls in between the two extremes.  The analysis also finds that the outgroup rankings stay essentially the same, even as comfort increases and intimacy decreases.  This supports other work indicating that the setting of intergroup contact is not neutral and can have a significant impact on the quality of contact (e.g., Maddux et al. 2005; Wittenbrink et al. 2001). Not only is the outgroup ranking stable across settings, it is stable across various levels of common attitudinal determinants.  This observation is the third important finding here: there is considerable consensus around the outgroup hierarchy, at least among white, mainline Christians.  Varying self-identities, education levels, income, residing in Quebec, sex, age, and so on produce similar rankings of outgroups.  This is true even with groups of people expected to be more progressive in their views, such as young people or the highly educated.  With few deviations, the ranking tends to be Blacks and Asians at the top, Jews and Muslims in the middle, and Atheists and fundamentalist Christians at the bottom.  Probably the biggest deviation is with the ranking of Atheists and fundamentalist Christians, or the two religiously different outgroups.  This might reflect the centrality of religious difference in perceptions of outgroup difference discussed above.  Because it is more important than ethno-racial difference, there is less consensus about its place in the outgroup hierarchy.  Or, again, it could be due to the lessons learned from the minority rights movements in Canada and  44 elsewhere in the ?60s and ?70s ? lessons that appear to have been learned by people across very different walks of life.  The contribution, then, of the present paper is its detailed examination of one of the factors that shape comfort with intergroup contact.  The current academic debate around the effects of contact, as well as the costs and benefits of ethnic diversity in general, would benefit from considering the determinants of the phenomenon they wish to link to a variety of political outcomes.   Moreover, with ethnic diversification in Western societies continuing with little abatement, clarity on these issues would be of interest to both citizens and policy makers.  Canada, with its expansive multicultural policy regime that has little to say about religious difference, would especially benefit from clarity.  This study shows that white, mainline Christians are still quite wary of religious difference.  Ethno-racial difference is perceived as problematic, as shown by the decrease in comfort ratings for Blacks and Asians as the setting becomes more intimate.  But, it is clear that religiously-defined groups bare the weight of social repudiation.11  Given that many of Canada?s recent immigrants come from countries with non-Western values, policy makers need to directly address this common parameter of outgroup difference by either encouraging newcomers to adopt mainstream values or by encouraging majority group Canadians to accept religious difference.  If majority group Canadians continue to be wary of religious difference and Canada?s immigration patterns remain the same, the ?equal status? condition of positive intergroup contact would be difficult to achieve.  Intergroup contact occurring between groups of unequal status are much more                                                 11 Reitz et al. (2009) find that Canadian minority groups who identify as ethnically and religiously different rely more on the former, rather than the latter, identity when forming attachments to the national community.  This finding hints at a radical disconnect between which minority identities are important for minorities and which minority identities are important for the majority.  45 likely to be deleterious to the ultimate goal of harmonious social and political relations in Canada and elsewhere.    46 Figure 2.1a-c: Predicted ?Very Comfortable? Response for Outgroups in Three Settings a. ?A teacher at a local school was someone who is [outgroup]?  b. ?Your boss was someone who is [outgroup]?  c. ?A close relative, like your sister or daughter, was going to marry someone who is [outgroup]?   . of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christians. of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christians. of ?very comfortable?Blacks Asians Jews MuslimsAtheists Fund.Christians 47 Table 2.1: Determinants of Comfort Ratings for Ethno-Racially and Religiously Defined Outgroups  Asian -0.194 Atheist*Teacher -0.226 F.Christian*income -0.0316  (0.253)  (0.190)  (0.0351) Jew -0.736** F.Christian*Boss 0.0809 Asian*urban 0.0971  (0.304)  (0.186)  (0.111) Muslim -1.366*** F.Christian*Teacher -0.459** Jew*urban 0.249*  (0.353)  (0.204)  (0.133) Atheist -2.581*** Asian*religious id 0.0405 Muslim*urban 0.00471  (0.439)  (0.0532)  (0.153) F.Christian 0.978** Jew*religious id 0.0361 Atheist*urban -0.151  (0.466)  (0.0605)  (0.180) Boss 0.353** Muslim*religious id 0.132* F.Christian*urban -0.371**  (0.152)  (0.0724)  (0.184) Teacher 0.668*** Atheist*religious id 0.666*** Asian*age 0.0462*  (0.154)  (0.0861)  (0.0257) religious id 0.0404 F.Christian*religious id -0.340*** Jew*age 0.104***  (0.0677)  (0.0974)  (0.0281) ethnic id 0.0117 Asian*ethnic id -0.0391 Muslim*age 0.00310  (0.0745)  (0.0584)  (0.0301) education 0.197*** Jew*ethnic id -0.00345 Atheist*age -0.0424  (0.0417)  (0.0644)  (0.0385) income 0.0350 Muslim*ethnic id 0.0593 F.Christian*age -0.0172  (0.0243)  (0.0839)  (0.0402) urban 0.287* Atheist*ethnic id 0.0511 Asian*sex -0.116  (0.147)  (0.0944)  (0.0926) age -0.196*** F.Christian*ethnic id 0.0806 Jew*sex 0.0170  (0.0297)  (0.107)  (0.100) sex 0.0681 Asian*education 0.0501 Muslim*sex -0.177  (0.121)  (0.0324)  (0.126) quebec -0.0409 Jew*education 0.0599 Atheist*sex -0.0991  (0.135)  (0.0374)  (0.148) Asian*Boss -0.177 Muslim*education 0.0280 F.Christian*sex -0.211  (0.112)  (0.0435)  (0.169) Asian*Teacher -0.104 Atheist*education 0.0267 Asian*quebec 0.0690  (0.117)  (0.0513)  (0.110) Jew*Boss -0.0293 F.Christian*education -0.241*** Jew*quebec -0.450***  (0.129)  (0.0609)  (0.129) Jew*Teacher -0.0984 Asian*income -0.0372** Muslim*quebec -0.693***  (0.133)  (0.0182)  (0.151) Muslim*Boss 0.470*** Jew*income -0.0502** Atheist*quebec 0.765***  (0.155)  (0.0217)  (0.159) Muslim*Teacher 0.560*** Muslim*income -0.0214 F.Christian*quebec -0.442**  (0.165)  (0.0247)  (0.186) Atheist*Boss 0.0838 Atheist*income -0.0238    (0.179)  (0.0316)   (cont.)  (cont.)  Observations 5880 Ordered logistic coefficients reported; Robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1  Independent Variable Scaling: 1) religious id: v. imp=1/v. not imp.=4, 2) ethnic id: v. imp=1/v. not imp.=4, 3) education: <grade 9=1/post-graduate=7, 4) income: <$10k=1/>$100k=10 5) urban: rural=0/urban=1, 6) age: 18-24=1/>74=8, 7) sex: male=0/female=1 8) quebec: no=0/yes=1   48 Figure 2.2: Predicted ?Very Comfortable? Response for Outgroups as In-Laws  a. Religious ID b. Ethnic ID c. Education    d. Income e. Urban f. Age    g. Sex h. Quebec  of 'very comfortable'Blacks Asians Jews Muslims Atheists Fund.Christiansnot at all important very important. of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christiansnot at all important very important0.2.4.6probability of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christianspost-graduate degree grade 8 or less0. of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christians$100 000 and over less than $10 of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christiansurban resident rural resident0.2.4.6probability of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christians75 years old and older 18-24 years old. of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christiansfemale male. of ?very comfortable?Blacks Asians Jews Muslims Atheists Fund.Christianslives in Quebec lives outside Quebec 49 Figure 2.3: Differences in Predicted ?Very Comfortable? Response for Outgroups as In-Laws  a. Religious ID b. Ethnic ID c. Education    d. Income e. Urban f. Age    g. Sex h. Quebec              -.4-.20.2.4difference in probability of comfortF.Christian/BlackBlack/AsianAsian/JewJew/MuslimMuslim/AtheistAtheist/F.Christianvery important not at all important-.20.2.4difference in probability of comfortF.Christian/BlackBlack/AsianAsian/JewJew/MuslimMuslim/AtheistAtheist/F.Christianvery important not at all important-.4-.20.2.4difference in probability of comfortF.Christian/BlackBlack/AsianAsian/JewJew/MuslimMuslim/AtheistAtheist/F.Christiangrade 8 or less post-graduate degree-.2-. in probability of comfortF.Christian/BlackBlack/AsianAsian/JewJew/MuslimMuslim/AtheistAtheist/F.Christianless than $10 000 $100 000 and over-.2-. in probability of comfortF.Christian/BlackBlack/AsianAsian/JewJew/MuslimMuslim/AtheistAtheist/F.Christianrural resident urban resident-.2-. in probability of comfortF.Christian/BlackBlack/AsianAsian/JewJew/MuslimMuslim/AtheistAtheist/F.Christian18-24 years old 75 years old and older-.2-. in probability of comfortF.Christian/BlackBlack/AsianAsian/JewJew/MuslimMuslim/AtheistAtheist/F.Christianmale female-.2-. in probability of comfortF.Christian/BlackBlack/AsianAsian/JewJew/MuslimMuslim/AtheistAtheist/F.Christianlives outside Quebec lives in Quebec 50  CHAPTER THREE  DETERMINANTS OF WHITE AND NON-WHITE ATTITUDES ABOUT ETHNIC DIVERSITY IN CANADA: THE ROLE OF ETHNIC PREJUDICE, MATERIAL SELF-INTEREST, AND POLITICAL VALUES  Due to changing immigration patterns over the past 50 years, a new kind of multiethnic society has emerged across North America and Western Europe.  Countries, such as England, France, Australia, and the Netherlands, have experienced violent inter-ethnic conflicts as their governments and citizenry grapple with this new ethnic diversity.  As one of the most ethnically diverse Western democracies (Patsiurko et al. 2012), Canada has surprisingly avoided the violent confrontations seen in other countries.  It has not, however, been immune to controversy.  For example, the fierce public backlash over proposed Sharia arbitration in Ontario in 2005 hints at an undercurrent of tension between Canada?s majority and minority ethnic groups.   The strained interethnic conditions in Quebec that led to the 2007 Bouchard-Taylor Commission on reasonable accommodation of minorities, the 2008 riots in Montreal after police shot an unarmed Honduran teen, and the public disagreement between the Prime Minister and the Chief Electoral Officer about Muslim women wearing niqabs in voting booths, are other examples. Despite these controversies, Canada?s recent history has been characterized by relative interethnic harmony.  Certainly, the durability of Canada?s interethnic harmony depends on the support of its citizens.  The attitudes Canadians hold are important predictors of whether their society continues its comparatively successful integration of its disparate ethnic groups.  Attitudinal research outside of Canada has identified key ingredients ? namely, prejudice,  51 material self-interest, and political values ? that underpin attitudes about the lived experience of ethnic diversity.  But, few attempts have been made to apply this research agenda to the Canadian context.   Even fewer attempts have been made to compare multiple attitudinal determinants in a single analysis in this country.   While Canadians are generally supportive of ethnic diversity, deep divisions can be found on more specific questions of exactly how immigrant communities should be integrated into the larger Canadian society.  This paper asks about the multiple motivations behind these divisions about ethnic diversity.  Are negative attitudes about, say, the contributions of immigrants or the preservation of ethnic minority culture motivated by prejudice, particularly since the public rhetoric around immigrants and multiculturalism is often shorthand for racialized communities?  Or, are negative attitudes fueled by the perception that newcomers themselves, or their demands on society and government, strain finite material resources, thus affecting people?s material self-interest?  Or, are attitudes about ethnic diversity shaped by a clash of deeply held values that have little to do with prejudice or material concerns?  Importantly, how might these factors depend on the individual?s ethnic minority or majority status? With individual-level survey data from the 1995, 2000, and 2005 waves of the Project Canada series, I investigate the relative influence of factors thought to shape the terrain of ethnic diversity attitudes.   In particular, I look at measures of ethnic prejudice, material self-interest, and political values.  I compare their relative effects on attitudes about the general perceived value of immigration and about cultural integration strategies.  I also test for variations across majority and minority ethnic groups.  Including rival factors in a single analysis allows a unique comparison of their relative power in driving attitudes in the  52 Canadian context.  Discovering the interplay of factors, especially how they might vary for majority and minority ethnic Canadians, will help predict the resiliency of the multicultural project in Canadian society, since the health of Canada?s new society depends on the harmonious relations between communities encouraged to be distinct by federal and provincial governments. To anticipate my results, I find that all three rival determinants ? ethnic prejudice, material self-interest, and political values ? shape attitudes about immigration and multiculturalism.  There are, however, important variations across majority and minority ethnic groups.  For example, while measures of ethnic prejudice dominate the attitudinal terrain of white respondents, it is not as clear cut for non-white respondents.  Moreover, there are important similarities between the two groups.  Education, for instance, has a strikingly similar impact on both white and non-white attitudes. This paper begins with a brief outline of the debate around the determinants of attitudes toward ethnic diversity in Canada and elsewhere.  After this, I empirically analyze the relative influence of ethnic prejudice, material self-interest, and political values using multivariate regression analysis.  The paper concludes with relevant discussion on the shape of attitudes and policy implications for this salient political realm in Western society. Determinants of Ethnic Attitudes  Like most political attitudes, there is no single factor that explains attitudes toward ethnic diversity.  Past research on ethnic diversity attitudes tends to focus on three factors ? ethnic prejudice, material self-interest, and the commitment to deeply-held political values.  These three factors, or ?primary ingredients? (Kinder and Sanders 1996), largely determine the terrain of individual attitudes about a variety of issues related to the political and social  53 integration of ethnic minority communities.  Each of these ingredients has advocates in the social science literature, and it is my assumption that they likely all have at least a small place in the attitudinal terrain.  The present paper endeavors instead to look at the relative influence of these primary ingredients and how their influence may depend on the ethnic status of the individual.  Identifying precisely how prejudice, material self-interest, and political values shape attitudes toward ethnic diversity can make an important contribution to the understanding and prediction of public opinion.  Predicting public opinion should be desirable for policymakers formulating policy in this highly divisive issue area. At a minimum, explanations of attitudes about multiculturalism, immigration, and ethnic diversity need to consider the role of ethnic prejudice.  Prejudice is typically defined as an antipathy derived from a faulty and inflexible generalization (Allport 1954).  Prejudice against different ethnic groups is generally thought to be a relentless undercurrent in real-world interethnic relations in Canada and elsewhere.12  However, the extent of its influence is debated (e.g., Sniderman and Hagen 1985).  The Canadian literature tends to paint an optimistic picture of Canadians? level of tolerance, suggesting that ethnic prejudice is comparatively low (e.g., Adams 2007; Berry and Kalin 1995; Lambert and Curtis 1984).  But relying on this optimism or rushing to label Canadian society as an interethnic relations success is short-sighted.  The psychology literature establishes that people everywhere have a propensity for prejudice against others they perceive as different.  Experiments in social psychology demonstrate how easily individuals divide themselves into groups and hold preferences for their perceived ingroup and feel threatened by perceived outgroups (e.g.,                                                 12 The concept of ethnicity is vague mainly because it depends on an individual?s subjective assessment of herself and others.  For analytical purposes here, I use Yinger?s standard definition of ethnicity: ?an ethnic group is a segment of a larger society whose members are thought, by themselves or others, to have a common origin and to share important segments of a common culture and who, in addition, participate in shared activities in which the common origin and culture are significant ingredients.? (1994, p. 3)  54 Tajfel and Turner 1979).  Social identity and self-categorization theories propose that ingroup membership is formed from a deep desire to achieve social and psychological order in a complex world.  This ingroup membership becomes part of an individual?s identity, creating a ?collective identity? for this individual.  As such, an individual?s identity is closely bound to collective-level orientations, which in turn have considerable impact on an individual?s attitudes, cognitions, and behaviour (see Duckitt 1992; Pettigrew 1998; Turner 1999).  Prejudiced attitudes, as well as harmful cognitions (i.e., stereotypes) and behaviour (i.e., discrimination), are products of this deep-seated investment in one?s ingroup, affecting interethnic relations in all societies. Theories of prejudice have become central to the political literature on race and ethnicity (e.g., Citrin et al. 2001; Kinder and Sears 1981; Mendelberg 2001).  They are the psychological underpinning of political manifestations of ?us? and ?them?.  Research has found that prejudice is a key determinant of a host of political behaviours and attitudes, such as voting for extreme right political parties (Husbands 1979), support or opposition to welfare policies (Gilens 1999; Quadagno 1994) and immigration levels (Brader et al 2008), preferences for multicultural policies (Citrin et al. 2001), and attitudes toward affirmative action and educational quotas (Kinder and Sanders 1996).  The vibrant ?contact vs conflict? debate also hinges on a disagreement over whether intergroup contact increases or decreases prejudice (e.g., Pettigrew and Tropp 2006; Putnam 2007; Stolle et al. 2008). While the Canadian literature on prejudice is underdeveloped compared to other countries, comparative studies find Canada to be a positive outlier (Hiebert 2006).  Domestically, there is regional variation in attitudes toward immigrants and ethnic minorities, particularly when Quebec is compared with the rest of Canada.  Attitudes tend to be more  55 exclusionary in Quebec ? though not always.  The exclusionary tendencies are chalked up to cultural insecurity (Gidengil et al. 2004; Sniderman et al. 1993), which is conceptually subsumed by social identity theory and the process of creating and protecting collective identities.13  Still, comparatively speaking, Canadians appear to overwhelmingly support the idea of ethnic diversity and take pride in their perceived collective tolerance (Adams 2007; Bilodeau et al. 2012; Reitz and Breton 1994).  However, more in-depth studies on Canadians? ethnic attitudes suggest that many Canadians perceive their society as composed of ethnically-defined groups.  This orientation is not, in and of itself, negative.  Past work, however, has shown that Canadians tend to arrange these groups in a social hierarchy from the most acceptable to the least acceptable, usually with their own group on top (e.g. Berry and Kalin 1995; Kalin and Berry 1996).  This hierarchy hints at a deeply embedded prejudice and ethnocentrism, at least historically.  The lack of current research on ethnic attitudes means that little is known if Canadians still view their society in a hierarchical fashion, or if more liberal attitudes brought on by generational effects or norms and policy changes have altered citizens? perceptions (important exceptions include Harell et al. 2012).   Canadians? policy attitudes also reflect a hierarchy of acceptable cultural expression encouraged by official multiculturalism policies at the federal and provincial level.  Relatively benign policies, such as government support for festivals and other cultural events, receive less public opposition than other, potentially more threatening, policies, such as third-language or faith-based education.  The 2007 Ontario provincial election was dominated by the question of whether to extend public funding for religious schools beyond the province?s traditional financial support for private Catholic schools.  A 2009 survey revealed that                                                 13 Gidengil et al. make the point that cultural insecurity influences Anglophone attitudes, as well.  See Gidengil et al. 2004.  56 Canadians were much less likely to agree with funding schools for Muslims, Hindus, Sihks, Buddhists, and Jews, as compared to Catholics (see Soroka and Roberton 2010).  Also, similar to other countries, intolerant attitudes about ethnic minorities are related to electoral support for Canada?s right-wing political parties (Nevitte et al. 2000).  Still, even after extensive analysis of tolerance, attachment to Canada, and views on multiculturalism, Berry and Kalin (1995) argue that prejudice in Canada is not generally a problem and that the country?s multicultural experiment is largely successful (also, see Adams 2007).  While this assertion may be true, their evidence still indicates that the interethnic status quo in Canada contains seeds that could become socially and politically disruptive in the right circumstances. While prejudice dominates explanations of ethnic attitudes, other determinants have been identified.  An individual?s material self-interest is, perhaps, the most prominent rival.  In fact, material self-interest is likely the most common ? and most controversial ? explanation for political behavior in general.  Assumptions about individual rationality, the pleasure-pain calculus, and egoism that privileges the self create the foundation for much of the political behaviour literature (e.g., Sears and Funk 1991).  The role of material self-interest in driving overtly political behaviour is intuitive.  For example, citizens might vote for a political candidate that offers them the most economic benefits.  Less intuitive is the role of material self-interest in explaining attitudes about ethnic diversity.  Some research has suggested that attitudes, which look prejudiced at first glance, are actually driven by material concerns.  Specifically, attitudes that appear to be targeting ethnic groups due to their ethnicity may simply be an individual?s concern about the impact of increased competition in the labour or the social welfare markets, resulting in increased material uncertainty (Citrin et al. 1997;  57 Kessler 2001; Kluegel and Smith 1983; Mayda 2006; Palmer 1996).14  These negative attitudes may target ethnic minorities in one scenario, but can shift to another group if the material stakes change.15  For example, Bobo (1983) argues that whites? opposition to the American desegregation policy of busing was driven, at least partially, by material self-interest.  The author claims that whites resented the change in their preferred state of affairs.  This preference had little to do with racial segregation and more to do with privileged access and influence.  Whites simply wanted to protect this perceived material advantage and would object to any policy that might threaten it.  Bobo finds that a family?s income is associated with attitudes about busing, namely, the higher the family income, the more opposition is expressed (but see Kinder and Sanders 1996).  A Canadian study also finds a negative association between material self-interest (i.e., household income) and pro-minority attitudes, after controlling for factors such as education levels and gender (Blake 2003).  Coenders et al. (2001) find the opposite in their examination of the relationship between employment status and general prejudice.  In their study, manual laborers, shopkeepers, retirees, and the unemployed tend to score higher on a prejudice scale than professionals, supervisors, and middle management.  So, economic insecurity, at least as defined by the authors, is associated with negative ethnic attitudes.   An influential study by Hainmueller and Hiscox (2007) refutes the interpretation that labour skills are a proxy for material interests at all.  Instead, the authors argue that the                                                 14 Realistic Group Conflict Theory, formulated by Sherif et al. (1961), examines the role of material interests at the group level.  The theory posits that individuals can be influenced by increased economic competition for their ingroup from outgroups, even if their own personal situation is unaffected. See, e.g., Coenders et al. 2004; Money 1999; Wilkes et al. 2008.  15 This literature is slightly different than the comparative literature examining the role of material self-interest in shaping anti-immigrant attitudes (e.g., Fetzer 2000; Mayda 2004).  The latter?s general focus is on attitudes about immigrants with related policy implications (e.g., yearly intake levels).  The former?s focus is on attitudes about interethnic relations and related policy implications (e.g., accommodating religious difference).  There is considerable overlap between the two approaches.  However, I do not intend to dwell on the comparative immigration literature because of its slightly different parameters.   58 positive relationship between skills (as measured by education levels) and attitudes about immigration are due to the progressive values inculcated by the higher education of highly skilled individuals.  In fact, the same logic could be applied to many typical measures of material self-interest, such as employment status or household income.  Muller and Tai (2010) re-examine Hainueller and Hiscox?s proposition, however, and argue that it only applies to individuals with higher academic education, rather than higher vocational education.  Muller and Tai find that individuals with high vocational education exhibit more anti-immigrant attitudes than individuals with low education in general, again providing evidence for the role of material self-interest.  In general, then, the evidence for the role of material self-interest in shaping attitudes about ethnic diversity is mixed.  As such, it is still a worthy contender to be investigated in the present study of attitudes in Canada. The third attitudinal ingredient under examination here is an individual?s deeply-held political values.  Sniderman and his co-authors have laid considerable groundwork for this research agenda, arguing that a substantial amount of negative ethnic attitudes can be explained by principled political stances rather than thoughtless prejudice or material considerations.  Sniderman and Hagen (1985), for example, argue that much of white American attitudes about race can be explained by a combination of principled stances on moralism (or viewing individual moral deficiencies as the cause of social ills) and individualism (lauding self-reliance and hard work).  As such, opposition to policies intended to help blacks is not about opposing blacks as a group, but opposing the policy as violating these values.  In Italy, Sniderman et al. (2000) identify the powerful role of adherence to ?authority values? (e.g., order and discipline, sacrifice and self-denial, and respect for authority) in shaping anti-immigrant sentiments, but also how anti-immigrant sentiments  59 increase the appeal of authority values.  Finally, in the Netherlands, Sniderman and Hagendoorn (2007) find that a commitment to liberalism and tolerance fosters acceptance of Muslims as a group, even though many Dutch are troubled by aspects of Muslim cultural practice.  So, the Dutch people?s deeply-held liberal values result in their acceptance of Muslims? illiberal practices (also see Sniderman and Carmines 1997; Sniderman and Hagendoorn 2007; Sniderman and Piazza 1993; and see Bobocel et al. 1998; Citrin et al. 2001; Kluegel and Smith 1986; Sears et al. 2000).  Literature on how political values inform Canadians? attitudes about multiculturalism, immigration, and ethnic diversity is scant.  Berry and Kalin?s concept of a ?multicultural ideology?, though, can be thought of as a principled position on ethnic relations.  The concept frames ethnic diversity as a normative good.  It supports ?having a culturally diverse society in Canada, in which ethnocultural groups maintain and share their cultures with others? (Berry and Kalin 1995, p. 306; also see Berry et al. 1977).  The authors fail to further develop the theoretical aspect of the concept, but their operationalization gives clues about their intention.  The ?multicultural ideology scale? is composed of ten survey items that measure whether a respondent agrees or disagrees that ethnic minorities should assimilate or maintain group distinctiveness, and whether ethnic diversity is valuable or harmful to Canadian society.  The scale has an intercorrelation alpha of .79 and includes items such as, ?It is best for Canada if all immigrants forgot their cultural background as soon as possible,? ?You can learn a lot from cultural groups,? and ?Cultural and racial diversity is a fundamental characteristic of Canadian society?. Berry and Kalin?s concept of a multicultural ideology, thus, reflects the broad value dimension of liberal egalitarianism, which shapes attitudes about the disadvantaged in society  60 and has been central to some attitudinal studies conducted outside Canada (e.g., Gaertner and Dovidio 1986; Sears et al. 2000).  It views ethnic minorities as often underprivileged and requiring assistance, typically from government, to be able to participate as equals in society.  This version of liberalism is often viewed as conflicting with classical liberalism, which posits that individuals ? not groups ? are the primary societal unit.  The multicultural ideology and, thus, liberal egalitarianism can be linked to recent normative political arguments asserting that multiculturalism is not in direct competition with the individualistic element of liberalism.  Instead, it reflects the value of equal opportunity that is at the core of liberal democracies (e.g., Kymlicka 1989).  Further, the official governmental position on multiculturalism takes a cue from Prime Minister Pierre Elliot Trudeau?s 1971 policy announcement.  Trudeau argued that, instead of inhibiting individual expression, multiculturalism?s recognition of distinct cultural identities ? seen as a critical psychological anchor ? provides individuals with the confidence needed to fulfill liberalism?s promise of individual liberty.16  So, a multicultural ideology, or liberal egalitarianism more generally, need not conflict with values associated with traditional liberalism in the minds of Canadians.  Being influenced by the multicultural ideology would certainly shape an individual?s attitudes about ethnic diversity whether she is a liberal or not.  In Canada, where multiculturalism has been politically salient for decades, the influence of this particular political principle will likely be readily visible.   Not only can political values have a direct influence on people?s attitudes about ethnic diversity, they can also act as powerful constraints on adverse psychological impulses.  Personality psychology has established that certain personality types are predisposed to ethnic prejudice (e.g., Sibley and Duckitt 2008).  Some individuals are able to use their beliefs to                                                 16 Trudeau?s speech can be found here: Accessed June 1, 2013.  61 override these entrenched dispositions with a process similar to breaking a bad habit.  Devine (1989) finds that individuals holding egalitarian beliefs can recognize that a conjured negative stereotype against an outgroup member, and the concomitant perceived threat, violates their beliefs and can decide to replace the negative feelings and attitudes with positive ones (also see Monteith 1993).  The work on how individuals can use their values to self-regulate their own prejudice tends to neglect prejudice against ethnic minorities.  Still, it is likely that liberal egalitarian values, or the multicultural ideology, can play a role through this ?self-regulation? in determining attitudes about immigrants and multiculturalism in Canada.   Majority and Minority Attitudes  Much of the empirical research on ethnic diversity focuses on predicting majority, rather than minority, attitudes.  Little attention is devoted to how these primary attitudinal ingredients, and others, may influence majority and minority groups differently (for exceptions, see Arends-Toth and Van de Vijver 2003; Bilodeau and Kanji 2006; Judd et al. 1995; Kalin and Berry 1996; White et al. 2006).  Clearly, the stakes for majorities and minorities in political policies addressing ethnic diversity will be different and, at times, opposed.  This dynamic alone makes it imperative that majorities and minorities are investigated and, if possible, investigated in the same study.   In reality or in perception, ethnic diversification often results in the strengthening of minority group social positions and the weakening of the majority group social advantage.  In Canada, ethnic diversification has been accompanied by a public consensus about multiculturalism and an extensive multicultural policy framework.  