Professional Documents
Culture Documents
Amy Erica Smith - Religion and Brazilian Democracy - Mobilizing The People of God-Cambridge University Press (2019)
Amy Erica Smith - Religion and Brazilian Democracy - Mobilizing The People of God-Cambridge University Press (2019)
BR AZILIAN
DEMOCR ACY
Mobilizing the People of God
Amy Erica Smith
Religion and Brazilian Democracy
Editors
David E. Campbell, University of Notre Dame
Anna M. Grzymala-Busse, Stanford University
Kenneth D. Wald, University of Florida, Gainesville
Richard L. Wood, University of New Mexico
Founding Editor
David C. Leege, University of Notre Dame
In societies around the world, dynamic changes are occurring at the intersection of religion
and politics. In some settings, these changes are driven by internal shifts within religions; in
others, by shifting political structures, institutional contexts, or by war or other upheavals.
Cambridge Studies in Social Theory, Religion, and Politics publishes books that seek to
understand and explain these changes to a wide audience, drawing on insight from social
theory and original empirical analysis. We welcome work built on strong theoretical framing,
careful research design, and rigorous methods using any social scientific method(s)
appropriate to the study. The series examines the relationship of religion and politics
broadly understood, including directly political behavior, action in civil society and in the
mediating institutions that undergird politics, and the ways religion shapes the cultural
dynamics underlying political and civil society.
Mikhail A. Alexseev and Sufian N. Zhemukhov, Mass Religious Ritual and Intergroup
Tolerance: The Muslim Pilgrims’ Paradox
Luke Bretherton, Resurrecting Democracy: Faith, Citizenship, and the Politics of a Common
Life
David E. Campbell, John C. Green, and J. Quin Monson, Seeking the Promised Land:
Mormons and American Politics
Ryan L. Claassen, Godless Democrats and Pious Republicans? Party Activists, Party Capture,
and the “God Gap”
Darren W. Davis and Donald Pope-Davis, Perseverance in the Parish? Religious Attitudes
from a Black Catholic Perspective
Paul A. Djupe and Christopher P. Gilbert, The Political Influence of Churches
Joel S. Fetzer and J. Christopher Soper, Muslims and the State in Britain, France, and
Germany
François Foret, Religion and Politics in the European Union: The Secular Canopy
Jonathan Fox, A World Survey of Religion and the State
Jonathan Fox, Political Secularism, Religion, and the State: A Time Series Analysis of
Worldwide Data
Anthony Gill, The Political Origins of Religious Liberty
Brian J. Grim and Roger Finke, The Price of Freedom Denied: Religious Persecution and
Conflict in the Twenty-First Century
Kees van Kersbergen and Philip Manow, editors, Religion, Class Coalitions, and Welfare States
Mirjam Kunkler, John Madeley, and Shylashri Shankar, A Secular Age beyond the West:
Religion, Law and the State in Asia, the Middle East and North Africa
www.cambridge.org
Information on this title: www.cambridge.org/9781108482110
doi: 10.1017/9781108699655
© Amy Erica Smith 2019
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2019
Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A.
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
names: Smith, Amy Erica, 1976– author.
title: Religion and Brazilian democracy : mobilizing the people of God / Amy Erica Smith.
description: Cambridge, UK ; New York, NY : Cambridge University Press, 2019. | Series:
Cambridge studies in social theory, religion and politics | Includes bibliographical references
and index.
identifiers: lccn 2018045104 | isbn 9781108482110 (hardback)
subjects: lcsh: Christianity and politics – Brazil. | Evangelicalism – Political aspects – Brazil.
| Catholic Church – Political activity – Brazil. | Democracy – Brazil. | Social change – Political
aspects – Brazil. | Brazil – Politics and government – 1985– | BISAC: POLITICAL SCIENCE /
Government / General.
classification: lcc br115.p7 s56 2019 | ddc 261.70981–dc23
LC record available at https://lccn.loc.gov/2018045104
isbn 978-1-108-48211-0 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
For Tibi, Oscar, and Adam
Com muito amor
Ş i cu drag
Contents
part i introduction 1
1 Introduction 3
2 Clergy, Congregants, and Religious Politicians 26
3 Methods and Case Studies 46
part ii what clergy think, say, and do 59
4 What Clergy Think and Say: Religious Teachings and Political Views 61
5 What Clergy Do: Encouraging Partisan and Electoral Politics 80
part iii how congregants respond 97
6 Church Influence on Citizens’ Policy Views and Partisanship 99
7 Church Influence on Voting Behavior 113
8 Church Influence on Citizen Support for Democracy 130
part iv representation 147
9 The Representational Triangle 149
10 Conclusion: Mobilizing the People of God 162
Afterword 176
Appendix A. List of Focus Groups and Church Observations 178
Appendix B. Focus-Group Protocol 181
Appendix C. Variable Coding and Information on Studies 183
References 184
Index 203
vii
Figures
viii
List of Figures ix
x
Acknowledgments
As always in a project of this scope, I have acquired legion debts to people and
institutions over the course of the research and writing of this book. First,
I gratefully acknowledge a number of sources of external funding. A Fulbright
Postdoctoral Fellowship funded my research stay in Brazil in July–November,
2014, and a Small Research Grant from the American Political Science
Association helped to fund the quantitative and qualitative studies I executed
in that period. A Regional Faculty Research-Travel Award from the Center for
Latin American and Caribbean Studies at the University of Wisconsin-
Milwaukee funded a planning visit in 2014, prior to my extended research
trip. A National Science Foundation Doctoral Dissertation Improvement
Grant (Award Number 0921716) helped to support both the 2008 Local
Elections Study and the 2010 Brazil Electoral Panel Study.
Straddling the line between internal and external funding is the support
I have received from the Kellogg Institute for International Studies at the
University of Notre Dame, where I was a Visiting Fellow in the 2016–2017
academic year. The Kellogg Institute provided material support for writing and
research over the course of that year, as well as opportunities to present and
workshop my work. The Institute also funded a book workshop at Notre Dame
in April, 2017. I am tremendously grateful to Director Paolo Carozza and
Associate Director Sharon Schierling, and to the Institute’s truly exceptional
and much-loved professional staff, including Denise Wright, Judy Bartlett,
Karen Clay, Therese Hanlon, and Elizabeth Rankin.
Finally, internally to Iowa State University (ISU), this research has been
supported by three Faculty Small Grant Awards from the College of Liberal
Arts and Sciences (2012–2013, 2014–2015, and 2016–2017), and by two
Foreign Travel Grants from the Provost’s Office. For help weaving all of these
resources together, I am very grateful to my chair, Mack Shelley, who has been
committed to . . . well, as one would say in Portuguese, “dar um jeito” – Mack
has always been sure we can find a way to make all the funding sources and
xi
xii Acknowledgments
travel plans and research leaves work out. I am also indebted to two highly
capable administrative support professionals in the Department of Political
Science at ISU, Shirley Barnes and Donna Burkhart, who have worked with
me – sometimes in spite of me – to try to make sure people get paid.
I also owe debts of gratitude to several people and institutions for access to
data. Thanks to the Latin American Public Opinion Project (LAPOP) and its
major supporters (the United States Agency for International Development, the
Inter-American Development Bank, and Vanderbilt University) for making the
AmericasBarometer data available. I am especially grateful for advanced access
to the 2017 Brazilian survey. Thanks also to Tim Power and César Zucco for
early access to the 2013 Brazilian Legislative Studies (BLS), and for introducing
me and Taylor Boas to those data by way of an invitation to a 2014 workshop at
Oxford. Last, Lucas Mingardi, Rafael Mucinhato, and Sergio Simoni, doctoral
candidates at the University of São Paulo, kindly shared with me data they had
painstakingly compiled on the religious affiliations of all Brazilian deputies
from the return to democracy through the 2007–2010 legislative period.
In Brazil, I have been very fortunate to have the help and friendship of many
people. My longest-running debt is to Ana Paula Evangelista Almeida, with
whom I have been working since 2008. Ana Paula and I first crossed paths when
she and Rafaela Reis – at the time both highly competent undergraduates in the
Department of Social Sciences at the Federal University of Juiz de Fora (UFJF) –
worked as my research assistants supervising a team of other undergraduate
interviewers in a study of the 2008 local election in Juiz de Fora. When
I returned to the same city in 2014 and 2017, Ana Paula worked as my
research coordinator and fieldwork supervisor. Much of the data presented in
this book bears Ana Paula’s thumbprints. Ana Paula is supersmart,
intellectually curious, hardworking, creative, funny, generous, and frugal with
other people’s money. Our work has been a true partnership, and I have learned
much from her. I am delighted that she will soon be finishing her own PhD in
rural sociology. She is already an excellent professor.
I have benefited from many other excellent research assistants in Brazil. Two
who stand out are Rafaela Reis (the abovementioned supervisor in 2008, and
herself now a professor at the UFJF) and Mariana Gonzales, who worked on the
2014 study. I am particularly grateful to Mariana for her extraordinarily good
summaries/transcripts of the focus groups we conducted. The 2014 research
also benefited from the research assistance of Pedro Makla, Suelen Masson,
Linecker Mauler, Marlon Moreira, and Aylla Prata in Juiz de Fora; Paulo Vítor
Del Rey da Silva and Júlia Vieira dos Santos in Rio de Janeiro; and Suiany Silva
de Moraes, Tuany Sousa, and Erivaldo Teixeira in Fortaleza. Thanks to Jakson
Alves de Aquino, professor at the Federal University of Ceará, and to Cesar
Zucco, professor at the Instituto Getúlio Vargas (IGV) in Rio, for their help
finding student research assistants in those two cities. Thanks also to Cesar and
Dani for their hospitality at the IGV.
Acknowledgments xiii
and Robin Globus Veldman gave extensive and very useful feedback on an early
version of the prospectus, and Jay McCann and Liz Zechmeister provided
insightful comments on a later version of the same. More generally, Jay, Liz,
and Paul have been exemplars of friendship and mentorship over these many
years. Though David Samuels has not commented on this project, discussions
with him about partisanship and voting behavior in Brazil shaped my thinking
as I was writing the book.
Earlier versions of pieces of this book were previously published in two
articles: “When Clergy Are Threatened: Catholic and Protestant Leaders and
Political Activism in Brazil,” Politics and Religion 9 (3): 431–55, in 2016; and
“Democratic Talk in Church: Religion and Political Socialization in the Context
of Urban Inequality,” World Development 99 (November 2017): 441–51.
Thanks to Cambridge University Press and to Elsevier.
Many people opened their homes and hearts to me and my family in Brazil.
José Luiz Britto Bastos and Dalila Freitas were extraordinarily helpful and
hospitable with my 2014 research visit, and became my children’s “Brazilian
grandparents.” Lilián Costa Magalhães, whom I met in 2014, became a good
friend, and she and seven-year-old Luiza were excellent hosts in 2017. Finally,
I am grateful for my time with Elisângela Andrade, a dear friend whom I met in
2008 and who became my children’s nanny when I returned in 2014. Elisângela
passed away in July 2015.
And last, my “real” family. Tibi Chelcea has been with me all the way from
my first visit to a Catholic parish in the city of Juiz de Fora in 2008. He spent
nine months with me in Brazil in 2008–2009, and four months in 2014. He has
managed with remarkable good humor my research trips away, my
conferencing, and my general distractedness and workaholic tendencies, and
he has never failed in his confidence in me and my work. I am so lucky to have
suckered this kind, cheerful feminist into putting up with me and taking care of
our two boys all these years. My mom, Esther Smith, and mother-in-law, Adina
Chelcea, have also been pillars of our little village, helping with childcare, cat-
sitting, plant-sitting, etc. Oscar and Adam joined our family on this journey,
and came along for the ride to Juiz de Fora in 2014 and South Bend in
2016–2017; they are the most wonderful distractions from work. This book is
dedicated to Tibi, Oscar, and Adam.
part i
INTRODUCTION
1
Introduction
3
4 Introduction
1
The second half of this definition is similar to that of Hunter (1992), emphasizing competing
groups defined by fundamental worldviews. However, thinking about the culture wars in a
comparative context reveals assumptions that are likely implicit in Hunter’s definition. To wit,
I differ from Hunter in emphasizing the methods – democratic politics – and the extension of the
conflict.
Introduction 5
secular democratic norms that lead them to resist some types of clergy influence.
Clergy are more influential on issues seen as core religious concerns, such as
those related to sexuality and the family, and less effective in guiding other
attitudes as well as vote choice. Doctrinally conservative citizens and
congregations are more readily influenced than others, yet even in the most
politically effective Pentecostal denomination, the Universal Church of the
Kingdom of God (UCKG), influence is far from automatic. In the early 2000s,
UCKG leaders projected that just 20 percent of “their” voters supported in-
group candidates (Conrado 2001). The partial and asymmetric influence of
clergy pushes Brazilian politics to the right, as religious conservatives are most
likely to be influenced.
Clergy also affect citizens’ democratic attitudes. On the one hand, clergy
convey their own robust support for democracy to citizens. On the other, clergy
who perceive the political system as biased against their religious in-group
undermine congregants’ confidence in that system. In addition, clergy who
promote dualistic, good-versus-evil visions of social conflict can contribute to
intolerance of out-groups such as atheists and gays.
Part IV argues that religious leaders also have substantial leverage over
politicians whom they choose to support. In Brazil’s highly permissive party
and electoral systems, hundreds of candidates run in most legislative districts,
and religious leaders have great latitude to get their chosen candidates onto
ballots. Furthermore, religious candidates attribute their electoral support more
to their grassroots religious base than to mass partisanship, elite party
organization, or ties to wealthy social groups. Thus, when religious
candidates get elected, they are strongly tied to religious patrons. Religious
institutions’ influence is intensified when “their” elected representatives are
themselves religious professionals.
This discussion provides many of the tools needed to address our second
puzzle: how are the culture wars shaping Brazilian democracy? Clergy-driven
politics both enhances and dulls representation, I argue. In classic theories in
political science, party leaders are supposed to help citizens understand how
issues fit together: “what goes with what” (Converse 1964). Parties may be
largely incapable of this kind of opinion leadership in Brazil, outside narrow
wedges of strong party identifiers (but see Samuels and Zucco 2014, 2018). By
acting as opinion leaders, however, clergy can help to align the views of religious
conservatives – both voters and legislators (Boas and Smith in press). At the
same time, they strengthen Brazil’s right more generally (Power and Rodrigues-
Silveira 2018). And in the context of the massive “Operation Car Wash”
corruption scandals (named after one money-laundering site) that have
unfolded across Brazil since 2014, religious middlemen have another positive
externality. Reliance on clergy as electoral intermediaries reduces candidates’
need for large campaign donations from wealthy individuals – the kinds of
transfers that feature prominently in corruption scandals, and that lead to
overrepresentation of business interests.
Introduction 7
behavior – that is, between explanations focusing on the social and political
circumstances stimulating doctrinal changes, and those focusing on the strategic
calculations of clergy. By contrast, I argue that explaining clergy political
activity requires us to consider the interaction between the religious supply
and demand sides; the strategic calculations of clergy respond to changes in
social and political conditions. Second, previous scholars have debated the
relative explanatory power of theologically based policy ideas and
institutional interests as incentives for clergy behavior. I argue, however, that
both ideas and institutional interests matter to clergy. Moreover, ideas shape
calculations of group interests by constraining the range of alternatives that can
be considered. Third, the richest studies of religion and politics in Latin America
have generally developed micro-level explanations of the political behavior
either of Catholic or of Protestant clergy; rarely have scholars incorporated
these two groups’ motivations and behaviors within a single study. Fully
understanding the ideological and institutional incentives clergy face,
however, requires incorporating the two groups within a single theoretical
framework.
But before we go further, let us introduce the actors who are the protagonists
of this story. What are Brazil’s major religious groups? Which citizens join
which groups? How have they taken part in Brazilian politics? Most of the
remainder of this chapter takes up these questions.
2
Throughout the text, the numbered codes beginning “CO” and “FG” denote specific visits to
congregations and other field sites, as listed in Appendix A.
10 Introduction
3
Within historical/mainline Protestantism, large religious traditions such as Presbyterianism or
Methodism have tended to be fragmented into many denominations, each with their own
organizational identity and hierarchical decision-making structures. For instance, three of the
many Presbyterian denominations in North and South America include the Presbyterian Church
of the USA (PCUSA), the Covenant Order of Evangelical Presbyterians (which largely operates in
the United States, in competition with the PCUSA), and the Presbyterian Church of Brazil.
Constituent denominations within one tradition often vary greatly in their theological and
political stances. In some traditions, constituent denominations continue largely to maintain a
shared identity and to work together within a larger federation. In the Church of England, for
example, denominations are for the most part organized territorially (by country or group of
countries); national Anglican denominations take part in a global body known as the Anglican
Communion that has no legal existence, but maintains a unified identity and shares much of its
doctrine.
Introduction 11
83.3
Percentage of Population
73.7
65
60
40
22.2
20
15.4
9.0
5.1 6.5
0 2.6 3.4 4.0
1.0
0
1872 1890 1940 1950 1960 1970 1980 1991 2000 2010
4
Robbins (2004) discusses these as defining traits of Pentecostalism and charismatic Protestantism.
However, they apply to evangelicalism more broadly.
12 Introduction
5
The AmericasBarometer is a periodic (typically biannual) series of surveys of public opinion,
democratic attitudes, and voting behavior, conducted in approximately two dozen countries of
North and South America by the Latin American Public Opinion Project (LAPOP). More
information as well as data downloads are available at www.americasbarometer.org.
14 Introduction
–2 –1 0 1 2 –2 –1 0 1 2 –2 –1 0 1 2
Coefficients estimate each variable’s impact on the likelihood of reporting a religious affiliation, versus Catholic.
When the 95 percent confidence interval does not overlap the 0 line, we say that the
variable is statistically significant.
Context – geography, time, and age group – is the most important factor
shaping who is likely to choose a religious affiliation other than Catholicism:
whether evangelical, nonreligious, or “other” (a heterogeneous category largely
comprised of small, non-Christian groups). The top row of coefficients in the
figure shows that over these nine years, respondents became much more likely
to report that they were all three non-Catholic religious affiliations. Age group
is coded in categories from “16–25” to “over 65,” and then recoded to run from
0 to 1. Older people were less likely to say they were Protestant, Pentecostal,
evangelical, or nonreligious rather than Catholic; they were more likely to be
members of other religions, such as Candomblé or Spiritist. The region of the
country also strongly influences religious affiliation. And switching away from
Catholicism is a decidedly urban phenomenon – people living in larger localities
were much more likely to choose all three non-Catholic religious affiliations
than people living in small and rural areas.
The bottom variables in the figure include a series of demographic
characteristics and personal attitudes. These variables are relatively
unimportant in determining who switched away from Catholicism. Women
were a little more likely than men to choose Protestantism, Pentecostalism, and
other religions over Catholicism, while they were less likely than men to say they
did not have a religion. People who identified as Afro-descendants were slightly
more likely to identify as Protestant/Pentecostal, over Catholic. Household wealth
did not matter, but those with university educations were slightly more likely to
say they were nonreligious or were a member of some other religion. Finally, those
who switched away from Catholicism held similar general political and social
16 Introduction
the drama
6
In Portuguese, “Crente não mexe em política” and “Irmão vota em irmão.”
Introduction 17
democracy, Catholicism was known for leftist political activism. Though the
National Conference of Bishops of Brazil had initially supported the military
regime, by the mid-1970s it turned against it and served as the most important
opposition force in civil society (Serbin 2000). Not only did Catholic leaders
campaign actively for democracy, but priests inspired by liberation theology
organized “ecclesiastical base communities” (comunidades eclesiais de base, or
CEBs) to promote political consciousness and organizing among the poor
(Bruneau 1980, 1982; Gill 1994, 1998; Mainwaring 1986). Some scholars
suggest that this activism was a response to incipient competition from
Protestants (Gill 1994, 1995; Hagopian 2008). Gill argues, “To prevent
nominal Catholics from choosing competitors, the episcopacy has advocated
(or at least tolerated) innovative reforms that better serve these individuals”
(1995, 405). During this period, the Church developed ties to the PT (Keck
1992; Mainwaring 1986; Mir 2007).7
Following the transition to democracy, however, the National Conference of
Bishops of Brazil became more quiescent. Gill (1995) argues that this transition,
too, was stimulated by institutional pressure to shore up state support for the
Church’s institutional overhead expenses. Catholic leaders today avoid overtly
partisan stances. Well-enforced norms prohibit clergy from running for office
and discourage politicking during campaigns, though pastoral letters
disseminated in parishes commonly promote nonpartisan civic norms such as
turnout and informed voting. The Church has also quietly maintained positions
of political power. For instance, it has been the guiding force behind the
interfaith National Forum for Religious Education, which advocates policy
related to public religious education under the 1988 Constitution (L. A.
Cunha 2009).
7
See also the unpublished 2018 working paper by Guadalupe Tuñón, “When the Church Votes
Left: How Progressive Religion Hurts Gender Equality.”
8
All English translations of Portuguese-language sources, interviews, surveys, and other commu-
nications in this book are my own.
18 Introduction
9
A further example of UCKG clergy disagreement with Macedo’s stance is found in [CO42].
10
Results from the 2014 Brazilian Electoral Panel Study. See Chapter 6 for further discussion.
Introduction 19
11
Between the 2012 and 2017 AmericasBarometer waves, support for the PT dropped from 17
percent to 9 percent.
20 Introduction
12
I adopt the Brazilian custom of referring to most politicians by their first names. Polling results
from “Tracking vox populi/Band/iG: Dilma tem 53% dos votos válidos,” Último Segundo,
October 2, 2010.
13
F. H. Cardoso almost certainly intended the word “demon” metaphorically, but the UCKG
authors of this newspaper article very likely intended for the word to be taken literally, given
UCKG cosmology. Thanks to Taylor Boas for sharing this newspaper, which he found in the
Columbia University archives.
Introduction 21
14
“Nos bens . . . de uso comum . . . é vedada a veiculação de propaganda de qualquer natureza,
inclusive pichação, inscrição a tinta, fixação de placas, estandartes, faixas e assemelhados.”
22 Introduction
2010s, leading to growing pessimism about the PT and dissatisfaction with the
incumbent president Dilma Rousseff. In June 2013, what Brazilians call the
“June Protests” began, as citizens nationwide took to the streets – many for
the first time in their lives – to take a stand in favor of public funding for social
services and basic infrastructure, and against large public-works projects in
preparation for the World Cup (Alonso and Mische 2016; Moseley and
Layton 2013). These protests were marked by rejection of parties and
partisanship (Alonso and Mische 2016). Then in late 2014, hints of the
Operation Car Wash corruption scandal began to appear. In this context, the
incumbent president squeaked over the finish line in first place in October 2014.
Within the first year of Dilma Rousseff’s second term, a congressional
movement for her impeachment arose. The long, political partisan process
culminated in seven-month proceedings that polarized civil society. On the
streets of most big cities, red-clad petistas (PT supporters) protesting what
they called a rightist “coup” clashed with pro-impeachment protesters decked
out in the colors of the Brazilian flag. People who made the mistake of wearing
the color red without political intent sometimes found themselves the target of
public ire. In April 2017, nine months after the completion of the impeachment,
Brazilians remained highly polarized over the legitimacy of the proceedings. In
that month, the AmericasBarometer asked Brazilians whether they thought it
had been fair. Responses were bimodal. More than half of respondents gave the
impeachment the very highest or very lowest rating: 21 percent reported a “1,”
indicating they strongly disagreed, and 32 percent reported a “7,” indicating
that they strongly agreed that the impeachment was fair. While there were small
differences between religious groups – with evangelicals on average slightly
more supportive of impeachment than Catholics, and the nonreligious slightly
less supportive – responses were bimodal within every religious group.
When President Michel Temer, Dilma’s former vice president and one of the
masterminds of the impeachment proceedings, took office in September 2016,
he was already highly unpopular. In 2017, his fiscal austerity policies and labor
reforms became the impetus for renewed protest. By June 2017, when the
AmericasBarometer went back into the field and I revisited many
congregations for follow-up interviews, Brazilians had spent two years in a
state of constant elite-level political crisis. The Operation Car Wash corruption
scandal had engulfed a large percentage of elected politicians of all parties.
Brazil was in many ways a changed country.
Brazil’s democratic troubles registered in many quantitative indicators. For
instance, the Varieties of Democracy project noted a drop in Brazil’s “Liberal
Democracy Index” from 0.78 on a 0–1 scale in 2014, to 0.57 in 2017.15 Indices
from Freedom House and the Economist have likewise registered democratic
declines since 2015. Meanwhile, public support for democracy among Brazilian
citizens has eroded even more rapidly. In 2012, 69 percent of Brazilians agreed
15
See https://www.v-dem.net/en/analysis/.
Introduction 23
with the statement that “democracy may have problems, but it is better than the
alternatives” – a level of support that had been essentially constant in 2007,
2008, and 2010. By 2014, support for democracy had dropped to 63 percent,
and by 2017 to 52 percent.16 Meanwhile, an index of citizens’ perceptions of the
legitimacy of the political system dropped from 0.50 on a 0–1 scale in 2010, to
0.34 by 2017. Levels of both partisanship overall and petismo (sympathy for the
PT) have declined in tandem with the legitimacy of democracy and the political
system.
What role have clergy, congregations, and other religious actors played?
Have they bolstered the political system? Or instead contributed to the slow,
partial erosion of Brazilian democracy? Throughout the book, we will find that
clergy have played mixed roles in this drama.
16
Data on support for democracy and the legitimacy of the political system are from the
AmericasBarometer. In the AmericasBarometer, the legitimacy of the political system is based
on responses to five questions asking about citizens’ perceptions of whether the courts, institu-
tions, and political system generally protect basic rights (see discussion in Online Appendix C).
17
This city has become a focus for a number of academic studies of social influence, analogous to
the US political-socialization studies focused on South Bend (Baker, Ames, and Renno 2006;
A. E. Smith 2016a).
24 Introduction
briefly introduced in the chapters in which they are used, and further details
presented in Appendix C.
Throughout the book, survey experiments help us get a better handle on how
congregational politicking works. A rapidly growing body of work applies
experimental methods to core questions in religion and politics.18
Experiments improve on individuals’ self-reports of what their clergy say;
they enable researchers to tease apart potential causal mechanisms for
observed correlations; and they can improve causal inference. The present
study is among the first to apply survey-experimental methods to study
religious elites (see also Calfano and Oldmixon 2016; Calfano, Michelson,
and Oldmixon 2017; Calfano, Oldmixon, and Suiter 2014). Experimental
methods are especially revealing in studying religious elites because clergy are
particularly likely to self-censor, given their leadership roles. For instance,
though very few clergy openly admitted to declining membership, priming
clergy to think about the threat of competition nonetheless affected the way
they responded to subsequent questions.
The “meat” of the book centers around six chapters describing how religious
elites, congregants, and politicians construct the culture wars and shape
Brazilian democracy. Part II focuses on clergy ideas and behavior. Chapter 4
examines what clergy think and say about policy issues, finding large gaps
between evangelical and Catholic clergy on one issue, sexuality, and smaller
gaps on a wide variety of issues. Chapter 5, then, shows that clergy widely
discuss neutral democratic norms related to political engagement. However,
they discuss candidates more sparingly.
Part III asks, how does congregational politics shape citizens’ attitudes and
behaviors? Chapter 6 finds that at the citizen level, religious gaps in policy views
are limited to gender, sexuality, and, to a growing extent, abortion. Chapter 7
then shows that churchgoers exposed to civic messages in church are more likely
to go to the polls, and that evangelical clergy’s campaigning can sway
evangelical voters. However, clergy influence is far from automatic: influence
is stronger among individuals and congregations with higher levels of doctrinal
conservatism. Chapter 8 demonstrates that religious groups have multivalent,
ambivalent impacts on congregants’ attachment to democracy.
Finally, Part IV assesses the upshot for Brazil’s representative democracy. As
shown in Chapter 9, evangelical legislators improve representation in certain
policy areas, since elected legislators as a whole have generally been
unrepresentative of Brazilian citizens on issues such as homosexuality and
abortion. However, evangelical legislators deviate from the interests of
evangelical citizens in some issue areas, promoting conservative positions on
economic policy and race that are not aligned with the evangelical base. In these
18
For a sampling of this new field, see Albertson (2011); Ben-Nun Bloom, Arikan, and
Courtemanche (2015); Boas (2014); Djupe and Calfano (2014); Glazier (2013); McClendon
and Riedl (2015); Weber and Thornton (2012).
Introduction 25
areas, the views of evangelical legislators more closely resemble those of clergy.
These policy deviations may be facilitated by personalistic ties forged between
politicians and citizens forged in church, and perhaps by clientelism.
Nonetheless, the concluding chapter argues that while Brazil’s culture wars
create polarization and push politics to the right, they may ultimately also
help to stabilize democracy by giving important civil-society groups a stake in
the Brazilian electoral game.
2
26
Clergy, Congregants, and Religious Politicians 27
exist within the party system. What leads the culture wars to develop in certain
times and places? And what are the democratic consequences when religious
groups get involved in politics? This chapter takes up these questions.
