Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Egalitarianism begins at home and in cohorts: egalitarianism in United States public opinion, 1952-2000
(USC Thesis Other)
Egalitarianism begins at home and in cohorts: egalitarianism in United States public opinion, 1952-2000
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
EGALITARIANISM BEGINS AT HOME AND IN COHORTS: EGALITARIANISM IN UNITED STATES PUBLIC OPINION, 1952-2000 by Jeffrey H. Whitten ____________________________________________________________________ A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (POLITICAL SCIENCE) August 2007 Copyright 2007 Jeffrey H. Whitten ii Table of Contents List of Tables iv List of Figures vii Abstract viii Chapter 1: Introduction 1 The Meaning of Equality Egalitarianism Across Time 10 The Formation of Egalitarianism 10 Chapter 2: Hypotheses 22 The Meaning of Equality 22 Predictors of Egalitarianism 27 Chapter 3: Data and Methods 38 Data 38 Dependent Variables 39 Independent Variables 41 In Defense of the Beleaguered Survey 42 The Identification Problem 47 Measurement Error 48 Rankings and Ratings 51 Chapter 4: The Meaning of Equality for the U. S. Public 53 Egalitarianism in the Aggregate 53 Egalitarianism Across Time: A Value Realignment? 62 Separate Grooves for Different Groups? 76 Chapter 5: Egalitarianism Begins at Home 91 Introduction 91 Results 92 Discussion 95 Chapter 6: Egalitarianism in Cohorts 97 Introduction 97 Results 98 Discussion 126 iii Chapter 7: Conclusion: Only the Young 138 Findings 138 Implications 146 Further Research 149 Bibliography 152 Appendices Appendix A: Question Wordings and Response Categories for Variables 188 Appendix B: Methodological Appendix 207 Appendix C: Measurement Model Parameter Estimates 209 iv List of Tables Table 1.1: Conceptualizations of Equality Mentioned in the Literature 3 Table 3.1. Measures of Types of Egalitarianism 40 Table 4.1. Measures of Fit for Egalitarianism Structural Equation Models 55 Table 4.2. General Egalitarianism Structural Equation Measurement Model Coefficients 57 Table 4.3. Opportunity Egalitarianism Structural Equation Measurement Model Coefficients 58 Table 4.4. Economic Egalitarianism Structural Equation Measurement Model Coefficients 59 Table 4.5. Racial Egalitarianism Structural Equation Measurement Model Coefficients 60 Table 4.6. Gender Egalitarianism Structural Equation Measurement Model Coefficients 61 Table 4.7. Means of Types of Egalitarianism in Various Years 63 Table 4.8. General Egalitarianism Structural Equation Model Fit Across Time 67 Table 4.9. Equal Opportunity Structural Equation Model Fit Across Time 69 Table 4.10. Economic Egalitarianism Structural Equation Model Fit Across Time 69 Table 4.11. Racial Egalitarianism Structural Equation Model Fit Across Time 70 Table 4.12. Gender Egalitarianism Structural Equation Model Fit Across Time 70 Table 4.13. Gay Rights Egalitarianism Structural Equation Model Fit Across Time 71 Table 4.14. Immigrant Egalitarianism Structural Equation Model Fit Across Time 71 v Table 4.15. Means and Structural Equation Model Fit of Types of Egalitarianism by Income 79 Table 4.16. Means and Structural Equation Model Fit of Types of Egalitarianism by Race 81 Table 4.17. Means and Structural Equation Model Fit of Types of Egalitarianism by Gender 86 Table 5.1. Theories and Hypotheses to Be Tested in Chapter 5 91 Table 5.2. Structural Equation Coefficients for Youth Egalitarianism, 1982: Political Knowledge/Group Identification 93 Table 5.3. Structural Equation Coefficients for Youth Egalitarianism, 1982: Demographics/Religiosity 94 Table 5.4. Structural Equation Coefficients for Youth Egalitarianism, 1997 96 Table 6.1. Additional Hypotheses Tested in Chapter 6 97 Table 6.2. Structural Equation Coefficients for NES Equal Opportunity Group Membership Models 99 Table 6.3. Structural Equation Coefficients for NES Equal Opportunity Group Identification Models 102 Table 6.4. Structural Equation Coefficients for NES Racial Egalitarianism Group Membership Models 105 Table 6.5. Structural Equation Coefficients for NES Racial Egalitarianism Group Identification Models 107 Table 6.6. Structural Equation Coefficients for NES Gender Egalitarianism Group Membership Models 110 Table 6.7. Structural Equation Coefficients for NES Gender Egalitarianism Group Identification Models 112 Table 6.8. Structural Equation Coefficients for NES General Egalitarianism Group Membership Models 114 Table 6.9. Structural Equation Coefficients for NES General Egalitarianism Group Identification Models 118 vi Table 6.10. Structural Equation Coefficients for NES Economic Egalitarianism Group Membership Models 120 Table 6.11. Structural Equation Coefficients for NES Economic Egalitarianism Group Identification Models 122 Table B.1. Measurement Model Parameters for Youth 1997 Egalitarianism 209 Table B.2. Measurement Model Parameters for Parent’s Egalitarianism (in Youth 1997 Egalitarianism Model) 210 Table B.3. Measurement Model Parameters for Youth 1982 Egalitarianism 211 Table B.4. Measurement Model Parameters for Youth 1973 Egalitarianism 212 Table B.5. Measurement Model Parameters for Parents’ 1973 Egalitarianism 213 Table B.6. Measurement Model Parameters for Youth 1973 Political Knowledge 214 Table B.7. Measurement Model Parameters for Media Exposure: 1982 Youth 215 Table B.8. Measurement Model Estimates for Religiosity 215 Table B.9. Egalitarianism Measurement Model Parameters for NES Equal Opportunity 216 Table B.10. Measurement Model Parameters for NES Racial Egalitarianism 217 Table B.11. Measurement Model Parameters for NES Gender Egalitarianism 219 Table B.12. Measurement Model Parameters for NES General Egalitarianism 220 Table B.13. Measurement Model Parameters for NES Economic Egalitarianism 221 vii List of Figures Figure 1. AIC for Indirect Models 73 Figure 2. CAIC for Indirect Models 74 viii ABSTRACT This dissertation investigates the mechanisms involved in the formation, maintenance and change in the level of egalitarianism in the United States public during the second half of the 20 th Century. This is an important topic in the study of American politics because egalitarianism is a core value and core values structure the policy attitudes of a public that, according to research, for the most part, does not follow politics closely and is often ambivalent about the political questions of the day. The two theories tested are the elite leadership theory and group identification and interest theory. Elite leadership theory predicts that on types of egalitarianism about which elites agree such as racial and gender egalitarianism and equal opportunity, exposure to elite discourse as measured by education, media exposure and political knowledge should predict these types of egalitarianism and group identity and interest theory explains the level of egalitarianism on types of egalitarianism about which elites disagree, such as equality in general and economic egalitarianism. In order to balance broad coverage in time and quality and quantity of indicators, data are drawn from two sources: the Youth Parents Political Socialization Panel Study and the American National Election Study Cumulative File. Because egalitarianism is a latent variable, data are analyzed by structural equation modeling. I find that both elite discourse and group identities and interests predict both kinds of egalitarianism. However, the family is an intervening variable and is one of the two strongest influences on egalitarianism. The other is the cohort or age group. The baby boomers, who were socialized during the 1960s and early 1970s, are the most ix egalitarian followed by Generation X. The generation who grew up during the New Deal and World War II is the next most egalitarian. The Post-World War II-1950s generation and the pre-New Deal generation are the least egalitarian. I interpret the cohort effect as the interaction of age group and history. Implications are that because I find robust bottom up formation of all types of egalitarianism, the concern over elite manipulation of the mass public to support policies against its interests is exaggerated, except on the issue of economic interests. Here, the evidence for group interest is meager. However, because egalitarianism is so stable, this renders democratic deliberation difficult where value conflict is implicated, although often there are choices as to which values can frame issues. 1 Chapter 1 Introduction Equality is the biggest lie ever told--Friedrich Nietzsche (qtd in Spiegelberg 1944, 101) The first and most intense passion which is engendered by the equality of conditions is, I need hardly say, the love of that same equality...Everybody has remarked that in our time…this passion for equality is every day gaining ground in the human heart--Alexis de Tocqueville (1863, 114) The research questions addressed here are: (1) What sources influence public attitudes about equality and (2) how stable is the level and meaning of equality over time? In order to answer these questions, I examine the influence of parents, age, cohort, the media, religion and group identities and memberships and other variables. I examine these questions by looking at American public opinion data, mostly from 1984 to 2000. These questions are important in the study of American politics because egalitarianism is a core value and core values structure the policy attitudes of a public that, according to research, for the most part, does not follow politics closely and is often ambivalent about the political questions of the day (Campbell et. Al [1960] 1976, 543; Converse 1964, 242; Glynn et. al. 1999, 251, Zaller and Feldman 1992, 579). For those who do not have a large store of political information, core values, such as egalitarianism insure that everyone cannot be persuaded of everything (Zaller 1991, 1215). According to this line of research, policy preferences reflect core values (Brewer 2003a, 1214; Feldman 1988, 416; Feldman and Steenbergen 2001, 658; Kellstedt 2000, 245; Sears 2000, 107). 2 In this chapter I first review the literature that generates hypotheses on the meaning of equality. Then I examine the work that has sought to explain how egalitarianism is formed, maintained and changed and conclude that we know very little about the specific mechanisms which affect the support for equality beyond that when people are made aware of an inconsistency between one value and other values, predispositions and behaviors, they will tend to resolve the inconsistency and that although there is evidence that the level of egalitarianism did change during the period from the late 1960s to the early 1980s, little is know what has happened since then or in general whether egalitarianism changes during the relatively quiet times of normal politics . Then I discuss the work that has been done on how other factors, such as party identification and liberalism/conservatism, are formed maintained and changed and conclude that this literature generates fruitful hypotheses on the questions addressed here as well as two useful theoretical perspectives: the elite leadership/political culture perspective and the group identification/interest perspective. The Meaning of Equality Before the mechanisms through which egalitarianism is socialized are investigated, however, the concept must be defined (For a summary of various definitions in the literature and the associated authors, see Table 1.1). In empirical work, the value of work, the value of equality has sometimes been treated a priori as a unitary concept (Ball- Rokeach, Rokeach and Grube 1984; Bishop Barclay and Rokeach 1972; Rokeach 1968; 1974; Rokeach and Cochrane 1972; Rokeach and Ball-Rokeach 1989), but Feldman’s (1988, 421-424) empirical determination resulted in a two factor solution: equality of 3 opportunity and equality of capabilities. This disagreement over whether equality is uni- or multi-dimensional is also evidenced by the literature on the philosophical and practical essence of the concept and its historical meaning. One view holds that equality has Table 1.1: Conceptualizations of Equality Mentioned in the Literature Conceptualizations Definitions Associated Authors Unidimensional A terminal value (an end state). Moral equality, equality of political participation, equality of economic opportunity, equality of result, racial and sexual equality. (Ball-Rokeach, Rokeach and Grube 1984; Bishop Barclay and Rokeach 1972; Rokeach 1968; 1974; Rokeach and Cochrane 1972; Rokeach and Ball-Rokeach 1989) Redenius (1981) 4 Table 1.1: Continued Conceptualizations Definitions Associated Authors Multi-dimensional Equality of Opportunity Formal Equality, non- discrimination, “parity in competition”, a means to advancement rather than the end state. One competing ideal of equality. Everyone should have an equal chance to strive for success. Rewards should depend on skills and motivation. Equal opportunity to get ahead. “Equal opportunity to make oneself what one can, to develop talents. Ability and effort should be rewarded. Everyone should have an equal chance to get ahead. Feldman (1988) Verba and Orren (1985b) Hochschild (1995) Sears, et. al. (2000) Westie (1965) Verba and Orren (1985a) 5 Table 1.1: Continued Conceptualizations Definitions Associated Authors Equality of Results The other competing ideal of equality. Differences in character, intelligence and knowledge should be ignored “in order to serve the well-being of a common humanity. Equality of condition, the reduction of all equality. Equality of result in income, status and power. Verba and Orren (1985b) McCloskey and Zaller (1984) Verba and Orren (1985a) Equal worth and dignity. Everyone is in some fundamental sense alike, no one has intrinsically greater worth; everyone should be prized as a member of a common humanity. Everyone should be judged according to individual worth. Equal intrinsic worth. All people are of equal worth. All people are of equal intrinsic worth, no one is intrinsically superior to anyone else, everyone’s interests should be given equal consideration. McCloskey and Zaller (1984). Westie (1965) Phillips (1999) Verba and Orren (1985a) Dahl (2006) 6 Table 1.1: Continued Conceptualizations Definitions Associated Authors Equality of Capabilities Everyone should have an equal say in running the country. People vary in skills and motivation. No persons are so much more qualified to run the country than everyone else that they should have absolute power. Feldman (1988) Sears, et. al. (2000) Dahl (2006) Individual Equality A competing unit of equality. Verba and Orren (1985b) Group Equality The other competing unit of equality. Equality for women, ethnic minorities, the handicapped and any other group who believe they are getting a “raw deal.” Removal of barriers to fair competition and compensation for past discrimination. Verba and Orren (1985b) Verba and Orren (1985a) Egalitarian Ideal One competing standard of judgment for equality. Whether people should be regarded as equal. “Nebulous, generalized ideal.” Verba and Orren (1985b) Verba and Orren (1985a) Redenius (1981) 7 Table 1.1 Continued Conceptualizations Definitions Associated Authors Equality of Process, Formal/Political/Legal Equality One competing domain of equality; differing levels of political influence. Equal treatment by the courts, equal right to vote and run for office, some economic rights such as the right to enter into contracts. Right to choose and hold accountable officeholders, “universal and inalienable nature of human rights, including minority rights.” Equal before the law and Constitution. Everyone treated equal in the eyes of the law, equal right to hold public office. Equal right to participate in politics. One person, one vote, “condemns gross disparity of power among individuals and groups. All people have identical rights. Verba and Orren (1985b) Sears, et. al. (2000) McCloskey and Zaller (1984). Westie (1965) Phillips (1999) Verba and Orren (1985a) Verba and Orren (1985a) 8 Table 1.1: Continued Conceptualizations Definitions Associated Authors Equal Treatment People should not be treated unequally because of group membership. Sears, et. al. (2000) Domain Specific Equality Economic Equality Income distributions flattens, floor under the poorest is raised The other competing domain of equality; a more equal distribution. Insuring everyone has the material necessities for a decent life. Refers to economic opportunities and life chances, patterns of social stratification. Outcomes ought to be somewhat equalized, should not depend so heavily on ability, opportunity or social origin. Equality in wealth, income and life chances, includes access to education and health. Equal opportunity with a floor. Social and economic citizenship, the right to equal treatment in economic matters. Lane (1959) Verba and Orren (1985b) McCloskey and Zaller (1984) Phillips (1999) Verba and Orren (1985a) 9 Table 1.1: Continued Conceptualizations Definitions Associated Authors Gender Equality Not defined. Attitudinal (sex roles, abortion, equal capabilities) and Behavioral (participating in groups or activities promoting sexual equality). Equal opportunities. Christenson and Dunlap (1984) Ball-Rokeach (1976) McCloskey and Zaller (1984) Racial Equality Not defined. Equal opportunities, integration, affirmative action. Christenson and Dunlap (1984) McCloskey and Zaller (1984) Gay Rights Equal job opportunities. McCloskey and Zaller (1984) many meanings, some of which contradict others (Ellis 1993; Pole [1978] 1993, 3; Rae et. al. 1981; Redenius 1981; Verba and Orren 1985a, 1; 1985b, 374). More specifically, these meanings, and the tensions between them, have been described as individual versus group equality, the egalitarian ideal versus the reality and process versus results, political (or formal/legal) versus economic equality (Ellis 1993, 43; Feldman and Zaller 1992, 272; Redenius 1981; Sowell 2002, ch. 6; Verba and Orren 1985b, 374). The analogy that has been made about the relationship between the various types of equality is that of a Rubik’s cube, with each face representing one type of equality (Verba and Orren 1985b, 374). According to this analogy, when one face is manipulated to achieve the desired result, the other faces are thrown into disarray and must likewise be adjusted (Verba and 10 Orren 1985b, 374). To the contrary, Phillips (1999, 19, 130-131) asserts that equalities are interrelated, for example, economic inequality undermines political equality. Not only is the meaning of equality contested in the case of the aggregate public, there is also some reason to believe that it is different among different groups and may change over time (Bell 1979, 27-30; Edsall 1984, 13; Fischer et. al. 1995, 3; Ladd and Bowman 1998, 1-2; Lee 2002, 132-136; McClosky and Zaller 1984, 86-90, 94-95; Phillips 1991, 1; Pole [1978] 1993, 411-421; Rorty 1998, 83; Schaefer 1979, 3; Verba and Orren 1985b, 369-371). 1 Egalitarianism Across Time For core values to stabilize policy preferences, core values themselves must be stable. Furthermore, if these values are to constrain elite manipulation or leadership of the public, it makes a difference how these values are formed. There is a paucity of work on the stability of egalitarianism and it extends only to the early 1980s. Egalitarianism was found to have increased during the late 1960s by Rokeach (1974, 227) but declined during the decade of the 1970s by Rokeach and Ball-Rokeach (1989, 783), but the reasons for these changes were not empirically investigated. The Formation of Egalitarianism Little scholarly work has addressed the mechanisms by which equality is socialized. The choice of a presidential candidate was found by McCann (1997, 564) to affect level of egalitarianism, but other mechanisms such as political communication, the family and other group identities and interests were not investigated. Ball-Rokeach, Rokeach and Grube (1984, 74-75) found subjects' rating of equality relative to other 1 For an alternative argument to Lee’s (2002, 132-136) “separate grooves” thesis, see Page and Shapiro’s (1992, 178-179) “parallel publics” theory. 11 values, particularly liberty, increased after watching a television program in which it was stated that those who felt sad, angry or ashamed about the assassination of Dr. Martin Luther King rated equality higher than those whose reaction was that “he brought it on himself.” When Rokeach (1968, 27) framed the choice between equality and freedom as in the case of the former desiring their own as well as the freedom of others, and in the case of the latter of desiring their freedom but not the freedom of others, ratings of equality among experimental subjects increased. Ratings of equality were also increased among those who were pro-civil rights but ranked equality low and were presented with a similar frame as above (Rokeach 1968, 30). However, the effects of other arguments or of all arguments made in the mass media were not addressed. One generalization that might be drawn from the above studies is that people will attempt to maintain consistency between levels of egalitarianism and other predispositions and preferences, particularly when inconsistencies are made salient (Rokeach 1968-1969, 552-553). However this line of research treats the original level of egalitarianism as a given and does not investigate how it arose. It also investigates only one type of argument and does not answer how media discourse in general affects how much people value equality. In order to fill this gap in the literature here I examine how the mass media, the family, social networks, schools, group identities and interests, peers and historical events shape egalitarianism. Below I review the relevant literature that addresses the formation of other core values and issues that are related to egalitarianism. 12 Although the role of the family in socializing egalitarianism has never been investigated, the role of the former in socializing other political predispositions is varied and depends on which predisposition is examined (Connell 1972; Davies 1965; Hess and Torney 1967; Hyman [1959] 1969; Jennings and Niemi 1968; 1974; Langton 1969; Levin 1961; Sigel 1970). According to Jennings, Stoker and Bowers (2001, 1), “the more concrete, affect-laden and central the object, the more successful was the transmission. More abstract, ephemeral, and historically conditioned attributes were much less successfully passed on.” The strongest relationship between parents and their children has been found in the case of party identification (Hyman [1959] 1969, ch. 4; Jennings and Niemi 1974, 39; 1968, 172; Levin 1961, 596-598; Tedin 1974, 1584). Not surprisingly, this relationship is stronger where children correctly perceive parents’ party identification and both parents identify with the same party (Jennings and Niemi 1974, 39; 1968, 172; Tedin 1974, 1586). Although there is disagreement in the literature on this question, substantial research supports the view that this is not a spurious relationship due to the common demographic characteristics of the children and parents (Dalton 1982, 139; Glass, Bengtson and Dunham 1986, 685; Jennings 1984, 261, 269,273; Levin 1961, 597; Tedin 1974, 1580; 1980, 142; Westholm and Niemi 1992, 25, 35). As is the case regarding most predispositions, the strength of the relationships declines over time (Jennings and Niemi 1981, 100, 113). However, the relationship between the parents and children’s perceptions of party differences was found to be very weak (Jennings and Niemi 1974, 105). A relationship of similar magnitude to that of party identification was found 13 between presidential candidate preferences of children and parents (Jennings and Niemi 1974, 54). The relationship between parents’ and children’s’ liberal-conservative political ideology is also positive, but weaker than in the case of party identification (Braungart 1971, 127; Glass, Bengtson and Dunham 1986, 692, 694; Greeley 1975, 193- 194). There is some evidence that one of the mechanisms through which this relationship is sustained is family argumentation. Braungart (1971, 127) found that frequency of political arguments between parents and children over civil rights, the distribution of wealth and American foreign policy was positively related to left-wing political group membership and identification. The family is significantly related to one’s placement on a humanism/materialism values dimension, but it is a much weaker agent in pro- and anti-establishment attitudes and postmaterialism than in the case of ideology (Bengtson 1975, 367; Jennings 1984, 261,267). The findings of past research on policy attitudes depend on the issue area and whether an attempt was made to purge the dependent variables of measurement error. Without correction for measurement error, past research indicates a weak relationship between parents and children’s support for civil liberties and a moderate relationship for racial issues (Acock and Bengtson 1980, 501, 506; Jennings and Niemi 1974, 78-79; 1968, 175; Tedin 1980, 1584). Studies that corrected for measurement error replicated the moderate relationship found above for racial issues and also found a moderate relationship on the issue of civil liberties and a strong relationship in the case of 14 gay rights but a low to moderate relationship on the question of government guaranteed jobs (Dalton 1980, 425, 427; Jennings, Stoker and Bowers 2001, 35). Regardless of the predisposition being examined, there are certain conclusions from the literature on when parental influence is likely to be significant. Jennings, Stoker and Bowers (2001, 14-19) found support for the notion that parental effects are likely to be greatest when the family is highly politicized and provides consistent cues over time. Tedin (1974, 1587) found that it is the salience of the issue to the parents rather than general politicization that enhances parental influence. The general role of education in the political socialization process has proved durable but the specific mechanism through which it works has proved elusive. The level of education is positively related to the acceptance of the liberal democratic creed and general political ideology (Glass, Bengtson and Dunham 1986, 692, 694; Hyman and Wright 1979; McClosky and Zaller 1984: McClosky, Zaller and Chang 1984; Nunn, Crocket and Williams 1978; Stouffer 1963 [1955]). More specifically, past research by Hyman and Wright (1979, 28, 54) and Smith (1981, 1265-66; 1982, 13, 16) found a positive relationship to support for civil liberties, equal opportunities for Catholics, Jews and African Americans and Campbell (1971) found a similar relationship in the case of racial egalitarianism, in general. There is substantial evidence of a positive relationship between education and support for the civil liberties of gays and egalitarian sex role attitudes (Loftus 2001,772; Mason and Lu 1988, 50; Smith and Fisher 1982, 82) Hasenfeld and Rafferty (1989, 1038-1039) found a negative relationship between level of education and support for the welfare state (though the relationship attained statistical 15 significance in only one of two models). However, no relationship was found between taking civics courses and political orientations, except in the case of African Americans, particularly those from less educated families (Jennings, Niemi and Langton 1974, 205; Langton 1969, 115, 117). Likewise, the effects of the predispositions of social studies teachers on those of their students were dwarfed by that of the parents, but homogeneity with the parents pushes the students in the direction of the parents and teachers, while students adopt an intermediate position between the parents and teachers, where they are in conflict. Demographic variables such as age, social class, race, gender, religion and income that have proven explanatory power for political predispositions that are like related to egalitarianism (Braungart 1971, 127; Hasenfeld and Rafferty 1989, 1038-1039; Hyman and Wright 1979, 33; Lee 2002 50, 58-59; Zaller 1992, 184). All might be interpreted as group consciousness or common socialization experiences or both. Age has been found to be a negative predictor towards support for civil liberties, authoritarianism and the welfare state and racial attitudes, though these relationships are likely the effect of the different socialization experiences of cohorts, and possibly period effects, rather than the inevitable effect of the life cycle (Condran 1979, 467, 474; Gilliam and Whitby 1989, 95; Hasenfeld and Rafferty 1989, 1038-1039; Hyman and Lee 2002, 58-59; Wright 1979, 33; Youmans 1973, 56; Zaller 1992, 184). 2 Bengtson (1975, 367) found that the generation to which one belongs was significantly related to one’s placement on a humanism/materialism values dimension. One possible mechanism through which these cohort and period effects might work is through the impact of formative events (Beck and 2 On the other hand, Bengtson Glass, and Dunham (1986, 692) found age was not significantly related to political, gender and religious ideology. 16 Jennings 1991; Sears and Valentine 1997). Markus (1979b, 338) found a combination of these effects in the case of the relationship on racial matters on partisan orientations, a relationship that he found to be stronger in the case of young adults. Not surprisingly, the social science literature supports the proposition that social class is related to ideological orientation (Braungart 1971, 127). Being minority, female, low income or having a lower level of education is a positive predictor of support for the welfare state (Hasenfeld and Rafferty 1989, 1038-1039). Past work also suggests that the highly educated view immigrants more favorably than the less educated (Fetzer 2000, 96, 100-101). In addition, those with low incomes are significantly more supportive of fair housing and employment but not school desegregation; the pattern of significance for education was reversed on these two issues (Braungart 1971, 127; Hasenfeld and Rafferty 1989, 1038-1039; Lee 2002, 58-59). In contrast, income is positively related to support for affirmative action (Jacobson 1985, 322). Other work finds that economic interests as measured by economic expectations are positively related to Democratic macropartisanship and policy liberalism (Durr 1993, 166; Erikson et. al. 2002, 135; Wleizien 1995, 992). 3 Findings on the relationship of religion to egalitarianism are mixed. The most unequivocal finding from past research is that the more religious tend to be more Republican and conservative. 4 In addition, Feldman (1988, 426) found religious 3 To the contrary, Gilens (1999, 46-47) found economic hardship made people more supportive of welfare spending. 4 See Green and Guth (1991, 212, 222); Kamieniecki (1985, 56); Kaufman (2004, 502); Kaufman and Petrocik (1999, 870); Knoke (1976, 29-35); Kohut et al. (2000, 74-77); Jelen and Smidt (1993, 207- 208); Layman (1997, 298); Lopatto (1985, 48-49); Miller and Wattenberg (1984, 302, 309). 17 attendance positively related to support for equal opportunity. Fundamentalist Protestants are less likely to support gay rights than other religions or those with no religion (Loftus’ 2001, 773) and Protestants in general are less likely to support civil liberties for gays than other religious groups or the non-religious. On the other hand, Rokeach (1969, 16) found no significant relationship between church attendance and importance of religion and ranking of equality. The view taken by the literature on the power of the mass media as a socialization agent has varied at different times. An older body of work characterized the extent of this influence as “minimal effects” (Berelson, Lazarsfeld and McPhee 1954, Dawson, Prewitt and Dawson 1977; Klapper 1960, Krause and Davis 1976, Lazarsfeld, Berelson and Gaudet 1965 [1948]). However, later work has also found that the news media can influence what people think as well as what they think about (Bartels 1993; Goidel, Shields and Peffley 1997; Kellstedt 2000; Nelson, Clawson and Oxley 1997; Page, Shapiro and Dempsey 1987; Price, Tewkesbury and Powers 1997). Rather than the old model of the media as a “‘hypodermic needle’ that could inject ideas into the body politic on command,” (Zaller 1992, 311) the newer body of research has much more sophisticated research designs, which examine the cumulative effects of the news media and control for resistance and opinion reinforcement (Fan 1988, 132; Nelson, Clawson and Oxley 1997, 567; Page, Shapiro and Dempsey 1987, 23). Published results at the aggregate level have tended to be strong, but one cannot make individual level inferences from these results (Brody 1991; Fan 1988; Fan and Tims 1989; Kellstedt 2000; Nadeau, et. al. 1999; Shah, et. al. 1999, but see Page, 18 Shapiro and Dempsey 1987 for more mixed results). Experimental results have also tended to be strong, but one cannot be sure that these results are generalizable outside the laboratory or if the experimental treatments are even representative of media content, unless parallel content analysis is also done (Ball-Rokeach, Rokeach and Grube 1984; Price, Tewksbury and Powers 1997; Neuman, Just and Crigler 1992; Nelson, Clawson and Oxley 1997; Terkildsen and Schnell 1997). Results at the individual level are mixed, but political awareness tends to perform better than direct measures of media exposure (Bartels 1993; Beck, et. al. 2002; Berelson, Lazarsfeld and McPhee 1954; Converse 1962; Goidel, Shields and Peffley 1997; McCloskey and Zaller 1984; Zaller 1991; 1992). At the aggregate level, Kellstedt (2000) found that egalitarian media framing was a significant cause of racial policy sentiment. On a wide variety of policy issues, Page, Shapiro and Dempsey (1987, 26, 28) found a correctly signed but insignificant relationship between change in news coverage of various foreign, domestic and defense issues and public preferences on such issues. However, they also investigated the possibility that different sources have differential impacts on the public and found that only such neutral sources as experts, editorial commentary and courts, who are perceived as less political than elected officials, have a significant effect (Page, Shapiro and Dempsey 1987, 30). In experimental work by Price, Tewksbury and Powers (1997, 491), frames were found to impact reported thoughts and emotional valence on the issues framed, while Nelson, Clawson and Oxley (1997, 567) found the experimental frames of public order and free speech affected tolerance for the Ku Klux Klan in the expected directions. 19 Terkildsen and Schnell (1997, 879) found a negative impact of economic equality, feminism and anti-feminism frames, but a positive impact of the political rights frame on subjects’ orientations toward gender equality, women’s rights, gender roles and the importance attached to “women’s issues.” Studies of media influence at the individual level have typically used three different measures, direct measures of content, political awareness and media exposure and analyzed two types of issues, consensus and contested (among political elites) (Bartels 1993; Beck, et. al. 2002; Berelson, Lazarsfeld and McPhee 1954; Converse 1962; Goidel, Shields and Peffley 1997; McCloskey and Zaller 1984; Zaller 1991; 1992). The first type of study is represented in the work of Beck, et. al. (2002, 65), who found that newspaper editorials were a significant predictor of the vote, but newspaper and television content, in general, was insignificant. Studies of the second type have met with mixed success. On the matter of liberal/conservative ideology, Bartels (1993, 278) failed to find any significant results for television exposure and mixed results for exposure to newspapers. Entman and Rojecki (2000, 233) found reliance on television for news was positively related to believing that individual poor Blacks were responsible for their poverty. Media exposure, though a straightforward measure, is not an unproblematic one. It has been criticized as being fraught with measurement error and social desirability bias and simply not performing empirically as well as political awareness (Zaller 1992, 21-22, 42-43, 292-293, 333-340). 5 McCloskey and Zaller (1984) examine various attitudes toward equality among community influentials and the mass public and find no 5 Bartels’ (1993) media exposure variable was corrected for measurement error. 20 differences in support among those aspects of equality over which political elites disagree (contested values) such as affirmative action and school busing. Elites agree on other aspects of equality such as equal opportunity, gender and racial equality (McCloskey and Zaller 1984, 80-97). On these consensus matters, community influentials are more supportive than the public in general (McCloskey and Zaller 1984, 90). In analyzing both mainstream and partisan issues, Zaller (1992, 111) found significant effects for political awareness on both types of issues. 6 An interaction between political awareness and ideology and political awareness and party identification was also posited by Zaller (1992, 111) and found both in the case of school busing and (surprising given Zaller’s categorization of it as a mainstream issue) equal accommodations. Past research has found a positive relationship between political sophistication and cultural and civil liberties liberalism but a negative one between the latter and economic liberalism (McCloskey 1958, 35; Neuman 1986, 81). However, these studies did not control for level of political discussion and the findings of a non- significant relationship between media exposure, but a positive and significant one for political discussion by Beck, et. al. (2002, 65) raises the possibility that the media relationships are spurious. However, since discussion networks tend to be homogeneous, merely cross-sectional evidence such as that discussed above is somewhat ambiguous on the question of causation (Berelson, Lazarsfeld and McPhee 1954, 116; Mutz and Martin 2001, 100-101). 6 Zaller’s (1992, 111) partisan issues were government job guarantees, the Vietnam War, and school busing, while the mainstream issues he examined were equal accommodations for all races, women’s rights, admission of China to the United Nations, trading with Communist-ruled nations and anti-isolationism. Mainstream and partisan issues are synonymous with consensus and contested issues. 21 It has been established in this chapter that there is a gap in the literature on the specific mechanisms by which egalitarianism is socialized and that the literature generates the hypotheses that the news media, group identification and interest, social context and other institutions should affect levels of egalitarianism. The next chapter sets forth the hypotheses to be tested by this research. Chapter 3 is devoted to methodology. Chapter 4 considers the meaning of equality for the United States public, as a whole, and various politically salient subgroups and whether the meaning of equality has changed over time. Chapter 5 begins the task of explaining and predicting the level of egalitarianism by examining the role of the family, the news media and group identification and interest. Chapter 6 adds age and cohort to the analysis. Chapter 7 concludes. 22 Chapter 2 Hypotheses The Meaning of Equality Dimensionality To quantitatively analyze a concept, it must first be measured. In order to measure it, it is necessary to empirically determine if it is really one concept or more than one concept, the question of dimensionality must be addressed. 1 It is common in social science research to use either a ranking procedure of a unidimensional notion of equality or to create a unidimensional scale of six statements in the agree-disagree format from the American National Election Study (Ball-Rokeach, Rokeach and Grube 1984, 41; Brewer 2003b, 198; Federico and Sidanius 2002, 174; Gilens 1995, 1004; Moore 1975, 405-406; Rokeach 1968, 23; 1973, 26-28; 1989, 776-778). 2 This work suggests: H1: Egalitarianism is unidimensional. On the other hand, historical and other work in political science also points to a multidimensional explanation (Kluegel and Smith 1986, 291-293; Pole [1978] 1993, 3; Redenius 1981, 4; Verba and Orren 1985b, 374). The alternative hypothesis suggested by this work is: H1A: Egalitarianism is multi-dimensional. 1 Since I was unable to identify an equation with all egalitarianism variables present, it is impossible to investigate this question directly. 2 However, Feldman (1988, 421-425) utilized the six items that are the precursor to the present NES scale to create a scale for equality of opportunity and Federico and Sidanius (2002, 1620 and Feldman and Steenbergen (2001, 666) utilized variables both for general egalitarianism and equality of opportunity. 23 Equality of Opportunity If egalitarianism is multidimensional, the literature clearly supports the hypothesis that equal opportunity would constitute one such dimension (McClosky and Zaller 1984, 80-86; Pole [1978] 1993, ch. 6; Redenius 1981, 4; Verba and Orren 1985a, 1, 5-20; 1985b, 374). 3 H2: Equal opportunity is one meaning of equality. Domain Specific Egalitarianism Some work in History and Political Science includes such domain specific types of egalitarianism as economic, race and gender but the existence of these types of egalitarianism has not been directly investigated quantitatively (Redenius 1981, 4; Verba and Orren 1985b, 374). Feldman and Zaller (1992, 285, 289) found little evidence of economic egalitarianism and Gilens (1995, 1008) found that egalitarianism was not significantly related to white Americans’ support for welfare spending. Over the past two decades or so, the question of economic inequality has been described as being off the agenda and the political power of the poor and working class has been considered to be in relative decline to the wealthy (Aronowitz 1992, xi, 9, 12, 47; Edsall 1984, 13; Edsall and Edsall 1991, 1-4; Elder 1993, 1, 5; Ladd 1970, 269; Ladd and Hadley 1978 [1975], 217-229; Rorty 1998, 83). Inglehart (1997, ch. 5) asserts there has been a shift from materialist to post-materialist values. In addition, class cleavages have been described as weakening and ideological cleavages have been described as being organized around social rather than economic issues (Clark 1998, 3; Clark and Inglehart 1998, 10-13; Dalton, Kuechler 3 The literature would also support the proposition that equal rights would constitute an important meaning for equality, but there were insufficient indicators to test this hypothesis. 24 and Bürklin 1990, 12; Elder 1993, 1; Ladd 1970, 269; Ladd and Hadley 1978 [1975], 217-229;) The implication of this work (the decline of class thesis) is that economic egalitarianism would be virtually non-existent. H3: There is no such thing as economic egalitarianism. 4 However, McCarty, Poole and Rosenthal (1997, 1: 2003, 4) take a diametrically opposed view—that over the past 30 years or so, American politics has become increasingly polarized around the redistribution of income (the “resurgence of class” thesis). Feldman and Steenbergen (2001, 667,669) revise the lack of economic egalitarianism thesis by arguing that egalitarianism does matter in the case of policies that are structured to include redistributive elements and are based on social rights rather than being means tested. H3A: Economic egalitarianism is one meaning of equality. For those who accept the decline of class thesis, class politics has been supplanted by identity politics. 5 Identities, such as those based on race, gender, ethnicity and sexual orientation, competing with class identities, and institutionalized by new social movements, have been viewed as severely undercutting class identity (Aronowitz 1992, 26-27, 29, 31; Dean 1996, 48; Elder 1993, 3, 5; Huckfeldt and Kohfeld 1989, 182; Kelly 2001, 5; Johnson, Laraña and Gusfield 1994, 6). H4: Racial egalitarianism is one meaning of equality. 4 This is an empirical, not a metaphysical claim and, obviously, it only applies to the United States during the time period of this study. Operationally, it means that the measurement model for economic egalitarianism should be rejected. 5 I use identity politics in the historical rather than the theoretical sense. A working class or blue- collar identity could be the most politically salient identity for individuals, but historically the term arose with and is customarily applied to the newer non-class political movements. 25 H5: Gender egalitarianism is one meaning of equality. H6: Gay rights are one meaning of equality H7: Immigrant rights are one meaning of equality. Change in Egalitarianism over Time Next I offer hypotheses on whether and how the meaning of equality has changed over time. The identity politics thesis suggests the following hypothesis on the change of the meaning of egalitarianism over time: H8: Identity politics based types of egalitarianism should exhibit increasing levels and tighter constraint over time. The resurgence of class thesis likewise generates an alternative hypothesis on the change of the meaning of equality over time. H8A: Economic egalitarianism should exhibit increasing levels and tighter constraint over time. Work on economic expectations, party identification and policy liberalism (Durr 1993, Erikson et. al. 2002, Wleizien 1995) suggests: H9A: Economic egalitarianism will increase in good economic times and decrease in hard economic times (the capability hypothesis). To the contrary, Gilens (1999, 46-47) found that Americans were less opposed to welfare spending during recessions than during economic recoveries and expansions, suggesting: H8B: Economic egalitarianism will increase in hard economic times and decrease in good ones (the need hypothesis). 