The social and political change in the status quo that began in the 1960s and 1970s might be perceived as weakening the majority position while strengthening the minority position.  The intention of Canada?s  62 multicultural framework is to integrate Canada?s numerous ethnic groups and share a common political and social experience.  This goal would, in fact, weaken the traditionally privileged position of Canadians with traditional British or French heritage.  The change in the societal status quo could drive majority group members to resist, and minority group members to embrace, liberalizing multicultural and immigration policies.   On the micro-level, interethnic contact may reveal practical differences in majority and minority attitudes.  Experiments suggest that interethnic contact may produce fewer negative effects for minority members than majority members, simply because the former tends to have more experience with intergroup contact (Hyers and Swim 1998).  This experience allows minorities to more readily develop cooperative strategies, such as the hyper-awareness and keen anticipation of conflict that characterizes the concept of intergroup ?mindfulness?.  These strategies, in turn, reduce intergroup anxiety.  Moreover, they may explain why ethnic minorities are less hostile to ethnic outgroups if their neighborhoods are mixed (Oliver and Wong 2003), or why ethnic minority trust levels are affected less negatively by neighborhood diversity than white trust levels (Stolle et al. 2008).   On the whole, then, it is reasonable to suggest that important attitudinal differences between majority and minority group members exist.  For example, outgroup hostility, a central feature in the prejudice literature and something that I look at here, might result in more assimilationist attitudes about cultural integration for someone who identifies with the ethnic majority.  In contrast, outgroup hostility would likely produce the opposite effect for someone who identifies with an ethnic minority, as it could be seen as strengthening their societal position vis-?-vis the ethnic majority.  Material self-interest is another example of how determinants might vary according to an individual?s ethnic identity.  Given that ethnic  63 minorities (and not just newcomers) tend to be in a more economically precarious position relative to the majority (e.g., Galabuzi 2006), their stakes with labour market competition or perceived strain on the social safety net would be higher.  Higher stakes would certainly amplify any influence that material self-interest would have on attitudes about immigration and multiculturalism.    Clearly, there is good theoretical reason why we should anticipate that the influence of attitudinal determinants will depend on the ethnic identification of the respondent.  Empirically examining these differences and what they mean in terms of the terrain underpinning ethnic attitudes should not be overlooked and is a central motivation here.   Data and Methods The role of ethnic prejudice, material self-interest and political values, as well as important differences between majority and minority ethnic groups, are the focus of the present empirical investigation into the terrain of ethnic attitudes in the Canadian context.  The attitudes to be explained are central to Canada?s ethnic diversity debate, tapping two different aspects of the question of ethnic minority integration.  The present study does not intend to settle the debate of whether these determinants influence ethnic attitudes; it is assumed that they do.  Instead, I seek to examine the relative influence of each across majority and minority ethnic status.  That is, the goal is to determine which factors matter more for the attitudes in question and how they are contingent on the ethnic identity of the respondent. To test the relative influence of these attitudinal determinants for both majority and minority groups, I draw on the 1995, 2000, and 2005 waves of the public opinion survey, Project Canada.17  The rich and diverse data are from a mail-in survey, which was designed to                                                 17 Reginald W. Bibby, Project Canada Survey Series, 1995, 2000, 2005.  University of Lethbridge.  The survey is conducted by mail. It has a sample size of approximately 1500 for each wave, and an average response rate of  64 measure Canadians? views on social issues, intergroup relations, and religion.  To facilitate an investigation into attitudinal differences between majority and minority ethnicities, the data are pooled so a sufficient number of non-white respondents are represented in the sample.  For the estimations, there are 557 non-white respondents and about 4406 white respondents.   Dependent variables The dependent variables used in the analysis measure two central aspects of ethnic diversity debate as it occurs in Canada: general attitudes about immigration and preferences for common cultural integration strategies.  I have rescaled each indicator so that high values reflect an attitude that is exclusionary toward these particular facets of ethnic diversity.  The statistical simulations following the regression analysis demonstrate which determinants improve negative attitudes and to what extent.  Each attitude, whether it be general attitudes toward immigration or how ethnic minorities might integrate into Canadian society, has clear policy implications.  Thus, investigating how negative attitudes can be improved, the analysis illuminates potential pools of public support for future policy decisions.   The first dependent variable, that I measure with ?immigration?, asks if the respondent agrees or disagrees that ?immigration is a good thing for Canada? with high scores indicating strong disagreement (see Appendix Two for precise wording and descriptive statistics for all variables).  This question solicits a vague position on immigration and does not press the respondent to expose interethnic resentments, self-interest, or value positions.  It might invoke retrospective or prospective assessments of Canada?s immigration levels, implicating identity, material, or value concerns.  If past Canadian evidence can be applied here, material self-interest may have a stronger association with this type of general statement about immigrants                                                                                                                                                    65%.  See  Accessed June 17, 2013.  The data was received directly from the investigator.  All analysis is my own.  65 (e.g., Palmer 1996; Wilkes et al. 2008).  Still, ethnic prejudice and political values likely play a role.  The distribution of the variable shows that about 28 percent of respondents disagree or disagree strongly that immigration is good.  About 69 percent agree or agree strongly. The second dependent variable ? mosaic ? measures a specific attitude about Canada?s multicultural society.  It asks if the respondent favours the ?melting pot? or ?mosaic? type of cultural integration.  High scores indicate the respondent favours a melting pot, or, as described in the questionnaire, the kind of society where immigrants ?should give up their cultural differences and become Canadians.?   This question clearly indicates that a melting pot requires cultural homogeneity.  The description of a mosaic encompasses the common understanding of Canada?s multicultural society, where ?people are loyal to Canada yet keep many of the customs of their previous countries.?  So, respondents are prompted to think about the dual affirmation of majority and minority culture, which is stressed in official multiculturalism policy.  There is less consensus in the data around this statement than with immigration, which is surprising given the oft-touted Canadian pride in multiculturalism.  Only 47 percent of respondents favoured a mosaic approach ? the type of multiculturalism supported by decades of government policy-making.  Thirty-six percent favoured the melting pot idea.  The question may be difficult for some respondents to answer, given the complicated nature of the issue of culture and cultural integration strategies, especially in light of prominent normative debates in academic circles about the precise relationship between ethnic minority group recognition and classical liberal individualism (e.g., Kymlicka 1989).  In fact, 17 percent of respondents stated they did not know how to answer the question, did  66 not respond, or made a qualified statement.18  Still, the question is used here because the concepts of a melting pot and a mosaic are often evoked in discussions of ethnic diversity, making the question an important indicator of public attitudes about this central concern in the diversity debate.  Moreover, a survey question explicitly mentioning the two buzzwords is rare, making it an attractive measure to explore.   Independent variables Multiple independent variables tap aspects of each of the three primary ingredients.  It would be virtually impossible to measure all aspects of ethnic prejudice, material self-interest, and political values in a single study.  Instead, a few salient aspects are measured that should advance an understanding of the primary ingredients? effects on individual attitudes.  The first ingredient ? ethnic prejudice ? is measured in part by two attitudinal scales.  The scales focus on the idea of outgroup resentment in the societal and personal realm.  An individual?s outgroup resentment can be expressed though resistance to an outgroup member?s participation in society and family life simply because the latter is perceived as ethnically different.  Resistance to an outgroup?s participation is a central behavioural outcome of an individual?s self-categorization and, thus, social identity.  It is a strong signal of prejudice and, since one?s social identity is formed early in life, it will precede the policy attitudes tapped in the dependent variables.   The first scale, power, combines attitudes about the perceived political power of four identifiable ethnic groups.  The question asks ?Do you think the following groups have too much power, too little power, or about the right amount of power in our nation's affairs??  The ethnic groups included in the scale are Blacks, Asians, East Indians and Pakistanis, and                                                 18 Qualified statements are of theoretical interest, but less than 2 percent of respondents made them.  Moreover, the survey does not record the qualifiers, so they may be a mix of more or less tolerant responses.  Accordingly, they are collapsed into the ?don?t know? category.  67 Whites.  The scale measures if the respondent thinks an ethnic outgroup has too much power, tapping the idea of outgroup resentment.  As such, it measures if white respondents think Blacks, Asians, and East Indians and Pakistanis have too much power, and if non-white respondents think Whites have too much power.  The dichotomous variable used to identify the ethnicity of the respondent, white, is generated from a self-identified ethnicity question in the survey.  The question asks the respondent?s ethnicity, offering a short list of possible responses.  The list is not detailed enough to get a thorough picture of the respondents? ethnic identity.  Still, a white/non-white binary is a common measure of ethnicity.  Moreover, it produces sufficient cases for a statistical analysis of non-white respondents.   Power is rescaled so both white and non-white responses run from 0 to 2; it has a Cronbach?s alpha of .68.  High scores indicate that the respondent thinks the identified outgroup has too much power.  I anticipate that the influence of power will be different for white respondents compared to non-white respondents. If outgroup resentment shapes Canadians? ethnic attitudes, power should be positively associated with the dependent variables for white respondents: higher outgroup resentment should be associated with more negative attitudes toward immigrants and multiculturalism.   For non-white respondents, power is expected to have a negative association with the dependent variables: higher outgroup resentment against whites should be associated with more positive attitudes toward immigrants and multiculturalism, as these should be viewed as contributing to the relative social and political influence of non-whites.  However, this influence may be muted because of the low end of the power scale.  There are two interpretations of non-whites stating whites have too little societal power (power=0).  Each interpretation has opposing implications for attitudes about immigration and multiculturalism.   68 Non-whites who think whites have too little power do exist: almost 16% of non-white respondents stated this opinion.  So, what might be the connection between this attitude and others about immigration and multiculturalism?  First, these non-whites may think many groups have too little power, whites included.  This could reflect a progressive, egalitarian spirit that desires a highly inclusive society, which would result in higher support for immigration and a mosaic.  Another possibility, however, is that non-whites who think whites have too little power might be reflecting a desire for an exclusive society.  These respondents may be categorizing whites as part of their ethnic ingroup.  In other words, whites are perceived as part of the ?us?.  So, non-whites who include whites as part of the ?us? and who think whites have too little power might be exclusionary in a similar manner to whites who think whites have too little power.  Immigrants, particularly non-white immigrants, would become ?them?, resulting in the respondent?s opposition to immigration and a preference for a melting pot.  The possible impact of these opposing motivations will be explored more once the results are discussed. The second scale measuring intergroup hostility is marriage.  It combines three survey questions asking if the respondent approves or disapproves of intermarriage between a) Blacks and Whites, b) Asians and Whites, or c) East Indians and Pakistanis and Whites.  The scale taps the respondent?s feelings about ethnic diversity in a small and intimate scenario.  It has been rescaled to run from 0 to 2, and it has a Cronbach?s alpha of .94.  The highest score indicates the respondent does not approve of any marriage between the identified ethnic groups, and the lowest score indicates approval of all three intermarriage scenarios.  Seventy-seven percent of respondents approve, and just over 15% disapprove, of all scenarios.  The remaining 7% fall in between.  Given that the question?s subject matter likely triggers social  69 desirability bias, it is surprising that so many respondents admitted to disapproving of the intermarriage scenarios proposed (e.g., Nederhof 1985).   For white respondents, strongly disagreeing with interethnic marriage should have a negative impact on their attitudes about immigrants and multiculturalism for similar reasons as power.  That is, disapproving of interethnic marriage, as an indicator of outgroup resentment, should result in more exclusionary attitudes about immigration and ethnic integration.  For non-whites, the impact is expected to be weaker, if there is an impact at all.  High scores on marriage, indicating high outgroup resentment, may contribute to negative attitudes about immigrants and multiculturalism.  For example, an ethnic minority who strongly disagrees with interethnic marriage may be intolerant of ethnic outgroups across the board.  This generalized intolerance might shape her attitudes about newcomers or ethnic integration.  However, because she is a minority, she may disapprove of interethnic marriage, yet still be supportive of immigration and multiculturalism because of a perceived personal or ingroup benefit.  The analysis is, thus, open to either possibility. The next two variables measure aspects of material self-interest.  The first is a dichotomous index of the respondent?s reported household income and employment status  ? typical objective measures of an individual?s material situation.  As the literature on material self-interest asserts, an individual?s tangible material situation is a good measure of the stakes in any situation that would potentially redistribute material advantage.  Thus, an individual?s material situation underpins other attitudes pertaining to potential shifts in her material position.  The index, named security, categorizes these objective measures so that ?0? indicates lower household income and unemployment (i.e., low material security) and ?1? indicates higher household income and employment (i.e., high material security).  The index was  70 created when preliminary tests demonstrated that household income and employment status did not contribute independently to the model.19  Theoretically, though, each variable captures different material stakes and, thus, different material vulnerabilities.  For example, someone could be unemployed and feel materially secure because they have a high household income.  So, a decision was made to include both variables with the security index.   Material self-interest is also investigated with a subjective measure of an individual?s material situation.  Satisfaction asks whether the respondent?s financial situation has gotten better, worse, or stayed the same in the past few years, with a positive response ? or high material security ? receiving a high score.  The literature investigating the role of material self-interest is divided about whether real or perceived financial instability produces negative or positive attitudes about competing groups (e.g., Bobo 1983; Coenders et al. 2001).  So, any influence uncovered in the analysis may be positive or negative.  The strength of the effect, however, will likely vary for whites and non-whites.  Non-whites in Canada tend to be more materially insecure compared to whites (Galabuzi 2006).  So, if material insecurity is associated with exclusionary attitudes about immigrants and multiculturalism, there could be a stronger effect for non-whites since they are, on average, characterized by relatively inferior material circumstances.  However, if material security is associated with exclusionary attitudes, there could be a stronger effect for whites, since they are, on average, characterized by relatively superior material circumstances. The influence of political values is indicated by chance and education.  Chance draws from a survey question that taps the debate between individualism and liberal egalitarianism, which is thought to shape attitudes about ethnic diversity (e.g., Kemmelmeier 2003; Sniderman et al. 2000).  The respondent is asked whether they agree or disagree that ?Anyone                                                 19 An F-test was nonsignificant.  71 who works hard will rise to the top? (strongly disagree=1, strongly agree=5).  Much of the political debate around ethnic integration in Canada and elsewhere seeks to determine the extent to which ethnic minority groups face systemic barriers to participation, and what (if anything) should be done to remove those barriers.  A liberal egalitarian political value would assert that these barriers exist, stymying opportunities for minority ethnic groups among others (Kinder and Sanders 1996; Strunk and Chang 1999).  Individualism, on the other hand, would discount these group-level barriers and, instead, emphasize the efficacy of the individual.  The wording of the variable itself is a strong indicator of the meritocratic aspect of individualism.  It asserts that success is a product of an individual?s abilities and those who work hard for success can achieve it.  Disagreeing with the survey question would indicate skepticism of the ?pull yourself up by your bootstraps? sentiment.  The indicator, as such, is a good measurement of general liberal egalitarianism, since it does not mention groups, ethnic or otherwise.  As such, it will not prime the respondent to think narrowly of societal barriers for ethnic minorities.  Instead, it will tap general values that are unencumbered of any possible bias related to the politicized nature of ethnic diversity issues.   In terms of how the variable might influence attitudes captured by the dependent variables, I anticipate that chance will work in the same direction for whites and non-whites.  Liberal egalitarians in each group should be more liberal in their views of immigration and multiculturalism.  The variable will likely have a stronger association with the variable measuring attitudes about cultural integration strategies (mosaic) since the latter more specifically implicates the debate around group rights and individual rights.  However, liberal egalitarianism also signals acceptance more generally of the idea of group-level interests.  Given that Canada?s discourse around immigration is tightly associated with the integration of  72 ethnic groups post-immigration, chance should also be associated with the dependent variable, immigration.  As such, high scores on chance ? reflecting individualistic values (or meritocracy more narrowly) ? should be positively associated with the dependent variables for both whites and non-whites.20 The second determinant in the political values battery is the respondent?s education level, measured with the variable education.  Education is thought to instill democratic norms of tolerance, with schools and universities deliberately working to reduce prejudice (e.g., Hainmueller and Hiscox 2007).  A higher education may also result in more critical assessments of immigration and multiculturalism that challenges the normative messages of the Canadian education system.  However, past research shows that education does tend to produce more progressive attitudes about immigrants and ethnic diversity (Bobo and Licari 1989; Blake 2003; Chandler and Tsai 2001; Oliver and Mendelberg 2000).  So, the more advanced a person?s education, the more likely they will be tolerant of perceived difference.  The effect of education should not differ significantly between white respondents and non-white respondents.  If the two groups have been through similar educational systems, the values inculcated should be similar.  However, the data cannot tell if some respondents, particularly non-whites, have attended schools outside Canada and thus be exposed to different value systems.  As such, I anticipate that the effects of education may be weaker for non-white respondents due to any diluting effects of other educational experiences.                                                 20 Perhaps chance is measuring attitudes about people?s raw capabilities, so disagreeing people can work their way to the top may reflect a respondent?s skepticism about others? intelligence or abilities.  The data provides no opportunity to test this possibility.  Fortunately, chance is related to other indicators measuring concern for minority groups in a way that is expected if my interpretation of the variable is correct.  For example, a crosstabulation shows that, compared to respondents who agree, respondents who disagree with chance are more likely to think ?People who are poor have a right to an income adequate to live on? and that ?Homosexuals are entitled to the same rights as other Canadians.?  So, individuals who, according to the present interpretation, are liberal egalitarians are more likely to support group rights with these two salient political issues.  This result signals that chance is a good measure of the ?liberal egalitarianism vs. individualism? values divide.  73 It should be noted that American work has suggested an interactive effect of political conservatism and education (Sidanius et al. 1996).  Namely, higher education among political conservatives serves to increase rather than decrease prejudice.  The argument is that political sophistication results in a better understanding of the zero-sum game of redistributive policies and, thus, the perceived weakening of the social and political dominance of the majority group.  In other words, political conservatives are better able to legitimize latent prejudices through the sophistication that comes with advanced education.  So, education is associated with prejudice through the intermediary of sophistication for political conservatives.  The present analysis will check for this possibility in the Canadian data.  The analysis controls for three additional variables shown to typically shape attitudes.  The dichotomous white has already been mentioned.  It indicates if the respondent is white or non-white and has been generated from the survey question asking respondents to self-identify their ethnicity.21   The respondent?s age and sex are also controlled.  Typically, older people and men tend to be more conservative and thus may hold more negative attitudes about Canada?s steadily diversifying population (Gidengil et al 2005; Schuman et al. 1997).  Moreover, in Canada, younger people will have been socialized into the multicultural consensus.  While there has been some criticism of multiculturalism, it has generally been the political norm since the early 1970s (e.g., Abu-Laban 2002).  So, the normative argument for ethnic diversity and multiculturalism will be normalized for young people growing up in the                                                 21 Canada?s ethnic diversity debate is only partially about different ethnic groups.  It is also about immigrant status.  I chose to define this central aspect of the debate by differences in ethnicity rather than immigrant status.  This decision was primarily guided by practical considerations.  Most notably, the number of cases was cut in half when the latter was substituted in the model.  There is considerable overlap between ethnic identity and country of birth in Canada?s popular discourse on ethnic diversity.  In fact, discussions around immigrants are often shorthand for visibly different immigrants; it is reasonable to assume that few Canadians are worried about the integration of British or German immigrants.  This overlap between ethnicity and immigrant status occurs in the data, as well: of the respondents born outside Canada, 43 percent identified as non-white, whereas only 19 percent identified as white.  As such, white is a suitable measure of the stakes tied to ethnic identity.  74 recent era.  So, respondents who are older or male are expected to have positive associations with the ethnic attitudes measured with the dependent variables.  Results Table 3.1 displays ordered logistic coefficients corresponding to the two dependent variables ? whether or not immigration is good (immigration) and if the respondent prefers a mosaic or melting pot approach to ethnic group integration (mosaic).  White and non-white respondents are separated so the relative influence of the determinants is more easily assessed.  Table 3.2 shows the marginal changes of each variable.  The coefficients are interpreted as the marginal change in the probability of strongly disagreeing with immigration or preferring a melting pot for a unit change in the independent variable of interest.  For the sake of simplicity, the controls age, sex, and year are not shown. Before delving into the direction of the associations, a few general patterns are worth noting.  With few exceptions, the determinants of outgroup resentment, material self-interest, and political values have a strong and consistent association with the dependent variables for white respondents.  Most are associated with both dependent variables.  A notable exception is the objective measure of material self-interest, which is the respondent?s reported income and employment situation (measured with the index stability).  The tables show that, for whites, it has no relationship with general attitudes about immigration.  Contrast this with the consistent relationship between subjective material self-interest (satisfaction) and the dependent variables.  This result suggests that the subjective perception of material security is more important to whites? attitudinal terrain than objective material security, at least with the ethnic attitudes measured here.  75  For non-whites, the associations are less consistent.  The only determinants that are associated with both dependent variables are attitudes about interethnic marriage (marriage) and education levels (education).  Attitudes about outgroup power (power), subjective material self-interest (satisfaction), and the measure of liberal egalitarian values (chance) are only statistically significant for the attitudes measured by immigration.  Objective material self-interest (security) is only statistically significant for attitudes measured by mosaic.  This relative inconsistency may be due to the different case numbers for whites and non-whites.  Or, it might be that opposing motivations for non-whites are cancelling out any effects.  I will address this in more detail below. For the most part, the variables are associated with the dependent variables in the expected directions.  For whites, negative attitudes about outgroup power and interethnic marriage are associated with more intolerant attitudes about immigration and a higher preference for a melting pot approach to ethnic minority integration.  As mentioned, social desirability bias may be at play here, given the sensitivity of the questions.  But, given that a sizable amount of respondents are willing to agree that various outgroups have too much power or that they disapprove of interethnic marriage, social desirability bias is not a concern here.  The survey?s mail-in approach also insulates the data from social desirability bias.  In fact, the impact of power is quite large, especially for attitudes about cultural integration: the marginal change in preferences for a melting pot is .15 for a unit change in concern about outgroup power (see fourth column in Table 3.2).  Notably, power is independent of material considerations, reinforcing the argument that it is indicating the outgroup resentment linked to ethnic prejudice.22                                                   22 Power and the indicator of subjective material self-interest, satisfaction, is correlated at -.02; power and the indicator of objective material self-interest, security, is correlated at -.04.  76 Non-whites have a similar association with attitudes about interethnic marriage: higher concern about interethnic marriage is associated with less support for immigration and a higher preference for a melting pot (predicted marginal change is .04 and .07, respectively).  Higher perceived material security is associated with more support for immigration and a mosaic for both groups (though this is only the case for non-whites with mosaic).  This result supports previous research finding a positive relationship between material security and positive ethnic attitudes (e.g., Citrin et al. 1997; Coenders et al 2001; Espenshade and Calhoun 1993).   Higher education and disagreeing with the individualistic sentiment that anyone working hard will rise to the top is also associated with more tolerant attitudes for whites.  Non-whites, too, have a similar association between education and the dependent variables.  The marginal changes coefficients suggest the effect of education is similar for both groups.  For chance, the effect is also similar, but small.  The associations for education and values (at least for whites) supports research arguing that they have a positive influence on ethnic tolerance (e.g., Espenshade and Calhoun 1993; Hainmueller and Hiscox 2007).  Sidanius et al. (1996) argue that higher education can be associated with more negative ethnic attitudes amongst conservative individuals in the US.  The analysis here finds no evidence for this proposal.  Education does have an interactive effect with a respondent?s political views, but only by serving to amplify positive ethnic attitudes with liberals.  It does not prompt conservatives to be more or less positive about ethnic diversity issues (analysis not shown). The tables hold a few surprises, as well.  First, for non-whites, power is negatively associated with immigration, measuring disapproval of immigration: when they feel that whites have too much power, they are more supportive of immigration.  The marginal impact  77 is -.04.  This relationship reflects the proposition that non-whites view immigration as a way to bolster their ingroup influence versus the white majority.  It also suggests that non-whites who think whites do not have enough power are categorizing whites as part of their ingroup and making exclusionary statements about immigration. 23  However, non-whites? attitudes on interethnic marriage are positively associated with immigration: when they disapprove of interethnic marriage, they are less supportive of immigration.  So, for non-whites, the two indicators of outgroup resentment (power and marriage) work in the opposite direction for general attitudes about immigration.   Another surprise is that power has no influence on non-white preferences for a melting pot over a mosaic.  It works as expected for whites: the more whites resent outgroup power, the more they prefer a melting pot.  I had speculated that, for ethnic minorities, resentment of outgroup power would increase preferences for a mosaic-style of ethnic integration, as a mosaic would give more societal influence to non-whites relative to whites.  However, this does not appear to be the case. The variable, chance, which indicates the contrast between individualism and liberal egalitarianism, also works differently for white and non-white respondents.  I had anticipated that chance would work in the same direction for both groups, though perhaps not at the same magnitude.  However, this is not the case.  For whites, the association between chance and the dependent variables is expected: agreeing that people just need to work hard to rise to the top is associated with higher disapproval of immigration and higher preferences for a melting pot.  