While the party-centric approach may have great appeal for explaining the
wealthy and established democracies where it was developed, it likewise fails to
explain Brazil’s culture wars. One problem is the country’s relatively low levels
of party identification. Parties cannot socialize citizens who fail to identify with
them. Though Samuels and Zucco (2018) argue that party identification in
Brazil is stronger than often assumed, partisanship and anti-partisanship have
been largely limited to one party, the center-left PT (the Partido dos
Trabalhadores, or Workers’ Party). The PT is Brazil’s only party for which
levels of mass partisanship measured by the 2014 AmericasBarometer were
roughly proportional to the party’s electoral success in voting for the National
Congress. The ideologically diffuse Party of the Brazilian Democratic
Movement (the Partido do Movimento Democrático Brasileiro, or PMDB)
had party identification at 3.9 percent in 2014 but 11.2 percent of the vote in
the Chamber of Deputies, meaning the ratio of the party’s percentage of
chamber seats to mass-level partisanship was 2.9.1 Among parties of the right
and center-right, the equivalent ratio of overrepresentation ranged from 4.0 to
well over 100.0.2 Meanwhile, politicians from the “embarrassed right” (direita
envergonhada) have tended to disown their own parties’ ideological positions in
the post-1985 era, mobilizing voters on the basis of personalistic and
clientelistic appeals, but not partisanship (Ames 2001; Mainwaring,
Meneguello, and Power 2000; Power and Zucco 2012).
A second problem with applying the party-centric approach to Brazil is the
absence of strong alliances between party and religious leaders. Despite the
PT–Catholic alliance in the party’s early years (Keck 1992; Mainwaring 1986),
today Catholic leaders eschew public partisan positions or candidate support.
The UCKG (the Igreja Universal do Reino de Deus, or Universal Church of the
Kingdom of God) also tended to support PT presidential candidates throughout
the 2000s, until the party imploded in the Operation Car Wash corruption
scandal in 2016. Other evangelical leaders have supported a wide range of
centrist and rightist parties, based on personalistic rather than ideological
criteria (Dantas 2011; Freston 1993; Lisboa 2010). Evangelical leaders today
deliberately avoid putting all their eggs in any partisan basket, scattering
support across the political spectrum and using parties strategically to
improve their own religious groups’ political standing (Dantas 2011).
The next section begins to develop a clergy-centered explanation of Brazil’s
culture wars. The Brazilian case is not unique, however; this approach can
1
Party ideology is estimated following Power and Zucco (2012). Sympathy in the electorate is
estimated based on the 2014 AmericasBarometer.
2
The Brazilian Social Democratic Party (PSDB) claimed 2.4 percent of AmericasBarometer respon-
dents but got 11.4 percent of the vote (ratio of 4.1). No other party claimed over 1.0 percent of
2014 survey respondents. Levels of party identification for some parties are estimated based on an
“Other” category that in total received 0.5 percent support. Even if the PMDB is recoded as
rightist, no more than 8 percent of interviewees identified with a rightist party, yet such parties
received well over half of votes in the 2014 legislative election.
30 Introduction
illuminate other developing country contexts. For instance, scholars debate the
causes of anti-homosexuality movements in sub-Saharan Africa. One
prominent example is Uganda’s 2009 and 2012 bills that sought to create
a “gay death penalty.” These bills had very high domestic approval, and
ultimately were defeated by means of international human rights intervention.
A party-centered approach offers little purchase in Uganda, where the party
system is relatively new and ethnically based (Conroy-Krutz 2013; Conroy-
Krutz, Moehler, and Aguilar 2016). As Grossman notes, “One implication of
the weak party system prevalent in Africa is that . . . opposition parties
commonly lack . . . the . . . capacity necessary to stimulate new issue
dimensions. Instead, they are more likely to react to cues from incumbents or
organized social groups” (2015, 341). Many scholars have instead debated
whether the culprit is local tradition, religion imposed by colonists, or other
global rightist forces (J. Anderson 2011; Cheney 2012; Sadgrove et al. 2012).
However, this focus on international networks of conservative Christians
minimizes the agency of local groups to accept, reject, or reinterpret messages.
By contrast, the activism of Pentecostal leaders and congregations better
explains observed patterns (Bompani and Terreni Brown 2015; Grossman
2015). As Grossman argues, “the upward trend in the issue saliency of LGBTs
is closely related to . . . (1) a rapid growth of Pentecostal, Evangelical, and
related Renewalist or Spirit-filled churches (demand factor) and (2)
a democratization process leading to heightened political competition (supply
factor)” (2015, 338). That is, it is the alliance of religious with political actors
that has created Uganda’s distinctive manifestation of culture-war politics.
3
Because this is a book about one country, I sometimes use terms, such as “church,” derived from
Christian traditions for the sake of simplicity and familiarity. However, I intend the approach to
apply broadly, including to non-Christian religious institutions. It should not necessarily apply to
local groups or clubs without religious content.
Clergy, Congregants, and Religious Politicians 31
Clergy take the political initiative first. In a seminal study, Verba, Schlozman,
and Brady (1995) noted the similarities between churches and labor unions as
sources of political socialization. Likewise, clergy play roles analogous to those
of union bosses in steering member opinion and in brokering between
politicians and voters. First, clergy take positions on many aspects of politics,
yet the secular norms of congregants sometimes lead clergy to self-censor.
Second, congregants adopt or resist clergy views. Third, congregants and
clergy vote for political representatives, who may or may not represent them
by sharing their views or responding to their policy demands. Representative
democracy and the culture wars emerge as higher-level properties of the system
of interactions among the three actors.
Observers of religion and politics often weigh the relative importance of
ideas and material concerns for explaining clergy behavior. Theology provides
the most obvious answers. Simply listening to religious leaders explain their
behavior, one might think their choices are straightforward responses to the
dictates of holy texts or core doctrines. However, this naïve view leads to
patently false assumptions: that religious traditions have fixed political
approaches (instead, temporal and geographic variation within groups is
tremendous); or that the political approaches of different traditions might be
inherently compatible or incompatible (instead, intergroup conflict is highly
contextually dependent) (Dowd 2015; Kuru 2009).
So why do religious leaders adopt different political ideas and practices in
different times and places? Scholars often note that theologies and practices are
socially constructed, but this does not imply that religious groups could adopt
any set of ideas. A loose analogy to Darwinian selection might help. Just as
individuals within a species may exhibit millions of possible trait mutations,
religious leaders could generate a very large variety of religious and political
ideas and practices. However, these new ideas only thrive if the individuals and
groups holding them thrive, or transmit them to new individuals and groups.
Some innovations spread more easily than others. Some innovations help
religious groups themselves survive and grow, while others inhibit survival of
the religious species.
Many scholars argue that material concerns affect which theological
innovations succeed. What “religious economy” scholars call “demand-side”
approaches emphasize social conditions, or the needs and demands of religious
adherents. As just one example, the Catholic Church’s leftist doctrine of
liberation theology may have been a response to intense poverty and
inequality in the developing world. However, this approach leads to the
obvious question of how, in a pluralistic world, religious leaders become
sensitized to varying potential constituencies whose interests are at odds:
slaveholders or slaves, landowners or peasants.
By contrast, the “supply-side” school focuses not on religious “consumers,”
but on the “opportunities and restrictions” facing religious leaders and
entrepreneurs (Finke and Iannaccone 1993, 27). Religious suppliers adapt
32 Introduction
their behavior to the opportunity structure created by the state, at times taking
oppositional stances and at others currying favor. Gill, for instance, argues that
the Catholic Church’s turn to the left in Latin America in the 1960s and 1970s
was not a response either to poverty or to theological innovation, but rather to
growing Protestant competition (1994, 1998). He also maintains that the Latin
American Church’s subsequent turn to the right in the 1980s was an attempt to
shore up state support for high overhead costs related to physical infrastructure
and training (Gill 1995).
The state-centric approach likely goes too far in the other direction, ignoring
clergy interest in addressing the societal conditions they face daily in their work.
Trejo brings the society-centered approach back in by contributing the insight
that religious demand shapes supply-side calculations (2009, 2014). That is,
interreligious competition for bodies in the pews made the Catholic Church
susceptible to constituents’ demands that clergy support pro-poor and pro-
indigenous activism – but only in places where the Church actually faced
competition.4 In this book, I conceive of clergy as rational actors who
maximize the perceived benefits and minimize the perceived costs of their
actions (e.g., Stark and Bainbridge 1996). Interreligious competition, and the
need to maintain and increase their group’s count of souls, critically shapes
clergy behavior. Thus, I coincide with Trejo in integrating the supply and
demand sides.
Yet a naïve interpretation of the supply-side approach leads to further
problems, promulgating a vision of clergy as cynical panderers who believe
that, as one critic put it, “the parishioner is always right” (Wilson Quarterly
2010). I do not assume that clergy adopt political beliefs opportunistically. Even
as clergy seek to maintain membership, they also hold sincere, yet evolving,
political and theological views that affect their incentive structures. Thus,
I concur with scholars who argue that political theology motivates clergy
action, which is still rational given complex and sometimes competing
objectives (Kuru 2009; Toft, Philpott, and Shah 2011).5 Hence, this book
weaves theological approaches together with integrated supply- and demand-
side explanations.
4
Hale (2018) further argues that Catholic parishes were only able to take advantage of increased
episcopal support for leftist activism in places where parish governance had previously been
decentralized.
5
Also see Gwyneth H. McClendon and Rachel Beatty Riedl’s manuscript “From Pews to Politics in
Africa and Beyond.” Unpublished, 2018.
Clergy, Congregants, and Religious Politicians 33
6
I believe most clergy care about their followers’ relationships to the transcendent, not just about
a tally of warm bodies in the pews. However, the reason why clergy want to gather souls is
unimportant for this explanation.
34 Introduction
souls, even nonpaying ones, complicates the calculus. As Mancur Olson (1971)
noted decades ago, targeting “selective incentives” to donors can stimulate
voluntary contributions. Prosperity theology doctrines that portray tithing as
a financial investment operate as a supernatural selective incentive. More
commonly, congregations provide a wide range of services without charge –
from mass, prayer, or worship service, to religious education, to social
connection with others in the congregation – and invite attendees to “give
back” (Falk and Fischbacher 2006; Gouldner 1960; Sugden 1984). Church
attendees may be more susceptible to calls for reciprocity, as religiosity is
associated with sustained cooperation, fairness, trust, and trustworthiness
(L. R. Anderson and Mellor 2009; Barrios and Gandelman 2014; Henrich
et al. 2010; Migheli 2017; Tan and Vogel 2008). Discussions of politics and
public issues can become one component of a “brand” cementing loyalty.
7
This is what Philpott (2007, 2009) terms “political theology.”
36 Introduction
clergy express their policy views. For instance, in a state where abortion is legal,
clergy who oppose abortion might lobby autocrats or legislators behind closed
doors (Grzymala-Busse 2015); they might organize marches in the streets; or
they might do nothing. Political process ideals are often embedded within
overtly nonpolitical theologies dealing with the nature of human agency and
society.8
What shapes clergy beliefs? Perhaps the most important influences include
internationally disseminated doctrine, childhood socialization, and adult
training in what Sandal calls “epistemic communities” (2017; Serbin 2006).
Yet views also slowly evolve in response to professional circumstances. Clergy
in hierarchical denominations are influenced and constrained by the views of
their superiors and peers (Calfano, Michelson, and Oldmixon 2017; Calfano
and Oldmixon 2016). Those whose ambitions are blocked in a hierarchy may
switch religious establishments and adopt new views (Nielsen 2017).
8
See McClendon and Riedl, “From Pews to Politics.”
Clergy, Congregants, and Religious Politicians 37
religions more or less equally.”9 Brazil is also coded in that data set as having
one of seventy-one possible restrictions on minority religions, and three of fifty-
six possible religious regulations. And the Association of Religion Data
Archives (ARDA) rates Brazil as a 0.6 on a 0–10 scale on the Government
Regulation of Religion Index, and a 0.7 on a 0–10 scale for Government
Favoritism of Religion, following Grim and Finke’s (2006) coding.10
In the past decade, observers have feared secularism might be in danger.
In 2008, the Lula administration approved a treaty with the Holy See; the
Chamber of Deputies and Senate subsequently ratified the concordat in late
2009. The concordat’s sponsors portrayed it as an innocuous ratification of
policies long in place, yet civil-society groups ranging from the Association of
Brazilian Magistrates to evangelical groups warned about infringement on
secularism (L. A. Cunha 2009; Folha Online 2009a, 2009b; Schiavon 2009;
L. M. F. de Souza 2016). Nonetheless, over the ensuing years the concordat does
not appear to have substantially shifted the Catholic–evangelical balance of
power.
As Menchik notes, around the world, “religious organizations and secular
state authority have coevolved over the course of the twentieth century” (2016,
11). What Mayrl (2016) calls a country’s “secular settlement” – the legal
relationship between church and state that is fixed in the medium term –
simultaneously shapes and is shaped by clergy activism. The Brazilian state’s
high level of secularism permits most forms of religious engagement in politics,
while its imperfect neutrality motivates engagement. First, Brazilian secularism
allows religious groups great latitude to take part in politics. Toft, Philpott, and
Shah (2011) argue that religious groups become more politically active in
secular states, as they are unconstrained by the need to curry favor. Indeed,
the Superior Electoral Tribunal publishes an electoral code in advance of each
election that typically prohibits churches from engaging in many forms of
explicit, public campaigning, yet penalties are small to nonexistent.
Second, the Brazilian state’s control of resources and policy generates
church–state conflicts, motivating clergy activism. The state can provide
religious groups with funding, ranging from contracts for drug-treatment
programs to support for religious facilities. Religious education in public
schools has been another major issue of contention. Ostensibly
nondenominational religious studies was reinstated in the 1934 constitution,
and replicated in subsequent constitutions, including the current 1988
constitution, yet today both Catholics and evangelicals hope to tailor religious
education to suit their own objectives. Further, relatively minor restrictions
9
This quote is from the RAS2 codebook. See www.religionandstate.org for the RAS2 data set and
Brazil’s coding.
10
See www.religionandstate.org for the Grim and Finke/ARDA coding. Measures of social ten-
sions indicate problems: Brazil is scored as a 5 on the 0–10 scale for Social Regulation of
Religion – that is, the presence of social tensions and intolerance of religious groups.
Clergy, Congregants, and Religious Politicians 39
such as the need for building permits and radio licenses have become sources of
aggravation for evangelical and Pentecostal congregations seeking to grow
(Gaskill 2002). Clergy also see restrictions on public speech such as the use of
megaphones in public spaces, and on evangelization during publicly funded
services, as limitations on religious liberty. These opportunities, constraints,
and grievances created by the state provide religious leaders with strong
incentives for many forms of political activism, ranging from legislative
advocacy to support for in-group political candidacies.
One form of political message is likely to be universally well received by
congregants and political authorities. Since re-democratization in the 1980s,
Brazilian political culture has emphasized building citizenship among the
historically excluded (Baiocchi 2005; Barros, Bernardes, and Macedo 2015;
Morrison 2010). Political, media, and civil-society elites publicize their support
for democratic principles and participation. Religious leaders likewise see both
moral and strategic reasons for supporting the democratic order. In the context
of elections, religious and nonreligious civic education discusses the need for
a voto consciente, or a conscientious vote.11 Conscientious voting involves
showing up to the polls and following voting procedures – meaning, in the
context of electronic voting and very high multipartism, typing correctly one’s
chosen candidates’ electoral codes (up to five digits in some races) on
a computer touch screen. But it also entails making informed, non-clientelistic
choices. Embedded within Brazilian political culture themselves, clergy
intuitively adopt the language of democracy and participation. At the same
time, the noncontroversial, consensual nature of citizenship norms allows
religious groups espousing them to be seen as disinterested benefactors of
society.
11
Voto consciente literally translates as “alert” or “aware vote.” However, the connotation in
Portuguese is closer to that of “conscientious vote” in English.
12
See McClendon and Riedl, “From Pews to Politics.”
40 Introduction
messages. Even when clergy talk about politics, congregants are not guaranteed
to recognize it. Congregants might be present during political sermons, yet
might fail to grasp the political content if they are uninterested or simply not
paying attention. Clergy may be intentionally or unintentionally vague if they
are concerned about annoying congregants. When parishioners and clergy
disagree at the outset, the need to keep bodies in the pews may discourage
clergy from voicing opinions forcefully (Novaes 2002). Thus, I expect some
decay in signals between the religious leader’s mouth and the congregant’s ears
and brain. This will especially be the case in Catholic parishes, where clergy may
be less inclined to express their views forcefully.
Moreover, even when congregants receive and recognize political messages
from clergy, they may not be influenced because they already agree with their
clergy at the outset. Citizens tend to self-select into congregations that are
politically compatible on core issues (Chesnut 2003b; Novaes 2002).
In addition, citizens and clergy within a congregation might often
independently favor the same politicians – ones identified with their religious
groups, or who share the same religious and political perspectives (Boas 2014;
Boas and Smith 2015; Campbell, Green, and Layman 2011; McDermott 2009).
Still, sometimes clergy have real opportunities for influence. When they do,
persuasion is likely to be partial – shaping some attitudes more strongly than
others. In his model of opinion change, Zaller (1992) notes that citizens are
more readily disposed to accept some kinds of elite messages than others; in
particular, views that elites consensually share are more likely to be persuasive.
As highly trusted, local civil-society elites, clergy may be quite influential when
they espouse democratic and participatory norms echoed by a wide range of
other elites.
Clergy also communicate information on candidates’ positions. For instance,
in church-based campaigns against Dilma Rousseff in Brazil in 2010, much
effort was spent on simply informing the religious public of her stances on
sexual and reproductive politics; directly partisan messages urging vote choice
in one direction or another were largely unnecessary. In such cases, again
I would expect clergy to be highly effective. Trust in clergy as local experts
and moral authorities is likely to lend credence to the presumably factual
information they disseminate. Furthermore, information may be conveyed
and reinforced not only through sermons, but indirectly, second-, third-, or
fourth-hand through congregant social networks.
Finally, clergy can seek to persuade congregants directly, imparting their
views on public-policy issues or candidates. Public religious speech can lead
congregants to consider issues and candidates in religious frames, and in terms
of sacred rhetoric (Bloom 2013; Marietta 2008; Ryan 2014). Topics related to
the dimensions of sexual and family traditionalism and church–state relations
might be expected to be especially susceptible to such framing effects (Layman
and Green 2006).
Clergy, Congregants, and Religious Politicians 41
between election periods than will congregants. Fourth, while candidates could
provide clientelistic side-payments to either clergy or congregants, congregants
may often be cheaper, though this is not necessarily the case. Side-payments can
reduce the degree of substantive representation. Thus, congregants may end up
well represented by religiously based politicians on core issues related to sexual
and family traditionalism, yet poorly represented on noncore issues.
It is important to remember that Brazil’s permissive party system and system
of open-list proportional representation create the background conditions
leading politicians to seek clergy alliances. These rules often entail hundreds
of candidacies for local, state, and federal legislative elections. The political
incorporation and representation of Brazilian evangelicals bears intriguing
parallels to Dancygier’s (2017) description of British and Belgian parties’
incorporation of Muslim immigrants in urban ethnic enclaves. In all three
countries, electoral rules give politicians an incentive to cater to voters from
religious minorities embedded in tightly knit communities. Yet also, in all three,
leaders in those minority enclaves hold political preferences that are
substantially more conservative than those of other community members.
Ultimately, the pattern of incorporation pulls politics in a conservative
direction.
summary
What kinds of forces produce Brazil’s distinctive version of the culture wars?
And what are the consequences for Brazilian democracy? This chapter outlines
the argument of the book. I begin by considering two prominent explanations
for why the culture wars arise in some settings but not others: the intervention of
international actors, and the behavior of partisan elites. Neither approach
provides much purchase to explain the Brazilian case. By contrast, a clergy-
driven explanation not only better fits that case, but it might also prove to
provide us with leverage to understand other cases in the developing world,
such as Uganda.
In the clergy-driven culture wars, I argue that clergy hold sincere political
views, yet their behavior is also shaped by the need to keep members in the pews
and to gather the monetary resources to continue their work. On the one hand,
the danger of alienating congregants leads clergy to restrain themselves in
expressing their views; this restraint reduces the political influence of clergy,
and ultimately dampens political polarization among citizens. On the other
hand, the need to gather monetary resources sometimes stimulates clergy
activism targeting the state, since the Brazilian state both directly controls
resources and regulates churches’ other growth-oriented activities. As a result,
particularly in evangelical congregations, clergy select and campaign for in-
group politicians. The fact that clergy both select candidates and serve as
brokers between candidates and congregants privileges the interests of clergy
over those of congregants.
44 Introduction
In-Group
Politicians
ce
en
policy representation
lu
constraint (voting)
inf
clientelism or
Clergy
inf
lue
co nc
ns e
tra
int
(th
re
at
of Congregants
ex
it)
including not just monetary but also policy gains. Simultaneously, the pastor’s
power in the intermediary role can hamper representation of congregants’
policy and material interests when those interests deviate from those of the
pastor.
On the other hand, clergy often enhance representation by serving as opinion
leaders, orienting the opinions of both masses and in-group elites. In some
policy domains, religious socialization has come to substitute for the elite
opinion leadership typically associated with political parties. The limits of this
opinion leadership, of course, depend on the persuasive power of the clergy.
Such persuasive powers likely vary by policy domain, with clergy more
effectively socializing citizens in areas that are “logically constrained” by
religious doctrine (Layman and Green 2006).
3
What produced Brazil’s culture wars? And how have those culture wars affected
Brazilian democracy? The evidence I have mustered to answer these questions
comes predominantly from a case study of the 2014 presidential election.
The analysis draws heavily on several large, national-level studies: the
nationally representative 2014 Brazilian Electoral Panel Study, which
interviewed citizens across the country in seven waves, from June
to November; the nationally representative 2014 AmericasBarometer,
conducted in April; and a September 2014 Internet-based experimental survey
that recruited subjects using Facebook ads. The most novel data from that
election, though, come from the Churches North and South project:
a collection of studies of clergy and congregations conducted between
early August and the end of October 2014. The ten congregations where this
project’s research team conducted in-depth fieldwork were all located in the city
of Juiz de Fora, which will be described below. In addition, about 200 Catholic,
Pentecostal, and evangelical clergy were interviewed in the cities of Juiz de Fora
and Rio de Janeiro in the Southeast region, as were over 200 evangelical and
Pentecostal clergy attending a professional development conference in the city
of Fortaleza in the Northeast region of the country. These clergy and
congregational studies are each briefly described below.
The rich case study from 2014 is complemented with a wide range of data
from Juiz de Fora and the country as a whole, spanning the period from 2002 to
2017. First, I returned to Juiz de Fora in June 2017 to reinterview many of the
same clergy (and a few new ones) from the congregations where in-depth case
studies had been conducted in 2014. For further context, the empirical analysis
also incorporates data from the 2008 mayoral race in Juiz de Fora, from the
previous three presidential elections (those of 2002, 2006, and 2010), and from
a national-level experimental study conducted during the 2012 local elections.
Finally, the book draws on surveys conducted outside election cycles among
both citizens and legislators: the 2007–2017 AmericasBarometer and the
1990–2013 Brazilian Legislative Surveys.
46
Methods and Case Studies 47
1
Barry Ames and Reynaldo T. Rojo-Mendoza, “Urban Context and Political Behavior:
Partisanship and Polarization in Two Brazilian Cities.” Unpublished paper, 2014.
48 Introduction
mayor and was the choice of most of the local political establishment. Though
Margarida took first place in the first round, Custódio managed to turn his
campaign around and won with a comfortable margin three weeks later.
Margarida’s sexuality featured prominently in the campaign, and stimulated
vehement evangelical opposition. Custódio’s campaign put out a controversial ad
on the Free Electoral Hour (Horário Eleitoral Gratuito, the federally mandated free
television time allocated to each candidate) showing the candidate eating lunch with
his family. This advertisement was widely interpreted as a jab at Margarida for her
homosexuality. She responded by putting out an ad of herself talking about her
relationship with her father and sisters, but the damage was done. Evangelical
churches also took an active role in opposing Margarida (Miranda 2008a,
2008b). The Council of Pastors of Juiz de Fora issued a letter to member churches
supporting Custódio, in part because he was “married and had children,” and
would not “damage the Christian family” (Miranda 2008a). Member pastors were
instructed to discuss the mayoral elections with parishioners. Catholic churches,
meanwhile, took no public position on the mayoral race, though a group of
Catholics had approved a list of candidates for city council.
There is every reason to believe that dynamics within both Catholic and
evangelical churches in this particular case are typical of those elsewhere in
Brazil over the past decade. The electoral mobilization of religious groups,
partially based on issues related to sexuality, has become increasingly
common in Brazil. Both the 2010 and 2014 presidential election campaigns
were marked by Catholic and Protestant mobilization related to homosexuality
and abortion; the eventual winner of both elections, Dilma Rousseff, is a female,
twice-divorced former guerrilla fighter. And politicians’ sexuality has been
prominent in other mayoral elections as well. For instance, the 2008 mayoral
elections in São Paulo featured a great deal of controversy surrounding the
sexual orientation of Gilberto Kassab, the eventual victor.
Chapter 7 incorporates survey data gathered following this election.
In November 2008, a team of students from the Federal University of Juiz de
Fora interviewed 1,089 Juiz de Fora residents about the recently concluded local
elections. Interviewers asked about religious behavior and experiences as part of
a larger questionnaire focusing on mechanisms of social influence, building on
the recently concluded 2002–2006 Two City Study. Respondents were
clustered within twenty-two neighborhoods that had been randomly sampled
for the prior study, with approximately fifty respondents per neighborhood.
offs. After a day of interviewing, it became clear that the face-to-face survey
interview was too long for administration within a conference (it lasted
about twelve minutes). Overnight, a reduced-length questionnaire was
developed (Version C) including only the twelve most theoretically
important survey items.
Table 3.1 presents basic characteristics of the sample, as well as numbers of
interviews conducted in each sample component. The non-traditional sample
design has obvious drawbacks, in that it is not clear to which extent the clergy
interviewed are representative of all clergy in Brazil. For that matter, we lack
a sampling frame or basis of comparison. Nonetheless, this is, to the best of my
knowledge, the first study to examine the political attitudes and behavior of
Brazilian clergy. Apart from the fact that clergy attending a professional
development conference are likely somewhat more resourceful and motivated
than average, in the context of small, upstart evangelical and Pentecostal
congregations, I am unaware of reasons why this sample would deviate from
the population of Brazilian clergy. Overall, the Version A/Version B split was
slightly imbalanced by city. To address this issue, analysis controls for the
location (i.e., includes fixed effects for city of interview).
As mentioned above, the survey included several questions about competition
and membership pressures. Directly after asking clerics about their attendance
levels, interviewers asked if their number of attendees had risen, declined, or
remained the same in the past two years. Next, they asked, “How active is your
church in conducting outreach to try to invite new people to attend? Are you
very active, somewhat active, not very active, or do you not try to recruit new
members?” Finally, they asked, “Do you worry that other churches might try to
attract members of your church?” Response options to this last question were “a
lot,” “a little,” and “not at all.” While the questions on attendance levels and
changes were included in all questionnaires, the questions on outreach efforts
and worry over competition were not asked in Version C.
Figure 3.1 presents responses to these questions. Reported membership
trends were on average quite sunny – implausibly so, given that church
attendance is largely a zero-sum game. That is, though Brazil’s population is
growing (albeit fairly slowly), the percentage identifying as nonreligious is also
slowly increasing. As a result, one congregation’s increased attendance will
usually need to be matched, at least approximately, by another congregation’s
attendance decline. However, only 7 respondents (all evangelical or
Pentecostal) reported a recent membership loss; 107 reported stability; and
276 reported growth. Given the low rate of reporting attendance loss, the
leftmost set of bars presents the proportion of clergy within each religious
tradition who report that attendance is either stable or declining. While
Catholic clergy were unwilling to report membership loss, they were much
more likely to report stagnant attendance than evangelicals and Pentecostals.
To the extent that any religious group would feel insecure about membership
despite the impressive self-reported numbers, it would be the Catholic Church.
Methods and Case Studies 51
Roman Entire
Catholic Evangelical Pentecostal Sample
* One survey was mistakenly conducted with a female Catholic lay leader who was assisting
a priest. Rather than eliminate the data, I include her interview.
Next, the middle set of bars presents the percentage of respondents within
each religious tradition who say that outreach efforts are “very active.” While
50.5 percent of evangelicals and 49.5 percent of Pentecostals chose the topmost
response, only 20 percent of Catholics did so. In all three religious traditions,
the bulk of the remaining respondents said their congregation was “somewhat
active.” Only about 20 percent of respondents in each group reported that their
congregation was “not very” or “not at all” active. Outreach efforts are
correlated with changes in weekly attendance at .38.
Finally, the last set of bars presents the percentage of respondents in each
group who said they worried “a lot” or “a little” about membership
competition. Reported worry does not correspond neatly to the more
objective measures. Overall, reported levels of worry are relatively similar
across the three groups, running from 36 percent to 47.5 percent. Very few
Catholic and evangelical respondents reported high levels of worry – just
4.9 percent and 5 percent of each group. By contrast, 21.6 percent of
Pentecostals reported high levels of worry. Worry is actually very high in
those congregations with the most active outreach and with rising attendance.
