Webpdf

Electoral Participation in Newly
Consolidated Democracies
This book examines why people vote in the newly consolidated democra-
cies of Africa, Latin America, East Asia, and Central and Eastern European
countries.
It addresses the question of how well models or theories of electoral par-
ticipation, initially developed in established democracies, “travel” to new
democracies. Based on recent cross-national survey data, it provides the first
systematic and comparative evaluation of this topic. Drawing on political
science, sociology, and psychology approaches, it reveals what is distinctive
about voting in new democracies and how they compare between themselves
and with more established democracies.
This book will be of key interest to scholars and students of political par-
ticipation, public opinion, voting behaviour, electoral politics, and political
parties as well as to international organisations and NGOs working in the
field of democracy promotion and in emerging democracies.
Elvis Bisong Tambe is Senior Lecturer in the Department of Political Science

at Linnaeus University, Sweden.
Routledge Research in Comparative Politics
Health and Political Engagement

Mikko Mattila, Lauri Rapeli, Hanna Wass and Peter Söderlund
Theorizing Cultures of Political Violence in Times of Austerity

Studying Social Movements in Comparative Perspective
Joanna Rak
Interest Group Organisation in the European Union

How Internal Organisational Structures Shape Interest Group Agency
Michelle Hollman
Coalition Government as a Reflection of a Nation’s Politics and Society

A Comparative Study of Parliamentary Parties and Cabinets in 12 Countries
Edited by Matt Evans
Transparency and Secrecy in European Democracies

Contested Trade-offs
Edited by Dorota Mokrosinska
Climate Governance across the Globe

Pioneers, Leaders and Followers
Edited by Rüdiger K.W. Wurzel, Mikael Skou Andersen and Paul Tobin
The Future of Political Leadership in the Digital Age

Neo-Leadership, Image and Influence
Edited by Agnieszka Kasińska-Metryka and Tomasz Gajewski
Electoral Participation in Newly Consolidated Democracies

Turnout in Africa, Latin America, East Asia, and Post-Communist Europe
Elvis Bisong Tambe
Electoral Participation in Newly
Consolidated Democracies
Turnout in Africa, Latin America, East
Asia, and Post-Communist Europe
Elvis Bisong Tambe

First published 2021
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2021 Elvis Bisong Tambe
The right of Elvis Bisong Tambe to be identified as author of this work has
been asserted by him in accordance with sections 77 and 78 of the Copyright,
Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised
in any form or by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying and recording, or in any information
storage or retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks,
and are used only for identification and explanation without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Names: Tambe, Elvis Bisong, author.
Title: Electoral participation in newly consolidated democracies: turnout in Africa,
Latin America, East Asia, and post-communist Europe / Elvis Bisong Tambe.
Description: Abingdon, Oxon; New York, NY: Routledge, 2021. |
Series: Routledge research in comparative politics |
Includes bibliographical references and index.
Identifiers: LCCN 2020043083 (print) | LCCN 2020043084 (ebook) |
ISBN 9780367468446 (hardback) | ISBN 9781003031482 (ebook)
Classification: LCC JQ1879.A15 T36 2021 (print) |
LCC JQ1879.A15 (ebook) | DDC 324.9–dc23
LC record available at https://lccn.loc.gov/2020043083
LC ebook record available at https://lccn.loc.gov/2020043084
ISBN: 978-0-367-46844-6 (hbk)
ISBN: 978-1-003-03148-2 (ebk)
Typeset in Times New Roman
by Newgen Publishing UK
Contents
List of tables vi
List of figures viii
Acknowledgements ix
1 Introduction 1
2 Why citizens vote 12
3 The comparative framework 59
4 Socio-demographics and the vote 84
5 Mobilisation and the vote 97
6 Political-psychological variables and the vote 116
7 Alternative explanations and the vote 130
8 The electoral context and the vote 146
9 A comparative overview 165
Conclusion 176
Appendices 181
References 209
Index 233
Tables
2.1 Overview of the expectation of the socio-demographic model

on vote choice 25
2.2 Overview of the expectation of the mobilisation model on
vote choice 30
2.3 Overview of the expectation of the political-psychological
model on vote choice 36
2.4 Overview of the expectation of the institutional model on
vote choice 40
2.5 Overview of the expectation of an alternative model on vote
choice 55
3.1 Selected cases 64
3.2 Political rights index for sub-Saharan Africa, Latin America,
East Asia, and post-communist states 65
3.3 Information about the dependent variable 71
4.1 Socio-demographic variables and vote choice in newly
consolidated democracies: bivariate relationship 86
4.2 Logistic regression of socio-demographic model and voting
in newly consolidated democracies 89
4.3 Theoretical matching of socio-demographic model on vote
choice –based on the multivariate results 92
5.1 Trends of the levels of party identification 99
5.2 Party identification, associational membership, and
clientelism: bivariate relationship 102
5.3 Multivariate models of voter turnout in newly consolidated
democracies: socio-demographic and mobilisation variables 103
5.4 Theoretical matching of mobilisation model on vote
choice: based on the multivariate results 108
5.5 Scope of clientelism practices in new democracies 111
5.6 The effect of clientelism on vote choice in new democracies 112
6.1 Political-psychological variables and electoral participation
in new democracies: bivariate associations 120
democracies: socio-demographic, mobilisation and
political-psychology variables 122
List of tables vii
6.3 Theoretical matching of political-psychological variables on
vote choice: based on the multivariate results 128
7.1 Transparency International Corruption Perception Index 131
7.2 Voters’ socio-tropic retrospective evaluation of national
economy in new democracies 133
7.3 Corruption, authoritarianism and the economy and voting in
new democracies: bivariate relationships 137
democracies: socio-demographic, mobilisation, political-
psychology and alternative variables 139
8.1 Trends in voter turnout in new democracies 148
8.2 Countries and values of their political-institutional
characteristics 151
8.3 The impact of political and institutional factors on
voting: bivariate relationship 154
democracies: individual and political-institutional variables 157
9.1 Summary of empirical findings 167
Figures
6.1 Interest in politics 117

6.2 Trust in government 118
6.3 Trust in people 119
7.1 Democratic erosion: number of symptoms by year and region 134
7.2 Democratic erosion: types of symptoms by region 135
Acknowledgements
Despite my personal efforts in writing this study on electoral participation in