The marginal change for strongly disagreeing with immigration is small, though (.006 and .02,                                                 23 Additional analysis lends support to this interpretation.  I replaced power with dummy variables for each category of the scale and reran the estimation for non-whites.  The dummy measuring the response ?whites have too little power? had a .38 probability of opposing immigration.  The dummy for ?whites have too much power? had a probability of .18.  The difference is statistically significant at p<.05.  78 respectively).  Still, the association indicates that people who agree that working hard results in success for everyone are individualists who are wary of policies with group rights implications.  The association for non-whites is unexpected, however.  Non-white respondents who have high scores on chance (reflecting individualistic or meritocratic values) have a lower probability of disagreeing that immigration is generally good for Canada.  The marginal change is -.01, which is small, but statistically significant.  So, non-whites are more supportive of immigration if they believe hard work results in success for all.  I have little explanation for this, other than perhaps some non-whites with a personal or family history of immigration view it through the lens of the hard work necessary to become established in a new country.  Perhaps this experience links the issue of immigration to the notion of hard work in the minds of some non-whites. Finally, for both whites and non-whites, being employed and having a high household income (security) is associated with a higher probability of preferring a melting pot to a mosaic integration approach.  So, higher objective material security is associated with more assimilationist preferences for cultural integration.  This supports past research about the association between material security (e.g., being employed, having a high income, etc.) and negative attitudes about ethnic minorities (e.g., Bobo 1983; Blake 2003).  This result also is a good reminder that even non-whites, who would be the primary target of multiculturalism policies, do not automatically support a mosaic-style integration strategy.  In fact, the marginal effect of security is stronger for non-whites than whites (.09 and .04, respectively). 24  Clearly, non-white material self-interest does not automatically translate into supporting government policies targeting their interests.                                                 24 This result for non-whites might be explained by length of stay.  That is, non-whites who have been in Canada for a relatively long time may be more materially secure and be happier to assimilate into the wider Canadian culture.  Unfortunately, the data does not enable a test of this possibility.    79 To add to this, the influence of objective material security is the opposite of the influence we observe for subjective material interests, measured with satisfaction.  In this case, at least for whites, thinking that one?s material situation has gotten better over the past few years is associated with a lower preference for a melting pot.  As mentioned, higher subjective material stability is associated with more support for immigration for both whites and non-whites. Predicted probabilities can help illustrate the relationships present in Table 3.1.25  By manipulating the values of each of the determinants and comparing the resulting predicted levels of the dependent variables, the relative importance of prejudice, material-self interest and political values for whites and non-whites can be inspected.  For instance, imagine a hypothetical person who identifies as white, who thinks ethnic outgroups have the too much power in society, strongly disapproves of interethnic marriage, has a negative judgment about her financial situation, strongly agrees that individuals who work hard will rise to the top, and has little formal education ? all indications that she should be relatively intolerant of ethnic diversity.26  The predicted probability of her strong disapproval of immigration in general (model 1) is .38, or 38 percent.  However, if this same person thinks ethnic outgroups have too little power ? or a shift in attitude from the negative extreme to the positive extreme ? the predicted probability of her disapproving of immigration drops to 11 percent.  A 27-point decrease in probability is a considerable improvement in general attitudes about immigration.  If she changes her mind on interethnic marriage instead, her probability drops to 22 percent, or 16 points.  Suppose this person changes her perception of her financial situation from negative                                                 25 Predicted probabilities are calculated using Stata?s margins command.  All values are statistically significant at p<0.001. 26 The manipulated values represent extreme negative responses for the respective variables.  All other variables are set to their means.  80 to positive.  This change in perceived material security is accompanied by a nine-point decrease in probability to 29 percent.  If she strongly disagrees with the individualistic idea that hard work results in individual success, the probability of disagreeing immigration is generally good also decreases to 29 percent.  Moving from the minimum to the maximum amount of education decreases the predicted probability to 16 percent.  Figures 3.1a-b illustrates these marginal changes for each model.  Again, the variables have been set to their negative extremes: high outgroup resentment, strong disapproval of interethnic marriage, low subjective material security, high individualistic values, and a low education.  Only the independent variables that are statistically significant in Table 3.1 are included.  This means that the number of determinants changes for each figure, but this gives a more accurate illustration of the predicted responses.  Each row corresponds to one of the dependent variables from Table 3.1.  The first column illustrates the relationships for white respondents and the second column illustrates the relationships for non-white respondents.  The horizontal axis indicates the base line hypothetical individual (point 1) and the attitudinal shift for each independent variable (point 2).  The change in predicted probability is recorded with its corresponding trend line at point 2. These figures reveal the complex terrain of Canadians? attitudes.  It is clear, however, that attitudes about outgroup power is a dominant influence for white respondents.  The consistently strong association between power and the dependent variables supports the literature arguing for the prevailing influence of prejudice on attitudes relevant to race and ethnicity in general, rather than material or value concerns (e.g. Gilens 1999; Kinder and Sanders 1996).  As mentioned, the influence of power for non-whites is not as straightforward.  It has no influence on preferences for a melting pot or mosaic.  And, it has a negative  81 association with general attitudes about immigration.  This association is illustrated in the box located in the upper right corner.  The line for power shows that a shift from being very concerned about outgroup power to not being concerned at all increases the probability of non-whites strongly disagreeing immigration is generally good by 26 points.  As mentioned, however, non-whites appear to be positively influenced by more liberal attitudes about interethnic marriage and education. Discussion The present paper has two broad goals.  One is to investigate the relative influence of ethnic prejudice, material self-interest, and political values on individual attitudes about ethnic diversity.  The second goal is to look at how these relationships differ depending on an individual?s majority or minority ethnic identity.  The attitudes under examination here are salient for the political debate around ethnic diversity in Canada, capturing citizens? assessments of whether immigration is good and, more specifically, how to approach the cultural integration of ethnic minorities.  The implications of the analysis build on comparative literature on attitudes about interethnic relations, offering to fill gaps about the relative influence of attitudinal determinants and how they might vary across ethnic status.  These insights are particularly useful for the Canadian context with its long history of ethnic diversification.  In Canada, there is a vibrant public debate about ethnic diversity and considerable government resources are committed annually to attracting immigrants and managing relations between the country?s disparate ethnic groups.  Given that ethnic diversification due to high immigration levels shows no sign of abating, understanding and explaining the terrain of Canadians? attitudes about ethnic diversity is crucial for the successful integration of immigrants and the maintenance of harmonious societal relations.  82 A prominent lesson from the analysis is that the attitudinal terrain is complex.  In fact, all of the primary ingredients of attitudes ? ethnic prejudice, material self-interest, and political values ? have an impact on Canadians? attitudes toward immigration and multiculturalism.  This may be good news for their academic advocates, but it points to the thorny, and somewhat mystifying, nature of attitudes.   However, there are determinants that, on balance, have a stronger impact than others.  Moreover, these determinants differ depending on the ethnic status of the respondent.  For whites, measures of prejudice (or, at least, outgroup resentment) have the strongest influence on attitudes about ethnic diversity.  This finding supports the literature arguing for the preponderant and long-lasting influence of ethnic intolerance (e.g., Kinder and Sanders 1996).  The measures of outgroup resentment are influential for non-whites, as well.  However, the influence of power, the variable with the strongest presence in the white attitudinal terrain, is inconsistent and unexpected.  It has no influence on preferences for a melting pot or a mosaic.  And, less concern about whites? power in society increases non-whites? concerns about immigration.  As I suggested, this could be due some non-whites identifying with whites in terms of their ethnic ingroup. Policymakers concerned with a harmonious society need to be aware of the resentment of perceived outgroup power, particularly for white Canadians.  In fact, addressing its effect may be particularly suited for policymakers.  Compared to the power scale, the analysis of the marriage scale implies that whites who are threatened by ethnic outgroups are less concerned about intergroup scenarios over which they exert more control.  While individuals have little control over interethnic marriage between others, they have complete control over the choice of their own spouse.  This could explain why concerns about interethnic marriage do not seem to influence the attitudinal terrain as much as concerns about outgroup power in society.  The  83 amount of outgroup power, on the other hand, is out of an individual?s control.  Instead, it is the hands of collective actors, such as policymakers.  Given that Canadian policy seeks to encourage ethnic minority engagement in society, and politics in particular, the negative impact of power for white respondent attitudes is troubling.  The challenge for policymakers, then, is to assure members of the ethnic majority that the engagement of ethnic minorities in politics and wider society is not threatening and actually contributes to the health of the democratic system.  In fact, it is heartening that the factor that most influences the attitudinal terrain of whites is a factor over which policymakers may be able to exert considerable influence. Another lesson derived from the analysis pertains to material self-interest.  For the most part, subjective, rather than objective, measures of material self-interest matter for whites.  This finding supports the literature that argues material interests reach beyond an individual?s measurable position in society.  How an individual perceives her position is an important attitudinal predictor and the position may involve relatively short-term calculations.  For non-whites, both objective and subjective measures matter, but only inconsistently.  More work needs to be done to determine exactly why this is the case.   The analysis also hints that subjective material interests and political values influence attitudes that, on their face, focus on social identity.  Specifically, statements about multiculturalism that seem rooted in identity concerns have complex underpinnings that thwart simple explanations.  This finding has implications for policymakers looking to shore up support for ethnic diversity.  Convincing citizens that newcomers are not a threat to their material security may result in positive opinions about overtly cultural issues.  A fruitful  84 policy approach would likely be an emphasis the collective economic benefits of immigrants ? an argument the Canadian government has recently stressed.27   The role of political values suggests other potentially fruitful policies.  White respondents who hold liberal egalitarian values that emphasize social responsibility tend to be supportive of immigration and mosaic-style integration.  Policymakers could, thus, frame the immigration and multicultural policies in a way that highlights egalitarian goals.  Appealing to other policy areas where social responsibility drives much of the political motivation (e.g., Canada?s universal health care policies) could convince wary Canadians to reframe how they think about the ethnic diversity issue.  Non-whites may take away a contrary message, given the influence of this particular value: liberal egalitarian values, at least how they are measured here, increase negative attitudes toward immigration.  However, highly educated non-whites hold positive immigration attitudes (similar to highly educated whites).  So, this may be a more dependable pool of support for policymakers. It is clear that the terrain underpinning Canadian attitudes about immigration and multiculturalism is a complex mix of factors and these factors vary across majority and minority ethnic groups.  Policymakers can be assured, however, that there are multiple ways in which to address fears about ethnic diversity.  To be sure, both white and non-white Canadians are generally positive about ethnic diversity.  But, the analysis here and elsewhere has shown the fissures in the public consensus, most notably among individuals concerned about the distribution of societal power.  The lessons offered here could be used to address weaknesses in the consensus in the hopes of preventing the kind of interethnic violence seen                                                 27 In the spring of 2012, the Minister of Citizenship, Immigration, and Multiculturalism, Jason Kenney, made a series of announcements about the economic benefit of immigrants.  For example, see Jason Kenney, ?Economic Growth and Prosperity the Focus of Immigration Changes.? Accessed January 5, 2013.  85 in other Western democracies.  Identifying and tackling weaknesses in this part of the social fabric could strengthen the health of Canada?s democratic process and the durability of its interethnic harmony into the future.     86 Table 3.1. Individual-Level Determinants of Disagreeing Immigration Has Been a Good Thing for Canada and Preferring a Melting Pot to a Mosaic   Immigration Mosaic  White Non-White White Non-White power 0.819*** -0.588*** 0.705*** -0.125  (0.0669) (0.176) (0.0573) (0.139) marriage 0.394*** 0.601*** 0.210*** 0.343***  (0.0502) (0.154) (0.0486) (0.128) security -5.44e-05 0.289 0.176*** 0.443**  (0.0691) (0.194) (0.0658) (0.193) satisfaction -0.195*** -0.423*** -0.116*** -0.0572  (0.0427) (0.121) (0.0407) (0.119) chance 0.105*** -0.131* 0.0883*** -0.0334  (0.0257) (0.0693) (0.0232) (0.0681) education -0.235*** -0.375*** -0.151*** -0.278***  (0.0250) (0.0705) (0.0245) (0.0686) Observations 4,406 557 4,407 557 Cell entries are ordered logit coefficients; Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1  Independent Variable Scaling: 1) power: too little power=0/too much power=2, 2) marriage: too little power=0/too much power=2, 3) security: low security=0/high security=1, 4) satisfaction: low satisfaction=0/high satisfaction=2, 5) chance: strongly disagree=1/strongly agree=5, 6) education: high school=1/university degree=6  Table 3.2. Marginal Changes for Predicted Strong Disagreement with Immigration and Preferences for a Melting Pot    Immigration Mosaic  White Non-White White Non-White power 0.0448*** -0.0411*** 0.154*** -0.0241  (0.00448) (0.0130) (0.0119) (0.0268) marriage 0.0216*** 0.0420*** 0.0459*** 0.0661***  (0.00300) (0.0118) (0.0105) (0.0242) security -2.97e-06 0.0202 0.0383*** 0.0854**  (0.00378) (0.0139) (0.0144) (0.0371) satisfaction -0.0107*** -0.0295*** -0.0252*** -0.0110  (0.00239) (0.00898) (0.00886) (0.0230) chance 0.00575*** -0.00918* 0.0193*** -0.00644  (0.00144) (0.00496) (0.00504) (0.0131) education -0.0129*** -0.0262*** -0.0330*** -0.0536***  (0.00152) (0.00583) (0.00528) (0.0130) Observations 4,406 557 4,407 557 Cell entries are marginal changes; Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1    87 Figure 3.1ai-bii. Predicted Probabilities after Extreme Changes in Prejudice, Material Self-Interest and Political Values for Whites and Non-Whites  a. Predicting strong disagreement that immigration is generally good for Canada i. Whites ii. Non-Whites    b. Predicting a preference for a melting pot  i. White ii. Non-White      . Probabilities1 2Attitude changepowermarriagesatisfactionchanceeducation0. Probabilities1 2Attitude changepowermarriagesatisfactionchanceeducation. Probabilities1 2Attitude changepowermarriagesecuritysatisfactionchanceeducation. Probabilities1 2Attitude changemarriagesecurityeducation 88 CHAPTER FOUR DETERMINANTS OF WHITE AND NON-WHITE ATTITUDES ABOUT ETHNIC DIVERSITY IN CANADA: THE ROLE OF CONTEXT  How does an individual?s lived context influence her attitudes about ethnic diversity?  And does this influence depend on whether she is from the ethnic majority or an ethnic minority group?  The current chapter is an extension of the analysis found in Chapter Three.  Here, I focus on possible contextual effects on Canadians? attitudinal terrain.  Canada?s relative interethnic harmony largely depends on understanding how people?s lived contexts affect their attitudes.  The diversification of Canadian society forces this issue onto the research agenda.  Fortunately, the Canadian literature has begun to devote attention to the effect of context on attitudes about ethnic diversity (e.g., Blake 2003; Soroka et al. 2006; Stolle et al. 2008).  Typical questions ask whether diversity ? or lack of diversity ? in an individual?s neighborhood has an effect on her attitudes, and, if yes, whether this effect is negative or positive.  The question of whether and how an individual?s environment shapes her attitudes, though, is far from settled.  This is true even outside Canada where the literature is considerably more developed (e.g., Pettigrew and Tropp 2006; Putnam 2007). The present chapter is interested in whether an individual?s local context shapes attitudes about immigration and multiculturalism policies.  Moreover, it is interested in whether any contextual effects depend on an individual?s identification with an ethnic minority or the ethnic majority.  For example, does living in an ethnically diverse area increase or decrease support for immigration?  Does it influence preferences for particular cultural integration strategies?  Does it change how individuals think about the obligations of newcomers?   Of course, an individual?s environment is much more than its ethnic  89 composition.  The economic well-being of the area might also shape people?s political and social views.  For example, how might local unemployment rates, a typical measure of an area?s material security, affect attitudes about immigration or multiculturalism?  Do local income rates have similar effects?  And, importantly, do these contextual-level determinants depend on an individual?s ethnic status?  Do whites and non-whites experience their contexts uniquely?  If they do, what does it mean in terms of studying attitudes about ethnic diversity?  What does it mean in terms of formulating multiculturalism policy? As with the third chapter, I rely on the 1995, 2000, and 2005 waves of the Project Canada public opinion series to investigate these questions.  Here, however, I also add relevant contextual data from the 1996, 2001, and 2006 Canadian census on the visible minority population, unemployment rate, and median household income in the respondent?s local area.  I find a complex relationship between an individual?s local context and her attitudes about ethnic diversity.  Notably, the effect of context appears to depend on whether an individual identifies as white or non-white.  In fact, I find no meaningful contextual effects for whites at all.  For non-whites, the results, albeit weak, are more nuanced.  Importantly, there is a hint that non-whites interact with their environments in a way that reflects the assumptions behind Canada?s multiculturalism policies.  Non-whites living in diverse areas appear to hold more positive attitudes about themselves and about integration into the wider national community.  This interpretation of the data is structured theoretically by the social psychological concepts of depersonalization, subjective uncertainty, and self-affirmation. The chapter begins with a brief outline of the debate around context and its effects on attitudes about ethnic diversity.  After, I review the social psychological theory of social identity and intergroup relations, with special attention to the mitigation of outgroup  90 resentment.  I empirically analyze hypotheses gleaned from the debate around the effects of contact with multivariate regression analysis and statistical predictions.  The chapter concludes with a relevant discussion on the relationship between lived contexts and the attitudinal terrain in Canada. Contextual Level Influences on Attitudes about Ethnic Diversity The effect of an individual?s context has taken on a central role in the research into ethnic attitudes.  Two debates inform much of this research: the contact vs. conflict debate and the realistic group conflict theory.28  The contact vs. conflict debate asks whether contact between ethnic groups influences perceptions of one?s group identity and, consequently, one?s attitudes about ethnic diversity.  The contact hypothesis posits that interaction between groups fosters tolerance by overcoming ignorance and wariness of people perceived as different (Ellison and Powers 1994; Kalin 1996; Oliver and Wong 2003; Sigelman and Welch 1993; Welch et al. 2001; Pettigrew 1998).  Thus, intergroup contact is thought to decrease the importance of divisive, subordinate group identities and increase the importance of inclusive, superordinate group identities.  It is argued that certain conditions need to be met for positive contact to occur, such as equal status and cooperative interdependence (e.g., Allport 1954).  However, some research has shown that contact can be quite casual to have a positive effect.  For example, simply knowing about an outgroup?s history is related to less prejudice against that group (Lee 2000; also, Stephan and Stephan 1984).  Regardless of how this positive contact occurs, the different strains of the contact hypothesis agree that positive contact results in perceived group divisions breaking down and a single shared identity forming (e.g., Gaertner et al. 1996; Hornsey and Hogg 2000; Pettigrew and Tropp 2000).                                                   28 The political science literature tends to use the contact vs. conflict debate as a catch-all for any study that looks at the effect of intergroup contact (e.g., Putnam 2007).  Other literatures ? sociology and social psychology, for example ? tend to make the distinction I make here.  91 By contrast, the conflict hypothesis asserts that interaction between groups breeds intolerance (Alesina and La Ferrara 2002; Blumer 1958; Giles and Buckner 1993; Taylor 1998).  The psychological wariness associated with the formation of ingroups and outgroups at the individual level is heightened as one?s ingroup comes into contact with ethnic outgroups.  Intergroup hostility occurs when the outgroup is perceived as a threat to the ingroup?s identity.  According to the hypothesis, intergroup hostility can stem simply from an outgroup increasing in size, but economic and political conditions can serve as triggers, as well (Oliver and Wong 2003; Sniderman et al. 2004).  Poor economic or political conditions can convert previously latent interethnic wariness into outright interethnic hostility.  The source of the hostility may appear to be economic or political, but the root is social psychological.   In an effort to clarify the debate, a meta-analysis in social psychology concluded that the majority of studies provide evidence that intergroup contact reduces prejudice, supporting the contact hypothesis (Pettigrew and Tropp 2006).  This is not to say that intergroup contact does not produce conflict ? the meta-analysis outlines numerous studies that find that it does.  Indeed, some argue that the effect of intergroup contact should change depending on the level of analysis and type of situation being explained.  It has been proposed that real or perceived outgroup threat will be more acute at higher levels of analysis (e.g., Oliver and Mendelberg 2000; Oliver and Wong 2003).  Others have argued the reverse (Arends-T?th and Van De Vijver 2003).  Yet others argue that the conditions of contact matter, such as cooperative or competitive situations, whether or not the opposing groups are perceived as equal, or if the contact is voluntary (Allport 1954; Pettigrew and Tropp 2006).  In a study of college dorm rooms, classrooms, and campuses, Haidt et al (2003) theorize that if students think the  92 function of the group is enhanced by diversity (like their classrooms), diversity is valued.  The specific conditions producing positive outcomes from contact can be difficult to achieve and may be interpreted differently by different groups (Robinson and Preston, 1976; Tropp and Pettigrew 2005).  This may explain why the conflict-contact debate endures.  To add to the puzzle, Putnam (2007) argues that both hypotheses are wrong, favouring a ?hunkering down? hypothesis where individuals become wary of outgroups and their own ingroup in the face of diversity.  Finally, some find no evidence of contextual effects on attitudes about ethnic relations.  Fetzer?s (2000) comparative study turns up no evidence of intergroup contact effects, even though the effects of group-level material interests and identity concerns are supported.  Clearly, the effect of intergroup contact on harmonious relations between ethnic groups is controversial and still worthy of study. The divergent results in the contact-conflict debate may also be partially due to the fundamental role of material interests in intergroup resentments, as argued by realistic group conflict theory.  In fact, the conflict hypothesis is similar to realistic group conflict theory, and the real-world manifestations of their predictions can appear the same (Coenders et al. 2004; Levine and Campbell 1972; Giles and Hertz 1994; Glaser 2003).29  Both theories predict that intergroup contact can produce negative attitudes toward the opposing group.  Whereas conflict hypothesis focuses on prejudice underpinning these attitudes, realistic group conflict theory focuses on material interests.  The theory asserts that clashes between groups, whether                                                 29 In fact, Coenders et al. (2004) point out how these theories are complementary.  Their Ethnic Competition Theory explains how a material struggle can turn into a social psychological one or a social psychological struggle can turn into a material one.  Similarly, Money (1999) argues that the mere presence of immigrants in the area is not enough to influence attitudes, but the presence of immigrants interacted with a rise in economic hardship is.  Fetzer (2000) neglects to make this connection when he finds no evidence that intergroup contact, measured by national immigration intake levels and percentage of foreign-born residents, has any effect on attitudes.  But, he does argue that majority versus minority cultural identity differences (his ?marginality? theory) explains long-term hostility toward immigrants, and economic interests explain short-term fluctuations in hostility toward immigrants.  93 ethnic or not, are driven by a struggle over limited resources.  Identity with an ingroup is a product of shared material interests with the members of the group, which in turn, creates a dependence with that group.  Thus, ingroup-outgroup divisions are formed from a conflict over resources, rather than a psychological need to simplify the world by identifying with a group.  Initially, the theory examined real conflict, but was later reformulated to include perceived conflict, as well (e.g., Blalock 1967).   The well-known Robbers Cave experiments, which spurred the realistic group conflict theory research agenda, demonstrated how relations between relatively equal groups of boys at a summer camp quickly deteriorated once competition over material goods started (Sherif et al. 1961).  Bonacich?s (1972) split labour market theory also demonstrates how the competition between high-skilled and low-skilled labour for jobs can fuel resentment and hostility.  This hostility can easily become ethnic-based if the labour groups are incidentally defined as such (e.g., native, unionized workers versus foreign, undocumented workers).  These resources can be overtly material, like financial wealth or tangible goods.  For example, a comparative study of the United States, France, and Germany found that the real disposable income per capita was inversely related to anti-immigrant attitudes (Fetzer 2000).  A Canadian study found that anti-immigrant sentiment was positively associated with the national unemployment rate, a clear link between the material interests of the majority group and attitudes about minorities (Wilkes et al. 2008).  But, resources can also be symbolic, such as leisure time, education, and social access (Bobo 1983; Sidanius et al. 1996).  The realistic group conflict theory asserts that prejudice is more than just the negative attitudes, feelings, and behaviour that are the focus of identity theories.  Instead, prejudice is fueled by the conflicts that arise when a group perceives a threat to accepted ways of life from another  94 group.  When the group is ethnically distinct, conflict may appear to be motivated by individual-level ethnic prejudice, but it is truly motivated by a need to protect the material status quo.30 Ingroup Size and Subjective Uncertainty In the third chapter, I outline several reasons why attitudes held by the majority ethnic group in a society might be different than attitudes held by a minority ethnic group, emphasizing the need to investigate them separately.  I touch on reasons, such as the perceived or real redistribution of social and political advantage to ethnic minorities that is implicit in multiculturalism policies.  Also, at the micro-level, minorities may develop different strategies for intergroup interaction simply because, by virtue of being in the minority, they are more used to encountering ethnic outgroups.  (see Chapter Three for a more detailed exposition). Tying together questions of contextual effects and the difference between majority and minority ethnic groups implicates questions of ingroup size.  What might be the consequences of different group sizes for majority and minority individuals?  It is common to look at the size of groups in contextual analyses, but few studies go beyond hypothesizing an additive effect of group size.  The size of the group, though, has more nuanced implications than                                                 30 There is a large literature arising at the intersection of realistic group conflict theory and the social psychological theories of intergroup conflict looking at if and how ethnic diversity affects popular support of the welfare state (for a summary, see Kymlicka and Banting 2006).  Typically, this literature?s empirical focus is institutional, as it looks at how immigration and/or multiculturalism policies influence welfare state policies.  Theoretically, it frames the effect of diversity as rooted in nationalistic sentiments, with increased diversity eroding the national cohesion that underpins public support for redistribution.  As such, the theory depends on the same group-level identity concerns as the contact-conflict debate.  It also implicates materialistic concerns, as native citizens are ostensibly reacting to a perceived threat to the material status quo.  While the literature does not typically locate itself in the realistic group conflict theory, it could plausibly do so, as any threat to redistribution (i.e., not just immigrants) could plausibly erode support for the welfare state.  95 simply increasing or decreasing a particular association.  In Canada where the amount of diversity changes considerably between communities, it is an important factor to consider.31   Social psychology posits that individuals join groups in an effort to reduce ?subjective uncertainty?, or a feeling that one?s environment cannot be controlled or predicted (e.g., Grieve and Hogg 1999; Hogg 2000).  Joining groups can reduce subjective uncertainty because they give cues to individuals about how to behave and think, providing a sense of order and control in the world.  For the current analysis, there are two important outcomes from the theory of subjective uncertainty.  The first is that the reduction of subjective uncertainty by self-identifying with a group results in ?self-affirmation?.  