For instance, 81.5 percent of those who reported high levels of worry also said
attendance was rising, while only 47.4 percent of who worried “a little” and
63.7 percent of those who worried “not at all” said attendance was rising.
Perhaps Catholic clergy failed to report declining attendance because the
Catholics who converted to evangelicalism had never attended mass to begin
with. If Catholicism were losing adherents “in name only,” we would find that
church attenders constituted a growing proportion of those identifying as
52 Introduction
80
Percentage of Clergy Reporting
60
40
20
0
though several were located downtown, drawing participants from across the
city.2 Quantitative exit interviews were conducted with approximately fifty
attendees at worship services in each church. Interviewers used gender quotas
and were told to approach every second person exiting the church. In addition,
to assess the extent to which socialization in churches imposes constraint
beyond that found in the broader population, interviews were also conducted
at five neighborhood sites near the churches: four public health clinics and, to
capture upper-income citizens who use private health providers, a shopping
mall catering to the upper-middle class.
The congregational data have two major limitations in terms of their
representativeness. First, a sample of this size cannot be assumed to be
representative of the entire population of congregations across the country, or
even within one city. Still, though random sampling of a large sample of
congregations would have been ideal, the congregations studied were
deliberately selected to represent a diverse array of religious approaches,
congregation sizes, socioeconomic levels, and – based on conversation with
local academic experts in the field – levels of political activism. That is, the
necessarily limited sample of cases was implicitly stratified based on
socioeconomic status and religious affiliation, and was selected for greater
variance on both the independent and dependent variables (Geddes 2003;
King, Keohane, and Verba 1994).
Second, people interviewed in church tend to be the kinds of people who go
to church. People encountered in focus groups in church tend to be the kinds of
people who are willing to stick around after service, or even come back on
a weeknight at the invitation of a friendly American researcher. Levels of church
attendance are high in Brazil in general, but we must assume that those levels are
somewhat overreported due to social desirability bias. However, respondents in
the congregational survey have given concrete behavioral evidence of their
willingness to attend. As a result of this second limitation, we cannot assume
that the congregational data are representative of the broader population, even
simply the population within each religious group in the city of Juiz de Fora.
To address this limitation, I use representative national-level surveys, as well as
a representative local-level study of the city of Juiz de Fora in 2008, to draw
more general inferences. I also use a pair of national-level but non-
representative Internet-based studies in which respondents were recruited
through Facebook advertisements to examine mechanisms of influence
experimentally. By contrast, the congregational data serve two purposes.
First, they enable us to examine diversity within congregations as well as
differences across congregations within a denomination. Second, the
quantitative congregational data can be merged with interviews of clergy
2
To be precise, the neighborhoods of Centro, Morro da Glória, and São Mateus, as defined by the
IBGE (Brazilian Institute of Geography and Statistics).
Methods and Case Studies 55
from seven of the eight congregations studied to examine the interplay between
clergy and their congregants.
The congregations referenced in the quantitative analysis are the following.
Note that congregation names are anonymized, and not linked to the
anonymized names used in the qualitative observations reported in Appendix
A and throughout the text, for the purpose of maintaining respondents’
confidentiality.
• “Catholic Site 1”: A parish in a low-income neighborhood, with a weekly
attendance of about 1,800–2,000.
• “Catholic Site 2”: A parish in an upper-middle-class neighborhood with
a weekly attendance of about 1,000.
• “Catholic Site 3”: A parish in a working-class neighborhood with a single
priest who serves two parishes, and a weekly attendance of about 150.
• “Catholic Site 4”: A Charismatic Catholic community with an estimated
weekly attendance over 1,000.
• “Evangelical Site 1”: A large, middle-class evangelical congregation belong-
ing to a historical Protestant denomination, with a weekly attendance of
between 2,000 and 2,500.
• “Evangelical Site 2”: A small, low-income storefront evangelical congrega-
tion belonging to a historical Protestant denomination, with a weekly atten-
dance of under 100.
• “Pentecostal Site 1”: A mixed-income Pentecostal congregation with
a weekly attendance of 350–400.
• “Pentecostal Site 2”: A low-income neo-Pentecostal congregation with
a weekly attendance of about 2,000.
Figure 3.2 further introduces the sites of the quantitative fieldwork by
presenting a few basic religious beliefs in the religious and community sites.
The top half of the figure presents views regarding the nature of the Bible:
whether it is “an ancient book recorded by men,” “the inspired word of
God,” or “the real word of God to be taken literally.” Levels of biblical
literalism varied dramatically across the religious sites, from 38.8 percent at
one evangelical Protestant site to 66.7 percent at a Pentecostal site. While it is
hard to detect patterns in this small sample of congregations, levels of biblical
literalism look to be highest at the Pentecostal sites and lowest in the middle-
class Catholic congregations. Indeed, in a multivariate analysis of both
congregational and community site interviews at the individual level, biblical
literalism is lower among upper-income respondents and higher among
evangelical and Pentecostal respondents. Negligible proportions of
respondents at any religious site said the Bible was an ancient book recorded
by men. Across the four evangelical sites, only one respondent chose this
answer, while six respondents at Catholic sites did so. By contrast, about one
in ten respondents at community sites chose this option.
56 Introduction
33.3
80
38.8 40.0
39.0 51.2
(Percentage)
55.7 56.3
60.2
47.4
60
40
66.7
57.1 56.1 60.0
48.8
20
figure 3.2 Core religious beliefs at the sites of the quantitative congregational study
The survey also asked respondents the extent to which they agreed or
disagreed on a five-point scale (that is, a scale running from “strongly
disagree” to “strongly agree”) with the statements “I frequently feel God’s
love” and “I often fear God’s wrath.” Belief in a punitive or wrathful divine
presence may affect behavior in a very different way from belief in a divinity that
is loving but not wrathful (for instance, Shariff and Rhemtulla 2012). Responses
to each of these questions are converted to a scale running from 0 to 1, so a score
of 0 corresponds to “strongly disagree,” 0.25 to “disagree,” 0.5 to “neither
agree nor disagree,” 0.75 to “agree,” and 1 to “strongly agree.” The bottom
half of Figure 3.2 presents the average level of agreement with these two
statements at each site. If a site average is statistically significantly above 0.5,
Methods and Case Studies 57
summary
This chapter has begun to introduce the empirical studies that will occupy most
of the rest of this book. Much of the data in this book comes from a series of
studies conducted in the city of Juiz de Fora, a medium-sized city in the
Southeast region that is broadly representative of key religious and political
trends. The Juiz de Fora studies include a neighborhood-based survey of the
2008 local election; a congregation-based study of the 2014 presidential
campaign; and 2017 reinterviews of clergy interviewed in 2014. After
describing these studies, this chapter presented quantitative data on the
sample characteristics, setting up the future analysis.
Beyond the studies described in this chapter, the analysis will incorporate
many other national-level studies, which will be introduced in the appropriate
chapters. Variable coding and sampling strategies are also described in further
detail in the Appendix.
part ii
What do Catholic and evangelical clergy tell their flocks about contemporary
policy issues and Brazilian democracy? Diving into the data from the 2014
clergy study, this chapter focuses on clergy teachings with respect to policy
issues such as homosexuality and abortion, and on their views of democracy
and the fairness of the political system. Chapter 5 then investigates clergy speech
with respect to elections and electoral participation.
Not surprisingly, clergy differ in their views. The study reveals very large
differences between Catholics and evangelicals in the priority given to one single
issue, homosexuality. As we will find in Chapters 6 and 9, the acceptability of
homosexuality and same-sex marriage constitutes the single policy issue tying
the views of evangelical citizens to those of their in-group elites at the pulpit and
in Congress. Beyond this one issue, Catholic and evangelical clergy also differ,
though less markedly and unwaveringly, in their positions on a broader group
of doctrinally conservative religious teachings and policy issues. Among the
policy issues on which Catholic and evangelical clergy differ are race, the
environment, and economic policy.
Second, in contrast to the religious gap in policy views, Catholic and
evangelical clergy nearly universally endorse democracy as a regime type.
The solid support for democracy on the part of clergy is particularly striking
when contrasted with the ambivalence among citizens. At the same time, clergy
profess greatly varying levels of internal (intra-congregational) and external
(societal) tolerance for diversity of opinion. While democratic attitudes are of
interest in themselves, I will show that they also shape other political behavior
of clergy, with profound downstream consequences.
Third, there are large gaps between religious groups in terms of their
grievances against the current Brazilian political system. While Catholic clergy
overwhelmingly believe the state is neutral toward their group, a substantial
minority of evangelical clergy believes that the country’s president and laws are
biased against them. Perceived fair treatment matters a great deal; clergy who
61
62 What Clergy Think, Say, and Do
think the system does not treat their group fairly tend to perceive the political
system in general as less legitimate.
What motivates clergy speech on political issues? The political teachings of
clergy, I argue in this and the next chapter, are driven by three factors. First are
democratic attitudes. In Chapter 8, we will find that Brazilian clergy’s firmly
rooted, nearly universally shared democratic commitments fortify democracy,
as pastors and priests transmit their attitudes to congregants. Nonetheless, this
chapter shows that political intolerance aggravates some clergy’s perceptions
that the political system is unfair – particularly evangelical clergy. Moreover,
intolerance also depresses support for political participation (see Chapter 5).
Second, institutional considerations matter. A novel survey experiment that
I conducted reveals that the objective of gathering souls affects what clergy say.
Likely because they actually experience dwindling attendance in real life,
Catholic clergy are uniquely sensitive to survey-experimental reminders that
they face membership pressures. Priests become less conservative about matters
of sexual morality and the family when reminded of the threat of losing
members. Evangelical clergy, by contrast, persist in talking about conservative
positions even when reminded of competitive pressures. However, as we will see
in the next chapter, institutional pressures shape both evangelical and Catholic
clergy advocacy of political participation.
Third, clergy policy views are closely tied to their theological perspectives.
Evangelicals’ conservatism is linked to deeply held, long-standing traditions of
doctrinal conservatism – traditions that are also correlated with lower internal
and external tolerance. This bundle of theological positions also motivates
clergy grievances against the political system, and, as we will see in the next
chapter, it drives politicking from the pulpit.
three scenes
1 In August 2014, about seventy clergy gather for the monthly meeting of the
Juiz de Fora Council of Pastors at the Comunidade Manancial, an evangelical
church in the Bairro Progresso. The worship space, like that of many evangelical
churches, is an unadorned, warehouse-like hall with white-painted cinderblock
walls and high ceilings, an oblong rectangle with a stage at the end farthest from the
street entrance. Two and a half hours into the monthly meeting, after we had
breakfasted and then prayed and sung together, the keynote speaker got up on
stage: Pastor Osésa Rodrigues, president of a very small national evangelical party,
the Liberal Christian Party.1 Pastor Osésa happened to be visiting from the Federal
District. In his hour-long talk, Pastor Osésa listed the threats to evangelical
churches today, threats both political and social. The LGBT movement loomed
large as a villain. The movement, he said, wanted to include “new genders” in
1
This spelling of Pastor Osésa’s name is based on media reports.
Religious Teachings and Political Views 63
topics.2 In the first two examples above, Pastor Osésa and José Luiz, the founder
of the Catholic prayer community, actually agreed on the sinfulness of
homosexuality. What distinguished their two very different congregational
approaches was not so much beliefs as the way they weighted the importance
of homosexuality relative to other issues. For José Luiz, the need for inclusion
outweighed opposition to homosexuality. Despite basic and important
theological distinctions, Catholics and evangelicals in Brazil often agree with
each other on many doctrinal and policy-related questions: that God loves
everyone, for example; that homosexuality and abortion are sins; and that it
is important to care for the poor. However, they discuss these topics with
greatly varying frequency. Clergy reports of the frequency with which they
discuss different topics give a better sense of what their congregants are more
or less likely to hear within church walls.
Religious Teachings
Before addressing policy issues, let us examine how clergy explain their faith to
congregants. Western Christian religious traditions vary greatly in perspectives
on core theological issues: God’s love and wrath; the question of whether faith
brings material rewards in the here and now; the need for evangelism to spread
the faith; and biblical prophecies regarding the End Times (that is, the
prophesied end of the current era and return of Jesus Christ). Surveys revealed
that three of these topics were quite common across religious traditions. First,
Catholic and evangelical clergy all reported that discussions of “God’s love and
pardon” are ubiquitous. Equally as important as God’s love, Pentecostal and
evangelical clergy also emphasized the need for evangelism (“the importance of
helping non-believers find God”) and to avoid sin. These latter two topics are
somewhat less common in Catholic congregations than in evangelical and
Pentecostal ones, but Catholic clergy still said they discuss these topics
frequently.
By contrast, three other issues are a source of differences within and across
religious traditions. “God’s wrath over the sinful people” was the topic on
which Catholic clergy differed the most from their Protestant colleagues.
While Catholic religious leaders on average “rarely” discuss this issue, it is on
average “frequent” in evangelical congregations, and somewhere between
occasional and frequent in Pentecostal ones. The End Times are also an
infrequent topic of discussion in Catholic congregations, but somewhere
between occasionally and frequently discussed by evangelicals and Pentecostals.
Finally, there are muted differences across religious traditions in emphasis on
the idea that “God will reward the faithful with prosperity and good health” –
a doctrine often called “prosperity theology.” However, the small inter-
2
Questions on topics of preaching were not asked in the third version of the questionnaire (see
Chapter 3 for a discussion of Version C).
Religious Teachings and Political Views 65
.25
Clergy Priorities in Discussion
0
–.25
–.5
Policy-Related Teachings
And what did clergy say about policy issues? Not surprisingly, some topics are
emphasized more in Catholic congregations, and others in evangelical ones.
Figure 4.1 assesses the reported priority given to preaching on various topics, by
the clergy member’s denomination. To produce the figure, responses to each
item were first recoded to run from 0 to 1, where 0 corresponds to discussing an
issue “very rarely,” and 1 corresponds to discussing it “very frequently.” Issues
were then ranked in terms of their order of priority in their religious tradition,
where 0 represents the average frequency of discussing all issues within that
tradition. Issues significantly above the 0 line are discussed more often than
66 What Clergy Think, Say, and Do
average, and issues significantly below the 0 line are discussed less frequently
than the average issue.
Figure 4.1 presents the eight potential topics of preaching in their order of
priority for Catholic clergy. Two survey topics relate to socioeconomic issues.
Clergy preaching on “the need for hard work and responsibility” might
encourage opposition to social assistance. By contrast, preaching on “ministry
to the poor” might encourage support for social assistance – especially in Brazil,
where the distinction between states and churches as providers of charity has
not been politicized in the way it has in the United States. Four other questions
relate to sexual and family traditionalism. Clergy who preach more frequently
on chastity and the traditional family, or on homosexuality or abortion as sins,
might push their congregants to the right on this second dimension. Finally,
clergy were asked about preaching on two other issues, “the need to care for the
environment” and “combating racism.” More frequent preaching on these two
issues would encourage left-leaning stances on these topics, though not
necessarily on the other dimensions.
In general, Catholic clergy appear to prioritize preaching on issues that
would push their congregants to the left. At the top end of the scale, Catholic
leaders say they give greatest priority to ministry to the poor, with combating
racism in second place. Those priorities are nearly inverted among evangelicals
and Pentecostals, for whom combating racism is – together with protecting the
environment – among the least important topics of discussion.
On the socioeconomic dimension, Catholic clergy give very high priority to
ministry to the poor, but only moderate emphasis to the need for hard work and
responsibility. By contrast, evangelical and Pentecostal clergy say they give
approximately equal emphasis to these two topics. Still, the differences
between Catholics and evangelicals/Pentecostals on socioeconomic concerns
are not dramatic. Many evangelical clergy strongly emphasize the importance
of ministry to the poor. For instance, one pastor affiliated with the progressive
global evangelical movement known as the “Integral Mission” gave a sermon in
which he repeatedly returned to a one-sentence refrain, Jesus’ command to his
disciples to feed an assembled crowd: “You give them something to eat”
[CO11].3
Figure 4.1 also addresses four topics related to family/sexual
traditionalism. Even highly conservative clergy may exhibit reticence about
discussing sensitive, personal issues with congregants. As Pastora Denise,
who runs a small storefront Pentecostal church on the border between
a lower-class and middle-class neighborhood, remarked, “These are
choices people make. If a person wants to talk with us about choices,
we’re going to talk about God’s will and we’re going to say that those
3
Nonetheless, the Integral Mission is viewed with suspicion by other evangelicals and Pentecostals.
One pastor reported that many of his colleagues perceived the movement as overly theologically
liberal, and insufficiently committed to a strict interpretation of the Bible [CO23].
Religious Teachings and Political Views 67
the environment rank last and second-to-last for both evangelicals and
Pentecostals.
God’s Love
Ministry to Poor
Prosperity Theology
Evangelism
Hard Work/Responsibility
Chastity
Homosexuality
Abortion
God’s Wrath
End Times
–.4 –.2 0 .2 .4
Catholic Evangelical/Pentecostal
figure 4.2 Competitive threat affects core and policy-related teachings – but only
among Catholics
70 What Clergy Think, Say, and Do
4
In fact, if we examine only respondents in the control condition group, Catholic responses in
Figure 4.1 are less distinct from their Protestant counterparts.
Religious Teachings and Political Views 71
5
Responses were on a five-point scale, ranging from “Strongly disagree” to “Strongly agree.”
Religious Teachings and Political Views 73
6
The question on internal tolerance was asked of all respondents, while the questions on external
tolerance and the legitimacy of democracy were asked in Version C of the questionnaire.
74 What Clergy Think, Say, and Do
up to all of us to pray for rain, and to teach our children to pray – but she also
implied that political leaders had extra responsibility. Referring to the PT, she
argued that, “One party in the country said that it was the salvation of the
nations, but it’s not like that . . . I said years ago, if God sends a drought to Brazil,
we’ll see if [former president] Lula is going to be able to make it rain.” She compared
the PT’s spiritual perfidy to that of the Islamic State: “The Islamic religion doesn’t
preach violence. It’s the Islamic State, falsely acting in the name of Islam. And in
Brazil, too, the LGBT movement is doing this in the name of the people” [CO34].
But the threat extended beyond the “gay movement.” In a large, middle-class
Baptist church across the city, for instance, congregants reported that the pastor
preached that “today in Congress, they’re considering a lot of laws that could hurt
the preaching of the gospel and our freedom of expression” [FG4].
As a final approach to understanding the political beliefs and attitudes of
clergy, I examine their perceptions of the way the political system treats
their groups. The clergy survey contained two questions about religious
groups’ relationships to politicians and the political system. The first
asked, “Talking about your religious group’s mission and its position in
society, do you think this country’s laws (1) favor your group, (2) hurt your
group, or (3) neither favor it nor hurt it?” Immediately following the clergy
member’s response, interviewers asked, “And the current president [Dilma
Rousseff], does she (1) favor your group, (2) hurt your group, or (3) neither
favor it nor hurt it?”
Figure 4.3 presents responses to these questions. Because answers to the two
questions are highly correlated, I created three index variables to summarize the
average extent to which each cleric believes that the country’s laws and current
president favor their group, threaten their group, or are neutral. A score of 1 on
a dimension would mean that the clergy member chose that response for both
questions, while a score of 0.5 indicates that the clergy member chose the given
response for one question, and 0 indicates they chose that response for neither
question. A priest who, for instance, says that the country’s laws are neutral but the
president is prejudiced against his group would get a score of 0.5 for “perceived
state harm,” 0.5 for “perceived state neutrality,” and 0 for “perceived state
favoritism.”
Figure 4.3 shows dramatic differences between Catholics, and
evangelicals and Pentecostals. Catholics perceive a much higher level of
state neutrality and a much lower level of state harm toward their group.
While the average level of perceived harm is .33 for evangelicals and .37 for
Pentecostals, it is only .03 for Catholics. By contrast, the average level of
perceived state neutrality is .81 for Catholics, but .47 and .48 for
evangelicals and Pentecostals, respectively. In fact, in the entire survey,
only three Catholic priests thought that either the country’s laws or the
current president hurt their group, while 73 percent thought the state was
neutral on both dimensions. Interestingly, levels of perceived favoritism are
similar across all three groups, between .16 and .21.
Religious Teachings and Political Views 75
1
.75
.5
.25
0
–5 0 5 10 –4 –2 0 2 –4 –2 0 2 4
may be more likely to want positive treatment than those who are highly
committed to maintaining state secularism. Third, religious traditions also
matter because the histories of groups’ relationships to the Brazilian state
vary dramatically. Despite the current religious free market, Catholicism
still has structural advantages. For instance, the Church has material assets
such as real estate that give it a leg up in competition, and some of the most
prominent symbols of the Brazilian nation are Catholic. As an evangelical
focus-group participant declared, “Our Lady of Aparecida [the patron saint
of Brazil] is theirs – she’s not mine!” [FG4].
Figure 4.4 presents determinants of grievances, perceived neutrality, and
favoritism; each of these dependent variables is on a three-point scale (0,
0.5, and 1). Once again, the dots and whiskers correspond to the estimated
impact of each of the independent variables (shown on the left-hand side of
the figure) on the dependent variables. The results show, first, that political
and religious beliefs matter. Clergy who say they preach more frequently
on conservative religious views, on God’s love, and on ministry to the poor
report higher levels of perceived state harm and lower levels of perceived
state neutrality, though not all of the coefficients are statistically
significantly different from zero. Second, more tolerant clergy, and those
who more strongly support democracy, are more likely to perceive the state
as neutral.7 Moreover, I find an interactive effect: clergy who are highly
7
Could believing the state favors or hurts one’s group affect the legitimacy of democracy, rather
than the reverse? It certainly seems plausible that causality works in the opposite direction.
I developed several versions of simultaneous equations models examining the mutual influence
Religious Teachings and Political Views 77
tolerant but less doctrinally conservative are also most likely to believe the
state is neutral. Third, after taking into account policy and process views,
religious traditions still matter: evangelicals and Pentecostals have higher
levels of grievances, and lower levels of perceived state neutrality.8
Determinants of perceived state favoritism are less consistent. Those who
preach more frequently on ministry to the poor are more likely to perceive the
state as favoring their group, but the other two topics of preaching are not
significantly associated with perceived favoritism. Support for democracy is
associated with higher, and external tolerance with lower, levels of perceived
favoritism. And finally, religious traditions do not differ significantly in terms of
this issue, after taking into account these other variables.
of democratic legitimacy and grievances. In all of the models, democratic legitimacy strongly
affected perceptions of the state’s neutrality or bias, while the state’s perceived neutrality or bias
had no impact on democratic legitimacy.
8
Membership pressures have no significant impact on perceived hurt/state neutrality. In several
alternative versions of the model not shown here, the membership-threat treatment has no impact
on perceived state bias or neutrality. Likewise, none of the three variables related to attendance
and membership – change in attendance, extent of outreach efforts, or weekly attendance –
significantly affects perceived state harm or neutrality. The one exception is that it appears that
clergy from larger congregations are less likely to believe the state favors their group.
9
This question was not asked in Version C of the questionnaire.
78 What Clergy Think, Say, and Do
summary
This chapter examined what clergy think about, and tell their congregants
about policy issues, the political process, and the neutrality of the political
system. Clergy differ first of all in terms of the policy issues they prioritize.
Evangelical and Pentecostal religious leaders talk with their congregants much
more frequently than do Catholics about a set of religious and policy-related
issues that I call, as a group, conservative religious teachings, such as God’s
wrath, the need to avoid sin, the need for chastity, and the “sin of
homosexuality.” When Catholic religious leaders are led to think about the
threat of losing members, they further de-emphasize these topics. By contrast,
Catholic leaders talk more frequently than do evangelicals and Pentecostals
about a series of left-leaning issues, including ministry to the poor, racism, and
the environment. While membership pressures do have some impact on what
Catholic clergy talk about, competition and the fear of membership loss do not
tell the whole story, either for Catholics or Pentecostals and evangelicals.
Instead, clerics’ choice of what to talk about is driven in large part by core
religious beliefs and religious traditions.
Second, Catholics also differ from Pentecostals and evangelicals in their
attitudes toward the political process – that is, toward how political decisions
should be made. While democracy is universally held in high regard in the
abstract, Catholic leaders are more supportive than evangelicals and
Pentecostals of diversity in opinions, both within their congregations and in
society at large. Again, the lower levels of tolerance in evangelical and
Pentecostal congregations appear to be due to basic differences between
religious traditions, rather than to differences in reactions to membership
pressures.
Third, Pentecostals and evangelicals perceive much greater levels of state bias
toward their groups than do Catholics. Three sets of factors affect perceived
state bias or neutrality: (a) policy attitudes matter – those who spend more time
talking with their congregants about either conservative policy and religious
views or ministry to the poor are less satisfied with the state’s relationship to
their religious group; (b) attitudes toward the political process also shape these
perceptions – those who are more committed to democracy and more externally
tolerant are likely to perceive the state as less biased toward their religious
group; and (c) even after accounting for policy views and attitudes toward the
political process, clergy in the Pentecostal and evangelical religious traditions
perceive greater state bias toward their religious groups. This may well be
Religious Teachings and Political Views 79
three scenes
80
Encouraging Partisan and Electoral Politics 81
were poorer and darker-skinned than the average resident. Most of the hour-
and-a-half-long service emphasized a prosperity theology message. The pastor
exhorted us to have faith that financial rewards would come to those who
prayed and who tithed to the church; tithes, in fact, were financial investments
in one’s own future earnings. As the service was ending, the pastor called all of
the congregants to the front of the room to pray. When the group was assembled
in a tight mass at the front of the hall, he began to talk with us about the election.
He spent ten minutes urging us to vote for the congregation’s candidates for
federal and state deputy – men who were supported by many local UCKG
congregations. We chanted the five-digit electoral code for these two candi-
dates, which voters would need to know in the voting booth. The pastor told us
we were free to support whomever we wanted for senator, governor, and
president. [CO30]
2. Several weeks later, I sat in a circle conducting a focus group with nine
Assembly of God (Pentecostal) pastors in a cold upstairs room of one of their
churches. Most agreed that pastors should condemn homosexuality, but not
speak against the Workers’ Party (PT) for being pro-gay. Instead, congregants
should inform themselves – as one pastor said, “the partisan aspect is obvious”
if you just look. However, a few disagreed. One held that “it’s public informa-
tion. You should tell it like it is.”1 They generally thought that clergy should not
endorse candidates, and approved of their Head Pastor André’s approach. He
“tells everyone there are candidates among the brothers and sisters [i.e., con-
gregants], and we should evaluate them and consider their proposals carefully,
and vote for them if we want to, but he doesn’t ever mention names.” Another
noted that, “the pulpit is to preach the word of God,” not to talk about
candidates. Most of the participants were uncomfortable with the very public
politicking of the celebrity Assembly of God Pastor Silas Malafaia from Rio de
Janeiro. Just one participant thought Pastor Malafaia’s behavior was appro-
priate. Still, even this participant clarified that, “You have to tell people who the
church’s candidates are, but you can’t tell them who to vote for.” [FG6]
3 In April 2016, as Brazil’s Congress was considering the impeachment of
President Dilma Rousseff, a friend in Juiz de Fora told me a story about her
small, leftist Santo Daime church. She told me that social life had become highly
polarized – both at the university and on the streets – and there was constant
danger of conflict exploding between people who were pro- and anti-
impeachment. At a congregation affiliated with her own, she said, the pastor
had come out in favor of impeachment in a sermon. Congregants who were
unhappy about that sermon had left to join her congregation. Her own pastor
had called on congregants not to talk about the impeachment at church, but “as
soon as church is over, that’s all they talk about” (personal communication).
1
Literally, “chamar os bois pelos nomes” (“call the bulls by their names”).
82 What Clergy Think, Say, and Do
100
Percentage Saying Church Leaders Are Likely to:
80
60
40
20
0
but I wouldn’t hand out materials for them. They brought a big box of campaign
materials to me, but I didn’t do anything with them. I already threw them away. [CO33]
This pastor’s resistance to allowing campaigning within church walls was not
ironclad, however. When I reinterviewed him in 2017, he described in a slightly
rueful tone how a coordinated group of supporters of one local 2016 city-
council candidate had managed to get their candidate onto the church’s events
agenda, and that the candidate had been booked to give talks and engage in
other activities during that campaign [CO39].
Figure 5.1 shows the percentage of clergy who said that it was “likely” or
“certain” (as opposed to “unlikely” or “very unlikely”) that leaders in their
congregations would “encourage members to vote”; “encourage members to
reflect faithfully on the election” (here called “conscientious voting”); and
“support a candidate to some office.” Majorities of evangelical and
Pentecostal clergy said their congregation would probably encourage both
turnout and conscientious voting. A little over 40 percent said their
congregation would likely also support a candidate.
Catholics were less likely than evangelicals and Pentecostals to agree that
their congregations would engage in any activity, though Catholic responses
varied a great deal according to the form of political engagement in question.
Fewer than one in five Catholic priests said church leaders were likely to support
any candidate for office. Catholic focus-group participants recounted a story
84 What Clergy Think, Say, and Do
they recalled of a Catholic clergy member who had tried a decade or so prior to
mobilize Catholics to support a list of Catholic candidates: “That didn’t work.
