 

.1 l: ,. «#1.!
. .
3.3.519. ‘
. .
:iﬁi... ‘

“4:

7‘20

Jr

r

i335

. .io' .

x. l 5:... x

2 .. If...
i

 

 

 

«9007

 

 

“LIBRARY
Michigan State
University

 

 

 

This is to certify that the
thesis entitled

SEPARATISM IN DECLINE:
HOW POLITICAL INVOLVEMENT HAS MODERATED THE
EVANGELICAL RIGHT

presented by

Robert G. Moore

has been accepted towards fulﬁllment
of the requirements for the

PhD. degree in Political Science

 

 

\JQSI m4 M\04$4l

IVlajo‘Pro‘fe s or’sSi na re

rim/0 7

Date

MSU is an aﬂVnnative-action, equal-opportunity employer

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DAIEDUE

DAIEDUE

DAIEDUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:lCIRC/DaleDuetindd-p.1

 

SEPARATISM IN DECLINE:
HOW POLITICAL INVOLVEMENT HAS MODERATED THE EVANGELICAL
RIGHT
By

Robert G. Moore

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Political Science

2007

ABSTRACT

SEPARATISM IN DECLINE: HOW POLITICAL INVOLVEMENT HAS
MODERATED THE EVANGELICAL RIGHT

By
Robert G. Moore

Despite the numerous studies exploring the effects of evangelical political
participation on American politics, few examine the reciprocal relationship of
participation’s effects on evangelicals. This study seeks to remedy this inattentiveness by
exploring the inﬂuence of increased political activism on American Evangelicals. It
measures changes for Evangelicals in both voting and campaign participation from 1976-
2000, and explores changes in issue accessibility for Evangelicals as well as Non-
Evangelicals. It delves into longitudinal changes in political, moral and tolerance
attitudes for Evangelicals, and evaluates the influence of participation on those attitudes.
Finally, an internal movement, new paradigm churches, is evaluated for its contribution
to political participation and attitude shifts.

A comparison of trends between 1976 and 2000 reveals evangelical and secular
social issue access is increasing at a rate signiﬁcantly above the remaining general
population. An analysis of social issue attitudes discloses that Evangelicals and Seculars
who more readily access social issues have more extreme views on these issues than
other Evangelicals and Seculars. Research also shows a reciprocal relationship between
the access of social and economic issues, and political participation for Evangelicals.

This study conﬁrms evidence of generational and period effects that suggest
permanent increases in evangelical political participation and a moderating of evangelical

attitudes in a liberal direction. Lastly, this dissertation demonstrates that new paradigm

church inﬂuences, as an internal movement, are found to increase political participation

and tolerance

For Marcia, Stephen, Ethan Luke and Drew

iv

ACKNOWLEDGEMENTS

Many people have contributed to making this dissertation possible. First I would
like to my high school English teacher, Dan Klink for providing the grammatical
background that made this writing possible.

I would also like to thank the members of my dissertation committee, Paul
Abramson, Ada Finiﬁer and Brian Silver for their service, support and suggestions. I
would especially like to acknowledge my chair, Bill Jacoby for his unfailing support and
constant encouragement through this process.

I would also like to thank my wife, Marcia, for her patience and love which made
some of the long, dark times less gloomy. Also for putting up with Dad’s long working
weekends and evenings, Stephen, Ethan, Luke and Drew -I love you guys. Finally I wish
to acknowledge my Lord who provides the inspiration for my life and the strength to

accomplish it.

Table of Contents

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

List of Tables . viii
List of Figures x
I. INTRODUCTION 1
II. EVANGELICALS AND SEPARATISM IN DECLINE 6
2.] Deﬁning Evangelicals ................................................................. 7
2.2 Measuring Evangelicals ............................................ 10
2.3 Evangelical Separatism and Political Participation ................................................. 13
2.4 The Socioeconomic Model .......................................... 16
2.5 Civic Volunteer Model ............................................................................................ 17
2.6 Mobilization Model ................................................................................................. 18
2.7 Participation and Religion ....................................................................................... 20
2.8 Evangelicals and Participation ................................................................................. 22
III. WHY CHANGING PARTICIPATION MODIFIES EVANGELICALS ............ 26
3.1 Secularization - ................................. 76
3.2 What are the Political Effects of Secularization? .................................................... 29
3.2.1 Eroding Religious Habits and Increasing Numbers of Seculars ....................... 30
3.2.2 Political and Moral Consequences. ....... - - ............. 31
3.2.3 Increasing Tolerance - ....................................... -34
3.3 Fighting the Culture War ......................................................................................... 37
3.4 New Paradigm Churches ......................................................................................... 39
3.5. Which Inﬂuences? ....................................... 44
IV. THE ORGANIZATION OF POLITICAL THINKING 46
4.1 Levels of Conceptualization .................................................................................... 47
4.2 Schema and Sophistication ...................................................................................... 53
4.3 Criticisms of Schema and the Resulting Chronicity Measurement ......................... 55
4.4 Participation, Sophistication, Information and Reciprocal Relationships ............... 56

V. POLITICAL PARTICIPATION AND ISSUE ACCESSIBILITY AMONG
EVANGELICALS 58
5.1 Theoretical Background ........................................................................................... 58
5.2 Hypotheses ............................................... 62
5.3 Data ..................................................................... 65
5.4 Evangelical Political Participation ........................................................................... 67
5.5 Chronicity . ..................................................... 71
5.6 Measuring the Chronicity- .............................................................. 74
5.7 Evidence of a (limited) Culture War ........................................................................ 76
5.8 Extremes in Attitudes ............................................................. 78
5.9 Methodology: ZSCML and ZSPLS ........................ 89
5.10 Results ............................... --_93

vi

 

5.11 Conclusions .......................... 106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

VI. PARTICIPATION AND EVANGELICAL MODERATION 107
6.1 Deﬁning Moderation .108
6.2 Hypotheses ..................................................................... 110
6.3 Age-Period-Cohort models- 114
6. 4 Data and Coding .......... 117
6. 5 Methodology Periods, Cohorts and Changes In NES Participation ...................... 119
6. 6 NES Participation Results ................................ 120
6. 7 NES Political Attitude Results .................................................... 127
6. 9 G88 Moral, Political and Tolerance Attitudes ...................................................... 137
6. 9 Conclusions ................... 156
VIII. PARTICIPATION AND NEW PARADIGM CHURCHES 158
7.1 Overview and Hypothesis 158
7. 2 Data - ...................................... 160
7.3 Conﬂicting Religious Traditions and Congregations ............................................ 161
7.4 The New Paradigm Scale ............................... 166
7.5 Methodology: Regression and Logistic Regression . .- . -- -. 169 _;
7.6 Results ...... .. .............................. 170 '1;
7.7 Conclusions ............................................................................... 185 I},
VIII. CONCLUSIONS 187 {ii
8.1 Future Directions ................................................................................................... 191 E
APPENDICES 193
REFERENCES 271

 

.‘un r... {3, :;‘_;;_';-.j-y.,s'_~' 'rmﬁihﬂﬂ—Hﬁ‘d’d’w

vii

 

List of Tables

Table 5.]. Comparing Evangelical Participation with Seculars and General Population
Political Participation: Difference of Means. - 69

 

Table 5.2. Trend Analysis: Evangelicals, Seculars and Non-Evangelicals Use of Social
Issue Chronicity, 1976-2000 77

 

Table 5.3. Social Issue Chronicity, Participation and Information : ZSCML Results ....... 92
Table 5.4. Economy Chronicity, Participation and Information : ZSCML Results ........... 99
Table 5.5. No Content Chronicity, Participation and Information : ZSCML Results ..... 103

Table. 6.1. NES Period and Ten Year Birth Cohorts (Number of Evangelicals 1n
parenthesis) ...................... - - - 118

 

Table 6. 2. General Social Survey Birth Cohorts (N umber of Evangelicals 1n Cohort 1n
Parenthesis) ...................................................................... 1 19

 

Table 6.3. Regression of Periods and Cohorts on Campaign Participation: NES 1972-
2000. 121

 

Table 6.4. Logistic Regression of Periods and Cohorts on Voter Participation: NES 1972-
2000. 125

 

Table 6.5. Regression of Periods, Cohorts and Campaign Participation on Environmental
Issues: NES 1980-2000. - -- .................................. 128

 

Table 6.6. Regression of Periods, Cohorts and Campaign Participation on Women’s
Issues: NES 1972-2000. ................................................................................ 132

Table 6. 7. Regression of Periods, Cohorts and Campaign Participation on Abortion
Support: NES 1972-2000.- 135

 

Table 6.8. Regression of Periods, and Cohorts on Sexual Attitudes: GSS 1974-2002. ..138

Table 6. 9. Regression of Periods, and Cohorts on Government Spending: GSS 1974-
2002.- .143

 

Table 6.10. Logistic Regression of Periods, and Cohorts on Tolerance of Atheists: GSS
1974-2002.- - - ................................. - - - - 147

 

 

Table 6.11. Logistic Regression of Periods, and Cohorts on Tolerance of Homosexuals:
GSS 1974-2002. - 150

 

viii

 

 

 

 

 

Table 6.12. Logistic Regression of Periods, and Cohorts on Tolerance of Communists:

 

GSS 1974-2002. ............................................................................................ 153
Table 7.1 Religious Identity Correlations for GSS Age Sixteen Preferences, Current GSS
Preferences and Current National Congregation Study Preferences. ........... 164
Table 7.2. Percentage of New Paradigm Church Attributes in Each Christian Tradition.
- - . . ........................................................................................ 166
Table 7.3. Factor Analysis of New Paradigm Church Attributes. .................................. 168

Table 7.4. Anova Test of Means: New Paradigm Scale Mean by Religious Tradition...169
Table 7.5. Logistic Regression: Paradigm Scales and Tolerance of Atheists ................. 171

Table 7.6. Logistic Regression: Paradigm Scales and Tolerance of Homosexuals. ........ 176

Table 7.7. Logistic Regression: Paradigm Scales and Tolerance of Communists. ......... 178
Table 7.8. Regression: Paradigm Scales and Abortion Support ...................................... 180
Table 7.9. Logistic Regression: Paradigm Scales and Probability of Voting .................. 183

ix

List of Figures

Figure 5.1. American Religious Traditions over Time. ...................................................... 66
Figure 5.2. Evangelicals and Rates of Political Participation, 1976-2000 .......................... 68
Figure 5.3. Religious Traditions and Vote Participation, 1976-2000. ................................ 71
Figure 5.4. Combined Tables of Evangelical, Secular and Non-Evangelical Chronicity
over Time: 1976-2000. ................................................................................... 75
Figure 5.5. Religious Groups, Social Issue Religious Groups and Aid to Blacks. ............. 79

Figure 5.6. Religious Groups, Social Issue Religious Groups and Assistance to Blacks. ..80
Figure 5.7. Religious Groups, Social Issue Religious Groups and Black Thermometer. ...81
Figure 5.8. Religious Groups, Social Issue Religious Groups and Gays in the Military....82
Figure 5.9. Religious Groups, Social Issue Religious Groups and Gay Adoption. ............ 83
Figure 5.10. Religious Groups, Social Issue Religious Groups and Women’s Roles. ....... 84

Figure 5.11. Religious Groups, Social Issue Religious Groups and Women’s Libber’s
Thermometer. ................................................................................................. 84

Figure 5.12. Religious Groups, Social Issue Religious Groups and Abortionist

 

Thermometer. ................................................................................................. 85
Figure 5.13. Religious Groups, Social Issue Religious Groups and Abortion Rights. ....... 86
Figure 5:14. Religious Groups and Social Issue Religious Groups, and Tolerance of
Differing Moral Standards. ............... - ........................................ 87
Figure 5:15 Religious Groups and Social Issue Religious Groups, and Adjusting Views of
New Moral Behavior ...................................................................................... 87
Figure 5.16. Religious Groups, Social Issue Religious Groups and the Environmental
Spending. ........................................................................................................ 88
Figure 5.17. Religious Groups, Social Issue Religious Groups and Environmentalist
Thermometer. ................................................................................................. 88
Figure 5.18. Representation of Simpliﬁed Chronicity-Participation Model ....................... 91

Figure 5.19. Increasing Political Participation and the Probability of Accessing the Social
Issue Chronicity by Religious Tradition ......................................................... 97

Figure 5.20. Increasing Religious Attendance and the Probability of Accessing the Social
Issue Chronicity by Religious Tradition... ..................................... 98

 

Figure 5.21. Increasing Political Information and the Probability of Accessing the Social
Issue Chronicity by Religious Tradition ......................................................... 99

Figure 5.22. Increasing Political Participation and the Probability of Accessing the
Economic Chronicity by Religious Tradition ............................................... 102

Figure 5.23. Increasing Political Information and the Probability of Accessing the
Economic Chronicity by Religious Tradition ............................................... 102

Figure 6.]. Campaign Participation by Period, Cohort and Evangelical Tradition ......... 121

Figure 6.2. Probability of Voting by Period, Cohort and Evangelical Tradition ............. 125
Figure 6.3. Environmental Spending by Period, Cohort, Campaign Participation and
Evangelical Tradition .................................................................................. 130
Figure 6.4. Environmentalist Thermometer by Period, Cohort, Campaign Participation
and Evangelical Tradition ........................................................................... 131
Figure 6.5. Feminist Thermometer by Period, Cohort, Campaign Participation and
Evangelical Tradition: NES 1972-2000 ...................................................... 133
Figure 6.6. Women’s Roles by Period, Cohort, Campaign Participation and Evangelical
Tradition: NES 1972-2000 .......................................................................... 134
Figure 6.7. Abortion Support by Period, Cohort, Campaign Participation and Evangelical
Tradition: NES 1980-2000 .......................................................................... 137
Figure 6.8. Sexual Morals by Period, and Evangelical Tradition: GSS 1974-2002 ........ 141
Figure 6.9. Sexual Morals by Cohort, and Evangelical Tradition: GSS 1974-2002 ....... 142

Figure 6.10. Government Spending by Period, and Evangelical Tradition: GSS 1974-
2002 ............................................................................................................. 145

Figure 6.11. Government Spending by Cohort, and Evangelical Tradition: GSS 1974-
2002 ............................................................................................................. 146

Figure 6.12. Tolerance of Atheists by Period, and Evangelical Tradition: GSS 1974-2002
............. - - - - - - - - - -..l49

 

xi

 

Figure 6.13. Tolerance of Atheists by Cohort and Evangelical Tradition: GSS 1974-2002
..................................................................................................................... 150

Figure 6.14. Tolerance of Homosexuals by Period and Evangelical Tradition: GSS 1974-
2002 ............................................................................................................. 152

Figure 6.15. Tolerance of Homosexuals by Cohort and Evangelical Tradition: GSS 1974-
2002 ............................................................................................................. 153

Figure 6.16. Tolerance of Communists by Period and Evangelical Tradition: GSS 1974-
2002 ............................................................................................................. 155

Figure 6.17. Tolerance of Communists by Cohort and Evangelical Tradition: GSS 1974-
2002 ............................................................................................................. 156

Figure 7.1. Probability of Tolerance of an Atheist Speaker by Religious Tradition New
Paradigm Attributes, and Attendance .......................................................... 173

Figure 7 .2. Probability of Tolerance of an Atheist Book by Religious Tradition New
Paradigm Attributes, and Attendance .......................................................... 174

Figure 7.3. Probability of Tolerance of an Atheist Teacher by Religious Tradition New
Paradigm Attributes, and Attendance .......................................................... 173

Figure 7.4. Probability of Tolerance of a Homosexual Teacher by Religious Tradition
New Paradigm Attributes, and Attendance. ................................................ 177

Figure 7.5. Probability of Tolerance of a Communist Speaker by Religious Tradition,
New Paradigm Attributes, and Attendance. ................................................ 179

Figure 7.6. Prochoice Support by Religious Tradition New Paradigrn Attributes, and
Attendance. .................................................................................................. 181

Figure 7.7. Probability of Voting by Religious Tradition New Paradigm Attributes, and
Attendance. .................................................................................................. 184

xii

CHAPTER I

INTRODUCTION

This dissertation is concerned broadly with how American Evangelical political
participation is changing over time and how, as a consequence, participation is changing
Evangelicals’ political thought. Changes in participation are theorized to affect how
individuals organize their political sophistication/political thinking. If Evangelical
political participation rose over the past two decades, as many sources assume, it
becomes a longitudinal laboratory in which theories about participation and political
sophistication are tested. Political participation, political knowledge and the structure of
political thought all inﬂuence one another. If one changes, repercussions should involve
the other two. A change increases the possibility of Evangelicals shifting in their
placement across the spectrum of all three of these variables. While perhaps not ﬁtting
the traditional deﬁnition of becoming more “politically sophisticated”, it is none the less
just as interesting if they are shifting while Non-Evangelicals are not.1

On the other hand, change arising from participation is not just limited to political
thought. Changing political participation requires new cues to motivate change and to
offset previous cues for avoiding participation. The receptiveness to cues may not be
uniform across the Evangelical tradition, and the effects may be more profound for some
segment of Evangelicals than for others. Furthermore, just as religious traditions are not
static institutions immune from outside inﬂuences; they are also not immune to internal
change. Understanding Evangelical political participation requires investigating internal

changes that Evangelicals have undergone over the past two decades. While the

 

' The best known hierarchy of political sophistication is the levels of conceptualization (Campbell et al.
1960). A thorough review of political sophistication is included in Chapter 4.

Evangelical tradition is not sufﬁciently monolithic to undergo the type of change Vatican
11 produced in the Roman Catholic Church, internal movements focusing on church
growth (particularly because of the entrepreneurial style of many Evangelical churches)
are widely disseminated within the tradition. One church growth movement in particular,
new paradigm churches, is evaluated for its effect on participation.

In 1979, Rev. Jerry Falwell founded the Moral Majority (the best known of the
many Evangelical interest and social movement groups that arose in this decade) in an
effort to bring about increased political activism among Evangelical Christians. Falwell,
at the founding of Moral Majority, addressed three areas of work for the new
organization: registration, information and mobilization of evangelicalsz. The subﬁeld of
religion and politics was rejuvenated by this latest turn in Evangelical political
involvement. Much effort in political science has gone into measuring differences
between Evangelical and Non-Evangelical attitudes, evaluating Evangelical effects on the
Republican Party coalition, examining churches as political communities, theorizing
about an ongoing or impending culture war and assessing the effects of religious
attendance and religious beliefs on political behavior and political attitudes.3

Previous research on Evangelical inﬂuence points to the electoral process and
resulting changes in evangelical voting patterns and partisanship. Notions of Evangelical
change have been limited to vague unmeasured assumptions about increases in
sophistication or shifts in political thought among evangelicals. These assumptions are
based on anecdotal evidence and the increase in the sophistication of interest groups such

as the Christian Coalition in the early 19905. Despite assumptions of increased

 

2 Thomas and Dobson, 1999
3 See Leege 2003 for an overview of the subﬁeld.

participation, there is no study that measures aggregate shifts in Evangelical political
issue accessibility or their political thought.

In the following chapters I attempt to explore this knowledge gap by measuring
how Evangelicals are changing in accessibility and thought. I measure participation
across religious traditions to see if Evangelicals have increased their participation. I
consider if and how Evangelical issue accessibility/political sophistication have changed
including looking at the trends in Evangelical political issue access. I assess if there is a
unique relationship between political participation, political information and political
issue access for Evangelicals. I also examine period and cohort effects to determine who
is being affected by changing participation within the Evangelical tradition. Finally, I see
if internal Evangelical alterations have contributed to the shifts in participation.

The following three chapters review the current literature and lay out the
theoretical basis for my research. Chapter 2 reviews measurement schemes for deﬁning
Evangelicals, and then, historically traces their attitudes toward political participation. It
also focuses on models of political participation and their inclusion or failure to include
religion. Chapter 3 highlights the internal consequences of increased participation apart
from political sophistication. It looks at the potential inﬂuence of secularization, the
culture war and new paradigm churches. Chapter 4 traces the development of various
measurements of political sophistication and deﬁnes the rationale for the measurement
used in this dissertation. It also summarizes the theories surrounding a reciprocal
relationship between sophistication and participation.

Chapter 5 investigates the historical trends of voting and campaign participation

for Evangelicals and their relationship to sophistication. Using national election studies

ﬁom 1976-2000, I focus on how changes in political sophistication and issue
accessibility, affect political participation. In addition, Chapter 5 demonstrates how
political participation is interacting with changes in accessibility. The chapter examines
the reciprocal relationship between various forms of participation and issue accessibility
as well as their relationships to political knowledge. By examining differing forms of
participation, I determine whether or not reciprocity is limited by the intensity of
participation.

Chapter 6 considers how increasing participation affects younger cohorts of
Evangelicals. Building on Zaller’s (1992) conclusion that older respondents will exhibit
greater partisan resistance to new patterns of participation, I evaluate voting and
campaign participation by period effects and birth cohorts to see if younger cohorts
present a unique response. Then I evaluate how this new participation may be shaping the
political and moral attitudes of these younger cohorts.

Chapter 7 steps away from the Evangelical versus Non-Evangelical comparisons
and focuses solely on Christian traditions (Evangelical, Mainline, Black and Catholic).
New paradigm church inﬂuences in these traditions are measured using the combined
1998 General Social Survey and National Congregation Study. I evaluate whether new
paradigm church inﬂuences explain participation and tolerance changes in these
traditions. The ﬁnal chapter outlines my conclusions and directions for future research.

By focusing on Evangelicals over time, I propose to conﬁrm supposed increases
in Evangelical participation and consider the prospects for future participation. In
reaffirming the reciprocal relationship between participation and political thought, I show

how both the scope and limitations of these reciprocal relationships are based on the type

of participation. I add to the understanding of political sophistication by showing that
religious identities have distinct effects on the reciprocal process of participation and
sophistication.

I also test the necessary inclusion of political information in Neuman’s (1986)
spiral of increasing sophistication. Neuman theorizes that political interest from one’s
environment increases participation, which then changes or raises the level of
sophistication. I provide a test of the effect of political information on changes in
participation and sophistication/issue accessibility.

Finally, I consider whether internal movements in religious traditions are of
sufﬁcient interest in their effect on political attitudes. When considering religious
traditions as political effecters and effectees over several decades, their internal
inalterability cannot be assumed. There may be a place in political science such as there
is in the sociology of religion for observing religious change in an effort to better

understand its effect on politics.

 

CHAPTER II

EVANGELICALS AND SEPARATISM IN DECLINE

The involvement of the Evangelical tradition in the American political scene has
received profuse attention since the rise of the Moral Majority in the late 19703 and the
establishment of numerous successor groups to the now defunct Jerry Falwell creation.
After twenty years of involvement, it is now not only possible to evaluate the effect of
Evangelicals on parties, policy and politicians, but it is also possible to consider the
changes within Evangelicals as a result of their political involvement. The effect of their
own participation on Evangelicals is potentially more profound than their participation’s
inﬂuence on current politics. While Evangelicals new found political participation has
been highly visible, these shifts in Evangelical political behavior may have weakened
some of the basic beliefs and practices once unique to this group. The move toward
becoming politically active could ultimately lead to the assimilation of Evangelicals into
the mainstream political culture.

As this movement takes on new roles of citizenship and political participation, it
experiences a change in the cues given by Evangelical elites. The message shifts from
one of separatism to one of involvement. This shift to activism was initially ignited by the
candidacy of Jimmy Carter, the ﬁrst self—avowed Evangelical to run for the presidency
since William Jennings Bryan. The subsequent mutual embrace of Evangelicals and the
Republican Party reinforces the shift. At the same time, conservative denominations are
being shaped by new internal religious trends. These trends seek to reduce sectarianism

and change the stance of these historically closed groups to one of increased openness to

 

the culture. In this chapter, I deﬁne how to measure Evangelicals, examine their tradition
of separatism, and survey the limited research focusing on changing rates of Evangelical

political participation.

2.1 Defining Evangelicals

American Protestantism holds within itself three major religious traditions,
Mainline Protestants, Black Protestants and Evangelical Protestants. Combined, they
represent over half of the American population with Evangelicals representing the largest
subgroup. Evangelical identiﬁers constitute approximately one quarter of all Americans.

Previous studies of American Evangelicals break down into two groups: those
focused on the Evangelical tradition and those examining Evangelical social movements.
Studies focusing on social movements identify their target as the Christian Right (also
referred to as the New Christian Right or the Religious Right). This research attempts to
measure political mobilization of Evangelicals as well as other conservative Christians.
These inquiries focus on organizations (the Christian Coalition, Religious Roundtable
and Family Research Council); leaders (Gary Bauer, Ralph Reed, Pat Robertson and
James Dobson) and their efforts to recruit and mobilize activists for a conservative
political agenda.4 These organizations tend to attract activists from several religious
traditions including Evangelicals, Mainline Protestants, Roman Catholics, and to some
extent, Black Protestants. However, they fail to encompass any tradition in its entirety.
Hood and Smith (2002) show the Religious Right as a religiopolitical movement

crosscutting major American religious traditions but not encompassing any one tradition.

 

4 For an overview of these organizations, see Guth, Green, Kellstedt and Smidt I996; Moen 1992 and
Wilcox 2000. For more detailed studies see Conover 1983; Green, Guth and Hill 1993; Guth 1996; and
Guth and Green 1990; Shupe and Stacey 1983; Wilcox 1986, 1988.

 

 

More importantly, for this study. they reveal that Religious Right identiﬁcation fails to
signiﬁcantly explain political behavior. They go so far as to suggest that the various
measurements of the Religious or Christian Right are so divergent and elusive that they
resemble a “snipe hunt” or a search for a mythical creature that no one ever bags. In part,
the Religious Right is difficult to identify because Evangelicals possess internal divisions
and mistrust that keep many of these organizations from becoming synonymous with the
entirety of the Evangelical tradition; although, they remain inﬂuential in the tradition.
They function as catalysts helping to explain shifts in Evangelical participation,s as a
segment of the tradition, rather than the whole. Measurements and analysis of these social
movement groups reveal that they do not represent the whole of Evangelical
mobilization, participation or attitude measurements.

This study focuses on Evangelicals as a religious tradition rather than a single
segment of a social movement. Penning and Smidt deﬁne a religious tradition as “a
characteristic way of interpreting and responding to the world”, (2003, p. 13) while
Kellstedt et al. deﬁnes it as “a group of religious communities that share a set of beliefs
that generates a distinct worldview” (1996, 176). This is not to assume that all
Evangelical identiﬁers are identical in their views, but rather, in the aggregate,
Evangelicals will respond and act in a similar fashion.

Deﬁning Evangelicals is simpliﬁed by focusing on the religious tradition rather
than their social movements. While they overlap, and in some isolated cases the social

movements may be encapsulated as segments of the traditioné, the movements represent

 

5 For example, the Christian Coalition claimed to have distributed 40 million voter guides in Evangelical
churches in 1998 election cycle (Wilcox 2000).
6 For example, the Inﬂuence of Concerned Women for America appears to be limited to Fundamentalists
within the Evangelical tradition (Jorstad 1993).

 

monolithic crusades seeking to inﬂuence a more diverse Evangelical tradition (Penning
and Smidt 2003). The Evangelical tradition is much broader than any single social
movement. The Evangelical tradition embraces a wide diversity of denominations and
theology representing charismatics and fundamentalists: Calvinists and Arminians, and
Baptists and Pentecostals, to name a few. Within the Evangelical tradition, these
differences appear as yawing chasms of demarcation; while to the outside observer, the
differences blur into a common heritage and a common identity. A survey of deﬁnitions

for Evangelicals demonstrates the common elements of this tradition.

“people professing conﬁdence in the Bible and preoccupied with the
message of God’s salvation of sinners through the death of Jesus Christ”
(Marsden 1980, 3)

“evangelicalism represents the school of Protestant Christianity that
emphasizes personal salvation through Jesus Christ and regards the Bible
as the ﬁnal authority concerning all matters of faith and practice.”
(Penning and Smidt 2002, 13)

“a desire to communicate the biblical message of salvation and to convert
others to Christianity, a strong belief in the Bible as a unique touchstone
for determining matters of truth and practice, an active commitment to
some branch of the institutional church, and a desire to integrate all of
one’s activities and thoughts as consistently as possible around one’s
religious convictions.” (Wuthnow 1989, 166)

“(1) the belief that the Bible is the inerrant Word of God, (2) the belief in
the divinity of Christ and (3) the belief in the efficacy of Christ’s life,
death and physical resurrection for the salvation of the human soul”
(Hunter, 1983 7).

Personal salvation, a literal view of the Bible and an emphasis on Evangelism combine to

form the common elements that unite these internal divisions into a single tradition.

   

2.2 Measuring Evangelicals

Over the past two decades segments within political science and sociology have
debated how best to measure Evangelicals. Recently, the subﬁeld of religion and politics
ﬁnally appears to have settled on a common measurement. The initial debate within the
religion and politics ﬁeld focused on whether beliefs, belonging or behaving served best
as the primary measurements for religious studies.7

The shortcomings of each of these measurements are exacerbated by the use of
survey data. Most surveys exclude questions of religious behavior. As a result, using
survey data required choosing between group deﬁnitions drawn from theological
distinctions (Smidt, 1988), or using a denominational afﬁliation, or some form of self-
identiﬁcation (Jelen, 1989a; Wilcox, 1987). However, using theological distinctions for
deﬁning Evangelicals is elusive because of the lack of diversity in survey questions. For
example, questions from the American National Election Study concerning the bible and
church attendance are not sufficient to distinguish between Mainline Protestants and
Evangelicals (Jelen, Wilcox and Smidt, 1990; Wilcox, 1986). Many Mainline Protestants
will have high attendance and traditional views of the Bible without embracing the need
to be born again to a personal relationship with Christ. In addition, despite
acknowledging orthodox views of scripture, their faith and practice will also be deﬁned
by greater reliance on reason, experience and contemporary thought than that of

Evangelicals (Campbell 2003; Kellstedt and Green 1993).

 

4 Beliefs work best for those faiths that deﬁne themselves through the use of doctrinal beliefs as the
bormdary for inclusion (such as pietistic groups). Belonging measurements focus on attendance,
participation in ritual, and daily exercises of personal devotions. However, most surveys do not provide
sufﬁcient information for this type of measurement. The defaeto standard has become belonging as
reported by the respondent’s self identiﬁcation. See Wald et al. 1988 for a more in depth discussion of
belief, behavior and belonging.

10

 

Self-identiﬁcation by survey respondents appears to solve the problem of using
theological deﬁnitions or denominational afﬁliation. However, Jelen reports that when
given the opportunity to choose between the labels of Fundamentalist, Evangelical,
Charismatic and Pentecostal, there are large percentages of overlap between the
categories as respondents choose more than one option. He ﬁnds that 71.2% of all
persons who consider themselves Fundamentalists, also consider themselves
Evangelicals. Similarly, 49.2% of the Evangelicals designate themselves as
Fundamentalists. In addition, while most Fundamentalists would consider Charismatics
to be an anathema, 62.9% of Charismatics labeled themselves as Fundamentalists and
69.3% said they were Evangelical (1993, 183-185). Jelen also notes that signiﬁcant
numbers of respondents from Mainline and Catholic denominations in the same way
report themselves as also being Evangelical or Fundamentalist.

Using denominational affiliations could potentially result in a sound measurement
of Evangelicals. Denominational afﬁliation has been faulted, though, for assuming a
uniformity of belief and practice within the churches (Wilcox, 1986). Kellstedt and Green
(1993) and Kellstedt et al. (1996) address this problem and have found solid evidence
that denominational distinctions remain strong even when controlling for demographics.
They conclude that careful examination of denominational beliefs and ethos allows for
the combination of similar denominations into uniﬁed religious traditions. Based on this
research, the most commonly accepted strategy for measuring religious affiliation is to
reduce the wide variety of denominations into a smaller number of religious traditions
(Campbell 2003; Kellstedt 1989; Kellstedt et al. 1996; Wald and Smidt 1993). Steensland

et al.’s (2000) classiﬁcation system for religious traditions has rapidly become the

11

defacto standard within the subﬁeld. Steensland et al reduces denominations into
traditions of Evangelical Protestant, Mainline Protestant, Black Protestants, Roman
Catholic, Jewish, Secular (those with no religious tradition, or an ill-deﬁned, [c.g.
“protestant”] tradition without a minimal religious commitment), Nontraditional Liberal,
Non-Traditional Conservative, and Other (a catchall for religious groups such as Greek
Orthodox). Because of the small percentage of respondents falling into the last three
categories (3-4% combined) they are often reduced to a single category of “Other”.

Some have tried to ﬁnd variance within denominational afﬁliation by using a
measurement of religious commitment. Kohut, Green, Keeter and Toth (2000) grouped
respondents with low levels of religious commitment (worship attendance, private prayer
and religious salience) with respondents with no religious afﬁliation. They found that
those with low religious commitment, regardless of denominational affiliation, held
similar political attitudes, political afﬁliations and political activities to those that they
deﬁned as “secularists”. This matches Lenski’s prior conclusion that “those who are more
active in a group usually conform to the norms of the group more faithfully than do
marginal and peripheral members” (1963, 23). Guth, Kellstedt, Smidt and Green (2003),
using proprietary data, also sought to create a tri-fold measurement (traditionalist, centrist
and modernist) within each major religious tradition through the use of a variety of
theological questions and numerous measurements of religious practices. These attempts

at greater speciﬁcity within religious traditions have not survived efforts to export them

 

8 Placement of Aﬁ‘ican Americans in this classiﬁcation is divided between those who put them in any
tradition based on reported denominational affiliation and those who relegate them to the Black Protestant
category regardless of denomination. Campbell points out that this is a minor difference as “almost all
blacks who affiliate with a church belong to one of seven predominately black denominations” (6). Many
studies exclude blacks from the Evangelical traditions as black Evangelicals, while theologically similar to
Evangelicals, tended to be more liberal on social domains such as law and order and national defense, and
have held distinct positions on party id and voting patterns (Kellstedt, I989; Cook and Wilcox, 1990).

12

I
ll

 

to time series datasets such as the General Social Survey and the National Election
Studies. Most national surveys lack sufﬁcient depth of theological questions and religious
practices to create such a tri-fold measurement. In addition, even with proprietary studies
using 3000 or more respondents, the Evangelical modernist category has suffered a small

N problem, and efforts to use this tri-fold measurement are declining.9

2.3 Evangelical Separatism and Political Participation

Historically many Evangelical denominations within the Evangelical tradition
have held to views of separation from modern culture, limiting their political and cultural
involvement. The call to “keep oneself unstained by the world” in some denominations
refers to a watchful attitude against secularization; while in others, it results in drastic
lifestyle choices involving a near total withdrawal from modern culture (Lenski, 1961;
Martin 1996; Whiteman 1990). While not uniform in strength or deﬁnition across all
Evangelical denominations, this separatism includes varying levels of the rejection of
ecumenicalism, of civic involvement, of political activism and of basic citizenship.
Historically, this separatism was enhanced by the rise of the Fundamentalist movement.
According to the historian George Marsden, the marks of Fundamentalism include
"separatism, insistence on strict doctrinal purity, and incivility towards persons with other
beliefs" (1991, 23). This movement has an effect on American denominations in three
forms: it results in fundamentalist denominations emphasizing separatism even from
other Evangelicals; it introduces fundamentalist inﬂuences into denominations (pietist,
holiness and Pentecostal denominations) that were not strict fundamentalists; and lastly,

it produces fundamentalist identiﬁers in mainstream churches who, despite their beliefs,

 

9 Personal conversation with Lyman Kellstedt, September 4, 2004

13

 

behaviorally give up on both the impetuous to battle modernity and the effort to embrace
separatism (Marsden 1980, 195).

The resulting effect of fundamentalism across denominations is a conservative
spectrum of separation. At one end, it is populated by denominations emphasizing a faith
that requires limiting their contact with the greater society. This extreme position not only
advocates little cultural involvement, but also eschews the basic roles of citizenship. This
view is balanced at the opposite end of the spectrum by Evangelical denominations that
encourage cultural and political interaction, while emphasizing the importance of a
distinct morality and worldview. Along this spectrum fall various other conservative
denominations emphasizing to a greater or lesser degree the need for community activism
and political involvement (Lienesch 1993; Smith 1998; Thomas and Dobson 1999).
Green (2000) shows two intermediate steps between extremes on the separation spectrum
in terms of political involvement and citizenship. One step reluctantly advocates limited
duties of citizenship (such as voting) and approaches political activities with skepticism.
A second intermediate step takes a more expansive view of citizenship but is also dubious
about involvement in politics. This spectrum matches the common assmnption that most
Evangelical groups have some aspect of separatism in relation to political activities
(Campbell 2003).

Historically, this separatism wanes and increases with variance in the
participation cues from both Evangelical leaders, and the increase or decrease of
mobilizing resources from Evangelical social movement organizations, and/or
Evangelical political organizations. A ﬂurry of such involvement and increased activism

surrounds Prohibition in the 1920s and the anti-communist crusades of the 19505 and

14

 

early 19603. These are bracketed by periods of withdrawal as the participation cues
decline and the movement experiences repudiation of its positions and politicians.'0

This separatism came under ﬁre in the late 19703 from a surprising source - the
Reverend Jerry Falwell. Falwell, a lifelong fundamentalist, reversed his aversion to
political activism, and in 1979, founded the Moral Majority based on the political goals
of “recruitment”, “information” and “mobilization” of Evangelicals (Wilcox 2000).
Similar groups including Christian Voice and Religious Roundtable were also founded
during this time period. These groups delivered the ﬁrst wave of new cues that oriented
Evangelicals from political inactivity to political participation. Also, this time period
experienced the election of the ﬁrst avowed “Evangelical” President, Jimmy Carter in
1976. Following his election, Carter called for Evangelicals to leave their separatism and
reengage in the political process (Wilcox 2000). While there is little evidence that
Evangelical participation increased with the 1976 election (Perkins, F airchild and Havens
1983), pundits and reporters give credit for the Reagan victory in 1980 to resurgent
Evangelical voters (Moen; Wilcox 2000).

Regenerus, Sikkink and Smith (1999), also demonstrate the effectiveness of these
new participation cues on Evangelicals. Even when breaking down the Evangelical

tradition into separate components of Evangelicals, Fundamentalists and Charismatics,

they ﬁnd that these religious identities (along with practicing Catholics) are the only

 

to The “Scopes Monkey Trial” in 1925 was a public humiliation of both Fundamentalism and its most
public leader, William Jennings Bryan. The trial, along with the repeal of Prohibition in 1933, led to a
retreat by conservative Christians that kept them out of public view until the late l970’s (Simpson 1983).
There is, however, a period of less noticed mobilization for the Evangelical Right in the 19503 and early
19603 as some ER groups aligned themselves with Senator Joseph McCarthy against communism. This
period of activism was limited in its scope within the ER and failed to attract attention outside the ER. It
declined in size and inﬂuence after the crushing defeat of its favored candidate for president, Barry
Goldwater in 1964 (Wilcox 1992). Hunter (1987) agrees with the patterns of decline and also deﬁnes each
of these time periods of involvement (19203, 19503-19603, 19803) as three distinct waves of political
activism.

15

identities relying on conservative Christian leaders or political organizations for
assistance in voting decisions. As the new cues become more entrenched within the
Evangelical tradition, along with new patterns of participation, most research continues to
center on the inﬂuence of Evangelicals on the political process. The effects of political
participation on Evangelicals remain the untold story of this time period. Much of this
untold story reﬂects a lack of interest in religion on the part of political participation

models.

2.4 The Socioeconomic Model

In political science, three overlapping and complementary models provide
the foundation for explaining political participation. The basic model of
participation in political science is the “standard socioeconomic model”. It
focuses on political participation arising from individuals’ socioeconomic status
(SES) mediated through their civic attitudes. Verba and Nie (1972) show that
higher levels of income, education and social status increase respondents’ free
time, money and skills, which in turn, increases campaign activity, voting,
communal activity and particularized contact. According to Verba and Nie, these
latter four forms of political participation are not a unitary concept, but rather four
distinct modes of participation. Socioeconomic status alone does not increase
participation, but rather works through the intervening effect of efﬁcacy, interest
in politics, a sense of civic duty and levels of political information. Social

environment also ﬁgures into the standard model as individuals with higher SES,

l6

in contrast to lower SES respondents, live and work in social environments that

increase civic skills, reinforce civic attitudes and encourage participation.11

2.5 Civic Volunteer Model

Verba, Schlozman and Brady (1995; Also Brady, Verba and Schlozman 1995;
Burns, Schlozman and Verba 2001) build on the SES model with their Civic
Volunteerism Model. By over-sampling political and civic activists, they elaborate on the
SES model and ﬁnd participation inﬂuenced by three factors: the resources to participate;
psychological engagement; and “recruitment networks” that encourage and equip citizens

to participate. Verba, Schlozman and Brady place more emphasis than Verba and Nie on

 

" Other researchers have built upon and reinforced the dominance of the SES model. Acock and Scott
(1980) found that social class affected attitudes towards political participation and higher social class
resulted in more highly visible political participation (voting, working for a candidate, wearing a political
button) versus low visibility activities (watching or listening to political programs, reading about the
election). Multiple studies using a variety of forms of participation have conﬁrmed the conclusion that
higher levels of education are more likely to result in greater levels of participation while lower levels of
education are not. While not as robust a ﬁnding, higher income over lower income also increases
participation (Bennett 1995, Burns, Schlozman and Verba 2001; Conway 2000; Dalton 2002; Kenny 1992;
Leighley 1990; Verba, Burns and Schlozman 1997; Verba, Schlozman and Brady 1995). While the
original SES model did not include age, increasing age has also been found to boost levels of participation;
though, there appears to be some drop off among the elderly for more active forms of participation
(Jankowski and Strate 1995; Jennings, 1979; Jennings and Markus 1988; Jennings and Niemi 1981;
Wolfﬁnder and Rosenstone 1980). Jankowski and Strate also conﬁrmed Vera and Nie’s distinction
between voting participation and campaign participation as separate modes of participation, and
demonstrated that the structure of participation is stable across cohorts, periods and age groups. Strate et al.
also conclude that civic competence (knowledge and habits of political knowledge acquisition), community
attachment, church attendance, government responsiveness and party attachment also mediate age effects.
They conclude that increases in each of these variables contribute to rising participation as individuals age.

Other factors found to increase participation include marriage, media usage, race, and sex (Bennett
1995; Eveland and Scheufele 2000; Jennings and Niemi 1981; Leighley 1991; McLeod and McDonald
1985; Scheufele and Nisbet 2002; Stoker and Jennings 1995; Verba and Nie 1972). However, these
predictors vary in their signiﬁcance and magnitude depending on the mode of participation being measured
(Bennett 1975, Dalton 1988; Eveland and Scheufele 2000; Leighley 1991, Neuman 1986). In particular,
these social status and civic orientations are least compelling for voter turnout. Regardless of the model
used, signiﬁcant inﬂuences on political participation are attributed to gender and race. For most modes of
participation, aside ﬁ’om voting, women demonstrate lower levels of involvement (Conway 1991;
Rosenstone and Hansen 1993; Verba and Nie; Verba, Schlozman and Brady 1995). Women, though, have
reversed a historical trend of lower rates of voting (Conway 2001). Blacks are less likely to vote than
whites though the differences in rates of registration and voting have declined over the past forty years.
Still when education and region are controlled for, Blacks of similar SES status still vote at lower levels
than whites (Abrarnson and Claggett 1984).

17

the social environment as a part of adult socialization/experiences which shape the skills
that citizens use in the civic and political realm. They ﬁnd that these skills increased by
the job environment, family structure, church involvement and organizational
participation. In addition, the social networks within these environments become
recruiting grounds for activity (See also Schlozman, Burns and Verba 1999). They
determine from this model that disparities in availability of political resources explain
different rates of political participation among Blacks, Latinos and Whites (V erba et al
1993). Verba, Schlozman and Brady conclude that the effect of activity in non-political
organizations, such as volunteer groups and the workplace, is similar between all
organization types. This has come under scrutiny as Ayala (2000) ﬁnds that activism in
volunteer organizations has a positive affect on political participation, while workplace

activities, which are more coercive in nature, have a diminished effect on participation.

2.6 Mobilization Model

The Mobilization Model focuses on the opportunities and cues provided for
political activism by the respondent’s environment. Mobilization serves as a substitute for
resources or civic orientations as an instrument of intervention between political
participation and socioeconomic status. Higher SES provides more opportunities for
participation through direct mobilization sponsored by political leaders and political
parties and indirect mobilization arising from social networks (Leighley 1996, 2001;
Rosenstone and Hansen 1993).

Direct mobilization also results from both, close competitive elections, and higher

campaign spending (Cox and Munger 1989; Gilliam 1985; Patterson and Caldiera 1983),

18

 

as well as party, candidate or group contact (Huckfeldt and Sprague 1992; Rosenstone,
Hansen, Freedman and Grabarek 1993). Rosenstone and Hansen ﬁnd that direct
mobilization involves the distribution of free information about issues and candidates. It
also includes the reduction of costs for participation through: offers of free childcare to
permit attendance at meetings; transportation to the polls; distribution of absentee ballots
and voter registration forms; and providing suggested text for letters to representatives
and media outlets.

Rosenstone and Hansen conclude that indirect mobilization relies upon the social
obligations found between friends, families and social group members. Calls for
participation, from informal contact, draw upon the individuals’ sense of commitment to
those persons with whom they most identify.

Cutting across each of these models is a similar view of the inﬂuence of volunteer
organizations in increasing political activity. As previously discussed, Verba, Schlozman
and Brady conclude that involvement in church and volunteer organizations provides a
training ground for participation skills and increases the likelihood of recruitment. Verba
and Nie also ﬁnd that involvement in volunteer organizations increases political
participation. Those volunteer organizations, with clearly expressed political goals,
providing opportunities for political dialogue, are the most effective in promoting
participation. Rosenstone and Hansen reach a similar view of voluntary organizations.
They conclude that involvement in such organizations “makes people susceptible to
mobilization” (87). Together, these participation models lay the theoretical foundations

for subsequent work concerning the inﬂuence of religious involvement on participation.

19

2.7 Participation and Religion

Despite these participation models, much of the research on political participation
and religion suffers from both a lack of attentiveness and a dearth of speciﬁcity
concerning religious traditions. Verba and Nie give little focus to religious affiliation in
their SES model, except to note slight differences in types of participation between
Protestants (higher) and Catholics (lower). They also evaluate the inﬂuences of church
related groups, such as Bible studies and women’s auxiliary groups, and uncover that
they encourage community activism while providing little forum for political discussion.

Secret, Johnson and Forrest (1990), using the SES model, conclude that
religiosity, as measured by attendance, and intensity of religious attendance increases
voting participation, regardless of race. They also ﬁnd non-traditional political activities
such as labor strikes, civil rights, and war protests are unaffected by religiosity.

Verba, Schlozman and Brady (1995) look at religious involvement serving as an
environment for gaining skills for civic and political participation. They discover that
Protestants have greater opportunities for church participation and acquisition of skills
than Catholics, while the exposure of both groups to political messages within their
denominations is equal. In support of numerous other studies, they also ﬁnd that church
attendance, in general, increases the likelihood of voting, but not more demanding forms
of political participation (See Calhoun-Brown 1996; Harris 1994; McKenzie 2001).
However, like Verba and Nie they provide little differentiation in their religious
categories except for the broad groupings of Protestant and Catholic. Rosenstone and

Hansen also acknowledge that church attendance has a large, positive effect on voter

20

turnout, but a very minor effect on campaign participation. They, nonetheless, repeatedly
fail to examine the inﬂuence of religious traditions in their models.

Fortunately, several studies have gone beyond the work of the foundational
models to broaden the understanding of religion and participation. Scheufele et al. (2003)
ﬁnd three dimensions to the inﬂuence of religion on participation. In keeping with
previously discussed models of participation, they show a positive effect of both
increased church based cues and political discussion on mobilizing attendees. This
mobilization, for obvious reasons, is also a function of attendance increasing accessibility
to the cues and discussion. Scheufele et al. ﬁnd doctrinal commitment or doctrinal
conservatism indirectly reduces participation, as those with the highest conservative
religious beliefs are less likely to read the newspaper, have lower levels of political
knowledge, and hold lower views of their own political efficacy. Scheufele et al. also
demonstrate that a signiﬁcant distinction within Protestant Christianity is affecting
participation. Exposure in church to political cues and the frequency of political
discussions is lower for Mainline Protestants than for Evangelical Protestants. While
doctrinal commitment lowers participation in most denominations, for Evangelicals, it
results in a “likely self selection process” that places them in churches with greater cues
and discussion and hence greater participation (316).

An examination of increasing participation, arising from civil skills gained in
churches, reaches a different conclusion from Verba, Schlozman and Brady. Djupe and
Grant (2001) show that church-gained civic skills for regular attendees fail to increase
political participation. They do conclude that religious traditions result in different levels

of participation, which mimics Beyerlein and Chaves (2003) results. Djupe and Grant

21

have additionally found: participation increases when a fellow religionist recruits
respondents for political activities; when religious leaders hold political meetings in the
church; and when the partisanship and political norms of the church are clearly perceived
by the respondent. Lastly, participation increases when respondents perceive their
political activism as closely tied to their church activism. In effect, those persons who see
their religious and political goals as synonymous increase their participation in both

realms.

2.8 Evangelicals and Participation

The emergence of Evangelicals as a political force does much to refocus the
various political participation models on the importance of religion. Few studies, though,
examine aggregate change in participation over time with the exception of those studies
examining shifts in party vote”. As Guth et al. note “one area that has remained
relatively neglected in this resurgent interest in the politics of religion has been the study
of political participation” (2000, 1). Much of the prior research targets results from
individual election cycles. For the most part, these studies focus on voting choice (which
candidate/party) as well as party identiﬁcation and ideology, rather than the choice to
vote. These studies almost exclusively focus on Evangelicals and the Republican Party.

The 1992 Presidential election reveals an increasing willingness on the part of

 

12 Manza and Brooks (1997) argue that that conservative Protestants are long-term GOP voters in
presidential elections from 1960-1992. They also ﬁnd no evidence of increased mobilization on the part of
conservative Protestants. Manza and Brooks divide the Protestant population into three categories,
theologically liberal, theologically moderate and theologically conservative. Layman (1997), using the
more common dichotomous conservative versus non-conservative measurement, for evaluating elections
from 1980 to 1994, concludes that conservative denominations have become more Republican, and that
religious commitment has increased GOP voting.

22

Evangelicals to vote for George Bush, particularly those for whom social issues are an
important inﬂuence on their vote (Kellstedt et al b. 1996). Layman and Carmines (1997)
show that religious traditionalism increases the probability of voting Republican,
identifying with the Republican party, and holding a conservative ideology. Layman
(2001) ﬁnds high orthodoxy (or religious traditionalism) and high religious commitment
increases the likelihood of voting Republican in Presidential elections from 1980-1996
(See also Kellstedt and Green 1993; Kellstedt et a1 a. 1996; Kohut et a1 2000; Mockabee
2002). Other studies spotlight the inﬂuence of Evangelicals/Christian Right on the GOP
at the state level (Green, Rozell and Wilcox 2000, 2003).

Given the popular assumption of increases in Evangelical participation, it is
surprising to ﬁnd how few studies examine historical voting and campaign participation
trends. In part, the studies are few because much of the focus is on the Christian Right or
Religious Right rather than the Evangelical Tradition. The earliest efforts at evaluating
Christian Right mobilization focus on social group memberships, ﬁnances and voter
registration drives (Jorstad 1987; Liebman and Wuthnow 1983). When the focus shiﬂs to
voting or campaign activity, the results are meager. Smidt (1987) for example, ﬁnds a
drop in Evangelical voting from 1980-1984 with one exception; he ﬁnds evidence that
young Evangelicals (ages 17-34) are increasing their voting participation and their
identiﬁcation with the GOP, compared to non-Evangelicals of the same age in both
elections. Wilcox (1992) ﬁnds hints that Moral Majority supporters are voting at higher
rates than predicted in 1980 when controlling for demographics, but he ﬁnds no such
change in 1984. Another study ﬁnds Evangelical voting in congressional elections falls

slightly between 1994 and 1998 (Kohut et a1 2000). Wald, Kellstedt and Leege (1993)

23

conclude that when demographic and political variables are controlled for, being a part of
the Evangelical traditional increases the likelihood of voting but does not affect campaign
activity. This is the only religious tradition they ﬁnd to signiﬁcantly affect voting or
participation. In one of the few studies to examine participation beyond voting, Guth et al
(2003) (using the Third National Survey of Religion and Politics) ﬁnd that conservative
Evangelicals are the only religious movement to signiﬁcantly increase political activism.
Unfortunately, the scale used for political activism in this study must be viewed with
caution, as respondents are asked to recall their activism for a four year (1996-2000) time
period. In addition, campaign participation is combined with protest activities, as well as
contact with public Ofﬁcials, to form the scale suggesting a mixing of dimensions of
participation and causal factors.

The Republican Party is also shown to beneﬁt from Evangelical participation.
Analysis of national convention delegates ﬁnds the percentage of GOP delegates
identifying themselves as “Evangelicals” increases from 10% in 1972 to 17% in 1992
while Evangelical identiﬁers among Democrats hold steady at 6%. Evangelical grth
among GOP delegates is exclusively found among Evangelicals with regular church
attendance while the percentage of non-regular attending Evangelicals remains static
(Layman 1999). Curiously, what remains unmeasured is the common assumption about
increases in Evangelical participation.

Another unmeasured assumption associated with participation is a change in
Evangelical sophistication and political thought. As participation increases, Evangelicals
are viewed as becoming more politically sophisticated. Changes in Evangelical rights

organizations, their agendas, and their levels of support, document this change in

24

sophistication in the 19903. Evangelicals widen their agenda from moral issues to using
“the language of liberalism” and focus on “rights, equality and opportunity” (Moen 1994
352). They also achieve greater ﬁnancial stability, broaden their internally acceptable
range of theological positions, move away from Capital Hill lobbying to the grassroots
focus of the early 19803, and experience an increase in new organizations, as well as a
revitalizing of older groups. (Moen 1992; Moen 1994; Wilcox, Jelen, and Linzey 1995;
Moen 1996; Rozell and Wilcox 1996; Wilcox, DeBell, and Sigelman 1999; Jelen 2000).
Evangelical insiders encouraging Evangelical leaders to communicate in language
applicable to the larger public, in an effort to distinguish between cultural pluralism and
cultural relativism, and to learn civility in public discourse (Cromartie 2001), further
reinforce this assumption of sophistication.

While these perceptions of sophistication do not match the measurements used in
political science, the discussion adds to the (largely unexplored) questions of how
evangelical participation has changed, what is the effect of that participation within the
Evangelical tradition and how has it made Evangelicals more politically sophisticated.
These two gaps, unmeasured changes in participation and the effects of those changes,
provide the impetus for this study. To lay the foundation for how increased participation
within the Evangelical tradition is affecting the tradition, I turn in the next two chapters to
how participation inﬂuences religious traditions and how participation changes

measurements of political sophistication.

25

CHAPTER III

WHY CHANGING PARTICIPATION MODIFIES EVANGELICALS

The increasing participation of Evangelicals is occurring across time and with
crosscutting inﬂuences that may aid the changing inﬂuence of the participation or negate
its affect. The openness of Evangelicals to political inﬂuences raises the possibility that
political and societal inﬂuences will change the Evangelical tradition by increasing
secularization, and by facilitating the adoption of democratic ideals such as tolerance. At
the same time, conservative, religious motives for participation (including the culture war
motif) may lend a counterbalance to the expected moderation of Evangelicals and cancel
out some of the expected liberalizing changes. Simultaneously, changes within the
Evangelical tradition, while not a consequence of political participation, have the
potential to inﬂuence the tradition in a manner similar to the increased political

involvement.

3.1 Secularization

Modernization/Secularization theory has predicted the decline of religion in
industrial and postindustrial society for over a century. The forces of modernity combine
to undermine the importance and inﬂuence of religion. Increasing rationalism, as
measured by increases in education, technology, and science, weakens the trust of the
masses in religion. A commitment to cultural diversity, arising from increased mobility
and the decline of traditional communities, as well as economic growth with its rejection

of religious separatism are additionally theorized to hasten the decline. In addition, liberal

26

democracy contributes to the process with its separation of church and state. By focusing
on harmony over religious orthodoxy, in conjunction with an emphasis on relativism,
liberal democracy undermines religion’s claims of authoritative truth. (See Weber 1948,
1976; Wilson 1966; Martin 1978; Bruce 2002; Norris and Inglehart 2004). Bruce paints
the results of this decline in very broad strokes. He states that secularization will result in:

(a) The declining importance of religion for the non-roles and institutions

such as those of the state and economy; (b) a declining of the social

standing of religious institutions and (c) a decline in the extent in which

people engage in religious practices and display beliefs of a religious kind,

and conduct other aspects of their lives in a manner informed by such

beliefs (Bruce 2002, 3).

Some argue that secularization is failing to live up to its predictions, and the
United States, along with Iran, Ireland and Poland are often cited as the prime cases for
rejection of the theory (See Berger 1999; Hadden 1987; Stark and F inke 2000; Martin
1999; Stark and Iannaccone 1994; Wallis and Bruce 1991). Many of these researchers
conclude that resurgent American and third-world Evangelicalism, and spreading Islamic
fundamentalism represent a defeat of secularization and call into question the underlying
theory. Martin, in particular, sees these movements as an adaptation (especially by
Evangelicals) to successfully create a political space as “inﬂuential commentators within
a pluralistic society” (1999, p. 48). This political space then becomes a buffer zone
around the Evangelical sub-culture. Martin, however, fails to analyze the possible
consequences of this new political interaction on the subculture. He ignores the long-term
inﬂuences of modernization resulting in this shift from separatism to activism.

Proponents of secularization argue that these resurgent movements are reactions to the

failure of separatism to isolate them from the modern world. These growing religious

27

movements are viewed as perhaps ﬁnal (ultimately unsuccessful?) efforts by adherents to
defend their religious faith and way of life.

Rodney Stark and others use Rational Choice Theory to promote the most serious
alternative theory to secularization. Stark and Bainbridge (1987) contend that a cycle of
secularization, revival, and religious innovation best explain the pattem of religious
inﬂuence. Older religious organizations tend to make accommodations with the secular
culture. New sects, in reaction to the secular drift, adopt a traditional view of the
supematural and, as a result, reject accommodation. As the sects gain popularity and
power in society, they begin the same drift toward accommodation and secularization.
Stark and Bainbridge believe that over centuries, and even millennium, entire religious
traditions may fall into disrepute from accommodation and be replaced by entirely new
traditions. This view does not see secularization as resulting in a decline of religion, but
rather a cyclical replacement of religious groups with new sects and new traditions. The
key argument against this view is that the cycle is limited primarily to the United States
(an admitted outlier in most measurements of religiosity and secularization). These cycles
of replacement may slow but not ultimately negate the secularization trend.