newly consolidated democracies, the success of this book cannot be attributed
entirely to my own efforts. Rather, it has been the outcome of the contributions
of various individuals who assisted me either directly or indirectly. First, spe-
cial thanks go to my PhD supervisors at the University of Sussex, Paul Webb
and Rekha Diwaker. Their contribution and support to the whole process of
my studies were just outstanding, and without their guidance, help, and direc-
tion, I could never have managed to write this book. Paul has been an excep-
tional supervisor and mentor, even after I completed my time at Sussex. He
had patience in reading through my book proposal and the final manuscript
providing me with extra suggestions and comments before I could submit it
to the publisher. Also, Paul was very encouraging when I sought an academic
position.
The academic job market is tough and quite depressing. As I moved from
one short-term position to another, Paul persistently wrote uncountable refer-
ence letters for me, while equally providing me with tips on how to approach
the few interviews I had in the UK, the US, and Sweden. Moreover, he checked
on me and my family over and over again to see how I was doing. A major dis-
advantage of undertaking a PhD study in the UK, as a self-sponsored student
without any sort of financial support (i.e. it truly feels like a miracle that all is
now completed!), is that I was unable to build a good research network during
my PhD time as I had to work full-time while studying. The implication is that
this somehow fell back on Paul, as I had to ask him to offer me comments/
suggestions for virtually every piece of writing I did and which he did every
time without hesitating. All said, words cannot describe my admiration and
respect for Paul.
Second, I am also thankful to Aleks Szczerbiak and Benjamin Stanley who
were actively involved in the supervisory process during the initial stages of
my PhD. Also, Aleks was kind enough to read my original book proposal that
was submitted at Routledge. The initial period of my doctoral project was
indeed the most challenging. I can still remember vividly my first supervisory
meeting with Paul Webb and Aleks Szczerbiak on 20 January 2014 at room
F16 (i.e. the old site of the Politics Department before it moved to Freeman
x Acknowledgements
Building), which was quite intimidating having read a profile of each of them
before moving to Sussex and discovering that they both possess a substantial
record in comparative research. Like a son taking the final advice from his
father before leaving home, I still remember vividly Aleks saying: Elvis, PhD
studies are an intellectual and psychological challenge, and I can now confirm
the truth of this statement as the whole process of completing my doctoral
dissertation and writing this book was indeed an intellectual and psycho-
logical hurdle.
Third, special thanks go to Paul Taggart and Susan Collard who were
both responsible for my PhD Annual Review at the University of Sussex.
Throughout my three years, not only did they assess my yearly progress,
but they made it a point to provide me with academic and non-academic
advice, which was important in my studies trajectory. Fourth, appreciation
also goes to Dan Hough at the Politics Department at Sussex and two of my
previous teachers, Nicholas Aylott and Mai-Brith Schartau (now retired), at
the University of Södertörn, Sweden, who were all very patient in reading
through my research proposal and giving me valuable comments before
I could submit it for my first-year evaluation at Sussex. Their initial comments
and suggestions gave me the psychological encouragement that I could navi-
gate myself through the journey of earning a doctoral degree.
Fifth, I am also grateful to my PhD viva voce examination committee,
most especially to Justin Fisher from Brunel University in London and
Francis Mcgowan from the Politics Department at Sussex. Although the
whole process of interrogating my thesis for two hours was quite a gruelling
experience, both Justin and Francis provided me with outstanding comments
and suggestions that later improved this book. My appreciation also goes
to Toygar Sinan Baykan (former PhD student at the University of Sussex),
who greatly encouraged me when I first contacted him about the idea of
writing this book. Given that he recently had the experience of publishing
a book with Cambridge University Press: The Justice and Development
Party in Turkey: Populism, Personalism and Organisation (i.e. a must-read for
comparativists interested in Turkeys party politics and new democracies), he
was very open in sharing with me his own experience and offering me some
advice on how to write a compelling book proposal.
Sixth, across the three years of my doctoral studies, I had the good for-
tune to attend and present parts of this project at two specialist schools on
political behaviour, elections, and party politics, notably the 24th ECPR PhD
Summer School on “political parties in new democracies” at the Center for
the Study of Democracy, Leuphana University Lüneburg, Germany, and
the 2nd Leuven-Montréal winter school on elections and voting behaviour,
Montréal, Canada (i.e. Leuven-Montréal winter school, an essential school
for every young scholar working on political behaviour, elections, and polit-
ical participation broadly defined). At these two specialist schools, I received
invaluable advice in the initial stages of my doctoral studies and which later
consisted much of what this book focuses on.
Acknowledgements xi
Also, I would like to thank the department of Political Science at Linnaeus
University, Sweden, for providing a stimulating and supporting environment
that led to the completion of this book. Additionally, a few colleagues from
my first academic position deserve special thanks, notably Henrique Carlos,
PI of the World Values Survey project in Brazil, Daniel Capistrano, and Sonia
Ranincheski (may her soul rest in peace). Upon completing my PhD from
Sussex University, I was blessed to pick up my first academic position as a
visiting assistant professor/postdoctoral fellow with the World Values Survey
Team at the Federal University of Rio Grande do Sul in Porto Alegre, Brazil.
Many thanks especially to Henrique who was not only my boss but even-
tually became a very good friend. He made so many efforts to ensure my
stay was intellectually stimulating through the many research conferences and
presentations we attended at UFRGS. He also considered my professional
development to be important; as such he took steps for me to be included in
a number of thesis examinations committees. Moreover, thanks to Henrique,
living in Brazil, during this time, provided me with so many fun memories
and experiences, like driving for almost 15 hours from Porto Alegre (Brazil)
to attend the VII Latin American Public Opinion Conference that was held
in Colonia del Sacramento, Uruguay, the ritual of drinking Chimarrão
(Brazilian mate), spending most of the weekend at Henrique’s home eating
homemade pizza and talking about Brazilian politics, most especially the rise
of authoritarian tendencies.
There are still many people whom I equally owe much gratitude to either
because of their encouragement, words of wisdom, or just because they
offered a listening ear to me. Many thanks go to Joakim Iseborg, my friend in
heart and spirit, without whom, I would not have embarked on this journey;
he was always there to confirm my path and directions, checking frequently to
see how I was holding together in body and soul. In fact, you mean so much
to me and I will not hesitate to say the Lord will bless you for your kindness
and love you have shown me, right back from the time our paths met in the
lovely city of Falun.
Finally, my family in Cameroon, Sweden, and the UK deserve the last
words of recognition and appreciation. As an African, the family is quite a
wide and expanded unit, and please do not ask me to mention everyone here.
Anyway, one thing I am sure is that most of them won’t be reading this book;
so I am 100 per cent certain none of them will be upset I didn’t mention
them here. Special thanks go to my father, Tambe Samuel Bisong, and mum,
Tambe Martha Bakume, and not forgetting about Jacob Mbah (Ni Mbah).
My parents grew up at a time in Cameroon when having a primary/secondary
education was just enough to see one through life and as such they did not
have the opportunity to acquire the “white-man education” as it was described
during that era. Yet, at a very young age, I was very much inspired by my
father’s love of reading. As the oldest in a family of six, I was always the
one who had to go out to buy newspapers for my dad, so my inspiration for
studies at this very young age was mainly built on this. Also, like I said, my
newgenprepdf
xii Acknowledgements
family is large. Without putting them in any order, I would like to thank the
following: Taku Samuel Bisong, Ma Comfort, Taboko Eveline, Andreas Tarh,
Melvis Tambe, Serge Ndi, Daniel Tanyi, Efu, Moses, Gertrude, Corazon,
Alex, Yvonne, Armstrong, Martha, Laetitia. Lastly, special thanks to my
partner, Margarete Anyi, for her constant love and support and our sons
who have been wonderful, understanding, and always supportive through the
writing process.
Finally, significant gratitude goes to the four anonymous referees who
provided constructive feedback on the original book proposal submitted to
Routledge. Without their insightful and constructive feedbacks and suggestions
notably on theory, research design, and structure, this book would have been
a less well conceived undertaking. Also, the analysis in this book was based
on data collected by the Afrobarometer, AsianBarometer, European Social
Survey, and AmericasBarometer surveys (the AmericasBarometer surveys
has a subscription fee, so I only had access to the free or reduced version of
the data). So many thanks to the Afrobarometer, AsianBarometer, European
Social Survey, and the Latin American Public Opinion Project (LAPOP) and
its various partners, sponsor, and supporters for making these data available.
Lastly, special thanks to the editor Andrew Taylor and Sophie Iddamalgoda
production editor and copyeditor Ting Baker at Routledge for being very
patient, friendly, and accommodating through this project.
1
Introduction
Did you vote in the last election? The question of why people turn out on
election day has been a longstanding issue of debate among political scientists
(Geys, 2006, p. 637) but the focus has typically been on the established dem-
ocracies of Western Europe and North America.1 This book examines the
question of why people vote in the new democracies of Africa, Latin America,
East Asia, and Central and Eastern European countries. Relatedly, it also
addresses the question of how well models or theories of electoral participa-
tion that were initially developed in established democracies “travel” to new
democracies.
These objectives are important because, over the past century, several waves
of democracy have swept across the globe, bringing about representative dem-
ocracy in places where it seemed inconceivable 50 years ago. The most recent
tide of democratisation was initiated in the mid-1970s, and, ever since, we
have witnessed a global spread of democracy. It began in Southern Europe in
the 1970s with the fall of the Portuguese dictatorship, then extended to Latin
America2 in the 1980s and culminated in the 1990s with the first democratic
elections for national representative institutions taking place in 44 countries
across Africa (22), Eastern Europe (19), and Asia (3) (International IDEA,
2002, p. 14).3 As a result of this change, the number of regimes in the world that
are presently referred to as liberal democracies has increased such that it has
become common to speak of a “third wave” of democratisation (Huntington,
1991). In this context, the question of electoral turnout is crucial as the future
of democracy in both established and emerging systems depends to a signifi-
cant extent on the electoral process. The existence of free and fair competitive
elections is invariably considered one of the critical features that defines a
nation as “democratic” (LeDuc et al., 1996; Katz, 1997). Indeed, Schumpeter
(1942, p. 265) famously offered a strictly procedural definition of democ-
racy: the democratic method is that institutional arrangement for arriving at
political decisions in which individual leaders acquire the power to decide by
means of a competitive struggle for the people’s vote. Huntington (1991, p. 7),
on the other hand, views democracy as a system in which decision-makers
are selected through fair, honest, and periodic elections in which candidates
freely compete for people’s votes. Similarly, Diamond (1999, p. 10) considers
2 Introduction
electoral democracy as a system characterised by regular, competitive, multi-
party elections with universal suffrage. Although it is possible to argue for
more stringent tests4 of a “true” democracy, it is, however, clear that elections
are a central institution of democratic governance because they reflect the
condition of democracy in as many ways as they define it (LeDuc et al.,
1996, p. 4).
Moreover, modern-day formulations of democratic theory recognise the
irreplaceable significance of the ballot box for the potential control of elites by
ordinary citizens by arguing that voting is democracy’s most fundamental act.
Bennett (1986) suggests that voting is the only form of political participation
undertaken by half the adult populace. One might rebuff this view by arguing
this rather depends on the country and election. For example, there are many
examples of elections that do not achieve a majority turnout, including US
presidential elections, European parliamentary elections in several European
countries, and local elections in countries such as the UK. Nevertheless,
an important point to note, as Roberts (2008, p. 3) argues, is that voting is
believed to keep community life vital and public institutions accountable. It
resolves conflict through “a participatory process of ongoing, proximate self-
legislation and the creation of a political community capable of transforming
dependent private individuals into free citizens and the partial and private
interest into public goods”. Also, voting is the main instrument by which citi-
zens can express their choices and influence policy in democracies. As a result,
Lanning (2008) argues that voter turnout is an exact and key indicator for a
functioning democracy.
With elections broadly regarded as the heartbeat of democracy, it is under-
standable that electoral participation has been a widely studied phenomenon
in the literature of comparative politics (Diwakar, 2008, p. 75). Nonetheless,
the question of why people vote has received greater attention among scholars
of Western democracies. These studies have not just been limited to voting at
the individual level but have also incorporated voting at the aggregate5 level
with a special focus on declining turnout and cross-national variations in
turnout (Kostelka, 2017; Blais and Carty, 1990; Blais and Dobrzynska, 1998;
Franklin, 2004; Jackman, 1987; Stockemer, 2015; 2016; Powell, 1980; Fornos,
Power, and Garand, 2004; Pérez-Liñán, 2001; Kostadinova, 2003; Kostadinova
and Power, 2007; Boulding and Brown, 2013; Pacek, Pop-Eleches, and Tucker,
2009; Lehoucq and Wall, 2004; Kuenzi and Lambright, 2007; Remmer, 2010).
There is certainly much of value in these studies and many interesting results
have been produced. However, the question of why people vote in newly emer-
ging democracies has yet to receive significant attention.
While, several studies have examined electoral participation within a single
country or region6 (Kuenzi and Lambright, 2011; Bratton et al., 2012; Bratton
et al., 2005; Bratton, 2013; Tambe, 2016; 2017a; 2018a; Orvista et al., 2005;
Flanagan et al., 1991; Fuh-Sheng Hsieh and Newman, 2002; Bernhagen and
March 2007; Barnes, 2006; Carlin, Singer, and Zechmeister, 2015; Carreras
and Castañeda-Angarita, 2014; Nadeau et al., 2017; Achen and Wang, 2017;
Introduction 3
Tsai Chia-Hung, 2001; Denton, 2016), little attention has been given to
cross-regional comparative research7 that explores and compares the nature
of electoral participation across new democracies after more than 20 years
of democratic transition and experience. In this volume, I examine turnout
at the individual level by asking: Why do people vote in newly consolidated
democracies? That is, what influences people’s decisions to vote? Relatedly it
also addresses the question of how well do models that were developed in
established democracies “travel” to new democracies. In order to furnish an
answer to this puzzle, I rely on recent cross-national public opinion surveys
from the Afrobarometer, AmericasBarometer, AsianBarometer, and the
European Social Survey to examine the factors that affect the individual
propensity to vote in eight countries in sub-Saharan Africa, nine countries
in Latin America, three in East Asia and six countries in Central/Eastern
Europe giving a grand total of 26 newly consolidated democracies across the
four geo-political regions.8 Thus, with the unique resources of cross-national
survey data, this book provides the first systematic and comparative evalu-
ation of this topic.
Although the focus of this study is on newly consolidated democracies, it is
equally important to examine what is happening in older established democra-
cies (i.e. Western European democracies) in relation as well, as this provides an
empirical benchmark against which to compare the voting behaviour in sub-
Saharan Africa, Latin America, East Asia, and Central/Eastern Europe: is
the process by which voters turn out in elections in the newly consolidated
democracies broadly similar to that we find in the established democracies of
Western Europe? The established democracies provide both theoretical and
empirical benchmarks against which to compare the new democracies.
Having established the research question that this book seeks to answer,
the rest of this introductory chapter deals with the expected contribution of
this study and the structure of the book. In addition, it is important to clarify
the term new democracies or newly consolidated democracies as it is applied in
this book. The terms new democracies, transitional democracies, emerging,
and young democracies tend to be used interchangeably by scholars and most
of the time these scholars are inclined to assume that readers understand their
usage. I shall therefore also dedicate a few words to this issue in this introduc-
tory chapter.
Purpose of the book

“Democracy is not just a spectator sport –it requires the active involve-
ment of its citizens” (Dalton, 2000, p. 57). This implies that a healthy dem-
ocracy will need the active participation of the citizenry. Thus, the question
of what influences people to participate in politics within a given polity
becomes important and warrants investigation.9 Specifically, this book, there-
fore, intends to make a contribution to political and electoral participation
research in three ways:
4 Introduction
First, understanding electoral participation at the individual level is in
itself an important research topic. Nevertheless, it is important to explain
why it is particularly interesting to focus on newly consolidated democracies.
The question of why or what influences people to vote has been fully explored
by scholars in advanced industrial democracies. Yet, even after more than
three decades of democratic experience, cross-national/regional comparative
research on voting in newly emerging democracies is rare or still developing.
This is an empirical gap this book fills by providing a broad examination
and explanation of electoral participation in young or newly consolidated
democracies.
Second, this book aims at theory –testing. With this ambition in mind, it
asks whether the standard models and explanations of vote choice that have
been developed in respect to established democracies work as well in newer
democracies. To be precise, some of the oldest electoral research took place
in the democracies of Western Europe, after that which was undertaken by
the Columbia and Michigan schools,10 US, where pioneering studies began
in the 1940s. National election studies were introduced across a handful of
these countries, beginning with Scandinavian countries: the Swedish national
election studies were initiated in 1956, followed by Norway in 1957 and then
other studies such as those in Britain, Germany and Netherlands (Curtice,
2002; Thomassen, 1994). Given that national election studies have now been
conducted in these countries and other Western European countries for more
than 50 years, it is normal that these countries have been inclined to spear-
head research in this area and, of course, the implication of this is that most
of the approaches or theories that have been postulated to explain why people
decide whether or not to vote have been influenced mostly by the experience
of European voters and or voters in Western advanced democracies. Thus, in
trying to explain what influences people to vote, this book draws on models
and theories that may be considered as Western constructed. How far can these
Western generated theories travel in the age of globalisation and the spread of
democracy? Does their explanation of voting fit well the emerging democra-
cies of sub-Saharan Africa, Latin America, East Asia, and post-Communist
European states?
Third, and from a broader comparative perspective, by focusing and
comparing sub-Saharan Africa, Latin America, East Asia, and Central and
Eastern European countries, this book explores whether there are similarities
of outcome, which might be expected given the relative newness of demo-
cratic elections in these regions, or if there are systematic differences between
the various geo-political regions. In a nutshell, the goal of this book is to
understand which of three patterns holds:
1. To see if there is something distinctive about voting in new democra-

cies as a whole –that is, to see if these newly consolidated democracies
broadly resemble each other but differ from the established democracies
of Western Europe; or
Introduction 5
2. to see if the new democracies are generally similar to Western European
countries or democracies and therefore confirm established research
truths; or
3. to see if each region is unique in its own way.
New democracies
What is the meaning of new-democracies? To address this question, the best
place to begin is with Samuel Huntington’s path-breaking study The Third
Wave: Democratisation in the Late Twentieth Century. Based on this study,
my operationalisation of countries I deem to be newly consolidated democra-
cies11 are defined by three broad criteria:
▪ Countries that have experienced the end of authoritarian rule.

▪ Countries that have experienced the installation of a democratic regime.
▪ Countries that are experiencing the consolidation of their democratic
regimes.
First, the countries referred to in this book as newly consolidated democ-