Not at all. Not everyone agreed with it” [FG1].
However, the one electoral activity with the participation of a majority of
Catholic clergy was the encouragement of conscientious voting. It is
common for Brazilian Catholic dioceses to produce handouts for clergy to
distribute to church attendees providing guidelines for voting (similarly, in
the US context, see Holman and Shockley 2017). Several versions of such
a handout circulated in Juiz de Fora during the 2014 campaign. One
developed by the Catholic communities of the Ecclesiastic Province of Juiz
de Fora for the 2014 elections began with a quote from Pope Francis:
“No one can demand that religion should be relegated to the inner
sanctum of personal life, without influence on societal and national life,
without concern for the soundness of civil institutions, without a right to
offer an opinion on events affecting society” (Pope Francis 2013). It then
stipulated twelve guidelines. The first numbered point simply described
which offices were being elected; several argued for the need to inform
oneself and to vote against corrupt candidates; and several final points
forbade both clergy candidacies and campaigning within congregational
walls, but called on Catholics to be civically engaged in other ways.
What candidates did religious leaders support? I did not broach this
sensitive topic directly with clergy. In my church observations,
congregational politicking on behalf of legislative or local-office candidates
was much more common than clergy support for presidential candidates
(see also Valle 2013). The Brazilian Electoral Panel Studies of 2010 and
2014 asked Brazilians which presidential candidates their clergy supported,
if any. Only about 15 percent of evangelicals and Pentecostals and 5 percent
of Catholics in each year reported their clergy member’s presidential
candidate. In the handful of cases where Catholic respondents knew their
priest’s candidate, Dilma Rousseff was the strong favorite in both years.
However, no single presidential candidate dominated among evangelical
clergy, and evangelical clergy did not exhibit strong preferences for
candidates of their own religious group. In 2010, only about a quarter of
evangelical support went to Marina Silva, the only major evangelical
candidate. Four years later, about half of evangelical clergy support went
to either her or Pastor Everaldo.
Though between a third and a quarter of evangelical clergy apparently
supported the Workers’ Party (PT) candidate Dilma Rousseff in each year,
I encountered a vehement strain of anti-petismo (that is, opposition to the PT)
among some Pentecostal and evangelical religious leaders (see also
A. D. Fonseca 2014). This anti-petismo was often expressed in dualistic
language. For instance, an Assembly of God pastor saw the PT as an
apocalyptic force supporting communism, totalitarianism, atheism, violence,
and the subversion of the Brazilian nation:
Encouraging Partisan and Electoral Politics 85
2
The saying is literally, “Politics isn’t a believer thing.” (Política não é coisa de crente.)
3
Literally, they “impose a halter vote” (“impõem um voto de cabresto”). This is a common idiom
referring to clientelistic voting with strong control and monitoring by a local boss, or
intermediary.
88 What Clergy Think, Say, and Do
politics often gave nuanced answers, along the lines of, “This is okay, but not
that.”
In a third evangelical focus group, respondents generally agreed that it was
fine for pastors to allow selected candidates to campaign within the
congregation, but the pastor should not actually officially endorse anyone
[FG3]. In the most permissive focus group, a participant drew the line at
allowing the candidate to give a sermon: “The candidate can’t get up at the
pulpit . . . If a candidate wants to be presented to [our congregation], the pastor
talks with him and presents him to us during worship . . . and prays for him, but
he doesn’t force anyone. We’re not being led around by the leash to vote.” Also,
he added, “the pastor doesn’t just tell us that he supports someone, he tells us
why he supports them . . . We’re not told to vote blindly. The pastor makes clear
his reasons.” Another participant added that “praying for the candidate is
biblical” [FG4]. Still, informal conversations with informants in other sites
indicated that the congregation of this permissive focus group had also lost
members who were unhappy with the intensity of clergy political mobilization.
Do congregants simply adjust their norms to the behaviors of their own
clergy? Some evidence suggests this does happen to a limited extent. In 2014,
two different national-level surveys included a question asking whether it was
appropriate for “a priest or pastor to support or campaign for a certain
candidate at election time.”4 In the LAPOP AmericasBarometer, conducted
in March and April of 2014, 49.8 percent of respondents gave this
hypothetical scenario the very lowest approval rating, and 85.4 percent
disapproved to some extent. At that time, surprisingly, Catholics and
evangelicals/Pentecostals reported indistinguishable levels of support for
clergy campaigning. In June 2014, though, just two to three months later,
a gap between religious groups appeared in the first wave of the 2014
Brazilian Electoral Panel Study (BEPS). Responses among Catholics were
nearly identical in the two surveys. However, evangelical approval of that
hypothetical scenario rose by eight percentage points between March/April
and June, to 23 percent. This suggests that as clergy began to form alliances in
preparation for the campaign that would soon begin, some evangelicals were
adjusting their norms to match the behavior of their clergy. Nonetheless, it is
also noteworthy that more than three quarters of evangelicals still disapproved
of clergy campaigning in June 2014, four months before the first- and second-
round elections.
4
In the AmericasBarometer 2014, the question is labeled BRAREL1; in BEPS 2014, it is REL1.
Encouraging Partisan and Electoral Politics 89
Social Movements
Legislative Advocacy
Turnout
Conscientious Voting
Candidate Endorsements
Catholic Evangelical/Pentecostal
figure 5.2 Competitive threat affects legislative advocacy and candidate endorsements
power (A. B. Fonseca 2008; Reich and dos Santos 2013). In the process,
they created what social-movement scholars would call new “repertoires of
participation” – new ways of organizing their communities to impact
politics (Dalton and Welzel 2014; Tarrow 1998). These styles of
campaigning are reminiscent of earlier methods of clientelistic organization
in that pastors see congregations as blocs of votes, and exert personal
influence on “their” voters. However, pastors are different from the rural
bosses, or coronéis, of old. There are no clientelistic rewards for voting with
the pastor except, perhaps, spiritual ones; nor do pastors attempt to monitor
whether voters comply. Still, these methods have proven highly effective in
promoting congregational interests.
Even if clergy try to shape the competitive landscape by seeking political
allies, religious leaders must also be aware of members’ wishes. As the third
example in this chapter’s introduction indicates, a pastor who misjudges her
audience might inadvertently drive away a third of her congregants. Clergy
political engagement is likely to be particularly risky in congregations where
congregants are politically divided, or where clergy and congregants disagree
(Djupe and Gilbert 2002, 2009).
Encouraging Partisan and Electoral Politics 91
appeals to defend the Christian family. That chapter also revealed that
doctrinally conservative clergy are more likely to believe the state is harming
their group interests, and less likely to believe the state is neutral. Now, I argue
that doctrinal conservatism motivates clergy calls to political activism, even
after accounting for perceptions of the state’s bias or neutrality. Even doctrinal
conservatives who perceive the state as neutral feel compelled to take action to
change policy. Third, democratic attitudes also shape calls to action. Clergy
who are more supportive of democracy and who are more tolerant, both
internally and externally, support democratic participation more strongly.
In the past two decades, Brazilian politics and public policy have moved to
the left on many issues related to sexual and family traditionalism. Most
salient is the legalization of same-sex marriage in a pair of high court
decisions in 2011 and 2013. But this move leftward is notable in other ways.
For instance, the controversy surrounding the sexual orientation of Margarida
Salomão in the 2008 Juiz de Fora mayoral race was a precursor of later
controversy over the sexual orientations of other leftist politicians in local
and legislative races. In the National Congress, leftist legislators increasingly
promote what their conservative critics call “gender ideology” – the notion
that gender roles are socially constructed, and that transgender individuals
deserve protection. These changes are no doubt influenced by global trends
leftward on such issues across Western democracies. Abortion is another hot-
button issue. While abortion is legal only in a handful of special circumstances,
rightist and leftist politicians both seek to move abortion policy away from the
status quo. After a December 2016 Supreme Federal Tribunal decision, some
observers speculated that the high court might soon legalize abortion. However,
late in 2017 the lower chamber took steps toward criminalizing abortion under
all circumstances by moving a controversial bill out of committee.
Leftward policy trends likely mobilize evangelical and Pentecostal clergy
activism. Not only do evangelicals and Pentecostals take conservative
positions on the dimension of sexual and family traditionalism, but as we saw
in the last chapter, their positions do not readily move leftward in response to
social pressure. The same methods of organizing that promote particularistic
group interests can also help clergy pursue policy goals. Moreover, as
evangelicals and Pentecostals grow in number, these methods of organizing
become increasingly effective.
Table 5.1 explores how these three sets of attitudes, in combination with
membership pressures, shape clergy political activity.5 For simplicity, the cells
marked with negative signs show that a given independent variable (listed in the
leftmost column) decreases clergy support for the form of political action found
5
The dependent variables are left on their original scales, and the analysis uses ordinal logistic
regression. Catholics and evangelicals/Pentecostals are presented together in a single model
because the multivariate models are underpowered among Catholics. I discuss results for the
separate groups when they differ meaningfully.
table 5.1 Characteristics associated with clergy support for political activity (all religious traditions combined)
at the top of the column; a positive sign indicates that the given independent
variable increases clergy support for that form of political activism. The table
only shows effects that are statistically significant at p<.10 (light gray) or p<.05
(darker gray).
This analysis again demonstrates the importance of membership pressures.
Political diversity within a congregation substantially decreases clergy support
for all five forms of activism. At the same time, candidate endorsements are
more common in growth-oriented congregations.6
Interestingly, perceiving the state as biased – either in favor of or against
one’s group – also motivates clergy political speech. Clergy who think the state
is neutral are less likely to encourage activism. The impact on candidate
endorsements is quite large – the probability of endorsing candidates drops
from .47 to .26 for those who say the state is neutral.
Results from a quasi-experiment that occurred in the Fortaleza component
of the clergy survey indicate that evangelical leaders become more supportive
of political activism when reminded of the state’s treatment of their religious
groups. As discussed in Chapter 3, on the second day of the evangelical
conference a new, shorter version of the survey was fielded, to address
logistical problems. The two questions on the state’s treatment of religious
groups had been included at the bottom of the original treatment version
(“Version A”) and close to the bottom, directly above the “membership-
threat” treatment, in the original control version (“Version B”). However,
in the new “Version C” questionnaire fielded on the second day of the
conference, these questions were moved to the top of the questionnaire,
following a single question on recent changes in attendance levels, and
directly before questions on political activities. Differences between
responses to Versions A and B on the first day of the conference, and those
to Version C on the second day, are instructive. The “political-grievance”
treatment in Version C is not a true experiment, since it was administered on
a single day of interviewing, and administration did not randomize between
Versions C and Versions A and B. Differences in responses between Version
C and Versions A and B might be the result of some other shock to political
attitudes on the third day of the conference, perhaps a politically motivating
talk. Nonetheless, I am not aware of any such shock that could explain
differences in effects.
Figure 5.3 presents the impacts of the grievance treatment from the second day
of the conference, relative to the control-condition responses on the first day.7
The sample is limited to Fortaleza respondents, since they may differ in various
ways from evangelicals and Pentecostals interviewed in Juiz de Fora or Rio de
Janeiro. Respondents to Version C on the second day of the conference were
6
Results hold if we also include a control for the membership-threat treatment.
7
The list of dependent variables omits conscientious-voting messages because this question was not
asked in Version C.
Encouraging Partisan and Electoral Politics 95
Social Movements
Legislative Advocacy
Turnout
Candidate Endorsements
0 .2 .4
figure 5.3 The quasi-experimental impact of the grievance treatment on clergy
support for political activism
substantially and significantly more likely to support legislative activism and to say
that they would promote turnout and endorse candidates. The apparent impact of
the questionnaire change on support for legislative activism is very large. Further
analysis reveals that effects are only statistically significant among those reporting
that the state is biased. Among those who believe the state is neutral, the
questionnaire change is not significantly associated with any form of political
activism. While we cannot conclusively say that these differences were due to the
placement of the grievance treatment at the beginning of the questionnaire, they are
suggestive.
Religious views also matter. In Table 5.1, doctrinal conservatism predicts
support for legislative activism and conscientious voting. Breaking out the
analysis by religious group, however, we find that doctrinal conservatism
strongly affects the behavior of evangelical and Pentecostal clergy, but has
little impact on Catholic clergy.
Finally, democratic orientations shape how clergy talk to congregants about
political participation. The most important such orientation is external
tolerance. Clergy who are more willing to accommodate disagreement within
society at large are more likely to support all five forms of political
participation. In addition, those who perceive democracy as more legitimate
are more supportive of social movements, while those who are more tolerant of
disagreement within their congregations more strongly support legislative
activism.
96 What Clergy Think, Say, and Do
summary
Part II of this book shows that the current period of religious-political
polarization in Brazil has resulted from two forces. First, clergy members’
strategic efforts to maintain and grow their membership in a constantly
evolving religious marketplace affect which policy views they discuss, and
which political actions they endorse. Different religious traditions respond
differently to competition. Sometimes membership pressures induce clergy to
change their religious and policy-related teachings. For instance, when Catholic
clergy are reminded of the threat of competition from other religious groups,
they become less doctrinally conservative. Membership pressures can also
induce clergy to support political activity, as religious leaders seek political
allies in their struggles against competitors. However, membership pressures
more often lead clergy to refrain from endorsing political activity, for fear of
offending members.
Second, however, sincere, nonstrategic theological and political views also
matter. Some teachings are impervious to membership threats, including
Catholic leaders’ reluctance to discuss homosexuality, and evangelical
emphasis on doctrinal conservatism. The result has been a widening gap
between Catholics and evangelicals in policy-related teachings, as Pentecostals
and evangelicals have been less likely to evolve in a liberal direction than
Catholics. As politicians and policy have likewise moved in a liberal direction,
conservative religious groups have perceived the shifting issue space as a threat.
This has led to heightened evangelical mobilization, using the methods of
political organizing that were first developed to help Pentecostals develop
elected state allies.
part iii
Part III of this book considers how citizens respond to the policy priorities and
political positions of clergy. Do we find evidence of the culture wars on the
ground among congregants and the nonreligious? That is, do political
engagement and religious polarization among clergy lead to a widening gap in
policy attitudes or vote choice among citizens? This chapter investigates
differences (or in many cases, the lack of differences) across religious groups
in partisanship, views on policy toward Christianity, family/sexual
traditionalism, and socioeconomic and environmental attitudes.
It reveals that religious influence on policy attitudes in Brazil is partial – limited
to a relatively narrow band of issues. The evidence confirms prior research
showing that Brazilian evangelicals are fairly liberal on issues such as poverty
policy and social insurance (McAdams and Lance 2013; Nishimura 2004; Pew
Research Center 2006). In many policy domains, the views of Catholics,
evangelicals, the nonreligious, and adherents to other religions are
indistinguishable. This is the case for socioeconomic attitudes, as well as
opinions on the environment and race. However, religious differences are large
and growing on the acceptability of homosexuality and support for same-sex
marriage. Religious gaps are also widening in attitudes related to abortion, while
members of different religious groups disagree strongly on the extent to which the
Brazilian state should privilege Christianity. Finally, though there have been no
religious differences in partisanship in much of the post-2002 period, a partisan
cleavage appeared in 2017 between Catholics and adherents to other religious
groups. Thus, at the mass level, the Brazilian culture wars revolve around
a limited but highly politically salient set of issue and partisan attitudes.
What explains these religious differences? Using the congregational surveys
and national-level studies, I tested for social influence within the churches
studied, looking for congruence within congregations and between
congregations and their broader neighborhoods. The qualitative and
quantitative results indicate that communities develop distinctive political
cultures on certain groups of issues – namely, those related to family and
99
100 How Congregants Respond
four scenes
1 In late October, six attendees of the São José Catholic Parish sat talking with
me in a brightly lit, newly furnished upstairs meeting room. Church was
important in the lives of all six focus-group participants. One was an active
member of the Cathedral’s Citizenship Committee; another took part in the
School of Faith, an adult Catholic education program; a third introduced herself
by declaring that, “My interest is God, Jesus, Our Lady, and my faith.” But
when the topic of homosexuality arose, the lack of a common framework for
understanding the issue was notable. Some opined that homosexuality was an
innate biological trait, others that it was a behavior. One person declared, “It’s
in the Bible, it’s a sin!” Another came out as a strong advocate for same-sex
Church Influence on Citizens’ Policy Views 101
marriage: “We need legislation that serves everyone. The government can’t
force the church to perform gay weddings, but for civil weddings, we’ve got
to resolve this issue once and for all.” Noting the diversity in perspectives,
I asked if they talked much about this topic at church. The response was
universally in the negative. As one participant said, “those topics could upset
someone and they’d never come back.” As the group wrapped up, the partici-
pant who had expressed the most negative views wanted to make sure her fellow
participants did not think she was opposed to gay people. Everyone agreed that
gay people should be allowed to attend church and participate in activ-
ities [FG1].
2 Four days earlier, I had talked with a group of young adults in a middle-class
evangelical church. Most participants had some post-secondary education, and
they were generally somewhat more tolerant of homosexuality than other
evangelicals I encountered. They agreed with each other that, as one participant
said, “The practice of homosexuality is a sin, but we should love them without
judging them and condemning them. Their sin is different from ours. We can’t
reject those people as a minority.” Despite the relatively tolerant message, the
participants clearly considered gays to be an out-group with which they con-
trasted their in-group. They expected gay people who attended their church to
reject “the homosexual lifestyle.” Another participant added, “Just like we
respect homosexuals as people, they have to respect our principles too” [FG3].
3 One Sunday in September, 2014, the wood-paneled sanctuary of the Juiz de
Fora Assembly of God held perhaps 100 adults, clustered in four groups, all
receiving Sunday school lessons. I sat with a group of about thirty women
listening to the lay teacher, Sister Enilda. Following the curriculum guide dis-
tributed in Assembly of God congregations throughout Brazil, the day’s lesson
was from the Epistle of James, and emphasized sins of omission such as greed
and lack of charity. Forcefully rejecting the prosperity gospel, Enilda declared
that, “There will always be poor people. It’s not their fault. We need to take care
to provide social services.” She exhorted those of us who had more than we
needed to donate our best things to others. “The church is one single body.
If one person is sick, we’re all sick. In the first church, we were all together,
eating together, doing things together. We need solidarity and charity inside the
church.” She framed social service not just as the right thing to do, but as
enlightened self-interest: “Today maybe I don’t need assistance, but tomorrow
I will . . . Wealth is not forever” [CO21].
4 In October 2014, at the Good News Baptist Church, ten focus-group
participants sat on folding chairs in a circle in a large, warehouse-style worship
space. Sitting alone underneath the high ceilings, the group seemed very small.
As participants introduced themselves, one man told us that he had come to
understand how important politics is because the country’s laws were prevent-
ing religious education in schools. Working as a hospital chaplain, he had
102 How Congregants Respond
found, was a very good “strategy” for evangelizing outside the church. He
would like to have school chaplains as well, but state secularism (“o Estado
Laico”) did not allow it. Because of that, he had become “a very politically
engaged person.” Two other participants chimed in, arguing that the Brazilian
state was only “quote-unquote, secular.” “Catholicism controls everything,”
they said, and “we can only teach Catholicism” in the schools [FG4]. (Note that
Catholic education is legally prohibited in public schools, but religious instruc-
tion that “respects cultural religious diversity” is part of the curriculum
[L. A. Cunha 2009]).
1
Online Appendix C explains the wording of all of these questions (or in some cases, indices).
2
Question GEN8 was as follows: “Now I’m going to talk about a controversial topic. What do you
think about men who have sex with men? Do you think they are free to do what they want, or do
you think they’re shameless, or do you think they’re mentally ill?”
Church Influence on Citizens’ Policy Views 103
.8
.6
.7
.4
.6
Level of Support
.2
.5
Economic Role of State Environment (Versus Economy)
.8
.55 .6 .65 .7 .75
.6
.4
2008 2010 2012 2014 2017 2008 2010 2012 2014 2017
Year
Brazil the change has been much faster among non-evangelicals. In the seven-
year period, support for same-sex marriage rose from .42 to .58 among
Catholics, and from .54 to .68 among “others”; but the rise was only from
.28 to .34 among evangelicals.
This growing religious gap suggests that different religious communities have
been constructing varying understandings of homosexuality. Indeed, I observed
very different language across congregations. Catholics often justified tolerance
by saying, “We are all children of God.” By contrast, one participant in a Baptist
church focus group declared that, “Gay people are not children of God. A child
of God is someone who receives the word” [FG4]. This was the only time I heard
such an opinion, and I suspect that this participant was in the minority in her
own congregation. Still, evangelical churches commonly attempt to “cure”
homosexuality through counseling and prayer, and insist that they will only
accept gays who undergo therapy. In the same focus group, a participant
reported that, “Right now in Congress, they’re planning laws to extinguish
the family as a social base.” Another participant added that, “the church is
going to be forced to perform this kind of union.” A third worried that the state
of Minas Gerais would pass laws so that “even in schools they teach that it’s
normal” [FG4].
The AmericasBarometer survey reveals smaller, but likewise growing, gaps
on abortion. Religious differences in opinion on whether abortion should be
allowed when the woman’s health is in danger were not statistically significant
104 How Congregants Respond
in 2012 or 2014, but the gap became statistically significant in 2017. This was
largely due to increasing conservatism among evangelicals.
However, the single abortion question from the cross-national
AmericasBarometer poorly matches the Brazilian criminal code, which permits
abortion when the mother’s life (not health) is in danger, and in the case of rape.
The inconsistency is problematic, since opinions on abortion are highly sensitive to
question wording (Lewis 2017). The 2010 and 2014 Brazilian Electoral Panel
Studies (BEPS) asked about scenarios matching legal conditions: “Should abortion
(1) never be permitted; (2) be permitted only in exceptional circumstances such as
rape or risk to the mother’s life; (3) be permitted at will only in the initial stages of
pregnancy; or (4) be permitted without restrictions?” In BEPS 2014, only 6 percent
of Catholics and 4 percent of evangelicals favored liberalizing abortion laws,
compared to 14 percent of those without a religion and 9 percent of those in
other religions. Majorities supported the status quo: 57 percent of Catholics,
51 percent of evangelicals, 60 percent of those without a religion, and 58 percent
of those in other religions. On the conservative end, 37 percent of Catholics and
45 percent of evangelicals favored making abortion completely illegal, while
26 percent of those without a religion and 33 percent of those in other religions did
so.3
Abortion and homosexuality are the two most important issues cleaving the
electorate by religion, and ultimately driving Brazil’s culture wars. Turning to
a third aspect of family and sexual traditionalism, I find few religious differences
in attitudes with respect to women’s roles in society and the labor force.
In 2007, Catholics were slightly more conservative than other groups on
women’s status in the labor force. However, between 2008 and 2017, surveys
reveal no religious differences in attitudes toward women’s roles in public office.
Survey data reveal virtually no religious differences on most other policies.
As the third scene in the opening of this chapter indicates, many evangelicals and
Pentecostals strongly support the social safety net. From 2008 to 2017,
AmericasBarometer asked two questions on socioeconomic attitudes: whether
the Brazilian state should own important industries, and whether the state should
attempt to redress economic inequality. While the electorate as a whole became
more conservative over this period, at no point were there significant gaps
between religious groups. Nor are there religious differences in self-placement
on the left–right ideological spectrum over the 2007–2017 period, nor in
environmental and racial attitudes. In 2014 and 2017, AmericasBarometer
interviewers asked respondents if they would prefer to protect the environment
or promote economic development. In three years, support for the environment
(versus economic development) declined dramatically, likely due to a severe
economic downturn. Yet there were essentially no differences between religious
groups. And in 2012 (the only year in which the question was asked), evangelicals
3
Satisfaction with the status quo declined somewhat between the BEPS 2010 and 2014.
Church Influence on Citizens’ Policy Views 105
25
Predicted Percentage Petista
20
15
10
5
figure 6.2 Religious affiliation and support for the Workers’ Party
were slightly, though not statistically significant, to the left of other citizens in
supporting university quotas for Afro-descendants.
Note that these views contrast dramatically with clergy priorities.
In Chapter 4, we saw that Catholic clergy are to the left of evangelical and
Pentecostal clergy on the economy, the environment, and racism. That is,
citizens’ attitudes fail to match local religious elites’ priorities on issues outside
the domain of family and sexual traditionalism. The lack of correspondence will
provide an important clue to understanding representation in Chapter 9.
What about partisanship? Chapter 5 found a streak of anti-petismo
(opposition to the Workers’ Party, or PT) among some – though certainly not
all – evangelical and Pentecostal clergy. Does this hostility translate to
congregants? Does the older Catholic–PT alliance register among citizens
today? Between 2007 and 2014, there were few statistically significant
differences in partisanship. However, Figure 6.2 shows that as the PT began
to decline in popularity between 2012 and 2014, evangelicals were quicker to
abandon the ship than were Catholics. By 2017, a statistically significant gap
had opened between Catholics and other religious groups.4
But one more policy issue drives religious divisions over politics. Scenes from
two focus groups illustrate debates over policies promoting secularism or
favoring Christianity. At the Good News Baptist Church, a participant who
worked in drug rehabilitation worried about suffering from “pressure from the
government” that prevented evangelization in public spaces and in publicly
4
This gap is robust to controls for demographics, region, and church attendance.
106 How Congregants Respond
funded rehab programs. Another participant was upset that Brazil’s anti-
discrimination and hate-speech laws punished some forms of speech against
homosexuality. “We can’t express ourselves at all. As soon as you say, ‘the
homosexuals,’ they want to shut us up” [FG4]. But not all evangelicals and
Pentecostals wanted restrictions loosened. At the Hope Church of the
Nazarene, one participant said, “We live in a secular state (‘um Estado
Laico’). I’m completely against evangelizing in schools. That’s dictatorship
and we don’t need it.” Another participant continued, “If the evangelical
church gets that opportunity, you have to open it up to everyone. If not, you
end up saying ‘you have to be this, you can’t be that.’ I’m not against it, but you
have to give everyone the opportunity” [FG5].
Unfortunately, nationally representative survey data on attitudes toward
state regulation of and privileges for Christianity are lacking. However, the
2014 congregational study provides a sense of religious divides. Church
members and citizens across Juiz de Fora were asked the extent to which they
agreed with “passing laws that recognize Christian values as the basis for our
nation.” Religious gaps were large: 74 percent of Catholics and 85 percent of
evangelicals agreed or strongly agreed with such policies, as opposed to
60 percent of the nonreligious and 50 percent of adherents to other religion.5
Differences among Catholics, evangelicals, and “others” are statistically
significant, even after controlling for frequency of religious participation and
the location/congregation of the interview.
5
The agreement among adherents of other religions may seem high. However, the most common
affiliations in the “other” category, such as Spiritism and Umbanda, incorporate elements of
Catholicism, and adherents often maintain affective ties to Catholicism.
6
See Gwyneth H. McClendon and Rachel Beatty Riedl’s manuscript “From Pews to Politics in
Africa and Beyond.” Unpublished, 2018.
Church Influence on Citizens’ Policy Views 107
Strongly
Agree
.75
Agreement
.5
.25
Strongly
Disagree
Recognize Different Same-Sex Abortion Environment Anti-Racism Anti-Poverty
Christianity Gender Marriage
Roles
figure 6.3 Clustering in policy views at eight congregations and community sites
on abortion – likely more so even than most people in their new congregations.
The conservatism and devoutness of both formerly Catholic evangelicals and
Catholic-to-Catholic switchers suggests that conservative, devout Catholics
leave congregations they view as permissive in search of stricter churches.
Congregational Results
Even if politics sometimes leads people to switch congregations, though,
congregations can still influence those in their midst. Both socialization and
selection likely occur simultaneously – sometimes even operating on the same
individual. The Churches North and South congregational study provides an
opportunity to peak into congregations to examine how opinions cohere and
change in different communities. Figure 6.3 examines support for various
policies in the eight congregations of the quantitative study, as well as
neighboring community sites. Responses are recoded to run from 0 to 1.
When the congregational average is above 0.5, the average respondent in that
congregation agrees with the policy (even if only slightly); when the
congregational average is below 0.5, the average respondent disagrees.
The confidence interval, represented by the whiskers surrounding each dot,
represents the variance in attitudes within each congregation. The larger the
confidence interval, the greater the diversity in attitudes within a congregation.
The figure reveals tremendous variation across and within congregations on
several issues core to Brazil’s culture wars. Though all groups supported laws
recognizing Christianity, the highest support was found in one evangelical
Church Influence on Citizens’ Policy Views 109
7
An index of the clergy member’s doctrinal conservatism turns out to be a better predictor of
congregant attitudes than the frequency with which the priest or pastor reports preaching on the
individual policy topic.
110 How Congregants Respond
0.25 to 0.01. Likewise, predicted support for different gender roles rises from
0.44 to 0.62, and the predicted probability of agreeing that Christianity should
receive special legal recognition rises from .77 to .92. Clergy conservatism also
has a smaller, marginally statistically significant association with attitudes
toward abortion and the environment. However, there is no association
between clergy attitudes and support for anti-racism or anti-poverty policy.