26 Phillips (1999, 74-75, 98) argues that different types of equality are related, which generates the following prediction: H10: There should be no significant differences in levels or constraint of the different types of egalitarianism. Equality of opportunity is a perennial favorite in the American political culture and should be relatively stable (Feldman 1988, 419, McCloskey and Zaller 1984, 80-81). H11: There should no be significant change in level or constraint of equality of opportunity. Support for all three possible movements in support for general egalitarianism can be found in the literature. Since there is no basis in logic or in fact for choosing among the three possible directions other than mere speculation, I entertain alternative hypotheses. The most relevant, though oldest finding is that egalitarianism increased from the late 1960s to early 1970s but declined from then to the early 1980s (Rokeach 1974, 227; Rokeach and Ball-Rokeach 1989, 783). Buttressing this prediction is the increasing popularity of the self-label of conservative, which is negatively related to egalitarianism (Ball-Rokeach, Rokeach and Grube 1984, 74-75; Christenson and Dunlap 1984, 864; Ellis 1993, 59-64; Glenn 1974, 179; Lipset 1979, xxiii-xxiv; McCloskey and Zaller 1984, 62, 90; Pole [1978] 1993, 411-421; Rokeach 1968, 27, 30; Schaefer 1979, 3). The assumption that this trend would continue throughout the 1980s and 1990s supports the following prediction: H12: General egalitarianism will decline over time. 27 Findings by other work of no significant trend on policy liberalism by Erikson et. al. (2002, 223, 225), Mayer (1992, 119-121) and Stimson (1999, 68, 139-140) generates the first of two alternative hypotheses: H12A: General egalitarianism will exhibit no significant trend in movement. Finally, work finding an increase in liberalism from World War II to the late- 1980s, particularly on issues of equal rights, supports the final possibility (Davis (1992, 261; Smith 1990, 479). H12B: General egalitarianism will increase over time. Predictors of Egalitarianism Next, I introduce hypotheses on which variables will predict egalitarianism. Here I follow the distinction between contested and consensus values and all its implications (McCloskey and Zaller 1984, 235-244; Sears et. al. 2000, 83-84; Zaller 1992, 98-102). 6 Consensus values are those about which elites agree and contested values are ones about which elites disagree. A one-sided information flow should be present in the case of consensus values while a two-sided information flow should be present in the case of contested values. In general, exposure to elite discourse should predict consensus values and group identification and interest should predict contested values. 7 Consensus Values 6 McCloskey and Zaller (1984, 235-237) refer to this distinction as clear versus contested values, while Zaller (1992, 98-102) describes it as a mainstream versus polarizing effect. 7 In cases where the two theories offer conflicting expectations I entertain alternative hypotheses. 28 The next task is to categorize the various types of egalitarianism. The literature is virtually unanimous in characterizing equal opportunity as a consensus value (McCloskey and Zaller 1984, 318; Sears, Henry and Kosterman 2000, 91-95). 8 Past research also strongly supports characterizing support for racial and gender equality, in the abstract, as consensus values (Brink and Harris [1966] 1967, 20; Jackman 1978, 303; 1981, 265; Kluegel and Smith 1983, 804; Knapp and Alston 1972-1973, 589; Smith 1982b, 374; White 1981, 562; Zaller 1992, 111). In contrast, economic equality is clearly controversial among elites and imbued with partisanship and ideology, which places it in the contested category (Feldman and Zaller 1992, 174-186, 292; Ladd and Bowman 1998, 104-105; McCloskey and Zaller 1984, 4). Equality, in general, might be a symbolic predisposition or a weighted sum of specific considerations, some consensus and some contested. Past work that has treated egalitarianism as a unitary concept has found that it is not exceedingly popular compared to other values, particularly freedom, with which it often conflicts (Rokeach and Ball-Rokeach 1989, 778). 9 Work which has treated egalitarianism as the sum of its components has found some, such as equal treatment and equal opportunity as consensus values, and others such as whether inequality is a problem in the U. S. or whether we worry too much about equality as contested (McCloskey and Zaller 1984, 318; Sears, Henry and Kosterman 2000, 91-95). Even if only some components are contested, that would mean the better practice would be to treat general egalitarianism as a contested value. 8 This is true with respect to elite discourse. In the mass public, Federico and Sidanius (2002, 164-165) found equality of opportunity was only a consensus value among liberals and moderates, but not conservatives. 29 Elite Leadership/Political Culture Hypotheses The elite/political culture theory of values predicts that consensus values should be equally distributed across all groups except high and low media exposure, political awareness, political discussion and education (McCloskey and Zaller 1984, ch 8). 10 H13: There will be a positive relationship between political communication 11 and support for equal opportunity, racial and gender egalitarianism. Group Identification and Interest Hypotheses Next, I present hypotheses suggested by group interest 12 and identification. Group interest and identification theory predicts that there will be differences in egalitarianism across groups defined other than by the level of political communication and education (Blumer 1958, 3; Bobo 1988, 91; Bobo 1998, 991-995; Bobo 2000, 139, 142, 143, 160 Campbell, et. al. [1960] 1980 121-122; Green, Palmquist and Shickler 2002, ix, 2, 4-7). Since women, people of color, the poor and working people have been denied opportunities offered to more favored groups, group interest and identity suggests the following hypotheses: H14: People who are not white will support equality of opportunity more strongly than whites. H15: Women will be more strongly supportive of equality of opportunity than men. 9 Equality was more popular in the late 1960s and early 1970s than the last measurements in the mid-1970s and early 1980s, though it never ranked as high as freedom (Christenson and Dunlap 1984, 864; Rokeach and Ball-Rokeach 1989, 778). 10 According to this theory, political culture is the cause of value choices, which are articulated and reinforced by elites and exposure to their discourse is merely an indicator (McCloskey and Zaller 1984). 11 I use political communication as a label for the variables media exposure, political awareness and political discussion. 30 H16: There will be a negative relationship between income and support for equality of opportunity. H17: Union households will be more supportive of equality of opportunity than non- union households. The most obvious predictions of group interest and identity are: H18: Non-whites will be more racially egalitarian than whites (Abramson 1975, 76, 87; Bobo 1988, 88-90, 91-98; Knoke 1979, 784). H19: Women will be more gender egalitarian than men (Mason and Lu 1988, 48). One pattern in the findings of past research might be characterized as group empathy. Traditionally disadvantaged groups tend to be more supportive of equality for other groups. This principle suggests the following hypothesis: H20: Women should be more racially egalitarian than men (Kluegel and Smith 1983, 810; Schuman, et. al. 1997, 235; Tuch and Hughes 1996, 735-736). H21: People who are not white should be more gender egalitarian than those who are white (Huber and Spitze 1980, 327; 1981, 163; Mason and Bumpass 1975, 1217; Mason and Lu 1988, 51) Another twist on this corollary to group interest and identification is where equality for one group poses no threat to the interests of another group; the second group is supportive of equality for the first group. This pattern specifically emerges in the case of income (Greeley and Sheatsley 1971, 16, 19; Huber and Spitze 1981, 163; Ladd 1976- 1977, 590, 595; Mason and Bumpass 1975, 1217; Peek and Brown 1980, 176, Tuch 12 I use group interest as synonymous with group self interest (Bobo and Kluegel 1993, 443) and realistic group conflict (Bobo 1998, 91-92). 31 1984, 66). Racial and gender equality poses less of a threat to white men at the top rung of the economic ladder, since non-whites and women are disproportionately poor, the economic competition between races and genders is most acute at the bottom. H22: There should be a positive relationship between income and racial egalitarianism and gender egalitarianism. The converse of this pattern also finds support in the literature. Some work describes the white working class, to which most union members belong, as often being in conflict with blacks and as being less supportive of civil rights proposals than professionals (Ladd 1970, 251; 1976-1977, 590, 595; Ladd with Hadley [1978] 1975, 230-231). This work suggests: H23: Union households will be less supportive of racial equality than non-union households. Past empirical analyses suggests that two other group memberships should predict racial and gender egalitarianism: age and cohort. A body of scholarship has racial and gender liberalism increasing in post World War II cohorts but leveling off or decreasing in cohorts coming of age since the 1970s. 13 This pattern offers support for the following hypothesis: H24: Racial and gender egalitarianism should be greater in middle cohorts than in younger and older ones. 13 See Davis (1992, 283, 285, 294); Firebaugh and Davis (1988, 251, 260-263, 267); Kiecolt (1988, 395); Mason and Lu 1988, 47; Steeh and Schuman (1992, 340, 350-351, 359); Schuman, et. al. (1997, 210-211); Smith (1981a, 1256, 1265-1266; 1982, 13-15; 1985, 205, 217, 220-221) and Wilson (1996, 253, 269-270). Smith (1982b, 377, 379) finds no cohort differences on the issue of residential integration. An earlier study by the same author finds cohort differences only among the Vietnam generation (Smith 1981b, 564-567). 32 The received wisdom also says the young are more racial and gender liberal than the old. 14 This creates the following hypothesis: H25: There will be a negative relationship between age and racial and gender egalitarianism. Contested Values If elite discourse reflects the political culture, exposure to elite discourse should not predict values about which elites disagree. On the other hand, there is a significant support for the proposition that group identification and interest predicts contested values. 15 If group identification and interests are relatively stable, values should also be relatively stable. H26: There will be a positive relationship between past and present egalitarianism. 16 The family is a group past research often finds to be an important factor in socializing values. 17 14 See Abramson (1975, 78); Bobo (1983, 1207); Bobo and Kluegel (1993, 452); Cherlin and Walters 1981, 456; Feree 1974, 394; Greeley and Sheatsley (1971, 16, 19); Hamilton (1972, 410); Huber and Spitze 1980, 327; 1981, 163; Jacobson (1985, 319); Mason and Lu 1988, 49; Maykovich (1975, 1014); Ransford [1977] 1994, 153; Ransford and Miller 1983, 52; Sears (1983, 99-100; 1988, 54); Sigelman and Welch 1982, 312; Smith (1981b, 558, 564-567); Tuch (1984, 56, 64-66; 1987, 506) and Tuch and Hughes (1996, 735-736). Kluegel and Smith (1983, 810) find no age differences on support for affirmative action and Bobo (1983, 1206) none for busing. Greeley and Sheatsley (1974, 243) found the young to be more racially tolerant, but that the middle aged were catching up. 15 See Campbell, Gurin and Miller (1954, 111); Campbell, et. al. ([1960] 1980 121-122); Knoke (1976, 3- 6); Knoke and Hout (1974, 700, 704, 711); Green and Palmquist (1990, 872-877, 880-881, 898; 1994, 437- 439, 441, 444-445, 447, 457); Green, Palmquist and Shickler (2002, ix, 2, 4-7). 16 The same prediction would be generated if values are influenced by information flow and information flow is constant (Achen 1992, 199). 17 See Acock and Bengston (1980, 501, 507); Connell (1972, 329); Dalton (1982, 145); Goldberg (1966, 919); Hess and Torney ([1967] 1968, 247); Hyman ([1959] 1969, ch. 4); Jennings (1984, 264- 265, 267, 270, 273); Jennings and Niemi (1968, 172-174; 1974, 39; 1968, 172); Jennings Stoker and Bowers (2001, 17); Kamieniecki (1985, 64); Knoke (1976, 95,105-106); Langton (1969, 52); Levin 33 H27: There will be a positive relationship between parents and children’s egalitarianism. However, research on both parental socialization and cohorts, suggests that socialization in the home is not leak proof. 18 Research on party identification and liberalism/conservatism, the quintessential contested values, finds the cohorts socialized during the New Deal/Fair Deal/New Frontier/Great Society periods are more liberal and Democratic than those socialized before and after. 19 The finding that egalitarianism is positively related to liberalism and identification with the Democratic Party, 20 suggests the following: H28: Middle cohorts will be more egalitarian in general than older and younger ones. Based upon group interests, underprivileged groups such as the poor, non-whites and women should be more supportive of contested egalitarianism than privileged groups such as whites, men and the affluent. H29: Underprivileged groups should be more supportive of contested egalitarianism than privileged ones. (1961, 596-598); Tedin (1974, 1584, 1586). I am unable to test the interaction between parents’ politicization and egalitarianism and the children’s’ egalitarianism due to linear dependency. 18 See Alwin (1997, 180); Converse (1976, 132); Carlsson and Karlsson (1970, 713); Cutler (1969-1970, 583, 587); Davis (1980, 1147-1148, 1154; 1992, 261, 295); Elder (1974, 262-264); Evan ([1965] 1968, 161); Feather (1975,143); Jennings, Stoker and Bowers (2001, 1); Knoke and Hout (1974, 700-701); Ladd (1970, 301-303); Lipset ([1959] 1960, 266-267) and Norpoth (1987, 376, 381-383, 385-387). 19 See Alwin (1997, 180); Converse (1976, 132); Carlsson and Karlsson (1970, 713); Cutler (1969-1970, 583, 587); Davis (1980, 1147-1148, 1154; 1992, 261, 295); Elder (1974, 262-264); Evan ([1965] 1968, 161); Knoke and Hout (1974, 700-701); Ladd (1970, 301-303); Lipset ([1959] 1960, 266-267) and Norpoth (1987, 376, 381-383, 385-387). The literature offers three substantive explanations for cohort effects: distinctive parental socialization, peer socialization and historical experience (Klecka 1971, 358-359; Ryder 1965, 843, 852-856). Carlsson and Karlsson (1970, 717) point out that because each birth cohort is associated with a range of parental cohorts, the initial shock would eventually dampen, resulting in inter- cohort stability. Even assuming a strong relationship between the values of the parent and child, parental socialization is not likely to be a strong explanatory factor for cohort effects. 20 See Feather (1975, 135), Federico and Sidanius (2002, 164-165). 34 Although the South has converged with the rest of the country in its partisan balance, it is also more conservative, less supportive of the welfare state and past scholarship has viewed its political culture as more deferential to authority than the rest of the country, which generates the following hypothesis. 21 H30: Southerners should be less supportive of contested equality than those living in other regions of the country. The hypotheses are different for general and economic egalitarianism for three variables: education, religiosity and trade unionism. On the question of the relationship of education to general egalitarianism, the verdict of past scholarship is mixed, finding education positively related to both conservatism and to post-materialism (Dalton 1977, 469; 22 Jacob 1957, 5, 34-35). Past work also finds education positively related to support for racial and gender liberal policies but negatively related to support for the welfare state (Bobo 1991, 80; Bobo and Kluegel 1993, 452, 455; Chafetz and Ebaugh 1983, 290; Jacob 1957, 4, 34-35; Ladd and Bowman 1998, 112-113). 23 Since past research offers conflicting results and elite discourse on general egalitarianism should be contested rather than consensus: 21 See Beck (1977, 486); Bobo (1991, 82); Chafez and Ebaugh (1983, 289); Firebaugh and Davis (1988, 260, 267); Gilliam and Whitby (1988, 65); Kluegel and Smith (1983, 810); Knapp and Alston (1972-1973, 587); Ladd with Hadley ([1975] 1978, 130, 151); Middleton (1976, 94, 102, 104, 106-107, 110); Miller (1981, 132); Norpoth (1987, 379); Peek and Brown (1980, 176); Pettigrew (1959, 28); Smith (1981a, 1256, 1262-1263; 1981b, 558, 564-567; 1982a, 3, 9-13; 1982b, 368-369, 375, 379; 1985, 205, 216-221); Tuch (1984, 63-65, 67; 1987, 506); Tuch and Hughes (1996, 735-736) and Wilson (1986, 180-181). 22 One of Dalton’s (1977, 462) indicators for post-materialism is giving people more say in important government decisions. 23 See Bobo (1983, 1207; 1991, 82); Bobo and Kluegel (1993, 452, 455); Campbell (1971, 54); Chafez and Ebaugh (1983, 289); Cherlin and Walters (1981, 456); Feree (1974, 394); Gilliam and Whitby (1988, 65); Greeley and Sheatsley (1971, 16, 19; 1974, 244-245); Huber and Spitze (1980, 327; 1981, 163); Jackman (1978, 302, 310-315, 320-321; 1981, 256-257); Jackman and Muha (1984, 758); Jacob (1957, 4, 29- 31, 34-35); Jacobson (1985, 319); Knoke (1979, 784); Ladd and Bowman (1998, 112-113); Mason 35 H31: Education should be unrelated to general egalitarianism. Although economic egalitarianism is a contested value, past work finds a negative relationship between economic egalitarianism and education (Bobo 1991, 80; Bobo and Kluegel 1993, 452, 455; Chafetz and Ebaugh 1983, 290; Jacob 1957, 4, 34-35; Ladd and Bowman 1998, 112-113). 24 H32: There will be a negative relationship between education and economic egalitarianism. Union members have become more conservative and are anxious to protect their status (Ladd 1970, 251; 1976-1977, 587-590; Ladd with Hadley [1978] 1975, 194-195, 217). Equality has come to be more closely associated with gender and racial issues and less with the economic issues that directly affect union members interests (Ladd 1976- 1977, 590; Ladd with Hadley [1978] 1975, 218-220). In addition, past empirical work finds that professionals are more likely to be liberal than manual workers and to reject the notion that there is too much concern with equality (Ladd 1976-1977, 595). H33: Union households are likely to be less supportive of general equality than non- union households. and Bumpass (1974, 1217); Mason and Lu (1988, 50); Maykovich (1975, 1014); Miller (1981, 131, 133); Ransford ([1977] 1994, 153); Ransford and Miller (1983, 52); Schreibner (1978, 178); Sears (1983, 99- 100); Shingles (1989, 957); Smith (1981a, 1256, 1266-1267; 1981b, 558, 564-567; 1982a, 15-17; 1982b, 377, 379-380; 1985, 205, 218-219, 221-222); Sniderman, Brody and Kuklinski (1984, 75-76); Schuman, et. al. (1997, 232); Schwartz (1967, 24, 30-31, 37-38, 44, 52, 60, 84-85, 90-91, 131-134); Stembler (1961, 168) and Tuch (1984, 56, 64-66; 1987, 506) and Welch and Sigelman (1982, 312). Kluegel and Smith (1983, 810) find a relationship with education for only some of the affirmative action policies they examined. Bobo (1983, 1206) finds no relationship in the case of busing. Shingles (1989, 957), Knoke (1979, 783) and Pollock (1983, 232) find no relationship between education and welfare state support. Peek and Brown (1980, 176) found a negative relationship between education and gender stereotyping. 24 Shingles (1989, 957), Knoke (1979, 783) and Pollock (1983, 232) find no relationship. 36 Although union members’ contribution to social spending has increased relative to their benefits from it, the core mission of labor unions, historically, has been lessening the economically unequal status of their members (Ladd 1976-1977, 590-591-595). 25 H34: Union households should be more economically egalitarian than non-union households. Past work reveals mixed results on the relationship of religiosity and values. On the one hand, the verdict from the research on religiosity and party identification is near unanimous in finding that the more religious are more Republican and conservative. 26 In addition, Feldman (1988, 426) found religious attendance positively related to support for equal opportunity. On the other hand, Rokeach (1969, 16) found no significant relationship between church attendance and importance of religion and ranking of equality. The conflicting nature of the evidence amassed from past work supports the following hypothesis: H35: Religious attendance will be unrelated to general egalitarianism. Since those who attend church often are more likely to be conservative and Republican: 27 H36: There should be a negative relationship between religiosity and economic egalitarianism. 25 Hamilton (1972, 322) finds little relationship between union membership and support for government provided health care and reasonable standard of living. 26 See Green and Guth (1991, 212, 222); Kamieniecki (1985, 56); Kaufman (2004, 502); Kaufman and Petrocik (1999, 870); Knoke (1976, 29-35); Kohut et al. (2000, 74-77); Jelen and Smidt (1993, 207-208); Layman (1997, 298); Lopatto (1985, 48-49); Miller and Wattenberg (1984, 302, 309). 27 See Green and Guth (1991, 212, 222); Jelen, Smidt and Wilcox (1993, 207-208); Kamieniecki (1985, 56); Kaufman (2004, 502); Kaufman and Petrocik (1999, 870); Knoke (1976, 29-35); Kohut 37 In the following chapter, I outline the methods I use to test these hypotheses and discuss the methodological issues raised by this analysis. et al. (2000, 74-77); Layman (1997, 298); Lopatto (1985, 48-49) and Miller and Wattenberg (1984, 302, 309). 38 Chapter 3 Data and Methods Public opinion does not exist--Pierre Bourdieu (1979, 129). To speak with precision of public opinion is a task not unlike coming to grips with the Holy Ghost--V. O. Key, Jr. (1961, 8) Public opinion is what opinion polls try to measure or measure with modest error--Philip E. Converse (1987, S14). Data Data are from three data sets. Study of Political Socialization: Parent-Child Pairs Based on Survey of Youth Panel and Their Offspring, 1997 (Jennings and Stoker 2004), Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined (Kent, Marcus and Niemi 1991) and National Election Studies 1948-2000 Cumulative Data File (Sapiro, Rosenstone and the National Election Studies 2001). The first two data sets are part of the same series of surveys. The respondents of the 1997 survey were the respondents of the prior surveys paired with their children who were 15 years of age or older. The reason for choosing these data sets are that the 1997 survey contains more indicators of egalitarianism and the indicators it contains are more direct, but the earlier surveys contain measures of past egalitarianism. Unfortunately, because these two data sets include only parents and children there is not full variance on age. They also do not include sufficient variables for multiple measurement of equal opportunity and domain specific egalitarianism. I add the 39 cumulative file from American National Election Study to the analysis in order to remedy these gaps. Dependent Variables The dependent variables for the parent-youth studies are latent variable estimates of students’ egalitarianism. Indicators for the 1997 data set are: two from the traditional NES egalitarianism scale, Push Equal Rights Too Far and Unequal Chances Not a Problem (See Appendix A for all question wordings and response categories and Table 3.1 for categories, types and measures of egalitarianism examined in this study). In addition, Women Equal and Anti-Gay Discrimination Laws (Appendix A) would seem, on their face, to be valid indicators, as well, a supposition that is supported by the results from the Fourth Chapter. The results from Chapter 4 also are consistent with the inclusion of two indicators of racial egalitarianism: Govt. Integrate Schools and Govt. Help Minorities (Appendix A). These indicators also measure equality of results, as does Govt. Provide Jobs, which also measures economic egalitarianism (Appendix A). For the 1973-1982 data set 1 , the Push Equal Rights Too Far and Unequal Chances Not a Problem variables are not present (Appendix A). Other than these two variables, the indicators for this data set are the same, with the addition of Bus to Achieve Integration, which measures racial egalitarianism and support for equality of results (Appendix A). Dependent variables from the NES Cumulative file are as follows: General Egalitarianism is a latent variable measured by the traditional 6 indicators from the American National Election Study (Appendix A). These are: Equal Opportunity, Push Equal Rights Too Far, Unequal Chances a Problem, Unequal Chances Okay, 1 There were insufficient indicators for egalitarianism in the 1965 wave. 40 Table 3.1. Measures of Types of Egalitarianism Type Definition Unidimensional-Contested Equal Treatment, Equal Rights, Equal Opportunity, Whether We Worry too Much about Equality. General Multidimensional-Consensus Equality of Opportunity A Continuum Running from Non- discrimination to Equal Life Chances. Multi-dimensional-Consensus-Domain Specific Racial Egalitarianism A Continuum Running from Equal Treatment, Equal Rights, Equal Opportunity and Equality of Results for People of All Races. Gender Egalitarianism A Continuum Running from Equal Treatment, Equal Rights, Equal Opportunity and Equality of Results for Men and Women Alike. Multi-dimensional-Contested-Domain Specific Economic A Continuum Running from Equal Treatment, Equal Rights, Equal Opportunity and Equality of Results for People of All Economic Circumstance. Immigrants Rights Equal Treatment, Equal Rights, Equal Opportunity for Citizens and Non-citizens Alike. Gay Rights Equal Treatment, Equal Rights, Equal Opportunity for Both Gays and Straights. Worry Less about Equality and Equal Treatment. Equal Opportunity is likewise measured by Equal Opportunity, Unequal Chances a Problem, Unequal Chances Okay and Worry Less about Equality (Appendix A). The general egalitarianism variables plus Guaranteed Jobs measure Economic Egalitarianism. Aid to Blacks plus the 6 general egalitarianism variables measure Racial Egalitarianism (Appendix A). 2 Finally, 2 As well as measuring two types of domain specific egalitarianism, Guaranteed Jobs and Aid to Blacks also measure social responsibility (Bobo 1988, 72, 74). 41 Gender Egalitarianism is measured with these 6 indicators, along with Women’s Equal Role (Appendix A). Independent Variables Students’ 1973 Egalitarianism and Parents’ Egalitarianism 3 are measured by the same indicators as above (Appendix A). Variables for testing group interest and identity are Race (coded non-white = 1), Gender (coded female = 1), Union Household (coded 1 for union household) and Religiosity (Appendix A). 4 There are four other group identity variables: Identification with Women’s Lib 5 (Gender multiplied by thermometer scores for Women’s Lib), Identification with Blacks (Race multiplied by thermometer scores for blacks), Identification with Unions (Union Household multiplied by thermometer score for labor unions) and Identification with the Poor (Income multiplied by thermometer score for the poor. See Appendix A). 6 Finally, Income is included in order to test self and group interest (Appendix A). There are also a series of dummy variables for cohort, which are categorized according to birth years as follows: 1895-1910 (the excluded, baseline category), 3 In order to eliminate the possibility of reciprocal causation, parents’ 1973 egalitarianism is entered into the equation as a predictor of students’ 1982 egalitarianism. 4 Religiosity is a latent variable estimated from the following indicators from the Parent-Youth Study: frequency of church attendance and interpretation of the Bible (more literal interpretation coded higher). Church Attendance measures Religiosity in the NES Cumulative File. 5 Identification with Women’s Groups in the NES Cumulative File. 6 These interaction variables are created by multiplying the group categories by thermometer scores for blacks, women or organized labor. The group categories are also included due to the problematic measurement properties of the thermometer scores, although they do tap psychological attachments (Conover 1988, 991; Krosnick 1991, 551, 563; Wilcox, Sigelman and Cook 1989, 251). 42 1911-1926, 1927-1942, 1943-1958, 1959-1974 and 1975-Present (Appendix A). Age is coded according to the following groups: 17-24, 25-34, 35-44, 45-54, 55-64, 65-74 and 75-99 (Appendix A). In order to test the elite leadership/political culture view, Political Knowledge, General Political Discussion, Education and Media Exposure are included in the models (Appendix A). In Defense of the Beleaguered Survey Because the only methodology utilized here is the analysis of surveys, before proceeding to a discussion of the methodological issues of this study, an examination of the criticisms of survey research is in order. Although polls and public opinion have come to be synonymous in the popular and academic mind, this view is not without its critics (Blumer 1948; Bourdieu 1979; Herbst 1992, 220; Lee 2002, 72). Two basic methodological objections shall be addressed here: (1), that polls do not measure public opinion at all and (2), that polls does not do a very good job of measuring public opinion. 7 The argument that polls do not measure public opinion is usually based on alternative definitions of public opinion. Some have argued public opinion is a group based phenomenon and to gather opinion at the individual level renders public opinion non-public (Bourdieu 1979,128; Blumer 1948, 543-544). However, to the extent that group based influences are at work (and they surely are), they will be manifest in the thinking of individuals. Although this definition may have been one of the most 7 Since a third type of criticism, that polls are bad for democracy, applies to all uses of polls and not specifically to their use in academic research, this view shall not be discussed here. For an 43 commonly accepted ones in past history, in the contemporary era, it is not the generally accepted one (Bogart 1972, 14-15; Lee 2002, 75-76; Ozouf 1988, S3, S5). It is also not the most useful definition of public opinion for purposes of research because it conflates the study of public opinion with that of interest groups and social movements and forecloses the posing of the question of the influence of the latter two on the former or renders such an analysis tautological by defining the dependent and independent variables as one and the same. The group-based definition also ignores the voices of the politically inactive and unorganized and could lead to the false impression that “noisy and overt ‘representatives’ of the public or of particular parts of it” speak for everyone (Verba 1996, 4, 6). By privileging the active, this definition also privileges those with higher education, higher incomes and higher status occupations, since the former are disproportionately composed of the latter (Verba 1996, 2, 4; Verba, Schlozman and Brady 1995, 493-494). A second alternative definition of public opinion is that opinion which is influential on public policy and is propounded by those who view public opinion as measured by surveys as politically inert and those who advocate weighting survey opinion according to influence (Blumer 1948, 544-545; Bourdieu 1979, 124). This definition is flawed in a similar fashion as the first. It is also not the generally accepted definition and forecloses investigating the influence of public opinion on public policy since it conflates the two (Newcomb 1948, 549). How one would non- tautologically weight public opinion other than by its influence on public policy is explication of this view see Beninger (1986); Bordieu (1979); Ginsburg (1986) and Habermas ([1962] 1989) and for a contrary view see Converse (1996) and Verba (1996). 44 unclear. Furthermore, inquiry into which groups are influential or not is foreclosed by this definition of public opinion. The factual assumption upon which it is based--that polls do not have influence with policymakers--is contradicted by some of the other criticisms of surveys and by empirical work in the political science literature (Converse 1987, S14; 1996, 651; Lee 2002, 81). The most extreme definition of public opinion is that it does not exist at all (Bourdieu 1979, 129-130). This is true, it is argued, because of the absence of three conditions: that everyone has an opinion, that everyone’s opinion is equal, and a consensus exists about what questions should be asked at a given time and that survey researchers accurately know this consensus or can know it (Bourdieu 1979, 124). This definition is not only even more idiosyncratic than the last two discussed, but raises the normatively disturbing possibility that in the absence of the idealized conditions listed, representative government, in the sense of representing people’s desires, would be impossible. The most telling criticism of surveys is that they do not measure public opinion very well. One fundamental criticism is that people may not have attitudes about the questions that pollsters ask but may have attitudes about questions that they do not (Bordieu 1979, 124, 125,127; Converse 1964; Dionne 1992, 164; Herbst 1992, 223; Lee 2002, 83; Neuman 1989, 219; Verba 1996, 6). Even if they do have such attitudes, the closed end questions commonly asked, lack richness and depth (Verba 1996, 6). Where respondents have attitudes that are not asked about, polls may not be able to distinguish elites leadership from that of non-elites (Lee 2002, 84). However, this is mainly a 45 problem in the case of new issues that arise rather than perennial ones, such as egalitarianism. A second methodological criticism of polls is that since they seek to measure “intangible, subjective phenomena--attitudes, feelings, beliefs, emotions” they are “plagued with considerable measurement noise—a mixture of true attitudes, quasi attitudes and artificially invented pseudo-opinions.” (Brady and Orren, 1995, 55; Neuman 1989, 219). However, some public opinion researchers admit that polls do not measure “true attitudes” but only opinion statements that consist of the balance of considerations that are the most salient at any given time (Zaller 1992, 35; Verba 1996, 6). 8 This means that what respondents tell pollsters at any given time may be influenced by media and governing elites (Verba 1996, 6). Compounding the problem of what pollsters ask when, is the fact that only those able to bear the considerable expense of conducting a probability survey may do so (Verba 1996, 6). It is not only the questions and answers to polls that are problematic but also the respondents. Those whom pollsters are able to reach and who will talk to them are not perfectly representative of those who are not (Brehm 1993, ch. 2; Brady and Orren 1995, 61; Singer 1987, S1; Verba 1996, 3). Polls under represent those who have no stable dwelling place and the rich, but overrepresent women, the elderly and African Americans and telephone polls miss the small percentage of Americans without telephones (Brehm 1993, 37; Brady and Orren 1995, 61; Verba 1996, 3). Furthermore, no one knows how high response rates must be to avoid bias and weighting only works if those who answer are representative of those who do not (Brady and Orren 1995, 62, 64). However, survey 8 A more detailed discussion of measurement error follows this section. 46 researchers have attempted and are still attempting to address the methodological problems of polls to which critics have pointed (Andrews 1984; Alvarez and Brehm 1995; 1997; Brehm 1993, 122-124; Lee 2002, 82; Schuman and Converse 1971). Criticism of polls extends even beyond the point at which the survey is completed, into the phase of data analysis. This phase has been described as “data- dredging” or obtaining results through the “power of prophecy” (Neuman 1989, 220; Selvin 1968, 418; Selvin and Stuart 1966, 20). However, this criticism is not uniquely applicable to surveys, but would apply to any quantitative non-experimental study in which strong theory is lacking. This need not impede knowledge building as long as past work is not ignored or exploratory results are tested in a different sample. Because of the problems of survey research some have argued that it should be supplemented in political communication studies by other methods such as content analysis, experiments, focus groups and depth interviewing (Neuman 1989, 254-255; Neuman, Just and Crigler 1992, 24; Lee 2002, chs. 3-4). The author agrees, that this injunction, if implemented in the aggregate (and it has begun to be), would add to, enrich and make more rigorous, knowledge gained by survey research alone. However, to insist on its implementation at the level of each researcher and each research project, ignores the unique resources, capabilities and circumstances of each researcher at any given point in time. Furthermore, survey research, done at the individual level, is a more conservative test of the effects of political communication on political predispositions than experiments and is also likely to be as conservative a test, if not more so, than aggregate time series studies and avoids the problem, in the latter case, of inability to 47 make inference to the individual level (Hovland 1959, 8). Because of the inability to infer cause and effect from content analysis to public attitudes, the former is low in internal validity (Neuman 1989, 223-224; Neuman, Just and Crigler 1992, 26). Because sample sizes are typically small and possibly unrepresentative, and the lack of structure makes causal inference problematic, depth interviewing or focus groups are also problematic (Feldman and Zaller 1992, 274-275; Neuman, Just and Crigler 1992, 27). In conclusion, the weaknesses of methods other than surveys are as serious, if not more so, than surveys. The Identification Problem Another methodological problem arises in testing hypotheses about the effects of age, cohort and period. Since cohort, age and period effects are perfectly collinear (the value of one is perfectly determined by the other two), equations including all three variables are unidentified without making further assumptions. 9 One solution is to transform one variable to make it nonlinear (Mason et. al. 1973, 246; Rodgers 1982, 782). However, in most social science research contexts the substantive foundation for choosing the particular transformation is often lacking, as is the case here (Mason et. al. 1973, 246). Another is to assume the effects of two or more age groups, cohorts or periods on the dependent variable are equal (Fienberg and Mason 1979, 3-4; Mason et. al. 1973, 247, 253). Like the first solution, the foundation for making such a choice is seldom there and the results are not robust to the choice made (Fienberg and Mason 1979, 3-4, 16; 1985, 4; Mason et. al. 1973, 248, 250, 253; Rodgers 1982, 775, 777-780, 782, 785; Sasaki and Suzuki 1989, 764). A third solution is to assume the effect of either age, 9 See Alwin (1997, 165); Baltes (1968, 152-153, 156-159); Evan (1959, 68); Fienberg and Mason (1979, 5, 13; 1985, 68); Glenn (1976, 900; 1981, 362-363; 1989, 755); Goldstein (1968, 96); Markus (1985, 262); Mason and Fienberg (1985, 2-3); Mason et. al. (1973, 242, 253); Rodgers (1982, 774) and Riley (1973, 41-42, 46). 48 cohort or period is zero, but this can lead to spurious relationships (Mason et. al. 1973, 242, 253; Rodgers 1982, 786). Since the results in Chapter 4 offer no evidence whatsoever of period effects during the time covered by this study, this is the solution adopted here. In addition, this is the theoretical prediction of the traditional view, which is supported by the results of the Fifth and Sixth chapters. Finally, the effects of the mass media and political discussion, which are likely mechanisms through which period effects would work, are already in the model. Measurement Error While Neuman (1989, 219) is correct that measurement error is a problem with survey research, this problem is not without solutions. There are three types of measurement error. The first type is bias, in which error is constant for all respondents and it does not affect the relationship among variables and is a problem only if one’s substantive interest is in measures of central tendency or in the y intercept (Andrews 1984, 410, 412; Wooldridge 2000, 292.). The only time this problem is relevant to this analysis is in Chapter 4 where I investigate the change in the mean of egalitarianism over time. However, since I examine only the means of latent variables and the scale of a latent variable is arbitrary, raising all means by a certain amount does not affect comparisons over time. The second type of measurement error is random error, in which errors in one variable are unrelated to those of others (Andrews 1984, 412; Asher 1974, 470; Wooldridge 2000, 292). This type of error, if found in dependent variables, can lead to inefficiency (inflated standard errors, which might result in a “false negative” finding on 49 the question of statistical significance) but will not lead to biased parameter estimates and would thus serve as a conservative test of the hypotheses investigated (Pindyck and Rubenfeld 1981, 176-177; Hu [1973] 1982, 101; Wooldridge 2000, 292). In addition, the multiple indicators, which I utilize here, will also ameliorate the magnitude of the problem of measurement error. Random measurement error, in independent variables, leads to both bias and inconsistency (estimated values of the parameter of interest will not converge to true values as sample size grows), of unknown magnitude and direction (Achen 1983, 72-73; Hu [1973] 1982, 102; Maddala 1977, 292-294; Wooldridge 2000, 296). In many cases, multiple indicators are also available for independent variables, particularly for self-reports of media exposure, for which measurement error is thought to be a considerable problem (Bartels 1993, 267; Zaller 1992, 334-335). Statistical techniques exist for utilizing these multiple measures to obtain an estimate of the true scores and will be utilized here (Blalock 1970; Wiley and Wiley 1970) Measurement is much more reliable for demographic attributes and for other core values such as party identification, for which multiple measures may not be available (Asher 1974, 473-484; Converse 1964, 240; Converse and Markus 1979, 37-38, 46; Krosnick 1991, 559-561; Norpoth and Lodge 1985, 301). The third type of measurement error is that which is correlated among variables (Achen 1983, 79; Andrews 1984, 412, Wiley and Wiley 1974, 172-176). Statistical techniques also exist to utilize multiple indicators to address this problem and will be used in this analysis (Blalock 1970; Wiley and Wiley 1974). 50 Some questions used in this analysis are in the agree-disagree format, which raises two specific types of measurement error: acquiescence/directional response bias and extreme response style. The first type of measurement error is the tendency to agree or disagree with any question, regardless of content (Bachman and O’Malley 1984, 491; Hui and Triandis 1985, 253; Jackson 1979, 394; Krosnick 1991, 552). Since many of the questions are balanced between those stated in the affirmative and negative, agreeing or disagreeing with all items would lead to random error (Hui and Triandis 1985, 253). This type of error has been considered to be random, not only among items, but also among individuals (Hui and Triandis 1985, 260; Krosnick 1991, 552). However, other studies have found that directional bias is correlated with race or education (Bachman and O’Malley 1984, 506; Schuman and Presser 1981, 229). Extreme response style is the tendency to choose the most extreme response categories, for example, strongly agree and strongly disagree, regardless of content (Bachman and O’Malley 1984, 491). This tendency has been related to age, education, income and race (Bachman and O’Malley 1984, 491; Greenleaf 1992, 328). This research suggests that the better practice is not to assume that the errors are uncorrelated and this is the practice I follow in this research. Whether correcting for measurement error in dependent variables removes the effects of vague questions and response categories, respondent vacillation or top of the head answers, it attains a better estimate of the underlying balance of considerations in the minds of the respondents than not to do so. 10 10 Correcting for measurement error substantially amounts to Zaller’s (1992, 198-199) explicit modeling of the recall function, but avoids the multicollinearity and resulting statistical imprecision caused by including political awareness three times in the equation, as well as in additional interaction variables. 51 I correct for measurement error by using structural equation modeling. This estimates both measurement and analysis models at the same time, giving it greater statistical efficiency than using factor scores or summed scales as dependent variables (Kim and Mueller 1978b, 51). 11 Unlike structural equation modeling, summed scales contain measurement error rather than removing it and estimating it separately (Bollen and Lennox 1991, 309-310). Rankings and Ratings There is disagreement in the literature about how best to measure egalitarianism. Some argue that values, including egalitarianism, are inherently comparative and utilize rankings of values to investigate value systems (Rokeach 1973). However, ranking is not without its problems. Ranking values is a difficult and time-consuming task for respondents (Alwin and Krosnik 1986, 536; Moore 1980, 405; Rokeach 1973, 28; Feather 1975; 37; 1973, 228). For this reason, the ranking method is expensive for researchers (Alwin and Krosnik 1986, 536; Munson and McIntyre 1979, 49). Rankings are also difficult to do over the telephone (Alwin and Krosnik 1986, 536). For these reasons ranking data is less widely available than rating data. The sum of rankings for each respondent is a constant, resulting in a linear dependency among the rankings and rendering conventional statistical techniques problematic (Alwin and Krosnik 1986, 536; Gorsuch 1970, 140-141; Jackson and Alwin 1980, 219, 221; Rokeach 1970, 159; 1973, 52). 12 Rankings also utilize single indicators and have been criticized as finding 11 There is seldom a perfect correlation between factor scores and the latent variable (Kim and Mueller 1978, 51). Summed scales assume the correlation between each indicator and latent variable is the same, which may or may not be true (Kim and Mueller 1978a, 65) 12 However, Rokeach (1970, 159; 1973, 43) argues that with a large number of values, such as the 18 in the Rokeach Values Surveys, the violation of the independence assumption is small. 