In addition, supporters of secularization argue that the theory is being applied too
rigidly. Bruce contends that secularization is not a universal law, but rather an
observation of patterns in Western Europe. Norris and Inglehart view secularization as “a
tendency, not an iron law” (p. 5). They readily accept the United States as an outlier
among postindustrial nations (along with Ireland and Italy) for not showing all of the
expected affects of secularization. They do, though, marshal evidence to show some

effects of secularization in the United States, pointing to rising percentages of seculars

28

(no religious affiliation) in the United States, and providing support for Himmelfarb’s
(1999) assertion that there are two distinct cultures in America.‘3 Bruce also gathers
evidence to show that the United States is not the outlier that others suppose. He focuses
on the declining popularity of Mainline Protestants and waning traditional morality

among conservative churches as evidence of secularization.

3.2 What are the Political Effects of Secularization?

Religious subgroups use separatism to isolate themselves from the affects of
secularization. The subgroup socializes its youth into its tenets and beliefs while using the
freedoms of the political system to avoid direct confrontation with the worldview of the
culture. Penning and Smidt point out that this strategy runs the risk of marginalizing the
subgroup, as its views are excluded from political and social discourse. Should the
subgroup choose to reject separatism, it is forced to confront diversity as a cultural and
political norm. Ultimately, claims to exclusive “truth” within the subgroup are weakened
by pressures to privatize religious beliefs. Furthermore, these claims are undermined by
exposure to competing values, beliefs and religious views; although, the rejection of
separatism does not necessarily result in adaptation to the culture. Berger indicates that a
“religious revolution” resulting in the hegemony of a mandated, state-sponsored religion
is also a possible replacement for separatism. Realistically, this approach has few chances
of success in liberal democracies valuing multiculturalism and plurality (Berger 1999, p.

3.4).

 

‘3 One culture is nationalistic, puritanical, religious, and family oriented. The other is self-indulgent,
secular, tolerant and multicultural (Himmelfarb 1999).

29

Evangelical rejection of separatism raises the question; in the political realm how
will this increased secularization pressure affect Evangelicals? Secularization theory
posits many consequences of increased secularization: eroding religious habits
accompanied by an increase in seculars, increasing moderation (liberalization) in social
and political attitudes (assumed to accompany a decline in the authority of religious

leaders), and gains in tolerance.

3.2.1 Eroding Religious Habits and Increasing Numbers of Seculars

An increasingly secular society will result in a decline of religious enthusiasm and
participation. Norris and Inglehart maintain that, “Religious habits will erode and the
public will become indifferent to spiritual appeals” (7). They find evidence of this theory
in Woodbeny et al.’s ﬂawed hypothesis that unchanging levels of church attendance over
the past 40 years are a reﬂection of increasing social desirability. Woodbeny et al. argue
that attendance has in fact dropped, and that their analysis of true attendance rates shows
a 40% over-report for current weekly religious attendance. Woodberry et al. assume,
however, that other historic attendance reports are more accurate. They fail to adequately
address the issue of why social desirability would inﬂate current attendance ﬁgures, but
not those ﬁgures ﬁ'om four decades prior.

Beyond the religious attendance ﬁgures, there is some evidence of a decline in
religious salience. Wuthnow (1989) states that religious indicators (participation, belief in
God, belief in the afterlife) are as strong today as they were forty years ago. In
contradiction to his own conclusion, Wuthnow (1988) ﬁnds evidence that Bible literalism

is weakening, as is the belief in hell. Applying these results to Evangelicals is difﬁcult, as

30

in the analysis of attendance, and in the analysis of religious indicators, none of the
studies break down the decline by religious tradition.

Furthermore, the percentage of Seculars in the United States has doubled over the
past decade, but the decline is not uniform across traditions. And, while the percentage of
Seculars increased over the past twenty years, it appears to be exclusively a shift from
Mainline Protestants into the Secular category. Evangelicals, Black Protestants and
Catholics have maintained their share of their population (even with population growth)
while Mainline denominations have lost both membership and attendance (Hout and
Fischer 2002). It is also important to note that while Seculars represent the segment of the
population who became indifferent to established religious traditions, that does not mean
they ﬁt a model of scientiﬁc rationalism. Martin’s discussion of secularization and
declining religion in Great Britain is tempered by the comment that, “the decline of
religion was not followed by rationality but by subterranean theologies and nonrational
sentiments and superstitions” (Martin 2000, 35). Seculars may still view themselves as
religious or spiritual; still, their behavior includes little evidence of religious belonging or

devotional practice, and it has no connection with established religious traditions.

3.2.2 Political and Moral Consequences

Norris and Inglehart argue that secularization results in a decline of the
preeminence of church leaders. With this decline comes a reduction of inﬂuence over the
moral and political issues of the day. They identify possible changes as occurring in
attitudes toward divorce, abortion, homosexuality, cultural change in general, and
equality for women. Chaves (1994) takes a slightly different tack in arguing that

secularization should be viewed, not necessarily as a reduction in religion itself, but as a

31

 

 

decline in religious authority; the religious beliefs of individuals can remain unaffected.
Meanwhile, the ability to mobilize and institutionalize those beliefs into social and
political action will decline as religious authority declines. While Norris and Inglehart
center their discussion on the waning inﬂuence to affect attitudinal shifts among
identiﬁers, Chaves focuses on the inﬂuence to mobilize them. In support of Norris and
Inglehart, Bruce (2002) points to some evidence of secularization in moral standards with
the increasing acceptance of divorce among Evangelicals.

Hunter also ﬁnds evidence of secularization. He uses a terminology of separatism
and moderation (as a substitute for “secularization”) to describe ongoing changes among
Evangelicals. He points out that “separatism and moderation are in some ways different
words for sectarian divisiveness and accommodation” (Hunter 1987, 195). Hunter makes
the important point (also iterated here) that some would interpret these terms as negative
assessments; when for the social scientist, they are merely the best value-neutral lexis
available. He believes that Evangelicals have taken both tacks (separatism and
moderation) simultaneously to confront a loss of protestant hegemony in the United
States. Despite the historical separatism, Hunter argues that the boundaries between
Evangelicals and the greater American community have become blurred. While not
settling for a simplistic deﬁnition that Evangelicals are becoming more liberal (though his
conclusions support this thesis), he argues that Evangelicals have broadened their
deﬁnitions for faith, belief and behavior to encompass previously rejected liberal
positions. This broadening occurs by adding to rather than rejecting more conservative
positions. Moderation, according to Hunter, is the acceptability of liberal positions as

valid expressions of orthodoxy without rejecting more conservative views. Hunter’s

32

 

deﬁnition of moderation also includes increasing tolerance and tolerability or “alignment
to the normative codes of civility” (47). For Hunter, moderation would involve following
a liberal path in political, moral and/or social attitudes, as well as increasing tolerance for
out-groups. Hunter’s theorized tolerance, though, would have to overcome J elen (1991)
and Wilcox’s (1992) ﬁndings that Christian Right supporters are motivated by negative
attitudes toward communists, gays and lesbians, feminists, and the American Civil
Liberties Union. These negative motives for participation may undermine the expected
moderating inﬂuence of participation.

Penning and Smidt (2002) reexamine Hunter’s conclusions with a new survey of
Evangelical college students, and argue that secularization is minimal for Evangelical
students. They ﬁnd little evidence that Evangelical students are less orthodox than
previous generations. Aside from women’s roles and support for an equal rights
amendment, Penning and Smidt ﬁnd no change on moral issues from Hunter’s study.
They also ﬁnd no shift in overall tolerance levels for these students (158). It should be
noted that both studies (Penning and Smidt, and Hunter) show strong secularizing
changes in comparison to the data collected in 1962. Evangelical attitudes have been
static since Hunter’s 1982 study, but there is no evidence to support that they reversed a
signiﬁcant secular leap that took place in the 19603. These results are applicable to only a
small portion of the American populous as both Hunter, and Penning and Smidt’s studies
observe a very small and unique portion of the Evangelical community. For example,
Penning and Smidt discover no change among Evangelical students in their attitudes
toward divorce, despite the Evangelical divorce rate now rivaling that of the rest of the

population. The theorized effects of secularization are for the entire religious tradition,

33

 

 

not just students whose commitment and background include a specialized Evangelical
college inﬂuence. Secularization as theorized by Martin, Bruce and others is an aggregate

process found among average Evangelical identiﬁers rather than among its elites.

3.2.3 Increasing Tolerance

Tolerance is also one of the theorized byproducts of secularization and liberal
democracy. Tolerance weakens religion, as claims to exclusive possession of the truth are
counterbalanced by the need to tolerate other viewpoints in the political arena. Strong
adherence to subculture boundaries allows a religious tradition to exclude, or at least
minimize, the inﬂuence of alternative viewpoints and beliefs. Diversity is neither valued
nor encouraged. This lack of diversity is demonstrated in Evangelical attitudes.
Evangelicals have selective tendencies toward greater intolerance than do other
Americans (e.g. Davis and Robinson 1996; Jelen and Wilcox 1990). While Evangelicals
are not more racist in their attitudes toward Blacks, Asians, Catholics, Hispanics or
immigrants (Guth et al. 1996; Smith 1998), they hold less tolerant views of Moslems,
homosexuals, atheists, feminists, communists, racists and militarists (Jelen and Wilcox
1990; Tamney and Johnson 1997; Smith 1996). Race is not the focus of Evangelical
intolerance; but alternatively, other religions (including a lack of religion), divergent
moral positions (homosexuality) and what could be termed as a broader category of isms,
feminism, racism and communism, are instead their targets.

As the barriers decline, (in this case through increased political activism) political
cooperation requires tolerance of previously rejected religious, ethical and moral views.

This cooperation is viewed as “cognitive contamination” as worldviews, ethics, and

34

 

lifestyles collide through political interaction (Berger 1992). Accommodation and
compromise then follow as individuals embrace a rapidly expanding acceptance of
relativism (Berger; Hunter 1991; Penning and Smidt). As Bruce succinctly points out, it
is difﬁth for a ﬁmdamentalist to rail against “anti-Christ Roman Catholics” on Sunday,
and then cooperate with Catholic representatives at a moral/political coalition meeting on
Monday (Bruce 29).

While not addressing secularization theory, Warren (2001) presents the
(unmeasured) theory that some groups do not facilitate this increase in tolerance as
participation increases. Although his argument is applied to the Christian Right as a
social movement rather than to Evangelicals as a tradition, he does raise concerns that the
monolithic nature of the Christian Right (and its cues to the Evangelical tradition) do not
result in conﬂict and contamination from other views. He sees the Christian Right
increasing political participation and possibly increasing political skills, but not
cultivating civility or tolerance. This view, though, is conditioned on the premise that
Christian Right political participation is limited solely to involvement in Christian Right
groups. This is a premise, given the alignment and participation of Evangelicals with the
Republican Party that seems dubious. Prior work focusing on the inﬂuence of
participation does show evidence of participation increasing the norm of tolerance.
Political elites and activists acquire a commitment to democratic norms through the act of
participation. Participation then teaches citizens about politics and increasing attention to
public issues. This process includes learning the “rules of the game” such as adopting the
norms of compromise and negotiation, and demonstrating increasing tolerance for

outgroups (Cook and Morgan 1971; Freie 1997; Lane 1959; Nagel 1987). Increased

35

exposure to the norms of civil liberties, and the desire to effectively participate in the
democratic arena increases the motivation to accept the norms. However, this
participation/political socialization process is tempered by the ideology of its participants.
Conservatives are less supportive of civil liberty norms for all groups regardless of their
ideological bent (Sniderman et al. 1989, 1991; Sullivan et al. 1993). The implications of
this last ﬁnding for conservative Evangelicals are clear. Their increased political
involvement (and cultural contamination) should result in a pattern of increasing
tolerance for “out” groups, though the conservative ideology of Evangelicals may temper
the increase.

Several of these consequences of secularization are applicable to this study as a
result of the increased political participation of Evangelicals. In particular, renewed
Evangelical participation raises the possibility of increasing moderation in social and
political attitudes, and increasing tolerance from learning the “rules of the game”. The
difﬁculty with all the measurements of secularization is that there is nothing to suggest
that all indices decline at the same rate, or that the decline is evenly spread across
political or social characteristics and attitudes. Some behaviors and attitudes may be more
resistant to secularization inﬂuences and may have a greater lag time for the inﬂuence of
secularization to appear. However, the shift in Evangelical political participation, and the
resulting decline in separatism, leaves open the possibility of a more rapid secularization

occurring over the past two decades.

36

3.3 Fighting the Culture War

The “culture war” may be another inﬂuence tempering the secularization of
Evangelicals. James Davison Hunter and Robert Wuthnow conclude that the American
electorate is increasingly divided in a culture war fought by orthodox and progressives
who hold divergent views of the education, morality, law, education, family, media and
politics (Hunter 1991, Wuthnow 1988).

Wuthnow sees the cultural war as an inter- and intra-denominational struggle. He
perceives the ﬁrst divide as existing between Evangelicals and liberal religious/
secularists. The second divide exists within most of the larger established denominations,
as Lutherans, Presbyterians, Methodists and Roman Catholics split somewhat equally
between religious conservatives and religious liberals. Curiously, Wuthnow also includes
Baptists in this divided group even though conservative Southern Baptists effectively
marginalized more moderate Baptists in a series of internal denominational maneuvers in
the 19703 and 19803. Wuthnow also makes the mistake of lumping northern, mainline
Baptist denominations with Evangelical Baptists.

Unlike Wuthnow, Hunter views the culture war as 301er an intra-denominational
conﬂict. He divides the participants in the war into “orthodox”, who hold traditional
religious beliefs and a commitment to transcendent truth, and “progressives”, who have
little or no religious commitment and a relativistic view of morality and truth. These two
camps are seen as crosscutting all religious traditions. Layman’s (2001) exploration of
these intra- and inter-denominational differences ﬁnds evidence for both. He concludes
that the most committed and orthodox within each religious tradition are more

conservative on social and political issues than their less committed and less orthodox

37

members. At the same time, inter-denominational differences are evident as all Seculars
and all Jews are more liberal than the least committed and least orthodox of any Christian
tradition. Also, both high and low commitment/orthodox Evangelicals are more
conservative than their respective high or low Mainline Protestant, Catholic or Black
Protestant counterparts. Low Evangelicals are more culturally and morally liberal than
high Evangelicals, but they are still more conservative than are low Christians of any
other tradition. It is important to note that this ﬁnding does not provide evidence of a
broadening cultural divide with increasing numbers enlisting on each side. Other
researchers while exploring this thesis ﬁnd limited evidence of a deep division between
traditionalists and modemists (Bolce and De Maio 1999; Davis and Robinson 1996;
Green, Guth, Snridt and Kellstedt 1996).

Others have suggested that this orthodox/progressive division only exists among
religious and secular elites, and activists (Layman and Carmines 1997). A growing body
of research has called this concept into question, or has attempted to modify its scope
(Evans 1997; Miller and Hoffman 1999; Williams 1997). DiMaggio, Evans and Bryson
(1996), in particular, conclude that there is little evidence of increasing polarization on
social issues. Fiorina (2004) marshals evidence to show that the culture war is an artifact
of political party polarization. He maintains that the polarized alternatives presented by
the Republican and Democratic parties create the appearance of a polarized electorate, as
they are forced to choose between two extremes. F iorina sees the culture war as an elite
conﬂict, comprising perhaps ten percent of the population.

Regardless of its veracity, usage of the culture war concept has become common

in Evangelical circles. It is used as a rallying cry and an organizing principle behind the

38

 

push for a conservative social agenda. The culture war, in effect, ﬂavors the participation
cues and provides much of the focus for the ideological direction that this increased
political participation should take. Since the messages of the culture war are attacks on
secularization, the political participation fueled by this perceived conﬂict may serve to

negate the secularizing inﬂuences of participating.

3.4 New Paradigm Churches

The inﬂuences on American Evangelicals over the past two decades are not
limited to external social and political forces. Evangelical Christianity is also inﬂuenced,
during this time of increased political involvement, by new sects that reverse the expected
pattern of cultural accommodation for new sects. Niebuhr (1962) traces the evolution of
newly formed sects from their inception to their accepted status as denominations. New
religious sects originate with the goal of purifying either a denomination or the entire
religious tradition. These radical sects are marked by high commitment and high
separatism. Successive generations begin a pattern of declining commitment. This decline
is often accompanied by a rising socio-economic status fueled by religious-based
industriousness and piety. Internal pressures for cultural acceptance result in a rejection
of separatism, and the sect transforms into a denomination that rejects claims of exclusive
doctrinal truth (See also Michels 1962; Troeltsch 1992). For Evangelicals, new inﬂuential
sects reversed this pattern by beginning with a rejection of exclusivity and separatism,
which is followed by a call for the Evangelical tradition to increase its cultural

acceptance.

39

A phenomenon called the “Jesus Movement” in the early 19703 opened
conservative Christians (alter a great deal of initial resistance) to rock music and
counterculture attitudes, as waves of young people entering conservative churches
brought their own distinctive culture with them. Interestingly, at the height of this
movement, one of its critics suggested that it was ripe for political absorption by the
conservative right (Nolan 1971). The Jesus Movement functioned as a harbinger of
changes by remolding established churches within the Evangelical Tradition. It also
spawned four new sects/denominations (Calvary Chapel, Gospel Outreach, HOpe Chapel
and the Association of Vineyard Churches), which continue to be on the forefront of
social activism among conservative denominations (Miller 1997). These groups are
known for their rapid growth (Penin, Kennedy and Miller 1997). Despite being relatively
small denominations, eighteen Calvary Chapels and ﬁve Vineyards are classiﬁed as
“megachurches” with weekly attendance of 2000 or more (Hartford Institute 2004). One
of the marks of these denominations is a lack of antagonism toward the surrounding
culture (Di Sabitino 1999; Miller 1997; Shipley 1996; 1998). For example, these
denominations have readily embraced what is termed “post-feminist” ideology, where
gender equality is assumed as the norm of family and public life (Shipley 1998; Stacey
1990).

Another new movement within Evangelical Christianity, having a similar embrace
of the surrounding culture but not arising from the Jesus Movement, is the “seeker
church” movement. With some 9500 afﬁliated churches in the Willow Creek
Association”, this movement is typiﬁed by the 20,000 member Willow Creek

Community Church in Chicago that began in 1975 (Hybels 1995; Shipley 1998).

 

" Willow Creek Association. http://www.willowcreek.com/wca_info/, September 1, 2003.

40

 

Additionally comprising this movement is Rick Warren’s equally large Saddleback
Valley Community Church. Kellstedt and Green (2003) also classify this movement
among new paradigm churches. These churches encourage an increasing involvement in
social issues including racial equality, the environment and care for the poor and
homeless.

All of these new groups, while begimring in the 19703, came to prominence
among Evangelicals in the 19803, and their increase in size has been dramatic. At just
three decades old, New Paradigm membership in 2000 made up 6 percent of the adult
American population (Kellstedt and Green 2003). These denominations and movements
are viewed by sociologist Donald Miller as “new paradigm churches”, which are
inﬁltrating the practices and attitudes of more established segments of several religious

traditions including Evangelical, Mainline and Roman Catholic.ls Miller describes new

paradigm church characteristics as:

“1. they were started after the mid-1960’s

2. the majority of the congregation born after 1945

3. seminary training of clergy is optional

4. worship is contemporary

5. lay leadership is highly valued

6. they have extensive small group ministries

7. clergy and congregants usually dress informally

8. tolerance of different personal styles is prized

9. pastors tend to be understated, humble and self-revealing

10. bodily, rather than merely cognitive, participation in worship is the norm

11. the “gifts of the Holy Spirit” are afﬁrmed

12. Bible-centered teaching predominated over topical sermonizing” (Miller
l997,p.20)

 

'5 The author, ﬁrsthand, observed the cross-denominational interest in these paradigm churches ﬁrst,
through attending, along with 5000 people, a Vineyard Conference (sponsored by a Roman Catholic
Church) held on the campus of the University of Michigan (October 1985), and second, through attending,
with several hundred pastors, a church growth conference featuring Rick Warren (seeker church
movement) at Fuller Theological Seminary (November 1986).

41

New paradigm churches take the initial stance of embracing the world around
them while still holding to Evangelical theology. Shipley writes of one paradigm
denomination:

“. . .Vineyard is liberal on some social issues and conservative on

others, but its disposition toward alternative moralities is consistently

marked by tolerance . . . its churches remain doctrinally conservative and

thus represent an innovative form of bom-again Christianity that weds

traditional theology and contemporary culture: members celebrate a

personal relationship with Jesus Christ and talk on cellular phones; they

believe in miracles and value modern science; they interpret scripture

literally and shop at the Gap; they are committed to evangelizing and yet

show remarkable tolerance for cultural diversity” (Shipley, 107).

These new paradigm churches are also known for operating food banks, collecting
clothing for distribution, providing free or low cost healthcare and operating gang and
prison programs (Miller 1997). Many of these churches hold liberal views on racial and
class issues and strive (with very limited success) for racially mixed congregations and
leadership (Shipley)l6. Kennedy ﬁnds that paradigm churches exhibit higher levels of
tolerance and greater support for the civil liberties of religious opponents as compared to
other Evangelicals (Kennedy cited in Shipley 1992). They (for churches in the 19803 and
19903) pioneered the use of technology in their services through state of the art sound
systems, video projection, multimedia with computer graphics, and electronic

instrumentation in their worship bands, as well as effective use of church websites, email

and database systems. In addition to contemporary worship, many new paradigm

 

'6 Miller’s congregational surveys found that Vineyard, Calvary Chapel and Hope Chapel congregations
averaged 2% Black attendees and 1% or less for Blacks in leadership. Hispanics (13%) and Asians (7%)
were more highly represented than Blacks in these congregations, but leadership percentages were still
lower (Hispanic 3.5% and Asian 1%). Wolfe (2003) does ﬁnd evidence that new paradigm churches are
effective in building churches among the Hispanic community.

42

 

churches include drama and performances by contemporary dance troupes in their
services (Hybels and Hybels 1995; Shipley; Miller 1997; Nathan 1995)

New paradigm inﬂuence in established Evangelical churches is seen in: the
adoption of contemporary worship styles”, the use of technology, increased informality
in dress and decor, establishment of weekly support/12-step groups and home fellowship
groups, increased nrinistry by the laity with a focus on using spiritual gifts, expanding
levels of community involvement, concern about racial issues, and positive attitudes
toward the surrounding culture18 (Miller 1997, 2000; Price 2002; Shipley; Wolfe 2003).
The Church Growth Movement, which has spawned a plethora of books and conferences
for church leaders on how to “grow” a larger church, has been heavily inﬂuenced by new
paradigm churches. C. Peter Wagner, the “guru” of church growth, frequently invited Bill
Hybels, Rick Warren and John Wimber (founder of the Vineyard) to speak at his
conferences. Many of Wagner’s principles of church grth are drawn from “successful”
paradigm churches (Jorstad 1993; Wagner 1984, 1987, 1994).

These new paradigm churches have not been limited in their inﬂuence on
Evangelical churches. Many of the more conservative renewal movements within

mainline denominations have turned to new paradigm churches as models for change

 

'7 For example, Vineyard Music International, an offshoot of the Association of Vineyard Churches,
specializes in producing recordings, sheet music, and accompaniment tapes of contemporary worship music
(all written and recorded by local Vineyard church “worship” or rock bands). Since its founding in 1984, it
has grown to become the second largest publisher of sacred music in the world. Similarly, Maranatha
Music arising from the Calvary Chapel churches has grown to become a leader in worship music. In 2002,
the two companies (publishing music from approximately 1000 member churches) combined to provide
over twenty ﬁve percent of all the contemporary worship music used worldwide (Integrity 2002).

" Bill Hybel, pastor of Willow Creek Church writes, “Everything ﬁom the music to the printed program is
designed speciﬁcally for unchurched people. . .and because we are trying to reach a different generation,
we’ve used cutting-edge communication methods; contemporary Christian music, drama, multimedia,
video and dance.” (I-lybel and Hybel, p. 173-174). Vineyard values include, “We value being CULTURE
CURRENT. We aim to develop an atmosphere of ease, and to speak, act, and dress in ways in which our
culture can respond positively. For example, we reﬂect this value through worship music that is of a
popular style.”(The Vineyard USA 2004).

43

(Kellstedt and Green 2003; Miller 2000). While the precise level of inﬂuence is difficult
to measure, one study of United Methodist clergy found that 31% supported the ideals of
a United Methodist Evangelical Renewal Group (Good News) and 16% supported the
ideals of the Willow Creek Association (McKinney and Finke 2002). Many mainline
churches have adopted contemporary worship styles or have added alternative worship
services to their schedules. In addition, many of these churches have added small group
ministries, including support groups, to their weekly schedule (Miller 1997). There is a
lack of evidence as to how new paradigm churches have inﬂuenced Black Protestant
denominations or Roman Catholic Churches. It is possible since these churches have
arisen within a predominately white Evangelical protestant tradition that they have
inﬂuenced predominately a white Protestant audience. However, Miller has expanded his
study of new paradigm churches to congregations in Asia, Africa and Latin America. He
ﬁnds many of the new paradigm attributes have been incorporated into the most rapidly
growing churches in these regions. The other unifying factor to these Asian, African and
Latin American churches is that they emphasize the gifts of the Holy Spirit in a
Pentecostal or Charismatic manner (Miller 2003). While no research in the United States
has measured new paradigm inﬂuence among Black congregations, there has been a
documented increase of Pentecostal and Charismatic practices among American Black

congregations, suggesting a possible new paradigm inﬂuence on them (Jorstad 1993).

3.5. Which Inﬂuences?

We are then left with decidedly mixed inﬂuences to potentially moderate

Evangelical attitudes. Increased political participation raises the possibility of moderation

44

occurring over the past two decades. Secularization theory suggests that such moderation
may have begun with changes in political participation. This moderation may also have
been aided by the stimulus of the new paradigm church movement. However, the culture
war ethos and the very nature of the cues leading to participation represent forces moving
in the opposite direction. As a result, for those issues most emphasized by the new
political cues, an ideological stalemate seems most likely. For those issues not given a
conservative emphasis within the Evangelical tradition, they, perhaps, hold the most

promise for ﬁnding patterns of attitude moderation and increased tolerance.

45

CHAPTER IV
THE ORGANIZATION OF POLITICAL THINKING

The concept of “political sophistication” has come to dominate the search for how
to organize the political thinking of the electorate. This overly broad, and at times not
easily deﬁned category, has become the catchall for a myriad of structures all proposing
to capture the underlying cognitive forms that guide the political thought of the American
public. It is also somewhat of a misnomer to label all efforts at measuring organization as
“political sophistication”, as researchers themselves have used a wide variety of terms
including; “political information”, “political expertise”, “political complexity”,
“schemata” and “levels of conceptualization” to describe various aspects of these
cognitive structures. Neuman (1986) notes:

“The political sophistication of the citizenry is the central issue in

theories of mass politics and democracy . . . The issue was once listed as

one of the central foci of political behavior research . . . Later it was listed

as one of the ﬁve central concerns in the ﬁeld . . . The sophistication issue

has also been labeled as the central unifying conception to emerge from

early voting studies . . . Yet the research tradition yields no agreed-upon

deﬁnition or approach to measurement (191).”

However, from simple additive measurements of political information to more
complex searches for an underlying continuum(s) of ideology, political sophistication has
become the expansive rubric that ties these concepts and terms together.

This chapter examines the development of the concept of political sophistication
from its origins in The American Voter to current controversial usages such as schemata.
It highlights differences and similarities in the concept across various subgroups in the

discipline of political science, and lays the groundwork for measuring the political

sophistication of American Evangelicals.

46

4.1 Levels of Conceptualization

Measuring changes in political thinking is dependent on identifying an underlying
structure to respondents’ thought processes. Campbell et al. (1960) and Converse (1964)
were the ﬁrst to seek such a structure across the electorate:

“We are interested in the presence or absence of certain abstractions

that have to do with ideology; but we are also interested in the degree

to which an individual’s political world is differentiated, and, most

important, in the nature of the degree of “connectedness” between the

elements that are successfully discriminated. In short, we are interested

in the structure of thought that the individual applies to politics; and

this interest forces us to deal in typologies and qualitative differences

(1960, 221—22).”

While not explicitly using the term “political sophistication,” their structure or
political conceptualization provides the underpinnings for the evaluation of political
thinking and continues to inﬂuence the discussion of sophistication. In theorizing about
this conceptualization, Campbell et al. concludes, “there must be surrogates for ideology
that bring large aggregates to act as though propelled by ideological concerns” (217).
Based on the responses to National Election Study open-ended questions concerning the
likes and dislikes about the parties and their presidential candidates, they conclude that
there are four categories of these surrogates. They classify these substitutes in a
descending order from highest level to lowest by measuring both the “remoteness in
content” and the distance from their ideal of ideological sophistication. As these
categories descend away from the ideal, even logical constraint decreases in the
individual’s belief system. For individuals “the range of relevant belief systems becomes

narrower and narrower” as they use fewer and fewer amounts of speciﬁc and organized

information (Converse, 213). Not only during this descent is there a decline from

47

ideological concepts, but the objects in the belief system are reallocated from these
abstract ideological concepts to familiar social groups or leaders, or even descend to
“objects of immediate experience as family, job or immediate associates” (Converse,
1964, 213-15).

The ﬁrst level of conceptualization is ideology and near-ideology, which
classiﬁes people by their reliance on abstract concepts, such as liberalism or
conservatism, and a broad underlying dimension for making political judgments. The
respondents are then able to apply these concepts to speciﬁc positions on public policy
issues. While this underlying dimension does not have to ﬁt a liberal-conservative
dichotomy, there must be some sense of a continuum upon which political objects shift
(Campbell et all, 1960; Converse, 1964).

The next level is group beneﬁts, which involves evaluating parties and candidates,
in terms of the positive or negative consequences of their election on a group within the
population. These individuals, according to Campbell et al. (1960, 219-220), may be
using an “ideology by proxy” by following the cues given by an ideological leadership.

The third level, nature of the times, involves references to broad social conditions,
or isolated and speciﬁc issues, without the beneﬁt of the ideological considerations of the
ﬁrst two categories. Individuals credit or blame one of the parties for a social or
economic state, or express their displeasure or agreement with a speciﬁc stance or
program.

The last or bottom level of conceptualization is called an absence of issue content.
In this category, there is little discrimination between the parties, and candidates tend to

be evaluated on personal characteristics. Varieties of individual stances populate this

48

level, including blind party allegiance, rejection of parties, and difﬁculty in following
politics. Regardless of the grouping, the result according to Converse is they have “no
shred of policy signiﬁcance whatsoever” (1964, 217).

Converse seeks to reinforce the levels of conceptualization in his original 1964
work by demonstrating the connection between: higher levels of conceptualization and
education; conceptualization and participation; conceptualization and political
knowledge; and conceptualization and elites. The connection between education and
stratiﬁcation of the electorate, however, is shown to be weaker than Converse initially
posited (Converse 2000). In addition, Jacoby (1988) ﬁnds the levels are not merely a
“surrogate for the same type of political sophistication that is tapped by education or by
other means of cognitive ability”. Converse expands the levels of conceptualization to
include the concept of a belief system which is “a conﬁguration of ideas and attitudes in
which the elements are bound together by some form or constraint or functional
interdependence” (207). For Converse, constraint is the ability to predict individuals’
attitudes based on the initial or prior knowledge of another of the individual’s attitudes.
This constraint can be based on “pure logic”, or an underlying psychological attachment
to a general idea that “serves as a sort of glue to bind together many more speciﬁc
attitudes and beliefs” (210). The constraint can also be based on affiliation with
“particular niches in the social structure” (211), which occur through afﬁliation with the
niche and/or the presentation of the group’s attitudes and beliefs as a coherent whole.

Converse examines sophistication in terms of both constraint and in active use of
a liberal to conservative ideology. Converse reinforces the conclusions of Campbell et al.

that sophistication-based use of ideological comments in National Election Study open-

49

ended questions, accurate placement of major political parties on the same scale, and
passable deﬁnitions of conservative and liberal, all relate to both higher levels of political
activism and higher levels of education. His analysis of constraint shows that much of the
American public exhibits little constraint among issue attitudes. When examining
correlations among issue attitudes, he concludes that most of the public demonstrates
little structure to their belief systems. Both of these measurements of sophistication, the
use of ideology (and the accompanying levels of conceptualization) and constraint, to
varying degrees, continue to inﬂuence the examination of sophistication.

However, the idea of sophistication as ideology, or levels of conceptualization,
has come under ﬁre by researchers. Some ﬁnd fault with the validity of the political
conceptualization as measured by Campbell et al. (1960), Converse (1964), Nie, Verba
and Petrocik19 and others. Smith (1980), in evaluating Nie et al.’s results, ﬁnds high
levels of individual movement between the levels. He concludes that the levels of
conceptualization are not measuring stable levels of sophistication, but rather, are
measuring the short-term impact of the rhetoric of political dialogue. Nie et al. faults this
conclusion and speculates that Smith’s usage of a different coding scheme may result in
the differing conclusions. Abramson (1981) also questions Smith’s results and suggests
that the measurements are unreliable, but that the underlying concept of political
conceptualization is stable. Cassel (1984), after evaluating Smith and Nie eta1.,
concludes that the levels of conceptualization are reliable, valid and stable. Additional
research expands on the concept of political conceptualization, and in the process,

provides validation for the original work. An analysis of cognitive structures ﬁnds that

 

'9 Nie Verba and Petrocik use the master codes rather than the original respondent’s comments to produce
an expanded version of the levels of conceptualization. They expand the levels to seven categories but still
maintain a hierarchical structure with ldeologues ranked highest and Apolitical lowest.

50

 

 

ideologues have high salience to the liberal-conservative continuum, while individuals at
lower levels of conceptualization do not use well-deﬁned cognitive structures or the
liberal-conservative continuum (Jacoby 1986). A liberal-conservative dimension is
operational for a portion of the public with higher levels of education, but for many
people, ideological identiﬁcation functions at the level of providing cues for voting or
choosing candidates (Jacoby 1991). Additional works have also helped to reestablish the
validity of the levels of political conceptualization (Carmines and Stimson 1982; Hanger
and Pierce 1982; Knight 1985; Luskin 1987, Wyckoff 1987; Feldman 1988).

Further research builds, to a lesser extent, upon the idea of sophistication as the
levels of conceptualization and, to a greater extent, upon Converse’s idea of constraint
indicating sophistication. These measurements often include the new element of political
knowledge. Neuman’s evaluation (1986) ﬁnds that sophistication includes three related
dimensions: political salience (including political interest, involvement, and attentiveness
to political news), political conceptualization (both integration by the levels of
conceptualization and differentiation), and political knowledge (based on measurements
of political factual items). Neuman’s view of political sophistication draws upon the
variables most commonly used for political sophistication, and its various other
manifestations, such as political awareness, political expertise, cognitive sophistication,
political involvement and political interest. Luskin (1990), though, faults Neuman for an
overly broad concept that mixes interest and attentiveness with sophistication. At the
same time, Luskin (1987) also tries to measure sophistication with a combined
measurement of the levels of conceptualization and an eleven-point political information

scale. Krosnick (1990) and Zaller (1990) both reﬁne sophistication as “political

51

 

expertise” with a focus on how respondents organize knowledge. Zaller measures
expertise using media exposure, political awareness, levels of participation and placement
of groups and candidates on a seven-point ideological scale. Luskin also faults Zaller as
he “commingles sophistication, which is what he really seems to have in mind, with
education, political interest, media use, and political participation” (2002, 235).

As can be seen from these later works on political sophistication, researchers have
begun to redeﬁne the concept, at a minimum, to include an element of stored information.
The amount of correct information stored is viewed as important, or more important, than
the organization of that information. With the work of Delli Carpirri and Keeter (1996),
and building on Zaller (1992), political knowledge is beginning to replace cognitive
structure as the preferred measurement.20 This, in no small part, is related to the success
of increasing levels of political knowledge as predictors of cognitive organization. This
research, however, does not have to be an either or proposition. Apart ﬁ‘om levels of
political information there is still much to be learned ﬁ'om the cognitive structure of
respondents. This is not to ignore the argument that political knowledge is the single best
indicator of political sophistication. However, seeking to measure changes in evangelical
thinking may require that the categories be more nuanced than the categories of The
American Voter (1960), more reliable than those of The Unchanging American Vote
(1976), and broader than political knowledge.

4.2 Schema and Sophistication

An additional effort to measure the organization of sophistication/political

thinking from political psychology involves how people organize their thoughts into

enduring broad categories or schema. These schemata are persistently available and

 

2° See also McGraw and Pinney; Smith 1989; Mondak 1999, 2001; Price 1999; Luskin 2002)

52

represent “long-term individual differences in perceptual readiness” (Bargh and Pratto
1986). Building upon the Information Processing Theory, researchers have concluded
that “one of the indisputable ﬁndings on long term memory is that information is not
stored in a random fashion but is organized meaningfully in packets of semantic
associations.” (Lodge and McGraw in Lodge et al. 1991, 1358).

Schema build on Converse’s view of constraint as cognitive organization (Jackson
and Marcus 1975; Fiske and Kinder 1981). These categories or schemata provide an
overarching structure or mental scaffolding through which new information is given
attention, processed and stored. This information processing depends on the ready access
of the structures for evaluation of new information. Schema, in effect, serve as mental
shortcuts for organizing the world around oneself (Lau, et al. 1991). For citizens these
shortcuts provide a “formula for comprehending patterns of political events” (Hamill et
al. 1985, 853).

Schemata can be determined by questioning, which reveals the structure in use
through the persistent or chronic accessing of similar responses. The accessibility of a
schema reveals the organizing principle(s) around which the individual is organizing their
political beliefs. Such main beliefs may take the form of party or issue categories
including economic concerns, or a more value driven schema like individuality, liberty or
equality (Conover and Feldman 1984). Some persons in the political realm are
aschematic while others utilize multiple schemata.

The accessibility of a schema also refers to the order in which a schema is
checked in response to information or a request for information. Highly accessible

schema are checked before less accessible ones and, if an appropriate ﬁt is found for the

53

information or request, the less accessible schema are not used (Axelrod 1973). The
accessibility of schemata also ﬁts with the general theory of attitude accessibility. As
Miller and Peterson write, “accessibility is the most ubiquitous measurement of attitude
strength in political science” (2004, 853). When asked to express a political attitude,
people quickly search their memory for the most accessible considerations, whose
accessibility is increased by the recency of use (Feldman 1995; Zaller 1992; Zaller and
Feldman 1992). The same principle applies to schema as it also depends on frequency
and the recency of prior activation (Fazio 1989).The accessibility of schemata are also
inﬂuenced by priming and saliency (Axelrod 1973; Iyengar and Kinder 1987; Lau 1986,
1989). Cues in the respondents’ environment and priming from media sources increase
the ease of retrieval of political attitudes and political information.

This accessibility does not preclude some citizens from using multiple schema
and readily switching from one schema to another with identical levels of access in the
hierarchy. A political schema is also not a necessity. Some people are aschematic and
demonstrate no organizing principle to their political thinking.

Schemata are viewed as being stable and resistant to change. The economy that
schemata provide make individuals less likely to change their schemata formulation in
response to limited disparate information. When change does occur, it tends to be slow
and incremental. Change can occur in terms of rejection or modiﬁcation of previously

formatted schema or in the restructuring of the hierarchy.

Mayer and Bower (1986) show that schemata can form from experience. Fiske
and Taylor (1991) also reveal that individuals can develop a schema “from abstract

communications of the schema’s general characteristics” (p 147). This suggests that,

54

apart ﬁom political knowledge, individuals lacking a schema for a particular political

concept will use their own experience or environment to develop one.

4.3 Criticisms of Schemata and the Resulting Chronicity Measurement

Not all political scientists view schemata as a positive development. Kuklinski,
Luskin and Bolland (1991) argue that schemata proponents have not demonstrated the
interrelationship between the components of schemata. They argue that proponents have
not produced experimental evidence for the connections between mental operations and
affect, that they have used overly expansive and ill deﬁned concepts, and they merely
have adopted a new name for old practices. While even proponents of schemata agree
with many of the criticisms”, opponents have also affirmed the validity of schemata or
issue accessibility as a valid measurement of political sophistication. Luskin ﬁnds Hamill
and Lodge(1986) and Hamill, Lodge and Blake’s (1985) schematic analysis of
sophistication similar to his own analysis. He concludes that their schema usage is
“almost a doppelganger” for segments of his own analysis and states that “the
development of different political schemata should be highly intercorrelated and highly

correlated with sophistication as a whole” (Luskin 1990, 353).

Lau’s work on schemata is especially appealing to this dissertation’s focus on
political sophistication. It has the primary advantage of providing a measurement of
accessibility. It also provides a ﬂexible nonhierarchical set of categories for evaluation. It

has the advantage of proven reliability and validity (Lau 1986, 1989; Erber and Lau

 

2] While acknowledging the validity of various critiques of schemata, Lau and Redlawski (2001) argue
that using party and ideology as schema has an established validity.

55

1990), and the added advantage of being redeﬁned as a chronicity or a measurement of
the chronic accessibility of each schema by the respondents. Lau’s chronicity
measurement does not propose to reveal the schematic structure of respondents, but
instead reveals how individuals consistently think (and chronically access) in terms of

one or more of the categories in the schemata.22

4.4 Participation, Sophistication, Information and Reciprocal Relationships
Looking at changes in sophistication necessitates looking at the reciprocal
relationship between sophistication, political knowledge and political participation.
Campbell et al. (1960) theorizes that political involvement and education will increase
levels of conceptualization and knowledge, though others ﬁnd a weaker than expected
correlation between education and conceptualization (Hagner and Pierce 1982; Jacoby
1988, 1991). Neuman (1986) and Delli Carpini and Keeter (1996) both demonstrate and
support the connection between participation, political knowledge and levels of
conceptualization. Neuman acknowledges that “political sophistication is a stable and
fundamental variable with considerable inertia” but he concludes that political
sophistication can be acquired through an upward spiral process over time, involving
interest (from a supportive, stimulating and active political environment) and
propagating knowledge which in turn increases interest (128-129). Neuman also posits
that education (formal and informal) and having a higher level of income are important in
changing conceptualization as well as participation in groups (which increases salience

but not the levels of conceptualization).

 

22 Even schema critics, Kuklinski, Luskin, and Bolland accept that chronic accessibility is “worth
distinguishing as a separate target” (1351).

56

Delli Carpini and Keeter show that political participation and political knowledge
inﬂuence each other reciprocally (See also Bennett 1994, 1995; Tan 1980). Delli Carpini
and Keeter (1996) conclude that when one’s environment encourages political
engagement, those individuals are most likely to search for information. Berggren (2001)
also ﬁnds that individuals who are limited in their SES and educational resources beneﬁt
disproportionably from free information provided by political groups, and that the
relationship between SES and sophistication weakens in that setting.

Bennett (1975) argues that the type of political activity exhibited is crucial to
determining shifts in political cognitive complexity. He theorizes that the activity will
promote change when: 1) it places the participant in direct conﬂict with the beliefs of
others, 2) when the stakes in the conﬂict were tangible beneﬁts, and 3) when the activity
required the participant’s individual initiative and autonomous input. Bennett argues that
voting is the political activity least likely to bring one into direct conﬂict with the beliefs
of others, thus not initiating change. Leighley’s (1991) examination, based on Bennett’s
assertions, conﬁrms that shifts in conceptualization resulted in shifts in participation and
vice versa.

The conclusion is that regardless of the type of measurement used for political
sophistication, a reciprocal relationship exists between participation and political

thinking. To accurately account for one will require controlling for the other.

57

 

CHAPTER V
POLITICAL PARTICIPATION AND ISSUE ACCESSIBILITY AMONG
EVANGELICALS
Has an emphasis on increased political involvement changed how Evangelicals

organize their political thinking? If Evangelicals have increased their political
participation beginning with cue changes at the beginning of the 19805, did this initiate
the theorized “spiral” relationship between participation and levels of political
sophistication? This chapter examines rates of Evangelical participation, and Evangelical
changes in political issue accessibility over time. Finally, this chapter evaluates the
reciprocal relationship between participation, political information and political
sophistication (as measured by accessibility) as a consequence of increases in Evangelical
political participation. It also provides an examination of the “culture war” motif as a
factor in the shifts in Evangelical issue accessibility. It begins with a brief discussion
tying together the theoretical background for the study by drawing ﬁ'om examinations of
Evangelical political participation, measurements of political sophistication, and the

intertwined effects of political participation and political sophistication.

5.1 Theoretical Background

Several key questions regarding Evangelicals are 1) over time, by what process is
stable political conceptualization changing for the Evangelical population, and 2) how
should those changes be measured?, Education, changes in participation, participation
producing conﬂict over beliefs, and cues from one’s environment have all been theorized
or shown to affect political sophistication, regardless of the deﬁnition of political

sophistication (See Chapter 3). Neuman’s (1986) and Delli Carpini and Keeter’s (1996)

58

work have demonstrated the connection between participation, political knowledge and
levels of conceptualization. Also, a similar relationship has been found among education
and income and levels of conceptualization (Campbell et al 1960; Hagner and Pierce
1982; Jacoby 1988, 1991). As previously mentioned, Leighley’s (1991) examination of
changes in conceptualization results in shifts in participation and vice versa. However,
Leighley’s study is limited to the differentiation and integration of The American Voter’s
four levels of conceptualization. Though building on Bennett’s (1976) arguments on the
criteria needed for cognitive change, she never explicitly tests for a reciprocal
relationship for Bennett’s assertion that voting would be the political activity least likely
to change political conceptualization.

There is little evidence to suggest that Evangelicals have undergone a major shift in
education levels or an increase in income over the past twenty-ﬁve years.l This leaves the
process of changing Evangelical political sophistication to the creation of a supportive
political environment within the Evangelical milieu, thus changing patterns of
participation and affecting concepts. As earlier examined, internal cue shifts changed
from nonparticipation to participation, because Christian Right political and social
organizations provided information, training and encouragement for participation. This
would theoretically trigger the reciprocal relationship of increased political participation,
increased political information and shifts in political sophistication.

For the purposes of this paper, shifts in sophistication will be measured using a
chronicity structure. The most venerable measure of political cognition from The

American Voter would require using levels of conceptualization that have fallen into

 

I One of the distinctions of Evangelicals is that their rates of college attendance and attaining a high school
diploma are lower than the general population (Ammerrnan, 1982; Kellstedt and Smidt, 1991).

59

disuse in political science. The levels are also limited essentially to three types of
response content and one lack of content, which could severely limit the ability to
measure change. Particularly if the expected changes in political conceptualization do not
ﬁt or are buried within the categories of the levels of conceptualization, the levels of
conceptualization would be ineffective. Lau (1986) also points out that the levels of
conceptualization are ordered from highest to lowest in terms of sophistication while
measurements of accessibility do try to make value judgments between categories or
assume a highest to lowest order. Though schemata usage has waned in political science,
Lau’s chronicity has been shown to be a valid and reliable means of showing the chronic
accessibility of political constructs (Erber and Lau 1990; DeSart 1995; Lau 1986, 1989).
An expansion of this chronicity can provide the means for measuring political
sophistication for both Evangelicals and non-evangelicals.

While Lau initially refers to chronicities as “schemata” in his earlier publications, for
the sake of continuity with Lau’s later work, I use the term chronicity to refer to his
political construct. Lau argues that individual political chronicities “facilitate the
processing, storage and later recall” of information about politicians and political events
(1986, 196). To measure these chronicities, he employs a “cue-free recall” using the
sixteen open-ended likes-and-dislikes questions from the American National Election
Studies (ANES) (1989). The responses given to these open-ended questions partially
represent an underlying cognitive structure or chronicity. Lau admits that they also
partially represent recent random exposure to political stimulus such as a news program
or conversation with a ﬁend. However, he does perform a number of tests of consistency

and reliability to demonstrate the validity of the underlying cognitive structure (discussed

6O

below). Lau categorizes the responses into four chronicities: issues, group relations, party
identiﬁcation and candidate personal factors, although he also suggests that several of the
chronicities could be further divided into smaller subgroups. By weighting the responses
in the order that they are recalled, he shows which chronicity is most accessible and
“therefore the most indicative of [chronicity] content” for each respondent (1986, 101).2

Lau measures these chronicities using ANES panel study data from election years
1956—1960 and also 1972-1976. To assess the stability of the chronicities, he uses a
LISREL analysis to evaluate the stability correlations of the later election year’s
chronicities being determined by the earlier year’s chronicities. The positive and highly
statistically signiﬁcant scores led Lau to conclude that the chronicities are reasonably
stable over time.

Lau then constructs ten validity criteria including knowledge of the candidate,
opinions on foreign and domestic policy issues, strength of party ID and reactions of the
Kennedy-Nixon debates. Again, using LISREL, Lau correlates these validity criteria with
the sixteen scores (four for each chronicity) generated in the measurement of the
chronicity (see footnote 1). All of the validity paths were signiﬁcant (p < .01). Lau

demonstrates that the chronicities are generally stable using two different panel studies,

 

2 Lau illustrates this process: “Consider a respondent in 1960 who says she likes Kennedy because (1) he is
a Democrat, (2) he proposed tax cuts and (3) he wants to help the poor. She dislikes Kennedy, on the other
hand, because (1) he is Catholic and she doesn’t want him to take orders from the Pope. The ﬁrst “like”
response falls into the party category, the second like falls into the issues category and the third like and
single “dislike” are both group responses. The ﬁrst like and the single dislike responses each receive a
weight of 3, the second like receives a weight of 2 and the third like receives a weight of 1. This respondent
would then have a score of .333 [3/(6+3 )] for the party category, .222 [2/(6+3 )] for the issues category, .444
for the groups category [(1+3)/(6+3)] and 0 for the candidate category [O/(6+3 )]. Comparable measures
would be constructed from the like and dislike questions about Nixon, the Democratic party and the
Republican Party. Hence every respondent receives 16 such scores, one for each of the four political
information categories for each of the four attitude object [chronicities]” (Lau 1989, 10). The actual method
of scoring as adapted from Lau is found on pages 70—7 1.

61

and four different presidential candidates. This stability reveals strong substantiation for
the chronicities existing as an underlying structure (1989).

Lau also makes a comparison of the chronicity scheme with the levels of
conceptualization from The American Voter (Campbell et al, 1960). He argues that the
chronicity “capture[s] the same sense as levels of conceptualization or belief systems”
(Lau 1986, 112). The “sameness” for both belief systems and Lau’s chronicity is
represented in a similar management of new information which provides cognitive
economy for making sense of the political world. As a result for this study, political
sophistication will be measured using a chronicity measure which reveals the primary

issue accessed by each respondent.

5.2 Hypotheses

Evangelical interest and social movement groups in the 19803 and 19905 pushed to
the forefront a collection of primarily social issues. The issues being addressed included
opposition to abortion, outcomes-based education, and sex education; defense of
traditional family values accompanied by a focus on traditional women’s roles; support
for decency in media; and rejection of gay lifestyles and homosexual civil rights. The
19903 saw a broadening of this agenda particularly by the Christian Coalition, to include
health care and economic issues such as tax policy. During this time period Evangelicals
expressed views that were distinctive from the general public on political issues including
affirmative action, defense spending, environmental issues, abortion, premarital sex,
extra-marital sex, homosexuality and feminism (Curry-Roper 1990; Guth et al 1996;

Wilcox 1987, Smidt et a1 2003).

62

This chorus of social concerns is also reinforced in the early 1990s by the
emergence of the “Culture War” theme. This motif has become common in Evangelical
circles and was the organizing theme of the 1992 Republican National Convention.
Hunter’s culture war term is thought to have been the inspiration for Patrick Buchanan’s
similarly themed convention address. That same election cycle is viewed as having ﬁrmly
established Evangelicals as a part of the Republican coalition (Kellstedt et a] 1994).
While the culture is a broadly quoted concept positing a sharp conﬂict between
traditionalists and progressives, there exists little empirical evidence for this division.
Increased polarization has not been found within the American public or by Evangelicals
or Seculars as a group. However, using a chronicity measurement and dividing religious
traditions into those accessing and not accessing social issues provides an alternative
means of measuring the culture war. The war may not be fought by broad religious/social
groups but by segments within those groups for whom social issues have a heightened
primacy.

Existing chronicities are brought to the forefront by salient elements in the
respondent’s environment. Previous research has shown that the candidate like-and-
dislike questions in the AN ES contribute to the emergence of the candidate chronicity but
that this emergence is also stimulated by more distant elements such as coverage of the
candidates by the news media (Miller et al. 1986). It raises the issue as to what would
activate or reinforce the social issues chronicity? The increased information and new set
of cues being released into the Evangelical world increase the possibility of a
restructuring of the accessibility of political chronicity for some Evangelicals. This

occurs as the content of the messages underlying these new cues is primarily concerned

63

with social issues. An increased emphasis on economic issues, as facilitated by the
Christian Coalition and adoption of Republican economic views also leave open the
possibility that a reciprocal relationship may exist between accessibility of economic
issues and political participation.

The trend of salience of social issues would not be limited to merely Evangelicals.
Increased media attention to crime, racial tensions, gay rights and abortion protests across
the 19805 and 19905 also heightens the likelihood of social issues’ accessibility being
reinforced across the entire population. As social issues are diffused by increased media
perhaps as a result of Evangelical political activity, by activities from counter-interest
groups such as People for the American Way and by the adoption of these issues by
political parties, they are no longer just the domain of Evangelicals. However,
Evangelical and Non-Evangelical issue accessibility is being stimulated by other actors in
the media/political realm. However, if Evangelicals are responding to heightened cues
distinct to their congregations and are recipients of messages unique to their Christian
media sources, they should show increasing saliency for social issues at a rate higher than
the Non-Evangelical population.

The following hypotheses and their corollaries are suggested by this analysis,
previous research and common sense:

H1. Increased emphasis on social issues in the American political arena will result in a
similar increase in accessibility of social issues for the entire population.

C1. Evangelical social issue access will increase at a higher rate than social issue

access for the Non-Evangelical population as Evangelical interest groups,

churches and social movements increase the salience for social issues. Access of
economic issues for the same reasons will also increase more for Evangelicals.

64

C2. In keeping with the theorized culture war, Seculars will also show a similar
pattern of increased social issue access above and beyond that of the Non-
Evangelical population.
H2. Also in keeping with the culture war theory, Evangelicals and Seculars primarily
accessing social issues will hold extreme positions on social issues when compared to the
rest of their respective religious categories.
H3. The explanations for social issue usage will vary between Evangelicals and other
religious groups as religious participation and religious orthodoxy affect Evangelical
access while political information and education will have an across-the-board inﬂuence
for all religious groups.
H4. A positive reciprocal relation will exist for Evangelicals for political participation,
political information and their usage of the social issue chronicity. The same reciprocal
relationship will also exist for political participation, political information and their
usage of the economy chronicity.
H5. In keeping with previous participation research, political participation will show a
reciprocal relationship with issue access while voting will not.
5.3 Data
Data for this study are taken from the 1976—2000 American National Elections
Study conducted by the Center for Political Studies at the University of Michigan (Sapiro
et al. 2002). As the primary focus of the accessibility measurement used in this study is
the open-ended questions used in Presidential election years, the data for off-year

congressional elections are excluded from the study. The cumulative ﬁle yields 13,909

respondents for the years in question and is treated as a pooled cross-sectional analysis.

65

Figure 5.1. American Religious Traditions over Time.