Thus, the congregational study reveals that religious communities develop
distinctive political cultures shared by clergy and congregants alike, yet
influence is partial – limited to some attitudes but not others.
Which individuals are most readily influenced? Influence turns out to be
asymmetric – limited to certain people. First, those who spend more time at
church, in worship service and other activities, are likely to be more strongly
influenced (Djupe and Gilbert 2009). Second, people with stronger secular
norms, as measured by opposition to enshrining “Christian values” in the
Brazilian nation, may resist religious influence. Third, doctrinal conservatism
may also matter. When the average opinion in a congregation diverges from
what doctrinal conservatives believe to be biblically correct, doctrinal
conservatives may actually resist influence. At other times, though, doctrinal
conservatives may be eager to follow religious leaders.8 Fifth, I control for
gender, since studies of socialization show that women are more sensitive to
social influence than are men (Djupe, Sokhey, and Gilbert 2007; Djupe,
McClurg, and Sokhey 2018); and for education and income, since both of
these resources may help citizens resist pressure to conform.
Figure 6.4 presents results from three variance function regression models.
Each model tries to predict which individuals are closest to the average opinion
in their congregations (Western and Bloome 2009; see also Boas and Smith in
press). The dependent variable is the individual’s deviation from the predicted
value for the congregation. When a coefficient for a given independent variable
along the left-hand side of the figure is statistically significantly below 0, this
means that that characteristic makes people more likely to agree with others in
their congregations. When a coefficient is statistically significantly above 0, by
contrast, it makes people less likely to agree with their congregations.
The analysis shows that people who spend more time in church come to agree
more strongly with their fellow congregants. However, only time spent in church
activities matters; attending worship service does not bring people closer to the
average opinion within their congregations. This suggests that sermons by
themselves are insufficient for social influence within congregations. Rather,
messages must be reinforced through repeated interaction with peers and
religious leaders. At the same time, those who more strongly adhere to secular
norms are more likely to diverge from their congregations on same-sex marriage
8
At the citizen level, by necessity I develop a different measure of doctrinal conservatism than the
one used for clergy. Here, doctrinal conservatism is the mean of citizens’ levels of biblical
literalism and fear of a wrathful God.
Church Influence on Citizens’ Policy Views 111
Different Same-Sex
Gender Roles Marriage Abortion
Frequency of Church Activities
Doctrinal Conservatism
Woman
Education
Income
Catholic Congregation
–.1 0 .1 –.1 0 .1 –.1 0 .1
summary
This chapter examines citizens’ policy views. Religious cleavages are pronounced
and expanding on two issues: same-sex marriage and abortion. By contrast, there
is little difference across religious groups in most other policy issues. The lack of
cleavages on issues such as racism, the environment, and the economy is
noteworthy – citizens fail to mirror the issue divides present among clergy.
Nonetheless, growing differences on the abortion and same-sex-marriage issues
may partially explain a new, though yet small, religious gap in partisanship.
112 How Congregants Respond
1
Also see Gwyneth H. McClendon and Rachel Beatty Riedl’s manuscript “From Pews to Politics in
Africa and Beyond.” Unpublished, 2018.
113
114 How Congregants Respond
two scenes
1 On the morning of September 7, 2014, I missed several calls from my
fieldwork supervisor. By the time I saw the cell-phone notifications and
called her back, things were going badly. She and a team of research
assistants were visiting the Universal Church of the Kingdom of God
(UCKG) in the Santa Emélia neighborhood to conduct interviews with
congregants as they left the morning’s prayer service. They had been stand-
ing in the dimly lit, carpeted front entry hall, as I had previously agreed with
the congregation’s head pastor. Unexpectedly, the doorman – one of a large
number of smartly uniformed, devoted attendants who made things run
around the congregation – had asked the research team to leave.
I managed to get the doorman on the phone and smooth things over.
About an hour later, though, I received another call from my fieldwork
supervisor: The head pastor had rescinded his authorization for them to
conduct interviews within the church. They could keep the completed
surveys, he agreed, but they would have to leave. Later, I heard that shortly
before the doorman had asked them to leave the first time, a couple of
interviewers had looked into a meeting room with an open door to the front
entry hall, where they had seen church members assembling large stacks of
campaign posters for the state and federal deputy candidates that were
supported by the local UCKG congregations [CO13].
2 In October 2014, the Hope Church of the Nazarene was renting a storefront
congregation in a dusty, low-income neighborhood of Juiz de Fora. The roll-
down, metal front door spanned the front of the church, which was perhaps
Church Influence on Voting Behavior 115
six meters wide. On Sundays, the music blasting from the church surely woke
any sleeping neighbors for a block or two in each direction – perhaps one reason
for the congregation’s later move (see Chapter 9). At a focus group one Sunday
afternoon that October, the fluorescent lights overhead were turned off, and the
indirect light coming in through the sunny entryway gave the meeting a mellow
atmosphere. Congregants disagreed with each other over whether they should
support coreligionist candidates. One participant held that, “without
Christians in charge, things are only going to get worse, and I’m afraid the
church will end up accepting the values of the world.” As an example, he
mentioned that “in the Vatican, the Catholic Church is wanting to support
homosexuality within church. We should pray and watch out so that doesn’t
happen.” But another participant disagreed: “When we say that we need
evangelical legislators to defend our rights, I think that’s selfish. And the other
[religious] institutions, what will happen to them? Is it just about our rights?”
Another participant chimed in: “we shouldn’t just worry about our rights, but
about the rights of society.” She pointed out that evangelicals in office too often
get involved in corruption. “Maybe some laws are going to affect us, but the
solution isn’t to have believers [evangelicals] there [in government], because the
evangelical church is losing its identity” [FG5].
2
The results discussed here are based on self-reports in the “Two City Study,” a six-wave, four-
year panel study focused on understanding how social context affected vote choice (for more
information on the study, and on the city of Caxias do Sul, see Online Appendix C).
116 How Congregants Respond
The major cleavage in that election was geographic. Those living in Juiz de Fora
preferred Lula, and those in Caxias do Sul preferred the candidate Geraldo
Alckmin.3
Further, consider 2010 and 2014 presidential vote choice, as reported in the
Brazilian Electoral Panel Studies. In both elections, Dilma Rousseff received
a plurality or majority of preferences in every religious group. Nonetheless,
religious affiliation still mattered, in that evangelicals were notably less
enthusiastic about the petista candidate than members of other groups. Yet,
despite the fact that most major candidates were consistent across the two
elections, the extent of in-group voting varied from one campaign to the
other. Assembly of God member Marina Silva captured greater support from
her coreligionists in 2014 than in 2010. In addition, the 2014 presidential
election included a minor candidate, Assembly of God Pastor Everaldo Dias
Pereira. As discussed in Chapter 5, Brazilian Electoral Panel Study (BEPS) data
indicate that Pastor Everaldo may have received more than a third of evangelical
clergy endorsements. However, he received less than 2 percent of evangelical
votes reported in BEPS 2014. One evangelical focus-group respondent reported
what he perceived as his congregation’s preference ordering, after Marina Silva
narrowly lost the first-round election in 2014: “Because he’s Catholic, Aécio
isn’t exactly our church’s ideal candidate . . . But if the choice is between an iron
rod and Dilma, we’ll vote for the iron rod. It’s bad for us now” [FG4].
Finally, Juiz de Fora’s 2008 local election presented yet another scenario:
religiously polarized voting despite the fact that all candidates shared the same
Catholic religious identification. In the first-round mayoral race, the
businessman Custódio Mattos was the strong preference of evangelicals and
those without a religion, while Margarida Salomão was the strong preference of
Catholics and those of other religions. Evangelical opposition to Margarida was
driven largely by her sexual orientation; the nonreligious preference for
Custódio may instead have been motivated by his business credentials.
Evangelicals’ united and fierce opposition to Margarida likely proved decisive
in that election, lending credence to a common evangelical claim: when
evangelical clergy are united, they can swing elections.
Taken together, these results urge a hard-to-resist conclusion: religious
affiliation matters in Brazil, but not due to any automatic tendency toward
religious in-group voting. Indeed, the second scene in the introduction of this
chapter suggests substantial ambivalence over the notion of voting based simply
on candidates’ religious identity. Instead, religious communities construct their
own stories of the political implications of their religious identities, beliefs, and
practices anew within the context of each election. Sometimes they interpret
elections in terms of group interests. At other times they focus on secular
3
The geographic divide relates to long-standing partisan differences in the two cities (for instance,
see Barry Ames and Reynaldo T. Rojo-Mendoza, “Urban Context and Political Behavior:
Partisanship and Polarization in Two Brazilian Cities.” Unpublished paper, 2014).
Church Influence on Voting Behavior 117
100
80
Percentage Reporting
60
40
20
0
figure 7.1 Campaigning and electoral discussion in eight congregations, Juiz de Fora,
2014
that the 2008 local election was marked by fierce opposition to the front-runner,
Margarida Salomão, on the basis of her sexual orientation. By contrast, though
the 2014 election featured two prominent evangelical candidates and a petista
female front-runner, neither in-group voting nor policy-related threat mobilized
evangelical clergy in the same way. The UCKG congregation discussed in the
first scene in the introduction of this chapter likely perceived my interviewers as
overly intrusive into the congregation’s campaign operations. However, the
congregation was exclusively focused on the legislative elections. As presented
in the first scene in Chapter 5, UCKG congregations in Juiz de Fora largely
ignored the executive-level races in 2014. In sum, these elections drive home the
conclusion that actual levels of evangelical mobilization vary dramatically by
race and over time, depending in large part on the nature of the candidates and
issues at stake in each election.
To what extent do congregations differ from each other? Once again, the
2014 Churches North and South congregational study provides the opportunity
to peek within religious communities. Figure 7.1 depicts dramatic variance in
clergy engagement across eight congregations. Similar percentages at all four
Catholic congregations reported hearing nonpartisan civic messages. In general,
Church Influence on Voting Behavior 119
pastor was said to support Pastor Everaldo followed their pastor’s direction,
and none of the people who said they had voted for Pastor Everaldo reported
clergy endorsement. Do clergy preferences have any influence on congregants?
Is the case of Pastor Everaldo an anomaly – either a real-life one or a result of
survey sampling error?
This section turns to the “A” in Zaller’s (1992) “R-A-S” (“Receive-Accept-
Sample”) model of opinion change. Do congregants adopt the political views
espoused by their clergy and fellow congregants? We begin by examining
congregation influence on turnout, civic engagement, and political knowledge.
The section then turns to vote choice. Throughout, three types of evidence
provide insight: nationally representative surveys; national-level, online
surveys with embedded survey experiments; and the Churches North and
South congregational study.
.95
Probability of Having Voted
.9
.85
.8
No Messages Messages
figure 7.2 Congregational messages and turnout in Juiz de Fora, 2008, and Brazil,
2014
Vote Choice
Not surprisingly, highly religious Brazilians tend to see their vote choice as
a problem to be resolved using religious methods and criteria. As a Catholic
122 How Congregants Respond
participant in a focus group reported, “When it’s time, I just go in there [to the
voting booth], and I ask the Holy Spirit to illuminate me” [FG1]. Still, as an
evangelical focus-group participant put it, “Prayer can take you down various
paths” [FG3]. This section investigates whether clergy influence election
outcomes – whether they subtly or unsubtly steer congregants down one path
or another.
Are clergy endorsements effective? In popular accounts, clergy politicking is
seen as prima facie evidence that religion affects politics. Novaes (2002)
describes the evangelical style of activism as “religious clientelism.” However,
this seems to imply that voters automatically follow the direction of their
leaders, with relatively little resistance. Yet there is actually little evidence on
the extent to which parishioners follow church leaders’ guidance. Studies in the
United States and Mexico have both indicated that churches are less able to
sway the vote than often thought (Díaz Domínguez 2006; Putnam and
Campbell 2011). As discussed in the first chapter, Conrado (2001) reported
that around the turn of the millennium, UCKG officials projected that they
could expect a 20 percent fidelity rate in congregant support for church
candidates – notable, but far from the fidelity rates often found for political
partisans.4 If UCKG organizers’ estimates remain accurate, this rate of
compliance implies that even within what is arguably Brazil’s most politically
mobilized denomination, influence is far from automatic.
We asked survey respondents whom their clergy endorsed in three
representative studies: the local election study of the 2008 Juiz de Fora mayoral
race and the national-level 2010 and 2014 Brazilian Electoral Panel Studies.5
Interestingly, in 2008, reported endorsements are statistically insignificant as
predictors of the vote, after controlling for evangelicalism. This is striking, as
this election was, out of the three, the one featuring the highest levels of
evangelical mobilization. In both 2010 and 2014, by contrast, there is some
indication that endorsements might have mattered: clergy endorsement of
Dilma is a statistically significant predictor of the vote in 2010, while clergy
support for Marina functions in the same way in 2014. However, clergy
endorsements of other candidates are statistically insignificant in vote-choice
models in the two elections. Overall, based on these three elections, it would
4
For example, in the 2014 BEPS, 73 percent of people who identified as petistas (PT supporters)
in June 2014 voted for Dilma Rousseff in the October 2014 first-round election, while just
39 percent of those who did not identify as petistas in June voted for her in October.
5
The analysis in this paragraph results from three multinomial logistic regression models regres-
sing self-reported first round vote choice on reported clergy endorsements, controlling for
religious affiliation. In 2008, due to perfect multicollinearity, four individuals are excluded who
said that both they and their pastors had supported the minor candidate Tarcísio Delgado.
The models from BEPS 2010 and 2014 are based on the post-election waves, and include
a control for the lagged vote choice. The lagged dependent variables improve causal inference
(Morgan and Winship 2007).
Church Influence on Voting Behavior 123
appear that perhaps clergy have some influence on congregants, but that
Brazilians were certainly not lemmings willing to follow their pastors off cliffs.
Nonetheless, there are reasons to suspect that these observational results
fail fully to capture clergy influence. First, these particular campaigns might
not be the most appropriate ones for detecting clergy influence. The leading
candidates in these three races were from the Workers’ Party (PT) and the
Party of Brazilian Social Democracy (PSDB), two parties without strong
religious linkages. Clergy endorsements may also matter more in legislative,
rather than executive, races. When evangelical and Pentecostal churches get
involved in campaigning, they often do so in support of in-group candidates in
state or federal legislative races.
Second and more importantly, the use of congregant reports presents
potential problems. Congregants may not be particularly good at identifying
whom their clergy support. Not only do messages decay; more importantly,
clergy endorsements may often reach congregants only secondhand, filtered
through church networks in ways that obscure the original source. The diffuse
effect of candidate endorsements may explain the results from Juiz de Fora in
2008. In that highly mobilized election, even the major evangelical clergy
association officially took a stance. Evangelicals who failed to report the
treatment were certainly, in many cases, nonetheless effectively treated,
whether directly by their pastors or indirectly by their fellow congregants in
the pews, in Bible study, or standing at the bus stop the morning following
a rousing worship service.
Third, in the context of actual election campaigns, many persuasive forces
operate simultaneously within congregations. Direct messages from clergy
either endorsing or opposing candidates appear in an environment in which
congregants are likewise talking about the election. Moreover, clergy and
congregants also transmit persuasive arguments about policy issues, and
factual information about candidate positions on those issues. It may be
difficult to disentangle the effects of covarying factors such as church
attendance, exposure to clergy endorsements, information transmittal,
religious group identity, and general conservatism.
A couple of examples illustrate the complexity of electoral mobilization
within congregations. Take, for instance, the 2010 election discussed in the
introduction of this book. As Catholics in the pews shared DVDs in which
Dilma Rousseff appeared to endorse abortion, arguably the most important
aspect of the subterranean congregational campaigns involved information
transmittal. Nonetheless, the information transmitted carried an implicitly
persuasive, partisan message. Or take another example from the Good News
Baptist Church, in which focus-group participants discussed their anger at
policies that they perceived were preventing them from proselytizing in the
course of their public and charitable activities [FG4]. One participant argued
that “the people of God need to open their eyes and recognize” the importance
of getting involved in politics in order to maintain “freedom of expression in
124 How Congregants Respond
.6
All Respondents
.6
.4
.5
.3
.4
.6
.3
.5
Control Gay Rights
.4
Activism
.3
Evangelical Activism &
Clergy Rejection Clergy Rejection
figure 7.3 Experimental impact of clergy campaigning and candidate issue stances on citizens
126 How Congregants Respond
6
In an analysis of yet another experiment fielded in the same study, I found that only evangelicals
low in secular norms support in-group candidates, while non-evangelicals high in secular norms
oppose evangelical candidates. However, analysis of the 2014 BEPS indicates that secular norms
did not condition responses to reported clergy endorsements.
Church Influence on Voting Behavior 127
.4
.6
Candidate Support (0–1 Scale)
.5
.35
.4
.3
.3
.2
.0 .2 .4 .6 .8 1
.25
Secular Norms
Control Evan- Cath- Active
gelical olic in Control Religious Treatments
Endorse- Endorse- Church
ment ment
summary
This chapter considered how clergy influence congregants’ electoral behavior.
As establishing social influence can be challenging, several forms of evidence
were brought to bear, including nationally representative surveys, survey
experimental data, and quantitative and qualitative data from congregational
studies. Across these studies, I conclude that clergy shape citizens’ political
behavior – turnout, civic norms, and vote choice. However, influence is far
from automatic. Once again, we find evidence that clergy influence is
asymmetric, affecting religious conservatives more than others. Evangelicals
may be more likely to take cues from clergy than Catholics, while members of
all religious groups who more strongly hold secular norms are less likely to
respond to clergy influence. Finally, doctrinally conservative congregations are
much more likely to coalesce in their candidate choices.
7
Congregants’ doctrinal conservatism is measured based on their levels of biblical literalism and
the extent to which they say they fear God’s wrath.
Church Influence on Voting Behavior 129
What is the upshot of this and the previous chapter for Brazil’s culture wars?
At the citizen level, religious politics in Brazil may be less polarized than
sometimes thought. Evangelicals, Catholics, and the nonreligious largely agree
on most major issues, with the important exceptions of church–state relations,
homosexuality, same-sex marriage, abortion, and perhaps gender roles. And
importantly, no party has developed strong voter loyalty within any particular
religious group. While religious groups often vote differently from each other,
no single pattern of religious cleavages has emerged in recent races. Sometimes
religious affiliation seems to affect vote choice; at other times clergy
mobilization appears to matter; and at still other times religion appears to be
trumped by other forces.
Still, we also find many signs of religious polarization. Though lay
evangelicals and Catholics feel similarly about many issues – ranging from anti-
poverty policy, to racism, to environmental protection – evangelicals are
becoming both increasingly conservative and increasingly distant from other
citizens on issues related to sexual and family traditionalism. Furthermore,
evangelical congregations exhibit a greater tendency toward homogeneity in
views on these issues. Evangelicals’ increasing conservatism may be one of the
forces driving a small, but potentially growing, religious gap in partisanship.
Furthermore, the growing conservatism of evangelicals is readily mobilized
into electoral politics. It bears repeating that the group now trending in
a conservative direction is precisely the one that has developed the most active
repertoires of political engagement in Brazil’s post-democratic period.
The results presented in this chapter indicate that religious groups, especially
evangelical ones, can and do influence “their” voters. Styles of campaigning that
clergy may have once deployed to promote in-group interests are readily
martialed in support of increasingly conservative candidates and issues.
Nonetheless, secular norms still matter. Brazilian citizens’ widespread
commitment to the separation of religion and politics constrains clergy speech
(Chapter 5), restricting the extent of politicking within churches. Moreover,
congregants with more secular views of society are more likely to resist
congregational influence on culture-war issues, potentially limiting the
ideological polarization of the campaigns that do occur within religious
communities (Chapter 6). Finally, secular norms lead some Brazilians – both
evangelical and Catholic – to resist the campaign-related messages they receive
within church walls.
Thus, we have a tension between growing ideological polarization and
religious engagement in politics on the one hand, and secular norms that
restrain the ferocity of such contests on the other. What is the end result for
Brazilian democracy? We turn to this question in the next two chapters.
8
130
Church Influence on Citizen Support for Democracy 131
five scenes
1 At the Good News Baptist Church, as voting returns rolled in on the Sunday
night of the 2014 first-round election, Pastor Willian sought to lift the spirits of
his congregants. Many people in attendance that night were disappointed that
Marina Silva, the presidential candidate most congregants had supported, had
narrowly missed going on to the second round, as the candidate Aécio Neves
had suddenly surged in the final two days of the race. Like a motivational
speaker, he called on congregants to shout “I am a winner!” several times,
each time getting louder. Even though Marina had narrowly missed second
place, he reminded us, the church had had some electoral victories at the level of
state and federal deputies and senator. Most importantly, he said, “you are a
candidate for the kingdom of God” [CO31].
2. In June 2017, I sat with Pastor Eduardo, head of the small storefront Hope
Church of the Nazarene, in his kitchen, eating pão de queijo (cheese bread) and
catching up on the changes in his church that had occurred since I last visited
(see Chapter 7). In general, Pastor Eduardo was not very interested in recent
political events, telling me: “What I want is to gather information about the
reign of God and the reign of Satan.” He did think politics sometimes affected
the reign of God. He told me that noise ordinances associated with environ-
mental laws inhibited his and other churches’ abilities to proselytize. His con-
gregation had had to move twice in the two years and eight months since I last
visited, in part due to complaints from the neighbors about noise volumes. He
recounted a story of a different church, where police had removed the sound
equipment. “Obviously you need to make sure people are following the rules,”
he said, “but the police are abusing the law . . . According to the Bible, according
132 How Congregants Respond
to the Word,” he continued, “Satan is the enemy. He always tries to inhibit the
reign of God through laws. Satan’s laws inhibit the reign of God on earth”
[CO43].
3 In June 2017, I also returned to the Disciples of Love Catholic prayer
community (see Chapter 4). Since I had last visited, the community had
grown from a humble one-room apartment into a three-story building, with a
large second-floor worship space and a third-floor chapel. Sitting on a crisp
leather couch in the newly constructed, high-ceilinged front entry room, I talked
with the eight assembled members about the community’s plans for
Charismatic Catholic evangelism throughout Brazil. As the conversation turned
to the state of the Brazilian nation, José Luiz, the group’s founder and leader,
situated recent events in apocalyptic terms. “This is not the end,” he explained,
“this is the beginning of the end . . . Democracy is already ending. They’re
removing the authority of the Constitution, but they continue to use the name
of democracy.” He agreed that the Brazilian people should fight to keep democ-
racy, but predicted that they would instead keep reelecting the same crooks as
always. Talking about the current president, he explained: “Temer is a
Satanist.” A bit taken aback, I asked, “How’s that?” José Luiz replied, “He
serves evil. Satan came to kill, steal, destroy. Temer has the attitude of Satan . . .
The people of God die in misery” [CO47; see also CO35].
4 On October 8, 2014, at 7:30 p.m., the Citizenship Committee convened at the
Catholic Cathedral downtown for what the chair formally announced was its
167th monthly meeting. Though the group is ostensibly ecumenical, all twelve
people in attendance were Catholic. The meeting began with the chair taking
attendance, noting several excused absences, and reading aloud the minutes from
the previous meeting. Attendees then reported on city-council activities over the
previous month – though one elderly member had entirely forgotten about the
city-council meeting to which she was assigned. A written report on that month’s
city-council activities would later be printed in the public newsletter the commit-
tee produces from members’ personal funds. The group also briefly discussed the
activities of the municipal Women’s Council, on which one of the members sat.
As the members discussed city-council business, they helped each other under-
stand the local legislative process and pending legislation.
5 Pastor Osésa Rodrigues, the president of the Liberal Christian Party, wrapped
up his hour-long talk at the Juiz de Fora Council of Pastors by calling on pastors
to take action (see Chapter 4). They should not just encourage congregants to
vote for evangelical candidates. Same-sex marriage was not a done deal, he said –
although the Supreme Federal Tribunal had legalized it, Congress could still
reverse the court’s decision legislatively. He exhorted us to work with the newly
formed, ecumenical Christian Alliance in Defense of the Family, which was
seeking to develop a popular initiative bill defining marriage as between a man
and a woman. A direct-democracy provision in Brazil’s 1988 Constitution
Church Influence on Citizen Support for Democracy 133
requires the legislature to consider any popular initiative that gathers 1.3 million
signatures. Pastor Osésa distributed pens printed with the new movement’s web
address, and called on each pastor to collect signatures within his or her con-
gregation. The talk concluded with his calling several members of the audience up
to the front to lay their hands on him to help him pray for the Christian family.
legitimate democracy
This section examines Brazilians’ perceptions of the legitimacy of democracy
and of the current political system. Have clergy and congregations exacerbated
or ameliorated the crisis of Brazilian democracy discussed in Chapter 1?
There are reasons to be hopeful that religious leaders can bolster congregant
support for democracy and the political system. As we saw in Chapter 4, clergy
express much higher levels of support for democracy than do citizens in general
in surveys. Even in 2017 in the midst of political crisis, clergy made clear their
conviction that Brazilian democracy should be defended. These attitudes could
lend legitimacy to a state that is losing its other sources of legitimacy. As highly
trusted, well-known civil-society elites, clergy could transmit their democratic
commitments to their flocks. Pastor Willian’s preaching at the Good News
Baptist Church, discussed in the first scene in the second section of this
chapter, provides an example of how clergy might help their congregants deal
with the inevitable disappointments of democracy.
There is also cause for concern that clergy could sometimes – perhaps
inadvertently – undermine congregants’ trust in political institutions. As we
also saw in Chapter 4, many evangelical and Pentecostal clergy perceive the
Brazilian political system as disadvantaging their own religious group. When
clergy preach about injustices against their religious in-group, congregants
might conclude that Brazilian democracy and the political system do not
work. Thus, as in the second and third scenes, citizens may extrapolate from
their personal dissatisfaction to the political system more broadly.
Brazilians’ levels of support for democracy and the political system fell
dramatically between 2012 and 2017 (see Chapter 1). What was the role of
religion? Strikingly, religious affiliation is not correlated with support for
democracy in the AmericasBarometer. That is, there are no consistent, statistically
significant gaps between religious groups in levels of support for democracy; nor
does church attendance predict support for democracy in any group.
However, religion is significantly correlated with citizens’ views on the
legitimacy of the political system. In every year of the AmericasBarometer,
legitimacy was higher among evangelicals and Catholics than among those
without a religion and in other religious groups.1 Moreover, in 2017 a
statistically significant gap opened between evangelicals and Catholics, such
1
These differences are robust to controls for education, wealth, gender, and size of place of
residence.
134 How Congregants Respond
Strongly
Agree
.75
Level of Support
.5
.25
Strongly
Disagree
Support for Democracy Legitimacy of Political System
figure 8.1 Attitudes toward the political system within eight congregations
that evangelicals were slightly more supportive of the legitimacy of the political
system; this difference is robust to controls for partisanship, political interest,
and demographics. Further, there is some evidence these religious differences
could be due to forces operating within congregations. In 2017 and across all
years, church attendance boosted levels of system support among both
Catholics and evangelicals, though not among those affiliated with other
religions.
To glimpse inside congregations, I turn once again to the 2014 Churches
North and South congregational study. Figure 8.1 examines variation in
legitimacy of democracy and the political system by congregation. Levels of
democratic legitimacy were fairly similar across congregations, though one
evangelical congregation reported significantly lower democratic legitimacy
than several others. The average legitimacy of the political system varied more
from congregation to congregation, with the two non-Pentecostal evangelical
sites a little lower than the other congregations.
What roles do clergy play? In Chapter 4, we found clergy who believe the
state is biased against their group’s interests tend to perceive the state as a whole
as less legitimate. These clergy attitudes may spill over into congregations, as
clergy talk about grievances with their congregations. Thus, I expect that when
clergy have group-related grievances, their congregants may become less
supportive of democracy and the political system; by contrast, clergy who
Church Influence on Citizen Support for Democracy 135
Evangelical
Educational Level
Income
Woman
Age Group
–.5 –.25 0 .25 .5 –.5 –.25 0 .25 .5
figure 8.2 Determinants of attitudes toward the state and democratic regime
ties to Satanism (as well as Masonry) (Agência O Globo 2016). Yet literal
demonization is not limited to perceptions of Temer; it is also evident in the
way many evangelicals discuss the role of the LGBT movement, and the way
some Catholics discuss abortion. Believing that people whose politics one
dislikes are deliberately serving evil and destruction presumably reduces
tolerance of those out-groups’ civil liberties. There is no need to compromise
with the Devil, after all.
Brazil’s 2014 presidential campaign often involved heated, polarized
rhetoric. Toward the end of that campaign, I began to wonder if the intensity
of the election campaign had damaged trust between members of different
religious and political groups. In the post-election wave of the 2014 Brazilian
Electoral Panel Study (BEPS), we included questions on respondents’ attitudes
toward five groups: petistas (supporters of the Workers’ Party), pesedebistas
(supporters of the Party of Brazilian Social Democracy, or PSDB), and
Catholics, evangelicals, and “atheists, or people who don’t believe in God.”
The survey asked the extent to which respondents believed the actions of each
group are “very wrong” or “very right” on a 0–10 scale. Unfortunately, we do
not have data on these attitudes from early in the campaign, so we cannot track
how they changed over time. Still, the results indicate the extent of affective
polarization post-election.