racies encompass countries that have experienced the end of authoritarian
rule during the “third wave” of democratisation. Huntington (1991) defines
a wave of democratisation as a group of transitions from non-democratic to
democratic regimes that occurs within a specific period of time and that sig-
nificantly outnumbers transitions in the opposite direction during that same
period of time. The third wave of democratisation has occurred at different
time intervals: first, it manifested itself in southern Europe in 1974, then
moved to Latin America in the late 1970s, then to Asia early in 1977, and
finally to post-Communist East European states and African countries in
1989 and the early 1990s.
Second, when I use the term newly consolidated democracies, I am refer-
ring not just to countries that have experienced the end of authoritarian
rule, but that have also witnessed the introduction of a democratic regime.
Huntington (1991, p. 9) argues that the inauguration of a democratic regime
was only confirmed in terms of the degree to which a country introduced pro-
cedural democracy. Since popular elections are considered to be the essence
of democracy, a critical point in the process of democratisation depends on
the replacement of the authoritarian leaders with a government that was
chosen or selected in a free, open, and fair election.
Third, new democracies are, by definition not consolidated (see, Kostelka,
2017), with Svolik (2015) arguing democratic consolidation can be expected
to take about 20 years (i.e. approximately five or six democratic elections). To
avoid ambiguity, Huntington (1991, p. 35) argues that “the circumstances that
contribute to the initial establishment of a democratic regime, may not con-
tribute to its consolidation and long-term stability”. Diamond (2002, p. 210)
suggests that it is one thing to inaugurate a democracy and quite another
6 Introduction
to maintain it in a stable fashion, thus implying that, for the future, a stable
functioning democracy requires a process that has been labelled democratic
consolidation. In this line, some scholars advance democratic consolidation can
be deem secured when ordinary citizens and the political elites accept and act
according to the rule of the game (see, e.g. Linz and Stepan, 1996; Przeworski,
1991) or as Schedler (1998 p. 91) puts it, a process where democracy becomes
the only game in town as the return to authoritarian rule becomes unthink-
able. Based on this, the operationalisation of countries as newly consolidated
democracies in this book refers to countries that have succeeded in consoli-
dating their democratic regimes. Huntington (1991, pp. 266–267) provides a
narrow definition of the term based on a two-turnover test. He suggests that
a country could be viewed as consolidated if the party that takes power in
the initial election at the time of transition loses a subsequent election and
the opposition party that ousts an incumbent is also defeated in a subsequent
election.
Boix, Miller, and Rosato (2013)12 on the other hand, define a country
as democratic if it satisfies conditions for both contestation and partici-
pation. Specifically, democracies feature political leaders chosen through
free and fair elections and satisfy a threshold value of suffrage. Cheibub
et al. (2010)13 judge a regime as democratic if the executive and the legis-
lature are directly or indirectly elected by popular vote, if multiple parties
are allowed, if there is de facto existence of multiple parties beyond the
appearance presented by the regime, if there are multiple parties within the
legislature, and if there has been no consolidation of incumbent advantage.
However, the turnover rule prescribed by Huntington (1991) and Cheibub
et al. (2010) makes the classification of democracy a bit complicated.14 For
example, in deciding which countries are democratic and which not, with
respect to sub-Saharan Africa these scholars do not consider Botswana
to be democratic because of the turnover rule, even though the country is
frequently described as a success story in terms of its democratic progress
and consolidation.
Stepan (1990, p. 44), on the other hand, provides a broad and argu-
ably more meaningful definition of the concept of democratic consolida-
tion that centres on the institutionalisation of the party system. To become
institutionalised, parties need to be competitive by demonstrating congruence
with the citizens’ policy priorities, and opposition parties, in particular, must,
therefore, represent credible democratic alternatives, which requires retaining
an independent ideological and institutional existence. The cases selected for
empirical analysis in this book will be defined by adherence to these three
criteria. In practical terms, this is achieved by limiting case-selection to tran-
sitional democracies that are considered democratic according to Freedom
House ratings,15 which capture the political rights in each country, especially
the fairness of their elections.16 This, therefore, reduces the number of coun-
tries17 I intend to study to 26 (i.e. eight in sub-Saharan Africa, nine in Latin
America, three in East Asia, and six in Central/Eastern Europe).
Introduction 7
The plan of the book

This book is divided into nine chapters. After this introductory chapter, the
second chapter provides a theoretical and empirical review of the existing
literature on political and electoral participation. The fact that these theories
motivate the research question helps me not just to use existing theories to
explain people’s voting decisions, but it also provides the leeway to evaluate
them critically as I see how far they can be stretched to explain voter turnout
in newly consolidated democracies. Similarly, though this book draws heavily
on previous studies of advanced democracies, it also discusses relevant studies
of electoral participation in Africa, Latin America, Asia, and Central/Eastern
Europe. In sum, the first part of this chapter is directed towards the gen-
eral features of the theories of voter turnout. Here, I discuss a number of
established theoretical and empirical approaches of turnout, summarising
the main assumptions and propositions of each, before moving on to state
the theoretical expectations that can be derived from these theories with
regards to electoral participation in newly consolidated democracies. In the
second part, I proceed by providing an alternative explanation of turnout in
newer democracies that focusses on issues such as corruption and authoritar-
ianism, etc. This section also considers an overview of the place of elections
in the democratisation process (i.e. how central are elections to democracy in
new democracies), followed with a brief discussion on why and how newly
consolidated democracies may differ from established democracies in terms
of individual-level voter turnout.
The third chapter asks: Why focus on the newly consolidated democra-
cies of Africa, Latin America, East Asia, and Central/Eastern Europe? To
answer this question, this chapter focuses on the comparative framework and
parameters for comparing the four regions. Next, a mention of case selection
and the comparative data and usefulness to the entire book/research question
will be highlighted. Finally, issues relating to the different methodological
challenges associated with studies of voting behaviour, the measurement of
dependent and independent variables will also be discussed.
Chapters 4 to 8 form the main empirical sections of the book. Here,
I move from description and research strategy to verifying explanations and
facts about voting behaviour. Using cross-national survey data from the
Afrobarometer, AmericasBarometer, Asianbarometer, and the European
Social Survey datasets, these chapters are structured along the primary/
major theories or theoretical framework. Also worth mentioning, I equally
provide a direct and explicit comparison with established democracies (i.e.
Western European democracies), as doing so allows an investigation of not
only whether but also by how much newly consolidated democracies differ
from older democracies of Western Europe. Chapters 4 through 7 examines
individual-level determinants of vote choice, while Chapter 8 examines the
role of electoral context/country-level or contextual characteristics on vote
choice. Finally, Chapter 9 concludes by providing a comparative overview of
8 Introduction
vote choice across the four regions by summarising the findings from the pre-
vious chapters (i.e. Chapters 4 through 8).
Chapter 4 commences with an interrogation of six socio-demographic or
descriptive characteristics, notably age, gender, religion, ethnicity/race, place
of residence and education on vote choice across four geopolitical regions
of the globe (i.e. sub-Saharan Africa, Latin America, East Asia, and post-
communist Europe). Overall, these results suggest quite a lot of agreement
across newly consolidated democracies but some inconsistency too. To be
specific, while the effect for some socio-demographics determinates on vote
choice across the four regions is more uniform (i.e. age, religion, and educa-
tion), but for others much less (gender and ethnicity/race). Overall, it would
suffice to say that socio-demographic models are strong determinants of
vote choice in Latin America, post-Communist European countries, and to
a lesser extent East Asia. While for sub-Saharan Africa I cannot say that the
socio-demographic predictors are generally powerful determinants of vote
choice.
Chapter 5 discusses the effect of three mobilisation variables on electoral
participation in new democracies notably party identification, association
membership, and clientelism. Overall, the results reveal a clear dominance
of mobilisation model when compared to the socio-demographic models/
variables. To be precise, two mobilisation variables, especially party identifi-
cation and association membership, play a more vital role in explaining elect-
oral behaviour in newer democracies. Albeit a more substantial impact of
party identification than associational membership, which is even more sur-
prising considering their limited experience with democracy and to a lesser
extent the instability of their party systems.
Chapter 6 examines the impact of political-psychological variables on vote
choice. It explores how political interest, political trust, internal and external
efficacy affect people’s voting decisions in newer democracies. The overall con-
tribution of this model to the vote choice model indicates it is quite modest
only in new democracies of post-Communist European countries, while for
the other regions, sub-Saharan Africa, Latin America, and East Asia, the gen-
eral contribution of the model to vote choice is relatively weak or tiny. But
among the political-psychological variables, political interest constitutes the
chief or principal driving force behind people’s decision to vote across all four
new democracies.
Chapter 7 discusses an alternative explanation of vote choice in newly
consolidated democracies that looks at corruption, authoritarianism, and
the economy. Three patterns emerge from the effect of these predictors on
vote choice. First, in three new democracies, particularly sub-Saharan Africa,
East Asia, and post-Communist European democracies, voters’ perception
of corruption tends to decrease voter turnout. Second, for the economy,
though, voters’ perception or evaluation of the economy does not fully attain
a significant level across most of the regions except in post- Communist
European countries where those who possess a more positive assessment
Introduction 9
of their country’s economy tend to vote more. Finally, considering authori-
tarianism, it suffices to say, people’s experience with authoritarian practices
tend to decrease participation in sub-Saharan Africa. Yet in post-Communist
European democracies, it is the reverse as the erosion of democratic principles
tends to mobilise voters to participate. Thus, the simple implication is that
voters in post-Communist European democracies will be more willing to
engage in voting and cast out such leaders as a way to protect democracies.
In contrast, in sub-Saharan Africa, any further increase with authoritarian
practices might pose a severe threat to electoral politics in this region as voters
might be less likely to participate.
Chapter 8 moves from individual-level explanations of vote choice to
country-level characteristics. This chapter examines the impact of political-
institutional variables on vote choice across four regions. Specifically, four
contextual predictors, namely disproportionality, the closeness of the
election, concurrent elections, and compulsory voting, are put to empirical
testing across the different regions, although the results across the different
regions reveal the impact of the institutional factors on vote choice gener-
ally tends to be distinct and limited. Yet, the effect of compulsory voting on
political behaviour stands out, as voters are more likely to turn out where
elections are mandatory and enforced with sanction than where they are not.
With regards to the other variables, I find that closeness of elections tends
to be influential determinants of vote choice in new democracies of sub-
Saharan Africa but not for the other regions. In sum, this chapter concludes
that while the political-institutional context is essential for vote choice, it is
far from decisive as most of the difference on whether or not a voter turns
out on election day in the different regions still very much depends on the
individual level.
Finally, in Chapter 9, the core objective of this book is to explain what
influences people to vote at an individual level and to establish how well the
models of voter turnout developed in established democracies can travel to
newer consolidated democracies. To do this, the book relied solely on models
and theories that may be considered as “Western constructed”; this, of course,
flags the importance of understanding how far these Western-generated the-
ories could travel to other contexts (i.e. newly consolidated democracies).
Having conducted the relevant empirical analysis, I am now able to address
the key question: Does the theoretical explanation of voting derived from the
experience of voters from established democracies fit well the emerging democ-
racies of sub-Saharan Africa, Latin America, East Asia, and Eastern Europe?
Drawing on the empirical results across the regions, this chapter looks sys-
tematically at whether the findings reported in the previous five chapters
are generally consistent with each other. This enables me to gauge how far
Western-derived models suit newly emerging democracies. Moreover, this
chapter also summarises and draw a broad conclusion from the analysis, in
light of which recommendations are made for future research. That said, the
major conclusion is that while some of the models are largely applicable in
10 Introduction
these emerging democracies, others are not. The approaches that perform best
are the mobilisation, socio-demographic, and, to a lesser extent, the alterna-
tive model, the various indicators of which significantly increase the chances
of an individual turning out to vote. On the other hand, our results across
the four geo-political regions offer some contradictory results with regards
to the impact of political-institutional variables (i.e. concurrent elections and
the closeness of election) and political-psychological variables (i.e. political
trust, internal and external efficacies) on voting. Overall, however, the bigger
picture in relation to voting in emerging democracies confirms that political
behaviour of voters or citizens in these regions is influenced by many of the
same factors that determine the electoral behaviour of citizens in established
democracies.
Notes
1 See Campbell et al. (1960); Campbell et al. (1954); Wolfinger and Rosenstone
(1980); Miller and Shanks (1996); Franklin (2004); Thomassen (2005); Leighley
and Nagler (2014); Lazarsfeld, Berelson, and Gaudet (1948); Berelson, Lazarsfeld,
and McPhee (1954); Blais and Daoust (2020); Heath, A. et al. (1993); Marsh et al.
(2008).
2 Cohen (2017, p. 1) observes that since the democratic transition in the 1980s in Latin
America, electoral democracy has been the status quo system of governments with
more than 100 (mostly) free, competitive, and fair elections for executive positions
having been held across the region.
3 Also see Carothers (2002) for a discussion on the transition paradigm.
4 In contrast to this minimalist definition of democracy, “maximalist” definitions
argue that the protection of civil liberties is necessary for democracy to flourish.
Dahl (1971, p. 7) theorised that inclusiveness, or public participation, and liberal-
isation, or public contestation, are key features of a democracy, or “polyarchy”.
5 For an overview of the current research on voter turnout at the aggregate level
see: Geys (2006); Blais (2006), Cancela and Geys (2016); Stockemer (2017).
6 While this single country or regional studies have produced excellent results,
their strength tend to lie in individual case studies rather than a truly thematic
comparison.
7 Bratton, Yun-han Chu, and Lagos’ (2010) article is one exception that examines
the “who votes” question across three regions of the global South (i.e. sub-Saharan
Africa, Latin America, and East Asia).
8 More on case selection and data justification and elaboration will be discussed in
Chapter 3.
9 See Thomassen (2000); Miller (1994) for an overview of some of the reasons for
doing comparative electoral research.
10 Two classical books that established the Columbia schools are: The People’s Votes
(Lazarsfeld, Berelson, and Gaudet, 1948) and Voting (Berelson, Lazarsfeld, and
McPhee, 1954). While The American Voter (Campbell et al. 1960) recognised the
primacy of the Michigan school.
11 In this book, although I tend to use the word new democracies, transitional dem-
ocracies, emerging or young democracies interchangeable, yet in this book, they
generally referred to countries that I classified as newly consolidated democracies.
Introduction 11
12 https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/
28468. [accessed 18/02/2019].
13 https:// s ites.google.com/ s ite/ j oseantoniocheibub/ d atasets/ d emocracy- a nd-
dictatorship-revisited.[accessed 18/02/2019].
14 Also, another shortcoming is that the data are limited to 2010.
15 Among comparativists, Freedom House scores or rating are generally considered
as the most common indicator to measure level of democracy.
16 Freedom House offers two indices of freedom: political rights and civil liberties,
which are then used to indicate “democracy”.
17 The criteria employed for selecting appropriate cases will be discuss in detail in
Chapter 3.
2
Why citizens vote
For many years, the question of why people vote has been of great interest
to political scientists but also to sociologists and political psychologists alike.
The need to provide an answer to this puzzle has generated the publication of
many articles, books, and handbooks,1 which are all aimed at throwing light
on this issue. This theoretical and empirical review of the existing literature
seeks to capture what has been written about the why and who votes questions
of electoral participation. As I show else where (see, Tambe, 2018a; 2017a),
the ‘’why vote?’’ question seeks to advance an explanation of people’s deci-
sion to vote, while the “who votes?” question deals with the characteristics
of those who vote and indeed, those who do not vote in elections. Though
these questions stand independently from each other, in an analytical sense,
in reality, they are closely dependent on each other: for instance, the social
characteristics approach that will be examined shortly, effectively addresses
both the why and the who questions at one and the same time. Therefore, in
this review, they will be dealt with simultaneously.
In reviewing the extensive literature on electoral participation, it can be seen
that Western scholars2 have dominated the debate so far in trying to explain
why people vote in elections. As Diwakar (2008, p. 76) rightly puts it, “empir-
ical literature on the determinants of turnout tends to focus on research in
Western democracies”. This chapter will therefore necessarily draw heavily on
previous studies of advanced democracies in Western Europe and the United
States of America; but, it also discusses relevant studies of electoral behaviour
in Africa, Latin America, East Asia, and Central/Eastern Europe. However,
before I move to these issues, the first part of this chapter is directed towards
the general features of the theories of voter turnout. Here, I discuss a number
of established theoretical and empirical approaches of turnout, summarising
the main assumptions and propositions of each, before moving on to state the
theoretical expectations that can be derived from these theories with regards
to electoral participation in newly consolidated democracies. In the second
part, I proceed by providing an alternative explanation of turnout in newer
democracies. This section also considers some of the findings from the dem-
ocratisation and social capital literature about electoral participation (i.e. how
central are elections to democracy in new democracies), followed with a brief
Why citizens vote 13
discussion on why and how newly consolidated democracies may differ from
established democracies in terms of individual-level voter turnout.
Contending theories –general features of the theories of