Percentage Religious Traditions

 

 

m 'l -~
~ _-—___
~~~—‘--— \
./‘\_ \ ”nu-ﬂ —'
' “ /'/ .M.
\
o d"
N \\-.'-v- “.‘ .
’o-o-o_.—.-o—I-’K
./'
./
./
o u'.’o’._.—.—U—I-O—O:;.-‘J-/- ..................
ﬂ 1 "’ ...................................................
———-'—.~__ _'___.
o "

 

 

1976 1980 1984 1988 1992 1996 2000

Election Year

 

 

Evangelical ----- Mainline Protestant
----------- Black Protestant — - - - - Roman Catholic
-— - — Jewish - ------ Secular

Other

 

 

 

Source: 1976-2000 ANES

Using multiple years of the ANES provides its own set of difficulties as others have

examined and debated (Abramson 1990; Miller 1990). Despite the goal of providing

over-time comparisons covering the 24 years under consideration, changes in AN ES

question wording and usage present several serious obstacles to pooling the data for this

study. In particular, the AN ES dropped the Bible attitudes question from the 1976 study

and changed the format of the question from four possible responses to three beginning in

1992. The 1976-1988 Bible question was recoded to provide compatibility with the later

question format The Two-Stage Auxiliary Instrumental Variables (ZSAIV) method of

“estimating across datasets” was used to extract the missing Bible question from the 1980

66

data set for use with the 1976 respondents.3 The NBS also changed the wording and
provided a necessary expansion of denominational codes in 1990 necessitating the
matching of the two differing lists to provide continuity (See Appendix A for coding).4
Figure 5.1 shows the breakdown of religious groupings found in the ANES, 1976-
2000. Evangelicals, Black Protestants and Catholics show a stable pattern over time while
Mainline Protestants declined and Seculars increased. While most of these categories are
self—evident, “Seculars” refer to those persons expressing no religious affiliation and the
“Other” category is a catchall for small religious traditions such as Eastern Orthodox,
non-traditional conservative denominations such as Jehovah Witnesses and non-

traditional liberal denominations such as Christian Scientists.

5.4 Evangelical Political Participation

As an entry point for this examination, I evaluate the historical patterns of
participation for Evangelicals as compared to other religious groups. These patterns are
evaluated using voter response and participation on a six-point campaign-participation

scale. Respondents received a score from 1 (no participation) to 6 (participation in all

 

3 ZSAIV extracts a variable ﬁ'om an auxiliary data set and regresses the variable using a set of explanatory
variables in the auxiliary dataset. The parameter estimated from this ﬁrst stage estimation is then used with
the same explanatory variables in the primary data set to create an “instrumental variable”. The IV is used
as a proxy for the missing variable. A key assumption of this method is that both the auxiliary and primary
dataset are sampled from the same population. The auxiliary model for the missing bible variable is: Bib1e=
.849 + religious tradition *(.098) + church attendance‘(.003) + education‘(.045) + race*(-.215) + south*(-
.053) + abortion’(.197) + campaign interest*(.038) + women’s equality*(-.049) + age*(.003) + politics
complicated“(.032). All included variables were signiﬁcant at .05 level or higher.

The inclusion of more speciﬁc denominations in 1990 allows for more clarity in deciding if a respondent
belongs in the Evangelical or Mainline categories. This may be reﬂected in a decrease in Mainline
percentages in 1992 and a slight increase in the Evangelical category. However, when using a similar
coding scheme in the General Social Survey (which did not change its denominational wording or coding
in 1992) and comparing years, 1984-2000, the Mainline category for the NES and GSS correlate at .91 (p <
.001). A similar concern over inﬂation of the Secular category beginning in 1992 with the use of new ﬁlter
questions for denominational affiliation arrives at a similar conclusion. Both the NES and GSS show a rise
in the percentages of Seculars from 1984-2000 (NES: 11.2%-17.1%, GSS: 9.9°/o-17.6%) and the two are
correlated at .78 (p < .01).

67

activities) for inﬂuencing others votes, attending political meetings, working for a party
or candidate, displaying a button/sticker or making a campaign contribution. Figure 5.2

shows the initial inﬂuence of separatism and the effect of changes in cues.

Figure 5.2. Evangelicals and Rates of Political Participation, 1976—2000.

0_
V W

30

Participation Percentage
10 20
1 J

1 /

O—r

I I I l l l
1976 1980 1984 1988 1992 1996 2000

 

 

 

 

 

Election Year
Inﬂuence Other's Vote - - - - Attend Political Meetings
---------- Work for Party/Candidate — - —' Display Button/Sticker
— -- - Donate Money

 

 

Source: 1976-2000 ANES

A difference of means test (Table 5.1) for political participation between
Evangelicals, Seculars and the remaining Non-Evangelicals shows that Evangelicals were
signiﬁcantly lower in their levels of participation in 1976, 1984 and 1988. There were no
signiﬁcant differences for the 1980, 1992, 1996 and 2000 elections. The changes in
participation in 1980 and 1992 may point to two watershed years. The 1980 election was
the ﬁrst election to combine both a “born again” candidate in incumbent Jimmy Carter (to

whom Evangelicals delivered a majority of their votes) and the inﬂuence of the ﬂedgling

68

Moral Majority. Evangelicals appeared to have responded to the single-year dynamics of
the 1980 election with increased involvement, but returned to their lower involvement in
the 1984 election. The 1992 election with the GOP’s emphasis on culture war solidiﬁed
new patterns of Evangelical participation, as Evangelicals matched Non-Evangelicals in
their political involvement in the two subsequent elections.

Table 5.1. Comparing Evangelical Participation with Seculars and General
Population in Political Participation: Difference of Means.l

 

 

Year Mean Standard N AN OVA F
Deviation
1976 7.79***
Evangelicals 1 , 57 a .82 544
Seculars 1,73 1.06 215
General Population 1.76 1.02 1643
1980 1.48
Evangelicals 1 .59 .85 292
Seculars 1.52 .82 163
General Population 1.65 .97 953
1984 3.98*
Evangelicals 1,50 3 .78 413
Seculars 1.65 1.00 210
General Population 1.64 .98 1319
1988 5.63"
Evangelicals 1,44 3 .76 386
Seculars 1, 51 .87 196
General Population 1.62 .99 1193
1992 .50
Evangelicals 1 .70 .93 526
Seculars 1 .63 .88 400
General Population 1.67 1.02 1330
1996 329*
Evangelicals 1.55 .75 171
Seculars 137 b .66 128
General Population 1. 59 .96 447
2000 2.11
Evangelicals 1 .66 .88 344
Seculars 1.53 .83 281
General Population 1.65 .94 930

 

Source: 1976~2000 ANES
‘ p < .05; "p < .01, "*p < .001

a Evangelical mean signiﬁcantly different from General Population, p < .05.
b Secular mean signiﬁcantly different from General Population, p < .05

69

' ANOVA analysis between the ﬁve largest religious groups, Evangelicals, Seculars, Mainline Protestants,
Black Protestants, and Roman Catholics produced similar results except that no signiﬁcant difference was
shown between Evangelicals and Black Protestants (with the exclusion of Evangelicals being signiﬁcantly
higher in 1992). Roman Catholics and Mainline Protestants showed the same signiﬁcant differences with
Evangelicals as the General Population.

Voter participation does not reveal an increase for Evangelicals in the same time
period. The bias created by over-reporting (Abramson and Aldrich 1982; Anderson and
Silver 1986; Burden 2000 and Teixeira 1992) is not an obstacle to an obstacle to
comparing groups.5 Despite the bias, over reported voting has been found to “closely
parallel” (Abramson and Aldrich, 503) ofﬁcial turnout statistics (also see Teixeira 1987;
for an opposing view see Burden 2000). Voter turnout was measured using nonvalidated
responses for 1976—2000. As Figure 5.3 shows, Evangelicals, Seculars and Black
Protestants remain below other traditions in their self-reported voting. However,
beginning in 1992, Evangelicals appear to have surpassed or matched voter turnout for

Seculars and Blacks.

 

5 For example Anderson and Silver found that despite the tendency to overreport voting there was no
signiﬁcant difference in rates of overreporting between Blacks and whites.

70

Figure 5.3. Religious Traditions and Vote Participation, 1976—2000.

 

8
a” ..r
'5 /\
g °~ / \
l: f‘“~~ \ /
g 00 .._____——-—.7"'_\-:———” ./ \,.V,—-
8)
O
S rx"
G
“L3
O
o '1
m \D

 

 

 

1976 1980 1984 1988 1992 1996 2000

Election Year

 

Evangelical ----- Mainline Protestant
----------- Black Protestant -— - ~— - - Roman Catholic
-- -- —- Jewish - ------ Secular

Other

 

 

 

Source: 1976—2000 ANES

5.5 Chronicity

To measure accessibility of political constructs, Lau’s formulation of political
chronicities was used with one important exception. The issues category was expanded to
include three areas: Economy, Social and Foreign issues. Though not measuring them
himself, Lau (1989) suggests that the issue category could be expanded into categories of
foreign policy, economy, racial and social issues.6 As Evangelical information and cues
have centered on social issues, a broad issue category would fail to capture the expected

nuances of change centering on social concerns. These chronicities are based on the

 

6An initial coding added a fourth issue category, Racial Issues. However, this chronicity failed to gain over
1% of the respondents in most of the election years in question and was dropped. Racial responses were
then included in the Social Issue category.

71

candidate and party open-ended comments from the AN ES presidential election year

studies from 1976—2000. The chronicities as adapted use a four-step process that weights

the responses both by the order in which they are recalled and by the number of responses

in order to control for verboseness:

l.

The open-ended responses using the AN ES master codes were categorized
according to one of the six types of political information: Candidates, Parties,
Groups, Economy Issues, Social Issues and Foreign Policy. Depending on the
year of the ANES study, each respondent could volunteer up to three (in 1972) or
ﬁve (all other years) likes and dislikes for the Democratic and Republican
Presidential Candidate and a similar number of likes and dislikes for each
respective political party. Following Lau’s example, only the ﬁrst three responses
were coded. The fourth and ﬁfth responses were excluded because the numbers of
responses were few, and because this also allows for the ﬁrture addition of the
1972 AN ES which only asked for three responses. Appendix B contains the
topical categories and a listing of the master codes used within each chronicity.
Each response is weighted in the reverse order of which it was recalled. A ﬁrst
response would receive a 3, a second response would be weighted 2 and a third
response was given a 1.

The sum of the weighted responses for each set of like and dislike questions is
divided by the weighted scores of the total number of answers given to the set of
likes and dislikes. This produces 24 scores between one and zero, six scores for

each chronicity drawn ﬁ'om each set of like-dislike questions.

72

4. The six matching chronicity scores were then added together to produce a single
score ranging from zero to six for each chronicity (e.g., all economic issue scores
from the Democratic Presidential Candidate likes and dislikes, the Republican
Presidential Candidate likes and dislikes and each respective party’s likes and
dislikes would be added for a single economic score). Each respondent was then
coded according to his/her highest chronicity category.

The following example illustrates this scoring procedure. A respondent in 1984 likes
Mondale because (1) he is experienced, (2) he has a progressive view of public morality,
and (3) he supports labor unions. The respondent dislikes Mondale because (1) he is too
well educated. The ﬁrst “like” falls into the candidate category, the second is found in the
social issue category and the third is in the group category. The ﬁrst “dislike” falls into
the candidate category. The ﬁrst like and dislike responses receive a weight of 3, the
second responses receive a weight of 2 and the third responses a weight of 1. Each
category total (candidates=6, social issues=2 and groups=l) is then divided by the total
weighting of all responses to control for verboseness. In this case that total would be 9 [3
+ 2 + 1 (from the likes) + 3 (from the dislikes)]. The respondent would then have a score
of .667 [6/(3+2+1+3)] for the candidate category, .222 [2/(3+2+1+3)] for the social issue
category and .111 [1/(3+2+1+3)] for the group category. The other three categories
(Foreign Policy, Economy Issues and Parties) are scored a zero. This results in six
chronicity scores (one for each category) derived from the Mondale likes and dislikes.
Similar measurements are computed for the likes and dislikes about Reagan, the
Democratic party and the Republican Party. This results in each respondent having 24

scores, six for each chronicity category from both of the candidates and both parties. The

73

ﬁnal step is to add the six scores for each chronicity category and to classify the

respondents based on their highest score.

5.6 Measuring the Chronicity

The chronicities or the categories of issue accessibility show a wide variation in
size. The Candidate chronicity is used at a rate over ten percentage points (27.5% -
16.9%) higher than the next largest chronicity, Economy. This is in keeping with
previous work showing that candidate question format and media coverage explain why
the Candidate chronicity is the largest category of response for these questions.
Nonschematic or No Content respondents who give no responses to any of the eight
open-ended questions make up nearly one in every ten citizens. This statistic is matched
by multiple chronicity individuals who make up 10.3% of the electorate7 For this last
group, two or more of the chronicity scores tied for the most frequently accessed
category.

Figure 5 .4 shows the stability or lack thereof for each chronicity over time for
Evangelicals, Seculars and Non-Evangelicals. For the sake of simplicity, graphic
representations of the chronicities exclude the No Content and Multiple chronicity
respondents.8 The Candidate chronicity shows the most volatility while Economy,
Groups, Parties and No Content are fairly consistent as a percentage of the population.
For the combined population, three chronicities exhibit a linear pattern over time. Usage
of the Foreign Policy chronicity is declining over time ([3 = -.34, p<.05, R2 = .52;). The

Social Issues chronicity shows consistent linear growth (13 = .37, p<.000, R2 = .95) for the

 

7 Analysis of the multiple schemata groups revealed no pattern to the multiple schemata accessed. See
Appendix C for the complete analysis.

' The No Content chronicity is stable over time at approximately 10% of the population while the Multiple
Chronicity shows a decline over time (Multiple Chronicity, B = -.07, p<.05, R2 = .59).

74

entire population. The Social Issues chronicity line matches the expectation of hypothesis
one that the access of social issues as the primary principle for organizing one’s political
world has increased.

Figure 5.4. Combined Tables of Evangelical, Secular and Non-Evangelical
Chronicities over Time: 1976—2000.

 

O_
v
Q)
5%”
5 5%“
8
0)
D-a
O
3.3.“ N“
.0
E
s e-
O

 

 

1976 1980 1984 1988 1992 1996 2000

Election Year

 

 

 

Candidates ----- Economic Issues
----------- Parties — - — ' - Groups
— —- - Social Issues - ------ Foreign Policy

 

Source: 1976—2000 ANES

Next, I turn to comparing these chronicities between Evangelicals, Seculars and

remaining Non-Evangelicals. I use Taylor’s (1980) four theories for evaluating trends

between groups to search for signiﬁcant differences these three religious traditions. 9

 

9 To evaluate trends between groups over time there must be underlying theories against which the
trends can be compared. D. Garth Taylor argues that “an array of possible theories can be used to describe a
set of data and the ultimate theoretical conclusion depends on the most parsimonious statistical model
which can be used to describe the data” (1980 87). Taylor suggests four possible theories with seven
possible outcomes which can be used for comparing two groups.

75

5.7 Evidence of a (Limited) Culture War

An analysis of the accessibility or Chronicity trends between Evangelicals and
Non-Evangelicals found the No Difference theory ﬁts most categories. [See Appendix D
discussion of Taylor’s trend analysis and for complete analysis of all trend lines] There
was no difference (analysis not shown) between the two groups on Candidates, Economy,
Foreign Policy and No Content. There was, however, a constant difference (analysis not
shown) over time for the Multiple and Group chronicities with Evangelicals averaging
1.7 percentage points more per election year in the Multiple category and 2.2 percentage

points less in the Group chronicity.

 

Theory I : No Change. This theory predicts that the changes observed between survey years are
merely random ﬂuctuations and that the percentage of change in each year is actually zero.

Theory 2: Enduring Drﬂerences. This theory would predict that the differences in
conceptualization are constant. While the difference between the groups is nonzero the differences are
enduring and the correlations between the groups remain the same from survey year to survey year.

Theory 3: Linear Patterns. This theory predicts that the changes in chronicity usage between
groups are occurring with a high degree of regularity. The end result is that the differences can be best
modeled by a linear pattern which shows an increasing or decreasing difference.

Theory 4: Social Catastrophic. This theory predicts changes in conceptualization that are erratic
perhaps as the result of sudden occurrences in the political world. This last theory in effect becomes a
catchall for data, while not specifying what model type might ﬁt the data.

76

Table 5.2. Trend Analysis: Evangelicals, Seculars and Non-Evangelicals’ Use of
Social Issue Chronicity, 1976-2000.

 

 

 

15 Social Issues Trends

03 /
219 ' I”
33
0
o
m
i”
:5.-
i
a.

0

 

 

1976 1980 1984 1988 1992 1996 2000
Electron Year

 

_ Evangelical """ Non-Evangelical
"""""" Secular

 

 

 

 

 

Trend Line: Evangelical Versus Non-Evangelical

 

 

 

Category diference
(Base=Non-
Evangelicals) Hypothesis Model X2 df p Decision
Evangelicals 1) no d=0 16.29 6 <.05 reject
difference
2) constant d= -1.86 15.04 5 <.05 reject
difference
3) linear d= -.69 + -.289 13.62 4 <.05 reject
4) linear 1.42 1 >.05 signiﬁcant
component
Trend Line: Evangelical Versus Secular
Category dﬁrence
(Base =Secular) Hypothesis Model X2 df p Decision
Evangelicals 1) no d=0 7.61 7 >.05 accept

difference

77

The comparison of Evangelicals with the Non-Evangelical population (excluding
Seculars) on social issues provides support for the corollary that Evangelicals would
increase their accessibility of social issues over that of Non-Evangelicals. Table 5.2
shows that the theories of “no difference” and “constant difference” between the two
groups are rejected. The linear theory is also rejected, but the analysis of the signiﬁcance
of the linear component reveals that the trend between Evangelical and Non-Evangelicals
contains a substantial linear factor or a linear segment. Social issues moved to the
forefront of both group’s political thought. However, during a segment of the time period,
Evangelicals increased their usage at a linear rate that exceeded the usage by Non-

Evangelicals.

5.8 Extremes in Attitudes

Both Hunter and Wuthnow agree that the culture war cuts across religious
traditions with each tradition experiencing conﬂict between internal factions of
conservatives and liberals. Wuthnow, also, sees the divide between two speciﬁc groups at
the extremes of the religious spectrum: Evangelical Christians and Secularists with
religious liberals. As these two groups are more homogeneous in their beliefs,
Evangelicals and Seculars experience far less internal division on their cultural attitudes,
and are therefore more reliably conservative (Evangelicals) and liberal (Seculars).

In keeping with Wuthnow’s view of Mainline Protestants, Black Protestant, and
Catholic social issue respondents will also be divided between conservative and liberal
views. In the aggregate these traditions would appear more moderate as the internal

extremes cancel each other out. Respondents giving primacy to Social Issues in these

78

traditions represent both conservative and liberal opinions and as a result they would be

expected to average out each other.

Figure 5.5. Religious Groups, Social Issue Religious Groups and Aid to Blacks.

 

Evangelical .......................................................... . .......................................
Social Issue Evangelical ........................................................ . ........................................
Black Protestant ................ . ...................................................

Social Issue Black Protestant .................................... . ............................................................
Mainline Protestant ......................................................... . .......................................

Catholic ...................................................... .. ..........................................
Social Issue Catholic ................................................... . .............................................
Secular _ ....................................................... . ..........................................

 

 

Source: 1976-2000 ANES

This would suggest a lack of a regular or distinct pattern between social issue and
non-social issue adherents for Mainline and Black Protestants and Roman Catholics.
There is, however, reason to believe that Mainline Protestants and Catholics accessing
Social Issues should move in a liberal direction as compared to Mainline Protestants and
Catholics who are accessing other chronicities. While Wuthnow is correct about the
conservative-liberal tensions within denominations, orthodox or conservative believers
within Mainline Protestants have historically (for the twentieth century) been a minority.

Despite recent gains (McKinney and Finke 2002; Smidt et al 2003), they still represent

79

the smaller group. Additionally Roman Catholics, even among the traditionalists, have
been more socially liberal (Kellstedt et a1 1996; Layman 2001; Welch and Leege 1991 ).
The most inconsistent pattern could then be expected from Black Protestants as they

demonstrate more conservative views on moral issues and more liberal views on social

issues.

Figure 5.6. Religious Groups, Social Issue Religious Groups and Assistance to
Blacks.

Evangelical . .

Social Issue Evangelical .
Black Protestant .......................................... . ......................................................

Social Issue Black Protestant .
Mainline Protestant .................................................................. . ..............................

 

Catholic - ..................... . ..................................
SCCUlaI' . ............................................................... .. .................................

Social Issue Secular ................ ........ .......

 

 

Source: 1984-2000 ANES

80

Figure 5.7. Religious Groups, Social Issue Religious Groups and Black
Thermometer.

Evangelical ........................................................ .. .......................................

Social Issue Evangelical ......................................................... . .......................................
Black Protestant .............................................................................. . ..................

Social Issue Black protestant ............................................................................. . ...................
Mainline Protestant ....................................................... . .........................................

 

 

CathOliC ......................................................... .. ......................................

Social Issue C atholic .............................................................. . ..................................

Secular ........................................................ . ........................................

Social Issue Secular ........................................................... ... ....................................
Unfavorable Neutral Unfavorable

Source: 1980-1988 ANES

This would suggest that on cultural issues there exists a spectrum of religious
traditions, with Evangelicals and Seculars anchoring the respective conservative and
liberal ends and other more evenly divided religious traditions existing in between. When
respondents who primarily access social issues are added to the spectrum and broken
down by tradition, social issue Evangelical and Seculars should be more extreme than the
remainder of Evangelicals and Seculars. Social issue respondents from other traditions
will fall near the rest of their respective tradition but to the liberal side.

The following analysis looks at whether those with heightened accessibility of

social issues within the Evangelical and Secular traditions also express extreme social

81

issue positions. To test the distribution, thirteen measurements of afﬁnity or political

attitudes are examined.

Figure 5.8. Religious Groups, Social Issue Religious Groups and Gays in the
Military.

Social Issue Evangelical .

Black Protestant .

Social Issue Black Protestant . ..
Mainline Protestant ......................................... . .......................................................

Social Issue Mainline Protestant .
Social Issue Catholic ............................. . ...................................................................

Secular ..................................... . ............................................................

 

 

Source: 1992—2000 ANES

82

Figure 5.9. Religious Groups, Social Issue Religious Groups and Gay Adoption.

Evangelical .

Social Issue Evangelical .

Social Issue Black Protestant .
Mainline Protestant .

Social Issue Mainline Protestant .

Catholic ............................................................... . .................................

Social Issue Catholic .

Secular ........................................................ . ........................................

 

 

Source: 1992—2000 ANES

Attitudinal responses were coded so that higher scores represented the
conservative position and lower scores represented the liberal one. Affinity to groups, as
measured by feelings thermometers, retained their normal coding with higher scores
expressing greater attraction. Figures 5.5 — 5.17 present the means for attitudes and
measures of affinity covering race, homosexuality, women’s roles, abortion, changing
o

morals and lifestyles, and the environment for the ﬁve largest religious traditions.1

AN OVA means tests matching these Figures are found in Appendix E.

 

'0 Jews and Social Issue Jews were excluded from the analysis because social issue Jew cells numbered
less than 10 for many questions.

83

Figure 5.10. Religious Groups, Social Issue Religious Groups and Women’s Roles.

Evangelical
Social Issue Evangelical
Black Protestant
Social Issue Black Protestant
Mainline Protestant
Social Issue Mainline Protestant
Catholic

Social Issue Catholic

Secular .. .. .. ..

Social Issue Secular

 

 

Equal Roles

Source: 1976—2000 ANES

l l I ﬂ

Midpoint Women in Home

Figure 5.11. Religious Groups, Social Issue Religious Groups and Women’s Libber’s

Thermometer.

Social Issue Evangelical ........................................... . .....................................................
BlaCk Protestant ..................................................................... . .............................

---------------------------------------------------------------------------------------------

Secular

Social Issue Secular ..

 

 

Unfavorable

84

Source: 1976—2000 ANES

As expected, social issue Evangelicals are consistently the most conservative,
except for race issues. Of thirteen attitudes or thermometers (excluding race), social issue
Evangelicals have the most conservative mean on ten measurements. They hold
conservative views both toward government aid to Blacks and in their afﬁnity on the
Black thermometer, but they did not hold the most extreme conservative positions among
the groups measured. They are also virtually tied for the most conservative position (.01
difference) with other Evangelicals in their opposition to gay adoption and support of

traditional values in society.

Figure 5.12. Religious Groups, Social Issue Religious Groups and Abortionist
Thermometer.

Social Issue Evangelical .............................................................. . .................................

Black Protestant .

Social Issue Black Protestant . ..................................
Mainline Protestant .

Social Issue Mainline Protestant .
Social Issue Catholic . ..

Secular .................................... . ............................................................

 

 

Unfavorable Neutral Favorable

Source: 1976—2000 ANES

85

Figure 5.13. Religious Groups, Social Issue Religious Groups and Abortion Rights.

Evangelical
Social Issue Evangelical
Black Protestant
Social Issue Black Protestant
Mainline Protestant
Social Issue Mainline Protestant
Catholic
Social Issue Catholic
Secular

Social Issue Secular

Source: 1980-2000 ANES

 

 

.................................................................. .
...................................................... .
........................................................... .
............................................ ,
....................................... .
................................................... .
................................ .
l l l I
Always Permit Never Permit

A similar pattern held true for social issue Seculars. Excluding race

measurements, social issue Seculars hold the most liberal position on ten of thirteen

measurements. Social issue Seculars are the most supportive toward abortion rights,

abortionists and spending on the environment and environmental activists. On race social

issues, Seculars maintain a liberal position but are surpassed by different categories of

Black Protestants, Mainline Protestants and Catholics.

86

Figure 5.14. Religious Groups and Social Issue Religious Groups, and Tolerance of

Differing Moral Standards.

social Issue Evangelical ................................

Black Protestant ................................

Social Issue Black Protestant

 

 

Social Issue Mainline Protestant .
Catholic ............................................ . ....................................................
Social Issue Catholic ............................................. . ...................................................
Secular .. ................ ........
Social Issue Secular .. ................................... . ...........................................................
5- A1930 I Neutral 1 S. Disagree

Source: 1988-2000 ANES

Figure 5:15 Religious Groups and Social Issue Religious Groups, and Adjusting

Views of New Moral Behavior.

 

 

 

 

Evangelical mu.-
Social Issue Evangelical -----o---

Black Protestant ............ . ................................................

Social Issue Black Protestant ------o
Mainline Protestant .......................................................... . .......................................

Social Issue Mainline Protestant -----o

Catholic --~-o---
Social Issue Catholic ......................................................... . ........................................
Secular ..................................................... 6 .............................................
Social Issue Secular ................................................. . ..............................................
5 Agree I Neutral 1 S. Disagree

Source: 1988-2000 ANES

87

The remaining religious traditions scored as expected. Mainline Protestants and
Catholics tended to score near, but on the conservative side of, the population mean for
most measurements. Mainline Protestants and Catholics giving primary access to the
social issue chronicity were on the liberal side of the population mean. Social issue Black
Protestants have mixed responses. As compared to Black Protestants, they were more
conservative on abortion, more liberal on women’s roles, and mixed on gay rights, race

issues, and values and morals.

Figure 5.16. Religious Groups, Social Issue Religious Groups and the
Environmental Spending.

Evangelical .................................................. O ..............................................

Social Issue Evangelical .................................................... . ............................................
Black Protestant .

Social Issue Black Protestant .
Mainline Protestant .

Social Issue Mainline Protestant .............................................. . ..................................................
Catholic . ............................................. .. ..................................................
Social Issue Catholic ............................................... . .................................................

Secular ..........................................u.....................................................

SOCial Issue Secular ..................................... . ...........................................................

 

 

Increase Keep Same Decrease

Source: 1984-2000 ANES

88

Figure 5.17. Religious Groups, Social Issue Religious Groups and Environmentalist
Thermometer.

Evangelical o ..
Social Issue Evangelical .

Black Protestant .

Social Issue Black Protestant o
Mainline Protestant o .. .

Social Issue Mainline Protestant o ..
Catholic . .. ..

Social Issue Catholic ................................................................ . ...............................

Secular .

Social Issue Secular nu-..u..unuunouunuu.no."non-noun.nun-"u..."on-......"”noun-nun u

 

 

Unfavorable Neutral Favorable

Next I turn to the query of the existence of a reciprocal relationship between new
patterns of Evangelical participation and the cognitive structure of their political thinking.
I examine shifts in issue accessibility for the major religious traditions and the

relationship of the shift with campaign participation and/or voting.

5.9 Methodology: ZSCML and ZSPLS

To explore how participation affects the chronicity type used by Evangelicals and
to explore the reciprocal relationship between the two, a variety of two-stage estimation
models are used to control for endogeneity. Figure 5.18 shows a simpliﬁed model of the

expected reciprocal relationship. The two-stage conditional maximum likelihood

89

(2SCML)ll estimation is used for two-stage estimation of the political participation scale
and chronicity type. ZSCML uses a binary choice-dependent variable and a continuous
independent variable. The ﬁrst-stage reduced-form equation contains variables which are
excluded from the second-stage probit estimation. The OLS coefficients ﬁ'om the ﬁrst
stage are used to estimate the reduced-form equations for inclusion in the probit analysis.
The residuals for each reduced-form equation are also included in the second stage as
additional variables. The results provide unbiased coefﬁcient estimates and more accurate
standard errors than other two-stage probit estimators (Alvarez 1994; Alvarez and
Glasgow 2000). The variables included in the model were suggested from previous
examinations of predictors of political participation (Jennings and Niemi 1981; Leighley
1991; Verba and Nie 1972) and political knowledge (Bennett 1995; Delli Carpini and
Keeter 1991, 1993, 1996). While a scale of political knowledge would be preferable to an
observation of political knowledge, inconsistency of ANES questions makes that
impossibility for the time period in question. Therefore, the political information variable
is the reviewer’s evaluation of the respondent’s political knowledge. 12 A reduced-form
equation for political participation and its residuals was computed along with a reduced-
form equation for political information and its residuals. The ﬁrst-stage OLS estimations

for the ZSCML are in Appendix G.

 

ll Rivers and Vuong (1988) provide a technical introduction to this procedure. See Alvarez and Glasgow

(2000) for an additional description and comparison of 2SCML with two-stage probit least squares
ZSPLS).

‘2 Luskin (1987) has demonstrated that the reviewer’s evaluations are a valid substitute for a political

knowledge scale.

90

Figure 5.18. Representation of Simplified Chronicity-Participation Model.

 

 

U1

1

 

 

 

 

Chronicity I
Issue
Accessibility

 

 

 

 

Control
Variables

 

 

 

/

A

 

 

Shared
Control
Variables

 

 

[312
[321

\.

A
r

 

"2

l

 

 

 

 

 

Political
Participation
I Voting

 

 

 

Control
Variables

 

 

 

 

Two-stage probit least squares (ZSPLS) is used for the examination of voting and

chronicity. ZSPLS differs from ZSCML in that it allows for both the dependent and

endogenous independent variables to be binary. Unlike ZSCML, no error terms are

included as variables in the second stage from the ﬁrst-stage estimation. Like the

ZSCML, ZSPLS coefﬁcients are shown to be unbiased but the ZSCML produces more
accurate standard errors. A ﬁrst-stage reduced-form equation was estimated for both
voting and political information. The variables used were drawn primarily from previous

analyses by Rosenstone and Hansen (1993), and Verba and Nie. The ﬁrst-stage probit

estimations for the ZSPLS are also included in Appendix G.

The ﬁrst-stage equations suggest a possible positive reciprocal relationship

between campaign participation and the Social Issues chronicity, and campaign

91

participation and the Economy chronicity. The equations also show a negative
relationship with the No Content chronicity for Evangelicals. The same relationship is
missing for political information (with the exception of No Content) in both the ZSCML
and ZSPLS models, while voting is unaffected by chronicity type.

The independent variable for the political participation model is the afore
mentioned six-point political participation scale. The variables included in the model
were suggested from previous examinations of predictors of political participation and
conceptualization (Bennett 1975; Leighley 1991; Luskin 1990; Neuman 1981, 1986)
Demographic variables (see Appendix F for descriptions of variables) include education,
income, gender, marital status, age (derivative and squared to control for a possible
curvilinear relationship between age and chronicity type), intelligence, urbanism of
residence, urbanism of upbringing, and living in the South.

Political variables include level of campaign interest, respondent interest in who
wins the election, reduced-form political participation and reduced-form political
information. Party identiﬁcation was measured as a quadratic regression with both a
seven-point scale for party identiﬁcation and the square of the scale included. This
controls for the parabolic shape of the relationship of strength of party identiﬁcation to
the dependent variable. This controls for the likelihood that not only is the direction of
party id from Democrat to Republican correlated with the dependent variable but also
that strength of partisanship regardless of party may also be correlated with the dependent
variable. Religious variables include attitude toward the Bible, church attendance, and
dummy variables for religious traditions (excluding Other). Interaction terms in the

model include religious traditions and attendance, religious traditions and Bible attitudes,

92

religious traditions and political participation, and religious traditions and political
information. Also, interaction terms for Evangelical education levels and income levels
are included.

As a control for serial correlation, dummy variables for each election year were
included in the models. As even a “failure to answer the open-ended questions” provides
an accessibility category, all 13,909 respondents were used in the chronicity dependent
variable. However, the N for the Chronicity-Participation Models is reduced to 11,999 in
the second stage of the model by the inclusion of the residuals variable from the ﬁrst
stage of the model. Of the 13,909 respondents, 11,889 gave usable responses for voter
turnout, which were then used as a dependent variable for the Chronicity-Voting Models.
Missing data in independent variables are set to the mean in the case of continuous
variables and to the mode for dichotomous variables. A model was run for each of the
eight chronicities: Social Issues, Foreign Policy, Groups, Parties, Economy, Candidates,

Multiple, and No Content.

5.10 Results

A total of sixteen models were run, eight searching for a reciprocal relationship
between political participation and chronicity access, and eight looking for a similar
relationship between voting and chronicity access. As only three models show a
possibility of a reciprocal relationship, only those three models are presented in this
chapter. The remaining models are found in Appendix H, though they are discussed in
passing in the chapter. Models include marginal effects, which were measured by holding
other dependent variables at their mean or mode. As anticipated, the social issue

chronicity (Table 5.3) and economic chronicity (Table 5.4) demonstrated a positive

93

relationship between political participation and accessing the chronicity.

Table 5.3. Social Issue Chronicity, Participation and Information: ZSCML Results.

 

 

 

Variables in Model Social Issues S. E. Marginal
Effects!
Demographics
Education -.549** (.054) -.05
Income -.207** (.029) -.02
Gender .722" (.053) .07
Married .072 (.044) .01
Intelligence .001 (.036) -.00
Age -.012 (.007) -.00
Age Squared -.000 (.000) -.00
Urban -.057"' (.029) -.01
Where Grew Up .072 (.046) .01
Political South -.031 (.045) -.00
Political Variables
Party Identiﬁcation .335" (.052) .03
Party Identiﬁcation Squared -.042** (.006) -.03
Campaign Interest -.582" (.050) -.05
Care Who Wins -.239** (.053) -.02
Political Participation -.404 (.432) -.05
Participation Residual -.010 (.022) -.00
Political Information 1.697" (.303) .17
Information Residual .001 (.028) .00
Religion Variables
Bible Attitudes .086 (.139) .01
Church Attendance -.066 (.066) -.00
Evangelical -.618 (.574) -.O4
Mainline Protestant -.333 (.573) -.00
Catholic -.670 (.577) -.05
Black Evangelical -.606 (.662) -.04
Secular -.l97 (.547) -.01
Jew -.400 (1.027) -.03
Election Year

94

 

 

Table 5.3 (cont’d).
1980
1984
1988
1992
1996
2000

Interaction Terms

Evangelical x Education
Evangelical x Income

Evangelical x Political Participation
Mainline x Political Participation

Black Protestant x Political Participation
Catholic x Political Participation
Secular x Political Participation

Jewish x Political Participation

Evangelical x Political Information
Mainline )1 Political Information

Black Protestant x Political Information
Catholic x Political Information
Secular x Political Information

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish )1 Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

95

.418**
.648**
.853“I
.599“I
.905**
.967**

.005

-.003

.821"I

-. 170
-.681
-.040

.031
.124

-.278

.160
.549
.142
.004

-.056

.158“
.093
.050
.065
.126

-.106
-.091
-.183
-.016

.053

.101

(096)
(088)
(088)
(092)
(105)
(090)

(.081)
(.051)

(482)
(467)
(562)
(470)
(493)
(729)

(314)
(290)
(343)
(290)
(304)
(445)

(074)
(075)
(091)
(074)
(148)

(159)
(155)
(188)
(157)
(154)
(252)

.04
.06
.08
.05
.08
.09

.00
.00

.08
-.02
-.06
-.01
-.01

.02

-.03
.02
.05
.02
.01

-.Ol

.01
.00
.00
.00
.01

-.01
-.Ol
-.02
-.00
.00
.01

 

 

Table 5.3 (cont’d).

Constant -4.364** (.597)
Observations 1 1,999

x2 Statistic 730.2?"

Pseudo R2 .12

 

 

Note: * p<.10, one-tail test; "p<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-Stage Conditional Maximum
Likelihood model (Rivers and Vuong 1988).

lMarginal effects are calculated at the mean/mode of the independent variables.

Reviewing the Social Issue model ﬁrst, a one-unit increase in Evangelical
participation increases the likelihood of using the Social Issue chronicity by 8%. While
church attendance is not signiﬁcant for any other religious tradition in this model,
Evangelical church attendance increases the likelihood of accessing Social Issues. The
coefficient for Evangelical Bible attitudes is not signiﬁcant. Beliefs about the Bible are
not motivating Evangelicals to increase the usage of social issues in the organization of
their political world. In one of the surprises of this study, Bible attitudes have no effect on
any of the chronicities regardless of religious tradition. For all respondents, rising
education and income decrease the likelihood of accessing social issues ﬁrst. Also
women, persons living in rural areas, moderate party identiﬁers and independents are
more likely to access Social Issues over other categories of political thought.

Figure 5.19 shows differences in probabilities for Evangelicals, Seculars and other
religious traditions as political participation rises. While controlling for the other
variables in the model, Evangelicals and Seculars double the likelihood of accessing
social issues from the lowest level of participation to the highest. All other religious
traditions increase their probability at most by two tenths of a percent when moving from

the lowest to highest participation levels. Just as the overall trend of social issue

96

chronicity use showed Evangelicals and Seculars as distinct in their increasing usage, so
they also are unique in how political participation increases their access of social issues.
Figure 5.21 shows the inﬂuence of political information on the likelihood of using the
social issues chronicity. Again, Seculars and Evangelicals show a similar trend as they
decline in chronicity use as political information increases. For Blacks, Mainlines and

Catholics, the trends also decline and the three trend lines are virtually identical.

Figure 5.19. Increasing Political Participation and the Probability of Accessing the
Social Issue Chronicity by Religious Tradition.

 

 

 

 

 

 

Probability of Using Social Chronicity

r
O
s
N
w
4:.
m4
E
“B-

Campaign Participation

 

 

Evangelical ---- Mainline
n- ....... - Black ................. Catholic
- ------ Secular

 

 

 

Figure 5.20 reveals the inﬂuence of church attendance on the usage of the social
issue chronicity. While the effect of attendance on other Christian traditions is nearly

non—existent, rising attendance has a strong inﬂuence on the probability of accessing

97

social issues.

Figure 5.20. Increasing Religious Attendance and the Probability of Accessing the
Social Issue Chronicity by Religious Tradition.

city
2
J_

l

.15

 

 

 

 

Probability of Using Social Chroni

 

_
-I
—
—

Religious Attendance

 

 

Evangelical ----- Mainline
_ ....... - BlaCk ................. Catholic

 

 

 

A similar positive relationship exists between the Economy chronicity and
political participation (Table 5.4). Both increasing education and perceptions of
intelligence decrease the usage of this chronicity. Female respondents, rising urban
inﬂuence and middle age increase the likelihood of using it. Both campaign and political
interest variables decrease the likelihood of using the Economy chronicity. Like the
Social Issue chronicity, strong party identiﬁcation decreases its usage, while weak and
independent identiﬁcation increases it. A unit increase in Evangelical political
participation raises the probability of using the economy chronicity by 42%. Neither

Evangelical church attendance nor Bible attitudes are affecting this chronicity choice.

98

Figure 5.21. Increasing Political Information and the Probability of Accessing the
Social Issue Chronicity by Religious Tradition.

”-1

o-l

 

Probability of Using Social Chronicity

 

 

 

 

Low

—
_

2 3

Political Information

 

 

 

Evangelical
._ ....... .. Black
- ------ Secular

----- Mainline
................. Catholic

 

 

Table 5.4. Economy Chronicity, Participation and Information: ZSCML Results.

 

 

 

Variables in Model Economy S. E. Marginal
Effectsl
Demographics
Education -.7l7** (.039) -.17
Income -.23 l *"' (.021) -.05
Gender .375" (.037) .09
Married .006 (.033) .00
Intelligence -.059"I (.027) -.01
Age .016" (.005) .00
Age Squared -.000** (.000) -.00
Urban .041 * (.021) .01
Where Grew Up -.034 (.034) .01
Political South .008 (.034) .00
Political Variables
Party Identiﬁcation . 152" (.038) .03
Party Identiﬁcation Squared -.013” (.005) -.03

99

 

 

Table 5.4 (cont’d).
Campaign Interest
Care Who Wins
Political Participation
Participation Residual
Political Information
Information Residual

Religion Variables
Bible Attitudes
Church Attendance
Evangelical
Mainline Protestant
Catholic

Black Evangelical
Secular

Jew

Election Year
1980
1984
1988
1992
1996
2000

Interaction Terms
Evangelical x Education
Evangelical x Income

Evangelical x Political Participation
Mainline Protestant x Political Participation
Black Protestant )1: Political Participation
Catholic x Political Palticipation

Secular x Political Participation

Jewish x Political Participation

Evangelical x Political Information
Mainline Protestant x Political Information

4411**
4174**
-2113**

-.027

2.661"

-.014

-. l 73
-.014
-.865
-.715
.269
.743
.135
-.754

-.259**
.206"

-.077

-.298"”"

-.050

-.201**

-.123*
-.033

1.696"

.425
.714
.427
.221
.316

-.509"'
-.104

100

(035)
(038)
(358)
(016)
(232)
(020)

(113)
(052)
(462)
(451)
(450)
(484)
(445)
(916)

(057)
(049)
(053)
(056)
(070)
(058)

(062)
(039)

(397)
(378)
(417)
(381)
(413)
(613)

(245)
(223)

-.10
-.05
-.SI
-.01

.63
-.00

-.04
-.00
-.16
-. l 3
.07
.24
.03
-. 14

-.06

.05
-.02
-.07
-.02
-.05

-.03
-.Ol

.42
.12
.l9
.10
.05
.06

-. l 3
-.03

 

 

Table 5.4 (cont’d).

Black Protestant x Political Information
Catholic x Political Information
Secular x Political Information

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

Constant

Observations
x2 Statistic

Pseudo R2

-.595*
-.337
-.203
-.172

-.003
.020
.045
.006

-.036

.097
.198
.061
.096
.121
.351

-3.063**

11999

1 159.78*"

.10

(244)
(223)
(240)
(364)

(058)
(057)
(065)
(057)
(129)

(128)
(122)
(138)
(123)
(125)
(215)

(470)

-.15
-.O8
-.05
-.O3

-.00
.00
.01

-.00
.00

.03
.05
.02
.02
.03
.09

 

Note: ‘p<. 10, one-tail test; "p<.05, two-tail test; ***; p<.Ol, two-tail test. This table reports

unstandardized coefﬁcients (standard errors in parentheses) from the 2-Stage Conditional Maximum

Likelihood model (Rivers and Vuong 1988).
lMarginal effects are calculated at the mean/mode of the independent variables.

Figures 5.22 and 5.23 show that political participation and political information
act in a similar manner for Evangelicals on the Economy chronicity as the variables did

on the Social Issue chronicity. Rising participation sharply raises the probability of using

this chronicity while rising political information decreases it.

101

 

Figure 5.22. Increasing Political Participation and the Probability of Accessing the
Economy Chronicity by Religious Tradition.

5H

 

 

 

 

 

Probability of Using Economic Chronic1ty

 

 

O -1
Low 2 3 4 5 High
Campaign Participation
Evangelical ----- Mainline
... ....... - BlaCk ................. Catholic
- ------ Secular

 

 

 

102

Figure 5.23. Increasing Political Information and the Probability of Accessing the
Economy Chronicity by Religious Tradition.

.4

 

 

O

 

_(

l

Low 2 3 4 High
Political Information

Probability of Using Economic Chronicity

—(

 

 

 

Evangelical ----- Mainline
_ ....... - Black ................. Cathol ic
- ------ Secular

 

 

The No Content chronicity presents an unexpected but theoretically sound result
for political information and political participation, decreasing the usage of this
chronicity (which is, in effect, an empty-response category). Table 5.5 does not include
the coefﬁcients or interaction terms for political information in the model with the
political information variables. Political information approaches being the perfect
predictor for No Content as increased political information decreases the likelihood of its
usage. This result is a model with an inﬂated pseudo-R2 of .93 and with nearly 55% of
the cases perfectly predicted.l3 Therefore, the present model only includes the political
participation variables. As expected, increased participation reduces the probability of
falling into this category by 14%. There is no signiﬁcant relationship between

Evangelical participation and this category of non-response but, surprisingly, increased

 

'3 Stata Help, httpz/lwww.statacom/s@porﬁaqs/staﬂlogitcd.hml. Accessed 6/11/2004.

103

participation in several religious traditions — Mainline Protestant, Black Protestant,

Catholic and Secular — increases the probability of giving no response.

Table 5.5. No Content Chronicity, Participation and Information: ZSCML Results.

 

 

 

Variables in Model No Content S. E. Marginal
Effectsi
Demographics
Education -.O84** (.037) .01
Income .012 (.025) .00
Gender .1 14*" (.043) .01
Married -.017 (.044) -.00
Intelligence 267*" (.033) .02
Age -.Ol9*" (.006) -.00
Age Squared .000“ (.000) .00
Urban .044 (.028) .00
Where Grew Up .102" (.045) .01
Political South .123‘" (.046) .01
Political Variables
Party Identiﬁcation -.868""‘ (.055) -.03
Party Identiﬁcation Squared -.107“" (.051) -.04
Campaign Interest -.131"‘” (.039) -.01
Care Who Wins -.348""' (.046) -.03
Political Participation -I 877“" (.334) -.14
Participation Residual -. 145*" (.035) -.01
Religion Variables
Bible Attitudes -.l 1 1 (.151) -.01
Church Attendance .01 l (.056) .00
Evangelical -.597 (.585) -.04
Mainline Protestant -I .552‘" (.569) -.12
Catholic -.796 (.570) -.06
Black Evangelical 4.359" (.615) -. IO
Secular -I.458"‘" (.524) -.ll
Jew .764 (1 .950) .06
Election Years
1980 -.058 (.074) -.00
1984 .01 1 (.068) .00
1988 .l 15 (.068) .01
1992 -.033 (.072) -.00

104

 

 

Table 5.5 (cont’d).
1996
2000

Interaction Terms
Evangelical x Education
Evangelical x Income

Evangelical x Political Participation
Mainline Protestant x Political Participation
Black Protestant x Political Participation
Catholic x Political Participation

Secular x Political Participation

Jewish x Political Participation

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish )1 Bible Attitudes

Constant

Observations
x2 Statistic

Pseudo R2

.060
.089

.082
.069

.057

1.015’”
.773"
.762“
.784"

.261

-.004
-.019
.052
-.039
.l 13

.01 1
.092
.089
-.069
.177
-.782

2565*"

12084

l936.61***

.27

(095)
(079)

(069)
(050)

(372)
(349)
(375)
(350)
(363)
(924)

(.063)
(.064)
(.074)
(.064)
(.2 l 8)

(.172)
(.164)
(.179)
(.164)
(.166)
(.492)

(.578)

.00
.01

.01
.01

.00
.08
.06
.06
.06
.02

-.00
-.00
.00
-.00
.01

.00
.00
.01
-.01
.01
-.O6

 

Note: ’p<. 10, one-tail test; ”p<.05, two-tail test; *"p<.01, two-tail test. This table reports unstandardized

coefﬁcients (standard errors in parentheses) from the 2-Stage Conditional Maximum Likelihood model

'vers and Vuong 1988).

Marginal effects are calculated at the mean/mode of the independent variables.

Bennett’s hypothesis, that voting does not affect political cognition, is not

supported by the ﬁrst stage of the voting models (Appendix H). The relationship,