52 differences in the meaning of the words rather than the underlying psychological construct (Gorsuch 1970, 139). Ratings are easier and less expensive to administer, can be used in a telephone survey and are not linear dependent (Alwin and Krosnik 1986, 536-537, Munson and McIntyre 1979, 49; Rankin and Grube 1980, 234). However, the ease of response to ratings may reduce data quality (Alwin and Krosnik 1986, 537). Ratings are susceptible to response sets and when all items are considered to be good or socially desirable, variance is deflated (Alwin and Krosnik 1986, 537; Block 1965; Cunningham, Cunningham and Green 1977, 379; Rokeach 1973, 52). Research comparing the two methods has found that aggregate preferences tend to be similar, but the similarity is less pronounced at the individual level (Alwin and Krosnik 1986, 537; Moore 1980, 407; Rankin and Grube 1980, 233, 236, 239-240). Construct validity appears to be similar for the two methods and predictive validity for ratings is somewhat higher (Alwin and Krosnik 1986, 537; Rankin and Grube 1980, 243). However, both of the approaches described above, assume the uni-dimensionality of egalitarianism without empirically demonstrating it. This may be a necessary expedient when investigating the structure of value systems, a process that requires that many values be included in the analysis, but is logically doubtful as well as empirically unsupported, and unnecessary, where, as here, only one value is being considered. Therefore, this analysis shall, by utilizing multiple indicators, attempt to determine empirically, the issue of whether, egalitarianism, in the mind of the United States public, for the period being investigated, is uni- or multi-dimensional (Feldman 1988, 420-421). 53 Because multiple indicators are available only for data in the rating format and the literature suggests that this format is not significantly different in many respects from the ranking format, data, in the rating format only, will be analyzed. In addition, rankings measure the relative importance of values and I am interested in the absolute importance of one value, which the rating format measures (Feather 1975, 31). 54 Chapter 4 The Meaning of Equality for the U. S. Public Before the question of mechanisms by which egalitarianism is formed can be investigated, the meaning of it must first be determined. Since the literature offers conflicting definitions, hypotheses as to these definitions will be tested empirically here. First, I consider the meaning of equality for the United States public, aggregated across time, to the extent possible 1 and groups. Then I consider the question of whether equality means the same across time and among various politically salient groupings of Americans. Here I test the meaning of equality for the American public in the following way: more constrained conceptualizations 2 of equality should exhibit better model fit than less constrained ones. Additionally, I test whether the levels of various types of egalitarianism vary across time and among groups. Egalitarianism in the Aggregate The fact that the all models have a RMSEA of less than .05, passing the close fit test, offers some evidence in their favor (Table 4.1). 3 However, because all except equality of opportunity have a significant Chi-Square and therefore fail the exact fit test, the strongest evidence is that equality means equality of opportunity. The domain specific 1 In this section, only, the same indicators are utilized at all times. Indicators were chosen to balance coverage across time and number of indicators, which were available from the ANES Cumulative File for the years 1984, 1986, 1990, 1992 and 1994. Aggregating across time and groups assumes homogeneity in the relationships between the variables, an assumption that may not be accurate and can be tested. 2 Since I am examining cross-sectional data, I investigate constraint as construct reliability rather than across time construct stability. 3 The rationale for the close fit test is trivial lack of fit can cause models to be rejected by the Chi-Square test in large samples (Bentler and Bonett 1980, 588, 591; Nevitt and Hancock 2000, 251, Ullman 2001, 698). 55 models do not fare well, all have higher AICs and all except gender have higher CAICS than general equality and equality of opportunity. 4 Table 4.1. Measures of Fit for Egalitarianism Structural Equation Models Egalitarianism Chi Square RMSEA AIC CAIC General 62.829**** 0.0296 50.829 1.165 Opportunity .836 .002 -1.163 -9.440 Economic 307.301**** .050 285.301 194.251 Racial 235.569**** .043 213.569 122.518 Gender 98.995**** .027 76.995 -14.055 N = 10693 **** = p < .0001 Source: National Election Study Cumulative File, 1948-2000. Table 4.1 offers support for both the resurgence of class and identity politics theses with both the economic and racial models passing the RMSEA test. The identity politics thesis is further supported by the superior fit of the racial and gender Turning to the coefficient estimates, it is unsurprising that the general indicators are statistically significant in all models (Tables 4.2-4.6). 5 The coefficient for the guaranteed jobs, aid to 4 The AIC is the Akaike Information Criterion and CAIC is Consistent Akaike Information Criterion. These measures of fit are included because they are the only ones that can be compared across non-nested models (Akaike 1987, 318-320). The AIC is included because it is better known. The formula is –2 Log Likelihood + logn x p, where logn is the natural logarithm of the number of cases and p is the number of parameters (Akaike 1987, 318-320). The CAIC, though less well known has the desirable statistical properties of consistency (converging to the true value as the sample size approaches infinity), unlike the AIC, and rewards parsimony more than the AIC (Bozdogan 1987, 358). The formula is –2 Log Likelihood + logn x p, where logn is the natural logarithm of the number of cases and p is the number of parameters. Lower values indicate better fit for both (Bozdogan 1987, 358). 5 Since the indicators for egalitarianism were categorical and the assumption that they were non-linear fit the data better than the assumption that they were linear, egalitarianism was estimated as a latent continuous normally distributed variable linked to the observed indicators by a probit function. Since this function is nonlinear, interpretation is not straightforward. As a result, I can make no specific claims about the magnitude of the coefficients. For every unit increase in the observed indicators, egalitarianism increases by the coefficient multiplied by the predicted probit score. The predicted probit score is the inverse z score--the predicted values for egalitarianism minus the mean divided by the standard deviation (Liao 1994, 22-24). 56 blacks and the women’s equal role are also correctly signed and statistically significant (Tables 4.2-4.6). Discussion Contrary to the historical and qualitative political science work, the results here support the assumption by much of the quantitative political science work that the six item NES general egalitarianism scale is an adequate measure of this concept— substantively that equality is unidimensional and means equality of opportunity, equal rights and equal treatment. However, contrary to the assumption that this is the only valid measure of egalitarianism, the results support treating egalitarianism as multi- dimensional—that separate dimensions may be found for the domain specific economic, gender and racial egalitarianism as well as for equality of opportunity. These results offer support for both the identity politics and resurgence of class theses. The strongest evidence for a meaning of equality is equality of opportunity. The research questions investigated in this chapter cannot be answered by the results reported here by a simple yes or no answer. The null hypothesis can be rejected in every case, but the evidence is stronger for some hypotheses than others. Substantively, what this suggests is that equality of opportunity is the predominant meaning of equality for the American public but that the other meanings also find adherents to a lesser degree. Similarly and paradoxically, egalitarianism is primarily uni-dimensional but is multi-dimensional to a lesser degree. To put it simply, egalitarianism means mostly one thing, but can mean other things to a lesser degree. In the next section I consider whether the meaning of egalitarianism is stable over time. 57 Table 4.2. General Egalitarianism Measurement Structural Equation Model Coefficients a (Robust Standard Errors in Parentheses) SHOULD SOCIETY INSURE EQUAL OPPORTUNITY 1.000 b HAVE GONE TOO FAR PUSHING EQUAL RIGHTS 1.580 (.05)*** BIG PROBLEM IF CHANCES NOT EQUAL 6.569 (.04)*** OK IF SOME HAVE MORE EQUAL CHANCE 1.479 (.05)*** SHOULD WORRY LESS ABOUT EQUALITY 1.771 (.06)*** FEWER PROBLEMS IF MORE EQUAL TREATMENT 1.055 (.06)*** N = 10693 *** = p < 0.001 Sources: American National Election Study Cumulative File, 1948-2000. c a All variables were coded on a 0-1 scale, with higher intensity of agreement egalitarian statements and disagreement with inegalitarian statements coded higher and higher intensity of disagreement with egalitarian statements and agreement with inegalitarian statements coded lower. All models are corrected for non-normal distribution (Satorra and Bentler 1994, 406-411). b On all measurement models, a variable was set to one in order to set the scale for the latent variable and to identify the equations (Bentler 1983, 18; Hayduk 1996, 25). c Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) y 5 = Φ -1 ( Λ 5 η + ε 5 ) y 6 = Φ -1 ( Λ 6 η + ε 6 ) where y 1 = SHOULD SOCIETY INSURE EQUAL OPPORTUNITY y 2 = HAVE GONE TOO FAR PUSHING EQUAL RIGHTS y 3 = BIG PROBLEM IF CHANCES NOT EQUAL y 4 = OK IF SOME HAVE MORE EQUAL CHANCE y 5 = SHOULD WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT Φ -1 = an inverse cumulative normal distribution function (inverse z score). η is a latent egalitarianism variable, the Λs are coefficients and the εs are error terms. 58 Table 4.3. Opportunity Egalitarianism Structural Equation Measurement Model Coefficients (Robust Standard Errors in Parentheses) SHOULD SOCIETY INSURE EQUAL OPPORTUNITY 1.000 BIG PROBLEM IF CHANCES NOT EQUAL .880 (.04)*** OK IF SOME HAVE MORE EQUAL CHANCE 1.613 (.07)*** SHOULD WORRY LESS ABOUT EQUALITY 1.526 (.06)*** N = 10693. *** = p < .0001. Source: American National Election Study Cumulative File, 1948-2000. d d Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) y 5 = Φ -1 ( Λ 5 η + ε 5 ) where y 1 = SHOULD SOCIETY INSURE EQUAL OPPORTUNITY y 3 = BIG PROBLEM IF CHANCES NOT EQUAL y 4 = OK IF SOME HAVE MORE EQUAL CHANCE y 5 = SHOULD WORRY LESS ABOUT EQUALITY Φ -1 = an inverse cumulative normal distribution function (inverse z score). η is a latent egalitarianism variable, the Λs are coefficients and the εs are error terms. 59 Table 4.4. Economic Egalitarianism Structural Equation Measurement Model Coefficients (Robust Standard Errors in Parentheses). SHOULD SOCIETY INSURE EQUAL OPPORTUNITY 1.000 HAVE GONE TOO FAR PUSHING EQUAL RIGHTS 1.538 (.05)*** BIG PROBLEM IF CHANCES NOT EQUAL .912 (.04)*** OK IF SOME HAVE MORE EQUAL CHANCE 1.447 (.05)*** SHOULD WORRY LESS ABOUT EQUALITY 1.689 (.05)*** FEWER PROBLEMS IF MORE EQUAL TREATMENT 1.069 (.04)*** 7PT GOVT GUARANTTED JOBS SCALE e .569 (.04)*** N = 10693 *** = p < 0.001 Source: American National Election Study Cumulative File, 1948-2000. f e This variable was coded on a 0-1 scale with favorability toward government jobs coded high. f Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) y 5 = Φ -1 ( Λ 5 η + ε 5 ) y 6 = Φ -1 ( Λ 6 η + ε 6 ) y 7 = Φ -1 ( Λ 7 η + ε 7 ) where y 1 = SHOULD SOCIETY INSURE EQUAL OPPORTUNITY y 2 = HAVE GONE TOO FAR PUSHING EQUAL RIGHTS y 3 = BIG PROBLEM IF CHANCES NOT EQUAL y 4 = OK IF SOME HAVE MORE EQUAL CHANCE y 5 = SHOULD WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 7 = 7PT GOVT GUARANTTED JOBS SCALE Φ -1 = a probit link function which is an inverse cumulative normal distribution function (inverse z score). η is a latent egalitarianism variable, the Λs are coefficients and the εs are error terms. 60 Table 4.5. Racial Egalitarianism Measurement Structural Equation Model Coefficients (Robust Standard Errors in Parentheses) SHOULD SOCIETY INSURE EQUAL OPPORTUNITY 1.000 HAVE GONE TOO FAR PUSHING EQUAL RIGHTS 1.562 (.08)*** BIG PROBLEM IF CHANCES NOT EQUAL .929 (.04)*** OK IF SOME HAVE MORE EQUAL CHANCE 1.434 (.05)*** SHOULD WORRY LESS ABOUT EQUALITY 1.702 (.06)*** FEWER PROBLEMS IF MORE EQUAL TREATMENT 1.076 (.04)*** 7PT AID TO BLACKS SCALE g .831 (.05)*** N = 10693. *** = p < 0.001 Source: American National Election Study Cumulative File, 1948-2000. h g This variable was coded on a 0-1 scale with favorability toward aid to blacks coded high. h Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) y 5 = Φ -1 ( Λ 5 η + ε 5 ) y 6 = Φ -1 ( Λ 6 η + ε 6 ) y 8 = Φ -1 ( Λ 8 η + ε 8 ) where y 1 = SHOULD SOCIETY INSURE EQUAL OPPORTUNITY y 2 = HAVE GONE TOO FAR PUSHING EQUAL RIGHTS y 3 = BIG PROBLEM IF CHANCES NOT EQUAL y 4 = OK IF SOME HAVE MORE EQUAL CHANCE y 5 = SHOULD WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 8 = 7PT AID TO BLACKS SCALE Φ -1 = a probit link function which is an inverse cumulative normal distribution function (inverse z score). η is a latent egalitarianism variable, the Λs are coefficients and the εs are error terms. 61 Table 4.6. Gender Egalitarianism Structural Equation Measurement Model Coefficients (Robust Standard Errors in Parentheses) Egalitarianism SHOULD SOCIETY INSURE EQUAL OPPORTUNITY 1.000 HAVE GONE TOO FAR PUSHING EQUAL RIGHTS 1.605 (.05)*** BIG PROBLEM IF CHANCES NOT EQUAL .897 (.04)*** OK IF SOME HAVE MORE EQUAL CHANCE 1.484 (.05)*** SHOULD WORRY LESS ABOUT EQUALITY 1.788 (.06)*** FEWER PROBLEMS IF MORE EQUAL TREATMENT 1.060 (.04)*** 7PT WOMENS EQUAL ROLE SCALE i .697 (.04)*** N = 10693. *** = p < 0.001 Source: American National Election Study Cumulative File, 1948-2000. j i This variable was coded on a 0-1 scale with favorability toward an equal role for women coded high. j Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) y 5 = Φ -1 ( Λ 5 η + ε 5 ) y 6 = Φ -1 ( Λ 6 η + ε 6 ) y 9 = Φ -1 ( Λ 9 η + ε 9 ) where y 1 = SHOULD SOCIETY INSURE EQUAL OPPORTUNITY y 2 = HAVE GONE TOO FAR PUSHING EQUAL RIGHTS y 3 = BIG PROBLEM IF CHANCES NOT EQUAL y 4 = OK IF SOME HAVE MORE EQUAL CHANCE y 5 = SHOULD WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 9 = 7PT WOMENS EQUAL ROLE SCALE Φ -1 = a probit link function which is an inverse cumulative normal distribution function (inverse z score). η is a latent egalitarianism variable, the Λs are coefficients and the εs are error terms. 62 Egalitarianism Across Time: A Value Realignment? Introduction In this section I consider whether and how the meaning of egalitarianism across time has changed since 1952 in the United States, this period being dictated by data availability. Since data for direct measures of egalitarianism are available only for about the past 20 years, this limits the length of time for the analysis of these measures. In addition, for two domain specific types of egalitarianism, gay rights and immigration, data are only available for the past decade to decade and a half, and as a result, the coverage in time is also limited for these analyses. Fortunately, data for three major domain specific kinds of egalitarianism—economic, racial and gender egalitarianism--are available for the full period, although the number of observations for equal civil liberties for unpopular groups is only four. The meaning of equality for the U. S. public is again operationalized as the constraint of particular types of egalitarianism at various times. 6 More specifically, I assume that when types of egalitarianism are more constrained, model fit will be better than otherwise. In addition, I am able to test the change in level of egalitarianism over time by conducting a Multiple Group Covariance and Mean Structure Analysis, with each year as a group. The changes in means over time strongly support the symbolic politics perspective and traditional theory of values (Table 4.7). None of the means are substantively different from the 1984 mean, which was fixed at zero in order to conduct the differences of means test (though the means for General Egalitarianism were statistically different from the 1984 mean). 6 Because the mean of a latent variable is dependent on its structural equation coefficients (or factor loadings), and since their scale is arbitrary, they cannot be compared out of sample and it is, 63 Table 4.7. Means of Types of Egalitarianism in Various Years General Opportunity Economic Race Gender 1984 0 k 0 0 0 0 1986 -.016 (.003)**** .000 (.002) (-.130) .000 (.003) (-.066) .000 (.002) (-.018) .000 (.004) (-.048) 1990 -.019 (.003)**** .000 (.002) (-.220) .000 (.003) (-.101) .000 (.002) (.014) -.000 (.004) (-.137) 1992 -.021 (.003)**** -.001 (.002) (-.346) -.000 (.003) (-.222) .000 (.002) (-.035) -.002 (.004) (-.472) 1994 -.017 (.003)**** -.001 (.002) (-.501) -.001 (.003) (-.365) .000 (.003) (-.006) -.002 (.004) (-.434) **** = p < 0.001 N = 10693 Source: American National Election Study Cumulative File, 1948-2000. l therefore, impossible to obtain estimates of the mean that are comparable out of sample. k The means for this year was set at zero in order to conduct a differences of means test l Equations are as follows: For General Egalitarianism: y 1,1984 = μ y1 + η + ε 1 y 2, 1984 = μ y2 + Λ 2 η + ε 2 y 3, 1984 = μ y3 + Λ 3 η + ε 3 y 4, 1984 = μ y4 + Λ 4 η + ε 4 y 5, 1984 = μ y5 + Λ 5 η + ε 5 y 6, 1984 = μ y6 + Λ 6 η + ε 6 η 1984 = ζ y 1,1986 = μ y1 + η + ε 1 y 2, 1986 = μ y2 + Λ 2 η + ε 2 y 3, 1986 = μ y3 + Λ 3 η + ε 3 y 4, 1986 = μ y4 + Λ 4 η + ε 4 y 5, 1986 = μ y5 + Λ 5 η + ε 5 y 6, 1986 = μ y6 + Λ 6 η + ε 6 η 1986 = μ η1986 + ζ y 1,1990 = μ y1 + η + ε 1 y 2, 1990 = μ y2 + Λ 2 η + ε 2 y 3, 1990 = μ y3 + Λ 3 η + ε 3 y 4, 1990 = μ y4 + Λ 4 η + ε 4 y 5, 1990 = μ y5 + Λ 5 η + ε 5 y 6, 1990 = μ y6 + Λ 6 η + ε 6 η 1990 = μ η1990 + ζ y 1,1992 = μ y1 + η + ε 1 y 2, 1992 = μ y2 + Λ 2 η + ε 2 64 Table 4.7: Continued y 3, 1992 = μ y3 + Λ 3 η + ε 3 y 4, 1992 = μ y4 + Λ 4 η + ε 4 y 5, 1992 = μ y5 + Λ 5 η + ε 5 y 6, 1992 = μ y6 + Λ 6 η + ε 6 η 1992 = μ η1992 + ζ y 1,1994 = μ y1 + η + ε 1 y 2, 1994 = μ y2 + Λ 2 η + ε 2 y 3, 1994 = μ y3 + Λ 3 η + ε 3 y 4, 1994 = μ y4 + Λ 4 η + ε 4 y 5, 1994 = μ y5 + Λ 5 η + ε 5 y 6, 1994 = μ y6 + Λ 6 η + ε 6 η 1994 = μ η1994 + ζ For Equal Opportunity: y 1,1984 = μ y1 + η + ε 1 y 3, 1984 = μ y3 + Λ 3 η + ε 3 y 4, 1984 = μ y4 + Λ 4 η + ε 4 y 5, 1984 = μ y5 + Λ 5 η + ε 5 η 1984 = ζ y 1,1986 = μ y1 + η + ε 1 y 3, 1986 = μ y3 + Λ 3 η + ε 3 y 4, 1986 = μ y4 + Λ 4 η + ε 4 y 5, 1986 = μ y5 + Λ 5 η + ε 5 η 1986 = μ η1986 + ζ y 1,1990 = μ y1 + η + ε 1 y 3, 1990 = μ y3 + Λ 3 η + ε 3 y 4, 1990 = μ y4 + Λ 4 η + ε 4 y 5, 1990 = μ y5 + Λ 5 η + ε 5 η 1990 = μ η1990 + ζ y 1,1992 = μ y1 + η + ε 1 y 3, 1992 = μ y3 + Λ 3 η + ε 3 y 4, 1992 = μ y4 + Λ 4 η + ε 4 y 5, 1992 = μ y5 + Λ 5 η + ε 5 η 1994 = μ η1992 + ζ y 1,1994 = μ y1 + η + ε 1 y 3, 1994 = μ y3 + Λ 3 η + ε 3 y 4, 1994 = μ y4 + Λ 4 η + ε 4 y 5, 1994 = μ y5 + Λ 5 η + ε 5 η 1994 = μ η1994 + ζ For Economic Egalitarianism: y 1,1984 = μ y1 + η + ε 1 y 2, 1984 = μ y2 + Λ 2 η + ε 2 y 3, 1984 = μ y3 + Λ 3 η + ε 3 y 4, 1984 = μ y4 + Λ 4 η + ε 4 y 5, 1984 = μ y5 + Λ 5 η + ε 5 y 6, 1984 = μ y6 + Λ 6 η + ε 6 y 7, 1984 = μ y7 + Λ 7 η + ε 7 η 1984 = ζ y 1,1986 = μ y1 + η + ε 1 y 2, 1986 = μ y2 + Λ 2 η + ε 2 y 3, 1986 = μ y3 + Λ 3 η + ε 3 y 4, 1986 = μ y4 + Λ 4 η + ε 4 65 Table 4.7: Continued y 5, 1986 = μ y5 + Λ 5 η + ε 5 y 6, 1986 = μ y6 + Λ 6 η + ε 6 y 7, 1986 = μ y7 + Λ 7 η + ε 7 η 1986 = μ η1986 + ζ y 1,1990 = μ y1 + η + ε 1 y 2, 1990 = μ y2 + Λ 2 η + ε 2 y 3, 1990 = μ y3 + Λ 3 η + ε 3 y 4, 1990 = μ y4 + Λ 4 η + ε 4 y 5, 1990 = μ y5 + Λ 5 η + ε 5 y 6, 1990 = μ y6 + Λ 6 η + ε 6 y 7, 1990 = μ y7 + Λ 7 η + ε 7 η 1990 = μ η1990 + ζ y 1,1992 = μ y1 + η + ε 1 y 2, 1992 = μ y2 + Λ 2 η + ε 2 y 3, 1992 = μ y3 + Λ 3 η + ε 3 y 4, 1992 = μ y4 + Λ 4 η + ε 4 y 5, 1992 = μ y5 + Λ 5 η + ε 5 y 6, 1992 = μ y6 + Λ 6 η + ε 6 y 7, 1992 = μ y7 + Λ 7 η + ε 7 η 1992 = μ η1992 + ζ y 1,1994 = μ y1 + η + ε 1 y 2, 1994 = μ y2 + Λ 2 η + ε 2 y 3, 1994 = μ y3 + Λ 3 η + ε 3 y 4, 1994 = μ y4 + Λ 4 η + ε 4 y 5, 1994 = μ y5 + Λ 5 η + ε 5 y 6, 1994 = μ y6 + Λ 6 η + ε 6 y 7, 1994 = μ y7 + Λ 7 η + ε 7 η 1994 = μ η1994 + ζ For Racial Egalitarianism: y 1,1984 = μ y1 + η + ε 1 y 2, 1984 = μ y2 + Λ 2 η + ε 2 y 3, 1984 = μ y3 + Λ 3 η + ε 3 y 4, 1984 = μ y4 + Λ 4 η + ε 4 y 5, 1984 = μ y5 + Λ 5 η + ε 5 y 6, 1984 = μ y6 + Λ 6 η + ε 6 y 8, 1984 = μ y8 + Λ 8 η + ε 8 η 1984 = ζ y 1,1986 = μ y1 + η + ε 1 y 2, 1986 = μ y2 + Λ 2 η + ε 2 y 3, 1986 = μ y3 + Λ 3 η + ε 3 y 4, 1986 = μ y4 + Λ 4 η + ε 4 y 5, 1986 = μ y5 + Λ 5 η + ε 5 y 6, 1986 = μ y6 + Λ 6 η + ε 6 y 8, 1986 = μ y8 + Λ 8 η + ε 8 η 1986 = μ η1986 + ζ y 1,1990 = μ y1 + η + ε 1 y 2, 1990 = μ y2 + Λ 2 η + ε 2 y 3, 1990 = μ y3 + Λ 3 η + ε 3 y 4, 1990 = μ y4 + Λ 4 η + ε 4 y 5, 1990 = μ y5 + Λ 5 η + ε 5 y 6, 1990 = μ y6 + Λ 6 η + ε 6 y 8, 1990 = μ y8 + Λ 8 η + ε 8 η 1990 = μ η1990 + ζ 66 Table 4.7: Continued y 1,1992 = μ y1 + η + ε 1 y 2, 1992 = μ y2 + Λ 2 η + ε 2 y 3, 1992 = μ y3 + Λ 3 η + ε 3 y 4, 1992 = μ y4 + Λ 4 η + ε 4 y 5, 1992 = μ y5 + Λ 5 η + ε 5 y 6, 1992 = μ y6 + Λ 6 η + ε 6 y 8, 1992 = μ y8 + Λ 8 η + ε 8 η 1992 = μ η1992 + ζ y 1,1994 = μ y1 + η + ε 1 y 2, 1994 = μ y2 + Λ 2 η + ε 2 y 3, 1994 = μ y3 + Λ 3 η + ε 3 y 4, 1994 = μ y4 + Λ 4 η + ε 4 y 5, 1994 = μ y5 + Λ 5 η + ε 5 y 6, 1994 = μ y6 + Λ 6 η + ε 6 y 8, 1994 = μ y8 + Λ 8 η + ε 8 η 1994 = μ η1994 + ζ For Gender Egalitarianism: y 1,1984 = μ y1 + η + ε 1 y 2, 1984 = μ y2 + Λ 2 η + ε 2 y 3, 1984 = μ y3 + Λ 3 η + ε 3 y 4, 1984 = μ y4 + Λ 4 η + ε 4 y 5, 1984 = μ y5 + Λ 5 η + ε 5 y 6, 1984 = μ y6 + Λ 6 η + ε 6 y 9, 1984 = μ y9 + Λ 9 η + ε 9 η 1984 = ζ y 1,1986 = μ y1 + η + ε 1 y 2, 1986 = μ y2 + Λ 2 η + ε 2 y 3, 1986 = μ y3 + Λ 3 η + ε 3 y 4, 1986 = μ y4 + Λ 4 η + ε 4 y 5, 1986 = μ y5 + Λ 5 η + ε 5 y 6, 1986 = μ y6 + Λ 6 η + ε 6 y 9, 1986 = μ y9 + Λ 9 η + ε 9 η 1986 = μ η1986 + ζ y 1,1990 = μ y1 + η + ε 1 y 2, 1990 = μ y2 + Λ 2 η + ε 2 y 3, 1990 = μ y3 + Λ 3 η + ε 3 y 4, 1990 = μ y4 + Λ 4 η + ε 4 y 5, 1990 = μ y5 + Λ 5 η + ε 5 y 6, 1990 = μ y6 + Λ 6 η + ε 6 y 9, 1990 = μ y9 + Λ 9 η + ε 9 η 1990 = μ η1990 + ζ y 1,1992 = μ y1 + η + ε 1 y 2, 1992 = μ y2 + Λ 2 η + ε 2 y 3, 1992 = μ y3 + Λ 3 η + ε 3 7y 4, 1992 = μ y4 + Λ 4 η + ε 4 y 5, 1992 = μ y5 + Λ 5 η + ε 5 y 6, 1992 = μ y6 + Λ 6 η + ε 6 y 9, 1992 = μ y9 + Λ 9 η + ε 9 η 1992 = μ η1992 + ζ y 1,1994 = μ y1 + η + ε 1 y 2, 1994 = μ y2 + Λ 2 η + ε 2 y 3, 1994 = μ y3 + Λ 3 η + ε 3 67 Table 4.8. General Egalitarianism Structural Equation Model Fit Across Time Chi-Square RMSEA AIC CAIC 1984 17.473** (6 df) .03 5.47 -34.83 1988 11.11 (7 df) .02 -2.89 -49.2 1992 31.36**** (6 df) .04 19.36 -21.5 2000 31.05**** (8 df) .08 83.21 31.22 * = p < .05 ** = p <. 01 **** = p < .0001 N = 10693 Source: American National Elections Studies, various years. Table 4.7: Continued y 4, 1994 = μ y4 + Λ 4 η + ε 4 y 5, 1994 = μ y5 + Λ 5 η + ε 5 y 6, 1994 = μ y6 + Λ 6 η + ε 6 y 9, 1994 = μ y9 + Λ 9 η + ε 9 η 1994 = μ η1994 + ζ where y 1 = SHOULD SOCIETY INSURE EQUAL OPPORTUNITY y 2 = HAVE GONE TOO FAR PUSHING EQUAL RIGHTS y 3 = BIG PROBLEM IF CHANCES NOT EQUAL y 4 = OK IF SOME HAVE MORE EQUAL CHANCE y 5 = SHOULD WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 7 = 7PT GOVT GUARANTTED JOBS SCALE y 8 = 7PT AID TO BLACKS SCALE y 9 = 7PT WOMENS EQUAL ROLE SCALE μ y1 = the sample mean for SHOULD SOCIETY INSURE EQUAL OPPORTUNITY μ y2 = the sample mean for HAVE GONE TOO FAR PUSHING EQUAL RIGHTS μ y3 = the sample mean for BIG PROBLEM IF CHANCES NOT EQUAL μ y4 = the sample mean for OK IF SOME HAVE MORE EQUAL CHANCE μ y5 = the sample mean for SHOULD WORRY LESS ABOUT EQUALITY μ y6 = the sample mean for FEWER PROBLEMS IF MORE EQUAL TREATMENT μ y7 = the sample mean for 7PT GOVT GUARANTTED JOBS SCALE μ y8 = the sample mean for 7PT AID TO BLACKS SCALE μ y9 = the sample mean for 7PT WOMENS EQUAL ROLE SCALE η = Egalitarianism (the yearly subscripts represent the subsamples) μ η = the mean of egalitarianism for each year (represented by the yearly subscripts) and the Λs are the coefficients, the εs are the error terms and ζ is the disturbance. 68 Next, I examine the results from the models with direct measures of egalitarianism, beginning with general egalitarianism. All of the models fit acceptably according to the RMSEA statistic, with the exception of the 2000 model, but only the 1988 model passes the Chi-Square test (Table 4.8). There is also a trend of weakening fit (Table 4.8). It is unsurprising that the equality of opportunity models pass all tests of fit and model fit is relatively stable over time. (Table 4.9). Table 4.10 reveals that the economic models fit acceptably according to RMSEA, but not the Chi-Square test. The evidence for increasing fit over time is mixed with the AIC increasing but the CAIC decreasing (Table 4.10). In the case of a conflict between the AIC and CAIC, I would find the CAIC more persuasive due to its superior statistical properties. The racial models again all pass the RMSEA test of fit but fail the Chi-Square test (Table 4.11). The evidence on trend is more mixed for racial egalitarianism than for previous types. The AIC again increases over time, but CAIC declines. (Table 4.11). All of the gender egalitarianism models again pass the RMSEA, but fail the Chi- Square test (Table 4.12). The evidence is again mixed, but the pattern is the opposite than that observed previously, with the AIC increasing slightly and the CAIC decreasing significantly (Table 4.12). 69 Table 4.9. Equal Opportunity Structural Equation Model Fit Across Time Chi-Square RMSEA AIC CAIC 1984 .401 (1 df) .00 -1.60 -8.32 1988 .25793 (1 df) .01 -1.07 -7.67 1992 .722 (1 df) .00 -1.28 -8.10 2000 1.198 (1 df) .007 -.801 -7.300 † = p < .10 **** = p < .0001 N = 10693 Source: American National Elections Studies, various years. Table 4.10. Economic Egalitarianism Structural Equation Model Fit Across Time Chi-Square RMSEA AIC CAIC 1972 72.97**** (14 df) .04 44.97 -51.7 1984 118.31**** (30 df) .04 58.31 -143 1988 133.21**** (37 df) .04 59.21 -186 1992 152.743**** (45 df) .03 62.743 -244 2000 99.87**** (21 df) .04 57.87 -78.6 **** = p < .0001 N = 10693 Source: American National Elections Studies, various years. 70 Table 4.11. Racial Egalitarianism Structural Equation Model Fit Across Time Chi-Square RMSEA AIC CAIC 1972 131.62**** (27 df) .04 77.62 -109 1984 106.81**** (23 df) .04 60.81 -93.7 1988 286.249**** (91 df) .03 104.25 -498 1992 277.53**** (78 df) .03 121.53 -410 2000 222**** (64 df) .04 94 -322 **** = p < .0001 N = 10693 Source: American National Elections Studies, various years. Table 4.12. Gender Egalitarianism Structural Equation Model Fit Across Time Chi-Square RMSEA AIC CAIC 1972 48.68**** (9 df) .04 30.68 -31.4 1984 136.05**** (25 df) .04 82.45 -97.6 1988 57.828**** (16 df) .04 25.83 -80 1992 356.873**** (46 df) .05 264.87 -48.8 2000 64.97**** (15 df) .04 34.97 -62.5 **** = p < .0001 N = 10693 Source: American National Elections Studies, various years. Every gay rights model fails the Chi-Square test but all pass the RMSEA test (Table 4.13). The evidence is again mixed on the question of fit over time, with the AIC stable, but the CAIC decreasing over time. 71 Table 4.13. Gay Rights Egalitarianism Structural Equation Model Fit Across Time Chi-Square RMSEA AIC CAIC 1988 65.14**** (22 df) .03 21.14 -124 1992 84.101*** (28 df) .03 28.10 -162 2000 78.79**** (28 df) .03 22.79 -159 **** = p < .0001 N = 10693 Source: American National Elections Studies, various years. Table 4.14. Immigrant Egalitarianism Structural Equation Model Fit Across Time Chi-Square RMSEA AIC CAIC 1992 129.638**** (45 df) .03 39.638 -267 2000 92.21**** (29 df) .03 34.21 -154 **** = p < .0001 N = 10693 Source: American National Elections Studies, various years. The immigrant models again pass the RMSEA test but fail the Chi-Square test (Table 4.14). Finally, the evidence is mixed on the question of whether model fit increased or decreased over time, though the CAIC revealed decreasing fit and it has more desirable statistical properties than the AIC (Table 4.14). To summarize the evidence from the direct measures, the AIC is relatively flat except for race and gender. After a large decrease in constraint in 1990s, the level of constraint for gender returns to near the point where sit started. Race has slightly 72 decreased in constraint since the mid 1980s. The CAICs for gay rights and gender are very stable. Though the economic CAIC varies more than the economic AIC, it still ends up about where it started. The CAIC for immigrant rights shows a slight decrease in constraint. Both measures show that there is little in the way of trends for economic and gender egalitarianism and gay rights. The race AIC and immigrant CAIC show slight decreases in constraint, so both measures are in large agreement in trend, though not in the level of variation. In absolute terms, the evidence from the direct measures clearly does not support the resurgence of class thesis because the level of economic constraint has not changed much since the mid 1980s. There is also a paucity of evidence for the identity politics thesis as measured by level of change in constraint. Only in the case of the decrease in the racial CAIC (showing increased constraint) is there evidence for this argument. On the question of which theory is a better explanation, there is evidence for both, with a bit more support for the identity politics thesis. There is mixed support for the economic theory, with only the CAIC (though it does have superior statistical properties than the AIC) showing an increase in economic constraint in 1992. The support for interrelationship is not as strong or robust as will be the case when I next discuss the indirect measures. It appears only in the case of the CAIC and is not as striking as that of the indirect measures. Now, turning to the models with only indirect measures of egalitarianism (the indirect models—see Chapter 3 and Appendix A for measures), the pattern for the period of time for which there are observations reinforces the earlier pattern—an increase in fit 73 Figure 1: AIC for Indirect Models Sources: National Elections Studies Youth-Parents Political Socialization Panel Study year 2000 1992 1988 1984 1982 1980 1973 1972 1964 1956 1952 AIC 600 400 200 0 -200 -400 Economic AIC Race AIC Gender AIC Civil Liberties AIC during the mid to late 1980s to the early 1990s, followed by a decline later in the 1990s (Figures 1 and 2). 7 Race and economic models tend to move in tandem with a less striking tendency for gender to move in tandem with the other two (Figures 1 and 2). For the earlier period of time, there is a spike from the mid to late 1960s to the early 1970s, followed by a decline, then another spike in the early 1980s for fit of race and economic models (Figures 1 and 2). Based on the few observations for civil liberties, it does not contribute much to the fit of egalitarianism models at all (Figures 1 and 2). There is 7 Since there were insufficient multiple indicators of domain specific egalitarianism prior to 1984, the statistics reported are the changes in the AIC or CAIC resulting from removing economic, race or gender variables from the models. For these charts, a larger number signifies increased, not decreased constraint as in the case of the direct measures. For instance, if when the economic variable is removed and the AIC or CAIC increases, that signifies more economic constraint than if they decreased when the economic variable was removed. 74 significant fluctuation, but clear trends across the time period examined here do not appear to be evident (Figures 1 and 2). Figure 2: CAIC for Indirect Models Sources: National Election Studies Youth-Parents Political Socialization Panel Study year 2000 1992 1988 1984 1982 1980 1973 1972 1964 1956 1952 CAIC 800 600 400 200 0 -200 Economic CAIC Race CAIC Gender CAIC Civil Liberties CAIC In terms of absolute change, Figures 1 and 2 do not offer much support for either the identity politics or the resurgence of class theses. The levels of constraint of economic, race and gender egalitarianism are either flat or decreasing over time. However, on the question of which is the better explanation, there is a crude pattern of economic constraint being greater than race or gender constraint until the mid 1980s, and then the pattern is reversed (except in the case of the racial AIC, which is always less constrained than the economic AIC). This offers some support to the identity politics thesis. 75 Discussion The most accurate characterization in the literature that the results presented here support is that the U. S. public perceives that equalities are interrelated (Phillips 1999). On the question of absolute change in constraint over time, the evidence is mixed on the question of the resurgence of class thesis and not robust to measurement of economic egalitarian constraint. Constraint only increases over time in the case of indirect egalitarian policy sentiment. However, no such increase is evident in the case of measures that directly mention equality. The evidence is also mixed, in the case of change of identity politics constraint over time, though the weight of it opposes the identity politics thesis. On the issue of whether resurgence of class or identity politics is the best explanation, racial and gender egalitarianism does not seem to supplant economic egalitarianism as predicted by the identity politics thesis, though based upon the limited number of observations, immigrants’ rights is more constrained than economic egalitarianism and gay rights is more constrained than economic egalitarianism until 2000. Contrary to the predictions of the identity politics thesis, economic egalitarianism does not supplant racial egalitarianism as predicted by the resurgence of class thesis, although there is evidence of economic egalitarianism becoming more constrained than racial egalitarianism after 1984. The various egalitarianisms tend to move in a loose tandem, sometimes responding to political leadership, as in the mid to late 1960s and sometimes running counter to it (or at least to the national administration) as in the early 1980s. Nor is there a clear relationship to the economic cycle, with increases in constraint 76 coming during the good economic times of the mid to late 1960s and the lean economic times of the early 1980s. The economic capabilities hypothesis, which predicts that economic egalitarianism should be more constrained in good economic times is supported about half the time, though the evidence is stronger when economic egalitarianism is measured directly and constraint is analyzed with the better statistic, the CAIC. The economic need hypothesis, which offers the reverse prediction to the one above, that economic egalitarianism should be more constrained during hard economic times, is supported about half the time. This is most likely because one indicator, economic growth, measures two opposing underlying variables, pulling in different directions. As a result, it is extremely difficult to distinguish the unique effects of each and results reveal only the net effect of the two opposing forces. One question these results do not answer is whether the egalitarianism varies among groups. It is to this question I turn in the next section. Separate Grooves for Different Groups? In the final section of this chapter, I test the “separate grooves” thesis, by investigating whether the meaning of equality varies across three politically salient groups: income, race and gender (Lee 2002, 189). One conclusion from Table 4.15 that jumps boldly from the page is that differences in means for all types of egalitarianism among all income groups are infinitesimal and statistically insignificant. Furthermore, for all types of egalitarianism, the constrained model fits better than the unconstrained one, which indicates that the 77 coefficients are not significantly different. Therefore the hypothesis that egalitarianism means different things to different income groups can be decisively rejected. However, this is not the case with respect to race. With the exception of economic egalitarianism (in a mild surprise, the difference in means are not statistically significant) all means are significantly higher for non-whites than for whites (Table 4.16). With the exception of gender, all of the unconstrained models fit better than the constrained ones, constituting evidence that the coefficients are significantly different for whites and non- whites (Table 4.16). Even in the case of gender egalitarianism two of the three measures of fit are lower for the unconstrained model (AIC and CAIC), and the RMSEA is the same. These results offer strong evidence for racial differences in the meaning of equality (Table 4.16). The earlier pattern evidenced among income groups is repeated in the case of gender differences of means, with none attaining conventional levels of statistical significance (Table 4.17). However, the verdict on model fit appears to be somewhat more mixed. The unconstrained model fits better in the case of economic egalitarianism; equally well for equality of opportunity but worse for all other types of egalitarianism. Discussion The evidence for group differences in the meaning or levels of egalitarianism is as weak among gender and income groups as it is strong among racial groups. This pattern is not entirely consistent with the group interest or identity nor does the class inversion thesis find any support from these results. In order to support the group interest and identity theory, women should be more egalitarian than men, non-whites more egalitarian 78 than whites and those with lower incomes more economic egalitarian than those with higher incomes. Although there are differences among races, as expected, there are not among men and women and those with more or less income. The class inversion thesis predicts that those with higher income will be more racial and gender egalitarian, but the results do not support those predictions, either. However, the results are consistent with the elite leadership/political culture perspective because, as predicted by this theory, there are no group differences in the level of race and gender egalitarianism and equality of opportunity. 79 Table 4.15. Means and Structural Equation Model Fit of Types of Egalitarianism by Income Model Constrained Unconstrained Measures of Fit Means Type Chi Square RMSEA AIC CAIC Low Income Middle Income High Income Chi Square RMSEA AIC CAIC General 173.13*** (37 df) .018 99.13 -207.12 .00 (.06) -.01 (.06) 0 132.77*** (17 df) .03 105.57 -35.13 Eq. Opp. 18.22 (15 df) .005 -11.77 -135.93 .00 (.095) .004 (.09) 0 17.63 (3 df) .02 .02 -13.20 Economic 1087.882* ** (76 df) .035 935.88 306.77 .00 (.068) -.03 (.066) 0 1086.43*** (48 df) .04 990.43 593.09 Race 368.800** * (70 df) .019 228.80 -350.61 .00 (.045) .014 (.044) 0 367.10*** (42 df) .03 283.10 -64.54 Gender 486.734** * (70 df) .023 346.73 -232.68 -.001 (.042) .007 (.042) 0 483.288*** (42 df) .03 399.29 51.64 *** = p < .001. N = 10693 Source: National Election Study Cumulative File, 1948-2000 m 80 m Table 4.15: Continued y 1,l = μ y1 + η + ε 1 y 2, l = μ y2 + Λ 2 η + ε 2 y 3, l = μ y3 + Λ 3 η + ε 3 y 4, l = μ y4 + Λ 4 η + ε 4 y 5, l = μ y5 + Λ 5 η + ε 5 y 6, l = μ y6 + Λ 6 η + ε 6 y 8, l = μ y8 + Λ 8 η + ε 8 η l = μ ηl + ζ y 1,m = μ y1 + η + ε 1 y 2, m = μ y2 + Λ 2 η + ε 2 y 3, m = μ y3 + Λ 3 η + ε 3 y 4, m = μ y4 + Λ 4 η + ε 4 y 5, m = μ y5 + Λ 5 η + ε 5 y 6, m = μ y6 + Λ 6 η + ε 6 y 8, m = μ y8 + Λ 8 η + ε 8 η m = μ ηm + ζ y 1,h = μ y1 + η + ε 1 y 2, h = μ y2 + Λ 2 η + ε 2 y 3, h = μ y3 + Λ 3 η + ε 3 y 4, h = μ y4 + Λ 4 η + ε 4 y 5, h = μ y5 + Λ 5 η + ε 5 y 6, h = μ y6 + Λ 6 η + ε 6 y 8, h = μ y8 + Λ 8 η + ε 8 η h = ζ 81 Table 4.16. Means and Structural Equation Model Fit of Types of Egalitarianism by Race Model Constrained Unconstrained Measures of Fit Means Type Chi Square RMSEA AIC CAIC Non-White Whites Chi Square RMSEA AIC CAIC General 746.72*** (22 df) .05 702.712 520.62 .33 (.068)*** 0 173.82*** (12 df) .03 149.82 50.49 Eq. Opp. 70.16*** (8 df) .03 54.2 -12.06 .42 (.103)*** 0 7.28 (2 df) .01 3.28 -13.27 Economic 1582.08*** (46 df) .05 1490.08 1109.32 .05 (.078) 0 761.26*** (32 df) .05 697.26 432.39 Race 639.323*** (42 df) .04 555.33 207.68 .22 (.058)*** 0 469.52*** (28 df) .04 413.52 181.76 Gender 625.87*** (42 df) .04 541.87 194.22 .14 (.058)** 0 413.76*** (28 df) .04 357.76 126 * = p < .05 **** = p < .0001 N = 10693 Source: National Election Study Cumulative File, 1948-2000. n n Equations are as follows: For General Egalitarianism: y 1,w = μ y1 + η + ε 1 y 2,w = μ y2 + Λ 2 η + ε 2 y 3,w = μ y3 + Λ 3 η + ε 3 82 Table 4.16: Continued y 4,w = μ y4 + Λ 4 η + ε 4 y 5,w = μ y5 + Λ 5 η + ε 5 y 6,w = μ y6 + Λ 6 η + ε 6 η w = ζ y 1nw = μ y1 + η + ε 1 y 2nw = μ y2 + Λ 2 η + ε 2 y 3,nw = μ y3 + Λ 3 η + ε 3 y 4,nw = μ y4 + Λ 4 η + ε 4 y 5,nw = μ y5 + Λ 5 η + ε 5 y 6,nw = μ y6 + Λ 6 η + ε 6 η nw = μ ηnw + ζ For Equal Opportunity: y 1,w = μ y1 + η + ε 1 y 3,w = μ y3 + Λ 3 η + ε 3 y 4,w = μ y4 + Λ 4 η + ε 4 y 5,w = μ y5 + Λ 5 η + ε 5 η w = ζ y 1nw = μ y1 + η + ε 1 y 3,nw = μ y3 + Λ 3 η + ε 3 y 4,nw = μ y4 + Λ 4 η + ε 4 y 5,nw = μ y5 + Λ 5 η + ε 5 η nw = μ ηnw + ζ For Economic Egalitarianism: y 1,w = μ y1 + η + ε 1 y 2,w = μ y2 + Λ 2 η + ε 2 y 3,w = μ y3 + Λ 3 η + ε 3 y 4,w = μ y4 + Λ 4 η + ε 4 y 5,w = μ y5 + Λ 5 η + ε 5 y 6,w = μ y6 + Λ 6 η + ε 6 y 7,w = μ y7 + Λ 7 η + ε 7 83 Table 4.16: Continued η w = ζ y 1nw = μ y1 + η + ε 1 y 2nw = μ y2 + Λ 2 η + ε 2 y 3,nw = μ y3 + Λ 3 η + ε 3 y 4,nw = μ y4 + Λ 4 η + ε 4 y 5,nw = μ y5 + Λ 5 η + ε 5 y 6,nw = μ y6 + Λ 6 η + ε 6 y 7,nw = μ y7 + Λ 7 η + ε 7 η nw = μ ηnw + ζ For Racial Egalitarianism: y 1,w = μ y1 + η + ε 1 y 2,w = μ y2 + Λ 2 η + ε 2 y 3,w = μ y3 + Λ 3 η + ε 3 y 4,w = μ y4 + Λ 4 η + ε 4 y 5,w = μ y5 + Λ 5 η + ε 5 y 6,w = μ y6 + Λ 6 η + ε 6 y 8,w = μ y8 + Λ 8 η + ε 8 η w = ζ y 1nw = μ y1 + η + ε 1 y 2nw = μ y2 + Λ 2 η + ε 2 y 3,nw = μ y3 + Λ 3 η + ε 3 y 4,nw = μ y4 + Λ 4 η + ε 4 y 5,nw = μ y5 + Λ 5 η + ε 5 y 6,nw = μ y6 + Λ 6 η + ε 6 y 8,nw = μ y8 + Λ 8 η + ε 8 η nw = μ ηnw + ζ 84 Table 4.16: Continued For Gender Egalitarianism: y 1,w = μ y1 + η + ε 1 y 2,w = μ y2 + Λ 2 η + ε 2 y 3,w = μ y3 + Λ 3 η + ε 3 y 4,w = μ y4 + Λ 4 η + ε 4 y 5,w = μ y5 + Λ 5 η + ε 5 y 6,w = μ y6 + Λ 6 η + ε 6 y 9,w = μ y9 + Λ 9 η + ε 8 η w = ζ y 1nw = μ y1 + η + ε 1 y 2nw = μ y2 + Λ 2 η + ε 2 y 3,nw = μ y3 + Λ 3 η + ε 3 y 4,nw = μ y4 + Λ 4 η + ε 4 y 5,nw = μ y5 + Λ 5 η + ε 5 y 6,nw = μ y6 + Λ 6 η + ε 6 y 8,nw = μ y8 + Λ 8 η + ε 8 η nw = μ ηnw + ζ where y 1 = SHOULD SOCIETY INSURE EQUAL OPPORTUNITY y 2 = HAVE GONE TOO FAR PUSHING EQUAL RIGHTS y 3 = BIG PROBLEM IF CHANCES NOT EQUAL y 4 = OK IF SOME HAVE MORE EQUAL CHANCE y 5 = SHOULD WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 7 = 7PT GOVT GUARANTTED JOBS SCALE y 8 = 7PT AID TO BLACKS SCALE y 9 = 7PT WOMENS EQUAL ROLE SCALE μ y1 = the sample mean for SHOULD SOCIETY INSURE EQUAL OPPORTUNITY 85 Table 4.16: Continued μ y2 = the sample mean for HAVE GONE TOO FAR PUSHING EQUAL RIGHTS μ y3 = the sample mean for BIG PROBLEM IF CHANCES NOT EQUAL μ y4 = the sample mean for OK IF SOME HAVE MORE EQUAL CHANCE μ y5 = the sample mean for SHOULD WORRY LESS ABOUT EQUALITY μ y6 = the sample mean for FEWER PROBLEMS IF MORE EQUAL TREATMENT μ y7 = the sample mean for 7PT GOVT GUARANTTED JOBS SCALE μ y8 = the sample mean for 7PT AID TO BLACKS SCALE μ y9 = the sample mean for 7PT WOMENS EQUAL ROLE SCALE η = Egalitarianism μ η = the mean of egalitarianism for each group (represented by the subscripts w = white, nw = nonwhite) and the Λs are the coefficients, the εs are the error terms and ζ is the disturbance. 86 Table 4.17. Means and Structural Equation Model Fit of Types of Egalitarianism by Gender Model Constrained Unconstrained Measures of Fit Means Type Chi Square RMSEA AIC CAIC Men Women Chi Square RMSEA AIC CAIC General 157.18*** (22 df) .02 113.18 -68.92 0 .06 (.05) 150.24*** (12 df) .03 126.24 26.91 Opportunity 10.53 (8 df) .00 -5.47 -71.69 0 .08 (.08) 2.35 (2 df) .00 -1.65 -18.21 Economic 2922.38*** (46 df) .07 2830.38 2449.62 0 -.06 (.06) 906.15*** (32 df) .05 842.15 577.28 Race 535.51*** (42 df) .03 451.51 103.87 0 -.02 (.04) 527.200*** (28 df) .04 471.20 239.44 Gender 493.03*** (42 df) .03 409.03 61.38 0 -.01 (.04) 481.664*** (28 df) .04 425.66 193.90 *** = p < .001 N = 10,693. Source: National Election Study Cumulative File, 1948-2000. o o For General Egalitarianism: y 1,m = μ y1 + η + ε 1 y 2,m = μ y2 + Λ 2 η + ε 2 y 3,m = μ y3 + Λ 3 η + ε 3 y 4,m = μ y4 + Λ 4 η + ε 4 y 5,m = μ y5 + Λ 5 η + ε 5 87 Table 4.17: Continued y 6,m = μ y6 + Λ 6 η + ε 6 η w = ζ y 1f = μ y1 + η + ε 1 y 2f = μ y2 + Λ 2 η + ε 2 y 3f = μ y3 + Λ 3 η + ε 3 y 4f = μ y4 + Λ 4 η + ε 4 y 5f = μ y5 + Λ 5 η + ε 5 y 6f = μ y6 + Λ 6 η + ε 6 η f = μ ηf + ζ For Equal Opportunity: y 1,m = μ y1 + η + ε 1 y 3,m = μ y3 + Λ 3 η + ε 3 y 4,m = μ y4 + Λ 4 η + ε 4 y 5,m = μ y5 + Λ 5 η + ε 5 η w = ζ y 1f = μ y1 + η + ε 1 y 3f = μ y3 + Λ 3 η + ε 3 y 4f = μ y4 + Λ 4 η + ε 4 y 5f = μ y5 + Λ 5 η + ε 5 η f = μ ηf + ζ For Economic Egalitarianism: y 1,m = μ y1 + η + ε 1 y 2,m = μ y2 + Λ 2 η + ε 2 y 3,m = μ y3 + Λ 3 η + ε 3 y 4,m = μ y4 + Λ 4 η + ε 4 y 5,m = μ y5 + Λ 5 η + ε 5 y 6,m = μ y6 + Λ 6 η + ε 6 88 Table 4.17: Continued y 7,m = μ y7 + Λ 7 η + ε 7 η w = ζ y 1f = μ y1 + η + ε 1 y 2f = μ y2 + Λ 2 η + ε 2 y 3f = μ y3 + Λ 3 η + ε 3 y 4f = μ y4 + Λ 4 η + ε 4 y 5f = μ y5 + Λ 5 η + ε 5 y 6f = μ y6 + Λ 6 η + ε 6 y 7,f = μ y7 + Λ 7 η + ε 7 η f = μ ηf + ζ For Racial Egalitarianism: y 1,m = μ y1 + η + ε 1 y 2,m = μ y2 + Λ 2 η + ε 2 y 3,m = μ y3 + Λ 3 η + ε 3 y 4,m = μ y4 + Λ 4 η + ε 4 y 5,m = μ y5 + Λ 5 η + ε 5 y 6,m = μ y6 + Λ 6 η + ε 6 y 8,m = μ y8 + Λ 8 η + ε 8 η w = ζ y 1f = μ y1 + η + ε 1 y 2f = μ y2 + Λ 2 η + ε 2 y 3f = μ y3 + Λ 3 η + ε 3 y 4f = μ y4 + Λ 4 η + ε 4 y 5f = μ y5 + Λ 5 η + ε 5 y 6f = μ y6 + Λ 6 η + ε 6 y 8,f = μ y8 + Λ 8 η + ε 8 η f = μ ηf + ζ 89 Table 4.17: Continued For Gender Egalitarianism: y 1,m = μ y1 + η + ε 1 y 2,m = μ y2 + Λ 2 η + ε 2 y 3,m = μ y3 + Λ 3 η + ε 3 y 4,m = μ y4 + Λ 4 η + ε 4 y 5,m = μ y5 + Λ 5 η + ε 5 y 6,m = μ y6 + Λ 6 η + ε 6 y 9,m = μ y9 + Λ 9 η + ε 9 η w = ζ y 1f = μ y1 + η + ε 1 y 2f = μ y2 + Λ 2 η + ε 2 y 3f = μ y3 + Λ 3 η + ε 3 y 4f = μ y4 + Λ 4 η + ε 4 y 5f = μ y5 + Λ 5 η + ε 5 y 6f = μ y6 + Λ 6 η + ε 6 y 8,f = μ y9 + Λ 9 η + ε 9 η f = μ ηf + ζ where y 1 = SHOULD SOCIETY INSURE EQUAL OPPORTUNITY y 2 = HAVE GONE TOO FAR PUSHING EQUAL RIGHTS y 3 = BIG PROBLEM IF CHANCES NOT EQUAL y 4 = OK IF SOME HAVE MORE EQUAL CHANCE y 5 = SHOULD WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 7 = 7PT GOVT GUARANTTED JOBS SCALE y 8 = 7PT AID TO BLACKS SCALE y 9 = 7PT WOMENS EQUAL ROLE SCALE μ y1 = the sample mean for SHOULD SOCIETY INSURE EQUAL OPPORTUNITY 90 Table 4.17: Continued μ y2 = the sample mean for HAVE GONE TOO FAR PUSHING EQUAL RIGHTS μ y3 = the sample mean for BIG PROBLEM IF CHANCES NOT EQUAL μ y4 = the sample mean for OK IF SOME HAVE MORE EQUAL CHANCE μ y5 = the sample mean for SHOULD WORRY LESS ABOUT EQUALITY μ y6 = the sample mean for FEWER PROBLEMS IF MORE EQUAL TREATMENT μ y7 = the sample mean for 7PT GOVT GUARANTTED JOBS SCALE μ y8 = the sample mean for 7PT AID TO BLACKS SCALE μ y9 = the sample mean for 7PT WOMENS EQUAL ROLE SCALE η = Egalitarianism μ η = the mean of egalitarianism for each group (represented by the subscripts m = male, f = female) and the Λs are the coefficients, the εs are the error terms and ζ is the disturbance. 