voter turnout
As briefly mentioned in the introductory chapter, not only have comparative
studies on voting been spearheaded by Western scholars but most import-
antly, the central question within the literature has been divided to capture
two very distinct levels that analysts wish to emphasise (i.e. individual and
aggregate level). At the individual level (which is the primary focus of this
book), the most important question centres on the following: why do people
vote or what influences people to participate in electoral politics the way they
do? On the other hand, the core question generally posed at the aggregate
level is the following: how can we account for the variations in voter turnout
across countries, within countries, and across geo-political regions (i.e. why do
some countries have high rates of voter turnout, whereas other have rates that
barely reach 50 per cent?). To provide answers to these questions, scholars
have typically relied on a range of contending theories that emphasise cultural
factors, resource (i.e. socioeconomic and demographic factors), institutional
factors, mobilisation, or rational choice.3 However, most of these theories
were largely developed based on the experience of voters from advanced dem-
ocracies, thus leading us to wonder whether their findings can be applicable
to newly consolidated democracies from other world regions. I will shortly
discuss these theoretical approaches to explaining individual-level electoral
participation. Next, I explore briefly why and how newly consolidated democ-
racies in sub-Saharan Africa, Latin America, East Asia, and Eastern Europe
may differ from established democracies in terms of individual-level voter
turnout. Moreover, I dedicate a few lines reflecting on the place of elections in
the democratisation process. How central are elections to democracy?
Theories of electoral participation

As briefly signposted in the opening section of this chapter, there has been
an extensive debate on the factors that are believed to affect people’s decision
of whether to vote or not to vote. Since the core objective of this volume is
to examine how well models or theories of electoral participation that were
initially developed in established democracies “travel” to newly consolidated
democracies, standing on the shoulder of two fundamental schools of
thoughts (i.e. Columbia and Michigan schools) that undergird most of the
theorising about vote choice and subsequent studies that precede it, I consider
three groups of factors as possible explanations of voting at the individual
level: socio-demographic, mobilisation, political-psychological, to this, I add an
alternative explanation. In the section that follows, I shall examine the general
features of these groups of theories, but most notably, I shall focus on the
14 Why citizens vote
extent to which these theories have been applied in newly consolidated dem-
ocracies. Nevertheless, before delving into these four theoretical perspectives
of vote choice, I find it necessary to note, the rational choice theory4 is per-
haps the leading paradigm for research on voting behaviour.5 To catalogue,
examine, and evaluate the field’s current understanding of the rational choice
approach with regards to explaining individuals’ voting decisions, one is com-
pelled to begin with the work of Anthony Downs, who is considered to be a
founding father on whose shoulders many subsequent scholars have stood.
The “calculus of voting” model was initially developed by Downs (1957) and
later extended by Riker and Ordeshook (1968). This model emphasises that
there is a cost–benefit calculus of voting whereby benefits should outweigh
costs in order for a person to turn out to vote. I can explain this model by
going back to an equation that was proposed by Downs.
R = (B) (P) – C + D
Where R is the total reward a citizen will gain from voting, B is the benefit
a person thinks he/she will accrue from having his/her preferred candidate
win, P is an individual perception of the probability that his/her single vote
will change the election outcome, C represents the cost of voting to the indi-
vidual, and D stands for the psychic satisfaction an individual would gain
from voting. The logical explanation of this equation is that if R is positive,
the citizen is assumed to gain a reward from voting that outweighs the cost
and will, therefore, participate in an election. As Harder and Krosnick (2008)
suggest, the more positive R is, the more likely an individual is to vote. A vari-
ation on this general approach has been proposed by Joshua Harder and Jon
A. Krosnick (2008) as follows:
Likelihood of voting = (Motivation to vote × Ability to vote) ÷ Difficulty

of voting
From this equation, it is clear that the decision of an individual of whether