105

 

however, is not reciprocal as the chronicity type does not inﬂuence the likelihood of
voting. However, voting, particularly for Evangelicals, does affect the likelihood of using
speciﬁc chronicities. When looking at the signiﬁcant marginal effects, Evangelicals who
vote decrease the probability of using the candidate (-15%), groups (-9%), and Social
Issues chronicities (-6%) and increase the likelihood of using the foreign policy
chronicity (3%). Unexpectedly, voting and campaign participation affect the Social Issue
chronicity in opposite directions. Campaign participation increases social issue usage and
voting decreases it. For the voting model as with the participation model the No Content
chronicity has to be treated as a distinct model with the exclusion of political information.
For Catholics, Mainlines and Evangelicals, voting increases (2%, 3%, 4%, respectively)
the probability of falling into this chronicity.
5.1 1 Conclusions

This study shows full or partial support for all ﬁve hypotheses and their
corollaries. As hypothesized, accessing social issues as an organizing concept has
increased for Evangelicals and Seculars at a rate above the remainder of the population.
Social issue access has also increased for the rest of the population at a lower level.
Evangelicals and Seculars using this chronicity represent opposite extremes on social
issues and afﬁnity toward groups. Also, theorized increases in participation by
Evangelicals show reciprocal relationships with the social issue and economy
chronicities.

The Social Issue model presents a picture of Evangelicals who attend church
reciprocally increasing their political participation and their access of the social issues.

Unlike Neuman’s hypothesized spiral though, the increased focus on social issues is not

106

being fueled by nor resulting in increases in political information. There are also
unanswered questions about using political information as a surrogate for political
sophistication. A serious disconnect between political information and political
sophistication is suggested when access of a cognitive structure such as the chronicities
and participation could both increase without an increase in political knowledge. This

appears to be an issue for future research.

107

CHAPTER VI

PARTICIPATION AND EVANGELICAL MODERATION

The act of political participation not only inﬂuences the political process, it
inﬂuences the participant. Aggregate changes in rates of Evangelicals’ participation can
result in aggregate changes in political attitudes. The very salience of motives (defense of
traditional values and a Culture War dichotomy) which encourages participation or lack
of salience and accompanying inactivity can help to shape the direction of attitude shifts.

There may also be unintended consequences of new participation. These attitude
shifts are not necessarily uniform across the Evangelical movement, but instead may be
more profound for younger birth cohorts. Declining separatism and secularization are
theorized to work hand in hand in reshaping Evangelical political participants. If younger
cohorts are most inﬂuenced by the new calls to participate, these new cues encouraging
participation may be shaping the future of the Evangelical in unexpected ways (at least
from the Evangelical perspective). Current political battles over divisions in social and
political issues may result in a future where Evangelical moderation has reduced the
differences in attitudes between Evangelicals and Non-Evangelicals.

This chapter focuses on ﬁrst, cohort and period changes for Evangelical
campaign and voter participation, and second, on cohort, period and participation effects
on Evangelical political and moral attitudes. The chapter covers a breadth of models and
attitudes and falls within the category of “mid-range studies” that “evaluate changes

across a range of phenomena” (Smith 1990, 480).

107

6.1 Deﬁning Moderation

What is meant by moderation of Evangelical views? As was demonstrated in
Chapter ﬁve, analysis of attitude means has shown a consistent conservative position on
moral, social and political attitudes for Evangelicals. Previous examinations of
moderation among Evangelicals have suggested that moderation would take the form of
moving away ﬁom conservative views (Hunter 1987; Penning and Smidt 2003). This
deﬁnition of moderation also ﬁts the expectations of proponents of secularization theory.

Prior studies and theories suggest that a liberalizing inclination among
Evangelicals would resist the overall trend of the time period under consideration.
National trends for this time period in question (1976—2000) do not suggest an overall
liberalizing of the American population. Smith provided an overview of trends on social,
economic and governmental attitudes. He examined 455 survey trends consisting of 419
attitudes from 1945 to 1987. Smith concluded that Americans generally moved in a
liberal1 direction, but that the movement was not consistent across all topics. The
strongest liberal changes were for individual freedoms such as feminism, abortion rights,
and race/ethnicity. The most conservative trends involved crime and increasing support
for social control. Economic issues had mixed trends. This liberal direction accelerates
until 1974 and then liberal and conservative trends balance out one another. The liberal
movement levels off in the 1980s but still remains above the pre-l974 levels. Smith

points out that the leveling of the liberal trend movement is not associated with the

 

lSmith cross-tabulated GSS issue questions with self-placement on a seven-point liberal/conservative scale,
presidential vote choice in 1972 and 1984 and three other indicators of liberal/conservative placement to
determine the liberal and conservative direction for each issue. Issues were also deﬁned on a
liberal/conservative spectrum by the positions taken by liberal groups such as the American Civil Liberties
Union or the Americans for Democratic Action and conservative groups such as the American
Conservative Union (Smith 1990, 482).

108

Reagan administration, but rather it predates Reagan’s 1980 victory by six years. Smith
found that post-1974 conservative shifts occurred most commonly on the topics of
abortion, economic regulation, foreign affairs and religion. A conservative turn was least
common for feminism, racial issues, and attitudes toward sex (Smith 1990).

A different perspective in longitudinal studies is taken by Stimson. His theory of
public mood “implies that the publics see every public issue through general
dispositions” (1999, 20). These general dispositions represent a global policy attitude in
which the public over time alternates between a liberal and conservative policy attitude.
Stimson argues that changes in public mood results in shifts on current policy within a
“zone of acquiescence.” This zone is created within the acceptable liberal and
conservative boundaries that bracket public preferences. Shifting mood alters the policy
choices within the boundaries of what is acceptable without shifting policy too far to the
right or left. Stimson pictures the process as a river of time with a course representing the
interaction of public mood meandering down the river. The only restraints on the course
are the far banks representing extreme liberal and conservative positions. Stimson argues
that the moves left are in response to conservative government policies and to the right in
response to liberal policies. These shifts in mood took a conservative turn in the 19605
and 19705 before turning liberal in the 19808 and early 19903 and returning to a
conservative direction in the middle 19908. Stimson’s measurement of public mood is an
aggregate measurement using the entire American population. He does not evaluate or
compare the public mood between segments of the population.

At ﬁrst glance this theory appears to make evaluating Evangelical changes

difficult to measure as it confounds a liberal shift in public mood with an expected liberal

109

or moderating trend for Evangelicals in the 19808 and 19908. There are several key
differences between Stimson’s longitudinal study and this chapter. First, three main areas
of this chapter’s evaluation, moral issues (excluding abortion), gender roles and tolerance
are not included in Stimson’s public mood studies. Second, Stimson’s measurement is
based on multiple dimensions combining a large number of attitudes. He acknowledges
that pieces (individual attitude questions) “usually—but not always—look much like the
whole” (93). Even if a model focusing on a single attitude reveals a trend deviating from
the public mood, it neither negates Stimson’s theory nor invalidates the individual model.
Third, Stimson does argue that for some issues such as abortion, the policy space between
the left and right extremes becomes too narrow for any acceptable government action.
While the policy position is not static, neither are the boundaries of the acceptable zone.
Liberal shifts in Evangelical attitudes in the 19808 and 19908 can be perceived as
responses similar to the rest of the population. These responses change current policy
choice by shifting the course taken through the acceptable policy space. It is also
theoretically possible that Evangelical moderation could shift the acceptable policy space
in a leftward direction without affecting the current policy. The models in this chapter are
not capable of differentiating between the two possible scenarios. Stimson’s public mood,
though, would not explain if Evangelical attitudes shift leftward at a greater rate than
those of Non-Evangelicals. If Evangelicals are closing the gap between themselves and

Non-Evangelicals, it suggests a dynamic at work that is separate from the public mood.

6.2 Hypotheses

There are two possible historical points at which evangelical attitudes begin to be

inﬂuenced by secularization. This ﬁrst starts with the 1980 election cycle when Moral

110

Majority cues increase and are accompanied by a brief surge in Evangelical campaign
participation. This election cycle also marks the beginning point for the shift in cues from
separatism to participation. This shift remains constant through the end of the twentieth
century to the point that only eight percent of Evangelicals espouse a separatist viewpoint
in 2000 (Green 2003, 16). However, it should be acknowledged that while cues change in
the 19808, campaign behavior shows no evidence of a permanent shift until the 1992
Presidential election (see Table 5.1). It is only when Evangelicals begin to align
themselves with the Republican Party that Evangelical campaign participation shows a
permanent increase.

At the same time, competing inﬂuences have the potential to negate the supposed
moderating inﬂuence of modernization and secularization. The agenda of the Republican
Party, the Culture War thesis and internal Evangelical cues all inﬂuence Evangelicals in a
conservative direction. This suggests that the salience of political and social issues should
vary among Evangelicals as the content of internal cues varies. Those messages with
consistent salience in the 19808 were opposition to abortion, support of school prayer,
opposition to gay rights, and opposition to the Equal Rights Amendment (Guth 1983;
Jorstad 1987; Hunter 1987; Moen 1992). In the 19908 while abortion2 and gay rights
remained front and center, the cues from social interest groups shifted to tax policy
toward families, a balanced-budget amendment, equal access of school facilities for
student religious groups, capital punishment and discrimination against traditional values

(Moen 1996; Wilcox 2000).

 

2 There were even minor shifts in abortion stances as some leaders moderated their views ﬁom opposition
to all abortion to acceptance of abortion in certain cases (e.g. rape, incest, health of the mother) (Moen).

111

Feminism, subsequently, stands out as one of the issues where mixed messages
were given within the Evangelical community. Moen shows that in the 19808, social
issue groups such as Concerned Women for America, which advocated the most extreme
rejection of feminism, had a very narrow audience and limited success with its agenda.
As a result, intensity of opposition to feminism gave way to a more laissez faire attitude
toward feminist beliefs in the 19908. Also during this time period, several Evangelical
denominations including Southern Baptist, Brethren and Christian Reformed also
increased or overturned historical opposition to women’s ordination. This is matched by
New Paradigm churches’ openness to women’s ordination, to feminist views of women
in the workplace, and to gender equality in family roles (Miller 1997).

Evangelicals have demonstrated views distinct from the general public on
political issues including affirmative action, defense spending, environmental issues,
abortion, premarital sex, extra-marital sex, homosexuality and feminism (Curry-Roper
1990; Guth et al. 1996; Wilcox, Jelen and Leege 1993). New patterns of socialization
affecting these views would include differing levels of parental involvement as well as
changes in the church environment toward participation. This increases the possibility
that moderating changes in attitudes would be most pronounced for younger birth cohorts
(Beck and Jennings 1982; Dalton 1980; Jennings and Niemi 1981). New paradigm
churches, or those within established denominations affected by paradigm
denominations, could also inﬂuence this younger group as they have younger attendees
than non-paradi gm churches.

Less salient issues are most likely to beneﬁt from the moderating inﬂuences of

political involvement. The effects of secularization would be most effective when

112

Evangelical political cues fail to provide a clear directional message on what ideological
position should be taken on an issue.

Hunter already provides limited support for a cohort change. His 1982—1985
study of Evangelical college and seminary students found a generally conservative trend
for political attitudes among this young post-1980 cohort3. They hold conservative views
on busing, the death penalty, social welfare, women’s roles, abortion and homosexuality.
Despite these views, he also ﬁnds that this cohort holds liberal views on gun registration,
prayer in schools, increased defense spending, nuclear power and a nuclear freeze. Based
on Hunter, there is some initial evidence of the liberalizing of younger Evangelical views.

Snridt (1987) also ﬁnds that young Evangelicals in the 1984 election increased
their political involvement both as compared to older Evangelicals and also in
comparison to Non-Evangelicals of similar age. This ﬁts with Zaller’s (1992) view of
partisan resistance that shows that the new dominant message of electoral involvement
would experience more resistance to internalization from older respondents. In Zaller’s
perspective, the younger respondents would have less predisposition toward separatism.
The participation cues of Evangelical organizations should socialize these younger
cohorts into higher levels of involvement". Based on this analysis, I hypothesize for
Evangelicals:

H1: A period effect of increasing participation for Evangelicals (arising from the effects

of new cues for political participation and political attitudes regardless of the theorized
cause, modernization, secularization or culture war) will begin with the 1980 election.

 

3 Hunter found 40% of Evangelical college students and 48% of seminarians identiﬁed themselves as
conservative. They also were more likely to identify themselves as Republicans (51% and 46%) and
Independents (34% and 32%) rather than Democrats (15% and 22%, respectively).

‘ Abramson points out that “we are likely to frnd consistent age-group differences only when we begin by
examining attitudes that, on theoretical grounds, should be affected either by a person’s position in the
lifecycle or by some major events that might affect the formative socialization of speciﬁc age groups”
(1975, 125).

113

H2: The rise of cues from Evangelical political and social movement groups encouraging
political activism in conjunction with other secularizing changes within the Evangelical
tradition will result in younger cohorts whose levels of political involvement will rise to
match or even exceed that of similarly aged Non-Evangelical cohorts.
H3: Increasing participation will result in a liberal shift on political issues. This shift,
though, will be limited to less salient, less controversial issues such as the enviromnent,
feminism and race, and will not affect cue issues such as abortion and homosexuality.
Cl: Many of these liberalizing effects would be accentuated in the youngest
cohorts who have “come of age” during the new cues and were never socialized
into the old separatism. These younger cohorts will demonstrate a less traditional
stance on political, cultural and moral issues as compared to the older Evangelical
cohorts and will close the “gap” on tolerance with Non-Evangelicals.

C2: Liberalizing or moderation will be increased by campaign participation
through contact with divergent viewpoints.

H4: Evangelicals will exhibit increased tolerance for out-groups. This increase would be
most evident for the out-group of atheists, the group most likely to be viewed as
threatening by religious Evangelicals.

C1: Younger Evangelical cohorts will also demonstrate the greatest levels of

tolerance for selected out-groups as compared to older Evangelical cohorts and
will again close the “gap” with Non-Evangelicals.

6.3 Age-Period-Cohort Models

If moderating or liberalizing changes have begun to occur among Evangelicals, as
a result of secularization brought on by a decline in separatism, an analysis of trends
across this time period should reveal the shifts. As there is an expectation of both period
and generational/cohort effects, this study draws upon the framework of age-period-
cohort (APC) analysis.

The difficulty with any APC model is the problem of identiﬁcation (F ienberg and
Mason 1985). The speciﬁcation of an APC model can lead to a model with perfect

multicolinearity, as the cohort is a linear function of age and period (cohort = period —-

114

age). Mason et al (1973) proposed to deal with the problem by omitting an additional
period, age or cohort dummy variable. This elimination sets the effects of the two omitted
variables as equal and allows for the estimation of age, period and cohort effects.
However, the Mason et a1 approach is only applicable to models that are nonlinear (284
fn). Glenn (2005) shows that this method only works for nonlinear models when the a
priori assumptions are exactly correct. Glenn also evaluates other methods for separating
age, period and cohort effects and concludes that “A deﬁnitive separation of age, period
and cohort effects is not just difﬁcult, but impossible” (2005, p. vii; See Wilmoth 1990
for a similar conclusion).5 As the period effects of Evangelical participation are
hypothesized to be linear, the Mason et al. approach is not applicable to this study.

There remain two options for properly identifying this model. One is the
substitution of part of the APC triad with a surrogate that captures a characteristic of the
removed periods, ages or cohorts. However, this type of model is not a true APC model
and the substituted variable is not likely to capture all of the effects of the missing APC
member (Glenn). The second option for identiﬁcation is to exclude one part of the APC
triad. Central to this option is the issue of whether there is any basis for expecting a
liberalizing inﬂuence from all three parts of the triad, age and periods and cohorts. Also
in terms of rates of participation, it must be determined if there is a theoretical basis for
expecting Evangelical aging to diminish the separatism which reduces their campaign
and voting participation.

Examinations of political attitudes over time have suggested that stability is the

norm with change possible at later ages, albeit at a much slower rate than in adolescence

 

5 Abramson (1978) tests generational differences using a method for estimating attitude means with and
without cohort replacement. However, Abramson acknowledges that his model overestimates period
effects—one of the key phenomena under evaluation here.

115

(Alwin, Cohen and Newcombe 1991; Jennings and Niemi 1981; Niemi and Jennings
1991; Sears and Funk [year]; Zaller 1992; see Krosnik and Alwin 1989 for an opposing
view). There is, however, no deﬁnitive research showing that aging produces a generally
conservative or generally liberal effect. Generational cohorts are found to be more liberal
or conservative than prior cohorts (Jennings and Niemi 1981; Smith 2000). As a prime
example of this, younger cohorts are found to be more tolerant while age effects are weak
(Glenn 1980; Jennings and Niemi 1981; Miller and Sears 1986; Sears 1975; Sullivan,
Pearson and Marcus 1982; see Hill 1997 for an opposing view).6 As the theorized effect
of secularization for moral, social and political issues is a liberal shift, there is no a priori
basis for expecting age to increase this liberalization or moderation.

A similar conclusion can be reached for political participation. A dividing of the
American electorate into three cohorts (Pre-New Deal, New Deal, Post-New Deal) shows
overall voter participation declining as the Post-New Deal replaces the Pre-New Deal.
Post-New Deal participation is lower than both preceding cohorts (Miller and Shanks
1996). Aging has a curvilinear effect on voting and participation with middle age
providing the highest levels of participation as compared to youth and elderly (Converse
and Niemi 1971; Rosenstone and Hansen 1993; Strate et al. 1989; Verba and Nie). When
examining participation there is no a priori basis for assuming that the inﬂuence of aging
is not uniform for Evangelicals and Non-Evangelicals. The aging trend should be
identical for both groups while the birth cohort trend is hypothesized to diminish the

distance between the younger Evangelical and Non-Evangelical cohorts.

 

6 The sheer number of attitudes being examined in this chapter forestalls the possibility of providing cohort
research for each of the political attitudes.

ll6

6.4 Data and Coding

Historical trends for Evangelicals and Non-Evangelicals are examined using data
from the 1972—2000 American National Election Studies (AN ES) and the 1974—2002
General Social Survey. While GSS surveys are available for 1972 and 1973, usage of
these two years would require dropping the only measurement of political ideology that is
not utilized until the 1974 survey. Given the scarcity of political variables in the GSS,
these years were dropped to keep the seven-point ideology scale. Religious traditions are
again coded based on the Steensland et a1 codings and divided into six categories
(Evangelical Protestant, Mainline Protestant, Black Protestant, Roman Catholic, Secular
and Other [which combines the Steensland et a1 non-traditional liberals, non-traditional
conservatives and others categories]). NES data is used to compare period and cohort
trends in campaign and voter participation for Evangelicals and Non-Evangelicals. It is
also used to compare period, cohort and campaign participation trends in political and
moral attitudes for Evangelicals and Non-Evangelicals. Based on Chapter Five’s results
and Bennett’s (1975) now-proven hypothesis that voting is not sufﬁciently interactive to
change political thinking, a similar analysis was not performed comparing periods,
cohorts, and voting participation trends in political and moral attitudes.

The GSS data does not include a campaign participation measurement. These
models are limited to examining period and cohort effects. The GSS is included, though,
because of its diversity of moral questions and variety of civil liberty/tolerance questions.
While participation effects cannot be shown, period and cohort trends can be compared to
NES trends to see if a similar outcome is observed. A lack of continuity in the NES (and

to a much lesser extent in the GSS) data severely limits the number of attitudes that can

117

be examined. Questions asked in 1980 or prior to 1980 were used as dependent variables
in the models whenever possible. In a few instances when Evangelicals have been shown
to hold different levels of support for attitudinal questions, recent shorter trends were

examined.

Table. 6.1. NES Period and Ten-Year Birth Cohorts (number of Evangelicals in

parentheses)
1903- 1913- 1923- 1933- 1943- 1953- 1963- 1973- Total
1912 1922 1932 1942 1952 1962 1972 1982

I972 177 200 246 234 308 46 O 0 1,211 (270)
1976 347 404 401 452 717 317 0 0 2,638 (577)
1980 159 229 223 233 355 352 O 0 1,551 (334)
1984 177 274 263 290 507 534 142 0 2,187 (471)
I988 109 211 248 252 430 515 248 0 2,013 (439)
1992 109 227 284 300 424 644 464 31 2,483 (573)
I996 51 122 203 210 273 436 322 96 1,713 (399)
2000 20 82 158 216 331 405 358 228 1,798 (394)

Total 1,149 1,749 2,026 2,187 3,345 3,249 1,534 355 15,594
(240) (398) (469) (526L (704) (730) (3 14) (77) (333 1)
Source: National Election Studies 1972-2000

Periods are measured in the NES using Presidential Election years from 1972—
2000. Eight birth cohorts, each covering ten years, are coded beginning with 1903—1912
and ending with 1973—1982. An identical coding for birth cohorts is used with the GSS.
Respondents with 1983 and 1984 birth years are available in the GSS but they only
represent 29 respondents. They are excluded to maintain an identical cohort coding with
the NES. GSS education and income variables were recoded to match NES education and

income categories. GSS periods represented individual survey years from 1974—2002.

118

Tables 6.1 and 6.2 show the number of respondents in each cohort and period for the
NES and GSS. The number of Evangelicals in each cohort and period is found in
parentheses following the totals. Evangelicals represent 21%—23% of each category.

Table 6.2. General Social Survey Birth Cohorts (number of Evangelicals in cohort in

parentheses)
1903- 1913- 1923- 1933- 1943— 1953- 1963- 1973- Total
1912 1922 1932 1942 1952 1962 1972 1982

1974 171 232 249 254 364 89 o 0 1,359 (287)
1975 185 205 242 260 353 138 0 0 1,383 (288)
1976 170 223 211 254 352 162 o 0 1,372 (308)
1977 164 238 251 275 335 192 o 0 1,455 (321)
1978 145 205 202 249 393 244 0 0 1,438 (329)
1980 139 214 213 200 354 285 0 0 1,405 (336)
1982 154 233 271 254 373 457 39 0 1,781 (317)
1983 112 191 225 222 353 403 45 0 1,551 (328)
1984 99 168 196 193 331 367 82 0 1,436 (358)
1985 103 193 217 212 303 366 100 0 1,494 (398)
1986 102 176 179 193 315 368 108 0 1,441 (370)
1987 95 208 207 264 362 439 203 0 1,778 (360)
1988 85 175 179 151 316 341 210 0 1,457 (367)
1989 72 162 177 223 288 371 223 0 1,516 (379)
1990 66 147 161 153 278 344 209 0 1,358 (316)
1991 62 158 169 172 293 374 275 3 1,506 (388)
1993 53 150 162 199 315 381 298 43 1,601 (454)
1994 83 235 297 385 582 740 565 99 2,986 (718)
1996 57 163 248 328 543 687 651 221 2,898 (673)
1998 39 144 268 302 472 654 656 292 2,827 (641)
2000 22 158 226 288 455 653 582 425 2,809 (604)
2002 0 115 239 298 434 545 590 501 2,722 (566)
Total 2,178 4,093 4,789 5,329 8,164 8,600 4,836 1,584 38,573

(541) (998) (1,149) (1,362) (1,865) (1,834) (1,071) (286) (9,106)
Source: General Social Survey, 1974-2002.

6.5 Methodology Periods, Cohorts and Changes in ANES Participation

The ﬁrst two models in this section use AN ES data and focus on period and

cohort changes in campaign and voter participation. The ﬁrst dependent variable is a six-

119

point campaign participation scale where respondents received a score from 1 (no
participation) to 6 (participation in all activities) for inﬂuencing others votes, attending
political meetings, working for a party or candidate, displaying a button/sticker or making
a campaign contribution. Following the SES model (V erba and Nie 1972; Rosenstone
and Hansen 1993), control variables were included for gender, education and income.7
Political variables included the respondent’s strength of partisanship, the respondent’s
concern for who wins the election, and if the respondent was contacted by a major party.
Additional control variables included religious attendance and dummy variables for
Evangelical Protestants, Mainline Protestants, Black Protestants, Roman Catholics, Jews
and Seculars (see Appendix I for descriptions of variables). Dummy variables are
included for each period and birth cohort. Interaction terms for Evangelical SES and
political variables as well as period and cohort variables, are also included. Missing data

in independent variables for all models in this chapter are set to the mean in the case of

continuous variables and to the mode for dichotomous variables.8

6.6 NES Participation Results

Table 6.3 contains the results for the campaign participation OLS model. With
exception of the South dummy variable, all of the SES and political variables are
signiﬁcant. Being female is the only SES/political variable that lowers the level of
campaign participation. The Evangelical SES and political variable interactions show that
the slopes of Evangelical education and partisan strength are lower than that of Non-

Evangelicals. Increasing education and partisan strength results in lower levels of

 

7 In keeping with Rosenstone and Hansen’s results, an initial model included a married variable, which
failed to be signiﬁcant, so it was dropped.

' Data was also weighted using vcﬂ1009 to produce a nationally representative pooled cross-sectional
analysis.

120

participation as compared to Non-Evangelicals. Most religious identities are not
signiﬁcant with the exception of Black Protestants and Roman Catholics which both
lower campaign participation. As previous studies have shown, church attendance raises
the level of participation and the effect is higher for Evangelical participants.

Table 6.3. Regression of Periods and Cohorts on Campaign Participation: NES
1972-2000.

 

Campaign Participation

 

Birth Cohorts

1903-1912 -.04 (.05)
1913-1922 .05 (.04)
1923-1932 .05 (.04)
1933-1942 .03 (.04)
1943-1952 .07 (.04)
1953-1962 .02 (.04)
1963-1972 .05 (.04)
Periods

1976 .08* (.03)
1980 -.04 (.04)
1984 -.08 (.04)
1988 -.07" (.04)
1992 -.11* (.04)
1996 -.26" (.04)
2000 -.25" (.05)
Gender -.12" (.02)
Education .18" (.01)
Income .06" (.01)
South .02 (.02)
Strength of Partisanship .12" (.01)
Care Who Wins .26" (.02)
Party Contact .39" (.02)
Church Attendance .02“ (.01)
Evangelical Protestant -.17 (.15)
Black Protestant -.12* (.05)
Mainline Protestant -.07 (.04)
Roman Catholic -.10* (.04)
Jewish .08 (.08)
Secular -.O7 (.04)
Interaction Terms

Evangelical X Gender .05 (.03)
Evangelical X Education -.07" (.02)
Evangelical X Income .02 (.02)
Evangelical X Church Attendance .03‘ (.01)
Evangelical X Strength of Partisanship -.06" (.02)
Evangelical X Care Who Wins .04 (.03)
Evangelical X Party Contact -.02 (.05)

121

Table 6.3 (cont’d).

Evangelical X South -.03 (.04)
Evangelical Birth Cohorts

1903-1912 -.03 (.09)
1913-1922 -.04 (.08)
1923-1932 .05 (.08)
1933-1942 .07 (.08)
1943-1952 .01 (.08)
1953-1962 .01 (.08)
1963-1972 -.06 (.09)
Evangelical Periods

1976 .09 (.07)
I980 .23“ (.08)
1984 .1 1 (.07)
1988 .12" (.07)
1992 .26" (.07)
1996 .29“ (.08)
1900 .33” (.08)
Constant .33" (.08)
N 13068

R2 . 15

 

*p<.05 (two tail test), I"“p<.01 (two tail test)

Notes: Entries are regression coefﬁcients with standard errors in parentheses.
The 1903-1912 cohort is dropped as the baseline variable.

The 1972 period variable is dropped as the baseline variable.

Period, cohort, and education effects on participation are evaluated by graphing
their effects while controlling for SES, political, and religious variables. The linear
relationship is not only examined, but loess smoothing9 is used to check for nonlinear
trends for birth cohorts. Where the loess smoothing reveals a signiﬁcant difference from
the linear plot, the loess line is also included in the graph. This is necessary, as I expect
the inﬂuence of birth cohorts to be nonlinear. As younger Evangelical cohorts are
expected to change in relationship to Non-Evangelical younger cohorts, these changes
may be limited to the youngest two or three cohorts. The changes may then be masked by

an OLS trend.

 

9 Loess or locally weighted regression is used to ﬁt a smooth curve to the points of a data scatterplot. Loess
gives more weight to neighboring points near the point whose response is being estimated and less weight
to points further away (Cleveland and Devlin 1988; Jacoby 1997).

122

Figure 6.1. Campaign Participation by Period, Cohort and Evangelical Traditionl

s e-I

w
e—

°°.

1.7

1.6

   

 

 

 

 

 

 

sr. .. 8r. -
(Q _‘ (ﬂ. -1
72 76 80 84 88 92 96 00 1903- 12 13-22 23-32 33-42 43-52 53-62 63-72 73-82
Election Year Birth Cohorts
Evangelical Evangelical Lowess Smoothing

 

_____ Other Other Lowess Smoothing

 

 

 

Note: OLS Data are frbm Table 6.3.
1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

Figure 6.1 demonstrates differing period and cohort effects for both Evangelicals
and Non-Evangelicals. Non-Evangelicals have declined in their campaign involvement
over the past three decades, while Evangelicals have increased their participation. The
coefficients for Evangelical periods (Table 6.3) uncover the points at which change
occurred. The 1980 election saw a one-time jump in participation, but an unbroken trend
of significant increase ran from 1988 to 2000. The data suggests that an initial response to
new cues in 1980 was followed by a slower acceptance of new patterns of participation.

The 1988 election was also a key year for Non-Evangelicals as campaign participation

began a signiﬁcant downward trend that lasted through 2000.

123

The results of the birth cohort trends are misleading when just the cohort
coefﬁcients are examined. None of the dummy cohort variables or interaction terms is
signiﬁcant. The graph of birth cohorts and campaign participation (Figure 6.1) does
reveal an interesting trend, showing a closing of the gap between younger Evangelical
and Non-Evangelical cohorts. The loess curve reveals that both Evangelical and Non-
Evangelicals birth cohort trends ﬁt the lifecycle theory. The oldest and youngest birth
cohorts have lower campaign participation than the middle cohorts. The true change in
Evangelical birth cohorts is not represented by the trend line but by the distance between
matching Evangelical and Non-Evangelical cohorts. Here we see that the four youngest
Evangelical cohorts (1943-52, 1953-62, 1963-72 and 1973-1982) have begun to close the
gap, and the youngest cohort matches Non-Evangelicals in campaign participation.
Younger Evangelical cohorts are now poised to match or exceed campaign participation
levels of Non-Evangelicals as they move into their middle ages.

The second model (Table 6.4) performs a similar evaluation of periods, cohorts
and voting. The second dependent variable is a measurement of voter participation coded
1 if the respondent voted in the Presidential election and 0 if the respondent did not vote.
The logistic regression model uses the same independent SES, religious, period, cohort
and interaction variables as the campaign model with a few additions. Again following
Verba and Nie (1972) and Rosenstone and Hansen (1993), the model includes strength of
partisanship, care who wins, contact by a major party as well as length of residence and
closeness of the election.

Table 6.4. Logistic Regression of Periods and Cohorts on Voter Participation: NES

1972-2000.
Voter Participation

Birth Cohorts

124

Table 6.4 (cont’d).
1913-1922
1923-1932
1933-1942
1943-1952
1953-1962
1963-1972
1973-1982

Periods
1976
1980
1984
1988
1992
1996
1900

Gender

Education

Income

South

Strength of Partisanship
Care Who Wins

Length of Residence
Closeness of Election
Party Contact

Religious Attendance
Evangelical Protestant
Black Protestant
Mainline Protestant
Roman Catholic
Jewish

Secular

Interaction Terms

Evangelical X Gender

Evangelical X Education
Evangelical X Income

Evangelical X Religious Attendance
Evangelical X Strength of Partisanship
Evangelical X Care Who Wins
Length of Residence

Evangelical X Closeness of Election
Evangelical X Party Contact
Evangelical X South

Evangelical Birth Cohorts
1913-1922
1923-1932
1933-1942
1943-1952
1953-1962
1963-1972
1973-1982

.66"
.69“
.56”
.30*
.10
-.35"'*
-.56"”"

-.30*
-.32*
-.03
-.20
.06
-.16
-. l 7

al7**
.68"”'I
.33**
a40“
.85**
.01“I
.21“
.87"”'I
.31“

.33**
.10

.39"
.60"
.57"
.85“
.42“

-.15
.03
.03
.04

-.02

.20
-.00*
-.15

.13

.10

.00
-.O8

.08
-. l 7
-.03

.1 l

125

(.15)
(.15)
(.14)
(.14)
(.13)
(.13)
(.14)

(.12)
(.13)
(.12)
(.12)
(.12)
(.13)
(.13)

(.05)
(-04)
(.03)
(.07)
(.06)
(.00)
(.06)
(.07)
(.03)

(.03)
(.53)
(.13)
(.11)
(.1 1)
(.23)
(.12)

(.1 1)
(.08)
(.06)
(-04)
(.06)
(.12)
(.00)
(.13)
(.16)
(.12)

(.32)
(.29)
(.29)
(.28)
(.27)
(.27)
(.30)

Table 6.4 (cont’d).

Evangelical Periods

1976 .69" (.24)
1980 .33 (.26)
1984 -.O9 (.23)
1988 -.03 (.24)
1992 .03 (.24)
1996 .15 (.25)
2000 .05 (.26)
Constant -5.59*"' (.38)
N 13,523

Pseudo R2 '23

 

*p<.05 (two-tail test), "p<.01 (two-tail test)

Notes: Entries are logistic regression coefﬁcients with standard errors in parentheses.
The 1903-1912 cohort is dropped as the baseline variable.

The 1972 period variable is dropped as the baseline variable.

As in the previous model, all of the political variables are signiﬁcant, and raise the
probability of voting. Being female and from the South both lower that probability.
Religious attendance and all of the religious identities, with the exception of Evangelical
Protestants, increase the likelihood of voting.

Figure 6.2 shows differing trends for the probability of voting. While both
Evangelicals and Non-Evangelicals have increased their likelihood of voting, just as with
campaign participation, Evangelicals have narrowed the gap over time. The loess line for
birth cohorts and voting shows a curvilinear trend similar to campaign participation.
Once again it is the younger Evangelical cohorts that are closing the participation gap
with Non-Evangelicals. There is little evidence that 1980 was a crucial year for voting.
Increases in Evangelical voting are most pronounced in the 1976 election and are

probably associated with the candidacy of J immy Carter.

126

Figure 6.2. Probability of Voting by Period, Cohort and Evangelical Traditionl

 

 

 

 

 

 

 

O} -1 O\ -4

or; a co _

‘0. -1 \D ..

W. - V). -

72 76 80 84 88 92 96 00 1903-12 13-2223-32 33-4243-52 53-62 63-7273-82
Election Year Birth Cohorts
Evangelical — Evangelical Lowess Smoothing
----- Other Other Lowess Smoothing

 

 

 

Note: Predicted probability data is from Table 6.4.
1Controlling for Gender, Education, Income, Cohorts, Religious attendance, Strength of partisanship, Care
who wins, Party contact and Religious identity.

6.7 NES Political Attitude Results

I next examine NES political attitudes for period, cohort and participation effects
on these attitudes. These attitudes cover topics for which Evangelicals have previously
shown differing levels of support as compared to Non-Evangelicals. While other
variables are included as controls, the discussion of the models will be limited to period,
cohort and campaign participation effects. The ﬁrst two regression models (Table 6.5)
look at environmental issues, a topic for which Evangelicals have shown lower levels of

support (Guth et a1 1996). The dependent variables are environmental spending (coded l

127

- Decrease, 2 - Keep the Same, 3 - Increase) and an environmentalist thermometer (O -
Unfavorable to 100 - Favorable). The control variables for these models and all
subsequent NES models are gender, education, income, South, party id, dummy variables
for religious traditions, church attendance and campaign participation. Evangelical
campaign participation and Evangelical church attendance are included as interaction
terms. Period, cohort variables, and their interaction terms remain the same as the voting
and campaign models.

Figures 6.3 and 6.4 show the OLS trend lines for periods, cohorts and
participation while controlling for the remaining variables in the models. Contrary to the
secularization theory, the period and participation effects have not only widened the gap
of environmental support between Evangelicals and Non-Evangelicals, but Evangelical
support has declined. Meanwhile Non-Evangelical support (with the exception of the
period effects in the environmentalist thermometer model) has increased. Only the
Evangelical birth cohorts show rising support for younger cohorts, but the rate of increase
is still below that of the rest of the population.

Table 6.5. Regression of Periods, Cohorts and Campaign Participation on
Environmental Issues: NES 1980—2000.

 

 

Environmental Environmentalist

Thermometerl Spending2
Cohorts
1913-1922 .74 (1.56) -.04 (.04)
1923-1932 .32 (1.49) -.03 (.04)
1933-1942 1.41 (1.48) -.05 (.04)
1943-1952 3.00* (1.41) -.06 (.04)
1953-1962 3.60“ (1.40) -.10* (.04)
1963-1972 3.61 * (1.46) -.13"' (.04)
1973-1982 4.12“ (1.96) -.23" (.05)
Periods
1988 4.38" (.81) -.32"‘ (.02)
1992 -4.27"' (.80) -.26* (.02)
1996 -8.51* (.85) -.02 (.02)
2000 -8.22"' (.90) -.09* (.02)

128

Table 6.5 (cont’d).

Gender 3.60“ (.44) .01 (.01)
Education 1.10“ (.27) -.04* (.01)
Income -.99* (.23) .03“ (.01)
South 1.41“ (.57) .03“ (.01)
Party ID -l.72"‘ (.1 1) .05“ (.00)
Church Attendance -.03 (.18) -.01 (.01)
Evangelical -8.77* (3.39) .29"I (.09)
Black Protestant -.88 (1.17) .13* (.03)
Mainline Protestant -.99 (.97) .10* (.03)
Roman Catholic 1.57 (.97) .08" (.03)
Secular .70 (1 .08) .06* (.03)
Campaign Participation .71” (.25) .00 (.01)
Interaction Terms

Evangelical X Campaign Participation -.02 (.62) .01 (.02)
Evangelical X Church Attendance .79* (.37) -.03* (.01)
Evangelical Cohorts

1913-1922 .57 (3.42) -.10 (.09)
1923-1932 3.77 (3.26) -.07 (.08)
1933-1942 -.31 (3.21) .01 (.08)
1943-1952 .70 (3.07) -.1 1 (.08)
1953-1962 2.57 (3.02) -.06 (.08)
1963-1972 4.70 (3.22) -.09 (.08)
1973-1982 11.17* (4.01) -.05 (.11)
Evangelical Periods

1988 4.37“ (1.84) .02 (.05)
1992 1.99 (1.78) -.02 (.04)
1996 1.12 (1.89) .09 (.05)
2000 .25 (2.00) .06 (.05)
Constant 67.69“ (2.21) 1.47“ (.06)
Observations 8062 9578

R2 .12 .10

 

‘p<.05 (two-tail test), l""‘p<.01 (two-tail test)

Notes: Entries are regression coefﬁcients with standard errors in parentheses.
The 1903-1912 cohort is dropped as the baseline variable.

1The 1980 period variable is dropped as the baseline variable. The question was not asked in 1984.
2the 1984 period variable is dropped as the baseline variable.

129

Figure 6.3. Environmental Spending by Period, Cohort, Campaign Participation
and Evangelical Tradition.l

 

 

 

 

 

 

 

 

 

 

g 0
s- i-
.5 .5 _ _ __ -
o 0 —————————— _—-
O 0

g E

g 1984 1988 1992 1996 2000 1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82

Election Year Ten Year Birth Cohorts

0

8

b ..

E: _________________

o _n

g ..

(I)

g ‘I I l T l l I

w 2 3 4 5 High
Campaign Participation
— Evangelical Trend OLS ----- Other Trend OLS

 

 

 

 

Note: OLS data is from Table 6.5.
‘Controlling for Gender, Education, Income, Cohorts, Religious attendance, Strength of partisanship, Care
who wins, Party contact and Religious identity.

130

 

Figure 6.4. Environmentalist Thermometer by Period, Cohort, Campaign
Participation and Evangelical Tradition.l

 

 

 

 

 

 

 

 

 

e - s -
a 3 g _

1988 1992 1996 2000 1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82

Election Year Ten Year Birth Cohorts

o -
00
E 4 : ——————————————
c a
\D
o ..
m f l I I l T

Low 2 3 4 5 High

Campaign Participation
Evangelical ----- Other

 

 

 

 

Note: OLS data is ﬁ'om Table 6.5.
lControlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

The next set of models (Table 6.6) examines attitudes towards feminism and
women’s roles using the feminist thermometer (0 - Unfavorable to 100 - Favorable) and a
seven-point women’s roles scale (1- Equal Roles to 7- Traditional Roles). The period,
cohort and campaign participation effects (Figure 6.5) reveal conﬂicting trends for the
feminist thermometer. The period trend shows a nearly parallel set of OLS lines for both

Evangelicals and Non-Evangelicals showing increasing support. Campaign participation,

perhaps as a reﬂection of Republican support for traditional roles, decreases afﬁnity for

131

feminists. The birth cohort trend, particularly when looking at the loess line, shows a
sharp liberalizing trend for the youngest Evangelical cohorts. Support for feminists,
regardless of the Evangelical/Non-Evangelical split, is nearly identical for the youngest
cohort.

Table 6.6. Regression of Periods, Cohorts and Campaign Participation on Women’s
Issues: NES 1972-2000.

 

 

Feminist ThermometerI Women’s Roles2
Cohorts
1913-1922 .94 (1.19) -.05 (.09)
1923-1932 1.43 (1.15) -.15 (.09)
1933-1942 .28 (1.15) -.43"‘ (.09)
1943-1952 2.51" (1.07) -.63‘ (.08)
1953-1962 .76 (1.09) -.70* (.08)
1963-1972 3.24" (1.21) -.74* (.09)
1973-1982 7.35“ (1.70) -.85* (.13)
Periods
1972 -8.92* (1.18)
1976 -2.46* (.95)
1984 3.23“ (.92)
1988 -.33* (.05)
1992 6.04"' (.90) -.57"' (.05)
1996 6.33“ (.94) -.52* (.05)
2000 4.82* (.98) -.57" (.07)
Gender 2.93" (.43) -.l 1 "' (.03)
Education 1.32“ (.26) -.38“' (.02)
Income -.76"' (.22) -.09"I (.02)
South 1.42“ (.55) .08 (.04)
Party ID -2.00"' (.1 1) .08‘ (.01)
Church Attendance 134' (.18) -.16“ (.01)
Evangelical -13.57* (3.03) .45“ (.21)
Black Protestant 6.89" (1.1 1) -.02 (.09)
Mainline Protestant -2.76* (.90) -.08 (.07)
Roman Catholic -.17 (.90) -.18* (.07)
Secular -1.61 (1.00) -.16" (.07)
Campaign Participation 1.06“ (.24) -.05* (.02)
Interaction Terms
Evangelical X Campaign -1.51"' (.64) .04 (.05)
Participation
Evangelical X Church Attendance 1.76* (.36) -.10* (.03)
Evangelical Cohorts
1913-1922 -1.10 (2.68) -.06 (.21)
1923-1932 -2.25 (2.52) .19 (.20)
1933-1942 -2.88 (2.54) .18 (.20)

132

Table 6.6 (cont’d).
1943-1952
1953-1962
1963-1972
1973-1982

Evangelical Periods
1976
1980
1984
1988
1992
1996
2000

Constant

Observations
R2

-3.12
-3.23
.81
4.07

5.45“
8.36"
4.63"

6.38"
8.06“
8.72“
49.89"

10958
.16

(2.41)
(2.44)
(2.71)
(3.73)

(2.69)
(2.20)
(2.25)

(2.10)
(2.23)
(2.26)

(2.01)

.27
.29
.14
-. 10

-.24
-.29"‘
-.19
-.51“

5.02“

13020
.16

(.19)
(.19)

(.21)
(.27)

(.12)
(.11)
(.13)
(.15)

(.14)

 

*p<.05 (two-tail test), “p<.01 (two-tail test)
Notes: Entries are regression coefﬁcients with standard errors in parentheses.
The 1903-1912 cohort is dropped as the baseline variable.

1The 1972 period variable is dropped as the baseline variable. The question was not asked in 1988.
2The 1980 period variable is dropped as the baseline variable.

133

Figure 6.5. Feminist Thermometer by Period, Cohort, Campaign Participation and
Evangelical Tradition: NES 1972—2000.l

" I
'..—’ .‘oo- -
...—"... ’oww
...—.....—
n”

I:

ngh
ngh

 

50
l
l
l
l
l
l
1
50

 

 

 

 

 

 

 

3 . 3 _
'3 r I I I I I I h '3 I I I I I I I l
72 76 80 84 88 92 96 00 1903-12 13-2223-32 33-4243-52 53-62 63-7273-82
Election Year Ten Year Birth Cohorts
a:
.994
:1:
O ..
V3 ﬂ
3
o —1
‘4 r l I I I I
Low 2 3 4 5 High
Campaign Participation

 

Evangelical —— Evangelical Loess Smoothing
"""""" Other ----- Other Loess Smoothing

 

 

 

Note: OLS data is ﬁ'om Table 6.6.
1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who

wins, Party contact and Religious identity.

The same cohort effect (Figure 6.7) is observed in the women’s roles scale. While
the overall differences are not large, the youngest Evangelical cohorts move in a more
liberal direction and the youngest cohort matches the Non-Evangelicals in their support
of role equality. The period trend also shows a closing of the gap between the two groups.
Unlike the feminist thermometer model, rising campaign participation moves respondents

in a liberal direction but does not close the gap with Non-Evangelicals.

134

Figure 6.6. Women’s Roles by Period, Cohort, Campaign Participation and
Evangelical Tradition: NES 1972-2000I

 

 

 

 

 

 

 

 

 

'23
C
.9
E
E
2 d N
.N.“
3 -
[3- l I 1 l l l l T
1903-1213-22 23-32 33-42 43-52 53-62 63-72 73-82
Election Year Birth Cohorts
'3
G
0
IE -
E
3 _
2
3.4
m r r I r I r
Low 2 3 4 5 ngh

Campaign Participation

 

Evangelical Trend OLS _ Evangelical Lowess Smoothing
----- Other Trend OLS Other Lowess Smoothing

 

 

 

 

Note: OLS data is ﬁom Table 6.6.
1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

The last NES model looks at abortion support. Table 6.7 and Figure 6.7 show that
abortion support has increased for Non-Evangelicals across periods, cohorts and
campaign participation. Without examining the cohorts of religious groups, Scott (1998),
using the GSS, has found evidence of a liberalizing trend for US. cohorts. This analysis
of the NES shows a variation on the cohort effect for Evangelicals. The cohort and
campaign effects show a slight increase in abortion support, while the period effect shows
a slight decline. For Non-Evangelicals all three trends show an increase in support. The

anti-abortion focus of the Evangelical tradition appears to have negated much of the

inﬂuences that raised the level of support for Non-Evangelicals. However, campaign

135

participation does increase abortion support, showing some evidence of the inﬂuence of

divergent viewpoints in the political realm.

Table 6.7. Regression of Periods, Cohorts and Campaign Participation on Abortion
Support: NES 1972—2000.

 

Abortion Support

Cohorts

1913-1922 .07 (.05)
1923-1932 .1 1* (.05)
1933-1942 .04 (.05)
1943-1952 .12“ (.05)
1953-1962 .24“ (.05)
1963-1972 .20* (.05)
1973-1982 .02 (.08)
Periods

1984 .01 (.04)
1988 -.04 (.04)
1992 .13“ (.04)
1996 .04 (.04)
2000 .00 (.04)
Gender . 12* (.02)
Education .22"I (.01)
Income .1 1" (.01)
South -.05" (.02)
Party ID -.05* (.00)
Church Attendance .22* (.01)
Evangelical -.43 ‘ (. 12)
Black Protestant -.25* (.05)
Mainline Protestant .02 (.04)
Roman Catholic -.23"' (.04)
Secular -.l 1" (.04)
Campaign Participation .03* (.01)
Interaction Terms

Evangelical X Campaign Participation -.02 (.02)
Evangelical X Church Attendance .06" (.01)
Evangelical Cohorts

1913-1922 .11 (.12)
1923-1932 .08 (.11)
1933-1942 .01 (.11)
1943-1952 -. 12 (.10)
1953-1962 -.21"‘ (.10)
1963-1972 -.11 (.11)
1973-1982 .10 (.16)
Evangelical Periods

1984 -.05 (.08)
1988 .05 (.08)

136

Table 6.7 (cont’d).

1992 .0 1 (.08)
1996 -.09 (.08)
2000 .01 (.09)
Constant 1 .33"‘ (.09)
Observations 1 1363

R2 .23

 

‘p<.05(two-tail test), “p<.01 (two-tail test)

Notes: Entries are regression coefﬁcients with standard errors in parentheses.
The 1903-1912 cohort is dropped as the baseline variable.

The 1980 period variable is dropped as the baseline variable.

Figure 6.7. Abortion Support by Period, Cohort, Campaign Participation and
Evangelical Tradition: NES 1980—2000.l

 

 

 

 

 

 

 

 

 

 

’e a-

E
m 4 _________________ m _ ______________
N - N 4
3 3
I3 I I I I I I .3 I I I I I f I l

80 84 88 92 96 00 1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82
Election Year Ten Year Birth Cohorts
fl»-
E
M _ _________________
N -
3 d
'3 I I I I l I
Low 2 3 4 5 High
Campaign Participation
Evangelical ----- Other

 

 

 

 

Note: OLS data is from Table 6.7.
1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

137

6.8 GSS Moral, Political and Tolerance Attitudes

The ﬁnal set of models use the General Social Survey to look at moral, political
and tolerance questions unavailable in the NES (See Appendix J for variable
descriptions). These models do not examine the effects of participation, as the GSS does
not contain a set of campaign participation variables similar to the NES. The control
variables for these models are similar to the NES models and include gender, education,
income, party identiﬁcation and political ideology. Peculiarities of the state coding in the
GSS make using a “sou ” variable similar to the NES impossible. Religious variables
are comprised of dummy variables for religious traditions and church attendance.
Evangelical campaign participation and Evangelical church attendance are included as
interaction terms. Cohorts are coded the same as the NES using 10-year birth cohorts
beginning in 1903 and running through 1982. Periods for the GSS are more numerous as
the GSS is conducted nearly every year from 1974 to 1990 and then every two years
following. Interaction terms for Evangelical periods and cohorts are included.

The ﬁrst set of models (Table 6.8) looks for change in the morals of Evangelicals
by examining sexual attitudes. Secularization triggered by political participation should
result in a closing of the gap between Evangelicals and Non-Evangelicals. The trends for
periods and cohorts (Figures 6.8 and 6.9) show little support for this hypothesis. While
Evangelicals are moving in a more liberal direction concerning homosexuality, the gap
between Evangelicals and others has actually widened in both the period effects and
cohort effects. Evangelicals are increasing their acceptance of homosexuality at a much
slower rate than the rest of the population. Premarital sex reveals a parallel trend between

the two groups. Evangelical cohorts, while starting at a lower rate of acceptance, have

138

increased their approval of premarital sex at rates similar to Non-Evangelical cohorts.
The same result also holds for premarital sex by periods. Both teen sex and extramarital

sex show a trend of near agreement on the undesirability of both.

Table 6.8. Regression of Periods and Cohorts on Sexual Attitudes: GSS 1974—2002.

 

 

Homosexuality Extramarital Sex1 Teen Sex2 Premarital Sell1
Birth Cohorts
1913-1922 .02 (.04) .08" (.02) .20" (.04)
1923-1932 .08" (.04) .12" (.02) .03 (.03) .32“ (.04)
1933-1942 .1 I“ (.04) .21" (.02) .1 1" (.03) .57" (.04)
1943-1952 .34" (.04) .24“ (.02) .17" (.03) .80" (.03)
1953-1962 .28" (.04) .15" (.02) .27" (.03) .88" (.03)
1963-1972 .39“ (.04) .07" (.03) .40“ (.03) .86" (.04)
1973-1982 .63“ (.05) .09'”I (.03) .49" (.04) .90" (.05)
Periods
1974 .18" (.05) -.O3 (.03) .09 (.05)
1975 .10 (.05)
1976 .1 1* (.05) .06* (.03)
1977 .10" (.05) .00 (.03) .19" (.05)
1978 .18" (.05)
1982 .00 (.04) -.05 (.03) .22" (.05)
1983 .00 (.00) .00 (.00) .25" (.06)
1984 -.02 (.05) -.03 (.03)
1985 -.07 (.05) -.08** (.03) .21" (.05)
1986 .06 (.04) .21" (.05)
1987 -.14""" (.05) -.09" (.03) .00 (.00)
1988 -. l 9’“I (.05) -.21"'* (.03) -.01 (.04) .08 (.06)
1989 -.14** (.05) -.18""" (.03) -.02 (.04) .06 (.06)
1990 -.15" (.05) -.18" (.03) .00 (.00) .1 1 (.06)
1991 -.10 (.05) -.14" (.03) .03 (.04) .14“ (.06)
1993 .10 (.05) -.19** (.03) -.04 (.04) .14“ (.06)
1994 .07 (.04) -.22" (.03) -.05 (.04) .09 (.05)
1996 .22" (.05) -.24" (.03) -.06 (.04) .09 (.05)
1998 .26" (.05) -.26" (.03) -. 19* (.04) -.01 (.05)
2000 .21" (.05) -.24" (.03) -.19* (.04) -.10* (.05)
2002 .28" (.05) -.31** (.03) -.18* (.04) .00 (.00)
Gender .20" (.01) -.07" (.01) -.14""" (.01) -.15“ (.01)
Education .27" (.01) .09" (.01) .06" (.01) .09" (.01)
Income .04" (.01) -.01 (.00) -.02"”" (.01) .03" (.01)
Ideology -.16** (.01) -.05 " (.00) -.08" (.00) -.12""' (.01)
Party ID -.15 (.01) -.00 (.00) -.00 (.00) -.01"”" (.00)
Religious -.09*‘ (.00) -.05" (.00) -.05** (.00) -.15** (.00)
Attendance
Evangelical -.26" (.10) -. 12 (.06) -.09 (.08) -.21 (.1 1)
Mainline -.27" (.03) -.1 1" (.02) -.04 (.03) .13" (.03)
Black -.50*" (.03) .05“ (.02) .01 (.03) .15" (.03)
Protestant
Catholic -.15" (.03) -.04’ (.02) .00 (.03) .27"”'I (.03)

139

Table 6.8 (cont’d).

Seculars .04 (.03) .10* (.02) .22' (.03) .12" (.03)

Interaction

Terms

Evangelical X .01 (.01) .01‘ (.00) -.00 (.01) -.04* (.01)

Church

Attendance

Evangelical

Birth Cohorts
1913-1922 -.02 (.08) -.06 (.05) -. 12 (.08)
1923-1932 .00 (.08) -.08 (.05) .01 (.06) .01 (.07)
1933-1942 -.08 (.07) -.15*" (.05) -.09 (.06) -.14"‘ (.07)
