|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 THE SEPARATION OF NATIONAL PRIDE AND CIVIC ENGAGEMENT IN THE UNITED STATES By MATTHEW FREDERICK VAN VOORHIS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUI REMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011
2 2011 Matthew Frederick Van Voorhis
3 ACKNOWLEDGMENTS I thank my chair, John C. Henretta, whose patient guidance and mentoring over many years tion of this degree. I am also grateful to my committee members, Tamir Sorek, Alin Ceobanu, and Michael Martinez, for their expertise and helpful su ggestions during this project. Finally, I wish to thank all of my family and friends for reasons mostly un related to this dissertation, but I will do so in person rather than on this required page.
4 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .............. 3 LIST OF TABLES ................................ ................................ ................................ .......................... 8 LIST OF FIGURES ................................ ................................ ................................ ...................... 10 ABSTRACT ................................ ................................ ................................ ................................ .. 12 CHAPTER 1 INTRODUCTION TO THE PROBLEM ................................ ................................ .............. 14 America as a Love Story ................................ ................................ ................................ ........ 14 National Pride Defined ................................ ................................ ................................ .......... 15 Civic Engagement Defined ................................ ................................ ................................ .... 16 Relevance ................................ ................................ ................................ ............................... 16 Roadmap ................................ ................................ ................................ ................................ 18 2 NATIONAL PRIDE LITER ATURE REVIEW ................................ ................................ .... 21 Theoretical Approaches to National Identity ................................ ................................ ......... 22 P rimordialism ................................ ................................ ................................ ................. 22 Ethno symbolism ................................ ................................ ................................ ............ 25 Modernism ................................ ................................ ................................ ...................... 27 Distinctions and Di chotomies ................................ ................................ ................................ 30 East vs. West ................................ ................................ ................................ .................. 31 Ethnic vs. Civic ................................ ................................ ................................ .............. 32 Essence vs. Const ruct ................................ ................................ ................................ ..... 32 Cultural vs. Political ................................ ................................ ................................ ....... 35 Nationalism vs. Patriotism ................................ ................................ .............................. 36 Dem ographic Predictors of National Pride ................................ ................................ ............ 37 Social Dominance Theory ................................ ................................ .............................. 37 Race, Ethnicity, and National Identity ................................ ................................ ........... 39 Gender and National Identity ................................ ................................ ......................... 39 Age ................................ ................................ ................................ ................................ 41 Education: A Vehicle for Nation Building or a Pa th to Critical Enlightenment? .......... 42 3 CIVIC ENGAGEMENT LIT ERATURE REVIEW ................................ ............................. 45 Good Citizenship and Civic Engagement ................................ ................................ .............. 45 Theoretical Understandings of Civic Engagement ................................ ................................ 46 ................................ ................................ ..................... 46 Elements of Social Capital ................................ ................................ ............................. 47 Informal sociability ................................ ................................ ................................ 48 Turnout in presidential elections ................................ ................................ ............. 49
5 Associational density ................................ ................................ ............................... 49 Interpersonal social trust ................................ ................................ ......................... 49 Volunteerism ................................ ................................ ................................ ........... 50 Criticisms of the Social Capital Framework ................................ ................................ .. 50 ................................ ................................ ....................... 52 nam ................................ ................................ .................... 53 ................................ ................................ ...................... 54 Elements of Civic Literacy ................................ ................................ ............................. 55 Willingness to engage in public and political affairs: Political interest .................. 56 Ability to engage in public and political affairs: Political knowledge .................... 56 Criticisms of the Civic Literacy Framework ................................ ................................ .. 57 Demographics of Civic Engagement ................................ ................................ ..................... 57 Race, Gender and Age ................................ ................................ ................................ .... 58 Education and Intellectual Ability ................................ ................................ .................. 60 Family Involvement ................................ ................................ ................................ ........ 61 Ope rationalizing Social Capital and Civic Literacy ................................ .............................. 61 Overlapping Features of Social Capital and Civic Literacy ................................ ........... 61 Political particip ation ................................ ................................ .............................. 62 Engagement in public affairs ................................ ................................ ................... 62 Summary of Civic Engagement Components and Subcomponents ............................... 66 4 DATA AND METHODS ................................ ................................ ................................ ...... 69 Dataset ................................ ................................ ................................ ................................ ... 69 Sample ................................ ................................ ................................ ................................ ... 70 ISSP Modules ................................ ................................ ................................ ........................ 72 Limitations of the GSS ................................ ................................ ................................ .......... 73 5 MEASURES ................................ ................................ ................................ .......................... 76 Natio nal Pride Variables ................................ ................................ ................................ ........ 76 Institutional Patriotism ................................ ................................ ................................ ... 76 Apolitical Patriotism ................................ ................................ ................................ ....... 77 Protectionist Nationalism ................................ ................................ ............................... 78 Nativist Nationalism ................................ ................................ ................................ ....... 78 Chauvinist Nationalism ................................ ................................ ................................ .. 79 Civic Engagement Variables ................................ ................................ ................................ 80 Informal Sociability Composite ................................ ................................ ..................... 80 Political Participation in Presidential Elect ions ................................ ............................. 80 Sum of Organizational Memberships ................................ ................................ ............. 81 Interpersonal Trust Index ................................ ................................ ............................... 81 Weekly Frequency of Newspaper Reading ................................ ................................ .... 82 Television Nondependency ................................ ................................ ............................ 82 Level of Political Interest ................................ ................................ ............................... 82 Political Knowledge Index ................................ ................................ ............................. 83 Engagement in Political Affairs ................................ ................................ ..................... 83 Control Variables ................................ ................................ ................................ ................... 84
6 Race ................................ ................................ ................................ ................................ 84 Gender ................................ ................................ ................................ ............................ 85 Age ................................ ................................ ................................ ................................ 85 Education ................................ ................................ ................................ ........................ 85 Family Status ................................ ................................ ................................ .................. 86 Intellectual Ability ................................ ................................ ................................ .......... 86 6 CONCEPTUAL MODEL ................................ ................................ ................................ ...... 89 Primary Concepts: Dependent Variables ................................ ................................ ............... 89 Exogenous and Intermediate Covariates ................................ ................................ ............... 90 7 RESEARCH QUESTIONS A ND HYPOTHESES ................................ ............................... 95 Research Questions ................................ ................................ ................................ ................ 95 Hypotheses ................................ ................................ ................................ ............................. 96 National Pride Hypotheses ................................ ................................ ............................. 97 Civic Engagement Hypotheses ................................ ................................ ....................... 98 8 ANALYSIS ................................ ................................ ................................ .......................... 100 Methodological Constraints ................................ ................................ ................................ 100 Regression Models ................................ ................................ ................................ ............... 100 9 UNIVARIATE AND BIVAR IATE R ESULTS ................................ ................................ 107 Univariate Statistics ................................ ................................ ................................ ............. 107 Civic Engagement ................................ ................................ ................................ ........ 108 Bivariate Relationships ................................ ................................ ................................ ........ 110 Correlation among Independent Variables ................................ ................................ ... 110 Exogenous correlation ................................ ................................ ........................... 110 Intermediate correlations ................................ ................................ ....................... 111 Correlations with National Pride ................................ ................................ .................. 112 Correlations with Civic Engagement ................................ ................................ ............ 114 Summary of Bivariate Findings ................................ ................................ .................... 123 10 REGRESSION RESULTS ................................ ................................ ................................ .. 145 National Pri de Regression Results (Models 1 3) ................................ ................................ 145 Civic Engagement Regression Results (Models 3 7) ................................ .......................... 148 Summary of Findings ................................ ................................ ................................ .......... 154 National Pride Results ................................ ................................ ................................ .. 154 Civic Engagement Results ................................ ................................ ............................ 155 On the Relationship betwe en National Pride and Civic Engagement .......................... 157 11 DISCUSSION AND CONCL USION ................................ ................................ ................. 174 National Pride Discussion ................................ ................................ ................................ .... 174
7 Social Dominance Theory and National Pride ................................ ............................. 17 4 Education and Nationalism ................................ ................................ ........................... 175 Are Patriotism an d Nationalism the Same? ................................ ................................ .. 175 Civic Engagement Discussion ................................ ................................ ............................. 177 Directions for Future Research ................................ ................................ ............................ 178 The Role of Religion ................................ ................................ ................................ .... 178 Tenure Status ................................ ................................ ................................ ................ 179 Geography ................................ ................................ ................................ .................... 181 APPENDIX A VARIABLE CODING, QUE STION WORDING, AND S CALE CONSTRUCTION ...... 184 Patriotism as Domain Specific National Pride ................................ ................................ .... 184 Nationalism Scales ................................ ................................ ................................ ............... 186 Social Capital Scales and Multi item Measures ................................ ................................ .. 188 Civic Literacy Scales and Multi item Measures ................................ ................................ .. 190 LIST OF REFERENCES ................................ ................................ ................................ ............ 197 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ...... 211
8 L IST OF TABLES Table page 4 1 Variable coverage: Concepts and years available ................................ ............................. 75 8 1 Outcome variables and types of regression: GSS, 1996 and 2004 ................................ 104 8 2 Models for civic engagement (1996): Informal sociability, political participation, newspaper reading, and TV nondependency ................................ ................................ .. 105 8 3 Models for civic engage ment (2004): Political participation, associational density, interpersonal trust, political interest, political knowledge, and engagement in public affairs ................................ ................................ ................................ .............................. 106 9 1 Descriptive statistics: GSS, 1996 and 2004 ................................ ................................ .... 127 9 2 .............. 128 9 3 coefficients of the variables in the analysis: GSS, 2004 .............. 129 9 4 Crosstabulation: Mean values for measures of patriotism, by covariates: GSS, 1996 and 2004 ................................ ................................ ................................ .......................... 130 9 5 Crosstabulation: Mean values for measures of nationalism, by covariates: GSS, 1996 and 2004 ................................ ................................ ................................ .......................... 131 9 6 Crosstabulation: Mean values for measures of social capita l, by covariates: GSS, 1996 and 2004 ................................ ................................ ................................ ................. 132 9 7 Crosstabulation: Mean values for measures of civic literacy, by covariates: GSS, 1996 and 2004 ................................ ................................ ................................ ................. 133 10 1 OLS regression of institutional patriotism: GSS, 1996 and 2004 combined .................. 160 10 2 OLS regression of apolitical patriotism: GSS, 1996 and 2004 combined ...................... 161 10 3 OLS regression of protectionism: GSS, 1996 and 2004 combined ................................ 162 10 4 OLS regression of nativism: GSS, 1996 and 2004 combined ................................ ........ 163 10 5 OLS regression of chauvinism: GSS, 1996 and 2004 combined ................................ .... 164 10 6 Logistic regression of informal sociability: GSS, 1996 ................................ .................. 165 10 7 Logistic regression of political participation: GSS, 1996 and 2004 combined .............. 166 10 8 Negative binomial regression models predicting associa tional density: GSS, 2004 ...... 167
9 10 9 OLS regression of interpersonal trust: GSS, 2004 ................................ .......................... 168 10 10 Logistic regression of newspaper rea ding: GSS, 1996 ................................ ................... 169 10 11 OLS regression of television nondependency: GSS, 1996 ................................ ............. 170 10 12 OLS regression of political interest: GS S, 2004 ................................ ............................. 171 10 13 OLS regression of political knowledge: GSS, 2004 ................................ ....................... 172 10 14 Negative binomial regression models for engagement in public affairs: GSS, 2004 ..... 173 11 1 combined ................................ ................................ ................................ ......................... 183 A 1 Patriotism scale constructions: GSS variable names, question wording, and coding for the indices of institutional and apolitical patriotism ................................ ................. 192 A 2 Nationalism scale constructions: GS S variable names, question wording, and coding for the indices of protectionism, nativism, and chauvinism ................................ ........... 193 A 3 Social capital measures: GSS variable names, question wording, and coding for i nformal sociability, political participation, associational density, and social trust ....... 194 A 4 Civic literacy measures: GSS variable names, question wording, and coding for newspaper reading, te levision nondependency, political interest, political knowledge, and engagement in public affairs ................................ ................................ .................... 195 A 5 Operationalization of television nondependency ................................ ............................ 196
10 LI ST OF FIGURES Figure page 3 1 Civic engagement Venn diagram: Conceptual components of social capital and civic literacy ................................ ................................ ................................ ............................... 68 5 1 Venn diagram: Variable coverage of civic engagement measures ................................ ... 88 6 1 Manifestations of national pride and civic engagement ................................ ................... 93 6 2 M anifestations of national pride and civic engagement ................................ ................... 93 6 3 Full conceptual model ................................ ................................ ................................ ....... 94 9 1 Voting in presidential elections by elect ion year. ................................ ........................... 134 9 2 Mean years of education by age group in 1996 and 2004. ................................ ............. 134 9 3 Mean score for intellectual ability by education level among nonwhites and whites. ... 135 9 4 Educational degrees obtained by number of children ................................ ..................... 135 9 5 Institutional patriotism by a ge group: 1996 and 2004 ................................ .................... 136 9 6 Apolitical patriotism by age group: 1996 and 2004 ................................ ........................ 136 9 7 Protectionist nationalism by level of in tellectual ability: 1996 and 2004 ....................... 137 9 8 Nativist nationalism by educational degree obtained: 1996 and 2004 ........................... 137 9 9 Chauvinist na tionalism by educational degree obtained: 1996 and 2004 ....................... 138 9 10 Informal sociability by number of children in 1996 ................................ ....................... 138 9 11 Infor mal sociability by age group in 1996 ................................ ................................ ...... 139 9 12 Electoral participation by educational degree obtained, combined for both waves ....... 139 9 13 Electoral participation by intellectual ability, combined for both waves ....................... 140 9 14 Mean number of associational affiliations, by educational degree in 2004. ................... 140 9 15 Political knowledge by age group in 2004. ................................ ................................ ..... 141 9 16 Scatterplot: Engagement in public affairs by years of education in 2004. ..................... 141 9 17 Institutional patriotism by education level across time periods ................................ ...... 142 9 18 Apolitical patriotism by education level across time periods ................................ ......... 142
11 9 19 Protectionism by education level across time periods ................................ .................... 143 9 20 Nativism by education level across time periods ................................ ............................ 143 9 21 Chauvinism by education level across time periods ................................ ....................... 144
12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE SEPARATION OF NATIONAL PRIDE AND CIVIC ENGAGEMENT IN THE UNITED STATES By Matthew Van Voorhis December 201 1 Chair: John C. Henretta Major : Sociology My project explored factors relating to national pride and civic engagement in the United States of America. To examine the relation between national pride and civic engagement, I consider ed multiple theoretical constructs of each. National pride was treated as having two components : patrio tism and nationalism Civic engagement, also multifaceted was measured using both Robert Putnam s social capital framework and Henry Milner s notion of civic literacy. I examined how national pride and civic engagement related to each other and to indep endent variables such as race, sex, age, education, family status, and intellectual ability. My goal was to answer four questions: ( 1 ) Who expresses high levels of national pride ? ( 2 ) Who has high levels of civic engagement? ( 3 ) H ow are national pride a nd civic engagement related, and in what ways is their association due to common correlation with measured covariates? ( 4 ) Finally, how does the nature of these relationships c hange and fluctuate over time? Since both national feelings and civic behavior in the United States are associated with such factors as race, age, and education, I employed a multivariate analysis to account for these and other exogenous influences affecting my dependent variables. The results of my analysis have several important i mplications for our understanding of American citizenship. First, the study s most consistent finding was that civic engagement levels increase as education increases;
13 this relationship was strong and significant across virtually every regression model th at I ran. Second, despite cultural myths and political propaganda about a deep and profound national devotion that all Americans share toward their nation, it turns out America s most patriotic citizens are no more engaged than the national average in mat ters of public and national concern. In fact, citizens with the strongest nationalistic tendencies tended to display lowest levels of civic engagement, particularly in 2004.
14 CHAPTER 1 INTRODUCTION TO THE PROBLEM America as a Love Story You can t really t ouch your country, but you can serve it. You can t really see your country, but you can love it. America is a love story. A love all Americans share. And from our midst rise extraordinary men and women who lead her to greatness Americans who found the ir love of country so profound that it naturally came first in their lives. Before work. Before friendship. Before self. Through war. Through peace. Despair and success. Our country is here. We are here. Because of their belief in all of us. My n ame is Robert Duvall. For three nights in early September of 2008, a Hollywood actor recited the above words on national television as part of the Republican National Convention Set to tranquil background music and slideshows tailored to each nightly the me, Robert Duvall s monologue was designed to evoke in US citizens not only a sense that we Americans must always put our country before all other matters in our lives, but also that t he best and most honorable citizens among us already did just that i nst inctively and without hesitation. T he party s official slogan that campaign season Country First activated a similar sentiment. A week earlier the Democrats had performed their own rendition of this American tale of love and sacrifice. T he 2008 Democ ratic National Platform was equally fulsome in its depiction of United States citizens (Democratic National Committee, 2008) Altogether, the 57 page platform contained 17 separate complimentary references to the American peop le including four i n the first three pages of text alone, where US citizens were commended for their personal responsibility, character, imagination, diligence, hard work, faith, candor, purposefulness, and compassion (pp. 6 7) The Democratic National Committee praised Americans for their decency and aspirations (p. 55) ; as well as for their dynamism, determination, and innovation (p. 16) while nonetheless being a welcoming and generous people (p. 45) Just
15 like their Republican adversaries the Democratic Party s uncritical portrayal of the US citizenry suggested to t he American people that they, all of them, transcended civic expectations. For Americans who saw themselves as devoted citizens, neither party s message dared to challenge that perception. Both narratives resonated with audiences, as did the imagery of a love story between Americans and their country. The unspecified terms of this romantic arrangement left much to the imagination, so that a wide range of citizens, whose respective contributions to the hallowed American love story varied greatly, could p lausibly imagine that they had adequately demonstrated to their country that she was truly, deeply loved. Flattering though they were, this message to US citizens from our political leaders carried with it an uneasy ambiguity. What did pu tting Country F irst entail? What discernible behavior had merited the effusive list of accolades bestowed on US citizens by Republicans and Democrats? In other words was the any connection between the kinds of Americans who professed to love their country and the kin ds who actively participated to promote its overall wellbeing? To examine this issue I investigated whether love of country defined analytically as national pride is associated with certain behaviors which many observers have viewed as central to a de mocratic society and generally labeled civic engagement. Does national pride in its various manifestations foster greater citizen involvement? Does active social and political participation coincide with national pride, or restrain it? I begin my discus sion with conceptual definitions of national pride and civic engagement. National Pride Defined National pride has been conceptualized in different ways. One of the more common approach es is to distinguish between patriotism and nationalism (Blank & Schmidt, 2003; Coenders & Scheepers, 2003; Doob, 1964; Feshbach, 1994; Sidanius, Feshbach, Levin, & Pratto, 1997; Viroli, 1995) These two terms are often used to describe an individual s passionate
16 attachment to his or her country (Blank & Schmidt, 2003; Coenders & Scheepers, 2003; Viroli, 1995) Hence, both patriotism and nationalism imply an existing subjective positive identification wit h the nation (Blank & Schmidt, 2003, p. 290) Patriotism is characterized by a love or dedication one feels for his country including its institutions, history, and founding principles. Nationalism is intense national devotion that locates one s own nation as superior to others Often, nationalism is motivated by actual fear of immigrants and foreigners; other times it can be classified as intolerance and prejudice Due to its implicit in group / out group social struc ture, nationalism is linked to such ideologies as xenophobia and ethnocentrism (Sidanius, et al. 1997) Civic Engagement Defined The notion of civic engagement is central to the theoretical and empirical project of assessing a good citizenry. Civic en gagement denotes citizens active involvement in community affairs. It entails working to make a difference in the civic life of our communities and developing the combination of knowledge, skills, values and motivation to make that difference (Ehrlich, 2000, p. vi) As with national pride, there are multiple and overlapping dimensions of civic engagement, and different ideas about how it should be measured and defined. One school of thought, led by Robert Putnam (1995; 2000) sees social capital as the prime indicator of civic engagement. A second camp favors a civic literacy framework (Milner, 2002) which stresses knowledgeable p articipation in civic affairs. These two notions, explored in greater depth in Chapter 3, informed my conceptualization of civic engagement measures for this project. Relevance American citizens levels of national pride are among the highest in the world (De Las Casas, 2008; Smith & Jarkko, 1998; Smith & Kim, 2006) particularly since the attacks of
17 September 11, 2001 (T.W. Smith 2005) For some, American pride manifests itself not only as self lov e but also as an aversion to foreigners, newcomers, and visitors. For others, it comes in the form of proudly displaying patriotic symbols on their homes, vehicles or clothing. So salient and socially desirable is patriotism in the US that in 2008, Ame rica s two major political parties vied to one up each other with regard to patriotic expression. Presidential candidates we re publically reprimanded by major news organizations for not wearing flag pins on their lapels not saluting the flag, and other a cts or non acts denoting insufficient patriotism. Are these public performances of patriotism and nationalism buttressed by any behind the scenes hard work on behalf of the nation? Do flag pins and patriotic bumper stickers signify an engaged citizenry or substitute for one ? Is a yellow ribbon an effective way to Support Our Troops, or merely an efficient way to excuse oneself from actively following much less participating in the public political discourse? Scholars from diverse ideological and discipl inary backgrounds have long debated the merits and pitfalls of national pride. Proponents of national pride credit it with decreasing crime, creating wealth, and fighting corruption (De Las Casas, 2008) Its critics argue tha t national pride should also take credit for an array of social problems ranging from unruly populism to genocide (Bar Tal & Staub, 1997; Blank & Schmidt, 2003) While debates about the virtues of national pride per sist, another group of social and political thinkers struggles with the question of what makes a good citizen. How is good citizenship defined, and how might we measure it? What circumstances or social environments tend to foster and sustain good citizen ship? Although answers vary depending on individual research agendas and disciplines, unifying themes do emerge in these ongoing scholarly debates. There is a consensus at least on what good citizenship is not ; few would suggest that societies are
18 better off when citizens are uninformed and lethargic, antisocial and contemptuous of one another, or socially disconnecte d and politically disengaged. Rather, most agree that a good citizen has to be socially involved or politically informed if not both (Milner, 2002; 2010; Putnam, 1996; 2000; Ricci, 2004) Yet, in the eyes of many voters, patriotic gestures that symbolize national affection and allegiance carry at least as much clout as actual civic action. Consider the rampant media criticism Barack Obama en dured during the 2008 presidential primaries for failing to place his hand on his chest during the national anthem, or failing to wear an American flag lap el pin Thus, the two debates about national pride and civic engagement have at least one thing in common ; both function at times as popular answers to the timeless ethical question: how should we live? Roadmap Chapter 1 has briefly introduced the central ideas explored in my project, national pride and civic engagement and clarified why understanding the association between them is worth our time. To investigate the relation between national pride and civic engagement in the United States, I used multip le conceptual models of each. Chapters 2 and 3 extensively sort through these competing conceptual definitions. Specifically, Chapter 2 provides a detailed review of the literature and scholarly debates pertaining to national pride. Chapter 3 does the s ame for civic engagement. Data for the current project were taken from the NORC General Social Survey (GSS). Chapter 4 introduces this dataset. I describe the different modules it includes, and discuss why 1996 and 2004 were the optimal waves for analys is in this stud y. Lastly, I touch on some of the dataset s limitations.
19 In C hapter 5, I discuss the concrete measures constructed for national pride and civic engagement. The se include d two measures of domain specific patriotism ( national pride in instit utional and apolitical achievements ), three measures of nationalism (protectionism, nativism, and chauvinism), and a total of nine measures of civic engagement (associational density, informal sociability, interpersonal trust, political participation, poli tical interest, political knowledge, newspaper reading, television nondependency, and engagement in public affairs) Additionally, I explain how covariates were measured for this project Chapter 6 describes the full conceptual model that informed my data analysis. It begins with a discussion of how my two primary concepts, national pride and civic engagement, are likely to relate to one another. After reviewing my fourteen dependent variables, I introduce my covariates, which consist of three exogenous variables (race, sex, and age) and intermediate controls for education and family status. Chapter 7 presents a series of research questions, and the hypotheses that guided my initial attempts to answer them. In addition to my primary research problem the relation between national pride and civic engagement I identify several secondary research questions. These secondary questions were necessary in order to facilitate a more nuanced understanding of the factors a ffecting national pride and civic engagement individually For each research question, I developed a hypothesis or set of hypotheses, informed by the literature review. I link each model to its respective research question in C hapter 8. Chapter s 9 and 10 present the results of my analysis. Wher e applicable, findings are displayed separately for 1996 and 2004. Chapter 9 lays out the bivariate results. Chapter 10 presents the multiple regression results.
20 Chapter 11 summarizes my findings and offers a series of conclusions. Included is a discu ssion of my project s limitations and suggestions for how future research should address these limitations.
21 CHAPTER 2 N ATIONAL PRIDE LITERA TURE REVIEW Existing research identifies two dimensions of national pride : patriotism and nationalism (Coenders & Scheepers, 2003; Doob, 1964; Feshbach, 1994; Viroli, 1995) B oth concepts signify attitudes of the individual affirming the nation (Blank & Schmidt, 2003, p. 290) P atriotism is a strong positive attachment felt toward one s own nation and its defining symbols (Kosterman & Feshbach, 1989) Nationalism is more of a sense of collective national superiority over other nations and peoples (De Figueiredo & Elkins, 2003; Feshbach, 1994) Some researchers further split the categories to allow for multiple forms of patriotism and nationalism to emerge (Sullivan, Fried, & Dietz, 199 2) For instance, several theorist s contend that patriotism can be divided further into blind patriotism and constructive patriotism (McCleary, Nalls, & Williams, 2009; Parker, 2010; Schatz, Staub & Lavine, 1999; Staub, 1997) While some social scientists see patriotism or nationalism as sub dimensions of more general notions of national attachment (Bar Tal, 1997; Blank & Schmidt, 2003) others work out side the nationalism patriotism spectrum altogether. Mikael Hjerm distinguished four clusters of national attachment civic, ethnic, multiple, and pluralist (1998) ; Frank L. Jones and Philip Smith labeled nationa l feelings as either ascriptive objectivist or subjective voluntarist (2001) This chapter pursues a deeper understanding of national pride, reviewing the theoretical and empirical literature on national identity, pa triotism, and nationalism. I begin with a discussion of the three theoretical paradigms of national identity: primordialism, ethno symbolism, and modernism. Next, I explore the scholarly tendency to divide nationalism into dichotomies such as East/West, e thnic/civic, essence/construct, and cultural/political I then address the fact that all of these dualisms ultimately depict the same basic matrix, and explain
22 which typology is most suitable for with my study s research aims and empirical questions. Th e chapter concludes with a n overview of past empirical findings on socio demographic predictors of national pride. Theoretical Approaches to National Identity Three broad paradigms shaped academic debates about national pride over the last century : primord ialism, ethno symbolism, and modernism. My overview of these competing theories begins with the more essentialist perspectives and progresses toward the more constructionist theories. Primordialism Primordialism was the earliest paradigm of nations and nationality. The umbrella term, primordialist, refers to scholars who contend ed that nations have existed since time immemorial. The common thread among primordialists is their belief in the antiquity of nations and especially in the idea that one s na tionality is an ascribed identity, much like to one s biological sex These scholars consider ed national identity to be a natural part of the human condition. Primordialists argued that every individual carries certain attachments throughout life that a re derived from birthplace, religious beliefs, family relations, linguistic traditions, and various social practices which are (a) natural to the individual; (b) spiritual (Geertz, 1973, pp. 259 260) in charac ter; and which (c) provide the foundation for an easy affinity with other individuals from one s shared background. Primordialists also usually view ethnicity as based on descent. Since few ethnic groups in the modern world can credibly claim that all its members share a known origin, it is not actual descent that defines ethnicity but simply the belief among group members in a common ancestry. Nationalism scholars often credit Edward Shils (1957) as the first to h ave used the term primordialism. From Shils perspective, the affinity one feels toward family as well as
23 toward members of a shared religion or ethnicity can only be characterized as primordial. Another primordialist, Clifford Geertz, employed a sim ilar definition: By the primordial attachment is meant one that stems from the givens social existence: immediate contiguity and kin connection mainly, but beyond them the givenness that stems from being born into a particular religious community, s peaking a particular language, or even a dialect of a language, and following particular social practices. These congruities of blood, speech, custom, and so on, are seen to have an ineffable, and at times overpowering, coerciveness in and of themselves (1973, p. 259) Primordialism developed out of romanticism and the notion of organic nationalism (Kohn, 1944) Th e organic nation w as an alternative to the voluntarist form of nat ionalism Voluntarist nationalism saw the nation as a rational association into which individuals entered based on their own free will. O rganic nationalism, in contrast, regarded nationality as a fixed, permanent attribute that one acquired involuntarily at birth. Anthony D. Smith portrayed the romanticists characterization of nationalism: Organic nationalism holds that the world consists of natural nations, and has always done so; that nations are the bedrock of history and the chief actors in the hist orical drama; that nations and their characters are organisms that can be easily ascertained by their cultural differentiae (1998, p. 146) Primordialist arguments about nationality we re inextricably linked to the scholarly debates on ethnicity. In fact, pure primordialist arguments were often more concerned with the latter. While they viewed ethnicity as the source of national allegiance, Shils and Geertz used the word primordial more to depict what they viewe d as the naturalness of ethnic ties than to advance a theory of nationalism. Indeed, many critics of the paradigm take greater issue with primordialism s view on the source and strength of ethnic identities than with its conception of nations. Rogers Br ubaker calls primordialism a long dead horse that writers on ethnicity and nationalism continue to flog (1996, p. 15) Brubaker adds that
24 [n]o serious scholar today holds the view that is routinely attributed to primordialists in straw man setups, namely that nations or ethnic groups are primordial, unchanging entities (1996, p. 15) But some social scientists continue to accept certain premises of primordialism, even if they reject the theory as a whole. For instance, they may adopt the view that nations have existed since time immemorial, but not that nationality results from any sort of primordial bonds. All of this le d some to suggest that two separate debates w e re actually occurring one between primordialists and instrumentalists 1 about the nature of ethnic attachments, and another between perennialists and modernists about the antiquity of nations (A.D. Smith 1994, p. 376) Such variability necessitates that we understand the difference between primordialism and perennialism, as well as the nuances among different forms of perennialism. Below I provide a brief synthesis of two branches of primordialism, and of its less radical cousin, perennialism. Sociobio logical primordialism. Sociobiologists regard nations and ethnicities as extended kinship groups. Spearheaded by Pierre van den Berghe (1978; 1979) this approach sees nations, ethnic groups, and race s as ultimately resulting from reproductive or genetic processes. From a sociobiological perspective, the nation is a superfamily stemming from common tribal origins. Cultural primordialism. A second variation of primordialism traces nationalism and eth nicity to a cultural essence of human society. Edward Shils (1957) explored the nature of social bonds between individuals of modern societies. He demarcated between the public, civil bonds of the modern state and more primordial bonds of kinship and religion. Shils argued that the latter remained vital, even within modern secular societies, marked by their myths, public ceremonies, and symbols. Perennialism. Some contemporary scholars accept certain premises of p rimordialism while discarding others. Specifically, they adopt the view that nations have existed since time immemorial, but not that nationality results from any sort of primordial bonds. Anthony Smith applied the label perennialism to this less ext reme strain of primordialism. Smith insists that perennialists need not be considered primordialists because it is possible to support the antiquity of ethnic and national attachments without endorsing the view that nationalities are natural. 1 I do not directly discuss instrumentalists in this text as their debate with primordialists about ethnicity is outside the scope of my project.