to vote or not to vote can be explained as a function of his/her motivation
to vote, his/her ability to vote, and the difficulty of the act of voting (Harder
and Krosnick, 2008, p. 526). Motivation here is therefore seen an important
factor that determines an individual propensity to vote. This motivation to
vote may derive from a strong preference for one candidate over competitors.
The motivation to vote here replaces B in Downs’ equation. The motivation
to vote is also explained either from the pressure that a voter received from
friends/family or from the belief that voting is a duty. All in all, this equation
suggests that the more motivation or ability that a citizen has, the more likely
he/she is to turn out to vote. The ability to vote generally refers to a citizen’s
aptitude and competence to absorb information about political events and
candidates. The ability of voters to process this political information enables
them to form candidate preferences (Harder and Krosnick, 2008, p. 527).
Finally, in their equation, the difficulty of voting simply points to certain
conditions outside the voter’s mind such as strictness of voting procedures,
the degree to which polling stations are publicised, the physical closeness of a
person’s polling station to his/her home, the availability of information about
the candidates contesting the election and, lastly, the availability of time at the
disposal of the would-be voters.
Unfortunately, the rational choice approach is subject to well- known
criticisms. Riker and Ordeshook (1968, p. 28) argue that the chances of any
individual vote affecting the outcome of an election for nationwide public
office are virtually zero, even in a close election. That is, one ballot cannot
make any difference. This, therefore, implies that the calculus of voting in its
purest form does not seem to work. The obvious question to ask is: does this
make the model redundant? To address this problem, several extensions and
amendments to the model have been developed by rational choice theorists.
These amendments propose that individuals may decide to vote to maintain
democracy, out of a sense of duty, because they are risk-averse and wish to
avoid the regret of having not voted and seeing their preferred candidate lose
by one vote, because of the reasoning that other citizens will not vote and
that their own vote could become decisive, because the cost of voting is prac-
tically nil and, finally, because they find it rational not to calculate benefits
and costs when both are very small (see Blais, 2000). However, as Blais (2000)
explains at length, these additions, in reality, violates some of the fundamental
assumptions of the original rational choice model.
Socio-demographic model
Demographics factors6
Beginning with the demographic determinants of vote choice, two classical
works by the Columbia school (e.g., The People’s Choice by Lazarsfeld,
Berelson, and Gaudet 1948; and Voting by Berelson, Lazarsfeld, and
Mcphee, 1954) established the supremacy of the demographic model in
explaining individual-level voting behaviour. According to these scholars,
citizens decisions of whether to participate or not are based on sociological
groups. Based on this line of arguments, the authors gave weight to demo-
graphic characteristics like age, gender, place of residence, and religion as
determinates of vote choice. In line with these studies, I start my examinations
of the effect of the demographic factors on electoral politics by looking at
five key indicators: age, gender, religion, ethnicity/race, and place of residence
(rural–urban divide).
Age
Beginning with age, theoretical, and empirical findings in Western democ-
racies suggest that age is an important explanatory variable that can predict
people’s voting participation at the individual level. Drawing on earlier
studies (see e.g. Lane, 1959; Braungart and Braungart, 1986), two types of
mechanism can be used to explain how age influence political behaviour.
First, age depicts a life-cycle development, with different stages of the life-
cycle (i.e. young, middle-aged, and elderly) staging different life experience
that can influence the way people see and think about politics. Second, age
also tells of generational or cohorts” effects that can affect participation rates
(Plutzer, 2002; Cutler and Bengtson, 1974, p. 163; O’Neill, 2003; Smets, 2012;
Blais et al., 2004; Franklin, 2004; Gallego, 2009; Persson et al., 2013; Wass,
2007). Contrary to the life-cycle approach, which places more emphases on
the dynamics of biological ageing that makes citizens gain political experi-
ence or grow into political life, the generational effects focus more on social,
cultural, and historical events that shape participatory patterns. For example,
military coup, the legacies of authoritarian rule, first democratic elections,
and or first multi-party elections can be viewed as important events that can
have effects on people’s political thinking, thus influencing the way they are
involved in electoral politics.7 However, I argue making generational effects
certainly will be more difficult or hard to identify in a comparative context
(i.e. most notably as this book focus on four geopolitical regions), given that
the regions and even countries have experienced different events at different
times. That said, in this book, my focus on age differences in participation
centres more on the life-cycle approach.8
According to the life-cycle effects, political participation, most especially
voting, is said to increase with the age of an individual. What this implies is
that young people turn out in lower numbers than older citizens. Kleimhenz
(1995), cited in Bühlmann and Freitag (2006), theorised that younger people
are more likely to abstain from voting as part of a typical life-cycle that predicts
that levels of participation tend to rise over an individual life span, as people
experience certain life events. The life-cycle theory is based on the idea that
young people have a low attachment to civic life since they either stay in edu-
cation or they are trying to establish a career for themselves after completing
their education, or because of higher mobility, which makes them participate
less (Strate et al., 1989, p. 443; Jankowski and Strate, 1995; Kinder, 2006). The
combination of these two traits deprives these young people of vital polit-
ical resources needed to be involved in politics; that is, they possess few rele-
vant political skills or knowledge. Jankowski and Strate (1995, p. 91) suggest
that the absence of these resources makes political participation difficult and
meaningless to young people. However, the life-cycle theory is not linear, for
it also predicts that participation rates of the oldest age group will tend to
be low, with a number of studies attesting to the existence of a curvilinear
relationship between age and turnout in Western European countries (Bhatti
and Hansen, 2012; Konzelmann et al., 2012; Wass, 2007; Blais et al., 2004;
Leighley and Nagler, 2014; Norris, 2003; Stolle and Hooghe, 2005; Al-bacete,
2014; Franklin 2004; Melo and Stockemer 2014). Cutler and Bengtson (1974,
p. 163) argue that not only does the social disengagement of older people
from social networks lead to a lower attachment to political life and voting
but the fact that older people are influenced by health problems and declining
income due to retirement all greatly hinder their involvement in voting. Thus,
while young individuals and older people tend to participate less, the life-cycle
theory suggests that the middle-aged have higher participation rates. This can
be explained by the fact that this age group tends to be more settled; they have
experience, knowledge, and responsibilities –that is, the resources and oppor-
tunities to participate in politics.
Turning to new or emerging democracies, the effect of age on voting
decisions has been shown by empirical research to adopt a familiar pattern
similar to countries in the global North (see Potgieter 2013). For example,
individual country studies in South Korea confirm age as a significant factor
influencing vote choice at the individual-level (Mo, Brady, and Ro, 1991).
Moreover, Park (2002, p. 140) argues that in Korea younger people (mostly
in their twenties) show more cynicism about politics and politicians due to
their keen sense of alienation from electoral politics and therefore tend to
participate at a lower rate. Furthermore, a striking characteristic of East
Asian countries is the high geographical mobility among young people, which
makes it hard for them to participate in electoral politics, as they are more
likely to fail to register to vote in the district to which they have recently
moved. Similarly, for sub-Saharan Africa and Latin America, several studies
also revealed a positive and significant relationship between age and voting
(Bratton, 1999, p. 565; Tambe, 2017a, 2018a; Kuenzi and Lambright, 2011;
Carreras and Castañeda-Angarita, 2014; Carlin et al., 2015; Seligson et al.,
1995). Furthermore, there is a connection between age and socioeconomic
status, given the high level of unemployment among young people in the con-
tinent; not only does this deprive them of a vital resource necessary for elect-
oral politics but it also makes them more dependent on older adults for their
basic subsistence. In fact, in this book, I argue that youth as a period of tran-
sition to adulthood has been prolonged to a period of waiting or “waithood”9
as younger people and millennials all over the world (i.e. especially those in
Africa, Latin America, East Asia, and Eastern Europe) are finding it harder
to advance through life, move away from home, and established themselves as
independent. I, therefore, expect young people to be less likely to vote than
older people.
Gender
An evaluation of previous studies demonstrates gender to be a variable that
affects people’s voting participation. As Beauregard (2014, p. 617) explains,
gender differences in political participation are found throughout the litera-
ture on gender and political behaviour. To be precise, the difference in polit-
ical participation between men and women first received much attention in the
United States in the 1980s, with Wirls (1986) identifying four gender-gaps10
in political engagement: the participation gap, the policy opinion gap, the
electoral gap, and the partisan gap. Specifically, empirical studies all tend to
show men participate more than women. Blais, Gidengil, Nevitte, and Nadeau
(2004) suggest that men are more likely to vote compared to women, while
Inglehart and Norris (2003) state that women continue to lag behind men in
many forms of participation. According to Burns et al. (2001), this disparity
in voter turnout is explained by the fact that men tend to have a higher desire
to vote than women. However, this merely begs the question, why? Their
explanation is therefore incomplete and unsatisfactory. Schlozman, Burns,
and Verba (1994) move away from the simple assumption of women having
less desire to participate and argue instead that the gender gap in participa-
tion can be explained by availability and access to political resources.
First, women are considered to participate at a lesser rate because they
are disadvantaged or lack the resources that are needed to facilitate political
participation. Second, another line of argument consistent with women’s lack
of resources is based on the logic of social learning and social roles attached
to both men and women. Duverger (1955) argues that adult roles assigned
to gender lead to a situation where women are more likely to be placed in
the private milieu; that is, at home. The challenge of the overwhelming share
of the housework, coupled with childcare, which may be exacerbated when
combined with the need to undertake paid employment, gives women less
free time, which is translated into a reluctance to participate in politics. By
contrast, he argues that men are often placed in the world outside the home.
Though, concerning gender gap in electoral participation, it is essen-
tial to note that in established democracies, no gender gap in voter turnout
has been observed at least since the 1990s. For example, Poole and Zeigler
(1985) drawing on the US (i.e. for a similar result for Australia and Britain
see Studlar, McAllister, and Hayes, 1998) showed that women have closed
the gender gap in voting and are now actually voting in higher numbers.
Also, recent studies indicate that the gender gap seems to have reduced or
even reversed, with women consistently being more likely to vote than men
(Carreras, 2018; Norris, 2002; Kostelka, Blais, and Gidengil, 2019; Alexander
and Coffé, 2018). Several explanations have been advanced to explain the
important change in gender difference. Among them, the most prominent
has been the development theory of gender gap that postulates that signifi-
cant changes in the sex role in post-industrial societies has had a substantial
effect on women values (Inglehart and Norris, 2000). By this, I mean that
the younger generation of women are more likely to be educated and in paid
employment, which makes their situation more like that of men than is true
of older generations of women. This argument is supported by Norris (2002)
who shows that the gender gap in voting has closed in established democra-
cies, thanks to this interaction between gender and generations.
Having established the expected direction of the effect of gender on voting
in advanced industrial democracies, the question is, will we find a gender
gap in participation across our new democracies of interest? Analysis of the
gender gap in East Asian and Latin America countries (Bratton et al., 2010)
reveals that the applicability of gender-sensitive laws (i.e. compulsory pri-
mary education) and the introduction of inclusive institutions (i.e. compul-
sory voting, proportional electoral system, gender quotas) encourage female
political participation and thus has enabled women to attain parity with men,
thus narrowing the gender gap (Córdova and Rangel, 2017; Kittilson and
Schwindt-Bayer, 2012; Jones, 2009; Schwindt-Bayer, 2009; Tripp and Kang,
2008; Barnes and Burchard, 2013; Burnet, 2011; Desposato and Norrander,
2009). Given this, I, therefore, might not expect men to participate at a higher
rate than women in East Asia and Latin America. However, for Africa, the
adult roles attached to gender (which may be changing) generally lead to a
situation where it is quite common to see men dominating the political sphere
than women. Based on this theoretical and empirical discussion, I would
hypothesise overall that the propensity for men to participate in voting is
higher than that of women.
Religion
Another demographic factor that has been theorised to influence people’s
decision whether or not to vote is religion. Although the world has become
more secularised over the decades, Chesnut (2003) and Broughton and Napel
(2000) argue that religion has always played an essential role in shaping elect-
oral politics both in established democracies and other parts of the world (i.e.
Latin America). Nedeau et al. (2017, p. 30) cite in many European countries
the link between religion and voting is direct and is driven from the establish-
ment of parties based on religious principles (i.e. many Christian democratic
parties). However, as pointed out by other studies from North America (the
US and Canada) the relationship between religion and voting is rather indirect
(e.g. for studies in Canada see Blais et al., 2002; for the US see Lewis-Beck
et al., 2008) where some parties do tend to attract more religious voters than
others. To be precise, relating to the two cases from North America, although
these parties are not created based on religious principles, yet they do tend to
attract voters or members due to the position that they take concerning reli-
gious incline issues such as abortion or same-sex marriage.
Moving away from established democracies, Bratton (2003) has advanced
the view that people of faith are globally more likely to vote than those who
say they are non-believers. The obvious explanation of this phenomenon
could be that churches, mosques, and temples act as agents of mobilisation
(e.g. see Segura and Bowler, 2012; Harris, 1994; Jones-Correa and Leal, 2001).
While I would expect the mobilisation effect of religious attendance to come
true in Eastern Europe, Latin America (i.e. Latin Americans and Eastern
Europeans have historically been heavily if not exclusively Catholic11) and to
some extent Africa countries where religiosity in these regions often affect
voters through parties emphasising the importance of issues that are often
debated in the public sphere.12 However, in East Asia, there is a mixed result
with regards to the role of religion in influencing citizens’ voting decisions.
Rood Steven (2002, pp. 157–158) shows that, surprisingly, religion had little
effect on who voters preferred in the Philippines (rather than whether or not
they actually voted). For example, Fidel Ramos won the 1992 elections despite
been a Protestant in a profoundly Catholic country, while Joseph Estrada won
the 1998 election despite being heavily criticised by the Catholic hierarchy on
his alleged immoral lifestyle. On the other hand, Chin (2002) reveals that in
Malaysia religion is an important factor in influencing how people vote, with
vote choice being between a secular modernist Islam and the more orthodox,
fundamentalist brand of Islam. Overall, I can plausibly expect individuals
who are more religious to have a higher propensity to vote.
Ethnicity or race
Ethnic identity is the inclination of an individual to define themselves in
terms of their cultural origin and descent-based traits (Olorunsola, 1972;
Horowitz, 1991; Berman et al., 2004). In line with this argument, Wolfinger
(1965, p. 896) suggests that ethnic identity or origin represents one of the
strongest dimensions of people’s perception of themselves and of others
and is therefore considered to play a very important role in politics. Based
on these findings, I move forward with the argument that ethnicity is one of
the most essential variables in the demographic model, but the question that
arises is how does ethnic identity/race influence people’s propensity to vote in
an election? When ethnic identity becomes sufficiently strong it can consti-
tute the basis for forming political opinion and in stimulating people’s voting
decisions; this, of course, usually leads to what some experts term “ethnic
voting”. Wolfinger (1965) explains ethnic voting as a situation where members
of a cultural group tend to show disproportionate affinity at the polls for a
particular political party. Because of the strong affinity that people share with
others, parties, interest groups, or other social groups may take advantage of
this and therefore mobilise individuals who then tend to vote as a bloc.
Posner (2005) stipulates that the logic of ethnic voting can be explained by
the fact that when people from the same descent are mobilised, they tend to
turn out to vote since they believe they are expressing their solidarity with their
group. This act of group solidarity might elevate leaders from their own ethnic
background to positions of power, most especially in the political executive,
thereby gaining collective representation. As some scholars have highlighted,
this can lead to election outcomes that are mere headcounts of ethnic groups,
that is, if turnout is high and if all voters chose parties associated with their
own communal identities then an election can resemble an ethnic “census”
(Lever, 1979; Horowitz, 1985; Ferree, 2006; McLaughlin, 2008). In many
African, Latin America, and East Asian countries, it is common for people
to show a strong affinity with a particular ethnic or tribal group. Because
of this strong affinity that people share with others, parties may take advan-
tage of this and therefore mobilise individuals who may then tend to vote as
a bloc. In sum, drawing these studies, it might seem that ethnicity has two
dimensions relevant to turnout.13 One is the nature of the particular ethnic
group (i.e. whether they have a distinct propensity to participate in the elect-
oral process). The other is the degree to which ethnicity is a political cleavage
in a given country. So, people from group A may vote more because people of
group B are mobilised to vote more. In new democracies, particularly those of
sub-Saharan Africa, this is likely to be an interesting line of analysis. Based on
this, I would, therefore, expect people who have strong ethnic attachments to
have a high propensity to vote.14
Place of residence
Looking at the last demographic variable, I consider the individual place of
residence (i.e. the distinction between rural and urban voters). Earlier studies,
at least in the context of established democracies, suggest that citizens in
rural areas were historically more likely to engage in electoral politics as a
result of the higher level of associational or informal network.15 However, a
survey of recent studies, in Western democracies, indicates that scholars are
still undecided if rural or urban voters tend to participate more. For example,
Karp et al. (2007) argue voters in urban or cites will be more likely to engage
since they are more attractive locations for parties to canvass. On the other
hand, Hoffmann-Martinot (1994) refutes the idea that turnout will be higher
in urban or cities centres by arguing that urbanisation tends to reduce inter-
personal bond and social networks thus making it less likely for people to par-
ticipate. To my knowledge, although the effect of rural and urban divide has
not been thoroughly examined in the context of newer consolidated democra-
cies, yet in a study that focuses on the determinants of non-voting across new
democracies in Central/Eastern European countries and sub-Saharan Africa,
Tambe (2018b) provides evidence that depicts non-voters were mostly those
in urban or cities areas. Similarly, in a stimulating study of individual-level
voting in Latin America, Carreras and Castañeda-Angarita (2014, p. 1092)
reveal that place of residence tends to influence electoral behaviour, with Latin
Americans who live in urban areas being less likely to vote compared to rural
dwellers.16 Turning our focus on the East Asia region, individual studies (i.e.
Mo, Brady, and Ru, 1991) and comparative studies (Bratton et al., 2010) con-
clude that rural voters are more likely to participate compared to the urban
dwellers. Based on this initial finding, I would expect voters in rural areas will
be more likely to participate than urban/city dwellers.
Socioeconomic factors
Having concluded with the demographic predictors of turnout, I now turn to
the socioeconomic dimension. This model centres on individual resources as
a potential explanation of citizens’ voting propensity. That said, researchers
have bundled together groups of variable relevant to personal resources,
which are said to influence people’s decision to vote at the individual level.
These variables consist of income, education, and/or social status. From its
inception, the socioeconomic/resource model has stood the test of time in
predicting voters’ decisions in Advanced Western democracies, in that each of
these factors has widely been found to be significantly associated with voting.17
How do these socioeconomic/resource variables dictate why and who votes?
To answer this question, I begin with the classic study of Sidney Verba and
Norman Nie. In their seminal work Participation in America, Verba and Nie
(1972) built their explanation of electoral participation on what they called
a “Baseline Model” consisting of income, occupation, and education. Later
renamed the Resource Model, this approach came to dominate explanations
of the individual turnout decision in the United States and elsewhere. They
suggest that specific resources are vital for there to be effective participation
of the citizenry in the political life of their societies: time, money, and civic
skills. First, two prime resources for investment in political participation are
time and money. A simple baseline condition for political involvement, espe-
cially more involved modes of participation, is the availability of time (see
Caramani, 2008). In democratic societies, individuals use time in the service
of political action in many ways: working in a campaign, writing a letter to a
political official, attending a community or party meeting, and, finally, during
voting. It is often argued that people who do not work, or who do not work full
time, tend to display higher levels of political participation (Brady et al., 1995,
p. 273). On the other hand, money can, of course, be donated to candidates,
parties or innumerable political organisations or causes. The major critique
here is that time and money differs in significant ways as resources. In com-
parison with money, time is both constrained and more equally distributed –
everyone has only 24 hours in a day. The upper limit on money is, of course,
much less constrained and the differences among individuals can be much
larger (Brady et al., 1995, p. 273).
Second, the next resource is civic skills. Civic skills here refer to those
communication and organisational capacities that are so essential to pol-
itical activity. Citizens who can speak, write well, or who are comfortable
organising and taking part in meetings are likely to be more effective when
they get involved in politics (Brady et al., 1995). The acquisition of civic skills
begins early in life, at home, and especially in school. This process can con-
tinue throughout adulthood. Adult civic skills relevant for politics can be
acquired and honed in the non-political institutions of adult life: the work-
place, voluntary associations, and churches. All these undertakings represent
opportunities in non-political settings to learn, maintain, and improve civic
skills. Drawing on the work of Verba and Nie, recent scholarship on electoral
participation has tended to group several variables that jointly correspond to
socioeconomic status (SES). As noted by Plutzer (2018, p. 78) socioeconomic
status, or SES, is not a variable but, rather, a catch-all term that encompasses
a wide range of potential characteristics. SES consists of an individual’s edu-
cational level, income, and/or occupational status. These variables are said to
play an important role in understanding people’s political behaviour, most
notably as it concerns their voting decisions. From a theoretical perspec-
tive, a high level of SES is considered to lead to a higher propensity to be
involved in electoral politics. This is because people with a higher level of
education and income are believed not just to possess more time, but also to
have the capacity to understand and comprehend the daily political debates
and discussion. Various scholars (Verba, Nie, and Kim, 1978; Wolfinger and
Rosenstone, 1980; Leighley and Nagler, 1992; Buhlmann and Freitag, 2006)
have empirically demonstrated that the socioeconomic status of an individual
is positively associated with electoral participation at the individual level.
Looking at the relationship between education18 and turnout separately, the
weight of evidence shows that in advanced democracies those with a higher
level of education are more likely to turnout (see e.g. Schlozman, Verba, and
Brady 2012; Verba, Schlozman, and Brady 1995; Wolfinger and Rosenstone
1980) with La Due Lake and Huckfeldt (1998) attesting the positive link
between education and political participation as “one of the most reliable
results in empirical social science”. Hillygus (2005) but also Gallego (2010)
hint about the fact that education is a primary driver of increased political
participation is “largely uncontested”. Putnam (1995), on the other hand,
posits education is the best individual predictor of participation. Also, in their
meta-analysis of individual-level research on voter turnout that drew mostly
from studies of Western democracies, Smets and van Ham (2013) show that
education is indeed positively related to individual-level turnout with most
studies having a success rate of 67 to 72 per cent. Relatedly, in his review of
education and political participation, Persson (2015) reaffirms that education
is of central importance in most studies of political behaviour, with studies
finding individuals with higher education participating at a higher rate. In
discussing the role of education on political participation, Persson (2015,
p. 690) demonstrates at length three main linkages that explain how educa-
tion is related to political participation: absolute education model, the pre-
adult socialisation model, and the relative education model. Beginning with
the absolute education model, Persson argues education has a direct causal
effect on political engagement. What this implies is that education tends to
increase civic skills, political knowledge, and/or political efficacy, which acts
as the causal mechanism triggering participation.
Next, concerning the pre-adult socialisation model, Persson contends edu-
cation works as a proxy for variables such as family socioeconomic status, pol-
itical socialisation that takes place in the home, and/or personal characteristics
such as cognitive ability, which are considered to affect participation rather
than education itself. Finally, the last linkage that exists between education
and political participation centres on the relative education model or sorting
model. According to this perspective, there is an indirect effect of educa-
tion on political participation, which can be explained via social status19 (see
e.g. Nie, Junn, and Stehlik-Barry, 1996). Within the literature, social status
is defined as having a central social network position in society, with Nie,
Junn, and Stehlik-Barry (1996) operationalising social network centrality as
proximity to governmental incumbents and political actors who make public
policy and to those in the mass media who disseminate. Drawing from this
perspective, it is, therefore maintained it is social status and not education that
increases participation.
Next, if I turn to the relationship between income and turnout, findings
by Teixiera (1992) suggest individuals with a low level of income are less
likely to vote compared to their wealthier counterparts, which also raise the
question why? Drawing on their now seminal work: Who Votes? Wolfinger and
Rosenstone (1980, p. 20–22) outline a theoretical and empirical connection of
why income tends to be associated with higher turnout, which is summarised
as follows:
• The poor have less time to devote to matters not essential to everyday
existence.
• Wealthy people have jobs that tend to increase their political engagement.
• Income affects one’s social context; thus, the wealthier are more likely to
be exposed to social networks with norms of civic duty.
• Lastly, wealthy individuals have a more significant stake in the political
system.
There is considerable evidence to support the socioeconomic/resource model