The group is chosen primarily because its norms and practices affirm the individual?s own worldview, and provides a psychological sense of being ?correct? (Hogg and Abrams 1993).  Self-affirmation protects an individual?s sense of self by legitimizing the categorizations she has made about who she thinks she is (Wichman 2010).  Secondly, if subjective uncertainty is reduced under specific circumstances, the bias generated by ingroup-outgroup formations can be mitigated.  Grieve and Hogg (1999) find that intergroup hostility can be reduced if subjective uncertainty is sufficiently low before individuals join groups.  In their experiments, giving subjects minimal information about the experimental task, as opposed to keeping the subjects ignorant, significantly reduced ingroup bias and outgroup resentment.   The real-world implication is that environments that offer individuals sufficient information about how they can fit in and achieve a sense of correctness will mitigate intergroup hostility.  Individuals will still identify with groups, but the need to reduce uncertainty is not as urgent and, thus, will not produce the hostility typically seen with the                                                 31 Many communities in Canada have little to no ethnic diversity.  The most diverse community, according to the 2006 census, is Vancouver BC.  The census reports that 51 percent of the city?s residents identify as non-white.   96 formation of groups.  While these social psychological experiments are conducted on small groups in artificial laboratory situations, their conclusions may be applicable to intergroup interactions at the local level.  It is proposed here that larger potential ingroups in an individual?s local area might provide such real-world conditions.  Compared to a small group, a large group that has the potential to be perceived as an ingroup by an individual would provide immediate, familiar information about the environment and an extensive basis for reducing subjective uncertainty.  An ethnic minority moving to a community where ethnic minorities compose 10 percent of the population would likely be more certain about their place in society and their control over the environment than if she moved to a community where ethnic minorities composed only one percent.  This dynamic is likely aided by the fact that many ethnic minorities in Canada concentrate in enclaves (e.g., Fong and Wilkes 2003; Hiebert et al. 2007). Notably, ethnic minorities do not have to be from an individual?s own ethnic group to have self-affirming effects.  Self-categorization theory posits that part of forming a collective identity involves going through ?depersonalization?, or a process where an individual perceives the members of her ingroup as interchangeable at some level (Hogg 2001).  This psychological process creates matching stereotypes for the salient ingroup and the individual herself.  The matching stereotypes, in turn, accentuate similarities between the individual and the salient ingroup, justifying her identification with the group.  These matching stereotypes can shift as the context shifts, making other dimensions of one?s social identity salient (Brewer 1991).  For example, people can experience this process when travelling in a foreign country.  The instant bond that can be felt with compatriots, even if nationality is not typically  97 a salient dimension in their self identity when at home, illustrates this social psychological effect.   I am suggesting here that ethnic minorities may reduce subjective uncertainty and gain self-affirmation by being around ethnic minorities in general and not necessarily from their own ethnic ingroup.  They will emphasize similarities with this new group to compensate for the absence of their own, more proximate, ethnic ingroup.  As Allport states, social identity can be envisioned as concentric rings, where the most important are closest to the self.  If one of these rings becomes unsustainable (e.g., there are no salient compatriots in the area), an individual will expand her social identity outwards to a more distant, yet achievable ring (Allport 1954).  Thus, if a particular ethnicity and being an ethnic minority are two nested, salient social identities for an ethnic minority individual, she will be motivated to identify with other ethnic minorities if there are no other members from her ethnic ingroup in the area.  An individual may shift her identity from being, say, Congolese to being African if there are no other Congolese in the area, but there are other Africans, for example.  So, she may come to identify with ethnic minorities outside her more proximate ingroup as a way to address the need of reducing subjective uncertainty.  This is not something I intend to directly test.  But it does shape expectations for ethnic minority attitudes in areas where the precise ethnic composition of the community is unknown. What might occur if an ethnic minority individual lives in an area that is homogenously characterized by the majority culture?  What would the Congolese woman do if she lived in a nearly all-white community and could not shift her ingroup identity to the African community?   The social psychology literature outlines two choices shaped by the basic human need to reduce subjective uncertainty by identifying with groups and  98 differentiating from others.  The Congolese woman may self-segregate from the wider white community.  That is, she would retain her ethnic minority identity and limit interaction with the majority group (Berry 2001).  However, she may choose to give up her ethnic identity and assimilate into the wider white community.  Her basic need to identify with a group may push her to adopt the majority white community as her ingroup.  As mentioned previously, adopting an ingroup identity requires depersonalization, or creating matching stereotypes for the individual and the prospective ingroup (Turner 1999).  So, an ethnic minority individual in a homogenous area may create matching stereotypes for herself and the majority group, requiring her to give up her minority identity and assume the majority identity.  Choosing to assimilate, then, signals a desire for a minority to ?fit in? with the majority group. From the viewpoint of a majority group member, the advantages of ingroup membership, most notably intragroup trust and cooperation, are typically bestowed to other ingroup members.  For example, whites will be more likely to socialize with other whites, simply because they are recognized as part of the ingroup (e.g., McPherson et al. 2001).  As such, symbols that differentiate the ingroup from outgroups become salient, since they limit unintentional distribution of ingroup benefits to outgroup members (Brewer 1999).  The Congolese woman living in a homogenously white area would be motivated to assimilate as a way of minimizing the possibility that the majority will perceive her as different and, thus, exclude her from the benefits of being an ingroup member.  She would not be able to adopt some of the ingroup symbols (e.g., physical appearance), but she could adopt cultural practices (e.g., celebrating Christmas if she is non-Christian).  In fact, so-called peripheral ingroup members have been found to be more derogatory of outgroups in public compared to core members, because of their need to prove themselves as ?one of the group? and, thus,  99 distinct from outgroup members (Noel et al. 1995).  So, an ethnic minority living in a white community might have a psychological need to assimilate into the majority, and also be willing to publicly express this need. What about the majority group member?  She already has a considerable basis for subjective certainty simply by virtue of belonging to the majority.  The majority?s norms and practices dominate the environment, providing a psychological sense of correctness and self-affirmation from birth.  Experiments have shown that, if only looking at numerical size, majority groups show less ingroup bias than minority groups, signaling more security in their majority ingroup identity (Mullen et al. 1999).  To be sure, majority group members divide themselves into ingroups and outgroups.  But, at the most general level of a national ethnicity, majority group members have a psychological advantage over minority group members.  Thus, compared to minority members, it is likely that majority members will not be as sensitive to the contextual effects of ethnic diversity.  The small shifts in the local ethnic composition will not influence majority attitudes to the same extent as minority attitudes.   Of course, the contact vs. conflict debate is predicated on the notion that contextual diversity does shape majority attitudes.  However, the debate is far from settled, with each approach claiming the opposite outcomes of intergroup contact.  Research that finds no contextual effects may look to the social psychological explanation outlined above as a way of interpreting non-findings.  It may be that, in certain situations or political climates, other factors will change the weighting of contextual influence.  Intergroup contact occurring in Canada may be such a situation.  For example, there is evidence that Canada?s strong multiculturalism policies resolve the ?progressive?s dilemma? and protect the county?s redistributive welfare policies despite high levels of ethnic diversity (e.g., Banting et al. 2006).   100 Indeed, the reciprocal nature of Canadian multiculturalism policy emphasizes the worth of both minority cultures and the majority culture.  Moreover, Canada is a behavioural outlier in terms of its generally supportive public attitudes toward immigrants and ethnic difference (Hiebert 2006).  In such an environment, where both majority and minority group identities are affirmed through government policy and social norms, majority group individuals may have less of an urgency to ?assimilate and differentiate? compared to environments where majority and minority identities are typically treated as zero-sum (Brewer 1999; Hewstone et al. 2002).  Data and Methods The role of contextual factors, and how they may vary between majority and minority ethnic groups, are the focus of the present empirical investigation into the terrain of ethnic attitudes in the Canadian context.  Two of the attitudes to be explained are the same as in the third chapter: statements about whether immigration to Canada is generally good (immigration) and preferences for a mosaic or melting pot approach to cultural integration (mosaic).  I also add the dependent variable obligations.  In terms of attitudinal specificity, this variable falls in between immigration and mosaic.  The survey question, from which it is derived, asks whether the respondent agrees or disagrees that immigrants ?have an obligation to learn Canadian ways.?  High scores indicate the respondent strongly agrees with the statement, implying that she embraces a compulsory cultural integration of immigrants into a greater Canadian identity.  This question asks specifically about a particular aspect of immigrant integration ? i.e., acculturation, or how an immigrant?s way of life can change due to contact with the host society.  It does not, however, ask if immigrants should give up their own ways.  It also does not specify which Canadian ways immigrants should be learning.  Thus, the  101 question leaves considerable room for interpretation by the respondent.  The distribution of obligations shows a strong consensus: almost 90 percent agree or strongly agree that immigrants have an obligation to learn Canadian ways, and only 10 percent disagree or disagree strongly. 32  The consensus is not surprising: it makes sense that even the most tolerant Canadians want newcomers to try to fit in.  This consensus does not render the indicator unusable, since illuminating differences are likely at the positive end of the response options.  While 42 percent of respondents strongly agreed immigrants have obligations to learn Canadian ways, almost half of those asked (48 per cent) stated they only somewhat agreed.  This distribution reflects important attitudinal differences with the former being rigid in their desire for immigrants to integrate into the mainstream.  It shows little tolerance for immigrants who decide to keep separate from the majority for a variety of reasons, such as distrust of the majority group (Berry 2001). Contextual-Level Independent Variables As in the third chapter, I am using the 1995, 2000, and 2005 waves of the Project Canada public opinion survey.33  Importantly, the Project Canada data identifies the respondent?s geographic area.  This area variable is a mixture of census metropolitan areas (e.g., Toronto, ON), census agglomerations (e.g., Cranbrook, BC), and census subdivisions (e.g., Selkirk,                                                 32 There was concern that the indicator may not be measuring an increasing intolerance of self-imposed segregation by immigrant groups.  The low end of the variable, or disagreeing that immigrants have obligations to learn Canadian ways, may indicate opposition to immigrants integrating into the Canadian mainstream, fueled by a dislike of immigrants in general rather than an extreme acceptance of self-imposed segregation.  To see if this was a possibility, I tested the association between obligations and other variables believed to measure general dislike of ethnic minorities, namely the extent to which the respondent feels uneasy around various identifiable ethnic minorities (Blacks, East Indians and Pakistanis, and Asians).  An association between the variables would cast doubt on the proposition that disagreeing that immigrants have obligations is being driven by positive attitudes toward immigrants.  Crosstabulations show that few respondents who disagree that immigrants have obligations also feel uneasy around ethnic minorities.  The largest association (3.9%) is between obligations and feeling uneasy around East Indians and Pakistanis.  It is clear that the variable is measuring ethnic attitudes as theorized.  33 Reginald W. Bibby, Project Canada Survey Series, 1995, 2000, 2005.  University of Lethbridge.  102 MB).  As such, the size of the respondent?s local context varies considerably.  Still, I was able to match the Project Canada data with population numbers from the 1996, 2001, and 2006 Canadian census, which should help control any variance related to population size. The geographic identifier in the Project Canada data also allowed census data on visible minority populations, unemployment rates, and household income levels to be matched to the respondent?s local context.  In the following analysis, these contextual variables measure the identity threat implicated in the contact-conflict debate and the material threat implicated in the realistic group conflict theory.   The variable, minorities, is a continuous measurement of the percentage of visible minorities living in the respondent?s geographic area (see Appendix Three for descriptive statistics).34  Minorities is also interacted with the binary variable for the respondent?s ethnicity, called white.  The interaction is to test my expectation that contextual effects will depend on the ethnicity of the respondent.  As a reminder, contact theory predicts that interethnic contact will improve ethnic attitudes, while conflict theory predicts the opposite.  However, my expectations are more complex given the theoretical discussion of the social psychological concepts of subjective uncertainty, self-affirmation, and depersonalization.  I expect that larger visible minority communities are able to provide non-whites with an immediate sense of control, creating the conditions for lower subjective uncertainty and, thus, lower outgroup resentment.  Depersonalization and the shifting nature of the social self means that non-whites can glean self-affirmation from ethnic minorities outside of their immediate ethnic ingroup.  As such, I expect that non-white respondents will have more positive ethnic                                                 34 The term ?visible minority? is officially defined in the Canadian government?s Employment Equity Act.  It defines visible minorities as ?persons, other than Aboriginal peoples, who are non-Caucasian in race or non-white in colour?.  See Accessed June 1, 2012.  103 attitudes in larger visible minority communities, regardless of the ethnic origins of these communities.   For whites, larger visible minority communities may be associated with more negative ethnic attitudes, as predicted by the conflict hypothesis.  In terms of social psychology, a larger visible minority presence could create more subjective uncertainty, thus result in more outgroup hostility.  On the other hand, the increased opportunity for positive contact with visible minorities that accompany larger visible minority presence may produce positive ethnic attitudes for whites, as predicted by the contact hypothesis.  It is also possible that larger visible minority communities will have little, if any, impact on white attitudes because whites are still the majority ethnicity in almost all the localities investigated here.  As mentioned, social psychology offers theoretical reasons for expecting a smaller contextual effect for whites: being in the majority provides considerable subjective certainty, compared to minorities (e.g., Brewer 1999; Grieve and Hogg 1999).  Moreover, the mediating impact of Canada?s multiculturalism norms and policies could further insulate whites from contextual effects, at least as they are measured here.  Regardless of the direction or strength of the relationship, it is expected that, at a minimum, the contextual effects of the local visible minority population will be dependent on the ethnic majority or minority status of the respondent.   Contextual measures of material interests are captured with the indicator material.  Realistic group conflict theory predicts that, all other things being equal, people from poorer neighborhoods will have more negative attitudes about ethnic diversity than people from richer neighborhoods.  People from poorer neighborhoods, regardless of their own financial situation, will sense more vulnerability to material competition in the form of disadvantageous  104 ethnic policies or immigrants perceived as increased competition in the job market.  People from more wealthy neighborhoods should not feel as vulnerable, regardless of their own finances.  The measure of material group interests ? material ? is a dichotomous index that categorizes levels of material security using the typical measures of household income and employment status.  Lower median household income and higher unemployment (i.e., low security) is ?0? and higher median household income and lower unemployment (i.e. high security) is ?1?.  So, if a respondent?s material context is associated with her ethnic attitudes, the analysis should find that respondents living in areas of low economic security should have more negative ethnic attitudes than respondents living in areas of high economic security. As mentioned, I add a variable controlling for the population of the respondent?s local context.  The population of the local contexts represented in the individual-level data vary considerably: the minimum is just over 200, the maximum is about 2.5 million, and the median is about 100 500.  As such, I control for this variation with population, which indicates if the localities are small (less than 10 000), medium (from 10 000 to 99 999), or large (over 100 000). 35  I also control for survey year (year) to account for possible period effects. Results Table 4.1 displays three ordered logistic regressions corresponding to the three dependent variables - whether or not immigration is good (immigration), whether or not immigrants are obligated to learn Canadian ways (obligations), and if the respondent prefers a mosaic or melting pot approach to ethnic group integration (mosaic).  The focus here is on the estimates                                                 35 There was a concern that the minorities variable would depend on community size, as interaction with outgroups may be more likely in smaller communities.  An interaction between minorities, population, and white was not significant, indicating that the influence of the visible minority population is not dependent on the size of the community.  105 for the effects of visible minority presence and local economic situation in the respondent?s geographic area.  However, I also include most of the individual-level variables from the third chapter as a way to provide a fuller picture of the respondent?s attitudinal terrain.  The variables marriage, security, and chance are dropped to simplify the analysis.  Their conceptual partners (i.e., power, satisfaction, and education) are retained.36  The individual-level variables are no longer interacted with the ethnic identity variable, again, for simplicity.  The individual-level controls (age and sex) and the contextual-level controls (population and year) are included in the calculations, but are not shown. I will not dwell on the individual-level associations; their analysis can be found in Chapter 3.  One quick observation can be made, though, between the relatively inconsistent influence of the contextual measures compared to the individual ones.  In fact, the measure of local unemployment rate and median household income (material) has no influence on the attitudes measured with the dependent variables.  This is in contrast with work in Canada and elsewhere that finds evidence for contextual material effects on attitudes about ethnic diversity and immigrants (e.g., Palmer 1996; Wilkes et al. 2008).  These studies use national indicators, rather than community-level ones, which may account for the difference in results. As for the measure of percentage of local visible minorities (minorities), it is associated with immigration, obligations, and mosaic.  The associations are statistically significant, but small.  Theoretically, though, the effect of visible minorities in the areas is expected to depend on the respondent?s ethnic identity, so the significant interactions are encouraging.   Predicted probabilities can be used to clarify the relationships observed in the table.  By setting variables of interest at particular levels, the precise nature of any possible effect can                                                 36 Omitting these variables did not meaningfully change the results.  106 be better examined.  Figure 4.1a-c displays predicted probabilities for a hypothetical white and non-white individual and their ethnic attitudes at five percent intervals of minorities.  The intervals are on the x-axis and run from zero to 50 percent. 37  As in Chapter Three, the individual-level variables ?  power, satisfaction, and education ? have been set at the maximum negative values to indicate an individual who plausibly holds extreme negative attitudes toward immigration, immigrant obligations, and multiculturalism.  My motivation for this manipulation is because I am interested to see if and how these individuals improve their attitudes.  See Chapter Three for details.  Recall that minorities measures the percentage of visible minorities living in the respondent?s area.  Since material, the measure of material group interest was not significant, I do not investigate it further. The figures offer a heterogeneous interpretation of the impact of context on white and non-white attitudes.  For whites, local ethnic diversity has virtually no impact of the attitudinal terrain.  There is simply no consistent evidence that points to either the contact hypothesis or the conflict hypothesis.  Notably, the associations between context and immigration, obligations, and mosaic do not notably improve when the individual-level variables are omitted from the model (not shown).  So, it is not a case of the individual-level variables obscuring subtle contextual effects for whites.  It appears, in the present analysis at least, that white attitudes about immigration, immigrant obligations, and multiculturalism are primarily determined at the individual level.  As suggested earlier, majority individuals already have a considerable basis for subjective certainty simply by virtue of belonging to the majority.                                                  37 While the census data categorizes some areas as having zero visible minorities, the Project Canada data has non-whites living in the same areas.  In fact, about 18% of the non-white respondents in the Project Canada data are living in an area that the census characterizes as having less than one percent visible minority presence.  The discrepancy is likely do to different data collection procedures.  107 Given that whites are the majority in almost all the areas studied here, white respondents may largely be insulated to the contextual effects of ethnic diversity in their local communities.  For non-whites, there is some evidence for contextual effects.  The evidence is not strong, but it is worth contemplating.  The first panel (Figure 4.1a) displays predicted negative attitudes about immigration (high scores on immigration).  It shows that a higher percentage of the local visible minority community is associated with a lower probability that non-whites will strongly disagree immigration is generally good.  That is, more visible minorities in the area appear to improve non-white attitudes about immigration.  The second panel (Figure 4.1b) presents predicted attitudes about immigrants being obligated to learn Canadian ways.  It shows that, in the most diverse areas, non-whites have the highest probability of strongly agreeing that immigrants must learn Canadian ways.  The third panel (Figure 4.1c) predicts preferences for a melting pot-style of cultural integration.  Here, more diversity on the ground is associated with a lower preference for a melting pot.   The patterns observed in Figure 4.1a-c offer a thought-provoking picture about the impact of ethnic diversity of a non-white?s context and her attitudes about ethnic diversity.  I interpret these results with caution, because the evidence is not overwhelming.  But, there is a hint of a social psychological process occurring that should be noted because of its potentially significant policy impact.  Interpreting Figures 4.1a and 4.1b together suggest a pattern similar to what social psychology predicts when it looks at social identity formation, which it typically does in smaller settings.  The process, involving subjective uncertainty, self-affirmation, and depersonalization, is discussed extensively above (e.g., Grieve and Hogg 1999; Wichman 2010).  These two figures suggest that larger visible minority communities are associated with more confident, self-affirmed non-white individuals, and these individuals  108 display higher ingroup bias and lower outgroup resentment.  Specifically, at high percentages of visible minorities, they are more likely to agree immigration is good (higher ingroup bias) and agree that immigrants have an obligation to learn Canadian ways (lower outgroup resentment).38 The third panel displaying preferences for melting pot-style cultural integration (Figure 4.1c) suggests that, if the social psychological process described above is occurring, the cumulative result is what policymakers and multiculturalism advocates intend.  The figure shows that non-whites are more likely to prefer a mosaic to a melting pot in highly diverse areas.  A higher preference for a mosaic is what one would expect if non-whites living amongst larger visible minority communities reduce their subjective certainty and increase their self-affirmation.  As for the left-hand side of the non-white trend line in Figure 4.1c, it appears that non-whites living in highly homogenous, white communities tend to choose to assimilate rather than self-segregate.  That is, they are much more likely to prefer a melting pot to a mosaic than non-whites living in ethnically diverse areas.  This suggests that, as above, they are shifting their ingroup identity from an ethnic minority to the ethnic majority.  Policy implications of this interpretation are discussed below. Discussion The present paper expands on the individual-level analysis in the third chapter.  Using the same data, it focuses on the effects of an individual?s lived context on her attitudes about                                                 38 There is additional support for this interpretation in the data.  Attitudes that measure self-confidence and a positive outlook in the broader social environment were associated with local context in an expected way if non-whites outgroup resentment decreases in large visible minority communities.  For instance, the predicted probability of non-whites strongly agreeing that ?the lot of people is getting worse, not better? decreases at higher levels of the contextual visible minority variable.  Also, strongly agreeing Canada is generally a ?highly compassionate country? increases.  And, perhaps the most telling, the predicted probabilities that they are ?very happy? these days increases as their context is characterized by more visible minorities.       109 immigration, immigrant obligations, and cultural integration strategies.  I also investigate the possibility that whites and non-whites interact with their contexts in unique ways.  There are good theoretical reasons to think that they will, and the analysis finds some evidence for this assertion. As mentioned, caution is required when assessing the results here.  Some of the results are straightforward.  For example, the indicator of material group interest has no influence on the attitudinal terrain.  The lack of influence reflects the inconsistent material measures at the individual level, as well.  (See Chapter Three for details.)  Moreover, the results counter recent Canadian research finding contextual economic effects on attitudes toward immigration, supporting realistic group conflict theory (e.g., Palmer 1996; Wilkes et al. 2008).  This research, however, uses national economic indicators, while I use neighborhood-level ones.  The measurement difference, in fact, may prompt a new hypothesis: Canadians pay less attention to local group interests compared to national group interests. Also, it is clear that white attitudes are not swayed by local ethnic diversity.  There is no meaningful association between the percentage of visible minorities in the area and whites? attitudes about immigration, immigrant obligations, and multiculturalism.  This is in contrast with previous Canadian work that has found contextual effects for whites (Blake 2003; Stolle et al. 2008).  These studies, however, are seeking to explain different attitudes and behaviour.  They also use different geographical sizes.  Given my results in Chapter Three, the source of white attitudes really appears to be at the individual-level, at least for the attitudes being investigated here.  This result may be particular to the Canadian setting, as its history of strong multiculturalism has been previously linked to comparatively positive attitudes about ethnic  110 diversity (e.g., Hiebert 2006) and the buttressing of redistributive policies against backlashes in public opinion (Banting et al. 2006).   While the associations between the contextual measures and the dependent variables are weak, I am not ready to dismiss contextual effects as some previous research has (e.g., Fetzer 2000).  My analysis produces associations between the visible minority contextual variables and non-white attitudes.  As discussed, the associations suggest that non-whites living in areas that are ethnically diverse have more ingroup bias and less outgroup resentment.  This observation supports the social psychological theories of subjective uncertainty, self-affirmation, and depersonalization (e.g., Grieve and Hogg 1999).  That is, living in an ethnically diverse area provides a greater opportunity for non-whites to confirm that their social identity and their worldview is ?correct?, thus, reducing subjective uncertainty.  Consequently, they become self-affirmed, and if they join another ethnic ingroup, they do some from a position of relative psychological security.  As such, their ingroup bias increases and their outgroup resentment decreases.  The higher preference for a mosaic at higher levels of local ethnic diversity also points to this interpretation of a confident ethnic minority individual who is willing to engage with the ethnic majority. If this social psychological process is actually occurring in diverse Canadian neighborhoods, the analysis here lends support to the assumptions behind Canada?s multiculturalism policy.  Namely, one of the central goals of multiculturalism policy is to protect and promote minority cultures as a way of creating confident and engaged ethnic minority citizens.  The idea is that individuals who feel secure in their cultural background will be more willing to integrate and identify with the national community (e.g., Abu-Laban and Stasiulis 1992).  These assumptions are elaborated through influential normative theory,  111 spearheaded in the Canadian context, about the importance of multicultural policies and values in an ethnically diverse, liberal democratic society.  The policies and values prescribed by these theorists stress the importance of recognizing multiple ways in which groups identify with the nation (Taylor 1992) and the necessity of ethnic group rights as a way of ensuring individual freedom in liberal democracies (Kymlicka 1989).  Individual freedom, achieved through ?cultural freedom?, is thought to create confident ethnic minorities who are willing to positively integrate into the national community.39  The analysis here offers some evidence that the rationale behind multiculturalism policy is sound.   The one hitch in this optimistic interpretation of non-white attitudes is that non-whites prefer a melting pot in homogenously white areas.  Certainly, policymakers intend for all ethnic minorities to gain confidence through multiculturalism recognition, regardless of the diversity in their area.  