25 Ethno symb olism E thno symbolism is a term applied to the theories of Anthony D. Smith, John Armstrong, and John Hutchinson. These scholars emphasize the cultural and ethno symbolic nature of nationalism and ethnicity. Whether they adhere to a more phenomenologic al approach and Barthian emphasis on the use of symbols and myths in maintaining ethnic boundaries, or to a more structural and ethno historical approach to the formation of nations, they are critical of what they see as the modernist failure to grasp the recurring nature of ethno symbolic ties and to ground their understanding of modern nations in the longue dure and in earlier ethnic myths, memories, symbols, and traditions (A.D. Smith 1998, p. 6) Ethno symbolism emphasizes the continuity between pre m odern communities and contemporary nations. Challenging the modernist premise that the nation is a distinctly modern phenomenon, this paradigm contends that many of its key elements predate the existence of nations themselves. Ethno symbolists argue th at the fundamental quality of nationalism is based on myths, values, memories, and symbols. From an ethno symbolist perspective: There can be no identity without memory (albeit selective), no collective purpose without myth, and identity, and purpose or d estiny are necessary elements of the very concept of a nation (A.D. Smith 1986, p. 2) Anthony D. Smith, the foremost scholar in the ethno symbolism camp argues that the spirit of what we now understand as national identity can for all nations, be trace d back t o pre modern ethnic sources or an ethnic core (1998, p. 192) which preceded not only contemporary nations but modernity itself (1986) According to Smith long before the emergence of modern nations there existed what he calls ethnies. These collective communities of the past had their own sets of unique characteristics, by which others identified them These distinctive features included : a distinguishing label or name; a historical link to a specific geographic territory or homeland ; common traditions and memories;
26 common cultural elements; a measure of solidarity, among elit es in particular; and, finally a myth of a shared ancestry. Presumably ethnies h ad coalesced in vastly different manners; but whether it was nostalgia, religion, or interstate warfare which had initially united them, the se groups provided a durable sense of identity that eventually became the foundation for the development of territor ialized nations. Smith s ideas were influenced considerably by John Armstrong s macrohistorical analysis (1982) about which Smith remarked: no other work attempts to bring together such a variety of evidence adminis trative, legal, military, architectural, religious, linguistic, social and mythological from which to construct a set of patterns in the slow formation of emergence of modern n ational identities on these patterns of ethnic persistence, and especially on the long term influence of myth symbol complexes (1998, p. 185) In Nations before Nationalism (198 2) Armstrong took an extended temporal perspective (1982, p. 3) that reaches back to the most primitive forms of organized collectivities. Armstrong s stated aim was to examine the emergence of the intens e group identification that today we term a nation (1982, p. 3) He concluded that traces of ethnic consciousness were present in antiquity and the medieval world, thereby situating contemporary nationalism as s imply the final stage in a long cycle of ethnic consciousness that stretched back to ancient civilizations. For Armstrong, the defining quality of this ethnic consciousness was its durability; thus, the development of ethnic identities must be analyzed ac ross a time dimension of multiple centuries. In addition to exploring the persistence of ethnic attachments across time, Armstrong s extended temporal perspective sought to explain the shifting si gnificance of boundaries for human identity (1982, p. 4) According to Armstrong, groups tend to define themselves not by
27 reference to their own characteristics but by exclusion, that is, by comparison to strangers (1 982, p. 5) With his emphasis on ethnic exclusion, and the self perceived borders that dictate the universal com parison with the stranger Armstrong adopted anthropological theorist Fredrik Barth s social interaction model, which allowed for modificat ions in the group s cultural content so long as the boundary mechanisms we re maintained. Much of ethno symbolism developed in an attempt to a straddle the line between primodrialism and a third major paradigm for describin g national identity, modernism. Ethno symbolism s foremost advocate, Anthony D. Smith, admitted that the modernist theoretical approach to nationalism as of the late 1990s, remain [ed] the dominant orthodoxy in the field (1998, p. xii) The f ollowing section details this dominant orthodoxy and outlines its key subcomponents. Modernism Modernism as a general paradigm views nations and nationalism as the product of modernization and of its accompanying transformation of society. This approach t o explaining national identity came about as a response to preceding generations of primordialists who tacitly accepted the basic premises of nationalists themselves. Unifying all scholars in the rather diverse modernist camp is the conviction that natio nalism is a uniquely modern phenomenon. Modernists argue that nations and nationalism came about only during the last few centuries and are the products of distinctly modern processes such as industrialism, capitalism, secularism, and the advent of the b ureaucratic state (1995, p. 29) Classical modernism. The original branch of modernism emerged largely as a project of anti perennialism (A.D. Smith, 1998, p. 18) Classical modernists initial set of proposi tions constituted primarily a rejection of the preceding essentialist paradigms that conceptualized
28 nations as the product of innate and natural historical forces. Classical modernists declared that nations were not ancient and immemorial but relatively r ecent; that one coul d not and should not project the defining components of modern nations onto historical collectives ; and, finally, that nations were the product of recent historical events and of the calculated, rational activity made possible by mode rn conditions. C ultural modernism. The sociocultural modernist perspective stresses the significance of social and cultural transformation in understanding nationalism. Theorists associated with cultural modernism include Ernest Gellner, Benedict Anderso n, Miroslav Hroch, and Rogers Brubaker. In Gellner s view, the emergence of nations at a particular modern moment was a product of civilization s need to transition from one societal form to the next i.e., from an agrarian society where only the rich were allowed to read, to a modern one where communication was paramount. G ellner attribute d nations and nationalism to the homogenizing imperatives of modern industrial society and to minority cultures react ions to such imperatives. So crucial was communica tion, Gellner argue d that it generated the need for nations (1983) Benedict Anderson s account of the advent of nationalism relate d the modern nation to the logical development of capitalism. Anderson argued that nations were imagined political communities, and that their emergence and development coincided with the gradual decline of other forms of imagined communities He attributed t his decline t o modern capitalism and the mass production of print media. Intri gued by his observation that since World War II every successful revolution has defined itself in national terms (1983, p. 2) Anderson set out to explain nationalism by focusing on its cultural origins, and, pr incipally on the transformations of consciousness which he believe d made nations possible. From his perspective, what ma de
29 nationalism truly unique was its ability to seduce people into perceiv ing themselves as intimately linked to strangers who they wou ld never meet. Political Modernism. A third variant of modernism explains nationalism through various political transformations associated with modernity, such as the expansion of suffrage or the emergence of the bureaucratic state. Theorists like John Breuilly, Eric Hobsbawm, Paul Brass, Anthony Giddens, Michael Mann, and Charles Tilly have analyze d nationalism s close relationship to various sources of power i.e., wealthy elites, war, and the modern state. I limit my summary to the ideas of Breuilly and Hobsbawm. John B reuilly (1993) wrote about nationalism as a form of politics. Breuilly s approach interpret s nationalism as a response to the necessities of the modern state. For Breuilly, nationalism refers to po litical movements seeking or exercising state power and justifying such action with nationalist arguments. By nationalist argument, he meant a political doctrine t hat was rooted in three basic assertions: (1) there exists a nation with an explicit and di stinctive character; (2) the values and interests of this nation supersede all other values and interests; and (3) the nation must be as independent as possible, which typically requires the attainment of political sovereignty. Eric Hobsbawm (1983; 1990) maintain ed that the nation was a prominent type of invented tradition. He declared that nationalism s primary historical relevance resided in its political capacity for state making In brief, nation alism comes before nations. Nations do not make states and nationalisms but the other way round (Hobsbawm, 1990, p. 10) Economic Modernism. The final group of modernist theories consider ed here emphasizes the role o f economic transformations in und erstanding national phenomena. Modernism s s ocioeconomic models derive nationalism from the rational workings of the world economy and
30 the economic and social interests of individuals. The two main theorists falling withi n this category are Tom Nairn and Michael Hechter. Nairn (1981) conceptualize d nationalism as a product of, and reaction to, the uneven development of capitalism. A neo Marxist, Nairn described nationalism in material ist terms. In seeking a useful explanatory framework within which to evaluate nationalism, Nairn concluded that the proper framework was that of world history as a whole. To Nairn, nationalism was not derived from a given society s idiosyncratic dynami cs; rather, it was determined by certain features of the world political economy, in the era between the French and industrial revolutions in the present day (1981, p. 332) Hechter s (1998) theories arose out of his detailed exploration of the economic and political development of the British Isles from the English Renaissance through the mid Twentieth century. H echter attribute d nationalism in the Celtic fringe of the Brit ish Isles to divisions of labor that reinforce d cultural differences in that region For Hechter, in an age of mass education and near universal literacy, cultural distinctions took on an elevated importance The social conditions of modernity were the m ain factor that compelled individual members of ethnic groups to stick together and Hechter saw these as problematic. Distinctions and Dichotomies Several binary distinctions emerge from the theoretical literature on nations and nationalism. The most con spicuous of these in the above theoretical discussion is the tension between modernity and antiquity; yet this is not the only dualistic framework for understanding nationalism. Spencer and Wollman identified as many as fourteen dualistic typologies intro duced by theorists attempting to explain nationalism (2005, p. 199) Anthony Smith s comprehensive comparison of theoretical scholarship on nations and nationalism noted that,
31 [ i ] n the writings of scholars on nations and nationalism, three antinomies are frequently proposed: the essence of the nation as opposed to its constructed quality; the antiquity of the nation versus its purely modern appearance; and the cultural basis of nationalism contrasted with its political aspirations and goals. These antinomies are built into both the theories of scholars and the historical scholarship and political activities of nationalists themselves (1998, p. 170) Indeed, the theor etical rift between ethno symbolists and modernists parallels other heuristic dichotomies such as east/west (Kohn, 1944) civic/ethnic (A.D. Smith, 2005) and even good / bad (Spencer & Wollman, 200 5) I will now explore several of these dualisms in greater depth. The long tradition of dividing nationalism into Weberian style ideal types dates back at least to the 1940s, when Hans Kohn (1944) first proposed the East West dichotomy. Since then, the tendency to partition issues of nationality into binary categories has remained more or less constant, while the nuances of the stated theoretical points of departure have shifted and evolved somewhat. Some theore tical frameworks state that there are two types of nations per se ; other s contend that nations were formed in two types of ways; still other theories propose dualistic typolog ies that refer to the two separate formulaic manners in which nationals display t heir allegiance and express their pride In the end, it seems all of these intricate distinctions are part of the same dichotomy. East vs. West The influential distinction between Eastern and Western nationalisms began with the work of Hans Kohn, wh o took the Rhine River to be the geographic dividing line between his two modes of national attachment. According to Kohn s typology, an organic, determinist brand of nationalism dominated the lands located to the east of that partition, in Germany, Italy Eastern Europe and Asia. By contrast, a more rational and voluntaristic brand of nationalism flourish ed exclusively in the Anglo Sax on West.
32 Many have pointed out that the east/west dichotomy is flawed or incomplete (Ceobanu & Escandell, 2008; Hutchi nson, 1987; A.D. Smith, 1983) Even though most contemporary scholars disregard Kohn s geographical demarcation as overly simplistic, there are some who retain the ideological distinction underlying Kohn s typology. Intertwined with the distinction betwe en East and West is the more nuanced (and more fashionable) civic/ethnic dichotomy. Ethnic vs. Civic Scholarly literature on nationalism often differentiates between ethnic and civic forms of national belonging (Brubaker, 199 6; Ignatieff, 1993; Jennings 1 997; Kellas, 1991; A.D. Smith 1986) Membership in an ethnic nation is based on the principle of descent ; the nation is seen as a fictive super family. In an ethnic nation, there is a perception or myth that all members of an ethnic citizenry can trac e their genealogy to shared starting point, and hence, all are related. In contrast to the ethnic model s stressing of (presumed) shared descent, membership in a civic nation is based on one s acceptance of basic principles, beliefs, or institutions. A c ivic conception of nationalism relies on the idea of a patria (Dietz, 1989; Viroli, 1995) A patria refers to a collective of institutions and law B oundaries for membership are presumed to be more permeable in a civi c nation as b elonging is associated with common customs and values. Most scholars categorize the US as a civic nation, along with most Western European nations. At the same time, as with the East/West typology, the ethnic and civic models of national id entity should always be viewed as ideal types rather than accurate descriptions of historical cases, for rhetorical self images and political realities often diverge (Citrin & Wright, 2009, p. 3) Essence vs. Construct The essence/construct dichotomy in theoretical debates about nationalism has to do with whether national identity is innate or a social construction. This debate can be simplified into a debate between essentialists and social constructionists. Essential ism is a multidisciplinary term
33 used to describe theories which argue that some universal or natural trait exists (or should exist) for all cases. Social constructionist theories stress that social life is actively produced among individuals and groups; such approaches emphasize the socially created aspects of society, and portray the social world as created or invented by groups and individual social actors as opposed to simply given or taken for granted (Berger & Luc kmann, 1966; Burr, 1995) The primordialist perspective constitutes a form of essentialism. Primordialist theories such as sociobiology deemphasize the socially created aspect of nations and national identity. Cultural primordialists, such as Clif ford Geert z 2 a nd Edward Shils, have been criticized by Jack Eller and Reed Coughlan (1993) who regard primordialism as incoherent and unsociological. For Eller and Coughlan, the idea of primordiality contains three u nique premises: (1) a given or underived essence of primordial attachments which predates all social interaction; (2) their coercive and overpowering qualities; (3) the emotional essence of primordial attachments. These three premises, according to Elle r and Coughlan, place primordial sentiments completely outside the realm of socially constructed emotions and bonds (1993, p. 187) At the other end of the nature nurture continuum, wherein one side marks the extreme essentialist view and the other represents radical constructivism, are various deconstructionist theoretical developments Included within this broad category are feminist and postmodernist conceptions of nationalism. Feminist contributions inclu de Cynthia Enloe s (1990) discussion of nationalism as a wholly masculine spectacle and Nira Yuval Davis (1997) discussion of the 2 Although Geertz himself never self identified with the primordialist paradigm, Anthony Smith classifies him as a (1998, p. 152)
34 gendered symbolism underlying national projects that personify the nation as a woman. 3 Postmodernist examples include Etienne Balibar s fictive ethnicity (Balibar & Wallerstein, 1992) Stuart Hall s ideas about voluntary ethnicity (Hall, 1992) and historian William H. McNeill s polyethnicity (McNeill, 1986) Anthony Smith attempted to synthesize all of these developments in a thin chapter at the back of his book (1998, pp. 199 220) Perhaps the lone common theme among them, however, is the characterization of nationality as an inherently problematic category. Some theories focus on globalization and the rise of multiculturalism ; others focus on the fragmentation of national identity; still others treat ethnicity as plastic and malleable so cial construction[s], deriving [their] meanings from the particular situations of those who invoke it and the relations of power between individuals and groups (A.D. Smith, 1998, p. 204) Between the two extremes of primordialism, on the one hand, and feminism and postmodernism, on the other are modernism and ethno symbolism. Among modernism and ethno symbolism, the latter embraces a more essentialist model than the former. Most modernis ts see the nation as purely a social construction, germane to a unique point in time. Benedict Anderson (1983) and Eric Hobsbawm (1983; 1990) discuss the nation as an imagined community and as a set of invented traditions, respectively. Within the ranks of modernism itself, some modernists are harsher than others on essentialist nationalisms (for instance, compare Anderson s choice of words, imagination to Ho bsbawm s somewhat more condescending, fabrication ). 3 Is feminism not cont racting itself by labeling nationalism an exclusively male phenomenon, yet also complaining that nationalism tendency to celebrate the nation in feminine terms? Quite oppositely, gender theorists not only are cognizant of the paradox, accentuating such in consistencies is precisely their goal. Yuval which is used selectively by different social agents in various (Yuval Davis 1997:43).
35 E thno symbolism emerged largely as an attempt to repudiate the modernist principle that nations and nationalism were constructed under unique circumstances during a specific historical time period. T he ethno symbolist emphasis on the endurance o f ethnic identities across vastly differing time periods and geographic locations marks it as an essentialist theoretical paradigm at least vis vis modernist perspectives on nationalism Pushing back agains t the modernist view that nations were creations of modernity, ethno symbolism s leading figure, Anthony Smith, wrote that Hobsbawm s and Anderson s theories can be regarded both as Marxian varieties of classical modernism, but also as moving beyond some o f the assumptions of that paradigm Their respective formulations of the invented traditions and the imagined community of the nation have provided the seedbed for more radical postmodernist developments in which the idea of national identity is tre ated as inherently problematic and broken down into its component narratives (1998, pp. 5 6) Cultural vs. Political Still other scholars of nationalism distinguish between political nations, made up of politically aware citizens who are equal before the law, and cultural nations, defined by common language, lineage, religious beliefs, and distinct geographic settlement (Alter, 1994) John Breuilly (1 993) advocated that use of the term nationalism be confined to a purely political movement. Eric Hobsbawm (1990) similarly asserted that nationalism s main relevance to historians was in its political capacity for state making. In response, Anthony D. Smith argued that such is an unduly restrictive approach to understanding nationality, in that it excludes key components of nationalism like culture, identity, and the homeland, resulting in a serious underestim ation of the scope and power of nationalism, and of its ethnic roots (1998, p. 177) John Hutchinson (1987; 1994) made the same point in his analysis of cultur al nationalism. Hutchinson acknowledged the occurrence of political nationalisms aimed at establishing
36 autonomous state institutions, but argued that such was only one small component of the phenomenon, and that t heorists of nationalism must give due weig ht to its cultural elements. Cultural nationalism, Hutchinson argued, seeks amoral regeneration of the community (1994, p. 41) Nationalism vs. Patriotism Lastly, a number of observers differentiate between patr iotism and nationalism (Connor, 1994; Doob, 1964; Feshbach, 1994; Kosterman & Feshbach, 1989; Viroli, 1995) Patriotism usually refers to a love or dedication felt for one s own country incl uding its institutions, history, and founding principles. Nationalism is characterized by intense feelings of national attachment that locate one s own nation as superior to other nations. The French statesman and military leader Charles de Gaulle famous ly said: Patriotism is when love of your own people comes first; nationalism is when hate for people other than your own comes first 4 Thus, while nationalism emphasizes the perceived ethnic inferiority of out groups, patriotism denotes a civic form of att achment involving self referential mechanisms (De Figueiredo & Elkins, 2003, p. 178) For Walker Connor, the distinction between patriotism and nationalism has to do with the difference between state and nation. Connor defines patriotism as the love of one s territorial state and nationalism as a love for one s ethnic nation According to Connor, in a world containing thousands of ethnonational groups and less than two hundred states, it is evident that fo r most people the sense of loyalty to one s nation and to one s state do not coincide. And they often compete for the allegiance of the individual (1994, p. 197) 4 As quoted in Herb Patriotism: Quotations from Around the World (2003, p. 14) and in George W. Nation, State, and Terri tory: Origins, Evolutions, and Relationships (2004, p. 119)
37 When the two loyalties collide, Connor asserte d that powerful allegiance (1994, p. 207) As Spencer and Wollman point ed out, several of the above dualisms may be understood in a sense as part of the same basic m atrix. The contrast specifically between West and East, between the political and cultural, between the civic and the ethnic, are all, we may argue, hewn from the same rock. They emerge to some degree sequentially and to some degree as successive reformu lations (2005, p. 199) For this study I adopted the often used typology conceptual typology that places patriotism or constructive patriotism on one side, and nationalism or blind nationalism on the other ( Blank & Schmidt, 2003; Davidov, 2011; McCleary, et al. 2009; Parker, 2010; Staub, 1997) D emographic Predictors of National Pride Little research exists on the association between national pride and civic engagement, but there is an abundance of data on the relation between national pride and the covariates used in my research model. To conclude my literature review on national pride, this section identifies several trends commonly observed in past studies of national pride. Analyzing the association b etween national pride and civic engagement necessitates an understanding of how both of these primary concepts are affected by socio demographic controls as race, sex, age, education, and marital status. Social dominance theory (Sidanius & Pratto, 1999) provides a starting point for making sense of such matters Social Dominance Theory Proponents of social dominance theory argue that members of society s dominant groups such as whites, males, and privileged social classes feel a greater sense of ownership and attachment toward their nation and its symbols. In any highly stratified social system or nation, pro country feelings will be associated with the dominant group s ideology because disproportionate amounts of power, control, and valued resources are allocated to members of
38 these dominant groups The principle has been applied primarily to exogenous status categories upon which social hierarchies are based, including race, ethnicity, gender, and age (Ishio, 2010; Levin, 2004; Snellman, Ekehammar, & Akrami, 2009; Foels & Reid, 2010) If there are racial or ethnic inequalities within a society, for instance, patriotic and nationalistic values will tend to coincide with the values of the dominant racial or ethnic group (Pea & Sidanius, 2002; Sidanius & Pratto, 1999) A social dominance orientation refers to the extent to which one desires that one s in group dominate and be superior to out groups (Pratto, Sidanius, Stallworth, & Malle, 1994, p. 742) Since those with elevated social status benefit greatly from established social hierarchies, members of high status groups have signif icant incentive to endorse such hierarchies Conversely, lower status groups tend to have lower levels of social dominance orientation (Sidanius & Pratto, 1999) Some social dominance theorists have speculated that one s s upport for group inequalities will increase if one s own group status increases (Foels & Reid, 2010; Levin, 2004) When one s group status increases, one s level of cognitive complexity decreases because one is less mo tivated to form accurate impressions of the social world. With higher status comes the luxury of not needing to fully process social factors, making it easier to endorse group inequalities (Foels & Reid, 2010, p. 690) Social dominance theorists argue that support for social hiera rchies is further aided by various legitimizing myths (Pratto, et al. 1994) one of which is nationalism. According to Pratto and her coauthors, legitimizing myths help to stabilize oppre ssion. That is, they minimize conflict among groups by indicating how individuals and social institutions should allocate things of positive or negative social value, such as jobs, gold, blankets, government appointments, prison terms, and disease (p. 741) Along with nationalism, o ther legitimizing myths include sexism, anti Black racism, ethnic prejudice, cultural elitism, heterosexism, stereotypes, notions of fate and karma, political
39 economic conservati sm, meritocracy, Manifest Destiny, Divine Right of Kings, and Social Darwinism (Pratto, Sidanius, & Levin, 2006, p. 275; Pratto, et al. 1994, pp. 742 743) Informed by a social dominance theory perspective, Yoshito Ishio (2010) recently explored the effects of various socio demographic variables on American patriotism, demonstrating the usefulness of the generalized group dominance perspective in explaining the effects of some socio demographic variables on patriotism. Ishio found that whether socially hierarchical structures are based on race/ethnicity, religion, age, or socioeconomic class, members of socially dominant groups are more emotionally attached to their country than are socially subordinate groups Subord inate groups may feel marginalized and have a weaker sense of ownership of their country and thus less patriotism (2010, pp. 85 86) Race, Ethnicity, and National Identity Studies have consistently shown that American national pride is strongest among white citizens (Carroll, 2007; Ishio, 2010; Pea & Sidanius, 2002; Pratto, et al. 2006) Sidanius and colleagues found that national pride levels were highest among whites and lowest among blacks (Sidanius, et a l. 1997; Sidanius & Pratto, 1999) In addition, social dominance orientation scores correlate with old fashioned and modern racism (Sidanius & Pratto, 1999; Snellman & Ekehammar, 2005) as well as ethnic ranki ng (Snellman & Ekehammar, 2005) These findings are consistent with social dominance theory, since whites constitute the dominant racial status in the United States (Pratto, et al. 2006; Sidanius, et al. 1997; Snellman, e t al. 2009) Both quantitative and qualitative findings suggest that the link is especially strong since 9/11 (Carroll, 2007; Ishio, 2010; Harlow & Dundes, 2004; Shaw, 2004) Gender and National Id entity The relationship between gender and national pride can be explored from a social dominance theory perspective. Men, in general, tend to be more social dominance oriented than women (Pratto, Sidanius, & Stallworth, 1993; Pratto, et al. 2006; Sidani us, Pratto, & Bobo,
40 1994) Links have also been found between social dominance orientation scores and negative opinions toward feminist politics (Bates & Heaven, 2001) Tom W. Smith and Lars Jarkko reported that m en in the Unit ed States tended to have higher levels of national pride than women (Smith & Jarkko, 1998) This finding is especially applicable to my study because out of 23 countries, the investigators observed this sort of gender disp arity i n only one other nation 5 suggesting there may be a gendered component to national attachment that is unique to the US context. Some scholars attribute the gender gap in national identification to a legacy of patriarchal oppression and sexism (Enloe, 1990) which over the years has corroded women s proclivity to honor and support their homeland (Craige, 1996) Others attribute women s comparatively low levels of national pride in America to the fact that men are more likely to join the US military and veterans are more patriotic than nonveterans (Butler & Johnson, 1991) Since the late 1980s a growing body of feminist literature has sought to address the mutual impact of gender and nation (Enloe, 1990; Nagel, 1998; Walby, 1996; Yuval Davis, 1997) Feminist scholars criticize mainstream theories of nationalism for failing to address both (a) the role of wome n in national projects, and (b) how the gender gap impacts our perceptions of nations and nationalism. Anthony D. Smith believes these criticisms are not without merit. [ I ] f the very nature of national projects is gendered, then a separate, or at least, a different kind of theory is required, one which takes this key attribute of the explanandum into account particularly from those who regard nationalism as linked to ethnicity and ethnicity to kinship, or from ethno symbolists for whom ethno history and m yth symbol complexes are central to the development of nations (1998, pp. 205 206) 5 West Germany was the only other nation among which this sort of disparity was observed between men and women.
41 Age Studying the effect of age on national pride is somewhat less conducive to social dominance theory approaches due to the n onlinear way in which age regulates people s access to social esteem and power. Research indicates that older individuals, in general, display greater levels of patriotism and nationalism than do their younger peers (Carroll, 200 7) This positive overall association does not mean the link between age and national pride is a simple linear relation. Neither the very youngest nor very oldest members of society occupy the absolute apex of the age based social hierarchy. Rather, t he age groups who benefit most from such hierarchies are typically somewhere in the middle. Ishio found that age had a parabolic effect (2010, p. 84) on whites patriotism in the US, such that as white Americans grow older from young adulthood to mature adulthood, their patriotism increases sharply Yet, after reaching the ages of 59 or 60, their patriotism declines substantially (2010, p. 84) By construing his cross se ctional results as an age effect, Ishio found support for the social dominance theory argument that members of dominant social groups are more patriotic than were their socially subordinate counterparts 6 Alternatively, one could interpret the variation a mong different age groups in Ishio s cross sectional sample as stemming from cohort differences, rather than age differences. Increasing levels of national pride may be a reflection of one s birth year as opposed to his or her age. Mikael Hjerm further e xplains this standard concern with regard to including age in a cross sectional analysis such as the present study: 6 Ishio s generalized group dominance perspective (2010, p. 67) social dominance theory (Pea & Sidanius, 2002; Sidanius & Pratto, 1999) which, Ishio claims, applied only to racial and ethnic groups. Ishio proposed the original theory be applied to additional social statuses (including age groups, socioeconomic classes, and religious affiliations) which also are characterized by dominant/subordinate social structures. It must be n debates surrounding social dominance theory, although a further discussion of the nature of his confusion is beyond the scope of this text.
42 It is somewhat problematic to analyse age in a cross sectional study The possible age effects can be derived from both cohort as well as l ife cycle effects The former refers to differences between age groups born at different times which are primarily due to a changing society, whereas the latter notion refers to differences between age groups that are due to changes over an individual s l ife. When using longitudinal data it may be possible to assess how likely the two different explanations are. However, in this case this is not possible, since the meaning of the nation is ever changing. This implies that the understanding of the nation could be different for different cohorts (2001) Furthermore, one must be careful about interpretation of cross sectional results when age is included as a covariate, especially in situations where the researcher does not control for education. Because different age cohorts also typically correspond to different levels of educational attainment, any observed relationship between age and nationalistic sentiment (or any other dependent variable) could just as easily be t he indirect effect of one s education as a direct consequence of being young versus old. Education : A Vehicle for Nation Building or a Path to Critical Enlightenment ? S cholars have long been divided on the nature of the relationship between education and n ational pride. The ostensible goal of education is to transmit knowledge to younger generations. In addition to this manifest function educational systems have often fulfilled a latent function : forging a sense of loyalty to the dominant group s culture s and socially constructed norms For Gellner, the public education system s central purpose was to implant in the mass citizenry an extreme and undying national loyalty (1983) F ostering a national consciousness via m ass public schooling fueled the development of s tandardized education systems in the French Third Republic (A.D. Smith 1998, p. 39) and more recently in Japan (Lehmann, 1982) S ome have also argued that American public educa tion self consciously promotes patriotism and the status quo (Loewen, 1995) and even that it should do exactly that (Bennett 1998)
43 But education in general and the environment of academia in particular may actually inhibit or restrain American patriotism. Berger spoke of the denigration of patriotism in important milieus of American intellectual and academic life (1977, p. 128) Brubaker acknowledges that few American schola rs wave flags, and many of us are suspicious of those who do. (2004, p. 118) Students socialized in such an environment may become critical of their own patriotism. Indeed, there seems to be a tension betw een the two roles of the educational system as, on the one hand, a mediator of the dominant cultural, commemoration of imagined nationality and, on the other hand, a promoter of democracy, multiculturalism and ethnic and cultural divergence (Hjerm, 2001, p. 38) To the average citizen, the project of consciously molding children into patriotic citizens may sound fairly innocuous, equivalent attempts to substantiate the more nationalistic, discriminatory forms of nati onal pride would no doubt need to be carried out more subtly. Nevertheless, s ince the educational system is part of the self perpetuating structure of a modern nation state ( T. Smith, 1997) some scholars regard education as an institution tha t facilitate s racism by transmitting the values of an intrinsic racist society (Robinson, 1994) or, at minimum, tacitly contribut es to the institutional discrimination of minorities ( T. Smith, 1997) From this perspective, education might be deemed an enabler of nationalistic sentiment to the extent that these alleged racis t and pro discrimination discourses border on ideologies of protectionism, nativism, or chauvinism. However, Klaus Schleicher s book, Nationalism in Education (1993) while acknowledging this aspect of the educational system, points out that schools are also expected to socialize students to embrace multicultural attitudes. Similarly, Hjerm notes that creating citizens who re the outspoken aim of the educational systems in many modern democracies today (2001, p. 38)
44 D espite theoretical tensions and ambiguities about the relatio nship between education and nationalism there is no such confusion empirically. W hen it comes to how education influences attitudes of ethnic prejudice, e mpirical r esearch consistently finds a negative correlation between educational attainment and vario us indicators of ethnic exclusionism, or ethnocentrism the interrelation between a favorable attitude toward the ethnic in group and unfavorable attitudes toward ethnic out groups (Adorno, Frenkel Brunswik, Lev inson, & Sanford, 1950; LeVine & Campbell, 1972; Sumner, 1959) Support for inequitable treatment of ethnic minorities is most prevalent among lower educated strata. Individuals with higher levels of education are less prejudiced toward ethnic out grou ps (Schuman, Steeh, Bobo, & Krysan, 1997; A.W. Smith 1981; 1985; Vogt, 1997) Similarly, Americans with higher levels of educational attainment are more likely to exhibit supportive attitudes toward racial integration (Taylor, Sheatsley, & Greeley, 1978)
45 CHAPTER 3 CIVIC ENGAGEMENT LIT ERATURE REVIEW Good Citizenship and Civic Engagement Citizenship is an elusive concept in the United States. Defining and assessing good citizenship (Ricci, 2004) has been the impetus of a vast body of literature in the social sciences and humanities. That project has proved challenging, not just for scholars but among the mass citizenry itself. Even though we Americans tend to respect and admire good citizenship we have failed to reach a national consensus as to what that entails. In a society noted for its national pride (De Las Casas, 2008; Smith & Kim, 2006) and self purported civic virtue, what is expected of the ind ividual? To be deemed a good citizen in the United States, what must one do for the sake of America, or on behalf of his or her local community? Despite Americans collectively lacking clear answers to these questions, many people believe that, when ci vic practice does not measure up to its ideal, a vital element is missing from the national landscape (Ricci, 2004, p. 3) David Selbourne (1994) argued that citizens who enjoy the comfort and safety of living within a civic order have a moral duty to promote policies that contribute to maintaining that order, as opposed to merely extracting benefits from it. In other words, as American citizens are entitled to certain rights, so t oo must they fulfill their obligation to use those rights to promote the greater good (Selbourne, 1994) Good citizenship is in many ways a social institution In a sociological context the term institution refers to the organi zed means that each society develops to meet its basic needs. Family, education, economics, politics, law, religion, medicine, science, the military, and the mass media are all considered social institutions. Each institution has its own unique set of no rms, statuses, and values. Like other social institutions, such as a stable economy or an
46 effective education al system society thrives or falters based on the strength and quality of citizenship. As Ricci noted: Americans understand not only that gover nment officials should work properly but also that citizens must help assure the quality of public life. In a democracy citizens rule, yet if they rule badly, all will suffer. Thus it is no exaggeration to say that not just constitutional checks and bala nces but also the practice of good citizenship has helped the nation to establish justice, insure domestic tranquility, provide for the common defense, promote general welfare, and secure liberty (2004, p. 3) Theor etical Understandings of Civic Engagement The scholarly notion of civic engagement is sometimes regarded as a barometer for good citizenship Civically engaged behavior promotes the quality of life in a community by using both political and nonpolitical means. A civically engaged person considers himself or herself part of a larger social structure, and thus treats social problems as personal concerns. Such an individual is willing to see the moral and civic dimensions of issues, to make and justify i nformed moral and civic judgments, and to take action when appropriate (Ehrlich, 2000, p. xxvi) This section discusses the theoretical approaches that inform my study, as well as the classical works from which they derive. My project draws from the theoretical work of two political scientists Robert Putnam (1995; 2000) and Henry Milner (2002) whose work cuts across the soci al science disciplines. Informed by these two thinkers, contemporary debates on the scholarly definition of civic engagement center on the notions of social capital and civic literacy. Putnam s Social Capital Framework The concept of social capital has enjoyed widespread use as a theoretical device (Coleman, 1988; Lesser, 2000; Pichler & Wallace, 2007; Portes, 1998; Putnam, 2000; Uslaner & Brown, 2005) Studies indicat e that societies with high levels of social capital and civic engagement also
47 have better schools, economies, and public services, as well as lower crime rates and more democratic governments (Coleman, 1988 ; Kawachi, Kennedy, & Loochner, 1997; Knack & Keefer, 1997) While this scholarly term has been applied in diverse ways (Boudieu, 1986; Coleman, 1988) I focus mainly on the writings of Robert Putnam (1995; 2000; 2001) as it is within this context that social capital is relevant to my project. Putnam s notion of social capital links the concept to civic engagement. In his formulation, soci al capital includes both formal and informal modes of social interaction among fellow citizens, along with civic attitudes such as trust, honesty, and reciprocity. Putnam s ideas are in some ways a Tocquevillian reworking of the American sociologist Jame s Coleman s work on social capital. Coleman (1990) treated social capital as an asset situated within social structural forces. Influenced by Tocqueville, Putnam shifts his focus away from the structure of social rel ations between individuals, instead analyzing the characteristics of those individuals themselves. Tocqueville believed citizens learn ed the skills of democratic participation as well as such civic virtues as reciprocity and interpersonal trust through pa rticipation in voluntary organizations (1969) Likewise, Putnam sees social capital as a public good inherent in the community in the form of stock from which any citizen may draw. The various components of social capital reinforce one another; their use builds up the stock, and vice versa. Elements of Social Capital Putnam s concept of social capital includes some combination of the following: informal sociability; electoral turnout in presidential elections; assoc iational density; interpersonal trust; and volunteerism.
48 Informal sociability Informal sociability, or what Putnam describes as schmoozing (2000, pp. 93 95) can include such activities as giving dinner part ies, hanging out with friends, playing cards, frequenting bars and nightspots, holding barbecues, visiting relatives, and sending greeting cards. Involvement in any one of these activities, Putnam contends, statistically increases one s likelihood of bein g involved with the others. Putnam explained that informal connections are very important in sustaining social networks. So in our inventory of social capital in America, we need to pay special attention to trends in schmoozing (Putnam, 2000, pp. 95, italics in original) P rior research has tended to focus largely on the contributions of formal social associations to civic engagement paying less attention to the functions of informal social aff iliations 1 As Kwak and coauthors explain ed While formal organizations are easier to observe and analyze, loose and amorphous networks of individuals who come together on a casual basis and at irregular times for leisure activities and socializing may be no less important than formal ones (Kwak, Shah, & Holbert, 2004) A potential benefit of informal associations is their horizontal structure, as opposed to the vertical structure that tends to define more formal organizations (Newton, 1997) For Putnam (1993a) horizontally structured arrangements facilitate the virtues of social capital, establish ing candid communications that produce cooperative actions These modes of social connection lead to an equity and openness in interactions (Newton, 1997) thereby enabling participants to be receptive to knowledge and opportunities that stem from these informal associations 1 However, there are exce ptions (Ganuza & Francs, 2008; Gray, 2009; La Due Lake & Huckfeldt, 1998; Leighley, 1990; Li, Pickles, & Savage, 2005; Newton, 1997; Perren, Arber, & Davidson, 2003; Wi lson & Musick, 1998)
49 Turnout in pres idential elections Putnam included voting in presidential elections as a main form of engaging in public affairs, which was one of five subcategories in his Comprehensive Social Capital Index (2000, p. 291) 2 Com pared to other items in the index, Putnam devoted considerably less time to discussing electoral participation. Nevertheless, of the 14 items used in calculating the overall measure, turnout in presidential elections had the third highest correlation with the final index. Associational density Putnam contended that social networks form the infrastructure of social capital by providing emotional and financial support for individuals and supplying political leverage and volunteers for community instituti ons (2000, p. 312) In Putnam s estimation, membership in voluntary organizations is crucial to accumulating the stock of social capital underpinning civic engagement. Associational membership fosters increased social trust, which thereby facilitates further participation. Putnam s belief that membership in organizations is vital to civic life isn t limited to his studies of social capital in the United States. In his analysis of Italy, Putnam regarded associa tional density as one of the main predictors of regional social capital, demonstrating the effects of differing degrees of membership density on effectiveness of government performance and various other societal outcomes ( 1993a) Interpersonal social trust In the social capital literature, social trust is both a cause and result of high levels of civic engagement (Brehm & Rahn, Individual Level Evidence for the Causes and Consequences of Social Capital, 1997; Putnam, 1993a) Putnam and other proponents of social capital theory regard generalized social trust as a product of interaction and social networking (Brehm & Rahn, 2 Turnout in presidential elections was one of two items Putnam included under this heading. Attending town meetings was the other, although this latter item had a lower correlation with the total score.