in the context of Western democracies.20 Yet, when this model is transported
to non-Western context, the results are even less clear, mixed, or directly
opposite to what conventional wisdom will expect. Studies of voting behav-
iour in Asia, Latin America, and Africa suggest that poor people participate
politically no less than more well-off citizens (Yadav, 2000; Krishna, 2002,
2008; Ahuja and Chhibber, 2012 on India; Booth and Seligson, 2008 on Latin
America; Isaksson, 2014 on Africa and Bratton et al., 2010, on East Asia).
For example, in a fascinating study conducted by Yogendra Yadav (2000,
p. 133) in India, it is shown that poorer people vote in higher numbers than
their more affluent peers. His findings reveal that “in 1990, a participatory
upsurge among the socially underprivileged, including scheduled castes and
tribes to the point that turnout of the lower orders of society was well above
that of the most privileged groups”.21
Similarly, in her innovative study of the role of resource in explaining pol-
itical participation, Isaksson (2014) provides empirical evidence showing the
resource model does a poor job in explaining political participation considering
that resource-poor Africans were significantly more likely to participate than
their resource-rich counterparts. Education, on the other hand, is found to
have mixed findings across new democracies. For instance, while education is
found to be positively associated with participation in India (Krishna, 2002,
2008), Latin America (Booth and Seligson, 2008; Carreras and Castañeda-
Angarita 2014; Carlin, Singer, and Zechmeister, 2015), and post-Communist
European countries (Orvista et al., 2005; Kostelka, 2014; Tambe, 2018b) but
not in East Asia and Africa. In East Asia, findings by Bratton et al. (2010,
p. 119) but also Tambe (2016, p. 169), reveal that people without formal edu-
cation are significantly more likely to vote.22 This suggests we might not expect
to find a strong relationship between education and voting in East Asia.
Concerning sub-Saharan Africa countries, individual country study on
the determinants of political participation in Zambia by Bratton (1999)
confirms education has no significant effect on participation. Likewise,
Croke et al. (2016) individual country study in Zimbabwe finds that edu-
cation decreases political participation, reducing the likelihood that better-
educated citizens vote or participate in other forms of political activities such
as contacting and meeting. Also, Blaydes (2011) drawing on the Egyptian
case, reports less-educated citizens are more than twice as likely to vote as the
educated.23 A few reasons why the socioeconomic status does not have much-
predicted power is that wealthier voters may want public goods and more
programmatic policies, which may be hard for parties to credibly promise to
deliver these goods, leading wealthier voters to abstain from voting (Nathan
2019). Also, as Croke et al. (2016) argue, it might be simply connected to
“deliberate disengagement”. That is, the educated may choose not to vote to
avoid legitimising flawed elections or because they believe there is no point
in voting. Furthermore, Kasara and Suryanarayan (2015) show rich voters
may vote more when the potential for redistribution is high or when the state
can tax.
Regardless of the mixed findings across the different geopolitical regions,
the theoretical assumption on which I will proceed in this book is that the
more a person climbs the educational ladder, and the higher they find them-
selves in terms of socioeconomic status, the more politically active they will
tend to be by virtue of their acquisition of civic skills. I, therefore, expect a
higher level of socioeconomic status (i.e. education, income, and/or occu-
pation) to be positively associated with higher propensity to vote. Table 2.1
indicates the overview expectation of the socio- demographic model on
vote choice.
Table 2.1 Overview of the expectation of the socio-demographic model on vote choice
Variables Expected effect for all regions

(AF, LEC, EA, and CEEC)
Socio-demographic Age +
model Gender (male) +
Religion +
Ethnicity/race +
Residence (rural) +
Education +
Note: Explains what variables within the four regions that are expected to take a positive/negative
sign. * AF (Africa), * LEC (Latin America), * EA (East Asia), and *CEEC (Central and Eastern
Europe).
Table 2.1 summarises this part of the discussion so far. Studies in advanced
Western democracies have so far dealt with six main socio-demographic
characteristics that explain people’s decision whether to vote or not to
vote: age, gender, religion, ethnicity/race, residence, and socioeconomic status
(i.e. education or income). Based on these studies, the standard hypothesises
that can be derived with regards to each of these variables are as follows:
H1a: Beginning with age, the older an individual is, the higher is his or her
propensity to turnout.
H1b: As concerns gender, I would expect that men will have a higher pro-
pensity to vote than women.
H1c: Individuals who are more religious will have a higher propensity
to vote.
H1d: Concerning ethnic identity, people who have strong ethnic
attachments will have a higher propensity to vote.
H1e: I would expect voters in rural areas will be more likely to participate
than the city or urban dwellers.
H1f: Moving to the socioeconomic status of an individual, the higher an
individual’s socioeconomic status (i.e. education and or income) the
higher is his or her propensity to vote.
Mobilisation model
The mobilisation model stipulates that the only way to explain political
participation is to move beyond the world of individuals to include social
networks, politicians, parties, activists, and interests groups (Rosenstone and
Hansen, 1993, p. 23). I have already discussed a number of demographic
influences that pertain to the socio-demographic model (age and gender), but
other sociological factors relate to the idea that people’s propensity to vote is
influenced by their associational life and this is what the mobilisation model is
concerned with. The central logic here is that social networks such as unions,
pressure/interest groups, and parties exert pressures on people to behave as
members of a group rather than as isolated individuals. In making people
act as groups rather as individuals, not only are these social networks able
to mount pressure on their members that enable them to comply, but these
networks have the potential to reduce the cost of political participation by
providing group members with political information and resources that are
necessary for electoral participation.
Apart from Rosenstone and Hansen, several other works have highlighted
the importance of the mobilisation model in getting out the vote. Uhlaner
(1989) found that group identification can make it more rational for par-
ticular segments of the electorate to show up at the polls even when the
cost-benefit incentive structure clearly predicts that they should not do so.
Further, the politicisation that occurs within parties, interest groups, and
social groups might increase the likelihood that an individual will vote
(Brady, Verba, and Schlozman, 1995; Leighley and Nagler, 2007). Based
on the theoretical literature, I can highlight the following variables that
will be empirically tested in this book. The variables are grouped into three
main categories: party identification, associational networks/civil society
organisations, and clientelism.
Party identification
Campbell, Gurin, and Miller (1954, pp. 88–89) define party identification as a
“sense of personal attachment which the individual feels towards the party of
his/her (sic) choice”. Previous studies have demonstrated the utility of party
identification as an important variable in understanding and explaining pol-
itical behaviour, especially voting in Western electoral politics (Dalton, 1984,
p. 264). According to this view, elections in most democracies are almost always
partisan contests, which therefore means that partisan attachments guide citi-
zens through the complexities of politics, providing vital information that
is necessary for making an informed electoral decision (see Brader, Tucker,
and Duell, 2013). Strong partisan identifiers are generally more involved in
electoral politics than non-identifiers (Dalton, 1984, p. 267). Partisanship,
therefore, helps in performing a mobilisational function;24 it is this mobilising
function that I am interested in, on the assumption that it might influence
individuals to vote at the individual level. King (1969, pp. 123–128) but also
Verba et al. (1978) have previously linked party identification to turnout.
These studies assume that attachment to a political party draws an individual
into the political process. In short, it has been shown that people with weak
or no party identification at all will show a lower propensity to participate in
elections, while those individuals with a strong party attachment or identifica-
tion possess a higher propensity to vote in elections.
Yet, concerning East Asia, studies of voter behaviour conducted in Hong
Kong and South Korea found party identification to be very weak (Scott,
1996; Park, 2002). Newman (2002, p. 57) argues that weak party identifica-
tion in Hong Kong could be explained by the fact that political parties are
all relatively new and subject to considerable change and weak consolidation
over a short period of time, and also because the Hong Kong population
now includes a large component of immigrants from the Mainland where
the cultural norm is to avoid politics. Given that parties in East Asia, Africa,
Latin America, and Eastern Europe are new and fluid, it may be that I will
not find a positive relationship between party identification and propensity to
vote at the individual level, as the comparative literature on Western democ-
racies would suggest. At the aggregate level, it is certainly possible that par-
tisan identification will not drive turnout strongly. It should also be said that
the aggregate-level mobilisational impact of partisan commitment may be
diminishing even in established democracies. There is now extensive empir-
ical literature that tends to show declining party memberships in Western
democracies (Van Biezen, Mair and Poguntke, 2012). For example, tracing
party membership and the extent of election organisation across UK con-
stituencies, Fisher et al. (2006, pp. 4–6) indicate a steady decline in member-
ship within the Conservative, Labour, Liberal Democrats, Plaid Cymru, and
SNP from 1992 to 2001.
Similarly, there is ample evidence of the weakening of partisan identifi-
cation across the established democracies (Dalton, 2000). This declining
membership and identification, coupled with the declining cost of obtaining
political information thanks to an information explosion that has been made
possible through the expansion and growth of mass and electronic media,
gives one cause to question whether party identification is still a viable explan-
ation of the voting decision at the individual level. That is, as party mem-
bership declined, so has partisan identification –and this should mean that
a significant source of electoral mobilisation has been eroded. From this
perspective, it is no surprise that electoral turnout has declined as well in
Western democracies. These developments are all consistent with one other.
Nevertheless, at the individual level, where people retain a strong partisan
identity, this may be expected to boost the propensity to vote. Thus, I hope to
find a positive association between party identification and propensity to vote.
That is, the more strongly an individual develops party ties, the higher will be
his or her propensity to vote.
Civil society organisations