91 Chapter 5 Egalitarianism Begins at Home Introduction In the next two chapters, I attempt to determine which mechanisms explain the socialization of egalitarianism. First, in this chapter, I analyze parental socialization, along with other control variables (See Table 5.1 for a recapitulation of Table 5.1 Theories and Hypotheses to Be Tested in Chapter 5 Theory Hypothesis Traditional View of Values (Campbell, et. al. 1960; Green, Schickler and Palmquist 2002). Symbolic Politics (Sears et. al. 1978, 1979,1980). Egalitarianism should be stable over time (more specifically, past egalitarianism should predict present egalitarianism). Parental Socialization (Jennings and Niemi 1968, 1974). Positive relationship between parents’ and child’s egalitarianism. Group Identification/Interest (Bobo, 1983, 1988, 1998, 2000; Bobo and Kluegel 1993, Bobo and Hutchins 1996; Campbell, et. al. 1960; Green, Schickler and Palmquist 2002). Women more egalitarian than men, non- whites more egalitarian than whites, people with lower income more egalitarian than people with higher income, union families more egalitarian than non-union families. Those who identify with women’s groups, Blacks, and labor unions more egalitarian than those who do not. Class Inversion (Ladd 1976-77) Those with higher income more egalitarian than those with lower income. Elite Leadership Contested Information Flow (McCloskey and Zaller 1984; Zaller 1991, 1992; Zaller and Feldman 1992) Minimal Effects (Berelson, Lazarsfeld and McPhee 1954, Dawson, Prewitt and Dawson 1977; Klapper 1960, Krause and Davis 1976, Lazarsfeld, Berelson and Gaudet 1965 [1948]) No relationship between contested egalitarianism and media exposure or political knowledge. 92 Results As Tables 5.2 and 5.3 reveal, 1973 egalitarianism is correctly signed and the only variable that has an effect on 1982 egalitarianism that is statistically different from zero. However, without the control for past egalitarianism, parents’ egalitarianism is also statistically significant, as are income and race. However, in the 1997 wave, parents’ egalitarianism is not statistically significant, when controlling for political discussion with parents (Table 5.4). The fact that this variable is correctly signed and statistically different from zero, suggests that talking politics with parents’ is one mechanism by which children learn and imitate parents egalitarian values. Unfortunately, since I was unable to control for parents’ politicization, 1 the possibility of learning through example cannot be entirely foreclosed. The lack of effect of non-parental political communication variables is reinforced by the findings in the 1997 wave (Table 5.4). Of the variables measuring identity and group interest, only gender is statistically significant, but identification with the women’s movement is not and neither is race. 2 1 Parents’ politicization rendered the polychoric correlation matrix not positive definite, making it impossible to analyze. This is often due to multi-collinearity or over-fitting. 2 In cross-sectional research, except for correctly specified simultaneous equation models, statistical models only establish association, not causation (Asher 1983, 11). For some variables such as race and gender, causation, where it exists, obviously runs from these variables to egalitarianism. For others such as group identities, media exposure and parents egalitarianism, causation could run the other way. There is little evidence in the literature of reverse causation from children to parents. It also seems much more likely that group identities cause egalitarianism for most people than the reverse because, first the causal mechanism is unclear, group memberships could not be caused by egalitarianism, are a similar construct to group identities and obtain similar results, and it seems unlikely that such an abstract concept would usually cause something so closely tied to concrete relationships and experiences as group identity. Even if only association can be shown, this is useful information because lack of association is evidence of lack of causation, which tends to falsify the hypothesis in question. Simultaneity only creates biased or inconsistent parameter estimates when an independent variable is correlated with the disturbance term (Kennedy 1998, 157). Respecifying the model by freeing the path from these variables where reverse causation might exist to the disturbance led to linear dependency, suggesting the freed path did not add anything to the model or explain any unique variance. Finally, the reason for the correlations with 93 Table 5.2. Structural Equation Coefficients for Youth Egalitarianism, 1982: Political Knowledge/Group Identification a Standardized Coefficients (Standard Errors in Parentheses) 1982 Egalitarianism 1973 Egalitarianism 1973 Egalitarianism 0.700 (.28)** - Parents’ Egalitarianism 0.049 (.234) 0.597 (.129) *** Political Knowledge -0.278 (.545) 0.309 (.495) Media Exposure 0.238 (.528) -0.094 (.429) Labor Identification -0.058 (.278) 0.164 (.242) Identification with Women’s Movement 0.061 (.149) -0.025 (.331) Identification with Blacks 0.138 (.293) -0.095 (.248) R 2 0.9502 0.5718 χ 2 = 473.180 (232 df)* RMSEA = .029 AIC = 9.180 CAIC = -1390.797 N = 769 * = p < 0.05 ** = p < .01 *** = p < .001 Source: Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined. b the disturbance term, since it represents omitted variables, is the correlation of the independent variables in the model with omitted variables (Kennedy 1998, 157). The high R 2 s and the problems of linear dependency and non-positive definite matrices in specifying the models suggests that there are few if any omitted variables that would explain any unique variance. a It was necessary to split the full model in order to render the polychoric correlation matrix positive definite. In order to conduct hypothesis tests on all coefficients, latent variable variances were fixed at one, rather than one coefficient, for all analysis models. b Equations are as follows: η 1 = Φ -1 ( Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + ζ 1 ) and η 2 = Φ -1 ( Γ 8 ξ 2 + Γ 9 ξ 3 + Γ 10 ξ 4 + Γ 11 ξ 5 + Γ 12 ξ 6 + Γ 13 ξ 7 + ζ 2 ) Where: η 1 = 1982 egalitarianism η 2 = 1973 egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = 1973 Egalitarianism 94 Table 5.3. Structural Equation Coefficients for Youth Egalitarianism, 1982: Demographics/Religiosity Standardized Coefficients (Standard Errors in Parentheses) 1982 Egalitarianism 1973 Egalitarianism 1973 Egalitarianism 0.677 (.217)*** Parents’ Egalitarianism 0.065 (.161) 0.544 (.074)*** Income 0.084 (.073) -0.170 (.080)* Gender 0.073 (.070) 0.047 (.076) Race 0.069 (.059) 0.170 (.052)*** Religiosity -0.046 (.056) 0.024 (.071) R 2 0.9198 0.4884 χ 2 = 142.677 (89 df)* RMSEA = .022 AIC = -35.322 CAIC = -572.382 N = 769 * = P < 0.05 Source: Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined. c Table 5.2: Continued ξ 2 = Parents’ Egalitarianism ξ 3 = Political Knowledge ξ 4 = Media Exposure ξ 5 = Labor Identification ξ 6 = Identification with Women’s Movement ξ 7 = Identification with Blacks the Γs are the coefficients, the ξs are dependent variables and the ζs are the disturbances. c Equations are as follows: η 1 = Φ -1 ( Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + ζ 1 ) and η 2 = Φ -1 ( Γ 7 ξ 2 + Γ 8 ξ 3 + Γ 9 ξ 4 + Γ 10 ξ 5 + Γ 11 ξ 6 + ζ 2 ) Where: η 1 = 1982 egalitarianism η 2 = 1973 egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = 1973 Egalitarianism ξ 2 = Parents’ Egalitarianism ξ 3 = Income ξ 4 = Gender ξ 5 = Race ξ 6 = Religiosity the Γs are the coefficients, the ξs are dependent variables and the ζs are the disturbances. 95 Discussion The results in this chapter offer support for the traditional perspective of values and the symbolic politics perspective. The stability of egalitarianism is evidenced by the fact that in models with a measure of past egalitarianism, it is the only variable with an effect different from zero. The hypothesis that egalitarianism is formed early in life also finds support in the statistically significant effect of parents’ egalitarianism and political discussion with parents. Support for group identification as a predictor of egalitarianism finds mixed support with race and gender significant predictors of egalitarianism, but not identification with blacks, women and organized labor. Since the effects of political knowledge, media exposure and general political discussion are not different from zero; the revisionist view of values would only be supported if information flow is a constant. The lack of effect of media exposure and political knowledge is consistent with the minimal effects and the contested elite leadership (Two-sided RAS—Receive-Accept- Sample—model) theories. Finally, respondents’ level of egalitarianism is consistent with their income, supporting a self-interest explanation. Economic egalitarianism benefits lower income people and comes at the expense of those with higher incomes. Men and variance on the age and age group variables. In the next chapter, I investigate the effect of these variables on egalitarianism utilizing the American National Election Study. 96 Table 5.4 Structural Equation Coefficients for Youth Egalitarianism, 1997 Standardized d Coefficients (Standard Errors and Test Statistics in Parentheses) Parents Egalitarianism 0.030 (.053) Political Discussion 0.283 (.058)*** Political Knowledge 0.053 (.043) Media Exposure 0.032 (.046) Identification with Women’s Movement 0.120 (.241) Race 0.006 (.046) Gender 0.138 (.13) R 2 0.389 χ 2 = 282.394 (138 df)* RMSEA = .034 AIC = 6.394 CAIC = -772.628 N = 769 * = P < 0.05 Source: Study of Political Socialization: Parent-Child Pairs Based on Survey of Youth Panel and Their Offspring, 1997. e d In order to conduct hypothesis tests on all coefficients, latent variable variances were fixed at one, rather than one coefficient. e The equation is as follows: η = Φ -1 ( Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + ζ) and where: η = Students’ Egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = Parents Egalitarianism ξ 2 = Political Discussion ξ 3 = Political Knowledge ξ 4 = Media Exposure ξ 5 = Identification with Women’s Movement ξ 6 = Race ξ 7 = Gender the Γs are the coefficients, the ξs are dependent variables and the ζ is the disturbance. 97 Chapter 6 Egalitarianism in Cohorts “…belonging to a generation is one of the lowest forms of solidarity.” (Rosenberg [1960] 1965, 244). Introduction In addition to the hypotheses tested in the previous chapter, I test hypotheses on the relationship between age and cohort and egalitarianism in this chapter. Since I am utilizing data from the American National Election Studies here, I am also able to conduct separate analyses on various types of egalitarianism: general egalitarianism, equality of opportunity, economic, racial and gender egalitarianism (See Table 6.1 for a recap and summary of additional hypotheses tested in this chapter). I begin with the results for consensus egalitaritianism: equal opportunity, then racial egalitarianism, followed by gender egalitarianism. Results for contested egalitarianism follow, with economic egalitarianism after general egalitarianism. The chapter concludes with a discussion of the results. Table 6.1. Additional Hypotheses Tested in Chapter 6 Theory Hypothesis Cohort (Peer interaction with history) Middle cohorts more egalitarian than earlier and later ones (for general, economic, race and gender egalitarianism; no difference for equality of opportunity). Age Younger more egalitarian than older ones (for general, economic, race and gender egalitarianism; no difference for equality of opportunity). 98 Results Consensus Equality Equal Opportunity The equal opportunity group membership models reveal both elite consensus and group contestation. Membership in the 1943-1958 cohort is the strongest predictor for support for equal opportunity, increasing opportunity egalitarianism by over 50 percent (Table 6.2). Belonging to the 1911-1926 or the 1959-76 increases equal opportunity egalitarianism by about one-fifth, with all other variables held constant, though belonging to the 1927-1942 cohort has no effect on level of support for equal opportunity. Democrats offer about 37 percent more support for equal opportunity than Republicans. 1 Women are more in favor of equal opportunity than men by a similar amount. Attending church every week decreases support for equal opportunity by about the same amount. 2 Non-whites support equal opportunity by about 13 percent more than whites. There is strong evidence for the power of political communication, through elite leadership and some evidence for the bottom up formation of support for equal opportunity. People who are politically knowledgeable, as well as heavy television news watchers are more egalitarian than the less politically informed and those who never watch the news on 1 I do not claim that party identification causes egalitarianism. Rather, it is included as a control variable in order to insure group identification and cohort results are not spurious. Reverse causation from party identification to egalitarianism could cause coefficients to be biased and inconsistent. (Kennedy 1998, 43- 44, 157). However, this is only a problem if party identification is correlated with the disturbance term (Kennedy 1998, 157). In a model respecification with a covariance added between party identification and the disturbance term, this model was linear dependent. In other words, this correlation did not add anything to the model, did not explain any unique variance. 2 This result is no doubt only partially due to the information flow in churches, with substantial legal and institutional limits on overt political messages. It is also partially the result of the mobilization campaign aimed at religious conservatives by Republicans and conservative groups. There are no doubt liberal denominations but the results make it clear that their socializing effect is overwhelmed by conservative churches. 99 Table 6.2. Structural Equation Coefficients for NES Equal Opportunity Group Membership Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership-Age Model Group Membership-Cohort Model South 0.1708 (0.0884)* 0.1784 (0.0742)** Union Household 0.1752 (0.0361)*** 0.1554 (0.0719)* Democratic Party Identification 0.3742 (0.0604)*** 0.3794 (0.0617)*** Age -0.0572 (0.0952) Female 0.3512 (0.1421)** 0.3584 (0.0919)*** Non-White 0.1366 (0.0462)** 0.1316 (0.0555)** Education -0.0678 (0.2213) -0.1304 (0.1863) Income 0.116 (0.306) 0.1452 (0.2656) 1911-1926 Cohort 0.2092 (0.0396)*** 1927-1942 Cohort 0.0392 (0.0981) 1943-1958 Cohort 0.5264 (0.1456)*** 1959-1976 Cohort 0.2182 (0.0944)* Church Attendance -0.3392 (0.149)* -0.3518 (0.0915)*** Watch TV News 0.3626 (0.1447)** 0.4388 (0.2615)* Read Newspaper -0.2114 (0.3152) -0.237 (0.4197) Political Knowledge 0.9854 (0.1536)*** 1.0888 (0.3678)** Political Discussion 0.4126 (0.1798)* 0.451 (0.2801) Chi-Square 55.22 (50 df) 943.1466 (71 df)*** AIC -30.8664 799.5466 CAIC -444.734 205.2332 RMSEA 0.006 0.034 R 2 0.9124 0.9638 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. a a The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ + Γ 12 ξ 12 + Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 1 and 100 television. 3 Those who talk politics a lot are more egalitarian, thought the estimate is only significant in one model, but there are no differences between newspaper readers and non-readers. Levels of support for equal opportunity are substantially the same among all levels of income and education. Replacing group memberships with group identification leaves the results largely undisturbed (Table 6.3). 4 Birth in the 1943-58 cohort again bears the strongest Table 6.2: Continued η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Γ 11 ξ 11 + Γ 12 ξ 12 Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 2 Where: η = Equal Opportunity Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Union Household ξ 3 = Democratic Party Identification ξ 4 = Age ξ 5 = Female ξ 6 = Non-White ξ 7 = Education ξ 8 = Income ξ 9 = 1911-1926 Cohort ξ 10 = 1927-1942 Cohort ξ 11 = 1943-1958 Cohort ξ 12 = 1959-1976 Cohort ξ 13 = Church Attendance ξ 14 = Watch TV News ξ 15 = Read Newspaper ξ 16 = Political Knowledge ξ 17 = Political Discussion the Γs are the coefficients and the ζs are the disturbances. 3 Since there is good reason to believe that the political communication variables bear a non-linear relationship to values, the observed responses were estimated as a function of a latent continuous normally distributed variable by a probit link function (Bartels 1993, 268; Zaller 1991,1220). As a result, interpretation of these coefficients is not straightforward. For every unit increase in the media variables, egalitarianism increases by the coefficient multiplied by the predicted probit score. The predicted probit score is the inverse z score--the predicted values for egalitarianism minus the mean divided by the standard deviation (Liao 1994, 22-24). 4 As an illustration of the difference between structural equation modeling (SEM) and ordinary least squares (OLS), and problems with using the latter in equations with summed scales rather than a latent variable, where there are multiple indicators for a variable, I attempted to replicate the model by estimating it with OLS and a summed scale dependent variable. The signs for 5 coefficients, income, education, television news, newspaper reading and political knowledge reversed. Education 101 relationship to support for equal opportunity, increasing this support by around 55 percent. Those in the 1911-1926 and 1958-1976 cohorts are around 22 or 23 percent more in favor of equal opportunity, with all other variables held constant, than the pre- 1911 baseline. Democrats are about 36 or 37 percentage points more egalitarian than Republicans, a difference similar in magnitude to that between those who never attend religious services and those who attend every week. Identifying most strongly with women’s liberation increases support for equal opportunity just a bit less. The effect of identifying strongly with blacks makes no difference, statistically. The evidence is mixed on the issue of whether identifying with the poor or labor unions increases opportunity egalitarianism. In the worse fitting analysis model, the effects of these two variables are not different from zero, but in the better fitting model, identification with labor unions increases the equal opportunity score by about 12 percent, but identifying with the poor decreases it a bit less than 20 percent. The estimate for Southerners is again that they support equal opportunity about 17 or 18 percent more than those living in other regions. 5 Political communication variables again perform strongly, with one exception, as does elite leadership and bottom up formation. Being politically knowledgeable, watching and income became significant, while television news and political knowledge became insignificant. Television news is fraught with measurement error and is correlated with education and income (Price and Zaller 1993, 135-136). Response set is always a possibility with questions in the agree-disagree format and is negatively correlated with education and income (Greenleaf 1992, 328). 5 A coding check revealed no error. This result is not due to composition effects—the South is disproportionately African American, but the addition of an interaction variable for Southern Blacks did not disturb the positive and significant relationship between Southern residence and support for equal opportunity and the interaction variable was almost never statistically significant. Support for equal opportunity is probably the result of the resolution of the conflict of group interest (the South has historically been the poorest section of the country) and ideology (it is also the most conservative section of the country). 102 Table 6.3. Structural Equation Coefficients for NES Equal Opportunity Group Identification Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Identification Age Model Group Identification Cohort Model South 0.17 (0.0561)** 0.1846 (0.0834)* Democratic Party Identification 0.3676 (0.1578)** 0.3744 (0.1176)*** Age -0.0216 (0.125) Education -0.0926 (0.1039) -0.158 (0.1337) 1911-1926 Cohort 0.2396 (0.0407)*** 1927-1942 Cohort 0.078 (0.1606) 1943-1958 Cohort 0.5512 (0.113)*** 1959-1976 Cohort 0.2258 (0.074)** Church Attendance -0.3314 (0.0711)*** -0.3426 (0.0822)*** Watch TV News 0.3446 (0.1445)** 0.427 (0.2011)* Read Newspaper -0.2144 (0.3155) -0.2396 (0.3967) Political Knowledge 0.9826 (0.2265)*** 1.091 (0.3804)** Political Discussion 0.4144 (0.1951)* 0.4584 (0.2429)* Identification with Blacks 0.1268 (0.1645) 0.1236 (0.097) Identification with Women’s Liberation 0.3486 (0.0478)*** 0.3526 (0.0782)*** Identification with Labor Unions 0.1364 (0.1426) 0.1248 (0.069)* Identification with the Poor -0.172 (0.2297) -0.1924 (0.0571)*** Chi-Square 50.832 (55 df) 746.4748 (74 df)*** AIC -59.168 598.4746 CAIC -514.422 -14.0486 RMSEA 0.0002 0.029 R 2 0.8828 0.9576 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. b b The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 1 and 103 television news and talking politics often increases support for equal opportunity, but reading the newspaper regularly or being highly educated again has no effect on level of support for equal opportunity. As expected, measures of political communication, both at the elite and mass levels, are positively related to how much people value equal opportunity. Unexpectedly, there are also differences among the major political cleavages in American society: region, political party, age group (though not age itself), religiosity, group membership and group identification. This suggests that the distinction between contested and consensus values is an oversimplification. Table 6.3: Continued η = Γ 1 ξ 1 + Γ 2 ξ + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 2 Where: η = Equal Opportunity Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Democratic Party Identification ξ 3 = Age ξ 4 = Education ξ 5 = 1911-1926 Cohort ξ 6 = 1927-1942 Cohort ξ 7 = 1943-1958 Cohort ξ 8 = 1959-1976 Cohort ξ 9 = Church Attendance ξ 10 = Watch TV News ξ 11 = Read Newspaper ξ 12 = Political Knowledge ξ 13 = Political Discussion ξ 14 = Identification with Blacks ξ 15 = Identification with Women’s Liberation ξ 16 = Identification with Labor Unions ξ 17 = Identification with the Poor the Γs are the coefficients and the ζs are the disturbances. 104 Racial Egalitarianism Table 6.4 reveals that the predictors for the group membership models for racial egalitarianism are similar to the ones for equality of opportunity, though some are not as robust. The strongest relationship is again that for the 1943-1958 cohort, which increases racial egalitarianism by about 36 percent. Those born in the 1959-1976 cohort are about 23 percent more racially egalitarian, while those in the 1911-1926 cohort are about 17 percent more egalitarian than the pre-1911 baseline, though there is no statistically significant difference in the case of the 1927-1942 cohort. The next strongest predictor is party identification, with Democrats 27 percent more racially egalitarian than Republicans. Women are about one-fifth more racially egalitarian than men and those who never attend religious services are more egalitarian than those who attend every week by a similar amount. Non-whites are about 13 percent more racially egalitarian than whites and the youngest respondents are about 8 percent more egalitarian than the oldest ones. People living in union households are 6 to 7 percent more racially egalitarian than those who do not, although the coefficient for that variable just misses statistical significance in one model. There are again no differences among levels of education or income. The effects of political communication are weaker for racial egalitarianism than for the types of egalitarianism previously examined, especially for bottom up formation. The politically knowledgeable are more racially egalitarian than those who are not and so are those who watch the news on television more often, though the estimate just misses statistical significance in one model. Reading the newspaper seems to make no difference to the level of racial egalitarianism and neither does level of political discussion. 105 Table 6.4. Structural Equation Coefficients for NES Racial Egalitarianism Group Membership Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership-Age Model Group Membership- Cohort Model South 0.0488 (0.026)* 0.0508 (0.0328) Union Household 0.071 (0.035)* 0.0602 (0.046) Democratic Party Identification 0.2708 (0.0335)*** 0.2704 (0.0322)*** Age -0.0834 (0.0338)** Female 0.2138 (0.0308)*** 0.2184 (0.0482)*** Non-White 0.1268 (0.0464)** 0.126 (0.051)** Education 0.0218 (0.0559) -0.0024 (0.0718) Income 0.0276 (0.1011) 0.0464 (0.0905) 1911-1926 Cohort 0.1654 (0.0271)*** 1927-1942 Cohort 0.083 (0.0632) 1943-1958 Cohort 0.3576 (0.0788)*** 1959-1976 Cohort 0.231 (0.0599)*** Church Attendance -0.2194 (0.0275)*** -0.2208 (0.0594)** Watch TV News 0.2314 (0.1506)* 0.2378 (0.1825) Read Newspaper -0.1306 (0.2615) -0.1394 (0.295) Political Knowledge 0.6478 (0.1954)*** 0.681 (0.2755)** Political Discussion 0.2442 (0.1735) 0.2588 (0.2171) Chi-Square 295.5585 (99df) 1775.97 (129 df)*** AIC 97.5584 1517.9702 CAIC -721.899 450.1924 RMSEA 0.0136 0.0344 R 2 0.8754 0.9292 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. c c The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ + Γ 12 ξ 12 + Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 1 and η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Γ 11 ξ 11 + Γ 12 ξ 12 Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 2 106 Replacing group memberships with group identification yields similar results (Table 6.5). Baby boomers are still the most racially egalitarian group. Those born between 1943 and 1958 are just shy of 40 percent more racially egalitarian than the pre- 1911 baseline, with all other variables held constant. Membership in the youngest cohort increases racial egalitarianism by about one-quarter, while the effect of being born in the New Deal cohort is about 20 percentage points. There is no statistical difference for the cohort growing up from the end of World II until the end of the 1950s. Democrats are about 27 percent more racially egalitarian. Those who never set foot in a church or synagogue are about one-fifth more in favor of racial equality than those who attend weekly. Women who identify strongly with women’s liberation are more racially egalitarian than those who do not by a similar magnitude. Identification with blacks increases racial egalitarianism by around 13 percentage points. Those who identify most Table 6.4: Continued Where: η = Racial Egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Union Household ξ 3 = Democratic Party Identification ξ 4 = Age ξ 5 = Female ξ 6 = Non-White ξ 7 = Education ξ 8 = Income ξ 9 = 1911-1926 Cohort ξ 10 = 1927-1942 Cohort ξ 11 = 1943-1958 Cohort ξ 12 = 1959-1976 Cohort ξ 13 = Church Attendance ξ 14 = Watch TV News ξ 15 = Read Newspaper ξ 16 = Political Knowledge ξ 17 = Political Discussion the Γs are the coefficients and the ζs are the disturbances. 107 Table 6.5. Structural Equation Coefficients for NES Racial Egalitarianism Group Identification Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Identification Age Model Group Identification Cohort Model South 0.047 (0.0436) 0.0478 (0.0518) Democratic Party Identification 0.2686 (0. 104)** 0.2682 (0.0461)** Age -0.0616 (0.0608) Education 0.015 (0.0453) -0.0146 (0.064) 1911-1926 Cohort 0.1976 (0.0393)*** 1927-1942 Cohort 0.1106 (0.0929) 1943-1958 Cohort 0.3874 (0.066)*** 1959-1976 Cohort 0.2442 (0.0507)*** Church Attendance -0.2138 (0.0278)*** -0.2164 (0.0759)** Watch TV News 0.237 (0.1463) 0.2632 (0.1761)* Read Newspaper -0.133 (0.2687) -0.1422 (0.2985) Political Knowledge 0.6402 (0.2683)** 0.6718 (0.3139)* Political Discussion 0.2426 (0.2599) 0.2586 (0.2222) Identification with Blacks 0.127 (0.0742)* 0.1266 (0.0567)* Identification with Women’s Liberation 0.202 (0.0411)*** 0.2242 (0.0608)*** Identification with Labor Unions 0.0576 (0.0384) 0.0466 (0.0653) Identification with the Poor -0.0574 (0.0925) -0.0734 (0.0265)** Chi-Square 185.2066 (104 df)*** 1223.17 (132 df) AIC -22.7932 740.648 CAIC -883.637 -133.44 RMSEA 0.0084 0.0278 R 2 0.8838 0.9212 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. d d The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 1 108 with the poor are 5 to 7 percentage points less egalitarian, though the coefficient is not statistically significant in one model. Region, age or education seems to make no difference. Elite leadership seems to have the strongest effect among the political communication variables. The political knowledgeable are more supportive of racial equality than the less well informed and those who watch the television news every day are more in favor of racial equality than those who never watch in only one of two models. There are no differences among levels of newspaper reading and political discussion. To summarize, there is evidence of the importance of both elite leadership and group identification in the formation of racial egalitarianism. Those socialized after the Table 6.5: Continued and η = Γ 1 ξ 1 + Γ 2 ξ + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 2 Where: η = Racial Egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Democratic Party Identification ξ 3 = Age ξ 4 = Education ξ 5 = 1911-1926 Cohort ξ 6 = 1927-1942 Cohort ξ 7 = 1943-1958 Cohort ξ 8 = 1959-1976 Cohort ξ 9 = Church Attendance ξ 10 = Watch TV News ξ 11 = Read Newspaper ξ 12 = Political Knowledge ξ 13 = Political Discussion ξ 14 = Identification with Blacks ξ 15 = Identification with Women’s Liberation ξ 16 = Identification with Labor Unions ξ 17 = Identification with the Poor the Γs are the coefficients and the ζs are the disturbances. 109 civil rights movement and Democrats are the most racially egalitarian groups. People who do not attend religious services or who are female or not white, or identify with women’s liberation or blacks are much more in favor of racial equality than those who are not a member of one of these groups. As well as contestation at the group level, there is consensus at the elite level, with the highly politically knowledgeable and heavy television news watchers more egalitarian than the politically unaware or the non-news watchers. There is no evidence of bottom-up formation through political discussion. Gender Egalitarianism The results for the group membership models support both the elite leadership and group identification theory in the formation of gender egalitarianism (Table 6.6). The strongest predictor of gender egalitarianism is the cohort to which respondents belong. The 1943-1958 cohort is again the most egalitarian--about 37 percentage points more egalitarian than the pre-1911 baseline. Membership in the youngest cohort raises the level of egalitarianism by about one-quarter and belonging to the 1911-1926 cohort increases gender egalitarianism by about 16 percentage points, with all other variables controlled. Class inversion theory receives no support whatsoever as there are no differences among income or education levels. The most religious are about a quarter less gender egalitarian than the least religious. There is a bit less evidence for the role of political communication and elite leadership in the case of gender egalitarianism than the other types of egalitarianism. Although those with high levels of political knowledge are more in favor of gender egalitarianism, there are no differences among levels of watching television news or reading the newspaper. There is also no evidence of bottom-up 110 Table 6.6. Structural Equation Coefficients for NES Gender Egalitarianism Group Membership Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership-Age Model Group Membership- Cohort Model South 0.0574 (0.0246)** 0.0612 (0.0304)* Union Household 0.0782 (0.0276)** 0.0692 (0.0426) Democratic Party Identification 0.2642 (0.0299)*** 0.2632 (0.037)*** Age -0.0912 (0.0375)** Female 0.223 (0.0388)*** 0.2224 (0.0436)*** Non-White 0.1008 (0.0455)* 0.096 (0.0468)* Education 0.0216 (0.0752) 0.0018 (0.0828) Income 0.0626 (0.1261) 0.0798 (0.1279) 1911-1926 Cohort 0.1634 (0.0284)*** 1927-1942 Cohort 0.0854 (0.0669) 1943-1958 Cohort 0.3768 (0.0792)*** 1959-1976 Cohort 0.2434 (0.0574)*** Church Attendance -0.2554 (0.0388)*** -0.2566 (0.0567)*** Watch TV News 0.268 (0.1412)* 0.2926 (0.2142) Read Newspaper -0.1368 (0.2632) -0.1442 (0.2944) Political Knowledge 0.6754 (0.1681)*** 0.7032 (0.2678)** Political Discussion 0.2396 (0.1657) 0.2702 (0.2155) Chi-Square 295.2944 (99 df)*** 1605.259 (129 df)*** AIC 97.2944 1347.259 CAIC -722.1633 279.4814 RMSEA 0.0136 0.0328 R 2 0.9392 0.9642 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. e e The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ + Γ 12 ξ 12 + Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 1 and η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Γ 11 ξ 11 + Γ 12 ξ 12 Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 2 111 formation of gender egalitarianism, as there are no differences between those who talk politics a lot and those who do not talk about politics at all. The group identification model results for gender egalitarianism are very similar to those of the group membership model. The largest coefficient is again for the 1943-1958 cohort, which increases gender egalitarianism by almost 40 percent, with all variables controlled (Table 6.7). Belonging to the youngest cohort increases egalitarianism by about one-fourth, while the oldest cohort is about one-fifth more egalitarian than the pre- 1911 cohort, with all other variables held constant, but the 1927-1942 cohort is again not statistically different than the baseline (Table 6.7). Democrats are again about one-quarter more gender egalitarian than Republicans (Table 6.7). People who never or seldom attend church services are more in favor of gender equality than those who attend every week by a similar magnitude. Identifying strongly with women’s liberation increases gender Table 6.6: Continued Where: η = Gender Egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Union Household ξ 3 = Democratic Party Identification ξ 4 = Age ξ 5 = Female ξ 6 = Non-White ξ 7 = Education ξ 8 = Income ξ 9 = 1911-1926 Cohort ξ 10 = 1927-1942 Cohort ξ 11 = 1943-1958 Cohort ξ 12 = 1959-1976 Cohort ξ 13 = Church Attendance ξ 14 = Watch TV News ξ 15 = Read Newspaper ξ 16 = Political Knowledge ξ 17 = Political Discussion the Γs are the coefficients and the ζs are the disturbances. 112 Table 6.7. Structural Equation Coefficients for NES Gender Egalitarianism Group Identification Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Identification Age Model Group Identification Cohort Model South 0.0358 (0.0226) 0.0542 (0.0477) Democratic Party Identification 0.2592 (0.0778)*** 0.26 (0.0611)*** Age -0.0594 (0.076) Education -0.0104 (0.0426) -0.0096 (0.0627) 1911-1926 Cohort 0.1858 (0.0305)*** 1927-1942 Cohort 0.1046 (0.1057) 1943-1958 Cohort 0.3922 (0.0667)*** 1959-1976 Cohort 0.24 (0.0508)*** Church Attendance -0.2338 (0.044)*** -0.2536 (0.0691)*** Watch TV News 0.2326 (0.1073)* 0.2722 (0.1726) Read Newspaper -0.134 (0.241) -0.1438 (0.2955) Political Knowledge 0.6278 (0.1762)*** 0.689 (0.3042)* Political Discussion 0.2444 (0.2291) 0.2706 (0.2196) Identification with Blacks 0.1398 (0.0744)* 0.1004 (0.0606)* Identification with Women’s Liberation 0.2118 (0.0241)*** 0.2286 (0.0512)*** Identification with Labor Unions 0.0736 (0.0937) 0.048 (0.0529) Identification with the Poor -0.0676 (0.1195) -0.0994 (0.0297)*** Chi-Square 283.0361 (104 df)*** 1393.224 (132 df)*** AIC 75.036 1129.2244 CAIC -785.808 36.615 RMSEA 0.0128 0.0298 R 2 0.7604 0.9456 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. f f The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 1 113 egalitarianism by about one-fifth while identifying strongly with blacks increases it by about half of that. Unlike the group membership models, the estimate for living in the South in the group identification models is not statistically different from zero and neither is the one for age, though it is close to statistical significance. Those who identify with the poor are about a tenth less gender egalitarian than those who do not in one model, although the estimate is not different in the other one. There are no differences among different levels of education. The only consistent evidence for the role of political communication is that those with a lot of political knowledge are more in favor of gender equality than those with less. Although the heavy viewers of television news are more in favor of gender equality than those who never watch the news in one model, the Table 6.7: Continued and η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 2 Where: η = Gender Egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Democratic Party Identification ξ 3 = Age ξ 4 = Education ξ 5 = 1911-1926 Cohort ξ 6 = 1927-1942 Cohort ξ 7 = 1943-1958 Cohort ξ 8 = 1959-1976 Cohort ξ 9 = Church Attendance ξ 10 = Watch TV News ξ 11 = Read Newspaper ξ 12 = Political Knowledge ξ 13 = Political Discussion ξ 14 = Identification with Blacks ξ 15 = Identification with Women’s Liberation ξ 16 = Identification with Labor Unions ξ 17 = Identification with the Poor the Γs are the coefficients and the ζs are the disturbances. 114 Table 6.8. Structural Equation Coefficients for NES General Egalitarianism Group Membership Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership-Age Model Group Membership- Cohort Model South 0.0634 (0.025)** 0.0668 (0.02912)* Union Household 0.0842 (0.0287)** 0.0718 (0.0437)* Democratic Party Identification 0.2642 (0.0287)*** 0.2676 (0.0335)*** Age -0.0792 (0.0796) Female 0.2256 (0.0358)*** 0.2262 (0.0421)*** Non-White 0.1144 (0.049)** 0.1094 (0.051)* Education 0.0054 (0.0703) -0.0162 (0.0759) Income 0.0556 (0.1252) 0.0748 (0.105) 1911-1926 Cohort 0.1588 (0.0291)*** 1927-1942 Cohort 0.067 (0.0633) 1943-1958 Cohort 0.3602 (0.0825)*** 1959-1976 Cohort 0.2204 (0.0567)*** Church Attendance -0.2456 (0.0431)*** -0.2468 (0.0569)*** Watch TV News 0.27 (0.1363)* 0.2974 (0.1861) Read Newspaper -0.143 (0.2612) -0.1522 (0.2987) Political Knowledge 0.6882 (0.1465)*** 0.7224 (0.2627)** Political Discussion 0.2622 (0.1551)* 0.2766 (0.2077) Chi-Square 238.4454 (81 df)*** 1588.043 AIC -54.6476 1372.043 CAIC -594.02 478.0894 RMSEA 0.0134 0.036 R 2 0.9168 0.9424 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. g g The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ + Γ 12 ξ 12 + Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 1 and η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Γ 11 ξ 11 + Γ 12 ξ 12 + Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 2 115 estimate just misses statistical significance in the other; there are no differences among levels of newspaper reading or political discussion. Contested Egalitarianism General Egalitarianism Table 6.8 offers support for group membership and political communication, elite leadership and bottom-up formation but the strongest effect on level of egalitarianism is the cohort to which respondents belong. Belonging to the 1942-1958 cohort increases general egalitarianism by about 36 percent and belonging to the 1959-1976 cohort increases egalitarianism by about 22 percent. Being born in the 1911-1926 cohort (the New Deal Cohort) increases egalitarianism by about 16 percent. The one cohort that does not increase egalitarianism at all is the 1927-1942 cohort, who grew up in the second half of the 1940s and the 1950s. Democrats are about 27 percent more egalitarian than Table 6.8: Continued Where: η = General Egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Union Household ξ 3 = Democratic Party Identification ξ 4 = Age ξ 5 = Female ξ 6 = Non-White ξ 7 = Education ξ 8 = Income ξ 9 = 1911-1926 Cohort ξ 10 = 1927-1942 Cohort ξ 11 = 1943-1958 Cohort ξ 12 = 1959-1976 Cohort ξ 13 = Church Attendance ξ 14 = Watch TV News ξ 15 = Read Newspaper ξ 16 = Political Knowledge ξ 17 = Political Discussion the Γs are the coefficients and the ζs are the disturbances. 116 Republicans. Those who attend church every week are about 27 percent less egalitarian than those who never do. Women are about 22 percent more egalitarian than men and non-whites about 11 percent more egalitarian than whites. Those living in a union household are about 7 or 8 percent more egalitarian than those who do not. The estimate for living in the South is about 6 percent more egalitarian than those living in other regions. One surprise is the strong performance of the political communication variables. Although regular newspaper reading made no difference, those who watch television news are estimated to be more egalitarian than those who are not, though the coefficient just misses statistical significance in one model, and the politically knowledgeable are more egalitarian than those who talk politics the least in one model and the effect just misses statistical significance in the other. The strongest predictors for general egalitarianism in the group identification models are again cohort, party, church attendance with mixed support for group identity and political communication (Table 6.9). Belonging to the cohort that grew up in the 1960s and early to mid 1970s adds about 39 percent to the egalitarianism score and Democrats are about 26 percent more egalitarian than Republicans. Membership in the cohort that has grown up since the mid 1970s or the one that grew up during the 1930s to mid 1940s raises general egalitarianism by about 20 percent. Age misses statistical significance by a wide margin. Those who attend church every week are about 25 percent less egalitarian than those who never do. The most committed feminists are about 23 percent more egalitarian than the least committed ones. A strong Black identity adds about 11 percent to general egalitarianism, this estimate is only statistically different from 117 zero in the better-fitting cohort model. 5 The effect of identifying with labor unions is not different in either model, though those who identify with the poor are estimated to be about 8 percent less egalitarian in the better-fitting model, though the estimate is not different from zero in the worse-fitting model. 6 There is more evidence of the potency of elite rather than bottom up leadership among the political communication variables, with political knowledge positively and significantly related to egalitarianism and watching television news statistically significant in one model and almost significant in the other. Reading the newspaper and talking politics are of no statistical effect. In summary, the strongest predictors for general egalitarianism are the cohorts who have grown up since the 1960s and party identification. The effect of age is non-existent. Being members of groups such as women and non-whites are also consistent predictors of general egalitarianism. Group identification did not perform as well as group membership, except in the case of identifying with women’s liberation. Those who attend church every week were consistently about 25 percent less egalitarian than those who do not attend church. Living in a union household made respondents more egalitarian, with no differences due to level of identification with unions. There are no statistical differences in egalitarianism among different levels of income or education. The level of elite leadership on this contested value is very surprising, with the effect of political 5 Identification with Blacks was positively correlated with four of the six indicators for general egalitarianism and the negative correlations tended to be much smaller than the positive ones. Since it correlated most strongly with party identification, perhaps this variable is suppressing the effect. Although the fact that it was insignificant in the age, but not the cohort model, would lead to the suspicion that age is suppressing the effect, the correlations between identification with Blacks and age were relatively small (around -0.06). 6 The possibility that the effect of identification with the poor is being suppressed by another variable is supported by the fact that it positively correlates with most of the indicators for general egalitarianism and its largest correlations are positive. Education is the likely culprit since it is the independent variable that correlates most strongly with identification with the poor (around –0.4). 