50 Individual Lev el Evidence for the Causes and Consequences of Social Capital, 1997; Gittell & Vidal, 1998; Putnam, 2000) 3 Generalized social trust can be seen as a standing decision (Rahn & Transue, 1998, p. 545) to give fel low citizens including those with whom one is not directly or personally acquainted the benefit of the doubt (Yamagishi & Yamagishi, 1994) High societal levels of social trust have been linked to a number of positive outcomes including strong economic performance (Fukuyama, 1995; Knack & Keefer, 1997; Yamagishi & Yamagishi, 1994) support for democratic principles (Muller & Seligson, 1994) and even life expectancy (Kawachi, et al. 1997) Finally, generalized social trust may function as a deterrent of immoral behavior, as citizens who believe others can be trusted are more trustworthy themselves less inclined to cheat, lie, and steal (Rotter, 1980) Volunteerism A fifth subcategory, volunteerism, was also included in Putnam s Comprehensive Social Capital Index (Putnam, 2000, p. 291) Due to data limitations, this component of soc ial capital was not measured in my analysis. While a series of questions was asked about voluntary associations, suitable variables were not available for the specific portion of the samples observed here. Criticisms of the Social Capital Framework Some critics challenge Putnam s emphasis on the correlation between social capital and various positive outcomes in democratic societies. The problem with this aspect of Putnam s theory is that populations with high levels of civic engagement are typically we althy and democratic in the first place, making it hard to ascertain which factor i n this association is the cause and which is the effect. Are societies wealthier and more democratic because their citizens 3 Trust in concrete others such as neighbors and co workers is correlated with high levels of interaction as well (Putnam, 2000)
51 are able to trust and engage in civic life? Or, do democracy and wealth explain why citizens are trusting and engaged (Inglehart, 1997) ? Others are skeptical of Putnam s underlying assumption that generalized social trust and membership in voluntary associations mutually re produce one another (Uslaner, The Moral Foundations of Trust, 2002) Putnam took this premise for granted but neglected to fully test it. When the relationship has been subjected to empirical scrutiny, citizens who scored hi ghest on civic attitudes were not necessarily the ones most likely to join voluntary associations (Claibourn & Martin, 2000; Stolle, 2001; Wollebk & Selle, 2003) In an effort to better understand the causal fl ow of the relationship, scholars have pursued a number of innovative approaches using the duration and intensity of organizational memberships as dynamic variables. To rule out the possibility of self selection effects required the use of longitudinal ana lyses, which were best equipped to test the contemporaneous and lagged effects of associational membership on interpersonal trust, and vice versa. Thus, Claibourn and Martin (2000) analyzed panel data from the Michiga n Socialization Studies (Jennings, Markus, & Niemi, 1991) The data indicated that respondents who joined a group in 1973 were not necessarily more civic in 1982 The authors ultimately concluded that there was no evidence su pporting the hypothesis that interpersonal trust encourages group memberships and only limited evidence suggesting that belonging to groups makes individuals more trusting (Claibourn & Martin, 2000, p. 267) Stolle (2001) explored whether the duration of group memberships influenced the intensity of social trust. If Putnam s proposed internal socialization function of voluntary associations w as valid, then longtime members would d isplay greater amounts of interpersonal t rust t han those who had joined only recently. But Stolle observed that established members of organization s were no more trusting than new members (Stolle, 2001)
52 Lastly, Wollebk and S elle (2003) recognized the need to transcend the simple dichotomy of member vs. nonmember. These authors compared members who were most active in voluntary associations with those who were least active. They discovered that passive organizational members were virtually indistinguishable from active members (Wollebk & Selle, 2003) In short, empirical research largely failed to affirm Putnam s hypothesis that the causal arrow points from joini ng associations to the desired civic behavioral or attitudinal outcomes. The absence of definitive causality leaves open the possibility that certain elements central to Putnam s definition of social capital, rather than enhancing prosperity among citizen s, are merely a byproduct of prosperous societies. Portes critique of Putnam One of the most extensive summaries of the social capital literature to date has been Alejandro Portes overview in Annual Review of Sociology (1998) Although Portes article was written before the book publication of Bowling Alone Putnam had already articulated his theory in a series of articles by that time (1993a; 1993b ; 1996) including one with the same title (1995) about which Portes recounted Putnam s article, Bowling Alone: America s Declining Social Capital, published in the Journal of Democracy in 1995, created something of a sensation, earning for its author a tte tte with President Clinton and a profile in People magazine. The nostalgic image evoked by the lonely bowler resonated with many powerful members of the American establishment and even inspired passages in Cli nton s State of the Union address (1998, pp. 18 19) Portes insisted that enthusiasm for Putnam s version of social capital, and for its expanding applicability to various social problems and processes, is la rgely unwarranted, in part because the set of processes encompassed by the concept are not new and have been studied under other labels in the past Calling them social capital is, to a large extent, just a means of presenting them in a more appealing con ceptual garb (1998, p. 21)
53 Portes called for a more neutral stance toward studying social capital. A more dispassionate approach, he claimed, would enable researchers to consider all facets of the event in ques tion and prevent turning the ensuing literature into an unmitigated celebration of community. Communitarian advocacy is a legitimate political stance; it is not good social science. As a label for the positive effects of sociability, social capital has, in my view, a place in theory and research provided that its different sources and effects are recognized and that their downsides are examined with equal attention (1998, p. 22) According to Portes, the processe s alluded to at the individual level by Putnam s conceptualization of social capital cut both ways Social ties can bring about greater control over wayward behavior and provide privileged access to resources; they can also restrict individual freedoms a nd bar outsiders from gaining access to the same resources through particularistic preferences. For this reason, it seems preferable to approach these manifold processes as social facts to be studied in all their complexity, rather than as examples of a v alue (1998, pp. 21 22) Ultimately Portes concluded: there is little ground to believe that social capital will provide a ready remedy for major social problems, as promised by its bolder proponents. Recent pr oclamations to that effect merely restate the original problems and have not been accompanied so far by any persuasive account of how to bring about the desired stocks of public civicness (1998, p. 21) Milner s critique of Putnam Henry Milner argued that the social capital literature downplayed the importance of political participation. Putnam, though including the steady decline in voter turnout in national elections in his litany of the symptoms of shrinking social capital in America, expends little analytical effort on factors affecting voter turnout. This is in keeping with the Tocquevillian approach in which participating in a voluntary organization is an act of a qualitatively different order than suppor ting a political party. The political participation literature, in contrast, makes no such distinction: voting is a form of participation, different in intensity but not substance from more active forms directed at the political choices of others, such as passing out campaign literature (Milner, 2002, p. 25)
54 Furthermore, Milner asserted that Putnam s social capital centered approach to studying civic engagement has left a gap that impedes our weaving together the str ands of philosophical debate over citizenship and community with the empirical findings of social scientists about social capital and political participation. Missing in the discussion is a clear conceptualization of the knowledge needed to exercise the r esponsibilities of citizenship as we enter the twenty first century (2002, p. 1) Milner sought to fill that scholarly gap with a concept he calls civic literacy : the knowledge and capacity of citizens to enga ge in public discourse and evaluate the performance of those in office. In order to more reliably measure civic engagement, Milner argued for additional indicators not instead of, but in addition to social capital measures. Milner s Civic Literacy Fram ework Expanding upon Putnam s ideas, Milner supplemented civic knowledge where Putnam had envisioned interpersonal trust. Milner explained his rationale behind selecting the phrase civic literacy for his new concept: The word civic is chosen because it combines in one word the notion of exercising one literacy is chosen because it implies that there is a known quantity that is attainable by each individual that cannot be stocked unl ike, say, knowledge or capital which can. A minority can build up a community s stock of physical capital and, by implication, social capital. However, though one person may read much more than another, both contribute equally to the overall literacy rat e of their community (2002, pp. 2 3, italics in original) Milner s framework emphasizes citizens knowledgeable participation in public affairs, a notion inspired by the political philosoph er William Galston (1991) Galston wrote about what he termed the specific political virtue of citizenship, or the willingness and ability to engage in public discourse and evaluate the performance of those in office (1991, p. 227) The intent of a civic literacy centered approach is to operationalize each of these two components willingness and ability as quantifiable measures of civic literacy. Ability,
55 Milner suggested, manifests itself as political knowledge. Willingness emerges in the form of political engagement and political interest. A second influence on Milner s civic literacy framework was the classical political theorist John Stuart Mill (1910) who taught that political participation was the only way for ordinary citizens to ensure that their interests would be accounted for in the decision making process. Through political participation, citizens acquire the skills and kno wledge to act as competent members of the community. Such skills and knowledge then contribute to the overall welfare of that community. Thus, from a civic literacy perspective, the well being of society at large depends on individual citizens willingne ss and ability to remain informed and engaged. At the same time, civic literacy isn t just about staying informed; it s also about how citizens go about informing themselves In societies marked by high civic literacy, citizens are less dependent on tel evision and more apt to learn about world events through reading newspapers, (Milner, 2002; Ricci, 2004) 4 Elements of Civic Literacy If civic literacy intends to assess citizens knowledgeable participation in publ ic affairs, then, following Galston (1991, p. 227) we must devise some way to gauge both (a) the extent to which citizens are willing to participate, and (b) their ability to do so. 4 Theories that assess citizenship on the basis of newspaper reading versus television media consumption may need t o be updated to account for changing forms of media usage. Now the Internet enables people to access current events from both the New York Times and Fox News for instance, is the distinction between periodicals and ew book, The Internet Generation (2010) takes a first step toward addressing for democracy and patt erns of civic participation. He fears that the Internet could exacerbate political inequities, but also acknowledges its potential to mobilize the younger generation in innovative ways.
56 Willingness to engage in publi c and political affairs: Political interest One s willingness to participate can be measured by whether or not one actually does so. Milner adopts a unidimensional approach to measuring political participation. In sharp contrast to Putnam, for whom natio nal electoral turnouts suffices as an adequate measure of political participation, Milner insists that local elections are a better barometer for evaluating participation in political affairs. At its heart, then, the willingness component to civic literac y is also a function of political interest. Ability to engage in public and political affairs: Political knowledge Possessing the ability to participate competently is a function of several factors, according to Milner. First and foremost, Milner argued t hat ability manifests itself in the form of political knowledge. But, as Milner acknowledged, evaluating political knowledge is a complex process, one that must, in a sense, occur indirectly. The effective distribution of political knowledge by the media and political institutions is crucial to Milner s notion of civic literacy. According to Milner, civic literacy is negatively correlated with high levels of television dependency and positively correlated with newspaper reading (2002, pp. 98 104) This perspective is backed by empirical research. For example, Fleming, Thorson, and Peng (2005) observed that newspaper use positively predicted associational membership, reg ardless of race Conversely, the authors found that citizens use of entertainment media, such as television, negatively affected both associational membership and volunteering (Fleming, Thorson, & Peng, 2005) In short, Miln er proposed that political knowledge was conditioned by the act of reading newspapers, as well as by refraining from excessive television consumption.
57 Criticisms of the Civic Literacy Framework The main limitation of the civic literacy approach is that Mil ner s conceptualization of citizenship is too narrow (Galston, 2004; McDevitt, 2003) At times, Milner becomes too enamored of his approach, and civic literacy verges on a single factor explanation for political p articipation (Galston, 2004, p. 1343) Galston further warned that civic literacy, as Milner operationalized it, would prove to be less effective in explaining changes across time than across space (2004, p. 1343) Indeed, a great deal of Milner s effort was aimed at developing measures that could travel well from one geographic location to another. If the appeal of the civic literacy framework is that it facilitates mean ing ful comparisons across cultures, this aspect of Milner s theory is also the source of its main caveat Michael McDevitt remarked that Milner comparative insights and predictive power of his model are formidable, the conceptualization of citizenship is surprisingly narrow. The focus on political knowledge as the essence of citizenship seems largely driven by the pragmatic need to find variables that (a) are measurable as objective ind icators, (b) can be used for national comparisons, and (c) are capable of predicting voting. Milner fails to seriously consider the conceptual limitations of his purely individualistic, cognitive (2003, pp. 433 434) Milner himself acknowledged that voter turnout was selected for its empirical expediency, as opposed to its being an optimal marker for civic literacy as he envisioned the concept. han voter turnout. Comparatively passive though it may be, voting is the only activity that serves the purpose of comparing broad and long term trends in political participation in different societies (Milner, 2002, p. 25) Demographics of Civic Engagement This section summarize s how various demographic factors predicted measures of social capital and civic literacy in past studies.
58 Race Gender and Age Data from a US national sample in 1995 indicated that over 50% of whites had volunteered over the last month, compared to just 35% of blacks (Hodgkinson & Weitzman, 1996, p. D148) As far back as the 1970s, race was identified as a key demographic factor influencing the formatio n of voluntary associations (Blau, 1977) How race mediates the effect of social trust on other civic engagement variables may vary depending on one s geographic region. Kohu t s analysis of Philadelphia residents found that t rust was unrelated to volunteering regardless of race (1998) But a separate study focusing on Alabama residents, indicated that trust had a stronger positive effect on volunteerism for blacks than for whites (Emig, Hesse, & Fisher, 1996) Verba, Schlozman and Brady explored various forms of civic engagement and found that men were more likely than women to make campaign contributions, to contact officials, to be affiliated with a politica l organization, and to work informally in the community (1995, pp. 254 255) By contrast, Putnam found that informal sociability was higher among women than among men (2000, pp. 94 95; 107 108) He reported that regardless of employment and marital status, keeping up with friends and relatives continues to be socially defined as women s work (2000, pp. 94 95) Women we re also more likely to express a sense of concern and responsibility for the welfare of others for example, by doing volunteer work more frequently (Putnam, 2000, p. 95) To be sure, informal connec tions generally do not build civic skills in the ways that involvement in a club, a political group, a union, or the church can, but informal connections are very important in sustaining social networks. So in our inventory of social capital in America, w e need to pay special attention to trends in schmoozing (Putnam, 2000, pp. 95, italics in original)
59 Some research suggests there is a gender gap in political interest. Bennett and Bennett (1989) found that, at all ages, women were less politically interested than men. However, Hayes and Bean (1993) found roughly equivalent levels of political interest among men and wome n in the US E vidence also suggests there is a persistent gender gap in political knowledge in North America Within any given occupation or educational category, women are less politically interested and informed than men (Milner, 2010; Verba, Burns, & Schlozman, 1997) O ne study found that college educated females were no more politically informed than males who had not completed high school (Gidengil, Blais, & Nevitt e, 2004, pp. 49 50) Women also consistently report lower levels of self perceived political knowledge (Gidengil, Giles, & Thomas, 2008) The link between gender and volunteering is complicated. In the United States, wom en are generally more inclined to volunteer than men (Mesch, Rooney, Steinberg, & Denton, 2006; Sundeen, Garcia, & Raskoff, 2009) Among those who do volunteer, women and men contribute roughly the same amount of time (Hodgkinson & Weitzman, 1996, p. D148) Moreover, the relation between gender and volunteering fluctuates across the life course. Among older people, males generally volunteer more hours than females (Wuthnow, 1995, p. 152) whereas the opposite is true among younger citizens (Gallagher, 1994) The relation between age and political involvement is complex as well. On the one hand, older individuals are typically not as politically attentive as are their younger counterparts (Bennett & Bennett, 1989) Despite this finding, people tend to become more likely to vote as they grow older (Coulson, 1999) When it comes to other forms of civic engagement, the relation with age is further confounded by the fact that the association is not linear. This is true for a number of indicators
60 of civic engagement. For example, informal sociability is lowest amo ng the middle aged, and more pronounced among young people and retired people. Likewise, Putnam points out, formal community involvement is relatively modest early in life, peaks in late middle age, and then declines with retirement. Informal social invo lvement follows the opposite path if over the life cycle, peaking among young adults, entering a long decline as family and community obligations press in, then rising again with retirement and widowhood (2000, p. 94 ) With regard to citizens engagement in public affairs, age is a much less significant predictor than cohort or birth year. Citing survey data over a twenty year period, Russell J. Dalton (1996) showed that the av erage age of all adults who d ever participated in a boycott, wildcat strike, sit in, or other lawful demonstration steadily increased from 35 to 46 between 1974 and 1995 (pp. 67 85) Similarly, in 2000, Putnam observed that protesting is less common among twenty somethings now than it was among people that age in the sixties and seventies, but protesting has become more common among middle aged and older people, as the sixties generation itself has aged (2000, pp. 164 165, italics in original) Education and Intellectual Ability Levels of political interest and political knowledge are positively correlated with educational attainment (Delli Carpini & Keeter, 1991; Milner, 2002) Moreover, more highly educated strata tend to vote more (Dopplet & Shearer, 1999, p. 18) In addition, one s intellectual ability may shape the nature of his or her civic engagement. Milner s discussion of the sources of civic literacy emphasized the empirical necessity of comparing cognitive proficiency (2002, p. 59) He used math and science scores as one indicator of civic literacy, and found that the countries with the highest levels of math and science literacy also displayed the highest levels of political participation and factual political knowledge (pp. 53 63)
61 Family Involvement For both women and men, getting married reduces time spent with friends, while increasing time spent in formal community associations and time spent at home. By contrast, single people show greater level s of informal sociability. Moreover, the effect of parenthood is that having children cuts further into informal social connectedness, while adding to formal community involvement Settling down ties for more formal ones, shi fting the balance between hanging out with friends and participating in community affairs (Putnam, 2000, p. 94) Operationalizing Social Capital and Civic Literacy Overlapping Features of Social Capital and Civic Lit eracy Putnam s formulation of social capital and Milner s notion of civic literacy both involved multiple, overlapping components (Figure 3 1). The aspects of civic engagement covered thus far were associated primarily with one of these perspectives or th e other. Informal sociability, associational density, and interpersonal trust were key features in Putnam s social capital model but were minimized by Milner. Similarly, newspaper reading, television nondependency, political interest, and political knowl edge were central themes of civic literacy that, at best, were peripheral to Putnam s discussion of social capital. Despite Milner s effort to put conceptual space between his ideas and those of Putnam, some items appeared in both authors models of civi c engagement, either directly or indirectly. Figure 3 1 diagrams the overlapping components of social capital and civic literacy, identifying two features that applied to both notions of civ ic engagement: (1) political participation ; and (2) engagement in public affairs. The terms political participatio n and engagement in public referred to two unique concepts for this study. Political participation limited its scope to voting in electi ons (Table A 3), while engagement in public affairs covered a broader range of activities
62 transcending one s basic electoral participation such as contacting political representatives, boycotting products, signing petitions, attending town meetings, or tak ing part in public protests (Table A 4). Let us now briefly explore each of these areas of overlap in further detail along with the rationale behind how I ulti mately decided to classify them. Political participation Political participation is central to both social capital and civic literacy. Putnam used voter turnout in national presidential elections to measure political participation. Milner not only criticized such an approach, but made it a key point of departure in his conscious break from Putnam. Citing Jane Junn (1991) Milner emphasized participation in local elections, arguing that a committed participation in local affairs represents a citizen s most vital interaction with the community and public sphere. Consistent with Putnam s measurement of political participation, the GSS only collected data on voting in national presidential elections. In light of the absence of GSS items assessing local electoral behavior, I categorized political participation as a social capital indicator for this study. Engagement in p ublic a ffairs Another overlap between Putnam s work on social capital and Milner s on civic literacy involve d citizens level s of engagement in public affairs Putnam use d two measures to assess en gagement in public affairs. As noted above, one o f these measures was voting in presidential elections 5 The other was an item asking if people had attended public meeting [s] on school or public affairs in [ the ] last year (Putnam, 2000, p. 291) In addition to the pair of activities classified as indicators of public engagement, Putnam s social capital scale contained items evaluating incidence of nonprofit organizations, community 5 specifically, w Comprehensive Social Capital Index (2000, p. 291) [m]
63 proje cts, and doing volunteer w ork. Putnam considered all three of these as markers of volunteerism. Alt hough community volunteerism earn ed its own subheading in Bowling Alone the items included under this banner can be regarded as indicators of an engaged public. Citizens undertak e volunteer work on community projects with social and political objectives in mind; often, these are the same objectives that motivate the activities Putnam considers as measures of engagement in public affairs (namely, participation in elections and to wn meetings). Thus, while Putnam demarcates between community voluntarism and engagement in public affairs (2000, p. 291) my study conceives of the former simply as an example of the latter. In other words, engagement in public affairs is an important feature of good citizenship and one of the many ways that an individual engages in public affairs is to volunteer within his or her community. Like voting, e ngagement in public affairs was integral to Milner s conception of civic literacy Milner s model may not explicitly include measures of engagement in public affairs, but his framing of civic literacy undoubtedly warrants doing so. Recall that to Milner civic literacy reflect ed both the cause and positi ve consequence of a population being willing and able to engage in public discourse (Milner, 2002, p. 1) as politically competent citizens. Given the centrality of willingness and ability to Milner s notion civic engagement, it seems an effective way to measure civic literacy would have been to locate measures that cover the exhibition of both of these complimentary aspects of citizenship at the same time. Yet, Milner s model contains measures that signify eithe r one aspect or the other either willingness or ability but not the concurrent appearance of both.
64 This is in some ways surprising, as it exposes Milner to the criticism that the variables he employed to measure willingness (or participation) may be enti rely unrelated to those used to measure ability (knowledge). That is, just because someone participates does not automatically mean she or he does so aptly and knowledgeably. Nor does being knowledgeable about issues of public concern necessarily always trigger a willingness or desire to participate politically 6 For a society to qualify as having high levels of civic literacy, its citizens must be both willing to participate and able to do so knowledgeably. In this sense, it is surprising that Milner o pt ed not to include measures signifying willingness and ability simultaneously. On the other hand, Milner clearly had good reasons for not doing so. A review of his conceptualization reminds us of the very specialized focus of Milner s research: While av ailable indicators do not always allow us to make the distinction as finely as we would wish, the intent behind our choice of the term civic literacy is clear: to compare societies on the basis of the proportion of their inhabitants demonstrating the knowl edge and skills to act as competent citizens (2002, p. 3) F rom this caveat, it seems Milner himself was cognizant that the specific measures he used to assess civic literacy did not conform ideally to the theoretic al constructs of the social scientific phenomenon he wished to study. Milner s omission of such items was likely due to the data limitations of his study s cross national orientation First of all, many variables were not available across the multitude o f nations Milner sought to study. Secondly, even when identical variables could be accessed across societies, survey items may have quite different meanings from one country to another due to differing national dynamics and cult ural contexts dictate For instance, suppose Milner 6 While caricatures of nonvoters often educe visions of apathy, ignorance, and disinterest, there is another form of civic disengagement whereby a conscientious theoretical justification is put forth by knowledgeable citizens who refuse to take part in elections. These nonparticipants are not only knowledgeabl e, but guided by a conscientious be political leaders
65 had wanted to study citizens protest or pamphleteering activities. Even if valid data existed and were collected identically in all the nations Milner studied, reliability concerns undoubtedly arise, as the act of protesting or pamphleteering inevitably is accompanied by certain culturally specific contexts, such as vastly differing levels of social risk involved with such acts, not to mention differing public perceptions of such acts which likely render them highly effective for ms of civic engagement in some societies and futile exercises in others. Given that Milner s main objective was to compare nations, rather than individual social actors inside a single nation, his research goals bound him to using very basic variables, one s which he could credibly claim traveled well across oceans and continents. The goal is to identify indicators for comparing countries level of civic engagement that overcome the limitations inherent in measures of trust and associational participation (Milner, 2002, p. 25) Milner declared that one such indicator was local electoral participation, which he tapped to operationalize political willingness His rationale was that voting in local elections represent ed a better indicator of voluntary political involvement within one s community than voting to elect a national leader. While this seems plausible it is also quite conceivable that more reliable items exist for measuring one s willingness to engage in p olitical affairs than the local election item Milner eventually settled upon. J ust like the citizen who votes for local school board candidates is more willing to participate in an informed manner than the citizen who partakes solely in nationwide electio ns, so too are there modes of participation that demonstrate a far greater willingness to be involved than merely casting a local ballot. Examples of such activities include data on whether or not respondents took part in various forms of protest aimed at triggering social change or raising awareness about matters of public concern. S uch measures were better suited to measure the concept of civic literacy as he defined it. T he problem was not that such measures were
66 inferior, nor even that they did not e xist (though that may well have been the case); the real reason Milner chose local electoral participation as his indicator of citizens willing to participate competently in public and political affairs was because his chosen measure needed to be reliable that is, measuring the same thing equivalently across societies. While measures of protest or pamphleteering behaviors might be more valid indicators of engagement in public affairs, they would likely be less reliable due to the diverse cultural contexts of the countries he analyzed. Thus, Milner s unit of analysis required that he sacrifice validity in order to salvage reliability. But my study was not fettered by the methodological dilemmas that restricted Mil ner s analytical possibilities. I examine d only a single nation, the United States. This allowed the freedom to explore and consider a wider assortment of measures than were at Milner s disposal, in order to more fully drill down into the actual definition of civic literacy. As such, I consider ed engagement in public affairs as a subcomponent of civic literacy. Summary of Civic Engagement Components and Subcomponents To recap, my study divide d civic engagement into two concepts, represented by a total of nine component measures. First, Putnam s social capital model of citizenship consist ed of the following four indicators : Informal sociability Gathering informally with neighbors or friends for casual events such as birthday parties, barbecues, tailgating, baby showers, card games, and Tupperwar e parties is crucial in maintaining the human networks that bolster social capital in the United States Such activities constitute what Putnam labeled schmoozing (2000, pp. 93 95) and are a key element t o his notion of civic engagement. Turnout in presidential elections. Putnam included voting in national elections as a component mea sure in his social capital index that informed his Bowling Alone thesis, Associational density. The number of organizational memberships is a vital component to Putnam s notion of social capital.
67 Interpersonal social trust. Generalized interpersonal trust is highly correlated with social capital. Second Milner s civic literacy framework consist ed of five variables measuring ci tizens willingness and ability to take part in public affairs. From the above discussion, we can infer that empirically assessing these abstract notions involves some combination of the following measures: Newspaper reading Weekly readership correspon ds to higher levels of civic literacy and correlates with political knowledge. Television nondependency Civic literacy is inversely related to citizens frequency of television watching. P olitical interest Civic literacy is greater when citizens are m ore interested in politics. P olitical knowledge Citizens who are knowledgeable about current events and politics are better equipped to perform competently as citizens. Engagement in public affairs Being involved in matters of public concern entails s uch activities as signing petitions, boycotting certain products to advance specific ethical or environmental causes, taking part in demonstrations, attending political rallies, contacting elected representatives or members of the media to voice concerns, or donating money or time to a political cause. All of the above behaviors are markers of civic literacy.
68 Figure 3 1 Civic e ngagement Venn d iagram: Conceptual c omponents of s ocial c apital and c ivic l iteracy A) Conceptual elements of Putnam s soc ial capital model. B) Common conceptual elements of social capital and civic literacy. C) Conceptual elements of Henry Milner s civic literacy framework.
69 CHAPTER 4 DATA AND METHODS Exploring the complexity of national pride and civic engagement required a schema that incorporated many sub measures of both concepts into a single model. Locating the right dataset was a challenge, given that notions of national pride and civic engagement are not only multidimensional but also relatively obscure. Most quanti tative datasets designed for the specific purpose of studying one of these concepts are not necessarily broad to enough to provide much more than a cursory glimpse of the other. Early in my project s planning stages it became clear that my research proble m demanded a dataset that covered a broad range of subjects as opposed to one that specialized narrowly in one of my two central concepts. This chapter introduces my dataset and describes my methodology. I discuss the differen t modules included in this data set and explain why 1996 an d 2004 were the optimal waves for analysis The chapter concludes by addressing some of the dataset s limitations, regarding both the survey in general and my project specifically. Dataset Data for this study were taken fro m the General Social Survey (GSS), which is conducted by the National Opinion Research Center (NORC) based in Chicago, Illinois (Davis, Smith, & Mardsen, 2005) With thousands of questions in a broad range of topical areas the GSS is one of the largest national public opinion survey s in the United States The majority of GSS data are collected via face to face interviews, using computer assisted personal interviewing (CAPI). As the omnibus survey for the academic social sc iences, the GSS was designed to accommodate secondary data analyses spanning a wide spectrum of potential empirical relationships According to its principal investigators James A. Davis and Tom W. Smith, the GSS s mission is to make timely, high quality scientifically relevant data available to the social
70 science research community (1992, p. 1) The GSS collects substantial data on political attitudes and social behaviors, along with the usual array of independ ent variables related to educational histories, family relations, and other demographic factors. As of 2011, NORC has fielded the survey 28 times since its inception 1972. Administered every two years, the GSS uses full probability sampling, and consiste ntly obtains a high response rate. In conducting the GSS, NORC adheres to the highest survey standards in terms of design, sampling, interviewing, processing, and documentation. Its survey methods are intended to produce high quality data for a represent ative sample. Individual GSS items are developed by leading specialists and then fully pretested. S everal data quality checks are employed, from validation to verification. Accordingly, it can be assumed that survey data are generalizable to the entire adult (non institutionalized) American population Since GSS samples are designed to be representative of the US non institutionalized population aged 18 and over the survey offers the ability to track social change in the US by monitoring the shifting tr ends in Americans social behaviors and attitudes over time. It also facilitates the comparison of GSS responses to similar surveys in other nations through the International Social Survey Program (ISSP) 1 Sample My study used data from 1996 and 2004. I analyzed two years data in order to consider nationwide trends over time on certain dimensions of national pride and civic engagement periods. These particular waves included the greatest combination of variables that were 1 In the early 1980s the GSS introduced a cross national co mponent. The cross national module has been developed as part of the International Social Survey Program (ISSP) since 1985. ISSP was founded in 1984 by research organizations from the US and three other countries: Australia, Great Britain, and West Germa ny. As of 2011 the ISSP has 48 members.
71 pertinent to my research questi ons. They were also optimally situated to explore differences in national sentiment and behavior before and after the attacks of September 11, 2001. The original 1996 and 2004 samples, taken together, were 80.2 % white 13.6% black, and 6.2% other 2 F ifty five percent of respondents in these samples were female and 45% were male, with a mean age of 45.36 years old, and an average of 13.52 years of education. R espondents in these samples earned a mean score of 6.11 on the WORDSUM test, the GSS s closes t approximation to an IQ measure (Huang & Hauser, 1998) Married respondents comprised the majority of the original combined sample (50.2 % ), followed by respondents who had never been married (22.3 % ), and, respectively, those who were divorced (15.2 % ), widowed (8.5 % ), and separated (3.7 % ). Seventy two percent of all respondents were parents; the average number of children among all people in the original, complete sample was 1.83. The above figures refer to complete original s urvey samples, which included 2,904 respondents in 1996 and 2,812 in 2004. I was unable to include all of these individuals in my study, as the GSS asked different blocks of questions to different parts of the sample, and n ot all measures are available to every respondent in a given wave. T he subsamples for which all relevant measures were non missing contained 791 respondents in 1996 and 1,10 4 in 2004. Thus, while the entire combined sample consisted of 5,176 respondents, my study was limited to the 1,8 9 5 individuals assigned to ISSP s National Identity and Citizenship Modules, and who provided data for all of the variables included in my regression models. 2 In 2004 it was possible to further differentiate race/ethnicity by cross referencing the racial categories white black and other with a dummy variable derived from a survey item that asked if respondents were of Spanish, Hispanic, or Latino/Latina ethnicity. When race was reclassified based on this measure, the racial breakdown f or other. Classifying respondents in this manner was not possible for 1996 data, due to the absence of a survey item probing Latino ethnicity.
72 ISSP Modules The GSS contains a core set of questions, which are repeated every time the survey is administered. Each wave also includes various topical modules with new sets of questions. Some of these topical questions are retired after a single administration; others are used again intermittently. Lastly, the GSS contains a cross national compone nt. As part of a cross national comparative survey research collaborative, t he annual ISSP module is administered in national surveys in nearly 50 participating nations throughout the world. 3 Many questions in ISSP modules are repeated items from previou s modules covering the same topic This was the case with the National Identity Module, which the GSS questions on patriotism and nationalism which I relied upon heavily for this project. Table 4 1 provides a complete list of measures used in this study and the years in which they were available. All exogenous and intermediate variables were present in both waves, as are the measures for national pride. The only affected variables were civic engagement measures. Of the nine civic engagement measures, e ight were found only in one wave or the other. Informal sociability, newspaper reading, and television nondependency were available solely in 1996; associational density, interpersonal trust, political interest, political knowledge, and engagement in publ ic affairs were only present in 2004; political participation was available in both waves. Due to my relatively small sample size further attrition due to missing data was a serious concern Indices and composite measures were constructed with an eye tow ard maximizing the number of cases in my analyses These s cales require d the use of multiple variables, but a number of respondents had missing values for one (or more) of the constituent measures. 3 Since the GSS is fielded every two years, all GSS administrations contain two international modules, ensuring that all ISSP modules are administered in t he US within one year of being fielded elsewhere throughout the world.
73 Therefore, in constructing the majority of my indices, I used the mean of the available measures (rather than the sum of all measures). 4 Similarly, in order to minimize missing data and thereby maximize the overall number of observations in my study, several survey items that were applicable to my topic had to be omitted. Omissions included measures for confidence in social institutions, charitable donations of time and money, and participation in local elections. Limitations of the GSS Despite its advantages the GSS has a few notable drawbacks. Because the GSS is a cross sectional survey, and has a relatively small sample size of approximately 3,000 respondents per wave, certain kinds of analyses are not possible for researchers using GSS datasets. T he GSS also suffers from the limitations that constraint a ny quantitative survey; though I briefly discuss the se below, they are unavoidable and NORC does their best to minimize them. Inability to a nalyze et hnic or g eographic s ubgroups The GSS is not an ideal instrument for the assessing trends among small sub populations, whether defined by geographic region, ethnicity, neighborhood, or sexual orientation. While many waves do contain survey items designed to collect such information, the GSS s sample size of just 3,000 people per wave means there will inevitab ly be too few respondents to make statistically significant conclusions about distinctive enclaves. Lack of longitudinal data It is not possible to examine changes across multiple time periods at the individual level using the GSS While the GSS is cond ucted every two years, questions are asked of a new set of respondents in each wave. R esearchers are therefore unable to utilize GSS data to conduct longitudinal analyses. 4 For a similar approach, see Dowley and Silver (2000)
74 Social d esirability b ias Administering surveys in person with each respondent le ads to self reported data This method of data collection is prone to criticism, as s ocial desirability may influence responses when the truth is conventionally unappealing. Nevertheless absent the ability to observe firsthand all of the behaviors and a ttitudes about which the survey asks, this mode of data collection is the most reliable method, as self reported data is the only form available.
75 Table 4 1. Variable coverage : Concepts and years available Concept GSS (1996) GSS (2004) In both waves? Exog enous variables X 1 Race (white = 1) Yes X 2 Sex (male = 1) Yes X 3 Age Yes Intermediate variables Z 1 Years of education Yes Z 2 Married (yes = 1) Yes Z 3 Number of children Yes Z 4 Intellectual ability Yes Patrio tism P 1 Institutional patriotism Yes P 2 Apolitical patriotism Yes Nationalism N 1 Protectionist nationalism Yes N 2 Nativist nationalism Yes N 3 Chauvinist nationalism Yes Social capital C 1 Informal sociability N o C 2 Political participation Yes C 3 Associational density No C 4 Interpersonal trust No Civic literacy L 1 Newspaper reading No L 2 TV nondependency No L 3 Political interest No L 4 Political knowledge No L 5 Engagement in public affairs No (N=791) (N=1 104)
76 CHAPTER 5 M EASURES National Pride Variables Drawing on a large body of work that dichotomizes good and bad forms of national attachment (Blank & Schmidt, 2003; Coenders & Scheepers, 2003; De Figueire do & Elkins, 2003; Habermas, 1992; Schatz, et al. 1999; Sidanius, et al. 1997; Spencer & Wollman, 2005; Staub, 1997) my study delineated between constructive, self oriented patriotism and destructive, other directed nationalism. F ive multi item indicat ors of national pride were used in my analysis These included two variations of domain specific national pride (Smith & Jarkko, 1998; Smith & Kim, 2006) to measure patriotism ; and three items taken from Ceoba nu and Escandell (2008) which included protectionism, nativism, and chauvinism, 1 to measure nationalism. 2 Institutional Patriotism Institutional p atriotism was a five item index measuring American citizens levels of patriotic pride in their country s specific achievements in various political and institutional arenas The index consisted of a series of survey questions probing how proud people were of the United States in five domain specific 3 categories including (1) American democracy; (2) America s political influence in the world; (3) America s economic achievements; (4) American 1 Though given different labels, these three indices for nationalism match up to indices used in other studies (Coenders & Scheepers, 2003; Hjerm, 2001) 2 Ceobanu and Escandell identif ied protectionism, nativism, and chauvinism as exclusive and other referential thereby constituting a dimension of nationalism 3 Both patriotism indices used in this study were subsets of the items employed by Smith and Jarkko (1998) to establish the domain specific patriotism index subsequently adapted by others (Dowley & Silver, 2000; Kersting, 2007; Smith & Kim, 2006) Using data from the International Social Survey Program s (ISSP) 1995 National Identity Study (NIS), Smith and Jarkko (1998) constructed an additive ten item scale ranging from ten to fifty purported to assess national pride in domain specific achievements.