Since Tocqueville (1994), there are those who have argued that clubs and
organisations can play an important role in the social and political integration
of individual. Almond and Verba (1963) thought of voluntary organisations
as the most essential foundation of stable democracy. In this regard, mem-
bership in a social organisation can encourage greater turnout by either
motivating or helping would-be voters by increasing their civic skills. This,
therefore, means that the more a person is involved in a cooperative network
with others, the more appealing casting a vote may appear to such persons in
the group. The reason why people who are involved in a voluntary or social
organisation participate more in electoral politics is the solidarity or “esprit
de corps” that exists within such groups.
Harder and Krosnick (2008) postulate that when members of a particular
social group identify strongly with their group, they developed a group con-
sciousness that may increase their decision to vote. Miller et al. (1981) add
that civil society organisations not only act as effective mediums of political
socialisation but also as a school for democracy since they bring together the
roles of group members and citizens. In a nutshell, people who are engaged
in social or voluntary groups develop solidarity through social bonding and
could, therefore, have higher motivation to participate in voting simply because
they share concern with similar issues that affect their group. Moreover, the
strong connection between group members gives them access to skills that
enable them to participate in politics.
Studies conducted in Latin American countries –that is, Argentina, Chile,
Mexico, and Peru –reveal that the involvement of people in non-political
organisations leads to greater participation in politics (Klesner, 2007), while
empirical evidence from sub-Saharan Africa equally indicates that member-
ship in voluntary organisations has a positive impact on people’s propensity
to turnout (Kuenzi and Lambright 2011; Tambe, 2017a). On East Asia, Scott
(1991, p. 152) hints that social network communications are the most essen-
tial factors in shaping voting behaviour in Japan. Other works (Curtis, 1971;
Flanagan and Richardson, 1977) also stress the role of social networks as
an important factor in predicting vote choice. In the case of Japan, the cru-
cial role of professional and community organisations that candidate and
personal support groups turn to has been highlighted. Based on this, I, there-
fore, expect countries with a higher membership of trade unions and or other
informal associations to reveal a higher level of participation. For this reason,
I hypothesise a positive association between close associational membership
and the propensity to vote in new democracies.
Clientelism
The practice of clientelism is relatively frequent in emerging democracies
(i.e. particularly those of Africa, Latin America, and East Asia) and is gen-
erally used by politicians and political entrepreneurs seeking for public office
to attract voters. Clientelism typically revolves around a network of direct
relationships between citizens and the patron (i.e., politicians) involving the
exchange of targeted goods or gifts in exchange for their support. Or as cited
by Kramon (2017, p. 7), clientelism tends to take on many different forms
ranging from politicians targeting resources directly to voters to sometimes
through direct vote-buying. Hicken (2011) argues the practice of clientelism
appears to be one of the main ailments affecting the quality of democracy
or democratic governance. First, clientelism tends to run counter to democ-
ratisation given that it does impede the development of political institutions,
undermines democratic consolidation by limiting the exercise of citizenship
(see e.g. Graziano, 1973; Fox, 1994) and as cited by Stokes (2005) but also Lyne
(2007) it endangers the secrecy of the vote. Second, Keefer (2007); Kitschelt
et al. (2010); Singer (2009); and Persson, Tabellini, and Trebbi (2003) argue
clientelism tends to be associated with corruption as it implies illegal practices
of vote-buying and encourages politicians to adopt illegal fundraising strat-
egies. In this volume, although the core objective of clientelism is generally to
alter the electoral results in a way that suits the patron, I argue that clientelism
does work as a tool of electoral mobilisation.
For instance, as cited by Tambe (forthcoming) in new democracies, particu-
larly those in Africa, the resource-poor compared to rich tend to have little
or no contact with the state or political parties except for the period leading
to elections. During these election periods, it is not uncommon to find polit-
ical parties, party heads, representative, and members of parliaments going
Table 2.2 Overview of the expectation of the mobilisation model on vote choice

Mobilisation Party identification +

model Associational membership +
Clientelism +
Europe).
back to their local communities to canvass local support. Thus, as pointed

by Nichter (2008), the existence of a strong clientelistic network is likely to
increase the incentive for citizens to go to the polls. In a context where cli-
entelism is widespread, I will expect voters not to base their voting decision
on government performance or party ideologies but rather on the personal
benefit they obtain from supporting a particular patron. Based on this,
I would expect citizens immersed in clientelistic networks to have a higher
probability of voting than the voters who are not. The variables to be tested
are once again are presented in Table 2.2.
In Table 2.2, my discussion of the mobilisation model generates the
following hypotheses:
H2a: There will be a positive association between party identification and

propensity to vote. That is, the more strongly an individual develops
party ties, the higher will be his or her propensity to vote.
H2b: Individuals who are involved in social organisation or are members
of political associations tend to have a higher propensity to vote.
H2c: I would expect voters that are targeted by material benefit will be
more likely to vote.
Political culture/political-psychological model

Political culture is seen as one of the most powerful concepts to emerge in
political behaviour research and is central to the study of citizens’ attitudes
(Dalton and Dieter, 2009, p. 6). It offers an insightful approach to daily polit-
ical events by identifying the underlying psychological forces that shape much
of civic life (Rosenbaum, 1975). Political culture has been defined either from
the individual or collective perspective depending upon the level at which
we want to study political life. From the individual perspective, political cul-
ture entails all the important ways in which a person is subjectively oriented
towards the essential elements in their political system. That is, we want to
know what individuals think and feel about the symbols, institutions, and
rules that constitute the fundamental political order of their society and how
they respond to them. From the collective perspective, on the other hand,
political culture seeks to understand the collective orientation of a group or
community of people towards the basic elements of their political system, its
institutions, and officials. This, therefore, means that when political culture is
discussed, it usually refers to the mass political orientation towards the whole
political system (Rosenbaum, 1975, pp. 3–4).
“Civic cultures” and political participation

Almond and Verba (1963, p. 26), in their seminal study The Civic Culture
defined political culture as consisting of cognitive, affective, and evaluative
orientations towards political phenomena that may be distributed across
national populations or subgroups. These combined in various ways, such
that three major cultural orientations –parochial, subject, and participa-
tory –could be identified. Each carried implications for the level and style of
political participation that citizens typically undertake. Beginning with the
parochial culture, Bennett (1986, p. 8) suggests a parochial person is an indi-
vidual who sees no relevance for himself and his family to what goes on in
the public realm since what the government does is irrelevant to his personal
existence. Such citizens have little or no cognitive understanding of the pol-
itical system. In countries where this type of culture is dominant, the citi-
zens tend not to expect anything positive from the government, nor do they
expect to participate in politics, since politics is considered as a domain of the
elites. Moreover, the government in such a polity is seen just as an enforcer
of rules and consequently the realm of politics is considered to be avoided at
all cost and means. In their analysis, Almond and Verba considered Mexico
as a perfect example that illustrates this culture. How does this culture relate
to electoral politics at the individual level? In a parochial society, I would
expect turnout to be very low since the awareness and expectation of what is
expected from the political system or government is low.
The next type is the subject culture. In societies with this culture, citizens
are believed to possess only cognitive orientations towards the output aspect
of the political system. Citizens in such societies, therefore, expect positive
“things” (policies) from the government, but they do not tend to be politically
active themselves. Italy and Germany were classified under the subject cul-
ture. The difference between the subject culture and parochial one discussed
above is that individuals in the subject culture are considered to have high
levels of awareness and expectation about the political system, but their par-
ticipation rate is low since they also consider politics, most especially voting,
as a domain of the elites. The implication of this culture with respect to voting
decisions is that people’s propensity to vote will be very low as long as the pol-
itical system or government meets the expectations of the citizenry.
The last of the cultural orientations is the participatory one. In soci-
eties where this culture predominates, individuals tend to have very high
governmental expectations and personally participate in electoral politics.
The participatory culture is considered as a central democratic feature and
portrays citizens as highly inclined to participate since they are well informed
of the input and output aspects of the political system. In their typology,
Almond and Verba include both the United States of America and the United
Kingdom as countries that have this participatory culture. Almond and
Verba’s works25 have been subject to various criticisms over the years –for
instance, it is difficult to assign countries a given cultural profile since this may
change and evolve over time, but their seminal approach provides this study
with a potentially useful cue in examining factors affecting electoral turnout.
Beer (1974, pp. 26–27) built on Almond and Verba’s work to argue that
political culture orients a people towards a polity and its processes, pro-
viding it with a system of belief, that is a cognitive map, a way of evaluating
its operations and set of expressive symbols. From the above-mentioned
definitions of political culture, it is clear political culture deals with individual
subjective orientations towards the basic elements of their political system.
This sounds like an accurate but not a satisfactory definition, as it is necessary
to define more specifically what the key subjective orientations are. This is no
easy task to undertake, as Rosenbaum (1975, p. 4) argues, scholars themselves
have never reached a consensus on the proper components of political cul-
ture. Some analysts begin by including in political culture all politically rele-
vant orientations including cognitive, evaluative, or expressive ones. This is so
unbounded that a researcher would have to spend an interminable time com-
piling an “elephantine list” of orientations so as to be sure nothing relevant is
left unnoticed Rosenbaum (1975). On the other hand, other analysts attempt
to make the list more manageable by limiting political culture to orientations
towards political institutions, which sounds like a useful strategy but, as
Geertz (1963, p. 212) argues, it may lead to omitting other dimensions of pol-
itical life that are instrumental in shaping the fundamental political order of
a society. But it is good to stress the abundance of definition need not mean
intellectual disorder since, in the proliferation of ideas about what constitutes
political culture, one can distil a set of common items on which most scholars
would agree. This point is supported by Lucian W. Pye (1966, p. 19) who
suggests that political culture must be limited to the attitudes, beliefs, and
sentiments that give order and meaning to the political process and provide
the underlying assumptions and rules that govern behaviour.
Similarly, Rosenbaum (1975) insists that whenever analysts offer any list
of what political culture is, they are, in effect, using a rule of thumb: those
dimensions of an individual’s thoughts, feelings or behaviour that are linked
to the creation and maintenance of a society’s fundamental political order
belong under the label “political culture”. In effect, this book relies on Almond
and Verba’s three core elements of cultural attitudes: orientation towards
governmental structures, orientation towards others in the political system,
and orientation towards one’s own political activity. Orientation towards
governmental structures is concerned with regime orientation and orienta-
tion towards government input and outputs. Orientation towards others
in the political system deals with issues relating to political trust, political
identification and rules of the game, and, finally, orientation towards one’s
own political activity deals with political competence and political efficacy
(Rosenbaum, 1975, pp. 6–8). From these orientations, I derive the following
variables: political interest, political knowledge political efficacy, political
trust, and social trust. The empirical connections between these factors
and people’s decision of whether to vote or not vote in newly consolidated
democracies will be examined in this book. These variables can be viewed
as cognitive characteristics that are expected to function as resources and
therefore increase people’s decision to participate. That is, in brief, it can be
hypothesised that the higher an individual’s political interest, political effi-
cacy, political trust, and/or social trust the higher their propensity to vote.
Political interest and political knowledge

I begin with political interest/knowledge as the first political culture variables
I intend to test empirically. Van Deth (1989, p. 278) defines interest as the
degree to which political activities or politics arouse citizens” curiosity.
However, I would argue that both political interest and knowledge should be
grouped in explaining people’s voting decisions. The rationale for grouping
these two variables is because research has proven that political interest
provides an incentive to acquire political knowledge Smets (2010). However,
as Brady et al. (1995, p. 271) suggest, someone who has political knowledge
is more likely to be interested in politics as well, or better, individuals who
are interested in political debates and who have enough political knowledge
to understand what is at stake are more likely to participate in politics. Do
political interest and knowledge help to explain people’s voting decisions at
the individual level? Classical studies undertaken by Lazarsfeld et al. (1948)
succeeded in demonstrating that political interest and political knowledge are
key explanatory variables for individuals” voting decisions. In this respect,
the lack of political interest and political knowledge is presumed to either
increase or lower people’s propensity to vote in elections. If individuals are
interested in politics, the logic is that such people will be involved in more
political debates either with friends, family, or colleagues, and the result is
that they will more likely to vote (Sheerin, 2008). On the other hand, a limited
amount of interest in politics and a lack of knowledge about political debates
will lower citizens” desire to participate in voting. Based on this logic, I expect
having a higher interest to be positively associated with a higher propensity
to vote in new democracies of Africa, Latin America, East Asia, and Eastern
Europe.
Political efficacy
Political efficacy is described as the perception citizens have of being cap-
able of acting effectively in the political arena. Hadjar and Beck (2010) define
political efficacy as the degree to which a person believes in his or her ability,
first to understand politics and second to influence the political processes
through action, including voting. Political efficacy comes in two parts: internal
efficacy generally refers to an individual’s competences, skills and resources to
deal with politics, while external efficacy is defined as an individual’s percep-
tion that political institutions are responsive to people’s attempts to exert pol-
itical influence (Clarke and Acock, 1989, p. 552). Studies by Campbell et al.
(1954), Abramson and Aldrich (1982), and Becker (2004) have shown that
political efficacy is the single most important variable in the psychological
model that tends to influence the decision whether to vote or not to vote. This,
therefore, means that widespread political efficacy throughout the electorate
will lead to a higher voter turnout, while the lack of political efficacy among
voters will lead to lower voter turnout. This line of argument is supported by
empirical studies by Craig and Maggiotto (1982), Karp and Banducci (2008),
and Norris (2002) who demonstrate that more efficacious voters are, the more
likely they are to vote. Consequently, I would hypothesise that voters in young
democracies who are politically efficacious will reveal a greater propensity to
participate in elections.
Political and social trust