1943-1952 -.29** (.07) -.18” (.05) -.06 (.05) -.14"' (.07)
1953-1962 -.22” (.07) -.06 (.05) -.13 "‘ (.05) -.10 (.07)
1963-1972 -.25** (.08) .00 (.05) -.19" (.06) -.06 (.08)
1973-1982 -.26* (.12) -.02 (.08) -.26“ (.08) -.22 (.12)

Evangelical

Periods
1974 -.21* (.10) -.05 (.06) .22 (.1 1)
1975 .28* (.12)
1976 -.17 (.10) -.11 (.06)
1977 -.13 (.10) -.14"' (.06) .17 (.11)
1978 .11 (.11)
1982 -.05 (.09) -.07 (.06) .08 (.1 1)
1983 .
1984 -.07 (.10) -.O4 (.06)
1985 -.11 (.09) -.01 (.06) .16 (.1 '1)
1986 .06 (.08) .17 (.1 1)
1987 -.01 (.09) -.05 (.06) .00 (.00) .
1988 .01 (.1 1) .05 (.07) .03 (.08) .31" (.12)
1989 .04 (.1 1) .02 (.07) .03 (.08) .34" (.12)
1990 -.08 (.1 1) -.01 (.07) .00 (.00) .26“ (.12)
1991 -.16 (.1 1) -.04 (.07) -.04 (.08) .18 (.12)
1993 -.07 (.10) .03 (.07) .03 (.08) .08 (.12)
1994 -.12 (.09) .06 (.06) .0'1 (.07) .19 (.11)
1996 -.28" (.09) .07 (.06) .01 (.07) .19 (.1 1)
1998 -.25" (.09) .05 (.06) .14 (.08) .18 (.1 1)
2000 -.22"' (.10) .07 (.06) .1 1 (.08) .28* (.1 1)
2002 -.31“ (.12) .10 (.07) .10 (.09) .09 (.13)

Constant 1.66" (.06) 1.73" (.04) 1.94* (.06) 2.98"

(.07)
Observations 23706 24594 14191 23378
R2 .25 .11 .15 .31

 

‘p<.05 (two-tail test), "p<.01 (two-tail test)

Notes: Entries are regression coefﬁcients with standard errors in parentheses.

lT'he 1903-1912 cohort is dropped as the baseline variable and the 1972 period variable is dropped as the
baseline variable.

2The 1985 period variable is dropped as the baseline variable and The 1913-1922 cohort is dropped as the
baseline variable.

140

ﬁgure 6.8. Sexual Morals by Period and Evangelical Tradition: GSS 1974—2002.l

 

 

 

 

 

 

 

 

 

 

 

 

 

NJ Cl)
8: I:
g Homosexuality g Extramarital Sex
’5 5 *5 "
Z Z
4 ———————— 4
m ———— w —————————
1: ————————
at i.

3 l r r r l r r r 3 r r r l r r l l
74 78 82 86 90 94 98 02 74 78 82 86 90 94 98 02
Year Year

DD 00
5 T s E P 'tal 8
6611 ex I‘CI'I‘IaI'I 8X
3 , a -
’5 ’5
Z Z
_ .1 ________________
- 4
no _________________ on
C G
2 “ e a
3 r I l 1 l r I r 3 I l l r l r r r
74 78 82 86 9O 94 98 02 74 78 82 86 9O 94 98 02
Year Year
Evangelical ----- Other

 

 

 

 

Note: OLS data is from Table 6.8.

1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who

wins, Party contact and Religious identity.

141

re 6.9. Sexual Morals by Cohort and Evangelical Tradition: GSS 1974—2002.l

 

 

 

 

 

 

 

 

 

 

 

 

 

E8"
8:” E”
g Homosexuality g Extramarital Sex
*5 ‘ a ‘
Z Z
-i ’— .— ..— ..v ‘‘‘‘ ’ 1
{0:0 a " .- -' —-— g) _________________
8 d ’ 8
3 F l T r I j l l 3 r l I T T ‘1 r T
1903-1213-2223-32 33-42 43-52 53-62 63-72 73-82 1903-1213-2223-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts Birth Cohorts
DD 00
.52: T S 8 P 't IS
8811 ex '- I'CI'I'laI’I a 6X
3 _ 3 _
’5 ’5 _,
Z Z ’ ‘, .— "’ ’-
-1 ‘ ct ’ d’ ’
on ———————— :_ on
r: —————— t:
2 ‘ “’“i 2 1
3 r r r l l l l r 3 l l r l l r I r
1903-1213-22 23-32 33-42 43-52 53-62 63-72 73-82 1903-1213-22 23-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts Birth Cohorts
Evangelical ----- Other

 

 

 

 

Note: OLS data is from Table 6.8.

1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who

wins, Party contact and Religious identity.

A set of models measuring changes in attitudes toward government spending

(Table 6.9) discloses very little difference between groups. Evangelicals are slightly less

favorable toward increased spending for foreign aid, education and the environment and

slightly more favorable toward military spending than Non-Evangelicals (Figures 6.10

and 6.11). The gap toward the latter three categories appears to be widening across the

periods. In keeping with Stimson’s public mood, all four measurements do show a liberal

trend, but Evangelical change is at a lower rate than Non-Evangelicals. Government

spending by cohorts reveals little difference between Evangelicals and Non-Evangelicals.

142

Tablel6.9. Regression of Periods and Cohorts on Government Spending: GSS 1974—
2002.

 

 

Spending on Spending on Spending on the Spending on

Foreign Aid Education Environment Military
Cohorts
1913-1922 -.01 (.02) -.12" (.02) -.09""" (.02) -.07" (.02)
1923-1932 -.06" (.02) -.15“ (.02) -.13*"' (.02) -.08*"‘ (.02)
1933-1942 -.06" (.02) -.22"”" (.02) -.22" (.02) -.01 (.02)
1943-1952 -.07"“" (.02) -.30" (.02) -.34** (.02) .07" (.02)
1953-1962 -.14*"‘ (.02) -.29" (.02) -.38** (.02) .10" (.02)
1963-1972 -.17"”" (.02) -.31"”" (.02) -.41“ (.02) .15" (.02)
1973-1982 -.19** (.03) -.26*"‘ (.03) -.46** (.04) .21" (.04)
Periods
1974 .06" (.02) -.03 (.03) -.24"“" (.03) .67" (.03)
1975 .03 (.02) .04 (.03) -.13*"' (.03) .63" (.03)
1976 .07" (.02) -.02 (.03) -.18""" (.03) .51'”I (.03)
1977 -.00 (.02) .05 (.03) -.04 (.03) .51" (.03)
1978 -.02 (.02) .05 (.03) -.08" (.03) .45“ (.03)
1982 .04 (.02) -.05" (.02) -.04 (.03) .51" (.03)
1983 .06“ (.03) -.10" (.03) -.10" (.03) .59" (.03)
1984 .00 (.04) -.15*" (.04) -.18“ (.04) .70" (.04)
1985 -.08* (.03) -.l4*"' (.03) -.12" (.03) .75” (.04)
1986 .04 (.03) -.12"”" (.03) -.l9** (.03) .72" (.04)
1987 .01 (.03) -.10** (.03) -.22" (.04) .71" (.04)
1988 .04 (.03) -.14"“" (.03) -.22" (.03) .70“ (.04)
1989 .05 (.03) -.21** (.03) -.29** (.03) .69" (.04)
1990 .01 (.03) -.23** (.03) -.29" (.03) .77“ (.04)
1991 .05 (.03) -.15** (.03) -.21** (.03) .58" (.04)
1993 .03 (.03) -.14*"' (.03) -.O6 (.03) .75“ (.04)
1994 .07" (.02) -.17"”" (.03) -.07" (.03) .58" (.03)
1996 .04 (.03) -.15"'* (.03) -.05 (.03) .56" (.03)
1998 -.05 (.03) -.15" (.03) -.07" (.03) .53” (.03)
2000 -.11" (.03) -.l7" (.03) -.05 (.03) .39" (.03)
2002 -.00 (.03) -.19"”" (.03) -.06 (.03) .26” (.03)
Gender -.02"' (.01) -.07" (.01) -.06"”" (.01) .04" (.01)
Education -.02"‘" (.00) -.03" (.00) -.02"“" (.01) .08" (.01)
Income .02" (.00) -.01"‘ (.00) .01“ (.00) -.00 (.00)
Ideology .02“ (.00) .05“ (.00) .06" (.00) -.08" (.00)
Party ID -.00 (.00) .03" (.00) .03" (.00) -.03" (.00)
Religious Attendance -.02*"‘ (.00) .00 (.00) .00 (.00) .00 (.00)
Evangelical .04 (.05) .02 (.05) .03 (.06) -.18" (.06)
Mainline Protestant .08“ (.02) .09" (.02) .05" (.02) -.13"”'I (.02)
Black Protestant -.00 (.02) -.09" (.02) -.04 (.02) -.05"‘ (.02)
Roman Catholic .07“ (.02) .08" (.02) .02 (.02) -.08*"‘ (.02)
Secular .03 (.02) .10" (.02) .02 (.02) .02 (.02)
Interaction Terms
EvangelicalX .01 (.00) .01 (.00) .02“ (.00) -.00 (.00)
Religious Attendance
Evangelical Cohorts
1913-1922 -.00 (.04) .07 (.04) .02 (.05) .11“ (.05)
1923-1932 .05 (.04) -.05 (.04) .00 (.04) .09 (.05)

143

Table 6.9 (cont’d).

1933-1942 .07* (.04) .02 (.04) .06 (.04) -.01 (.05)
1943-1952 .04 (.04) -.01 (.04) .02 (.04) -.02 (.04)
1953-1962 .01 (.04) -.05 (.04) .01 (.04) .00 (.05)
1963-1972 .04 (.05) -.06 (.05) -.OO (.05) .1 1 (.06)
1973-1982 -.09 (.07) -.19* (.07) -.03 (.08) .07 (.08)
Evangelical Periods

1974 -.03 (.05) .08 (.06) .02 (.06) -.12 (.06)
1975 -.O6 (.05) -.04 (.06) -.O4 (.06) .00 (.06)
1976 -.05 (.05) .06 (.06) .07 (.06) .01 (.06)
1977 -.05 (.05) -.05 (.05) -.08 (.06) -.l 1 (.06)
1978 .02 (.05) -.1 1* (.05) -.08 (.06) -.14* (.06)
1982 -.O4 (.05) .02 (.05) -.04 (.06) -. 14* (.06)
1983 .03 (.06) -.03 (.07) -.08 (.07) -.22* (.07)
1984 .05 (.07) -.04 (.08) -.10 (.08) -.12 (.09)
1985 .02 (.06) .06 (.06) -.13* (.07) .01 (.07)
1986 -.08 (.06) .07 (.06) -.01 (.07) -.01 (.07)
1987 -.01 (.07) .00 (.07) -.09 (.08) -.10 (.08)
1988 -.10 (.06) .01 (.07) -.05 (.07) -. 12 (.08)
1989 -.07 (.06) .10 (.07) -.09 (.07) -.00 (.07)
1990 -.09 (.06) .10 (.07) -.05 (.07) -.06 (.08)
1991 .08 (.06) .05 (.07) -.10 (.07) -.1 1 (.07)
1993 .02 (.06) .09 (.06) .00 (.07) -.10 (.07)
1994 -.O4 (.05) .07 (.05) -.03 (.06) -.03 (.06)
1996 .04 (.05) .15" (.06) .06 (.06) -.11 (.06)
1998 .07 (.05) .12" (.06) -.02 (.06) -. 12 (.06)
2000 .09 (.05) .12“ (.06) -.06 (.06) -.05 (.07)
2002 -.02 (.07) .22" (.07) .08 (.08) -.01 (.08)
Constant 266* (.03) 1.69* (.04) 1.69" (.04)

Observations 221 16 22509 22 137 22003

R2 .04 .09 .10 .13

 

‘p<.05 (two-tail test), *p<.01 (two-tail test)
Notes: Entries are regression coefﬁcients with standard errors in parentheses.

1The 1903-1912 cohort is dropped as the baseline variable and the 1980 period variable is dropped as the
baseline variable.

144

Figure 6.10. Government Spending by Period and Evangelical Tradition: GSS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1974—2002}
§ Foreign Aid g Education Spending
3 1 3 -1
i """"=“' 8
E- S 1
m m
8’ 0 ~ " — -- rd: ______
i. 31
g l r 1 T T r l f g r I l l l T r
74 78 82 86 90 94 98 02 "‘ 7 82 86 90 94 98 02
Year Year
{5 Environmental Spending :3: Military Spending
2 - 23 -
8 a
o o _______________
5- S 4
w _ m
8 ““““““““ 8
2 ~ 2 4
g r I r r I T r l g l r r f I r
74 78 82 86 90 94 98 02 82 86 90 9 98 02
Year Year
Evangelical ----- Other

 

 

 

 

Note: OLS data is from Table 6.9.
1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

145

Figure 6.11. Government Spending by Cohort and Evangelical Tradition: GSS

 

 

 

 

 

 

 

 

 

 

 

 

 

1974—2002.l

g Foreign Aid § Education Spending

r. 7 8 —

_ _____ O

8 ———————— 8

i E

m m \N
g r I f 1 1 T r r E r r r r I r r r

1903-1213-22 23-32 33-42 43-52 53-62 63-72 73-82 1903-1213-2223-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts Birth Cohorts

“g Environmental Spending § Military Spending

0 0

g -1 1— 1

. i ........

54 g . ————————— r“
m m __

0 0

.3- 8 -

.‘5’ r r r l r T l f g I I I 1 m I l l

1903-1213-22 23-32 33-42 43-52 53-62 63-72 73-82 “1903-1213-2223-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts Birth Cohorts
Evangelical ----- Other

 

 

 

 

Note: OLS data is from Table 6.9.

1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who

wins, Party contact and Religious identity.

A change in the probability of expressing tolerance is examined in the last three

sets of models. The logistic regressions look at support for keeping an out-group book in

the library, allowing a speech by an out-group member and keeping a college teacher

who is affiliated with an out-group. The ﬁrst set of models looks at atheists (Table 6.10),

the second at homosexuals (Table 6.11) and the third at communists (Table 6.12). All of

the dependent variables are coded so that 1 equals tolerant and 0 equals intolerant.

146

Table 6.10. Logistic Regression of Periods and Cohorts on Tolerance of Atheists:
GSS 1974-2002.l

 

 

Atheist Book Atheist Speech Atheist Teacher
Cohorts
1913-1922 .28" (.08) -.31" (.08) -.31"‘ (.09)
1923-1932 .44" (.08) -.63* (.08) -.73* (.08)
1933-1942 .65" (.08) -.88* (.08) -1.07* (.08)
1943-1952 1.00" (.08) -1.23* (.08) -l.51* (.08)
1953-1962 .91" (.08) -l.21"‘ (.08) -1.56* (.08)
1963-1972 .93" (.09) -1.12* (.09) -1.50* (.09)
1973-1982 1.12” (.12) 4.30” (.13) -1.53* (.12)
Periods
1974 .00 (. 10) .02 (.10) .03 (. 10)
1976 -.12 (.10) -.04 (.10) .17 (.10)
1977 -.06 (.10) .04 (.10) .25" (.10)
1982 -.03 (.10) .14 (. 10) .07 (.09)
1984 -.05 (.10) .10 (.1 1) .20" (. 10)
1985 -.19 (.10) .16 (.10) .20" (.10)
1987 -.03 (.10) .07 (.10) .26“I (.09)
1988 -.09 (.12) -.14 (.12) .30" (.11)
1989 -.00 (.12) -.03 (.12) .07 (.1 1)
1990 .05 (.12) -.16 (.13) .07 (.11)
1991 .19 (.12) -.21 (.13) .08 (.11)
1993 -.07 (.12) .11 (.12) .18 (.11)
1994 -. 12 (.10) .16 (.10) .33" (.09)
1996 -.26"”" (.10) .21" (.10) .16 (.10)
1998 -.16 (.10) .04 (.1 1) .07 (.10)
2000 -.36** (.10) .19 (.10) .22“ (.10)
2002 -.13 (.12) .17 (.13) .15 (.11)
Gender .00 (.03) .16“ (.03) .12" (.03)
Education .66" (.02) -.61** (.02) -.57** (.02)
Income .15" (.01) -.14** (.01) -.07"”" (.01)
Ideology -.11" (.01) .08" (.01) .14" (.01)
Party ID .03“ (.01) .03" (.01) .01 (.01)
Religious -.1 1" (.01) .08” (.01) .07" (.01)
Attendance
Evangelical -.72** (.20) .63" (.20) .63" (.23)
Mainline -.34" (.07) .35" (.07) .29" (.06)
Protestant
Black Protestant -.81** (.08) .77“ (.08) .54" (.07)
Roman Catholic -.27** (.07) .31" (.07) .15" (.06)
Secular .19“ (.09) -.21 "' (.09) -.32"' "‘ (.07)
Interaction Terms
Evangelical X -.03* (.01) .01 (.01) .01 (.01)
Religious
Attendance
Evangelical
Cohorts
1913-1922 -.13 (.16) .24 (.16) .37 (.20)
1923-1932 -.17 (.16) .27 (.15) .45“ (.19)
1933-1942 -.18 (.16) .23 (.15) .33 (.18)

147

Table 6.10 (cont’d).

1943-1952 -.27 (.15) .25 (.15) .29 (.18)
1953-1962 -.18 (.15) .31“ (.15) .03 (.18)
1963-1972 -.35* (.17) .42“ (.17) .16 (.19)
1973-1982 .00 (.27) .45 (.26) .09 (.27)

Evangelical

Periods
1974 -.1 1 (.20) .20 (.20) -. 10 (.21)
1976 .24 (.20) .08 (.20) -.24 (.21)
1977 -.07 (.20) .04 (.20) -. 12 (.21)
1982 -.37 (.19) -.10 (.19) -.04 (.20)
1984 .14 (.20) -.32 (.20) -.25 (.20)
1985 -.07 (.19) -.01 (.19) -. 16 (.20)
1987 .33 (.19) -.31 (. 19) -.41* (.19)
1988 .08 (.22) . 12 (.22) -.16 (.23)
1989 .26 (.22) -.37 (.22) -.30 (.22)
1990 -.09 (.23) .02 (.23) -.25 (.23)
1991 -.08 (.22) .35 (.22) -.18 (.22)
1993 .23 (.21) -.25 (.22) -.35 (.21)
1994 .26 (.19) -.38* (.19) -.58** (.19)
1996 .38“ (.19) -.45* (.19) -.22 (.19)
1998 .33 (.19) -.16 (.20) -.41* (.20)
2000 .57" (.20) -.54" (.20) -.67*"' (.20)
2002 .37 (.24) -.58* (.25) -.57"' (.24)

Constant -.56" (.14) .48" (. 14) 1.29" (.14)

Observations 24336 247 15 24091

2
Pseudo- R

 

*p<.05 (two-tail test), *p<.01 (two-tail test)
Notes: Entries are logistic regression coefﬁcients with standard errors in parentheses.

'rhe 1903-1912 cohort is dropped as the baseline variable and the 1980 period variable is dropped as the
baseline variable.

All three sets of models show period and cohort increases in tolerance by
Evangelicals. As Evangelical belief and flmdamentalism have been found to decrease
tolerance for out-groups, these trends represents a signiﬁcant shift for Evangelicals
(F ilsinger 1976; Laythe, Finkel and Kirkpatrick 1999; Laythe et a1 2002; Karpov 2002)").
An additional study (Wilson 1994) previously found a positive trend for cohorts but did

not include religious variables. The period effects, though, for Evangelicals demonstrate

 

'° It should be noted that the Laythe, Finkel and Kirkpatrick (1999) and Laythe et a1 (2002) studies would
not pick up a period or cohort effect as each study sampled only college students. The Karpov study uses
two survey years of the GSS (1991 and 1998) to examine changes in tolerance. However, Karpov uses the
GSS’s own coding of fundamentalist, moderate and liberal respondents rather than a more speciﬁc
measurement of religious traditions.

148

the largest effect in closing the tolerance gap. All nine period trends (Figures 6.11, 6.13
and 6.15) show Evangelical identiﬁers increasing their probability of tolerance over time,
and with a higher slope of increase than Non-Evangelicals. The cohort trends (Figures
6.12, 6.14 and 6.16) show a decline in the gap between groups but at a lower rate than the
period trends.

Figurle 6.12. Tolerance of Atheists by Period and Evangelical Tradition: GSS 1974-
2002.

 

 

 

 

 

Atheist Book Atheist Speaker

Q —l ________ Q .1 ___________
‘2 ~ “I -
61 -1 N 1 '
c - c -l

I I I I j j I I I I I I I T r I

74 78 82 86 90 94 98 02 74 78 82 86 90 94 98 02
Year Year
Atheist Teacher

Q ..l
(o. _ ____________
.. :;//
Q .1
o

I I I ﬂ I I

 

 

 

 

 

Evangelical ----- Other

 

 

Note: Predicted probabilities are from Table 6.11.
|Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

149

Figurle 6.13. Tolerance of Atheists by Cohort and Evangelical Tradition: GSS 1974—
2002.

 

 

 

 

 

 

 

 

 

 

Atheist Book Atheist Speaker
.39 _ ‘— __, .2 ,,,,, 09 _ '— .. __ ,,,,,,
‘9 - """""" . ’’’’’’’
a. 4 -
6! -1
O - O '5
I I I I I I I I I I I I I I I j
1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82 1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts Birth Cohorts
Atheist Teacher
0°. ..
\q ..
Q: -.
0! -t
c J
I j I I T I I
1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts
Evangelical ----- Other

 

 

 

 

Note: Predicted probabilities are from Table 6.11.
IControlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

Table 6.11. Logistic Regression of Periods and Cohorts on Tolerance of
Homosexuals: GSS 1974-2002.l

 

 

Homosexual Book Homosexual Speech Homosexual Teacher
Cohort
1913-1922 .39" (.08) -.51** (.08) -.37“ (.08)
1923-1932 .65" (.08) -.68** (.08) -.51"”" (.08)
1933-1942 .77" (.08) -.84" (.08) -.81" (.08)
1943-1952 1.08" (.07) -1.05*"' (.08) -1.19" (.08)
1953-1962 1.08“ (.07) -1.02" (.08) -1.14** (.08)
1963-1972 1.05" (.09) -.93 " (.09) -1.21** (.09)
1973-1982 1.50" (.13) 4.26” (.15) 4.58" (.14)
Periods
1974 .05 (.10) .1 1 (.11) .06 (. 10)
1976 -.04 (.10) .19 (.1 1) .08 (.10)
1977 .10 (.10) .11 (.10) .10 (.10)
1982 -.09 (.09) .06 (. 10) -.02 (.09)

150

Table 6.11 (cont’d).

1984 -.05
1985 -. 17
1987 -.20"
1988 .05
1989 .04
1990 .08
1991 .41 "' "'
1993 . 17
1994 .14
1996 .06
1998 .27"
2000 .16
2002 .34"
Gender .22"
Education .62"
Income .17"
Ideology -.12* "
Party ID .01
Religious -.1 1"
Attendance
Evangelical -.69"
Mainline -.17“"‘I
Protestant

Black Protestant -.43 "
Roman Catholic -.02

Secular .08

Interaction

Terms

Evangelical X -.03*

Religious

Attendance

Evangelical

Cohorts
1913-1922 .13
1923-1932 -.06
1933-1942 .01
1943-1952 -.06
1953-1962 -.05
1963-1972 -.04
1973-1982 .10

Evangelical

Periods
1974 -.16
1976 .27
1977 -.37
1982 -.1 l
1984 . 15
1985 -.29
1987 .17
1988 -. 14
1989 .28

(10)
(10)
(10)
(12)
(12)
(12)
(12)
(12)
(10)
(10)
(10)
(10)
(13)

(03)
(02)
(01)
(01)
(01)
(01)

(.21)
(.07)

(08)
(07)
(08)

(.01)

(18)
(17)
(17)
(16)
(16)
(18)
(28)

(20)
(20)
(20)
(19)
(20)
(19)
(19)
(22)
(22)

.08
.05
.06
a13
a33‘
a34*
240“
a32‘
n36**
-.34"”'I
n55**
a3l*'
n37**

-.25"

-.66**

-.18**
.14"
.00
.10"

.73"
.16"

.35""'I
-.09
-.15

.06"

-.05
-.10
-. 10

-.10
-.05
.05

-. 16
-.33
-.04
.01
-.48"‘
.05
-.26
.06
-.27

151

(11)
(11)
(11)
(13)
(13)
(14)
(14)
(13)
(11)
(11)
(12)
(12)
(14)

(03)
(02)
(02)
(01)
(00)
(01)

(21)
(08)

(09)
(08)
(10)

(.01)

(17)
(17)
(16)
(16)
(16)
(18)
(29)

(21)
(21)
(20)
(20)
(20)
(20)
(19)
(23)
(23)

n07
«03
.13
.07
n16
«17
«06
n36“
n33“
a51"
-.56"”'I
--.49""I
m55**

-.32**
_.62*‘
-.l4"
.19"
-.03"
.09"

.91**
.24**

.30“I
-.10
-.07

.08"

-.21
-.33
-.25
-.18
-.44"‘
-.28
-.37

.01
-.27
.22
.24
-. 13
.04
-.28
-.04
-.35

(10)
(10)
(10)
(12)
(12)
(12)
(12)
(12)
(10)
(11)
(11)
(11)
(13)

(03)
(02)
(01)
(01)
(01)
(01)

(22)
(07)

(08)
(07)
(09)

(01)

(19)
(19)
(18)
(18)
(18)
(20)
(30)

(21)
(21)
(21)
(20)
(20)
(20)
(19)
(23)
(22)

Table 6.11 (cont’d).

1990 -.02 (.23) .14 (.24) -.03 (.23)
1991 -.22 (.22) .07 (.23) -.29 (.22)
1993 .07 (.21) -.47* (.23) -.14 (.22)
1994 .1 1 (.19) -.29 (.20) -.26 (. 19)
1996 .16 (.19) -.40 (.21) -.23 (.20)
1998 -.16 (.19) .09 (.21) -.00 (.20)
2000 .06 (.20) -.44"‘ (.21) -.47* (.21)
2002 .05 (.24) -.62* (.27) -.47 (.25)
Constant -1.32** (.14) 1.04" (. 15) 1.34" (.14)
Observations 24262 24278 24074
2
Pseudo-R

 

*p<.05 (two-tail test), *p<.01 (two-tail test)
Notes: Entries are logistic regression coefﬁcients with standard errors in parentheses.

1The 1903-1912 cohort is dropped as the baseline variable and the 1980 period variable is dropped as the
baseline variable.

Figure 6.14. Tolerance of Homosexuals by Period and Evangelical Tradition: GSS
1974—2002}

 

 

 

 

 

Homosexual Book Homosexual Speaker

_ -- II-i I1
Q d —————— “i _ ‘‘‘‘‘‘‘‘‘‘‘‘‘‘
‘0. "' ——————— \D' ‘ -;-://
V: - // V. '-
01 - N -
O - c ..

I I I W I I l I I I I I I I I I

74 78 82 86 90 94 98 02 74 78 82 86 90 94 98 02
Year Year

Homosexual Teacher

_-

a v“,-
.-‘ ﬁ-‘v’v’a

we‘ -’————"—

V:—

(‘1—

c-

 

 

I T I I I I j

74 78 82 86 90 94 98 02

Year

 

 

 

Evangelical ----- Other

 

 

Note: Predicted probabilities are ﬁ'om Table 6.12.
1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

152

Figure 6.15. Tolerance of Homosexuals by Cohort and Evangelical Tradition: GSS

1974—2002.l

 

Homosexual Book

...-
e94
~94
1...
Q-
o-r

—’
”
..—
t
”a,
ﬂu.”
—’
f
’

 

"

 

I l I I I I

I
l903-l2 13-22 23-32 33-42 43-52 53-62 63-72 73-82

Birth Cohorts

Homosexual Teacher

......
06.4
q.
-:r_.
or.

 

°-I

I

 

 

I I I I I I I I
l903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82

Birth Cohorts

 

 

 

 

Evangelical

Homosexual Speaker

«2 - ,,,,,,,,,
q _ -/
V; ..
6! ..
o q

I I I I I I I I

1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts

----- Other

 

 

 

Note: Predicted probabilities are ﬁ'om Table 6.12.
1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who

wins, Party contact and Religious identity.

Table 6.12. Logistic Regression of Periods and Cohorts on Tolerance of

Communists: GSS 1974—2002.

 

Ban Communist Book Communist Speech Communist Teacher
Birth Cohorts
1913-1922 .29" (.08) -.38" (.08) .24" (.08)
1923-1932 .47" (.08) -.56" (.08) .46" (.08)
1933-1942 .61" (.08) -.71"‘* (.08) .66" (.08)
1943-1952 .88" (.07) -.91" (.07) 1.04" (.08)
1953-1962 .95" (.08) -.94" (.07) 1.12“ (.08)
1963-1972 .93“ (.09) -.87" (.08) 1.22“ (.08)
1973-1982 1.17” (.12) -1.03** (.12) 1.20" (.11)
Periods
1974 .28" (.10) -.28"‘* (.10) .28" (.10)
1976 .02 (. 10) -.03 (.10) .14 (.10)
1977 .07 (.10) -.15 (. 10) .06 (.09)

153

Table 6.12 (cont’d).

1982
1984
1985
1987
1988
1989
1990
1991
1993
1994
1996
1998
2000
2002

Gender
Education
Income
Ideology

Party ID
Religious
Attendance
Evangelical
Mainline Protestant
Black Protestant
Roman Catholic
Secuhu

Interaction Terms
Evangelical X
Religious
Attendance

Evangelical Birth

Cohorts
1913-1922
1923-1932
1933-1942
1943-1952
1953-1962
1963-1972
1973-1982

Evangelical
Periods
1974
1976
1977
1982
1984
1985
1987
1988
1989
1990

.01
-.04
-.15
-.06
-.10
-.08

.04

.27*

.07
-.08
-.20"‘
-.O6
-.22*
-. 13

405
.70**
.15**

410**

.O3****

-.07*"'

-.69**
-.43**
-.90“"I
-.42**
.06

-.05"""

-.18
-.39*
-.20
-.25
-.26
-. 19
-.O9

-.3 l
.24
-.22
-. 14
.18
.09
.35
.20
.42
.2 l

(10)
(10)
(10)
(10)
(11)
(12)
(12)
(12)
(12)
(10)
(10)
(10)
(10)
(12)

(03)
(02)
(01)
(01)
(01)
(01)

(21)
(07)
(08)
(07)
(09)

(.01)

(17)
(16)
(16)
(15)
(15)
(17)
(26)

(20)
(20)
(20)
(19)
(20)
(19)
(19)
(22)
(22)
(23)

-.06
-.03
.05
.01
-. 15
-. l 3
-.17
-.45**
-. l 8
-.07
.14
-.06
.1 1
.03

.24"
-.69**
-.13**

.12"

.03****

.06”

.98"
.53"
.79"
.47"
-.07

.03"

.03
.21
.06
.03
.l I
.08
.33

.03
-.l l
.01
.09
-.29
-.15
-. 19
.17
-.48*
-.35

154

(09)
(10)
(10)
(10)
(11)
(11)
(12)
(12)
(11)
(10)
(10)
(10)
(10)
(12)

(03)
(02)
(01)
(01)
(01)
(01)

(21)
(07)
(08)
(07)
(08)

(.01)

(17)
(16)
(16)
(15)
(16)
(17)
(25)

(20)
(20)
(20)
(19)
(19)
(19)
(19)
(22)
(22)
(23)

.19“
.15
.10
.10
.20
.34”
.28*
.41"
.44*#
.21“
.25"
.27Ml
.18
.19

-.14**
.49"
.06M
-.14'Ml
.01
”04"

-.13
-.33**
-.42**
-.21"”'I
.08

-.O3 "'

-.31
-.36*
-.38*
-.34"'
-.19
-.28
-. 12

-.41
-.06
-.35
-.53"
-. 14
-.40"‘
-.24
-.35
-.26
.06

(09)
(10)
(10)
(09)
(11)
(11)
(11)
(11)
(11)
(09)
(09)
(09)
(09)
(11)

(03)
(02)
(01)
(01)
(01)
(01)

(21)
(06)
(07)
(06)
(07)

(.01)

(18)
(17)
(17)
(16)
(16)
(18)
(26)

(21)
(20)
(20)
(20)
(20)
(20)
(19)
(22)
(22)
(22)

Table 6.12 (cont’d).

1 991
l 993
1 994
l 996
1 998
2000
2002

Constant

Observatigns
Pseudo-R

.03
.45*
.35
.43"I
.31
.43*
.42

-.88**

24180

(.22)
(.22)
(.19)
(.19)
(.19)
(.20)
(.24)

(.14)

.16
-.60**
-.41 *
-.43 *
-.24
-.60"‘*
-.61 *

.64*#

24440

(.22)
(.21 )
(.18)
(.19)
(.19)
(.19)
(.24)

(.14)

-.28
.00
.17
.07
.05
.29
.37

-1.08**

23610

(.22)
(.21)
(.18)
(.19)
(.19)
(.19)
(.23)

(.13)

 

*p<.05(two-tail test), *p<.01 (two-tail test)
Notes: Entries are regression coefﬁcients with standard errors in parentheses.
The 1903-1912 cohort is dropped as the baseline variable

1The 1980 period variable is dropped as the baseline variable.

Figure 6.16. Tolerance of Communists by Period and Evangelical Tradition: GSS

 

 

 

 

 

 

 

 

1974-2002}
Communist Book Communist Speaker
09 - _ __ __ _ °°. ‘ ... ..
we - ———————————— e - -------------
v. - // er. - ———-//——
‘3 - e:
O - G
I I I I I I I I I I I I I I
74 78 86 90 94 98 02 74 78 82 86 90 94 98 02
Year Year
Communist Teacher

Q -1
~42 d ___________
e- V/
q d
a —1

I I I I I I

74 78 86 90 94 98 02

Year
Evangelical ----- Other

 

 

 

 

Note: Predicted probabilities are from Table 6.13.
lControlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

155

Figure 6.17. Tolerance of Communists by Cohort and Evangelical Tradition: GSS

 

 

 

 

 

 

 

 

1974—2002}
Communist Book Communist Speaker
on 4 e— .1
ea 4 ___ __. .. -- -- " “' 99 a ——————
‘0. '1 ... av ﬁﬁﬁﬁﬁﬁ \O. 'l _— _— --' "' ﬂﬂﬂﬂﬂﬂ
e. - e. - '—
N, 4 or _
O -' O -'
I I F j I I r I I I I I I I l I
l903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82 1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts Birth Cohorts
Communist Teacher
"" '1
”i ‘ .— ’
‘0. "‘ ..1- ’ —' .l- ‘I’ a
V: "' _. —— “" i ’- ’ ’
(N! ..
o
I I I I I I I I
1903-12 13-22 23-32 33-42 43-52 53-62 63-72 73-82
Birth Cohorts
Evangelical ----- Other

 

 

 

 

Note: Predicted probabilities are ﬁ'om Table 6.13.
1Controlling for Gender, Education, Income, Religious attendance, Strength of partisanship, Care who
wins, Party contact and Religious identity.

6.9 Conclusions

Examination of period and cohort trends for campaign participation and voting
ﬁnds split support for Hypotheses 1 and 2. Campaign participation reveals 1980 to be a
crucial year that began the Evangelical trend of change, while voting shows no such
trend. Younger Evangelical cohorts have increased both their rate of voting and campaign

participation to match or exceed younger Non-Evangelical cohorts.

156

The liberalizing effect of campaign participation receives mixed support as
participation increases liberal views of abortion but decreases support for environmental
issues and feminism. The “direction” of the participation may be an intervening factor in
the effect of participation as increasing GOP identiﬁcation also reduces support for
environmental issues and feminism. Divergent viewpoints from campaign participation
only change respondents if they truly are divergent. The period effects also failed to show
a consistent liberal trend. The pattern of Evangelical birth cohorts reveals a liberal trend
for nearly all of the models, though liberalizing effects on sexual issues are occurring at a
much lower rate than for the Non-Evangelical population.

The most dramatic conclusion of the chapter is the increasing support of tolerance
for out-groups across Evangelical periods and among younger cohorts. Evangelicals are
beginning to approach Non-Evangelicals in their support for civil liberties. This
reorientation is not limited to one out-group but is across the board for atheists,
homosexuals and communists. As homosexuality is one of the targeted groups of
Evangelical political and social movement organizations, this ability to embrace civil
liberties while at the same time not increasing acceptance of homosexual morality
suggests that a very complicated balancing of moral values and civil norms is taking root
in the Evangelical tradition. Evangelicals are increasingly willing to accept the civil
liberties’ claims of homosexuals while maintaining their historical views of the morality
of homosexuality. The question remains whether they can maintain this balancing act

with increased cultural/political involvement.

157

CHAPTER VII
PARTICIPATION AND NEW PARADIGM CHURCHES

As separatism declined through the end of the twentieth century, there is little
doubt that the cues of Evangelical political and social movement organizations facilitated
much of the new participation. This decline, though, was not the only internal change that
had the potential to affect Evangelical participation and political attitudes. The 19808
brought a dramatic rise in the number of Evangelical political groups and saw more
fundamental change in the nature of the Evangelical church. The rise of a broad range of
“new paradigm churches” that rejected separatism, not only in politics but on a wide
range of cultural fronts (music, dress, technology), became catalysts for change. These
new denominations affected more established Evangelical churches by advocating greater
acceptance of current cultural norms. This chapter looks into the inﬂuence of new
paradigm groups within American Christianity and looks for a connection between

tolerance, moderation, participation and new paradigm characteristics.

7.] Overview and Hypothesis

As Evangelicals embrace political participation, an internal movement may help
to speed the adoption of this activism. New paradigm churches have been hailed as a
“New Reformation,” changing how Evangelicals view and interact with a modern secular
culture. As formerly discussed, these churches are more tolerant of anti-religious views,
more liberal on race and class issues, more focused on racial equality, more involved in
providing social services and more willing to adopt the music, arts, dress and language of

the surrounding culture.

158

The cultural acceptance of these churches raises the possibility that new paradigm
attitudes increase political participation. While the evidence of such a pattern is sparse,
Willow Creek Association clergy demonstrate high levels of political activism. Over a
third (36%) tried to persuade others how to vote, 19 percent gave money to a campaign,
16 percent displayed a campaign sign and 13 percent attended a campaign rally. A
majority (56%) urged their congregations to register or vote and 27% publicly supported
a candidate (Kellsted and Green 2003). Miller (1997) reports that 49 percent of Calvary
Chapel, Vineyard and Hope Chapel attendees often or sometimes try to persuade another
how to vote. Over two thirds (68%) believe that it is important to volunteer time and/or
give money to political causes.

This new openness to the culture and increasing levels of tolerance also suggests
that secularization or political moderation would ﬁnd a faster reception in new paradigrn-
inﬂuenced churches than in more traditional Evangelical churches. The interaction with
divergent viewpoints producing cognitive contamination would appear to be highest for
new paradigm attendees within the Evangelical tradition. Based on this analysis, I
hypothesize that:

H1: As new paradigm churches encourage involvement in an attendee’s culture,
increasing new paradigm attributes among Evangelicals will increase political
partrcrpatron.

H2: As new paradigm churches have a greater tolerance for anti-religious persons and
organizations, increasing new paradigm attributes among Evangelicals will increase
tolerance for Atheists as an out-group.

H3: As new paradigm churches exhibit greater tolerance for moral diversity, increasing

new paradigm attributes among Evangelicals will increase political moderation as
measured by a liberal direction.

159

The one caveat to this last hypothesis is that while new paradigm churches
express and demonstrate greater tolerance for those not like themselves, new paradigm
attendees have been shown to hold traditional Evangelical views on morals. Additionally,
they are more politically conservative than Non-Evangelicals (Miller 1997; Shipley).
Moral and political tolerance may be a norm for interaction with the culture, but internal
tenets for members are traditional Evangelical views. However, these previous studies
did not compare new paradigm moral and political views with those of other Evangelicals

to see if they signiﬁcantly differed.

7.2 Data

The National Congregation Study (“NCS”) was conducted in conjunction with the
1998 General Social Survey. GSS respondents (Total N=2862) who attended a religious
service at least once a year were asked to identify a speciﬁc congregation and its location.
The congregations were then contacted and a representative leader was interviewed.

The result is a survey of 1236 congregations who were nominated for the study by
GSS respondents. Approximately half of the GSS respondents (N =l476) can be matched
to their congregation. The number of congregation attendees being larger than the
number of congregations is a result of larger congregations being identiﬁed by more than
one respondent. The NCS covers facets of church structure, church demographics,
theology, worship structure, internal church programs and activities (including political
activity), and community involvement. For this study, the congregation questions were

matched with the appropriate General Social Survey respondents and their GSS answers.l

 

' I am deeply indebted to Professor Mark Chaves at the University of Arizona for making the combined
GSS-NCS dataset available to me.

160

7.3 Conflicting Religious Traditions and Congregations

The NCS’s administrators produced their own set of denominational codings
separate from the GSS responses. These were based on affiliations obtained from the
individual congregations. The NCS has its own religious-tradition coding based on its
own denominational coding. Neither the NCS denominational coding nor the religious-
tradition coding is used in this study.

Several major differences exist between the NCS traditions coding and the
Steensland et al. coding used in this and previous chapters. First, the NCS traditions
coding divides Black Protestants across all of the Christian traditions including
Evangelicals in opposition to the most commonly used classiﬁcation schemes (Campbell
2003; Kellstedt 1989; Cook and Wilcox 1990). Evangelicals in particular are usually
coded as a white tradition. Furthermore, persons who have behaviorally and attitudinally
been identiﬁed as “Seculars” (low attendance and vague or nonexistent identiﬁcation of a
tradition) by Steensland et a1 (as well as Kellstedt and Wald 1993; Kellstedt et a1 1996;
Kohut et a1 2000) are included in Evangelical, Mainline, Black and Catholic traditions
and the denominational coding by the NCS. This is despite these persons showing that
they hold very distinct political, social and moral attitudes and appear to have little
identiﬁcation with a religious tradition. Respondents with the lowest level of attendance
(once or twice a year), even when reporting “None” for their General Social Survey
religious identiﬁcation, were asked to name the congregation they attended and were
coded into that denomination. For the sake of compatibility with previous chapters and
out of conviction about the necessity of coding Seculars and Black Protestants into

separate categories, Steensland et al’s coding was followed based on the GSS religious

161

variables. Then, since new paradigm churches are a movement within American
Christianity, non-traditional and secular respondents were dropped leaving Evangelicals,
Mainline, Blacks and Catholics (N=1298).

In addition, use of actual congregations versus religious self-identiﬁcation appears
to violate assumptions about “belonging” and the inﬂuence of group identiﬁcation.
Identiﬁcation is not necessarily synonymous with the congregation of current attendance.
Attendance may be a reﬂection of a spousal or family choice or even a lack of available
choices. Previous research has shown the powerful effect that identiﬁcation with a label
has in creating and maintaining group-based attitudes. This research on social
categorization suggests that one’s self-perception as a part of a group insufﬁcient for
respondents to act as a group (Hogg and Abrams 1988; Tajfel 1981, 1982; Turner 1982).
As Hogg and Abrams state, "the group is thus within the individual ..." rather than
existing as an external interaction (p. 19). This self-identiﬁcation has increasingly
become the de facto standard of politics and religion research (Wald and Martinez 2001).

Two comparisons of GSS identiﬁcation and NCS congregations appear to support
previous work on identiﬁcation. In both the GSS and the National Congregation Study,
respondents are asked to identify their religious preference as Catholic, Protestant,
Jewish, Other (with DK in NES) or None (only in the GSS). If they respond “Protestant”
they are asked what church or denomination (GSS “if any”). I recoded both the GSS
religious classiﬁcations variable (“relig”) and the National Congregation Study
classiﬁcations variable (“trad2”) using a very simple three-part identiﬁcation: Protestant,

Catholic and Other.2 I then checked the percentage of matching identiﬁcations between

 

2 For this analysis, those responding “None” (N=35 dropped) for their GSS religious preference were
excluded to maintain continuity between the GSS and NCS. The NCS takes “none” respondents who report

162

the GSS and NCS by each level of religious attendance. Different answers for their GSS
Protestant/Catholic/Other identiﬁcation and for their NCS Protestant/Catholic/Other
affiliation are the case for over one-sixth of the respondents in the database. This
response is somewhat mitigated by attendance as 11% differ at the highest level of
attendance while 18% differ at the lowest level. It still remains that, of those persons who
attend religious services more than weekly, one in nine identify with a tradition that is
different from where they attend.

Further examination of religious identity suggests that the most accurate identity
is that which respondents report, and not the church they attend. Again, using a simple
Protestant, Catholic, and Other triad, I recoded respondents’ reported religious preference
at the age of 16. This I correlated with both their current GSS and NCS
Protestant/Catholic/Other religious preference.3 Table 7.1 shows the results broken down
by attendance. At every level of reported religious attendance, with the exception of
“nearly every week,” the “age 16 preference” is more highly correlated with the GSS
current preference than with the congregation the respondents are currently attending.
The age-l 6 current-GSS correlation is also always higher that the correlation between the
current GSS identiﬁcation and the NCS congregation identiﬁcation. This may be an
artifact produced by the respondents’ recall of their earlier religious preference or it may
show the permanence of early religious identiﬁcation. It would appear to be a mistake to
classify respondents by their NCS denominations if the researcher is looking for the

respondents’ religious identity. It also produces the curious result that the attributes from

 

attending a congregation the previous year and places them into Catholic or Protestant denominations based
on the church they attended.
3 Again the “none” category was dropped ﬁ'om both the Age-16 and current-GSS preferences (N =79

dmpped).

163

one’s current congregation in one religious tradition, and self-identiﬁcation with a
different religious tradition, are working in conjunction for some respondents to affect
political attitudes. While beyond the scope of this chapter, this is a question that deserves
greater future investigation.

Table 7.1. Religious Identity Correlations for GSS Age Sixteen Preferences, Current
GSS Preferences and Current National Congregation Study Preferences.

 

 

 

 

Correlations
Religious Attendance N Age 16 - GSS Age 16 - NCS GSS - NCS
Once or twice a year 152 .74 .66 .63
Several times a year 211 .83 .62 .60
Once a month 143 .72 .53 .62
2-3 times a month 187 .88 .58 .59
Nearly every week 144 .46 .54 .53
Weekly 388 .70 .53 .65
More than weekly 161 .74 .59 .66
Total 1,461 .60 .50 .57

 

Note: all correlations are signiﬁcant at p < .001

7.4 The New Paradigm Scale

New paradigm churches through their music, cultural sensitivity and church
growth have had an expanding inﬂuence on Evangelical Christianity. As examined
previously, their inﬂuence in established Evangelical churches is seen in adoption of
contemporary worship styles, use of the arts in worship, increased informality in dress
and decor, increased use of technology, establishment of weekly support/12-step groups
and home-fellowship groups, increased ministry by the laity with a focus on using
spiritual gifts, changed levels of community involvement, and concern about racial
issues. While new paradigm churches also have the characteristic of younger attendees
and lower education requirements for clergy, the former is a consequence of new

paradigm attributes, and the latter is less likely than other attributes to be exported to or

164

approved by established Evangelical denominations. It is important to note that none of
these groupings of attributes by themselves capture new paradigm inﬂuence and by
themselves may be more inﬂuential in a religious tradition other than the Evangelical
tradition. Mainline Protestants have traditionally used more technology and had a higher
receptivity to the arts than Evangelicals have but with less participatory worship.
Charismatics and Pentecostals (within all four traditions) have had a more expressive
worship style without the concerns of community involvement or therapeutic (support
group) outreaches.

While there are no explicit new paradigm church questions in the NCS, several
church measurements match new paradi gm characteristics. I created an additional new
paradigm scale using sixteen measurements which combine six smaller categories :
informality (laughter or applause during worship), contemporary and participatory
worship (electric guitar, drums, raising hands, jumping, shouting or dancing during
worship), performing arts (skit or dance performance), technology (use of multimedia, a
church website or email), spiritual giﬁs (speaking in tongues, groups for practicing
spiritual gifts, groups meeting for healing)‘, and groups for community involvement
(groups to discuss racial relations, various support groups, groups meeting to assess
community needs). This created a scale of new paradigm characteristics (“NPC”) coded

from 1 to 17 (1 - no NPC characteristics to 17 - all NPC characteristics).5

 

4 An initial scale excluded speaking in tongues as a part of the spiritual gifts dimension. Removal of the
tongues variable reduced the alpha reliability to .69. As many new paradigm churches have charismatic or
Pentecostal tendencies and this emphasis on the experiential is intermingled with many of their attributes,
tongues were included in the scale.

5 Even with this many variables the NCS does not measure every new paradigm attribute. The NCS does
not measure the percentage of congregants involved in ministry which goes hand in hand with lay ministry
in new paradigrn churches. A possible surrogate is questions about spiritual gift usage which can provide a
rough ministry replacement. Also the breakdown of church groups as recorded by the NCS was
problematic. The NCS recorded groups into categories of fellowship-social, religious-spiritual, religious-

165

Table 7 .2 shows the percentage of respondents whose church gave an afﬁrmative
response to each variable. While New Paradigm church inﬂuence is unmeasured among
Black churches, the table shows that Black churches may score high on the paradigm
scale because of the traditionally demonstrative worship in their churches and large
numbers of Black churches in the 19808 adding charismatic and Pentecostal practices to
their churches. Black churches scored the highest percentage of drum and guitar usage
and highest in jumping, laughter, applause and speaking in tongues. The possibility exists
that the new paradigm scale for Evangelicals, Mainline and Catholics represents an
accurate measurement of new paradigm inﬂuence (greater cultural accommodation,
participation and moderation) while the same scale for Black Protestants may represent a

conservative shift arising from traditional Pentecostal inﬂuences.

Table 7.2. Percentage of New Paradigm Church Attributes in Each Christian
Tradition.

 

 

 

Variables Evangelical Mainline Black Roman
Protestant Protestant Protestant Catholic
Drums used durlng mam serv1ce 34% 15% 59% 13%
Electrrc guitar used during mam serv1ce 36% 22% 33% 31%
Adults jumped, shouted or danced durrng 14% 3% 57% 2%
main serv1ce
Applause during mam serv1ce 66% 56% 88% 43%
Congregants ralsed hands during main 47% 27% 85% 53%
servrce
Used overhead projector durrng marn 29% 12% 14% 3%
servrce

 

education-fellowship, religious-fellowship-social, religious-service-fellowship, religious-devotional-
education, homeowners-neighborhood groups, activities in homes, bible-scripture-study-social, and
education-social and others. Identifying the type of small groups/home fellowships germane to new
paradigm congregations was impossible so this attribute was excluded. Support groups were identiﬁable
despite falling into three categories, substance abuse, divorce and support group unspeciﬁed. All substance
abuse and divorce respondents with afﬁrmative responses also gave afﬁrmative responses in the
unspeciﬁed variable so that variable was included.

166

Table 7.2 (cont’d).

Group met to receive or practice gifts of

the spirit in past 12 months 14% 16% 26% 28%
Tgrtlnggrelihipoken during worship in past 19% 6% 45% 20%
ggilgrsmet for physical healing in past 12 36% 47% 37% 50%
35am; [[31:2ng during worship within 25% 30% 42% 29%
Congregation has a website 30% 36% 15% 25%
Congregation uses email 36% 41% 21% 19%
Earsciulpzngg 1:3] gliscuss race relations in 15% 27% 30% 20%
Group met to plan or conduct an

35:33:? of community needs in past 40% 48% 53% 52%
$133131: mmget to support people with speciﬁc 30% 50% 27% 49%
8120:“ Eggsrmed during worship within last 79% 82% 72% 57%
Laughter during main service 82% 86% 77% 58%

 

Note: N varies between 1286 and 1298 for respective variables.

The new paradigm scale was then examined for reliability. The alpha reliability
test for the scale is .73 and the exclusion of any of the variables failed to increase the
alpha level. Exploratory factor analysis was performed on the data to see if its dimensions
matched the categories of the scale (informality, participatory worship, technology,
spiritual gifts, performing arts, community involvement groups). The factor analysis was
performed using principal component analysis with a varimax rotation. The dimensions
were determined by examining the scree plot and the eigenvalues. The exploratory factor
analysis found ﬁve dimensions to the data rather than six (Table 7.3). Inforrnality did not

appear as a separate dimension but instead was split between the participatory worship

167

and performing arts. Dance performance was also split between performing arts and
spiritual gifts. Given charismatic and Pentecostal affinity for banners, ﬂags and dance,
this division has a sound theoretical basis (Ward 2003; Webber 1995). For Pentecostal
churches, a trained dance performance is still an emotional worship expression tied to the
exuberance of the Holy Spirit and spiritual gifts. For less Pentecostal churches, dance
performance is tied to an appreciation of the arts.

Table 7.3. Factor Analysis of New Paradigm Church Attributes. l

Participatory Spiritual Technology Community Performance
Variables Worship Gifts Involvement

 

Drums used during main

. .806
serv1ce

Electric guitar used during
main service

Adults jumped, shouted or
danced during main service

.682

.647

Applause during main

. .61 7
servrce

Congregants raised hands

. . . .615
during mam serv1ce

Used overhead projector

during main service '504 .338

 

 

Group met to receive or
practice gifts of the spirit in .722
past 12 months

Tongues spoken during

worship in past 12 months '426 ~637

Group met for physical
healing in past 12 months '556 357

Dance performed during
worship within last 12 .498 .368
months

Congregation has a website .793

 

 

Congregation uses email .731

 

 

Group met to discuss race 7 1 7
relations in past 12 months '

Group met to plan /conduct
an assessment of community .679
needs in past 12 months

 

 

168

Table 7.3 (cont’d).

Group met to support people

with speciﬁc problems 522

 

Skit performed during
worship within last 12 .796
months

Laughter during main
service .678

 

 

 

Eigenvalue 2.80 1.77 1.59 1.48 1.35
Percentage of Variance 16.5% 10.4% 9.4% 8.7% 8.0%

 

1A" factor loadings above .30 are shown. Extraction Method: Principal Component Analysis. Rotation
Method: Varimax with Kaiser Normalization.

A comparison of the new paradigm scale means by religious tradition (Table 7.4)
reveals that the new paradigm levels for Evangelicals and Mainline Protestants are very
similar. Black Protestants, perhaps because of their worship styles, score signiﬁcantly
higher than the other two Protestant groups, while Roman Catholics show the lowest

average level of new paradigm attributes.

Table 7.4. Anova Test of Means: New Paradigm Scale Means by Religious
Tradition.

 

 

Religious New Paradigm Standard N ANOVA F
Tradition Meanl Deviation

Evangelical 6.39 a 3.31 406 2133*"
Mainline 6.09 e b 2.86 291

Black 7.81 c 3.02 168

Catholic 5.63 b 2.64 429

 

1Means with the same subscript are not signiﬁcantly different from each other (p < .05).
tit
p < .001.

7.5 Methodology: Regression and Logistic Regression

While the N for the four Christian traditions in the NCS is 1298 respondents,
many of the dependent variables under consideration were split ballot questions. This

resulted in many dependent variables with an approximated N of 800. Control variables

169

(See Appendix K for descriptions.) include demographics (education, income, gender,
age, number of children, marital status, and urban), political variables (party
identiﬁcation or strength of partisanship, and political ideology) and religious variables
(religious attendance, and dummy variables for Evangelicals, Mainline and Black;
Catholics are the excluded category). Interaction terms take in the new paradigm scale
and attendance, new paradigm scale and religious traditions, and religious traditions and
attendance.6
An initial set of twenty-seven regression and logistic regression models (not
shown) using tolerance, political, and moral attitudes plus participation as dependent
variables, failed to ﬁnd any signiﬁcant effects for Evangelical X paradigm interaction. A
second set of models was run with an additional set of interaction variables using a three-
way interaction for religious traditions X paradigm scale X church attendance. This tests
for a new paradigm inﬂuence that is mediated by attendance and tradition. Seven models
showed signiﬁcant effects for Evangelicals with increasing attendance in churches with

higher paradigm scores. Five models fall into the category of tolerance, one model is a

moral/political attitude and the last model is a measurement of participation.

7.6 Results

Evangelicals with increased attendance interacting with increasing levels of new

paradigm characteristics raised their probability of tolerance of atheists, homosexuals and

 

6 For an initial set of models, ﬁrst order terms, paradigm and attendance, were centered by subtracting the
variable mean. This has the advantage of reducing colinearity between the ﬁrst order terms and interaction
terms and it also improves the ability to interpret the ﬁrst order terms when adding the interaction terms
(Aiken and West 1991). Despite reducing the VIF scores, no new variables became signiﬁcant (p < .10,
one-tail test) or increased their signiﬁcance level. Given the ongoing debate that centering produces nearly
identical results as non-centered regression and the possibility of systematically biased estimates of the
main effects, the models are presented without centered ﬁrst order terms (Katrichis 1992; Kromrey and
Foster-Johnson 1998).

170

communists in ﬁve of nine tolerance models. Models for each of the three out-groups
include measuring support for the civil rights of an out-group member as measured by
giving a public speech, keeping an out-group-authored book in the library, or allowing an

out-group member to teach at the college level.

Table 7.5. Logistic Regression: Paradign Scales and Tolerance of Atheists

 

 

Atheist Speaker Atheist Book Atheist Teacher
Political Variables
Party ID .05 (.05) .04 (.04) .07 (.04)
Political Ideology -.O7 (.07) -. l 3 * (.07) -.12* (.06)
Demographics
Education .58‘" (. 10) 60*" (.10) 66"" (.10)
Income .10 (.09) .09 (.09) .05 (.09)
Gender -.36“ (. 18) .06 (.17) -.01 (. 17)
Age -.02"‘" (.01) -.01"* (.01) -.04*" (.01)
Children in Home .00 (.05) -.01 (.05) .07 (.06)
Urban .04 (.03) .00 (.03) .04 (.03)
Marital Status -.02 (.07) . 10 (.07) -.00 (.06)
Religious Variables
Religious Attendance -.19“ (.08) -. 14* (.08) -.02 (.07)
Evangelical -.41“’ (.22) -.67* ** (.22) -.22 (.21)
Mainline .28 (.25) -.46" (.23) .08 (.22)
Black -.34 (.30) -.51+ (.30) -.50"‘ (.29)
New Paradigm Scale .01 (.06) .04 (.06) .00 (.05)
Interaction Terms
Evangelical X Paradigm -.39* (.22) -.59"* (.22) -.32 (.20)
Mainline X Paradigm -.48* (.26) -.48** (.24) -.01 (.21)
Black X Paradigm -.24 (.32) .13 (.31) -.01 (.29)
Paradigm X Attendance -.02 (.03) -.02 (.03) -.01 (.03)
Evangelical X Attendance .04 (.11) -.05 (.11) -.09 (.11)
Mainline X Attendance .29“ (.13) .07 (.12) .20" (.11)
Black X Attendance -.02 (.15) .10 (.15) -.16 (.15)
Evangelical X Attendance X Paradigm .08" (.04) .09" (.04) .06‘ (.03)
Mainline X Attendance X Paradigm .09“ (.05) .07 (.04) -.Ol (.04)
Black X Attendance X Paradigm .05 (.05) -.04 (.05) .02 (.05)
Constant .46 (.70) -.O9 (.70) -.20 (.67)
Observations 85 l 829 821
Pseudo-R2 .14 .13 .15

 

Notes: Entries are logistic regression coefﬁcients with standard errors in parentheses. The Roman Catholic
tradition is dropped as the baseline variable.
‘p<.10 (one-tail test), "p<.05 (two-tail test), "'p<.01 (two-tail test)

171

Table 7.5 gives the results for measuring tolerance of atheists, an out-group for
which new paradigm churches have already been shown to have increased support for
their civil rights. In keeping with research beginning with Stouffer (1954), education
increases tolerance in all three atheist models and age decreases tolerance. In support of
Kennedy’s conclusion about new paradigm churches increasing tolerance for atheists,
Evangelicals with higher levels of attendance, and whose congregation has higher new
paradigm characteristics, have higher probabilities of tolerance in all three models. As the
ﬁrst set of models ﬁnds, signiﬁcant change in tolerance requires the interaction of
attendance with new paradigm levels. Figures 7.1 through 7.3 show a uniform pattern for
Evangelical attendance and new paradigm levels. Occasional attendance produces a ﬂat
to slight increase in the probability of tolerance, regardless of increasing new paradigm
levels. Those with increasing new paradigm levels and little or no attendance drop in
their probability of expressing tolerance. The probability of tolerance increases for
Evangelicals with increasing attendance and increasing new paradigm levels. The level of
new paradigm inﬂuence appears to reverse the effects of commitment in an Evangelical
congregation. Higher levels of commitment (as measured by attendance) reduce the
probability of tolerance in a traditionally Evangelical church, while decreased
involvement in the traditional church raises the probability of tolerance. The new cultural

ethos produced in a new paradigm church turns this on its head.

172

Figure 7.1. Probability of Tolerance of an Atheist Speaker by Religious Tradition,
New Paraﬂrn Attributes, and Attendance.

Probability of Tolerance of Atheist Speaker

   

 

 

 

 

 

 

 

 

 

"‘1
”-1
\Q-
V:—
or-
Low2468101214High Low2468101214Hi
Evangelical New Paradigm Scale Mainline New Paradigm Scale
“3‘ ....................................... ea .. ___.__--—--"""
e- y a- “4"
V:- v-
“in N
I I I I I I I I I I I I I I I I I I
Low2468101214High Low2468101214Hi
Black New Paradigm Scale Catholic New Paradigm Scale

 

Attend Weekly or More
----- Once to Several Times a Month
----------------- Attend Once or Twice a Year or Less

 

 

 

 

Note: Predicted probability data is from Table 7.5.
lControlling for Strength of partisanship, Party ID, Gender, Children, Marriage, Education, Income, and
Urban.

The most puzzling part of this pattern is that lower attendance at a new paradigm
church reduces tolerance, while lower attendance increases tolerance for traditional
Evangelical churches. The answer may be in the nature of the new paradigm churches.
Many of them are relatively new congregations, or traditional churches that have
undergone radical changes in their structure and style. Frequently established churches