77 history; and (5) America s fair and equal treatment of all groups in society. Five a nswer choices ranged from not proud at all to very proud 4 I nstitutional patriotism encompasse d proud sentiments felt or expressed toward specific institutional domains Maurizio Viroli asserted that t he language of patriotism has been used over the centuries to strengthen or invoke love of the poli tical institutions and the way of life that sustain the common liberty of the people, that is love of the republic (1995, p. 1) Institutional patriotism as measured here contained two of the three items comprising institutional legitimacy, as operationalized by Ceobanu and Escandell (2008) All three of the component items of institutional legitimacy were also included as part of Tom W. Smith s construct for domain specific patri otism, and Ceobanu and Escandell classified institutional legitimacy as self referential (2008, p. 1151) or a form of national attachment not conditioned by the existence of outsiders. As such pride in nationa l institutions is conceptually closer to patriotism than to nationalism ; I called my index institutional patriotism to reflect that fact. Others have constructed and analyzed similar measures, applying to them such labels as pride in state s institutio ns (Sorek & Ceobanu, 2009) system legitimacy (Knudsen, 1997) constructive patriotism (Blank & Schmidt, 2003; Davidov, 2009) political legitimacy (Dogan, 2009) or political national pride (Hjerm, 1998) Apolitical Patriotism Apolitical patriotism was an index comprised of four survey items assessing Americans national pride in several apol itical domains including (1) science and technology; (2) sports; (3) arts and literature; and (4) military. As with institutional patriotism, answer choices ranged from not proud at all to very proud 4 Appendix A discusses question wording and coding of answer choices in greater depth
78 The concept of apolitical patriotism involves a s trong sense of admiration or national pride toward American accomplishments that are outside the scope of traditional political or institutional realms. Protectionis t Nationalism Protectionist nationalism ( or simply p rotectionism ) was the first of the t hree nationalistic attitudes analyzed for this project. To measure different levels of protectionism, my study used a scale consisting of three agree disagree items from the GSS: Foreigners should not be allowed to buy land in America. American televis ion should give preference to American films and programs. America should limit the import of foreign products in order to protect its national identity. Respondents were included in the study if they had at least one valid measure. My scale again use d the means (rather than sums) of these responses, in order to maximize the number of valid cases. Thus, I took the mean score rather than rely on an additive model. (Appendix A and Table A 2 offer further detail on the structure of missing data ). Nativi s t Nationalism Nativist Nationalism (or just nativism) refers to the strength of one s desire to preserve the dominance of a nation s established inhabitants My project measured nativism using a four item scale. The survey items used to construct this index inquired about respondents views on the importance of various traits for qualifying as truly American including one s (1) birth location; (2) citizenship status; (3) duration of residency within the US; and (4) ability to speak English. Once agai n, cases were included in my analysis for all respondents who answered at least one question. The scale range d from 1 (i.e., all four of the above items were n ot i mportant at
79 a ll ) to 4 (all were v ery i mportant ) 5 I ndividual cases were retained as lon g as the respondent provided a valid response to at least one of the four questions in this model. Thus, I took the mean score rather than rely on an additive model, including respondents with at least one valid response. Chauvinis t Nationalism To measure chauvinist nationalism ( or simply chauvinism) I created a multi item index replicating the measure devised by Ceobanu and Escandell (2008) and similar to the pre eminence index subsequently used by Sorek and Ceobanu (2009) 6 It consisted of four agree disagree items dealing with matters of allegiance and national superiority. The items were phrased as follows: I would rather be a citizen of America than of any other country in t he world. Generally speaking, America is a better country than most other countries. People should support their country even if the country is in the wrong. The world would be a better place if people from other countries were more like the America ns. The scale ranged from 1 to 4, with 1 reserved for those who averaged the most extreme anti national sentiments across the board, and 4 designating that someone gave the extreme pro national response ( a gree s trongly ) to every question he or she answe red. Those who tended to answer less decisively ( disagree somewhat agree somewhat ) would have a score that lies somewhere between 2 to 3. The Appendix provides further detail about question wording, answer choices, and coding structure. 5 Appendix A and Table A 2 give further detail on the exact items. 6 items (2009, p. 493) but omitted the item which probed respon
80 Civic Enga gement Variables Measures of civic engagement we re informed both by Robert Putnam s social capital framework 7 and by Henry Milner s civic literacy centered approach ( Chapter 3). I use d a total of nine items to measure these elements of civic engagement. These include d fou r 8 measures of social capital ( informal sociability, political participation in national elections, associational density, and interpersonal trust ) and five civic literacy indicators ( newspaper reading, TV non dependency, political interes t, political knowledge, and engagement in political affairs ) Informal Sociability Composite Informal sociability was a dummy variable that flags whether or not respondents socialize with neighbors or friends at least several times monthly. This composit e measure was constructed from a pair of survey items asking respondents to estimate how often they: (1) spend a social evening with someone who lives in your neighborhood ; and (2) spend a social evening with friends who live outside the neighborhood. The dichotomous item was coded as 1 if respondents indicated that they engaged in either one of these activities several times per month (or more). Since less than a third of survey respondents were asked these items in 2004, the informal sociability com posite was used only in 1996. Political Participation in Presidential Elections Political participation was measured with a dummy variable indicating whether or not a respondent took part in the most recent United States presidential election The variab le was 7 Comprehensive Social Capital Index (2000, p. 291) measures for four of these (1) associational density, (2) engagement in public affairs, (3) informal sociability, and ( 4 ) interpersonal trust are available in the General Social Survey. Although a series of que stions are asked about voluntary associations, suitable variables were not available for the specific portion of the samples observed here. 8 availab
81 c oded as 1 for respondents who voted; 0 for those who did not. Although voting is considered an indicator of both civic literacy and social capital, I counted it as a social capital measure for this project, due to the GSS s reliance on data from p residential election s consistent with Putnam s approach Milner criticized the use of national election data, and focused instead on local electoral behavior (for which data were not available in the GSS in 1996 or 2004). Sum of Organizational Membersh ips The GSS contained a set of items pertaining to memberships in various kinds of groups. A ssociational density was the total number of organizational memberships each respondent held. Altogether fifteen specific membership types were listed, along wi th an all inclusive sixteenth category, Any other groups People who said they held memberships were asked if they belonged to more than one gr oup of that type, and if so, how many. Examples of possible membership types included labor unions, study gro ups, political clubs, and church affiliated group (Appendix A) Equivalent information was unavailable for 1996; thus, my analysis for that year could not include associational density as one of its measures. Interpersonal Trust Index My study measure d s ocial trust using an interpersonal trust index derived from a set of two q uestions that probe d the extent to which respondents believe d other human beings to be: fair; and trustworthy These items we re found in the National Priorities section of the Environ ment Module, originally conducted in 1993 with follow ups in 1998 and 2004. Because the Environmental
82 Module was not administered in 1996, it was not possible to compute a corresponding interpersonal trust index for that wave 9 Weekly Frequency of Newspap er Reading Frequency of newspaper reading was computed from a single item from a single GSS item, which asked : How often do you read the newspaper every day, a few times a week, once a week, less than once a week, or never? The ordinal item was colla psed into a dummy variable, coded as 1 for respondents who answered at least once a week and 0 for those who said less than once a week or never Television Nondependency The lack of dependency on television was derived from a survey question that a sked, On the average day, about how many hours do you personally watch television? Original response data from 0 to 24, where each value represented the number of hours per day that respondents spent watching television. As very small numbers of respon dents spent more than nine hours watching television each day, those responses were collapsed into a single category. Moreover, since theories suggest low levels of television watching correspond with high social capital (Put nam, 2000) this item was reverse coded. 10 Level of Political Interest Political interest was an ordinal variable taken from the 2004 Citizenship Module. While the variable was measured in 1996 it could not be used for that year because the portion of the sample asked the question did not overlap with those asked the national pride questions. 9 One subsection in 1996 contained a series of questions that asked if respondents thought others were helpful, fair, onal Identity Module (i.e., the portion of the 1996 survey that probed national pride). 10 Table A 5 in Appendix A provides specifics on how this item was coded.
83 Respondents were asked How interested would you say you personally are in politics? In 2004, four Likert scale response categories included Very interested, Fairly interested, Not very interested, and Not at all interested. (Figure 5 1 displays all civic engagement measures used in this study and the years in which they were available). Political Knowledge Index Political knowledge was a scale derived from a pair of self reported measures that gauge d respondents knowledge about politics Respondents we re asked whether or not (and how strongly) they agree d with the following two statements: I feel that I have a pretty good understanding of the import ant political issues facing our country I think most people are better informed about politics and government than I am. Scoring of the second variable was reversed before combining the pair to create a p olitical k nowledge i ndex (Table A 5 ). Engageme nt in Political Affairs For 2004, a set of items was worded as follows: Here are some different forms of political and social action that people can take. Please indicate, for each one, [w] hether you have done any of these things in the past year, [w] hethe r you have done it in the more distant past, [w] hether you have not done it but might do it, [o] r have not done it and would never, under any circumstances, do it. There we re then eight separate items : Signed a petition Boycotted, or deliberately bought, c ertain products for political, ethical or environmental reasons Took part in a demonstration Attended a political meeting or rally Contacted, or attempted to contact, a politician or a civil servant to express your views
84 Donated money or raised funds for a social or political activity Contacted or appeared in the media to express your views Joined an Internet political forum or discussion group For each of the eight individual forms of civic action, respondents were coded as 1 if indicated they had taken part in such an action in the last year. All other respondents with nonmissing data they were coded as 0. The eight dummy variables were then counted up, producing for each respondent a measure of engagement in public affairs that ranged from 0 to 8 and denoting how many separate types of citizen action respondents had taken part in during the past year. This variable was not collected in 1996. ( Figure 5 1 displays all civic engagement measures and the years in which they we re available). Control Variabl e s Race Race was a dummy variable coded as 1 if the respondent was white; 0 otherwise I experimented with several different coding schemas in an effort to account for black, Asian and Latino ethnicities. Ultimately, however, there were too few cases of nonwhite ethnicities in the dataset to differentiate among multiple racial and ethnic minorities. As of 1996, the GSS still measured race ba sed on interviewers observations. Survey administrators were instructed to assess whether respondents appeared t o be white, black, or other; respondents were then officially categorized based on these subjective evaluations. This method was overhauled in 2002, when the GSS began measuring race according to the procedures used in the decennial Census which relied o n respondents racial and ethnic self identifications.
85 Gender Gender was coded as 1 for male respondents and as 0 for female respondents using the GSS variable SEX To garner this measure, interviewers we re instructed to record their personal observat ions of each respondent s sex (or gender ) rather than to explicitly ask people if they were male or female. Technically, then, th is survey item represents neither the biological sex nor gendered identity of a respondent but a researcher s best guess as to both. With this caveat in mind, I use the term gender throughout this study. Age Age was an interva l variable measuring the respondent s age in years at the time of the survey The original GSS item wa s a measure of cohort ; i n an effort to maximize the validity of responses interviewers inquired about one s birth year, as opposed to one s actual age. The GSS then computed each respondent s age by subtracting the individual s birth year of birth from the survey year. All respondents over the age of 90 were r ecoded as 89 Education Education was an interval variable measuring the respondent s total years of formal schooling completed ranging from 0 to 20 with a code of 20 denoting twenty or more years of education The GSS also collected educa tion as an ordinal variable referring to respondents scholastic milestones and degrees obtained. My intent was to explore the cumulative influence of additional years of education, rather than the impact of a credential. Thus, for my statistical correla tions and multiple regression analyses, I employed the more detailed numerical measure of actual years of formal schooling. My use of the ordinal measure was limited to a few charts, presented in Chapter 9 to clarify the nature of certain bivariate associ ations (Tables 9 4 through 9 7; Figures 9 3, 9 4, 9 8, 9 9, 9 12, and 9 14).
86 Family Status Marital status was collected as a nominal variable to account for five possible categories: married, divorced, widowed, separated, and never married. In addition, a quantitative variable counted the number of children each respondent has. For the regression analyses, the marital status variable was collapsed into a dichotomous variable, coded as 1 if the respondent was married and as 0 if he or she was not. Intel lectual Ability To account for Milner s emphasis on relevance of cognitive abilit y (2002, pp. 53 77) I included i ntellectual ability in my model in order to explore the impact of cognitive skills attitudinal and behavioral outcomes. The GSS contains a variable called WORDSUM, which purports to assess respondents cognitive abilities. WORDSSUM is an annual intelligence assessment given to half of all GSS respondents. Respondents were chosen randomly to take the ten word vocabulary subtest derived from the Wechsler Adult Intelligence Scale (WAIS), a popular IQ test (Wechsler, 1958; Zhu & Weiss, 2005) For each word, the survey presented five possible definitions and instr ucted respondents to select the correct one. Individual correct responses were then coded as 1 and then summed across all ten words. Thus, WORDSUM ranges from 1 to 10, providing each person s total number of correct responses. 11 Empirical literature exami ning the validity of this GSS measure of intelligence has found overall that the verbal test score variable WORDSUM is a reasonably good proxy for IQ (Alwin, 11 The raw GSS datasets included outcomes for each of the ten individual words, each of which d istinguished between non attempts and wrong answ ers. For the final WORDSUM tally, respondents who did not attempt a single answer were coded as missing, whereas those who attempted some but not all answers were coded based on their number of correct answers. (In short, skipped responses were treated i dentically to wrong answers so long as the respondent tried to answer at least one question).
87 1991; Huang & Hauser, 1998; Wolfe, 1980) Lee Wolf e (1980, p. 110) found a correlation of 0.71 between the WORDSUM vocabulary assessment and the Army General Classification Test (AGCT). As Bryan Caplan and Stephen C. Miller acknowledge, WORDSUM cannot be taken as literal gauge of intelligence After all, [i]f intelligence is thought of as the ability to think or acquire knowledge, the WORDSUM subtest is not a direct test of intelligence, but rather a test of knowledge (20 10, p. 639) Nevertheless, citing British psychologist Raymond Cattell s (1987) theories of fluid and crystalized intelligence, Caplan and Miller argue d that while not a direct measure of intelligence, the WORDSUM sub test is essentially a measure of crystallized intelligence, i.e. the knowledge the subtest s results show re fl ect not innate intelligence but the level of knowledge that respondents needed their intelligence to acquire The development of vocabulary depen ds crucially on fl uid intelligence and therefore can be seen as a proxy for general intelligence (2010, p. 640) More broadly, vocabulary knowledge has tended to correlate positively with tests of general intellig ence (Alwin, 1991; Huang & Hauser, 1998; Miner, 1957; 1961; Wechsler, 1958) Miner (1961) found a correlation of 0.75 between general intelligence and results on a 20 word vocabul ary. David Wechsler, creator of the WAIS, offered a theoretical explanation for this empirical link between vocabulary and IQ: the size of a man s vocabulary is not only an index of his schooling, but also an excellent measure of his general intelligence. Its excellence as a test of intelligence may stem from the fact that the number of words a man knows is at once a measure of his learning ability, his fund of verbal information and the general range of his ideas. (1958: 84).
88 Figure 5 1. V enn d iagra m: Variable coverage of c ivic e ngagement m easures A) Social capital indicators available only in 1996. B) Social capital indicators available in both 1996 and 2004. C) Social capital indicators available only in 2004. D) Civic literacy indicators avai lable only in 1996. E) Civic literacy indicators available only in 2004.
89 CHAPTER 6 CONCEPTUAL MODEL Primary Concepts: Dependent Variables When assessing national pride, one question to consider should be whether or not individuals who show high levels of patriotism and nationalism also tend to be good citizens. A number of observers (Ehrlich, 2000; Milner, 2002; Ricci, 2004; Selbourne, 1994) argue in various ways that civic engagement is an asp ect of good citizenship. This issue may turn, however, on the association between national pride and civic engagement. Perhaps engaged individuals hold patriotic sentiments, such as being proud of one s country and its institutions. Conversely, civic i nvolvement may be more of a defensive posture associated with insiders nationalistic feelings of scorn and suspicion toward outsiders whom they view as a threat to their homeland. Conceptually, national pride constitutes a set of beliefs and attitudes, while civic engagement mainly deals with a set of behaviors. As Chapters 2 and 3 illustrate d, national pride and civic engagement are multidimensional, each manifesting itself in myriad ways. Dividing these primary concepts into subcomponents allowed me to test whether different forms of national feelings and civic minded conduct were r elated to each other in divergent ways. Suppose for instance that patriotism correlated positively with both social capital and civic literacy. One could not infer from t hose observations alone that national pride enhanced civic engagement Because both national pride and civic engagement are measured in a cross sectional study, it is not possible to infer causation. Equally as important, patriotism is but one variety of national pride; the same civic measures that exhibited a positive correlation to patriotism might correlate negatively with other, more nationalistic forms of national pride.
90 Hence the examination of the association between component parts of national pr ide and civic engagement offers greater insight than examining the association of the overall measures. My study considered two subcomponents each of national pride and civic engagement. National pride can be expressed as either patriotism or nationalism; c ivic engagement invo lves social capital and civic literacy. Figure 6 1 shows these subclasses of national pride and civic engagement, and the possible associations among them. Because my analysis relied on cross sectional data, it was not possible to address the issue of causality. Though not pictured in the Figure, my analysis allowed for but did not analyze the correlation between patriotism and nationalism, or b etween social capital and civic literacy. F igu re 6 2 illustrates the specific sub measures comprising each of my primary concepts Observed correlations between measures of national pride and those of civic engagement represent the association between specific aspects of nation oriented feelings and community oriented behavior. Each curved line connects one subcategory of national pride to one subcategory of civi c engagement. Exogenous and Intermediate Covariates A number of covariates might account for national pride and civic engagement or modify the nature of the relationship between these two concepts. Sex, race, and age we re exogenous variables, in that they are determined externally. Education, marital status, parenthood, and intellectual ability were considered intermediate variables because they often depend on values of the exogenous variables. Incorporating these covariates into my model allowed for th e conceptual possibility that the exogenous variables would have both direct and indirect effects on national pride and civic engagement, with the indirect effects occurring via the intermediate variables. Figure 6 3 presents the full conceptual model wit h all variables included and with
91 covariates divided into exogenous and intermediate categories. The ensuing discussion of covariates clarifies my rationale for including each as a control variable in the model. Racial identifiers were included as pa rt of the general conceptual model to test the validity of recent empirical findings on the effects of r ace/ethnicity on national pride. Evidence suggests national pride is strongest among whites, especially after 9/11 ( Carroll, 2007; Ishio, 2010) Additionally, a ccounting for race in my regression equations allowed me to explore the extent to which American national identity is perceived or defined along ethnic/racial lines, as theorized by ethno symbolism (A.D. Smith 1986) I also included gender in my overall conceptual framework for a number of reasons Men have been consistently underrepresented in the GSS ( T.W. Smith 1979) but this is consistent with most public opinion surveys (Brehm, 1993, pp. 23 Including gender in my model was required in order to assess: the empirical finding that males in the United States show higher levels of national pride than females in the United States (Smith & Jarkko, 1998) ; theoretical notions of patriotism and nationalism as uniquely masculine qualities (Connell, 1995; Enloe, 1990; Nagel, 1998) ; the potential gender gap in civic engagement (Bennett & Bennett, 1989) along with the differential ways in which men and women choose to participate in their communities (Putnam, 2000; Verba, et al., 1995) ; and potential interactions between gender and other independent variables. Age is likely to correlate positively with both manifestations of national pride (Carroll, 2007) That is, older individuals tend to exhibit stronger expressions of national pride. Some sugges t this is more the result of cohort differences than age differences. Increasing levels of patriotism and nationalism may reflect cohort rather than age differences because younger cohorts came of age during a cosmopolitan era of globalization and increas ed multiculturalism (Camilleri & Falk, 1992; Craige, 1996) Older persons also are often less politically attentive
92 than younger ones (Bennett & Bennett, 1989) In the absence of s trong assumptions about the size and direction of one of the effects, it is not possible to distinguish among age, period, and cohort effects in this research E ducation level was expected to influence the nature and strength of the relationship between na tional pride and civic engagement. Some social scientists contend that the education system has long served the functional purpose of fostering loyalty to the nation state (Mitter, 1993; Tamir, 1993) Others have poi nted out that the outspoken aim of the educational system is to foster a respect for human rights, democratic values, and fo rward thinking multiculturalism (Schleicher, 1993, pp. 13 38) Indeed, M ikael Hjerm s emp irical analysis of ten diverse countries revealed a decrease in levels of nationalist sentiment as years of education increased (2001) Education also tends to be associated with higher levels of political knowledge an d political interest (Delli Carpini & Keeter, 1991; Milner, 2002) Lastly, intellectual ability was included in the conceptual model to see whether such abilities shaped respondents levels of national pride, or th e nature of their civic engagement. Conceptually, the intent was to consider whether cognitive ability contributes to attitudinal and behavioral outcomes in a manner that is substantively distinct from the effect of education. This conceptual approach ad heres to Milner s ideas about the need to compar [e] cognitive proficiency (2002, p. 59) Milner considered math and science scores as possible sources of civic literacy (pp. 53 77)
93 Figure 6 1. Manifestations of national pride and civic engagement. A) National pride takes the form of patriotism or nationalism. B) Civic engagement involves social capital and civic literacy. C) The curved lines linking t he subclasses denote that these links were conceived as correlations, not as causal relationships. Figure 6 2 Manifestations of national pride and civic engagement. A) Measures of patriotism B) Measures of nationalism. C ) Measures of social ca pital D ) Measures of civic literacy. E ) The curved lines linking the subclasses denote that these links were conceived as correlations, not as causal relationships. A A C B B C D E
94 Figure 6 3. Full conceptual model A) Exogenous and intermediate covariates. B) Independent variables. B A
95 CHAPTER 7 RESEARCH QUESTIONS A ND HYPOTHESES Research Questions The goals of this research project were both to illuminate the relationship between national pride and civic engagement and to understand each of thes e broad concepts ind ividually. T o that end I focused on four research questions Question 1: Who expresses high levels of national pride? Question 2: Who has high levels of civic engagement? Question 3: How are national pride and civic engagement related? Question 4: How does the nature of these relationships change and fluctuate over time? Because both national pride and civic engagement are multifaceted, each involving separate subtypes, my research extended beyond the formal research questions specified at the ons et of this project. E ach of my first two research questions essentially constituted a pair of subquestions. To understand national pride, for instance, required a comprehension of the basic factors influencing both patriotism and nationalism. I n framin g Question 1 therefore, I sought to uncover: Wh at kinds of US citizens express high levels of patriotism (measured as national pride in specific institutional and apolitical achievements)? a nd, What kinds of US citizens express high levels of nationalis m ( measured as protectionism, nativism, and chauvinism )? Question 2 also involved of multiple inquiries; to answer it required drilling down into civic engagement s component parts. Specifically, What sorts of US citizens have high levels of social capi tal (measured using a set of four subcomponents: informal sociability, political participation, associational density, and interpersonal trust)? and,
96 What kinds of US citizens express high levels of civic literacy (measured using a series of five subcomp onents: newspaper reading, television nondependency, political interest, political knowledge, and engagement in public affairs)? Question 3 attempted to expand upon the first two questions by pursuing an understanding of the nature of the relationship betw een national pride and civic engagement. Accordingly, it required scrutinizing the connection of both forms of national pride to both forms of civic engagement (Figures 6 1 and 6 2). M y study s use of data from two separate decades allowed me to describe differences across two separate time periods In comparing 1996 data to 2004 data, Question 4 assess ed how the nature of the relation between national pride and civic engagement changed over time. For each research question, I developed a h ypothesis or se t of hypotheses, based on the theories and empirical evidence reviewed in Chapters 2 and 3 Some questions were linked to more than one hypothesis ; similarly, hypotheses several hypotheses correspond ed to more than one of the original research questions. Hypotheses Informed by theoretical considerations and past empirical findings I developed a series of hypotheses related to each research question. The following hypotheses address the predicted relationships between the concepts in my model s S pecific equation s testing each hypothesis are presented in C hapter 8. Hypothesis 1: Members of socially dominant groups exhibit higher levels of national pride than members of socially subordinate groups. Hypothesis 2: Higher levels of education lead to lower lev els of national pride. Hypothesis 3: Patriotism and nationalism are positively correlated to each other, but their relation to covariates and civic engagement outcomes differs. Hypothesis 4: Social capital increases with increasing levels of education.
97 Hyp othesis 5: Civic literacy increases with increasing levels of education and intellectual ability. Hypothesis 6: The different components of national pride differ in their relationship to civic engagement. Hypothesis 7: National pride levels were higher in 2004 than in 1996. National Pride Hypotheses Hypothesis 1 : Social status and national pride I hypothesized that members of socially dominant groups exhibit higher levels of national pride than members of socially subordinate groups According to social dominance theory, members of higher status groups experience a greater sense of ownership and attachment toward their nation and its symbols than do members of subordinate groups. All other things being equal, the hegemonic or dominant group in any socia l hierarchy will show greater support for national ideologies. Dominant social groups in the United States include whites, males, and older people, although the relation between age and social status is not directly linear (Chapter 3). A dhering to a so c ial dominance theory approach (Ishio, 2010; Sidanius & Pratto, 1999) and empirical findings connecting social dominance orientation to attitudes about race (Pratto, et al. 2006; Sidanius, et al. 1997; Snellman & E kehammar, 2005) Hypothesis 1 involve d an assumption that national pride levels are higher among whites than among nonwhites. Also implied by Hypothesis 1 is the assertion that men display greater levels of national pride than women in accordance with em pirical findings (Smith & Jarkko, 1998) along with theoretical work which regards nationalism as a distinctly masculine phenomenon (Enloe, 1990) This hypothesis was consistent with the assumpt ions of social dominance theorists (Ishio, 2010; Pea & Sidanius, 2002; Sidanius & Pratto, 1999) and evidence that social dominance orientation scores correlate with sexist and anti feminist views (Bates & Heaven, 2001)
98 Hypothesis 2: Education and national pride. Second, I posited that levels of national pride decrease as education levels increase. This hypothesis was influenced by the notion that y ears of formal schooling tend to be negatively associated with national pride (Schuman, et al. 1997; Vogt, 1997) Hypothesis 3 : Patriotism and nationalism. R esearch questions that inquired about what kinds of people are patriot ic and nationalistic compel one to consider how these tw o forms of national pride relate to one another. My third hypothesis posited that patriotism and nationalism are positively correlated to each other, but th at covariates have differential effects on patriotism and nationalism Since national pride implie s both patriotic and nationalistic sentiments, indicators of patriotism and nationalism were assumed to correlate positively with one another (Davidov, 2011; Pratto, et al. 1993) However their relationships to covariates were expected to diverge (Raijman, Davidov, Schmidt, & Hochman, 2008) as was their associations to civic engagement outcomes Civic Engagement Hypotheses Hypothesis 4: Education and social capital. Fourth, I hypothesized that s ocial capital increases with increasing levels of education Me asures of social capital are positively associated with education (Putnam, 2000; Verba, et al., 1995) Hypothesis 5: Education, intellectual ability, and civic literacy. Taking into account scholarship showing that e duc ation and intellectual ability are positively correlated to measures of civic literacy (Milner, 2002) I hypothesized that civic literacy increases with increasing levels of education and intellectual ability. Hypothesis 6: National pride and civic engagement. I hypothesized that m easures of civic engagement relate differently to measures of national pride. In particular, I expected to find the strongest negative associations existed between nationalism (i.e., protectionism
99 nationalism, and chauvinism) and civic literacy. Persons with high levels of patriotism (i.e., institutional and apolitical patriotism) were expected to exhibit greater levels of both types of civic engagement than those with high levels of nationalism (i.e., protectionism, nationalism, and chauvinism). Hypothesis 7: National pride and civic engagement over time. I hypothesized that national pride is stronger in 2004 than in 1996 I expected that this was due, at least in part, to the attacks of Sept ember 11, 2001. O nly one of the nine civic engagement measures was collected in both waves of the survey. Thus, I was not able to hypothesize about changes in civic engagement levels over this time period.
100 CHAPTER 8 ANALYSIS Methodological Constraints T he issue of the causal order of national pride and civic engagement cannot be resolved in the present study. In addition, several key variables of interest were discrete variables or limited dependent variables; consequently, I could not assume a normal error, a required precondition for ordinary least squares regression My regression model is a descriptive model that does not and cannot, assume any causal link between national feelings and civic engagement. Although I position civic engagement as an outcome variable for the sake of the regression analysis, this was done as a descriptive regression, not as a structural equation model. It is just as plausible that civic involvement influences national pride as the other way around. Several of my measu res for national pride and civic engagement were either indices derived from Likert scales, or summed counts of several individual discrete variables. This means that they were limited dependent variables, for which I could not assume normal errors. As a result, I could not use OLS regression for these variables, as that method assumes a normal distribution of the errors. Instead, I used a negative binomial regression model, a method that accommodates limited dependent variables. For political participat ion a dichotomous variable measuring whether or not respondents voted in the most recent presidential election I estimated a logistic regression (Menard, 2001) Regression Models In this section, I link each individual regress ion model to its corresponding research question, tying together the material presented in C hapters 6 and 7 In addition to this
101 discussion, the models to be estimated are summarized in Table 8 1. Each regression equation addressed a unique research ques tion. My main objective was to shed light on the association between national pride and civic engagement. To what extent is this relationship accounted for by the exogenous and intermediate variables? A second and closely related goal was to identify t he major predictors of national pride and civic engagement. The first four research questions posed above seek to pinpoint variables that trigger unusually high or low levels of national pride and civic engagement. Questions A and B deal with two forms of national pride patriotism and nationalism, respectively. The next two questions (C and D) allude to notions of civic engagement. With the final question (E), I return to the central aim of my project, which was to gain a better understanding of the re lation between national pride and civic engagement. It also addressed the issue of covariates. To address the initial four research questions and preliminary hypotheses, I devised two sets of equations each for national pride and civic engagement. Models 1 and 2 investigated national pride. Model 1 estimated the total effects of exogenous variables on measures of national pride. Model 2 estimated the direct effects of exogenous covariates and total effects of intermediate covariates on national pride. 1 These regression equations can be written mathematically as: (8 1) where N s (s = 1 to 3: protectionism, nativism, chauvinism) refers to measures of nationalism, X m (m=1 to 3: race sex, age) is an exogenous covariate, is the intercept, and is the random error effect. 1 For national pride measures, my regression outputs include a third model that adds survey year to the model to compare differences in overall trends between 199 6 and 2004.
102 (8 2) where Z q (q=1 to 4: education, marital status, number of children, intellectual ability) is an intermediate covariate. An identical set of equations was estimated for measures of patriotism, P j (j= 1 to 2: institutional patriotism, apolitical patriotism). For civic engage ment, there were five sets of models altogether. Models 3 and 4 estimated total and direct effects, parallel to Models 1 and 2 above; the lone difference was that social capital and civic engagement were substituted for patriotism and nationalism, as: (8 3) and (8 4) where C k (k = 1 to 4: informal sociability political participation, associational density, interpersonal trust) refers to measures of social capital. In addition, three statistical models addressed the primary hypothes es. Model 5 estimated the association of patriotism and civic engagement, controlling for exogenous and intermediate covariates. (8 5) Model 6 estimated the association of nationalism and civic engagement, controlling for exogenous and intermediate covaria tes. (8 6) Model 7 estimated the association of patriotism and nationalism with civic engagement, controlling for exogenous an d intermediate covariates. (8 7)
103 where C k (k = 1 to 4: informal sociability political participation, associational density, interpe rsonal trust) refers to measures of social capital. An identical set of equations were estimated for measures of civic literacy, L r (r= 1 to 5: newspaper reading, TV nondependency, political interest, political knowledge, engagement in political affairs).
104 Table 8 1. Outcome variables and types of regression: GSS, 1996 and 2004 Name of outcome variable Outcome variable type Regression type Patriotism P1 Institutional patriotism Scale/index OLS P2 Apolitical patriotism Scale/index OLS Nationalism N 1 Protectionism Scale/index OLS N2 Nativism Scale/index OLS N3 Chauvinism Scale/index OLS Social Capital C1 Informal sociability Dummy Logistic C2 Political participation Dummy Logistic C3 Associational density Count Negative binomial C4 Interpers onal trust Scale/index OLS Civic Literacy L1 Newspaper reading Dummy Logistic L2 TV dependency Interval OLS regression L3 Political interest Ordinal OLS L4 Political knowledge Scale/index OLS L5 Engagement in public affairs Ratio Negative binomial
105 Table 8 2. Models for civic engagement (1996): Informal sociability, political participation, newspaper reading, and TV nondependency Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Exogenous variables Race (white=1) X 1 X 1 X 1 X 1 X 1 X 1 X 1 Sex (male=1) X 2 X 2 X 2 X 2 X 2 X 2 X 2 Age X 3 X 3 X 3 X 3 X 3 X 3 X 3 Intermediate variables Education Z 1 Z 1 Z 1 Z 1 Z 1 Married (yes=1) Z 2 Z 2 Z 2 Z 2 Z 2 Number of children Z 3 Z 3 Z 3 Z 3 Z 3 Intellectual ability Z 4 Z 4 Z 4 Z 4 Z 4 National pride Patriotism Institutional patriotism P 1 P 1 P 1 P 1 Apolitical patriotism P 2 P 2 P 2 P 2 Nationalism Protectionism N 1 N 1 N 1 N 1 Nativism N 2 N 2 N 2 N 2 Chauvinism N 3 N 3 N 3 N 3 Civic engagement Social capital Informal sociability C 1 C 1 C 1 C 1 C 1 Political participation C 2 C 2 C 2 C 2 C 2 Associational density Interpersonal trust Civic literacy Newspap er reading L 1 L 1 L 1 L 1 L 1 TV nondependency L 2 L 2 L 2 L 2 L 2 Political interest Political knowledge Engagement in public affairs
106 Table 8 3 Models for civic engagement (2004): Political partici pation, associational density, interpersonal trust, political interest, political knowledge, and engagement in public affairs Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Exogenous variables Race (white=1) X 1 X 1 X 1 X 1 X 1 X 1 X 1 Sex (male=1) X 2 X 2 X 2 X 2 X 2 X 2 X 2 Age X 3 X 3 X 3 X 3 X 3 X 3 X 3 Intermediate variables Education Z 1 Z 1 Z 1 Z 1 Z 1 Married (yes=1) Z 2 Z 2 Z 2 Z 2 Z 2 Number of children Z 3 Z 3 Z 3 Z 3 Z 3 Intellectual ability Z 4 Z 4 Z 4 Z 4 Z 4 National pride Patriotism Institutional patriotism P 1 P 1 P 1 P 1 Apolitical patriotism P 2 P 2 P 2 P 2 Nationalism Protectionism N 1 N 1 N 1 N 1 Nativism N 2 N 2 N 2 N 2 Chauvinism N 3 N 3 N 3 N 3 C ivic engagement Social capital Informal sociability Political participation C 2 C 2 C 2 C 2 C 2 Associational density C 3 C 3 C 3 C 3 C 3 Interpersonal trust C 4 C 4 C 4 C 4 C 4 Civic literacy Newspaper re ading TV nondependency Political interest L 3 L 3 L 3 L 3 L 3 Political knowledge L 4 L 4 L 4 L 4 L 4 Engagement in public affairs L 5 L 5 L 5 L 5 L 5
107 CHAPTER 9 UNIVARIATE AND BIVARIATE R ESULTS Analysis results a re reported in C hapters 9 and 10. The focus in C hapter 9 is limited to univariate and bivariate results while C hapter 10 reports multiple regression results. Univariate Statistics In the 1996 sample, respondents had a mean age of 45. 2 years old. Eighty three percent were coded 1 as white, 1 3 % as black, and 5% as other. 2 Fifty nine percent were female and 41 % were male. Respondents in the 1996 sample had an averag e of 13. 6 years of education. Twenty six percent of respondents in the 1996 subsample had a Bachelor s degree and 88% had a high school diploma Fifty one percent were married, 22 % were divorced or separated, 19 % had never married, 3 and 9% were widowed. On average, 1996 respondents had 1. 9 children and an intellectual ability score of 6.1 on the WORDSUM vocabulary assessment (Huang & Hauser, 1998) which is derived from a popular IQ test (Chapter 5). Respondents in the 2004 sample had a mean age of 46.9 years old. Eighty one percent were white and 19% were nonwhit e. Fifty six percent were female and 44% were male. The 200 4 sample had an average of 13.9 years of education. Eighty nine percent had high school degrees and 31% had college diplomas. More than half of respondents were married. Specifically, 55% were married, 19% were divorced or separated, 19% had never married, and 7% were widowed. On average, 2004 respondents had 1. 9 children and an intellectual ability score of 6.4. 1 Prior to 2002, race was coded based on the observations of GSS interviewers. 2 Due to rounding, percentages will not always round to 100%. 3 Th e rather high percentage that many in the sample we re youn g although less than one half of respondents in this category were under the age of 30.
108 Overall, respondents in the 1996 GSS subsample used for this study were younger, less educated, and scored lower on the intellectual ability measure than their 2004 counterparts. Table 9 1 displays univariate statistics for all measures analyzed in the present study s multivariate regression models. National Pride Both component me asures of patriotism were higher among the 2004 sample than among the 1996 sample. On the institutional patriotism scale, respondents averaged just 3.81 in 1996, compared to an average of 3.99 for respondents in 2004. On the apolitical patriotism scale, th e average score for 1996 survey respondents was 4.18, whereas survey respondents in the 2004 sample had a mean institutional patriotism score of 4.36. Mean scores for two of the three forms of nationalism were significantly higher in 2004 than in 1996. On the nativist nationalism scale, survey respondents averaged just 4.15 in 1996, compared to 4.45 in 2004. On the chauvinist nationalism scale, survey respondents averaged 3.68 in 1996, compared to 3.76 in 2004. Scores on the protectionist nationalism scal e were slightly higher in 1996 than in 2004; respondents averaged 3.24 in 1996 and 3.18 in 2004. Civic Engagement Political participation was the only civic engagement measure collected in both survey years examined in this study. O perationalized as voti ng in presidential elections the survey asked whether or not respondents had v oted in the most recent presidential election ; the 1996 measure therefore refers to the 1992 elections between Bill Clinton and George H. W. Bush, and the 2004 measure refers to the 2000 elections between George W. Bush and Al Gore As a
109 consequence, the political participation item differed notably from other civic engagement measures collected in 2004, in that it could not have been influenced by the attacks of September 11, 2 001. Figure 9 1 compares survey findings with national voting statistics in order to consider the results within the context of broader nationwide trends. Consistent with national trends, a slightly higher proportion of 1996 respondents reported having v oted in 1992 than the proportion of 2004 respondents who said they d voted in 2000. Overall, 73% of 1996 respondents reported having voted in 1992, compared to 69% of 2004 respondents reported having voted in 2000. In both waves, significantly greater prop ortions of respondents reported having voted than the corresponding national voter turnouts for the years about which they were asked. The other e ight indicators for civic engagement were available only in either one wave or the other. Three were measured only in 1996: informal sociability, newspaper reading, and television nondependency. Fifty eight percent of people in the 1996 sample had recently socialized informally with either friends or neighbors. Eighty three percent of survey respondents were at least weekly newspaper readers On the television nondependency item, which ranged from 1 to 10, 4 respondents had an average tally of 7.13 Five measures of civic engagement were measured exclusively in 2004: associational density, interpersonal trust, political interest, political knowledge, and engagement in public affairs. The 2004 respondents belonged to an average of 2.14 organizations, and had a mean score of 3.15 on the interpersonal trust index. Nearly half of all respondents in 2004 said they were fairly interested in politics (49.55 % ), and an additional 21.74 % claimed they were very interested ; the rest admitted they were either not very interested (19.57 % ), or not at all 4 Appendix A explains the operationalization of television nondependency.