The results here show that minorities living in homogenously white areas still prefer to assimilate.  This finding indicates a behavioural limit to multiculturalism policy: the basic psychological need to identify with groups and differentiate from others means that minorities surrounded by the majority will tend to give up their ethnic minority identity to ?fit in?.  It may be that ethnic minorities are happy to assimilate.  If so, the policy goal of a strong national community would be achieved, at least with these individuals.  But, a strong national community through assimilation is not what multiculturalism policy intends.  And, ethnic minorities might not be happy to assimilate.  Given the theoretical viewpoint that ethnic minorities are often limited in the majority ingroup symbols they can adopt (e.g., physical appearance vs. cultural practice), they may find themselves relegated to the periphery                                                 39 The idea that cultural freedom, ensured by official recognition, will result in confident and integrated ethnic minority citizens comes from Trudeau?s 1971 multicultural policy announcement. Accessed June 1, 2013.  112 of the majority ingroup if they are accepted into the ingroup at all.  In this case, assimilation would not provide the same psychological security as the original ethnic ingroup identity.  Thus, assimilation would not produce happy minority citizens.  Unfortunately, the data does not allow for a test the mindset of minorities who choose to assimilate.   Again, though, I interpret these results with caution.  The evidence is not strong.  But the question of how Canadian citizens actually behave in their diverse contexts is a question that should be pursued.  The results here suggest at a minimum that majority Canadians interact with their environment in a substantively different way than minority Canadians.  At a maximum, the results suggest that multiculturalism policy is working, at least mostly, as intended.     113 Table 4.1. Contextual-Level Determinants of Disagreeing Immigration Has Been a Good Thing for Canada, Immigrants are Obligated to Learn Canadian Ways, and Preferring a Melting Pot to a Mosaic   Immigration Obligations Mosaic white -0.153 0.310* 0.217  (0.204) (0.168) (0.155) minorities -0.0124*** 0.0105*** -0.0118**  (0.00289) (0.00274) (0.00517) white*minorities 0.0101* -0.00997*** 0.0137***  (0.00594) (0.00226) (0.00314) material -0.105 -0.0640 0.190  (0.151) (0.157) (0.150) white*material 0.115 0.00752 -0.191  (0.209) (0.125) (0.189) power 0.610*** 0.541*** 0.633***  (0.0919) (0.123) (0.0702) satisfaction -0.192*** -0.178*** -0.133***  (0.0544) (0.0494) (0.0485) education -0.302*** -0.173*** -0.159***  (0.0611) (0.0440) (0.0387) Observations 3,510 3,472 3,524 Cell entries are ordered logistic coefficients; Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1  Independent Variable Scaling: 1) white: non-white=0/white=1, 2) minorities: % visible minorities, 3) material: low security=0/high security=1, 4) power: too little power=0/too much power=2, 5) satisfaction: low satisfaction=0/high satisfaction=2, 6) education: high school=1/university degree=6     114 Figure 4.1a-c. Predicted Probabilities after Changes in the Local Visible  Minority Presence for White and Non-White Respondents  a. Strongly disagree that immigration is a good thing.    b. Strongly agree that immigrants have an obligation to learn Canadian ways.   c. Prefer a melting pot to a mosaic.       . of strongly disagree0% 50%Percentage of visible minoritieswhite non-white. of strongly agree0% 50%Percentage of visible minoritieswhite non-white. of preferring melting pot0% 50%Percentage of visible minoritieswhite non-white 115 CHAPTER FIVE PERSONALITY, POLITICAL VALUES, AND ATTITUDES  ABOUT ETHNIC MINORITIES IN CANADA  Decades of research in psychology have built the consensus that personality influences prejudice.  Recent breakthroughs in the so-called Big Five personality factors refine this idea by establishing that the personality dimensions Openness to Experience and Agreeableness consistently influence attitudes toward others perceived as different (Duckitt 2001).  What has not been established is how personality influences attitudes toward public policies regulating ethnic relations.  Past research has identified numerous determinants that shape political attitudes about these policies, such as values, material self-interest, political sophistication, and media consumption.  But, the fact that political attitudes are fundamentally shaped by personality is often overlooked in the political science literature.  As Mondak (2010) states, citizens are not simply ?blank slates? onto which political determinants imprint.  Personalities influence how these political determinants shape the attitudes political scientists investigate.  The focus of the current paper is on the influence of Openness to Experience and Agreeableness personality dimensions on individual ethnic attitudes.  These dimensions are consistently linked to tolerant attitudes and behaviour in the psychological literature (e.g., Sibley and Duckitt 2008).  Is their influence in the real world as predicted in the smaller-scale, more intimate scenarios examined by psychologists?  Are these personality dimensions stable influences on ethnic policy attitudes or can they be overridden?  What happens when negative stereotypes are encountered?  Are personalities predisposed to tolerance able to disregard stereotypes?  Or do stereotypes affect these personalities in a similar manner as personalities  116 predisposed to intolerance?  What other determinants might help these typically tolerant individuals overcome any negative reactions to express positive ethnic attitudes more aligned with their personalities?  Research on citizens? attitudes about policies related to ethnic relations rarely looks at the link between citizens? personality and their policy preferences.  This study endeavors to do so in a Canadian setting.  Amongst Western democracies, Canada is a particularly interesting case where ethnic diversity policies are concerned.  The country has been de facto ethnically diverse since its inception in 1867, with sizable English, French, and Aboriginal populations.  In 1971, de facto ethnic diversity became a political reality when Canada became the first country to adopt an official policy of multiculturalism.  Since then, millions of tax dollars have been dedicated annually to federal multiculturalism programs, and a large policy framework has been built with the objective of interethnic understanding, civic participation, and interaction between Canada?s disparate ethnic groups (Abu-Laban and Stasiulis 1992).  Moreover, it is widely acknowledged that Canadians are proud of their multicultural society, and a normative consensus has emerged amongst the public and political leaders about the value of ethnic diversity (e.g., Adams 2007; Bilodeau et al. 2012).  However, cracks in the consensus exist (e.g., Bissoondath 1994; Seidle 2008).  So, the Canadian context is highly relevant for an analysis attempting to untangle the effects of personality on ethnic diversity policy preferences.  The theoretical and empirical focus here is about the complex relationship between political attitudes and personality traits.  The analysis hopes to contribute to the academic debate in two ways.  First, it asks if the personality traits that consistently predict prejudice in small-scale psychological studies predict ethnic policy attitudes in a large national study of  117 Canadians in their everyday lives.  The analysis finds that the insights of psychological experiments can only go so far.  When a controversial symbol of ethnic difference ? the Muslim hijab ? is primed, personality traits no longer predict tolerant and intolerant attitudes.  The typical gap between personalities prone to prejudice and personalities prone to tolerance closes.  The second goal of the paper is to investigate how this gap can be opened again ? how might individuals with personalities prone to tolerance overcome negative reactions to controversial ethnic symbols and support ethnic policy as they should, given the evidence from experimental psychology?  That is, how can they achieve attitude-personality consistency?  The analysis demonstrates that people with tolerant personalities can achieve attitudinal coherence if they subscribe to liberal egalitarian political values.  Liberal egalitarian values, or valuing the equality of groups, the welfare of others, etc., can override exclusionary reactions of typically tolerant individuals to a potent stereotype of ethnic difference.   An important insight of the analysis, then, is that attitudes expressed by people with tolerant personalities have a dual source: their personality and their values.  When the former fails to produce ethnic tolerance, the latter can do so.  Egalitarian values act as a behavioural ?corrective? when a negative stereotype triggers a mismatch between typically tolerant personalities and ethnic attitudes.  This insight ties into ground-breaking psychological research on the dual-process of behaviour (e.g., Devine 1989).  The data for the empirical analysis come from the 2011 Canadian Election Study (CES).  The national public opinion survey contains a unique experiment on multicultural policy attitudes and ethnic cues.   The CES also includes measures of the Big Five personality dimensions ? a rarity in Canadian public opinion polls.  Together, along with a typical  118 indicator of liberal egalitarian values, these measures allow a statistical analysis of dual nature of ethnic attitudes, driven by personality and political values and shaped by highly salient ethnic cues. The paper begins with a review of the psychological research on personality and prejudice.  Then, social psychological insights about how negative stereotypes can override personality compulsions toward ethnic tolerance are discussed.  The role of political values as attitudinal determinants follows.  An empirical analysis of the relationship between personality and values in the face of a controversial ethnic cue is then conducted with multivariate regression analysis.  I conclude with a discussion of the implications for explaining the terrain of ethnic attitudes in Canada and elsewhere.   Psychology and Prejudice Numerous factors influence how people respond to circumstances they encounter every day.  The psychology literature identifies factors that, due to their deep and broad influence, lay the foundation for many social and political attitudes.  The literature is divided into several branches, but two are particularly relevant to the study of prejudice.  Personality psychology looks at specific personality traits that predispose individuals to various attitudes, including prejudice.  Social psychology, on the other hand, looks at the interaction between an individual?s perception of the self and the perception of her environment.  They each provide important insights into the ?faulty and inflexible? generalizations that are the core of prejudice (Allport 1954).  I will treat each briefly in turn. Personality and prejudice Rooted in biology and, thus, influencing people?s fundamental worldviews, the impact of personality predates the social and political factors typically used in explanations of individual  119 attitudes (Bergeman et al. 1993; Costa and McCrae 1992; McCrae and Costa 2008; Sniderman 1975).40  Personality psychology research shows that people reacting to situations are not simply ?blank slates? on which social and political influences imprint (Mondak 2010, p. 2).  Instead, their attitudes and behaviour are both directly and indirectly influenced by their psychological nature. Personality psychology made an indelible mark on psychological explanations of prejudice in the 1950s, when Adorno et al.?s (1950) The Authoritarian Personality was published.  Until then, much of the focus in psychology was on processes that resulted in antipathy against groups, such as scapegoating or projection (Duckitt 1992).  Adorno and his colleagues drew attention to the role of individual traits to explain the actions of Nazis and the complicity of their sympathizers.  Criticisms of Adorno?s work soon appeared (e.g., Stewart and Hoult 1959; Titus and Hollander 1957) and explanations of prejudice that focused on personality traits fell out of theoretical favour. In the 1980s, breakthroughs in personality psychology research renewed interest in the relationship between personality types and prejudice.  The Five Factor Model, or the Big Five, dominates theories of this relationship.  The model proposes that the universe of personalities can be principally explained with five factors (e.g., McCrae and John 1992).  The factors ? Openness to Experience, Agreeableness, Conscientiousness, Extraversion, and Emotional Stability ? combine with external influences (e.g., culture, life events) to produce the observable aspect of personality, or the so-called ?characteristic adaptations? (e.g., self-esteem, habits, interests, motives, etc.) (McCrae and Costa 2008).  The five factors are thought                                                 40 Gosling and colleagues have found similar personality dimensions in animals as in humans, bolstering the case for biological roots of personality.  See Gosling and John 1999; Gosling et al. 2003.   120 to be universal across cultures (McCrae and Terracciano 2005) and remain relatively stable across lifespans (Srivastava et al. 2003).  Numerous studies consistently linked two factors ? Agreeableness and Openness to Experience (hereafter Openness) ? to prejudiced attitudes (e.g., Sibley and Duckitt 2008). Personalities high in Agreeableness tend to be modest, cooperative, and altruistic.  Conversely, personalities low in Agreeableness tend to be egocentric, antagonistic, and critical (Costa and McCrae 1992; Goldberg 1990).  These people tend to be more prejudiced because their critical and antagonistic nature often results in quick, negative judgments of others.  Their egocentrism results in the degradation of people or groups perceived to be lower in social status, such as many ethnic minority groups. On the other hand, personalities high in Openness are described as intellectually curious, open to new information, and self-reflective; low in Openness is described as traditional, conventional, and comfortable in familiar routines (Flynn 2005).  Personalities low in Openness are associated with prejudiced attitudes because of their reliance on convention and tradition for psychological well-being: people or groups who are perceived as violating convention or tradition are regarded as threatening.  Minority groups are often differentiated by how they lie at the margins of convention and tradition. So, people low in Openness view these groups as dangerous and a threat to society?s order and stability, which is a gross violation of their psychological security. Social psychology and prejudice Social psychology looks beyond personality to environmental factors that shape people?s mental state.  The social psychological theory of social identity, especially, demonstrates how an individual?s sense of self and her perceptions of the world around her interact to shape  121 attitudes and behaviour (e.g., Tajfel and Turner 1979).  The most relevant aspect of social identity theory with respect to the formation of prejudice is the idea of social categorization.  The theory states that individuals categorize themselves and others into ingroups and outgroups to simplify the complexity of their social environment (Macrae et al. 1994).  Importantly, social categorization leads to intergroup competition.  Because individuals have a basic need for positive self-esteem, they look to bolster the status of their ingroup and undermine the status of outgroups (e.g., Turner et al. 1979).  They pursue this need through intergroup comparisons in which they exaggerate comparisons that favour the ingroup and minimize the comparisons that favour the outgroup.  These evaluative comparisons prompt individuals to compete on behalf of their ingroup with the ultimate goal of psychological security and self-worth. Social psychology posits that stereotypes are a product of social categorization and shape competition between ingroups and outgroups (e.g., Brown 2010; Greenwald and Banaji 1995).  As part of the cognitive component of attitudes, stereotypes influence the subjective understanding of what is true or not about a target.  Individuals use them to categorize people into groups as part of the social identity formation process.  The evaluation of stereotypes, part of the affective component of attitudes, generates the negative or positive generalizations associated with prejudice.  Stereotypes exert considerable influence on attitudes because they do not require much cognitive effort to access (Brewer 1988; Devine 1989; Fiske and Neuberg 1990).  Moreover, they are learned early on in life, are repeatedly encountered, and thus are deeply embedded in the cognitive structure people access for information when interacting with their daily environment.  122 According to social identity theory, intergroup competition is triggered when a particular group parameter is salient.  Stereotypes help define these parameters.  Ethnicity is a salient parameter in virtually all societies (Mullen 2001; Verkuyten 2005).  Stereotypes about ethnic groups fuel intergroup competition as they are used as a shortcut for information about outgroups.  As will be demonstrated, the Muslim hijab invokes a powerful stereotype about Muslim group difference in Canada and certainly elsewhere.  The mere depiction of it is enough for negative, exclusionary reactions to be elicited not only from a pool of respondents with personalities prone to prejudice, but from respondents across the board. Political Values and Prejudice Stereotypes are only one instrument that people use to assess others and decide on subsequent behaviour.  Political values also shape the way a person understands and approaches experiences in her everyday life.  The rigorous study of political values can be traced to Rokeach?s The Nature of Human Values (1973).  Values are cognitive exemplifications of abstract goals.  They are transsituational, desirable, and act to motivate behaviour (Bilsky and Schwartz 1994).  As mentioned, stereotypes are powerful constraints on attitudes because they are part of the fundamental identity formation process and, thus undergird much of an individual?s cognitive structure.  Political values, on the other hand, are less embedded in the cognitive structure because they are acquired later on in life.  They are not ?overlearned? (or automatic) like stereotypes, but take effort by the perceiver to influence attitudes relative to stereotypes (Devine 1989; Monteith 1993).  To be sure, some values are learned early in childhood when parents, consciously or not, transmit their own ideals and goals (Hughes et al. 2006).  Bubeck and Bilsky (2004) find that the ten universal values identified by Schwartz (1992) can be already found amongst value systems in adolescents.   But, compared to  123 stereotypes, which start to take root in infancy, values are relatively new in the cognitive structure (Eichstedt et al. 2002; Serbin et al. 2001).   The function of values also makes them less constraining on attitudes, relative to stereotypes.  Values are prescriptive abstractions, informing desirable goals and modes of conduct (Schwartz 1994).  Stereotypes, conversely, are heuristics, giving people a shortcut to making sense of the world (Bodenhausen and Wyer 1985; Macrae et al. 1994).  They are reductive, rather than idealized.  The cognitive component of stereotypes gives them the potential for more sway on an individual?s attitude formation than values.  That is, it is relatively easy for an individual to rely on stereotypes, rather than values, when making judgments.  Still, values are important behavioural motivators because they prescribe goals and modes of conduct, which can generate self-esteem (Tesser 2000).  They create a behavioural standard that individuals will try to achieve and so will have an important role in forming attitudes.  The question of the present analysis is whether values are influential enough to overcome the weight of stereotypes. Why might values work together with personality traits to overcome deep-seated psychological impulses?  Theoretical and empirical work have offered evidence that values are closely related to an individual?s personality.  For example, McCrae and Costa?s (2008) personality system theory links values with personality by arguing that values are produced by the confluence of basic psychological tendencies (i.e., the Big Five personality factors) and environmental influences.  So, values emerge at the crossroads of nature (i.e., personality) and nurture (i.e., environment) (Olver and Mooradian 2003).  Empirically, values and personality have also been correlated (Dollinger et al. 1996; Luk and Bond 1993).  Roccas et al. (2002), for example, find that the Neuroticism personality trait correlates with stimulation and  124 tradition values.  The Extraversion trait, on the other hand, correlates with achievement and hedonism values.   Importantly, however, analysis of the relationship between values and personality insist on their theoretical and empirical independence.  They may overlap, but they are distinct attitudinal determinants.  Personality traits explain attitudes through a lens of a general pattern, whereas values explain attitudes by the goal for which it strives (Winter et al. 1998).   Also, because values are idealized goals, they are generally positive.  Personality traits, on the other hand, can be positive and negative (Hitlin and Piliavin 2004).  Concrete examples of the difference between personality and values include the individual who is predisposed to physical aggression (e.g., high in Neuroticism), but values harmonious interpersonal relationships.  Or, an individual who is a social butterfly (high in Extraversion), but values peace and solitude.  Indeed, the data used in the current study provides additional examples of this distinction.  The 2011 Canadian Election Study (CES) demonstrates that people high in Openness or Agreeableness do not always hold values congruent with their personalities.  For example, people high in Agreeableness are thought to be especially trusting.  But in the CES data, there is no significant difference between respondents high in Agreeableness who say ?most people can be trusted? and people ?need to be very careful? when dealing with others.41  So, while their personalities predispose them to be trusting, they do not always exhibit trust in hypothetical scenarios.  Also, people high in Openness are more likely to be independent-minded and unconventional.  Yet, no significant difference exists between respondents high in Openness who think it is more important for children to learn independence as opposed to                                                 41 The question is: ?Generally speaking would you say that most people can be trusted, or, that you need to be very careful when dealing with people??  125 respect for authority.42   So, while they may be closely related, values and personality should be treated as independent influences on ethnic attitudes. What are the values that influence attitudes about ethnic diversity?  Scholars disagree on the number of basic value dimensions people use when formulating attitudes (e.g., Rokeach 1973; Braithwaite 1997; Schwartz 1994).  However, they do agree that one dimension can be described with the values of liberal egalitarianism, which stresses the liberty and equality of all people.  The concept of the individual is debated within this philosophy, having implications for individual and group rights in democratic societies.  But, the liberal egalitarian emphasis on equality should result increase positive attitudes about people who may be perceived as different.  This association is in contrast to values that have been previously linked to prejudice.  Conservative or classically liberal values, such as individualism, self-reliance, moral traditionalism, and deference of authority, have been implicated as underpinning ethnic and racial tensions, either through a principled commitment to these values or through so-called symbolic racism (e.g., Kinder and Sears 1981; Sniderman and Hagen 1985).  The emphasis of the present analysis, however, is about the reformulation of ethnic attitudes by individuals who are psychologically predisposed to ethnic tolerance.  Liberal egalitarian values seem like a good candidate to do just that, as a commitment to valuing all people should help an individual overcome negative attitudinal triggers.  In fact, the present analytic approach supports other work arguing that the focus in values research should be on egalitarian rather than individualistic values (Hurwitz and Peffley 1992; Sears et al 1999).  Also, in response to the critique that group rights and recognitions are contrary to liberal tenets, Kymlicka (1989) asserts that a commitment to liberal egalitarianism requires a                                                 42 The question is: ?Now some questions about your values. Here are some qualities that children can be encouraged to learn. Which one do you think is more important? Independence, or respect for authority??  126 commitment to multiculturalism if cultural membership is understood as largely involuntary yet central to individual fulfillment.  Normatively, then, it is the hope that liberal egalitarianism can result in positive interethnic relations.  Canada?s political and normative multiculturalism is, at least in part, about protecting ethnic identity and creating a sense of belonging for ethnic minorities.  As such, liberal egalitarian values have the potential to shape attitudes about ethnic minorities above and beyond the deep-seated influences of personality traits and stereotypes triggered by an individual?s context.  Moreover, it is the goal of Canadian multiculturalism that these values have a positive impact.  Past psychological research on the link between values and personality traits adds to the appeal of investigating liberal egalitarian values, rather than the absence of individualistic or conservative values, as a way to overcome negative situational triggers specifically with the personality traits under scrutiny here.  McCrae (1996) has found that Openness to Experience, one of the two personality dimensions consistently linked to general prejudice, is associated with preferences for particular political value systems.  He asserts that personalities high in Openness are more likely to adhere to liberal values and, conversely, personalities low in Openness are more likely to adhere to conservative ones (also Trapnell 1994; but see Saucier 1994).  Van Hiel et al. (2000) find similar evidence in Belgium and Poland.  They add that the Values component of a disaggregated Openness index (defined as a ?readiness to re-examine own values and those of authority figures?43) has the highest negative correlation with right-wing political ideologies.  In contrast, the components of Ideas, Actions, and Fantasy have considerably smaller correlations; the components Aesthetics and Feelings perform relatively poorly.  Using the Rokeach Values Survey (RVS), Dollinger et al. (1996) find that Openness correlates with values such as ?world of beauty?, ?imaginative?, and ?broad-mindedness?.                                                  43 From the authoritative NEO Personality Inventory-Revised manual (Costa and McCrae 2008).  127 Personalities characterized by high Openness, then, should be more likely to question their own ethnic prejudices and the prejudices they see in their environment, which suggests that their idealized goals motivating their attitudes are likely egalitarian.  Agreeableness, too, correlates with values that should shape ethnic attitudes. Dollinger et al. (1996) find that Agreeableness is correlated with the RVS values of being ?helpful?, ?cheerful?, and ?loving?.  Individuals who subscribe to these values may also subscribe more broadly to egalitarian ones, since egalitarianism is, in part, about helping others.  The authors also find that Agreeableness negatively correlates with the value of ?social recognition?, which could influence the strength of an individual?s social identity and her reliance on stereotypes to make judgments.  Devaluing social recognition likely reduces the individual need to demarcate one?s ingroup from others, which is at the core of stereotype usage (Turner et al. 1979).  That is, the individual may still have a psychological need for social recognition, but she does not value it.  Luk and Bond?s (1993) study of Chinese students observes that Agreeableness is strongly correlated with values related to benevolence and restricting one?s impulses.  They posit that these values would be attractive to personalities high in Agreeableness because the values would create a happy and peaceful environment.  Again, this empirical link suggests that it makes good sense for the present investigation of liberal egalitarian values and their role in moderating negative ethnic stereotypes for personalities predisposed to tolerance. Personality, Stereotypes, Political Values and  Prejudice against Muslims  This study examines how personality traits, negative stereotypes, and liberal egalitarian values work together to shape attitudes about ethnic diversity.  Past research has found that values can override personality impulses when it comes to attitudes about minority groups.  Devine?s  128 (1989) study of American adults, for example, finds that low prejudiced individuals agreeing with egalitarian principles about race are aware of negative stereotypes about Blacks but make an effort to reject these stereotypes (also see Fazio and Dunton 1997; Monteith 1993).  Her ground-breaking study demonstrated the dual process of ?automatic? (i.e., personality) and ?controlled? (i.e., principles) aspects of stereotypes.  While these studies show how values can override negative stereotypes about Blacks, they make no claims about how values might override negative stereotypes about other ethnic minorities.  It is possible that the case of Black-White race relations in America is not generalizable to other ethnic minority experience.  There is a mainstream consensus against anti-Black prejudice, even if the consensus is only lip-service (e.g., Kinder and Sears 1981; Kinder and Sanders 1996).  This mainstream consensus does not exist, at least not to the same extent, for other ethnic minorities.  Still, Devine?s study suggests that investigating the possible role of values in controlling negative reactions to ethnic minorities is a good analytical point of departure. Currently, the most salient ethnic minority target of negative attitudes in the West is the Arab Muslim community.  Even before the high profile terrorist attacks of September 11, 2001 and the subsequent ?War on Terror?, the stereotype of Arab Muslims as terrorists or extremists was pervasive (Shaheen 1997; Strabac and Listhaug 2008).  Not surprisingly, social psychologists have found evidence that negative stereotypes and symbolic threats to identity are important predictors of prejudice against Arab Muslims (Kalkan et al. 2009; Kam and Kinder 2007; Sniderman and Hagendoorn 2007).  In fact, simple intergroup divisions based on negative stereotypes of Arab Muslims occurs with children as young as four years old (Bar-Tal 1996; Teichman 2001).  Visual cues have been found to trigger anti-Muslim hostility  129 (Unkelbach et al., 2008; but see Harrell et al. 2012), but contact with Muslims can mitigate negative attitudes (Gonzales et al. 2008). There has been little research on personality type and anti-Muslim prejudice.  Evidence does link Social Dominance Orientation and anti-Arab prejudice (Pratto et al 1994) and Right-Wing Authoritarianism and anti-Muslim prejudice (Echebarria-Echabe and Guede 2007).  However, these individual-level orientations are not typically viewed as personality traits, but as worldviews (Duckitt 2001).  Still, they theoretically mediate between traits and attitudes.  Social Dominance Orientation is thought to mediate personalities low in Agreeableness, and Right-wing Authoritarianism mediates personalities low in Openness to Experience, to produce prejudice (Sibley and Duckitt 2008).  So, it is likely that, similar to studies of other types of prejudice, low Agreeableness and low Openness produce negative attitudes about Muslims in Canada. The literature examining the relationship between political values and prejudice against Muslims is also slight.  Sniderman and his colleagues find that authoritarian values are a powerful predictor of anti-Muslim attitudes (Sniderman et al 2004; Sniderman and Hagendoorn 2007).  They admit, though, that their theoretical conception of authoritarian values is vague, but their empirical measure includes items that reflect politically conservative values (e.g., ?It is better to live in an orderly society in which the laws are vigorously enforced than to give people too much freedom.?).  Others find that education can reduce anti-Muslim prejudice by instilling democratic norms of tolerance in the individual (Fetzer and Soper 2003; Strabac and Listhaug 2008).  So, values do shape attitudes about Muslims and, thus, liberal egalitarian values may help override anti-Muslim stereotypes.   130 Data and Methods The current analysis investigates the relationship between personality traits, stereotypes, political values, and attitudes about Muslims in Canada.  The purpose of the analysis is three-fold.  First, I look at the role of personality traits in shaping attitudes about ethnic diversity.  Does the relationship between personality and ethnic attitudes consistently predicted in the psychology literature appear in the case of Muslims in the Canadian context?  Second, the role of negative stereotypes in shaping attitudes about ethnic diversity is examined.  How do negative stereotypes mediate the relationship between personality traits and ethnic attitudes?  