have undergone signiﬁcant turmoil with membership loss or even church splits when

adopting new paradigm styles. The likely possibility is that infrequent new paradigm

173

attendees represent evangelicals whose attendance has declined in opposition to the new

paradigm inﬂuences. While these disgruntled attendees would represent the church

stalst in traditional Evangelical churches, new paradigm directions (seen as “worldly”

compromise) have increased dissatisfaction and reduced attendance, but not affected their

“normal” lower tolerance. It becomes a possibility that the two most intolerant

Evangelical groups, residing at opposing ends of the scales in Figures 7.1 through 7.3, are

in effect drawn from the same Evangelical subgroup.

Figure 7.2. Probability of Tolerance of an Atheist Book by Religious Tradition, New
Paradigm Attributes, and Attendance.

 

 

Probability of Tolerance of Atheist Book

 

 

 

 

 

 

 

 

 

 

... 4
sq .
‘°- 1
v. 4
. NJ
Low 6 8 10 12 14 High Low 2 4 6 8 10 12 14 Hi
Evangelical New Paradigm Scale Mainline New Paradigm Scale
— —( —1 -1 -
“H “3 * _._..—..._..-—-‘-‘-"“""
e- e-*"'
v. - e. -
N -1 (V. cl
. I I I1 T I I I I I I I I I I I I
Low 6 8 10 12 14Hi Low 2 4 6 8 10 12 14111
Black New Paradigm Scale Catholic New Paradigm Scale

 

 

Attend Weekly or More
----- Once to Several Times a Month

----------------- Attend Once or Twice a Year or Less

 

 

 

Note: Predicted probability data is from Table 7.5.
I Controlling for Strength of partisanship, Party ID, Gender, Children, Marriage, Education, Income, and

Urban.

174

A similar pattern also exists for Mainline attendees. Figures 7.1 and 7.2 show that
paradigm inﬂuence increases the tolerance of higher attending Mainline respondents,
while Figure 7.3 shows that Mainline new paradigm attributes increase tolerance
regardless of attendance levels. Adoption of this new Evangelical style by Roman
Catholics appears to move them in a more conservative direction on tolerance. Also in
keeping with the possibility of a Pentecostal inﬂuence rather than a cultural-embracing
inﬂuence, high attending Black Protestants either decline or have the lowest increase of
tolerance of the three attendance levels. Increased tolerance appears to only occur for new
paradigm Evangelicals or Mainline Protestants.

Figure 7.3. Probability of Tolerance of an Atheist Teacher by Religious Tradition,
New Paradigm Attributes, and Attendance.

Probability of Tolerance of Atheist Teacher

 

 

 

 

 

 

 

 

 

 

 

 

00. ..
‘0' q -— .fﬂ.vre~*"°‘“w;.;
V: .. .....................
N ..
Low 2 4 6 8 10 12 14Hi Low 2 4 6 8 10 1214Hi
Evangelical New Paradigm Scale Mainline New Paradigm Scale
«2 - ee. -
‘0. " \O. .. ....;._.._._._.._._._.u._._..—-.m.-r:'..'.'.'.:1'.."' -
‘l: - V. ‘
er a e:
I I I I I I I I I I I I I I I I f I
Low 2 4 6 8 10 12 14 High Low 2 4 6 8 10 12 14 Hi
Black New Paradigm Scale Catholic New Paradigm Scale

 

Attend Weekly or More
----- Once to Several Times a Month
----------------- Attend Once or Twice a Year or Less

 

 

 

 

175

Note: Predicted probability data is from Table 7 .5.
' Controlling for Strength of partisanship, Party ID, Gender, Children, Marriage, Education, Income, and

Urban.

While only one of each of the three models for Homosexuals and Communists
(Tables 7.6 and 7.7) shows a signiﬁcant attendance/new paradigm/Evangelical
interaction, the conﬁgurations remain the same. For Evangelical and Mainline
Protestants, increasing exposure to an increasing number of new paradigm attributes
raises the probability of supporting civil liberties and tolerance for out-groups activities

(Figures 7.4 and 7.5).

Table 7.6. Logistic Regression: Paradigm Scales and Tolerance of Homosexuals.

 

 

 

Homosexual Speaker Homosexual Book Homosexual Teacher
Political Variables
Party ID -.00 (.05) .07 (.05) -.02 (.05)
Political Ideology -.20** (.08) -. 13+ (.07) -.1 l (.07)
Demographics
Education .67*** (.12) 53*" (.11) .72"* (.12)
Income .09 (.1 1) .00 (. 10) .09 (. 10)
Gender -.07 (.20) .09 (. 18) 60*" (.20)
Age -.02**" (.01) -.02"* (.01) -.04*** (.01)
Children in Home .06 (.06) -.04 (.06) .00 (.06)
Urban .05 (.04) -.02 (.03) .01 (.04)
Marital Status .02 (.08) .07 (.07) .03 (.08)
Religious Variables
Religious Attendance -.12 (.10) -.16+ (.09) -.07 (.09)
Evangelical .26 (.18) .26+ (.15) .34" (.17)
Mainline -.68*" (.26) -1.18*** (.23) -.83 “l“ (.25)
Black .45 (.32) -.39 (.26) . 15 (.31)
Political Variables -.36 (.36) -1.03"’" (.31) -.52 (.34)
New Paradigm Scale .03 (.07) .03 (.06) .01 (.07)
Interaction Terms
Evangelical X Paradigm -.06 (.26) -.30 (.23) -.39 (.24)
Mainline X Paradigm .11 (.33) -.42 (.28) .49 (.32)
Black X Paradigm 1.05" (.44) -.05 (.32) .31 (.37)
Paradigm X Attendance .02 (.04) -.02 (.03) -.O3 (.03)
Evangelical X Attendance -.09 (.13) -.08 (.12) -.17 (.12)
Mainline X Attendance .14 (.16) -.16 (.13) -.20 (.15)
Black X Attendance .22 (.18) .01 (.16) .08 (.17)

176

Table 7.6 (cont’d).

Evangelical X Attendance .01 (.04) .05 (.04) .O7* (.04)
X Paradigm

Mainline X Attendance X -.01 (.06) .06 (.05) -.05 (.05)
Paradigm

Black X Attendance X -.18"'* (.07) -.00 (.06) -.O6 (.06)
Paradigm

Constant .81 (.84) .65 (.75) -.12 (.80)
Observations 820 807 807

PseudwRZ .19 .17 .22

 

Notes: Entries are logistic regression coefﬁcients with standard errors in parentheses. The Roman Catholic
tradition is dropped as the baseline variable.
*p<.10 (one-tail test), "p<.05 (two-tail test), ***p<.01 (two-tail test)

Figure 7.4. Probability of Tolerance of a Homosexual Teacher by Religious-
Tradition New-Paradigm Attributes, and Attendance.

 

Probability of Tolerance of Homosexual Teacher

 

 

 

 

 

 

 

 

 

 

 

Low 2 4 6 8 10 12 14 Hi Low 2 4 6 8 10 12 14 High
Evangelical New Paradigm Scale Mainline New Paradigm Scale
-— -1 ~ ‘1 “In..." "
e2 - ~e. -
”I d “I a
N - or a
I I I I I I l I I I I I I I I I I 1
Low 2 4 6 8 10 12 14 Hi Low 2 4 6 8 10 12 14 Hi
Black New Paradigm Scale Catholic New Paradigm Scale

 

Attend Weekly or More
----- Once to Several Times a Month
----------------- Attend Once or Twice a Year or Less

 

 

 

 

Note: Predicted probability data is from Table 7.6.
1Controlling for Strength of partisanship, Party ID, Gender, Children, Marriage, Education, Income, and

Urban.

177

Table 7.7. Lgistic Regression: Paradigm Scales and Tolerance of Communists.

 

Communist Speaker Communist Book Communist Teacher

 

Political Variables

Party ID .08‘ (.05) .15*" (.05) .01 (.04)
Political Ideology -.13 ** (.07) -.15** (.07) -.06 (.06)
Demographics

Education .70**" (.10) 72*" (.10) 37*" (.10)
Income .10 (.09) .01 (.09) .02 (.09)
Gender -.30* (.17) -.03 (.17) .05 (. 16)
Age -.01‘”“ (.01) -.01 (.01) -.02*** (.01)
Children in Home -.03 (.05) -.02 (.05) .03 (.05)
Urban .03 (.03) -.01 (.03) .01 (.03)
Marital Status -.01 (.07) .07 (.07) .02 (.06)
Religious Variables

Religious Attendance -.11 (.07) -.13+ (.08) -.06 (.07)
Church Theology .17 (.14) .28+ (.14) .22" (.13)
Evangelical -.34 (.22) -.58*** (.22) -.40“ (.21)
Mainline .23 (.24) -.32 (.24) -.32 (.22)
Black -.21 (.29) -.54* (.30) -.56* (.29)
New Paradigm Scale .01 (.05) -.O4 (.06) -.05 (.05)
Interaction Terms

Evangelical X Paradigm -.38* (.21) -.28 (.22) -.15 (.20)
Mainline X Paradigm -.42* (.25) -.44* (.25) .14 (.22)
Black X Paradigm -.13 (.31) .08 (.31) -.28 (.30)
Paradigm X Attendance -.02 (.03) -.02 (.03) -.02 (.03)
Evangelical X Attendance .05 (.11) .05 (.11) .IO (.10)
Mainline X Attendance -.05 (.12) -.01 (.12) .06 (.11)
Black X Attendance -.06 (.15) -.OO (. 15) -.24 (.15)
Evangelical X Attendance X .06" (.04) .05 (.04) .03 (.03)
Paradigm

Mainline X Attendance X .07“ (.04) .08“ (.04) .00 (.04)
Paradigm

Black X Attendance X Paradigm .04 (.05) -.Ol (.05) .08 (.05)
Constant -.75 (.77) -1.16 (.73) .21 (.68)
Observations 8 1 5 803 785

Fund”? .14 .14 .09

 

Notes: Entries are logistic regression coefficients with standard errors in parentheses. The Roman Catholic
tradition is dropped as the baseline variable.
‘p<.10 (one-tail test), "p<.05 (two-tail test), *"p<.01 (two-tail test)

178

Figure 7.5. Probability of Tolerance of a Communist Speaker by Religious-
Tradition New-Paradigm Attributes, and Attendance.

 

Probability of Tolerance of Communist Speaker

e—n-t ~-l

 

 

   

 

 

 

 

 

 

 

 

Low 2 4 6 8 10 12 14 Hi Low 2 4 6 8 10 12 14 Hi
Evangelical New Paradigm Scale Mainline New Paradigm Scale
°°. a °°. - _...... -
\O. 4 ~42 - __.—__...———--B "‘
er. - e. -
e: e: a
I I I I T T j l I I I I I I f I I I
Low 2 4 6 8 10 12 14 High Low 2 4 6 8 10 12 14 Hi
Black New Paradigm Scale Catholic New Paradigm Scale

 

Attend Weekly or More
----- Once to Several Times a Month
----------------- Attend Once or Twice a Year or Less

 

 

 

 

Note: Predicted probability data is from Table 7.7.
l Controlling for Strength of partisanship, Party ID, Gender, Children, Marriage, Education, Income, and
Urban.

The next model looks at the only moral/political attitude beyond tolerance to have
a signiﬁcant Evangelical new paradigm effect. The abortion scale used as the dependent
variable was produced ﬁ'om six GSS questions measuring support for abortion under a
variety of circumstances. The scale is coded O - no support for abortion to 6 - highest
support for abortion. The regression model (Table 7.8), using the same control variables

as the tolerance model, ﬁnds that new paradigm attributes decrease abortion support at

nearly every level of attendance for Evangelical, Mainline and Black Protestants. Only

179

the lowest levels of Evangelical attendees are unchanged by increasing new paradigm
levels. Among new paradigm Roman Catholics (Figure 7.6) with high and occasional
attendance, a moderation or secularization of abortion attitudes appears to be taking
place. For the moment, Evangelical new paradigm churches appear to be maintaining a
balancing act of conservative theology with increasing cultural openness. The lack of
signiﬁcance for new paradigm attributes in so many models (not shown) suggests that NP
churches are not yet inﬂuencing their attendees in a more secular direction for a variety
of attitudes. However, with the exception of abortion attitudes, they are not signiﬁcantly
inﬂuencing Evangelicals in a conservative direction either. The surprise is that they are
not producing moderation, but a conservative shift (at least for abortion) in Mainline and

Black traditions.

Table 7.8. Regr_ession: Paradigm Scales and Abortion Support.

 

 

Abortion Scale
Political Variables
Party ID -.09* (.06)
Political Ideology -.21 ** (.09)
Demographics
Education .07 (.14)
Income -.02 (. 1 3)
Gender .20 (.24)
Age .00 (.01)
Children in Home -.09 (.08)
Urban .11" (.04)
Marital Status .02 (.09)
Religious Variables
Religious Attendance -.34*" (.11)
Evangelical 1.31“” (.43)
Mainline 260*" (.85)
Black .24 (.41)
New Paradigm Scale -.05 (.07)
Interaction Terms
Evangelical X Paradigm 1.01" (.45)
Mainline X Paradigm -.24 (.78)
Black X Paradigm .38 (.39)

180

Table 7.8 (cont’d).

Paradigm X Attendance .07*
Evangelical X Attendance -.19
Mainline X Attendance -.59
Black X Attendance .56***
Evangelical X Attendance X Paradigm -.15**
Mainline X Attendance X Paradigm -.01
Black X Attendance X Paradigm -.O7
Constant 1.46
Observations 737

R2 .19

(.04)
(.21)
(.40)
(.20)

(.07)
(.1 1)
(.07)

(1.02)

 

Notes: Entries are regression coefﬁcients with standard errors in parentheses. The Roman Catholic tradition

is dropped as the baseline variable.

I"p<.10 (one-tail test), ”p<.05 (two-tail test), I""“'p<.01 (two-tail test)

Figure 7.6. Pro-choice Support by Religious-Tradition New-Paradigm Attributes,

and Attendance.

 

 

 

 

 

 

 

 

 

 

 

 

Prochoice Support

fa fl»-
1"! :E ..
v .. .......................................................... e - T_‘----—--..__~ .............
... r — ““““““““““““ N - _— ‘
3 .4 i 3 ..
3 T T l r r l l I r ,3 I r I I I r I r I

Low 2 4 6 8 10 12 14Hi Low 2 4 6 8 10 12 14111

Evangelical New Paradigm Scale Mainline New Paradigm Scale
.5 in
E ‘ m '
‘t 1 — V _ ...........................
W ——————— W

N - _ ' N - __.
3 a 3
.3 I I I I I I I I I .3 I I I I I I I I I

Low 2 4 6 8 10 12 14 High Low 2 4 6 8 10 12 14 High

Black New Paradigm Scale Catholic New Paradigm Scale
Attend Weekly or More
----- Once to Several Times a Month
----------------- Attend Once or Twice a Year or Less

 

 

 

Note: Predicted probability data is from Table 7.8.

1Controlling for Strength of partisanship, Party ID, Gender, Children, Marriage, Education, Income, and

Urban.

181

 

The last model looks for a new paradigm effect on political participation in the
form of voting. The dependent variable for this model is whether the respondent reported
voting in the 1996 Presidential Election. Dependent variables are the same as the
tolerance and abortion models with the exception of the addition of a number of
congregation-speciﬁc political variables. Using the NCS also allows for inclusion of
church variables not normally found in political models. As the strictness or theological
conservativeness of a church has been shown to affect participation, a liberal to
conservative theology variable is included in the voter participation model (Campbell
2000). In addition, the voter participation model includes variables covering candidates
speaking to the congregation, voter-guide distribution, opportunities for political activism
and groups meeting to discuss politics. As Black Protestants historically have the lowest
voting tendency among the religious groups, they are dropped as the comparison
category.

As Table 7.9 shows, the model ﬁnds signiﬁcant support for the strict church
hypothesis. As a congregation’s theology moves from liberal to moderate to conservative,
the probability of voting decreases 9%. At the same time, if the perceived political stance
of the church is conservative, the probability of voting increases by 11% over a liberal
congregation. None of the other congregational political variables are signiﬁcant. The
control variables perform as expected with strength of partisanship, education and income
increasing the likelihood voting, being female decreasing the likelihood, and age showing

a curvilinear effect of increase then decrease.

182

Table 7.9. Logistic Regression: PamﬂggScales and Probability of Voting.

 

Voted in 96
Marginal Effects

Political Variables
Strength of Partisanship .61*** (.08) .09
Church Political Ideology .33" (. 15) .05
Opportunities for Political Activism -.04 (.17) -.00
Group to discuss Politics -.30 (.27) -.02
Voter Guides Distributed .09 (.18) .00
Candidate spoke to congregation .30 (.34) .03
Demographics
Education 57*" (.10) .10
Income .12‘ (.08) .02
Gender -.27* (.16) -.04
Age .13"‘" (.03) .02
Age-squared -.09*** (.02) .01
Urban -.04 (.03) -.01
Religious Variables
Church Theology -.31" (.15) -.05
Religious Attendance .00 (.11) .00
Evangelical -.15 (1 .05) -.01
Mainline .64 (.99) .09
Catholic -1 .07 (.65) -.03
New Paradigm Scale .03 (.06) .00
Interaction Terms
Evangelical X Paradigm -.48"‘* (.24) -.08
Mainline X Paradigm -.28 (.27) -.03
Catholic X Paradigm -.25 (.25) -.06
Paradigm X Attendance -.04 (.03) -.01
Evangelical X Attendance .03 (.13) .01
Mainline X Attendance .41 "'** (.16) .08
Catholic X Attendance . 12 (.13) .02
Evangelical X Attendance X Paradigm .08" (.04) .01
Mainline X Attendance X Paradigrn .06 (.05) .01
Catholic X Attendance X Paradigm .05 (.04) .01
Constant 6.14"" (1.40)
Observations l 156
Pseudo-R2 '21

 

Notes: Entries are logistic regression coefficients with standard errors in parentheses. The Black Protestant
tradition is dropped as the baseline variable
*p<.10 (one-tail test), "p<.05 (two-tail test), l'“'”"p<.01 (two-tail test)

The interaction of attendance, new paradigm characteristics and being Evangelical

is again signiﬁcant. Evangelicals with high attendance in congregations with high new

183

paradigm levels slightly increase their probability of voting (Figure 7.7). Other
Evangelical levels of attendance dropped in their probability of voting as the new
paradigm attributes increased. Increasing attendance in a new paradigm church for the
other three traditions decreases the likelihood of voting. Given that new paradigm
characteristics are intertwined with Evangelical or conservative theology (having started
in the Evangelical tradition), a congregation that is not already Evangelical is often
becoming more theologically conservative by adopting this style and message. In effect
for Evangelicals, new paradigm increases are a liberalizing inﬂuence while for every

other tradition new paradigm increases are a conservative shift.

Figure 7.7. Probability of Voting by Religious-Tradition New-Paradigm Attributes,
and Attendance.

 

 

 

 

 

 

 

 

 

 

 

 

Probability of Voting

u—t-i —
”3““_r~—— ”j”"'—-”‘i::r
‘0. q .........'.:':.:: _____ ‘0. 1 .......................................
Vt ‘ ................... q. .1
e: - N. -

Low 2 4 6 8 10 12 14 Hi Low 2 4 6 8 10 12 14 High

Evangelical New Paradigm Scale Mainline New Paradigm Scale
ee. - e9 - M
\o. _ ‘0. - _.'_'pa-.~I'\ﬁ?.=.:_::- —————
v. _l v. - ..............
N. - e! -
I I I I I I I I j I I I I I I I I I
Low 2 4 6 8 10 12 14 High Low 2 4 6 8 10 12 14 High

Black New Paradigm Scale Catholic New Paradigm Scale

 

Attend Weekly or More
----- Once to Several Times a Month
----------------- Attend Once or Twice a Year or Less

 

 

 

 

184

Note: Predicted probability data is from Table 7.9.
1Controlling for Strength of partisanship, Party ID, Gender, Children, Marriage, Education, Income, and
Urban.

7.7 Conclusions

Under the surface of changing Evangelical political participation has lurked a
growing often unnoticed inﬂuence among Evangelicals and to a lesser extent Mainline
Protestants. The new paradigm movement has begun to transform these two traditions
from the inside with increasing participation and tolerance for Evangelicals and
increasing tolerance for Mainline Protestants. While both Evangelical and Mainline show
some new paradigm effect, why is there a clearer pattern of increased tolerance for
Evangelicals than Mainline Protestants? Mainline churches already embrace tolerance at
a higher rate than Evangelicals. For Mainliners, adopting new paradigm styles merely
moves them along in the same direction they were heading while for Evangelical
churches it is a new tack, a change in orientation on tolerance and participation.

Conversely, it is not clear from the data that new paradigm attributes apply to
congregations outside of Evangelical and Mainline Protestants. The results show clear
changes in Evangelical and Mainline respondents as their attendance and NP attributes
increase. The evidence is less clear for Black Protestants and Roman Catholics. What is a
scale of new paradigm attributes for Evangelicals and Mainline congregations may be a
scale of Pentecostal/charismatic inﬂuence for Roman Catholics and Black Protestants.
The new paradigm scale carries with it an ethos of openness to the culture, tolerance, and
participation while the Pentecostal scale may be closing congregations to the culture. The

constant inﬂuence of attendance interacting with NP levels suggests that new paradigm

185

inﬂuence requires exposure in order to produce change among respondents. This effect is
most evident in new paradigm effects on tolerance.

The results of this model support Djupe and Grant’s (2001) ﬁnding that religious
traditions do not directly promote different levels of participation bur rather participation
depends on how churches connect their religious practices and their political activism. In
this case though, participation is not tied to political events (political speaker, political
group) held within the church, but rather, it is a reﬂection of Evangelicals rejecting the
old separatism standard. The mantra of cultural relevance helps new paradigm attendees
embrace the civil liberty standards of the greater community, and raises voting
participation. Still, the theorized secularization of embracing cultural relevance does not
yet appear to be occurring. This is not to suggest that it will not, merely that it bears

further watching as new paradigm inﬂuence continues to grow.

186

CHAPTER VIII

CONCLUSIONS

Evangelical participation has changed and as a result Evangelicals are changing.
The campaign participation and voting trends in this study suggest a gradual acceptance
of new levels of political participation. The 1992 Presidential election may have been a
key election for Evangelical participation in the Republican Party, but it did not trigger
changes in Evangelical participation. The mutual embrace of Evangelicals and the GOP
must be assumed to have been built upon the foundation of sporadic, yet increasing,
Evangelical political activism. Evangelicals have effectively shed all signs of separatism
in politics. As a result, the GOP ﬁnds a new constituency and Evangelicals ﬁnd a home in
a major party because Evangelicals are both conditioned to participate and are looking for
a participatory outlet that signals receptiveness to their concerns. The changes in 1992 are
a consequence of trends starting in 1980. This is a tale of a gradual process that covers
two decades, involves a reordering of Evangelical political thought, and focuses mostly
on changes among younger Evangelicals.

Chapter Five shows that the American electorate underwent a fundamental shift in
its political sophistication as measured by issues accessibility. From one percent of the
total population in 1976, the accessibility of Social Issues has steadily grown to the point
that by the year 2000 it was accessed ﬁrst by one out of eight Americans. This
demonstrates a shift in political thinking country, and also suggests the effectiveness of
new cues in restructuring the hierarchy of chronicities for Evangelicals and Non-
Evangelicals alike. These results also hint at some validity to the culture war motif. The

increased political involvement of Evangelicals is a reﬂection of increased concern about

187

social issues accompanied to a lesser extent by more participation by Evangelicals using
the economic chronicity. A similar shift among Seculars completes the drawing of lines
for the cultural conﬂict. As a result, Evangelicals and Seculars appear to be increasing
their access to the Social Issues chronicity in a manner that reﬂects the culture war
expectations of Hunter (1991) and Wuthnow (1998). The lagging increase of the
chronicities’ usage for the rest of the population reveals the widening inﬂuence of this
war beyond Evangelicals and Seculars.

Fiorina (2004) may be partially right in his belief that the culture war is being
fought by ten percent of the American electorate. However, the trends of increasing
numbers of seculars, increasing participation of Evangelicals and growing usage of social
issue access imply that the culture war is not yet fully engaged. The conﬂicts of the past
two decades may only be the opening skirmishes of a larger war yet to come. Conversely,
it should be acknowledged that Evangelical participation has not exceeded that of other
religious traditions to any signiﬁcant degree. Evangelicals have not become super-
participants. The past two decades have merely seen the return of religious conservatives
to the political arena. Their participation is only noticeable because of their previous
avoidance of the political realm.

The Social Issues chronicity usage also shows the inﬂuence of Evangelical church
environment as attendance increases its accessibility, while economic chronicity remains
unaffected by attendance. Additionally, orthodoxy as measured by views for the Bible
fails to be signiﬁcant for either reciprocal model. The inﬂuence of attendance on political

sophistication not only reinforces prior conclusions about churches as political

188

communities, but suggests another area of effect for these communities that deserves
greater consideration (Wald, Owen and Hill 1988; Djupe and Grant 2001).

Social Issues is not the only chronicity affecting Evangelical participation. It is
possible that the usage of the economic chronicity may also provide good news for the
GOP, as a segment of Evangelicals is beginning to reﬂect traditional Republican
emphasis on economic issues. Since orthodoxy and attendance are not affecting the
access of economic issues, it suggests that Evangelicals are recipients of economic cues
fiom outside the religious tradition. While this study cannot identify those cues, it
suggests that Evangelicals are open to more traditional, political inﬂuences such as the
media and political parties.

As expected, social issues and the economy cues within the Evangelical
community result in a reciprocal relationship between participation and chronicities. The
study supports the ﬁndings of Leighley (1991), Junn (1991) and Bennett (1975) that
participation and political sophistication have a reciprocal relationship. Surprisingly, I do
not ﬁnd support for Delli, Carpini and Keeter (1996), Bennett (1994, 1995) and Tan’s
(1980) conclusions that changes in sophistication are dependent on shifts in political
information. Evangelicals who are shifting their issue accessibility and increasing their
participation are not doing so as a result of increases in their levels of political
knowledge. There is a curious unexplained disconnect between increased participation,
changes in political issue access and increases in political information. Neuman’s (1986)
spiral based on increasing political information does not appear to have been initiated by
the positive Evangelical political atmosphere. The closest explanation for this

phenomenon is Breggren’s ﬁndings about free information from political groups

189

weakening the relationship between socio-economic status and sophistication. Perhaps
the free information within the Evangelical environment also short-circuits the
relationship between sophistication, participation and the seeking of new information. It
remains one of the most puzzling ﬁndings of this study and one worthy of fluther inquiry.
The examination of participation and political information shows that they affect
Evangelicals in a unique manner. The effects of both in increasing or decreasing
chronicity usage are more severe than for other religious traditions. Participation sharply
increases social issues and economic chronicities as compared to other traditions, and
increasing political information more sharply reduces them (with the exception of
Mainlines). While proving that these two variables are unique for Evangelicals in their
chronicity usage, it is beyond the scope of this study to suggest an answer as to why they
affect Evangelicals differently. What is left is a picture of Evangelicals with increasing
participation and shifts in sophistication that are changing without beneﬁt of increasing
political information. It is a disturbing picture of an increasingly mobilized portion of the
electorate that still chooses to remain ignorant of the realm in which it is participating.
Furthermore, Evangelicals are changing in that they are learning the “rules of the
game.” Chapter Six reveals that participation and moderation are increasing among the
youngest cohorts. The future of Evangelical participation looks bright based on younger
cohort trends. The future of civic norms is positive as younger Evangelicals are
demonstrating signiﬁcant increases in tolerance. The cohort participation trend provides
support for Zaller’s (1992) “partisan resistance” that older respondents will be less open

to new political messages and are more likely to reject participation.

190

The liberal trend of these younger cohorts illustrates that studies such as Hunter
(1982) and Penning and Smidt (2003), which focus on secularization among Evangelical
elites, are failing to grasp what is occurring for the tradition as a whole. Evangelical
leaders, trained in the tradition’s colleges and seminaries, are holding to traditional
beliefs while the non-elites are drifting away from those positions. The future of this
tradition is a highly traditional leadership core leading an increasingly secular ﬂock.
Secularization is occurring within the Evangelical tradition and political moderation is
accompanying it.

This dissertation also brings to light that Evangelicalism, like all religious
traditions, is not a static movement. Internal inﬂuences can be just as profound as
external inﬂuences. The new paradigm church movement demonstrates the effects of a
growing movement within Evangelicals in both increasing tolerance and increasing
participation. It suggests that there is a need in political science for a subﬁeld much like
the sociology of religion where religious change is examined and catalogued because of

its potential political consequences.

8.1 Future Directions

Among the possible directions for future work is a more comprehensive
examination of the Social Issues and Economic chronicity respondents. While in some
election years they are a very small N, it may be possible to evaluate these groups by SES
and other demographic factors. In addition to political participation, more work is needed
to determine the effect that these chronicities may have on shifts in Evangelical voting
(party choice, not participation) and party identiﬁcation as well as efficacy. As

Evangelicals are increasing in participation and in learning civic norms, is there a similar

19]

effect on their political efﬁcacy? In particular, the effects of the Social Issues chronicity
on party identiﬁcation may contribute to the literature concerning party realignment
around traditional issues. As the Social Issues chronicity shows no sign of abating its
linear increase, it deserves further attention in future ANES surveys.

This study also opens an avenue to considering how to deﬁne an accurate
religious identity. The discord between the GSS codes and NCS codes provides fertile
ground for evaluating the inﬂuence of one’s current congregation versus one’s religious
identiﬁcation.

This study shows that religious identities have a signiﬁcant role in political
sophistication, levels of participation and civic norms. It is but the ﬁrst step of many in
evaluating internal changes within religious identities as a result of their interaction in the

political realm.

192

Appendix A

Classifying Religious Traditions

Historically measuring evangelicals involved a debate between using on a
theological distinctives (Smidt, 1988) or a denominational affiliation (Wilcox, 1987;
J elen, 1989a). Hovever, most surveys do not include nuanced theological questions that
would allow for determining religious traditions. Using denominations has been faulted
for assuming a uniformity of belief and practice within the churches and ignores those
who attend non-Evangelical churches who may hold similar views (Wilcox, 1986).

Over time the subﬁeld of politics and religion has come to accept the
denominational measurements as the defacto standard for evaluation religious traditions
in America. Steensland et al.’s (2000) coding for the General Social Survey has also been
used with the NES and in a very short time has become the most often referenced
standard for denominational classiﬁcation. It is an adaptation of the Steensland et al.

classiﬁcation that is used in this paper.l The codings are:

Chapter 5 and 6 (National Election Study 1972-1988)
1) Respondents were assigned as follows:

Evangelical Protestants: Non-black respondents affiliated with Baptist, Primitive Baptist,
Free Will Baptist, Southern Baptist, Pentecostal, Assemblies of God, Church of God,
Disciples of Christ, Nazarene, United Missionary, Free Methodist, Protestant Missionary,
Church of God in Christ, Plymouth Brethren, Church of Christ, Salvation Army, Seventh
Day Adventists, Evangelical and Reformed, Reformed, Dutch Reformed, Christian

 

' I am indebted to my instructors and fellow participations at the Pollsters and Parishioners: Seminar on
Survey Research and American Religion sponsored by the Henry Institute at Calvin College, 2003 for their
input in our cooperative effect in converting the Steensland et a1 measurement from the GSS to the NES
and for the correction of errors (albeit few) and new GSS additions in Steensland et al.’s classiﬁcations.
Denominations not covered in the Steensland et a1. measurement were classiﬁed with J. Gordon Melton’s
Encyclopedia of American Religions (6th edition), Frank S. Mead and Samuel S. Hill’s Handbook of
Denominations in the United States (9th edition) and when available denominational websites.

193

Reformed, Missouri Synod Lutheran, Mennonite, Amish, Church of the Brethren and
Other Fundamentalists.

Mainline Protestant: Presbyterian, Lutheran, Congregational, United Church of Christ
Episcopal, Anglican, Church of England, United Methodist/Methodist/Evangelical
United Brethem, Disciples of Christ, Christian and white respondents falling into the
Baptist and Southern Baptist categories residing outside the political South.

Black Protestant: Aﬁican Methodist Episcopal, all black members of Evangelical and
Mainline denominations and black respondents falling into the Baptist and Southern
Baptist categories, and black respondents, missing their denominational response or
refusing to respond, with church attendance more frequent that yearly.

Roman Catholic.
Jewish.
Secular: Agnostic, Atheist

Other: Orthodox, Christian Scientist, Spiritualist, Mormon Latter Day Saints, Unitarian,
Jehovah’s Witness, Quaker, Unity, Muslim, Buddhist, Hindu, Bahai, Other non-Judeo-
Christian religions, Other religions.

1) Respondents with unclear responses are classiﬁed based on other characteristics.
Those with the high levels of attendance were coded evangelical. Those with low (about
once a month or less) or no reported attendance were coded secular.

2) White respondents answering Protestant, Non-denominational or Community church
with the high levels of attendance were coded evangelical. Those with low or no reported
attendance were coded secular.

3) Black respondents answering Protestant, Non-denominational or Community church
with the high levels of attendance were coded black protestant. Those with low or no
reported attendance were coded secular.

Chapter 5 and 6 (National Election Study 1992-2000)
1) Respondents were assigned as follows:

Evangelical Protestants: Non-black respondents afﬁliated with 7th Day Adventist,
Fundamentalist Adventists (Worldwide Church of God), Adventist , American Baptist
Association, Baptist Bible Fellowship, Baptist Missionary Association of America,
Conservative Baptist Association of America, General Association of Regular Baptist
Churches (G,A,R,B,), National Association of Free Will Baptists (United Free Will
Baptist Church), Primitive Baptists, National Baptist Convention in the U,S,A,, United

194

F ree-Will Baptist Church, Reformed Baptist (Calvinist), Southern Baptist Convention,
Fundamental Baptist (no denom, ties), Local (independent) Baptist churches with no
denominational ties or links to a national fellowship, Baptist , Old Order Amish,
Evangelical Covenant Church (not Anabaptist in tradition), Evangelical Free Church (not
Anabaptist in tradition), Brethren in Christ, Mennonite Brethren, Brethren , Mennonite
Church, Christian and Missionary Alliance (CMA), Church of God (Anderson, IN),
Church of the Nazarene, Free Methodist Church, Salvation Army, Wesleyan Church,
Church of God of Findlay, OH, Holiness, Church of God, Plymouth Brethren,
Independent Fundamentalist Churches of America, Independent-Fundamentalist ,
Lutheran Church--Missouri Synod; LC-MS, Wisconsin Evangelical Lutheran Synod;
WELS, Other Conservative Lutheran, Primitive Methodist, Congregational Methodist
(Fundamentalist), Methodist , Assemblies of God, Church of God (Cleveland, TN),
Church of God (Huntsville, AL), International Church of the Four Square Gospel,
Pentecostal Church of God, Pentecostal Holiness Church, United Pentecostal Church
International, Church of God in Christ (incl. NA whether), Church of God in Christ
(International), Church of God of the Apostolic Faith, Church of God in Prophecy,
Vineyard Fellowship, Open Bible Standard Churches, Full Gospel, Apostolic Pentecostal,
Spanish Pentecostal, Pentecostal ; Church of God, Presbyterian Church in American
(PCA), Evangelical Presbyterian, Reformed Presbyterian , Christian Reformed Church
(inaccurately known as "Dutch Reformed"), Free Hungarian Reformed Church, Christian
Churches and Churches of Christ, Churches of Christ; "Church of Christ" , Christian
Congregation

Mainline Protestant: Episcopalian; Anglican, Independent Anglican, Episcopalian,
United Church of Christ (includes Congregational, Evangelical and Reformed),
Congregational Christian, Church of the Brethren, Moravian Church, Quakers (Friends),
Evangelical Lutheran Church in America (formerly Lutheran Church in America and The
American Lutheran Church); ELCA, Lutheran, United Methodist Church; Evangelical
United Brethren, Presbyterian Church in the U.S.A., Cumberland Presbyterian Church,
Presbyterian, Reformed Church in America, Reformed, Christian Church (Disciples of
Christ)and white respondents falling into the Baptist and Southern Baptist categories
residing outside the political South.

Black Protestant: American Baptist Churches U.S.A., Baptist General Conference,
National Baptist Convention of America, National Primitive Baptist Convention of the
U.S.A., Progressive National Baptist Convention, Aﬁican Methodist Episcopal Church,
African Methodist Episcopal Zion Church, Christian Methodist Episcopal Church, all
black members of Evangelical and Mainline denominations and black respondents falling
into the Baptist and Southern Baptist categories, and black respondents, missing their
denominational response or refusing to respond, with church attendance more frequent
that yearly.

Roman Catholic.

Jewish. Jewish, no preference, Orthodox, Conservative, Reformed, Jewish other.

195

Secular: Agnostic, Atheist.

Other: Greek Rite Catholic, Greek Orthodox, Russian Orthodox, Rumanian Orthodox,
Serbian Orthodox, Syrian Orthodox, Armenian Orthodox, Georgian Orthodox, Ukranian
Orthodox, Eastern Orthodox, Muslim; Mohammedan; Islam, Buddhist, Hindu, Bahai,
American Indian Religions (Native American Religions), New Age, Wicca (Wiccan),
Pagan, Other non-Christian/non-Jewish, Scientology, Religious/ethical cultsMore than 1
major religion (e. g., Christian, Jewish, Moslem, etc.), Christian Scientists, Mormons;
Latter Day Saints, Spiritualists, Unitarian; Universalist, Jehovah's Witnesses, Unity;
Unity Church; Christ Church Unity, Fundamentalist Adventist (Worldwide Church of
God), Non-traditional Protestant.

1) Respondents with unclear responses are classiﬁed based on other characteristics.
Those with the high levels of attendance were coded evangelical. Those with low (about
once a month or less) or no reported attendance were coded secular.

2) White respondents answering Protestant, no denomination given, Non-denominational
Protestant, Community church, Inter-denominational Protestant, Christian, or "j ust
Christian" with the high levels of attendance were coded evangelical. Those with low or
no reported attendance were coded secular.

3) Black respondents answering Protestant, Non-denominational or Community church
with the high levels of attendance were coded black protestant. Those with low or no
reported attendance were coded secular.

Chapter 6 General Social Survey (1972-2002)
(Using variables DENOM and OTHER)

Evangelical Protestant: American Baptist Association, Baptist, Don't Know Which,
Advent Christian, Amish, Apostolic Christian, Apostolic Church, Assembly of God,
Bible Missionary, Brethren Church, Brethren, Brethren, Plymouth, Brother of ChriSt,
Calvary Bible, Chapel of Faith, Charismatic, Chinese Gospel Church, Christ Cathedral of
Truth, Christ Church Unity, Christian and Missionary Alliance, Christian Calvary
Chapel, Christian Catholic, Christian; Central Christian, Christian Reformed, Christ in
Christian Union, Christ in God, Churches of God Except with Christ and Holiness,
Church of Christ, Church of Christ, Evangelical, Church of Daniel’s Band, The Church of
God of Prophecy, Church of Prophecy, Church of the First Born, Church of the Living
God, Community Church, Covenant, Dutch Reformed, Evangelical Congregational,
Evangelical Covenant, Evangelical, Evangelist, Evangelical Free Church, Evangelical
Methodist, Evangelical United Brethren, Faith Christian, Faith Gospel Tabernacle, First
Christian, Four Square Gospel, Free Methodist, Free Will Baptist, Full Gospel, Grace
Brethren, Holiness Church of God, Holiness Nazarene, Holy Roller, Independent,
Independent Bible, Bible, Bible Fellowship, Independent Fundamental Church of

196

America, Laotian Christian, Living Word, Lutheran Church-Missouri Synod,

Macedonia, Mennonite, Mennonite Brethren, Missionary Baptist, Missionary Church,
Mission Covenant, Nazarene, New Testament Christian, No Denomination Given or
Nondenominational, Open Bible, Other Baptist Churches,Other Fundamentalist,

Other Lutheran Churches, Other Methodist Churches, Other Presbyterian Churches,
Pentecostal, Pentecostal Assembly of God, Pentecostal Church of God, Pentecostal
Holiness, Holiness, Pentecostal, People's Church, Pilgrim Holiness, Primitive Baptist,
Salvation Army, Seventh Day Adventist, Southern Baptist Convention, Swedish Mission,
Triumph Church of God, The Way Ministry, Wesleyan, Wesleyan Methodist-Pilgrim,
Wisconsin Evangelical Lutheran Synod, Vineyard Fellowship.

Black Protestant: Aﬁican Methodist, Aﬁican Methodist Episcopal Church, Aﬁican
Methodist Episcopal Zion Church, American Baptist Association', American Baptist
Churches in the U.S.A.', Apostolic Faith, Baptist, Don't Know Which', Christian
Tabernacle, Church of God in Christ, Church of God in Christ Holiness, Church of God,
Saint & Christ, Disciples of God, Federated Church, Holiness; Church of Holiness,
House of Prayer, Methodist, Don't Know Which I, National Baptist Convention of
America, National Baptist Convention, U.S.A., Inc.', Other Baptist Churches', Other
Methodist Churches', Missionary Baptist',Pentecostal Apostolic, Primitive Baptist,
Sanctiﬁed, Sanctiﬁcation, Southern Baptist Convention', United Holiness, Witness
Holiness, Zion Union, Zion Union Apostolic, and Zion Union Apostolic-Reformed.

Mainline Protestant: American Baptist Churches in the U.S.A., American Lutheran
Church, American Reformed, Baptist Northern, Christian Disciples, (Congregationalist,
First, Congregationalist), Disciples of Christ, Episcopal Church, Evangelical Lutheran,
Evangelical Reformed, First Christian Disciples of Christ, First Church, First Reformed,
Friends, Grace Reformed, Hungarian Reformed, Latvian Lutheran, Lutheran Church in
America, Lutheran, Don't Know Which, Methodist, Don't Know Which, Moravian,
Presbyterian Church in the U.S.A, Presbyterian, Don't Know Which, Presbyterian,
Merged, Quaker, Reformed, Reformed Church of Christ, Reformed United Church of
Christ, Schwenkfelder, United Brethren, United Brethren in Christ, United Church of
Canada, United Church of Christ, United Church of Christianity, United Methodist
Church, United Presbyterian Church in the U.S.A,

Other: Christadelphians, Christ Church Unity, Christian Scientist, Church of Jesus Christ
of the Restoration, Church Universal and Triumphant, Eden Evangelist, Jehovah's
Witnesses, Jesus Latter Day Saints, Latter Day Saints, Latter Day Saints -Morrnon, Latter
Day Saints -Reorganized, Mind Science, Mormon, New Age Spirituality, New Birth
Christian, Religious Science, True Light Church of Christ, Spiritualist, Unitarian,
Universalist, United Church, Unity Church, Unity, Worldwide Church of God
Roman Catholic.

Jewish. Jewish, no preference, Orthodox, Conservative, Reformed, Jewish other.

Secular: Agnostic, Atheist.

197

1) Respondents with unclear responses are classiﬁed based on other characteristics.
Those with the high levels of attendance were coded evangelical. Those with low (about
once a month or less) or no reported attendance were coded secular.

2) White respondents answering Protestant, Non-denominational or Community church
with the high levels of attendance were coded evangelical. Those with low or no reported
attendance were coded secular.

3) Black respondents answering Protestant, Non-denominational or Community church
with the high levels of attendance were coded black protestant. Those with low or no
reported attendance were coded secular. Those Black respondents answering “Other
<Baptist, Methodist, ect.. or “<Baptist, Methodist, etc.> -Don’t know which” were coded
Black Protestant. White respondents were coded Evangelical if “Other Baptist, Other
Lutheran, Other Methodist or Other Presbyterian”. White respondents were coded
Mainline Protestant if “Baptist -Don’t know which, Lutheran -Don’t know which,
Methodist, -Don’t know which or Presbyterian -Don’t know which”.

198

Appendix B
Chronicities and NES Master Codes

Candidate Qualities
201, 211-214 215-221, 301, 302, 305-320, 401-404, 407, 408, 411-422, 423-426, 431-
454, 601-604, 609, 703-710, 719

Parties
101,102,111,112,121,122,131-135,141,151,161-174,197, 501-520, 531-536, 597,
723, 728

Groups
531-536, 1201-1248, 1297

Social Issues (with topics)
Education

Values and Principles (including religion)
Abortion

Feminism

Health Care

Gun Control

Crime

Drugs

Environment
Homosexuality

Race

809-838, 847-849, 897, 914-925, 945-1045, 949-951, 962-964, 968-990, 991-993, 994-
1003, 1013-1024, 1059-1061

Economic (with topics)
Budget and the Economy, Government Reform, Tax Reform, Social Security, Welfare
and Poverty, Inﬁ'astructure, Jobs, Energy policy

553, 554, 605, 606, 805-808, 841-846, 901-913, 926-944, 952-961, 965-967, 1004-1012,
1025-1058

Foreign Policy( with topics)
Defense, War and peace

1101-1199

199

Appendix C

Multiple Chronicity Respondents

The multiple chronicity category represents one tenth of the US. population
which fail to provide a dominant chronicity in their responses to open-ended questions.
Theoretically these respondents may be drawing from a chronicity structure that allows
equal access to several chronicity, thus resulting in tied scores between two or more
chronicity categories. This category may also represent those persons lacking in a
political chronicity structure but unlike no issue content respondents they are willing to
give random responses which also resulting in tied categories.

Multiple Chronicity respondents were examined with exploratory factor analysis
to determine if their chronicity scores demonstrate an underlying structure. Candidate,
economy, parties, groups foreign policy and social issues scores were evaluated using
principle components analysis and varimax rotation for each election year to search for a
structure to the multiple chronicity respondent’s responses. The data itself has
questionable adequacy for using factor analysis as the Kaiser-Meyer-Olkin statistic for
the individual years ranged from a high of .500 in 2000 to a low of .318 in 1972. All of
the years ﬁt Kaiser’s category of unacceptable for factor analysis (Kaiser quoted in Kim
and Meuller 1978, p. 54).

Analysis of the diagonal elements on the anti-image correlation matrix for the
KMO individual statistics for each variable found a uniformity of scores that showed that
the unsuitability of the data for factor analysis is consistent across the variables and not
limited to one or two variables with low individual KMO values. For most years the near

linear shape of the plottings complicated analysis of the scree plot to determine the

200

number of factors. Therefore all eigenvalues above 1.0 were used on the examination
Comparison of rotated components (Table Cl) between years failed to reveal a consistent
identiﬁable pattern to the structure of the factors. The factor analysis reveal that the
multiple chronicity respondents are truly individualistic in their responses and the

category harbors no underlying aggregate structure.

201

.gonm 0.8 M. O>O£m wouOom wﬁocaaoo ~2—

 

cow.

o\ch
_N. _

e\ovm
_v. _

cea—
N..—

.Xém
2.—

$34.

33
mm.—

e\ow_
ma. _

_mm.

c\c_N
em.—

.XLN
mm.—

Axecm
_ N. _

cem—
ho. _

24m.

e\e_N
hm.—

gmm
om. _

cea—
3.—

34m.

{can
NN. _

o\c_m
mm.—

e\eoN
mm. _

o\e~N
3.—

m3.

KEN
en.—

{ce—
2..

o\o¢N o\o_N
om; am.—

2m.

28m
05

3§te>
.«e e\..
3328me

 

am.
no.
3..

cm.

on.

mm.

mm.

mm.

5.

8.

cc.

me.
we. mm.

an.

mm.

S.

3.
mm. _m.

it.

_w.

em.

3.

an.

No.

me.

ac.

we.

an.

 

 

=8"

 

ea—

 

": _

 

awe—

 

3a—

cae—

 

e3—

 

monum—
38m

8:8
cameo...—

83.5
been.
mquoom

essence

353.6%

 

 

 

.5835 saga: been eaceom 432:2 .5 asap

202

Appendix D: Trend Lines

To evaluate trends between groups over time there must be underlying theories
against which the trends can be compared. D. Garth Taylor argues that “an array of
possible theories can be used to describe a set of data and the ultimate theoretical
conclusion depends on the most parsimonious statistical model which can be used to
describe the data” (1980 87). Taylor (Table D1) suggests four possible theories with
seven possible outcomes which can be used for comparing two groups.

Theory 1 : No Change. This theory predicts that the changes observed between
survey years are merely random ﬂuctuations and that the percentage of change in each
year is actually zero. Differences in chronicities accessed between Evangelicals and the
Nonevangelical according to this theory do not exist.

Theory 2: Enduring Dijferences. This theory would predict that the differences in
conceptualization are constant. While the difference between the groups is nonzero the
differences are enduring and the correlations between the groups remain the same from
survey year to survey year.

Theory 3: Linear Patterns. This theory predicts that the changes in chronicity
usage between groups are occurring with a high degree of regularity. The end result is
that the differences can be best modeled by a linear pattern which shows an increasing or
decreasing difference.

Theory 4: Social Catastrophic. This theory predicts changes in conceptualization
that are erratic perhaps as the result of sudden occurrences in the political world. This last
theory in effect becomes a catchall for data, while not specifying what model type might

ﬁt the data.

203

Table D1. Taylor’s Decision Sequence for Evaluating Trends in Public Opinion1

 

 

Hypothesis Decision One Decision Two Model Type
No difference Accept if X2 N.S. > 1. d = 0
l Reject
Constant value Accept if X2 N.S.
Reject

Test reduction X2 2. Signiﬁcant: d=constant

from No difference
model on one df 3. NS: d=constant

borderline signiﬁcant

 

Linear Trend Accept if X2 N.S. Test reduction X2 4- Signiﬁcant:
from Constant =a+b(t1me)

difference model
on one df 5. NS: Borderline linear
Reject trend

Test reduction X2 6~ Signiﬁcant linear

 

from Constant component
difference model .
on one df 7. Nonlrnear trend

Explore further
nonlinear or
multiple linear
models

 

These theories move from the simplest theory which explains the data to the most
complex. The emphasis on parsimony requires moving to a more complex model only

when the data cannot be explained by any of the previous simpler models. Each of the

 

' Adapted from Taylor, 1980 p. 94.

204

ﬁrst three theories uses a Chi test built around a predicted value of the percentage of
change between years on each pairing between religious groups and the population. The
no change theory posits that the rate of change equals 0 (d=0). The enduring diﬁiarences
uses the average rate of change over the time period of the trend (d=average change). The
linear patterns uses a least squares estimate (d=constant + bx [time]). Three additional
theories are also tested by examining the difference between Chi test scores for each
model to determine if there is a borderline or signiﬁcant linear or constant component to
the trend. Trend lines for each chroncity were examined between Evangelicals and
Nonevangelicals (excluding seculars). Tables 2D-8D show the trend analysis over seven
elections from 1976 to 2000. With the exception of Social Issues the chronicities are
remarkably similar. Five of the trends (Candidates, Economy, Foreign policy, Parties and
No Content) exhibit no difference between Evangelicals and Nonevangelicals while

Groups and Multiple chronicities have a constant difference.

205

Table D2. Trend Analysis Evangelicals and Non-evangelicals in “Foreign Policy
Chronicity”: 1976-2000

Foreign Policy Trends

 

 

 

 

 

 

2 _
E
E 2 —
E»
8
o
u.
0
DD
8

m —l
i
0..

o _—

l l l l 1 l 1
l 976 1980 l 984 1988 l 992 1996 2000
Election Year
Evangelical ----- NonevangelicaLI
Category difference
(Base=Nonevangelical) Hypothesis Model X2 df p Decision
Evangelicals 1) no difference d=0 7.96 6 >.05 accept

 

206

Table D3. Trend Analysis Evangelicals and Non-evangelicals in “Groups
Chronicity”: 1976-2000

 

 

 

 

 

 

 

 

 

Groups Trends
2 —1

U) sf __.

Q u—t

:1

8

O

0

8°

1:

i? N ~

o 1—1
a.

2 4
l I f I l I
1 976 1980 1984 1988 1992 1996 2000
Election Year
Evangelical ----- Nonevangelical
Category diﬂerence
(Base =Nonevangelical) Hypothesis Model X2 df p Decision
Evangelicals 1) no difference d=0 19.88 6 <.05 reject
2) constant d= 2.17 6.10 5 >.05 accept
difference

 

207

Table D4. Trend Analysis Evangelicals and Non-evangelicals in “Parties
Chroniciy”: 1976-2000

Party Trends

12 14
1 1

Percentage Party
1 0
I

 

 

 

 

 

 

 

 

oo —1

\o .—

l I l I I l I
l 976 1980 1984 1988 1992 1996 2000
Election Year
Evangelical ----- Nonevangelical
Category diﬂerence

(Base =Nonevangelical) Hypothesis Model X2 df p Decision
Evangelicals 1) no difference d=0 8.56 6 >.05 accept

 

208

Table D5. Trend Analysis Evangelicals and Non-evangelicals in “Economy
Chronicity”: 1976-2000

 

Economy Trends

20 22 24
1 1 1

18

Percentage Economy

16
1

 

 

 

I l I I l T I
1976 1980 1984 1988 1992 1996 2000
Election Year

 

 

 

Evangelical ----- Nonevangelical

 

 

Category diﬂerence
(Base =Nonevangelical) Hypothesis Model X2 df p Decision

Evangelicals I) no difference =0 6.23 6 >.05 accept

 

209

Table D6. Trend Analysis Evangelicals and Non-evangelicals in “Candidate
Chronicity”: 1976-2000

Candidate Trends

Percentage Candidate
30 40 50
l L 1

20
1

 

 

 

 

 

 

 

 

C .
e— I I f I I I
1976 1980 1984 1988 1992 1996 2000
Election Year
Evangelical ----- Nonevangelical
Category diﬂ'erence

(Base=Nonevangelical) Hypothesis Model 712 df p Decision
Evangelicals 1) no difference d=0 8.71 6 >.05 accept

 

210

Table D7. Trend Analysis Evangelicals and Non-evangelicals in “Multiple
Chronicity”: 1976-2000

    

 

 

 

 

 

Multiple Chronicity Trends

2 ..

3 _
0
2i
'3
2
o N _
no —
S
t:
8
5
:1.

g a

\ .. — r ’ T ' -
00 —1
l 1 l l I l l
1976 1980 1984 1988 1992 1996 2000
Election Year
Evangelical ----- Nonevangelical
Category difference
(Base=NonevangelicaI) Hypothesis Model X2 if p Decision
Evangelicals 1) no difference d=0 12.68 6 <.05 reject
2) constant d= 7.91 5 >.05 accept
difference -1 .95

 

 

211

Table D8. Trend Analysis Evangelicals and Non-evangelicals in “No Content
Chronicity”: 1976-2000

 

 

 

 

 

 

 

 

No Content Trends

3 _

N _
E —1
8
t:
c
U
e
Z o _.
o '—
DD
.‘3
c:
i
a. go _

to J

l l 1 l I l l
1976 1980 1984 1988 1992 1996 2000
Election Year
Evangelical ----- Nonevangelical
Category dlﬂerence
(Base=Nonevangelica/) Hypothesis Model X2 df p Decision
Evangelicals 1) no difference d=0 6.58 6 >.05 accept

 

212

Appendix E
Religious Groups, Social Issue Religious Groups and Issue Means

Table El: Religious Groups, Social Issue Religious Groups and Race.‘l

 

 

Year Mean Standard N ANOVA F
Deviation
Aid to Blacks 43.41”.
Evangelical 5.55 a 2.15 2458
Social Issue Evangelical 5.10 a, b 1.76 218
Black Protestant 4.06 C 2.72 1453
Social Issue Black Protestant 4.56 c 2.54 135
Mainline Protestants 5.20 e b 2.03 2855
Social Issue Mainline Protestants 4.80 b 1.87 167