110 interested (9.15 % ). The 2004 sample had a mean score of 3.60 on the political knowledge index, and on average had taken part in 1.49 of the eight activities counted as different modes of engagement in public affairs. Bivariate Relationships The above statistics refer to the measures as they were described in C hapter 5 and calculated for the multivariate analyses reported in C hapter 10. There was a dditional information available for some variables; although this added level of detail was ultimately lost when ordinal and nominal measures were collapsed into dummy variab les, it is nonetheless worth reporting here. Nonwhite respondents were collapsed from the nonwhite categories black and other In 2004 only, the racial categories white black and other can be further stratified by cross referencing with a dummy variable that indicates whether or not respondents were of Latino ethnicity. Unmarried respondents were collapsed from a total of four unmarried categories: widowed divorced separated and never married. Bivariate crosstabs will be displayed bas ed on these categories. Also, in order to make better sense of the data, for the bivariate section only, the continuous variables age, education, number of children, and intellectual ability will be reported at the ordinal level. Correlation among Indepen dent Variables Table 9 2 displays Pearson s Correlation coefficients for variables in the analysis in 1996. Table 9 3 displays the same information for variables present in 2004. Exogenous correlation Race and age. Whites were on average older than nonw hites in both 1996 and 2004. In 1996, whites were an average of 46.19 years old and nonwhites were an average of 40.59 years
111 old. In 2004, whites were an average of 48.03 years old and nonwhites were an average of 41.90 years old. Intermediate correlat ions Education and exogenous controls. Race was a significant predictor of education in both waves. (Tables 9 2 and 9 3 present P values and coefficients). Whites in the 1996 sample had a mean of 13.69 years of education, compared in 13.04 years for non whites. In 2004, white respondents had a mean of 14.08 years of education, compared to 13.24 years for nonwhites. For both whites and nonwhites, mean years of education was higher in 2004 than in 1996. Older respondents tended to have fewer years of ed ucation, on average, than younger respondents (Figure 9 2). A negative correlation between age and education was observed both in 1996 ( r = .223; P value <.0001) and in 2004 ( r = .109; P value = .0003). Marital status and exogenous controls. Race was a significant predictor of marital status, both in 1996 and in 2004. Specifically, in the 1996 subsample, 53 % of whites were married compared to 38 % of nonwhites; likewise, 58 % of white respondents were married in 2004, compared to 42 % of nonwhites. Paren thood and exogenous controls. Not surprisingly, in both 1996 and 2004, age significantly predicted not only parental status but also the how many children respondents had. Parents, on average, were 47.75 years of age in 1996 and 49.50 years of age in 200 4. By contrast, childless respondents were 37.65 and 38.91 years old respectively. In general, the older respondents were, the more children they were likely to have even when all nonparents were removed from the analysis. Intellectual ability and exogen ous controls. On the measure for intellectual ability, whites scored higher than nonwhites in both waves. The average score for whites was 6.38 in 1996 and 6.65 in 2004, whereas nonwhites averaged 4.94 and 5.37 respectively. Some, but not
112 all, of this d isparity can be explained by educational disparities in years of education However, even when education is held constant, whites, on average, still scored a point higher than nonwhites on nearly every level of education (Figure 9 3). Correlation among in termediate controls. Education was negatively correlated with number of children, in both 1996 ( r = .255; P value <.0001) and 2004 ( r = .184; P value <.0001) On average, childless respondents had 14.59 years of education, compared to an average of 13. 51 years among all who were parents. Nearly one half of respondents with less than a high school education had three or more children, compared to just 20 % of respondents with at least a college degree. By contrast, greater than one third of all college educated respondents had no children, whereas only 15 % of those without a high school degree were childless (Figure 9 4). Correlations with National Pride This section presents the significant bivariate correlations involving the study s national pride i ndices. Tables 9 4 and 9 5 display crosstabulations of mean values for measures of patriotism and nationalism by covariates. Institutional patriotism. Tables 9 2 and 9 3 show that in 1996 and 2004, respectively, a significant correlation existed between institutional patriotism and all three exogenous independent variables. Institutional patriotism scores were highest in white males ( Table 9 4 ). In addition, institutional patriotism was strongest among older respondents; however, this trend was more con spicuous in 1996 than in 2004. That is, the gap in institutional patriotism between the old and young was much wider in 1996 than it was in 2004. For instance, in 1996, the oldest age group scored .62 higher than the youngest age group in institutional p atriotism. In 2004, that difference was just .15. This trend is illustrated in Figure 9 5
113 Overall, married people were significantly more institutionally patriotic than unmarried people during both years. However, in 1996, when unmarried categories w ere further broken down into subcategories, widows had the highest levels of institutional patriotism (3.96), followed by respondents who were married (3.87), divorced or separated (3.79), and those who had never been married (3.60). In 2004, institutiona l patriotism was strongest among married respondents (4.10), followed by widowed (3.95), divorced/separated (3.89), and never married (3.81). Apolitical patriotism. As shown in Tables 9 2 and 9 3, a significant correlation existed between apolitical patri otism and gender, with men displaying significantly higher levels of apolitical patriotism than women in both years of the survey. Apolitical patriotism was strongest in the older age groups. However, as with institutional patriotism, the correlation bet ween age and apolitical patriotism was somewhat stronger in 1996, and the differences across age groups once again converged somewhat in 2004 (Figure 9 6). Two correlations were observed only in 2004. First, whites in 2004 reported significantly higher levels of apolitical patriotism than nonwhites. Second, apolitical patriotism was significantly higher among married respondents than among those who were unmarried in 2004. Protectionist nationalism. The strongest bivariate predictors of protectionist nationalism were education and intellectual ability. Each was negatively correlated with protectionism ( Tables 9 2 and 9 3 ). During both years analyzed for this project, protectionism decreased as years of education increased. Furthermore, it appears p rotectionism flourishes where intellectual ability is scarce. This trend was fairly consistent across years ( Figure 9 7 ). Protectionism was also significantly correlated with how many children respondents had. In general, the more children they had, the more likely they were to agree with protectionist
114 statements. This positive correlation was especially strong in 1996, when parents of five or more children scored 3.49 on this study s protectionism scale, compared to rates of 3.30 for people who had betw een two and four children, 3.24 for those with only one child, and just 3.08 for childless respondents (Table 9 5). Nativist nationalism. Both education and intellectual ability correlated negatively with nativist nationalism. ( Tables 9 2 and 9 3 provide correlation coefficients and P values). Figure 9 8 shows not only that nativism declined as education level increased, but also that nativism was stronger in 2004 than in 1996 for all education levels. Additionally, in both years, nativism was generally more pronounced among older individuals than among younger ones (Table 9 5). Chauvinist nationalism. As with other nationalism indices, education and intellectual ability were among the strongest bivariate predictors of chauvinism. In both waves, levels of chauvinism dropped markedly and steadily with each additional level of education obtained (Figure 9 9). Another trend observed in both waves was that older people were typically more likely than younger people to adhere to chauvinist sentiments (Table s 9 2 and 9 3). Also in both waves, the more children respondents had, the stronger their expressions of chauvinism. Two significant associations with chauvinism were observed only in 2004. First, levels of chauvinist nationalism were greater among white s than among nonwhites. Second, chauvinism was stronger among married respondents than among unmarried respondents. Correlations with Civic Engagement This section presents the significant bivariate correlations involving the study s civic engagement indi cators. Tables 9 6 and 9 7 display present the mean values for measures of social capital and civic literacy, respectively, by covariates. Informal sociability. Informal sociability, a dummy variable for recurrent socialization with neighbors and/or frie nds, correlated negatively with family factors. On average, as a
115 respondent s number of children increased, his or her likelihood of informal socializing declined. Table 9 6 shows that 76 % of childless respondents socialized with friends or neighbors at least several times a month; among parents who had just one child, 63 % managed to socialize more than once monthly; and fewer than half of respondents with two or more children found social time for friends or neighbors more than one time per month. Figur e 9 10 demonstrates that married people were least likely to see neighbors or friends on an informal social basis more than once a month. By contrast, likelihood of informal socializing was highest for those who had never been married; 74% of never marrie d individuals reported socializing informally with neighbors or friends multiple times per month or more. Next likely were those who were divorced or separated, 68 % of whom said they socialized with neighbors or friends several times monthly at minimum. Even widows and widowers reported greater likelihood of informal socializing (59 % ) than married people (48 % ). Informal sociability was also negatively correlated with age. This was not a linear relationship, however, as Figure 9 11 reveals. Probability o f socializing multiple times per month was highest among respondents age 29 and under, over three quarters of whom (77 % ) indicated they socialized at least several times per month. Average rates of informal sociability then gradually declined from ages th irty to fifty, with each successive age group reporting lower likelihood of neighborly socializing than its predecessor. Only 44 % of people in their fifties socialized with neighbors or friends more than once in a typical month. After the age of 60, howe ver, informal sociability rebounds to some extent, with a little more than half of all respondents over the age of sixty reporting that they socialized with neighbors and/or friends multiple times each month.
116 Informal sociability was negatively correlated with protectionist nationalism ( r = .098; P value =.0060) This was the only measure among the five national pride indicators that was significantly related to informal sociability. Political participation. Political participation, measured as whether o r not a respondent voted in the most recent presidential election, was the only civic engagement submeasure assessed in both years of the present analysis. The 1996 survey item asks whether respondents took part in the 1992 presidential elections; the 200 4 measure probes participation in the 2000 presidential elections. Older respondents were generally more likely to report having voted in the previous presidential elections. For example, in both waves, the age group under the age of 30 was least likely t o say they d voted in the most recent presidential elections; just 61 % of 1996 respondents said they participated in the 1992 elections, and only 40 % of 2004 survey respondents stated reported that they voted in 2000. By contrast, among the oldest age gro up (those 70 and older), 87 % of 1996 respondents and 90 % of 2004 respondents claimed they had voted in their respective most recent presidential elections ( Table 9 6 for further details). Both education and intellectual ability were positively correlated with electoral participation in both waves. Figure 9 12 combines the two years data to chart the effect of educational attainment on likelihood of having voted in the last election. Each educational milestone carried with it an increased tendency to vot e. Fewer than half of those without a high school degree said they voted in the most recent presidential election. About two thirds (67 % ) of respondents with high school diplomas, 73 % of those with some college, 83 % of college graduates, and 91 % of respo ndents with schooling beyond college participated in presidential elections.
117 The bivariate relation between electoral participation and intellectual ability scores is presented in Figure 9 12. On average, respondents with the highest scores were most like ly to vote, and those with the lowest scores were least likely to do so. For example, among people who scored eight or above on this ten point assessment, 86 % voted. By contrast, fewer than half (47 % ) of individuals scoring three or below participated in the last election. Race/ethnicity appeared to affect electoral turnout, especially among 2004 survey respondents. Seventy four percent of whites claimed they voted in the 2000 presidential elections, compared to 58 % of blacks, and 42 % of Latinos. Also in 2004, marital status played a role in predicting likelihood of having voted in the most recent presidential elections. Over three quarters (76 % ) of married respondents in the 2004 sample said they d voted in the 2000 elections, compared to 60 % of divo rced/separated people, and barely one half (51 % ) of never married individuals. Political participation exhibited several statistically significant bivariate correlations with submeasures of national pride: Institutional patriotism was positively correlat ed with political participation in 2004 ( r = .090; P value = .0027). Although no correlation was found between these two variables in 1996, the positive bivariate association nonetheless remained significant when both waves data were combined ( r = .073; P value =.0014). In 2004 only, apolitical patriotism was positively correlated with political participation ( r = .091; P value = .0024). Protectionist nationalism was negatively correlated with political participation in 1996 ( r = .142; P value <.0001). Even though these two variables were uncorrelated in 2004, the association nonetheless remained significant when both waves data were combined ( r = .089; P value =.0001). In 1996 only, nativist nationalism was negatively correlated with political parti cipation ( r = .076; P value =.0336). In 1996, chauvinist nationalism was negatively correlated with political participation ( r = .079; P value =.0268).
118 Associational density. Three covariates were significantly correlated with associational density. Fi rst, r espondents with higher levels of education tended to join a greater number of community groups. Respondents with a graduate education, on average, belonged to the largest number of community associations (4.00), followed by respondents with college degrees (3.06), those who had taken some college classes (2.20), those with no education after earning their high school diploma (1.63), and those who never graduated high school, who on average held a mere 0.91 organizational memberships (Figure 9 14). Se cond, scoring well on the survey s intellectual assessment was strongly correlated with holding a greater number of voluntary group memberships. Respondent who scored particularly poorly in terms of cognitive skills (three or below) belonged to a mean of 0.93 community groups. By contrast, those who scored exceptionally high (eight or above) belonged to nearly three (2.83) organizations, on average ( Table 9 6 ). Marriage was the third and final independent variable significantly affecting the number of org anized groups to which individuals voluntarily belonged. On average, married respondents belonged to the greatest number of voluntary organizations (2.44), followed by people who had never married (1.86), and those who were divorced or separated (1.85). The lowest prevalence of associational density was found among people were widowed, averaging just 1.46 group memberships ( Table 9 6 ). Associational density was significantly related to all five national pride items in my study. The patriotism indices a nd nationalism indices had opposite relationships with organizational participation, however. Number of group memberships was positively correlated with institutional patriotism ( r =.101; P value =.0008) and apolitical patriotism ( r =.109; P value =.0003) By contrast, group membership was negatively correlated with protectionism ( r =
119 .115; P value =.0001), nativism ( r = .093; P value =.0019) and chauvinism ( r = .070; P value =.0198). Interpersonal trust. Several independent variables exhibited stati stically significant bivariate correlations with the interpersonal trust index. As was observed with other social capital measures, education and intellectual ability both had positive correlations with trust. Trust tended to be higher among whites than among nonwhites. In general, older respondents displayed greater degrees of interpersonal trust than younger respondents. Lastly, trust levels were highest among married respondents (3.28) and widows/widowers (3.24), and lowest among divorced or separate d individuals (2.99) and those who never married (2.91). All of these correlations are presented in Table 9 6 Trust was positively correlated with both patriotic attitudes and negatively correlated with two of the three nationalistic attitudes in my an alysis. Both institutional patriotism ( r =.214; P value <.0001) and apolitical patriotism ( r =.154; P value <.0001) had positive correlations with social trust; but trust was negatively correlated with protectionism ( r = .144; P value <.0001) and nativis m ( r = .095; P value =.0015) Newspaper reading. Education and intellectual ability were positively correlated to weekly newspaper reading. Ninety one percent of those with beyond a college education and 88 % of people with college degrees read newspaper s on at least a weekly basis, compared to 82% of high school grads and just 71% of those with less than a high school education (Table 9 7). As for how reading newspapers relates to national pride, the only significant association is a negative correlation between weekly newspaper readership and protectionist nationalism ( r = .110; P value =.0019)
120 Television nondependency. The notion of television nondependency denotes a sense of autonomy from television. One who has high levels of television nondepende ncy exhibits no reliance on, or habitual desire to watch TV. Recall from Chapter 5 that for the purposes of this study, TV nondependency is operationalized as the amount of time per day not watching television. Though this may seem counterintuitive, the researcher deemed it preferable because hours spent watching TV would have been the only inverted measure in the analysis. Findings shown in Table 9 2 indicate that television nondependency correlated negatively with age and positively with both educatio n and intellectual ability. In addition, high levels of nondependency were more common among white respondents. Television nondependency is negatively correlated with all three f orms of nationalistic attitudes (Table 9 2). Political interest. Respondents levels of political interest were positively correlated with how old they were, how educated they were, and where they fell along the intellectual ability spectrum. Those who said they were very interested in politics had a mean age of 50.08 years. O n average, they had 14.88 years of education and averaged 6.84 on the survey s intellectual assessment component. By contrast, respondents who said they were either not very interested or not at all interested were seven years younger, with a mean age of 42.97 years old. They also had two fewer years of education (12.79) and scored over a full point lower (5.75) on the cognitive examination. Political interest was significantly associated with all five national pride measures in my study. Specificall y, it was positively correlated with the patriotism indices institutional patriotism ( r =.114; P value =.0001) and apolitical patriotism ( r =.130; P value <.0001) and
121 negatively correlated with nationalism indices, protectionism ( r = .150; P value <.0001) nativism ( r = .065; P value =.0308) and chauvinism ( r = .078; P value =.0099). Political knowledge. The political knowledge index utilized a pair of survey items that probed the extent to which respondents perceived themselves as knowledgeable about politics vis vis their peers. Age correlated positively with this civic engagement measure, although once again not in a strictly linear fashion. Self reported political knowledge was lowest for respondents in their twenties (averaging 3.33 on the po litical knowledge index), and highest among respondents in their fifties and sixties (3.73 and 3.77, respectively). Political knowledge then dropped off a bit for people 70 and above (3.64). Men s self assessed political knowledge (3.73) was higher than that of women (3.50), and married respondents (3.67) higher than that of their unmarried counterparts (3.51). Among the various unmarried categories of respondents, those who were divorced or separated ended up with significantly the higher scores on th e political knowledge index (3.55), followed by people who had never been married (3.49), and widows (3.45). Lastly, respondents with greater levels of educational attainment and intellectual ability reported having high levels of political knowledge (Tab le 9 7). Political knowledge was positively correlated with both patriotism indices i.e., institutional patriotism ( r =.091; P value =.0026) and apolitical patriotism ( r =.096; P value =.0014) and negatively correlated with nationalism indices, protectioni sm ( r = .161; P value <.0001), nativism ( r = .150; P value <.0001) and chauvinism ( r = .090; P value =.0028). Engagement in public affairs. There was a significant correlation between being white and engagement in public affairs, with whites participa ting in a greater number of different forms of political and social action than nonwhites. Moving beyond the simple white/nonwhite
122 dichotomy, as Table 9 7 does, white respondents still had the highest mean number of social/political acts (1.64 ), but consi derable variation was observed among the different categories of nonwhite individuals Specifically, black respondents took part in an average of 1.12 activities, while Latinos engaged in 0.85 activities, and those classified as other took part in 0.82 acts, on average. 5 While considering public action in sociopolitical matters, I observed statistically significant bivariate correlations with two of the study s intermediate covariates. First, education was positively correlated with engagement in publi c affairs ; Table 9 7 shows that activity increases significantly with each successive educational milestone. Those with less than a high school education took part in the lowest average number of acts of public engagement (0.50). People with only a high school diploma engaged in an average of just 1.24 activities, and those with some college took part in 1.44. College educated respondents averaged 2.01 social/political acts. The individuals most engaged in public affairs were those who had received addi tional schooling beyond college; such individuals, on average, engaged in 2.67 public activities associated with social/political action. Figure 9 1 6 depicts a scatterplot of engagement in public affairs by years of education. Engagement in public affairs was also positively correlated with intellectual ability. The respondents who scored lowest on the intellectual ability assessment were also the least engaged in matters of public concern; those with intellectual ability scores of three or below engaged in 0.59 social or political activities, on average. Respondents who scored in the middle ranges on the ten point intellectual assessment took part in a moderate number of acts of public engagement; those who scored 4 or 5 engaged in a mean of 0.94 acts, w hile those who scored 6 5 As Chapter 5 discusses, there were not enough ethnic and racial minorities to further addre ss divisions of nonwhite respondents in multiple regression analyses
123 or 7 engaged in a mean of 1.42 acts. Finally, the top performers on the intellectual assessment were also the most active in social/political matters, as people with scores of eight or above averaged 2.25 acts of engagement in publ ic affairs. Engagement in public affairs was negatively correlated with a total of four national pride indicators, including one measure of patriotism institutional patriotism ( r = .064; P value = .0328) and the three nationalism indices, protectionism ( r = .245; P value <.0001), nativism ( r = .228; P value <.0001), and chauvinism ( r = .249; P value <.0001). Summary of Bivariate Findings Bivariate correlations between race and national pride revealed that whites tended to be significantly more patrioti c than nonwhites. Nationalism was largely unaffected by race, although in one of the waves, 2004, whites were significantly more chauvinist than nonwhites. Men expressed significantly higher levels of both forms of patriotism than women. N ationalism was largely equivalent for both genders; the one exception was that men displayed slightly higher levels of chauvinism than women in 2004. Age was positively correlated with both types of patriotism. It was also positively correlated with all three forms of nationalism in 1996, and with two of the three forms (nativism and chauvinism) i n 2004 Education and patriotism were generally unrelated, although a weak negative correlation existed in 1996 between years of education and apolitical patriotism. In striki ng contrast, I observed highly significant, negative correlations between education and all three forms of nationalism, both in 1996 and 2004. Marital status was significantly correlated with patriotism in several instances. Comparing ma rried and unmarrie d respondents those who were married expressed greater levels of
124 institutional patriotism during both years and higher levels of apolitical patriotism in 2004. Marriage and nationalism were largely uncorrelated, except that in 2004 married people had si gnificantly higher rates of chauvinist nationalism than unmarried people. Number of children was positively correlated to respondents levels of institutional patriotism in 1996. In addition, having higher quantities of children correlated positively with protectionism and chauvinism in 1996, and with all three nationalistic indices in 2004 (although in the latter wave correlations with protectionism and nativism were considerably weaker than correlations with chauvinism). The only statistically significan t association of intellectual ability to patriotism was a weak positive correlation with apolitical patriotism in 2004. However, in both waves, intellectual ability was show n to have highly significant, negative correlations with all three forms of nation alism. Race was significantly correlated with two of the four social capital indicators with w hites significantly more likely than nonwhite s to have participated in the mo st recent presidential election s, and to view strangers as generally fair and trustw orthy Whites also exhibited higher levels of civic literacy on four of the five indicators television nondependency, political interest, political knowledge, and engagement in public affairs Gender was uncorrelated with social capital, except that amon g 1996 respondents women were more likely than men to report that they had voted in the last presidential election. Men reported significantly higher levels than women on two of the five measures of civic literacy (political interest and political knowled ge).
125 Age was correlated with several forms of social capital, but the direction of these correlations was inconsistent. 6 B ivariate findings on the relationship between age and civic literacy were in conclusive in 1996, as newspaper reading was slightly mor e prevalent among older respondents but television nondependency was significantly higher among younger respondents. On the other hand, in 2004, two civic literacy indicators were positively correlated with age; older citizens reported that they were more interested in and knowledgeable about political matters than younger citizens The amount of education respondents had was highly correlated with nearly all measures of civic engagement. With the exception of informal sociability ( with which education was uncorrelated ), more education was correlated with greater levels of civic engagement. Married people had higher amounts of social capital than unmarried people in three of the four social capital categories ( political participation, associational density and interpersonal trust ) 7 Respondents who were married also scored higher than unmarried respondents on three of the five civic literacy indicators television nondependency, political knowledge, and engagement in public affairs. Number of children had v irtually no association with social capital, although it was negatively correlated with informal sociability and had a very slight positive correlation with voting in one of the waves. The number of children respondents had was completely uncorrelated wit h civic literacy. Nearly every civic engagement measure was positively influenced by intellectual ability. The lone exception was informal sociability, upon which intellectual ability had no effect. 6 Older people had lower levels of informal sociability but higher levels of interpersonal trust; they were significantly more likely to have voted in the prior presidential elections. 7 Married people, however, were less likely than unmarried people to socialize informally with neighbors or friends.
126 Of course, the findings presented in Chapter 9 refer onl y to bivariate correlations. While bivariate relationships can be a good starting point, to get a clearer picture of what is going on, one must consider multiple factors. Chapter 10 reports the findings from my multiple regression models.
127 Table 9 1. Des criptive s tatistics: GSS, 1996 and 2004 Variable Variable type In wave 1? Mean (1996) S.D. (1996) Range (1996) In wave 2? Mean (2004) S.D. (2004) Range (2004) Exogenous variables X1 Race (white = 1) Dummy 0.8293 0 1 0.8107 0 1 X2 Sex (male = 1) Dummy 0.4147 0 1 0.4366 0 1 X3 Age Ratio/censored 45.2364 15.9082 22 89+ 46.8723 15.7579 22 89+ Intermediate variables Z1 Education Ratio/censored 13.5778 2.8602 0 20+ 13.9194 2.81 68 0 20 Z2 Married (yes = 1) Dummy 0.507 0 0 1 0.5453 0 1 Z3 Number of children Ratio/censored 1.8571 1.5626 0 7 1.8605 1.5583 0 8+ Z4 Intellectual ability Interval 6.1353 2.0471 0 10 6.4058 1.9777 0 10 Patriotism P1 Institutional patriotism Scale/index 3.8092 0.7324 1 5 3.9922 0.7216 1 5 P2 Apolitical patriotism Scale/index 4.1777 0.6414 1 5 4.3612 0.5821 1 5 Nationalism N1 Protectionism Scale/index 3.2448 0.8021 1 5 3.1803 0.8143 1 5 N2 Nativism Scale/index 4.1467 0.8973 1 5 4.4549 0.7353 1 5 N3 Chauvinism Scale/index 3.6835 0.6867 1 5 3.7592 0.6979 1 5 Social capital C1 Informal sociability Dummy 0.5803 0 1 C2 Political participation Dummy 0.7320 0 1 0.6866 0 1 C3 Associational density Count 2.1458 2.4949 0 14 C4 Interpersonal trust Scale/index 3.1526 0.9529 1 4 Civic literacy L1 Newspaper reading Dummy 0.8306 0 1 L2 TV dependency Interval/censored 7.1327 1.8360 1 10 L3 Political interest Ordinal 2.8388 0.8681 1 5 L4 Political knowledge Scale/index 3.5974 0.8059 1 5 L5 Engagement in public affairs Ratio 1.4909 1.6775 0 8 (N=791) (N=1,104)
128 Table 9 2. Pearson s c orrelation c oefficients of the v ariables in the a nalysis : GSS 1996 (X2) (X3) (Z1) (Z2) (Z3) (Z4) (P1) (P2) (N1) (N2) (N3) (C1) (C2) (L1) (L2) (X1) Race (white=1) 0.06 0. 13 *** 0.09 0.12 *** 0.05 0.26 *** 0.19 *** 0.04 0.06 0.07 0.02 0.02 0.09 0.00 0.14 *** (X2) Sex (male=1) 0.04 0.05 0.04 0.08 0.03 0.10 ** 0.09 ** 0.00 0.04 0.05 0.02 0.07 0.00 0.03 (X3) Age 0.22 *** 0.03 0.34 *** 0.03 0.26 *** 0.13 *** 0.08 0.13 *** 0.23 *** 0.17 *** 0.18 *** 0.09 0.17 *** (Z1) Education 0.04 0.26 *** 0.54 *** 0.06 0.08 0.40 *** 0.30 *** 0.32 *** 0.05 0.26 *** 0.16 *** 0.28 *** (Z2) Marital status (married=1) 0.19 *** 0.03 0.08 0.00 0.01 0.06 0.03 0.21 *** 0.0 5 0.07 0.08 (Z3) Number of children 0.14 *** 0.09 ** 0.04 0.15 *** 0.06 0.18 *** 0.18 *** 0.01 0.02 0.07 (Z4) Intellectual ability 0.02 0.02 0.32 *** 0.29 *** 0.24 *** 0.01 0.27 *** 0.12 ** 0.18 *** (P1) Institutional patriotism 0.48 *** 0.02 0.22 *** 0.35 *** 0.02 0.07 0.05 0.01 (P2) Apolitical patriotism 0.05 0.17 *** 0.26 *** 0.04 0.00 0.01 0.06 (N1) Protectionism 0.35 *** 0.40 *** 0.10 ** 0.14 *** 0.11 ** 0.16 *** (N2) Nativism 0.43 *** 0.02 0.08 0.03 0.14 *** (N3) Chauvinism 0.05 0.08 0.00 0.20 *** (C1) Informal sociability 0.02 0.01 0.01 (C2) Political participation 0.12 *** 0.11 ** (L1) Newspaper reading 0.02 (L2) TV nondependency *p< .05; **p < .01; ***p< .001 (two tailed tests)
129 Table 9 3. Pearson s c orrelation c oefficients of the v ariables in the a nalysis : GSS 2004 (X2) (X3) (Z1) (Z2) (Z3) (Z4) (P1) (P2) (N1) (N2) (N3) (C2) (C3) (C4) (L3) (L4) (L5) (X1) Race (white=1) 0.01 0.15 *** 0.12 *** 0.13 *** 0.00 0.25 *** 0.22 *** 0.14 *** 0.02 0.02 0.10 *** 0.15 *** 0.03 0.21 *** 0.07 ** 0.09 ** 0.11 *** (X2) Sex (male=1) 0.05 0.02 0.05 0.07 0.04 0.09 ** 0.08 ** 0.01 0.04 0.07 0.02 0.05 0.03 0.09 ** 0.14 *** 0.02 (X3) Age 0.11 *** 0.07 0.37 *** 0.10 ** 0.09 ** 0.09 ** 0.05 0.10 *** 0.17 *** 0.31 *** 0.04 0.16 *** 0.14 *** 0.13 *** 0.04 (Z1) Education 0.08 ** 0.18 *** 0.44 *** 0.02 0.04 0.26 *** 0.27 *** 0.24 *** 0.29 *** 0.36 *** 0.25 *** 0.29 *** 0.30 *** 0.36 *** (Z2) Marital status (married=1) 0.17 *** 0.06 0.16 *** 0.11 *** 0.00 0.06 0.08 ** 0.18 *** 0.13 *** 0.15 *** 0.05 0.10 *** 0.07 (Z3) N umber of children 0.07 0.00 0.00 0.07 0.07 0.13 *** 0.07 0.03 0.03 0.00 0.02 0.04 (Z4) Intellectual ability 0.05 0.06 0.24 *** 0.21 *** 0.18 *** 0.27 *** 0.20 *** 0.27 *** 0.24 *** 0.30 *** 0.33 *** (P1) Institutional patriotis m 0.53 *** 0.09 ** 0.23 *** 0.44 *** 0.09 ** 0.10 *** 0.21 *** 0.11 *** 0.09 ** 0.06 (P2) Apolitical patriotism 0.10 *** 0.21 *** 0.32 *** 0.09 ** 0.11 *** 0.15 *** 0.13 *** 0.10 ** 0.02 (N1) Protectionism 0.33 *** 0.39 ** 0.06 0.11 *** 0.14 *** 0.15 *** 0.16 *** 0.25 *** (N2) Nativism 0.42 *** 0.00 0.09 ** 0.10 ** 0.07 0.15 *** 0.23 *** (N3) Chauvinism 0.01 0.07 0.01 0.08 0.09 ** 0.25 *** (C2) Political participation 0.26 *** 0.18 *** 0.33 *** 0.33 *** 0.28 *** (C3) Associational density 0.20 *** 0.23 *** 0.26 *** 0.38 *** (C4) Interpersonal trust 0.17 *** 0.17 *** 0.16 *** (L3) Political interest 0.47 *** 0.32 *** (L4) Political knowledge 0.36 *** (L5) Engagement in public affairs *p< .05; **p < .01; ***p< .001 (two tailed tests)
130 Table 9 4. Crosstabulation: Mean values for measures of patriotism, by covariates : GSS 1996 and 2004 Institutional patriotism Apolitical patriotism 1996 2004 1996 2004 Exogenous covariates Race White 3.87 4.07 4.19 4.40 Black 3.42 3.59 4.10 4.14 Latino 3.97 4.38 Other 3.73 3.87 4.19 4.20 Sex Male 3.90 4.07 4.25 4.42 Female 3.75 3.94 4.13 4.32 Age 29 and under 3.52 3.90 4.12 4.28 30 39 3.69 3.92 4.14 4.33 40 49 3.83 3.99 4.11 4.32 50 59 3. 99 4.05 4.21 4.43 60 69 3.98 4.08 4.25 4.45 70 and over 4.14 4.05 4.40 4.40 Intermediate covariates Education LT high school 3.92 3.86 4.15 4.12 High school 3.79 3.99 4.24 4.39 Junior college 3.75 3.97 4.12 4.43 College 3.81 4. 08 4.14 4.40 Grad school 3.82 3.99 3.97 4.33 Marital status Married 3.87 4.10 4.18 4.42 Widowed 3.96 3.95 4.22 4.30 Divorced/ s eparated 3.79 3.89 4.22 4.35 Never married 3.60 3.81 4.10 4.24 Number of children None 3.70 3.94 4.1 6 4.33 One 3.78 4.01 4.26 4.34 2 4 3.87 4.02 4.17 4.40 5 or more 3.81 3.90 4.09 4.21 Intellectual ability Scored 3 or below 3.73 3.87 4.02 4.02 Scored 4 or 5 3.82 3.98 4.24 4.39 Scored 6 or 7 3.79 3.99 4.23 4.41 Scored 8 or abo ve 3.85 4.03 4.09 4.35
131 Table 9 5. Crosstabulation: Mean values for measures of nationalism, by covariates : GSS 1996 and 2004 Protectionism Nativism Chauvinism 1996 2004 1996 2004 1996 2004 Exogenous covariates Race White 3.22 3.19 4.12 4.43 3.68 3.79 Black 3.48 3.31 4.40 4.58 3.76 3.61 Latino 2.97 4.68 3.71 Other 2.99 2.79 3.94 4.02 3.61 3.67 Sex Male 3.24 3.17 4.11 4.42 3.73 3.81 Female 3.25 3.19 4.17 4.48 3.65 3.72 Age 29 and under 3.20 3.08 4.11 4.39 3.50 3.55 30 39 3.25 3.21 4.09 4.43 3.67 3.69 40 49 3.14 3.16 4.00 4.40 3.55 3.79 50 59 3.20 3.21 4.09 4.42 3.68 3.76 60 69 3.42 3.19 4.34 4.51 3.84 3.89 70 and over 3.46 3.27 4.57 4.74 4.18 3.99 Intermediate covariates Education LT high school 3.66 3.42 4.63 4.70 4.07 3.95 High school 3.39 3.31 4.21 4.56 3.75 3.86 Junior college 3.26 3.31 4.15 4.55 3.53 3.74 College 2.84 2.92 3.88 4.19 3.45 3.58 Grad school 2.55 2.72 3.57 4.13 3.33 3.46 Marital status Married 3.25 3.18 4.10 4.49 3.70 3.81 Widowed 3.46 3.38 4.57 4.69 4.04 3.91 Divorced/ s eparated 3.24 3.21 4.18 4.34 3.66 3.76 Never married 3.13 3.08 4.05 4.38 3.49 3.56 Number of children None 3.08 3.07 4.04 4.36 3.46 3 .62 One 3.24 3.22 4.14 4.48 3.71 3.71 2 4 3.30 3.20 4.20 4.48 3.77 3.82 5 or more 3.49 3.38 4.11 4.60 3.75 3.98 Intellectual ability Scored 3 or below 3.53 3.38 4.47 4.54 3.87 3.87 Scored 4 or 5 3.48 3.40 4.38 4.66 3.83 3.93 Score d 6 or 7 3.27 3.21 4.18 4.49 3.68 3.79 Scored 8 or above 2.85 2.91 3.73 4.21 3.46 3.55
1 32 Table 9 6. Crosstabulation: Mean values for measures of social capital, by covariates : GSS 1996 and 2004 Informal sociability Political participation Assoc iational density Interpersonal trust 1996 2004 1996 2004 1996 2004 1996 2004 Exogenous covariates Race White 0.58 0.75 0.74 2.24 3.27 Black 0.55 0.69 0.58 2.17 2.62 Latino 0.42 1.14 2.84 Other 0.61 0.53 0.52 2.15 3.30 Sex Male 0.57 0.70 0.68 2.28 3.18 Female 0.59 0.76 0.69 2.04 3.13 Age 29 and under 0.77 0.61 0.40 1.67 2.93 30 39 0.62 0.68 0.61 2.10 3.02 40 49 0.53 0.76 0.71 2.37 3. 16 50 59 0.44 0.79 0.77 2.36 3.30 60 69 0.53 0.78 0.85 2.19 3.23 70 and over 0.51 0.87 0.90 2.02 3.42 Intermediate covariates Education LT high school 0.55 0.54 0.38 0.91 2.60 High school 0.56 0.7 0 0.65 1.63 3.04 Junior college 0.62 0.76 0.72 2.20 3.27 College 0.67 0.85 0.82 3.06 3.44 Grad school 0.56 0.94 0.89 4.00 3.59 Marital status Married 0.48 0.76 0.76 2.44 3.28 Widowed 0.59 0.76 0.79 1 .46 3.24 Divorced/ s eparated 0.68 0.72 0.60 1.85 2.99 Never married 0.74 0.66 0.51 1.86 2.91 Number of children None 0.76 0.75 0.63 2.14 3.09 One 0.63 0.68 0.63 2.11 3.10 2 4 0.49 0.75 0.73 2.22 3.20 5 or more 0.49 0.65 0.65 1.58 3.11 Intellectual ability Scored 3 or below 0.56 0.49 0.45 0.93 2.64 Scored 4 or 5 0.62 0.64 0.54 1.76 2.87 Scored 6 or 7 0.55 0.76 0.69 2.08 3.16 Scored 8 or above 0.59 0.8 8 0.85 2.83 3.48
133 Table 9 7. Crosstabulation: Mean values for measures of civic literacy, by covariates : GSS 1996 and 2004 Newspaper reading Television nondependency Political interest Political knowledge Engagement in public affairs 1996 20 04 1996 2004 1996 2004 1996 2004 1996 2004 Exogenous covariates Race White 0.83 7.25 2.88 3.66 1.64 Black 0.82 6.54 2.76 3.48 1.16 Latino 2.66 3.24 0.85 Other 0.86 6.64 2.67 3.47 0.82 Sex Male 0.83 7.19 2.93 3.73 1.45 Female 0.83 7.09 2.77 3.50 1.52 Age 29 and under 0.75 7.35 2.68 3.33 1.24 30 39 0.84 7.31 2.69 3.55 1.45 40 49 0.84 7.41 2.86 3 .60 1.52 50 59 0.86 7.06 2.89 3.73 1.80 60 69 0.81 6.89 3.03 3.77 1.57 70 and over 0.90 6.08 3.01 3.64 1.26 Intermediate covariates Education LT high school 0.71 6.18 2.38 3.22 0. 50 High school 0.82 7.03 2.75 3.46 1.24 Junior college 0.87 7.42 2.81 3.59 1.44 College 0.88 7.58 3.05 3.88 2.01 Grad school 0.91 8.02 3.34 4.10 2.67 Marital status Married 0.86 7.28 2.88 3.67 1.60 Widowed 0.83 6.13 2.77 3.45 1.22 Divorced/ s eparated 0.83 7.13 2.72 3.55 1.34 Never married 0.77 7.21 2.86 3.49 1.44 Number of children None 0.79 7.38 2.88 3.59 1.52 One 0 .82 7.07 2.75 3.53 1.61 2 4 0.86 7.05 2.86 3.63 1.49 5 or more 0.78 6.98 2.78 3.59 0.89 Intellectual ability Scored 3 or below 0.81 6.66 2.51 3.23 0.59 Scored 4 or 5 0.76 6.80 2.64 3.3 5 0.94 Scored 6 or 7 0.85 7.19 2.82 3.56 1.42 Scored 8 or above 0.88 7.57 3.11 3.94 2.25
134 Figure 9 1. Voting in presidential elections by election year. Figure 9 2. Mean years of education by age group in 1996 and 2004
135 Figure 9 3. Mean score for intellectual ability by education level among nonwhites and whites. Figure 9 4. Educational degrees obtained by number of children. The GSS variable that collapsed years of schooling into a series of formal educationa l milestones helps to illustrate the negative correlation between years of education and number of children. Nearly one third (31%) of respondents with less than a high school education had four or more children, compared to just 8% of those with at least a college degree.