Figure 8.3 shows the average rating of all five groups, by the respondent’s
religious affiliation. Responses have been recoded to run from 0 to 1, so the line
at 0.5 denotes a neutral rating. All four religious groups reported either neutral
or positive attitudes toward both Catholics and evangelicals. Catholics and
evangelicals, in fact, approved of each other fairly highly: Catholics rated
evangelicals at .69, and evangelicals rated Catholics .64. Apart from
Catholics’ and evangelicals’ ratings of their own groups (.83 for Catholics and
.87 for evangelicals), these were among the highest group ratings in the study.
Hence, there is very little evidence of social polarization between Catholics and
evangelicals. Furthermore, gaps between evangelicals and Catholics may be
closing over time. Comparing the ratings from 2014 with survey ratings of
evangelicals in 2002 and 2006, it appears that Catholics are gradually becoming
warmer toward evangelicals.2
Catholics’ and evangelicals’ mutual approval contrasts sharply with
partisans’ attitudes toward their party rivals. Returning to the 2014 data,
petistas scored their own political party at .84, and the PSDB at .41 – a gap of
nearly half the scale. Meanwhile, pesedebistas scored their own party as .75,
and the Workers’ Party (PT) at .43. Reassuringly, though, the partisan
polarization did not extend to religious groups. For instance, religious groups
basically coincided in their ratings of political parties. Every religious group
2
Data from 2002 and 2006 are from the Two City Study. Respondents rated only evangelicals in
those two years. Because the Two City Study was limited in geographic scope, we cannot draw
definite conclusions about change.
138 How Congregants Respond
Catholic Respondents
Evangelical Respondents
Nonreligious Respondents
Other Religion
0 .25 .5 .75 1
Levels of Approval of:
Atheists Evangelicals Catholics
PSDB PT
rated the PT more highly than the PSDB, though the PT–PSDB gap in ratings
was largest among the nonreligious and smallest among evangelicals. Likewise,
partisans were not highly polarized in their ratings of religious groups. Petistas
rated evangelicals a .73 and Catholics a .78, while pesedebistas rated
evangelicals a .78 and Catholics a .67. Thus, there is minor evidence of a PT-
Catholic versus PSDB-evangelical alliance in terms of in-group/out-group
attitudes, but the links are weak.
The absence of any serious Catholic–evangelical social tension is good news
for Brazilian democracy. Two aspects of Brazil’s religious demographics and
culture wars help to explain this felicitous finding. First is the lack of a religious
group–party alliance, which has prevented the affective polarization
characterizing the culture wars in the United States. That is, partisan divisions
do not consistently cleave religious groups. Second, the frequency of religious
conversion means that nearly all Brazilians have both evangelical and Catholic
friends and family members. Mixed-religion social networks have likely
reduced tension, just as scholars have found in the US case (Putnam and
Campbell 2012).
Nonetheless, we do find a different, and troublesome, form of affective
polarization. All groups fairly strongly disapproved of atheists. Atheists
received an average score of .09 from evangelicals, .14 from Catholics, and
Church Influence on Citizen Support for Democracy 139
.24 from adherents to other religions. Strikingly, even the nonreligious only
gave atheists a score of .30 – likely because the great majority of the nonreligious
say they believe in God. In a country where most people normatively value
religion, being nonreligious can be hard. An atheist in a focus group described
the social obstacles she faces: “In the case of people who already know me when
I mention that I’m an atheist, I see that they’re conflicted between the image they
have of atheists as something from the Devil, as people who are going to hell . . .
and the person they know me to be” [FG7].
Do forces within congregations foster mutual dislike between various
groups? Though the congregational study did not measure intergroup
attitudes, further data from the 2014 BEPS illuminate the sources of tensions
between groups. Based on the data presented in Figure 8.3, I calculate affective
polarization: the difference between the scores given to the highest- and the
lowest-rated groups (Iyengar and Westwood 2015). The measure ranges from 0
to 1, with higher values indicating a larger gap in group ratings. One score
corresponds to affective polarization between religious groups, and a second to
political parties.
In a multivariate analysis, religious variables affected partisan affective
polarization, and vice versa. Respondents who said their clergy supported a
candidate perceived bigger differences between the PT and the PSDB.
Meanwhile, people who sympathized with the PT were more affectively
polarized in both religious and partisan preferences. Using structural equation
modeling, I attempt to tease apart which form of polarization has the greater
influence on the other. Though the results must be taken with a bit of caution, it
appears that partisan polarization fuels religious polarization, but that religious
polarization has no impact on partisan polarization.
A critical aspect of democracy is that partisan contestants must recognize the
civil rights of groups they dislike or – perhaps with certain exceptions – perceive
as dangerous. There are both theological and social reasons to expect gaps
between Catholics and evangelicals in political tolerance. First, dualistic or
Manichaean religious doctrine identifying individuals, groups, or behaviors as
evil may encourage some citizens to deny equal participation to out-groups.
Second, recall that evangelical and Pentecostal clergy expressed substantially
lower levels of internal (intra-congregational) and external (extra-
congregational) tolerance than did Catholic clergy in the clergy study
(Chapter 4). If evangelicals take cues from clergy, we might expect the
religious gap in political tolerance to be greatest among those who attend
worship service more frequently.
To assess tolerance, we turn to the LAPOP AmericasBarometer, which has
regularly asked respondents whether homosexual candidates should be allowed
to run for office. This question is immediately relevant for real-world politics,
since religious groups in Brazil have often mobilized precisely to oppose gay
politicians – as in, for instance, the 2008 mayoral race in Juiz de Fora. Figure 8.4
examines the relationships among religious affiliation, congregational
.7
Tolerance of the Rights of Gays
.7
to Run for Office (0–1 scale)
.6
.6
.5
.5
.4
Never Once/ Monthly Weekly >
Twice Weekly
.4
a Year
Catholic Evan- None Other Frequency of Church Attendance
gelical/ Religion
Pentecostal Other Religion No Religion
Religious Affiliation
Catholic Evangelical/
Pentecostal
figure 8.4 Religion, church attendance, and political tolerance
Church Influence on Citizen Support for Democracy 141
attendance, and tolerance for gays running for office. Consistent with prior
research in other countries, the nonreligious are the most tolerant, and there are
sizable differences among various religious groups (Beatty and Walter 1984).
Catholics are significantly less tolerant than are the nonreligious and those
within “other” religions, while evangelicals are less tolerant than Catholics.
The right pane of the figure examines how church attendance affects this form
of political tolerance. Religious attendance has a small, statistically significant
impact on political tolerance among Catholics. However, church attendance
more substantially reduces tolerance among evangelicals and “others.” Among
those who never attend church, the nonreligious and those in other religious
groups have nearly identical levels of tolerance, as do evangelicals and
Catholics. Yet levels of tolerance are substantially lower among evangelicals
and adherents of other religions who attend church regularly.
Thus, congregations have mixed impacts on Brazilians’ social and political
attitudes toward their fellow citizens. There is positive news. Although there is
some evidence of tension between religious groups in the highly charged
political climate in 2014, evangelicals and Catholics still had fairly positive
attitudes toward each other. Yet there is also reason to worry. Partisanship – in
particular, petismo – hurts intergroup trust, and clergy who get involved in
election campaigns contribute to partisan polarization. Moreover, acceptance
of atheists is extremely low. Finally, religious citizens are much less willing than
the nonreligious to extend civil rights to groups that they dislike, and attending
church erodes tolerance further, most prominently among evangelicals.
participatory democracy
Citizens’ ideas and behaviors support or undermine democracy on an ongoing
basis, this chapter argues; electoral participation alone is insufficient. The
previous sections investigated citizens’ attitudes toward the democratic regime
and each other. This section turns back to citizen behaviors, focusing on
participation outside of elections and campaigns. Eminent political theorist
Robert Dahl proposed that “a key characteristic of a democracy is the
continuing responsiveness of government to the preferences of its citizens,
considered as political equals” (1971, 1). Such “continuing responsiveness”
would require “institutions for making government policies depend on votes
and other expressions of preferences” (3).3 Implicit in these propositions is the
requirement that citizens regularly communicate their political desires to each
other or to government, beyond the limits of any electoral cycle.
Do congregations and clergy encourage political participation beyond
campaigns? In the previous chapter, we saw that religious leaders effectively
promote turnout and conscientious voting. Now we turn to several other forms
3
This is a tall order; Dahl recognized that the actually existing more-or-less-democratic regimes,
which he preferred to call “polyarchies,” had never fully fulfilled it.
142 How Congregants Respond
Political Protest
Since 2013, contentious forms of politics such as protest have become
increasingly common in Brazil. Like the other forms of participation, protest
constitutes an important way for citizens to express policy views. Protests, in
fact, may be distinctively effective in bringing new issues to the policy agenda
(Moseley 2015). Nonetheless, protest can have an ambivalent relationship with
democracy. Not only does protest often express grievances against the political
system, but in its extreme form it can destabilize elected governments or even
regimes. Thus, protests uniquely reflect the democratic troubles Brazil has faced
since 2013.
Do religious groups boost or discourage protest participation? Religious
communities could serve as a ready site for mobilizing protests. Catholics and
evangelicals heading to the streets might recruit a few church friends to go with
them, just as someone headed to a shift at the congregation’s charity shop might
also ask a friend from church for help. Asked what he would tell a parishioner
considering going to a protest, Father Miguel from the Santa Fé parish responded,
“Go! And take someone with you!” [CO36]. Though clergy largely ignored the
topic of protest when interviewed in 2014, three years later a number of clergy
explicitly acknowledged protests as legitimate. Pastor Eric from the Vila Bela
Methodist Church opined that citizens should each take part in the activism
corresponding to their social groups: teachers should take part in protests by
teachers’ groups, and union workers in workers’ strikes [CO37].
Nonetheless, evangelicalism could also discourage protest. Even in 2017,
most evangelical clergy expressed discomfort with protests. As Pastor Djalma
explained, “People don’t understand that when you’re a Christian, you don’t
have time for those things. We’re too busy building the kingdom of God”
[CO40]. A 2014 focus group for evangelical young adults indicates how
patterns of interaction within congregations and beyond church walls can
discourage protesting. A college student reported that she knew a lot of
evangelicals at the university who had gone out to the streets, but that many
other young people at church urged each other not to participate. Another
worried that if she participated in protests, she might “damage biblical
principles.” And a third explained that she had gone out to the streets, but
had been turned off by foul language and indecent behavior [FG3]. Thus, both
social influence in congregations and standards of behavior socialized in those
settings discouraged youth in this congregation from protest participation.
Figure 8.5 presents levels of protest participation, by religious group, across
the six waves of the AmericasBarometer. Across almost all of this period, the
nonreligious and those of “other” religious affiliations were most likely to take
part in protest. Nonetheless, we find intriguing religious dynamics. Between
2010 and 2014, religious gaps grew, as protest participation rose more quickly
among “nones” and “others” than among Catholics or evangelicals. However,
between 2014 and 2017, religious gaps closed, as protest participation grew
144 How Congregants Respond
15
Protest in Past Year (Percent)
10
5
0
summary
How do clergy and citizens’ interactions around politics shape Brazil’s
prospects for deliberative, liberal, legitimate, or participatory democracy?
This chapter shows that churches have multivalent impacts on Brazilian
democracy. First, congregational life shapes citizens’ trust in the basic
Church Influence on Citizen Support for Democracy 145
institutions of the nation state. Clergy support for democracy can critically
bolster the system’s legitimacy, in a period when Brazilian citizens have
become increasingly disenchanted with their current regime. Nonetheless,
when clergy believe the political system is biased against their group, they
undermine support for democracy and the political regime among citizens.
Second, congregational life affects citizens’ attitudes toward out-groups. The
good news is that, despite the “clergy skirmishes” that often drive Brazil’s
culture wars, Catholics and evangelicals have fairly positive attitudes toward
each other. Still, there is some evidence that religious politicking fuels partisan
tensions, and that partisan tensions in turn fuel religious ones. While the
connection between affective polarization on religious and partisan lines was
weak in 2014, repeated battles across the same fault lines in subsequent years
could exacerbate social tensions. In addition, dualistic religious doctrines more
generally affect attitudes toward atheists and gays. Finally, both evangelical
religious affiliation and church attendance decrease Brazilians’ willingness to
extend civil liberties to groups they dislike.
Third, congregations boost participation in the many venues through which
citizens seek to shape public policy beyond elections. Evangelicals and Catholics
who attend religious services more frequently are more likely to contact their
representatives, to sign petitions, and to participate in local level efforts to
improve the community. Many forces within religious institutions foster
participation. Not only do clergy often explicitly encourage participation, but
congregational social networks provide the human resources necessary for
effective mobilization. Moreover, civil-society activities are often sponsored
within congregations, and provide a gateway to further mobilization. Protest
participation is a partial exception, however. While Catholic clergy often
encourage participation in contentious politics, evangelical clergy and
congregants tend to discourage it.
In Parts II and III of this book, we have seen how Brazil’s culture wars have
developed among clergy and citizens. Clergy, motivated by a combination of
ideas and organizational interests, polarize around a set of issues related to the
family, gender, sexuality, and the rights and responsibilities of churches; and
they sometimes get involved in partisan electoral politics. Citizens adopt some
of their attitudes on the same set of issues. Citizens take cues for electoral
behavior from their clergy, and more doctrinally conservative churches tend
to coalesce more strongly in their vote choices. Moreover, as we saw in this
chapter, religious polarization is not just a matter of abstract policy attitudes.
Rather, affective polarization extends to people’s attitudes toward their fellow
citizens and the political system. Nonetheless, secular norms lead citizens
partially to resist clergy influence and political activism, and they boost the
legitimacy of democracy and the political system. If ideological, electoral, and
social conflict are the symptoms of the Brazilian culture wars, secular norms –
even, or especially, the secular norms of highly religious citizens – are the
potential cure.
part iv
REPRESENTATION
9
149
150 Representation
officials also appear to represent evangelicals’ desire for greater defense of in-
group religious prerogatives. However, electing in-group representatives
imposes policy costs on evangelical citizens as well. Outside of the two
aforementioned dimensions, evangelical legislators are not very good
representatives of their constituencies’ policy, ideological, or partisan
positions. In particular, elected politicians are substantially more
conservative, and more likely to be aligned with rightist parties, than are the
citizens who tend to vote for them.
What drives these partial representational failures? I argue that the answer
lies in large part in the ways religious communities build personal relations with
the politicians they tend to support for office. Religiously based politicians bring
new blood into office; they are much less likely to rely on oligarchic family ties to
get there, and they are more strongly tied to what they perceive as their
grassroots base. Moreover, they rely to a much lower degree on large
campaign donations (Netto 2016). Instead, personal ties appear to drive
evangelical politicians’ campaigns. Religious communities are one of the most
important places citizens come into contact with politicians, and the close
networks fostered in congregations enable closer relations between politicians
and their voters.
The personalization of evangelical campaigns has many positive benefits.
As a result of their personal ties, religious citizens are more likely to contact
politicians to address community or personal problems. Moreover, evangelical
politicians’ lower reliance on dynastic family ties and on large campaign
donations could both democratize the candidate recruitment process and keep
it cleaner. Nonetheless, this personalization also has negative effects. In church,
evangelicals and Pentecostals are more likely to be targeted with clientelistic
offers. Personal relations may also undermine citizens’ attention to politicians’
policy positions, particularly on issues of lower relevance within the religious
community. Furthermore, in the congregation–politician relationship, clergy
often have an informational advantage over citizens. This advantage leads
congregations to center their mobilization on politicians who align more
closely with the (conservative) views of religious leaders than the (relatively
progressive) views of religious followers.
two scenes
1 On April 17, 2016, the Chamber of Deputy’s roll-call vote to impeach
then-president Dilma Rousseff lasted six hours, as each deputy spoke in turn
before casting a vote. The first “yes” vote was cast by an evangelical,
Deputy Washington Reis from the Party of the Brazilian Democratic
Movement (PMDB), who hoped that God would “pour out blessings on
this nation” (Chagas 2016b). A little while later, evangelical Deputy
Eduardo Bolsonaro, wrapped in the flag of his state of São Paulo, declared
his vote: “For the people of São Paulo in the streets . . . for the military in
The Representational Triangle 151
1964 [the date of the military coup], today and always, for the police, in the
name of God and the Brazilian family, Yes! And Lula and Dilma in jail!”
(Chagas 2016b). Eduardo Cunha, a member of the Assembly of God and
president of the Chamber of Deputies, likewise voted yes. The diverse
dedications for the “yes” vote, ranging from “for Peace in Jerusalem” to
“for the Masons,” became the stuff of satire, with one humorous newspaper
observing, “After being cited by every pro-impeachment deputy, God will be
investigated by the Public Ministry” – a reference to the fact that many
deputies voting yes were themselves being interviewed by public prosecutors
in connection with the Operation Car Wash scandal (Carrapatoso 2016;
Zorzanelli 2016). With a qualified majority of 2/3 required, the impeach-
ment would have narrowly passed to the Senate if only non-evangelicals had
voted; 289 non-evangelicals voted in favor of impeachment, and 134
against. However, evangelicals gave the “yes” vote a comfortable margin.
Of the Chamber’s eighty-one evangelical members, 93 percent voted in
favor of the impeachment, following the decision of their caucus leadership
as well as the Universal Church of the Kingdom of God (UCKG)-linked
Brazilian Republican Party (PRB) (Chagas 2016a; Galindo 2016; Redação
Pragmatismo 2016). Immediately after the Chamber of Deputies had voted
to impeach then-president Dilma Rousseff, Pastor Marco Feliciano, an
Assembly of God pastor and federal deputy from the state of São Paulo,
posted to Twitter, “VICTORY FOR BRAZIL! BYE-BYE, DARLING!”1
2 In October 2016, UCKG Bishop Marcelo Crivella, nephew of founder
Edir Macedo, was elected mayor of Rio de Janeiro. This victory was widely
described as a sign of the growing political power of evangelicals. While
evangelicals have been increasingly successful in legislative races relying on
open-list proportional representation, winning executive races has been
harder. Yet in an interview, anthropologist Ronaldo Almeida cautioned
against making too much of this triumph. Crivella, he explained,
would now:
have to negotiate with the City Council, where the UCKG does not have a strong
base. Crivella may have an electoral base, but the political base depends on alliances
with other politicians, evangelicals or not . . . [I]t is an exaggeration to say that with
the election of Crivella, the UCKG will rule Rio. Society is pluralistic and the UCKG
is only a part of the entire evangelical sector. Most evangelicals are uncomfortable
with the UCKG . . . That’s why he was constantly working to distance himself from
the brand, while still keeping a discourse connected to the UCKG. It is very subtle.
(Charleaux 2016)
1
“Bye-bye, darling” became a pro-impeachment slogan during the Chamber and Senate votes.
The expression came from a quote from former president Lula in a wire-tapped, recorded phone
call with then-president Dilma.
152 Representation
2
This counts only those whom Cunha reported were actually holding office at that moment, and
who were not on leave (that is, including suplentes and excluding licenciados). The list of official
signatories can be found at: http://www.camara.leg.br/internet/deputado/frenteDetalhe.asp?
id=53658.
3
The leftist evangelicals not registered with the caucus include Chico Alencar (Partido do
Socialismo e Liberade), Fabiano Horta and Rejane Dias (PT), and Humberto de Lucena
(Partido Verde). The leftist evangelicals who are registered with the EPF are Benedita da Silva
(PT) and João Derly (Rede Sustentabilidade).
The Representational Triangle 153
4
Religious affiliation comes primarily from data supplied by Mucinhato and coauthors
(Mucinhato 2014; Simoni Junior, Mucinhato, and Mingardi 2015, 2016). Where those data are
missing, I use Agência DIAP (2010), Gonçalves (2011), and Cunha (M. N. Cunha 2016a).
154 Representation
1
Legislator Support for Policy
.75
.5
.25
0
Catholic Pentecostal/Evangelical
None Other
5
Among all legislators, attitudes toward abortion, same-sex marriage, spending for the poor, and
affirmative action load highly on a single dimension with an eigenvalue of 1.13 (unrotated factor
analysis). Among evangelical legislators alone, attitudes toward abortion, same-sex marriage, and
state economic policy load highly on a single dimension, with an eigenvalue of 1.82. However, the
evangelical-only factor analysis is based on an extremely small number of cases, because the
seventh wave of the BLS is the only wave for which all variables are present.
156 Representation
1
Evangelical Divergence from Catholic
Positions (Standard Deviations)
.5
0
–.5
–1
–1.5
figure 9.2 Differences between Catholics and evangelicals in policy attitudes, for
clergy, citizens, and legislators
significantly more conservative than are Catholic clergy on all five issues.
By contrast, evangelical citizens are slightly to the left of Catholics, though
not statistically significantly so, on economics, race, and the environment.
Evangelical legislators resemble clergy to a greater extent than they resemble
citizens on social spending and race-related policy. Only on the environment do
evangelical legislators appear closer to citizens than to clergy. These patterns
indicate that clergy and candidates together mobilize citizens on the basis of
abortion and same-sex marriage, yet elected officials deviate from their base on
other issues.
least on issues such as economic policy. Third, clergy might hand-pick candidates –
often members of their own religious communities – who they know share their
policy priorities. Fourth, evangelicals who choose to run for office – or at least those
who win elections – may simply be more reliably conservative. Adjudicating among
these four explanations using statistical methods is beyond the scope of what it is
possible to do with the present data. However, knowledge of the Brazilian case
leads to the conclusion that all four mechanisms likely sometimes occur, but the
two selection mechanisms play an outsized role.
In Brazil’s electoral and party systems, clergy have great latitude to select
compatible candidates. Under the rules of open-list proportional
representation, used for legislative elections to the lower chamber (in federal
elections) or only chamber (in local and state elections), parties field slates of
candidates, with candidates unranked within slates. Citizens can vote for either
candidates or parties, but most citizens choose to vote for candidates; candidate
votes are aggregated at the party level to determine the number of seats each
party wins. The candidate-level votes also serve to rank candidates within party
lists, determining which candidates win the seats allocated to each party. Thus,
the spare votes of a candidate who does extremely well – that is, the votes over
what would be necessary for that candidate to win by himself or herself – can
boost the chances of other candidates on his or her party list. Likewise,
moderately popular but unelected candidates help their parties win seats
because their votes are added to the party total. Hence, parties are eager to
field candidates who promise to draw votes. Moreover, in most local, state, and
federal legislative districts, dozens of parties seek to fill candidate slates, yielding
hundreds of candidates in each race. As a result, parties are more than willing to
accept candidates with credible support from clergy. This mechanism may
explain how the UCKG hierarchy or some Assembly of God pastors field
candidates, strategically choosing the parties whose lists they believe give their
candidates the best chances of attaining office (Conrado 2001; Dantas 2011).
In other circumstances, the fourth mechanism may provide a better explanation.
Though recruiting candidates from within the religious group’s own ranks is
common in some denominations, a model of candidate recruitment in which each
congregation selects its own candidate is unsustainable. Few congregations, or even
groups of affiliated congregations, would be able to muster enough voters to get
their own candidates elected, and many would lack a high quality pool of potential
candidates. In the UCKG, the denominational hierarchy solves these problems by
coordinating candidacies across congregations. Often, though, congregations
support evangelical candidates from outside their immediate religious group.
The local politician Noraldino Lúcio Dias Junior from Juiz de Fora is
prototypical. Noraldino was elected to the Minas Gerais state legislature in 2014.
Prior to that he was a member of the city council, and in 2016 he ran a strong but
ultimately unsuccessful campaign for mayor of Juiz de Fora. Noraldino is himself
a member of a Methodist congregation. Though Methodist clergy neither recruit
candidates to run nor openly campaign from the pulpit, Noraldino relies on
158 Representation
1
Priority Given Each Electoral Base
.75
.5
.25
0
summary
How have Brazil’s clergy-driven culture wars shaped representative democracy?
Evangelical legislators dramatically expand the political power of evangelical
clergy and congregations. Not only do evangelical legislators respond to the
6
Unfortunately, I also lack information from the qualitative research on this issue. I have not
observed clientelistic offers in evangelical or Catholic congregations.
The Representational Triangle 161
same church leaders as lay evangelicals, but they also become a receptive ear for
evangelical social movements and legislative activism.
There are many ways in which the cultural struggles stemming from the rapid
entry of religious groups into electoral politics have improved representation,
particularly on the political right, where substantive representation has
historically been poor in Brazil. Legislators elected with the support of
evangelical congregations match their constituents closely in their views on
key issues, especially abortion, same-sex marriage, and policy toward
evangelical churches. Since non-evangelical legislators have tended to be to
the left of non-evangelical citizens on these issues, evangelical legislators
improve congruence between citizens and elites as a whole on same-sex
marriage and abortion (Boas and Smith in press). Moreover, the evangelical
style of campaigning provides many evangelicals with personal ties to the
political world that they would otherwise lack. More generally, churches
provide a venue for citizens of all religious backgrounds to meet politicians.
These political ties enable citizens to contact politicians to express their policy
views and request personal help when needed.
However, evangelical legislators diverge from the interests of evangelical
citizens on the economy and race. The racial conservatism of evangelical
legislators is particularly notable. On these issues, evangelical politicians’
views resemble those of in-group clergy more closely than those of voters.
I argue that this divergence is due to the asymmetric influence of clergy over
politicians. Clergy constitute one kind of electoral broker, but with an unusually
high level of independent power; this autonomy gives clergy unusual leverage in
capturing policy rents. Evangelical legislators’ divergence from their
constituents’ views on these issues may also be facilitated by clientelism.
The evangelical style of campaigning also exacerbates party fragmentation,
I argue. As noted in Chapter 1, this argument illuminates old debates over the
causes of party fragmentation – whether institutional and incentive-based, or
sociological and cleavage-based (Duverger 1972; Sartori 1976). The argument
presented here brings the two sides together by suggesting that sociological
cleavages can sometimes create institutional incentives that foster
fragmentation.
10
162
Conclusion: Mobilizing the People of God 163
1
Taken from Taylor Boas, “Expanding the Public Square: Evangelicals and Electoral Politics in
Latin America,” 2018, 3. Prepared for The Inclusionary Turn in Contemporary Latin America,
editors Diana Kapiszewski, Steven Levitsky, and Deborah Yashar (unpublished).
164 Representation
Like their Brazilian counterparts, Chile’s evangelicals have gained inclusion . . . Yet
inclusion has taken very different forms in the two countries. In Brazil, evangelicals
have sought and achieved influence within the halls of power, whereas in Chile, they have
remained primarily on the sidelines.2
The case of Costa Rica constitutes yet another puzzle. Though evangelicals
and Pentecostals constituted 26 percent of respondents in the 2016
AmericasBarometer, evangelical candidates have historically received only
a small fraction of that vote. In fact, one observer marveled that legislative
candidates from evangelical parties had received 8 percent of the 2016 vote –
admittedly, a dramatic rise from 3 percent in the 1998 elections (Salazar 2017).
It is also worth remembering that until the surprise event of the ICHR ruling,
fewer than one in twenty likely voters in Costa Rica intended to support the
evangelical candidate.
Thus, thinking about the broader Latin American context brings the
arguments driving this book into new relief. In this chapter, we revisit the
book’s central questions. First, what explains religious groups’ entry into
politics and the polarization of politics along religious lines? Second, what are
the consequences for Latin American democracy?
2
Boas, “Expanding the Public Square,” 2–3.
Conclusion: Mobilizing the People of God 165
Though this point may seem obvious, it is worth remembering in the Latin
American context, where evangelical and Pentecostal activism has largely been
a phenomenon of “third-wave” (i.e., post-1980) democracies. If Latin America
enters a new era in which liberal democratic freedoms erode, the recession of
civil liberties could affect religious groups’ approaches to politics.
Beyond the impact of regime type, the electoral and party systems can also
facilitate or hinder groups’ entry into politics. In Brazil, both have been
exceptionally porous, guaranteeing many opportunities for evangelical
engagement in electoral politics. The combination of extreme multipartism
and high-magnitude electoral districts (that is, districts in which many
representatives are elected at once) has led party leaders under pressure to fill
out their slates on ballots to accept new (evangelical) blood eagerly. In addition,
Brazil’s form of open-list proportional representation in high-magnitude
districts enables candidates to target their mobilization efforts on specific
electoral corrals such as congregations (Ames 2001). The significance of these
rules is evidenced by the fact that evangelicals have had a much harder time
getting candidates elected to executive office or the Senate.
Finally, religious groups need resources: perhaps most importantly, human
ones. Demographics are not destiny, but they are correlated with groups’
fortunes. In countries with larger evangelical populations, evangelicals are
more likely to succeed in electoral politics, and successfully to mobilize
resistance to policy changes such as the promotion of LGBT rights (Corrales
2017; Mora Torres 2010). High levels of religious attendance and devotion
among evangelicals and Pentecostals increase the value of these human
resources. Resources can also be ideational and psychological. When
religiously based candidates and clergy campaign, they draw on shared
language, theology, religious symbols, and in-group identities to mobilize in-
group turnout and activism (Albertson 2014; Calfano and Djupe 2009; Chapp
2012; Oro 2003b, 2006).