Roman Catholic 4.90 a. b 2.17 2797
Social Issue Roman Catholic 4.77 b 2.04 160
Secular 4.87 a, b 2.26 1453
Social Issue Secular 4.93 a: b 2.22 135
Population Mean 5.04 2.23 11426
Assistance to Blacks 95.59
Evangelicals 2.20 a .77 1567
Social Issue Evangelicals 2.23 a, b .81 188
Black Protestant 1.38 c .54 750
Social Issue Black Protestant 1.35 C .56 54
Mainline Protestants 2.16 a» b .66 1681
Social Issue Mainline Protestants 2.19 b .72 142
Roman Catholic 2.03 e b .74 1800
Social Issue Roman Catholic 2.02 b .64 132
Seculars 2.05 a, b .75 1008
Social Issue Seculars 2.02 e b .71 126
Population Mean 2. 04 . 75 7448
Black Thermometer 159.14""‘
Evangelicals 61.51 a: b 19.04 2237
Social Issue Evangelicals 62.35 a. b 19.90 208
Black Protestant 84.05 c 17.08 997
Social Issue Black Protestant 82.79 c 18.08 53
Mainline Protestants 59.69 a 18.61 2669
Social Issue Mainline Protestants 61.41 s b 19.05 168
Roman Catholic 6252 e b 19.29 2522
Social Issue Roman Catholic 67.02 b 19.48 153
Secular 5100 a 19.48 1330

213

Table E1 (cont’d).

Social Issue Secular 64.69 a, b 19.83 135

Population Mean 63. 62 I 9. 95 l l 4 72

 

aMeans with the same subscript are not signiﬁcantly different from each other. *** p < .001.

Table E2: Religious Groups, Social Issue Religious Groups and Homosexuality.‘

 

 

214

Year Mean Standard N ANOVA F
Deviation

Gays in the Military l6.59***
Evangelicals 3.16 a, b 2.19 954
Social Issue Evangelicals 3.34 a 2.21 129
Black Protestant 2.96 a, b 2.35 451
Social Issue Black Protestant 2.35 b, c, d 2.18 40
Mainline Protestants 2.50 b, c, d 2.13 819
Social Issue Mainline Protestants 2.13 d 2.00 89
Roman Catholic 239 b, ca d 2.1 1 1050
Social Issue Roman Catholic 1.92 d 1.93 96
Seculars 2.30 c, d 2.09 750
Social Issue Seculars 1.76 d 1.84 95
Population Mean 2. 62 2.18 44 73

Gay Adoption 1930*"
Evangelicals 4.40 ' 1.66 807
Social Issue Evangelicals 4.39 ' 1.63 110
Black Protestant 4.35 a 2.01 389
Social Issue Black Protestant 3.91 a, b 1.96 34
Mainline Protestants 3.79 a» b 2.1 1 681
Social Issue Mainline Protestants 3.24 c, d 2.35 71
Roman Catholic 2.76 a, b 2.21 878
Social Issue Roman Catholic 3.44 b 2.44 82
Seculars f 3.33 °’ d 2-23 629
Social Issue Seculars 2.58 c 2.03 88
Population Mean 3.86 2.10 3769

Gay and Lesbian Thermometer 58.69“"
Evangelicals 2435 a, b, c 25.92 1569
Social Issue Evangelicals 21.88 a 24.59 190
Black Protestant 38.2., c, 4 e 28.34 678
Social Issue Black Protestant 30.35 a, b, 0 27,89 49
Mainline Protestants 33.13 b, Ct d 25.73 1727

Table E2 (cont’d).

Social Issue Mainline Protestants
Roman Catholic

Social Issue Roman Catholic
Seculars

Social Issue Seculars

39.30 d’ e
40.50 d’ c
46.41 °’ f
42.20 °

51.00 f
35.22

Population Mean

26.79
25.98
27.58
27.28
27.53
27.27

146
1789
128
1019
126
7421

 

8Means with the same subscript are not different ﬁ'om each other. *"‘* p < .001.

Table E3: Religious Groups, Social Issue Religious Groups and Women’s

 

 

Issues.‘I
Year Mean Standard N ANOVA F
Deviation
Women’s Libber Thermometer 6614*"
Evangelicals 51.87 a, b 24.11 1966
Social Issue Evangelicals 46.28 a 26.05 166
Black Protestant 7012 d 22.60 814
Social Issue Black Protestant 76.15 9 19.70 45
Mainline Protestants 54.95 b, c 21.40 2283
Social Issue Mainline Protestants 60.73 C 22.17 129
Roman Catholic 59.65 c 21.00 2255
Social Issue Roman Catholic 60.19 c 22.73 135
Seculars 60.63 c 21.11 1164
Social Issue Seculars 70.78 9 20.91 116
Population Mean 58.19 22. 77 9310
Women Equal Roles 2870*"
Evangelicals 3.56 a 2.50 2395
Social Issue Evangelicals 3.23 a, b 2.04 217
Black Protestant 3.37 a 2.70 1097
Social Issue Black Protestant 3.04 a, b, C 2.30 57
Mainline Protestants 3.0, a b, c. d 2.23 2778
Social Issue Mainline Protestants 239 c, d, c 1.98 168
Roman Catholic 2.89 a b, c, d 2.30 2715
Social Issue Roman Catholic 2.33 d, e 1.97 161
Seculars 2.57 b, cs d, e 2.31 1427
Social Issue Seculars 2.02 c 1.91 134
Population Mean 3.05 2.38 11 l 49

 

215

a"Means with the same subscript are not different from each other. *** p < .001.

Table E4: Religious Groups, Social Issue Religious Groups and Abortion.‘

 

 

Year Mean Standard N ANOVA F
Deviation
Anti-Abortionist Thermometer 19.08“”
Evangelicals 55.54 as ba 6 29.97 692
Social Issue Evangelicals 66.03 a 32.20 71
Black Protestant 52.55 a, b. C» d 28.45 291
Social Issue Black Protestant 64.54 a, b 22.45 13
Mainline Protestants 47.79 c, e, f 26.62 963
Social Issue Mainline Protestants 49.06 b, c, d, e 26.15 67
Roman Catholic 56.28 a, b, C 25.72 817
Social Issue Roman Catholic 56.05 a, b: c 32.71 44
Seculars 33,59 4 f 26.37 342
Social Issue Seculars 34.67 f 31.06 39
Population Mean 51.38 28.18 3339
When Abortion Should Be Allowed 9522*"
Evangelicals 2.5190 1.03 2192
Social Issue Evangelicals 2.86 a 1.04 233
Black Protestant 2.36 b, c, d 1.17 990
Social Issue Black Protestant 2.56 a, b 1.18 64
Mainline Protestants 1.95 e, f .98 2350
Social Issue Mainline Protestants 1.76 f .96 174
Roman Catholic 2.24 c. d. e 1.08 2429
Social Issue Roman Catholic 2.19 d e 1.09 169
Seculars 1.71 f, 8 .97 1362
Social Issue Seculars 1.43 g .80 152
Population Mean 2.16 1.08 l 00 72

 

aMeans with the same subscript are not different from each other. "* p < .001.

216

Table E5: Religious Groups and Social Issue Religious Groups, and Values,

 

 

Morals and Lifestyles.a
Year Mean Standard N ANOVA F
Deviation
New Lifestyles Breakdown Society 3296*"
Evangelicals 4.16 a’ b 1.19 1263
Social Issue Evangelicals 4.23 3 1.06 163
Black Protestant 3.34 b, c 1.46 582
Social Issue Black Protestant 3.93 a, b» C 1,29 45
Mainline Protestants 3.88 a, b, C 1.17 1283
Social Issue Mainline Protestants 3.59 Ct d 1.32 116
Roman Catholic 3.79 b, C 1.24 1405
Social Issue Roman Catholic 3.62 C d 1.19 108
Seculars 3.36 d, e 1.44 885
Social Issue Seculars 3.08 C 1.41 113
Population Mean 3.82 1.1 7 5903
Must Change Views of Moral Behavior 2875""
Evangelicals 3.49 a» b 1.49 1264
Social Issue Evangelicals 3.85 a 1.46 163
Black Protestant 2.74 C 1.53 582
Social Issue Black Protestant 3.09 b, C 1.59 45
Mainline Protestants 3.15 b, C 1.40 1283
Social Issue Mainline Protestants 3.12 b, C 1.42 118
Roman Catholic 2.97 C 1.43 1406
Social Issue Roman Catholic M» b, C 1.32 107
Seculars 2.71 C 1.39 887
Social Issue Seculars 2.78 C 1.39 112
Population Mean 2.78 1.39 5967
Need for Traditional Values 2684*"
Evangelicals 4.43 a .92 1264
Social Issue Evangelicals 4.47 a .85 163
Black Protestant 4.219.th 1.32 582
Social Issue Black Protestant 4.09 b, C 1.20 45
Mainline Protestants 4.30 a, b .90 1282
Social Issue Mainline Protestants 406 b, c 1.07 118
Roman Catholic 4.27 a, b .99 1406
Social Issue Roman Catholic 4.06 b, C 1.05 108
Seculars 3.92 Ct d 1.20 886
Social Issue Seculars 3.64 d 1.28 113

217

Table E5 (cont’d).

 

 

 

Population Mean 4.24 .95 5936
Tolerant of Differing Morals 27.30‘“
Evangelicals 2.9, b 1.38 1263
Social Issue Evangelicals 3.37 a 1.41 163
Black Protestant 2.44 d, e 1.37 582
Social Issue Black Protestant 2.98 a, b 1.50 45
Mainline Protestants 2.66 b, C d 1.28 1283
Social Issue Mainline Protestants 259 b, c, d, C 1.16 118
Roman Catholic 2.47 C d, C 1.22 1404
Social Issue Roman Catholic 2.50 C d, C 1.20 107
Seculars 2.24 d, C 1.19 886
Social Issue Seculars 2.18 C 1.29 113
Population Mean 2.59 1.31 5964
3Means with the same subscript are not different from each other. *** p < .001.
Table E6: Religious Groups, Social Issue Religious Groups and the
Environment.a
Year Mean Standard N ANOVA F
Deviation
Environmental Spending 1613*"
Evangelicals 1.65 ab .65 1755
Social Issue Evangelicals 1.72 a .74 214
Black Protestant 1. 49 C .61 820
Social Issue Black Protestant 1. 45 b, C .59 60
Mainline Protestants 1. 57 a. b. c .61 1869
Social Issue Mainline Protestants 11.51 b, C .58 164
Roman Catholic 1. 50 b, c .59 2022
Social Issue Roman Catholic 1.54 a, b, C .61 148
Seculars 1.47 b. c .60 1168
Social Issue Seculars 1.23 d .47 141
Population Mean [.54 .62 836]
Environmentalist Thermometer 10.81 "*
Evangelicals 66. 13 a. b 20.68 1464
Social Issue Evangelicals 63.54 a 21.61 170
Black Protestant 72.28 b» c. d 20.28 643
Social Issue Black Protestant 73.04 C d 18.24 45

218

Table E6 (cont’d).

Mainline Protestants

Social Issue Mainline Protestants
Roman Catholic

Social lssue Roman Catholic
Seculars

Social Issue Seculars

Population Mean

67.61 8"”

68.41
71.12
67.81
69.80

69.51 d
68.80

amgd
b,c,d
a,b,c
a,b,c,d

20.10
21.15
18.98
19.32
19.75
19.71
20.38

1627
130
1660
1 19
4325
340
6973

 

2‘Means with the same subscript are not different from each other. *" p < .001.

219

Appendix F

Table F1. Variables Used in Chapter Five NES Models and Reduced Form

 

 

 

 

 

 

 

 

 

Eguations.
Variable Coding
Political 1. Lowest level of participation (none)
Participation Scale 6. Highest level of participation in campaign activities
(vcf0723)
Political Respondents general level of information about politics and
Information public affairs seemed:
(vcf0050a)
5. Very High
4. Fairly High
3. Average
2. Fairly Low
1. Very Low
Voted I. Voted '
(based on vcf702) 0. Did not vote
Chronicity Dummy Each Chronicity was coded:
Variables l. Chronicity
0. Otherwise
Party Identification Generally Speaking, do you usually think of yourself as a
(vcf0301) Republican, a Democrat, an Independent, or what?
1. Strong Democrat
2. Weak Democrat
3. Independent - Democrat
4. Independent - Independent
5. Independent - Republican
6. Weak Republican
7. Strong Republican
Party Identification
Squared
Internal Efficacy Scale created from “People like me don't have any say about what
(vcf0613 & the government does" (0- agree l-disagree) and “'Sometimes
vcf06l4) politics and government seem so complicated that a person like

 

me can't really understand what's going on” (0— agree l—disagree).

2. Highest Efﬁcacy
1 .
0. Lowest Efﬁcacy

220

 

Table F1 (cont’d).

 

Care who wins
(vcf031 l)

1. Don't care very much or DK, Pro-Con, Depends, and Other
2. Care a good deal

 

General Political
Interest
(vcf03l3)

Pay attention to Politics/Public Afﬁars
l. Hardly at all

2. Only now and then

3. Some of the time

4. Most of the time

 

Campaign Interest
(vcf0310)

1. Not much interested
2. Somewhat interested
3. Very much interested

 

 

 

 

 

 

 

 

Party Contact Did anyone from one of the political parties call you up or come
(vcf9030) around and talk to you about the campaign?
1. Yes
0. No
Strength of 1. Independent or Apolitical
Partisanship 2. Leaning Independent
(vcf0305) 3. Weak Partisan
4. Strong Partisan
TV Did the respondent watch TV programs about the campaign?
(vcf0725) 1. No
2. Yes
Radio Did the respondent hear about the campaign on the radio?
(vcf0726) 1. No
2. Yes
Newspapers Did the respondent read about the campaign in newspapers?
(vcf0727) 1. No
2. Yes
Age (vcf0101) Respondent age
Education 1. Grade school or less (0-8 grades)
(vcﬂll 10) 2. High school (12 grades or fewer, incl. non-college

 

training if applicable)

3. Some College (13 grades or more but no degree;
college, no identiﬁcation of degree status)

4. College or advanced degree

221

 

Table F 1 (cont’d).

 

 

 

 

 

 

 

 

 

Gender 1. Male
(vcf0104) 2. Female
Income 1. O to 16 percentile
(vcf0114) 2. 17 to 33 percentile
3. 34 to 67 percentile
4. 68 to 95 percentile
5. 96 to 100 percentile
Unemployed 1. Yes
(vcf0116) O. No
Urbanism: 1. Central cities
Currently live in 2. Suburban areas
(vcf0111) 3. Rural, small towns, outlying and adjacent areas
Where Grew Up 1. Urban
(vcf0134) 2. Rural
Church Attendance 5. Every week
(vcf0130 & 4. Almost every week
vcf0130a) 3. Once or twice a month
2. A few times a year
1. Never
Bible 1. The Bible is the actual Word of God and is to be taken literally,
(vcf0845 & word for word
vcf0850) 2. The Bible is the Word of God but not everything in it Should be
taken literally, word for word
3. The Bible is a book written by men and is not the Word of God
Political South 1. South
(vcf0113) O. Nonsouth
Religious Tradition Each tradition was coded:
Dummy Variables 1. Religious tradition
0. Otherwise

 

 

 

 

222

 

 

 

 

 

 

 

 

 

223

 

 

 

$3 $3 $3 $3 $3 $3 $8 $3
8.- 8.- 8.- 8.- 8.- 8.- 8.- 8.- Nau-mm=a>m-§.<
63 G3 A83 A88 33 A83 A83 38.3
.2 .- 8.- 8.- 8.. 8.- 8. 8.- .2. £028.20
$3 $3 $3 $3 $3 $3 $3 $3
.8.- ..8: 68.- a8.- ..8.- ...8.- a8.- ..8.- Aoeuwﬁm 3585
$3 $3 $3 $3 $3 $3 $3 $3
{hi—m.- ..»me ......Lmr ...:Nmr aha—Mr 2.5—m.- _.:.i~m.- .25—m.- BQEQU Nanak
$3 $3 $3 $3 $3 $3 $3 $3
3.38. :38. .232. _.:.:..N_. :38. ESE. :32. .....iN—. 39.23 39.53%
$3 $3 $3 $3 $3 $3 $3 $3
3.1.2. «......ON. ...»...ON. .....iom. ......aON. 2.32. ...:ON. :33. 8:5 0:3 0.80
A89 A83 A83 A83 A89 A83 A83 A88
:82. ...:2. 2.3:. 3.3.2. .....32. :32. 8.32. 3.5:. amp—35 HEEQEQU
:3 :3 :3 :3 :3 :3 :3 :3
8.- 8.- 8.- 8.- 8.- 8.- 8.- 8.- =o.§§80
$3 $3 $3 $3 $3 $3 $3 $3
.130. :50. 3.39. ......hc. ......ho. ......ho. Sync. .31-o. 2:005
$3 $3 $3 $3 $3 $3 $3 $3
8.- 8.- 8.- 8.- 8.- 8.- 8.- 8.- 5980
$3 $3 $3 $3 $3 $3 $3 $3
*8. a8. *8. a8. a8. *8. a8. a8. acumen—om
2.8.80 haze.— 853— .382 E 833...;
oz 2.5—...). 8385.0 maeaeum 85.35 3.5.5 :39...“— 33m
8.3.5:...—
3082 8.8.... coo—60¢ 522—6339 uni—.55 E 2.3. agar—9:5 nag—ESQ 833...; 33:35

 

853...; maeuoauam ..8 $535er 8.8% 533.0% ”nemuanmownam 53.830 40 ~33.

aaaoﬁocu as... .835. .8. ocean—Em who 8.5 .2...— "420mm
w 55...?

 

 

«mam.
...-8.8
2.8
N. .ua
8.~&
$8
**§¢o;
$3
8.
$3
8.-
$3
8.-
$3
8.
$3
8.-
$3
«go—-
:3
8.
$3
8.-
$3
8.-
$3
*oo.

 

mwoN.
......Nhéw
8.8
.88
8?".
$8
*n*mo;
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
«*o_.
:3
8.
$3
8.-
$3
8.-
$3
*mo.

 

mw8~_
tinge”
a...
Run
82"...
$8
***©o._
$3
8.
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
**o_.
:3
8.
$3
8.-
$3
8.-
$3
*mo.

 

«mom.
silk-ow
2.8
$8.8
3?...
$8
«ignc;
$3
aN—R
$3
8.-
$3
8.-
$3
8.
$3
8.-
$3
inc—-
:3
8.
$3
8.-
$3
8.-
$3
.mo.

 

mwom_
3.32.6”
2.8
8.8
swan“.
$3
u**mc;
$3

. _.
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
§*o_.
:3
8.
$3
8.-
$3
8.-
$3
inc.

 

mw8~_
3360.0”
2.8
8.";
8.8"...
$8
§§*vo;
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
§§O_.
:3
8.
$3
8.-
$3
8.-
$3
§mc.

 

mwcm.
...-8.8
8....
8.8
8.8"...
$3
***vo;
$3
8.
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
uuc—.
:3
8.
$3
8.-
$3
8.-
$3
«no.

 

anon.
2.3.55.0”
8....
8m.
88$ um
$3
«**mo;
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
8.-
$3
**c_.
:3
8.
$3
8.-
$3
8.-
$3
*oc.

 

Z
owmﬂﬁm-m

.M

cnbmoEoEU x 5.5-=02 N
as. 88.2820 82..
88800
b_u_=a-.._0 n ...;m-acz
383.35.» 0 Shack H SEW-.82
88.60 8.6% H Sim-.82
89.8.: 39.59% a. ream-20>.
8.59 on? 0.80 x §>méoz
$22:— cwmwaEmU x §>méoz
nozquuc X 29.0.82
0885 x §>méoz

hove—0G X §>moﬁoz

cozaosum x §>m-=oz
...—Ena- 25223.:

$883 .0 2an

 

224

A859: 85 .038 05 E cog—05 20>» corn—2.50 Eton you .828 8 moﬁatg be» 5:85
.8835 _ansbm 05 Eat @3388 03 83st? R433: .420mm 05 .«o 03% 383. 05 5 32:2: 3:8 3:28.. 5890 8 wow: 838:3 05 you
4885:on 5 Echo 2%ng $56508 mAO wonmuavqﬁmg 3.8%: 038 ma: .33 33.95 ._oo.va_.._:. ”3.3 :86}: ._o.va ”a... 3.0.3 :56?» .novn—a 6qu

225

 

 

 

 

 

 

 

 

 

 

$3 a... $3 a... $3 $3 a... $3
8. no. mo. we. no. mo. co. no. Buzmngm-noz
$3 $3 $3 $3 $3 $3 G3 83
9:6.- 8.- 2. ...-2. 8.- 8. 8. .$.. 3.9.8.5
$3 $3 $3 $3 $3 $3 $3 $3
2.....w_. 22.—N. 2.3.3. 22.—N. .15—N. 2.5—N. ......LN. 2.3—N. Snag-352
$3 $3 $3 $3 $3 $3 $3 $3
*2. 2.:LN. 2.3—N. 2.3—N. 2.3—N. ......ON. 22.—N. 2.3—N. 283.635
$3 $3 $3 $3 $3 $3 $3 $3
...wo. ...wo. 3o. ...wo. ..wo. .8. .8. .8. 2.8%
$3 $3 $3 $3 $3 $3 $3 $3
3.- nor 3.. co: 3.- 3.. 3.- m5.- BS§U are...
$3 $3 $3 $3 $3 $3 $3 $3
2.....mo. 2.2.06. 223°. 2.3.00. 2.3.09 2.2.06. 2.....co. 2.58. A§wncﬂtchw «9.4285
$3 $3 $3 $3 $3 $3 $3 $3
2.1.2. 2.5.2. 2.3.2. 2.32. 2.2.2. 2.1.2. 2.3.2. {2.2. ram-25 30.53%
$3 $3 $3 $3 $3 $3 $3 $3
...»...om. ....iNN. 2.....NN. 22.. — N. 2.....NN. 2.....NN. 2....NN. ...»...NN. $285 :wEQEmU
$3 $3 $3 $3 $3 $3 $3 $3
2.....3. :32. 21.2. 2.1.2. ’25—. 2.1.2. 2.3.2. .....ie. 2:3 0:3 0.80
$3 $3 $3 $3 $3 $3 $3 $3
:32. ...:2. 2.3.2. 2.2.2. 2.3.2. 2.2.2. 2.5.2. 22.2. 0:50...—
$3 $3 $3 $3 $3 $3 $3 $3
...:omr ......om... 22.6”. 33ri 22.5”- ......wN... .....inr ...:am.- 80:00
$3 $3 $3 $3 $3 $3 $3 $3
2.33”. 2.....cm. 22.0mm. 2.30m. 2.30m. 2.30m. 2.....em. atom. cotwosum
2.3.50 5:...— moamﬂ .252 ... 835...;
oz «Eu—=2 Sag—=30 hacnoom v.3... .5 3.5.0 5.95% 32%
scan—Ea...—

3932 ...-Sh v3.53— :SﬁﬁuEE— :ui—EaU ... 25,—. .33—330

.82.... 333.; 2.9.52.5

 

 

m2£¢€¢> nzcaowccum .8 35.9525 5......— ..ouscom unctaécunu .332...— .N0 033—.

226

 

883
38?.

$3
8.
$3
8.
$3
8.
$3
. ..-
$3
8.-
$3
8.-
$3

8.-
$3
8.
$3
*oc.
$3
8.
$3
8.-
$3
8.
$3
8.-

 

$3

8..

$3
8.

$8..
{©o.
96.0

co.
ANA...
95..
«we.

$3
8.-

 

89"..
88"..

$3

$3
8.
$3
8.
$3
*m_c
$3
8.-
$3
8.-
$3

8.-
$3
8.
$3
‘vo.
$3
8.
$3
8.-
$3
*wo.

$3
8.-

 

88.8
8.. .u.

$3
8.
$3
§c_r
$3
8.
$3
*w—f
$3
8.-
$3
8.-
$3

5.-

$3
8.

A8,.
55o.

$3
8.
$3
8.
$3
85¢.
$3
8.-

 

8E.
88"..
$3
8.
$3
$3
8.
$3
.m—f
$3
8.-
$3
8.-
$3

8.-
$3
8.
$3
.8.
$3
8.
$3
8.
$3
*wc.
$3
8.-

 

88.8
8a".

Goo

$3
8.-
$3
8.
$3
*m~r
$3
8.-
$3
8.-
$3

8.-

$8..
8.

$3
.8.
$3
8.
$3
8.-
$3
**ho.

$3
8.-

 

8i.
8..."...
$3
8.
$3
8.
$3
8.
$3
*m_r
$3
8.-
$3
8.-
$3

8.-
$3
8.
$3
ico-
$3
8.
$3
8.-
$3
8*wc.
$3
8.-

 

88.".
8... u.

$3

E...
co.-
$0..

$3
*m_r
$3

8.-
$3

8.-
$3

Na.-
5...
8.
$5..
.00.
$3
8.
$3
cc.-
$3
{‘56.
ANA...
5.-

 

onb_o_..o...0 x 5>méoz
...... 85.0.8.5 .8...

288.80

5.3.82.0 a ...a>m-..oz
.mmumuimz H ...aaM-toz
.85835 R £95.ch
0.83% a. ﬂaw-=02

88.6 0 kick a ﬂaw-=02

883388..
\c 58.225 x gnaw-=02

89.85 Banach. x .8318?
888.... $89.80 x .§>m-=oz
8:3 239 0.80 x .§>m-..oz

2:8... x .§>m-:oz

.0280 x .§>m-:oz

5.80:3. x .§>m-..oz
match. 5:228..—

.$....8. $0 08.8

 

227

A8523 85 388 2: E @3305 803 noun—oboe 3.8m .8 35:8 9 833.5, .80» so:
.8335 _annbm 05 Eat tons—88 0.8 333:8 83.53% ..dzomm 05 ..o cm?” 289. 05 E 39:05 258 3:23.. 58% 8 wow: 838:3 on
68855.8; 5 88.5 22.533 $58508 mqo Evuavaﬁmﬁ. $.82 038 mi..- 83 :86?“ .~oo.vq..:.;_ Umﬂ =86?“ .SVQ ”...... “88 53.93 .noVa... x

 

 

 

 

 

 

 

 

 

 

aahﬂ a2. 2 aahg $2.2 oak-2 @052 thM— Qahg Z

*
:nﬁmeN .31-ch ***ah.owN «SK-N. SN :iwmdem .....ithbN ......» GdON Egg-CON 03$qu
mv. 3... _v. :V. _v. :u. _v. 3.. W—

$883 No use.

 

228

229

 

 

 

 

 

 

 

 

 

8. 8. 8. 8. 8. 8.- 8. 8. 885.333
$3 $3 $3 $3 $3 $3 $3 $3
333hmr 3.350- 333hmr uuohmr 33350- 3.36m.- 333hmr 3.3be 35:00 buds—
Amoo AMOQ A2: 39v Amoo Ame-V Ana-v Ame-V
3.32. .332. .333. 3.32. 3.32. 3.32. 2.32. 3.32. A~&§u&~§k\o £3555.
Ame-v 30¢ Amog Ana-v Amoo Ana-v Ame-v Ame-v
35o. Lb. 35c. 35¢. Kb. 350. 35c. 35o. “am-.35 Bottom
Gog 33 33 Goo Goo Gee Gag Gog
333cm. 33;. 3.3—m. .33—m. .33—m. 3.3—m. .33—m. .33—m. mi? 0:3 0.30
$3 $3 $3 $3 $3 $3 $3 $3
33VN. 33VN. 339N- 33VN. 33¢” .33?” 33V” 3...?” “mo-.85 $39.30
$3 $3 $3 $3 $3 $3 $3 $3
No.- Ncr No.- No.- mcr No.- No.- No.- 29.8me 3.30
$3 $3 $3 $3 $3 $3 $3 $3
3...:- 33:. 3.3:. 3.3:. .33:- _.:.3:. 3.3:- ..33:. 00:50:83. nous—.0
$3 $3 $3 $3 $3 $3 $3 $3
8.- 8.- 8.- 8..- 8.- 8.- 8.- 8.- 383.5%:
:3 :3 :3 :3 :3 :3 :3 :3
_o.- oo.- 8.- co.- co.- cor 8.- Sr 2.8: S Ema-5
$3 $3 $3 83 $3 $3 83 83
.338.- 338.- .....3oov 3.38.- 3338.- 236°.- 33369- .338.- wanna-owlx
Soc :3 :3 £3 :3 29V :3 :3
33cc. 3330c. 333cc. 333cc. 33cc. 330°. 3309. 330c- ow<
Amco A89 Ame-v Ame-v Ame-v Ame-v AMA: AmCQ
332. 333. 333. .333. 333. 333&—. 333. 3.33. 2:00:—
$o.v $3 $90 $3 $3 $90 $3 $3
33mN. 33cm. 33cm. 330m. 330m. 3.3mm. 333cm. 330N- corn-“.58
2.3.30 hug—o..— noamm— _ovo—Z E 833.3,
cz 2.5—=2 8%.—=30 ...—.853. 85.....— mnschw 5.9....— ....on
...—252 ...-8...— cousvom «58> E 2.3. 3.3.5.. :0 «£25 83......5 Eon—.095

 

 

 

833.3.» mscuowacam ...... £56530 ...-S..— 3953“ Hunt...» .m0 035—.

 

$3

$3
8.-
$3
8.-
$3
3m _ .
$3
3.-
$3
8.-
:3
8.
$3
3°C.
:3
3N9-
$3
8.
$3
8.-

$3
S.
83
8.-
$3

 

$3
8.
$3
8.-
$3
8.-
$3
30 _ .
$3
8..
$3
8.-
:3
8.
:3
...oo.
:3
3N6.-
$3

$3
8.-

$3
G.
$3

$3

 

$3

$3
$3
8.-
$3
32 .
$3

8.-
$3
8.-
:3

8.
$3
306.
:3
.8.-

$3

$3
8.-

$3
8.
$3
8.-
$3

 

$3
8.
$3
8.-
$3
8.-
$3
3m _ .
$3
8..
$3
8.-
:3
8.
83
...oo.
:3
3N9-
$3
5.
$3
8.-

A23
an.
$3

$3

 

$3

$3
8.-
$3
8.-
$3
30 _ .
$3
8.-
$3
8.-
:3
8.
$3
.8.
:3
...Nor
$3

$3
8.-

$3

$3
8..
$3

 

$3

$3
8.-
$3

 

$3
3.
$3
8.-
$3
8.-
$3
32 .
$3
8..
$3
8.-
:3
3.
$3
360.
:3
3N9-
$3

$3
3.-

$3
: 3

no.
$3

 

$3

$3
$3
8.-
$3
3n — .
$3
8.-
$3
8.-
:3
8.
$3
.8.
:3
3N0.-
$3

$3
8.-

$3

: 3
8.
$3

 

8.9.85 Enzack a ...?m-noz
$53 055 0.50 x .§>m-:oz
$285 38950 x ﬁgm-coz
20.28% 865 H gnaw-=02
8.8323. 55:0 x .cgm-coz
nmaoismnb x «SP-«.202
was: .... 395 R gnaw-.62
uoaaum-ow< x .53-7.82

3.... x .53-“$52

2:85 x .§>m-:oz

5:825 x .:m>m-:oz
mar—oh. £5223:—
Nuuamngm-zoz
8.2820

$2.83 8 28$

 

230

A552?- 85 _uvoE 05 5 nous—2: 89$ coca—9:8 _mtom ..8 .9280 3 mozutg an»
sauce—m gowns—um 3.562% 05 EB,“ cog—88 2a BEE-a; «$33: .m-Emm 2: mo 093 9.88 05 E cog—2: 2:8 3:28.. 5520 8 com: 22:53 05

 

 

.6.“ 330553...— E ﬂoto 2%ng $56508 uQNEHmcqﬁmcs atone 038 $5- 38 __Séz ..oonIin uma =86»: ..o.vm n1; :3. 29-95 .mon... 6qu
an: as: a”: a»: an: as: am: a»: z
2. 2. 2. M2. 2. 2. 2. 2. E833
..m—AEWM ....ithVoN SIR-NWVON &u§hM.N$N ......»wVAVON ......NMNVON ....iowAvoN {*ivoévoN omumﬁﬁm Nx
$3 $3 $3 $3 23 $3 $3 $3
.....iwnﬁ- five-m- 3.3—9m.- ..ERQM- Etyvod- .....ioodu ......Vod- i**vw.Mn ESE—CU
a: a: .3 E3 33 c 3 2: .3 a. .3 93
m _ .- E .- 2. 3.- 8. 3. 3.- 2. $2820 x $5-82
$3 $3 $3 $3 $3 $3 $3 $3
8. 3. s. 8. S. 3. 8. S. ...-983% 3525 a 32.3-52
33 $3 $3 $3 Q3 $3 $3 $3
3. 8. 8. 3. 8. 8. 8. 8. 88:8 ban x $5.82
$33 $3 $3 $3 $3 $3 $3 $3
38%:3
3.- 3.- 8.- 3.- 3.- 3.- 3.- 3.- 3 £323. a .=§m-§2

 

 

 

 

 

 

 

 

 

.5203 8 29¢

231

Appendix H

Chapter Five ZSCML and ZSPLS Tables

Table H1. Foreign Policy Chronicity, Participation and Information : ZSCML
Results

 

 

 

Variables in Model Foreign Policy S. E. Marginal
Effectsl

Demographics

Education -.693 * * (.05 7) -.05
Income -.262** (.030) -.02
Gender .350" (.053) .03
Married .003 (.046) -.00
Intelligence -.032 (.040) -.00
Age -.024** (.007) -.Ol
Age Squared .000 (.000) .00
Urban .059* (.029) .00
Where Grew Up -,049 (.048) -.00
Political South .069 (.047) .00

Political Variables

Party Identiﬁcation .278" (.056) -.00
Party Identiﬁcation Squared -.03 I ** (.007) -.01
Campaign Interest -.515** (.051) -.04
Care Who Wins -.308** (.052) -.03
Political Participation -l .609“ * (.550) -. 12
Participation Residual -.020 (.024) -.00
Political Information 2.519" (.355) .19
Information Residual —.003 (.030) .00
Religion Variables

Bible Attitudes -.001 (.163) -.00
Church Attendance -. l 64* (.078) -.02
Evangelical ,3 83 (.661) .0]
Mainline Protestant .400 (.656) .00
Catholic .149 (.661) -.01
Black Evangelical -.S87 (.769) -.O4
Secular .534 (_652) .06
Jew .056 (1.323) .01

Election Year

232

 

 

Table H1 (cont’d).
1980
1984
1988
1992
1996
2000

Interaction Terms

Evangelical x Education

Evangelical x Income

Evangelical x Political Participation
Mainline x Political Participation

Black Protestant x Political Participation
Catholic x Political Participation
Secular x Political Participation

Jewish x Political Participation

Evangelical x Political Information
Mainline x Political Information

Black Protestant x Political Information
Catholic x Political Information
Secular x Political Information

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

233

.613“
.593"
.479"
-.163

-.280*

-.416**

.058
.043

l .215*
1.042
-.276
.466
.041
.622

-.858*
-.668*
.282
-.330
-.136
-.422

.129
.146
.051
.126
.157

.043

-.086
-.082
-.029
-.098
-.002

(075)
(.072)
(076)
(089)
(135)
(110)

(084)
(052)

(596)
(577)
(705)
(584)
(617)
(909)

(362)
(339)
(409)
(341)
(360)
(531)

(085)
(085)
(110)
(085)
(197)

(180)
(176)
(219)
(179)
(179)
(306)

.08
.07
.05
-.01
-.02
-.02

.00
.00

.09
.09
-.02
.04
-.00
.05

-.06
-.06

.02
-.03
-.01
-.03

.01
.02
.01
.01
.01

.00
-.00
-.00

.00

.00
-.00

 

 

Table H1 (cont’d).

Constant -3. l 85* * (.689)
Observations l 1999

x2 Statistic 683.96"*

Pseudo R2 .13

 

Note: "' p<.10%, one tail test; "p<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-Stage Conditional Maximum
Likelihood model (Rivers and Vuong 1988).

lMarginal effects are calculated at the mean/mode of the independent variables.

Table H2. Groups Chronicity, Participation and Information : ZSCML Results

 

 

 

Variables in Model Groups S. E. Marginal
Effectsl

Demographics

Education -.630* * (.042) -. 12
Income -200" (.023) -.04
Gender .405" (.041) .07
Married .103" (.036) .02
Intelligence .100“ (.030) .02
Age .021" (.005) .004
Age Squared -.000** (.000) -.00
Urban -.032 (.022) -.01
Where Grew Up .003 (.03 7) -.00
Political South -.203 * * (.03 8) -.04

Political Variables

Party Identiﬁcation -.019 (.041) -.08
Party Identiﬁcation Squared -.022** (.005) .07
Campaign Interest -,454** (.03 8) -.08
Care Who Wins -.130** (.041) -.03
Political Participation -1 .427" (.369) -.25
Participation Residual -.012 (.018) -.00
Political Information 2. 148* * (.240) .39
Information Residual .061M (.022) .01
Religion Variables

Bible Attitudes -.140 (.1 18) -.04
Church Attendance .019 (.055) -.00
Evangelical -.3 82 (.474) -.08

234

 

 

 

Table H2 (cont’d).
Mainline Protestant
Catholic

Black Evangelical
Secular

Jew

Election Year
1980
1984
1988
1992
1996
2000

Interaction Terms
Evangelical x Education

Evangelical x Income

Evangelical x Political Participation
Mainline x Political Participation

Black Protestant x Political Participation
Catholic x Political Participation
Secular x Political Participation

Jewish )1 Political Participation

Evangelical x Political Information
Mainline x Political Information

Black Protestant x Political Information
Catholic x Political Information
Secular x Political Information

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

235

-.3 19
-.026
.031
-.248
.190

-.192**

.005
.016

-.378**

-.190*

-.226**

-.065
-.O67

.788
-.214
.729
.107
.678
.389

-.256
.153

-.359
-.098
-.352
-.408

-.048
-.022
-.O7l
-.042
-.O37

(464)
(461)
(493)
(460)
(838)

(060)
(054)
(055)
(062)
(078)
(062)

(.068)
(.043)

(418)
(396)
(422)
(392)
(427)
(590)

(259)
(232)
(247)
(229)
(249)
(347)

(062)
(062)
(067)
(060)
(119)

-.O7
-.02
-.03
-.05

.04

-.03
.00
.00

-.06

-.O3

-.03

-.01
-.01

.15
-.05
.10
.01
.12
.04

-.06

.02
-.05
-.02
-.O7
-.07

.00
.00
-.00
-.00
-.00

 

 

Table H2 (cont’d).

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

Constant

Observations
x2 Statistic

Pseudo R2

.183
.053
.130
.085
.136
.245

-2.875**

11999

1285.22***

.13

(.135)
(.130)
(.141)
(129)
(.133)
(.208)

(.485)

.05
.02
.04
.03
.04
.05

 

Note: * p<.10%, one tail test; Mp<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefficients (standard errors in parentheses) from the 2-Stage Conditional Maximum

Likelihood model (Rivers and Vuong 1988).

lMarginal effects are calculated at the mean/mode of the independent variables.

Table H3. Parties Chronicity, Participation and Information : ZSCML Results

 

 

 

Variables in Model Parties S. E. Marginal
Effectsl
Demographics
Education .163" (.045) .02
Income -.005 (.025) -.00
Gender -.187** (.044) -.02
Married -.068 (.039) -.01
Intelligence -.O67* (.032) -.01
Age -.004 (.006) -.00
Age Squared .000 (.000) .00
Urban -.012 (.025) -.00
Where Grew Up -.008 (.041) -.00
Political South -.006 (.040) -.00
Political Variables
Party Identiﬁcation -.327** (.045) .04
Party Identiﬁcation Squared .039" (.006) .05
Campaign Interest -.1 15" (.042) -.02
Care Who Wins .036 (.046) .01
Political Participation 1.3 78** (.441) .18

 

 

 

Table H3 (cont’d).
Participation Residual
Political Information

Information Residual

Religion Variables
Bible Attitudes
Church Attendance
Evangelical
Mainline Protestant
Catholic

Black Evangelical
Secular

Jew

Election Year
1980
1984
1988
1992
1996
2000

Interaction Terms
Evangelical x Education

Evangelical x Income

Evangelical x Political Participation
Mainline x Political Participation

Black Protestant x Political Participation
Catholic x Political Participation
Secular x Political Participation

Jewish x Political Participation

Evangelical x Political Information
Mainline x Political Information

Black Protestant x Political Information
Catholic x Political Information

Secular x Political Information

237

-.001

-.857**

-.015

.367*
-.O93
.584
.651
.673
.357
.358
.548

.493**
.455**
.598**
.713**
.712**
.875“I

.023
-.053

-.048
1.01 1*
.907
.938“
.821
.600

.014

-.517
-.305
-.467
-.297

(019)
(285)
(024)

(146)
(066)
(556)
(546)
(548)
(591)
(543)
(956)

(068)
(062)
(063)
(073)
(085)
(067)

(072)
(046)

(483)
(464)
(505)
(467)
(499)
(668)

(300)
(278)
(302)
(279)
(298)

.00
-.l l
-.00

.05
-.01
.11
.10
.11
.07
.09
.18

.09
.08
.ll
.13
.16
.19

.01
-.01

.01
.15
.13
.14
.13
.09

-.01
-.07
-.04
-.07
-.05

 

 

Table H3 (cont’d).

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

Constant

Observations
x2 Statistic

Pseudo R2

-.070

.143 *
.073
.036
.079
-.O48

-.373*
-.320"‘
-.333

-.380*
-.361*
-.541*

-l.140*

11999
607.53***

.08

(401)

(072)
(072)
(080)
(071)
(145)

(162)
(156)
(171)
(157)
(160)
(234)

(.565)

-.01

.01
.01
.00
.01
-.02

-.06
-.05
-.05
-.05
-.05
-.09

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-Stage Conditional Maximum

Likelihood model (Rivers and Vuong I988).

lMarginal effects are calculated at the mean/mode of the independent variables.

Table H4. Candidate Chronicity, Participation and Information: ZSCML Results

 

 

 

Variables in Model Candidate S. E. Marginal
Effectsl

Demographics

Education -.408** (.035) -.I I

Income -.149“‘* (.019) -.04

Gender .412“ (.034) . 12

Married -.068* (.030) -.02

Intelligence .020 (.025) .00

Age -.006 (.004) -.00

Age Squared .000" (.000) .00

Urban -.005 (.019) -.00

Where Grew Up -.075* (.031) -.02

Political South -.036 (.031) -.01

238

 

 

 

Table H4 (cont’d).
Political Variables

Party Identiﬁcation

Party Identiﬁcation Squared
Campaign Interest

Care Who Wins

Political Participation
Participation Residual
Political Information
Information Residual

Religion Variables
Bible Attitudes
Church Attendance
Evangelical
Mainline Protestant
Catholic

Black Evangelical
Secular

Jew

Election Year
1980
1984
1988
1992
1996
2000

Interaction Terms
Evangelical x Education

Evangelical x Income

Evangelical x Political Participation
Mainline Protestant x Political Participation
Black Protestant x Political Participation
Catholic x Political Participation

Secular x Political Participation

Jewish x Political Participation

.4l4**
m044**
a243**
n236**
a652*
.007
L606**
a023

.038
.030
.599
.997*
.555
.199
.509
1.347

2593**
n97l**
a730**
-L029**
n641**
«854**

.175"
.097“

.910"
-.633
-.489
-.605
-.573
-.570

239

(035)
(004)
(.032)
(.034)
(316)
(015)
(204)
(018)

(102)
(046)
(412)
(404)
(407)
(466)
(404)
(744)

(047)
(046)
(046)
(050)
(062)
(051)

(054)
(.035)

(350)
(333)
(403)
(336)
(366)
(514)

-.02
-.10
-.07
-.08
-.15
-.00
.45

-.00

.01
.01
.24
.38
.19
.10
.16
.56

-.15
-.23
-. 18
-.24
-. 16
-.20

.05
.03

.23

-.23
-. I 6
-.22
-.22
-.22

 

 

Table H4 (cont’d).

Evangelical x Political Information
Mainline Protestant x Political Information
Black Protestant x Political Information
Catholic x Political Information

Secular x Political Information

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

Constant

Observations
x2 Statistic

Pseudo R2

-.811**

.078
.057
.179
.147
.240

-.032
-.038
-.014
-.019
-.138

-.001
-.008
.139
.016
-.012
-.370*

-3.l49**

1 1999

1326.6?"

.09

(216)
(196)
(232)
(197)
(214)
(301)

(051)
(051)
(064)
(050)
(106)

(115)
(110)
(132)
(111)
(114)
(181)

(.423)

-.22
.04
.02
.08
.07
.09

-.Ol
-.02
-.01
-.01
-.05

.00
-.00
.04
.00
-.01
-. 12

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; "“; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-Stage Conditional Maximum

Likelihood model (Rivers and Vuong 1988). Marginal effects are calculated at the mean/mode of the

independent variables.

Table H5. Multiple Chronicity, Participation and Information : ZSCML Results

 

 

 

Variables in Model Multiple S. E. Marginal
Enema!

Demographics

Education -.046 (.042) -.01

Income .047* (.023) .01

Gender .073 (.042) .01

Manied .041 (.03 6) .0 I

240

 

 

 

Table H5 (cont’d).
Intelligence

Age

Age Squared

Urban

Where Grew Up
Political South

Political Variables

Party Identiﬁcation

Party Identiﬁcation Squared
Campaign Interest

Care Who Wins

Political Participation
Participation Residual
Political Information

Information Residual

Religion Variables
Bible Attitudes
Church Attendance
Evangelical
Mainline Protestant
Catholic

Black Evangelical
Secular

Jew

Election Year
1980
I984
1988
1992
1996
2000

Interaction Terms
Evangelical x Education

Evangelical x Income

.026
-.003
.000
.005
.050
.01 1

.003

.001
.060
.167"
-1.1 l8M
-.005
.468
-.002

-.061
.055
.523
.853
.075
.051
.488
.453

-.O60

-.167**
-.234**
-.364**
-.398**
-.247**

.054
-.017

241

(030)
(005)
(000)
(023)
(037)
(038)

(043)
(005)
(038)
(041)
(390)
(020)
(245)
(022)

(123)
(057)
(468)
(463)
(465)
(513)
(457)

(L007)

(058)
(055)
(057)
(064)
(081)
(063)

(064)
(041)

.00
-.00
.00
.00
.01
.00

.01
.00
.01
.03
-.19
-.00
.07
-.00

-.01
.01
.1 l
.21
.02
.03
.09
.13

-.01
-.03
-.04
-.05
-.05
-.04

.01
-.00

 

 

Table H5 (cont’d).

Evangelical x Political Participation .162 (.431)
Mainline Protestant x Political Participation -.278 (.417)
Black Protestant x Political Participation -.3 66 (.470)
Catholic x Political Participation -.151 (.417)
Secular x Political Participation -.202 (.456)
Jewish x Political Participation -.5 88 (.770)
Evangelical x Political Information -.268 (.258)
Mainline Protestant x Political Information -.118 (.240)
Black Protestant x Political Information .122 (.266)
Catholic x Political Information -.067 (.239)
Secular x Political Information -.093 (.259)
Jewish x Political Information -.162 (.414)
Evangelical x Attendance —.O71 (.062)
Mainline Protestant x Attendance -.050 (.063)
Black Protestant x Attendance -.009 (.072)
Catholic x Attendance -.006 (.062)
Jewish x Attendance .096 (.141)
Evangelical x Bible Attitudes .1 14 (.138)
Mainline Protestant x Bible Attitudes .009 (.135)
Black Protestant x Bible Attitudes .108 (.152)
Catholic x Bible Attitudes .204 (.135)
Secular x Bible Attitudes .067 (.137)
Jewish x Bible Attitudes .341 (.25 8)
Constant -1.473** (.484)
Observations I 1999

x2 Statistic 329.93****

Pseudo R2 .04

.03
-.04
-.06
-.02
-.04
-.O9

-.O4
-.02

.02
-.01
-.01
-.03

-.01
-.01
-.01
-.00

.01

.02

.02
.03
.01
.05

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-Stage Conditional Maximum
Likelihood model (Rivers and Vuong I988).

lMarginal effects are calculated at the mean/mode of the independent variables.

242

 

Table H6. Social Issue Chronicity, Voting and Information : ZSPLS Results

 

 

 

Variables in Model Social Issues S. E. Marginal
Effectsl
Demographics
Education -.470*"‘* (.046) -.04
Income -.169"' "‘ * (.027) -.02
Gender 140*" (.048) .07
Married .052 (.041) .00
Intelligence -.012 (.029) —.00
Age -.004 (.007) -.00
Age Squared -.000 (.000) -.00
Urban -.041 (.027) -.00
Where Grew Up .050 (.044) .00
Political South -.01 7 (.042) -.00
Political Variables
Party Identiﬁcation Squared -.176"* (.060) -.02
Party Identiﬁcation -.207*** (.059) -.02
Campaign Interest -.586*** (.046) -.05
Care Who Wins -.246*" (.050) -.02
Voting -.082 (. I 79) -.01
Political Information I .482“ * * (.238) .14
Religion Variables
Bible Attitudes .101 (.134) .01
Church Attendance -.009 (.054) -.00
Evangelical -2.847* * "' (.981) -.26
Mainline Protestant -.l97 (.925) -.03
Catholic -.61 1 (.93 l) -.06
Black Evangelical -1.098 (1.052) -.10
Secular .455 (.870) .03
Jew -.113 (1.376) -.01
Election Year
1980 434*" (.089) .04
I984 637*" (.080) .06
1988 812*" (.080) .07
I992 .601 "* (.081) .05
1996 1081*" (.095) .10
2000 1.078" *"‘ (.082) .10

Interaction Terms

 

 

Table H6 (cont’d).

Evangelical x Education
Evangelical x Income

Evangelical x Voting
Mainline x Voting

Black Protestant x Voting
Catholic x Voting

Secular x Voting

Jewish x Voting

Evangelical x Political Information
Mainline x Political Information

Black Protestant x Political Information
Catholic x Political Information
Secular x Political Information

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

Constant

N

x2 Statistic

Pseudo R2

-.219***
.006

-.691***
-.012
-.154
-.031
.122
-.052

959*"
.072
.337
.145
-.096
.169

.238"”‘“'I
.025
.030
.032
.012

-.153
-.O80
-.189
-.038
.035

-. 148

4454*"

13909
877.43*"

.13

(.074)
(.050)

(201)
(181)
(199)
(183)
(189)
(230)

(257)
(230)
(261)
(231)
(235)
(334)

(064)
(062)
(080)
(062)
(114)

(152)
(149)
(178)
(151)
(149)
(221)

(928)

-.02
.00

—.06
-.00
-.01
-.00

.01
-.00

.09
.01
.03
.02
-.Ol
.02

.02
.00
.00
.00
.00

-.02
-.01
-.02
-.01

.00
-.Ol

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; "*; p<.01, two-tail test. This table reports

unstandardized coefﬁcients (standard errors in parentheses) from the 2-stage probit least squares model.
lMarginal effects are calculated at the mean/mode of the independent variables.

244

 

Table H7. Foreign Policy Chronicity, Votingand Information : ZSPLS Results

 

 

 

Variables in Model Foreign Policy S. E. Marginal
Effectsl
Demographics
Education -.54*** (.05) -.04
Income -. 14*" (.03) -.01
Gender 27*" (.05) .02
Married -.00 (.04) -.00
Intelligence -.04 (.03) -.00
Age .01 (.01) .00
Age Squared -.00*** (.00) -.00
Urban .05 (.03) .00
Where Grew Up -.00 (.04) -.00
Political South .06 (.04) .00
Political Variables
Party Identiﬁcation Squared .07 (.06) .01
Party Identiﬁcation -.09 (.06) -.01
Campaign Interest -.42*" (.05) -.03
Care Who Wins -.21*" (.05) -.02
Voting -.78*** (.25) -.06
Political Information 227* * * (.34) .1 8
Religion Variables
Bible Attitudes -.O7 (.16) -.01
Church Attendance -.03 (.05) -.OO
Evangelical 2.64“ (1.21) .19
Mainline Protestant .84 (1.20) .06
Catholic 1.29 (1.21) .09
Black Evangelical -.25 (1.48) -.03
Secular 1.45 (1.17) .10
Jew .51 (2.06) .03
Election Year
1980 67*" (.07) .05
1984 63*“ (.06) .05
1988 46*" (.07) .04
1992 .03 (.07) .002
I996 -. I 7 (.12) -.02
2000 -.02 (.10) -.OO

Interaction Terms

 

 

Table H7 (cont’d).

Evangelical x Education -.00 (.08) -.00
Evangelical x Income -.04 (.05) -.OO
Evangelical x Voting .46 (.27) .03
Mainline x Voting .02 (.26) .00
Black Protestant x Voting .03 (.32) .00
Catholic x Voting .23 (.27) .02
Secular x Voting .12 (.27) .01
Jewish x Voting .16 (.42) .01
Evangelical x Political Information -.77* (.36) -.05
Mainline x Political Information -.33 (.34) -.03
Black Protestant x Political Information -.02 (.41) .00
Catholic x Political Information -.45 (.34) -.O3
Secular x Political Information -.38 (.36) -.03
Jewish x Political Information -.28 (.54) -.02
Evangelical x Attendance .03 (.06) .00
Mainline Protestant x Attendance .I3* (.06) .01
Black Protestant x Attendance .03 (.09) .00
Catholic x Attendance .08 (.06) .01
Jewish x Attendance .07 (.16) .01
Evangelical x Bible Attitudes .06 (.17) .Ol
Mainline Protestant x Bible Attitudes -.05 (.17) -.00
Black Protestant x Bible Attitudes .02 (.21) .00
Catholic x Bible Attitudes .01 (.17) .00
Secular x Bible Attitudes -.02 (.17) .00
Jewish x Bible Attitudes .05 (.29) .00
Constant ~6.5 l ** * (1.20)

N 13909

x2 Statistic 721.19"

Pseudo R2 .12

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-stage probit least squares model.

lMarginal effects are calculated at the mean/mode of the independent variables.

246

 

Table H8. Groups Chronicity, Voting and Information : ZSPLS Results

 

 

 

Variables in Model Groups S. E. Marginal
Effectsl
Demographics
Education -.478*** (.036) -.09
Income -.155** * (.022) -.03
Gender .287 *** (.036) .05
Married .093*** (.033) .02
Intelligence .056* (.024) .01
Age 025*" (.006) .00
Age Squared -.000*** (.000) -.00
Urban -.041‘I (.020) -.01
Where Grew Up .019 (.034) .00
Political South -.183 * * * (.03 5) -.03
Political Variables
Party Identiﬁcation Squared -.390*** (.056) -.07
Party Identiﬁcation 447*" (.050) .08
Campaign Interest -.416* * * (.035) -.08
Care Who Wins -.115"* (.039) -.02
Voting .071 (.169) .02
Political Information 1071*" (.223) .20
Religion Variables
Bible Attitudes -.054 (.108) -.01
Church Attendance -.011 (.044) -.00
Evangelical -2.211*** (.806) -.41
Mainline Protestant —1.052 (.785) -.18
Catholic -.223 (.775) -.04
Black Evangelical .149 (.844) .02
Secular -.448 (.743) -.09
Jew -.275 (1.177) -.06
Election Year
1980 -.104 (.054) -.02
1984 .105“ (.048) .02
1988 .090 (.050) .02
1992 -.076 (.050) -.0 I
1996 -.024 (.070) -.00
2000 -.005 (.061) .00

Interaction Terms

247

 

 

Table H8 (cont’d).

Evangelical x Education
Evangelical x Income

Evangelical x Voting
Mainline x Voting

Black Protestant x Voting
Catholic x Voting
Secular x Voting

Jewish x Voting

Evangelical x Political Information
Mainline x Political Information

Black Protestant x Political Information
Catholic x Political Information
Secular x Political Information

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

Constant

N
x2 Statistic

Pseudo R2

-.201*"
-.069

-.478** *
-.208
.043
-.038
-.051
-.106

823*"
.289
.030
.049
.075
-.129

.025
.01 7
-.054
-.022
.042

.070
-.O42
.020
.021
.051
.246

-2.881***

13909
1354.16****

.12

(062)
(041)

(185)
(179)
(191)
(177)
(183)
(254)

(245)
(228)
(242)
(225)
(235)
(326)

(051)
(051)
(058)
(050)
(094)

(125)
(120)
(130)
(119)
(124)
(188)

(.783)

-.03
-.01

-.09
-.04

.01
-.Ol
-.01
-.03

. 15
.05
.00
.01
.01
-.02

.00
.00
-.Ol
-.01
.01

.02
-.00
.01
.01
.02
.05

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; "*; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-stage probit least squares model.

lMarginal effects are calculated at the mean/mode of the independent variables.

248

 

Table H9. Parties Chronicity, Voting and Information : ZSPLS Results

 

 

 

Variables in Model Parties S. E. Marginal
Effectsl
Demographics
Education -. I48" * (.03 8) -.02
Income -.103*** (.023) -.02
Gender .046 (.039) .01
Married -.026 (.035) -.00
Intelligence -.O4l (.025) -.01
Age -.014* (.006) -.00
Age Squared .000“ (.000) .00
Urban -.030 (.022) -.01
Where Grew Up -.044 (.037) -.01
Political South -.019 (.03 6) -.00
Political Variables
Party Identiﬁcation Squared .155*** (.058) .02
Party Identiﬁcation 223*" (.056) .03
Campaign Interest -.163"* (.038) -.03
Care Who Wins .029 (.042) .00
Voting .340 (.191) .05
Political Information .01 l (.243) .01
Religion Variables
Bible Attitudes .267“ (.131) .04
Church Attendance -.120* (.048) -.02
Evangelical 1 .070 (.862) . 19
Mainline Protestant -.696 (.842) -.I 1
Catholic -.402 (.844) -.04
Black Evangelical -.370 (.927) -.03
Secular -.175 (.799) -.01
Jew - I .479 (1.345) -.22
Election Year
1980 215*" (.059) .03
I984 239*" (.054) .04
1988 363*" (.055) .06
1992 .075 (.058) .01
I996 295*" (.073) .05
2000 .371 ** * (.062) .06

Interaction Terms

 

 

Table H9 (cont’d).

Evangelical x Education
Evangelical x Income

Evangelical x Voting
Mainline x Voting

Black Protestant x Voting
Catholic x Voting
Secular x Voting

Jewish x Voting

Evangelical x Political Information
Mainline x Political Information

Black Protestant x Political Information
Catholic x Political Information
Secular x Political Information

Jewish x Political Information

Evangelical x Attendance
Mainline Protestant x Attendance
Black Protestant x Attendance
Catholic x Attendance

Jewish x Attendance

Evangelical x Bible Attitudes
Mainline Protestant x Bible Attitudes
Black Protestant x Bible Attitudes
Catholic x Bible Attitudes

Secular x Bible Attitudes

Jewish x Bible Attitudes

Constant

N
x2 Statistic

Pseudo R2

.138*
-.005

.152

-.293
-. 130
-.21 1
-. I46
-.506

-.362
.286
.290
.201
.23 l
.799“

.071
.117"
.051
.114“
-.O72

-.243
-.216
-.219
-.260
-.237
-.455“'

-.606

l 3909

3575.82****

.22

(064)
(043)

(205)
(199)
(215)
(199)
(203)
(290)

(263)
(246)
(266)
(246)
(253)
(368)

(055)
(054)
(065)
(055)
(114)

(146)
(141)
(154)
(142)
(145)
(211)

(.843)

.02
-.00

.03
-.05
-.02
-.O3
-.02
-.O8

-.06
.05
.04
.03
.03
. 12

.01
.02
.01
.02
-.01

-.04
-.03
-.04
-.04
-.04
-.07

 

Note: "' p<. 10%, one tail test; "p<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-stage probit least squares model.

lMarginal effects are calculated at the mean/mode of the independent variables.

250

 

Table H10. Economy Chroniciy, Votinggnd Information : ZSPLS Results

 

 

 

Variables in Model Economy S. E. Marginal
Effectsl
Demographics
Education -.47 l ** * (.033) -.I 1
Income -. I37" * (.020) -.03
Gender 253*“ (.033) .06
Married -.01 l (.030) -.00
Intelligence -.052* (.022) -.01
Age .031* *" (.005) .01
Age Squared -.000*** (.000) -.00
Urban 049*" (.019) .01
Where Grew Up -.008 (.032) -.00
Political South .023 (.03 l) .01
Political Variables
Party Identiﬁcation Squared 218"” (.047) .06
Party Identiﬁcation .020 (.046) .01
Campaign Interest -.367*** (.032) -.09
Care Who Wins -.l35*** (.036) -.04
Voting -.226 (.157) -.06
Political Information 1465*" (.206) .36
Religion Variables
Bible Attitudes -. I61 (.104) -.04
Church Attendance .052 (.041) .01
Evangelical -l .460 (.757) -.33
Mainline Protestant -.469 (.736) -.08
Catholic .276 (.735) .08
Black Evangelical .273 (.814) .06
Secular -.253 (.702) -.05
Jew .647 (1.250) .16
Election Year
1980 -.O64 (.051) -.02
1984 343*" (.044) .08
1988 .013 (.048) -.00
I992 .090“ (.045) .02
I996 220““ (.062) .05
2000 .187*** (.056) .04

Interaction Terms

251

 

 

Table H10 (cont’d).

Evangelical x Education -.361 *** (.057) -.O9
Evangelical x Income -.074* (.037) -.02
Evangelical x Voting -.210 (.169) -.04
Mainline Protestant x Voting .002 (.164) .01
Black Protestant x Voting -.104 (.181) -.02
Catholic x Voting -.008 (.164) .00
Secular x Voting -.136 (.170) -.03
Jewish x Voting .198 (.255) .05
Evangelical x Political Information 781*" (.226) .18
Mainline Protestant x Political Information .084 (.209) .01
Black Protestant x Political Information -.11 I (.229) -.03
Catholic x Political Information -.090 (.208) -.03
Secular x Political Information .063 (.217) .01
Jewish x Political Information -.31 l (.330) -.08
Evangelical x Attendance -.022 (.047) -.01
Mainline Protestant x Attendance -.033 (.046) -.01
Black Protestant x Attendance .034 (.056) .01
Catholic x Attendance -.032 (.046) -.01
Jewish x Attendance -.082 (.105) -.02
Evangelical x Bible Attitudes .063 (.1 19) .02
Mainline Protestant x Bible Attitudes .169 (.114) .04
Black Protestant x Bible Attitudes .067 (.127) .02
Catholic x Bible Attitudes .083 (.115) .02
Secular x Bible Attitudes .073 (.117) .02
Jewish x Bible Attitudes .259 (.191) .06
Constant -4.017"‘** (.738)

N 13909

x2 Statistic 1079.59*“

Pseudo R2 .08

 

Note: "‘ p<.10%, one tail test; "p<.05, two-tail test; *"; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-stage probit least squares model.

lMarginal effects are calculated at the mean/mode of the independent variables.

252

 

Table H11. Candidate Chronicﬂy, Voting and Information : ZSPLS Results

 

 

 

Variables in Model Candidate S. E. Marginal
Effectsl
Demographics
Education -.234*** (.030) -.07
Income -.095“* (.018) -.03
Gender 320*" (.030) .10
Married -.O67* (.027) -.02
Intelligence .018 (.020) .01
Age -.005 (.005) -.00
Age Squared .000“ (.000) .00
Urban .003 (.017) -.00
Where Grew Up -.068* (.029) -.02
Political South -.03 1 (.029) -.01
Political Variables
Party Identiﬁcation Squared -.009 (.040) -.00
Party Identiﬁcation -.268*** (.039) -.08
Campaign Interest -.22 I " * (.029) -.07
Care Who Wins —.270*" (.032) -.08
Voting -.033 (.143) -.00
Political Information 1043*" (.186) .32
Religion Variables
Bible Attitudes -.001 (.095) .01
Church Attendance .021 (.03 7) .01
Evangelical -I .023 (.682) -.32
Mainline Protestant 1.294 (.668) .39
Catholic .746 (.670) .21
Black Evangelical .242 (.804) .10
Secular .875 (.639) .26
Jew 1.396 (1.065) .45
Election Year
1980 -.513*" (.042) -. I6
1984 -.898"’" (.041) -.28
1988 -.665*“ (.041) -.21
1992 -.756"* (.041) -.24
1996 -.447"‘""" (.055) -. I4
2000 -.715"* (.049) -.22

Interaction Terms

253

 

 

Table H11 (cont’d).

Evangelical x Education -.036 (.048) -.01
Evangelical x Income .071* (.032) .02
Evangelical x Voting -.466*"* (.153) -.15
Mainline Protestant x Voting .106 (.149) .03
Black Protestant x Voting .066 (.178) .02
Catholic x Voting .1 10 (.150) .03
Secular x Voting .127 (.155) .03
Jewish x Voting .161 (.224) .05
Evangelical x Political Information .270 (.200) .09
Mainline Protestant x Political Information -.336 (.188) -.10
Black Protestant x Political Information -.l77 (.222) -.05
Catholic x Political Information -.l68 (.188) -.04
Secular x Political Information -.243 (.196) -.06
Jewish x Political Information -.142 (.285) -.04
Evangelical x Attendance .044 (.042) .01
Mainline Protestant x Attendance -.O47 (.042) -.02
Black Protestant x Attendance -.052 (.057) -.02
Catholic x Attendance -.050 (.042) -.02
Jewish x Attendance -.O87 (.084) -.02
Evangelical x Bible Attitudes .043 (.107) .01
Mainline Protestant x Bible Attitudes .038 (.103) .01
Black Protestant x Bible Attitudes .140 (.124) .03
Catholic x Bible Attitudes .051 (.105) .01
Secular x Bible Attitudes -.004 (.107) -.01
Jewish x Bible Attitudes -.258 (.164) -.09
Constant -2.064*** (.671)

N 13909

x2 Statistic 1365.3 1 ***

Pseudo R2 .08

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; *"; p<.OI, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-stage probit least squares model.

lMarginal effects are calculated at the mean/mode of the independent variables.

254

 

Table H12. Multiple Chronicity, Votingand Information : ZSPLS Results

 

 

 

Variables in Model Multiple S. E. Marginal
Effectsl
Demographics
Education —.190* * * (.03 7) -.03
Income .071“ * * (.022) .01
Gender . 132* * * (.03 7) .02
Married .004 (.034) .00
Intelligence .045 (.025) .01
Age .021" * (.006) .00
Age Squared -.000*** (.000) -.00
Urban .032 (.021) .01
Where Grew Up .057 (.035) .01
Political South -.007 (.035) -.00
Political Variables
Party Identiﬁcation Squared .110* (.050) .02
Party Identiﬁcation . 124* * * ( .048) .02
Campaign Interest -. I 73 * * * (.035) -.03
Care Who Wins .076 (.039) .01
Voting -.854*** (.177) -.14
Political Information 1 . 164* * * (.226) . 19
Religion Variables
Bible Attitudes -.071 (.117) -.01
Church Attendance .I76*** (.047) .03
Evangelical 5662*" (.837) .93
Mainline Protestant L688" (.834) .26
Catholic -.078 (.838) -.03
Black Evangelical .565 (.926) .07
Secular .182 (.791) .02
Jew .680 (1.448) .14
Election Year
1980 .088 (.052) .02
1984 .069 (.050) .01
1988 -.01 l (.052) -.00
I992 -. 126* (.051) -.02
1996 .030 (.075) .01
2000 .261 ** * (.061) .05

Interaction Terms

255

 

Table H12 (cont’d).

 

Evangelical x Education 207*" (.059) .04
Evangelical x Income -.105*** (.039) -.02
Evangelical x Voting 1.357*** (.190) .22
Mainline Protestant x Voting .222 (.187) .03
Black Protestant x Voting .131 (.205) .02
Catholic x Voting -.031 (.188) -.01
Secular x Voting -.053 (.193) -.01
Jewish x Voting -.097 (.309) -.01
Evangelical x Political Information -1.648*** (.244) -.27
Mainline Protestant x Political Information -.450 (.232) -.07
Black Protestant x Political Information -.151 (.254) -.02
Catholic x Political Information -.O99 (.232) -.01
Secular x Political Information -.108 (.240) -.01
Jewish x Political Information -.4 l 8 (.378) -.O8
Evangelical x Attendance -.248*** (.053) -.04
Mainline Protestant x Attendance -.087 (.053) -.01
Black Protestant x Attendance -.O49 (.064) -.01
Catholic x Attendance .006 (.053) .00
Jewish x Attendance .073 (.1 15) .01
Evangelical x Bible Attitudes .126 (.131) .02
Mainline Protestant x Bible Attitudes -.003 (.128) .00
Black Protestant x Bible Attitudes .090 (.143) .02
Catholic x Bible Attitudes .184 (.129) .03
Secular x Bible Attitudes .064 (.131) .01
Jewish x Bible Attitudes .197 (.225) .04
Constant -5.922*** (.829)

N 13909

x2 Statistic 30.86""

Pseudo R2 .02

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-stage probit least squares model.

lMarginal effects are calculated at the mean/mode of the independent variables.

256

 

Table H13. No Content Chronicity, Votingand Information : ZSPLS Results

 

 

 

Variables in Model No Content S. E. Marginal
Effectsl
Demographics
Education 340*" (.038) .02
Income 360*" (.028) .02
Gender -.004 (.040) -.00
Married -.079 (.042) -.00
Intelligence .221" * (.032) .01
Age .073*** (.007) .00
Age Squared -.001 *** (.000) -.00
Urban .029 (.027) .00
Where Grew Up .101" (.043) .01
Political South .140* * * (.043) .01
Political Variables
Party Identiﬁcation Squared .012 (.054) .00
Party Identiﬁcation -.01 l (.051) -.00
Campaign Interest .145*** (.036) .01
Care Who Wins .016 (.045) .00
Voting -2.018*** (.172) -.13
Religion Variables
Bible Attitudes -.O39 (.151) -.00
Church Attendance 357*" (.058) .02
Evangelical 3205*" (.565) .20
Mainline Protestant .864 (.526) .06
Catholic 1.258" (.527) .08
Black Evangelical .450 (.572) .03
Secular .262 (.437) .02
Jew 2.361 (1.345) .17
Election Years
1980 .056 (.071) .00
1984 300*“ (.065) .02
1988 .191 *** (.064) .01
1992 .018 (.067) .00
1996 .135 (.091) .01
2000 I . 162*" (.079) .07
Interaction Terms
Evangelical x Education -.298*** (.067) -.02

257

 

 

Table H13 (cont’d).

Evangelical x Income -.190*** (.047) -.01
Evangelical x Voting 656*" (.179) .04
Mainline Protestant x Voting .420" (.173) .03
Black Protestant x Voting .275 (.181) .02
Catholic x Voting .354" (.172) .02
Secular x Voting .256 (.176) .02
Jewish x Voting -.035 (.444) .00
Evangelical x Attendance -.143** (.064) -.01
Mainline Protestant x Attendance -.142** (.066) -.01
Black Protestant x Attendance -.015 (.076) -.00
Catholic x Attendance -.147** (.065) -.01
Jewish x Attendance -.O33 (.225) -.01
Evangelical x Bible Attitudes -.098 (.169) -.01
Mainline Protestant x Bible Attitudes .036 (.165) -.00
Black Protestant x Bible Attitudes .063 (.179) .00
Catholic x Bible Attitudes -.163 (.165) -.01
Secular x Bible Attitudes .030 (.167) .00
Jewish x Bible Attitudes -l.219** (.608) -.08
Constant -8.976*** (.619)

N 13909

x2 Statistic $15.88*"

Pseudo R2 .80

 

 

Note: * p<.10%, one tail test; "p<.05, two-tail test; ***; p<.01, two-tail test. This table reports
unstandardized coefﬁcients (standard errors in parentheses) from the 2-stage probit least squares model.

lMarginal effects are calculated at the mean/mode of the independent variables.

258

Appendix I

Table I1. Variables Used in Chapter Six NES Models.

 

 

 

 

 

 

 

 

 

 

Variable Coding
(NES variable No.)
Political 1. Lowest level of participation (none)
Participation Scale 6. Highest level of participation in campaign activities
(vcf0723)
Political Respondents general level of information about politics and
Information public affairs seemed:
(vcf0050a)
5. Very High
4. Fairly High
3. Average
2. Fairly Low
1. Very Low
Voted I. Voted
(based on vcf702) 0. Did not vote
Abortion There has been some discussion about abortion during recent
(vcf0838) years. Which one of the opinions on this page best agrees with
your view?
1. Abortion should never be permitted.
2. Abortion should be permitted only if the life and health of the
woman is in danger.
3. Abortion should be permitted if, due to personal reasons, the
woman would have difﬁculty in caring for the child.
4. Abortion should never be forbidden, since one should not
require a woman to have a child she doesn't want.
Environment Should federal spending on improving and protecting the
(vcf9047) environment be increased, decreased or kept about the same?
1. Increased
2. Same
3. Decreased
Environmentalists 1(cold)-97(warm)
Thermometer
(vcf0229)
Women’s Libber’s l(cold)-97(warm)
Thermometer
(vcf0225)

 

259

 

Table I 1 (cont’d).

 

Women’s Roles
Scale

Some people feel that women should have an equal role with men
in running business, industry and government. Others feel that a

 

 

(vcﬂl834) women's place is in the home.
Where would you place yourself on this scale?
1. Women and men should have an equal role
2
3.
4.
5
6.
7. Women's place is in the home
Strength of 1. Independent or Apolitical
partisanship 2. Leaning Independent
(vcf0305) 3. Weak Partisan
4. Strong Partisan
Length of
Residence

 

 

 

 