118 Table 6.9. Structural Equation Coefficients for NES General Egalitarianism Group Identification Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Identification Age Model Group Identification Cohort Model South 0.06 (0.0373) 0.0624 (0.0488) Democratic Party Identification 0.264 (0.1003)** 0.2672 (0.057)*** Age -0.0508 (0.0672) Education -0.007 (0.0424) -0.031 (0.0708) 1911-1926 Cohort 0.1856 (0.0384)*** 1927-1942 Cohort 0.0892 (0.1052) 1943-1958 Cohort 0.3876 (0.0733)*** 1959-1976 Cohort 0.2228 (0.0555)*** Church Attendance -0.244 (0.0311)*** -0.246 (0.0838)** Watch TV News 0.255 (0.1379)* 0.2882 (0.193) Read Newspaper -0.1466 (0.2656) -0.1556 (0.3053) Political Knowledge 0.6866 (0.2172)*** 0.7202 (0.347)* Political Discussion 0.2646 (0.2217) 0.2804 (0.2516) Identification with Blacks 0.1194 (0.1055) 0.1132 (0.0635)* Identification with Women’s Liberation 0.2322 (0.0266)*** 0.2316 (0.0665)*** Identification with Labor Unions 0.073 (0.0939) 0.0546 (0.0646) Identification with the Poor -0.0808 (0.1283) -0.0992 (0.0312)*** Chi-Square 136.0445 (86 df)*** 1061.115 (111 df) AIC -35.9556 838.1148 CAIC -747.807 -80.674 RMSEA 0.0074 0.0284 R 2 0.8904 0.9248 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. h h The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 1 119 knowledge significant and positive in all tests and watching television news significant and positive in three of four models. However, this role seemed to be limited to television, since reading the newspaper was insignificant in all tests. Support for bottom- up formation of general egalitarianism was limited as political discussion was only significant in one of four models, though it was close in the other three models. Table 6.9: Continued and η = Γ 1 ξ 1 + Γ 2 ξ + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 2 Where: η = General Egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Democratic Party Identification ξ 3 = Age ξ 4 = Education ξ 5 = 1911-1926 Cohort ξ 6 = 1927-1942 Cohort ξ 7 = 1943-1958 Cohort ξ 8 = 1959-1976 Cohort ξ 9 = Church Attendance ξ 10 = Watch TV News ξ 11 = Read Newspaper ξ 12 = Political Knowledge ξ 13 = Political Discussion ξ 14 = Identification with Blacks ξ 15 = Identification with Women’s Liberation ξ 16 = Identification with Labor Unions ξ 17 = Identification with the Poor the Γs are the coefficients and the ζs are the disturbances. 120 Table 6.10. Structural Equation Coefficients for NES Economic Egalitarianism Group Membership Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership-Age Model Group Membership- Cohort Model South 0.0376 (0.0241) 0.0356 (0.0306) Union Household 0.0796 (0.0348)* 0.0656 (0.0446) Democratic Party Identification 0.2668 (0.0266)*** 0.2732 (0.0329)*** Age -0.095 (0.038)** Female 0.2022 (0.0284)*** 0.2086 (0.0471)*** Non-White 0.1372 (0.042)*** 0.1332 (0.0487)** Education -0.001 (0.0501) -0.0168 (0.0663) Income 0.0492 (0.1052) 0.0718 (0.0893) 1911-1926 Cohort 0.153 (0.0272)*** 1927-1942 Cohort 0.0868 (0.0541)* 1943-1958 Cohort 0.3748 (0.0872)*** 1959-1976 Cohort 0.2494 (0.0576)*** Church Attendance -0.2328 (0.0378)*** -0.235 (0.0599)*** Watch TV News 0.2484 (0.1207)* 0.2784 (0.1756) Read Newspaper -0.129 (0.2414) -0.138 (0.2808) Political Knowledge 0.6428 (0.1424) *** 0.666 (0.2639)** Political Discussion 0.24 (0.1611) 0.2568 (0.2171) Chi-Square 570.1692 2194.947 (129 df)*** AIC 370.9692 1936.947 CAIC -453.454 869.1698 RMSEA 0.0208 0.0388 R 2 0.7514 0.864 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 N = 10693 Source: National Election Study Cumulative File, 1948-98. i i The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ + Γ 12 ξ 12 + Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 1 and η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Γ 11 ξ 11 + Γ 12 ξ 12 Γ 13 ξ 13 + Φ -1 ( Γ 14 ξ 14 + Γ 15 ξ 15 + Γ 16 ξ 16 ) + ζ 2 121 Economic Egalitarianism The results from the economic egalitarianism models reveal a similar pattern to the other types of egalitarianism examined so far. In the group membership models, the strongest predictor is again being born between 1943 and 1958, which increases egalitarianism by around 37 percent (Table 6.10). A strong effect is also evident for the 1959-1976 cohort, which increases egalitarianism by almost 25 percent and the 1911- 1926 cohort, which increases it by around 15 percent. The 1927-1942 cohort is a bit less than 9 percent more egalitarian than the pre-1911 baseline, controlling for all other variables, and unlike in other models, is statistically significant in this one. Party identification is again a strong and robust predictor of egalitarianism, with Democrats estimated to be around 26 or 27 percentage points more egalitarian than Republicans. Those who never attend religious services are about 23 percent more in favor of Table 6.10: Continued Where: η = Economic Egalitarianism Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Union Household ξ 3 = Democratic Party Identification ξ 4 = Age ξ 5 = Female ξ 6 = Non-White ξ 7 = Education ξ 8 = Income ξ 9 = 1911-1926 Cohort ξ 10 = 1927-1942 Cohort ξ 11 = 1943-1958 Cohort ξ 12 = 1959-1976 Cohort ξ 13 = Church Attendance ξ 14 = Watch TV News ξ 15 = Read Newspaper ξ 16 = Political Knowledge ξ 17 = Political Discussion the Γs are the coefficients and the ζs are the disturbances. 122 Table 6.11. Structural Equation Coefficients for NES Economic Egalitarianism Group Identification Models Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Identification Age Model Group Identification Cohort Model South 0.0334 (0.0245) 0.0348 (0.0293) Democratic Party Identification 0.246 (0.0878)** 0.2648 (0.0408)*** Age -0.0848 (0.0755) Education -0.006 (0.0538) -0.0322 (0.0753) 1911-1926 Cohort 0.1726 (0.0391)*** 1927-1942 Cohort 0.1006 (0.0862) 1943-1958 Cohort 0.3868 (0.0765)*** 1959-1976 Cohort 0.2356 (0.057)*** Church Attendance -0.2332 (0.0408)*** -0.2342 (0.0742)*** Watch TV News 0.2252 (0.1008)* 0.2658 (0.1581)* Read Newspaper -0.1286 (0.2498) -0.1408 (0.2735) Political Knowledge 0.6144 (0.1775)*** 0.6616 (0.2815)** Political Discussion 0.24 (0.2284) 0.2592 (0.2204) Identification with Blacks 0.1306 (0.0602)* 0.1348 (0.0458)** Identification with Women’s Liberation 0.2058 (0.124) 0.216 (0.0491)*** Identification with Labor Unions 0.0724 (0.0565) 0.0526 (0.0655) Identification with the Poor -0.0582 (0.081) -0.0864 (0.0347)** Chi-Square 334.6222 (104 df)*** 1526.304 (132 df)*** AIC 126.6222 1262.3038 CAIC -734.222 169.694 RMSEA 0.0146 0.0316 R 2 0.7554 0.8592 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 Source: National Election Study Cumulative File, 1948-98. j j The equations are as follows: η = Γ 1 ξ 1 + Γ 2 ξ 2 + Γ 3 ξ 3 + Γ 4 ξ 4 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 1 and 123 Table 6.11: Continued N = 10693 economic equality than those who attend weekly. Women are about 20 percent more economically egalitarian than men and non-whites are about 13 percent more supportive of economic equality than whites. The youngest respondents support economic equality by about 9 or 10 percentage points more than the oldest. Southerners are estimated to support economic equality by about 3 or 4 more points than those living in other regions, though due to a higher estimate for the standard error (a measure of the diversity of opinion within the two groups), is just misses statistical significance in the group membership-cohort model. A similar is evident for those living in union households, who are estimated to favor economic equality by about 7 or 8 percentage points more than those not living in a union household, but due to a larger estimate of the within group Table 6.11: Continued η = Γ 1 ξ 1 + Γ 2 ξ + Γ 4 ξ 4 + Γ 5 ξ 5 + Γ 6 ξ 6 + Γ 7 ξ 7 + Γ 8 ξ 8 + Γ 9 ξ 9 + Γ 10 ξ 10 + Φ -1 ( Γ 11 ξ 11 + Γ 12 ξ 12 + ( Γ 13 ξ 13 + Γ 14 ξ 14 ) + Γ 15 ξ 15 + Γ 16 ξ 16 + Γ 17 ξ 17 + ζ 2 Where: η = Equal Opportunity Φ -1 = an inverse cumulative normal distribution function (an inverse z score) ξ 1 = South ξ 2 = Democratic Party Identification ξ 3 = Age ξ 4 = Education ξ 5 = 1911-1926 Cohort ξ 6 = 1927-1942 Cohort ξ 7 = 1943-1958 Cohort ξ 8 = 1959-1976 Cohort ξ 9 = Church Attendance ξ 10 = Watch TV News ξ 11 = Read Newspaper ξ 12 = Political Knowledge ξ 13 = Political Discussion ξ 14 = Identification with Blacks ξ 15 = Identification with Women’s Liberation ξ 16 = Identification with Labor Unions ξ 17 = Identification with the Poor the Γs are the coefficients and the ζs are the disturbances. 124 diversity, just misses statistical significance in the second model. As in previous types of egalitarianism examined, elite leadership and political communication is in evidence. Those who regularly watch television news (though the estimate just misses statistical significance in one model) and are politically knowledgeable are more economically egalitarian than those who do not watch the news or are not politically knowledgeable, though there are no differences among levels of newspaper reading. Bottom-up political communication is not supported by the results of Table 6.10, as those who talk politics a lot are no more supportive of economic equality than those who do not. Cohort and party identification are again the strongest predictors for economic egalitarianism in the group identification models (Table 6.11). Belonging to the 1943- 1958 cohort increases economic egalitarianism by almost 39 percent, while belonging to the youngest one increases economic egalitarianism by almost one-quarter. Being born in the 1911-1926 cohort increases economic egalitarianism by about 17 percent, but the level of support for economic equality for those growing up from the mid-1940s to the late 1950s is not statistically different from the pre-1911 baseline, with all other variables held constant. Democrats are estimated to be about one-fourth more supportive of economic equality than Republicans. The unchurched are about 25 percent more economically egalitarian than the churched. The women who identify most strongly with women’s liberation are about 21 to 22 percent more economically egalitarian than the ones who identify least with it, although the estimate just misses statistical significance in one model. Blacks who identify the most strongly with Blacks are about 13 percent more in favor economic equality than Blacks with no sense of group identity at all. The effect of identifying with labor unions is not statistically different from zero. There are no 125 differences in age, education levels or level of identification with the poor in support for economic equality. Once again, exposure to elite political communication (political knowledge and watching television news), except for reading newspapers seems to increase economic egalitarianism. However, there is little evidence for bottom up formation of economic egalitarianism through political communication in Table 6.11, with no differences among levels of political discussion. The results for economic egalitarianism might be summarized as follows. It is formed both by contestation and consensus. As expected, women and minorities, and especially those with the strongest sense of group identification, are more economically egalitarian than men and whites. It is unsurprising that Democrats are in favor of more economic equality than Republicans. Those who grew up during the Great Depression, World War II and during the 1960s to the mid-1970s are more economically egalitarian than those coming of age before the Great Depression, the Post-World War II-1950s period or after the mid-1970s. Living in a union household or identifying with labor unions only matters about half of the time and there are usually no regional differences in levels of support for economic equality. Those who do not attend religious services are much more economically egalitarian than those who attend every week. However, there is no evidence of differences between levels of income, education or identifying with the poor. Despite all of the evidence of contestation at the group level, there is also substantial evidence of elite consensus. Heavy news watchers and the most politically knowledgeable are more in favor of economic equality than those who never watch the news or who know little or nothing about politics, though there are no differences between heavy and non-newspaper readers. The results for economic equality 126 contained no evidence for bottom up political communication, with no differences among levels of political discussion. Discussion Two general propositions have guided this research: (1) group identification and interest should predict contested egalitarianism, and (2) political communication should predict consensus egalitarianism. In this section I summarize the degree to which this analysis supports these propositions. Consensus Egalitarianism The expectation that political communication and elite leadership would be involved in the formation of consensus egalitarianism is strongly supported in the case of political knowledge, which is widely thought to be the best measure of elite political communication (Price and Zaller 1993, 133, 138-139, 148; Zaller 1992, 21-22, 42-43, 292-293, 333-340). Although there is reason to believe people over report how often they read newspapers or watch the news on television and even accurate reports of these activities do not measure how much information from these sources is retained, these variables were included in the models because they are straightforward measures of exposure to political communication and as a hedge against the possibility that those with high levels of political knowledge have unobserved and unmeasured characteristics, other than exposure and retention of elite discourse, in common (Price and Zaller 1993, 135- 136). The effect of these media exposure variables on consensus egalitarianism varies from mixed to nonexistent. Of these variables, watching television news performed the strongest. It was positive and significant in every equal opportunity model and positive and significant half of the time in race and gender egalitarianism models and close to 127 significance in the other four models. There was a negative relationship in all models between reading the newspaper and consensus egalitarianism, but the large standard error, which indicates a wide diversity of opinion within levels of newspaper reading, rendered the estimates statistically significant. The role of bottom-up political communication, measured by political discussion, varied according to type of consensus egalitarianism. Political discussion performed strongest in the case of equal opportunity, being positively and significantly related to it in every model. The opposite was true in the racial egalitarianism models. Political discussion was insignificant (though close to significance in some models), though still positive in every model and it performed only slightly better in the gender models. It was only significant in one of four tests for this type of egalitarianism, though it was also close in some models. Even though there is elite consensus for these types of egalitarianism and it does matter, the results suggest that, consistent with an alternative body of empirical work and theory on group identification and interest, there is also contestation among the group divisions in society. In other words, elite consensus is not necessarily incompatible with group divisions on these types of values. For instance, people of color and women have traditionally been afforded less opportunity and have historically been denied civil and political rights. Because people often value what they have been deprived of, as expected, the results of this chapter are consistent, with the hypothesis that women and people who are not white would offer stronger support, even for types of egalitarianism, that are supported by elite consensus. In every single test, women and those women who strongly identify with women’s liberation, were more in favor of consensus egalitarianism 128 than men. A similar pattern is evident in the case of non-whites and those who identify strongest with Blacks. In every test, non-whites are more consensus egalitarian than whites. However, like the identification with women’s liberation variable, the results for identification with Blacks are weaker. This variable was insignificant in both equal opportunity models, suggesting those with the strongest Black identity are at least very ambivalent about whether equal opportunity is sufficient or whether it is anti-affirmative action, and therefore against group interest. Ambivalence was also evident in the gender equality models, with identification with Blacks, statistically significant in only one of two models, though there was a positive relationship in both. Expectedly, there was no evidence of ambivalence in the racial equality models, because identification with Blacks was positively related to support for racial equality and statistically significant in both models. Surprisingly, the group most supportive of equal opportunity, gender or racial equality is not women or minorities, but baby boomers, the group some would argue that is the most privileged in history. 7 Strong positive relationships were also found among the post-baby boom generation and the age group that grew up during the New Deal. The literature is rife with findings relating cohort to various other policy attitudes and political 7 This ranking does not include the media variables. As stated earlier, the relationships are non-linear and about all I can say about them is whether they are statistically different from zero and whether they are positive or negative. It also does not include party identification, which is included as a control variable to insure that relationships between cohort and group membership and identity and egalitarianism are not spurious. There are three theoretical perspectives that are relevant to this question. If preferences form party identification, either egalitarianism causes party identification or other preferences cause both. If group identity and interest forms party identification, and causes egalitarianism as well, causation of both is parallel. Finally, the only theoretical perspective that would predict that party identification would cause egalitarianism is the institutional one. If there is any validity to the first two perspectives, the effect of party identification on egalitarianism, is at least in part, a feedback effect. Since determining this question was not the primary purpose of this research, and I have been unable empirically to do so, the answer to this question will have to await further research. 129 values, 8 but is a bit murky on the substantive agent which cohort measures. Since what cohorts have in common is their age and the period in which they grew up and I was unable to include parents’ egalitarianism in the models in this chapter, this and the strong results for parental socialization in the previous chapter might raise the suspicion that the cohort effects found in this chapter are really parental socialization in disguise. However, since parents come from different cohorts, cohort differences due to distinctive parental socialization during different periods of time, would eventually dampen over time (Carlsson and Karlsson 1970, 717). The relatively short time covered by this study and the fact that age and cohort could not be included in the same models due to collinearity might suggest that the cohort effect could not be disentangled from an aging effect were it not for three things. First, age performs nowhere nearly as strong as cohort. The variable is significant only in two of six models, unlike cohort, which is statistically significant in every model. Second, an inspection of the cohort results makes it clear that if it is an aging effect in disguise, it is a non-linear one, rather than the familiar linear one hypothesized (that people get more conservative—or inegalitarian in this case—as they get older). Third, it is unclear why being at the age the baby boomers, post baby boomers and New Deal generation were during the time of this period would make one more egalitarian, while being at the age the post-World War II-1950s generation were during the time covered by this study would make one less so. Another clue is that whatever substantive agent cohort represents is that since media exposure, political knowledge, group membership and identification are already included in the models, whatever the cohort effect is, it works 8 See the sources cited in the literature review and hypotheses sections. 130 relatively uniformly across levels of media exposure, political knowledge and group memberships and group identification. When these clues are combined with the findings of chapters 4 and 5 that egalitarianism does not change in the aggregate over time 9 and is most strongly related to parents egalitarianism and parents’ discussion of politics with children and given the knowledge of the historical contexts during which each of these cohorts grew up, it leads to the conclusion that what the cohort effect is most likely substantively is an interaction of a peer effect with history (Davis 2004 162, Jennings 1987, 367, Ryder 1965, 847, Smith 1982a, 8; 1985, 211). 10 These results could also be viewed as too easy a test, whether post-New Deal cohorts are more egalitarian than pre-New Deal ones. Although it is necessary to exclude one category to avoid perfect collinearity and since there was good reason to believe pre- New Deal cohorts would be less egalitarian than post-New Deal ones, I chose to exclude the former. Consequently it is true that whatever the level of egalitarianism for the pre- New Deal cohort, it goes into the y intercept and the coefficients for the others are added to obtain their levels of egalitarianism, but this is also controlling for all of the other variables in the model, so the effect for being born in each of these cohorts are above and beyond the effects for the other variables in the model and the results become more impressive viewed in that light. 11 9 This does not conclusively prove it does not change at the individual level over time, but if it did during the period covered by this dissertation, the individual level changes would have to cancel out, leaving the mean the same. If this is what really occurred, it is a pretty strong coincidence, to say the least. 10 In addition to family socialization, which I have discussed above, rising levels of formal education have been suggested as accounting for cohort differences in attitudes (Ryder 1965, 845). 11 One exception to this is the post-World War II-1950s cohort, who although the estimate for their egalitarianism ranges between 3 and 11 percent more egalitarian than the pre-New Deal cohort, the 131 Another group cleavage that yields a large and persistent difference in level of support for consensus egalitarianism is religiosity. Those who never attend church are significantly more egalitarian than those who attend weekly. The substantive explanation for this is less clear. Part of the explanation might be the information flow. It is true that the overtly political content encountered in church is limited and some of that occurs in churches in liberal denominations which would likely be egalitarian in nature, but it is clear from the results that to the extent the reason why people who attend church are less egalitarian than those who do not is information flow, the egalitarian information flow is being overwhelmed by inegalitarian information flow. The results are also in part the result of the mobilization campaign directed at religious conservatives conducted by the Republican Party, but since party identification is already in the model, the effect of church attendance is in addition to that of Republican Party identification. Another factor is conservative, evangelical churches seem to be more vibrant and faster growing than the mainline churches, which are more likely to be liberal. Although the relationship for union membership or identity to consensus equality is mixed, the effect of these variables were strongest for equal opportunity and when measured by membership in a union household. The union household variable was strongest for equal opportunity—the coefficient was largest and was significant in both models. It fared less well in the gender and race models being statistically significant only half the time. The group identification variable was statistically different from zero half of the time for equal opportunity, but in none of the gender or racial equality models. There are a couple of possible explanations for this pattern of results, both having to do magnitude of the standard errors renders the estimates not statistically different from the level of egalitarianism for the pre-New Deal cohort. 132 with the existence of both cooperation and conflict between men and whites, who have historically been dominated labor unions, and women and minorities. The strong support is likely seen as a mechanism to resolve these conflicts between groups that are in coalition. Equal opportunity is likely being interpreted as supporting such policies as seniority and last-hired first-fired when they come into conflict with affirmative action. The mixed results for race and gender probably reflect the existence of both inter-group conflict and cooperation. There are some group identities and memberships for which the results are more mixed. One is region. I hypothesized that due to hierarchical, conservative political culture of the South, its residents would be less egalitarian than those residing in other regions, but the relationships were, without exception, positive, though some were not statistically significant. This is not due to the South being disproportionately African American. In unreported results, I found that the inclusion of an interaction variable for race and region (Southern Black) did not cause the sign for the Southern variable to reverse and was almost never significant. The reason why the South is not more inegalitarian than the rest of the country is probably because it has converged with it. The South is more supportive of equal opportunity than other regions is probably because it has historically been the poorest section of the country and also the most conservative. Support for equal opportunity reconciles the opposing forces of group interest and ideology. The evidence for the class inversion thesis is weak. The strongest evidence for class inversion and against group interest is that the identification with the poor variable was negatively related to all types of consensus egalitarianism, but was statistically 133 significant only half of the time. The effect of education and income were never statistically different from zero in any model. Contested Egalitarianism The results for contested egalitarianism are strikingly similar to those for consensus egalitarianism: elite consensus coexisting with group divisions in the mass public. The single strongest predictor for both general and economic egalitarianism is the same as for consensus egalitarianism: the 1943-58 cohort, with the youngest cohort not far behind. The s-curved pattern is also evident. Egalitarianism increases with the New Deal generation, decreases with the post-World War II-1950s generation, increases with the 1960s-early 1970s generation and then decreases again with the youngest cohort. The performance of age was more mixed. It was statistically insignificant in both general egalitarianism models, but younger people were more in favor of economic equality than older ones. As expected, there is strong support for the role of group identity and interest in the formation of contested egalitarianism. Democrats are substantially more in favor of contested egalitarianism than Republicans in all tests, as are women as compared to men and non-Whites compared to Whites. The results are a bit more mixed in the case of group identification variables. Those who identify strongly with women’s liberation are always more in favor of general egalitarianism than those who do not, but only half of the time for economic egalitarianism. The reverse pattern was evident in the case of those who strongly identify with Blacks. They were always more supportive of economic egalitarianism, this variable was only statistically significant in one of two general egalitarianism models, probably due to ambivalence, ambiguity and the possibility 134 that equality in general is incompatible with affirmative action and insufficient to advance group interests. The performance of labor unions as agents of socialization for equality was weaker than expected—the effect was only statistically different from zero half of the time. The results differed based upon both type of equality and measurement of the construct. Union households were always more in favor of general equality than non-union households, but only more in favor or economic equality in one of two models. There were no differences in the amount of support offered to contested egalitarianism among levels of identification with labor unions. This pattern likely reflects the traditional role of American unions of advancement rather than equality, or the conflicting interests and identities of union members (more white collar, more affluent than they used to be). As in the case of consensus equality, religiosity was also one of the strongest predictors of contested equality, being substantively large and statistically significant in all models (those who never attend religious services again more egalitarian than those who attend weekly). Group interest and class inversion seemed to battle to about a draw in the case of education, income and identification with the poor. Education and income never even came close to being statistically significant in any model. Identification with the poor was always negatively related to egalitarianism, but was never significant for economic egalitarianism and only statistically significant in one of two general equality models, perhaps reflecting the lack of class-consciousness in the United States. The biggest surprise in the results for contested egalitarianism is the performance of elite leadership and political communication. I reasoned that because economic egalitarianism was imbued with partisanship and that because general egalitarianism 135 was a weighted sum of different types of egalitarianism, some of which are popular and some of which are unpopular, there would be a two-sided information flow and each side would check the other. Even though there is a two-sided information flow, there is substantial support for the view that the egalitarian forces were winning the argument during the period of this study. As in the case of consensus egalitarianism, political knowledge was statistically significant and positively related to contested egalitarianism in every test. 12 The results for watching television news were almost as impressive. It was statistically significant and positively related to contested egalitarianism in every model but one. However, the same could not be said for reading the newspaper. This variable was negatively related to general and economic egalitarianism in every test but was never statistically different from zero, possible because newspapers are more elite oriented than television news and their readers are more polarized than television news viewers. Evidence for bottom up formation of contested egalitarianism by political communication was meager. Political discussion was always positively related to contested egalitarianism, but was statistically significant in only one-fourth of the models. It is unsurprising that one side or the other has the advantage at times in a two sided information flow, but because the relationships are non-linear, it is difficult to tell how large this advantage is. These results undermine the distinction between contested and consensus values or at least the assumption that elite consensus leads to mass consensus. Both elite leadership and group divisions were found in both consensus and contested 12 Past research usually finds a negative relationship between political knowledge and liberalism on economic issues. There was a negative bivariate relationship between political knowledge and most indicators for economic egalitarianism in my models. This suggests the possibility that one or more omitted variables were suppressing the positive relationship in past work. 136 egalitarianism. There were some close calls to be made in characterizing the types of equality, and perhaps I erred in some that I made. But even if I missed it in all cases, these results are hard to square with prior beliefs about values, because elite leadership and group division existed in all. The similar results for both types of egalitarianism might be attributed that in all of the dependent variables except one, the indicators were the same except for one. In the other, equal opportunity, four of the six indicators are the same. All models except this one might be viewed as general egalitarianism, though for instance, it is reasonable to view the construct of racial equality to consist of equal opportunity, equal treatment, equal rights and equality of results (that is the concept the aid to Blacks variable captures) and general statements about equality thrown in to capture residual concepts. But even if all dependent variables except one measure general equality, and it is a contested value, there should not be elite consensus. If I was incorrect in my categorization of it and it is really a consensus value, and if elite consensus was all that mattered in the mass public, the group divisions evident in the results should not exist. In the case of equal opportunity, three of the four indicators specifically refer to equal chances or equal opportunity, and adding the statement that refers to equality in general, was necessary to identify the measurement model, and would seem to be harmless substantively. If these indicators measure equal opportunity adequately, it is a consensus value among elites, and elite consensus is all that matters in determining how the mass public divides on the question, the group divisions should not be there. Even if I have erred in categorizing equal opportunity as a consensus value, and that is unlikely, the elite consensus should not then exist. 137 In the final chapter, I synthesize the results from the past three chapters and discuss their implications for the big picture and for future research. 138 Chapter 7 Conclusion: Only the Young For many people, the very conception of a democratic political system rests fundamentally upon processes of political persuasion (Sears and Whitney 1973, 253). In this chapter, I first review the findings of previous chapters, then discuss the implications of these findings and finally suggest avenues for further research. Findings In Chapter 4, I posed the question of the meaning of equality. The literature offers two views on each of two dimensions. On one dimension, egalitarianism runs along a continuum between several multidimensional, domain specific concepts and a unitary, symbolic concept. Although I found that the six item NES general egalitarianism scale fits the data better than the domain specific scales, the equal opportunity scale fit the data best and is the only model that would be accepted under the most conservative test of fit. In the second section of that chapter, I investigated the question of whether the meaning of equality changes over time. Due to the lack of explicit and identical measures of egalitarianism before 1984, I first conceptualized this as value constraint (similar to ideological constraint) and operationalized it by comparing model fit over time. Judged by this criterion, contrary to two views in the literature, economic egalitarianism has not supplanted by race or gender egalitarianism, nor has the reverse happened. Rather, value constraint for the three types of egalitarianism seemed to be inter-related and to move in tandem over time. Secondly, I examined the change in the mean level of egalitarianism since 1984 and found none. 139 In the final section of Chapter 4, I looked for differences in the level of egalitarianism among three politically salient groups: race, gender and income. Surprisingly, there were no differences between men and women or among any of the income groups. However, unsurprisingly, non-whites were significantly more egalitarian than whites. After addressing the meaning of equality, in Chapter 4, I began the task of explaining the level of support for equality in the next chapter. In this chapter, the picture that emerges is that children learn egalitarian values from parents, and once this occurs, the level of egalitarianism is stable over time. Beyond this, race, gender and income also predict egalitarianism but measures of political communication (other than with parents) do not. Since there was much less variation in age in the data examined in Chapter 5 than in the U. S. population and only a measure of general egalitarianism, in Chapter 6 I broadened my focus to include measures of consensus as well as contested types of egalitarianism and to include age and cohort as predictors. As expected, I found that elite discourse and political communication variables positively related to consensus egalitarianism (equal opportunity, racial and gender egalitarianism) and group membership and identification was positively related to contested egalitarianism (general and economic egalitarianism). However, contrary to my hypotheses, elite discourse and political communication variables were positively related to contested egalitarianism and group membership and identification variables were positively related to consensus egalitarianism. Political knowledge was positively and significantly related to every type of egalitarianism in every model. Television news 140 watching was always positively related to every type of egalitarianism and statistically significant three-fourths of the time. However, television news did perform stronger in the consensus models than in the contested ones. It was statistically significant in seven of eight consensus models but only eight of twelve contested models. Reading the newspaper was always negatively related to all types of egalitarianism in all models, but the coefficient was never statistically different from zero. The evidence for bottom up formation of egalitarianism through political communication was weak for consensus egalitarianism and even weaker for contested egalitarianism. Political discussion always bore a positive relationship to all types of egalitarianism but was statistically significant one-third of the time for consensus egalitarianism and only one-fourth of the time for contested egalitarianism. However, political discussion was statistically significant for every equal opportunity model. Group memberships were among the most robust predictors of egalitarianism of all kinds. People who are not white were significantly more egalitarian than whites and women were more egalitarian than men in every model. Respondents living in a union household were significantly more egalitarian in seven of ten models. Two of the three exceptions occurred in consensus models (gender and race). The evidence in the case of the strongest group identifiers was more mixed. Those who never attend religious services were always significantly more egalitarian than those who attend weekly. Identification with women’s liberation was positively related to egalitarianism in every model and statistically significant in every model but one. Black identification was always positively related to egalitarianism but was only statistically significant a bit more than two-thirds of the time. Meager evidence for the contested-consensus theory was 141 offered by this variable never being significant in the case of equal opportunity. Among the weaker variables, identification with labor unions was always positively related to egalitarianism, but was only statistically significant in one equal opportunity model. The sign for the identification with the poor variable was always negative, supporting class inversion, but not group interest, but was statistically significant only half of the time. The results for income offered support for neither theory, as it was statistically insignificant in every model, as was education. The results for regional differences were mixed, except in the case of equal opportunity, where Southerners were always more supportive. Being born in the 1943-1958 cohort (the baby boomers) bore the single strongest relationship to all types of egalitarianism. The familiar curvilinear cohort relationship that exists in the case of other political values and policy attitudes was evident here. I had expected that the pattern of egalitarianism of cohorts would form an upside down “u” over time, with younger and older cohorts being less egalitarian than middle ones. However, the pattern I found was a bit more complicated: an s-curve rather than an upside down u-curve. The pre- and post-New Deal generations were less egalitarian than the New Deal generation. The one surprise was the drop in the level of egalitarianism among the post-World War II-1950s generation. Compared with the pre-New Deal baseline, the New Deal generation (the 1911- 1926 cohort) was significantly more egalitarian, but the post-World War II-1950s cohort (born from 1927 to 1942) was not. Although not as egalitarian as the baby boomers, the youngest cohort (born from 1959 to 1976) examined in this study, was more egalitarian than the pre-1911 baseline. For scientific instrumentalists, those who seek 142 only to predict dependent variables, the inquiry would end here. On the other hand, for scientific realists, those who are interested in the substantive meaning of cohort effects that would be somewhat unsatisfying. As noted in the prior chapter, three substantive explanations have been offered for cohort effects: distinctive parental socialization, peer socialization and historical experience. Because parents come from different cohorts, distinctive parental socialization is not likely to be the best explanation (Carlsson and Karlsson 1970, 717). 1 Since the level of egalitarianism in each generation seems to follow a neat pattern, reflecting the social and political context of the times, cohort effects are most likely an interaction between peer effects and history (Beck and Jennings 1991, 743; Bengtson 1975, 368; Davis 2004, 162; Jennings 1987, 367; Jennings and Niemi 1975, 1329; Ryder 1965, 847, Smith 1985, 211; Strass and Howe 1991, 64-65). In the late 19 th and early 20 th century, during the period of conservative and Republican dominance and when Social Darwinism was the dominant ideology, people socialized during this period were less egalitarian (Burnham 1975, 342; CQ Press 2002, 40-41, 46; Galbraith [1958] 1984, 48-54; Ladd 1997, 4; Pole [1978] 1993, 264-265; Redenius 1981, 83-84, 107; Schlesinger [1986] 1999, 232-235). On the other hand, cohorts coming of age from the 1930s to the mid-1940s, when the ruling coalition included the working class and minorities and when institutions extending opportunity were created and legitimized, are more egalitarian (CQ Press 2002, 55-56; Edsall and Edsall 1919, 5, 26, 33; Hofstader 1948, 334-335; Isserman and Kazin 2000, 48; Ladd 1970, 190-192, 237-238, 303; 1976- 77, 577; Redenius 1981, 106; Romasco 1983, vii, 8-9, 11, 29, 30 ch 4, ch 9, 241, 244; Schlesinger [1986] 1999, 32, 240; Strass and Howe 1991, 271). 1 Bengtson (1975, 358) finds both at work in the formation of collectivism and individualism. 143 The institutional background from the mid-1940s to the end of the 1950s changes little. The Republican successes in capturing the Presidency during the 1950s and majorities in Congress a couple of times in the late 1940s and early 1950s leave the institutional setting largely undisturbed, other than some weakening of labor law (CQ Press 2002, 61; Isserman and Kazin 2000, 302). The Democratic Party is still the majority party and still controls the Congress for all but about four years during the period from 1945 to 1960 and controlled the Presidency about half of the time, with the wildly popular national hero Dwight D. Eisenhower occupying the White House the rest of the time (CQ Press 2002, 55, 64; Isserman and Kazin 2000, 7, 48). However, the reason for decline in egalitarianism, in this generation, seems subtler, more diffuse, rather than being a product of the institutional setting, a product of a public mood. There is a sense that the United States public had gone through two of the most serious crises in its history, the Great Depression and World War II and was suffering from crisis and reform exhaustion (CQ Press 2002, 63; Isserman and Kazin 2000, 7; Romasco 1983, vii, 241; Schlesinger [1986] 1999, 32; Stimson 2004, 23; Strass and Howe 1991, 372). Not only had prosperity returned with war production, but also there is a sense of turning inward (Isserman and Kazin 2000, 17; Stimson 2004, 23; Strass and Howe 1991, 372; 1997, 67, 162). Racial inequality begins to be contested and watered-down civil rights legislation is passed in the late 1950s, but it seems to lack the urgency of the following decade (Isserman and Kazin 2000, 30-32). There seems to be a moderate consensus that is shattered later (Isserman and Kazin 2000, 17-18; Stimson 2004, 23). The crisis of the 1960s is in some ways the mirror image of that of the 1930s, but its effect on the level of egalitarianism is similar. Poverty and racial and gender 144 inequality in an affluent society are the crises of this period (CQ Press 2002, 69; Galbraith [1958] 1984, 245, 250; Ladd 1970, 243-248; Isserman and Kazin 2000, 12, 16, 49-50, 54, 198-199). New social movements modeled on the Civil Rights movement arise and they and their demands for more equality are institutionalized in the last days of the majority Democratic coalition (Alwin 1998, 46, 51; Edsall and Edsall 1991, 35; Isserman and Kazin 2000, ch 6, 140; Redenius 1981, 129, 133; Schlesinger [1986] 1999, 32). Coming full circle, cohorts coming of age in the last quarter of the 20 th century are substantially less egalitarian than the middle cohorts. These younger cohorts were socialized in a time when the New Deal coalition was weakening and Republicans and conservatives attempted to supplant this former ruling coalition with a new ruling coalition of their own (CQ Press 2002, 78; Craig and Halfacre 1997, 64; Isserman and Kazin 2000, 294; Ladd 1997, 2; Redenius 1981, 145; Sloan 1999, xi). Institutionally, labor unions, a major player in the New Deal coalition, and along with the Democratic Party, the two significant forces supporting more economic equality, weakened as manufacturing jobs declined and changed as new members came disproportionately from the public sector (Edsall 1984, 21, ch 4; Stiglitz 2003, 183). The Democratic Party, faced with an affluent majority, faced the strategic dilemma of either moving to the center and seeking the votes of the suburban middle class or finding itself as a permanent minority (Edsall 1984, 20, 49; Edsall and Edsall 1991, 29). The acceleration of global economic competition offered economic egalitarians the dilemma of concessions to multinational corporations or the loss of jobs (Cohen 2001, 19; Edsall and Edsall 1991, 26, 217, 229; Herman 1995, 25-27). Ideologically, during this time, government is distrusted and seen as incompetent (Burnham 1975, 345; CQ Press 2002, 70; Ladd 145 1997, 5). Redistribution of wealth is on the defensive, philosophically, being seen as violating individual rights and destroying incentives for work, savings and investment (CQ Press 2002, 76; Friedman and Friedman [1979] 1980, 7, 55, 65-69, 128; Galbraith [1958] 1984, 68; Isserman and Kazin 2000, 207-208, 220; Murray 1984, 9; Wanniski 1989, 71). Formal legal racial and gender equality is achieved, leaving only the more difficult problems for which there are no easy solutions that do not threaten the interests of other groups (Edsall and Edsall 1991, 13; Gelpi, et al 1986, 1; Holtz 1995, 189; Isserman and Kazin 2000, 296; Mason and Lu 1988, 41). Poverty is feminized and race- coded (Gilens 1999, 67-68, 100-101; Edsall and Edsall 1991, 6; Isserman and Kazin 2000, 199; Yousefi 1997, 1). The results can be synthesized as follows: Although there is evidence for other meanings of equality, the strongest evidence is that it means equality of opportunity, since this is the only model that passed the Chi-Square (exact fit) test. There is no evidence from the results reported here that the level of equality changes over time. 2 However, the level of constraint does. Although the results are not conclusive, it likely is primed by political leadership. This is probably done by arguments based on need and capability. Egalitarianism was most constrained both during the economic good times of the 1960s and the lean times of the early 1980s. In the 1960s, capability was high, but the key to increasing egalitarian constraint was that political leaders and events convinced the public that need was also great (inequality in an affluent society). The situation was 2 At least not during the period of this study, but caution should be exercised in generalizing this finding to other time periods, especially times of great upheaval in American politics, such as the period from the 1960s-early 1980s not to mention the 1930s, the 1890s and the 1850s and 1860s. 146 the reverse in the early 1980s: need was great but capability was limited (by the worst recession since the Great Depression). My results and the results of other research (see Erikson, et. al. 2002, 204, for instance) suggest at least on the question of domestic policy preferences in the mass American public, if not on character and competence, Ronald Reagan’s critics carried the day. Rather than one type of egalitarianism displacing another as suggested by the identity politics or resurgence of class theses, I find that the level of egalitarian constraint of economic, gender and racial egalitarianism rise and fall in tandem. This is likely due to some combination of the race-coding and feminization of poverty and the common interests of groups allied in a political coalition. Consistent with the finding that the level of aggregate egalitarianism changes very little over time, a strong argument could be made based on these results that the best predictor of a person’s level of egalitarianism is his or her past level of egalitarianism. Not only is a person’s level of egalitarianism relatively stable over time, but it is formed by early adulthood. The most important influences in its formation are how egalitarian a person’s parents are, how much they talk politics and the times in which he or she grows up. Although group identification and interests and elite leadership are no doubt important, the results reported here suggest that egalitarian influences at home are an intervening variable. Implications In this section, I discuss the theoretical and practical implications of these findings. There has been a long debate in American politics on the capabilities of the public on two dimensions: whether the political ideas they hold immunize them from 147 elite manipulation and whether these ideas are held with sufficient flexibility to make democratic deliberation possible. 