136 Figure 9 5. Institutional patriotism by age group: 1996 and 2004. During both survey years, i nstitutional patriotism scores were generally higher among o lder age groups than among younger ones O verall institutional patriotism lev els were higher in 2004 than in 1996, and this is reflected in all age groups except those over 70. Figure 9 6. Apolitical patriotism by age group: 1996 and 2004. For nearly every age group, mean levels of apolitical patriotism were higher in 2004 than in 1996, as reflected in all age groups except those older than 70.
137 Figure 9 7. Protectionist nationalism by level of intellectual ability: 1996 and 2004. Levels of protectionism were highest for respondents who scored lowest on the GSS s intellectu al ability assessment. Figure 9 8. Nativist nationalism by educational degree obtained: 1996 and 2004. Low educational attainment corresponded to high levels of nativism in both waves. Nativist nationalism was higher in 2004 than in 1996, irrespective of education.
138 Figure 9 9. Chauvinist nationalism by educational degree obtained: 1996 and 2004. Low educational attainment corresponded to high levels of chauvinism in both waves. Chauvinism was higher in 2004 than in 1996 for all but the least educa ted citizens. Figure 9 10. Informal sociability by number of children in 1996. Respondents with fewer children were generally more likely to socialize informally with friends or neighbors.
139 Figure 9 11. Informal sociability by age group in 1996. In general, being young was correlated with greater having a likelihood of socializing informally Among all age groups, people between 50 and 59 were least likely to socialize with friends or neighbors. Figure 9 12. Electoral participation by educational degree obtained, combined for both waves. Likelihood of voting increased with each successive level of educational attainment.
140 Figure 9 13. Electoral participation by intellectual ability, combined for both waves. The better respondents performed on the survey s intellectual assessment, the more likely they were to have voted in the previous US presidential election. Figure 9 14. Mean number of associational affiliations, by educational degree in 2004.
141 Figure 9 15. Political knowledge by age group in 2004. Figure 9 1 6 Scatterplot: Engagement in public affairs by years of education in 2004.
142 Figure 9 1 7 Institutional patriotism by education level across time periods. Overall, levels of institutional patriotism increased between 1996 an d 2004, and this trend was reflected at every level of educational attainment with the exception of those who had achieved less than a high school education. Figure 9 18 Apolitical p atriotism by e ducation l evel across t ime p eriods Overall, levels of a political patriotism increased between 1996 and 2004, and this trend was reflected at every level of educational attainment with the exception of those who had achieved less than a high school education. 3.70 3.75 3.80 3.85 3.90 3.95 4.00 4.05 4.10 1996 2004 Less than HS High school Junior college Bachelor Graduate 3.90 4.00 4.10 4.20 4.30 4.40 4.50 1996 2004 Less than HS High school Junior college Bachelor Graduate
143 Figure 9 19 Protectionism by e ducation l evel acro ss t ime p eriods Overall, levels of protectionist nationalism remained fairly stable between 1996 and 2004, with a slight increase among the most educated strata, and a slight decrease among the least educated strata. Figure 9 20. Nativism by e ducation l evel across t ime p eriods Between 1996 and 2004, nativism saw its sharpest increase among respondents who were most educated; levels of nativism did not significantly change during that time period for those who had less than a high school education. 2.50 2.70 2.90 3.10 3.30 3.50 3.70 1996 2004 Less than HS High school Junior college Bachelor Graduate 3.50 3.70 3.90 4.10 4.30 4.50 4.70 1996 2004 Less than HS High school Junior college Bachelor Graduate
144 F igure 9 21. Chauvinism by e ducation l evel across t ime p eriods Overall, levels of chauvinist nationalism remained fairly stable between 1996 and 2004, with a slight increase among the most educated strata, and a slight decrease among the least educated st rata. 3.30 3.40 3.50 3.60 3.70 3.80 3.90 4.00 4.10 4.20 1996 2004 Less than HS High school Junior college Bachelor Graduate
145 CHAPTER 10 REGRESSION RESULTS National Pri de Regression Results (Models 1 3 ) This section summarizes regression analyses of national pride. Models 1 2, and 3 examined the effects of exogenous and intermediate covariates on each of my five nationa l pride indices. Results are shown in Tables 10 1 through 10 5. Institutional patriotism. Table 10 1 displays the OLS regression results for institutional patriotism. In Model 1, whites and males had significantly higher net levels of institutional patr iotism than nonwhites and f emales respectively There was also a positive net association between age and institutional patriotism controlling for other covariates The strength and statistical significance of these M odel 1 associations did not dimini sh when the intermediate covariates, education, marital status, number of children, and intellectual ability, were added to the equation in M odel 2. Marital status had a weak, but statistically significant, association with institutional patriotism. Mod el 3 shows that i nstitutional patriotism was stronger in 2004 than in 1996 controlling for the other covariates. Additionally, an interaction was observed between age and survey year ; while age had a positive net effect on institutional patriotism in b oth waves this effect was substantially stronger in 1996 than in 2004. Apolitical patriotism. Table 10 2 displays the OLS regression results for apolitical patriotism. As with institutional patriotism, net levels of apolitical patriotism were significant ly higher among white and males than among nonwhites and females, respectively Age had a significant positive influence on apolitical patriotism. There was also a positive association
146 between being white and apolitical patriotism, though weaker and les s significant than the association of whiteness with institutional patriotism. The net association of gender and race with apolitical patriotism weakened only slightly with the additional inclusion of intermediate covariates in M odel 2. Apolitical patriot ism had a weak positive association with being married and a weak negative association with number of children The relation of education and intellectual ability to apolitical patriotism was not statistically significant. Apolitical patriotism was strong er in 2004 than in 1996 controlling for the other covariates. In Model 3, no significant interactions were observed involving the dummy variable for year and any other covariate. Hence Model 2 is the correct model for this outcome. Protectionis t national ism Table 10 3 shows the OLS regression results for protectionism. In M odel 1, older people generally expressed significantly higher net levels protectionist nationalism than younger people. In M odel 2, both education and intellectual ability had negati ve net associations with protectionism, while marital status and parental involvement had null net associations. Controlling for the M odel 2 covariates produced a weak positive association between whiteness and protectionism Model 3 indicated that the n egative relationship between education and protectionism was smaller in 2004. Nativis t nationalism Table 10 4 displays the OLS regression results for nativism. Levels of nativism were significantly higher among older respondents, controlling for other co variates. Nonwhites displayed higher levels of nativism than whites ( although r ace s significance as a predictor of nativism disappeared when intermediate covariates were taken into account in M odel 2 ) Both education and intellectual ability had signifi cantly negative net effects on nativism. Taken together, these results suggest that the lower levels of nativism observed among
147 whites in Model 1 can be explained largely by their comparatively higher levels of education and intellectual ability. 1 Lastly n ativism was stronger in 2004 than in 1996, controlling for other relevant covariates. Chauvinis t nationalism Table 10 5 shows the OLS regression results for chauvinism. C h auvinism was significantly higher among men than among women, and generally high er among older individuals. Both e ducation and intellectual ability were negatively associated with chauvinism. Married people were slightly more likely to express chauvinist sentiments than were unmarried people Although race did not significantly inf luence chauvinism in Model 1, whites showed significantly higher levels of chauvinism than nonwhites when education and intellectual ability were considered in Model 2 Both education and intellectual ability displayed significant negative associations wi th chauvinism, controlling for the other covariates. Taken together, these results indicate that the non significance of race in Model 1 was likely due to whites comparatively higher levels of education and intellectual ability which may have served to suppress their higher levels of chauvinism 2 Model 2 results also indicated chauvinism was stronger in 1996 than in 2004 controlling for other covariates. However, w hen Model 3 introduced interactions the dummy variable for year became insignificant w hile the interaction variable for year and race was significant at the .01 level. The interaction indicated that the null 1 While education and intellectual ability were themselves highly correlated (Tables 9 2 and 9 3), it was also the case that intellectual scores were higher for whites than for nonwhites at each successive level of education (Figure 9 3). To further investigate, I reran Model 2 for nativism, once with education deleted and once with intellectual ability deleted. It turned out the model wit h education deleted changes race more than the one with intellectual ability. Both versions caused race to become statistically insignificant, but deleting education changed race more than deleting intellectual ability. (The P value for race was .86 with education deleted, versus .18 with intellectual ability was deleted). 2 To further investigate, I reran Model 2 for chauvinism, once with education deleted and once with intellectual ability deleted. Both versions had similar influences on the associatio n between race and chauvinism, with a P value of .0004 when education was deleted, versus 0198 when intellectual ability was deleted
148 net association between race and chauvinism is replaced by a positive relationship in 2004 in which whites had higher levels of chauv inism net of covariates. Civic Engagement Regression Results (Models 3 7) In Models 3 and 4, I investigated the effects of exogenous and intermediate covariates on social capital and civic literacy. Finally, in M odels 5 through 7, I added the national pri de measures presented above to the models for each of the measures of social capital and civic literacy. These results are shown in Tables 10 6 through 10 14. Recall that, unlike the national pride scales, which were computable in both 1996 and 2004, eig ht of the nine measures of civic engagement in this study were assessed in either one wave or the other. Voting in presidential elections was the only civic engagement variable assessed in both waves. Informal sociability. Logistic regression results in Table 10 6 reveal that impromptu socializing with neighbors and friends was most likely to occur among respondents who were young, unmarried, and had fewer children. Race, sex, education, and intellectual ability had no significant net association with in formal sociability. I nformal sociability was greater among those who displayed low l evels of protectioni sm, controlling for the other covariates in the model. Nativism and chauvinism had no significant association Neither institutional patriotism nor ap olitical patriotism was significantly related to informal sociability. Political participation. Among the nine civic engagement variables in my study, electoral participation was the only measure to appear in both data years. Table 10 7 shows logistic r egression results for voting in presidential elections. I pooled the data from both waves to analyze this phenomenon, as was done with the national pride measures. Likelihood of having voted in the previous presidential election was greater for responden ts in the 1996 sample than
149 for those in the 2004 sample There were no significant interactions with year, and therefore the interaction model was not reported in the table. Age and gender were significant predictors of political participation in all five regression models ; w omen in the sample had a slightly greater probability of voting than men, and likelihood of voting tended to be greater for older respondents In M odel 3, voter turnout was greater among whites than among nonwhites ; but th e associatio n disap peared after controlling for intermediate covariates in later models Respondents with more education were significantly more likely to have voted in the last election. Likewise, ma rried individuals were more likely than their unmarried peers to t ake part in the electoral process. Voter turnout was positively associated with intellectual ability, even after controlling for years of education. National pride measures had no net association with voting in presidential elections. Associational densi ty. Negative binomial regression models predicting associational density are presented in Table 10 8. I used a negative binomial model for this outcome measure because associational density was a count variable with a limited range. Two intermediate cov ariates were significantly associated with group membership: years of education and marital status. Married respondents tended to be involved with a greater numbers of group affiliations than married respondents Associational participation tended to be most prevalent among individuals with higher levels of education. Education was a strong predictor of associational density. Its impact on associational density was significant at the .001 level for all four of the regression mod els in which it was consi dered, whereas the predictive power of all other covariates fluctuated based on which national pride measures were included in the equation. Results for M odels 5 and 7 indicated a statistically significant net association between apolitical patriotism and number of organizational memberships. Individuals who exhibited high
150 levels of apolitical patriotism typically had greater involvement in organizations. None of the nationalism indices were significantly related to group membership, controlling for other variables in the analysis. Interpersonal trust. Table 10 9 shows the OLS regression results for interpersonal trust in 2004. In Model 3, race and age are associated with interpersonal trust. The association of older age remains consistent throughout th e models whereas the association of race decreases slightly (though remaining significant) as additional measures are introduced into the equation. Specifically, in Model 4, education, intellectual ability, and being married all are positively associated with high degrees of interpersonal trust. All three of these intermediate covariates remain significant as national pride measures are added to subsequent iterations of the regression model. Model 5 results show that institutional patriotism exhibits a positive association with trust. By contrast, in Models 6 and 7, nationalism exhibits a negative relationship with trust 3 Furthermore, the positive association observed between trust and institutional patriotism becomes even stronger in Model 7 when pro tectionism, nativism, and chauvinism are included in the analysis. Newspaper reading. Logistic regression was performed to examine the differences between newspaper readers and non newspaper readers as a measure of civic engagement. Table 10 10 shows the logistic regression of likelihood of newspaper reading on covariates and national pride indices in 1996. In Model 3, there was a weak positive relationship between respondents age and likelihood of reading newspapers. The positive association of age wi th 3 The negative association between chauvinism and trust is significant at the 0.10 level, while Table 10 9 only marks sig nificance past the .05 threshold.
151 newspaper readership grew stronger in Model 4, when exogenous and intermediate covariates were considered together. A similar relationship was observed between education and newspaper readership, with years of education positively affecting individuals likelihood of reading newspapers. In Model 5 there was no significant relationship between the patriotism indices and newspaper reading, but protectionism was negatively associated with readership in Models 6 and 7. Television nondependency. Table 1 0 11 displays the OLS regression results for television nondependency in 1996 a function of the amount of time per day spent not watching TV. Television nondependency was strongest among white people and among younger people. These effects diminished sl ightly with the inclusion of education and other intermediate covariates in Model 2. In fact the moderate positive influence of education on a respondent s television nondependency was stronger than the effects of either age or race. However, intellectua l ability had no impact on television nondependency, controlling for education. In Model 5, patriotism items were included in the model. Patriotic attitudes had no significant impact on respondents television watching behavior. When nationalism measures were included in Model 6, it was observed that chauvinism was negatively associated with television nondependency. In other words, people who displayed higher levels of chauvinism also watched greater amounts of television. Political interest. Table 10 12 shows the regression results for political interest in 2004. In Model 1, respondents self reported level of interest in politics was regressed on their race, sex, and age. Men in the study reported having significantly higher levels of political int erest than women. Age was also significantly related to political interest, as older individuals, in
152 general, assessed their level of interest in politics as higher than their younger counterparts. Race had no significant effect on political interest, co ntrolling for the other covariates. Model 4 controlled for four additional covariates. Education and intellectual ability were both found to be positively associated with political interest. Neither marriage nor parenthood was significantly related to su rvey respondents purported interest levels. Model 5 controlled for two patriotism indices, institutional patriotism and apolitical patriotism. Neither significantly affected the extent to which respondents said they were interested in political matters. Model 6 removed the patriotism scales from the equation and added three nationalism indices. Protectionism was negatively associated with interest in politics. The other two measures of nationalistic sentiment, nativism and chauvinism, were unrelated to political interests. Model 7 retained the three nationalism indices from Model 6 and reintroduced the patriotism scales from Model 5. In this final model, both institutional and apolitical patriotism were positively related to political interest. Also, the association between protectionism and political interest, observed in Model 6, disappeared in Model 7. However, chauvinism, which was not significant Model 6, exhibited a statistically significant negative association with political interest in the fi nal model. Political knowledge. Table 10 13 displays OLS regression results for political knowledge in 2004. In Model 3, males and whites, respectively, were more likely than females and nonwhites to report themselves as politically knowledgeable. There was also a positive association between age and levels of self reported political knowledge. The effects of sex and age were consistently significant throughout the models, but, similar to the results for other civic engagement indicators, race was only statistically significant in Model 3, where intermediate
153 covariates omitted from the model. In Models 4 through 7, both education and intellectual ability were positively associated with respondents levels of political knowledge. In Model 5, there was no association between patriotism measures and political knowledge. Similarly, Model 6 uncovered no direct relationship between nationalism scales and knowledge, although when protectionism, nativism, and chauvinism were controlled for, the relationship b etween marital status and political knowledge (previously insignificant) became significant. More precisely, married people rated themselves as more politically knowledgeable than unmarried people, controlling for the nationalism indices and the other cov ariates in Model 6. The significant relationships differed in Models 6 and 7. When both patriotic and nationalistic indices were considered in the final model, a weak negative association materialized between nativism and levels of political knowledge. Also, marriage was significant only in Model 6, but showed no statistically significant relationship with political knowledge in Model 7. Engagement in public affairs. Table 10 14 shows the regression results for engagement in public affairs in 2004. W hites were more likely than nonwhites to engage in public affairs, but this association was only significant in Model 3. When education and other intermediate covariates were included, race did not significantly affect one s level of engagement in public affairs. Individuals with more years of education were more engaged in public matters than were those with fewer years of education. Intellectual ability was positively associated with engagement as well. Institutional patriotism was negatively associat ed with engagement in public affairs when patriotic attitudes were first considered in Model 5. Patriotism scales were removed in Model 6 and replaced with nationalism scales. Both protectionism and chauvinism were negatively
154 related to engagement in pub lic affairs. That is, high levels of protectionism and chauvinism corresponded with low amounts of public engagement. Nativism had no significant effect. In Model 7 the patriotism and nationalism scales were considered together. Neither variation of pa triotism was statistically significant when nationalism indices were held constant. On the other hand, the negative effects of protectionism and chauvinism that were observed in Model 6 remained statistically significant in Model 7. Summary of Findings Na tional Pride Results Summary of patriotism findings. My first research question asked, Who expresses high levels of national pride? To answer this question, I first sought to uncover what sorts of citizens we re highly patriotic. Patriotism was the fir st of two national pride variables examined in the present study. Statistically significant f indings for two indices measuring domain specific national pride i nstitutional and apolitical patriotism are summarized as follows, controlling for other relevant covariates : Being white or male, older age, and marriage were positively associated with both institutional and apolitical patriotism Number of children had a positive association with one of the two forms of patriotism (apolitical patriotism). L evels o f both kinds of patriotism were higher in 2004 than in 1996. Summary of nationalism findings. The task of investigating who expresses high levels of national pride also involved finding out what sorts of citizens were prone to expressing its more national istic variety. I analyzed three elements of nationalism including protectionism, nativism, and chauvinism After co ntrolling for relevant covariates statistically significant f indings for we re as follows
155 E ducation and intellectual ability were negativ ely associated with all three forms of nationalism In other words, levels of nationalism generally increased as levels of education and intellectual ability decreased. A ge was positively associated with two of the three forms of nationalism (nativism and chauvinism) B ivariate correlation s were observed between age and all three forms of nationalism in 1996 M ales displayed higher levels than females of all three forms of nationalism, although in the case of protectionism this relationship was only signi ficant in M odel 1 (i.e., when education, marital status, number of children, and intellectual ability were not controlled ). F or all three forms of nationalism, being white had a more positive effect on the outcomes in M odel 2 than in M odel 1 For protecti onism and chauvinism, there was no association in M odel 1 and a significant positive association with whiteness in M odel 2. The regression of nativism on being white yielded a significant negative relation in M odel 1 and no relation in M odel 2. All of th is implies that the absence of education in Model 1 suppressed the relationship between race and nationalism. Levels of two of the three forms of nationalism (nativism and chauvinism) were higher in 2004 than in 1996. Civic Engagement Results Summary of s ocial capital findings. My second research question asked, Who has hig h levels of civic engagement? The first step toward resolving this question involved asking what sorts of citizens ha d high amounts of social capital, operationalized in this study a s informal sociability, political participation, associational density, and interpersonal trust. Below are my statistically significant social capital findings controlling for relevant covariates Education was positively associated with three out of fou r social capital measures (political participation, associational density, and interpersonal trust). Married respondents had higher levels of three kinds of social capital (political participation, associational density, and interpersonal trust) but unmar ried respondents enjoyed higher levels of informal sociability. Intellectual ability was positively associated with two social capital indicators (political participation and interpersonal trust).
156 Being white was associated with higher levels of two of the four social capital measures (political participation and interpersonal trust) 4 Age was positively associated with three social capital measures (political participation, associational density, and interpersonal trust ), and negative ly associated with inf ormal sociability. The positive associations were stronger for political participation and interpersonal trust than for associational density. In fact, the positive relation of age to associational density was significant only in Model 2 (i.e., when I co ntrolled for education, marital status, number of children, and intellectual ability). The number of children respondents had was negatively associated with informal sociability. Political participation was significantly higher among women than among men. A significantly higher proportion of 1996 respondents voted in the 1992 presidential election compared to the proportion of 2004 respondents that had voted in 2000. Summary of civic literacy findings. The next step to understanding the traits associated with high civic engagement was to inquire which people typically had high levels of civic literacy. Civic literacy was the second component of civic engagement explored in this research and consisted of newspaper reading, television nondependency, politi cal interest, political knowledge, and engagement in public affairs. Statistically significant c ivic literacy findings for civic literacy controlling for other relevant covariates, are summarized as follows: Higher levels of education were associated wit h scoring higher in all five civic literacy measures (newspaper reading, television nondependency, political interest, political knowledge, and engagement in public affairs). Intellectual ability was positively associated with three civic literacy indicato rs (political interest, political knowledge, and engagement in public affairs). Age was positively related to three of the five civic literacy measures (newspaper reading, political interest, and political knowledge), but negatively related to television n ondependency, as older people spent more time watching TV than their younger peers The positive relation of age to newspaper reading was stronger in Model 2, when I controlled for education, marital status, number of children, and intellectual ability. 4 For political participation, I observed this relationship only in model 1 (i.e., when education, marital status, number of children, and intellectual ability were not controlled) ;
157 M en were significantly more interested in and knowledgeable about politics than women according to self reports of political interest and knowledge. Whites had greater levels of television nondependency than nonwhites, as the latter spent more time watchin g TV. Being white also was associated with higher levels political knowledge and engagement in public affairs, although both of these relationships were present only in Model 1 which omitted education, marital status, number of children, and intellectual ability as covariates. On the Relationship between National Pride and Civic Engagement The third research question for this study involve d my central analytical problem broadly formulated to contemplate the nature of the relationship between national pri de and civic engagement. Since e ach of these concepts was measured in multiple ways, interesting subquestions emerge d Do patriotism and nationalism relate to civic engagement in different ways ? Do es social capital have the same association as civic lit eracy with national pride? To what extent are the observed associations indicative of direct associations between national pride and civic engagement, as opposed to common correlation with independent variables? Patriotism and social capital. Are patriot ic attitudes associated with social capital outcomes? Below I summarize my significant findings controlling for covariates, regarding the relationship between patriotism and social capital. Institutional patriotism exhibited a statistically significant, positive associat ion with interpersonal trust. Apolitical patriotism and associational density were significantly related, with higher levels of a political patriotism corresponding to greater quantities of organizational memberships. Patriotism and civic literacy. Are levels of patriotism and civic literacy mutual predictors of one an other? Findings on the nature of this relationship are summarized below, controlling for covariates.
158 Institutional patriotism was negatively related to engagement in public affairs, such that those expressing higher levels of institutional patriotism were less likely to become actively engaged in matters of public concern This association was observed only in Model 5 however. A fter controlling for nationalistic indices, i n Model 7, the relationship was no longer statistically significant Both institutional patriotism and apolitical patriotism were positively associated with political interest levels but only when I controlled for nationalistic indicators. Nationalism and social capital. To what extent is nationalism associated with measures of social capital? Controlling for the relevant covariates in my model, I present and summarize finding s below concerning the association between nationalism and social capital Prot ectionist nationalism was negatively associated with both interpersonal trust and informal sociability. In other words, those who displayed stronger attitudes of protectionism had less trust in their fellow citizens and were significantly less likely to s ocialize with neighbors or friends Neither nativism nor chauvinism was significantly associated with any of the four social capital measures. Nationalism and civic literacy. How are nationalism and civic literacy mutually related? Below I synthesize my significant findings on the nature of this relationship, controlling for covariates. Protectionis t nationalism was negatively associated with newspaper reading, political interest, and engagement in public affairs. Those with the strongest protectionist attitudes, in other words, were less likely to regularly read newspapers, reported significantly lower levels of political interest, 5 and were less engaged in matters of public concern. Nativis t nationalism was negatively associated with political knowledg e. That is, people who demonstrated greater degrees of nativist sentiment reported being less knowledgeable about politics than those who tended to reject nativism. 6 Chauvinis t nationalism was negatively associated with television nondependency, political interest, 7 and, engagement in public affairs In other words, respondents who displayed 5 The negative association between prote ctionism and political interest was statistically significant only in model 5. 6 The negative association between nativism and political knowledge was statistically significant only in model 7. 7 The negative association between chauvinism and political in terest was statistically significant only in model 7.
159 strong chauvinist attitudes had greater dependency on television, expressed less interest in politics, and were less active in public affairs.