Finally, the role of political values, specifically liberal egalitarianism, is considered.  Can political values overcome triggered responses driven by negative stereotypes so attitude-personality coherence can be achieved? The first analysis: Personality in the ?Fatma? experiment The data come from the 2011 Canadian Elections Study (CES).  The study embedded an experiment in its web-based questionnaire, which was the fourth and final wave.  The dependent variables here are the respondent?s attitudes captured by this experiment.  The experiment solicited responses to one of two randomly assigned conditions.44  The number of respondents for each condition is no more than 370, but this is sufficient for the present analysis.  Respondents were asked to read a short text about a woman named ?Fatma?.  Fatma is said to represent the Canadian Turkish-Muslim Action Network, a religious outreach organization.  The text states that Fatma is applying for $80 000 from the federal Multiculturalism Grants and Contributions Program.  A picture of the woman ostensibly applying for the public grant accompanies each text.                                                   44 A third condition about a Portuguese Catholic woman has been omitted from the current study.  131 The conditions are exactly the same except for one crucial difference.  In the first condition, the picture is of a woman wearing Western dress.  In the second condition, the picture is of the same woman, except that she is wearing a hijab.  The hijab leaves the woman?s face exposed, but covers her hair and shoulders.  The respondent is asked if they support or oppose the government funding the grant application.  The dependent variables ? called Fatma 1 and Fatma 2 ? are four-category ordinal variables where the highest score indicates the respondent ?strongly opposes? the application (all question wording and descriptive statistics can be found in Appendix Four).  For Fatma 1, about 26 percent of respondents strongly oppose, 30 percent oppose, 36 percent support, and 9 percent strongly support.  For Fatma 2, about 23 percent strongly oppose, 33 percent oppose, 39 percent support, and 5 percent strongly support.    Taken at its face, the distribution of responses shows little difference between the conditions.  Over-all distributions, however, reveal little about the sub-groups of respondents who make up the different response categories.  The present paper is interested in how personality dimensions influence ethnic attitudes, particularly when negative stereotypes are primed.  The embedded experiment suggests how negative stereotypes can push typically tolerant personalities to become reactionary.  There are several ethnic cues in the experiment that could trigger negative stereotypes of Arab Muslims.  Both experimental conditions include the woman?s name (Fatma), her country (Turkey), and religion (Islam) as cues.  These cues may be enough to trigger exclusionary responses not just from personalities predisposed to intolerance, but from across the board.  The cues may also be relatively benign and fail to elicit a negative reaction.  The depiction of the hijab in the second condition, however, may amplify the effect of ethnic cues in a way ?Fatma?, ?Turkey?, or ?Muslim? do not.  In the  132 discourse around ethnic diversity in Canada and elsewhere in the West, the hijab is more than just a head covering.  It is a potent symbol of ethnic difference (e.g., Bullock and Jafri 2001; Strabac and Listhaug 2008; Unkelbach et al. 2010).  So, comparing responses to both conditions can shed light on the role that negative stereotypes ? and which negative stereotypes ? play in shaping attitudes. Before examining the influence of liberal egalitarian values, an initial investigation was performed on the direct relationship between personality variables and the dependent variables.  That is, whether or not the respondent?s personality is associated with her support or opposition to Fatma?s multiculturalism grant application in the assigned condition.  Looking at the direct relationship tests the influence of personality versus the influence of negative stereotypes rooted in one?s social identity.  For the first time, the CES data contains measures of the Big Five personality dimensions, using the Ten Item Personality Index (TIPI).  Developed by Gosling et al. (2003), TIPI is a brief measure of the Big Five that taps each dimension with four traits.  Two traits are from the low end of the dimensions? spectrum and two are from the high end.  For each side of the spectrum for each personality dimension, the respondent is asked whether agrees or disagrees if the traits apply to her.  This results in ten questions in total.  The index of the TIPI personality traits is outlined in Table 5.1.  Respondents are asked the extent to which they agree or disagree that the pair of traits describes their personalities.  The scores are reversed where necessary and summed to create measures of individual personality.  High scores indicate a personality high in the given dimension. As mentioned, the consensus in the personality psychology literature is that the dimensions of Openness to Experience and Agreeableness independently and consistently  133 predict prejudice against others perceived as different.  The experiment here only partially supports this consensus.  Table 5.2 displays the results of an ordered logistic regression for each of the simple models testing the direct relationship between personality and attitudes toward multiculturalism spending as measured by the dependent variable.  The first estimation using respondents who saw the woman not wearing a hijab (Fatma 1) is in line with previous personality psychology research.  Namely, having a personality high in Openness or high in Agreeableness is associated with less opposition to Fatma?s grant application.  Also as expected, the other personality traits have no impact on the results.   In the second model, for respondents who saw Fatma wearing a hijab, none of the personality indicators are significant.  This is in contrast to the typical expectations of personality psychology.  Clearly, the visual cue of the hijab has an impact on attitudes toward Fatma?s multicultural grant application.  The collapse of support from high Openness and high Agreeableness individuals does not mean that these respondents are less supportive of the application.  It may be that people low in these personality dimensions are becoming more positive. 45  Further tests, however, demonstrated that respondents who are expected to be tolerant given their personalities are decreasing their support for Fatma?s application (not shown).  When the hijab is primed, the probability that these respondents oppose the application increases by nine points for each personality dimension.46  This increase is statistically significant.  So, experimental cues of Fatma?s name, country, and religion, present in the first condition, are not enough to provoke individuals either high in Openness or Agreeableness to react negatively to her grant application.  However, the hijab cue is.                                                 45 Individuals low in Agreeableness and Openness are, in fact, influenced by the depiction of the hijab.  Their attitudes are predicted to become more favourable to Fatma?s grant application when the hijab is depicted. This finding is provocative and deserves further investigation. 46 Predicted probabilities estimated with Stata?s margins command.  134 The second analysis: The expanded model The exclusionary reaction to the hijab by people with typically tolerant personalities prompts further questions about how might these people control incongruent reactions to stereotypes.  That is, what might motivate people with typically tolerant personalities to look positively on Fatma 2?s grant application if their personalities do not produce expected attitudes?  The hypothesis here is that political values, particularly liberal egalitarian ones, are able to override negative attitudes triggered by the simple depiction of a hijab.  If this is the case, the support for ethnic policies expressed by individuals high in Openness or Agreeableness has a dual source: the ?automatic? source of their personality traits and the ?controlled? source of their liberal egalitarian values (Devine 1989).   A second analysis to investigate this hypothesis was performed with an expanded model and a new dependent variable.  The dependent variable ? grant ? combines the responses of respondents who saw Fatma 1 and Fatma 2 into a single variable, allowing comparisons between the two conditions.  The variable still indicates whether the respondent supports or opposes the multiculturalism grant application with the highest score measuring strong opposition.  A control called condition indicates the respondent?s assigned condition: 0 is Fatma 1 (without hijab) and 1 is Fatma 2 (with hijab).   The independent variable measuring liberal egalitarian values is an index, which I call egalitarian.  It contains three variables that tap separate elements of the broad umbrella of egalitarian values, without explicitly mentioning ethnic minorities.  The first variable in the index asks respondents if the government should ?leave people to get ahead on their own? or ?see to it that everyone has a decent standard of living?.  The latter response suggests that the individual values the idea of equality of outcome that is part of egalitarianism.  The second  135 variable asks if the most important element of a democracy is ?letting the majority decide? or ?protecting the needs and rights of minorities?.  This measure is about the use of political processes to ensure equality of disadvantaged groups.  It cues the idea of minorities, though it does not mention which minority is the subject of the question.47  The third variable asks if respondents agree or disagree that the welfare state makes people less willing to look after themselves.  This is similar to the first measure in that it refers to some people needing more help than others.  However, it is specifically about individuals who are on the extreme end of the economic spectrum, rather than the relatively vague subject in the first measure.   The index sums scores on each variable, so lower scores measure low egalitarian values and higher scores measure higher egalitarian values.  It is then interacted with the variable measuring the experimental condition (condition) and a personality variable, either Openness or Agreeableness.  The three-way interactions should identify if liberal egalitarian values provide an extra push for respondents high in Openness or Agreeableness to control their negative reactions to the hijab in Fatma 2.  I also include several controls that have been previously shown to influence attitudes about ethnic minorities.  First, I include the respondent?s household income as a measure of objective material self-interest.  There is evidence that material considerations can shape attitudes about ethnic minority groups if they are perceived as materially threatening (e.g., Bobo 1983; Kluegel and Smith 1983; Palmer 1996; though, see Kinder and Sanders 1996).  Second, I include a measure of media consumption, which sums the weekly usage of television, radio, and newspaper specifically for news.  Because of its agenda-setting powers, exposure to mass media is believed to have an effect on attitudes (e.g., Bartels 1993;                                                 47 Additional analysis removed this item from the index to see if it was driving any regression effects.  The test produced similar regression results with larger errors.  This indicates that the item does not drive the effect of the index, but still contributes to its predictive power.  136 McCombs et al. 1997), especially attitudes about Muslims (e.g., Saeed 2007).  Third, individuals who are more politically sophisticated should be more aware of the norms of tolerance in society, which should differentiate their attitudes from individuals with less sophistication.  I control for this with an index of typical political knowledge questions.  Contact with outgroups has been shown to positively influence ethnic attitudes (Pettigrew and Tropp 2006).  As such, I control for whether the respondent states she socializes with  Muslims.  Finally, the respondent?s gender, age, and ethnicity are added.   Numerous studies have shown that men tend to display more prejudiced attitudes than women, particularly when the target is seen as a threat to the social hierarchy (Altemeyer 1998; Whitley 1999).  Age should also be associated with attitudes toward multicultural funding.  Older respondents may be more likely to view ethnic diversity as threatening because it is relatively recent demographic development.  Younger people, by contrast, have grown up with Canada?s multicultural norm and thus will be more likely to be at ease with ethnic diversification.  The respondent?s ethnicity is measured with the CES self-reported ethnicity question, which I recode as whether the respondent is white or not.  Table 5.3 displays the results of two ordered logistic regressions for the expanded model.  The analysis has separated the personality traits, so the second column investigates the role of Openness and the third column investigates the role of Agreeableness.  To keep the analysis manageable, the controls are not shown.  As a reminder, the dependent variable (grant) runs from 1 to 4, with 4 being strong opposition to Fatma?s grant application.  Not surprisingly, the table does not offer much insight.  The present analysis is concerned only with the high end of the personality scales.  So, the coefficients provide little insight into the role liberal egalitarian values have in helping people high in Openness or high  137 in Agreeableness control their exclusionary reaction to the hijab.  As such, the analytical focus shifts to predicted probabilities.  By allowing the personality variables to be set to their highest values, predicted probabilities can be used to untangle the reactions of people high in Openness or high in Agreeableness at different levels of liberal egalitarian values.   Results are displayed in Figure 5.1.  The figure illustrates the attitudinal change for people high in Openness at low, medium, and high levels of liberal egalitarian values for each experimental condition.  (Personalities high in Agreeableness are omitted for the sake of space and will be discussed below.)  Recall that Table 5.2 shows that these people should be predisposed to be favourable toward the grant but that the hijab seems to quash this personality influence.  Figure 5.1 clearly illustrates how liberal egalitarian values shape support for Fatma?s grant application amongst these respondents.  The negative effect of the hijab is much stronger for these individuals if they have low liberal egalitarian values.  The gap between the conditions at the left-hand side of the x-axis demonstrates this large hijab effect.  Specifically, for respondents high in Openness with low liberal egalitarian values who are exposed to the hijab, the predicted probability of opposing the grant application is .65.  These same respondents, but in the non-hijab condition, have a predicted probability of .31.  On the other hand, respondents high in Openness with either medium or high liberal egalitarian values have the same predicted response, regardless of condition.   Figure 5.1 is a clear demonstration of the power of the negative stereotype around the hijab.  It is also a clear demonstration of how liberal egalitarian values can override negative reactions to this stereotype for people high in Openness.  These values are not simply highly correlated with Openness.  The Pearson correlation coefficient between Openness and the liberal egalitarian values index is .09.  Liberal egalitarian values, at least how they are  138 measured here, are independent from the respondent?s personality in shaping attitudes toward Fatma.  It is also unlikely that respondents high in Openness who hold high liberal egalitarian values simply do not perceive the hijab as a negative stereotype.  An additional check of the data shows that almost 25 percent of high Openness respondents who hold high liberal egalitarian values rate Muslims at 50 or lower on a feelings thermometer where 0 is ?really don?t like? and 100 is ?really like?.  Also, controlling for general feelings about Muslims in the model with this thermometer variable did not change the results.  So, Figure 5.1 is clearly an illustration of how these individuals use their values to control negative reactions to the Muslim hijab.  For individuals high in Agreeableness, the results are different.  Once all the variables are added to the model, Agreeableness ceases to predict attitudes in either condition.  Notably, liberal egalitarian values had a strong direct effect with the dependent variable (grant): higher liberal egalitarian values decrease the probability of opposition to Fatma?s application by .41.  However, the disappearance of Agreeableness as a predictor of attitudes was perplexing.  To test if the effect of Agreeableness observed in the first condition (Fatma 1) observed in Table 2 was being overwhelmed by the interaction, the estimations for both personality dimensions were rerun without any interactions (not shown).  Openness still functioned as before: it predicted attitudes in Fatma 1 and failed to predict in Fatma 2.  However, Agreeableness was not statistically significant in either estimation.  So, the attitudinal gap observed between low and high Agreeableness individuals for Fatma 1 in Table 2 failed to materialize once other attitudinal determinants were added.  The implications of this finding will be discussed below.   Discussion Research has suggested that Openness to Experience and Agreeableness influences prejudice.   139 Their influence specifically on ethnic prejudice has only been recently investigated.  The present analysis contributes to this debate by showing that one of these personality dimensions ? Openness to Experience ? does influence prejudice toward Muslims in terms of preferences about funding multiculturalism projects.  The study goes further by demonstrating how negative stereotypes alter this expected relationship between personality and prejudice.  The depiction of the hijab ? a potent ethnic symbol ? spurred individuals with personalities predisposed to tolerance to increase their opposition of multiculturalism funding.  If liberal egalitarian values are held, however, the analysis showed how individuals high in Openness can overcome the exclusionary reaction that the hijab triggers.  My findings support the argument that political values matter to ethnic attitudes.  The respondents here are not simply paying lip service to values.  Instead, their values shape their ethnic attitudes in expected ways.  The analysis shows that liberal egalitarian values, in particular, are able to override a potent, negative ethnic stereotype.  These values are associated with more tolerant attitudes in both experimental conditions for both personality dimensions.  But, these values also give respondents high in Openness an extra push to be tolerant of Fatma when she is wearing a hijab.  This finding supports the valuable insights of attitude formation derived from the psychological studies of the ?automatic? and ?controlled? aspects of stereotypes (e.g., Devine 1989; Monteith 1993).  But instead of demonstrating this dual process in an laboratory setting, typical of psychology research, the current analysis demonstrates it with people in their every day environments. The analysis also highlights the differences between the Openness to Experience and Agreeableness personality dimensions.  Openness directly predicts attitudes toward Fatma in the first experiment, and interactively predicts attitudes in the second experiment.  Any initial  140 influence from Agreeableness, however, is diminished to non-significance once the model is expanded.  Considering that Agreeableness is a common determinant of tolerant attitudes, its failure to predict support for Fatma?s grant application in either experimental condition certainly demands future investigation.  There is evidence that individuals high in Agreeableness may not be as well equipped to disregard stereotypes as individuals high in Openness, especially in an anonymous survey setting.  Personalities high in Openness are curious and open-minded and, thus, are more likely to be attentive to information that disconfirms stereotypes (Flynn 2005; Jost et al 2003).  Their open-mindedness would also motivate them to accept ethnic difference on its face, regardless of the setting.  In contrast, personalities high in Agreeableness are primarily concerned with the social or experiential life, where interpersonal harmony is critical (John and Srivastava, 1999).  In general, this concern for harmony motivates high Agreeableness individuals to be accepting of difference.  But, in an anonymous opinion survey where interpersonal harmony is not a factor, the personality impulse toward tolerance may be suppressed by other attitudinal determinants.  The influence of Agreeableness, as well as its interaction with liberal egalitarian values, may be different if the experiment took place in a setting where interpersonal relations are emphasized.  For example, if the high Agreeableness respondent knew she was being watched or the experiment took place publicly, she may be driven more by her desire to create harmony and, thus, be more accepting of Fatma?s ethnic difference. The insights of the study also suggest ways in which policymakers can shore up positive attitudes about ethnic diversity.  In countries such as Canada that have comparatively inclusive multiculturalism policies, and where there is no prohibition against the hijab, the current analysis demonstrates how attitudes can become negative simply because this symbol  141 of ethnic difference is observed.  The analysis also shows how liberal egalitarian values, with their emphasis on equality can help dampen this negative effect.  Given that stereotypes are so deeply rooted and difficult to change, policymakers interested in societal harmony should be encouraged to inculcate liberal egalitarian values to counteract reactionary attitudes.  It may be enough to inculcate these values to prompt citizens high in Openness to reject their negative reactions to Muslim stereotypes, creating a sort-of spill-over effect for the management of ethnic relations.  Indeed, the analysis shows that individuals do not have to be strident liberal egalitarians.  Even moderate liberal egalitarian beliefs seem to be enough to override the hijab effect.   In all, the present analysis goes some way to demonstrate the interplay between personalities and political values with the salient issue of ethnic attitudes.  The experiment provides a strong test of the influence of personality and values.  Admittedly, the test is about a rather particular situation (Muslims and multicultural funding), which could reduce the generalizability of the results.  Still, the political environment as it is, especially in Canada, has questions of Muslims, multicultural norms, and the role of public policy in the forefront.  Understanding how personalities and political values influence ethnic attitudes seems worthy of further study.       142 Table 5.1: Ten Item Personality Index (TIPI)  Personality Dimension Low High Openness to Experience Conventional, uncreative Open to new experiences, complex Agreeableness Critical, quarrelsome Sympathetic, warm Extraversion Reserved, quiet Extraverted, enthusiastic Emotional Stability Anxious, easily upset Calm, emotionally stable Conscientiousness Dependable, self-disciplined Disorganized, careless  Table 5.2: Personality Traits and Attitudes toward Fatma?s Multicultural Grant Application, Both Conditions   Fatma 1 Fatma 2 openness -0.887*** -0.202  (0.238) (0.228) agreeable -0.542** 0.133  (0.218) (0.229) extraversion 0.0894 0.117  (0.171) (0.173) conscientious 0.144 0.181  (0.219) (0.212) stability 0.318 0.154  (0.202) (0.217) Observations 247 259 Cells display ordered logistic coefficients; Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1  Note: Each personality dimension is scaled from 1 to 4 with 4 indicating the respondent is high on the particular dimension.   143 Table 5.3: Predictors of Opposition to Fatma?s Multiculturalism Grant Application for Openness to Experience and Agreeableness Personalities, Combined Conditions   Openness Agreeableness condition -5.631 0.817  (3.834) (3.932) egalitarian -0.397 -0.199  (0.441) (0.429) condition*egalitarian 0.555 -0.250  (0.596) (0.595) openness -0.720   (0.919)  condition*openness 2.008   (1.242)  egalitarian*openness 0.0205   (0.142)  condition*egalitarian*openness -0.214   (0.191)  agreeable  -0.0831   (0.819) condition*agreeable  -0.103   (1.236) egalitarian*agreeable  -0.0534   (0.134) condition*egalitarian*agreeable  0.0509   (0.187)    Observations 325 327 Cells display ordered logistic coefficients; standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1  Independent Variable Scaling: 1) condition: condition 1=1/condition 2=2, 2) egalitarian: low=3/high=9    144 Figure 5.1: High in Openness Personalities and Opposition to Fatma?s Multicultural Grant Application at Different Levels of Liberal Egalitarian Values of opposing Fatma's applicationLow Medium HighFatma 1Fatma 2 145 CHAPTER SIX CONCLUSION  On October 8, 1971, Prime Minister Pierre Elliot Trudeau stood in the Canadian House of Commons and introduced the federal policy of multiculturalism with these bold words:  A policy of multiculturalism within a bilingual framework commends itself to the government as the most suitable means of assuring the cultural freedom of Canadians. Such a policy should help break down discriminatory attitudes and cultural jealousies. National unity if it is to mean anything in the deeply personal sense, must be founded on confidence in one's own individual identity; out of this can grow respect for that of others and a willingness to share ideas, attitudes and assumptions. A vigorous policy of multiculturalism will help create this initial confidence. It can form the base of a society which is based on fair play for all.48  Trudeau?s response to the recommendations made by the Royal Commission on Bilingualism and Biculturalism would fundamentally reshape government policy and public discussion about ethnic diversity in Canada to the present day.  The idea that ?cultural freedom? empowers the individual and results in a stronger, more cohesive, society dovetails with more recent writings of a cadre of Canadian political philosophers that insists the majority culture protect or recognize minority group rights as a way of fulfilling liberal democratic values of equal opportunity and individual liberty (e.g., Kymlicka 1989; Taylor 1992).  Trudeau?s speech, as well as the prescriptions of these philosophers, offers a way to empirically assess the thorny issues of ethnic diversity and modern liberal democratic societies.  While the assumptions about ?cultural freedom? have been adopted by successive Canadian governments, as well as a number of influential academics, are they realistic considering the                                                 48 Trudeau?s speech can be found here: Accessed June 1, 2013.  146 behavioural tendencies of actual people?  Given how Canadians formulate their attitudes about ethnic diversity, multiculturalism, immigration, and so on, can Trudeau?s hope for a strong, cohesive, and diverse society be realized?  This dissertation addressed the question of how people actually behave within the context of an ethnically diverse society.  It places the normative goals of Trudeau and others aside and examines the foundations of people?s attitudes about issues that are at the heart of the ethnic diversity debate in Canada.  In doing so, it bridges a significant gap in the political science literature, particularly in Canada.  Very little is known about how individuals form attitudes about their ethnically distinct neighbors, a surprising lacuna given the long and relatively distinguished track record Canada has in managing ethnic diversity.  To be sure, the political and social issue of Canadian ethnic diversity has garnered considerable attention recently.  However, many of the recent studies, while offering important insights, neglect the foundations of attitudes about ethnic diversity issues.  We still have little understanding of why Canadians say what they do about topics such as the consequences of immigration, approaches to integration, or the social acceptance of minority outgroups (exceptions include Citrin et al. 2012; Harell et al. 2012).    In fact, the issue of Canadian ethnic diversity has been a source of academic debate for decades, even if its rigorous empirical testing is new.  The title of the dissertation references both Gibbon?s (1938) The Canadian Mosaic and Porter?s (1965) The Vertical Mosaic, two seminal texts assessing Canada?s social diversity.  Porter wrote his book in response to Gibbon and, specifically, what Porter thought were Gibbon?s grossly misguided assumptions about Canada?s interethnic harmony.   My goal here is similar: to test assumptions and fill  147 gaps that still exist in the Canadian political science literature on the complex ?mosaic? of citizens? attitudes in the context of ethnic diversity. Summary of Main Findings The second chapter seeks to draw a fundamental map of the outgroup perceptions of Canadians, or at least white, mainstream Christian Canadians.  It does so with the use of comfort ratings, which is an empirical indicator of the social acceptance of ethnic minorities.  It builds on seminal Canadian sociological research in the 1970s and 1990s on the ?ethnic hierarchy? and indicates that outgroup rankings are still a reality more than 40 years into Canada?s multicultural experiment (e.g., Berry et al 1977; Berry and Kalin 1995).   In political science, this basic aspect of intergroup relations has been neglected.  Instead, perceptions of ethnic minorities are treated as essentially the same (important exceptions include Sniderman et al. 2000; Harell et al. 2012).  For example, Stolle et al.?s (2008) look at the effects of ethnic diversity on generalized trust suggests that the insights from American studies ? that diversity decreases trust ? may matter more in Canada because of the comparatively high heterogeneity of minority groups.  The authors fail to develop specific reasons why high ethnic diversity may amplify any negative effects observed in the US, where the most salient ethnic divisions are between whites, blacks, and Hispanics.  My chapter on the outgroup hierarchy would provide a compelling explanation.  It indicates that the reason why Canadian diversity may have more negative effects than American diversity is not just because there are more ethnic minority groups, but because there are more cross-categorized ethnic minority groups.   Many of Canada?s salient ethnic groups are defined not only by their ethno-racial differences (such as blacks and Hispanics in the US), but also their religious differences.  The  148 large Punjabi community in BC, Moroccans in Montreal, or South Asians in Toronto all would be plausibly cross-categorized as ethno-racially and religiously different than the white, mainstream Christian Canadian.  My analysis clearly demonstrates how cross-categorized outgroups are less socially accepted than single-category ethnic outgroups.  The inclusion of single-category religious groups in my analysis places the consequence of perceptions of religious difference into stark relief.  The chapter?s identification of the outgroup hierarchy, as well as its stability over subpopulations typically differentiated by their ethnic attitudes, emphasizes the importance of studying the impact of different outgroup perceptions.  Outgroups are clearly not perceived the same and this perceptional difference holds not only for groups who tend to be more wary of outgroup difference, but for groups across the board. The third chapter looks at individual-level determinants of white and non-white Canadians? attitudes about immigration and multiculturalism.  It seeks to address two weaknesses in the literature: the lack of studies comparing a wide range of possible attitudinal determinants and the lack of studies comparing majority and minority attitudes.  The chapter takes a cue from Kinder and Sanders?, Divided By Color (1996), a ground-breaking attitudinal study on race in America.  Like them, I look at the relative effects of the ?primary ingredients? of attitudes ? social identity, material self-interest, and political values ? but in the Canadian context.  These three factors reflect the three dominant explanations in the literature on ethnic and racial attitudes (e.g., Kinder and Sears 1981; Bobo 1983; Sniderman and Piazza 1993, respectively).   Given Canada?s different history with immigration and diversity, as well as its different institutions and political discourse, it is not obvious that the lessons of Kinder and Sanders would apply here.  However, I find that they largely do: social identity, specifically  149 measures of outgroup resentment, are the dominant shapers of the attitudinal terrain in Canada, at least with the attitudes investigated in the analysis.  Importantly, though, I find that the measure of outgroup resentment that matters most for whites (attitudes about outgroup power) is different than the measure of outgroup resentment that matters most for non-whites (attitudes about interethnic marriage).  Material self-interest (at least, subjective material self-interest) and political values are also found to influence the attitudinal terrain, supporting past research (e.g., Esses et al 2001; Hainmueller and Hiscox 2007).  This finding is certainly a helpful reminder of the complexity of the attitudinal terrain, and that it is a ?mosaic? of factors.  Comparing multiple factors helps clarify their relative importance.  