3 The picture of the American public that emerges from this research is that they are quite fixed in the values they hold, and thus, the concern over the manipulation of values is exaggerated. On the other hand, the results reported here, call into question the likelihood of democratic deliberation effecting attitude change where value conflict is implicated. However, one exception to this generalization exists in the case of the young. I find that their egalitarian values mirror that of their parents and are shaped by the historical experience of their politically formative years. Due to the limitations of existing data, I was unable to test both in the same model, so the possibility that one is spurious remains a live one. On the other hand, as stated in an earlier chapter, since parents come from differing cohorts, it is unlikely that the cohort effect is really the effect of distinctive parental socialization at different times in disguise (Carlsson and Karlsson 1970, 717). A related debate is whether and to what extent public opinion is formed from the bottom up or top down. These results support both arguments. Such 3 See Abramson and Ostrom (1991, 189-190); Achen (1975, 1229, 1231); Bartels (1993, 267); Bennett (1973, 544); Boyd (1972, 429); Brown (1970, 61); Campbell, et. al. ([1960] 1980, chs 8-10); Carmines and Stimson (1980, 177-178); Converse (1964, 255-256; 1974, 660); Converse and Markus (1979, 48); Converse, et. al. (1969, 1095-1101); Durr (1993, 159); Erikson (1979, 113); Erikson, MacKuen and Stimson (2002, xvii-xxii); Feldman (1988, 416-418); Field and Anderson (1969, 396-398); Fiorina (1981, 199); Flanagan and Zingale (2002, 137-145); Goldberg (1969, 20); Green and Palmquist (1990, 872); Green et. al. (2002, 11-13); Holm and Robinson (1978, 244); Jackson (1975, 183-184; 1983, 857-859); Judd and Milburn (1980, 627); Kessel (1972, 465); Key (1966, 7); Knight (1985, 851); Kramer (1971, 140); Krosnick (1991, 570); Lee (2002, 199-209); Levitin and Miller (1973, 751); Lippman (1993, 3-5, 29); Luttbeg (1966, 406); Markus (1979, 338); Miller et. al. (1976, 758, 767-770); Neuman (1986, 185-189); Neuman et. al. (1992, 13-19, 110-117, 119-120); Nie et. al. (chs 7-10, 18); Nie and Anderson (1974, 549- 550); Norpoth and Lodge (1985, 301); Page and Jones (1979, 122-128); Page and Shapiro (1982, 26-40; 1992, 1); Page, Shapiro and Dempsey (1987, 23); Pierce (1970, 39-42) Pierce and Rose (1974, 646); Pomper (1972, 416-420); Pomper with Lederman (1980, 55-78); Popkin (1991, 7); RePass (1971, 400); Sears and Funk (1999, 1); Sears, et. al. (1979, 370-371, 1978, 59; 1980, 671); Sears and Whitney (1973, 235-236, 257-258, 262-265); Shapiro (1969, 1118-1119); Sniderman et. al. (1986, 424-430); Stimson (1975, 393); Sullivan et. al. (1978, 233); Wleizien (1995, 981-983); Yeric and Todd (1989, 102-116); Zaller and Feldman 1992, 611-612). 148 bottom up mechanisms as the family and social groupings performed well, though political discussion did not, possibly due to the homogeneity of discussion groups. Those groups that have historically been denied equality usually value it more, although I found little evidence of class interest. Possibly this reflects the state of the world or the deficiency of existing measures. The lack of class-consciousness is a characteristic of the American polity that has been often noted. Things like income and education perhaps measure two conflicting tendencies: class interest and identity, as well as the substantive phenomena that class inversion describes (conflicts between professionals who are often more influenced by the dictates of their expertise than with making money and those in organizations whose mandate is to maximize profit as well as between lower middle class white men and minorities and women). However, the lack of class-consciousness explanation is strengthened by the weak performance of identification with the poor. This variable was often statistically insignificant and always negatively related to egalitarianism. These findings support the identity politics thesis but not the resurgence of class thesis in the context of egalitarianism. These results have implications for another longstanding debate in American Politics—over the traditional and revisionist view of values. I find egalitarianism to be consistent with the traditional view in that it is relatively stable over time. If the mix of identities of individuals are relatively stable and value formation is driven by identity, the stability observed here would be unsurprising. In summary, the implications of this study to the health of democracy is mixed. Because although I do find evidence of elite leadership, there is also robust evidence of bottom up formation of egalitarianism offering stern resistance to elite leadership 149 when it runs counter to group interest, except in the case of economic issues. This mitigates the level of concern over elite manipulation. However, egalitarianism does seem exceedingly stable over time. The implication of this is in cases of value conflict, it could be a case of an immovable force meeting an unstoppable object and will lead to people talking past each other rather than true deliberation. One way this is mitigated is that political leaders and political groups often have leeway to frame political issues in terms of more than one core value, or not to frame them in these terms at all. Further Research It would be nice to have the data to test the cohort and parental socialization theories against each other. This would require a random sample of parents and children as well as the voting age public in general. However, the difficulties in maintaining funding for existing surveys and the expense of conducting such a survey make obtaining funding an uphill battle. More likely scenarios for improved data include new variables in existing surveys and new surveys among smaller subpopulations. As well as utilizing existing surveys on Asians, Latinos and African Americans, new surveys that target immigrants specifically would be helpful. The process by which they form values and how the values they bring from their original country interact with those of their new country is greatly under- studied. Such surveys would be helpful in identifying when socialization of parents by children occurs. One deficiency in existing data that makes testing the traditional view of values as updated by Green, Palmquist and Shickler (2002) difficult is that in political science we do not measure group identification as well as they do in social psychology, for example. In this analysis, I was unable to distinguish group interest from group 150 identity with the data studied. Better measures of both would be helpful. Peer influence on value formation has also been understudied. Direct measures of this influence would be desirable. Direct measures of information flow of a variety of types of egalitarianism obtained through content analysis would also be useful. Further experiments could build upon the work of Milton Rokeach and Sandra Ball-Rokeach, by strengthening the evidence for causation in the case of the news media and identifying which specific arguments for or against equality work and which do not, beyond those that point out an inconsistency between how much someone values equality and other salient attitudes, beliefs and behaviors. In closing, I have investigated the formation of egalitarianism because it stabilizes policy preferences and offers resistance to elite discourse. I have contributed to the literature in the following ways: I find that egalitarianism is exceedingly stable in the aggregate, at least during the period covered by this dissertation, and probably during times of normal politics, in general. Egalitarianism is formed in young adulthood through two basic sources: First, by how egalitarian a person’s parents are and how much they talk politics with their children and second, by the historical context in which they grow up. Combining the empirical finding that there is a neat pattern by how cohorts differ in level of egalitarianism (pre-New Deal cohorts and the post-World War II generations less egalitarian than the 1960s-early 1970s generation (the baby boomers), with the succeeding generation (sometime called Generation X) and the New Deal generation somewhere in between, with the above findings that egalitarianism is stable and formed by young adulthood, the most plausible substantive interpretation for cohort effects is 151 peer effects interacting with historical context. Finally, I have offered evidence that there is both consensus at the elite 4 level coexisting with group divisions that are consistent with group interest at the mass level, for all types of egalitarianism that are examined. This suggests that elite discourse does not filter down uncritically to the mass public. 4 Remember that the elites I am examining here are those people who are highly informed about politics. I do not extend this claim beyond that. It seems implausible that there would be consensus about all types of egalitarianism among, for instance, political officeholders and political activists. 152 Bibliography Abramson, Paul. 1975. Generational Change in American Politics. Lexington, MA: D. C. Heath. Abramson, Paul R. and Charles W. Ostrom, Jr. 1991. “Macropartisanship: A Reassessment.” American Political Science Review 85 (1): 181-192. Abramson, Paul R. and Ronald Inglehart. 1995. Value Change in Global Perspective. Ann Arbor, MI: University of Michigan Press. Achen, Christopher H. 1975. “Mass Political Attitudes and the Survey Response.” American Political Science Review 69 (4): 1218-1231. ------1992. “Social Psychology, Demographic Variables and Linear Regression: Breaking the Iron Triangle in Voting Research.” Political Behavior 14 (3): 195-211. Acock, Alan C. and Vern L. Bengtson. 1980. “Socialization and Attribution Processes: Actual Versus Perceived Similarity Among Parents and Youth.” Journal of Marriage and the Family 42 (3): 501-514. Akaike, Hirotugu. 1987. “Factor Analysis and AIC.” Psychrometrika 52 (3): 317-332. Allsop, Dee and Herbert F. Weisberg. 1988. “Measuring Change in Party Identification in a Campaign.” American Journal of Political Science 32 (4): 996-1017. Alvarez, R. Michael and John Brehm. 1995. “American Ambivalence Towards Abortion Policy: Development of a Heteroskedastic Probit Model of Competing Values.” American Journal of Political Science 39: 1055-1082. ------1997. “Are Americans Ambivalent Towards Racial Policies?” American Journal of Political Science 41: 345-375. Alwin, Duane F. 1993. “Attitude Development in Adulthood: The Role of Generational and Life Cycle Factors.” In New Directions in Attitude Measurement. Dagmar Krebs and Peter Schmidt. Eds. New York: Walter de Gruyter. ------1997. “Aging, Social Change and Conservatism.” In Studying Aging and Social Change: Conceptual and Methodological Issues. Melissa A. Hardy. Ed. Thousand Oaks, CA: Sage. ------1998. “The Political Impact of the Baby Boom: Are There Persistent Generational Differences in Political Beliefs and Behavior?” Generations 22 (1): 46-54. Alwin, Duane and Jon Krosnik. 1986. “The Measurement of Values in Surveys: A Comparison of Ratings and Rankings.” Public Opinion Quarterly 49 (4): 535-552. 153 Andrews, Frank M. 1984. “Construct Validity and Error Components of Survey Measures: A Structural Modeling Approach.” Public Opinion Quarterly 48 (2): 409-442. APSA Task Force on Inequality and American Democracy. 2004. “American Democracy in an Age of Rising Inequality.” Perspective on Politics 2 (4): 651-666. Aronowitz, Stanley. 1992. The Politics of Identity: Class, Culture, Social Movements. New York: Routledge. Asher, Herbert B. 1974. “Some Consequences of Measurement Error in Survey Data.” American Journal of Political Science 18 (2): 469-485. ------1983. Causal Modeling. 2 nd Edition. Newbury Park, CA: Sage Publications. Bachman, Jerald G. and Patrick M. O’Malley. 1984. “Yea-Saying, Nay-Saying, and Going to Extremes: Black-White Differences in Response Styles.” Public Opinion Quarterly 48 (2): 491-509. Ball-Rokeach, Sandra J., Milton Rokeach and Joel W. Grube. 1984. The Great American Values Test. New York: Free Press. Baltes, P. B. 1968. “Longitudinal and Cross-Sectional Sequences in the Study of Age and Generation Effects.” Human Development 11:145-171. Bartels, Larry M. 1993. “Messages Received: The Political Impact of Political Exposure.” American Political Science Review 87 (2): 267-285. Beck, Paul Allen. 1977. “Partisan Dealignment in the Postwar South.” American Political Science Review 71 (2): 477-496. Beck, Paul Allen and M. Kent Jennings. 1991. “Family Traditions, Political Periods, and the Development of Partisan Orientations.” Journal of Politics 53 (3): 742-763. Beck, Paul Allen, Russell J. Dalton, Steven Greene and Robert Huckfeldt. 2002. “The Social Calculus of Voting: Interpersonal, Media and Organizational Influences on Presidential Choices.” American Political Science Review 96 (1): 57-73. Bell, Daniel. 1979. “On Meritocracy and Equality.” In The New Egalitarianism: Questions and Challenges. David Lewis Schaefer. Ed. Port Washington, N. Y.: Kennikat Press. Bengtson, Vern L. 1975. “Generation and Family Effects in Value Socialization.” American Sociological Review 40 (3): 358-371. 154 Beninger, James R. 1986. The Social Control Revolution. Cambridge: Harvard University Press. Bennett, Stephen Earl. 1973. “Consistency Among the Public’s Social Welfare Policy Attitudes in the 1960s.” American Journal of Political Science 17 (3): 544-570. Bentler, Peter M. 1983. “Simultaneous Equation Systems as Moment Structure Models: With an Introduction to Latent Variable Models.” Journal of Econometrics 22: 13-42. Bentler, P. M. and Douglas G. Bonett. 1980. “Significance Tests and Goodness of Fit in the Analysis of Covariance Structures.” Psychological Bulletin 88 (3): 588-606. Berelson, Bernard R., Paul F. Lazarsfeld and William N. McPhee. 1954. Voting: A Study of Opinion Formation in a Presidential Campaign. Chicago: University of Chicago Press. Berry, William D. 1984. Nonrecursive Causal Models. Thousand Oaks: Sage Publications. Bishop, George F., Andrew M. Barclay and Milton Rokeach. 1972. “Presidential Preferences and Freedom-Equality Value Patterns in the 1968 American Campaign.” Journal of Social Psychology 88: 207-212. Bishop, George F., Alfred J. Tuchfarber and Robert W. Oldendick. 1978. “Change in the Structure of American Political Attitudes: The Nagging Question of Question Wording.” American Journal of Political Science 22 (2): 250-269. Blumer, Herbert. 1948. “Public Opinion and Public Opinion Polling.” American Sociological Review 13 (5): 542-549. Bobo, Lawrence. 1983. “Whites’ Opposition to Busing: Symbolic Racism or Realistic Group Conflict?” Journal of Personality and Social Psychology 45 (6): 1196- 1210. ------1988. “Group Conflict, Prejudice, and the Paradox of Contemporary Racial Attitudes.” in Eliminating Racism: Profiles in Controversy. Phyllis A. Katz and Dalmas Taylor. Eds. New York: Plenum Press. ------1991. “Social Responsibility, Individualism, and Redistributive Policies.” Sociological Forum 6 (1): 71-92. ------1998. “Race, Interests, and Beliefs About Affirmative Action: Unanswered Questions and New Directions.” American Behavioral Scientist 41 (7): 985-1003. 155 ------ 2000. “Race and Beliefs about Affirmative Action: Assessing the Effects of Interests, Group Threat, Ideology, and Racism.” In Racialized Politics: The Debate about Racism in America. David O. Sears, Jim Sidanius and Lawrence Bobo. Eds. Chicago: University of Chicago Press. Bobo, Lawrence and James R. Kluegel. 1993. “Opposition to Race-Targeting: Self- Interest, Stratification Ideology, or Racial Attitudes.” American Sociological Review 58 (4): 443-464. Bobo, Lawrence, James R. Kluegel and Ryan A. Smith. 1997. “Laissez-Faire Racism: The Crystallization of a Kinder, Gentler, Antiblack Ideology.” In Racial Attitudes in the 1990s: Continuity and Change. Steven A. Tuch and Jack K. Martin. Eds. Westport, CN: Praeger. Bobo, Lawrence and Vincent L. Hutchings. 1996. “Perceptions of Racial Group Competition: Extending Blumer's Theory of Group Position to a Multiracial Social Context.” American Sociological Review 61 (6): 951-972. Bogart, Leo. 1972. Silent Politics: Polls and the Awareness of Public Opinion. New York: Wiley-Interscience. Bollen, Kenneth and Richard Lennox. 1991. “Conventional Wisdom on Measurement: A Structural Equation Perspective.” Psychological Bulletin 110 (2): 305-314. Borrooah, Vani K. 2002. Logit and Probit: Ordered and Multinomial Models. Thousand Oaks, CA: Sage Publications. Bourdieu, Pierre. 1979. “Public Opinion Does Not Exist.” M. C. Axtmann. Trans. In Communication and Class Struggle. Vol. 1: Capitalism, Imperialism. Armand Mattelart and Seth Siegelaub. Eds. New York: International General. Boyd, Richard W. 1972. “Popular Control of Public Policy: A Normal Vote Analysis of the 1968 Election.” American Political Science Review 66 (2): 429-449. Bozdogan, Hamparsum. 1987. “Model Selection and Akaike’s Information Criterion (AIC): The General Theory and Its Analytical Extensions.” Psychrometrika 52 (3): 345-370. Brady, Henry E. and Gary R. Orren. 1992. “Polling Pitfalls: Sources of Error in Public Opinion Surveys.” in Media Polls in American Politics. Thomas E. Mann and Gary R. Orren. Eds. Washington, D. C.: Brookings Institution. Braungart, Richard G. 1971. “Family Status, Socialization and Student Politics: A Multivariate Analysis.” American Journal of Sociology 77 (1): 108-130. 156 Brehm, John. 1993. The Phantom Respondents: Opinion Surveys and Political Representation. Ann Arbor, MI: University of Michigan Press. Brewer, Paul R. 2003a. “The Shifting Foundations of Public Opinion about Gay Rights.” Journal of Politics 65 (4): 1208-1220. Brewer, Paul R. 2003b. “Values, Political Knowledge and Public Opinions on Gay Rights: A Framing-Based Account.” Public Opinion Quarterly 67 (2): 173-201. Brink, William and Louis Harris. [1966] 1967. Black and White: A Study of U. S. Racial Attitudes Today. New York: Simon and Schuster. Brody, Richard A. 1991. Assessing the President: The Media, Elite Opinion, and Public Support. Stanford, CA: Stanford University Press. Brody, Richard A. and Lawrence S. Rothenberg. 1988. “The Instability of Partisanship: An Analysis of the 1980 Presidential Election.” British Journal of Political Science 18 (4): 445-465. Brown, Stephen R. “Consistency and Persistence of Ideology: Some Experimental Results.” Public Opinion Quarterly 34 (1): 60-68. Browne, Michael W. and Robert Cudeck. 1993. “Alternative Ways of Assessing Model Fit.” In Testing Structural Equation Models. Kenneth A. Bollen and J. Scott Long. Eds. Newbury Park, CA: Sage. Burnham, Walter Dean. 1975. “American Politics in the 1970s: Beyond Party.” In The American Party Systems: Stages of Political Development. New York: Oxford University Press. Byrne, Barbara M. 1994. Structural Equation Modeling with EQS and EQS Windows: Basic Concepts, Applications and Programming. Thousand Oaks, CA: Sage. Campbell, Angus. 1971. White Attitudes Toward Black People. Ann Arbor: Institute for Social Research. Campbell, Angus, Gerald Gurin and Warren E. Miller. 1954. The Voter Decides. Evanston, IL: Row, Peterson. Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes. [1960] 1980. The American Voter. Unabridged Ed. Midway Reprint. Chicago: University of Chicago Press. Carlsson, Gosta and Katarina Karlsson. 1970. “Age, Cohorts, and the Generation of Generations.” American Sociological Review 35 (4): 973-988. 157 Carmines, Edward G. and James A. Stimson. “The Two Faces of Issue Voting.” American Political Science Review 74 (1): 78-91. ------1981. “Issue Evolution, Population Replacement, and Normal Partisan Change.” American Political Science Review 75 (1): 107-118. ------1984. “The Dynamics of Issue Evolution: The United States.” In Electoral Change in Advanced Industrial Societies: Realignment or Dealignment. Russell J. Dalton, Scott C. Flanagan and Paul Allen Beck. Eds. Princeton, NJ: Princeton University Press. ------1986. “On the Structure and Sequence of Issue Evolution.” American Political Science Review 80 (3): 901-920. Carmines, Edward G. and Paul Sniderman. 2002. “The Structure of Racial Attitudes: Issue Pluralism and the Changing American Dilemma.” In Understanding Public Opinion. 2 nd ed. Barbara Norrander and Clyde Wilcox. Eds. Clark, Terry Nichols. 1998. “Overview of the Book.” In The New Political Culture. Terry Nichols Clark and Vincent Hoffman-Martinot. Eds. Boulder, CO: Westview. Clark, Terry Nichols and Ronald Inglehart. 1998. “The New Political Culture: Changing Dynamics of Support for the Welfare State and Other Policies in Postindustrial Societies.” In The New Political Culture. Terry Nichols Clark and Vincent Hoffman-Martinot. Eds. Boulder, CO: Westview. Clark, Terry Nichols and Seymour Martin Lipset. 1991. “Are Social Classes Dying?” International Sociology 4: 397-410. Chafez, Janet Saltzman and Helen Rose Fuchs Ebaugh. 1983. “Growing Conservatism in the United States: An Examination of Trends in Political Opinion Between 1972 and 1980.” Sociological Perspectives 26 (3): 275-298. Cherlin, Andrew and Pamela Barnhouse Walters. 1981. “Trends in United States Men’s and Women’s Sex Roles: 1972 to 1978.” American Sociological Review 46 (4): 453-460. Christenson, James A. and Riley E. Dunlap. 1984. “Freedom and Equality in American Political Ideology: Race and Gender Preferences.” Social Science Quarterly 65 (3): 861-867. Cohen, Edward S. 2001. The Politics of Globalization in the United States. Washington, DC: Georgetown University Press. Condran, John G. 1979. “Changes in White Attitudes Toward Blacks: 1963-1977.” Public Opinion Quarterly 43 (4): 463-476. 158 Connell, R. W. 1972. “Political Socialization in the American Family: The Evidence Re- Examined.” Public Opinion Quarterly 36 (3): 323-333. Conover, Pamela Johnston. 1988. “Feminists and the Gender Gap.” Journal of Politics 50 (4): 985-1010. Conover, Pamela Johnston and Stanley Feldman. “The Origins and Meanings of Liberal/Conservative Self Identification.” American Journal of Political Science 25 (4): 617-645. Converse, Philip E. 1964. “The Nature of Belief Systems in Mass Publics.” In Ideology and Discontent. David E. Apter. Ed. New York: Free Press. ------1974. “Nonattitudes and American Public Opinion: Comment: The Status of Nonattitudes.” American Political Science Review 68(2): 650-660. ------1975. “Public Opinion and Voting Behavior.” In Handbook of Political Science. Vol. 4. Fred I. Greenstein and Nelson W. Polsby. Eds. Reading, MA: Addison- Wesley. ------ 1976. The Dynamics of Party Support: Cohort-Analyzing Party Identification. Beverly Hills: Sage Publications. ------1987. “Changing Conceptions of Public Opinion in the Political Process.” Public Opinion Quarterly 51 (4 pt. 2): S12-S24. ------1996. “The Advent of Polling and Political Representation.” PS: Political Science and Politics 29 (4): 649-657. Converse, Philip E. and Gregory B. Markus. 1979. “Plus ca change…: The New CPS Election Study Panel.” American Political Science Review 73: 32-49. Converse, Philip E. and Roy Pierce. 1985. “Measuring Partisanship.” Political Methodology 11 (3-4): 143-166. Converse, Philip E., Warren E. Miller, Jerrold G. Rusk and Arthur C. Wolfe. 1969. “Continuity and Change in American Politics: Parties and Issues in the 1968 Election.” American Political Science Review 63 (4): 1083-1105. Cook, Elizabeth Adell and Clyde Wilcox. 1991. “Feminism and the Gender Gap—A Second Look.” Journal of Politics 53 (4): 1111-1122. Coverdill, James E. 1997. “White Ethnic Identification and Racial Attitudes.” In Racial Attitudes in the 1990s: Continuity and Change. Steven A. Tuch and Jack K. Martin. Eds. Westport, CN: Praeger. 159 CQ Press. 2002. Presidential Elections, 1789-2000. Washington, D. C.: CQ Press. Craig, Stephen C. and Angela C. Halfacre. 1997. “Political Issues and Political Choice: Generations and the Potential for Realignment in American Politics.” In After the Boom: The Politics of Generation X. Lanham: Rowan and Littlefield. Crittenden, John. 1962. “Aging and Party Affiliation.” Public Opinion Quarterly 26 (4): 648-657. Cutler, Neal E. 1969-1970. “Generation, Maturation and Party Affiliation: A Cohort Analysis.” Public Opinion Quarterly 33 (4): 583-588. Cunningham, William H., Isabella C. M. Cunningham and Robert T. Green. 1977. “The Ipsative Process to Reduce Response Set Bias.” Public Opinion Quarterly 41 (3): 379-384. Curran, Patrick J., Stephen G. West and John J. Finch. 1996. “The Robustness of Test Statistics to Nonnormality and Specification Error in Confirmatory Factor Analysis.” Psychological Methods 1 (1): 16-29. Dahl, Robert A. 2006. On Political Equality. New Haven, CT: Yale University Press. Dalton, Russell J. 1977. “Was There a Revolution: A Note on Generational Versus Life- Cycle Explanations of Value Differences.” Comparative Political Studies 9 (4): 459-471. ------1980. “Reassessing Parental Socialization: Indicator Unreliability Versus Generational Transfer.” American Political Science Review 74 (2): 421-431. ------ 1982. “The Pathways of Parental Socialization.” American Politics Quarterly 10 (2): 139-157. Dalton, Russell J, Manfred Keuchler and Wilhelm Bürklin. 1990. “The Challenge of the New Movements.” In Challenging the Political Order: New Social and Political Movements in Western Democracies. Russell J. Dalton and Manfred Kuechler. Eds. New York: Oxford University Press. Davies, James C. 1965. “The Family’s Role in Political Socialization.” The Annals of the American Academy of Political and Social Science 361: 11-19. Davis, James A. 1980. “Conservative Weather in a Liberalizing Climate: Change in Selected NORC General Social Survey Items, 1972-1978.” Social Forces 58 (4): 1129-1156. 160 ------1992. “Changeable Weather in a Cooling Climate Atop the Liberal Plateau: Conversion and Replacement in Forty-Two General Social Survey Items, 1972- 1989.” Public Opinion Quarterly 56 (3): 261-306. ------2004. “Did Growing Up in the 1960s Leave a Permanent Mark? Evidence from the General Social Survey.” Public Opinion Quarterly 68 (2): 161-183. Dawson, Richard E., Kenneth Prewitt and Karen S. Dawson. 1977. Political Socialization: An Analytic Study. 2 nd Ed. Boston: Little, Brown and Company. Dionne, E. J., Jr. 1992. “The Illusion of Technique: The Impact of Polls on Reporters and Democracy.” in Media Polls in American Politics. Thomas E. Mann and Gary R. Orren. Eds. Washington, D. C.: Brookings Institution. Dobson, Douglas and Douglas St. Angelo. 1975. “Party Identification and the Floating Vote: Some Dynamics.” American Political Science Review 69 (2): 481-490. Downs, Anthony. 1957. An Economic Theory of Democracy. New York: HarperCollins. Dreyer, Edward C. 1973. “Changes and Stability in Party Identification.” Journal of Politics 35 (3): 712-722. Durr, Robert H. 1993. “What Moves Policy Sentiment?” American Political Science Review 87 (1): 158-170. Edsall, Thomas Byrne. 1984. The New Politics of Inequality. New York: W.W. Norton. Edsall, Thomas Byrne and Mary D. Edsall. 1991. Chain Reaction: The Impact of Race, Rights, and Taxes on American Politics. New York: Norton. Eismeier, Theodore J. 1982. “Public Preferences about Government Spending: Partisan, Social, and Attitudinal Sources of Policy Differences.” Political Behavior 4 (2): 133-145. Elder, Glen H., Jr. 1974. Children of the Great Depression: Social Change and Life Experience. Chicago: University of Chicago Press. Ellis, Richard. 1993. American Political Cultures. New York: Oxford University Press. Entman, Robert M. and Andrew Rojecki. 2000. The Black Image in the White Mind. Chicago: University of Chicago Press. Erikson, Robert S. 1979, “The SRC Panel Data and Mass Political Attitudes.” British Journal of Political Science 9 (1): 89-114. 161 Erikson, Robert S., Michael B. MacKuen and James Stimson. 2002. The Macro Polity. Cambridge: Cambridge University Press. Erikson, Robert S., Norman R. Luttbeg and Kent L. Tedin. 1991. American Public Opinion: Its Origins, Content, and Impact. 4 th ed. New York: Macmillan. Erskine, Hazel. 1971. “The Polls: Women's Role.” Public Opinion Quarterly 35 (2): 275-290. Evan, William M. 1959. “Cohort Analysis of Survey Data: A Procedure for Studying Long-Term Opinion Change.” Public Opinion Quarterly 63-72. ------[1965] 1968. “Cohort Analysis of Attitude Data” in Computer Methods in the Analysis of Large Scale Social Systems. 2 nd ed. James M. Besher. Ed. Cambridge, MA: M. I. T. Press. Fan, David P. 1988. Predictions of Public Opinion from the Mass Media: Computer Content Analysis and Mathematical Modeling. New York: Greenwood Press. Fan, David P. and Albert R. Tims. 1989. “The Impact of the News Media on Public Opinion: American Presidential Election, 1987-1988.” International Journal of Public Opinion Research 1 (2): 151-163. Feagin, Joe R. 1975. Subordinating the Poor: Welfare and American Beliefs. Englewood Cliffs, NJ: Prentice-Hall. Federico, Christopher M. and Jim Sidanius. 2002. “Sophistication and the Antecedents of Whites’ Racial Policy Attitudes: Racism, Ideology, and Affirmative Action in America.” Public Opinion Quarterly 66 (2): 145-176. Feather, Norman T. 1973. “The Measurement of Values: Effects of Different Assessment Procedures.” Australian Journal of Psychology 25 (3): 221-231. ------1975. Values in Education and Society. New York: Free Press Feldman, Stanley. 1988. “Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values.” American Journal of Political Science 32 (2): 416-440. Feldman, Stanley and John Zaller. 1992. “The Political Culture of Ambivalence: Ideological Responses to the Welfare State.” American Journal of Political Science 36 (1): 268-307. Feldman, Stanley and Marco R. Steenbergen. 2001. “The Humanitarian Foundation of Public Support for Social Welfare.” American Journal of Political Science 45 (3): 658-677. 162 Feree, Myra Marx. 1974. “A Woman for President? Changing Responses: 1958-1972.” Public Opinion Quarterly 38 (3): 390-399. Fetzer, Joel S. 2000. Public Attitudes toward Immigration in the United States, France and Germany. Cambridge, UK: Cambridge University Press. Field, John Osgood Anderson and Ronald E. Anderson. 1969. “Ideology in the Public’s Conceptualization of the 1964 Election.” Public Opinion Quarterly 33 (3): 390- 398. Fienberg, Steven E. and William M. Mason. 1979. “Identification and Estimation of Age- Period-Cohort Models in the Analysis of Discrete Archival Data.” Sociological Methodology 10: 1-67. -------1985. “Specification and Implementation of Age, Period and Cohort Models.” In Cohort Analysis in Social Research: Beyond the Identification Problem. William M. Mason and Stephen E. Fienberg. Eds. Finkel, Steven E. 1995. Causal Analysis with Panel Data. Thousand Oaks, CA: Sage Publications. Fiorina, Morris P. Retrospective Voting in American National Elections. New Haven, CT: Yale University Press. Firebaugh, Glenn and Kenneth E. Davis. 1988. “Trends in Antiblack Prejudice, 1972- 1984: Region and Cohort Effects.” American Journal of Sociology 94 (2): 251- 272. Fischer, Claude S., Michael Hout, Martin Sánchez Jankowski, Samuel R. Lucas, Ann Swidler and Kim Voss. 1996. Inequality by Design: Cracking the Bell Curve Myth. Princeton, NJ: Princeton University Press. Flanigan, William H. and Nancy Zingale. 2002. Political Behavior of the American Electorate. 10 th ed. Washington, DC: CQ Press. Franklin, Charles H. 1989. “Measurement and the Dynamics of Party Identification.” Political Behavior 14 (3): 297-309). Franklin, Charles H. and John E. Jackson. 1983. “The Dynamics of Party Identification.” American Political Science Review 77 (4): 957-973. Friedman, Milton and Rose Friedman. [1979] 1980. Free to Choose: A Personal Statement. New York: Harcourt, Brace, Jovanovich. Froman, Lewis A., Jr., and James K. Skipper, Jr. 1963. “An Approach to the Learning of Party Identification.” Public Opinion Quarterly 27 (3): 473-480. 163 Galbraith, John Kenneth. [1958] 1984. The Affluent Society. 4 th ed. Boston, MA: Houghton, Mifflin. Gatlin, Douglas S., Michael W. Giles and Everett F. Cataldo. 1978. “Policy Support within a Target Group: The Case of School Desegregation.” American Political Science Review 72 (3): 985-995. Gelpi, Barbara C., Nancy C. M. Hartsock, Claire C. Novak and Myra H. Strober. 1986. Women and Poverty. University of Chicago Press. Gerber, Alan and Donald P. Green. 1998. “Rational Learning and Partisan Attitudes.” American Journal of Political Science 42 (3): 794-818. Gilens, Martin. 1988. “Gender and Support of Reagan: A Comprehensive Model of Presidential Approval.” American Journal of Political Science 32 (1): 19-49. ------1995. “Racial Attitudes and Opposition to Welfare.” Journal of Politics 57 (4): 994- 1014. ------ 1999. Why Americans Hate Welfare: Race, Media and the Politics of Antipoverty Policy. Chicago: University of Chicago Press. Giles, Michael W. and Arthur Evans. 1986. “The Power Approach to Intergroup Hostility.” Journal of Conflict Resolution 30 (3): 469-486. Gilliam, Franklin D. and Kenny J. Whitby. 1989. “Race, Class, and Attitudes toward Social Welfare Spending: An Ethclass Interpretation.” Social Science Quarterly 70 (1): 88-100. Ginsburg, Benjamin. 1986. The Captive Public: How Mass Opinion Promotes State Power. New York: Basic Books. Glass, Jennifer, Vern L. Bengtson and Charlotte Chorn Dunham. 1986. “Attitude Similarity in Three-Generation Families: Socialization, Status Inheritance or Reciprocal Influence?” American Sociological Review 51 (5): 685-698. Glenn, Norval D. 1974. “Aging and Conservatism.” Annals of the American Academy of Political and Social Science 415 (Sep.): 176-186. ------1976. “Cohort Analysts’ Futile Quest: Statistical Attempts to Separate Age, Period and Cohort Effects.” American Sociological Review 41 (5): 900-904. ------1981. “Age, Birth Cohorts, and Drinking: An Illustration of the Hazards of Inferring Effects from Cohort Data.” Journal of Gerontology 36 (3): 362-369. 164 ------1989. “A Caution About Mechanical Solutions to the Identification Problem in Cohort Analysis: Comment on Sasaki and Suzuki.” American Journal of Sociology 95 (3): 754-761. Glenn, Norval D. and J. L. Simmons. 1967. “Are Regional Cultural Differences Diminishing?” Public Opinion Quarterly 31 (2): 176-193. Glenn, Norval D. and Ted Hefner. 1972. “Further Evidence on Aging and Party Identification. “ Public Opinion Quarterly 36 (1): 31-47. Glynn, Carroll J., Susan Herbst, Garrett J. O’Keefe, Robert Y. Shapiro and Mark Lindeman. 1999. “Public Opinion and Democratic Competence.” In Public Opinion. Carroll J. Glynn, Susan Herbst and Garrett J. O’Keefe and Robert Y. Shapiro. Boulder, CO: Westview. Goidel, Robert K., Todd K. Shields and Mark Peffley. 1997. “Priming Theory and RAS Models: Toward an Integrated Perspective of Media Influence.” American Politics Quarterly 25 (3): 287-318. Golay, Michael. 1997. Where America Stands 1997. New York: John Wiley and Sons. Goldberg, Arthur S. 1966. “Discerning a Causal Pattern on Data on Voting Behavior.” American Political Science Review 60 (4): 913-922. ------1969. “Social Determinism and Rationality as Bases of Party Identification.” American Political Science Review 63 (1): 5-25. Goldstein, H. 1968. “Longitudinal Studies and the Measurement of Change.” Statistician 18 (2): 93-117. Gorsuch, Richard L. 1970. “Rokeach’s Approach to Value Systems and Social Compassion.” Review of Religious Research 11 (2): 139-143. Greeley, Andrew M. 1975. “Model for Ethnic Political Socialization.” American Journal of Political Science 19 (2): 187-206. Greeley, Andrew M. and Paul B. Sheatsley. 1971. “Attitudes Toward Racial Integration.” Scientific American 225 (6): 13-19. ------1974. “Attitudes Toward Racial Integration.” In Social Problems and Public Policy: Inequality and Justice. Lee Rainwater. Ed. Chicago: Aldine. Green, Donald Philip and Bradley Palmquist. 1990. “Of Artifacts and Partisan Instability.” American Journal of Political Science 34 (3): 872-902. 165 ------1994. “How Stable is Party Identification?” Political Behavior 16 (4): 437-466. Green, Donald, Bradley Palmquist and Eric Shickler. 2002. Partisan Hearts and Minds: Political Parties and the Social Identities of Voters. New Haven: Yale University Press. Green, John C. and James L. Guth. 1991. “The Bible and the Ballot Box: The Shape of Things to Come.” In The Bible and the Ballot Box: Religion and Politics in the 1988 Election. James L. Guth and John C. Green. Eds. Boulder, CO: Westview Press. Greenleaf, Eric A. 1992. “Measuring Extreme Response Style.” Public Opinion Quarterly 56 (3): 328-351. Gurin, Patricia, Arthur H. Miller and Gerald Gurin. 1980. “Stratum Identification and Consciousness” Social Psychology Quarterly 43 (1): 30-47. Habermas Jürgen. [1962] 1989. The Structural Transformation of the Public Sphere. T. Burger. Trans. Cambridge, MA: MIT Press. Hamilton, Richard F. 1972. Class and Politics in the United States. New York: John Wiley and Sons. Hanushek, Eric A. and John E. Jackson. 1977. Statistical Methods for the Social Sciences. New York: Academic Press. Hasenfeld, Yeheskel and Jane A. Rafferty. 1989. “The Determinants of Public Attitudes Toward the Welfare State.” Social Forces 67 (4): 1027-1048. Hayduk, Leslie A. 1996. LISREL: Issues, Debates and Strategies. Baltimore: Johns Hopkins University Press. Herbst, Susan. 1992. “Surveys in the Public Sphere: Applying Bourdieu’s Critique of Opinion Polls.” International Journal of Public Opinion Research 4(3): 220-229. Herman, Edward S. 1995. Triumph of the Market: Essays on Economics, Politics, and the Media. Boston, MA: South End Press. Hess, Robert D. and Judith V. Torney. [1967] 1968. The Development of Political Attitudes in Children. Chicago: Aldine. Hochschild, Jennifer L. 1981. What’s Fair: American Beliefs about Distributive Justice. Cambridge, MA: Harvard University Press. ------1995. Facing Up to the American Dream: Race, Class and the Soul of the Nation. Princeton, NJ: Princeton University Press. 166 ------ 2004. “Introduction and Comments.” Perspective on Politics 2 (4): 647-649. Hofstadter, Richard. 1948. The American Political Tradition: and the Men Who Made It. New York: Alfred A. Knopf. Holm, John D. and John P. Robinson. 1978. “Ideological Identification and the American Voter.” Public Opinion Quarterly 42 (2): 235-246. Holtz, Geoffrey T. 1995. Welcome to the Jungle: The Why Behind “Generation X”. New York: St. Martin’s Griffin. Hovland, Carl I. 1959. “Reconciling Conflicting Results Derived from Experimental and Survey Studies of Attitude Change.” American Psychologist 14 (1) 8-17. Hoyle, Rick H. and Abigail T. Panter. 1995. “Writing about Structural Equation Models.” In Structural Equation Modeling: Concepts, Issues and Applications. Thousand Oaks, CA: Sage. Huber, Joan and Glenna Spitze. 1981. “Wives’ Employment, Household Behaviors, and Sex-Role Attitudes.” Social Forces 60 (1): 150-169. Huber, Joan and William H. Form. 1973. Income and Ideology: An Analysis of the American Political Formula. New York: Free Press. Huckfeldt, Robert and Carol Weitzel Kohfeld. 1989. Race and the Decline of Class in American Politics. Urbana, IL: University of Illinois Press. Hughes, Michael. 1997. “Symbolic Racism, Old-Fashioned Racism, and Whites’ Opposition to Affirmative Action.” In Racial Attitudes in the 1990s: Continuity and Change. Steven A. Tuch and Jack K. Martin. Eds. Westport, CN :Praeger. Hughes, Michael and Steven A. Tuch. 2000. “How Beliefs about Poverty Influence Racial Policy Attitudes: A Study of Whites, African Americans, Hispanics and Asians in the United States.” In Racialized Politics: The Debate about Racism in America. David O. Sears, Jim Sidanius and Lawrence Bobo. Eds. Chicago: University of Chicago Press. Hui, C. Harry and Harry C. Triandis. 1985. “The Instability of Response Sets.” Public Opinion Quarterly 49 (2): 253-260. Hunt, Matthew O. 1996. “The Individual, Society, or Both? A Comparison of Black, Latino and White Beliefs about the Causes of Poverty.” Social Forces 75 (1): 293-322. Hyman, Herbert. [1959] 1969. Political Socialization. New York: Free Press. 167 Hyman, Herbert H. and Charles R. Wright. 1979. Education’s Lasting Influence on Values. Chicago: University of Chicago Press. Inglehart, Ronald. 1997. Modernization and Postmodernization: Cultural, Economic and Political Change in 43 Societies. Princeton, NJ: Princeton University Press. ------1990. Culture Shift in Advanced Industrial Society. Princeton, NJ: Princeton University Press. Isserman, Maurice and Michael Kazin. 2000. American Divided: The Civil War of the 1960s. New York: Oxford University Press. Jackman, Mary R. 1978. “General and Applied Tolerance: Does Education Increase Commitment to Racial Integration?” American Journal of Political Science 22 (2): 302-324. ------1981. “Education and Policy Commitment to Racial Integration.” American Journal of Political Science 25 (2): 256-269. Jackman, Mary R. and Michael J. Muha. 1984. “Education and Intergroup Attitudes: Moral Enlightenment, Superficial Democratic Commitment or Ideological Refinement?” American Sociological Review 49 (6): 751-769. Jackson, David J. and Duane F. Alwin. 1980. “The Factor Analysis of Ipsative Measures.” Sociological Methods and Research 9 (2): 218-238. Jackson, John E. 1975. “Issues, Party Choices and Presidential Votes.” American Journal of Political Science 19 (2): 161-185. ------ 1979. “Statistical Estimation of Possible Response Bias in Close-Ended Questions.” Political Methodology 6 (4): 393-423. ------1983. “The Systematic Beliefs of the Mass Public: Estimating Policy Preferences with Survey Data.” Journal of Politics 45 (4): 840-865. Jacob, Philip E. 1957. Changing Values in College: An Exploratory Study of the Impact of College Teaching. New York: Harper and Brothers. Jacobson, Cardell K. 1985. “Resistance to Affirmative Action: Self-Interest or Racism?” Journal of Conflict Resolution 29: 306-329. Jelen, Ted G., Corwin E. Smidt and Clyde Wilcox. 1993. “The Political Effects of the Born-Again Phenomenon.” In Rediscovering the Religious Factor in American Politics. David C. Leege and Lyman A. Kellstedt. Eds. Armonk, NY: M. E. Sharpe. 168 Jennings, M. Kent. 1984. “The Intergenerational Transfer of Political Ideologies in Eight Western Nations.” European Journal of Political Research 12 (3): 261-276. Jennings, M. Kent and Richard Niemi. 1968. “The Transmission of Political Values from Parents to Child.” American Political Science Review 62 (1): 169-184. ------ 1974. The Political Character of Adolescence: The Influence of Families and Schools. Princeton, NJ: Princeton University Press. ------1975. “Continuity and Change in Political Orientations: A Longitudinal Study of Two Generations.” American Political Science Review 69 (4): 1316-1335. Jennings, M. Kent, Gregory B. Markus, and Richard G. Niemi. [1983] 1991. Youth- Parent Socialization Panel Study, 1965-1982: Three Waves Combined [Computer file]. Ann Arbor, MI: University of Michigan. Center for Political Studies/Survey Research Center [producers]. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor]. Jennings, M. Kent, and Laura Stoker. 2004. Study of Political Socialization: Parent-Child Pairs Based on Survey of Youth Panel and Their Offspring, 1997 [Computer file]. ICPSR version. Ann Arbor, MI: University of Michigan, Center for Political Studies/Survey Research Center [producer], 1998. Ann Arbor, MI: Inter- university Consortium for Political and Social Research [distributor]. Jennings, M. Kent, Laura Stoker and Jake Bowers. 2001. “Politics Across Generations: Family Transmission Reexamined.” (unpublished manuscript). Jennings, M. Kent, Richard Niemi and Kenneth P. Langton. 1974. “Effects of the High School Civics Curriculum.” in M. Kent Jennings and Richard Niemi. The Political Character of Adolescence: The Influence of Families and Schools. Princeton, NJ: Princeton University Press. Jennings, M. Kent and Richard Niemi. 1968. “The Transmission of Political Values from Parents to Child.” American Political Science Review 62 (1): 169-184. ------1974. The Political Character of Adolescence: The Influence of Families and Schools. Princeton, NJ: Princeton University Press. ------1975. “Continuity and Change in Political Orientations: A Longitudinal Study of Two Generations.” American Political Science Review 69 (4): 1316-1335. ------1978. “The Persistence of Political Orientations: An Over-Time Analysis of Two Generations.” British Journal of Political Science 8 (3): 333-363. 169 Jennings, M. Kent, Richard Niemi and Lee H. Ehman. 1974. “Social Studies Teachers and Their Pupils.” in M. Kent Jennings and Richard Niemi. The Political Character of Adolescence: The Influence of Families and Schools. Princeton, NJ: Princeton University Press. Jennings, M. Kent, Richard Niemi and Suzanne Koprince Sebert. 1974. “The Political Texture of Peer Groups.” in M. Kent Jennings and Richard Niemi. The Political Character of Adolescence: The Influence of Families and Schools. Princeton, NJ: Princeton University Press. Judd, Charles M. and Michael A. Milburn. 1980. “The Structure of Attitude Systems in the General Public: Comparisons of a Structural Equation Model.” American Sociological Review 45 (4): 627-643. Kamieniecki, Sheldon. 1985. Party Identification, Political Behavior and the American Electorate. Westport, CT: Greenwood Press. Kaufman, Karen M. 2002. “Culture Wars, Secular Realignment, and the Gender Gap in Party Identification.” Political Behavior 24 (3): 283-307. ------ 2004. “The Partisan Paradox: Religious Commitment and the Gender Gap in Party Identification.” Public Opinion Quarterly 68 (4): 491-511. Kaufmann, Karen M. and John R. Petrocik. 1999. “The Changing Politics of American Men: Understanding the Sources of the Gender Gap.” American Journal of Political Science 43 (3): 864-887. Kessel, John H. 1972. “Comment: The Issues in Issue Voting.” American Political Science Review 66 (2): 459-465. Kellstedt, Lyman A. 1989. “Evangelicals and Political Realignment.” In Contemporary Evangelical Political Involvement: An Analysis and Reassessment.” Corwin A. Smidt. Ed. Lanham, MD: University Press of America. Kellstedt, Paul M. 2000. “Media Framing and the Dynamics of Racial Policy Preferences.” American Journal of Political Science 44 (2): 245-260. Kennedy, Peter. 1998. A Guide to Econometrics. Cambridge, MA: MIT Press. Kenski, Henry C. 1988. “The Gender Factor in a Changing Electorate.” In The Politics of the Gender Gap: The Social Construction of Political Influence. Carol M. Muller. Ed. Newbury Park, CA: Sage Publications. Kerr, W. A. 1944. “Correlates of Politico-Economic Liberalism-Conservatism.” Journal of Social Psychiatry 20: 61-77. 170 Kiecolt, K. Jill. 1988. “Recent Developments in Attitudes and Social Structure.” Annual Review of Sociology 14: 381-403. Key, V. O., Jr. 1961. Public Opinion and American Democracy. New York: Alfred Knopf. -------1966. The Responsible Electorate: Rationality in Presidential Voting, 1936-1960. Cambridge, MA: Harvard University Press. Kim, Jae-On and Charles W. Mueller. 1978a. Factor Analysis: Statistical Methods and Practical Issues. Newbury Park, CA: Sage Publications. ------1978b. Introduction to Factor Analysis: What It Is and How to Do It. Newbury Park, CA: Sage Publications. Kinder, Donald R. 1983. “Diversity and Complexity in American Public Opinion.” In Political Science: The State of the Discipline. Ada F. Finifter. Ed. Washington, D. C.: American Political Science Association. Kinder, Donald R. and David O. Sears. 1985. “Public Opinion and Political Action.” In Handbook of Social Psychology: Volume II. Special Fields and Applications. Gardner Lindzey and Elliot Aronson. Eds. New York: Random House. Kinder, Donald R. and Laurie A. Rhodebeck. 1982. “Continuities in Support for Racial Equality, 1972 to 1976.” Public Opinion Quarterly 46 (2): 195-215. Kinder, Donald R. and Lynn M. Sanders. 1996. Divided by Color: Racial Politics and Democratic Ideals. Chicago: University of Chicago Press. Kinder, Donald R. and Nicholas Winter. 2001. “Exploring the Racial Divide: Blacks, Whites, and Opinions on National Policy.” American Journal of Political Science 45 (2): 439-456. King, Desmond S. and Rogers M. Smith. 2005. “Racial Orders in American Political Development.” American Political Science Review 99 (1): 75-92. King, Gary, James Honaker, Anne Joseph and Kenneth Scheve. 2001. “Analyzing Incomplete Political Science Data: An Alternative Algorithm for Multiple Imputation.” American Political Science Review 95 (1): 49-69. Klapper, Joseph T. 1960. The Effects of Mass Communication. Glencoe, IL: Free Press. Klecka, William R. 1971. “Applying Political Generations to the Study of Political Behavior: A Cohort Analysis.” Public Opinion Quarterly 35 (3): 358-373. Kline, Rex B. 1998. Principles and Practice of Structural Equation Modeling. New York: Guilford Press. 171 Kluegel, James R. 1990. “Trends in Whites' Explanations of the Black-White Gap in Socioeconomic Status, 1977-1989.” American Sociological Review 55 (4): 512- 525. Kluegel, James R. and Eliot R. Smith. 1982. “Whites' Beliefs about Blacks' Opportunity.” American Sociological Review 47 (4): 518-532. -------1983. “Affirmative Action Attitudes: Effects of Self-Interest, Racial Affect and Stratification Beliefs on Whites’ Views.” Social Forces 61 (1): 797-824. ------ 1986. Beliefs About Inequality: Americans View of What Is and What Ought To Be. New York: Aldine de Gruyter. Knapp, Melvin J. and Jon P. Alston. 1972-1973. “White Parental Acceptance of Varying Degrees of School Desegregation, 1965 and 1970.” Public Opinion Quarterly 36 (4): 585-591. Knight, Kathleen. 1985. “Ideology in the 1980 Election: Ideological Sophistication Does Matter.” Journal of Politics 47 (3): 828-853. Knoke, David. 1976. Change and Continuity in American Politics: The Social Bases of Political Parties. Baltimore: Johns Hopkins Press. -------1979. “Stratification and the Dimensions of American Political Orientations.” American Journal of Political Science 23 (4): 772-791. Knoke, David and Michael Hout. 1974. “Social and Demographic Factors in American Political Party Affiliation, 1952-1972.” American Sociological Review 39 (5): 700-713. Kohut, Andrew, John C. Green, Scott Keeter and Robert C. Toth. 2000. The Diminishing Divide: Religion’s Changing Role in American Politics. Washington, D. C.: Brookings Institution. Kramer, Gerald H. 1971. “Short-Term Fluctuations in U. S. Voting Behavior, 1896- 1964.” American Political Science Review 65 (1): 131-143. Krause, Sidney and Dennis Davis. 1976. The Effects of Mass Communication on Political Behavior. University Park: Pennsylvania State University Press. Krosnick, Jon A. 1991. “The Stability of Political Preferences: Comparisons of Symbolic and Nonsymbolic Attitudes.” American Journal of Political Science 35 (3): 547-576. Ladd, Everett Carll, Jr. 1970. American Political Parties: Social Change and Political Response. New York: W. W. Norton. 172 ------ 1976-1977. “Liberalism Upside Down: The Inversion of the New Deal Order.” Political Science Quarterly 91 (4): 577-600. ------1993. “The 1992 Vote for Clinton: Another Brittle Mandate.” Political Science Quarterly 108 (1): 1-28. ------1997. “1996 Vote: The “No Majority” Realignment Continues.” Political Science Quarterly 112 (1): 1-28. Ladd, Everett Carll, Jr. and Charles Hadley. 1978 [1975]. Transformations of the American Party System: Political Coalitions from the New Deal to the 1970s. New York: W. W. Norton. Ladd, Everett Carll and Karlyn H. Bowman. 1998. Attitudes Toward Economic Inequality. Washington, D. C.: American Enterprise Institute for Public Policy Research. Lane, Robert. 1959. “The Fear of Equality.” American Political Science Review 53 (1): 35-51. ------1962. Political Ideology: Why the American Common Man Believes What He Does. New York: Free Press. Langton, Kenneth P. 1969. Political Socialization. New York: Oxford University Press. Lau, Richard R., Thad A. Brown, and David O. Sears. 1978. “Self Interest and Civilians’ Attitudes Toward the War in Vietnam.” Public Opinion Quarterly 42: 464-483. Layman, Geoffrey C. 1997. “Religion and Political Behavior in the United States: The Impact of Beliefs, Affiliations, and Commitment from 1980 to 1994.” Public Opinion Quarterly 61 (2): 288-316. Lazarsfeld, Paul F., Bernard Berelson, and Hazel Gaudet. [1948] 1965. The People’s Choice, How the Voter Makes Up His Mind in a Campaign. 2 nd Ed. New York: Columbia University Press. Lee, Sik-Yum, Wai-Yin Poon, and P. M. Bentler. 1995. “A Two-Stage Estimation of Structural Equation Models with Continuous and Polytonomous Variables.” British Journal of Mathematical and Statistical Psychology 48: 339-358. Lee, Taeku. 2002. Mobilizing Public Opinion: Black Insurgency and Racial Attitudes in the Civil Rights Era. Chicago: University of Chicago Press. Levin. Martin L. 1961. “Climate of Opinion, Social Climates and Political Socialization.” Public Opinion Quarterly 25 (4): 596-606. 173 Levitin, Teresa E. and Warren E. Miller. 1979. “Ideological Interpretations of Presidential Elections.” American Political Science Review 73 (3): 751-771. Liao, Tim Futing. 1994. Interpreting Probability Models: Logit, Probit and Other Generalized Linear Models. Thousand Oaks, CA: Sage Publications. Lippman, Walter. 1993. The Phantom Public. New Brunswick, NJ: Transaction Publishers. Lipset, Seymour Martin. [1959] 1960. Political Man: The Social Bases of Politics. Garden City, NY: Doubleday. ------1979. The First New Nation. New York: Norton. Loftus, Jeni. 2001. “America’s Liberalization in Attitudes Toward Homosexuality, 1973 to 1998.” American Sociological Review 66 (5): 762-782. Lomax, Fay. 1979. Who Should Be Helped: Public Support for Social Services. Beverly Hills, CA: Sage. Long, J. Scott. 1983a. Confirmatory Factor Analysis: A Preface to LISREL. Newbury Park, CA: Sage Publications. ------1983b. Covariance Structures: An Introduction to LISREL. Newbury Park, CA: Sage Publications. Lopatto, Paul. 1985. Religion and the Presidential Vote. New York: Praeger Publishers. Lowery, David and Lee Sigelman. 1981. “Understanding the Tax Revolt: Eight Explanations.” American Political Science Review 75 (4): 963-974. Luttbeg, Norman R. 1968. “The Structure of Beliefs Among Leaders and the Public.” Public Opinion Quarterly 32 (3): 398-409. Luttbeg, Norman R. and Michael M. Gant. 1985. “The Failure of Liberal/Conservative Ideology as a Cognitive Structure.” Public Opinion Quarterly 49 (1): 80-93. MacCallum, R. C., Browne, M. W. and Sugawara, H. M. 1996. “Power Analysis and Determination of Sample Size for Covariance Structure Modeling.” Psychological Methods 1 (2): 130-149. MacKuen, Michael B., Robert S. Erickson and James S. Stimson. 1989. “Macropartisanship.” American Political Science Review 83 (4): 1125-1142. Maddala, G. S. 1983. Limited Dependent and Qualitative Variables in Econometrics. Cambridge: Cambridge University Press. 174 Manza, Jeff and Clem Brooks. 1999. Social Cleavages and Political Change: Voter Alignments and U. S. Party Coalitions. New York: Oxford University Press. Markus, Gregory B. 1979a. Analyzing Panel Data. Newbury, CA: Sage Publications. Markus, Gregory B. 1979b. “The Political Environment and the Dynamics of Public Attitudes: A Panel Study.” American Journal of Political Science 23 (2): 338- 359. Markus, Gregory B. 1985. “Dynamic Modeling of Cohort Change: The Case of Political Partisanship.” In Cohort Analysis in Social Research: Beyond the Identification Problem. William M. Mason and Stephen E. Fienberg. Eds. Markus, Gregory B. and Philip E. Converse. 1979. “A Dynamic Simultaneous Equation Model of Electoral Choice.” American Political Science Review 73 (4): 1055- 1070. Mason, Karen Oppenheim and Larry L. Bumpass. 1975. “U. S. Women’s Sex-Role Ideology, 1970” American Journal of Sociology 80 (5): 1212-1219. Mason, Karen Oppenheim and Yu-Hsia Lu. 1988. “Attitudes Toward Women’s Familial Roles: Changes in the United States, 1977-1985.” Gender and Society 2 (1): 39- 57. Mason, Karen Oppenheim, John L. Czaijka and Sara Arber. 1976. “Change in U.S. Women’s Sex-Role Attitudes, 1964-1974.” American Sociological Review 41 (4): 573-596. Mason, Karen Oppenheim, William H. Mason, H. H. Winsborough and W. Kenneth Poole. 1973. “Some Methodological Issues in Cohort Analysis of Archival Data.” American Sociological Review 38 (2): 242-258. Mason, William M. and Stephen E. Fienberg. 1985. “Introduction: Beyond the Identification Problem.” In Cohort Analysis in Social Research: Beyond the Identification Problem. William M. Mason and Stephen E. Fienberg. Eds. Mayer, William G. 1992. The Changing American Mind: How and Why Public Opinion Changed Between 1960 and 1980. Ann Arbor, MI: University of Michigan Press. Maykovich, Minako K. 1975. “Correlates of Racial Prejudice.” Journal of Personality and Social Psychology 32 (6): 1014-1020. McCann, James A. 1997. “Electoral Choices and Core Value Change: The 1992 Presidential Campaign.” American Journal of Political Science 41 (2): 564-683. 175 McCarty, Nolan M., Keith T. Poole and Howard Rosenthal. 1997. Income Redistribution and the Realignment of American Politics. Washington, D. C.: AEI Press. ------2003. “Political Polarization and Income Inequality.” Unpublished manuscript. McClendon, McKee J. and Fred P. Pestello. 1983. “Self-Interest and Public Policy Attitude Formation: Busing for School Desegregation.” Sociological Focus 16 (1): 1-12. McCloskey, Herbert. 1958. “Conservatism and Personality.” American Political Science Review 52 (1): 27-48. ------1964. “Consensus and Ideology in American Politics.” American Political Science Review 58 (2): 361-382. McCloskey, Herbert and John Zaller. 1984. The American Ethos: Public Attitudes Toward Capitalism and Democracy. Cambridge, MA: Harvard University Press. McCloskey, Herbert, John Zaller and Dennis Chang. 1984. “Social Learning and the Acquisition of Political Norms.” In Herbert McCloskey and John Zaller. The American Ethos: Public Attitudes Toward Capitalism and Democracy. Cambridge, MA: Harvard University Press. Meier, Kenneth J. 1975. “Party Identification and the Vote Choice: The Causal Relationship. “ Western Political Quarterly 28 (3): 496-505. Middleton, Russell. 1976. “Regional Differences in Prejudice.” American Sociological Review 41 (1): 94-117. Miller Arthur H. and Martin P. Wattenberg. 1984. “Politics From the Pulpit: Religiosity and the 1980 Elections.” Public Opinion Quarterly 48 (1): 301-317. Miller, Arthur H., Warren E. Miller, Alden S. Raine and Thad E. Brown. 1976. “A Majority Party in Disarray: Policy Polarization in the 1972 Election.” American Political Science Review 70 (3): 753-778. Miller, Richard B. 1981. “Regional Difference in White Attitudes to Desegregation.” Free Inquiry in Creative Sociology 9 (2): 131-134. Miller, Warren E. 1991. “Party Identification, Realignment, and Party Voting: Back to the Basics.” American Political Science Review 85 (2): 557-568. Miller, Warren E. and J. Merrill Shanks. 1996. The New American Voter. Cambridge, MA: Harvard University Press. Moore, Michael. 1975. “Rating Versus Ranking in the Rokeach Value Survey: An Israeli Comparison.” European Journal of Social Psychology 5 (3): 405-408. 176 Munson, J. Michael and Shelby H. McIntyre. 1979. “Developing Practical Procedures for the Measurement of Personal Values in Cross-Cultural Marketing.” Journal of Marketing Research 16: 48-52. Murray, Charles. 1984. Losing Ground: American Social Policy, 1950-1980. New York: Basic Books. Mutz, Diana C., and Paul S. Martin. 2001. “Facilitating Communication Across Lines of Political Difference: The Role of the Mass Media.” American Political Science Review 95 (1): 97-114. Myrdahl, Gunnar. [1944] 1962. An American Dilemma, Vol. 1. New York:Harper and Row. Nadeau, Richard, Richard G. Niemi, David P. Fan and Timothy Amato. 1999. “Elite Economic Forecasts, Economic News, Mass Economic Judgments, and Presidential Approval.” Journal of Politics 61 (1): 109-135. Nelson, Thomas E., Rosalee A. Clawson and Zoe M. Oxley. 1997. “Media Framing of a Civil Liberties Controversy and Its Effect on Tolerance.” American Political Science Review 91 (3): 567-583. Neuman, W. Russell. 1986. The Paradox of Mass Politics: Knowledge and Opinion in the American Electorate. Cambridge, MA: Harvard University Press. Neuman, W. Russell. 1989. “Parallel Content Analysis: Old Paradigms and New Proposals.” In Public Communication and Behavior. Vol. 2. George Comstock. Ed. Neuman, W. Russell, Marion R. Just and Ann N. Crigler. 1992. Common Knowledge: News and the Construction of Political Meaning. Chicago: University of Chicago Press. Nevitt, Jonathan and Gregory R. Hancock. 2000. “Improving the Root Mean Square Error of Approximation for Nonnormal Conditions in Structural Equation Modeling.” Journal of Experimental Education 68 (3): 251-168. Newcomb, Theodore M. 1948. “Public Opinion and Public Opinion Polling: Discussion.” American Sociological Review 13 (5): 549-552. Nie, Norman and Kristi Anderson. 1974. “Mass Belief Systems Revisited: Political Change and Attitude Structure.” Journal of Politics 36 (3): 540-591. Nie, Norman, Sidney Verba and John R. Petrocik. 1976. The Changing American Voter. Cambridge, MA: Harvard University Press. 177 Norpoth, Helmut. 1987. “Under Way and Here to Stay: Party Realignment in the 1980s?” Public Opinion Quarterly 51 (3): 376-391. Norpoth, Helmut and Milton Lodge. 1985. “The Difference between Attitudes and Nonattitudes in the Mass Public: Just Measurements.” American Journal of Political Science 29 (2): 291-307. Norrander, Barbara. 1999. “The Evolution of the Gender Gap. “ Public Opinion Quarterly 63 (4): 566-576. Nunn, Clyde Z., Harry J. Crockett, Jr., and J. Allen Williams. 1978. Tolerance for Nonconformity. San Francisco: Jossey-Bass. Ozouf, Mona. 1988. “Public Opinion at the End of the Old Regime.” Journal of Modern History 60: S1-S21. Page, Benjamin I. and Calvin C. Jones. 1979. “Reciprocal Effects of Policy Preferences, Party Loyalties and the Vote.” American Political Science Review 73 (4): 1071- 1089. Page, Benjamin I. and Robert Y. Shapiro. 1982. “Changes in Americans’ Policy Preferences, 1935-1979.” Public Opinion Quarterly 46 (1): 24-42. ------ 1992. The Rational Public: Fifty Years of Trends in Americans Policy Preferences. Chicago: University of Chicago Press. Page, Benjamin I., Robert Y. Shapiro and Glenn R. Dempsey. 1987. “What Moves Public Opinion?” American Political Science Review 81 (1): 23-43. Patterson, Thomas E. and Robert D. McClure. 1976. The Unseeing Eye: The Myth of Television Power in National Elections. New York: G. P. Putnam’s Sons. Peek, Charles W. 1980. “Sex Prejudice Among White Protestants: Like or Unlike Ethnic Prejudice?” Social Forces 59 (3): 169-185. Pettigrew, Thomas F. 1959. “Regional Differences in Anti-Negro Prejudice.” Journal of Abnormal and Social Psychology 59: 28-36. Phillips, Anne. 1999. Which Equalities Matter? Malden, MA: Blackwell. Pierce, John C. and Douglas D. Rose. 1974. “Nonattitudes and American Public Opinion: The Examination of a Thesis.” American Political Science Review 68 (2): 626- 649. 178 Pole, J. R. [1978] 1993. The Pursuit of Equality in American History. Rev. Ed. Berkeley: University of California Press. Pollock III, Philip H. 1983. “Are There Two Liberalisms? The Partisan and Demographic Influences on Two Dimensions of Political Belief.” Sociological Focus 16 (4): 227-237. Pomper, Gerald M. 1972. “From Clarity to Confusion: Issues and the American Voters, 1956-1968.” American Political Science Review 66 (2): 415-428. Pomper, Gerald with Susan Lederman. 1980. Elections in America: Control and Influence. 2 nd Ed. New York: Longman’s. Popkin, Samuel L. 1991. The Reasoning Voter: Communication and Persuasion in Presidential Campaigns. Chicago: University of Chicago Press. Price, Vincent and John Zaller. 1993. “Who Gets the News? Alternative Measures of News Reception and Their Implications for Research.” Public Opinion Quarterly 57 (2): 133-164. Price, Vincent, David Tewksbury and Elizabeth Powers. 1997. “Switching Trains of Thought: The Impact of News Frames in Readers’ Cognitive Responses.” Communication Research 24 (5): 481-506. Rae, Douglas, Douglas Yates, Jennifer Hochschild, Joseph Morone and Carol Fessler. 1981. Equalities. Cambridge, MA: Harvard University Press. Rankin, William L. and Joel W. Grube. 1980. “A Comparison of Ranking and Rating Procedures for Value System Measurement.” European Journal of Social Psychology 10 (3): 233-246. Ransford, H. Edward. [1977] 1994. Race and Class in American Society. 2 nd rev. ed. Rochester, VT: Schenkman Books. Ransford, H. Edward and Jon Miller. 1983. “Race, Sex and Feminist Outlooks.” American Sociological Review 48 (1): 46-59. Redenius, Charles. 1981. The American Ideal of Equality: From Jefferson's Declaration to the Burger Court. Port Washington, NY: Kennikat Press. RePass, David E. 1971. “Issue Salience and Party Choice.” American Political Science Review 65 (2): 389-400. Riley, Matilda White. 1973. “Aging and Cohort Succession: Interpretations and Misinterpretations.” Public Opinion Quarterly 37 (1): 35-49. 179 Robinson, Robert V. and Wendell Bell. 1978. “Equality, Success, and Social Justice in England and the United States.” American Sociological Review 43 (2): 125-143. Rodgers, William L. 1982. “Estimable Functions of Age, Period and Cohort Effects.” American Sociological Review 47 (6): 774-787. Rokeach, Milton. 1968. “A Theory of Organization and Change Within Value-Attitude Systems.” Journal of Social Issues 24 (1): 13-33. ------1968-1969. “The Role of Values in Public Opinion Research.” Public Opinion Quarterly 32 (4): 547-559. ------1969. “Value Systems in Religion.” Review of Religious Research 11 (1): 3-23. ------1970. “Commentary on the Commentaries.” Review of Religious Research 11 (2): 155-162. ------1973. The Nature of Human Values. New York: Free Press. ------1979. Understanding Human Values: Individual and Social. New York: Free Press. Rokeach, Milton and Sandra Ball-Rokeach. 1989. “Stability and Change in American Value Priorities, 1968-1981.” American Psychologist 44 (5): 775-784. Romasco, Albert U. 1983. The Politics of Recovery: Roosevelt’s New Deal. New York: Oxford University Press. Rorty, Richard. 1998. Achieving Our Country: Leftist Thought in Twentieth-Century America. Cambridge, MA: Harvard University Press. Rosenberg, Harold. [1960] 1965. The Tradition of the New. New York: McGraw-Hill. Rosenstone, Steven J., Donald R. Kinder, Warren E. Miller, and the National Election Studies. American National Election Study. 2000. American National Election Study, 1996: Pre- and Post-Election Survey [Computer file]. 4th version. Ann Arbor, MI: University of Michigan, Center for Political Studies [producer], 1999. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2000. Ryder, Norman B. 1965. “The Cohort as a Concept in the Study of Social Change.” American Sociological Review 30 (6): 843-861. Sapiro, Virginia, Steven J. Rosenstone, and the National Election Studies. 2001. 1948- 2000 Cumulative Data File [dataset]. Ann Arbor, MI: University of Michigan, Center for Political Studies [producer and distributor]. 180 Sasaki, Masamichi and Tatsuzo Suzuki. 1989. “A Caution About the Data to Be Used for Cohort Analysis: Reply to Glenn.” American Journal of Sociology 95 (3): 761- 765. Satorra, Albert, and Bentler, Peter M. 1994. “Corrections to Test Statistics and Standard Errors in Covariance Structure Analysis.” In Latent Variables Analysis: Applications for Developmental Research. Alexander von Eye and Clifford C. Clogg. Eds. Thousand Oaks, CA: Sage. Schaefer, David L. 1979. “Introduction.” In The New Egalitarianism: Questions and Challenges. David Lewis Schaefer. Ed. Port Washington, N. Y.: Kennikat Press. Schlesinger, Arthur M., Jr. [1986] 1999. The Cycles of American History. 1 st Mariner Books ed. Boston, MA: Houghton, Mifflin. Schriebner, E. M. 1978. “Education and Change in American Opinions on a Woman for President.” Public Opinion Quarterly 42 (2): 171-182. Schuman, Howard and Jean M. Converse. 1971. “The Effects of Black and White Interviewers on Black Responses.” Public Opinion Quarterly 35 (1): 44-68. Schuman, Howard and Stanley Presser. 1981. Questions and Answers in Attitude Surveys: Experiments on Question Form, Wording and Content. New York: Academic Press. Schuman, Howard, Charlotte Steeh and Lawrence Bobo. 1985. Racial Attitudes in America: Trends and Interpretations. Cambridge, MA: Cambridge University Press. Schuman, Howard, Charlotte Steeh, Lawrence Bobo and Maria Kryson. 1997. Racial Attitudes in America: Trends and Interpretations. Cambridge, MA: Harvard University Press. Schwartz, Mildred A. 1967. Trends in White Attitudes Toward Negroes. Chicago: National Opinion Research Center. Selvin, Hanan C. 1968. “Survey Analysis: Methods of Survey Analysis.” in International Encyclopedia of the Social Sciences. Vol. 15. David S. Sills. Ed. Searing, Donald, Gerald Wright and George Rabinowitz. 1976. “The Primacy Principle: Attitude Change and Political Socialization.” British Journal of Political Science 6 (1): 83-113. Sears, David O. 1975. “Political Socialization.” In Handbook of Political Science. Fred I. Greenstein and Nelson Polsby. Eds. Reading, MA: Addison-Wesley. 181 ------1983. “The Persistence of Early Political Predispositions: The Roles of Attitude Object and Life Stage.” In Review of Personality and Social Psychology, Vol. 4. Ladd Wheeler and Phillip Shaver. Eds. Beverly Hills: Sage. ------1988. “Symbolic Racism.” In Eliminating Racism: Profiles in Controversy. Phyllis A. Katz and Dalmas A. Taylor. Eds. New York: Plenum Press. Sears, David O. and Carolyn Funk. 1999. “Evidence of the Long-Term Persistence of Adults Political Predispositions.” Journal of Politics 61 (1): 1-28. Sears, David O. and Jack Citrin. 1982. Tax Revolt: Something for Nothing in California. Cambridge, MA: Harvard University Press. Sears, David O. and Nicholas Valentine. 1997. “Politics Matters: Political Events as Catalysts in Preadult Socialization.” American Political Science Review 91 (1): 45-65. Sears, David O. and Richard A. Whitney. 1973. “Political Persuasion.” In Handbook of Communication. Ithiel de Sola Pool and Wilbur Schramm. Eds. Chicago: Rand McNally. Sears, David O., Carl P. Hensler and Leslie K. Speer. 1979. “Whites’ Opposition to ‘Busing’: Self-Interest or Symbolic Politics.” American Political Science Review 73 (2): 369-384. Sears, David O., John Hetts, Jim Sidanius and Lawrence Bobo. 2000. “Race in American Politics.” In Racialized Politics: The Debate about Racism in America. David O. Sears, Jim Sidanius and Lawrence Bobo. Eds. Chicago: University of Chicago Press. Sears, David O., P. J. Henry and Rick Kosterman. 2000. “Egalitarian Values and Contemporary Racial Politics.” In Racialized Politics: The Debate About Racism in America. David O. Sears, Jim Sidanius and Lawrence Bobo. Eds. Sears, David O., Richard R. Lau, Tom R. Tyler and Harris M. Allen. 1980. “ Self-Interest vs. Symbolic Politics in Policy Attitudes and Presidential Voting.” American Political Science Review 74: 670-684. Sears, David O., Tom R. Tyler, Jack Citrin and Donald R. Kinder. 1978. “Political System Support and Public Response to the Energy Crisis.” American Journal of Political Science 22 (1): 56-82. Selvin, Hanan C. and Alan Stuart. 1966. “Data-Dredging Procedures in Survey Analysis.” American Statistician 20 (3): 20-23. 182 Shah, Dhavan V., Mark D. Watts, David Domke, David P. Fan and Michael Fibison. 1999. “News Coverage, Economic Cues, and the Public’s Presidential Preferences, 1984-1996.” Journal of Politics 61 (4): 914-943. Shapiro, Michael J. 1969. “Rational Political Man: A Synthesis of Economic and Social- Psychological Perspectives.” American Political Science Review 63 (4): 1106- 1119. Shapiro, Robert Y. and Harpreet Mahajan. 1986. “Gender Differences in Policy Preferences: A Summary of Trends From the 1960s to the 1980s.” Public Opinion Quarterly 50 (1): 42-61. Shapiro, Robert Y. and John T. Young. 1989. “Public Opinion and the Welfare State: The United States in Comparative Perspective.” Political Science Quarterly 104 (1): 59-89. Shingles, Richard D. 1989. “Class, Status and Support for Government Aid to Disadvantaged Groups.” Journal of Politics 51 (4): 933-962. Sigel, Roberta. 1970. Learning About Politics. New York: Random House. Sigelman, Lee and Susan Welch. 1991. Black Americans’ Views on Racial Inequality: The Dream Deferred. New York: Cambridge University Press. Singer, Eleanor. 1987. “Editor's Introduction.” Public Opinion Quarterly 51 (4 pt. 2): S1- S3. Sloan, John W. 1999. The Reagan Effect: Economics and Presidential Leadership. Lawrence, KS: University of Kansas Press. Smith, A. Wade. 1981a. “Tolerance of School Desegregation, 1954-77.” Social Forces 59 (4): 1256-1274. ------1981b. “Racial Tolerance as a Function of Group Position.” American Sociological Review 46 (5): 558-573. ------1982a. “White Attitudes Toward School Desegregation, 1954-1980: An Update on Continuing Trends.” Pacific Sociological Review 25 (1): 3-25. ------1982b. “Attitudes of Whites Toward Residential Integration.” Phylon 43 (4): 368- 384. ------1985. “Cohorts, Education and the Evolution of Tolerance.” Social Science Research 14: 205-225. 183 Smith, Dwayne M. and Lynne J. Fisher. 1982. “Sex Role Attitudes and Social Class: A Reanalysis and Clarification.” Journal of Comparative Family Studies 8 (1): 77- 88. Smith, Elliot R. and James R. Kluegel. 1984. “Beliefs and Attitudes about Women’s Opportunity: Comparisons with Beliefs About Blacks and a General Perspective.” Social Psychology Quarterly 47 (1): 81-94. Smith, Tom W. 1990. “Liberal and Conservative Trends in the United States Since World War II.” Public Opinion Quarterly 54 (4): 479-507. Sniderman, Paul M. and Michael Gray Hagen. 1985. Race and Inequality: A Study in American Values. Chatham, NJ: Chatham House. Sniderman, Paul M., Michael G. Hagen, Philip E. Tetlock and Henry Brady. 1986. “Reasoning Chains: Causal Models of Policy Reasoning in Mass Publics.” British Journal of Political Science 16 (4): 405-430. Sniderman, Paul M., Richard Brody and James H. Kuklinski. 1984. “Policy Reasoning and Political Values: The Problem of Racial Equality.” American Journal of Political Science 28 (1): 75-94. Sowell, Thomas. 2002. A Conflict of Visions: Ideological Origins of Political Struggles. New York: Basic Books. Spiegelberg, Herbert. 1944. “A Defense of Human Equality.” Philosophical Review 53 (2): 101-124. Spitze, Glenna and Joan Huber. 1980. “Changing Attitudes Toward Women’s Nonfamily Roles.” Sociology of Work and Occupations 7 (3): 317-335. Steeh, Charlotte and Howard Schuman. 1992. “Young White Adults: Did Racial Attitudes Change in the 1980s?” American Journal of Sociology 98 (2): 340-367. Steeh, Charlotte and Maria Krysan. 1996. “The Polls-Trends: Affirmative Action and the Public, 1970-1995.” Public Opinion Quarterly 60 (1): 128-158. Steiger, J. H. 1990. “Structural Model Evaluation and Modification: An Interval Estimation Approach.” Multivariate Behavioral Research 25: 173-180. Stembler, Charles Herbert. 1961. Education and Attitude Change: The Effect of Schooling on Prejudice Against Minorities. New York: Institute of Human Relations Press. Stiglitz, Joseph E. 2003. The Roaring Nineties: A New History of the World’s Most Prosperous Decade. New York: W. W. Norton. 184 Stimson, James A. 1975. “Belief Systems: Constraint, Complexity and the 1972 Election.” American Journal of Political Science 19 (3): 393-417. ------2004. Tides of Consent: How Public Opinion Shapes American Politics. New York: Cambridge University Press. Stonecash, Jeffrey M., Mark D. Brewer, R. Eric Petersen, Mary P. McGuire and Lori Beth Way. 2000. “Class and Party: Secular Realignment and the Survival of Democrats Outside the South.” Political Research Quarterly 53 (4): 731-752. Stouffer, Samuel A. 1963 [1955]. Communism, Conformity and Civil Liberties: A Cross- Section of the Nation Speaks Its Mind. New York:Doubleday. Strass, William and Neil Howe. 1991. Generations: The History of Americas Future, 1584-2069. New York: Morrow. ------1997. The Fourth Turning: An American Prophecy. New York: Broadway Books. Sullivan, John L., James E. Piereson and George E. Marcus. 1978. “Ideological Constraint in the Mass Public: A Methodological Critique and Some New Findings.” American Journal of Political Science 22 (2): 233-249. Tabachnick, Barbara G. and Linda S. Fidell. 2001. Using Multivariate Statistics. 4 th ed. Boston: Allyn and Bacon. Tedin, Kent L. 1974. “The Influence of Parents on the Political Attitudes of Adolescents.” American Political Science Review 68 (4): 1579-1592. ------1980. “Assessing Peer and Parent Influence on Adolescent Political Attitudes.” American Journal of Political Science 24 (1): 136-154. Terkildsen, Nayda and Frank Schnell. 1997. “How the Media Frames Public Opinion: An Analysis of the Women's Movement.” Political Research Quarterly 50 (4): 879- 900. Tessler, Mark, Carrie Konold and Megan Reif. 2004. “Political Generations in Developing Countries: Evidence and Insights from Algeria.” Public Opinion Quarterly 68 (2): 184-216. Tocqueville, Alexis de. 1863. Democracy in America, Vol. II. 2 nd Ed. Trans. Henry Reeve and Francis Bowen. Cambridge, MA: Sever and Francis. Available at <http://books.google.com> Tuch, Steven A. 1984. “A Multivariate Analysis of Response Structure: Race Attitudes, 1972-1977. Social Science Research 13: 55-71. 185 ------1987. “Urbanism, Region, and Tolerance Revisited: The Case of Racial Prejudice.” American Sociological Review 52 (2): 504-510. Tuch, Steven A. and Michael Hughes. 1996. “Whites’ Racial Policy Attitudes.” Social Science Quarterly 77 (4): 723-745. Tufte, Edward R. 1974. Data Analysis for Politics and Policy. Englewood Cliffs, NJ: Prentice-Hall. Tumin, Melvin M. 1957. “Exposure to Mass Media and Readiness for Desegregation.” Public Opinion Quarterly 21 (2): 237-251. Valentino, Nicholas and David O. Sears. 1998. “Event-Driven Political Communication and the Preadult Socialization of Partisanship.” Political Behavior 20 (2): 127- 154. Verba, Sidney. 1996. “The Citizen as Respondent: Sample Surveys and American Democracy Presidential Address, American Political Science Association, 1995.” American Political Science Review 90 (1): 1-7. Verba, Sidney and Gary Orren. 1985a. Equality in America: The View from the Top. Cambridge, MA: Harvard University Press. Verba, Sidney and Gary Orren. 1985b. “The Meaning of Equality in America.” Political Science Quarterly 100 (3): 369-387. Verba, Sidney and Norman H. Nie. 1972. Participation in America: Social Equality and Political Democracy. New York: Harper, Row. Verba, Sidney, Kay Lehman Schlozman and Henry E. Brady. 1995. Voice and Equality: Civic Voluntarism in American Democracy. Cambridge: Harvard University Press. Wald, Kenneth D. and Corwin D. Smidt. 1993. “Measurement Strategies in the Study of Religion and Politics.” In Rediscovering the Religious Factor in American Politics. David C. Leege and Lyman A. Kellstedt. Eds. Armonk, NY: M. E. Sharpe. Wanniski, Jude. 1989. The Way the World Works. 3 rd ed. Morristown, NJ: Polyconomics. Welch, Susan. 1982. “Changes in Public Attitudes Toward Women in Politics.” Social Science Quarterly 63: (2): 312-322. Westie, Frank R. 1965. “The American Dilemma: An Empirical Test.” American Sociological Review 30: 527-538. 186 Westholm, Anders and Richard G. Niemi. 1992. “Political Institutions and Political Socialization: A Cross-National Study.” Comparative Politics 25 (1): 25-41. Wilcox, Clyde and Barbara Norrander. 2002. “Of Moods and Morals: The Dynamics of Opinion on Abortion and Gay Rights.” In Understanding Public Opinion. 2 nd ed. Barbara Norrander and Clyde Wilcox. Eds. Wilcox, Clyde, Lee Sigelman and Elizabeth Cook. 1989. “Some Like It Hot: Individual Differences in Responses to Group Feeling Thermometers.” Public Opinion Quarterly 53 (2): 246-257. Wiley, David E. and James A. 1970. “The Estimation of Measurement Error in Panel Data.” American Sociological Review 35 (1): 112-116. Wiley, James A. and Mary Glenn Wiley. 1974. “A Note on Correlated Errors in Repeated Measurements.” Sociological Methods and Research 3 (2): 172-188. Williams, Linda Faye. 2004. “The Issue of Our Time: Economic Inequality and Political Power in America.” Perspective on Politics 2 (4): 683-689. Wilson, Thomas C. 1986. “Interregional Migration and Racial Attitudes.” Social Forces 65 (1): 177-186. ------1996. “Cohort and Prejudice: Whites’ Attitudes Toward Blacks, Hispanics, Jews and Asians.” Public Opinion Quarterly 60 (2): 253-274. Wirls, Daniel. 1986. “Reinterpreting the Gender Gap.” Public Opinion Quarterly 50 (3): 316-330. Wleizien, Christopher. 1995. “The Public as a Thermostat: Dynamics of Preferences for Spending.” American Journal of Political Science 39 (4): 981-1000. Yeric, Jerry L. and John R. Todd. 1989. Public Opinion: The Visible Politics. 2 nd ed. Itasca, IL: F. E. Peacock. Youmans, E. Grant. 1973. “Age Stratification and Value Orientations.” International Journal of Aging and Human Development 4 (1): 53-65. Yousefi, Mahmoud. 1997. “An International Overview of Gender Inequality.” In Economic Dimensions of Gender Inequality: A Global Perspective. Jane M. Rivers and Mahmoud Yousefi. Eds. Westport, CT: Praeger. Zaller, John. 1991. “Information, Values and Opinion.” American Political Science Review 85 (4): 1215-1237. 187 ------1992. The Nature and Origins of Mass Opinion. New York: Cambridge University Press. Zaller, John and Stanley Feldman. 1992. “A Simple Theory of the Survey Response: Answering Questions Versus Revealing Preferences.” American Journal of Political Science 36 (3): 579-61. 188 Appendix A Question Wordings and Response Categories for Variables Study of Political Socialization: Parent-Child Pairs Based on Survey of Youth Panel and Their Offspring, 1997 (Jennings and Stoker 2004) Egalitarianism PANEL RESPONDENT OFFSPRING (THIRD GENERATION) V7700 46. TOO PUSHNG EQUAL RGHTS 1 V7705 51. PROB NOT EQUAL CHANCE 2 V7804 60.GOVT PRVD JOBS 3 1 46. We have gone too far in pushing equal rights in this country. 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 8. DK (“don’t know” response written in, or marked between response options) 9. NA (No response given) 2 51. One of the big problems in this country is that we don't give everyone an equal chance. 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 8. DK (“don’t know” response written in, or marked between response options) 9. NA (No response given) 3 60. Some people feel that the government in Washington should see to it that every person has a job and agood standard of living. Others think the government should just let each person get ahead on his own. And other people have opinions somewhere in between. 1 2 3 4 5 6 7 If you haven’t thought much about this issue, check this box and GO TO Question 61. If you have thought about it . . . Where would you place yourself on this scale? ________ 1. GOVT SEE TO A JOB AND GOOD STD OF LIVING 2. 3. 4. 5. 6. 7. GOVT LET EACH PERSON GET AHEAD ON THEIR OWN 96. R CHECKED ‘HAVEN’T THOUGHT ABOUT THIS’ BOX. 189 V7551 32.SHD FD GOV INT SCH 4 V7802 58.GVT HLP MNTY GPS 5 V7553 34.ANTI-GAY DISC LAWS 6 V7801 57.WOMEN EQUAL 7 98. DK (“Don’t know” response written in, or marked between response boxes) 99. NA (No response given) 4 32. Which do you think: (CHECK ONLY ONE BOX) 1. The government in Washington should see to it that white and black children go to the same schools. 2. The government in Washington should stay out of this area as it is none of its business. 0. I have no opinion on this issue. 8. DK (“don’t know” response written in, or response marked between response boxes) 5 58. Some people feel that the government in Washington should make every possible effort to improve the social and economic position of Blacks and other minority groups. Others feel that the government should not make any special effort to help minorities because they should help themselves. And other people have opinions somewhere in between. 1 2 3 4 5 6 7 If you haven’t thought much about this issue, check this box and GO TO Question 59. If you have thought about it . . . Where would you place yourself on this scale? ________ 1. GOVT SHOULD HELP MINORITY GROUPS 2. 3. 4. 5. 6. 7. MINORITY GROUPS SHOULD HELP THEMSELVES 96. R CHECKED ‘HAVEN’T THOUGHT ABOUT THIS’ BOX. 98. DK (“Don’t know” response written in, or marked between response boxes) 99. NA (No response given) 6 34. Do you favor or oppose laws that protect homosexuals against job discrimination? 1. Favor, strongly 2. Favor, but not strongly 3. Oppose, but not strongly 5. Oppose, strongly 8. DK (“don’t know” response written in, or response marked between response boxes) 9. NA (no response given) 7 57. Recently there has been a lot of talk about women's rights. Some people feel that women should have an equal role with men in running business, industry and government. Others feel that women's place is in the home. And other people have opinions somewhere in between. 1 2 3 4 5 6 7 If you haven’t thought much about this issue, check this box and GO TO Question 58. If you have thought about it . . . Where would you place yourself on this scale? ________ 1. WOMEN AND MEN SHOULD HAVE AN EQUAL ROLE 2. 3. 4. 5. 6. 190 Political Discussion V7854 63.FRQ TALK W/ PRNTS PUB AFF 8 Media Exposure V7250 11.READ POL MAGS REG 9 Political Knowledge V7310 18.WCH PRTY/HOW MRE CNSVTV 10 Group Identification V7654 44e.WOMEN S MOVEMENT 11 7. WOMEN’S PLACE IS IN THE HOME 96. R CHECKED ‘HAVEN’T THOUGHT ABOUT THIS’ BOX. 98. DK (“Don’t know” response written in, or marked between response boxes) 99. NA (No response given) 8 63. How often do you and your parents talk about any kind of public affairs and politics, that is, anything having to do with local, state, national, or international affairs? 1. Very often 2. Pretty often 3. Not very often 4. Never 8. DK (“don’t know” response written in, or marked between response options) 9. NA (No response given) 9 11. Are there any magazines that you read regularly about public affairs and politics? 1. Yes 3. Response marked between response boxes. 5. No 8. DK (“don’t know” response written in, or response marked between response boxes) 9. NA (no response given) 10 18. Would you say that either one of the two major parties is more conservative or more liberal than the other? 4. Yes, Democrats are a lot more conservative 3. Yes, Democrats are a little more conservative 2. Yes, Republicans are a little more conservative 1. Yes, Republicans are a lot more conservative 8. DK (“don’t know” response written 11 44. Please rate the following groups and individuals using the thermometer on the right. You may use any number between 0 and 100 degrees to show how favorable or unfavorable you feel. If you don’t know too much about a group, put an “X” on the line and go on to the next one. The following code was used for each of the groups or individuals listed: 000. . . . 191 Demographics V7954 77. RACE 12 V7955 78. SEX 13 Religion V7951 74.FRQNC CHRCH ATND 14 V7952 75. R INTERP OF BIBLE 15 V7953 76.RELIGION GUIDANCE 16 . 100. 998. DK (marked X or “don’t know” ) 999. NA The Women’s Movement 12 77. What is your race? 1. White 2. Black 3. Other (PLEASE SPECIFY) 9. NA (No response given) 13 78. What is your sex? 1. Male 2. Female 9. NA (No response given) 14 74. How often do you go to church or synagogue? 1. Every week 2. Almost every week 3. Once or twice a month 4. A few times a year 5. Never 8. DK (“don’t know” response written in, or marked between response options) 9. NA (No response given) 15 75. Here are four statements about the Bible. Please check the one that comes closest to your own view. 1. The Bible is God’s word and all it says is true. 2. The Bible was written by men inspired by God but it contains some human errors. 3. The Bible is a good book because it was written by wise men, but God had nothing to do with it. 4. The Bible was written by men who lived so long ago that it is worth very little today. 8. DK (“don’t know” response written in, or marked between response options) 9. NA (No response given) 16 76. Would you say that religion provides some guidance in your day-to-day living, quite a bit of guidance, or a great deal of guidance in your day-to-day life? 1. Some 3. Quite a bit 5. A great deal 192 PARENTS OF PANEL OFFSPRING (FROM 1997 WAVE OF HIGH SCHOOL SENIORS FROM CLASS OF 1965) Egalitarianism V6100 K1a.TOO PSHNG EQUL RGHTS 17 V6105 K1f.PROB NOT EQUL CHANCE 18 V5250 C1.GOVT PRVD JOBS 19 7. None 8. DK (“Don’t know” response written in, or marked between response options) 9. NA (No response given) 17 (R.B., P.15) Now, please look at page 15. I'm going to read some more of the kinds of things people tell us when we interview them and ask you whether you agree or disagree with them. Here is the first statement: We have gone too far in pushing equal rights in this country. Do you agree or disagree with this? 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 8. DK 9. NA 0. INAP (partial iws) 18 (R.B., still on P.15) One of the big problems in this country is that we don't give everyone an equal chance. (Do you agree or disagree?) 