160 Table 10 1. OLS r egres sion of institutional patriotism: GSS, 1996 and 2004 combined Model 1 Model 2 Model 3 Constant 3.188 3.231 3.022 (0.061) (0.104) (0.164) *** *** *** Race (white=1) 0.355 0.338 0.293 (0.042) (0.044) (0.070) *** *** *** Sex (male=1) 0.127 0. 118 0.147 (0.033) (0.033) (0.051) *** *** ** Age 0.006 0.006 0.011 (0.001) (0.001) (0.002) *** *** *** Education 0.005 0.003 (0.007) (0.011) Married 0.154 0.095 (0.033) (0.051) *** Number of children 0.012 0.005 (0.011) (0.018) Intellectual ability 0.004 0.004 (0.010) (0.015) Year=2004 0.177 0.173 0.501 (0.033) (0.033) (0.211) *** *** Race*( Y ear=2004) 0.075 (0.090) Sex*( Y ear =2004) 0.039 (0.066) Age*(Year=2004) 0.008 (0.002) *** Education*(Year=2004) 0.002 (0.014) Married*(Year=2004) 0.109 (0.067) Parenthood*(Year=2004) 0.027 (0.023) Intell ectual a bility*(Year=2004) 0.004 (0.019) *** p <= .001, *** p <= .01, *** p <= .05
161 Table 10 2. OLS regression of apolitical patriotism: GSS, 1996 and 2004 combined Model 1 Model 2 Model 3 Constant 3.865 3.927 4.130 (0.052) (0.089 ) (0.141) *** *** *** Race (white=1) 0.127 0.112 0.021 (0.036) (0.038) (0.060) *** ** Sex (male=1) 0.103 0.096 0.120 (0.028) (0.028) (0.044) *** *** ** Age 0.004 0.004 0.006 (0.001) (0.001) (0.001) *** *** *** Education 0.006 0.017 (0.006) (0.009) Married 0.071 0.036 (0.029) (0.044) Number of children 0.025 0.043 (0.010) (0.015) ** Intellectual ability 0.002 0.002 (0.008) (0.013) Year=2004 0.178 0.175 0.158 (0.028) (0.028) (0.182) *** *** Race*(Year=2004) 0.145 (0.077) Sex*(Year=2004) 0.032 (0.057) Age*(Year=2004) 0.003 (0.002) Education*(Year=2004) 0.019 (0.012) Marri ed*(Year=2004) 0.065 (0.058) Parenthood*(Year=2004) 0.031 (0.020) Intellectual a bility*(Year=2004) 0.006 (0.017) *** p <= .001, *** p <= .01, *** p <= .05
162 Table 10 3. OLS regression of protecti onism: GSS, 1996 and 2004 combined Model 1 Model 2 Model 3 Constant 3.140 4.400 4.695 (0.070) (0.113) (0.179) *** *** *** Race (white=1) 0.047 0.110 0.009 (0.049) (0.048) (0.076) Sex (male=1) 0.014 0.015 0.020 (0.038) (0.035) ( 0.055) Age 0.003 0.001 0.000 (0.001) (0.001) (0.002) ** Education 0.065 0.086 (0.007) (0.012) *** *** Married 0.021 0.022 (0.036) (0.056) Number of children 0.016 0.027 (0.012) (0.019) I ntellectual ability 0.071 0.071 (0.010) (0.016) *** *** Year=2004 0.071 0.023 0.504 (0.038) (0.036) (0.230) Race*(Year=2004) 0.161 (0.098) Sex*(Year=2004) 0.055 (0.072) Age*(Year=200 4) 0.002 (0.002) Education*(Year=2004) 0.034 (0.015) Married*(Year=2004) 0.007 (0.073) Parenthood*(Year=2004) 0.021 (0.025) Intellectual a bility*(Year=2004) 0.021 (0.021) *** p <= .001, *** p <= .01, *** p <= .05
163 Table 10 4. OLS regression of nativism: GSS, 1996 and 2004 combined Model 1 Model 2 Model 3 Constant 3.986 5.090 5.277 (0.069) (0.114) (0.180) *** *** *** Race (white=1) 0.124 0.00 4 0.031 (0.048) (0.048) (0.077) Sex (male=1) 0.060 0.071 0.055 (0.037) (0.036) (0.056) Age 0.006 0.006 0.006 (0.001) (0.001) (0.002) *** *** *** Education 0.055 0.057 (0.007) (0.012) *** *** Married 0.0 48 0.053 (0.036) (0.056) Number of children 0.020 0.027 (0.013) (0.020) Intellectual ability 0.067 0.067 (0.010) (0.017) *** *** Year=2004 0.297 0.336 0.019 (0.037) (0.036) (0.232) *** *** Race*(Ye ar=2004) 0.057 (0.098) Sex*(Year=2004) 0.023 (0.073) Age*(Year=2004) 0.002 (0.002) Education*(Year=2004) 0.003 (0.015) Married*(Year=2004) 0.174 (0.074) Parenthood*(Year=2004) 0.011 (0.026) Intellectual a bility*(Year=2004) 0.034 (0.021) *** p <= .001, *** p <= .01, *** p <= .05
164 Table 10 5. OLS r egression of chauvinism: GSS, 1996 and 2004 combined Model 1 Mode l 2 Model 3 Constant 3.233 4.076 4.148 (0.059) (0.096) (0.153) *** *** *** Race (white=1) 0.041 0.144 0.002 (0.041) (0.041) (0.065) *** Sex (male=1) 0.084 0.082 0.097 (0.032) (0.030) (0.047) ** ** Age 0.008 0.007 0.008 (0.00 1) (0.001) (0.002) *** *** *** Education 0.044 0.048 (0.006) (0.010) *** *** Married 0.074 0.041 (0.031) (0.048) Number of children 0.018 0.024 (0.011) (0.017) Intellectual ability 0.051 0.051 ( 0.009) (0.014) *** ** Year=2004 0.061 0.092 0.016 (0.032) (0.030) (0.197) ** Race*(Year=2004) 0.241 (0.084) ** Sex*(Year=2004) 0.020 (0.062) Age*(Year=2004) 0.002 (0.002) Educ ation*(Year=2004) 0.005 (0.013) Married*(Year=2004) 0.058 (0.063) Parenthood*(Year=2004) 0.011 (0.022) Intellectual a bility*(Year=2004) 0.011 (0.018) *** p <= .001, *** p <= .01, *** p <= .05
165 Table 10 6. Logistic regression of informal sociability: GSS, 1996 Model 3 Model 4 Model 5 Model 6 Model 7 Constant 1.227 1.818 2.000 2.675 2.977 (0.267) (0.502) (0.714) (0.846) (0.932) *** *** ** ** ** Race (white=1) 0.229 0.389 0.331 0.395 0.345 (0.196) (0.213) (0.216) (0.213) (0.217) Sex (male=1) 0.118 0.119 0.127 0.127 0.131 (0.149) (0.155) (0.156) (0.157) (0.158) Age 0.023 0.022 0.023 0.023 0.023 (0.005) (0.005) (0.005) (0.0 05) (0.006) *** *** *** *** *** Education 0.009 0.007 0.010 0.012 (0.033) (0.033) (0.034) (0.035) Married 0.908 0.924 0.917 0.928 (0.158) (0.159) (0.159) (0.160) *** *** *** *** Number of children 0.120 0.130 0.117 0.126 (0.053) (0.054) (0.054) (0.055) * Intellectual ability 0.033 0.032 0.044 0.044 (0.046) (0.046) (0.047) (0.047) Institutional patriotism 0.204 0.176 (0.125) (0.130) Apolit ical patriotism 0.193 0.190 (0.138) (0.139) Protectionism 0.277 0.266 (0.112) (0.113) Nativism 0.010 0.017 (0.099) (0.100) Chauvinism 0.124 0.107 (0 .132) (0.139) *** p <= .001, *** p <= .01, *** p <= .05
166 Table 10 7. Logistic regression of political participation: GSS, 1996 and 2004 combined Model 3 Model 4 Model 5 Model 6 Model 7 Constant 1.002 5.865 6.020 6.262 6.290 (0.199) (0.435) (0.569) (0.648) (0.698) *** *** *** *** *** Race (white=1) 0.500 0.036 0.019 0.042 0.033 (0.128) (0.146) (0.148) (0.147) (0.149) *** Sex (male=1) 0.233 0.277 0.282 0.265 0.267 (0.106) (0.116) (0.116) (0.117) (0.117) * Age 0.040 0.055 0.054 0.054 0.054 (0.004) (0.005) (0.005) (0.005) (0.005) *** *** *** *** *** Education 0.266 0.266 0.270 0.269 (0.027) (0.027) (0.027) (0.027) *** *** *** *** Married 0.503 0.496 0.499 0.495 (0.117) (0.1 18) (0.117) (0.118) *** *** *** *** Parenthood 0.048 0.046 0.044 0.044 (0.041) (0.041) (0.041) (0.042) Intellectual a bility 0.162 0.162 0.166 0.166 (0.034) (0.034) (0.035) (0.035) *** *** *** *** Year=2004 0.294 0.501 0.510 0.543 0.546 (0.108) (0.118) (0.119) (0.121) (0.121) ** *** *** *** *** Institutional patriotism 0.050 0.034 (0.093) (0.097) Apolitical patriotism 0.000 0.008 (0.107) (0.108) Protectionism 0.053 0.050 (0.081) (0.082) Nativism 0.127 0.124 (0.082) (0.083) Chauvinism 0.009 0.021 (0.097) (0.103) *** p <= .001, *** p <= .01, *** p <= .05
167 Tabl e 10 8. Negative binomial regression models predicting associational density: GSS, 2004 Model 3 Model 4 Model 5 Model 6 Model 7 Constant 0.514 1.924 3.034 2.102 2.791 (0.137) (0.232) (0.358) (0.394) (0.428) *** *** *** *** *** Race (white=1) 0.082 0.163 0.215 0.171 0.212 (0.096) (0.093) (0.095) (0.094) (0.095) Sex (male=1) 0.103 0.070 0.048 0.067 0.048 (0.075) (0.070) (0.069) (0.070) (0.069) Age 0.003 0.005 0.004 0.005 0.005 (0.002) (0.002) (0.002) (0.0 02) (0.002) Education 0.149 0.151 0.151 0.147 (0.014) (0.014) (0.015) (0.015) *** *** *** *** Married 0.267 0.225 0.263 0.226 (0.072) (0.072) (0.072) (0.072) *** ** *** ** Number of children 0.007 0.003 0.007 0 .002 (0.025) (0.025) (0.025) (0.025) Intellectual ability 0.036 0.039 0.039 0.033 (0.021) (0.021) (0.022) (0.022) Institutional patriotism 0.089 0.114 (0.058) (0.062) Apolitical patriotis m 0.184 0.200 (0.071) (0.072) ** ** Protectionism 0.010 0.007 (0.048) (0.048) Nativism 0.009 0.021 (0.052) (0.052) Chauvinism 0.039 0.057 (0.057) (0.062) Log Likelihood 2169.00 2086.32 2077.81 2086.01 2077.06 *** p <= .001, *** p <= .01, *** p <= .05
168 Table 10 9. OLS r egression of interpersonal trust: GSS, 2004 Model 3 Model 4 Model 5 Model 6 Model 7 Constant 2.403 1.142 0.284 1.513 0. 998 (0.101) (0.170) (0.252) (0.290) (0.310) *** *** *** ** Race (white=1) 0.463 0.292 0.215 0.294 0.225 (0.072) (0.071) (0.072) (0.072) (0.072) *** *** ** *** ** Sex (male=1) 0.039 0.038 0.013 0.025 0.001 (0.056) (0.054) (0.054) (0.054) (0.054) Age 0.008 0.008 0.007 0.008 0.008 (0.002) (0.002) (0.002) (0.002) (0.002) *** *** *** *** *** Education 0.060 0.061 0.055 0.050 (0.011) (0.011) (0.011) (0.011) *** *** *** *** Married 0.196 0.153 0.196 0.158 (0.055) (0.055) (0.055) (0.055) *** ** *** ** Number of children 0.002 0.007 0.001 0.008 (0.019) (0.019) (0.019) (0.019) Intellectual ability 0.067 0.067 0.061 0.054 (0.016) (0.015) (0.016) (0.016) *** *** *** *** I nstitutional patriotism 0.188 0.219 (0.044) (0.047) *** *** Apolitical patriotism 0.049 0.081 (0.054) (0.054) Protectionism 0.103 0.086 (0.037) (0.036) ** Nativism 0. 046 0.077 (0.041) (0.041) Chauvinism 0.073 0.043 (0.045) (0.048) *** p <= .001, *** p <= .01, *** p <= .05
169 Table 10 10. Logistic regression of newspaper reading: GSS, 1996 Model 3 Model 4 Model 5 Mode l 6 Model 7 Constant 0.932 2.036 2.196 1.866 1.764 (0.347) (0.669) (0.917) (1.057) (1.166) ** ** Race (white=1) 0.077 0.339 0.375 0.341 0.359 (0.255) (0.276) (0.278) (0.277) (0.280) Sex (male=1) 0.005 0.002 0.013 0 .004 0.005 (0.193) (0.200) (0.201) (0.202) (0.203) Age 0.016 0.025 0.024 0.024 0.023 (0.006) (0.007) (0.007) (0.007) (0.007) *** ** ** ** Education 0.184 0.183 0.176 0.175 (0.045) (0.045) (0.046) (0.046) *** *** *** ** Married 0.367 0.355 0.383 0.378 (0.203) (0.203) (0.204) (0.205) Number of children 0.015 0.011 0.018 0.015 (0.071) (0.072) (0.071) (0.072) Intellectual ability 0.033 0.034 0.027 0.028 (0.060) (0.059) (0.0 61) (0.061) Institutional patriotism 0.136 0.074 (0.156) (0.163) Apolitical patriotism 0.061 0.072 (0.173) (0.173) Protectionism 0.310 0.305 (0.148) (0.149) Nativism 0.073 0.069 (0.126) (0.127) Chauvinism 0.200 0.190 (0.167) (0.175) *** p <= .001, *** p <= .01, *** p <= .05
170 Table 10 11. OLS r egression of television nondependency: GSS, 1996 Model 3 Model 4 Model 5 Model 6 Model 7 Constant 7.462 5.104 5.338 6.476 6.477 (0.231) (0.410) (0.584) (0.676) (0.747) *** *** *** *** *** Race (white=1) 0.810 0.638 0.624 0.638 0.602 (0.171) (0.174) (0.177) (0.174) (0.176) *** *** *** *** *** Se x (male=1) 0.027 0.015 0.018 0.041 0.032 (0.129) (0.127) (0.127) (0.127) (0.127) Age 0.022 0.018 0.018 0.016 0.017 (0.004) (0.004) (0.004) (0.004) (0.004) *** *** *** *** *** Education 0.143 0.141 0.124 0.123 (0.027) (0. 027) (0.028) (0.028) *** *** *** *** Married 0.162 0.160 0.174 0.164 (0.128) (0.129) (0.128) (0.129) Number of children 0.057 0.052 0.065 0.061 (0.045) (0.045) (0.045) (0.045) Intellectual ability 0.030 0.0 30 0.015 0.015 (0.038) (0.038) (0.038) (0.038) Institutional patriotism 0.056 0.129 (0.101) (0.105) Apolitical patriotism 0.098 0.065 (0.111) (0.111) Protectionism 0.071 0.059 (0.090) (0.091) Nativism 0.002 0.008 (0.080) (0.081) Chauvinism 0.253 0.283 (0.108) (0.113) *** p <= .001, *** p <= .01, *** p <= .05
171 Table 10 12. OLS regression of political interest: GSS, 2004 Model 3 Model 4 Model 5 Model 6 Model 7 Constant 2.774 0.773 0.093 0.992 0.396 (0.136) (0.228) (0.343) (0.391) (0.421) *** *** Race (white=1) 0.183 0.022 0.080 0.006 0.051 (0.097) (0.096) (0.098) (0.097 ) (0.098) Sex (male=1) 0.217 0.217 0.195 0.224 0.200 (0.076) (0.073) (0.073) (0.073) (0.073) ** ** ** ** ** Age 0.011 0.013 0.013 0.014 0.013 (0.002) (0.003) (0.003) (0.003) (0.003) *** *** *** *** *** Education 0.114 0.115 0. 111 0.106 (0.015) (0.015) (0.015) (0.015) *** *** *** *** Married 0.021 0.012 0.018 0.014 (0.074) (0.075) (0.074) (0.074) Number of children 0.000 0.004 0.003 0.008 (0.026) (0.026) (0.026) (0.025) Inte llectual ability 0.073 0.072 0.066 0.059 (0.021) (0.021) (0.021) (0.021) *** *** ** ** Institutional patriotism 0.100 0.148 (0.060) (0.064) Apolitical patriotism 0.130 0.156 (0.073) (0.073) Protectionism 0.108 0.096 (0.050) (0.050) Nativism 0.083 0.051 (0.056) (0.056) Chauvinism 0.050 0.152 (0.061) (0.066) *** p <= .001, *** p <= .01, *** p <= .05
172 Table 10 13. OLS regression of political knowledge: GSS, 2004 Model 3 Model 4 Model 5 Model 6 Model 7 Constant 3.121 1.746 1.420 2.272 1.987 (0.086) (0.141) (0.213) (0.242) (0.261) *** *** *** *** *** Race (white=1) 0.151 0.020 0.041 0.008 0. 035 (0.061) (0.060) (0.061) (0.060) (0.061) Sex (male=1) 0.221 0.231 0.222 0.225 0.213 (0.048) (0.045) (0.045) (0.045) (0.045) *** *** *** *** *** Age 0.006 0.006 0.005 0.006 0.006 (0.002) (0.002) (0.002) (0.002) (0.002) *** ** *** *** *** Education 0.065 0.066 0.060 0.057 (0.009) (0.009) (0.009) (0.009) *** *** *** *** Married 0.083 0.071 0.091 0.076 (0.046) (0.046) (0.046) (0.046) Number of children 0.020 0.022 0.020 0.023 (0.016) (0 .016) (0.016) (0.016) Intellectual ability 0.080 0.080 0.073 0.070 (0.013) (0.013) (0.013) (0.013) *** *** *** *** Institutional patriotism 0.037 0.069 (0.037) (0.040) Apolitical patriotism 0.049 0.076 (0.045) (0.046) Protectionism 0.050 0.045 (0.031) (0.031) Nativism 0.056 0.071 (0.035) (0.035) Chauvinism 0.005 0.053 (0.038) (0.041) *** p <= .001, *** p <= .01, *** p <= .05
173 Table 10 14. Negative binomial regression models for engagement in public affairs: GSS, 2004 Model 3 Model 4 Model 5 Model 6 Model 7 Constant 0.048 2.274 2.031 0.786 1.014 (0.130) (0.214) (0.323) (0.338 ) (0.374) *** *** ** Race (white=1) 0.351 0.077 0.116 0.132 0.132 (0.096) (0.091) (0.092) (0.090) (0.091) *** Sex (male=1) 0.050 0.058 0.046 0.047 0.052 (0.070) (0.065) (0.065) (0.064) (0.064) Age 0.002 0.003 0. 003 0.004 0.003 (0.002) (0.002) (0.002) (0.002) (0.002) Education 0.114 0.111 0.094 0.093 (0.013) (0.013) (0.013) (0.013) *** *** *** *** Married 0.102 0.128 0.131 0.129 (0.066) (0.067) (0.065) (0.066) Num ber of children 0.010 0.010 0.004 0.003 (0.024) (0.023) (0.023) (0.023) Intellectual ability 0.121 0.121 0.098 0.098 (0.019) (0.019) (0.019) (0.019) *** *** *** *** Institutional patriotism 0.124 0.038 ( 0.052) (0.056) Apolitical patriotism 0.052 0.105 (0.067) (0.066) Protectionism 0.106 0.108 (0.043) (0.043) Nativism 0.061 0.067 (0.045) (0.046) Chauv inism 0.154 0.163 (0.051) (0.058) ** ** Log Likelihood 1835.94 1740.63 1737.64 1724.45 1723.19 *** p <= .001, *** p <= .01, *** p <= .05
174 CHAPTER 11 DISCUSSION AND CONCL USION National Pride Discussion What kinds of citizens express high levels of national pride? Informed by social dominance theory (Pratto, et al. 1994; Sidanius & Pratto, 1999) I hypothesized that levels of national pride are greatest among members of dominant social groups namely, whites, males, and older people. This series of hypotheses was only partially affirmed. Race, sex, and age all were fairly good predictors of patriotism but yielded somewhat mixed results in terms of their ability to predict nationalism. I also hypothesized that people with fe wer years of education express greater levels of national pride. Education performed well as a (negative) predictor of nationalism, but had minimal influence on patriotism. Lastly, I predicted that patriotism and nationalism indices would correlate highl y with one another, but that patriotism and nationalism would relate differently to other variables in the model. Findings supported both parts of this hypothesis. Patriotism and nationalism scales were significantly correlated. However, the two types o f national pride were associated with separate sets of independent variables, and patriotism s association with civic engagement was often the direct opposite to that of nationalism. Social Dominance Theory and National Pride Social dominance theory propos es that higher status groups within society are more likely to adopt hierarchy enhancing positions than are lower status groups (Pratto, et al. 1994; Sidanius & Pratto, 1999) Proponents of this viewpoint have applied a social dominance theory perspectiv e to the study of national attachment, often collapsing patriotism, nationalism, and other national pride dimensions into a singular concept.
175 On the whole, my findings largely supported the hypothesis that members of socially dominant groups are more pat riotic than members of subordinate social groups. As expected, whites, males, and older people had respectively higher levels of institutional patriotism than did nonwhites, females, and younger people. Apolitical patriotism was linked to being older a nd being male in both waves, and to being white in 2004. While race, sex, and age all influenced patriotism levels in the ways I forecasted, these demographic factors were somewhat less effective as predictors of nationalism. Nonetheless, I found some s upport for the hypothesized connection between socially dominant statuses and nationalism. Older people had significantly higher levels of nationalism on two of the three nationalistic indices, nativism and chauvinism. Chauvinism was also higher for mal es and for white people. Education and Nationalism One of the most consistent findings in my study was the link between education and nationalism, which was consistent with my second hypothesis that national pride levels decrease as education increases. F or all forms of nationalistic sentiment, and in both survey years, nationalism was strongest among respondents with lower levels of education, and levels of nationalism decreased with increasing levels of education. On the other hand, I failed to find sup port for the hypothesized link between patriotism and low levels of education. Both bivariate and multiple regression analyses showed no indication that education was associated with either institutional or apolitical patriotism. Are Patriotism and Nation alism the Same? Social dominance theory sometimes treats patriotism and nationalism as sub dimensions of a single phenomenon, or considers the former to be an element of the latter (Pratto, et al. 1994)
176 Felicia Pratto and her colleagues acknowledged tha t other schools of scholarly thought distinguished between patriotism and nationalism, and that different forms of national attachment ranged greatly from pro country feelings to comparative prejudice and beliefs that one s own country should dominate inferior foreign countries. But the social dominance theorists declared that these different forms of national attachment could all be handled interchangeably because group, and thus we postulate d that patriotism, nationalism, and chauvinism would all be significantly related to social dominance orientation (Pratto, et al. 1994, p. 742) Is social dominance theory s indiscriminate treatment of patriotism and nationalism appropriate? For this stu dy, I hypothesized that measures of patriotism and nationalism were positively correlated with one another, but that the two forms of national pride would differ in their associations with other variables in the model (Davidov 2011) The two patriotism measures, institutional patriotism and apolitical patriotism, were found to be highly correlated in both waves. Likewise all three nationalism measures were highly correlated with one another in both waves as well. Finally the patriotism and nationalism measures were all positively correlated in 2004; the same was true in 1996, with one exception: neither institutional nor apolitical patriotism were significantly related to protectionism (Tables 9 2 and 9 3). When data fr om both waves were pooled, all bivariate correlations within and between patriotism and nationalism were significant and positive, including the correlation between patriotism indices and protectionism (Table 11 1). Also consistent with the third hypothe sis, several covariates affected patriotism and nationalism differently. As discussed above, socially dominant racial, gender, and age categories were related to higher levels of patriotism but, on the whole, had little significance to nationalism
177 outcome s. On the other hand, increased levels of education were related to decreased nationalism levels but had virtually no impact on patriotism levels. The differential correlation of education with patriotism and nationalism supports the notion that these t wo forms of national pride are conceptually distinct (Blank & Schmidt, 2003; Connor, 1994; Davidov, 2009; Viroli, 1995) Overall, nationalism tended to have negative associations with civic engag ement, whereas patriotism exhibited weaker, more scattered relationships. This further illustrates that patriotism and nationalism may be fundamentally different phenomena. Civic Engagement Discussion What kinds of citizens display high levels of civic en gagement, in the form of social capital and civic literacy? I hypothesized that Americans with high levels of education display highest levels of both social capital and civic literacy. Controlling for other covariates, this hypothesis was supported for three of the four measures of social capital, 1 and for all five measures of civic literacy. I also predicted civic literacy scores are highest for Americans with highest levels of intellectual ability, which was operationalized as having high cognitive s cores on the GSS s WORDSUM assessment. This hypothesis was confirmed for three of the five measures of civic literacy. Intellectual ability had a positive net effect on respondents purported levels of political interest, political knowledge, and engagem ent in public affairs, but was not significantly associated with newspaper reading or television nondependency. One interpretation of these findings involves the years in which the respective measures were collected. The three civic 1 socializing informally on a regular basis with neighbors or friends.
178 literacy measures lin ked to high levels of intellectual ability (political interest, political knowledge, and engagement in public affairs) were collected only in 2004, whereas newspaper reading and television nondependency (for which I observed no significant net associations with intellectual ability) were collected only in 1996. Assuming that all variables were valid and reliable measures of civic literacy, these findings suggest intellectual ability influenced civic literacy in 2004, but not in 1996. The results of my an alysis have several important implications for our understanding of American citizenship. First, the study s most consistent finding was that civic engagement levels increase as education increases; this relationship was strong and significant across virt ually every regression model that I ran Second, despite cultural myths and political propaganda about a deep and profound national devotion that all Americans share toward their nation, it turns out America s most patriotic citizens are no more engaged t han the national average in matters of public and national concern. In fact, citizens with the strongest nationalistic tendencies tended to display lowest levels of civic engagement Directions for Future Research Future research needs to more fully explo re three areas which were not thoroughly discussed in this study: (1) religion, (2) homeownership, and (3) geography. The Role of Religion Religion and religiosity are central to scholarly debates on national identity. For some, national attachment is a secular replacement of religion; in other words, it is a secular form of consciousness, one that, indeed, sacralized the secular (Greenfield, 1996) For others, the foundations of modern national ideologies are rooted in rel igion.
179 Churches and religious institutions also have the potential to ameliorate the class bias in political participation. This is due to the combination of, on the one hand, the negative correlation between church attendance and social class, and, on t he other, the effectiveness of religious institutions in politically mobilizing their members. When offering explanations for why they volunteer, African Americans are more influenced by their church than are whites (Musick, Wil son, & Bynum, 2000) Religion s influence on how race affects civic engagement was not limited to realm of community volunteering. On average, blacks belonged to more associations than whites, largely because they were more likely to belong to ethnic a nd religious organizations than whites at the same socioeconomic level (Verba, et al. 1995) Again, members of churches, work related environments, and voluntary associations tend to include citizens with lower than average levels of education and income E ach newly mobilized voter from that demographic represents another step toward equality of political participation. For all of the above reasons, future analyses should further investigate this phenomenon by comparing models of associational density th at include religious participation to those that do not. In addition, subsequent studies might control for such religious indicators as a respondent s childhood religion, denomination, church attendance, strength of affiliation, and fundamentalism. Tenure Status Tenure status, defined as the distinction between renting versus owning one s domicile, likely has important implications for this research topic. Homeownership has been examined within the context of social stratification (Lee, 1998; Masnick, 2004; Newman, 2008) but also as a catalyst of civic engagement (DiPasqualea & Glaeser, 1999; Rotolo, Wilson, & Hughes, 2010) The decision to purchase a home signifies a lo ng term commitment to be part of a single community. Although in this case the commitment is obviously a financial one, it also
180 represents the probability that one will remain stationery in a given community for at least a number of years. In turn, a gre ater likelihood exists that he or she will form a bond or attachment to that community. Accordingly, studies have found positive correlations between tenure status and various civic engagement indicators (Bo wers, 2004; DiPasqualea & Glaeser, 1999; Rohe & Watson, 2007) Stakes theory (Rotolo, et al. 2010, pp. 572 573) suggests that civic engagement is higher among citizens who have a stake in their community. The theory assumes property ownership prov ides citizens with a stake in their community that causes them to care more about its wellbeing. Owning a home is a basic measure of social class (Rohe, van Zandt, & McCarthy, The Social Costs and Benefits of Homeownership, 2001) Citizens tend to cite economic motives for buying a home, and often their home is their only monetary asset (Saunders, 1990, p. 87) Yet, purchasing a home differs from other sorts of financial investments. Unli ke stocks or bonds, for instance, one s home is affixed to a specific geographic site. Furthermore, unlike other types of ownership, owning a home should create incentives for householders to improve the quality of their communities since community qualit y is capitalized into the value of their homes (DiPasqualea & Glaeser, 1999, p. 354) Owning a home is also associated with various family related characteristics (Aaronson, 2000) Studies s uggest homeownership functions as a deterrent against disengaged parenting (Grinstein Weiss, Williams Shanks, Manturuk, Key, Jong Gyu, & Greeson, 2010) 2 and facilitates positive child outcomes, such as higher educational attai nment, lower likelihood of 2 A recent study by Mi chal Grinstein Weiss and her colleagues found that homeowners were more likely to provide structured activities to their children than were renters and that television nondependency was higher among children of homeowners (2010)
181 poverty later in life and other undesirable outcomes later in life, fewer emotional and behavioral problems, and lower rates of teen pregnancy (Aaronson, 2000; Cairn ey, 2005; Green & White, 1997; Harkness & Newman, 2002; Haurin, Parcel, & Haurin, 2002) Tenure status would undoubtedly provide instructive clarification of my research findings, due to its unique dual status as an indicator of both socioeconomic factor s and observable family characteristics, as well as its relation to various measures of civic engagement including political involvement, associational density, engagement in public affairs, and television dependency. Several factors contributed to homeow nership s absence from my regression analyses. A major challenge to measuring the effects of homeownership on civic engagement has to do with the distinction between tenure status, residential mobility, and duration of residence. Putnam found that citize ns who expected to move within the next five years were significantly less likely to volunteer in their communities than those who did not anticipate moving (2000, p. 204) Renters tend to have lived in their com munity for shorter periods of time than homeowners, and it is widely understood that longstanding residents typically are more civically and politically engaged (Bowe rs, 2004; Lane, 1959; Rohe & Stewart, 1996, p. 51; Rosenstone & Hansen, 1993; Verba & Nie, 1972; Wolfinger & Rosenstone, 1980) It may be so it may be that length of residency, and not tenure status, is the primary force behind some of the results often attributed to homeownership in the literature (Rotolo, et al. 2010) Data on homeownership and length of residency was only sporadically available in the GSS, especially in 2004, but future studies should strive to account for both of these factors. Geo graphy Future studies should also take geographic data into account. How race mediates the relation among various civic engagement variables may fluctuate based on geography. In the
182 south, for instance, past research indicates that high levels of social trust are more likely to trigger volunteerism among blacks than among whites (Emig, et al. 1996) But in the northeast, there was no correlation between volunteerism and social trust, irrespective of race (Kohut, 1998)
183 Table 11 1. Pearson s correlation coefficients of all national pride variables: GSS, 1996 and 2004 combined (P2) (N1) (N2) (N3) (P1) Institutional patriotism 0.52 *** 0.05 0.24 *** 0.40 *** (P2) Apolitical patriotism 0.07 ** 0.21 *** 0.29 *** (N1) Protectionis m 0.33 *** 0.39 *** (N2) Nativism 0.43 *** (N3) Chauvinism *p< .05; **p < .01; ***p< .001 (two tailed tests)
184 APPENDIX A VARIABLE CODING, QUE STION WORDING, AND S CALE CONSTRUCTION Patriotism as Domain Specific National Pride Cal culating patriotism scales. I used a variation of the domain specific patriotism index established by Smith and Jarkko (Smith & Jarkko, 1998; Smith & Kim, 2006) and subsequently adapted by others (Dowley & Silver, 2000; Kersting, 2007; Smith & Kim, 2006) In their original articulation of this index, Smith and Jarkko (1998) simply called the concept national pride in s pecific achievements. Smith and colleagues later tweaked the terminology slightly, as a subsequent paper unveiled the phrase domain specific national pride to refer to the same ten item scale (Smith & Kim, 2006) Othe rs have referred to this index simply as patriotism (Dowley & Silver, 2000) reflecting its conceptual distance from the more nationalistic, generalized forms of national pride. Using data from the International Social Surv ey Program s (ISSP) 1995 National Identity Study (NIS), Smith and Jarkko (1998) constructed an additive ten item scale ranging from ten to fifty. The ten items, all of which were available in the GSS questionnaires in 1996 and 2004, probed respondents national pride in specific achievements by asking them how proud they were of the United States across ten realms: (1) the way democracy works; (2) political influence in the world; (3) economic achievements; (4) the soc ial security system; (5) science and technology; (6) sports; (7) arts and literature; (8) armed forces; (9) history; and (10) fair and equal treatment of all groups. For the original survey items, answer choices ranged from (1) very proud to (5) not pro ud at all. These response orders were reverse coded for each single item construct, with a high score (measured on 1 5 Likert scales) indicative of increased pride.
185 In my initial attempt to replicate Smith s ten item scale (Smith & Jarkko, 1998; Smith & Kim, 2006) reliability checks revealed acceptable Cronbach s alpha scores (0.79). However, the scale was problematic for reasons that likely stemmed from the large number of composite items. Multiple factors loaded, and inter item correlations revealed a higher alpha when certain measures were removed. Thus, it was necessary to assess patriotism using two separate measures. Based on observations of eigenvalues, I divided Smith s and Jarkko s additive index i nto a pair of indices which I referred to as institutional patriotism and apolitical patriotism. The two new scales of institutional patriotism and apolitical patriotism covered nine of the ten measures included in Smith s original index of domain specifi c national pride (Table A 1). 1 The first of my two measures of patriotism, which I labeled, institutional patriotism, consisted of five items: 2 Proud of the way democracy works in America Proud of America s political influence in the world Proud of America s economic achievements Proud of America s history Proud of America s fair and equal treatment of all groups in society My other measure of patriotism, which I called apolitical patriotism, included four composite items: 3 Proud of America s scientific and technological achievements Proud of America s achievements in sports Proud of America s achievements in the arts and literature Proud of America s armed forces 1 2 = 0. 73 ) 3 = 0. 66 )
186 Criteria for inclusion. Smith and Jarkko included in their scale only those respondents who an swered all ten items. However, following Dowley and Silver (2000) my much smaller sample necessitated using a mean score, keeping all respondents who provided data for one or more of the items. This strategy enabled m e to utilize patriotic attitudinal data for a greater number of respondents, whereas an additive model would have thrown out cases even if the information was missing for only a single item. Dowley and Silver (2000) f ound themselves in a similar situation; wanting to maximize the number of racial and ethnic minorities in their study, they opted to use a mean score on the ten items. Treatment of ambiguous/indecisive responses. Some respondents (about sixty cases alt ogether) answered that they Can t Choose for one of the composite items used in my patriotism scales. Since there was no available middle category for these items, I adopted Smith s recommended approach, which was to interpret Can t choose as an attem pted neutral answer on the part of respondents (Smith & Jarkko, 1998) Nationalism Scales Range and calculation of nationalism scales. All nationalism scales were constructed from items originally appearing in the ISSP NIS. The indices were calculated on a scale of 1 to 5, with 1 representing lowest levels of nationalism, and 5 representing its highest levels. To measure protectionist nationalism, I replicated the index constructed by Ceobanu and Escandell (2008) Protectionism was the mean of the answers to the following questions: 4 Now we would like to ask you about relations between America and other countries. How much do you agree or disagree with each of the following statem ents. 4 = 0. 61 )
187 America should limit the import of foreign products in order to protect its national identity Foreigners should not be allowed to buy land in America American television should give preference to American films and programs To measure nativist nationalism, I replicated the four item scale used by Ceobanu and Escandell (2008) Items were worded as follows: 5 Some people say the following things are important for being truly American. Others say they are not i mportant. How important do you think each of the following is... Following Ceobanu and Escandell (2008) my study s nativism scale included four of the seven possible response options 6 to this GSS item: To have been born in America? To have American citizenship? To have lived in America for most of one s life? To be able to speak English? To measure chauvinist nationalism, I constructed a multi item index identical to the one devised by Ceobanu and Escandell (2008) My chauvinism scale used four agree disagree items addressing matters of allegiance and national superiority. Respondents were asked, How much do you agree or disagree with each of the following statements. The stateme nts were read as follows: 7 I would rather be a citizen of America than of any other country in the world The world would be a better place if people from other countries were more like Americans 5 = 0. 75 ) 6 The other three answer categories wer e: (1) to be a Christian; (2) to respect America s political institutions and laws to respect America s political institutions and laws; (3) to feel American. 7 = 0. 63 )
188 Generally speaking, America is a better country than mo st other countries People should support their country even if the country is in the wrong Criteria for inclusion. Again, I took the mean score rather than rely on an additive model, and individual cases were retained as long as respondents provided a valid response to at least one question in each model (Dowley & Silver, 2000) Negative items were reverse coded and then summed; each sum was then divided by the denominator for that respondent. Social Capital Scales and Multi item Measures Associational density was computed by counting the total number of voluntary groups or organizations to which an individual belonged. The 2004 Topical Module for Socio Political Participation contained a set of items pertaining to me mberships in various kinds of groups and associations, including Fraternal Groups Service clubs Veterans groups Political clubs Labor unions Sports groups Youth groups School service groups Hobby or garden clubs School fraternities or sororities Nationali ty groups Farm organizations Literary, art, discussion or study groups Professional or academic societies Church affiliated groups Any other groups For each of the sixteen organizational categories, respondents were asked a series of questions.
189 Now we wo uld like to know something about the groups or organizations to which individuals belong. Here is a list of various organizations. Could you tell me whether or not you are a member of each type? People could hold memberships in multiple organizations withi n a given category. Respondents who indicated they participated in one of the sixteen types of organizations listed above were asked a follow up: Do you belong to more than one (organization of that type)? Those who again answered in the affirmative w ere asked a third question: How many (organizations of that type) do you belong to? At times, the data were incomplete or unclear. For instance, some respondents indicated (in the first question) that they were group members, but did not specify (in th e second question) whether they took part in multiple organizations or just one. 8 For all sixteen of the organizational categories, respondents identifying themselves as group participants were more likely to be participants in a single group than in mult iple groups. 9 Therefore, respondents with missing data on the indicator for one vs. multiple groups were coded as belonging to a single organization within that category. Other respondents indicated (in response to the second question) that they were me mbers of multiple organizations but failed to provide an exact number (in response to the third question). For those respondents, I imputed the median response among respondents who gave valid data for all three questions. 10 The difference between mean an d median for these measures 8 The highest incidence of this scenario was five cases for church affi liate groups 9 The exact proportions ranged from 65% (for professional societies) to 96% (for school fraternities). That is, 65% of people participants in professional societies belonged to just one professional society; 96% of school fraternity member s were only members of a single fraternity.
190 is negligible, given the relatively low incidence missing data (never more than nine cases in any one category), as well the shape of the distributions. For all sixteen classifications of group membership, distributions are ske wed to the right and the median response equals the mode. The median response among respondents with multiple affiliations was two. Thus, if someone confirmed he or she belonged to multiple garden clubs (for instance) but then refused or was unable to pr ovide a precise tally, I coded that person as being a member of two garden clubs. My measure of interpersonal trust is derived from two questions probing the extent to which respondents believed other human beings to be (a) fair, and (b) trustworthy. 11 The se items were found in the National Priorities section of the Environment Module, originally conducted in 1993 with follow ups in 1998 and 2004. As this portion of the survey was not administered in 1996, I am not able to compute a corresponding social tr ust measure for that wave. One subsection in 1996 contained a series of questions that asked if respondents thought others were helpful, fair, and trustworthy; however, that subsection has no overlap with the parts of the survey that asked about national pride. Civic Literacy Scales and Multi item Measures Political knowledge was a scale derived from a pair of self reported measures from the ISSP Role of Government module, which explores attitudes about political interest, trust, and efficacy. The politic al knowledge index was the mean of the answers to a pair of questions about 10 variable counting. The diffe rence between mean and median for these measures was negligible, given the relatively low incidence missing data (never more than nine cases in any one category), as well the shape of the distributions. For all sixteen classifications of group membership, distributions were skewed to the right and the median response equaled the mode. 11 = 0. 52 )
191 respondents self reported levels of political knowledge. Respondents were asked the extent to which they agreed with the following statements: 12 I feel that I have a pretty good understanding of the important political issues facing our country I think most people are better informed about politics and government than I am The second item was inverted before combining the pair to create a political knowledge index. The scale ran from 1 to 5. For 2004, a set of items was worded as follows: 13 Here are some different forms of political and social action that people can take. Please indicate, for each one, [w]hether you have done any of these things in the past year, [w]hether you have done it in the more distant past, [w]hether you have not done it but might do it, [o]r have not done it and would never, under any circumstances, do it. There were then eight items which included: Signed a petition Boycotted, or deliberately bought, certain products for political, ethical or environmental reasons Took part in a demonstration Attended a political meeting or rally Contacted, or attempted to contact, a politician or a civil servant to express your views Donated money or raised funds for a social or political activity Contacted or appeared in the media to express your views Joined an Internet political forum or discussion group 12 = 0. 58 ) 13 = 0. 28 )
192 Table A 1. Patriotism scale constructions: GSS variable names, question wording, and coding for the indices of institutional and apolitical patriotism Variables Definition (GSS item name) Correlation with total Institutional patriotism a How proud are you of America in each of the following? The way democracy works (PROUDDEM) 0.532 Its political achievements in the world (PROUDPOL) 0.551 America s economic achievements (PROUDECO) 0.520 Its history (PROUDHIS) 0.429 Its fair and equal treatment of all groups in society (PROUDGRP) 0.454 Apolitical patriotism a How proud are you of America in each of the following? Its scientific and technological achievements (PROUDSCI) 0.388 Its achievements in sports (PROUDSPT) 0.503 Its achievements in the arts and literature (PROUDART) 0.480 America s armed forces (PROUDMIL) 0.378 Source. GSS 1996 and 2004 a Recoded as follows: (5) very proud; (4) somewhat proud; (3) can t choose; (2) not very proud; (1) not proud at all.
193 Table A 2. Nationalism scale constructions: GSS variable names, question wording, and coding for the indices of protectionism nativism, and chauvinism Variables Definition (GSS item name) Correlation with total Protectionism a How much do you agree or disagree with each of the following statements? America should limit the import of foreign products in order to protect its national economy (IMPORTS) 0.382 American television should give preference to American films and programs (AMTV) 0.431 Foreigners should not be allowed to buy land in America (FORLAND) 0.456 Nativism b How important do you think each of the followin g is (for being truly American)? To have been born in America (AMBORNIN) 0.603 To have American citizenship (AMCIT) 0.520 To have lived in America for most of one s life (AMLIVED) 0.655 To be able to speak English (AMENGLSH) 0.525 Chauvinism a How much do you agree or disagree with the following statements? I would rather be a citizen of America than of any other country in the world (AMCITIZN) 0.406 The world would be a better place if people from other countries were more like the Ameri cans (BELIKEUS) 0.467 Generally speaking, America is a better country than most other countries (AMBETTER) 0.534 People should support their country even if the country is in the wrong (IFWRONG) 0.338 Source. GSS 1996 and 2004 a Recoded as foll ows: (5) agree strongly; (4) agree; (3) can t choose/neither agree nor disagree; (2) disagree; (1) disagree strongly. b Recoded as follows: (5) very important; (4) fairly important; (3) can t choose; (2) not very important; (1) not important at all.
194 Ta ble A 3. Social capital measures: GSS variable names, question wording, and coding for informal sociability, political participation, associational density, and social trust Variables Definition (GSS item name) Informal sociability a Would you use this car d and tell me which answer comes closest to how often you do the following things: Spend a social evening with someone who lives in your neighborhood (SOCOMMUN) Spend a social evening with friends who live outside the neighborhood (SOCFREND) Po litical participation b In 2000, you remember that Gore ran for President on the Democratic ticket against Bush for the Republicans. Do you remember for sure whether or not you voted in that election? (VOTE00) In 1992, you remember that Clinton ran for P resident on the Democratic ticket against Bush for the Republicans and Perot as an independent. Do you remember for sure whether or not you voted in that election? (VOTE92) Associational density c Now we would like to know something about the groups or org anizations to which individuals belong. Here is a list of various organizations. Could you tell me whether or not you are a member of each type? / [If yes]: Do you belong to more than one [group of that type]? / [If yes]: How many [groups of that type] do you belong to? Fraternal groups (MEMFRAT / TYPFRAT / NUMFRAT) Service clubs (MEMSERV / TYPSERV / NUMSERV) Veterans groups (MEMVET / TYPVET / NUMVET) Political clubs (MEMPOLIT / TYPPOLIT / NUMPOLIT) Labor unions (MEMUNION / TYPUNI ON / NUMUNION) Sports groups (MEMSPORT / TYPSPORT / NUMSPORT) Youth groups (MEMYOUTH / TYPYOUTH / NUMYOUTH) School service groups (MEMSCHL / TYPSCHL / NUMSCHL) Hobby or garden clubs (MEMHOBBY / TYPHOBBY / NUMHOBBY) School fraternit ies or sororities (MEMGREEK / TYPGREEK / NUMGREEK) Nationality groups (MEMNAT / TYPNAT / NUMNAT) Farm organizations (MEMFARM / TYPFARM / NUMFARM) Literary, art, discussion or study groups (MEMLIT / TYPLIT / NUMLIT) Professional or acade mic societies (MEMPROF / TYPPROF / NUMPROF) Church affiliated groups (MEMCHURH / TYPCHURH / NUMCHURH) Any other groups (MEMOTHER / TYPOTHER / NUMOTHER) Interpersonal trust d How often do you think that people would try to take advantage of you if they got the chance and how often would they try to be fair? (BEFAIR) Generally speaking, would you say that people can be trusted or that you can t be too careful in dealing with people? (CANTRUST) Source. GSS 1996 and 2004 a Recoded as follows: (1 ) almost daily/several times a week/several times a month; (0) once a month/several times a year/once a year/never/don t know. b VOTE92 was used in 1996, and VOTE00 was used in 2004. Item was recoded as follows: (1) voted; (0) did not vote/refused to an swer/don t know or can t remember. c Sum. d BEFAIR was coded as follows: (1) try to take advantage almost all of the time; (2) try to take advantage most of the time; (3) try to be fair most of the time; (4) try to be fair almost all of the time. CANTRUST was recoded as follows: (4) people can almost always be trusted; (3) people can almost always be trusted; (2) you usually can t be too careful in dealing with people; (1) almost always can t be too careful in dealing with people.