This discussion highlights at least three important questions that merit future
reflection, theorizing, and investigation. First, why are religious groups so
readily motivated by these particular issues: policy related to family and
sexual traditionalism, and in-group threats? What is it about abortion and
homosexuality that makes them the most common triggers for culture wars
across the globe? One possible answer could be that these issues uniquely
violate what many evangelicals and Pentecostals see as the core of their
religious tradition. In Chapter 1, we saw that moral asceticism and frequent
evangelism are defining characteristics of these religious traditions. Abortion
and homosexuality violate dictates of moral asceticism; encroachment by the
state or other religious groups threatens the dictate to evangelize. Moreover, the
tendency toward a dualistic, good-versus-evil worldview in evangelicalism and
Pentecostalism may make adherents more susceptible to mobilization when
moral asceticism and evangelism are threatened. This is just one possibility;
answers should draw on qualitative and quantitative research in a variety of
Conclusion: Mobilizing the People of God 167
However, there are significant gaps in the importance evangelical and Catholic
clergy place on such issues.
Evangelical and Pentecostal forms of political organizing will tend to
privilege the issue priorities of clergy over those of congregants, as discussed
in Chapters 2 and 9. That is, clergy’s role as brokers – choosing candidates and
mobilizing voters – enables them to capture policy “rents.” As a result, when the
political preferences of clergy and congregants diverge, candidates elected with
congregational support will tend to be closer to the views of clergy. Religious
leaders can control the information that reaches congregants, focusing on issues
where candidates and congregants agree, such as abortion and homosexuality.
In addition, candidates may be more likely to provide clientelistic side-payments
to congregants than to clergy. Finally, the asymmetry in information and
sophistication between clergy and congregants will make clergy more likely to
monitor politicians’ behavior.
Electing religious rightists strengthens the power of the right at the elite level
more generally. Politicians on the religious right form alliances with other
rightist politicians, including those whom Power and Rodrigues-Silveira term
the “clientelistic right,” the “law-and-order right,” and the “economic right”
(2018). The dramatic alignment of evangelicals with other rightist groups in
supporting the impeachment of Dilma Rousseff signals the importance of
evangelicals to the right at the elite level more broadly. At the level of civil
society, the Free Brazil Movement (MBL, or Movimento Brasil Livre), a large,
new libertarian social movement organization, has begun connecting the
various facets of rightism. And in Colombia, one finds various rightisms
instrumentally linked in the campaign against the peace deal.
Once again, questions remain for future research. For instance, why are
clergy more reliably conservative than citizens on so many issues? Tentatively,
it seems likely the answer has two parts. First, clergy have more fully absorbed
and internalized internationally disseminated doctrine constructed within
globalized religious communities that contains ideological assumptions about
“what goes with what.” That is, there is no single, logically necessary story
connecting, for instance, attitudes on abortion and tariffs. (Indeed, a naïve
reading of the issues could even suggest that support for abortion rights and
opposition to tariffs should go together.) However, clergy are more likely than
citizens to grasp and accept ideological programs defining “left” and “right,”
and dictating that people who oppose abortion rights should also oppose tariffs
(e.g., Converse 1964). Second, conservative clergy are likely to be more aware
than citizens that in national politics, religious rightists support right-leaning
causes more generally.
Nonetheless, there are reasons to hope that most churches will, despite their
rightist proclivities, avoid mobilizing behind the antidemocratic far right.
Clergy support for democracy in the abstract will reduce enthusiasm for
overtly authoritarian options. Perhaps even more importantly, religious
groups’ engagement in electoral politics gives clergy who might be tempted by
authoritarian options a stake in the electoral game. Moreover, incentives to
electoral disunity – the very incentives that prevent evangelical denominations
from functioning effectively as a partisan front – may also prevent a unified
evangelical coalition behind any particular politician who promises to rig the
game to in-group advantage. If so, Brazil’s clergy-driven culture wars could
ultimately help to stabilize democracy.
Afterword
176
Afterword 177
references
Alves, José Eustáquio Diniz. 2018. “O voto evangélico garantiu a eleição de Jair
Bolsonaro.” EcoDebate (blog). October 31, 2018. www.ecodebate.com.br/2018/10/
31/o-voto-evangelico-garantiu-a-eleicao-de-jair-bolsonaro-artigo-de-jose-eustaquio-
diniz-alves/.
appendix a
Unless otherwise noted below, all interviewee and church names used in the
book are changed for the purpose of anonymizing the qualitative data.
Quotation marks in the third columns below indicate that a congregation’s
name has been changed.
Focus Groups
178
List of Focus Groups and Church Observations 179
Church Observations
(continued)
Focus-Group Protocol
In the fictitious city of Bela Vista (RJ), in a universe not very far away, the
following things happen . . .
1. In an evangelical church, Pastor Sérgio distributes a pamphlet instructing
church members on the importance of voting.
2. In another evangelical church in the city, Pastor João preaches to church
members, “Politics isn’t for believers.” He’s waiting for the End Times
and says they should concentrate on saving souls.
3. In a Catholic church in the Santo Tomás neighborhood, Father Luiz often
talks about the sin of homosexuality and the Workers’ Party (PT) in the
Catholic formation class, saying that the PT is very misguided in its
legislative activism.
4. In an evangelical church in the same neighborhood, Pastor Eunice is
collecting signatures for a popular initiative law to define marriage as
between a man and a woman.
5. In the Catholic cathedral in the center of the city, Father Flávio has
a private meeting with some deacons he knows well. He tells them that
he supports Dilma and is opposed to the ideology of the Party of Brazilian
Social Democracy (PSDB).
6. In an evangelical church in the city, an evangelical candidate asks Pastor
Ricardo for his support. The pastor tells the candidate that he personally
supports him, but that he can’t campaign for him inside the church.
The candidate stays outside the church after service distributing
materials.
7. At the end of May, a Catholic church in the São Leopoldo neighborhood
organizes a march against crime. At the march, Father José Luiz preaches
to the participants that the mayor isn’t doing enough to combat crime.
8. In a Four-Square Baptist church in the city, Pastor William strongly
supports the presidential candidate Marina, a state deputy, and
a federal deputy. He puts up posters for them outside the church.
181
182 Appendix B
183
References
Abramowitz, Alan. 2010. The Disappearing Center: Engaged Citizens, Polarization, and
American Democracy. New Haven, CT: Yale University Press.
Abril.com. 2010. “Dilma pede direito de resposta a Canção Nova por sermão de padre.”
Abril, October 7.
Agência DIAP. 2010. “Evangélicos crescem no Congresso; PSC tem mais
representantes.” DIAP – Departamento Intersindical de Assessoria Parlamentar.
Agência O Globo. 2016. “Deputado pede que Temer ‘abandone satanismo’ e
‘a maçonaria.’” Epoca Negocios, November 18. http://epocanegocios.globo.com/
Brasil/noticia/2016/11/deputado-pede-que-temer-abandone-satanismo-e-maconaria
.html.
Albertson, Bethany L. 2011. “Religious Appeals and Implicit Attitudes.” Political
Psychology 32 (1): 109–30.
2014. “Dog-Whistle Politics: Multivocal Communication and Religious Appeals.”
Political Behavior, January, 1–24.
Aldrich, John H. 1995. Why Parties? The Origin and Transformation of Political Parties
in America. Chicago: University of Chicago Press.
Alex-Assensoh, Yvette, and A. B. Assensoh. 2001. “Inner-City Contexts, Church
Attendance, and African-American Political Participation.” Journal of Politics
63 (3): 886–901.
Allport, Gordon Willard. 1979. The Nature of Prejudice. New York: Basic Books.
Alonso, Angela, and Ann Mische. 2016. “Changing Repertoires and Partisan Ambivalence in
the New Brazilian Protests.” Bulletin of Latin American Research, March.
Ames, Barry. 2001. The Deadlock of Democracy in Brazil. Ann Arbor, MI: University of
Michigan Press.
Ames, Barry, Alyssa Huberts, Fabiana Machado, Lucio Rennó, David Samuels, Amy
Erica Smith, and Cesar Zucco. 2016. “Brazilian Electoral Panel Study: 2014 Results.”
IDB-TN-915. Washington, DC: Inter-American Development Bank. www.iadb.org/en/
research-and-data/publication-details,3169.html?pub_id=IDB-TN-915.
Ames, Barry, Fabiana Machado, Lucio Rennó, David Samuels, Amy Erica Smith, and
Cesar Zucco. 2013. “Brazilian Electoral Panel Survey.” Technical Note IDB-TN
-508. Washington, DC: Inter-American Development Bank. www.iadb.org/en/
research-and-data/publication-details,3169.html?pub_id=IDB-DB-105.
184
References 185
2014b. The Politics of Evangelical Identity: Local Churches and Partisan Divides in
the United States and Canada. Princeton, NJ: Princeton University Press.
Bean, Lydia, Marco Gonzalez, and Jason Kaufman. 2008. “Why Doesn’t Canada Have
an American-Style Christian Right? A Comparative Framework for Analyzing the
Political Effects of Evangelical Subcultural Identity.” The Canadian Journal of
Sociology / Cahiers Canadiens de Sociologie 33 (4): 899–943.
Beatty, Kathleen Murphy, and Oliver Walter. 1984. “Religious Preference and Practice:
Reevaluating Their Impact on Political Tolerance.” Public Opinion Quarterly 48
(1B): 318–29.
Bebbington, David William. 1989. Evangelicalism in Modern Britain: A History from
the 1730s to the 1980s. London: Unwin Hyman.
Bello, Jason, and Meredith Rolfe. 2014. “Is Influence Mightier than Selection? Forging
Agreement in Political Discussion Networks during a Campaign.” Social Networks,
Special Issue on Political Networks, 36 (January): 134–46.
Ben-Nun Bloom, Pazit. 2013. “The Public’s Compass: Moral Conviction and Political
Attitudes.” American Politics Research 41 (6): 937–64.
Ben-Nun Bloom, Pazit, Gizem Arikan, and Marie Courtemanche. 2015. “Religious
Social Identity, Religious Belief, and Anti-Immigration Sentiment.” American
Political Science Review 109 (02): 203–21.
Berelson, Bernard R., Paul F. Lazarsfeld, and William N. McPhee. 1954. Voting: A Study of
Opinion Formation in a Presidential Campaign. Chicago: University of Chicago Press.
Bineham, Jeffery L. 1988. “A Historical Account of the Hypodermic Model in Mass
Communication.” Communication Monographs 55 (3): 230–46.
Boas, Taylor C. 2014. “Pastor Paulo vs. Doctor Carlos: Professional Titles as Voting
Heuristics in Brazil.” Journal of Politics in Latin America 6 (2): 39–72.
Boas, Taylor C., and Amy Erica Smith. 2015. “Religion and the Latin American Voter.”
In The Latin American Voter, edited by Ryan E. Carlin, Matthew Singer, and
Elizabeth Zechmeister, 99–121. University of Michigan Press.
In press. “Looks Like Me, Thinks Like Me? Descriptive Representation and Opinion
Congruence in Brazil.” Latin American Research Review 54 (3).
Bob, Clifford. 2012. The Global Right Wing and the Clash of World Politics.
Cambridge, UK: Cambridge University Press.
Bohn, Simone. 2004. “Evangélicos no Brasil: Perfil sócio-econômico, afinidades
ideológicas e determinantes do comportamento eleitoral.” Opinião Pública 10
(2): 288–338.
2007. “Contexto político-eleitoral, minorias religiosas e voto em pleitos presidenciais.
(2002–2006).” Opinião Pública 13 (2): 366–87.
Bom, Klaas. 2015. “‘I Feel the Presence of God in My Tears’: On the Theological
Contribution to the Research of Latin American Pentecostalism.” Exchange 44
(2): 177–200.
Bompani, Barbara, and S. Terreni Brown. 2015. “A ‘Religious Revolution’? Print Media,
Sexuality, and Religious Discourse in Uganda.” Journal of Eastern African Studies 9
(1): 110–26.
Booth, John A., and Mitchell A. Seligson. 2009. The Legitimacy Puzzle in Latin America:
Political Support and Democracy in Eight Latin American Nations. Cambridge,
UK: Cambridge University Press.
Borges, Laryssa. 2010. “TV católica Canção Nova exibe sermão contra voto em Dilma.”
Terra.com, October 6.
References 187
Bornschier, Simon. 2010. “The New Cultural Divide and the Two-Dimensional Political
Space in Western Europe.” West European Politics 33 (3): 419–44. https://doi.org/
10.1080/01402381003654387.
Bruneau, Thomas C. 1980. “The Catholic Church and Development in Latin America:
The Role of the Basic Christian Communities.” World Development 8 (7): 535–44.
1982. The Church in Brazil: The Politics of Religion. Austin, TX: University of Texas
Press.
Burdick, John. 1993a. Looking for God in Brazil: The Progressive Catholic Church in
Urban Brazil’s Religious Arena. Berkeley, CA: University of California Press.
1993b. “Struggling against the Devil: Pentecostalism and Social Movements in Urban
Brazil.” In Rethinking Protestantism in Latin America, edited by Virginia Garrard-
Burnett and David Stoll. Philadelphia: Temple University Press.
Calegari, Luiza. 2017. “O que pensam os manifestantes contra e a favor de Judith
Butler.” Exame, November 15. https://exame.abril.com.br/brasil/o-que-pensam
-os-manifestantes-contra-e-a-favor-de-judith-butler/.
Calfano, Brian Robert, and Paul A. Djupe. 2009. “God Talk: Religious Cues and
Electoral Support.” Political Research Quarterly 62 (2): 329–39.
Calfano, Brian Robert, Melissa R. Michelson, and Elizabeth A. Oldmixon. 2017.
A Matter of Discretion: The Politics of Catholic Priests in the United States and
Ireland. London; Lanham, MD: Rowman & Littlefield.
Calfano, Brian Robert, and Elizabeth A. Oldmixon. 2016. “Remembering to Ask the
Boss: Priming and the Dynamics of Priest Reliance on Bishop Cues.” Religions 7
(3): 21.
Calfano, Brian Robert, Elizabeth A. Oldmixon, and Jane Suiter. 2014. “Who and What
Affects the First Estate? An Analysis of Clergy Attitudes on Cultural and Economic
Issues.” Politics 34 (4): 391–404.
Campbell, David E., John C. Green, and Geoffrey C. Layman. 2011. “The Party Faithful:
Partisan Images, Candidate Religion, and the Electoral Impact of Party
Identification.” American Journal of Political Science 55 (1): 42–58.
Campbell, David E., John C. Green, and J. Quin Monson. 2014. Seeking the Promised
Land: Mormons and American Politics. Cambridge, UK: Cambridge University
Press.
Carranza, Brenda. 2006. “Catolicismo midiático.” In As Religiões no Brasil, edited by
Faustino Teixeira and Renata Menezes, 69–88. Petrópolis, Brazil: Editora Vozes.
Carrapatoso, Miguel Santos. 2016. “Impeachment Dilma. As melhores pérolas dos
deputados para justificarem o voto.” Observador, April 18. http://observador.pt/
2016/04/18/impeachment-dilma-as-melhores-perolas-dos-deputados-justificarem
-voto/.
Castro, Mônica Mata Machado de. 2007. “Eleitorado brasileiro: Composição e grau de
participação.” In Sistema Político Brasileiro: Uma introdução, 25–42. Rio de
Janeiro: Fundação Konrad Adenauer.
Chagas, Tiago. 2016a. “‘Partido da universal’ anuncia que votará a favor do impeachment.”
Notícias Gospel (blog). April 14. https://noticias.gospelmais.com.br/partido-universal
-anuncia-votara-favor-impeachment-82222.html.
2016b. “Primeiro voto do impeachment foi de um evangélico.” Notícias Gospel
(blog). April 18. https://noticias.gospelmais.com.br/primeiro-voto-impeachment
-foi-evangelico-82289.html.
188 References
Chapp, Christopher B. 2012. Religious Rhetoric and American Politics: The Endurance
of Civil Religion in Electoral Campaigns. Ithaca, NY: Cornell University Press.
Charleaux, João Paulo. 2016. “Qual a influência das igrejas evangélicas na política
brasileira.” Nexo Jornal, October 28. https://www.nexojornal.com.br/entrevista/
2016/10/28/Qual-a-influ%C3%AAncia-das-igrejas-evang%C3%A9licas-na-pol%
C3%ADtica-brasileira.
Cheney, Kristen. 2012. “Locating Neocolonialism, ‘Tradition,’ and Human Rights in
Uganda’s ‘Gay Death Penalty.’” African Studies Review 55 (2): 77–95.
Chesnut, R. Andrew. 1999. “The Salvation Army or the Army’s Salvation?: Pentecostal
Politics in Amazonian Brazil, 1962–1992.” Luso-Brazilian Review 36 (2): 33–49.
2003a. “A Preferential Option for the Spirit: The Catholic Charismatic Renewal in
Latin America’s New Religious Economy.” Latin American Politics and Society 45
(1): 55–85.
2003b. Competitive Spirits: Latin America’s New Religious Economy: Latin
America’s New Religious Economy. New York: Oxford University Press.
2009. “Charismatic Competitors: Protestant Pentecostals and Catholic Charismatics
in Latin America’s New Religious Marketplace.” In Religion and Society in Latin
America, edited by Lee M. Penyak and Walter J. Petry, 207–23. Maryknoll, NY:
Orbis Books.
Conrado, Flávio Cesar. 2001. “Política e Mídia: A Igreja Universal do Reino de Deus nas
eleições.” Religião & Sociedade 21 (2): 85–111.
Conroy-Krutz, Jeffrey. 2013. “Information and Ethnic Politics in Africa.” British
Journal of Political Science 43 (02): 345–73.
Conroy-Krutz, Jeffrey, Devra C. Moehler, and Rosario Aguilar. 2016. “Partisan Cues
and Vote Choice in New Multiparty Systems.” Comparative Political Studies, 49
(1): 3–35.
Converse, Philip E. 1964. “The Nature of Belief Systems in Mass Publics.” In Ideology
and Discontent, edited by David Apter, 75–169. London: Free Press of Glencoe.
Corrales, Javier. 2017. “Understanding the Uneven Spread of LGBT Rights in Latin
America and the Caribbean, 1999–2013.” Journal of Research in Gender Studies 7
(1): 52–82.
Corten, André. 1999. Pentecostalism in Brazil: Emotion of the Poor and Theological
Romanticism. New York: St. Martin’s Press.
Coutinho, Raquel Zanatta, and André Braz Golgher. 2014. “The Changing Landscape
of Religious Affiliation in Brazil between 1980 and 2010: Age, Period, and Cohort
Perspectives.” Revista Brasileira de Estudos de População 31 (1): 73–98.
Cowie, Sam. 2018. “Violent Deaths of LGBT People in Brazil Hit All-Time High.”
The Guardian, January 22. www.theguardian.com/world/2018/jan/22/brazil-lgbt
-violence-deaths-all-time-high-new-research.
Cozzens, Donald B. 2000. The Changing Face of the Priesthood: A Reflection on the
Priest’s Crisis of Soul. Collegeville, MN: Liturgical Press.
Cunha, Luiz Antônio. 2009. “A educação na concordata Brasil-Vaticano.” Educação &
Sociedade 30 (106): 263–80. https://doi.org/10.1590/S0101-73302009000100013.
Cunha, Magali Nascimento. 2016a. “Composição da bancada evangélica.” Mídia,
Religião E Cultura (blog).
2016b. “Frentes parlamentares de cunho religioso.” Mídia, Religião e Política (blog).
2016c. “Bancada evangélica em processo de ampliação: Bolsonaro batizado em
Israel.” Mídia, Religião e Política (blog). May 17.
References 189
Dahl, Robert. 1971. Polyarchy: Participation and Opposition. New Haven, CT: Yale
University Press.
Dalton, Russell J., and Christian Welzel. 2014. The Civic Culture Transformed: From
Allegiant to Assertive Citizens. Cambridge, UK: Cambridge University Press.
Dancygier, Rafaela M. 2017. Dilemmas of Inclusion: Muslims in European Politics.
Princeton, NJ: Princeton University Press.
Dantas, Bruna Suruagy do Amaral. 2011. “Religião e política: ideologia e ação da
bancada evangélica na Câmara Federal.” Doctoral dissertation, São Paulo:
Pontifícia Universidade Católica de São Paulo.
Demartini, Maria. 2017. “Mackenzie abre núcleo de estudos que contesta teoria da
evolução.” Exame, May 11. http://exame.abril.com.br/ciencia/mackenzie-abre
-nucleo-de-estudos-que-contesta-teoria-da-evolucao/.
Díaz Domínguez, Alejandro. 2006. “¿Influyen los ministros de culto sobre la intención
de voto?” Perfiles Latinoamericanos 28: 33–57.
Djupe, Paul A., Anand E. Sokhey, and Christopher P. Gilbert. 2007. “Present but Not
Accounted For? Gender Differences in Civic Resource Acquisition.” American
Journal of Political Science 51 (4): 906–20.
Djupe, Paul A., and Brian Calfano. 2014. God Talk: Experimenting with the Religious
Causes of Public Opinion. Philadelphia: Temple University Press.
Djupe, Paul A., and Christopher P. Gilbert. 2002. “The Political Voice of Clergy.”
Journal of Politics 64 (2): 596–609.
2003. The Prophetic Pulpit: Clergy, Churches, and Communities in American
Politics. Lanham, MD: Rowman & Littlefield Publishers.
2006. “The Resourceful Believer: Generating Civic Skills in Church.” Journal of
Politics 68 (1): 116–27.
2008. “Politics and Church: Byproduct or Central Mission?” Journal for the Scientific
Study of Religion 47 (1): 45–62.
2009. The Political Influence of Churches. Cambridge, UK; New York: Cambridge
University Press.
Djupe, Paul A., Jacob R. Neiheisel, and Kimberly H. Conger. 2018. “Are the Politics of
the Christian Right Linked to State Rates of the Nonreligious? The Importance of
Salient Controversy.” Political Research Quarterly, 71 (4): 910–22.
Djupe, Paul A., Scott D. McClurg, and Anand E. Sokhey. 2018. “The Political
Consequences of Gender in Social Networks.” British Journal of Political Science
48 (3): 637–58.
Dowd, Robert A. 2015. Christianity, Islam, and Liberal Democracy: Lessons from Sub-
Saharan Africa. Oxford: Oxford University Press.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper
& Row.
Duverger, Maurice. 1972. Party Politics and Pressure Groups: A Comparative
Introduction. New York: Thomas Y. Crowell.
The Economist. 2018. “Latin America’s Human-Rights Court Moves into Touchy
Territory.” The Economist, February 1. https://www.economist.com/news/americas
/21736199-ruling-favour-gay-marriage-will-provoke-resistance-latin-americas-human
-rights-court.
Engeli, Isabelle, Christoffer Green-Pedersen, and Lars Thorup Larsen. 2012. Morality
Politics in Western Europe: Parties, Agendas and Policy Choices. London: Palgrave
Macmillan.
190 References
Freston, Paul. 1993. “Brother Votes for Brother: The New Politics of Protestantism in
Brazil.” In Rethinking Protestantism in Latin America, edited by Virginia Garrard-
Burnett and David Stoll, 66–110. Philadelphia: Temple University Press.
2004. “Evangelical Protestantism and Democratization in Contemporary Latin
America and Asia.” Democratization 11 (4): 21–41.
G1. 2015. “71% dos futuros deputados se dizem Católicos e 16%, Evangélicos.” G1,
January 31. http://g1.globo.com/politica/noticia/2015/01/71-dos-futuros-deputados
-se-dizem-catolicos-e-16-evangelicos.html.
Galindo, Rogerio Waldrigues. 2016. “93% dos deputados da bancada evangélica votaram
pelo impeachment.” Caixa Zero, April 17. www.gazetadopovo.com.br/blogs/caixa
-zero/93-dos-deputados-da-bancada-evangelica-votaram-pelo-impeachment/.
Gandini, Arthur. 2015. “Cerca de 75% do clero é gay, estima ex-sacerdote católico.”
A Igreja Paulistana, May 30. www.igrejapaulistana.com/2015/05/cerca-de-75-do
-clero-e-gay-diz-ex.html.
Gans-Morse, Jordan, Sebastián Mazzuca, and Simeon Nichter. 2014. “Varieties of
Clientelism: Machine Politics during Elections.” American Journal of Political
Science 58 (2): 415–32.
Garrard-Burnett, Virginia. 2009. “‘Like a Mighty Rushing Wind’: The Growth of
Pentecostalism in Contemporary Latin America.” In Religion and Society in Latin
America, edited by Lee M. Penyak and Walter J. Petry, 190–202. Maryknoll,
New York: Orbis Books.
Gaskill, Newton Jeffrey. 2002. “Power from on High: The Political Mobilization of
Brazilian Evangelical Protestantism.” Ph.D. Thesis, Austin, Texas: University of
Texas at Austin. http://repositories.lib.utexas.edu/handle/2152/11613.
Geddes, Barbara. 2003. Paradigms and Sand Castles: Theory Building and Research
Design in Comparative Politics. Ann Arbor, MI: University of Michigan Press.
Gill, Anthony. 1994. “Rendering unto Caesar: Religious Competition and Catholic
Political Strategy in Latin America, 1962–79.” American Journal of Political
Science 38 (2): 403–25.
1995. “The Institutional Limitations of Catholic Progressivism: An Economic
Approach.” International Journal of Social Economics 22 (9/10/11): 135–48.
1998. Rendering unto Caesar: The Catholic Church and the State in Latin America.
Chicago: University of Chicago Press.
Glazier, Rebecca A. 2013. “Divine Direction: How Providential Religious Beliefs Shape
Foreign Policy Attitudes.” Foreign Policy Analysis 9 (2): 127–42.
Gonçalves, Juliana. 2017. “‘Queimem a bruxa!’ Visita de Judith Butler
provoca manifestações nas ruas de São Paulo.” The Intercept (blog). November 7.
https://theintercept.com/2017/11/07/judith-butler-bruxa-manifestacoes-sao-paulo
-ideologia-genero/.
Gonçalves, Rafael Bruno. 2011. “Bancada evangélica?: Uma análise do discurso
parlamentar evangélico durante a 52a Legislatura da Câmara Federal.” Master’s
Thesis, Universidade Federal de Pelotas. http://repositorio.ufpel.edu.br:8080/
handle/123456789/1591.
Goren, Paul, and Christopher Chapp. 2017. “Moral Power: How Public Opinion on
Culture War Issues Shapes Partisan Predispositions and Religious Orientations.”
American Political Science Review 111 (1): 110–28.
Gouldner, Alvin W. 1960. “The Norm of Reciprocity: A Preliminary Statement.”
American Sociological Review 25 (2): 161–78. https://doi.org/10.2307/2092623.
192 References
Huckfeldt, Robert, and John Sprague. 1995. Citizens, Politics, and Social
Communication: Information and Influence in an Election Campaign.
Cambridge, UK: Cambridge University Press.
Hunter, James Davison. 1992. Culture Wars: The Struggle to Control the Family, Art,
Education, Law, and Politics in America. New York: Basic Books.
Huntington, Samuel P. 1991. The Third Wave: Democratization in the Late Twentieth
Century. Norman, OK: University of Oklahoma Press.
Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society. Princeton, NJ:
Princeton University Press.
1997. Modernization and Postmodernization. Princeton, NJ: Princeton University
Press.
Inglehart, Ronald, and Pippa Norris. 2003. Rising Tide: Gender Equality & Cultural
Change around the World. Cambridge, UK: Cambridge University Press.
Ireland, Rowan. 1993. “The Crentes of Campo Alegre and the Religious Construction of
Brazilian Politics.” In Rethinking Protestantism in Latin America, edited by
Virginia Garrard-Burnett and David Stoll, 45–65. Philadelphia: Temple University
Press.
1999. “Popular Religions and the Building of Democracy in Latin America: Saving the
Tocquevillian Parallel.” Journal of Interamerican Studies and World Affairs 41 (4):
vi–136.
Iyengar, Shanto, and Sean J. Westwood. 2015. “Fear and Loathing across Party Lines:
New Evidence on Group Polarization.” American Journal of Political Science 59
(3): 690–707.
Jacob, Cesar Romero, Dora Rodrigues Hees, and Philippe Waniez. 2013. Religião
e Território no Brasil, 1990–2010. Rio de Janeiro: Editora PUC-Rio.
Jacobs, Andrew. 2016. “Brazil Is Confronting an Epidemic of Anti-Gay Violence.”
The New York Times, July 5, sec. Americas. www.nytimes.com/2016/07/06/
world/americas/brazil-anti-gay-violence.html.
Jacoby, William G. 2014. “Is There a Culture War? Conflicting Value Structures in
American Public Opinion.” American Political Science Review 108 (04): 754–71.
Jelen, Ted G. 1992. “Political Christianity: A Contextual Analysis.” American Journal of
Political Science 36 (3): 692–714.
Jenkins, Philip. 2002. The Next Christendom: The Rise of Global Christianity.
New York: Oxford University Press.
Kaoma, Kapya. 2009. Globalizing the Culture Wars: US Conservatives, African
Churches, and Homophobia. Somerville, MA: Political Research Associates.