The chapter also highlights the utility of comparing majority attitudes with minority attitudes, since the impact of variables for the two groups can be considerably different. The fourth chapter extends beyond Kinder and Sanders and other work to include contextual-level variables.  I look at variables that measure the identity threats implicated by the ?contact vs. conflict? debate and the material threats implicated by realistic group conflict theory.  I find no evidence that people?s material context shapes their attitudes about immigration, immigrant obligations, and multiculturalism.  The results counter recent Canadian work finding contextual economic effects (e.g., Palmer 1996; Wilkes et al. 2008).  These studies, however, use national economic indicators, while I use more local ones. The analysis of local ethnic diversity, however, hints at a process that supports the assumptions behind multicultural theory.  This finding, I think, is the most significant of the chapter.  The evidence is weak, but I think it is suggestive.  It appears that, for non-whites, the more ethnically diverse their environment, the more they hold attitudes congruent with positive engagement with the majority community yet retention of their own cultural  150 distinctiveness.  This is precisely what the advocates of multiculturalism would want to see in the real world.  It is also theoretically supported by social psychological theories of subjective uncertainty, depersonalization, and self-affirmation.  The only caveat is that ethnic minorities tend to desire the assimilationist-style melting pot in homogenously white areas.  I explain this with reference, again, to social psychology.  But, this finding places a behavioural limit to the goals of multiculturalism.  To be sure, my analysis is limited due to practical data concerns, but it is an important step in empirically examining the assumptions underpinning multiculturalism in Canada. For whites, there is little evidence that local diversity has an impact on their attitudes about immigration and multiculturalism.  Given the results in Chapter Three, the source of white attitudes really appears to be at the individual-level.  Whites are the majority in virtually all the areas examined here, so perhaps they are simply indifferent to local diversity when individual-level determinants are considered.  They also would have ample opportunity to receive the message that their identity and worldview is ?correct?.  Finally, Canada?s strong multiculturalism policies and norms may curb the influence of contextual diversity for whites. The fifth chapter contributes to the ethnic diversity literature in two ways.  First, it introduces the role of personality into the process of attitude formation.  There is virtually no research in Canada looking at personality and attitudes about ethnic minorities.  As Mondak (2010) states, citizens are not simply ?blank slates? onto which political determinants imprint.  Personality shapes commonly investigated attitudinal determinants.  This chapter demonstrates how certain personality dimensions influence attitudes about Muslims in Canada.  My analysis shows how the insights of personality psychology do explain some of the attitudes toward Muslims, but not all.  Some negative stereotypes, specifically the Muslim  151 hijab, drive people with typically tolerant personalities to hold attitudes indistinguishable from their less tolerant counterparts.  The analysis, too, shows how not all stereotypes are equal in their potency.  It is seeing the hijab, and not the mention of  ?Fatma?, ?Muslim? or ?Turkey?, that triggers intolerance in this instance. The second, and more significant, contribution of the article is its demonstration of the role of liberal egalitarian values in shaping attitudes.  Kymlicka has argued extensively that the autonomy and equality aspects of liberal egalitarianism necessitate a multicultural approach to ethnic group relations in liberal democracies (e.g., Kymlicka 1995).  The analysis empirically shows how this prescription functions in the real world: these values can override exclusionary reactions to stereotypes about Muslims.  Canadians do not simply pay lip service to these values.  They are used, at least by people with a particular personality type, to control negative reactions and produce ethnic tolerance that is congruent with their personalities.  As such, my analysis shows why the promotion of these values is important.  It is not just that these values are normatively good, but that they actually function to correct intolerance. Taken together, the contributions of the fifth chapter highlight the psychological theory of the dual source of attitudes (e.g., Devine 1989).  This theory hypothesizes that the process of attitude formation contains an unconscious, or ?automatic?, component and a conscious, or ?controlled?, component.  In my analysis, the automatic component is the individual?s personality and the controlled component is her political values.  A central prediction of this theory is that the controlled component can work to correct behaviour if the individual judges it to be undesirable or incongruent with the automatic component.  My analysis confirms that liberal egalitarian values can correct negative reactions by people with typically tolerant personalities.  Typically, the dual source theory is demonstrated in a  152 laboratory setting with highly controlled conditions.  My analysis demonstrates it in a real-world setting, which is a significant methodological contribution to the literature. Limitations This dissertation, thus, contributes in multiple ways to the literature of attitudes about ethnic diversity in Canada.  Certainly, there are weaknesses.  Generalizability across space is hampered because of my focus on Canada.  This focus is intentional.  Canada?s history with ethnic diversity on the ground, as well as its multicultural norms and policy framework, makes it a unique case in which to investigate the contours of citizens? attitudes.  However, the focus limits the lessons that can be applied to other countries or regions.  There are implications for other locations, to be sure.  The stress on empirical comparison, particularly of the majority and minority groups, is sound advice for anyone interested in explaining ethnic attitudes.  So is the observation that religious difference should not be conflated with ethno-racial difference.  And, certainly, similar personality influences would be seen in other countries.  However, the history, institutions, and political cultures of other locales will influence the attitudinal terrain of citizens in unique ways. Another weakness is comparability across the chapters.  Because of data limitations, I necessarily relied on different datasets.  There was no single source of data that contained enough information for me to dig as deeply into the attitudinal terrain as I wanted.  This methodological weakness means that key concepts are operationalized differently across the chapters.  For example, social identity is measured with strength of religious and ethnic identities in the first article, outgroup resentment and opposition to interethnic marriage in the second article, and identifying as white in the third article.  The lack of data sources enabling a deep analysis into Canadians? attitudes signifies, I think, the lack of scholarly attention to the  153 deep analysis of attitudes.  As mentioned, there is a growing literature on attitudes and ethnic diversity in Canada and researchers are beginning to collect useful data.  In fact, the latest Canadian Election Study (2011), which I use in the fifth chapter, has the greatest potential yet for a single-source study of the fundamental structure of attitudes in Canada.   It is unfortunate that I was near completion of the dissertation when the data became publicly available.  Still, the question of ethno-racial and religious differences in the outgroup hierarchy would not have been answered in the CES, so the Canadian field is still waiting for a comprehensive dataset for the study of ethnic attitudes. Implications Notwithstanding the limits of the dissertation, there is much to be gleaned.  Future research could pursue a host of implications found in the chapters.  For example, how do non-whites assess single-categorized and cross-categorized outgroups?  Work in the 1970s and 1990s suggests that the ordering of outgroups by non-whites will be remarkably similar to whites, with the exception of the minority group in question, rather than whites, at the top of the hierarchy (e.g., Berry et al. 1977; Berry and Kalin 1995).  But, is this still the case, especially with second- or third-generation visible minority individuals?  And how might it matter to the outgroup hierarchy if the perceiver is an atheist or agnostic?  Given the growing number of non-believers in Canada, their outgroup rankings could become the social norm in the near future.49  What about the effect of an ethnically diverse context on non-white attitudes in other countries with strong multiculturalism policies, such as Australia (e.g., Banting et al. 2006)?  Would we find similar evidence for the reduction of subjective uncertainty?  Does this mean, as the diversity and the welfare state literature suggests, that institutions matter for the                                                 49 According to Statistics Canada, 24% of Canadians reported having no religious affiliation in 2011.  In 2001, the number of non-believers was 17%.  See 2011 National Household Survey.  154 ?contact vs. conflict? debate?  Do liberal egalitarian values work to override negative stereotypes of ethnic groups other than Muslims?  Do these values work in this manner in other political cultures, such as in countries with low multicultural policy contexts?  Or do these values need to be commonly lauded parts of the national ethos or prominent in government policy structures, legitimizing and clarifying them in the minds of citizens?  Finally, how do attitudes about and held by Canada?s indigenous peoples fit?  As mentioned, I omitted indigenous peoples from the minorities considered here, since their place in Canada?s political history is significantly different.  But, an excellent future study would compare the foundations of indigenous, immigrant minority and majority Canadian attitudes. Clearly, there are many ways to apply and extend the lessons found here.  These suggested routes of future research are hardly the only ones.  I would only encourage researchers to continue pursuing this relatively new interest into the roots of attitudes about ethnic diversity in Canada.  There is much work to be done if we are to fully understand why Canadians say what they do about these issues.  In a country where ethnic diversification shows no signs of abating, this is certainly a worthy goal.    155 BIBLIOGRAPHY  Abu-Laban, Yasmeen. 2002. ?Liberalism, Multiculturalism and the Problem of Essentialism.? Citizenship Studies, 6, 459-482.  Abu-Laban, Yasmeen and Daiva Stasiulis. 1992. ?Ethnic Pluralism under Siege: Popular and Partisan Opposition to Multiculturalism.? Canadian Public Policy, 18, 365-386.  Adams, Michael. 2007. Unlikely Utopia: The Surprising Triumph of Canadian Pluralism. Toronto: Penguin Canada.  Adorno, Theodor W., Else Frenkel-Brunswik, Daniel Levinson, and Nevitt Sanford. 1950. The Authoritarian Personality. New York: Harper.  Alba, Richard D., and Victor G. Nee. 2005. Remaking the American Mainstream: Assimilation and Contemporary Immigration. Harvard University Press.  Alcoba, Natalie. ?Ontario rejects sharia law,? National Post. September 12 2005, A1.  Alesina, Alberto and Eliana La Ferrara. 2002. ?Who Trusts Others?? Journal of Public Economics, 85, 207-234.  Allport, Gordon. 1954. The Nature of Prejudice. Cambridge: Perseus Books.  Altemeyer, Bob. 1998. ?The ??Other?? Authoritarian Personality.? In Mark P. Zanna (ed.), Advances in experimental social psychology, Vol. 30, San Diego: Academic Press, 47-91.  Arends-T?th, Judit and Fons J.R. Van De Vijver. 2003. ?Multiculturalism and Acculturation: Views of Dutch and Turkish-Dutch.? European Journal of Social Psychology, 33, 249-266.  Banting, Keith, Richard Johnston, Will Kymlicka, and Stuart Soroka. 2006 ?Do Multiculturalism Policies Erode the Welfare State? An Empirical Analysis.? In Will Kymlicka and Keith G. Banting, (eds.). Multiculturalism and the Welfare State: Recognition and Redistribution in Contemporary Democracies. Oxford University Press, 49-90.  Bar-Tal, Daniel. 1996. ?Development of Social Categories and Stereotypes in Early Childhood: The Case of ?The Arab? Concept Formation, Stereotype and Attitudes by Jewish Children in Israel.? International Journal of Intercultural Religion, 20, 341-370.  Bartels, Larry M. 1996. ?Uninformed Votes: Informational Effects in Presidential Elections.? American Journal of Political Science, 40, 194-230.  Bergeman, C.S., Heather M. Chipuer, Robert Plomin, Nancy L. Pedersen, G. E. McClearn, John R. Nesselroade, Paul T. Costa, Jr., and Robert R. McCrae. 1993. ?Genetic and Environmental Effects on Openness to Experience, Agreeableness, and Conscientiousness: An Adoption/Twin Study.? Journal of Personality, 61, 159-179.  156  Berry, John W. 2001. "A Psychology of Immigration." Journal of Social Issues, 57: 615-631.  Berry, John W. and Rudolf Kalin 1995. ?Multicultural and Ethnic Attitudes in Canada: An Overview of the 1991 National Survey.? Canadian Journal of Behavioural Science, 27: 301-320.  Berry, John W., Rudolf Kalin and Donald Taylor. 1977. Multiculturalism and Ethnic Attitudes in Canada. Ottawa: Ministry of Supply and Services.  Biles, John, and Humera Ibrahim. 2005. ?Religion and Public Policy: Immigration, Citizenship, and Multiculturalism?Guess Who?s Coming to Dinner.? In Paul Bramadat and David Seljak, (eds.), Religion and Ethnicity in Canada. University of Toronto Press, 154-177.  Bilodeau, Antoine and Mebs Kanji. 2006. ?Political Engagement Among Immigrants in Four Anglo-Democracies.? Electoral Insight, 8: 43-49.  Bilodeau, Antoine, Luc Turgeon, and Ekrem Karakoc. 2012. ?Small Worlds of Diversity: Views toward Immigration and Racial Minorities in Canadian Provinces.? Canadian Journal of Political Science, 45, 579-605.  Bilsky, Wolfgang and Shalom H. Schwartz. 1994. ?Values and Personality.? Journal of Personality, 8, 163-181.  Bissoondath, Neil. 1994. Selling Illusions: The Cult of Multiculturalism in Canada. Toronto: Penguin Press.  Blake, Donald E. 2003. ?Environmental Determinants of Racial Attitudes Among White Canadians.? Canadian Journal of Political Science, 36, 491-509.  Blalock, Hubert M. 1967. Toward a Theory of Minority-Group Relations. New York, Wiley.  Blumer, Herbert. 1958. ?Race Prejudice as a Sense of Group Position.? The Pacific Sociological Review, 1, 3-7.  Bobo, Lawrence and Frederick C. Licari. 1989. ?Education and Political Tolerance: Testing the Effects of Cognitive Sophistication and Target Group Affect.? Public Opinion Quarterly, 53, 285-308.  Bobo, Lawrence. 1983. ?Whites' Opposition to Busing: Symbolic Racism or Realistic Group Conflict?? Journal of Personality and Social Psychology, 45, 1196-1219.  Bobo, Lawrence. 1991. ?Social Responsibility, Individualism, and Redistributive Policies.? Sociological Forum, 6, 71-92.   157 Bobo, Lawrence. 1999. ?Prejudice as Group Position: Microfoundations of a Sociological Approach to Racism and Race Relations.? Journal of Social Issues, 55: 445-72.  Bobocel, D. Ramona, Leanne S. Son Hing, Liane M. Davey, David J. Stanley, and Mark P. Zanna. 1998. ?Justice-Based Opposition to Social Policies: Is It Genuine?? Journal of Personality and Social Psychology, 75, 653-669.  Bodenhausen, Galen V., and Robert S. Wyer. 1985. ?Effects of Stereotypes in Decision Making and Information-Processing Strategies.? Journal of Personality and Social Psychology, 48, 2, 267-282.  Bolce, Louis, and Gerald De Maio. 2008. ?A Prejudice for the Thinking Classes Media Exposure, Political Sophistication, and the Anti-Christian Fundamentalist." American Politics Research, 36, 155-185.  Bonacich, Edna. 1972. ?A Theory of Ethnic Antagonism: The Split Labor Market.? American Sociological Review, 37, 547-559  Brader, Ted, Nicholas A. Valentino, Elizabeth Suhay. 2008. ?What Triggers Public Opposition to Immigration? Anxiety, Group Cues, and Immigration Threat.? American Journal of Political Science, 52, 959-978.  Braithwaite, Valerie. 1997. ?Harmony and Security Value Orientations in Political Evaluation.? Personality and Social Psychology Bulletin, 23, 401-416.  Brewer, Marylinn B. 1979. ?In-Group Bias in the Minimal Intergroup Situation: A Cognitive-Motivational Analysis.? Psychological Bulletin, 86: 307-324.  Brewer, Marilynn B. 1988. ?A Dual Process Model of Impression Formation.? In Thomas Srull and Robert Wyer. (eds.), Advances in Social Cognition, vol. 1, 1-36. Hillsdale, NJ: Erlbaum.  Brewer, Marilynn B. 1991. ?The Social Self: On Being the Same and Different at the Same Time." Personality and Social Psychology Bulletin, 17,, 475-482.  Brewer, Marilynn B. 1999. ?The Psychology of Prejudice: Ingroup Love or Outgroup Hate?? Journal of Social Issues, 55, 429-444.  Brewer, Marilynn B., and Norman Miller. 1988. "Contact and Cooperation: When Do They Work?" In Phyllis Katz and Dalmas Taylor, (eds.), Eliminating Racism, New York: Plenum Press.  Brewer, Marilynn B. and Rupert Brown. 1998. ?Intergroup Relations.? In Daniel Gilbert, Susan Fiske, and Gardner Lindsey, (eds.) Handbook of Social Psychology, vol. 2, New York: Oxford University Press.   158 Brown, Rupert. 2010. Prejudice: Its Social Psychology. (2nd ed.) West Sussex: Wiley and Sons.  Bubeck, Maike and Wolfgang Bilsky. 2004. ?Value Structure at an Early Age.? Swiss Journal of Psychology, 63, 1, 31-41.  Bullock, Katherine H. and Gul Joya Jafri. 2001. ?Media (Mis)Representations of Muslim Women in the Canadian Nation.? Canadian Woman Studies, 20, 2, 35-41.  Burton, Jonathan, Alita Nandi and Lucinda Platt. ?Measuring Ethnicity: Challenges and Opportunities for Survey Research.? Ethnic and Racial Studies, 33:1332-1349.  Campbell, David E. 2006. "Religious 'Threat' in Contemporary Presidential Elections," Journal of Politics, 68: 104-115.  Chandler, Charles R. and Yung-mei Tsai. 2001. ?Social Factors Influencing Immigration Attitudes: An Analysis of Data from the General Social Survey?. Social Science Journal, 38, 177-188.  Chui, Tina, Kelly Tran, and H?l?ne Maheux. 2007. Immigration in Canada: A Portrait of the Foreign-born Population, 2006 Census, Ottawa: Ministry of Industry.   Citrin, Jack, Beth Reingold and Donald P. Green. 1990. ?American Identity and the Politics of Ethnic Change.? Journal of Politics, 52: 1124-1154.  Citrin, Jack, Donald P. Green, Christopher Muste, and Cara Wong. 1997. ?Public Opinion toward Immigration Reform: The Role of Economic Motivations." Journal of Politics, 59, 858-881.  Citrin, Jack, David O. Sears, Christopher Muste, and Cara Wong. 2001. "Multiculturalism in American Public Opinion." British Journal of Political Science, 31, 247-275.  Citrin, Jack, Richard Johnston, and Matthew Wright. 2012. ?Do Patriotism and Multiculturalism Collide? Competing Perspectives from Canada and the United States.? Canadian Journal of Political Science, 45, 531-552.  Coenders, Marcel, Peer Scheepers, Paul M. Sniderman, and Genevi?ve Verberk. 2001. ?Blatant and Subtle Prejudice: Dimensions, Determinants, and Consequences; Some Comments on Pettigrew and Meertens.? European Journal of Social Psychology, 31, 281-297.  Coenders, Marcel, Marcel Lubbers, and Peer Scheepers. 2004. ?Majority Populations? Attitudes towards Migrants and Minorities.? Report for the European Monitoring Centre on Racism and Xenophobia 4, no. 1.  Costa, Paul T. and Robert R. McCrae. 2008. ?The Revised NEO Personality Inventory (NEO-PI-R).? In Gregory Boyle, Gerald Matthews, and Donald Saklofske (eds.), The SAGE  159 Handbook of Personality Theory and Assessment: Personality Measurement and Testing. London: SAGE Publications Ltd., 179-199.  Costa, Paul T., and Robert R. McCrae. 1992. ?Four Ways Five Factors are Basic.? Personality and Individual Differences, 13, 653-665.  Crisp, Richard J. and Miles Hewstone. 2007. ?Multiple Social Categorization.? Advances in Experimental Social Psychology, 39: 163-254.  Devine, Patricia G. 1989. ?Stereotypes and Prejudice: Their Automatic and Controlled Components.? Journal of Personality and Social Psychology, 56, 5-18.  Dollinger, Stephen J., Frederick T. L. Leong, and Shawna K. Ulicni. 1996. ?On Traits and Values: With Special Reference to Openness to Experience.? Journal of Research in Personality, 30, 23-41.  Donakowski, Darrell W., and Victoria M. Esses. 1996. ?Native Canadians, First Nations, or Aboriginals: The Effect of Labels on Attitudes toward Native Peoples.? Canadian Journal of Behavioural Science, 28, 2, 86-91.  Duckitt, John. 2001. ?A Dual-Process Cognitive-Motivational Theory of Ideology and Prejudice.? Advances in Exerimental Social Psychology, 33, 41-113.  Duckitt, John. 1992. The Social Psychology of Prejudice. Westport, CT: Praeger.  Dustmann, Christian and Ian Preston. 2001. ?Attitudes to Ethnic Minorities, Ethnic Context and Location Decisions.? The Economic Journal, 111: 353-373.  Dustmann, Christian, and Ian Preston. 2007 ?Racial and Economic Factors in Attitudes to Immigration." The BE Journal of Economic Analysis & Policy, 7, 1-39.  Echebarria-Echabe, Agustin and Emilia Fernandez Guede 2007. ?A New Measure of Anti-Arab Prejudice: Reliability and Validity Evidence.? Journal of Applied Social Psychology, 37, 1077-1091.  Eichstedt, Julie A. Lisa A. Serbin, Diane Poulin-Dubois, Maya G. Sen. 2002. ?Of Bears and Men: Infants? Knowledge of Conventional and Metaphorical Gender Stereotypes.? Infant Behavior and Development, 25, 296-310.  Ellison, Christopher G. and Daniel A. Powers. 1994. ?The Contact Hypothesis and Racial Attitudes among Black Americans.? Social Science Quarterly, 75, 385-400.  Espenshade, Thomas J., and Charles A. Calhoun. 1993. "An Analysis of Public Opinion toward Undocumented Immigration." Population Research and Policy Review, 12, 189-224.   160 Esses, Victoria M. and R. C. Gardner. 1996. ?Multiculturalism in Canada: Context and Current Status.? Canadian Journal of Behavioural Science, 28:145-152.  Esses, Victoria M., John F. Dovidio, Lynne M. Jackson and Tamara L. Armstrong. 2001. ?The Immigration Dilemma: The Role of Perceived Group Competition, Ethnic Prejudice, and National Identity.? Journal of Social Issues, 57: 389-412.  Fazio, Russell H. and Bridget C. Dunton 1997. ?Categorization by Race: The Impact of Automatic and Controlled Components of Racial Prejudice.? Journal of Experimental Social Psychology, 33, 451-470.  Fetzer, Joel S. 2000. Public Attitudes toward Immigration in the United States, France, and Germany. Cambridge University Press.  Fetzer, Joel S. and J. Christopher Soper. 2003. ?The Roots of Public Attitudes toward State Accommodation of European Muslims' Religious Practices Before and After September 11.? Journal for the Scientific Study of Religion, 42, 2, 247-258.  Fiske, Susan T., and Steven L. Neuberg. 1990. ?A Continuum Model of Impression Formation, from Attention and Interpretation.? In Mark P. Zanna (ed.) Advances in Experimental Social Psychology, New York: Academic Press.  Flynn, Francis J. 2005. ?Having an Open Mind: The Impact of Openness to Experience on Interracial Attitudes and Impression Formation.? Journal of Personality and Social Psychology, 88, 5, 816-826.  Fong, Eric and Rima Wilkes. 2003. ?Racial and Ethnic Residential Patterns in Canada.? Sociological Forum, 18, 577-602.  Gaertner, Samuel L. and John F. Dovidio 1986. ?The Aversive Form of Racism.? In John F. Dovidio and Samuel L. Gaertner (eds.), Prejudice, discrimination, and racism. Orlando, FL: Academic Press, 61-89.  Gaertner, Samuel L., John Dovidio, and Betty A. Bachman. 1996. ?Revisiting the Contact Hypothesis: The Induction of a Common Ingroup Identity.? International Journal of Intercultural Relations. 20, 271-290.  Galabuzi, Grace-Edward. 2006. Canada's Economic Apartheid: The Social Exclusion of Racialized Groups in the New Century. Canadian Scholars Press.  Gibbon, John M. 1938. The Canadian Mosaic: The Making of a Northern Nation, Toronto. Gilroy.  Gidengil, Elisabeth, Andr? Blais, Richard Nadeau, and Neil Nevitte. 2004. ?Language and Cultural Insecurity.? In Alain-G. Gagnon (ed), Qu?bec: State and Society (3rd ed.), Peterborough, Broadview Press, 345-368.  161  Gidengil, Elisabeth, Matthew Hennigar, Andr? Blais, and Neil Nevitte. 2005. ?Explaining the Gender Gap in Support for the New Right: The Case of Canada.? Comparative Political Studies, 38, 1-25.  Gilens, Martin. 1999. Why Americans Hate Welfare: Race, Media, and the Politics of Antipoverty Policy. Chicago: The University of Chicago Press.  Giles, Michael and Melanie A. Buckner. 1993. ?David Duke and Black Threat: An Old Hypothesis Revisited.? Journal of Politics, 55, 702-713.  Giles, Micheal and Kaenan Hertz. 1994. ?Racial Threat and Partisan Identification.? The American Political Science Review, 88, 317-326.  Glaser, James M. 2003. ?Social Context and Inter-Group Political Attitudes: Experiments in Group Conflict Theory.? British Journal of Political Science, 33, 607-620.  Goldberg, Lewis R. 1990. ?An Alternative "Description of Personality": The Big-Five Factor Structure.? Journal of Personality and Social Psychology, 59, 1216-1229.  Goldstein, Jay E. 1985. ?The Prestige Dimension of Ethnic Stratification.? In Rita M. Bienvenue and Jay E. Goldstein (eds.), Ethnicity and Ethnic Relations in Canada (2nd ed.). Toronto: Butterworths, 181-198.  Gonzalez, Karina Velasco, Maykel Verkuyten, Jeroen Weesie and Edwin Poppe. 2008. ?Prejudice towards Muslims in The Netherlands: Testing Integrated Threat Theory.? British Journal of Social Psychology, 47, 667-685.  Gosling, Samuel D. and Oliver P. John. 1999. ?Personality Dimensions in Nonhuman Animals: A Cross-Species Review.? Current Directions in Psychological Science, 8, 3, 69-78.  Gosling, Samuel D., Peter J. Rentfrow, and William B. Swann Jr. 2003. ?A Very Brief Measure of the Big-Five Personality Domains.? Journal of Research in Personality, 37, 504-528.  Greenwald, Anthony G. and Mahzarin R. Banaji. 1995. ?Implicit Social Cognition: Attitudes, Self-Esteem, and Stereotypes.? Psychological Review, 102, 4-27.  Grieve, Paul G. and Michael A. Hogg. 1999. ?Subjective Uncertainty and Intergroup Discrimination in the Minimal Group Situation.? Personality and Social Psychology Bulletin, 25, 926-940.  Ha, Shang E. 2010. ?The Consequences of Multiracial Contexts on Public Attitudes toward Immigration.? Political Research Quarterly, 63, 29-42.   162 Hagendoorn, Louk. 1995. ?Intergroup Biases in Multiple Group Systems: The Perception of Ethnic Hierarchies.? European Review of Social Psychology, 6, 199-228.  Haidt, Jonathan, Evan Rosenberg, and Holly Hom. 2003. ?Differentiating Diversities: Moral Diversity Is Not Like Other Kinds.? Journal of Applied Social Psychology, 33, 1-36.  Hainmueller, Jens, and Dominik Hangartner. 2013. ?Who Gets a Swiss Passport? A Natural Experiment in Immigrant Discrimination.? American Political Science Review, 107, 159-187.  Harell, Allison, Stuart Soroka, Shanto Iyengar, and Nicholas Valentino. 2012. ?The Impact of Economic and Cultural Cues on Support for Immigration in Canada and the United States.? Canadian Journal of Political Science, 45, 499-530.  Hewstone, Miles, Mark Rubin, and Hazel Willis. 2002. ?Intergroup Bias.? Annual Review of Psychology 53, 575-604.  Hiebert Daniel, Nadine Schuurman, and Heather Smith. 2007. Multiculturalism ?On the Ground?: The Social Geography of Immigrant and Visible Minority Populations in Montreal, Toronto, and Vancouver, Projected to 2017. Gatineau, Quebec: Department of Citizenship and Immigration Canada.  Hiebert, Daniel. 2006. ?Winning, Losing, and Still Playing the Game: The Political Economy of Immigration in Canada? Tijdschrift voor Economische en Sociale Geografie, 97, 38-48.  Hitlin, Steven and Jane Allyn Piliavin. 2004. ?Values: Reviving a Dormant Concept.? Annual Review of Sociology, 30, 359-393.  Hoedl, Klaus. 2000. ?Physical Characteristics of the Jews.? Jewish Studies at the Central European University: Public Lectures (1996-1999), Budapest: CEU Press, 59-70.  Hogg, Michael A. 2000. ?Subjective Uncertainty Reduction through Self-Categorization: A Motivational Theory of Social Identity Processes." European Review of Social Psychology, 11, 223-255.  Hogg, Michael A. 2001. ?Social Categorization, Depersonalization, and Group Behavior.? In Michael A. Hogg and R. Scott Tindale, (eds.), Blackwell Handbook of Social Psychology: Group Processes, Malden, MA: Blackwell Publishing, 56-87.  Hogg, Michael A. and Dominic Abrams. 1993. ?Towards a Single-Process Uncertainty-Reduction Model of Social Motivation in Groups.? In Michael A. Hogg and Dominic Abrams, (eds.), Group Motivation: Social Psychological Perspectives, Hemel Hempstead, UK: Harvester Wheatsheaf, 173-190.  Hood, M.V. and Irwin L. Morris. 1998. ?Give Us Your Tired, Your Poor,?But Make Sure They Have a Green Card: The Effects of Documented and Undocumented Migrant Context on Anglo Opinion Toward Immigration.? Political Behaviour, 20, 1-15.  163  Hornsey, Matthew J. and Michael A. Hogg. 2000. ?Assimilation and Diversity: An Integrative Model of Subgroup Relations.? Personality and Social Psychology Review, 4, 143-156.  Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press.  Hughes, Diane, James Rodriguez, Emilie P. Smith, Deborah J. Johnson, Howard C. Stevenson, and Paul Spicer. 2006. ?Parents? Ethnic-Racial Socialization Practices: A Review of Research and Directions for Future Study.? Developmental Psychology, 42, 747-770.  Hurwitz, Jon and Mark Peffley. 1992. ?Traditional versus Social Values as Antecedents of Racial Stereotyping and Policy Conservatism.? Political Behavior, 14, 4, 395-421.  Husbands, Christopher T. 1979. ?The "Threat" Hypothesis and Racist Voting in England and the United States.? In Robert Miles and Annie Phizacklea, (eds.), Racism and Political Action in Britain. London: Routledge, 147-183.  Hyers, Lauri L. and Janet K. Swim. 1998. ?A Comparison of the Experiences of Dominant and Minority Group Members during an Intergroup Encounter.? Group Processes and Intergroup Relations, 1, 143-163.  Jackson, Lynne M. and Victoria M. Esses. 2000. ?Effects of Perceived Economic Competition on People?s Willingness To Help Empower Immigrants.? Group Processes and Intergroup Relations, 3, 419-435.  Jackson, Susan E., Karen E. May, and Kristina Whitney. 1995. ?Understanding the Dynamics of Diversity in Decision-Making Teams.? In Richard A.Guzzo and Eduardo Salas (eds.), Team Effectiveness and Decision Making in Organizations, San Franscico: Jossey Bass, 204-261.  Jehn, Karen A., Gregory B. Northcraft and Margaret A. Neale. 1999. "Why Differences Make a Difference: A Field Study of Diversity, Conflict, and Performance in Workgroups," Administrative Science Quarterly, 44: 741-763.  John, Oliver P. and Sanjay Srivastava. 1999. ?The Big Five Trait Taxonomy: History, Measurement, and Theoretical Perspectives.? In Lawrence A. Pervin and Oliver P. John (eds.), Handbook of Personality: Theory and Research (2nd ed.), New York: Guilford Press, 102-138.  Johnston, Richard, Keith Banting, Will Kymlicka, and Stuart Soroka. 2010. ?National Identity and Support for the Welfare State.? Canadian Journal of Political Science, 43, 2, 349-377.  Jost, John T., Jack Glaser, Arie W. Kruglanski, and Frank J. Sulloway. 2003. ?Political Conservatism as Motivated Social Cognition.? Psychological Bulletin, 129, 3, 339-375.   164 Kalin, Rudolf, and John W. Berry. 1982. ?The Social Ecology of Ethnic Attitudes in Canada.? Canadian Journal of Behavioural Science, 14, 97-109.  Kalin, Rudolf. 1996. ?Ethnic Attitudes as a Function of Ethnic Presence.? Canadian Journal of Behavioural Science, 28, 171-179.  Kalin, Rudolph and John W. Berry. 1996. ?Interethnic Attitudes in Canada: Ethnocentrism, Consensual Hierarchy and Reciprocity.? Canadian Journal of Behavioural Science, 28, 253-261.  Kalkan, Kerem Ozan, Geoffrey C. Layman, Eric M. Uslaner. 2009. ??Bands of Others?? Attitudes toward Muslims in Contemporary American Society.? The Journal of Politics, 71, 3, 1-16.  Kam, Cindy D. and Donald R. Kinder. 2007. ?Terror and Ethnocentrism: Foundations of American Support for the War on Terrorism.? The Journal of Politics, 69, 320-338.  Kemmelmeier, Markus. 2003. ?Individualism and Attitudes Toward Affirmative Action: Evidence From Priming Experiments.? Basic and Applied Social Psychology. 25, 111-119.  Kessler, Alan. 2001. ?Immigration, Economic Insecurity, and the ?Ambivalent? American Public.? Working Paper No. 41. Center for Comparative Immigration Studies, University of California-San Diego.  Kinder, Donald R. and David O. Sears. 1981. ?Prejudice and Politics: Symbolic Racism Versus Racial Threats to the Good Life.? Journal of Personality and Social Psychology, 40, 414-431.  Kinder, Donald R. and Lynn M. Sanders. 1996. Divided By Color: Racial Politics and Democratic Ideals. Chicago: University of Chicago Press.  Kluegel, James R. and Eliot R. Smith. 1983. ?Affirmative Action Attitudes: Effects of Self-Interest, Racial Affect, and Stratification Beliefs on Whites' Views.? Social Forces. 61, 797-824.  Kluegel, James R. and Eliot R. Smith. 1986. Beliefs about Inequality: Americans? Views of What is and What ought to Be. Aldine: New York.  Kymlicka, Will. 1989. Liberalism, Community and Culture. Oxford: Oxford University Press.  Kymlicka, Will. 1995. Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford: Clarendon Press.   Kymlicka, Will. 2002. ?Civil Society and Government: A Liberal-Egalitarian Perspective.? In Nancy L. Rosenblum and Robert C. Post (eds.), Civil Society and Government, Princeton: Princeton University Press, 2002, 79-110.  165  Kymlicka, Will and Keith Banting. 2006. ?Immigration, Multiculturalism, and the Welfare State.? Ethics & International Affairs, 20, 281-304.  Lambert, Ronald D. and James E. Curtis. 1984. ?Quebecois and English Canadian Opposition to Racial and Religious Intermarriage, 1968-1983.? Canadian Ethnic Studies, 16, 30-46.  Lee, Taeku. 