195 Table A 4. Civi c literacy measures: GSS variable names, question wording, and coding for newspaper reading, television nondependency, political interest, political knowledge, and engagement in public affairs Variables Definition (GSS variable name) Newspaper reading a Ho w often do you read the newspaper every day, a few times a week, once a week, less than once a week, or never? (NEWS) Television nondependency b On the average day, about how many hours do you personally watch television? (TVHOURS) Political interest c H ow interested would you say you personally are in politics? (POLINT1) Political knowledge d To what extent do you agree or disagree with the following statements? I feel I have a pretty good understanding of the important political issues facing America (POLEFF19) I think most people in America are better informed about politics and government than I am (POLEFF20) Engagement in public affairs e Here are some different forms of political and social action that people can take. Please indicate, for each one, [w]hether you have done any of these things in the past year, [w]hether you have done it in the more distant past, [w]hether you have not done it but might do it, [o]r have not done it and would never, under any circumstances, do it. Signed a peti tion (SIGNDPET) Boycotted, or deliberately bought, certain products for political, ethical or environmental reasons (AVOIDBUY) Took part in a demonstration (JOINDEM) Attended a political meeting or rally (ATTRALLY) Contacted, or attempted to co ntact, a politician or a civil servant to express your views (CNTCTGOV) Donated money or raised funds for a social or political activity (POLFUNDS) Contacted or appeared in the media to express your views (USEMEDIA) Joined an Internet political forum or discussion group (INTERPOL) Source GSS 1996 and 2004 a Recoded as follows: (1) every day/a few times a week/once a week; (0) less than once a week/never. b Table A 5 illustrates the operationalization of television nondependency. c Recod ed as follows: (5) very interested; (4) fairly interested; (3) can t choose; (2) not very interested; (1) not at all interested. d Recoded as follows: (5) strongly agree; (4) agree; (3) can t choose/neither agree nor disagree; (2) disagree; (1) strongly disagree. e Sum: (1) have done it in the past year; (0) have done it in the more distant past/have not done it but might do it/have not done it and would never do it/can t choose.
196 Table A 5. Operationalization of television nondependency Hours per day spent watching television Hours per day spent not watching television N (out of 2,904 respondents from original 1996 sample) N (out of 791 respondents analyzed in regression models) Television nondependency score 0 24 85 25 10 1 23 384 142 9 2 22 564 24 3 8 3 21 343 152 7 4 20 261 115 6 5 19 123 46 5 6 18 80 30 4 7 17 19 6 3 8 16 31 16 2 9 15 3 1 1 10 14 22 7 1 11 13 5 1 1 12 12 12 5 1 13 11 2 1 1 14 10 2 0 1 15 9 3 1 1 16 8 1 0 1 18 6 3 0 1 20 4 1 0 1 22 2 2 0 1 24 0 1 0 1 Missing 95 7 0
197 LIST OF REFERENCES Aaronson, D. (2000). A Note on the Benefits of Homeownership. Journal of Urban Economics, 47 (3), 356 369. Adorno, T. W., Frenkel Brunswik, E., Levinson, D. J., & Sanford, R. N. (1950). The Authoritaria n Personality. New York: Harper. Alter, P. (1994). Nationalism. New York: Oxford University Press. Alwin, D. (1991). Family of Origin and Cohort Differences in Verbal Ability. American Sociological Review, 56 (5), 625 638. Anderson, B. (1983). Imagined Comm unities: Reflections on the Origin and Spread of Nationalism. London: Verso. Armstrong, J. A. (1982). Nations before Nationalism. Chapel Hill, NC: University of North Carolina Press. Balibar, E., & Wallerstein, I. (1992). Race, Nation, Class: Ambiguous Ide ntities. London: Verso. Bar Tal, D. (1997). The Monopolization of Patriotism. In D. Bar Tal, & E. Staub (Eds.), Patriotism: In the Lives of Individuals and Nations (pp. 246 270). Chicago: Nelson Hall. Bar Tal, D., & Staub, E. (1997). Patriotism: In the Liv es of Individuals and Nations. Chicago: Nelson Hall. Bates, C., & Heaven, P. C. (2001). Attitudes to Women in Society: The Role of Social Dominance Orientation and Social Values. Journal of Community and Applied Social Psychology, 11 (1), 43 49. Bennett, L. L., & Bennett, S. E. (1989). Enduring Gender Differences in Political Interest: The Impact of Socialization and Political Dispositions. American Politics Quarterly, 17 (1), 105 122. Bennett, W. J. (1998). The Place to Harvest Patriots. School Administrator 55 (5), 38 40. Berger, P. L. (1977). Facing up to Modernity: Excursions in Society, Politics, and Religion. New York: Basic Books. Berger, P. L., & Luckmann, T. (1966). The Social Construction of Reality: A Treatise in the Sociology of Knowledge. New York : Doubleday. Blank, T., & Schmidt, P. (2003). National Identity in a United Germany: Nationalism or Patriotism? An Empirical Test With Representative Data. Political Psychology, 24 (2), 289 312.
198 Blau, P. M. (1977). Inequality and Heterogeneity. New York: Fr ee Press. Boudieu, P. (1986). The Forms of Social Capital. In J. G. Richardson (Ed.), Handbook of Theory and Research for the Sociology of Education (pp. 241 258). New York: Greenwood. Bowers, J. (2004). Does Moving Disrupt Campaign Activity? Political Psy chology 525 543. Brehm, J. (1993). The Phantom Respondents: Opinion Surveys and Political Representation. Ann Arbor, MI: University of Michigan Press. Brehm, J., & Rahn, W. (1997). Individual Level Evidence for the Causes and Consequences of Social Capita l. American Journal of Political Science, 41 (3), 999 1023. Breuilly, J. (1993). Nationalism and the State (2nd ed.). Manchester: Manchester University Press. Brubaker, R. (1996). Nationalism Reframed: Nationhood and the National Question in the New Europe. Cambridge: Cambridge University Press. Brubaker, R. (2004). In the Name of the Nation: Reflections on Nationalism and Patriotism. Citizenship Studies, 8 (2), 115 127. Burr, V. (1995). An Introduction to Social Constructionism. New York: Routledge. Butler, J. S., & Johnson, M. A. (1991). An Overview of the Relationships Between Demographic Characteristics of Americans and Their Attitudes Towards Military Issues. Journal of Political and Military Sociology, 19 273 291. Cairney, J. (2005). Housing Tenure and Psychological Well Being During Adolescence. Environment and Behavior, 37 (4), 552 564. Camilleri, J. A., & Falk, J. (1992). The End of Sovereignty? The Politics of a Shrinking and Fragmenting World. Aldershot: Edward Elgar. Caplan, B., & Miller, S. C. (201 0). Intelligence Makes People Think like Economists: Evidence from the General Social Survey. Intelligence, 38 (6), 636 647. Carroll, J. (2007). Post 9/11 Patriotism Remains Steadfast: Nonwhites Least Likely to Feel Highly Patriotic. In A. M. Gallup, & F. N ewport, The Gallup Poll: Public Opinion 2005 (pp. 266 267). Lanham, MD: Rowman and Littlefield. Cattell, R. B. (1987). Intelligence: Its Structure, Growth, and Action. New York, NY: Elsevier. Ceobanu, A. M., & Escandell, X. (2008). East is West? National F eelings and Anti immigrant Sentiment in Europe. Social Science Research, 37 (4), 1147 1170.
199 Citrin, J., & Wright, M. (2009). Defining the Circle of We: American Identity and Immigration Policy. The Forum, 7 (3), Article 6. Claibourn, M. P., & Martin, P. S. ( 2000). Trusting and Joining? An EmpiricalTest of the Reciprocal Nature of Social Capital. Political Behavior, 22 (4), 267 291. Coenders, M., & Scheepers, P. (2003). The Effect of Education on Nationalism and Ethnic Exclusionism: An International Comparison. Political Psychology, 24 (2), 313 343. Coleman, J. S. (1988). Social Capital in the Creation of Human Capital. American Journal of Sociology, 94 (Suppl), S95 S120. Coleman, J. S. (1990). Foundations of Social Theory. Cambridge, MA: Harvard University Press. Connell, R. W. (1995). Gender and Power: Society, the Person, and Sexual Politics. Berkeley, CA: University of California Press. Connor, W. (1994). Ethnonationalism: The Quest for Understanding. Princeton, NJ: Princeton University Press. Coulson, T. (1999 ). Voter Turnout in Canada: Findings from the 1997 Canadian Election Study. Electoral Insight, 1 18 22. Craige, B. J. (1996). Society, American Patriotism in a Global. Albany, NY: SUNY Press. Dalton, R. J. (1996). Citizen Politics: Public Opinion and Poli tical Parties in Advanced Industrial Democracies. Chatham, NJ: Chatham House. Davidov, E. (2009). Measurement Equivalence of Nationalism and Constructive Patriotism in the ISSP: 34 Countries in a Comparative Perspective. Political Analysis 64 82. Davidov, E. (2011). Nationalism and Constructive Patriotism: A Longitudinal Test of Comparability in 22 Countries with the ISSP. International Journal of Public Opinion Research, 23 (1), 88 103. Davis, J. A., & Smith, T. W. (1992). The NORC General Social Survey: A User's Guide. Newbury Park, CA: Sage. Davis, J. A., Smith, T. W., & Mardsen, P. V. (2005). General Social Survey, 1972 2004: Cumulative Codebook. Chicago: National Opinion Research Center. d e Figueiredo, R. J., & Elkins, Z. (2003). Are Patriots Bigots? An Inquiry into the Vices of In group Pride. American Journal of Political Science, 47 (1), 171 188. d e Las Casas, G. (2008, May/June). Is Nationalism Good for You? Foreign Policy (165), 50 56.
200 Delli Carpini, M. X., & Keeter, S. (1991). Stability and Change in the US Public's Knowledge of Politics. Public Opinion Quarterly, 55 (4), 583 612. Democratic National Committee. (2008). Renewing America's Promise: The 2008 Democratic National Platform. Washington, DC: Democratic National Convention Committee, Inc. Dietz M. G. (1989). Patriotism. In T. Ball, J. Farr, & R. L. Hanson (Eds.), Political Innovation and Conceptual Change (pp. 177 193). Cambridge: Cambridge University Press. DiPasqualea, D., & Glaeser, E. L. (1999). Incentives and Social Capital: Are Homeowners Better Citizens? Journal of Urban Economics, 45 (2), 354 384. Dogan, M. (2009). Political Legitimacy: New Criteria and Anachronistic Theories. International Social Science Journal, 60 (196), 195 210. Doob, L. W. (1964). Patriotism and Nationalism: Their Psy chological Foundations. New Haven, CT: Yale University Press. Dopplet, J., & Shearer, E. (1999). Nonvoters: America's No Shows. Thousand Oaks, CA: Sage. Dowley, K. M., & Silver, B. D. (2000). Subnational and National Loyalty: Cross National Comparisons. In ternational Journal of Public Opinion Research, 12 (4), 357 351. Ehrlich, T. (2000). Civic Responsibility and Higher Education. Phoenix: Oryx Press. Eller, J., & Coughlan, R. (1993). The Poverty of Primordialism: The Demystification of Ethnic Attachments. E thnic and Racial Studies, 16 (2), 183 202. Emig, A. G., Hesse, M. B., & Fisher, S. H. (1996). Black White Differences in Political Efficacy, Trust, and Sociopolitical Participation: A Critique of the Empowerment Hypothesis. Urban Affairs Review, 32 (2), 264 276. Enloe, C. (1990). Bananas, Beaches, and Bases: Making Feminist Sense of International Politics. Berkeley, CA: University of California Press. Feshbach, S. (1994). Nationalism, Patriotism and Aggression: A Clarification of Functional Differences. In L. R. Huesmann (Ed.), Aggressive Behavior: Current Perspectives (pp. 275 288). New York: Plenum. Fleming, K., Thorson, E., & Peng, Z. (2005). Associational Membership as a Source of Social Capital: Its Links to Use of Local Newspaper, Interpersonal Communica tion, Entertainment Media, and Volunteering. Mass Communication & Society, 8 (3), 219 240. Foels, R., & Reid, L. D. (2010). Gender Differences in Social Dominance Orientation: The Role of Cognitive Complexity. Sex Roles, 62 (9 10), 684 692.
201 Fukuyama, F. (199 5). Trust: The Social Virtues and the Creation of Prosperity. New York: Free Press. Galewitz, H. (Ed.). (2003). Patriotism: Quotations from Around the World. Mineola, NY: Dover Publications. Gallagher, S. K. (1994). Doing Their Share: Comparing Patterns of Help Given by Older and Younger Adults. Journal of Marriage and the Family, 56 (3), 567 578. Galston, W. A. (1991). Liberal Purposes: Goods, Virtues, and Diversity in the Liberal State. New York: Cambridge University Press. Galston, W. A. (2004). Review: C ivic Literacy: How Informed Citizens Make Democracy Work by Henry Milner. The Journal of Politics, 66 (4), 1341 1344. Ganuza, E., & Francs, F. (2008). Rethinking the Sources of Participation: A Case Study of Spain. European Sociological Review, 24 (4), 479 493. Geertz, C. (1973). The Interpretation of Cultures: Selected Essays (2nd ed.). London: Fontana. Gellner, E. (1983). Nations and Nationalism. Ithaca, NY: Cornell University Press. Gidengil, E., Blais, A., & Nevitte, N. (2004). Citizens. Vancouver: Unive rsity of British Columbia Press. Gidengil, E., Giles, J., & Thomas, M. (2008). The Gender Gap in Self Perceived Understanding of Politics in Canada and the United States. Politics & Gender, 4 (4), 535 561. Gittell, R. J., & Vidal, A. (1998). Community Organ izing: Building Social Capital as a Development Strategy. Thousand Oaks, CA: Sage. Gray, A. (2009). The Social Capital of Older People. Ageing & Society, 29 (1), 5 31. Green, R. K., & White, M. J. (1997). Measuring the Benefits of Homeowning: Effects on Chi ldren. Journal of Urban Economics, 41 (3), 441 461. Greenfield, L. (1996). The Modern Religion. Critical Review, 10 (2), 169 191. Grinstein Weiss, M., Williams Shanks, T. R., Manturuk, K. R., Key, C. C., Jong Gyu, P., & Greeson, J. K. (2010). Homeownership a nd Parenting Practices: Evidence from the Community Advantage Panel. Children and Youth Services Review, 32 (5), 774 782. Habermas, J. (1992). Citizenship and National Identity: Some Reflections on the Future of Europe. Praxis International, 12 (1), 1 19. Ha ll, S. (1992). The New Ethnicities. In J. Donald, & A. Rattansi (Eds.), Race, Culture and Difference (pp. 252 269). London: Sage.
202 Harkness, J., & Newman, S. J. (2002). Homeownership for the Poor in Distressed Neighborhoods: Does This Make Sense? Housing Po licy Debate, 13 (3), 597 630. Harlow, R., & Dundes, L. (2004). 'United' We Stand: Responses to the September 11 Attacks in Black and White. Sociological Perspectives, 47 (4), 439 464. Haurin, D. R., Parcel, T. L., & Haurin, R. J. (2002). Does Homeownership A ffect Child Outcomes? Real Estate Economics, 30 (4), 635 666. Hayes, B. C., & Bean, C. S. (1993). Gender and Local Political Interest: Some International Comparisons. Political Studies, 41 (4), 672 682. Hechter, M. (1998). Internal Colonialism: The Celtic Fr inge in British National Development, 1536 1966 (2nd Revised ed.). New Brunswick, NJ: Transaction Publishers. Hjerm, M. (1998). National Identities, National Pride and Xenophobia: A Comparison of Four Western Countries. Acta Sociologica, 41 (4), 335 347. Hj erm, M. (2001). Education, Xenophobia and Nationalism: A Comparative Analysis. Journal of Ethnic and Migration Studies, 27 (1), 37 60. Hobsbawm, E. (1983). Introduction: Inventing Traditions. In E. Hobsbawm, & T. Ranger (Eds.), The Invention of Tradition (p p. 1 14). Cambridge: Cambridge University Press. Hobsbawm, E. (1990). Nations and Nationalism Since 1780: Programme, Myth, Reality. Cambridge: Cambridge University Press. Hodgkinson, V. A., & Weitzman, M. S. (1996). Giving and Volunteering in the United St ates. Washington, D. C.: Independent Sector. Huang, M. H., & Hauser, R. M. (1998). Trends in Black White Test Score Differentials: II. The WORDSUM Vocabulary Test. In U. Neisser, The Rising Curve: Long Term Gains in IQ and Related Measure (pp. 303 332). Wa shington, DC, DC: American Psychological Association. Hutchinson, J. (1987). The Dynamics of Cultural Nationalism: The Gaelic Revival and the Creation of the Irish Nation State. London: Allen & Unwin. Hutchinson, J. (1994). Modern Nationalism. London: Font ana. Ignatieff, M. (1993). Blood and Belonging: Journeys into the New Nationalisms. London: Chatto and Windus. Inglehart, R. (1997). Modernization and Postmodernization. Princeton, NJ: Princeton University Press.
203 Ishio, Y. (2010). Social Bases of American Patriotism: Examining Effects of Dominant Social Statuses and Socialization. Current Sociology, 58 (1), 67 93. Jennings, M. K., Markus, G. B., & Niemi, R. G. (1991). Youth parent Socialization Panel Study, 1965 1982: Three Waves Combined. Ann Arbor, MI: Uni versity of Michigan,Center for Political Studies/Survey Research Center [producers], 1983. Ann Arbor: Inter university Consortium for Political and Social Research [distributor]. Jennings, R. (1997). Rethinking Ethnicity: Arguments and Explorations. London : Sage. Jones, F. L., & Smith, P. (2001). Individual and Societal Bases of National Identity: A Comparative Multi Level Analysis. European Sociological Review, 17 (2), 103 118. Junn, J. (1991). Participation and Political Knowledge. In W. Crotty (Ed.), Poli tical Participation and American Democracy (pp. 193 212). New York: Greenwood. Kawachi, I., Kennedy, B. P., & Loochner, K. (1997). Long Live Community: Social Capital as Public Health. The American Prospect (35), 56 59. Kellas, J. G. (1991). The Politics of Nationalism and Ethnicity. London: Macmillan. Kersting, N. (2007). Sport and National Identity: A Comparison of the 2006 and 2010 FIFA World Cups. Politikon South African Journal of Political Studies, 34 (3), 277 293. Knack, S., & Keefer, P. (1997). Does Social Capital Have an Economic Payoff? A Cross Country Investigation. The Quarterly Journal of Economics, 112 (4), 1251 1288. Knudsen, K. (1997). Scandinavian Neighbours with Different Character? Attitudes toward Immigrants and National Identity in Norway and Sweden. Acta Sociologica, 40 (3), 223 243. Kohn, H. (1944). The Idea of Nationalism. New York: Macmillan. Kohut, A. (1998). Trust and Citizen Engagement in Metropolitan Philadelphia. Washington, DC: Pew Research Center. Kosterman, R., & Feshbach, S. (1 989). Towards a Measure of Patriotic and Nationalistic Attitudes. Political Psychology, 10 (2), 257 274. Kwak, N., Shah, D. V., & Holbert, R. L. (2004). Connecting, Trusting, and Participating: The Direct and Interactive Effects of Social Associations. Poli tical Research Quarterly, 57 (4), 643 652. La Due Lake, R., & Huckfeldt, R. (1998). Social Capital, Social Networks, and Political Participation. Political Psychology, 19 (3), 567 584.
204 Lane, R. E. (1959). Political Life: Why People Get Involved in Politics. Glencoe, NY: Free Press of Glencoe. Lee, P. (1998). Housing Policy, Citizenship, and Social Exclusion. In A. Marsh, & D. Mullins (Eds.), Housing and Public Policy: Citizenship, Choice, and Control (pp. 57 78). Philadelphia, PA: Open University Press. Lehma nn, J. P. (1982). The Roots of Modern Japan. London: Macmillan. Leighley, J. E. (1990). Social Interaction and Contextual Influences on Political Participation. American Politics Quarterly, 18 (4), 459 475. Lesser, E. L. (Ed.). (2000). Knowledge and Social Capital: Foundations and Applications. Woburn, MA: Butterworth Heinemann. Levin, S. (2004). Perceived Group Status Differences and the Effects of Gender, Ethnicity, and Religion on Social Dominance Orientation. Political Psychology, 25 (1), 31 48. LeVine, R A., & Campbell, D. T. (1972). Ethnocentrism: Theories of Conflict, Ethnic Attitudes, and Group Behavior. New York: Wiley. Li, Y., Pickles, A., & Savage, M. (2005). Social Capital and Social Trust in Britain. European Sociological Review, 21 (2), 109 123. Loewen, J. W. (1995). Lies My Teacher Told Me. New York: The New Press. Masnick, G. C. (2004). Home Ownership and Social Inequality in the United States. In K. Kurz, & H. P. Blossfeld (Eds.), Home Ownership and Social Inequality in Comparative Perspective (pp. 304 337). Stanford, CA: Stanford University Press. McCleary, D. F., Nalls, M. L., & Williams, R. L. (2009). Types Of Patriotism As Primary Predictors Of Continuing Support For The Iraq War. Journal of Political and Military Sociology, 37 (1), 77 94. Mc Devitt, M. (2003). Review: Civic Literacy: How Informed Citizens Make Democracy Work by Henry Milner. Public Opinion Quarterly, 67 (3), 432 434. McNeill, W. H. (1986). Polyethnicity and National Unity in World History. Toronto: University of Toronto Press. Menard, S. W. (2001). Applied Logistic Regression Analysis. (Sage University Papers Series on Quantitative Applications in the Social Sciences, series no. 07 106). Thousand Oaks, CA: Sage. Mesch, D. J., Rooney, P. M., Steinberg, K. S., & Denton, B. (2006). The Effects of Race, Gender, and Marital Status on Giving and Volunteering in Indiana. Nonprofit and Voluntary Sector Quarterly, 35 (4), 565 587.
205 Mill, J. S. (1910). Representative Government. London: Dent and Sons. Milner, H. (2002). Civic Literacy: How I nformed Citizens Make Democracy Work. Medford, MA: Tufts University Press. Milner, H. (2010). The Internet Generation: Engaged Citizens or Political Dropouts. Medford, MA: Tufts University Press. Miner, J. B. (1957). Intelligence in the United States. New York: Springer Publishing. Miner, J. B. (1961). On the Use of a Short Vocabulary Test to Measure General Intelligence. Journal of Educational Psychology, 52 (3), 157 160. Mitter, W. (1993). Nationalism, Regionalism, and Internationalism in Europe. In K. Sch leicher, Nationalism in Education (pp. 107 131). Frankfurt: Peter Lang. Muller, E. N., & Seligson, M. A. (1994). Civic Culture and Democracy: The Question of Causal. American Political Science Review, 88 (3), 635 652. Musick, M. A., Wilson, J., & Bynum, W. B. (2000). Race and Formal Volunteering: The Differential Effects of Class and Religion. Social Forces, 78 (4), 1539 1570. Nagel, J. (1998). Masculinity and Nationalism: Gender and Sexuality in the Making of Nations. Ethnic and Racial Studies, 21 (2), 242 26 9. Nairn, T. (1981). The Break up of Britain: Crisis and Nationalism (2nd ed.). London: Verso. Newman, S. J. (2008). Does Housing Matter for Poor Families: A Critical Summary of Research and Issues Still to be Resolved. Journal of Policy Analysis and Manag ement, 27 (4), 895 925. Newton, K. (1997). Social Capital and Democracy. American Behavioral Scientist, 40 (5), 575 586. Underrepresentation. International Journal of Public Opi nion Research, 1 (2), 164 176. Parker, C. S. (2010). Symbolic versus Blind Patriotism: Distinction without Difference? Political Research Quarterly, 63 (1), 97 114. Pea, Y., & Sidanius, J. (2002). U.S. Patriotism and Ideologies of Group Dominance: A Tale of Asymmetry. The Journal of Social Psychology, 142 (6), 782 790. Perren, K., Arber, S., & Davidson, K. (2003). Men's Organisational Affiliations in Later Life: The Influence of Social Class and Marital Status on Informal Group Membership. Ageing and Society, 23 (1), 69 82.
206 Pichler, F., & Wallace, C. (2007). Patterns of Formal and Informal Social Capital in Europe. European Sociological Review, 23 (4), 423 435. Portes, A. (1998). Social Capital: Its Origins and Applications in Modern Sociology. American Sociolog ical Review, 24 1 24. Pratto, F., Sidanius, J., & Levin, S. (2006). Social Dominance Theory and the Dynamics of Intergroup Relations: Taking Stock and Looking Forward. European Review of Social Psychology, 17 (1), 271 320. Pratto, F., Sidanius, J., & Stall worth, L. M. (1993). Sexual Selection and the Sexual and Ethnic Basis of Social Hierarchy. In L. Ellis (Ed.), Social Stratification and Socioeconomic Inequality: A Comparative Biosocial Analysis (pp. 111 137). Westport, CT: Praeger. Pratto, F., Sidanius, J ., Stallworth, L. M., & Malle, B. F. (1994). Social Dominance Orientation: A Personality Variable Predicting Social and Political Attitudes. Journal of Personality and Social Psychology, 67 (4), 741 763. Putnam, R. D. (1993a). Making Democracy Work: Civic T raditions in Modern Italy. Princeton, NJ: Princeton University Press. Putnam, R. D. (1993b). The Prosperous Community: Social Capital and Public Life. The American Prospect (13), 35 42. Putnam, R. D. (1995). Bowling Alone: America's Declining Social Capital Journal of Democracy, 6 (1), 65 78. Putnam, R. D. (1996). The Strange Disappearance of Civic Life in America. The American Prospect (24), 34 38. Putnam, R. D. (2000). Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster Putnam, R. D. (2001). Civic Disengagement in Contemporary America. Government and Opposition, 36 (2), 135 156. Rahn, W. M., & Transue, J. E. (1998). Social Trust and Value Change: The Decline of Social in American Youth, 1976 1995. Political Psychology, 1 9 (3), 545 565. Raijman, R., Davidov, E., Schmidt, P., & Hochman, O. (2008). What Does a Nation Owe Non citizens? National Attachments, Perception of Threat and Attitudes toward Granting Citizenship Rights in Comparative Perspective. International Journal o f Comparative Sociology, 49 (2 3), 195 200. Ricci, D. M. (2004). Good Citizenship in America. New York: Cambridge University Press. Robinson, D. (1994). Teaching Whole People. National Forum, 74 (1), 34 36.
207 Rohe, W. M., & Stewart, L. S. (1996). Homeownership and Neighborhood Stability. Housing Policy Debate, 7 (1), 37 81. Rohe, W. M., & Watson, H. L. (Eds.). (2007). Chasing the American Dream: New Perspectives on Affordable Homeownership. Ithaca, NY: Cornell University Press. Rohe, W. M., van Zandt, S., & McCa rthy, G. (2001). The Social Costs and Benefits of Homeownership. Cambridge, MA: Joint Center for Housing Studies of Harvard University. Rosenstone, S. J., & Hansen, J. M. (1993). Mobilization, Participation and Democracy in America. New York: Macmillan. Ro tolo, T., Wilson, J., & Hughes, M. E. (2010). Homeownership and Volunteering: An Alternative Approach to Studying Social Inequality and Civic Engagement. Sociological Forum, 25 (3), 570 587. Rotter, J. B. (1980). Interpersonal Trust, Trustworthiness, and Gu llibility. American Psychologist, 35 (1), 1 7. Saunders, P. R. (1990). A Nation of Home Owners. London: Unwin Hyman. Schatz, R. T., Staub, E., & Lavine, H. (1999). On the Varieties of National Attachment: Blind Versus Constructive Patriotism. Political Psyc hology, 20 (1), 151 174. Schleicher, K. (1993). Nationalism in Education. Frankfurt: Peter Lang. Schuman, H., Steeh, C., Bobo, L. D., & Krysan, M. (1997). Racial Attitudes in America: Trends and Interpretations. Cambridge, MA: Harvard University Press. Selb ourne, D. (1994). The Principle of Duty: An Essay on the Foundations of Civic Order. London: Sinclair Stevenson. Shaw, T. C. (2004). 'Two Warring Ideals': Double Consciousness, Dialogue, and African American Patriotism Post 9/11. Journal of African America n Studies, 8 (1), 20 37. Shils, E. (1957). Primordial, Personal, Sacred, and Civil Ties. British Journal of Sociology, 7 13 45. Sidanius, J., & Pratto, F. (1999). Social Dominance: An Intergroup Theory of Social Hierarchy and Oppression. Cambridge: Cambrid ge University Press. Sidanius, J., Feshbach, S., Levin, S., & Pratto, F. (1997). The Interface Between Ethnic and National Attachment: Ethnic Pluralism or Ethnic Dominance? The Public Opinion Quarterly, 61 (1), 102 133.
208 Sidanius, J., Pratto, F., & Bobo, L. D. (1994). Social Dominance Orientation and the Political Psychology of Gender: A Case of Invariance? Journal of Personality and Social Psychology, 67 (6), 998 1011. Smith, A. D. (1983). Theories of Nationalism (2nd ed.). London: Duckworth. Smith, A. D. (19 86). The Ethnic Origins of Nations. Oxford: Blackwell. Smith, A. D. (1994). The Problem of National Identity: Ancient, Medieval and Modern? Ethnic and Racial Studies, 17 (3), 375 399. Smith, A. D. (1995). Nations and Nationalism in a Global Era. Cambridge: Polity Press. Smith, A. D. (1998). Nationalism and Modernism. New York: Routledge. Smith, A. D. (2005). Civic and Ethnic Nationalism. In P. Spencer, & H. Wollman (Eds.), Nations and Nationalism: A Reader (pp. 177 183). New Brunswick, NJ: Rutgers University Press. Smith, A. W. (1981). Racial Tolerance as a Function of Group Position. American Sociological Review, 46 (5), 558 573. Smith, A. W. (1985). Cohorts, Education, and the Evolution of Tolerance. Social Science Research, 14 (3), 205 225. Smith, T. (1997). Recognising Difference: The Romani 'Gypsy' Child Socialisation and Education. British Journal of Education, 18 (2), 243 256. Smith, T. W. (1979). Sex and the GSS: Nonresponse Differences. GSS Methodological Report No. 9 Smith, T. W. (2005). National pride : Over Time and Across Countries. GSSNews, 19 3 4. Smith, T. W., & Jarkko, L. (1998). National Pride: A Cross national Analysis. Chicago: NORC. Smith, T. W., & Kim, S. (2006). National Pride in Comparative Perspective: 1995/96 and 2003/04. International J ournal of Public Opinion Research, 18 (1), 127 136. Snellman, A., & Ekehammar, B. (2005). Ethnic Hierarchies, Racial Prejudice, and Social Dominance Orientation. Journal of Community and Applied Social Psychology, 15 (2), 83 94. Snellman, A., Ekehammar, B., & Akrami, N. (2009). The Role of Gender Identification in Social Dominance Orientation: Mediating or Moderating the Effect of Sex? Journal of Applied Social Psychology, 39 (4), 999 1012.
209 Sorek, T., & Ceobanu, A. M. (2009). Religiosity, National Identity and Legitimacy: Israel as an Extreme Case. Sociology, 43 (3), 477 496. Spencer, P., & Wollman, H. (2005). Good and Bad Nationalisms. In P. Spencer, & H. Wollman (Eds.), Nations and Nationalism: A Reader (pp. 197 217). New Brunswick, NJ: Rutgers University Pres s. Staub, E. (1997). Blind Versus Constructive Patriotism: Moving from Embeddedness in the Group to Critical Loyalty and Action. In D. Bar Tal, & E. Staub, Patriotism: In the Lives of Individuals and Nations (pp. 213 228). Chicago, IL: Nelson Hall. Stolle, D. (2001). Getting to Trust: An Analysis of the Importance of Institutions, Families, Personal Experiences and Group Membership. In P. Dekker, & E. M. Uslaner (Eds.), Social Capital and Participation in Everyday Life (pp. 118 133). London: Routledge. Sull ivan, J. L., Fried, A., & Dietz, M. G. (1992). Patriotism, Politics, and the Presidential Election of 1988. American Journal of Political Science, 36 (1), 200 234. Sumner, W. G. (1959). Folkways. New York: Dover. Sundeen, R. A., Garcia, C., & Raskoff, S. A. (2009). Ethnicity, Acculturation, and Volunteering to Organizations: A Comparison of African Americans, Asians, Hispanics, and Whites. Nonprofit and Voluntary Sector Quarterly, 38 (6), 929 955. Tamir, Y. (1993). United We Stand? The Educational Implication s of the Politics of Difference. Studies in Philosophy and Education, 12 (1), 57 70. Taylor, D. G., Sheatsley, P. B., & Greeley, A. M. (1978). Attitudes Toward Racial Integration. Scientific American, 238 (6), 30 37. Tocqueville, A. d. (1969). Democracy in A merica. Garden City, NY: Doubleday Anchor. Uslaner, E. M. (2002). The Moral Foundations of Trust. Cambridge: Cambridge University Press. Uslaner, E. M., & Brown, M. (2005). Inequality, Trust, and Civic Engagement. American Politics Research, 33 (6), 868 894 v an den Berghe, P. (1978). Race and Ethnicity: A Sociobiological Perspective. Ethnic and Racial Studies, 1 (4), 401 411. v an den Berghe, P. (1979). The Ethnic Phenomenon. New York: Elsevier. Verba, S., & Nie, N. H. (1972). Participation in America: Politi cal Democracy and Social Equality. New York: Harper and Row.
210 Verba, S., Burns, N., & Schlozman, K. L. (1997). Knowing and Caring about Politics: Gender and Political Engagement. The Journal of Politics, 59 (4), 1051 1072. Verba, S., Schlozman, K. L., & Brad y, H. E. (1995). Voice and Equality. Cambridge, MA: Harvard University Press. Viroli, M. (1995). For Love of Country: An Essay on Patriotism and Naitonalism. Oxford: Clarendon. Vogt, W. P. (1997). Tolerance and Education: Learning to Live With Diversity an d Difference. Thousand Oaks, CA: Sage. Walby, S. (1996). Woman and Nation. In G. Balakrishnan (Ed.), Mapping the Nation (pp. 235 254). New York: Verso. Wechsler, D. (1958). The Measurement and Appraisal of Adult Intelligence. Baltimore, MD: Williams & Witk ins. White, G. W. (2004). Nation, State, and Territory: Origins, Evolutions, and Relationships. Lanham, MD: Rowman and Littlefield. Wilson, J., & Musick, M. A. (1998). The Contribution of Social Resources to Volunteering. Social Science Quarterly, 79 (4), 7 99 814. Wolfe, L. M. (1980). The Enduring Effects of Education on Verbal Skills. Sociology of Education, 53 (2), 104 114. Wolfinger, R. E., & Rosenstone, S. J. (1980). Who Votes? New Haven, CT: Yale University Press. Wollebk, D., & Selle, P. (2003). The Im portance of Passive Membership for Social Capital Formation. In D. Stolle, & M. Hooghe (Eds.), Generating Social Capital (pp. 67 88). New York: Palgrave. Wuthnow, R. (1995). Learning to Care. New York: Oxford University Press. Yamagishi, T., & Yamagishi, M (1994). Trust and Commitment in the United States and Japan Motivation and Emotion, 18 (2), 129 166. Yuval Davis, N. (1997). Gender and Nation. London: Sage. Zhu, J., & Weiss, L. (2005). The Wechsler Scales. In D. P. Flanagan, & P. L. Harrison (Eds.), C ontemporary Intellectual Assessment: Theories, Tests, and Issues (pp. 297 324). New York, NY: The Guilford Press.
211 BIOGRAPHICAL SKETCH Matthew Van Voorhis received his Ph.D. from the University of Florida in 2011. While completing his dissertation, he w as a fulltime Research Coordinator at the external quality review organ ization for the Health and Human Services Commission of the state of Texas.