Care who wins 1. Don't care very much or DK, Pro-Con, Depends, and Other
(vcf0311) 2. Care a good deal
Campaign Interest 1. Not much interested
(vcf0310) 2. Somewhat interested
3. Very much interested
Party Contact Did anyone ﬁ'om one of the political parties call you up or come
(vcf9030) around and talk to you about the campaign?
1. Yes
0. No
Strength of 1. Independent or Apolitical
Partisanship 2. Leaning Independent
(vcf0305) 3. Weak Partisan

4. Strong Partisan

 

 

Closeness of
election
(vcf07l4)

 

Do you think the presidential race will be close or will one
candidate win by quite a bit?

0. Will win by quite a bit

1. Pro-con, depends

2. Close Race

260

 

Table II (cont’d).

 

 

 

 

 

 

 

 

 

 

 

Length of Actual years coded 0-90
Residence
Education 1. Grade school or less (0-8 grades)
(vcf0110) 2. High school (12 grades or fewer, incl. non-college
training if applicable)
3. Some College (13 grades or more but no degree;
college, no identiﬁcation of degree status)
4. College or advanced degree
Gender 1. Male
(vcf0104) 2. Female
Income 1. 0 to 16 percentile
(vcf0114) 2. 17 to 33 percentile
3. 34 to 67 percentile
4. 68 to 95 percentile
5. 96 to 100 percentile
Church Attendance 5. Every week
(vcf0130 & 4. Almost every week
vcf0130a) 3. Once or twice a month
2. A few times a year
1. Never
Bible 1. The Bible is the actual Word of God and is to be taken literally,
(vcf0845 & word for word
vcf0850) 2. The Bible is the Word of God but not everything in it should be
taken literally, word for word
3. The Bible is a book written by men and is not the Word of God
Political South 1. South
(vcf0113) O. Nonsouth
Religious Tradition Each tradition was coded:
Dummy Variables 1. Religious tradition
0. Otherwise
Birth Cohorts 1. 1903-1912 7. 1963-1972
2. 1913-1922 8. 1973-1982
3. 1923-1932
4. 1933-1942
5. 1943-1952
6. 1953-1962

261

 

Table II (cont’d).

 

 

Periods

 

Presidential election years ﬁ'om 1972-2000
. 1972
. 1976
. 1980
. 1984
. 1988
. 1992
. 1996
8. 2000

\IQMAUJN—I

 

262

 

Appendix J

Table J1. Variables Used in Chapter Six General Social Survey Models.

 

Variable

Coding

 

Dependent

 

Extramarital Sex

What is your opinion about a married person having sexual
relations with someone other than the marriage partner--is it
always wrong, almost always wrong, wrong only sometimes, or
not wrong at all?

I. Always Wrong

2. Ahnost Always Wrong

3. Sometimes Wrong

4. Not Wrong

 

Homosexuality

What about sexual relations between two adults of the same sex--
do you think it is always wrong, almost always wrong, wrong
only sometimes, or not wrong at all?

1. Always Wrong

2. Almost Always Wrong

3. Sometimes Wrong

4. Not Wrong

 

Teen Sex

What if they are in their early teens, say 14 to 16 years old? In
that case, do you think sex relations before marriage are always
wrong, almost always wrong, wrong only sometimes, or not
wrong at all?

1. Always Wrong

2. Almost Always Wrong

3. Sometimes Wrong

4. Not Wrong

 

Premarital Sex

If a man and woman have sex relations before marriage, do you
think it is always wrong, almost always wrong, wrong only
sometimes, or not wrong at all?

1. Always Wrong

2. Almost Always Wrong

3. Sometimes Wrong

 

 

 

4. Not Wrong
Government 1. Increase
Spending: Foreign 2. Same
Aid 3. Decrease
Government 1. Increase
Spending: 2. Same
Education 3. Decrease

 

263

 

Table J1 (cont’d).

 

Government
Spending:
Environmental

1. Increase
2. Same
3. Decrease

 

Government
Spending: Military

1. Increase
2. Same
3. Decrease

 

Atheist Speaker

There are always some people whose ideas are considered bad or
dangerous by other people. For instance, somebody who is against
all churches and religion...

If such a person wanted to make a speech in your
(city/town/community) against churches and religion, should he
be allowed to speak, or not?

1. Yes, Allowed to speak

0. No, not allowed to speak

 

Atheist Teacher

Suppose he is teaching in a college. Should he be ﬁred, or not?
0. Yes, ﬁred
1. Not ﬁre

 

Atheist Book

If some people in your community suggested that a book he wrote
against churches and religion should be taken out of your public
library, would you favor removing this book, or not?

0. Favor

1. Not Favor

 

Communist
Speaker

Now, I should like to ask you some questions about a man who
admits he is a Communist.

Suppose an admitted Communist wanted to make a speech in
your community. Should he be allowed to speak, or not?

1. Yes, Allowed to speak

0. No, not allowed to speak

 

Communist
Teacher

Suppose he is teaching in a college. Should he be ﬁred, or not?
0. Yes, ﬁred
1. Not ﬁre

 

Communist Book

Suppose he wrote a book, which is in your public library.
Somebody in your community suggests that the book should be
removed from the library. Would you favor removing it, or not?
0. Favor

1. Not Favor

 

 

Homosexual
Speaker

 

And what about a man who admits that he is a homosexual?

A. Suppose this admitted homosexual wanted to make a speech in
your community. Should he be allowed to speak, or not?

1. Yes, Allowed to speak

0. No, not allowed to speak

264

 

Table J1 (cont’d).

 

Homosexual
Teacher

Suppose he is teaching in a college. Should he be ﬁred, or not?
0. Yes, ﬁred
1. Not ﬁre

 

Homosexual Book

If some people in your community suggested that a book he wrote
in favor of homosexuality should be taken out of your public
library, would you favor removing this book, or not?

0. Favor

1. Not Favor

 

Independent

 

Party Identification

Generally speaking, do you usually think of yourself as a
Republican, Democrat, Independent or what?

Strong Democrat

Independent

Strong Republican

 

Political Ideology

Extremely Liberal
Liberal

Slightly Liberal
Moderate

Slightly Conservative
Conservative
Extremely Conservative

 

Education

“HQPPP’PE‘NPI‘

. Grade school or less (0-8 grades)

2. High school (12 grades or fewer, incl. non-college
training if applicable)

3. Some College (13 grades or more but no degree;
college, no identiﬁcation of degree status)

4. College or advanced degree

 

Income

1. O to 16 percentile

2. 17 to 33 percentile
3. 34 to 67 percentile
4. 68 to 95 percentile
5. 96 to 100 percentile

 

Gender

1. Male
2. Female

 

 

Age

 

Actual age, 18-80

265

 

Table J1 (cont’d).

 

Age-squared

Actual age squared * .001

 

Children in Home

Number of children the respondent has ever had.

 

Urbanism:
Currently live in

Population of Residency

1. Open country

2. 1000-2499

3. 2500-9,999

4. 10,000-49,000

5. Unincorporated area of a Medium Central City
6. Unincorporated of a Large Central City
7. Suburb of a Medium Central City

8. Suburb of a Large Central City

9. Medium Central City

10. Large Central City

 

Marital Status

1. Not Married
0. Married

 

Religious
Attendance

How often do you attend religious services?
8. More than weekly

7. Every week

6. Almost every week

5. Two to Three times a month

4. Once or less a month

3. Two to three times a year

2. Once a year or less

 

 

Religious Tradition
Dummy Variables

 

Each tradition was coded:
1. Religious tradition
0. Otherwise

 

266

 

Appendix K

Table Kl. Variables Used in National Congregation —1998 General Social Survey

 

 

 

 

 

 

 

 

 

Models.

Variable Coding

Dependent .

Abortion Support ' 0. Low

Scale 6. High

1996 Presidential I. Voted

Vote 0. No Vote

Atheist Speaker There are always some people whose ideas are considered bad or
dangerous by other people. For instance, somebody who is against
all churches and religion...
If such a person wanted to make a speech in your
(city/town/community) against churches and religion, should he
be allowed to speak, or not?
1. Yes, Allowed to speak
0. No, not allowed to speak

Atheist Teacher Suppose he is teaching in a college. Should he be ﬁred, or not?
0. Yes, ﬁred
1. Not ﬁre

Atheist Book If some people in your community suggested that a book he wrote
against churches and religion should be taken out of your public
library, would you favor removing this book, or not?
0. Favor
1. Not Favor

Communist Now, I should like to ask you some questions about a man who

Speaker admits he is a Communist.
Suppose an admitted Communist wanted to make a speech in
your community. Should he be allowed to speak, or not?
1. Yes, Allowed to speak
0. No, not allowed to speak

Communist Suppose he is teaching in a college. Should he be ﬁred, or not?

Teacher 0. Yes, ﬁred

1. Not ﬁre

 

 

Communist Book

 

Suppose he wrote a book, which is in your public library.
Somebody in your community suggests that the book should be
removed from the library. Would you favor removing it, or not?
0. Favor

1. Not Favor

267

 

Table Kl (cont’d).

 

 

Homosexual And what about a man who admits that he is a homosexual?
Speaker
A. Suppose this admitted homosexual wanted to make a speech in
your community. Should he be allowed to speak, or not?
1. Yes, Allowed to speak
0. No, not allowed to speak
Homosexual Suppose he is teaching in a college. Should he be ﬁred, or not?
Teacher 0. Yes, ﬁred

1. Not ﬁre

 

Homosexual Book

If some people in your community suggested that a book he wrote
in favor of homosexuality should be taken out of your public
library, would you favor removing this book, or not?

0. Favor

1. Not Favor

 

Independent

 

Party Identification

Generally speaking, do you usually think of yourself as a
Republican, Democrat, Independent or what?

. Strong Democrat

. Independent

. Strong Republican

 

Strength of
Partisanship

. Independent

. Weak Partisan
. Partisan

. Strong Partisan

 

Political Ideology

Extremely Liberal
Liberal

Slightly Liberal
Moderate

Slightly Conservative
Conservative
Extremely Conservative

 

Education

HSQMPVNrewN—vh~

. Grade school or less (0-8 grades)

2. High school (12 grades or fewer, incl. non-college
training if applicable)

3. Some College (13 grades or more but no degree;
college, no identiﬁcation of degree status)

4. College or advanced degree

 

 

Income

 

l. 0 to 16 percentile
2. 17 to 33 percentile
3. 34 to 67 percentile

5. 96 to 100 percentile
4. 68 to 95 percentile

268

 

Table Kl (cont’d).

 

 

 

 

 

Gender 1. Male
2. Female
Age Actual age, 18-80
Age-squared Actual age squared * .001
Children in Home Number of children the respondent has ever had.
Urbanism: Population of Residency

Currently live in

1. Open country

2. 1000-2499

3. 2500-9,999

4. 10,000-49,000

5. Unincorporated area of a Medium Central City
6. Unincorporated of a Large Central City

7. Suburb of a Medium Central City

8. Suburb of a Large Central City

9. Medium Central City

10. Large Central City

 

Marital Status

1. Not Married
0. Married

 

Religious
Attendance

How often do you attend religious services?
8. More than weekly

7. Every week

6. Almost every week

5. Two to Three times a month

4. Once or less a month

3. Two to three times a year

2. Once a year or less

 

Religious Tradition
Dummy Variables

Each tradition was coded:
1. Religious tradition
0. Otherwise

 

 

Church Theology

 

Theologically speaking would your congregation be considered
more on the conservative side, more on the liberal side, or right in
the middle?

1. More Liberal

2. Right in the Middle

3. More Conservative

 

269

 

Table K1 (cont’d).

 

Church Ideology

Politically speaking, would your congregation be considered
more on the conservative side, more on the liberal side, or right in
the middle?

1. More Liberal

2. Right in the Middle

3. More Conservative

 

Political Activism
in Church

Within the past 12 months, have people at worship services been
told of opportunities for political activity, including petition
campaigns, lobbying, or demonstrating?

1. Yes

2. No

 

Group to Discuss
Politics

Within the past 12 months, have there been any groups or
meetings or classes or events speciﬁcally focused on discussing
politics?

1. Yes

2. No

 

Voter Guides
Distributed

Have voter guides ever been distributed to people through your
congregation?

1. Yes

2. No

 

Candidate Spoke to
Congregation

Have any political candidates have spoken to your congregation
in the past 12 months

1. Yes

2. No

 

 

New Paradigm
Scale

 

New Paradigm Church attributes (1 through 16)
1. Low
16. High

 

270

 

REFERENCES

Abramson, Paul R. 1975. Generational Change in American Politics. Lexington: D.C.
Heath.

Abramson, Paul R. 1976. Generational Change and the Decline of Party Identiﬁcation in
America, 1952-1974. American Political Science Review 70: 469-78.

Abramson, Paul R. 1978. Generational Replacement and Partisan Dealignment in Britain
and the United States. British Journal of Political Science 8: 505-9.

Abramson, Paul R. 1981. “Comment on Smith.” American Political Science Review.
75:146-149.

Abramson, Paul R. 1990. The Decline of Over-Time Comparability in the National
Election Studies. Public Opinion Quarterly 54: 177-90.

Abramson, Paul R., and William Claggett. 1984. Race-Related Differences in Self-
Reported and Validated Turnout. Journal of Politics 46 (3):71 9-73 8.

Aiken, Leona S. and Stephen G. West. 1991. Multiple Regression: Testing and
Interpreting Interactions. Newbury Park: Sage Publications.

Alvarez, R. Michael and Garrett Glasgow. 2000. “Two-Stage Estimation of Nonrecursive
Choice Models”. Political Analysis. 8:147-165.

Ammerman, Nancy T. 1982. “Operationalizing Evangelicalism: An Amendment.”
Sociological Analysis. 43: 1 70-1 72.

Anderson Barbara, and Brian Silver. 1986. “Measurements and Mismeasurements of the
Validity of the Self-Reported Vote. American Journal of Political Science.
30:771-785.

Axelrod, Robert, 1973. Schema Theory: An Information Processing Model of Perception
and Cognition. American Political Science Review. 67:1248-1266.

Ayala, Louis J. 2000. "Trained for Democracy: The Differing Effects of Voluntary and
Involuntary Organizations on Political Participation". Political Research
Quarterly, 53 (1): 99-115.

Bargh, John A. and Felicia Pratto. 1986. Individual Construct Accessibility and
Perceptual Selections. Journal of Experimental Social Psychology. 22:293-311.

271

Bennett. W. Lance. 1975. The Political Mind and the Political Environment. Lexington:
Lexington Books.

Bennett, Stephen Earl. 1994. “Changing Levels of Political Information in 1988 and
l 990.” Political Behavior. 1 6: 1-19.

Bennett, Stephen E. 1995. "Comparing Americans' Political Information in 1988 and
1992." Journal of Politics 57: 521-532.

Berger, Peter. 1992. “Secularism in Retreat” The National Interest. 46:3-12.

Berger, Peter L. 1999. “The Desecularization of the World: A Global Overview.” In
Peter L. Berger’s (ed) The Desecularization of the World: Resurgent Religion and
World Politics. Grand Rapids, MI: Eerdmans Publishing Company.

Berggren, Heidi M. 2001. “Institutional Context and Reduction of the Resource Bias in
Political Sophistication.” Political Research Quarterly 542531—52.

Beyerlein, Kraid and Mark Chaves. “The Political Activities of Religious Congregations
in the United States”. Journal for the Scientiﬁc Study of Religion. 42:299-246.

Bolce, Louis and Gerald De Maio. 1999.“Religious Outlook, Culture War Politics and
Antipathy Toward Christian Fundamentalists. Public Opinion Quarterly. 63:29-
61.

Brady, Henry, Sydney Verba, and Kau Lehman Schlozman. 2005. “Beyond SES: A
Resource Model of Political Participation,” American Political Science Review.
89:271-294.

Bruce, Steve. 2002. God is Dead: Secularization In the West. New York: Blackwell
Publishing.

Burns, Nancy, Kay Lehman Schlozman, and Sidney Verba. 2001. The Private Roots of
Public Action. Cambridge, MA: Harvard University Press.

C. Peter Wagner. 1984. Leading Your Church to Growth. Ventura, CA: Regal Books.

C. Peter Wagner. 1987. Strategies for church growth: Tools for eﬂective mission and
evangelism. Ventura, CA: Regal Books.

C. Peter Wagner. 1994. Your Spiritual Gifts Can Help Your Church Grow. Ventura, CA:
Regal Books.

Calhoun-Brown, Allison. 1996. "Aﬁican American Churches and Political Mobilization:

The Psychological Impact of Organizational Resources". The Journal of Politics.
58: 93 5-953.

272

Campbell, Angus, Philip E. Converse, Warren E. Miller and Donald E. Stokes. [1960]
1980. The America Voter. Chicago: University of Chicago Press.

Campbell, David E. 2000. Acts of faith: Strict churches and political mobilization. Paper
presented at the annual meeting of the American Political Science Association,
Washington, DC.

Campbell, David E. 2003. "Measurement of Religious Afﬁliation and Commitment."
Unpublished Manuscript.

Carmines, Edward G. and James A Stimson. 1980. “The Two Faces of Issue Voting.”
American Political Science Review. 74:78-91.

Cassel Carol A.1984. “Issues in Measurement: The ‘Levels of Conceptualization’ Index
of Ideological Sophistication.” American Journal of Political Science. 28:418-
429.

Chaves, Mark. 1994. "Secularization As Declining Religious Authority." Social Forces
72(3):749-74.

Chaves, Mark. 1998. NATIONAL CONGREGATIONS STUDY. DATA FILE AND
CODEBOOK [computer ﬁle]. Tucson, Arizona: University of Arizona,
Department of Sociology [producer]. University Park, PA: American Religion
Data Archive [distributor].

Chaves, Mark, Mary Ellen Konieczny, Kraig Beyerlein, and Emily Barman. 1999. "The
National Congregations Study: Background, Methods, and Selected Results."
Journal for the Scientiﬁc Study of Religion. 38: 458-476.

Cleveland, William S. and Devlin, SJ. 1988. "Locally Weighted Regression: An
Approach to Regression Analysis by Local Fitting." Journal of the American
Statistical Association. 83: 596-610.

Conover, Pamela Johnston. 1982. “The Mobilization of the New Right: A Test of Various
Explanations.” Western Political Quarterly. :632-649.

Conover, Pamela Johnson. 1984a. “The Inﬂuence of Group Identiﬁcation on Political
Perception and Evaluation.” Journal of Politics. 46:760-85.

Conover, Pamela Johnson. 1984b. “Issues in Measurement: The ‘Levels of
Conceptualization’ Index of Ideological Sophistication.” American Journal of
Political Science. 28:418-429.

Conover, Pamela Johnston, and Stanley Feldman. 1984a. “Group Identiﬁcation, Values,
and the Nature of Political Beliefs.” American Politics Quarterly 12: 1 51-75.

273

Conover, Pamela Johnston, and Stanley F eldman. 1984b. “How People Organize the
Political World: A Schematic Model.” American Journal of Political Science.
28:98-127.

Conway, M. Margaret. 1991. Political Participation in the United States. Washington,
DC: Congressional Quarterly Inc.

Conway, Margaret. 2000. Political Participation in the United States. 3rd ed. Washington
DC: Congressional Quarterly Press.

Converse, Philip E. and Richard Niemi. 1971. “Non-voting among Young Adults in the
United States”. In Political Parties and Political Behavior, ed. William J. Crotty,
et al. 2nd ed., Boston: Allyn and Bacon.

Converse, Philip E. 1964. “The Nature of Belief Systems in Mass Publics” in David
Apter’s (ed) Ideology and Discontent. London: Free Press of Glencoe.

Converse, Philip E. 2000. “Assessing the Capacity of Mass Electorates," Annual Review
of Political Science. . 3 :33 1-353.

Cook, Elizabeth Adel]. And Clyde Wilcox. 1990. “Religion and Political Attitudes
among Black Americans in Washington, DC.” Polity 22:527-39.

Cook, Terrance C. and Patrick Morgan. 1971. “An Introduction to Participatory
Democracy”. In Participatory Democracy, eds. Terrance C. Cook and Patrick
Morgan. San Francisco: Canﬁeld.

Cox, Gary W. and Michael C. Munger. 1989 “Closeness, Expenditures, and Turnout in
the 1982 US. House Elections,” American Political Science Review 83: 217-231.

Cromartie, Michael, ed. 2001. A Public Faith: Evangelicals and Civic Engagement. Ed.
MMD: Rowman and Littleﬁeld.

Curry-Roper, Joyce 1990. "Contemporary Christian Eschatologies and Their
Relationship to Environmental Stewardship." Professional Geographer. 42 (2):
157-169.

Dalton, Russell. 2002. Citizen Politics: Public Opinion and Political Parties in Advanced
Industrial Democracies. 3rd ed. NY: Chatham House Publishers.

Davis, Nancy J ., and Robert V. Robinson. 1996. Are the rumors of war exaggerated?

Religious orthodoxy and moral progressivism in America. American Journal of
Sociology 102 (3):756-787.

274

Davis, James A., Tom W. Smith, and Peter V. Marsden. 2003. GENERAL SOCIAL
SURVEYS, 1972-2002: [CUMULATIVE FILE] [Computer ﬁle]. 2nd ICPSR
version. Chicago, IL: National Opinion Research Center [producer], 2003. Storrs,
CT: Roper Center for Public Opinion Research, University of Connecticut/Ann
Arbor, MI: Inter-university Consortium for Political and Social Research
[distributors].

Delli Carpini, Michael, and Scott Keeter. 1991. Stability and Change in the United-States
Publics Knowledge of Politics. Public Opinion Quarterly 55 (4):583-612.

Delli Carpini, Michael X., and Scott Keeter. 1993. "Measuring Political Knowledge:
Putting First Things First." American Journal of Political Science 37: 1179-1206.

Delli Carpini, Michael, and Scott Keeter. 1996. What American Know About Politics and
Why It Matters. New Haven: Yale University Press.

DeSart, Jay A. 1995. “Information Processing and Partisan Neutrality: a Reexamination
of the Party Decline Thesis. The Journal of Politics. 57:776-795.

DiMaggio, Paul, John Evans, and Bethany Bryson. 1996. Have American's Social
Attitudes Become More Polarized? American Journal of Sociology 102 (3):690-
755.

DiSabatino, David. 1999. The Jesus People Movement: An Annotated Bibliography and
General Resource. Connecticut: Greenwood Press.

Djupe, Paul A and J. Tobin Grant. 2001. "Religious Institutions and Political
Participation in America". Journal for the Scientiﬁc Study of Religion. 40:303-314

Eber, Ralph and Richard R. Lau 1990. "Political Cynicism Revisited: An Information
Processing Reconciliation of Policy-based and Incumbency-based Interpretations
of Changes in Trust in Government. American Journal of Political Science.
34:236-53.

Evand John H. 1997. "Worldviews or Social Groups as the Source of Moral Value
Attitudes: Implications for the Culture Wars Thesis.” Sociological Forum 12:
371-404.

Eveland, William P., Jr., and Dietram A. Scheufele. 2000. "Connecting News Media Use
with Gaps in Knowledge and Participation." Political Communication 17: 215-
237.

Fazio, Russell H. 1989. “On the Power and Functionality of Attitudes: The Role of
Attitude Accessibility.” In A. R. Pratkanis, S. J. Breckler, and A. G. Greenwald,
eds. Attitude Structure and Function. Hillsdale, NJ: Lawrence Erlbaurn and
Associates.

275

Feldman, Stanley. 1988. “Structure and Consistency in Public Opinion: the Role of Core
Beliefs and Values.” American Journal of Political Science. 32:416-440.

Feldman, Stanley. 1995. “Answering Survey Questions: The Measurement and Meaning
of Public Opinion,” in Milton Lodge and Kathleen M. McGraw, eds., Political
Judgment: Structure and Process. Ann Arbor: University of Michigan Press.

Fienberg, S., and W. M. Mason 1985. "Speciﬁcation and Implementation of Age, Period
and Cohort Models," in eds. W.M. Mason and S. Fienberg. Cohort Analysis in
Social Research. New York: Springer-Verlag.

Fiorina, Morris. 2004. Culture War? The Myth of a Polarized America. New York:
Longrnan.

Fiske, Susan T. and Donald R. Kinder. 1981. “Involvement, Expertise, and Schema Use:
Evidence from Political Cognition.” In Personality, Cognition, and Social
Interaction, ed. Nancy Cantor and John F. Kihlstrom. L. Erlbaum Associates.

F iske, Susan T. and Shelley E. Taylor. 1991. Social Cognition. New York, NY: Random
House.

Freie, John F. 1997. "The Effects of Campaign Participation on Political Attitudes."
Political Behavior. 19: 133 - 156.

Gilliam, Franklin D. 1985. “Inﬂuences on Voter Turnout for US House Elections in Non
Presidential Years.” Legislative Studies Quarterly 10:339-51.

Glenn, Norval D. 1980. "Values, Attitudes, and Beliefs." In Constancy and Change in
Human Development, ed. by O.G. Brim and J. Kegan. Cambridge, Massachusetts:
Harvard University Press.

Glenn, Norval D. 2005. Cohort Analysis (2“d ed.) Thousand Oaks, CA: Sage
Publications.

Goren, Paul. 2004. “Political Sophistication and Policy Reasoning: A Reconsideration.”
American Journal of Political Science 48:462-78

Green, John C. 2000. "Evangelicals and Civic Engagement: An Overview" Unpublished
Manuscript.

Green, John C. 2003. "Evangelical Protestants and Civic Engagement: An Overview. A
Public Faith. ed. Michael Cromartie. New York: Rowman and Littleﬁeld.

276

Green, John C., James L. Guth, and Kevin Hill. 1993. "Faith and Election: The Christian
Right in Congressional Campaigns, 1978-1988." Journal of Politics 55:80-91.

Green, John C., James L. Guth, Corwin E. Smidt and Lyman A. Kellstedt, eds. 1996.
Religion and the Culture Wars: Dispatches From the Front. New York: Rowman
and Littleﬁeld.

Green, John C., Mark J. Rozell and Clyde Wilcox., eds. 2000. Prayers in the Precincts:
The Christian Right in the I 998 Elections. Washington D. C.: Georgetown
University Press.

Green, John C., Mark J. Rozell and Clyde Wilcox., eds. 2003Jhe Christian Right in
American Politics: Marching to the Millennium. Washington, DC: Georgetown
University Press.

Guth, J arnes L.. 1996. "The Politics of the Christian Right" Religion and the Culture
Wars: Dispatches From the Front. ed. John C. Green, James L. Guth, Corwin E.
Smidt and Lyman A. Kellstedt. New York: Rowman and Littleﬁeld.

Guth, James L. and John C. Green. 1990. “Politics in a New Key: Religiosity and
Participation Among Political Activists.” Western Political Quarterly. 43: 153-
1 79.

Guth, James L. and John C. Green 1993. “Salience: The Core Concept?” In David C.
Leege and Lyman A. Kellstedt’s (ed) Rediscovering the Religious Factor in
American Politics. New York: M. E. Sharpe.

Guth, James L., John C. Green, Lyman A. Kellstedt, and Corwin E. Smidt. 1995. “Faith
and the Environment: Religious Beliefs and Attitudes on Environmental Policy.”
American Journal of Political Science. 39: 364-3 82.

Guth, James L., John C. Green, Corwin E. Smidt and Lyman A. Kellstedt, eds. 1996.
“The Political Relevance of Religion” In Religion and the Culture Wars:
Dispatches From the Front, ed. Green, John C., James L. Guth, Corwin E. Smidt
and Lyman A. Kellstedt. New York: Rowman and Littleﬁeld.

Guth, James L., John C. Green, Lyman A. Kellstedt, and Corwin E. Smidt. 2002.
“Religion and Political Participation.” Unpublished paper presented at the annual
meeting of the American Political Science Association, Boston.

Guth, James L., Lyman A. Kellstedt, Corwin E. Smidt. And John C. Green. 2003. “The
Religious Basis of Contemporary American Party Alignments.” Unpublished
paper presented at the Midwest Political Science Conference, Chicago.

Hadden, Jeffery, K. 1987. Toward Desacralizing Secularization Theory. Social Forces
65: 587-611.

277

Hagner, Paul R. and John C. Pierce. 1982. “Correlative Characteristics of Levels of
Conceptualization in the American Public.” Journal of Politics. 44:779-807.

Hagner, Paul R. and John C. Pierce. 1984. “Racial Differences in Political
Conceptualization.” Western Political Quarterly. 37:212-235.

Hagner, Paul R., John C. Pierce and Kathleen Knight. CONTENT CODINGS OF
LEVELS OF POLITICAL CONCEPTUALIZATION. 1956-1988 [UNITED
STATES] [Computer ﬁle]. 4th release. Ann Arbor, MI: Inter-intiversity
Consortium for Political and Social Research [distributor], 1989.

Hamill, Ruth, Milton Lodge and Fredrick Blake. 1985. The Breath, Depth and Utility of
Class, Partisan and Ideological Schemata. American Journal of Political Science.
29:850-870.

Hamill Ruth, and Milton Lodge 1986. "Cognitive Consequences of Political
Sophistication." In Richard R. Lau and Donald 0. Sears, (Eds), Political
Cognition: The 19th Annual Carnegie Symposium on Cognition. Hillsdale, NJ:
Erlbaum.

Harris, Fredrick C. 1994. “Something Within: Religion as a Mobilizer of African-
American Political Activism. Journal of Politics. 56:42-68.

Hartford Institute for Religion Research. 2004. http://hirr.hartsem.edu/org/
faith_megachurches_research.htm1. Accessed 11 October 2004.

Hill, Kevin A. 1997. “Generations and Tolerance: Is Youth Really a Liberalizing
Factor?” In Aﬁer the Boom, Stephen Bennett and Steven Craig, editors. Lanham,
MD: Rowman and Littleﬁeld.

Himmelfarb, Gertrude. 1999. One Nation, Two Cultures. New York: Knopf.

Hogg, Michael A., and Abrams, Dominic 1988. Social Identifications: A Social
Psychology of Intergroup Relations and Group Processes. New York: Routledge.

Hood, M. V. and Mark Caleb Smith. 2002. “On the Prospect of Linking Religious-Right
Identiﬁcation with Political Behavior: Panacea or Snipe Hunt?” Journal for the
Scientific Study of Religion. 41 :697-710.

Hout, Michal and Claude S. Fischer. 2002. “Why more Americans Have No Religious
Preference: Politics and Generations. American Sociological Review. 67: 165-190.

Huckfeldt, Robert and John Sprague. 1992. “Political Parties and Election Mobilization:

Political Structure, Social Structure, and the Party Canvass.” American Political
Science Review 86:70—86

278

Hunter, James Davidson. 1983. American Evangelicalism: Conservative Religion and the
Quandary of Modernity. New Jersey: Rutgers University Press.

Hunter, James Davidson. 1987. Evangelicalism: the Coming Generation. Chicago:
University of Chicago Press.

Hunter, James Davidson. 1991. Culture Wars: the Struggle to Deﬁne America. New
York: Basicbooks.

Integrity Media: 2002 Annual Report. 2002. www.integritymusic.com/company/
investors/annua102.pdf.

Iyengar, Shanto. and Donald Kinder. 1987. News that Matters: Television and American
Opinion. Chicago, University of Chicago Press.

J acoby William G. 1986. “Levels of Conceptualization and Reliance on the Liberal-
Conservative Continuum” Journal of Politics. 48: 423-431.

J acoby William G. 1988. “The sources of Liberal-conservative Thinking: Education and
Conceptualization. Political Behavior. 10:316-332.

Jacoby, William G. 1991. “Ideological Identiﬁcation and Issue Attitudes.” American
Journal of Political Science. 35: 1 78-205.

J acoby, William G. 1997. Statistical Graphics for Univariate and Bivariate Data.
Thousand Oaks, CA: Sage Publications.

J acoby, William G. 2002. “Core Values and Political Attitudes.” In Barbara Norrander
and Clyde Wilcox, eds., Understanding Public Opinion. 2"d ed. Washington, DC:
CQ Press.

Jackson, Thomas H. and George E. Marcus. 1975. “Political Competence and Ideological
Constraint.” Social Science Research 4:93-11 1.

Jankowski, Thomas B., and John M. Strate. 1995. "Modes of Participation Over the Adult
Life Span." Political Behavior 17: 89-106

Jelen, Ted G. 1987. “The Effects of Religious Separatism on White Protestants in the
1984 Presidential Election.” Sociological Analysis. 48:30-45.

Jelen, Ted G. 1989a. “Biblical Literalism and Inerrancy: Does the Difference Make a
Difference?” Sociological Analysis 49:421-29.

279

J elen, Ted G. (ed). 1989b. Religion and Political Behavior in the United States. New
York: Praeger.

J elen. Ted G. 1991. The Political Mobilization of Religious Beliefs. New York: Praeger.

Jelen, Ted G. 1992. “Political Christianity: A Contextual Analysis.” American Journal of
Political Science. 36:692-714.

J elen, Ted G. 1993. “The Political Consequences of Religious Group Attitudes.” Journal
of Politics. 55:178-190.

J elen, Ted G. 2000. Religious Priorities and Attitudes Toward Church and State. Review
of Religious Research 42: 87-95.

Jelen, Ted G., Clyde Wilcox and Corwin E. Smidt. 1990. “Biblical Literalism and
Inerrancy: A Methodological Investigation.” Sociological Analysis 5 1 :307-13.

J elen, Ted G., and Clyde Wilcox. 1990. “Denominational preferenCe and the dimensions
of political tolerance”. Sociological Analysis 51:69-80.

J elen, Ted G., and Clyde Wilcox. 1992. “The Effects of Religious Self-Identiﬁcations on
Support for the New Christian Right: An Analysis of political Activists.” Social
Science Journal 29: 199-210.

Jelen, Ted G., and Clude Wilcox. 1997. “Conscientious Objectors in the Culture War? A
Typologu of Attitudes Toward Church-State Relations.” Sociology of Religion.
58:277-288.

Jennings M. Kent and Richard G. Niemi. 1981. Generations and Politics: A Panel Study
of Young Adults and Their Parents. Princeton, NJ: Princeton University Press.

J orstad, Erling. 1987. The New Christian Right. Lewiston, NY: Edwin Mellen Press.

Jorstad, Erling. 1993. Popular Religion in America: The Evangelical Voice. Westport,
CT: Greenwood Press.

J unn, Jane. 1991. Participation and Political Knowledge. In Political Participation and
American Democracy. ed. William Crotty. New York: Greenwood Press.

Karpov, Vyacheslav. 2002. “Religiosity and Tolerance in the United States and Poland.”
Journal for the Scientiﬁc Study of Religion. 42:267-288.

Kellstedt, Lyman. 1989. “The Meaning and Measurement of Evangelicalism: Problems

and Prospects.” In Ted G. Jelen’s (ed) Religion and Political Behavior in the
United States. New York: Praeger.

280

Kellstedt, Lyman and John C. Green. 1993 “Religious Group Identiﬁcation: Toward a
Cognitive Theory of Religious Mobilization”. In David C. Leege and Lyman A.
Kellstedt’s (ed) Rediscovering the Religious Factor in American Politics. New
York: M. E. Sharpe.

Kellstedt, Lyman A., John C. Green, James L. Guth, and Corwin . Smidt. 1996.
"Grasping the Essentials: The Social Embodiment of Religion and Political
Behavior." In Religion and the Culture Wars, edited by John C. Green, James L.
Guth, Corwin E. Smidt and Lyman A. Kellstedt. MD: Rowman and Littleﬁeld

Kellstedt, Lyman A., John C. Green, James L. Guth, and Corwin E. Smidt, 1994.
"Religious Voting Blocs in the 1992 Election: The Year of the Evangelical?"
Sociology of Religion. 55:307-326.

Kellstedt, Lyman and Corwin Smidt. 1991. “Measuring Fundamentalism: An Analysis of
Different Operational Strategies.” Journal for the Scientiﬁc Study of Religion.
30:259-278.

Kenny, Christopher B. 1992.“Political Participation and Effects from the Social
Environment.” American Journal of Political Science 36: 259—267.

Kim, Jae-On and Charles W. Mueller 1978. Factor Analysis: Statistical Methods and
Practical Issues. Thousand Oaks, CA: Sage Publications, Quantitative
Applications in the Social Sciences Series, No. 14.

Knight, Kathleen. 1985. “Ideology in the 1980 Election: Ideological Sophistication Does
Matter”. Journal of Politics. 47: 828-853.

Kohut Andrew, John C. Green, Scott Keeter, Robert C. Toth. 2000. The Diminishing
Divide: Religion 's Changing Role in American Politics. Washington DC:
Brookings Institution Press.

Krosnick, Jon A. and Duane F. Alwin. 1989. “Aging and Susceptibility to Attitude.
Change.” Journal of Personality and Social Psychology, 57:416-425.

Krosnick, Jon A. 1990. “Expertise and Political Psychology.” Social Cognition 8: 1—8.

Kuklinski, James H., Robert Luskin and John M. Bollen. 1991. “Where's the Schema?
Going Beyond the 'S' Word in Political Psychology,” American Political Science
Review, 85: 1341-1356.

Lane, Robert E. 1959. Political Life. New York: The Free Press.

Lau, Richard R. 1986. Political Schemata, Candidate Evaluations and Voting Behavior.

In Political Cognition: The 19'” Annual Carnegie Symposium on Cognition, ed
Richard R. Lau and David 0. Sears. Hillsdale, NJ: Lawrence Erlbaum.

281

Lau, Richard R. 1989. Construct Accessibility and Electoral Choice. Political Behavior.
11:5-32.

Lau, Richard R. and David P. Redlawski. 2001. "Advantages and Disadvantages of
Cognitive Heuristics in Political Decision Making." American Journal of Political
Science, 45: 951 - 971.

Lau, Richard R., Richard A. Smith, and Susan T. F iske. 1991. Political Beliefs, Policy
Interpretations and Political Persuasion. Journal of Politics. 53:644-6 75.

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: 288 - 316.

Layman, Geoffrey C. 1999. '"Culture Wars' in the American Party System: Religious and
Cultural Change Among Partisan Activists Since 1972." American Politics
Research, 27:1, pp. 89-121.

Layman, Geoffrey C. 2001. The Great Divide: Religious and Cultural Conﬂict in
American Party Politics. New York: Columbia University Press

Layman Geoffrey and Edward Carmine. 1997. “Cultural Conﬂict in American Politics:
Religious Traditionalism, Postrnaterialism, and US. Political Behavior.” Journal
of Politics. 59.3:751-757

Laythe, Brian, Deborah G. F inkel, and Lee A. Kirkpatrick. 1999. “Predicting Prejudice
From Religious Fundamentalism and Right-Wing Authoritarianism.” Journal for
the Scientiﬁc Study of Religion. 3821-10.

Laythe, Brian, Deborah G. Finkel, Robert Bringle and Lee A. Kirkpatrick. 2002.
“Religious Fundamentalism as a Predictor of Prejudice: A Two-Component
Model.” Journal for the Scientiﬁc Study of Religion. 41: 623-635.

Leege, David C. 2003. “Methodological Advances in the Study of American Religion
and Politics.” Unpublished paper presented at the annual meeting of the American

Political Science Association, Philadelphia.

Leighley, Jan 1991. “Participation as a Stimulus of Political Conceptualization.” The
Journal of Politics. 53: 198-211.

Leighley, Jan E. 1996. "Group Membership and the Mobilization of Political
Participation." The Journal of Politics 582447-463

Leighley, Jan E. 2001. “Attitudes, Opportunities and Incentives: A Field Essay on
Political Participation”. Political Research Quarterly 48: 181-209

282

Leinesch, Michael. 1993. Redeeming America: Piety and Politics in the New Christian
Right. Chapel Hill, NC: University of North Carolina Press.

Lenski, Gerhard. 1963. The Religious Factor: A Sociological Study of Religion ’s Impact
on Politics, Economics and Family Life. Garden City, NY: Doubleday-Anchor.

Liebman, Robert C. and Robert Wuthnow (eds). 1983. The New Christian Right. New
York: Aldine Publishing Company.

Lodge, Milton, Kathleen M. McGraw, Pamela Johnston Conover, Stanley Feldman, and
Arthur H. Miller. 1991. “Where is the Schema? Critiques” (in Symposium).
American Political Science Review, 85: 1357-1380.

Luskin, Robert C. 1987. “Measuring Political Sophistication”. American Journal of
Political Science. 31: 856-899.

Luskin Robert C., 1990. “Explaining Political Sophistication”. Political Behavior.
12:331-336

Luskin, Robert C. 2002. “Political Psychology, Political Behavior, and Politics:
Questions of Aggregation, Causal Distance, and Taste.” In Thinking About
Political Psychology, ed. James H. Kuklinski. New York: Cambridge University
Press.

Marsden, George M. 1980. Fundamentalism and American Culture. New York: Oxford
University Press.

Martin, William. 1996. With God on Our Side. New York: Broadway Books.

McKenzie, Brian D. 2001. “Self-Selection, Church Attendance, and Local Civic
Participation”. Journal for the Scientific Study of Religion. 40:479-488.

McKinney, Jennifer and Roger Finke 2002. “Reviving the Mainlines: An Overview of
Clergy Support for Evangelical Renewal Movements. Journal for the Scientific
Stuay of Religion. 41 :771-783

Michels, Robert 1962. Political Parties: A Sociological Study of the Oligarchic
Tendencies of Modern Democracy. New York: Free Press.

Miller, Arthur H., Patricia Gurin, Gerald Gurin, and Oksana Malanchuk. 1981. “Group
Consciousness and Political Participation.” American Journal of Political Science.
25:494-511.

Miller, Allan S., and John P. Hofﬁnan. 1999. “The Growing Divisiveness: Culture Wars
or a War of Words?” Social Forces 78:721—746.

283

Miller, Arthur H., and Martin P. Wattenberg. 1984. “Politics ﬁom the Pulpit: Religiosity
and the 1980 Elections.” Public Opinion Quarterly. 48:301-317.

Miller, Arthur H., Martin P. Wattenburg and Oksana Malanchuk. 1986 Schematic
Assessment of Presidential Candidate. American Political Science Review.
80:521-540.

Miller, Donald E. 2000. “Rethinking Church: Learning from New Paradigm Churches.”
Good News Magazine. J anuary/F ebruary.

Miller Donald E. 2003. “Emergent Patterns of Congregational Life and Leadership in the
Developing World: Personal Reﬂections from a Research Odyssey” Pulpit and
Pew research Reports. No. 3: 1-40

Miller, Joanne M. and David A. M. Peterson. 2004. “Theoretical and Empirical
Implications of Attitude Strength” Journal of Politics. 66:847-867.

Miller, Steven D. and David 0. Sears. 1986. “Stability and Change in Social Tolerance:
A Test of the Persistence Hypothesis.” American Journal of Political Science. 30:
214-236.

Miller, Warren E. 1990. “Response: Continuity in NES Studies.” Public Opinion
Quarterly, 54:191-194.

Miller, Warren and J. Merrill Shanks. 1996. The New American Voter. Cambridge, MA:
Harvard University Press.

Mockabee, Stephen T. 2002. “Religion and the 2000 Vote.” Unpublished paper presented
at the annual meeting of the American Political Science Association, Boston.

Moen, Matthew C. 1992. The Transformation of the Christian Right. Tuscaloosa:
University of Alabama Press.

Moen, Matthew C. 1996. “From Revolution to Evolution: The Changing Nature of the
Christian Right” in The New Christian Right. ed. Melvin I, Urofsky and Martha

May. Tuscaloosa: New York: Garland Publishing,Inc.

Mondak, Jeffrey J. 1999. “Reconsidering the Measurement of Political Knowledge.”
Political Analysis 8:57-82.

Mondak, Jeffrey J. 2001. “Developing Valid Knowledge Scales.” American Journal of
Political Science 45:224—238.

284

Mutz, Diana C. and Jeffery J. Mondak. 1997. “Dimensions of Sociotropic Behavior:

Group-Based Judgments of Fairness and Well-Being.” American Journal of
Political Science. 41 :284-308.

Nagel, Jack H. 1987. Participation. New York: Prentice Hall.

National Association of Evangelicals. 2004. For the Health of the Nation: An Evangelical
Call to Civil Responsibility. Accessed 20 December 2004. http://www.nae.net/
images/civic_responsibility2.pdf.

Nathan, Rich 1995. Empowered Evangelicals. Anaheim, CA:Vine Books.

Nie, Norman, Sidney S Verba. and John R. Petrocik. 1976. The Changing American
Voter. Cambridge, Mass: Harvard University Press.

Niebuhr, H. Richard. 1962. The Meaning of Revelation, New York: Macmillan.
Nolan, John. 1971. “Jesus Now: Hogwash and Holy Water.” Ramparts. 10:20-26.

Neuman, W. Russell 1981. “Differentiation and Integration: Two Dimensions of Political
Thinking.” American Journal of Sociology, 86: 1236-1268.

Neuman, W. Russell. 1986. The Paradox of Mass Politics. Cambridge: Harvard
University Press.

Norris, Pippa and Ronald Inglehart. 2004. Sacred and Secular: Religion and Politics
Worldwide. Cambridge University Press.

Paolino, Phillip. 1995. “Group-Salient Issues and Group Representation: Support for
Women Candidates in the 1992 Senate Elections.” American Journal of Political
Science. 39:294-313.

Patterson, Samuel C., and Gregory Calderia 1983.Getting Out the Vote: Participation in
Gubernatorial Elections. American Political Science Review 77:675-689.

Penning, James M. and Corwin Smidt. 2002. Evangelicalism: the Next Generation.
Grand Rapids, MI: Baker Academic.

Perkins, Jerry, Donald Fairchild, and Murray Havens. 1983. “The Effects of
Evangelicalism on Southern Black and White Political Attitudes and Voting
Behavior”. In Religion and Politics in the South: Mass and Elite Perspectives.
Tod A. Baker, Robert P. Steed, and Laurence W. Moreland, New York: Praeger.

Perrin, Robin D., Kennedy, Paul, and Miller, Donald E. 1997. "Examining the Sources of

Conservative Church Growth: Where are the New Evangelical Movements
Getting Their Numbers?," The Journal for the Scientific Study of Religion 36:1.

285

Price, Matthew J. 2002. AFTER THE REVOLUTION: A Review of Mainline Protestant
Clergy Leadership Literature Since the 1960s. Theology Today. 59: 428-50.

Price, Vincent. 1999. “Political Information.” In Measures of Political Attitudes, ed.
John P. Robinson, Phillip R. Shaver, and Lawrence S. Wrightsman. San Diego:
Academic Press.

Regnerus, Makr D., David Sikkink and Christian Smith. 1999. “Voting with the Christian
Right: Contextual and Individual Patterns of Electoral Inﬂuence.” Social Forces.
77: 1375-1402.

Rivers, Douglas and Quang H. Vuong 1988. “Limited Information Estimators and
Exogeneity Tests for Simultaneous Probit Models” Journal of Econometrics. 39:
347-366

Rosenstone, Stephen J ., and John Mark Hansen. 1993. Mobilization. Participation. and
Democracy in America. New York: Macmillan.

Rosenstone, Steven J ., John Mark Hansen, Paul Freedman, and Marguerite Grabarek.
1993. "Voter Turnout: Myth and Reality in the 1992 Election'" Paper presented at
the annual meeting of the American Political Science Association, Washington,
DC.

Rozell, Mark J. and Clyde Wilcox, eds. 1996. God at the Grassroots: The Christian Right
in the American Elections. Lanharn, Md.: Rowman and Littleﬁeld.

Sapiro, Virginia, Steven J. Rosenstone, and the National Election Studies. AMERICAN
NATIONAL ELECTION STUDIES CUMULATIVE DATA FILE, 1948-2000
[Computer ﬁle]. 11th ICPSR version. Ann Arbor, MI: University of Michigan,
Center for Political Studies [producer], 2002. Ann Arbor, MI: Inter-university
Consortium for Political and Social Research [distributor], 2002.

Scheufele, Dietram A., Matthew C. Nisbet, and Dominique Brossard 2003. “Pathways to
Political Participation? Religion, Communication Contexts, and Mass Media”
International Journal of Public Opinion Research. 15: 300 - 324.

Schlozman, Kay Lehman; Burns, Nancy; Verba, Sidney, "What happened at work

today?: A multistage model of gender, employment, and political participation."
Journal of Politics. 61: 29 - 53.

Scott, Jacqueline. 1998. “Generational Changes in Attitudes to Abortion: a Cross-national
Comparison.” European Sociological Review. 14:177-190.

Sears, David O. 1975. Political Socialization. In Fred 1. Greenstein and Nelson W. Polsby
(eds.), Handbook of Political Science Vol. 2. Reading, MA: Addison.

286

Sears, David O. and Carolyn L. Funk. 1999. “Evidence of the Long-Term Persistence of
Adults' Political Predispositions.” The Journal of Politics. 61: 1-28.

Secret, Philip E., James B. Johnson and Audrey W. Forrest. 1990. “The Impact of
Religiosity on Political Participation and Membership in Voluntary Associations
Among Black and White Americans”. Journal of Black Studies. 21: 87-102.

Shibley, Mark A. 1996. Resurgent Evangelicalism in the United States. Columbia, SC:
University of South Carolina Press.

Shibley, Mark A. 1998. "Contemporary Evangelicals: Bom-Again and World
Afﬁrming." Annals 558:67-87.

Shupe, Anson, and William Stacy. 1983. "The Moral Majority Constituency." In The
New Christian Right, ed. Robert C. Liebman and Robert Wuthnow. New York:
Aldine.

Simpson, John H. 1983. “Moral Issues and Status Politics.” In The New Christian Right,
ed. Robert Liebman and Robert Wuthnow. New York: Aldine.

Smidt, Corwin. 1987. “Evangelicals and the 1984 Election: Continuity or Change?”
American Politics Quarterly. 15:419-444.

Smidt, Corwin L. 1989. ”Praise the Lord’ Politics: A Comparative Analysis of the Social
Characteristics and Political Views of American Evangelical and Charismatic
Christians.” Sociological Analysis 50:53-72.

Smidt, Corwin L. 1988. "’Evangelicals within Contemporary American Politics:
Differentiating between Fundamentalist and Non-Fundamentalist Evangelicals?”
Western Political Quarterly 41: 601-20.

Smith, Christian. 1998. American Evangelicalism: Embattled and Thriving. Chicago:
University of Chicago Press.

Smith, Eric R. A. N. 1980. “The Levels of Conceptualization: False Measurements of
Ideological Sophistication.” American Political Science Review. 74:685-96.

Smith, Tom W. 2000. “Changes in the Generation Gap, 1972-1998” GSS Social Change
Report No. 43. National Opinion Research Center/University of Chicago.
Accessed 12 October 2004. http://cloud9.norc.uchicago.edu/dlib/sc-43.htm.

Sniderrnan, Paul M. , Philip E. Tetlock, James M. Glaser, Donald Philip Green, and

Michael Hout. 1989. Principled Tolerance and the American Mass Public. British
Journal of Political Science 19:25-46.

287

Sniderrnan, Paul M., Joseph F. Fletcher, Peter H. Russell, Philip E. Tetlock, and Brian J.
Gaines. 1991. The Fallacy of Democratic Elitism: Elie Competition and
Commitment to Civil Liberties. British Journal of Political Science 21 :349-71.

Stein, Janice Gross. 1994. Political Learning by doing: Gorbachev as Uncommitted
thinker and motivated learner. International Organization. 48:155-83.

Stark, Rodney and William Bainbridge. 1985. The Future of Religion: Secularization,
Revival and Cult Formation Berkeley, CA: University of California Press.

Stark, Rodney and Laurence Iannaccone. 1994. "A Supply-side Reinterpretation of the
'Secularization' of Europe." Journal for the Scientific Study of Religion. 33 :3
(Sept): 230-253

Stark, Rodney, Laurence R. Iannaccone, Monica Turci, and Marco Zecchi. 2002. "How
Much Has Europe Been Secularized?" Inchiesta 32(136):99-112.

Stimson James A. 1999. Boulder, Public Opinion in American: Moods, Cycles and
Swings. CO: Westview Press.

Steensland, Brian, Jerry Z. Park, Mark Regnerus, Lynn D. Robinson, W. Bradford
Wilcox, and Robert D. Woodberry 2000 "The Measure of American Religion:
Toward Improving the State of the Art" Social Forces. 79: 291-318.

Stein, Janice Gross. 1994. “Political Learning by doing: Gorbachev as Uncommitted
thinker and motivated learner.” International Organization. 48:155-83.

Strate, John M., Charles J. Parrish, Charles D. Elder and Coit C. Ford 111. 1989. “Life
Span Civic Development and Voting Participation. American Political Science
Review. 83:443-464.

Sullivan, John L. James Piereson, and George E. Marcus. 1982. Political Tolerance and
American Democracy. Chicago: University of Chicago Press.

Sullivan, John L., Pat Walsh, Michal Shamir, David G. Barnum, and James L. Gibson.
1993. Why Politicians are More Tolerant: Selective Recruitment and Socialization
Among Political Elites in Britain, Israel, New Zealand, and the United States.
British Journal of Political Science 23:51-77.

Taj fel, Henri, ed. 1981. Human Groups and Social Categories: Studies in Social
Psychology. New York: Cambridge University Press.

Taj fel, Henri, ed. 1982. Social Identity and Intergroup Relations. Cambridge: Cambridge
University Press.

288

Tan, Alexis S. 1980. ‘Mass Media Use, Issue Knowledge and Political Involvement.
Public Opinion Quarterly. 44: 241-48.

Taylor, D. Garth. 1980. “Procedures for Evaluating Trends in Public Opinion.” Public
Opinion Quarterly. 44: 86-100.

Tamney, Joseph B., and Stephen D. Johnson. 1997. The popularity of strict churches.
Review of Religious Research 39: 209-23.

Teixeira, Ruy. 1992. The Disappearing American Voter. Washington, DC: Brookings.

Troeltsch, Ernst. 1992. The Social Teaching of the Christian Churches. Louisville:
Westminster John Knox Press.

Turner, John C. 1981. “Toward a Cognitive Redeﬁnition of the Social Group.” In Social
Identity and Intergroup Relations, ed. Henri Taj fel. New York: Cambridge
University Press.

The Vineyard USA. 2004. Theological and Philosophical Statements. Accessed 10
February 2005. http://www.vineyardusa.org/aboutlbeliefs.aspx .

Thomas, Cal and Ed Dobson. 1999. Blinded By Might: Why the Religious Right Can’t
Save America. Grand Rapids, Michigan: Zondervan Publishing House.

Troeltsch, Ernst. 1992. The Social Teaching of the Christian Churches, 2 Volumes.
Louisville, Kentucky: Westminster/John Knox Press.

Verba, Sidney, Nancy Burns, and Kay Lehman Schlozman. 1997. Knowing and Caring
about Politics: Gender and Political Engagement. Journal of Politics 59 (4):]051-
1072.

Verba Sidney, Kay Lehman Schlozman, Henry E. Brady. 1993. "Race, Ethnicity and
Political Resources: Participation in the United States." British Journal of
Political Science 23: 453-97.

Verba Sidney, Kay Lehman Schlozman, Henry E. Brady. 1995. Voice and Equality:
Civic Voluntarism in American Politics. Cambridge, MA: Harvard University

Press.

Verba Sidney and Norman H. Nie. 1972. Participation In American: Political
Democracy and Social Equality. New York: Harper and Row.

Wagner, C. Peter. 1984. Your Church Can Grow. Ventura, CA: Regal Books.

Wagner, C. Peter. 1987. Strategies for Church Growth. Ventura, CA: Regal Books.

289

Wagner, C. Peter. 1994. Your Spiritual Gifts Can Help Your Church Grow. Ventura, Ca:
Regal Books.

Wald, Kenneth D., Lyman A. Kellstedt and David C. Leege. 1993. “Church Involvement
and Political Behavior” In Rediscovering the Religious Factor in American
Politics, ed. David C. Leege and Lyman Kellstedt, Armonk, NU: ME Sharpe

Wald, Kenneth D. and Michael D. Martinez. 2001. “Jewish Religiosity and Political
Attitudes in the United States and Israel. Political Behavior. 23: 377-399.

Wald, Kenneth D., Dennis E. Owen, and Samuel S. Hill, Jr. 1988. "Churches as Political
Communities." American Political Science Review 82 :531-48.

Wald, Kenneth W. and Corwin E. Smidt. 1993. “Measurement Strategies in the Study of
Religion and Politics”. In Rediscovering the Religious Factor in American
Politics, ed. David C. Leege and Lyman A. Kellstedt. New York: M. E. Sharpe.

Wallis, Roy and Steve Bruce. 1991. Secularization: Trends, data, and theory. Research in
the Social Scientific Study of Religion 3: 1-31.

Warren, Mark E. 2001 . Dry Bones Rattling. Princeton: Princeton University Press.

Ward, Peter. 2003. “The Production and Consumption of Contemporary Charismatic
Worship in Britain as Investment and Affective Alliance.” Journal of Religion
and Popular Culture. http://www.usask.ca/relst/jrpc/art4-britconsumption.html
Accessed 3 April 2005.

Webber, Robert E. 1995. “Biblical Dance and Contemporary Worship.” In Robert E.
Webber’s (ed.) The Biblical Foundations of Christian Worship. Peabody
Massachusetts: Hendrickson Publishers.

Weber, Max. 1948, From Max Weber: Essays in Sociology, ed. Hans Gerth and C.
Wright Mills. London: Routledge.

Welch, Michael R., and David C. Leege. “Dual Reference Groups and Political
Orientations: An Examination of Evangelically Oriented Catholics.” American
Journal of Political Science. 35:28-56.

White, John Kenneth. 2003. The Values Divide: American Politics and Culture in
Transition. New York: Chatham House Publishers.

Whiteman, Curtis, w. 1990. “Separatism” In Daniel G. Reid’s (ed.) Dictionary of
Christianity in America. Downers Grove, 111.: Intervarsity Press.

Wilcox, Clyde. 1986. “Fundamentalists and Politics: An Analysis of the Effects of
differing Operational Deﬁnitions.” Journal of Politics. 48:1041-1051.

290

Wilcox, Clyde. 1987. “Religious Orientations and Political Attitudes: Variations Within
the New Christian Right.” American Politics Quarterly. 15: 274-296.

Wilcox, Clyde. 1988. “Political Action Committees of the New Christian Right: a
Longitudinal Analysis.” Journal for the Scientific Study of Religion 31 :76-87.

Wilcox, Clyde. 1992. God ’s Warriors. Baltimore: John Hopkins University Press.

Wilcox, Clyde. 2000. Onward Christian Soldiers: The Religious Right in American
Politics. Boulder, Colorado: Westview Press.

Wilcox, Clyde, Ted G. Jelen, and Sharon Linzey. 1995.. Rethinking the Reasonableness
of the Religious Right. Review of Religious Research. 36: 263-276.

Wilcox, Clyde, Matthew DeBell, and Lee Sigelman. 1999. "The Second Coming of the
New Christian Right: Patterns of P0pular Support in 1984 and 1996." Social
Science Quarterly. 80: 181 - 192.

Wilson, Thomas C. 1994. "Trends in Tolerance Toward Rightist and Leftist Groups,
1976-1988 - Effects of Attitude-Change and Cohort Succession." Public Opinion
Quarterly. 58: 539-556.

Wuthnow, Robert. 1988. The Restructuring of American Religion. Princeton: Princeton
University Press.

Wyckoff, Mikel L. 1987. “Measures of Attitudinal Consistency as Indicators of
Ideological Sophistication: A Reliability and Validity Assessment”. Journal of
Politics. 49: 148-168

Young, Robert L. 1992. “Religious Orientation, Race and Support for the Death Penalty.”
Journal for the Scientific Study of Religion 31:76-87.

Zaller, John. 1990. “Political Awareness, Elite Opinion Leadership, and the Mass Survey
Response.” Social Cognition 8: 125—53.

Zaller, John. 1992. The Nature and Origins of Mass Opinion. New York: Cambridge
University Press.

291

 

011111111111 00