2014. “The Paradox and Tension of Moral Claims: Evangelical Christianity, the
Politicization and Globalization of Sexual Politics in Sub-Saharan Africa.”
Critical Research on Religion 2 (3): 227–45.
Keck, Margaret E. 1992. The Workers’ Party and Democratization in Brazil. New
Haven, CT: Yale University Press.
King, Gary, Robert O. Keohane, and Sidney Verba. 1994. Designing Social Inquiry:
Scientific Inference in Qualitative Research. Princeton, NJ: Princeton University
Press.
Kitschelt, Herbert. 2000. “Linkages between Citizens and Politicians in Democratic
Polities.” Comparative Political Studies 33 (6–7): 845–79.
194 References
Koleva, Spassena P., Jesse Graham, Ravi Iyer, Peter H. Ditto, and Jonathan Haidt. 2012.
“Tracing the Threads: How Five Moral Concerns (Especially Purity) Help Explain
Culture War Attitudes.” Journal of Research in Personality 46 (2): 184–94.
Kuru, Ahmet T. 2008. “Passive and Assertive Secularism: Historical Conditions,
Ideological Struggles, and State Policies toward Religion.” World Politics 59 (4):
568–94.
2009. Secularism and State Policies toward Religion: The United States, France, and
Turkey. New York: Cambridge University Press.
Laakso, Markku, and Rein Taagepera. 1979. “‘Effective’ Number of Parties: A Measure
with Application to West Europe.” Comparative Political Studies 12: 3–27.
Lankina, Tomila, and Lullit Getachew. 2012. “Mission or Empire, Word or Sword?
The Human Capital Legacy in Postcolonial Democratic Development.” American
Journal of Political Science 56 (2): 465–83.
Lasswell, Harold Dwight. 2011. Politics: Who Gets What, When, How. Whitefish, MT:
Literary Licensing, LLC.
Layman, Geoffrey C., and Thomas M. Carsey. 2002. “Party Polarization and ‘Conflict
Extension’ in the American Electorate.” American Journal of Political Science 46
(4): 786–802.
Layman, Geoffrey C., and John C. Green. 2006. “Wars and Rumours of Wars:
The Contexts of Cultural Conflict in American Political Behaviour.” British
Journal of Political Science 36 (1): 61–89.
Layman, Geoffrey C., Thomas M. Carsey, and Juliana Menasce Horowitz. 2006. “Party
Polarization in American Politics: Characteristics, Causes, and Consequences.”
Annual Review of Political Science 9 (1): 83–110.
Lazarsfeld, Paul F., Bernard R. Berelson, and Hazel Gaudet. 1948. The People’s Choice:
How the Voter Makes Up His Mind in a Presidential Campaign. New York:
Columbia University Press.
Lehmann, David. 1996. Struggle for the Spirit: Religious Transformation and Popular
Culture in Brazil and Latin America. Oxford: Polity Press.
Lemos, Christina, and Douglas Tavolaro. 2007. O Bispo: A História Revelada de Edir
Macedo. São Paulo: Editora Larousse do Brasil.
Levine, Daniel H. 2009. “The Future of Christianity in Latin America.” Journal of Latin
American Studies 41: 121–45.
2012. Politics, Religion & Society in Latin America. Boulder, CO: Lynne Rienner
Publishers.
Levitsky, Steven, and Daniel Ziblatt. 2018. How Democracies Die. New York: Crown/
Archetype.
Lewis, Andrew R. 2017. The Rights Turn in Conservative Christian Politics: How
Abortion Transformed the Culture Wars. Cambridge, UK: Cambridge University
Press.
Lisboa, Fábio Aguiar. 2010. “Qual o papel da religião nas últimas eleições?” O Jornal
Batista, November 6. www.adiberj.org/portal/2010/11/06/qual-o-papel-da-religiao
-nas-ultimas-eleicoes/.
Løland, Ole Jakob. 2015. “The Position of the Biblical Canon in Brazil: From Catholic
Rediscovery to Neo-Pentecostal Marginalisation.” Studies in World Christianity 21
(2): 98–118.
References 195
Lupu, Noam. 2013. “Party Brands and Partisanship: Theory with Evidence from
a Survey Experiment in Argentina.” American Journal of Political Science 57 (1):
49–64.
Mainwaring, Scott. 1986. The Catholic Church and Politics in Brazil, 1916–1985.
Stanford University Press.
Mainwaring, Scott, Rachel Meneguello, and Timothy J. Power. 2000. Partidos
Conservadores No Brasil Contemporâneo: Quais São, O Que Defendem, Quais
São Suas Bases. São Paulo: Paz e Terra.
Maldonado, Arturo. 2011. “Compulsory Voting and the Decision to Vote.” 63.
AmericasBarometer Insights. Latin American Public Opinion Project. Vanderbilt
University.
Mali, Tiago. 2014. “Candidatos evangélicos crescem quase 50%.” Época, September
16. http://epoca.globo.com/tempo/eleicoes/noticia/2014/09/bcandidatos-evangelicosb
-crescem-quase-50.html.
Malta, Magno. 2014. “Speech to the Brazilian Federal Senate.” Brasília, August 20.
www25.senado.leg.br/web/atividade/pronunciamentos/-/p/texto/409272.
Margolis, Michele F. 2016. “Cognitive Dissonance, Elections, and Religion: How
Partisanship and the Political Landscape Shape Religious Behaviors.” Public
Opinion Quarterly 80 (3): 717–40.
2018. From Politics to the Pews: How Partisanship and the Political Environment
Shape Religious Identity. Chicago: University of Chicago Press.
Mariano, Ricardo. 2011. “Laicidade à brasileira: católicos, pentecostais e laicos em
disputa na esfera pública.” Civitas – Revista de Ciências Sociais 11 (2): 238–58.
Mariano, Ricardo, and Antônio Flávio Pierucci. 1992. “O envolvimento dos
pentecostais na eleição de Collor.” Novos Estudos Cebrap 34: 92–106.
Marietta, Morgan. 2008. “From My Cold, Dead Hands: Democratic Consequences of
Sacred Rhetoric.” The Journal of Politics 70 (3): 767–79.
Mariz, Cecília Loreto. 2006. “Catolicismo no Brasil contemporâneo: Reavivamento
e diversidade.” In As Religiões No Brasil, edited by Faustino Teixeira and
Renata Menezes, 53–68. Petrópolis, Brazil: Editora Vozes.
Mariz, Cecília Loreto, and Maria das Dores Campos Machado. 1997. “Pentecostalism
and Women in Brazil.” In Power, Politics, and Pentecostals in Latin America, edited
by Edward L. Cleary and Hannah W. Stewart-Gambino, 41–54. Boulder, CO:
Westview Press.
Marshall, Ruth. 2009. Political Spiritualities: The Pentecostal Revolution in Nigeria. 1st
ed. Chicago and London: University of Chicago Press.
Martin, Bernice. 2006. “Pentecostal Conversion and the Limits of the Market
Metaphor.” Exchange 35 (1): 61–91.
Martin, David. 1993. Tongues of Fire: The Explosion of Protestantism in Latin America.
Oxford: Wiley.
Mason, Lilliana. 2015. “‘I Disrespectfully Agree’: The Differential Effects of Partisan
Sorting on Social and Issue Polarization.” American Journal of Political Science 59
(1): 128–45.
2018. Uncivil Agreement: How Politics Became Our Identity. Chicago: University of
Chicago Press.
Mayrl, Damon. 2016. Secular Conversions: Political Institutions and Religious
Education in the United States and Australia, 1800–2000. Cambridge, UK:
Cambridge University Press.
196 References
McAdams, Erin S., and Justin Earl Lance. 2013. “Religion’s Impact on the Divergent
Political Attitudes of Evangelical Protestants in the United States and Brazil.”
Politics and Religion 6 (03): 483–511.
McClendon, Gwyneth, and Rachel Beatty Riedl. 2015. “Religion as a Stimulant of
Political Participation: Experimental Evidence from Nairobi, Kenya.” Journal of
Politics 77 (4): 1045–57.
McClurg, Scott D. 2006. “The Electoral Relevance of Political Talk: Examining
Disagreement and Expertise Effects in Social Networks on Political
Participation.” American Journal of Political Science 50 (3): 737–54.
McCrudden, Christopher. 2015. “Transnational Culture Wars.” International Journal
of Constitutional Law 13 (2): 434–62.
McDermott, Monika L. 2009. “Religious Stereotyping and Voter Support for
Evangelical Candidates.” Political Research Quarterly 62 (2): 340–54.
Menchik, Jeremy. 2016. Islam and Democracy in Indonesia: Tolerance without
Liberalism. Cambridge, UK: Cambridge University Press.
Mendonça, Antônio Gouvea. 2006. “Evangélicos e pentecostais: Um campo religioso em
ebulição.” In As Religiões No Brasil, edited by Faustino Teixeira and
Renata Menezes, 89–110. Petrópolis, Brazil: Editora Vozes.
Migheli, Matteo. 2017. “The Gospel and Economic Behaviour: Experimental Evidence
from a Trust Game.” Annals of Public and Cooperative Economics 33 (1): 33–45.
Mir, Luís. 2007. Partido de Deus: Fé, Poder e Política. São Paulo: Alaúde Editorial.
Miranda, Ricardo. 2008a. “ELEIÇÕES 2008: Carta Divulgada.” Tribuna de Minas.
2008b. “Evangélicos querem um quarto das cadeiras da Câmara.” Tribuna de Minas.
Mora, G. Cristina. 2008. “Marketing the ‘Health and Wealth Gospel’ across National
Borders: Evidence from Brazil and the United States.” Poetics, Religion and Culture
36 (5–6): 404–20.
Mora Torres, Jose Enrique. 2010. “The Political Incorporation of Pentecostals in
Panama, Puerto Rico and Brazil: A Comparative Analysis.” Unpublished doctoral
dissertation, Storrs, CT: University of Connecticut. https://opencommons
.uconn.edu/dissertations/AAI3451309.
Moraes, Manoel Ribeiro de. 2010. “ELEIÇÕES 2010: Democracia versus iniquidade:
O purismo religioso à disposição do retrocesso democrático.” Novos Diálogos:
Jornal de Justiça, Espiritualidade e Cultura (online).
Morgan, William, and Christopher Winship. 2007. Counterfactuals and Causal
Inference. Cambridge, UK: Cambridge University Press.
Morrison, Penelope Kay. 2010. “Democracy ‘At Risk’? Governmental and Non-
Governmental Organizations, ‘At Risk’ Youth and Programming in Juiz de Fora,
Brazil.” Unpublished doctoral dissertation. Pittsburgh, PA: University of
Pittsburgh.
Moseley, Mason W. 2015. “Contentious Engagement: Understanding Protest
Participation in Latin American Democracies.” Journal of Politics in Latin
America 7 (3): 3–48. http://d-scholarship.pitt.edu/9046/.
Moseley, Mason W., and Matthew L. Layton. 2013. “Prosperity and Protest in Brazil:
The Wave of the Future for Latin America?” 93. AmericasBarometer Insights.
Mucinhato, Rafael Moreira Dardaque. 2014. “Quem são os deputados brasileiros? Um
balanço do perfil biográfico de 1986 a 2012.” In O Congresso Nacional, Os
Partidos Políticos e o Sistema de Integridade: Representação, Participação
References 197
Pew Research Center. 2006. Spirit and Power: A 10-Country Survey of Pentecostals.
Washington, DC: Pew Forum on Religion and Public Life. www.pewforum.org/
Christian/Evangelical-Protestant-Churches/Spirit-and-Power.aspx.
2011. Global Christianity: A Report on the Size and Distribution of the World’s
Christian Denominations. www.pewforum.org/2011/12/19/global-christianity
-traditions/.
Phillips, Dom. 2015. “Afro-Brazilian Religions Struggle against Evangelical Hostility.”
The Washington Post, February 5. www.washingtonpost.com/world/the_americas/
afro-brazilian-religions-struggle-against-evangelical-hostility/2015/02/05/
b6a30c6e-aaf9-11e4-8876-460b1144cbc1_story.html.
Philpott, Daniel. 2007. “Explaining the Political Ambivalence of Religion.”
The American Political Science Review 101 (3): 505–25.
2009. “Has the Study of Global Politics Found Religion?” SSRN Scholarly Paper ID
1416404. Rochester, NY: Social Science Research Network.
Pierucci, Antônio Flávio, and Reginaldo Prandi. 1995. “Religiões e voto: a eleição
presidencial de 1994.” Opinião Pública 3 (1): 32–63.
Pitkin, Hanna Fenichel. 1967. The Concept of Representation. Berkeley, CA: University
of California Press.
Polêmica Paraíba. 2016. “VEJA VÍDEO – Marcelo Crivella Diz ‘gay é fruto de aborto
malsucedido.’” Polêmica Paraíba. www.polemicaparaiba.com.br/politica/veja
-video-marcelo-crivella-diz-gay-e-fruto-de-aborto-malsucedido/.
Pope Francis. 2013. “Evangelii Gaudium: Apostolic Exhortation on the Proclamation
of the Gospel in Today’s World.” Vatican: The Holy See. http://w2.vatican.va
/content/francesco/en/apost_exhortations/documents/papa-francesco_esortazione
-ap_201311-24_evangelii-gaudium.html.
Power, Timothy J. 2009. “Compulsory for Whom? Mandatory Voting and Electoral
Participation in Brazil, 1986–2006.” Journal of Politics in Latin America 1 (1):
97–122.
Power, Timothy J., and Cesar Zucco. 2012. “Elite Preferences in a Consolidating
Democracy: The Brazilian Legislative Surveys, 1990–2009.” Latin American
Politics and Society 54 (4): 1–27.
Power, Timothy J., and Rodrigo Rodrigues-Silveira. 2018. “The Political Right and
Party Politics.” In The Routledge Handbook of Brazilian Politics, edited by
Barry Ames. New York and London: Routledge.
Przeworski, Adam. 1999. Democracy and the Market Political and Economic Reforms
in Eastern Europe and Latin America. New York: Cambridge University Press.
Putnam, Robert D., and David E. Campbell. 2012. American Grace: How Religion
Divides and Unites Us. 2nd ed. New York: Simon & Schuster.
Ranquetat Júnior, Cesar Alberto. 2016. Laicidade à brasileira: Um estudo sobre
a controvérsia em torno da presença de símbolos religiosos em espaços públicos.
Jundiaí, São Paulo: Paco Editorial.
Rayside, David, and Clyde Wilcox. 2011. Faith, Politics, and Sexual Diversity in Canada
and the United States. Vancouver: UBC Press.
Redação Pragmatismo. 2016. “Maluf e bancada evangélica declaram apoio ao
impeachment de Dilma.” Pragmatismo Político (blog). April 7. https://www
.pragmatismopolitico.com.br/2016/04/maluf-e-bancada-evangelica-declaram-apoio
-ao-impeachment-de-dilma.html.
References 199
Reich, Gary, and Pedro dos Santos. 2013. “The Rise (and Frequent Fall) of Evangelical
Politicians: Organization, Theology, and Church Politics.” Latin American Politics
and Society 55 (4): 1–22.
Rink, Anselm. 2018. “Do Protestant Missionaries Undermine Political Authority?
Evidence from Peru.” Comparative Political Studies 51 (4): 477–513.
Robbins, Joel. 2004. “The Globalization of Pentecostal and Charismatic Christianity.”
Annual Review of Anthropology 33 (1): 117–43.
Ryan, Timothy J. 2014. “Reconsidering Moral Issues in Politics.” Journal of Politics 76
(02): 380–97.
Ryfe, David M. 2005. “Does Deliberative Democracy Work?” Annual Review of
Political Science 8 (1): 49–71.
Sadgrove, Joanna, Robert M. Vanderbeck, Johan Andersson, Gill Valentine, and
Kevin Ward. 2012. “Morality Plays and Money Matters: Towards a Situated
Understanding of the Politics of Homosexuality in Uganda.” The Journal of
Modern African Studies 50 (1): 103–29.
Salazar, Camila. 2017. “Voto por diputados evangélicos se triplicó en cinco elecciones.”
La Nación, June 26. www.nacion.com/gnfactory/investigacion/2017/partidos
-evangelicos/index.html?desktop=true.
Sampaio, Tatiana de Souza. 2016. “A influência das organizações não governamentais
na política brasileira de direitos humanos LGBT+.” Paper presented at Entre
o global e o local, Conference XVII, Encontro de História da Anpuh-Rio. Rio de
Janeiro, August 8–11.
Samuels, David, and Cesar Zucco. 2014. “The Power of Partisanship in Brazil: Evidence
from Survey Experiments.” American Journal of Political Science 58 (1): 212–25.
2018. Partisans, Anti-Partisans, and Non-Partisans: Voting Behavior in Brazil.
Cambridge, UK: Cambridge University Press.
Sandal, Nukhet A. 2017. Religious Leaders and Conflict Transformation: Northern
Ireland and Beyond. Cambridge, UK: Cambridge University Press.
Sant’Anna, Lourival. 2010. “Eleição mostra influência das igrejas.” O Estado de São
Paulo, October 10. https://brasil.estadao.com.br/noticias/geral,eleicao-mostra
-influencia-das-igrejas-imp-,623104.
Santos, André Luis dos. 2009. “O processo de formação das atitudes políticas dos jovens
da Assembléia de Deus em Porto Alegre/RS.” Revista historiador 2(2): 364–98.
Sartori, Giovanni. 1976. Parties and Party Systems: A Framework for Analysis. Vol. 1.
Cambridge, UK: Cambridge University Press.
Schiavon, Fabiana. 2009. “Acordo de Brasil e Vaticano é aprovado em comissão.”
Consultor Jurídico, August 13. www.conjur.com.br/2009-ago-13/acordo-entre
-brasil-vaticano-aprovado-comissao-camara.
Schwartz, Mildred A., and Raymond Tatalovich. 2009. “Cultural and Institutional
Factors Affecting Political Contention over Moral Issues.” Comparative Sociology
8 (1): 76–104.
Serbin, Kenneth P. 2000. Secret Dialogues: Church-State Relations, Torture, and Social
Justice in Authoritarian Brazil. Pittsburgh, PA: University of Pittsburgh Press.
Serbin, Ken. 2006. Needs of the Heart: A Social and Cultural History of Brazil’s Clergy
and Seminaries. Notre Dame, IN: University of Notre Dame Press.
Shariff, Azim F., and Mijke Rhemtulla. 2012. “Divergent Effects of Beliefs in Heaven
and Hell on National Crime Rates.” PLoS ONE 7 (6): e39048.
200 References
Verba, Sidney, Kay Lehman Schlozman, and Henry E. Brady. 1995. Voice and Equality:
Civic Voluntarism in American Politics. Cambridge, MA: Harvard University Press.
Vianna, Cláudia Pereira. 2015. “O movimento LGBT e as políticas de educação de
gênero e diversidade sexual: Perdas, ganhos e desafios.” Educação e Pesquisa 41 (3):
791–806. https://doi.org/10.1590/s1517-97022015031914.
Wainscott, Ann Marie. 2017. Bureaucratizing Islam: Morocco and the War on Terror.
Cambridge, UK: Cambridge University Press.
Wald, Kenneth D., Dennis E. Owen, and Samuel S. Hill. 1988. “Churches as Political
Communities.” American Political Science Review 82 (2): 531–48.
1990. “Political Cohesion in Churches.” Journal of Politics 52 (1): 197–215.
Weber, Christopher, and Matthew Thornton. 2012. “Courting Christians: How
Political Candidates Prime Religious Considerations in Campaign Ads.” Journal
of Politics 74 (02): 400–13.
Webster, Steven W., and Alan I. Abramowitz. 2017. “The Ideological Foundations of
Affective Polarization in the US Electorate.” American Politics Research 45 (4):
621–47.
Western, Bruce, and Deirdre Bloome. 2009. “Variance Function Regressions for
Studying Inequality.” Sociological Methodology 39 (1): 293–326.
Wilson Quarterly. 2010. “The Parishioner Is Always Right.” Wilson Quarterly 34 (1):
76–7.
Wolf, James G. 1989. Gay Priests. 1st ed. San Francisco: HarperCollins.
Wolfe, Alan. 1999. One Nation, After All: What Middle-Class Americans Really Think
About: God, Country, Family, Racism, Welfare, Immigration, Homosexuality,
Work, the Right, the Left, and Each Other. New York: Penguin Books.
Zaller, John R. 1992. The Nature and Origins of Mass Opinion. Cambridge, UK:
Cambridge University Press.
Zorzanelli, Marcelo. 2016. “Após ser citado por todos deputados pró-impeachment,
Deus será investigado pelo Ministério Público.” Sensacionalista, April 17. https://
www.sensacionalista.com.br/2016/04/17/apos-ser-citado-por-todos-deputados-
pro-impeac-hment-deus-sera-investigado-pelo-ministerio-publico/.
Index
abortion, 18, 20, 166 clergy involvement, 42, 43, 80, 81, 82, 83,
clergy views on, 67, 117, 119
and Dilma Rousseff, 20 electoral rules, 21, 38, 86, 157
legal status in Brazil, 3, 92, 104 and evangelical support, 7, 21, 29, 88, 90,
legislator views, 154, 155, 161 114, 116, 118, 123, 131, 157, 160
and public opinion, 3, 18, 92, 99, 103, 104, Roman Catholic Church support, 20, 29, 123
109, 110, 111 Campos, Eduardo, 3
religious-elite views on, 67, 68 Canada, 28
and religious switching/conversion, 107 Cardoso, Fernando Henrique, 20, 153
and the Roman Catholic Church, 63 Catholicism
affective polarization; see polarization, affective clergy endorsements, 84
Afro-Brazilian religions, 12, 14, 19, 152 demographic trends, 11, 12, 14, 70
agnosticism: see atheism and agnosticism demographic trends in Latin America, 162
Alckmin, Geraldo, 116 in Chile, 165
Aldrich, John H., 171 party ties, 29, 84, 85, 105
Almeida, Ronaldo, 151 and transition to democracy, 17
Alvarado, President Carlos (of Costa Rica), 162 and voting guides, 84
American Convention on Human Rights, 162 caucuses in National Congress, 152; see also
Assembly of God, 16, 39, 81, 82, 84, 101, 157 Roman Catholic Apostolic Parliamentary
atheism and agnosticism, 70, 139, 153 Front; evangelical caucus
Caxias do Sul, 115
Belgium, 43 Chamber of Deputies, 4, 20, 29, 150, 153, 157
Benedict XVI, Pope, 20 leadership, 153
Bolsonaro, Eduardo, 150 Charismatic Catholicism, 13, 34, 70, 109,
Bolsonaro, Jair, 154, 174 132, 152
brokers, electoral, 42, 44, 160, 161, 170 Charismatic Protestantism, 13, 34
chastity,
campaign finance, 6, 150, 158 clergy views on, 67
campaigns Chile, 163, 164, 165
2002 general election, 115 church–state relations, 3, 35, 76, 102, 166
2006 general election, 115 in Chile, 165
2008 local election, Juiz de Fora, 47, 116, 117 legislator preferences, 155
2010 general election, 19, 85, 116, 122, 123 and noise ordinances, 131
2014 general election, 20, 21, 116, 122 perceptions of, 76, 94
203
204 Index
church–state relations, (cont.) public support for, 22, 72, 131, 133, 134,
and public opinion, 99, 106, 109 150, 167, 171, 174
civic education, 39, 149, 167 “third wave,” 149
civil society; see political participation representative democracy, 149
clergy transition to, 16, 17
constraints on speech of, 66 Demócratas Party, 155
influence on candidate electability, 126 Dias, Pastor Everaldo, 119, 155
influence on congregant vote choice, 122, doctrinal conservatism, 41, 42, 62, 95, 111
123, 126, 126, Downs, Anthony, 70
motivations of, 31, 68, 71 drought, 73
policy views of, 66 dualism; see evil, perceptions of
religious teachings of, 35, 64, 71
and socialization, 36 ecclesiastical base communities, 17
clientelism, 90, 149, 159, 160, 161, 170 economic policy
and citizens’ norms, 87 and legislator views, 154
in congregations, 150 and public opinion, 104
Colombia, 163, 165, 170 education in public schools
communism, core curriculum of, 18
opposition to, 85 and evolution, 19
compulsory and voluntary voting rules, 87, 120 religious education in, 17, 38, 101
Concordat (Brazil and Holy See, 2009), 38 sexual education in, 3, 18
congregations electoral system, 6, 43, 151, 157, 160, 166
competition among, 51, 89 End Times theology (eschatology), 64, 71, 132
growth and decline of, 50, 52, 62, 68, 70 endorsements
hosting politicians, 158 from clergy, 80, 81, 82, 83, 84, 86, 88, 89,
and outreach, 51 117, 122
political diversity (internal) of, 91, 128 and citizen norms, 86, 87, 88, 91
and social influence, 99, 108, 110, 119, environmental protection
123, 128 clergy views on, 67
conscientious voting, 39 and public opinion, 99, 104
and citizen norms, 87 evangelical caucus (bancada evangélica), 16,
clergy discussion of, 80, 83, 92, 117 115, 152
Constitution of Brazil, 17, 37, 133 Evangelical Parliamentary Front (Frente
Costa Rica, 163, 164, 165 Parlamentar Evangélica); see
conversion; see religious switching Parliamentary Front for Defense of
Crivella, Bishop Marcelo, 18, 19, Traditional Peoples of African Origin;
151 evangelical caucus
culture wars evangelicalism
causes, 7, 27, 28 and Afro-Brazilian religions, 14, 19, 152
definition of, 4, 27 definition of, 10, 100, 166
Cunha, Eduardo, 20, 151 and demographic trends, 11, 12, 14
and demographic trends in Latin
Dancygier, Raphaela, 43 America, 162
democracy, 165; see representative democracy; ideological position in Brazil, 99
participatory democracy and National Constituent Assembly
breakdown of, 136, 143 (1987–1988), 16
clergy support for, 5, 61, 72, 133 and religious attendance, 13
deliberative democracy, 130, 136 evangelization
indicators of level of, 22 restrictions on, 106
liberal democracy, 130, 136 evil, perceptions of, 63, 132, 136, 166, 169
participatory democracy, 130, 141 evolution, 19
Index 205
Feliciano, Pastor Marco, 151 Lula (President Luiz Inácio Lula da Silva),
frames, religious, 40 38, 115
France, 26, 33
Francis, Pope, 84 Macedo, Bishop Edir, 18, 67
Free Electoral Hour (Horário Eleitoral Malafaia, Pastor Silas, 81
Gratuito), 48 Malta, Senator Magno, 18
Manichaean theology; see evil, perceptions of
Garotinho, Anthony, 115 Mattos, Custódio, 47, 116
“gender ideology,” 3, 75, 92 Methodism, 157
gender roles Methodist Church, 9
public opinion, 109, 111 Mexico, 122
Gill, Anthony, 17, 32, 165 military regime, 174
Guatemala, 163 ministry to poor
clergy teachings on, 66
homosexuality, 166; see also same-sex marriage MBL (Free Brazil Movement), 170
among Catholic clergy, 71
anti-hate-speech laws, 106 National Conference of Bishops of Brazil,
clergy views on, 5, 61, 63, 64, 67, 71, 81 17, 18
conversion therapy, 103 National Constituent Assembly
discrimination and hate crimes, 18, 75 (1987–1988), 16
gay candidates, attitudes toward, 139 National Forum for Religious Education, 17
gay-rights movement, 27, 73, 162 neo-Pentecostalism, 12, 34
gay-friendly churches in Brazil, 169 Neves, Aécio, 21
politician views on, 62 Nicaragua, 165
public opinion on, 99, 100, 101, 102, 103 Noraldino Lúcio Dias Junior, 157
Hunter, James Davison, 4, 27
hypodermic models of persuasion, 113 open-list proportional representation, 43, 151,
157, 166
ideological identification Operation Car Wash, 6, 19, 22, 29, 151
and legislator views, 155
and public opinion, 104 Parliamentary Front for Defense of Traditional
Integral Mission (evangelical movement), 66 Peoples of African Origin, 152
Inter-American Court of Human Rights, 162 participatory democracy; see democracy
partisanship, 23, 29, 137
Juiz de Fora, 47, 157 anti-petismo (anti-PT sentiment), 73, 81,
Council of Pastors, 48, 62 84
local elections of 2008, 47 Catholic, 8, 19, 84, 85, 99, 105
politics of, 17 evangelical, 19, 84, 105
petismo (PT support), 22, 23, 129
Kassab, Gilberto, 48 party system, 4, 6, 8, 160, 161, 166; see also
Workers’ Party (PT)
legislators, see politicians evangelical parties, 7, 155
legitimacy of political system rightist parties, 29, 155
public opinion on, 136 Pentecostalism
legitimacy, perceptions of state, 130 and prosperity theology, 65
clergy views on, 61, 74, 77 definition of, 10, 12, 100
public opinion on, 23, 131, 133, 134, 135 and demographic trends, 13
liberal democracy; see democracy in Africa, 30
liberation theology, 17, 31 and styles of campaigning, 91
lobbying (legislative advocacy) by churches, 17, Pereira, Pastor Everaldo Dias, 21, 116
19, 89, 142, 158 Pitkin, Hanna Fenichel, 149
206 Index