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 8. DK 9. NA 0. INAP (partial iws) 19 C1. (R.B., P.2) Now please look at page 2 of the booklet. Some people feel that the government in Washington should see to it that every person has a job and a good standard of living. Suppose these people are at one end of this scale—at point number 1. Others believe that the government should let each person get ahead on his or her own. Suppose these people are at the other end—at point number 7. And other people have opinions somewhere in between, at points 2,3,4,5,or 6. Where would you place yourself on this scale, or haven't you thought much about this? 1. GOVT SEE TO A JOB AND GOOD STD OF LIVING 2. 3. 4. 5. 6. 7. GOVT LET EACH PERSON GET AHEAD ON THEIR OWN 96. HAVEN’T THOUGHT ABOUT THIS 98. DK 99. NA 193 V5252 C3.GVT HLP MNTY GP 20 V5314 C11a.SHD FD GOV INT SCH 21 V5254 C5.WOMEN EQUAL 22 V5360 C16k.SCHL BD NOT HIRE GAY 23 20 C3. (R.B., P.4) Now, page 4. Some people feel that the government in Washington should make every possible effort to improve the social and economic position of blacks and other minority groups. Others feel that the government should not make any special effort to help minorities because they should help themselves. (And other people have opinions somewhere in between). Where would you place yourself on this scale, or haven't you thought much about this? 1. GOVT SHOULD HELP MINORITY GROUPS 2. 3. 4. 5. 6. 7. MINORITY GROUPS SHOULD HELP THEMSELVES 96. HAVEN’T THOUGHT ABOUT THIS 98. DK 99. NA 21 Do you think the government in Washington should see to it that white and black children go to the same schools or stay out of the area as it is none of its business? 1. SEE TO IT THAT WHITE AND BLACK CHILDREN GO TO THE SAME SCHOOLS 2. R SELECTED “1” BUT ALSO MADE A COMMENT AGAINST FORCED INTEGRATION/BUSING OR R SAID “GOV’T SHOULD SEE TO IT THAT WHITE AND BLACK CHILDREN ARE ALLOWED TO GO TO THE SAME SCHOOLS BUT NOT FORCE THEM TO DO SO” 3. (VOLUNTEERED) DEPENDS, IN BETWEEN, SOME GOVERNMENT HELP IS NEEDED 4. (VOLUNTEERED) AGAINST FORCED INTEGRATION/BUSING (DID NOT PICK 1 OR 5) 5. STAY OUT OF THE AREA AS IT IS NONE OF ITS BUSINESS. 7. OTHER (SPECIFY) 8. DK 9. NA 0. INAP (5, 8 or 9 in C11) 22 C5. (R.B., P.6) Now, page 6. Recently there has been a lot of talk about women's rights. Some people feel that women should have an equal role with men in running business, industry and government. Others feel that women's place is in the home. (And other people have opinions somewhere in between). Where would you place yourself on this scale, or haven't you thought much about this? 1. WOMEN AND MEN SHOULD HAVE AN EQUAL ROLE 2. 3. 4. 5. 6. 7. WOMEN’S PLACE IS IN THE HOME 96. HAVEN’T THOUGHT ABOUT THIS 98. DK 23 A school board should not hire a person to teach if that person is an admitted homosexual. (Do you agree or disagree with this?) 1. AGREE 3. (VOLUNTEERED) DEPENDS, BOTH 5. DISAGREE 8. DK 194 Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined (Kent et. al. [1983] 1991) Dependent Variables Egalitarianism R'S POSITION ON GUARANTEED JOB AND LIVING STANDARD SCALE: Var 356 (Parents 1973), Var 1221 (Students 1982). 24 R'S POSITION ON GOVERNMENT AID TO MINORITIES SCALE: Var 0343 (Students 1973), Var 371 (Parents 1973), Var 1233 (Students 1982). 25 R'S POSITION ON SCHOOL BUSING SCALE: Var 0357 (Students 1973), Var 385 (Parents 1973). 26 9. NA 24 Q.C1. SOME PEOPLE FEEL THAT THE GOVERNMENT IN WASHINGTON SHOULD SEE TO IT THAT EVERY PERSON HAS A JOB AND A GOOD STANDARD OF LIVING. OTHERS THINK THE GOVERNMENT SHOULD JUST LET EACH PERSON GET AHEAD ON HIS OWN. AND, OF COURSE, OTHER PEOPLE HAVE OPINIONS SOMEWHERE IN BETWEEN. SUPPOSE PEOPLE WHO BELIEVE THAT THE GOVERNMENT SHOULD SEE TO IT THAT EVERY PERSON HAS A JOB AND A GOOD STANDARD OF LIVING ARE AT ONE END OF THIS SCALE--AT POINT NUMBER 1. AND SUPPOSE THAT THE PEOPLE WHO BELIEVE THAT THE GOVERNMENT SHOULD LET EACH PERSON GET AHEAD ON HIS OWN ARE AT THE OTHER END--AT POINT 7. Q.C1A. WHERE WOULD YOU PLACE YOURSELF ON THIS SCALE, OR HAVEN'T YOU THOUGHT MUCH ABOUT THIS? 1. GOVERNMENT SHOULD SEE TO JOB AND GOOD STANDARD OF LIVING 2. 3. 4. 5. 6. 7. GOVERNMENT SHOULD LET EACH PERSON GET AHEAD ON HIS OWN 25 Q.C3. SOME PEOPLE FEEL THAT THE GOVERNMENT IN WASHINGTON SHOULD MAKE EVERY POSSIBLE EFFORT TO IMPROVE THE SOCIAL AND ECONOMIC POSITION OF BLACKS AND OTHER MINORITY GROUPS. OTHERS FEEL THAT THE GOVERNMENT SHOULD NOT MAKE ANY SPECIAL EFFORT TO HELP MINORITIES BECAUSE THEY SHOULD HELP THEMSELVES. Q.C3A. WHERE WOULD YOU PLACE YOURSELF ON THIS SCALE, OR HAVEN'T YOU THOUGHT MUCH ABOUT IT? 1. GOVERNMENT SHOULD HELP MINORITY GROUPS 2. 3. 4. 5. 6. 7. MINORITY GROUPS SHOULD HELP THEMSELVES 195 SHOULD FEDERAL GOVERNMENT ENSURE SCHOOL INTEGRATION: Var 389 (Students 1973), Var 417 (Parents 1973), Var 0121 (Students 1982). 27 R'S POSITION ON EQUAL RIGHTS FOR WOMEN SCALE: Var 0364 (Students 1973), Var 392 (Parents 1973), Var 0098 (Students, Parents 1982). 28 Independent Variables POLITICAL KNOWLEDGE LENGTH OF U.S. SENATE TERM: Var 0473 (Students 1973). 29 26 Q.C5. THERE IS MUCH DISCUSSION ABOUT THE BEST WAY TO DEAL WITH RACIAL PROBLEMS. SOME PEOPLE THINK ACHIEVING RACIAL INTEGRATION OF SCHOOLS IS SO IMPORTANT THAT IT JUSTIFIES BUSING CHILDREN TO SCHOOLS OUT OF THEIR NEIGHBORHOODS. OTHERS THINK LETTING CHILDREN GO TO THEIR NEIGHBORHOOD SCHOOLS IS SO IMPORTANT THAT THEY OPPOSE BUSING. Q.C5A. WHERE WOULD YOU PLACE YOURSELF ON THIS SCALE, OR HAVEN'T YOU THOUGHT MUCH ABOUT THIS? 1. BUS TO ACHIEVE INTEGRATION 2. 3. 4. 5. 6. 7. KEEP CHILDREN IN NEIGHBORHOOD SCHOOLS 27 Q.C10. "SOME PEOPLE SAY THAT THE GOVERNMENT IN WASHINGTON SHOULD SEE TO IT THAT WHITE AND BLACK CHILDREN ARE ALLOWED TO GO TO THE SAME SCHOOLS. OTHERS CLAIM THAT THIS IS NOT THE GOVERNMENT'S BUSINESS." HAVE YOU BEEN CONCERNED ENOUGH ABOUT THIS QUESTION TO FAVOR ONE SIDE OVER THE OTHER? Q.C10A. (IF YES) DO YOU THINK THAT TH GOVERNMENT IN \ WASHINGTON SHOULD -- 1. (YES) SEE TO IT THAT WHITE AND BLACK CHILDREN GO TO THE SAME SCHOOLS 3. (YES) OTHER, DEPENDS, BOTH BOXES CHECKED 5. (YES) STAY OUT OF THIS AREA AS IT IS NONE OF ITS BUSINESS 28 Q.C6. RECENTLY THERE HAS BEEN A LOT OF TALK ABOUT WOMEN'S RIGHTS. SOME PEOPLE FEEL THAT WOMEN SHOULD HAVE AN EQUAL ROLE WITH MEN IN RUNNING BUSINESS, INDUSTRY, AND GOVERNMENT. OTHERS FEEL THAT WOMEN'S PLACE IS IN THE HOME. Q.C6A. WHERE WOULD YOU PLACE YOURSELF ON THIS SCALE, OR HAVEN'T YOU THOUGHT MUCH ABOUT THIS? 1. WOMEN AND MEN SHOULD HAVE AN EQUAL ROLE 2. 3. 4. 5. 6. 7. WOMEN'S PLACE IS IN THE HOME 196 WHERE IS TITO A LEADER: Var 0474 (Students 1973). 30 NUMBER OF MEMBERS OF U.S. SUPREME COURT: Var 0475 (Students 1973). 31 NAME OF GOVERNOR OF R'S STATE: Var 0476 (Students 1973). 32 LOCATION OF W.W.II CONCENTRATION CAMPS: Var 0477 (Students 1973). 33 WAS F.D.R. A DEMOCRAT OR REPUBLICAN: Var 0478 (Students 1973). 34 29 Q.E4. NEXT, I'D LIKE TO ASK YOU A FEW QUESTIONS THAT YOU MAY OR MAY NOT BE ABLE TO ANSWER. WE DON'T EXPECT PEOPLE TO KNOW ALL THE ANSWERS. Q.E4A. ABOUT HOW MANY YEARS DOES A U.S. SENATOR SERVE? 1. ONE 2. TWO 3. THREE 4. FOUR 5. FIVE 6. SIX 7. SEVEN OR ABOVE 30 Q.E4B. MARSHALL TITO IS A LEADER IN WHAT COUNTRY? 1. YUGOSLAVIA 2. OTHER EASTERN EUROPEAN SATELLITE CZECHOSLOVAKIA RUMANIA BULGARIA HUNGARY POLAND ALBANIA E. GERMANY 3. USSR (RUSSIA) 4. CHINA 5. ANY OTHER COUNTRY 31 Q.E4C. DO YOU HAPPEN TO KNOW ABOUT HOW MANY MEMBERS THERE ARE ON THE UNITED STATES SUPREME COURT? 01. . . 97. 97 AND ABOVE 32 Q.E4D. WHO IS THE GOVERNOR OF (THIS STATE) NOW? 1. CORRECT ANSWER 5. INCORRECT ANSWER 33 Q.E4E. DURING WORLD WAR II, WHICH NATION HAD A GREAT MANY CONCENTRATION CAMPS FOR JEWS? 1. GERMANY 2. RUSSIA 3. OTHER COUNTRY 34 Q.E4F. DO YOU HAPPEN TO RECALL WHETHER PRESIDENT FRANKLIN ROOSEVELT WAS A REPUBLICAN OR A DEMOCRAT? WHICH? 1. REPUBLICAN 2. DEMOCRAT 3. OTHER 197 MEDIA EXPOSURE FREQUENCY WATCH TV ABOUT POLITICS: Var 314 (Students 1973). 35 FREQUENCY/LISTEN RADIO POLITICS: Var 315 (Students 1973). 36 REGULARLY READ CERTAIN MAGAZINES ON POLITICS Var 0316 (Students 1973). 37 R'S ESTIMATED FAMILY INCOME: Var 0602 (Students 1982). 38 35 Q.B2. HOW ABOUT TELEVISION - DO YOU WATCH ANY PROGRAMS ABOUT PUBLIC AFFAIRS, POLITICS, AND THE NEWS ON TELEVISION? Q.B2A. (IF YES) ABOUT HOW OFTEN DO YOU WATCH SUCH PROGRAMS - ALMOST DAILY, TWO OR THREE TIMES A WEEK, THREE OR FOUR TIMES A MONTH, OR A FEW TIMES A YEAR? 1. YES, ALMOST DAILY 2. YES, TWO OR THREE TIMES A WEEK 3. YES, THREE OR FOUR TIMES A MONTH 4. YES, A FEW TIMES A YEAR 5. NO 36 Q.B3. HOW ABOUT RADIO - DO YOU LISTEN TO ANY PROGRAMS ABOUT PUBLIC AFFAIRS, POLITICS, AND THE NEWS ON THE RADIO. Q.B3A. (IF YES) HOW OFTEN DO YOU LISTEN TO THEM ON RADIO - ALMOST DAILY, TWO OR THREE TIMES A WEEK, THREE OR FOUR TIMES A MONTH, OR A FEW TIMES A YEAR. 1. YES, ALMOST DAILY 2. YES, TWO OR THREE TIMES A WEEK 3. YES, THREE OR FOUR TIMES A MONTH 4. YES, A FEW TIMES A YEAR 5. NO 37 Q.B4. FINALLY, HOW ABOUT MAGAZINES - DO YOU READ ABOUT PUBLIC AFFAIRS AND POLITICS IN ANY MAGAZINES. 1. YES 5. NO 38 Q.M30. PLEASE LOOK AT THIS CARD AND TELL ME THE LETTER OF THE INCOME GROUP THAT INCLUDES THE INCOME OF ALL MEMBERS OF YOUR FAMILY IN 1972 BEFORE TAXES. THIS FIGURE SHOULD INCLUDE SALARIES, WAGES, DIVIDENDS, INTEREST, PENSIONS, AND ALL OTHER INCOME. WHAT WOULD BE YOUR BEST GUESS? 01. NONE 02. LESS THAN $1,000 03. $1,000 TO $1,999 04. $2,000 TO $2,999 05. $3,000 TO $3,999 06. $4,000 TO $4,999 07. $5,000 TO $5,999 08. $6,000 TO $6,999 09. $7,000 TO $7,999 10. $8,000 TO $8,999 198 R'S RELIGIOUS BELIEFS HOW OFTEN DOES R GO TO CHURCH: Var 0749 (Students 1973). 39 R'S INTERPRETATION OF THE BIBLE: Var 0750 (Students 1973). 40 RACE, SEX R'S SEX: Var 0609 (Students 1982). 41 RESPONDENT'S RACE: Var 0610 (Students 1982). 42 11. $9,000 TO $9,999 12. $10,000 TO $10,999 13. $11,000 TO $11,999 14. $12,000 TO $14,999 15. $15,000 TO $19,999 16. $20,000 TO $24,999 17. $25,000 TO $34,999 18. $35,000 AND OVER 39 Q.M33C. (IF ANY RELIGIOUS PREFERENCE) WOULD YOU SAY YOU GO TO (CHURCH/SYNAGOGUE) EVERY WEEK, ALMOST EVERY WEEK, ONCE OR TWICE A MONTH, A FEW TIMES A YEAR, OR NEVER? 1. EVERY WEEK 2. ALMOST EVERY WEEK 3. ONCE OR TWICE A MONTH 4. A FEW TIMES A YEAR 5. NEVER 40 Q.M34. HERE ARE FOUR STATEMENTS ABOUT THE BIBLE AND I’D LIKE YOU TO TELL ME WHICH IS CLOSEST TO YOUR OWN VIEW. 1. THE BIBLE IS GOD'S WORD AND ALL IT SAYS IS TRUE 2. THE BIBLE WAS WRITTEN BY MEN INSPIRED BY GOD BUT IT CONTAINS SOME HUMAN ERROR 3. THE BIBLE IS A GOOD BOOK BECAUSE IT WAS WRITTEN BY WISE MEN BUT GOD HAD NOTHING TO DO WITH IT 4. THE BIBLE WAS WRITTEN BY MEN WHO LIVED SO LONG AGO THAT IT IS WORTH VERY LITTLE TODAY 41 1. MALE 2. FEMALE 42 1. WHITE 2. BLACK 199 American National Election Studies Cumulative Data File, 1948–2000 (Sapiro et. al. 2001) Dependent Variables Egalitarianism SHOULD SOCIETY INSURE EQUAL OPPORTUNITY (VCF9013) 43 HAVE GONE TOO FAR PUSHING EQUAL RIGHTS (VCF9014) 44 BIG PROBLEM IF CHANCES NOT EQUAL (VCF9015) 45 43 1988 and later: Using page XXX for our choice of answers, I would like you to tell me how much you agree or disagree with each statement. (RESPONDENT BOOKLET SHOWN TO R) 1984, 1986, 1990: I am going to read several statements. After each one, I would like you to tell me whether you agree strongly, agree somewhat, neither agree nor disagree, disagree somewhat, or disagree strongly. (RESPONDENT BOOKLET SHOWN TO R) 'Our society should do whatever is necessary to make sure that everyone has an equal opportunity to succeed.' 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 44 1988 and later: Using page XXX for our choice of answers, I would like you to tell me how much you agree or disagree with each statement. (RESPONDENT BOOKLET SHOWN TO R) 1984, 1986, 1990: I am going to read several statements. After each one, I would like you to tell me whether you agree strongly, agree somewhat, neither agree nor disagree, disagree somewhat, or disagree strongly. (RESPONDENT BOOKLET SHOWN TO R) 'We have gone too far in pushing equal rights in this country.' 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 45 1988 and later: Using page XXX for our choice of answers, I would like you to tell me how much you agree or disagree with each statement. (RESPONDENT BOOKLET SHOWN TO R) 1984, 1986, 1990: I am going to read several statements. After\ each one, I would like you to tell me whether you agree strongly, agree somewhat, neither agree nor disagree, disagree somewhat, or disagree strongly. (RESPONDENT BOOKLET SHOWN TO R) 'One of the big problems in this country is that we don't give everyone an equal chance.' 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 200 OK IF SOME HAVE MORE EQUAL CHANCE (VCF9016) 46 SHOULD WORRY LESS ABOUT EQUALITY (VCF9017) 47 FEWER PROBLEMS IF MORE EQUAL TREATMENT (VCF9018) 48 7PT GOVT GUARANTEED JOBS SCALE (VCF0809) 49 46 1988 and later: Using page XXX for our choice of answers, I would like you to tell me how much you agree or disagree with each statement. (RESPONDENT BOOKLET SHOWN TO R) 1984, 1986, 1990: I am going to read several statements. After each one, I would like you to tell me whether you agree strongly, agree somewhat, neither agree nor disagree, disagree somewhat, or disagree strongly. (RESPONDENT BOOKLET SHOWN TO R) 'It is not really that big a problem if some people have more of a chance in life than others.' 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 47 1988 and later: Using page XXX for our choice of answers, I would like you to tell me how much you agree or disagree with each statement. (RESPONDENT BOOKLET SHOWN TO R) 1984, 1986, 1990: I am going to read several statements. After each one, I would like you to tell me whether you agree strongly, agree somewhat, neither agree nor disagree, disagree somewhat, or disagree strongly. (RESPONDENT BOOKLET SHOWN TO R) 'The country would be better off if we worried less about how equal people are.' 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 48 1988 and later: Using page XXX for our choice of answers, I would like you to tell me how much you agree or disagree with each statement. (RESPONDENT BOOKLET SHOWN TO R) 1984, 1986, 1990: I am going to read several statements. After each one, I would like you to tell me whether you agree strongly, agree somewhat, neither agree nor disagree, disagree somewhat, or disagree strongly. (RESPONDENT BOOKLET SHOWN TO R) 'If people were treated more equally in this country we would have many fewer problems.' 1. AGREE STRONGLY 2. AGREE SOMEWHAT 3. NEITHER AGREE NOR DISAGREE 4. DISAGREE SOMEWHAT 5. DISAGREE STRONGLY 49 Some people feel that the government in Washington should see to it that every person has a job and a good standard of living. Suppose these people are at one end of a scale, at point 1. Others think the government should just let each person get ahead on his/their own. Suppose these people are at the other end, at point 7. And, of course, some other people have opinions somewhere in between, at pints 2, 3, 4, 5 or 6. Where would you place yourself on this scale, or haven't you thought much about this? (7- POINT SCALE SHOWN TO R) 1. Government see to job and good standard of living 2. 201 7PT AID TO BLACKS SCALE (VCF0830) 50 7PT WOMENS EQUAL ROLE SCALE (VCF0834) 51 Independent Variables RESPONDENT AGE GROUP (VCF0102) 52 3. 4. 5. 6. 7. Government let each person get ahead on his own 50 1970-1984, 1986 FORM B, 1988 FORM B: Some people feel that the government in Washington should make every possible effort to improve the social and economic position of blacks and other minority groups. Others feel that the government should not make any special effort to help minorities because they should help themselves. 1986 FORM A, 1988 FORM A, 1990 and later: Some people feel that the government in Washington should make every possible effort to improve the social and economic position of blacks. Suppose these people are at one end of a scale, at point 1. Others feel that the government should not make any special effort to help blacks because they should help themselves. Suppose these people are at the other end, at point 7. And, of course, some other people have opinions somewhere in between, at points 2, 3, 4, 5 or 6. ALL YEARS: Where would you place yourself on this scale, or haven't you thought much about it? (7- POINT SCALE SHOWN TO R) 1. Government should help minority groups/blacks 2. 3. 4. 5. 6. 7. Minority groups/ blacks should help themselves 51 Recently there has been a lot of talk about women's rights. Some people feel that women should have an equal role with men in running business, industry and government. Others feel that a women's place is in the home. Where would you place yourself on this scale or haven't you thought much about this? (7-POINT SCALE SHOWN TO R) 1. Women and men should have an equal role 2. 3. 4. 5. 6. 7. Women's place is in the home 52 1. 17 - 24 2. 25 - 34 3. 35 - 44 4. 45 - 54 5. 55 - 64 6. 65 - 74 202 RESPONDENT COHORT (VCF0103) 53 RESPONDENT GENDER (VCF0104) 54 RESPONDENT RACE (3) (VCF0106a) 55 R EDUCATION (3) (VCF0140a) 56 INCOME (VCF0114) 57 7. 75 - 99 53 1. 1975 - Present 2. 1959 - 1974 3. 1943 - 1958 4. 1927 - 1942 5. 1911 - 1926 6. 1895 - 1910 7. Before 1895 54 1. Male 2. Female 55 1. White 2. Black 3. Asian 4. Native American 5. Hispanic 7. Other 56 1. 8 grades or less ('grade school') 2. 9-12 grades ('high school'), no diploma/equivalency 3. 12 grades, diploma or equivalency 4. 12 grades, diploma or equivalency plus non-academic training 5. Some college, no degree; junior/community college level degree (AA degree) 6. BA level degrees 7. Advanced degrees incl. LLB 57 1972-1990, 1992 long-form, 1994-later: Please look at this card/page and tell me the letter of the income group that includes the income of all members of your family living here in <previous year> before taxes. This figure should include salaries, wages, pensions, dividends, interest, and all other income. (IF UNCERTAIN:) What would be your best guess? 1992 short form: Can you give us an estimate of your total family income in 1991 before taxes? This figure should include salaries, wages, pensions, dividends, interest and all other income for every member of your family living in your house in 1991. First could you tell me if that was above or below $24,999? (IF UNCERTAIN: what would be your best guess?) (IF ABOVE/BELOW $24,999:) I will read you some income categories, could you please stop me when I reach the category that corresponds to your family situation? The percentiles appearing here correspond to the following: 0-16 17-33 34-67 68-95 96-100 YEAR PERCENTILE PERCENTILE PERCENTILE PERCENTILE PERCENTILE 1984 none-$6999 $7000-12999 $13000-29999 $30000-59999 $60000 + 1986 none-$8999 $9000-14999 $15000-34999 $35000-74999 $75000 + 1990 none-$9999 $10000-16999 $17000-34999 $35000-89999 $90000 + 1992 none-$9999 $10000-19999 $20000-39999 $40000-89999 $90000 + 1994 none-$11999 $12000-21999 $22000-44999 $45000-104999 $105000 + 203 HOUSEHOLD UNION MEMBERSHIP (VCF0127) 58 CHURCH ATTENDANCE (1) (VCF0130) 59 POLITICAL SOUTH/NONSOUTH (VCF0113) 60 1. 0 to 16 percentile 2. 17 to 33 percentile 3. 34 to 67 percentile 4. 68 to 95 percentile 5. 96 to 100 percentile 58 (1956-1984: Does anyone) (1986-later: Do you or else) in this household belong to a labor union? (IF YES:) Who is it that belongs? 1. Yes, someone in household belongs to a labor union 2. No, no one in household belongs to a labor union 59 1970-1988: (IF ANY RELIGIOUS PREFERENCE) Would you say you/do you go to (church/synagogue) every week, almost every week, once or twice a month, a few times a year, or never? 1990 and later: Lots of things come up that keep people from attending religious services even if they want to. Thinking about your life these days, do you ever attend religious services, apart from occasional weddings, baptisms or funerals? (IF YES:) Do you go to religious services every week, almost every month, a few times a year, or never? Beginning in 1990, question flow was reversed, so that the religious service attendance question preceded the religious affiliation questions, whereas in previous years religious service attendance followed the religious affiliation questions. Also note that beginning in 1990 the determination of frequency of religious service attendance was preceded by a general attendance filter question (yes/no); a "no" response to the filter question 1990 and later has been coded 5. 1. Every week 2. Almost every week 3. Once or twice a month 4. A few times a year 5. Never (1990 and later: 'No' in filter) 7. No religious preference (1970-1988) 60 QUESTION: --------- Region - political south VALID_CODES: ------------ 1. South 2. Nonsouth NOTES: ------ GENERAL NOTE: This var identifies the 11 secession states: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, and Virginia. These states cannot be identified for 1954. This identification of 'south' does NOT conform to the definition 204 HOW MANY DAYS R WATCHED NATIONAL TV NEWS (VCF9035) 61 HOW OFTEN R READ NEWSPAPER IN LAST WEEK (VCF9033) 62 SUMMARY- RECALL OF HOUSE CANDIDATE NAMES (VCF0976) 63 used in VCF0112 (Census-defined region). 61 HOW MANY DAYS R WATCHED NATIONAL TV NEWS (VCF9035) How many days in the past week did you watch national news (1986: a national news program) (1984,1990-1994: the news; 1996-2000: the national news) on TV? Filter: in 1984 only, this variable was preceded by the question, "How often do you watch the national network news on TV...?" and if R replied that s/he watched national network news at all, then R was asked how many days in the past week s/he watched the news. NOTE: distributions vary considerably when the filter is administered (1984). Actual number of days coded. 0. NONE 1. ONE DAY . . 7. EVERY DAY 62 HOW OFTEN R READ NEWSPAPER IN LAST WEEK (VCF9033) How many days in the past week did you read a daily newspaper? Actual number of days coded. 0. NONE 1. ONE DAY . . 7. EVERY DAY 63 SUMMARY- RECALL OF HOUSE CANDIDATE NAMES (VCF0976) Do you happen to remember the names of the candidates for Congress-- that is, for the House of Representatives in Washington--[1996-later: Who ran in the November election district in November? NO NAME CORRECT 00. R's recall of any name(s) incorrect/DK/NA; any party mention(s) OR party mentions DK/NA; R says s/he recalls no names in VCF0972 ONE CORRECT NAME 10. R's recall correct for one name; party mention incorrect/ DK/NA 11. R's recall correct for one name; party mention correct TWO CORRECT NAMES 205 HOW OFTEN R DISCUSSES POLITICS (VCF0732) 64 7-PT SCALE PARTY IDENTIFICATION (VCF0301) 65 BLACKS THERMOMETER (VCF0206) 66 20. R's recall correct for 2 names; both party mentions incorrect/ DK/NA 21. R's recall correct for 2 names; one party mention correct 22. R's recall correct for 2 names; both party mentions correct THREE CORRECT NAMES 30. R's recall correct for 3 names; all party mentions incorrect/DK/NA 31. R's recall correct for 3 names; one party mention correct 32. R's recall correct for 3 names; two party mentions correct 33. R's recall correct for 3 names; all party mentions correct 64 Do you ever discuss politics with your family or friends? (RESPONDENT BOOKLET) How often do you discuss politics with your family or friends -- every day, 3 or 4 times a week, once or twice a week, or less often than that? 1. Every day 2. 3 or 4 times a week 3. Once or twice a week 4. Less often 5. Never / No (to lead-in) 65 QUESTION: --------- Generally speaking, do you usually think of yourself as a Republican, a Democrat, an Independent, or what? (IF REPUBLICAN OR DEMOCRAT) Would you call yourself a strong (REP/DEM) or a not very strong (REP/DEM)? (IF INDEPENDENT, OTHER [1966 AND LATER: OR NO PREFERENCE]:) Do you think of yourself as closer to the Republican or Democratic party? VALID_CODES: ------------ 1. Strong Democrat 2. Weak Democrat 3. Independent - Democrat 4. Independent - Independent 5. Independent - Republican 6. Weak Republican 7. Strong Republican 9. Apolitical (1966 only: and DK) 66 1978-1984: (1980-1984: And still using the thermometer, how would you rate the following.) PRECEDING THERMOMETER DESCRIPTION FOR POLITICAL LEADERS, 1978-1984: I'd like to get your feelings toward some of our political leaders and other people who are in the news these days. I'll read the name of a person and I'd like you to rate that person using the feeling 206 LABOR UNIONS THERMOMETER (VCF0210) 67 POOR PEOPLE THERMOMETER (VCF0223) 68 WOMENS LIBBERS THERMOMETER (VCF0225) 69 thermometer. Ratings between 50 degrees and 100 degrees mean that you feel favorably and warm toward the person; ratings between 0 and 50 degrees mean that you don't feel favorably toward the person and that you don't care too much for that person. If we come to a person whose name you don't recognize, you don't need to rate that person. Just tell me and we'll move on to the next one. If you do recognize the name, but you don't feel particularly warm or cold toward the person, then you would rate the person at the 50 degree mark. 1986-later: (And still using the thermometer, how would you rate the following) PRECEDING THERMOMETER DESCRIPTION FOR POLITICAL LEADERS, 1986-later: I'd like to get your feelings toward some of our political leaders and other people who are in the news these days (1990: who have been in the news). I'll read the name of a person and I'd like you to rate that person using something we call the feeling thermometer. Ratings between 50 and 100 degrees mean that you feel favorably and warm toward the person; ratings between 0 and 50 degrees mean that you don't feel favorably toward the person and that you don't care too much for that person. You would rate the person at the 50 degree mark if you don't feel particularly warm or cold toward the person. If we come to a person whose name you don't recognize, you don't need to rate that person. Just tell me and we'll move on to the next one. In all presidential years except 1968 and 1976, group thermometers followed, within the same (wave) interview if not immediately, the series of thermometers for specific political figures: R was asked to recall for the group thermometers the "explanation" of the thermometer itself which was given for the individual thermometers. 00-96. Degrees as coded 97. 97-100 Degrees 98. DK (exc. 1964-1968: SEE VCF0201 note) 99. NA; INAP, no post IW 1972,1976,1980,1984,1988,1992,1996); form III, IV (1972); question not used 67 00-96. Degrees as coded 97. 97-100 Degrees 98. DK (exc. 1964-1968: SEE VCF0201 note) 99. NA; INAP, no post IW (1972,1976,1980,1984,1988,1992,1996); form III, IV (1972); question not used 68 00-96. Degrees as coded 97. 97-100 Degrees 98. DK 99. NA; INAP, no post IW (1972,1976,1980,1984,1988,1992,1996); form III, IV (1972); question not used 69 00-96. Degrees as coded 97. 97-100 Degrees 98. DK 99. NA; INAP, no post IW (1972,1976,1980,1984,1988,1992,1996); form III, IV (1972); question not used 207 Appendix B Methodological Appendix The structural equation model is as follows: 1 (1). The measurement model, a system of equations: (a). For dependent variables: y i = Λ yi η + ε i where the y i s are the observed variable, the Λ yi s are the coefficients (factor loadings), η is the latent variable and the ε i s are the error terms. (a). For independent variables: x i = Λ yi ξ + δ i where the x i s are the observed variables, the Λ yi s are again the coefficients, ξ is the latent variable and the δ i s are the error terms. (2). The analysis model (or structural equations): η = Γ i ξ i + ζ, where η is the dependent variable, Γs are the coefficients, ξ i s are the independent variables and ζ is the error term (Long 1983, 15-19). The above model assumes the variables are continuous. Rather than analyzing a covariance matrix for categorical variables, as is done in the case of continuous variables, polychoric (correlations between two categorical variables) or polyserial correlations (those between categorical and continuous variables) are analyzed. Polychoric and polyserial correlations are computed based on the assumption that the observed 1 For those not familiar with structural equation models, think of them as regression and factor models solved simultaneously, or regression models with latent variables. 208 categorical variables are a function of unobserved continuous normally distributed latent variables linked to the observed variables by a probit link function: η = Φ -1 ( Σ β k x k ) where η is the latent normally distributed variable, Φ -1 is an inverse cumulative normal distribution (inverse z score), the x k s are the observed variables and the β k s are their coefficients (Liao 1994, 21). The equation for multi-group mean and covariance structure is derived as follows: Taking the expectations of an ordinary structural equation model for an uncentered variable: y i = α yi + Λ yi η + ε i gives: μ yi = α yi + Λ yi μ ηij + ε i where the means are for the ith indicator for the jth group, the Λs are the coefficients, ε i s are the error terms and α is the y intercept (Byrne 1994, 192) 209 Appendix C Measurement Model Parameter Estimates Table B.1. Measurement Model Parameters for Youth 1997 Egalitarianism Coefficients (Standard Errors and Test Statistics in Parentheses) Push Equal Rights Too Far 1 Unequal Chances Not a Problem .993 (.086) *** Govt. Provide Jobs .871 (.086) *** Fed. Govt. Integrate Schools .604 (.085) *** Govt. Help Minorities 1.217 (.147) *** Women Equal -.309 (.100) ** Anti-Gay Discrimination Laws .157 (.086) N = 769 * = p < 0.05 ** = p < 0.01 *** = p > 0.001 Source: Study of Political Socialization: Parent-Child Pairs Based on Survey of Youth Panel and Their Offspring, 1997. a a Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) y 5 = Φ -1 ( Λ 5 η + ε 5 ) y 6 = Φ -1 ( Λ 6 η + ε 6 ) y 7 = Φ -1 ( Λ 6 η + ε 6 ) where η = Egalitarianism y 1 = Push Equal Rights Too Far y 2 = Unequal Chances Not a Problem y 3 = Govt. Provide Jobs y 4 = Fed. Govt. Integrate Schools y 5 = Govt. Help Minorities y 6 = Women Equal y 7 = Anti-Gay Discrimination Laws Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 210 Table B.2. Measurement Model Parameters for Parents’ Egalitarianism (in Youth 1997 Egalitarianism Model) Coefficients (Standard Errors and Test Statistics in Parentheses) Push Equal Rights Too Far .199 (.034) *** Govt. Provide Jobs .405 (.034) *** Govt. Help Minorities .512 (.038) *** Women Equal 1 School Bd. Should Not Hire Gay .387 (.034) *** Fed. Govt. Integrate Schools .361 (.033) *** N = 769 * = p < 0.05 ** = p < 0.01 *** = p > 0.001 Source: Study of Political Socialization: Parent-Child Pairs Based on Survey of Youth Panel and Their Offspring, 1997. b b Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) y 5 = Φ -1 ( Λ 5 η + ε 5 ) y 6 = Φ -1 ( Λ 6 η + ε 6 ) where η = Egalitarianism y 1 = Push Equal Rights Too Far y 2 = Govt. Provide Jobs y 3 = Govt. Help Minorities y 4 = Women Equal y 5 = School Bd. Should Not Hire Gay y 6 = Fed. Govt. Integrate Schools Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 211 Table B.3. Measurement Model Parameters for Youth 1982 Egalitarianism Coefficients (Standard Errors and Test Statistics in Parentheses) Political Knowledge/Group Identification Model Demographics/Religiosit y Model Govt. Provide Jobs c 1 - Fed. Govt. Integrate Schools .737 (.076) *** .779 (.079) *** Women Equal .539 (.083) *** .589 (.084) *** Should Govt. Help Minorities d - 1 N = 769 * = p < 0.05 ** = p < 0.01 *** = p > 0.001 Source: Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined. e c Excluded from the model to render the polychoric correlation matrix positive definite. d Excluded from the model to render the polychoric correlation matrix positive definite. e Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) where η = Egalitarianism y 1 = Govt. Provide Jobs y 2 = Fed. Govt. Integrate Schools y 3 = Women Equal y 4 = Should Govt. Help Minorities Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 212 Table B.4. Measurement Model Parameters for Youth 1973 Egalitarianism Coefficients (Standard Errors and Test Statistics in Parentheses) Political Knowledge/Group Identification Model Demographics/Religiosity Model Govt. Provide Jobs .830 (.064) *** .826 (.071) *** Bus to Achieve Integration 1 1 Women Equal .650 (.061) *** .656 (.065) *** Should Govt. Help Minorities .869 (.064) *** .860 (.063) *** N = 1135 * = p < 0.05 ** = p < 0.01 *** = p > 0.001 Source: Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined. f f Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) where η = Egalitarianism y 1 = Govt. Provide Jobs y 2 = Bus to Achieve Integration y 3 = Women Equal y 4 = Should Govt. Help Minorities Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 213 Table B.5. Measurement Model Parameters for Parents’ 1973 Egalitarianism Coefficients (Standard Errors and Test Statistics in Parentheses) Political Knowledge/Group Identification Model Demographics /Religiosity Model Govt. Provide Jobs 1 1 Govt. Help Minorities .876 (.067) *** .919 (.065) *** Women Equal .307 (.060) *** .302 (.066) *** Should Federal Govt. Integrate Schools .540 (.055) *** .508 (.055) *** N = 1135 * = p < 0.05 ** = p < 0.01 *** = p > 0.001 Source: Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined. g g Equations are as follows: y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) where η = Egalitarianism y 1 = Govt. Provide Jobs y 2 = Govt. Help Minorities y 3 = Women Equal y 4 = Should Federal Govt. Integrate Schools Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 214 Table B.6. Measurement Model Parameters for Youth 1973 Political Knowledge Coefficients (Standard Errors and Test Statistics in Parentheses) Length of Senate Term 1 Where is Tito Leader .300 (.029) *** Number of Members on Supreme Court -.021 (.029) Governor of R’s State .060 (.027) * Location of Concentration Camps .050 (.029) Was Roosevelt Dem. Or Rep. .211 (.029) *** Which Party More Conservative .253 (.029) *** N = 1135 * = p < 0.05 ** = p < 0.01 *** = p > 0.001 Source: Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined. h h Equations are as follows y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) y 4 = Φ -1 ( Λ 4 η + ε 4 ) y 5 = Φ -1 ( Λ 5 η + ε 5 ) y 6 = Φ -1 ( Λ 6 η + ε 6 ) y 7 = Φ -1 ( Λ 7 η + ε 7 ) where η = Political Knowledge y 1 = Length of Senate Term y 2 = Where is Tito Leader y 3 = Number of Members on Supreme Court y 4 = Governor of R’s State y 5 = Location of Concentration Camps y 6 = Was Roosevelt Dem. Or Rep. y 7 = Which Party More Conservative Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 215 Table B.7. Measurement Model Parameters for Media Exposure: 1982 Youth Coefficients (Standard Errors and Test Statistics in Parentheses) Watch TV About Politics 1 Listen to Radio About Politics .528 (.040) *** Read Magazines About Politics .680 (.042) *** N = 1135 * = p < 0.05 ** = p < 0.01 *** = p > 0.001 Source: Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined. i Table B.8. Measurement Model Estimates for Religiosity Coefficients (Standard Errors and Test Statistics in Parentheses) Frequency Church Attendance 1 Interpretation of Bible .203 (.030) *** N = 1135 * = p < 0.05 ** = p < 0.01 *** = p > 0.001 Source: Youth-Parent Socialization Panel Study, 1965-1982: Three Waves Combined. j i Equations are as follows y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) y 3 = Φ -1 ( Λ 3 η + ε 3 ) where η = Media Exposure y 1 = Watch TV About Politics y 2 = Listen to Radio About Politics y 3 = Read Magazines About Politics Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. j Equations are as follows y 1 = Φ -1 ( η + ε 1 ) y 2 = Φ -1 ( Λ 2 η + ε 2 ) 216 Table B.9. Egalitarianism Measurement Model Parameters for NES Equal Opportunity Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership- Age Model Group Membership- Cohort Model Group Identification- Age Model Group Identification- Cohort Model INSURE EQUAL OPPORTUNITY 0.224 (0.652) 0.1828 (0.0228)*** 0.24 (0.1218)* 0.1906 (0.0607)*** UNEQUAL CHANCES A PROBLEM 0.3674 (0.0742)** 0.3704 (0.0386)*** 0.3448 (0.3248) 0.3558 (0.131)** UNEQUAL CHANCES OKAY 0.8256 (0.026)*** 0.8444 (0.0182)*** 0.783 (0.1892)*** 0.8182 (0.0716)*** WORRY LESS ABOUT EQUALITY k 1 1 1 1 N = 10693 * = P < 0.05 ** = P < 0.01 *** = P < 0.001 Source: National Election Study Cumulative File, 1948-2000. l Table B.8: Continued where η = Religiosity y 1 = Frequency Church Attendance y 2 = Interpretation of Bible Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. k This is admittedly a general equality indicator rather than a specific indicator for equality of opportunity, but it was necessary to include this variable in the model to identify the measurement model. l Equations are as follows y 1 = Λ 1 η + ε 1 y 2 = Λ 2 η + ε 2 y 3 = Λ 3 η + ε 3 y 4 = η + ε 4 where 217 Table B.10. Measurement Model Parameters for NES Racial Egalitarianism Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership- Age Model Group Membership- Cohort Model Group Identification- Age Model Group Identification- Cohort Model INSURE EQUAL OPPORTUNITY 0.2886 (0.0764)*** 0.268 (0.0295)*** 0.2908 (0.1309)* 0.2824 (0.055)*** PUSH EQUAL RIGHTS TOO FAR 1.2262 (0.0857)*** 1.2418 (0.0411)*** 1.245 (0.0661)*** 1.2286 (0.0281)*** UNEQUAL CHANCES A PROBLEM 0.481 (0.1595)** 0.4748 (0.077)*** 0.4654 (0.2197)* 0.4636 (0.09)*** UNEQUAL CHANCES OKAY 0.8334 (0.075)*** 0.8748 (0.0389)*** 0.7998 (0.1619)*** 0.8366 (0.0742)*** WORRY LESS ABOUT EQUALITY 1 1 1 1 FEWER PROBLEMS IF MORE EQUAL TREATMENT 0.167 (0.0453)*** 0.1682 (0.0247)*** 0.2018 (0.0534)*** 0.2046 (0.0211)*** 7PT AID TO BLACKS SCALE 0.7932 (0.0724)*** 0.78 (0.0344)*** 0.7856 (0.1141)*** 0.7738 (0.0622)*** N = 10693* = P < 0.05 ** = P < 0.01 *** = P < 0.001 Source: National Election Study Cumulative File, 1948-98. m Table B.9: Continued η = Equal Opportunity y 1 = INSURE EQUAL OPPORTUNITY y 2 = UNEQUAL CHANCES A PROBLEM y 3 = UNEQUAL CHANCES OKAY y 4 = WORRY LESS ABOUT EQUALITY Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 218 m Table B.10: Continued Equations are as follows y 1 = Λ 1 η + ε 1 y 2 = Λ 2 η + ε 2 y 3 = Λ 3 η + ε 3 y 4 = Λ 4 η + ε 4 y 5 = η + ε 5 y 6 = Λ 6 η + ε 6 y 7 = Λ 7 η + ε where η = Racial Egalitarianism y 1 = INSURE EQUAL OPPORTUNITY y 2 = PUSH EQUAL RIGHTS TOO FAR y 3 = UNEQUAL CHANCES A PROBLEM y 4 = UNEQUAL CHANCES OKAY y 5 = WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 7 = 7PT AID TO BLACKS SCALE Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 219 Table B.11. Measurement Model Parameters for NES Gender Egalitarianism Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership- Age Model Group Membership- Cohort Model Group Identification- Age Model Group Identification- Cohort Model INSURE EQUAL OPPORTUNITY 0.221 (0.1118)* 0.2068 (0.0425)*** 0.243 (0.1684) 0.2354 (0.0771)* PUSH EQUAL RIGHTS TOO FAR 1.2382 (0.1108)*** 1.2294 (0.0559)*** 1.2324 (0.0893)*** 1.2252 (0.0316)*** UNEQUAL CHANCES A PROBLEM 0.403 (0.1914)* 0.3298 (0.0783)*** 0.3906 (0.251) 0.3838 (0.1004)*** UNEQUALCHANCES OKAY 0.8524 (0.0739)*** 0.891 (0.0479)*** 0.8228 (0.1668)*** 0.8508 (0.0796)*** WORRY LESS ABOUT EQUALITY 1 1 1 1 FEWER PROBLEMS IF MORE EQUAL TREATMENT 0.0532 (0.0822) 0.0564 (0.2562) 0.1026 (0.0654) 0.1134 (0.0318)*** 7PT WOMENS EQUAL ROLE SCALE 0.6768 (0.0643)*** 0.715 (0.0365)*** 0.679 (0.0644)*** 0.7122 (0.0228)*** N = 10693 * = p < 0.05 ** = p < 0.01 *** = p < 0.001 Source: National Election Study Cumulative File, 1948-2000. n n Equations are as follows y 1 = Λ 1 η + ε 1 y 2 = Λ 2 η + ε 2 y 3 = Λ 3 η + ε 3 y 4 = Λ 4 η + ε 4 y 5 = η + ε 5 y 6 = Λ 6 η + ε 6 y 7 = Λ 7 η + ε 7 where η = Gender Egalitarianism y 1 = INSURE EQUAL OPPORTUNITY y 2 = PUSH EQUAL RIGHTS TOO FAR y 3 = UNEQUAL CHANCES A PROBLEM y 4 = UNEQUAL CHANCES OKAY 220 Table B.12. Measurement Model Parameters for NES General Egalitarianism Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership- Age Model Group Membership- Cohort Model Group Identification- Age Model Group Identification- Cohort Model EQUAL OPPORTUNITY 0.2464 (0.0962)** 0.2136 (0.0371)*** 0.2368 (0.1669) 0.2326 (0.11)* PUSH EQUAL RIGHTS TOO FAR 1.266 (0.0823)*** 1.2442 (0.0421)*** 1.2394 (0.0629)* 1.234 (0.0441)*** UNEQUAL CHANCES A PROBLEM 0.4296 (0.196)* 0.409 (0.0905)*** 0.3954 (0.28) 0.3948 (0.1823)* UNEQUAL CHANCES OKAY 0.8396 (0.0994)*** 0.8722 (0.0493)*** 0.8006 (0.2116)*** 0.8336 (0.133)*** WORRY LESS ABOUT EQUALITY 1 1 1 1 FEWER PROBLEMS IF MORE EQUAL TREATMENT 0.1102 (0.0595)* 0.0924 (0.0343)** 0.1144 (0.0431)** 0.1318 (0.0339)*** N = 10693 * = P < 0.05 ** = P < 0.01 *** = P < 0.001 Source: National Election Study Cumulative File, 1948-2000. o Table B.11: Continued y 5 = WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 7 = 7PT WOMENS EQUAL ROLE SCALE Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. o Equations are as follows y 1 = Λ 1 η + ε 1 y 2 = Λ 2 η + ε 2 y 3 = Λ 3 η + ε 3 y 4 = Λ 4 η + ε 4 y 5 = η + ε 5 y 6 = Λ 6 η + ε 6 221 Table B.13. Measurement Model Parameters for NES Economic Egalitarianism Structural Equation Modeling Coefficients (Robust Standard Errors in Parentheses) Group Membership- Age Model Group Membership- Cohort Model Group Identification- Age Model Group Identification- Cohort Model EQUAL OPPORTUNITY 0.3052 (0.0521)*** 0.2742 (0.021)*** 0.2896 (0.0922)*** 0.2842 (0.0458)*** PUSH EQUAL RIGHTS TOO FAR 1.2252 (0.0971)*** 1.2072 (0.0497)*** 1.2184 (0.092)*** 1.2186 (0.0376)*** UNEQUAL CHANCES A PROBLEM 0.4384 (0.1247)*** 0.4152 (0.0558)*** 0.405 (0.1733)** 0.4038 (0.0728)*** UNEQUAL CHANCES OKAY 0.8342 (0.06)*** 0.8736 (0.0348)*** 0.796 (0.1235)*** 0.8276 (0.0584)*** WORRY LESS ABOUT EQUALITY 1 1 1 1 FEWER PROBLEMS IF MORE EQUAL TREATMENT 0.2492 (0.0393)*** 0.2092 (0.0201)*** 0.2374 (0.0506)*** 0.2376 (0.0184)*** GUARANTEED JOBS 0.2338 (0.0643)*** 0.2004 (0.025)*** 0.2388 (0.0865)** 0.2464 (0.0476)*** N = 10693 * = P < 0.05 ** = P < 0.01 Table B.12: Continued where η = Gender Egalitarianism y 1 = INSURE EQUAL OPPORTUNITY y 2 = PUSH EQUAL RIGHTS TOO FAR y 3 = UNEQUAL CHANCES A PROBLEM y 4 = UNEQUAL CHANCES OKAY y 5 = WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms. 222 Table B.13: Continued *** = p < 0.001 Source: National Election Study Cumulative File, 1948-2000. p p Table B.13: Continued Equations are as follows y 1 = Λ 1 η + ε 1 y 2 = Λ 2 η + ε 2 y 3 = Λ 3 η + ε 3 y 4 = Λ 4 η + ε 4 y 5 = η + ε 5 y 6 = Λ 6 η + ε 6 y 7 = Λ 7 η + ε 7 where η = Gender Egalitarianism y 1 = INSURE EQUAL OPPORTUNITY y 2 = PUSH EQUAL RIGHTS TOO FAR y 3 = UNEQUAL CHANCES A PROBLEM y 4 = UNEQUAL CHANCES OKAY y 5 = WORRY LESS ABOUT EQUALITY y 6 = FEWER PROBLEMS IF MORE EQUAL TREATMENT y 7 = GUARANTEED JOBS Φ -1 = a probit link function, which is an inverse cumulative normal probability distribution (an inverse z score) The Λs are coefficients and the εs are error terms.
Abstract (if available)
Abstract
This dissertation investigates the mechanisms involved in the formation, maintenance and change in the level of egalitarianism in the United States public during the second half of the 20th Century. This is an important topic in the study of American politics because egalitarianism is a core value and core values structure the policy attitudes of a public that, according to research, for the most part, does not follow politics closely and is often ambivalent about the political questions of the day. The two theories tested are the elite leadership theory and group identification and interest theory. Elite leadership theory predicts that on types of egalitarianism about which elites agree such as racial and gender egalitarianism and equal opportunity, exposure to elite discourse as measured by education, media exposure and political knowledge should predict these types of egalitarianism and group identity and interest theory explains the level of egalitarianism on types of egalitarianism about which elites disagree, such as equality in general and economic egalitarianism.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Political incorporation and transnationalism: a study of South Asian immigrants in the United States
PDF
Health care for all? Anti-Latino and anti-immigrant attitudes, health care policy, and the Latino community
PDF
The role of skin tone on candidate evaluation: an analysis of racial socialization among Latino immigrants
PDF
Death and politics in the United States Courts of Appeals
PDF
Private school choice at the state level: diversity and success
PDF
Pathways of local economic development: tales of cities in the United States and South Korea
PDF
Adversaries and statecraft: explaining U.S. foreign policy toward rogue states
PDF
The political psychology of scarcity in American society: a study of culture, discourse, and public opinion
PDF
Enterprising citizenship: Mexican immigrant empowerment and public-private partnerships in Santa Ana, California
PDF
Beyond nationalism: a history of leisure discourse in and between the United States and Japan, 1910-1940
PDF
Assessing the implementation of the RECLAIM cap-and-trade market for pollution: measurement issues in counterfactuals, goal attainment, and command-and-control alternatives
PDF
The changing geographies of concentrated poverty and concentrated affluence in the United States, 1990-2000
PDF
The Catholic church in Latin America: an evaluation of the institutional and political impacts of progressive church reforms
PDF
A comparative study of food safety-related public relations practices in China and the United States
PDF
The resiliency of Arab authoritarianism and the Arab-Israeli conflict: the United States' role in the cases of Egypt and Jordan
PDF
Organized intraethnic conflict and cooperation in divided stateless nations: the competitive dynamics of rival ethnonationalist factions
PDF
Learning from our global competitors: a comparative analysis of science, technology, engineering and mathematics (STEM) education pipelines in the United States, Mainland China and Taiwan
PDF
Understanding the role of population and place in the dynamics of seasonal influenza outbreaks in the United States
PDF
Toward counteralgorithms: the contestation of interpretability in machine learning
Asset Metadata
Creator
Whitten, Jeffrey H.
(author)
Core Title
Egalitarianism begins at home and in cohorts: egalitarianism in United States public opinion, 1952-2000
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Political Science
Publication Date
08/08/2007
Defense Date
04/23/2007
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
equality,OAI-PMH Harvest,United States public opinion,Values
Place Name
USA
(countries)
Language
English
Advisor
Crigler, Ann (
committee chair
), [illegible] (
committee member
), Wong, Janelle S. (
committee member
)
Creator Email
jwhitten@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m779
Unique identifier
UC1491516
Identifier
etd-Whitten-20070808 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-541550 (legacy record id),usctheses-m779 (legacy record id)
Legacy Identifier
etd-Whitten-20070808.pdf
Dmrecord
541550
Document Type
Dissertation
Rights
Whitten, Jeffrey H.
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
United States public opinion