2000. "Racial Attitudes and the Color Line(s) at the Close of the Twentieth Century." In Paul M. Ong, (ed.), The State of Asian Pacific America: Transforming Race Relations, Los Angeles: LEAP Asian Pacific American Policy Institute and UCLA Asian American Studies Center, 103-158.  LeVine, Robert A., and Donald T. Campbell. 1972. Ethnocentrism: Theories of Conflict, Ethnic Attitudes and Group Behaviour. New York: Wiley Mayda, Anna Maria. 2006. Who Is Against Immigration? A Cross-Country Investigation of Individual Attitudes toward Immigrants. Discussion Paper for the Institute for the Study of Labor, Bonn, Germany.  Luk, Chung Leung and Michael Harris Bond. 1993. ?Personality Variation and Values Endorsement in Chinese University Students.? Personality and Individual Differences, 14, 3, 429-437.  Macrae, C. Neil, Alan B. Milne, and Galen V. Bodenhausen. 1994. ?Stereotypes as Energy-Saving Devices: A Peek Inside the Cognitive Toolbox.? Journal of Personality and Social Psychology, 66, 1, 37-47.  Macrae, C. Neil, Galen V. Bodenhausen, and Alan B. Milne. 1995. ?The Dissection of Selection in Person Perception: Inhibitory Processes in Social Stereotyping.? Journal of Personality and Social Psychology, 69, 397-407.  Maddux, William W., Jamie Barden, Marilynn B. Brewer, Richard E. Petty. 2005. ?Saying No to Negativity: The Effects of Context and Motivation to Control Prejudice on Automatic Evaluative Responses.? Journal of Experimental Social Psychology, 41,19-35.  Mayda, Anna Maria. 2006. "Who Is Against Immigration? A Cross-Country Investigation of Individual Attitudes toward Immigrants." Review of Economics and Statistics, 88, 510-530.  McCombs, Maxwell E., Donald L. Shaw, David Weaver. 1997. Communication and Democracy: Exploring the Intellectual Frontiers in Agenda Setting Theory. New Jersey: Lawrence Erlbaum.  McCrae, Robert R. 1996. ?Social Consequences of Experiential Openness.? Psychological Bulletin, 120, 3, 323-337.  McCrae, Robert R. and Antonio Terracciano. 2005. ?Universal Features of Personality Traits From the Observer?s Perspective: Data From 50 Cultures.? Journal of Personality and Social Psychology, 88, 3, 547-561.  166  McCrae, Robert R. and Oliver P. John 1992. ?An Introduction to the Five-Factor Model and Its Applications.? Journal of Personality, 60, 175-215.  McCrae, Robert R. and Paul T. Costa. 2008. ?The Five-Factor Theory of Personality.? In Oliver P. John, Richard W. Robins, & Lawrence A. Pervin (eds.), Handbook of Personality: Theory and Research (3rd ed.) New York: Guilford Press, 159-181.  McPherson, Miller, Lynn Smith-Lovin, and James M. Cook. 2001. ?Birds of a Feather: Homophily in Social Networks." Annual Review of Sociology, 27, 415-444.  Meadwell, Hudson. 1993. ?The Politics of Nationalism in Quebec.? World Politics, 45, 203-241.  Mendelberg, Tali. 2001. The Race Card: Campaign Strategy, Implicit Messages, and the Norm of Equality. Princeton: Princeton University Press.  Mondak, Jeffry J. 2010. Personality and the Foundations of Political Behavior. New York: Cambridge University Press.  Money, Jeannette. 1999. Fences and Neighbors: The Political Geography of Immigration Control. Cornell University Press.  Monteith, Margo J. 1993. ?Self-Regulation of Prejudiced Responses: Implications for Progress in Prejudice-Reduction Efforts.? Journal of Personality and Social Psychology, 65, 469-485.  Mullen, Brian, Rupert Brown, and Colleen Smith. 1992. ?Ingroup Bias as a Function of Salience, Relevance, and Status: An Integration.? European Journal of Social Psychology, 22, 103-122.  Mullen, Brian. 2001. ?Ethnophaulisms for Ethnic Immigrant Groups.? Journal of Social Issues, 57, 3, 457-475.  M?ller, Tobias and Silvio H. T. Tai. 2010. ?Who Opposes Immigrants' Integration into the Labor Market? The Swiss Case,? Swiss Journal of Economics and Statistics, 146, 741-767.  Mutz, Diana C., and Jeffery J. Mondak. 2006. ?The Workplace as a Context for Cross?Cutting Political Discourse.? Journal of Politics, 68, 140-155.  Nevitte, Neil, Andre Blais, Elisabeth Gidengil, and Richard Nadeau. 2000. Unsteady State: The 1997 Canadian Federal Election. Don Mills: Oxford University Press.   Noel, Jeffrey G., Daniel L. Wann, and Nyla R. Branscombe. 1995. ?Peripheral Ingroup Membership Status and Public Negativity toward Outgroups.? Journal of Personality and Social Psychology, 68, 127-138.  167  Oliner, Samuel P. and Pearl M. Oliner. 1988. The Altruistic Personality: Rescuers Of Jews In Nazi Europe.  New York: Free Press.  Oliver, J. Eric and Janelle Wong. 2003. ?Intergroup Prejudice in Multiethnic Settings.? American Journal of Political Science, 47, 567-582.  Oliver, J. Eric and Tali Mendelberg. 2000. ?Reconsidering the Environmental Determinants of White Racial Attitudes.? American Journal of Political Science, 44, 574-589.  Palmer, Douglas L. 1996. ?Determinants of Canadian Attitudes Toward Immigration: More Than Just Racism?? Canadian Journal of Behavioural Science, 28, 180-192.  Parrillo, Vincent N. and Christopher Donoghue. 2005. ?Updating the Bogardus Social Distance Studies: A New National Survey.? Social Science Journal, 42, 257-271.  Patchen, Martin, Gerhard Hofmann and William R. Brown. 1980. ?Academic Performance of Black High School Students under Different Conditions of Contact with White Peers.? Sociology of Education, 53, 33-51.  Patsiurko, Natalka, John L. Campbell, and John A. Hall. 2012. ?Measuring Cultural Diversity: Ethnic, Linguistic, and Religious Fractionalization in the OECD.? Ethnic and Racial Studies, 35, 195-217.  Pettigrew, Thomas F. 1998. ?Intergroup Contact Theory.? Annual Review of Psychology. 49, 65-85.  Pettigrew, Thomas F. and Linda R. Tropp. 2000. ?Does Intergroup Contact Reduce Prejudice? Recent Meta-Analytic Findings.? In Stuart Oskamp (ed.), Reducing Prejudice and Discrimination: Social Psychological Perspectives. Mahwah, NJ: Erlbaum, 93-114.  Pettigrew, Thomas F. and Linda R. Tropp. 2006. ?A Meta-Analytic Test of Intergroup Contact Theory.? Journal of Personality and Social Psychology, 90, 751-783.  Pineo, Peter C. 1977. ?The Social Standing of Ethnic and Racial Groupings.? Canadian Review of Sociology, 14, 147-157.  Porter, John. 1965. The Vertical Mosaic: An Analysis of Social Class and Power in Canada. Toronto: University of Toronto Press, 1965.  Pratto, Felicia, Jim Sidanius, and Shana Levin. 2006. ?Social Dominance Theory and the Dynamics of Intergroup Relations: Taking stock and looking forward.? European Review of Social Psychology, 17, 271-320.   168 Pratto, Felicia, Jim Sidanius, Lisa M. Stallworth, and Bertram F. Malle. 1994. ?Social Dominance Orientation: A Personality Variable Predicting Social and Political Attitudes.? Journal of Personality and Social Psychology, 67, 4, 741-763.  Putnam, Robert D. 2007. ?E Pluribus Unum: Diversity and Community in the Twenty-First Century.? Scandinavian Political Studies, 30, 137-174.  Qian, Zhenchao. 2005 ?Breaking the Last Taboo: Interracial Marriage in America." Contexts, 4, 33-37.  Quadagno, Jill. 1994. The Color of Welfare: How Racism Undermined the War on Poverty. New York: Oxford University Press.  Quillian, Lincoln. 1995. "Prejudice as a Response to Perceived Group Threat: Population Composition and Anti-Immigrant and Racial Prejudice in Europe." American Sociological Review, 60, 586-611.  Reitz, Jeffrey G. and Raymond Breton. 1994. The Illusion of Difference: Realities of Ethnicity in Canada and the United States. Toronto: CD. Howe Institute.  Reitz, Jeffrey G., and Rupa Banerjee. 2007. ?Racial Inequality, Social Cohesion and Policy Issues in Canada.? In Keith G. Banting, Thomas J. Courchene, and F. Leslie Seidle (eds). Belonging? Diversity, Recognition and Shared Citizenship in Canada. Institute for Research on Public Policy, 1-57.  Reitz, Jeffrey G., Mai B. Phan, and Rupa Banerjee. 2009. Multiculturalism and Social Cohesion: Potentials and Challenges of Diversity. Springer.  Reitz, Jeffrey G., Rupa Banerjee, Mai Phan, and Jordan Thompson. 2009. ?Race, Religion, and the Social Integration of New Immigrant Minorities in Canada.? International Migration Review, 43, 695-726.  Robinson Jr., Jerry W. and James D. Preston. 1976. ?Equal-Status Contact and Modification of Racial Prejudice: A Reexamination of the Contact.? Social Forces, 54, 911-924.  Roccas, Sonia, Lilach Sagiv, Shalom H. Schwartz, and Ariel Knafo. 2002. ?The Big Five Personality Factors and Personal Values.? Personality and Social Psychology Bulletin, 28, 789-803.  Rokeach, Milton and Mezei, Louis. 1966. ?Race and Shared Belief as Factors in Social Choice.? Science, 151, 167-172.  Rokeach, Milton. 1973. The Nature of Human Values. New York: Free Press.  Rokeach, Milton. 1979. Understanding Human Values. New York: Simon & Schuster.   169 Saeed, Amir. 2007. ?Media, Racism and Islamophobia: The Representation of Islam and Muslims in the Media.? Sociology Compass, 1, 443-462.  Saucier, Gerard. 1994. ?Trapnell verses the Lexical Factor: More Ado about Nothing?? European Journal of Personality, 8, 291-298.  Schissel, Bernard, Richard Wanner and James S. Frideres. 1989. ?Social and Economic Context and Attitudes toward Immigrants in Canadian Cities.? International Migration Review, 23, 289-308.  Schuman, Howard, Charlotte Steeh, Lawrence Bobo, and Maria Krysan. 1997. Racial Attitudes in America: Trends and Interpretations. Cambridge: Harvard University Press.  Schwartz, Shalom H. 1992. ?Universals in the Content and Structure of Values: Theory and Empirical Tests in 20 Countries.? In Mark P. Zanna (ed.), Advances in Experimental Social Psychology, Vol. 25, New York: Academic Press, 1-65.  Schwartz, Shalom H. 1994. ?Are there Universal Aspects in the Structure and Contents of Human Values.? Journal of Social Issues, 50, 4, 19-45.  Sears, David O. and Carolyn L. Funk. 1991. ?The Role of Self-Interest in Social and Political Attitudes.? Advances in Experimental Social Psychology. Vol 24. San Diego: Academic Press Inc., 1-92.  Sears, David O., P.J. Henry, and Rick Kosterman. 2000. ?Egalitarian Values and Contemporary Racial Politics.? In David O. Sears, Jim Sidanius, and Lawrence Bobo, (eds.), Racialized Politics: The Debate about Racism in America, Chicago: University of Chicago Press, 75-117.  Seidle, F. Leslie. 2008. ?Testing the Limits of Minority Accommodation in Quebec: The Bouchard-Taylor Commission.? In John Erik Fossum, Johanne Poirier, Paul Magnette (eds.) The Ties that Bind: Accommodating Diversity in Canada and the European Union, Brussels, Belgium: Peter Lang, 77-104.  Serbin, Lisa A., Diane Poulin-Dubois, Karen A. Colburne, Maya G. Sen and Julie A. Eichstedt. 2001. ?Gender Stereotyping in Infancy: Visual Preferences for and Knowledge of Gender-Stereotyped Toys in the Second Year.? International Journal of Behavioral Development, 25, 1, 7-17.  Shaheen, Jack G. 1997. ?Arab and Muslim Stereotyping in American Popular Culture.? Occasional Papers Series. Center for Muslim-Christian Understanding, Georgetown University.  Sherif, Muzafer, O.J. Harvey, B. Jack White, William R. Hood, and Carolyn W. Sherif. 1961. Intergroup Conflict and Cooperation: The Robbers Cave Experiment. Norman, Oklahoma: University Book Exchange.  170  Sibley, Chris G. and John Duckitt. 2008. ?Personality and Prejudice: A Meta-Analysis and Theoretical Review.? Personality and Social Psychology Review, 12, 248-279.  Sidanius, Jim, Felicia Pratto, and Lawrence Bobo. 1996. ?Racism, Conservatism, Affirmative Action, and Intellectual Sophistication.? Journal of Personality and Social Psychology, 70, 476-490.  Sigelman, Lee and Susan Welch. 1993. ?Contact Hypothesis Revisited: Black-White Interaction and Positive Racial Attitudes.? Social Forces, 71, 781-795.  Skitka, Linda J. and Elizabeth Mullen. 2002. ?The Dark Side of Moral Conviction.? Analyses of Social Issues and Public Policy, 2, 35-41.  Sniderman, Paul M. 1975. Personality and Democratic Politics. Berkeley and Los Angeles: University of California Press.  Sniderman, Paul M. and Michael G. Hagen 1985. Race and Inequality: A Study in American Values. Chatham, NJ: Chatham House.  Sniderman, Paul M. David A. Northrup, Joseph F. Fletcher, Peter H. Russell, and Philip E. Tetlock. 1993. ?Psychological and Cultural Foundations of Prejudice: The Case of Anti-Semitism in Quebec.? Canadian Review of Sociology and Anthropology, 30, 242-271.  Sniderman, Paul M. and Thomas Piazza. 1993. The Scar of Race. Cambridge: Harvard University Press.  Sniderman, Paul M. and Edward G. Carmines, 1997. "Reaching Beyond Race," Political Science and Politics. 30, 466-471.  Sniderman, Paul M., Pierangelo Peri, Rui J. P. De Figueiredo, Rui J.P. de Figueiredo Jr., and Thomas Piazza. 2000. The Outsider: Prejudice and Politics in Italy. Princeton: Princeton University Press.  Sniderman, Paul M., Louk Hagendoorn, and Markus Prior. 2004. ?Predisposing Factors and Situational Triggers: Exclusionary Reactions to Immigrant Minorities.? American Political Science Review, 98, 35-49.  Sniderman, Paul M. and Louk Hagendoorn. 2007. When Ways of Life Collide: Multiculturalism and Its Discontents in the Netherlands. Princeton: Princeton University Press.  Soroka, Stuart and Sarah Roberton. 2010. ?A Literature Review of Public Opinion Research on Canadian Attitudes towards Multiculturalism and Immigration, 2006-2009.? Citizenship and Immigration Canada Report.   171 Srivastava, Sanjay, Oliver P. John, Samuel D. Gosling, and Jeff Potter. 2003. ?Development of Personality in Early and Middle Adulthood: Set Like Plaster or Persistent Change?? Journal of Personality and Social Psychology, 84, 1041-1053.  Stephan, Walter G., Cookie White Stephan. 1984. ?The Role of Ignorance in Intergroup Relations.? In Norman Miller and Marilynn B. Brewer, (eds.), Groups in Contact: The Psychology of Desegregation. Orlando, FL: Academic, 229-256.  Stewart, Don and Thomas Hoult. 1959. ?A Social-Psychological Theory of the Authoritarian Personality.? American Journal of Sociology, 65, 3, 274-279.  Stolle, Dietlind, Stuart Soroka and Richard Johnston. 2008. ?When Does Diversity Erode Trust? Neighborhood Diversity, Interpersonal Trust and the Mediating Effect of Social Interactions.? Political Studies, 58, 57-75.  Strabac, Zan and Ola Listhaug. 2008. ?Anti-Muslim Prejudice in Europe: A Multilevel Analysis of Survey Data from 30 Countries.? Social Science Research, 37, 268-286  Strunk, Daniel R and Edward C. Chang. 1999. ?Distinguishing between Fundamental Dimensions of Individualism-Collectivism: Relations to Sociopolitical Attitudes and Beliefs.? Personality and Individual Differences, 27, 665-671.  Tajfel, Henri and John C. Turner. 1979. ?An Integrative Theory of Intergroup Conflict.? In William G. Austin and Steven Worchel (eds.), The Social Psychology of Intergroup Relations, Monterey, CA: Brooks/Cole, 33-47.  Tajfel, Henri. 1982. ?Social Psychology of Intergroup Relations.? Annual Review of Psychology, 33, 1-39.   Taylor, Charles. 1992. ?The Politics of Recognition,? In Amy Gutman, (ed.), Multiculturalism and the "Politics of Recognition". Princeton: Princeton University Press, 25-75.  Taylor, Marylee C. 1998. ?How White Attitudes Vary with the Racial Composition of Local Populations: Numbers Count.? American Sociological Review, 63, 512-535.  Teichman, Yona. 2001. ?The Development of Israeli Children's Images of Jews and Arabs and Their Expression in Human Figure Drawings.? Developmental Psychology, 37, 6, 749-761.  Tesser, Abraham. 2000. ?On the Confluence of Self-Esteem Maintenance Mechanisms.? Personality and Social Psychology Review, 4, 4, 290-299.  Titus, H. Edwin and E. P. Hollander. 1957. ?California F Scale in Psychological Research.? Psychological Bulletin, 54, 1, 47-64.  Trapnell, Paul. 1994. ?Openness versus Intellect: A Lexical Left Turn.? European Journal of Personality, 8, 273-290.  172  Triandis, Harry C. and Earl E. Davis. 1965. ?Race and Belief as Determinants of Behavioural Intentions.? Journal of Personality and Social Psychology, 2, 715-725.  Tropp, Linda R. and Thomas F. Pettigrew. 2005. ?Relationships between Intergroup Contact and Prejudice among Minority and Majority Status.? Psychological Science, 16, 951-957.  Turner, John C., R. J. Brown, Henri Tajfel. 1979. ?Social Comparison and Group Interest in Ingroup Favouritism.? European Journal of Social Psychology, 9, 2, 187-204.  Turner, John. 1999. ?Current Issues in Research ?n Social Identity and Self-Categorization Theories.? In Naomi Ellemers, Russell Spears, and Bertjan Doojse (eds.), Social Identity: Context, Commitment, Content, Oxford: Blackwell, 6-34.  Unkelbach, Christian, Hella Schneider, Kai Gode, Miriam Senft. 2010. ?A Turban Effect, Too: Selection Biases Against Women Wearing Muslim Headscarves". Social Psychological and Personality Science. 1, 378-383.  Unkelbach, Christian, Joseph Forgas, Thomas Denson. 2008. ?The Turban Effect: The Influence of Muslim Headgear and Induced Affect on Aggressive Responses in the Shooter Bias Paradigm.? Journal of Experimental Social Psychology, 44, 1409-1413.  Urban, Lynn M. and Norman Miller. 1998. ?A Theoretical Analysis of Crossed Categorization Effects: A Meta-Analysis.? Journal of Personality and Social Psychology, 74, 894-908.  Uslaner, Eric M. 2010. ?Segregation, Mistrust, and Minorities.? Ethnicities, 10, 415-434.  van Hiela, Alain, Malgorzata Kossowskab, and Ivan Mervielde. 2000. ?The Relationship between Openness to Experience and Political Ideology.? Personality and Individual Differences, 28, 741-751.  van Oudenhoven, Jan Pieter, Charles M. Judd, and Miles Hewstone. 2000. ?Additive and Interactive Models of Crossed Categorization in Correlated Social Categories.? Group Processes and Intergroup Relations, 3, 285-295.  Verkuyten, Maykel, and Barbara Kinket. 2000. ?Social Distances in a Multi Ethnic Society: The Ethnic Hierarchy among Dutch Preadolescents.? Social Psychology Quarterly, 63, 1, 75-85.  Verkuyten, Maykel. 2005. The Social Psychology of Ethnic Identity. New York: Routledge.  Wagner, Ulrich, Oliver Christ, Thomas F. Pettigrew, Jost Stellmacher and Carina Wolf. 2006. ?Prejudice And Minority Proportion: Contact Instead Of Threat Effects.? Social Psychology Quarterly, 69, 380-390.   173 Wainryb, Cecilia, Leigh A. Shaw, Marta Laupa, and Ken R. Smith. 2001. ?Children?s, Adolescents?, and Young Adults? Thinking about Different Types of Disagreements.? Developmental Psychology, 37, 373-386.  Welch, Susan, Lee Sigelman, Timothy Bledsoe, and Michael Combs. 2001. Race and Place: Race Relations in an American City. New York: Cambridge University Press.  Wells, Marlene, and John W. Berry. 1992. ?Attitudes Towards Aboriginal Self-Government: The Influence of Knowledge, and Cultural and Economic Security.? Canadian Journal of Native Studies, 12, 75-93.  White, Stephen, Neil Nevitte, Andr? Blais, Joanna Everitt, Patrick Fournier, Elisabeth Gidengil. 2006. ?Making Up for Lost Time: Immigrant Voter Turnout in Canada.? Electoral Insight, 8, 10-16.  Whitley, Bernard E. 1999. ?Right-Wing Authoritarianism, Social Dominance Orientation, and Prejudice.? Journal of Personality and Social Psychology. 77, 1, 126-134.  Wichman, Aaron L. 2010. ?Uncertainty and Religious Reactivity: Uncertainty Compensation, Repair, and Inoculation.? European Journal of Social Psychology, 40, 35-42.  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, 302-329.  Winter, David G., Oliver P. John, Abigail J. Stewart, Eva C. Klohnen, Lauren E. Duncan. 1998. ?Traits and Motives: Toward an Integration of Two Traditions in Personality Research.? Psychological Review, 105, 230-250.  Wittenbrink, Bernd, Charles M. Judd, and Bernadette Park. 2001. ?Spontaneous Prejudice in Context: Variability in Automatically Activated Attitudes.? Journal of Personality and Social Psychology, 81, 815-827.  Yinger, John Milton. 1994. Ethnicity: Source of Strength? Source of Conflict? Albany: University of New York Press.  Yzerbyt, Vincent, Y., Steve Rocher and Georges Schadron. 1997. ?Stereotypes as Explanations: A Subjective Essentialistic View of Group Perception.? In Russell Spears, Penelope Oakes, Naomi Ellemers and S. Alexander Haslam, (eds.), The Psychology of Stereotyping and Group Life. London: Basil Blackwell.  Online References  Canada. Employment Equity Act. Ottawa. Accessed June 1, 2012.   174 Citizenship and Immigration Canada departmental website.  Accessed February 20, 2013  Curry, Bill. ?Forum dredges up Quebeckers' ire toward minorities.? Globe and Mail. Retrieved March 1, 2012.  Kenney, Jason. ?Economic Growth and Prosperity the Focus of Immigration Changes.? Accessed January 5, 2013.  Statistics Canada. 2011. National Household Survey.  Figure 2.  Ottawa. May 1, 2013.  Trudeau, Pierre Elliot. Parliamentary Speech on Multiculturalism. Accessed June 1, 2013.    175 APPENDIX ONE  CHAPTER 2: QUESTION WORDING  Dependent Variables  Would you feel very comfortable, comfortable, uncomfortable or very uncomfortable if:   Sample a: your boss was Sample b: a teacher in your local school was Sample c: a close relative, like your sister or daughter, was going to marry  Someone who is a fundamentalist Christian Someone who is Jewish Someone who is Black Someone who is Muslim Someone who is Asian  Someone who is an atheist  Independent Variables (% in brackets)  Religious id I will read you a number of factors which may contribute to one's personal feeling of identity.  For each, please tell me whether it is very important, important, not very important, or not at all important to your own sense of identity?  How about: religion?  1 very important   (23)    2  important   (8.9) 3 not very important   (31.2)     4 not at all important   (16.9)   Ethnic id I will read you a number of factors which may contribute to one's personal feeling of identity.  For each, please tell me whether it is very important, important, not very important, or not at all important to your own sense of identity?  How about: ethnicity or race?  1 very important   (21)    2  important   (37.2) 3 not very important   (31.7)     4 not at all important   (10.3)   Education What is the highest level of formal education that you have completed? Grade 8 or less   (1.9) Some high school   (11.5) Complete high school   (19.9) Technical, vocational post-secondary  (24.8) Some university  (15.39)  176 Complete university degree  (17.3) Post graduate degree  (9.3)  Income Which of the following categories best describes your total household income?  That is, the total income of all persons in your household combined, before taxes. Please stop me when I reach your category.  Under $10,000   (5) $10,000 to just under $20,000   (9.8) $20,000 to just under $30,000   (11.3) $30,000 to just under $40,000   (12.3) $40,000 to just under $50,000   (18.5) $50,000 to just under $60,000   (10.9) $60,000 to just under $70,000   (7.9) $70,000 to just under $80,000   (6.4) $80,000 to just under $100,000   (6.7) $100,000 and over   (11.5)  Age What age group do you fall into? 18 to 24   (27.6) 25 to 30   (18.7) 31 to 34   (6.1) 35 to 44   (16.7) 45 to 54   (14) 55 to 64   (9.6) 65 to 74   (5)  75+     (2.4)  Urban [Recorded by interviewer]   0 rural   (18)  1 urban   (82.1)  Sex [Recorded by interviewer]   0 male   (43.1)  1 female   (57)  Quebec [recorded by interviewer]  0 no   (75)     1 yes   (25)     177 Construction of white, mainline Christian sample  Two variables were used to measure if the respondent was a white, mainline Christian.  The first variable measured if the respondent identified with an ancestry judged by the author to be white.  The second variable measured if the respondent identified with a religious denomination judged by the author to be mainline Christian.  The original version of the two variables are as follows:  White We are all Canadians, but our ancestors come from all over the world. How would you describe your own ethnic background?    White respondents were judged to be from the following backgrounds: English/Irish/Scottish/Welsh, French/French Canadian, German, Greek, Italian, Mennonite, Polish, Portuguese, Russian, Scandinavian/Swedish/Norwegian/Denmark, Ukrainian  Non-white respondents were judged to be from the following backgrounds: Chinese/Taiwanese/Hong Kong, East Indian/Pakistani, Filipino, Aboriginal/Native/Metis, ?other?, and ?don?t know? or ?refused?.  The collapsed variable is: White  0 no   (19.7)  1 yes   (80.3)  Mainline Christian  (As mentioned in footnote #12, respondents who identify as ?Christian?, ?Other Christian?, or ?Protestant? are classified as non-mainline because devotees who eschew specific denominational labels are often fundamentalists.)  And what church or religious organization - if any - do you, yourself, identify with?  Mainline Christians were judged to be from the following: Anglican, Baptist, Lutheran, Presbyterian, Catholic, United  Non-mainline Christians were judged to be from the following: Protestant, Christian, other Christian, Atheist/Agnostic, ?other?, ?don?t know?, or ?refused?.  The collapsed variable is: christian  0 no   (41.2)  1 yes   (58.9)     178   CHAPTER 2: DESCRIPTIVES  Variable Obs Mean Std. Dev. Min Max comfort 5880 3.104 0.712 1 4 religious id 5880 2.335 0.946 1 4 ethnic id 5880 2.254 0.872 1 4 education 5880 4.255 1.545 1 7 income 5880 5.533 2.578 1 10 urban 5880 1.798 0.402 1 2 age 5880 3.749 2.051 1 8 sex 5880 1.603 0.489 1 2 Quebec 5880 0.299 0.458 0 1                                179  APPENDIX TWO  CHAPTER 3: QUESTION WORDING  Dependent Variables  immigration In general, immigration is a good thing for Canada.    1 Strongly disagree    2 Disagree    3 DK    4 Agree    5 Strongly agree mosaic Some people say that Canada should be a "melting pot" for people coming here from other countries ? they should give up their cultural differences and become Canadians. Others say that Canada should be a "mosaic", where people are loyal to Canada yet keep many of the customs of their previous countries. How do you feel about this?    1 I favour the ?melting pot? idea    2 No preference/qualified statement/don?t know    3 I favour the ?mosaic? idea  Individual-Level Independent Variables (% in brackets):  power Do you think the following groups have too much power, too little power, or about the right amount of power in our nation's affairs? [Blacks/Asians (Orientals)/East Indians and Pakistanis/Whites    0 Too little    (19.7)    1 Right amount   (70)    2 Too much   (10.3) marriage Do you approve of marriages between: [Whites and Blacks/Whites and Asians (Orientals)/Whites and East Indians or Pakistanis]    0 Yes   (77.1)    1 Undecided/don?t know   (7.2)    2 No   (15.7) security 0 Low household income, unemployed   (42) 1 High household income, employed   (58)  Index of household income and employment status, security, is derived from the following two questions:   Which of the figures below comes closest to your total family income, before      taxes, in [year]?      1 less than $10,000      2 $10,000-19,999       3 $20,000-29,999      4 $30,000-39,999      5 $40,000-49,999  180      6 $50,000-59,999      7 $60,000-69,999      8 $70,000-100,000      9 Over $100,000  At present, are you:     0 Other     1 Employed full-time  satisfaction During the last few years, would you say your financial situation has been:    1 Getting worse   (25.2)    2 Staying about the same   (46.7)    3 Getting better   (28.1) chance Anyone who works hard will rise to the top.    1 Strongly disagree   (7.7)    2 Disagree   (42.3)    3 DK   (2.1)    4 Agree   (34.6)    5 Strongly agree   (13.3) education In terms of formal education, what is the highest level you have completed?    1 Grade School   (7.6)    2 High school   (30.4)    3 Technical or business school   (26.1)    4 Undergraduate degree   (17.8)    5 Graduate or professional school   (15.8)    6 Doctorate degree   (2.4)  white  Are you:   0  Non-white*   (12.5)   1  White   (87.5)   *Non-white includes: Black, Asian (Oriental), East Indian or Pakistani, Other,       Varied, No Answer  CHAPTER 3: DESCRIPTIVES  Variable Obs Mean Std. Dev. Min Max immigration 4406 2.499 1.143 1 5 mosaic 4407 1.924 .906 1 3 power 4406 0.878 0.528 0 2 marriage 4406 0.330 0.680 0 2 security 4406 0.600 0.490 0 1 satisfaction 4406 2.046 0.743 1 3 chance 4406 3.027 1.272 1 5 education 4406 3.160 1.266 1 6 age 4406 1947.352 16.364 1904 1987 sex 4406 1.599 0.490 1 2   181  APPENDIX THREE  CHAPTER 3: DESCRIPTIVES  Variable Obs Mean Std. Dev. Min Max obligations 3472 4.233 .899 1 5 vismin 3472 12.173 14.595 0 51.04 material 3472 .524 .499 0 1      182  APPENDIX FOUR  CHAPTER 5: QUESTION WORDING  Dependent Variable  Fatma 1/Fatma 2 Now we would like to know what you think about multiculturalism programs in Canada. For example, please consider the following story: Fatma is the president of the Canadian Turkish-Muslim Action Network. Her group has recently applied to Canada's Multiculturalism Grants and Contributions Program for $80,000 to fund an outreach project to raise awareness of Turkish-Muslim contributions to Canada's culture. Do you support or oppose the government funding Fatma's outreach project?  1 strongly support  2 somewhat support  3 somewhat oppose  4 strongly oppose  Independent Variables (% in brackets):  open/agreeable  Here are a number of personality traits that may or may not apply to you. Please indicate the extent to which you agree or disagree with each pair of traits. You should rate the extent to which each pair of traits applies to you, even if one characteristic applies more strongly than the other. I see myself as ... [see Table 1 for list of TIPI traits] 1 strongly agree 2 somewhat agree 3 somewhat disagree 4 strongly disagree  The open and agreeable variables are constructed by averaging the two responses to the relevant TIPI questions.  The distributions are as follows:  open low openness (0)  1.5 (0.5)  2 (5.1)  2.5 (19.3)  3 (36)  3.5 (24.7) high openness  (14.4)  agreeable low agreeable (0.13)  183  1.5 (0.5)  2 (4.9)  2.5 (18.4)  3 (30.5)  3.5 (29.3) high agreeable   (16.2)  egalitarian  low egalitarian   (11)  2 (6.5)  3 (24)  4 (9.3)  5 (25.5)  6 (6.9) high egalitarian   (16.8)  [index of egalitarian made from the following:] The government should: 1 See to it that everyone has a decent standard of living  2 Leave people to get ahead on their own Which is more important in a democratic society? 1 Letting the majority decide  2 Protecting the needs and rights of minorities  The welfare state makes people less willing to look after themselves. 1 strongly/somewhat agree 2 strongly/somewhat disagree  household Could you please tell me your total household income before taxes for the year 2010?   0 less than $10 000   (1)  1 $10 000 to $19 999   (6)  2 $20 000 to $29 999   (6.7)  3 $30 000 to $39 999   (10.1)  4 $40 000 to $49 999   (9.3)  5 $50 000 to $59 999   (9.7)  6 $60 000 to $69 999   (10.1)  7 $70 000 to $79 999   (6.5)  8 $80 000 to $89 999   (7.8)  9 $90 000 to $99 999   (4.7)  10 >$99 999   (28)  job Are you currently self employed, working for pay, retired, unemployed or looking for work, a student, caring for a family, or something else?  [original categories are combined into the following:]  0 retired; unemployed/looking for work; student; caring for a family; disabled   (44)  184 1 self employed (with or without employees); working for pay (full or part time,    includes on paid leave); R volunteers works at two or more jobs; student and    working for pay; caring for family and working for pay; retired and working for    pay   (56.2)  security  low chance of losing income,easy to replace   (7.9)  3 (16.7)  4 (8.6)  5 (16.6)  6 (32.5)  7 (8.4)  8 (4)  9 (3.8) high chance of losing income, hard to replace   (1.5)  [index for security made from the following:] How likely is it that this income will be lost in the next year?  1 very likely  2 somewhat likely  3 somewhat unlikely  4 very unlikely  If this income were lost, how easy or difficult would it be to find another source of income or a comparable job?  1 very easy  2 somewhat easy  3 somewhat difficult  4 very difficult  traditional media   [index made of the following:] Generally speaking, how many days in a week do you do the following things? [Watch the news on TV?  Read the news in the newspaper? Listen to news on the radio?]  0 none/never   (3.6) 1 one    (5.4) 2 two   (15.4) 3 three   (24.1) 4 four   (16.2) 5 five   (8.3) 6 six   (8.3) 7 seven/everyday  (11.1)  sophistication  0   (18.9) 1   (23.5)  185 2   (33.9) 3   (23.7)   [index for sophistication made of the following:] Do you happen to recall the name of the Premier of your Province?  0 [incorrect answer]  1 [correct answer] And the name of the federal Minister of Finance?  0 [incorrect answer]  1 [correct answer] And the name of the Governor General of Canada who just finished her term last December?  0 [incorrect answer]  1 [correct answer]  socialize Now thinking about all the people you socialize with, including close friends as well as others from work or elsewhere, are any of them: [Muslim]  0 no   (73.1)  1 yes   (27)  white To what ethnic or cultural group do you belong?  0 non-white   Bangladeshi, Black/African, Chinese, Guyanese, Haitian, Indian, Israeli,    Jamaican, Japanese, Korean, Lebanese, Pakistani, Filipino, Sikh, Sri    Lankan, Tamil, Other Asian, Other South American, Other African, Other    Caribbean, Arabic/Middle Eastern, Inuit/Metis/Aboriginal/Native  1 white   Canada, Australian, Austrian, British, Croation, Czech, Danish, Dutch,    English, French, Finnish, German, Greek, Holland, Hungarian, Irish,    Italian, Jewish/Hebrew, Norwegian, Polish, Portuguese, Russian, Scottish,    Serbian, Slovakian, Spanish, Swedish, Ukranian, Welsh, American, Other    European, Mennonite, Anglo Saxon/WASP/Caucasian, Acadian,    Quebecois/French Canadian/Francophone [if respondent answered ?Canada?, she had a chance to provide another ethnicity.  This second option is combined in the same manner.]            186  CHAPTER 5: DESCRIPTIVES  Variable Obs Mean Std. Dev. Min Max grant 325 2.748 0.905 1 4 condition 325 1.492 0.501 1 2 egalitarian 325 6.178 1.929 3 9 openness 325 3.094 0.572 1.5 4 agreeable 327 3.109 .567 1 4 household 325 7.009 2.938 0 10 traditional  media  325  4.454  1.653  0  7 sophistication 325 0.678 0.308 0 1 socialize 325 0.345 0.476 0 1 age 325 1954.985 13.876 1924 1993 gender 325 0.486 0.501 0 1 white 325 0.957 0.203 0 1   


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items