Trust in political institutions has been theorised as a psychological vari-
able that generally affects people’s involvement in political activities. In this
regard, trust deals with citizens” evaluations of political institutions, political
entrepreneurs, and the political system. Political or public trust in govern-
ment has been assessed by the extent to which citizens have confidence in
public institutions to operate in the best interest of the society and its citi-
zenry (Cleary and Stokes, 2006; Kim, 2005; Thomas, 1998). In this book,
I will focus on citizens” trust in the political system and how it influences
their propensity to vote. Miller and Listhaug (1990, p. 358) and Bühlmann
and Freitag (2006) suggest that trust in a country’s political system reflects
one’s evaluation of whether or not the political authorities and institutions
at the national or local level are performing in accordance with one’s nor-
mative expectations, as people always expect a certain amount of perform-
ance from public institutions with regard to transparency, accountability and
integrity. Ragsdale and Rusk (1993), and Pattie and Johnston (2001) argue
that if citizens do not trust the political system, then the likelihood he or
she will participate in voting will be lowered. Putnam (2000) provides a more
robust theoretical backing for this assumption, arguing that trust is the main
component of a democratic society. He stipulates that people will only vote if
they are confident that the political system is responding in some way to their
voting behaviour.
In Latin America, individual countries studies in Bolivia (Smith, 2009),
Chile (Carlin, 2006), and Costa Rica (Seligson, 2002) indicate that citizens
with a high level of trust for democratic institutions are more likely to engage
in electoral politics. Based on these considerations, the theoretical propos-
ition that I would put forward with regards to political trust is that the higher
the political trust, the higher the propensity to vote. At the aggregate level,
it might be that low political trust weakens turnout the new democracies of
East Asia; using survey data from the East AsianBarometer covering Japan,
Philippines, South Korea, Taiwan, and Thailand, Chang and Chu (2006)
demonstrate that citizens” perception of corruption is inversely related to
their trust in government institutions. However, this aggregate-level finding
still rests on the hypothesised positive relationship between trust and voting
at the individual level. But is trust in the political system the sole type of
trust that can influence people’s voting decision? How about inter-personal
or social trust? That is, does having trust in individuals within a given society
in any way influence a citizen’s decision to participate? Social trust is defined
as the extent to which a person is confident in and willing to act on the basis
of the words, action, and decision of another (McAllister, 1995, p. 25). Cox
(2003) and Timpone (1998) find that individuals who have trust in others are
more likely to vote, while those who do not are less likely to vote. They argue
that distrustful people may be less interested in the act of voting since they
perceived the political system and other individuals as corrupt. It is possible
that social trust is generally lower in countries (e.g. some in East Asia, Africa,
and Latin America) that are multi-ethnic, multi-religious, or multilingual.
The implication of this is that voters may be divided along these lines and
therefore inclined to distrust each other. Regardless of this, I still expect that
in general terms the greater an individual’s sense of social trust, the higher his
or her propensity to vote.
Although these orientations (political trust, social trust, interest, and effi-
cacy) are useful in explaining why people decide to vote, they have, nonethe-
less, come in for strong criticism. Almond and Verba (1963, p. 519) remark
that these political culture orientations are implicit and often unconscious
in an individual life –so basic that the person hardly reflects on them. In
this sense, many are primitive orientations because they are so implicit and
taken for granted that each individual holds them and believes that all others
hold them; they become unstable assumptions or postulates about politics.
Regardless of these aforementioned criticisms, I nevertheless argue that
given the newness of democracy in these emerging countries, it is impera-
tive for scholars to study how political culture/political-psychological might
explain electoral behaviour in these societies. So, from this discussion of pol-
itical culture effects, I derive the expectations already alluded to –that I will
find positive associations between political interest, trust, efficacy (external
and internal), and the decision to vote in newly consolidated democracies.
Table 2.3 shows the variables that I have considered for empirical testing.
Our theoretical discussion of the political culture or political-psychological
model has revealed the following variables as essential in explaining people’s
decision to vote or not to vote. As shown in Table 2.3, these variables consist of
political interest, political efficacy, and political trust. The literature surveying
Table 2.3 Overview of the expectation of the political-
psychological model on
vote choice

(AF; LEC; EA, and CEEC)
Political-Psychological Political trust +

Model Political interest +
Political efficacy +
Europe).
people’s decision to vote in Western democracies has so far enabled us to for-

mulate the following hypothesis with regards to each of these variables:
H3a: The higher an individual’s political interest, the higher is his or her
propensity to vote in an election.
H3b: The higher an individual’s sense of internal or external efficacy, the
higher will be his or her propensity to participate in electoral politics.
H3c: The greater an individual’s trust in political institutions, the higher
will be his or her propensity to vote.
The electoral or political-institutional context

The previous section has focused on a discussion on why citizens tend to
engage in electoral politics by relying on individual-level indicators drawn
from three theoretical perspectives (i.e. socio- demographic, mobilisation,
political-psychology). However, an important body of literature on electoral
participation points to the fact that people’s decisions whether to vote or not
does not only depend on individual-level factors but also on the context in
which individuals find themselves (Gallego, 2015; Franklin 2004; Dalton and
Anderson, 2011; Dalton et al., 2011; Thomassen, 2005; Evans and de Graaf,
2013; Anderson and Singer, 2008). Thus, in this volume, a key point that
I consider is how the political-institutional context will explain differences in
levels of voter turnout in newly consolidated democracies. Since the lists of
contextual, country, and/or political-institutional factors that could theoret-
ically condition voting behaviour is extensive,26 before discussing the effects of
particular institutional variables, I will introduce some influential accounts of
institutional explanations.
A survey of institutional explanations

I begin with the work of Powell, who argues that turnout tends to be higher
in those countries where the government assumes responsibility for voter
registration than in those where it is up to the citizen to get registered on the
electoral list. Moreover, turnout is (not surprisingly) higher in countries where
voting is made compulsory by law (Powell, 1982, p. 116). Crewe (1981, p. 251)
agrees with Powell about the effect of compulsory voting laws while adding
that the type of electoral system also has an effect, in that turnout is higher in
proportional systems. Jackman (1987, p. 45) found turnout to be related to five
institutional variables: compulsory voting, nationally competitive districts,
electoral disproportionality, multipartism, and unicameralism. Turnout is
higher when voting is compulsory, in unicameral countries and more pro-
portional systems, and lower as the number of parties increases. Blais and
Carty (1990, p. 15) found turnout to be higher with compulsory voting when
the population is smaller and under proportional electoral systems. Black
(1991, p. 101) agrees that compulsory voting is the most significant variable,
although he found other institutional effects to be weak, including the type
of electoral system, degree of disproportionality, or unicameralism. Franklin
(1996, pp. 36–50) shows that turnout is affected positively by the degree of
proportionality of electoral outcomes and by the presence of compulsory
voting, postal voting, and Sunday voting and negatively by the number of
polling days. Salient elections give rise to some 30 per cent greater turnout
than non-salient elections. This is because when an election is considered
essential and many issues are at stake, people are more inclined to partici-
pate in such elections. Compulsory voting increases turnout by about 6–7 per
cent in countries that make voting obligatory. Postal voting is worth another
5–6 per cent, and disproportionality is worth about half a percentage point
in turnout for every percent by which the distribution of seats in the legis-
lature approaches proportionality with the distribution of votes (Franklin,
2002, p. 158).
Finally, Birch (2010) has produced a fascinating study that examines the
relationship between the perception of electoral fairness and people’s deci-
sion of whether to vote or not. Central to her study is the argument that
citizens” perceptions of the competitive context in which elections take
place matter (Jackman and Miller, 1995; Blaise, 2000; Franklin, 2002). This
research has broken important ground in conceptualising the competitive-
ness of elections, especially in terms of the decisiveness of the contest and the
choice on offer. More specifically, her unique contribution is to identify and
examine the degree to which an election is perceived as being a fair contest
and the impact of this upon turnout. Exploring the possible consequences
of the perception of electoral misconduct for voting decisions, the argu-
ment she proposes is that when voters perceive that an election will be free
and fair, they are more likely to vote. On the other hand, when voters have
reservations about the ability or willingness of those conducting the election
to maintain electoral integrity, people’s propensity to vote will be less. These
hypotheses are put to the test in an analysis of 31 elections across the globe
and her results confirm that, if voters fear that polls are corrupt, they are less
likely to cast their votes.
Institutional factors
Having surveyed some of the relevant literature that focuses on the polit-
ical and institutional factors, it is possible to identify a set of variables that
I intend to test empirically. They are as follows:
Electoral system
According to Farrell (2001, p. 4), the electoral system is defined as “the means
by which votes are translated into seats in the process of electing politicians
into office”. A good number of comparative studies have succeeded in
showing that the electoral system impacts on political participation, particu-
larly as it relates to turnout (Franklin, 2002; Norris, 2002). The argument here
is that cross-national differences in turnout can be explained by the type of
electoral system that is adopted by a country. What kind of electoral system
is theorised to positively affect citizens’ desire to participate? The literature
suggests that proportional electoral systems (i.e. PR systems) are most likely
to have a positive effect on voters’ turnout, as opposed to majoritarian or
plurality systems (Blais and Aarts, 2006), with Blais and Carty (1990) pro-
posing this is because PR systems tend to reduce distortion in the conversion
of ballots into seats. Essentially, proportional systems tend to enhance the
likelihood that a single vote might contribute to the election of a winning
candidate, so there is a greater incentive to vote. Similarly, Ladner and Milner
(1999) argue that majoritarian electoral systems tend to hinder electoral par-
ticipation for two reasons: First, supporters or voters of a previously defeated
party may either stay away from the election since they consider their vote to
have no effect on the election.
Second, they argue that majoritarian electoral systems tend to hinder voting
because previously defeated parties may reduce their mobilisational efforts.
With regards to these considerations, therefore, the theoretical proposition
which I will make is that we expect proportional representation to enhance
people’s propensity to vote. This is supported by the fact that PR systems
encourage citizens to vote since they know that votes transfer more directly into
party representation and they thus perceive the system as fair (Endersby and
Krieckhaus, 2008, p. 602; Stockemer, 2015). Based on this reasoning, I hypothe-
sise a positive relationship between the level of proportionality of an electoral
system and the propensity of voters to participate in elections. However, taking
into consideration that a bulk of these studies were conducted in Western
democracies, we cannot be sure that in new democracies we are still going to
find such as strong correlation between the electoral system variable and voter
turnout, given the newness of the electoral systems in these countries.27
Compulsory voting
Birch (2009, p. 2) defines compulsory voting as the legal obligation to
attend the polls at election time/day. Studies in Western democracies have
categorically established compulsory voting as having a positive effect on
people’s voting decisions. Blais (2000) suggest the effect of compulsory voting
on turnout is one of the robust findings in studies that study cross-national
variation in voter turnout.28 Though compulsory voting clearly increases
voter turnout, most especially in countries where it is enforced, little is known
with regards to the type of sanction that is applied to non-voters. The uncer-
tainty about the exact influence of compulsory voting exists because research
findings on the contribution of this variable usually have mixed results. One
brand of research reveals turnout to be high in countries where compulsory
voting exists and is complemented with sanctions. However, another group of
studies shows that in countries where voting is compulsory by law, but where
the law is not enforced by sanctions, voter turnout is equally high since the
duty to vote is internalised by the citizens (Hadjar and Beck, 2010).
Concurrent elections
Concurrent elections is another contextual factor that will foster people’s pro-
pensity to vote. Cox and Munger (1989) contend that the possibility of many
elections taking place at the same time will arouse greater media coverage in at
least one of the elections, which invariably boosts the amount of money spent
on campaigning and as such increases the amount of information and public
awareness about the issues at stake in these elections. Additionally, Filer and
Kenny (1980) and Carter (1984) suggest concurrent elections tend to increase
people’s propensity to vote given that the cost of going to the polling booth is
fixed and unrelated to the number of elections in which a voter needs to cast
a vote. I, therefore, expect people would participate more in countries where
multiple elections are held concurrently than in countries where they are not.
The closeness of the election

Blais (2000) contends that among the institutional variables that are said to
influence voting decisions at the individual level, the closeness of the election
has produced the most consistent results and findings. The approach assumes
that the closer an election appears to be prior to the election day, the more
likely people will turn out to vote. The rationale behind this is that voters vote
in such elections since they believe their vote is more likely to make a diffe-
rence to an election outcome. By contrast, in a situation where an election
outcome is regarded as “predetermined” –that is, a foregone conclusion
–such a contest will lead to a high degree of abstention and apathy from
voters who will perceive that their vote has no influence in such an election.
As argued by Bühlmann and Freitag (2006), in a close competitive election an
individual’s propensity to vote is higher since the weight of each individual
vote is believed to increase with the supposed marginality of an electoral out-
come. Thies (2002) contends that the closeness of an election tends to drive
citizens to participate in Japan as political elites (party and factional leaders)
increase their efforts at mobilising people to go to the polls. Furthermore,
Table 2.4 Overview of the expectation of the institutional model on vote choice

Political/institutional Disproportionality -
model Compulsory voting +
Closeness of the election +
Concurrent elections +
Europe).
scholars such as Cox and Munger (1989) and Shachar and Nalebuff (1999)
have succeeded in showing empirically that campaign efforts by parties and
candidates are usually focused more on districts where outcomes are expected
to be close. This means that there is an added mobilisational effect in such
contests as well. Each of the variables discussed in the previous paragraph
might be considered a factor relating to the institutional context of the polit-
ical system in question: proportional electoral systems increase the propensity
to vote; the more competitive an election, the higher the propensity to vote
in such countries; countries having compulsory voting will produce a higher
propensity for people to vote and individuals have a higher propensity to vote
when legislative and presidential elections are held concurrently. I would,
therefore, expect such factors to come into play in Africa, Latin America,
East Asia, and Eastern Europe. Table 2.4 summarises the variables and the
expected effect across the four regions.
H4a: Disproportional electoral systems decrease the propensity to vote.

H4b: The more competitive an election, the higher the propensity to vote.
H4c: Individuals have a higher propensity to vote when legislative and
presidential elections are held concurrently.
H4d: Countries having compulsory voting will produce a higher propen-
sity for people to vote.
Democratic theories, founding elections, social capital, and electoral

turnout in new democracies
Elections and the democratisation process: how

central is election to democracy?
The third of wave of democratisation process saw a large number of countries
across Africa, Latin America, East Asia, and Eastern Europe organising com-
petitive multiparty elections for the very first time, with citizens allowed to
choose their preferred leaders in presidential, legislative, and local elections.
Drawing on this development, minimalist and procedural definitions of dem-
ocracy lay strong emphasis on the process of election, viewing democratisa-
tion as a process whereby uncompetitive authoritarian regimes give way to
regimes based on the choice of leaders.29 As cited by Lindberg (2006a, 2006b),
following the onset of the third wave of democratisation, elections have been
hailed as the hallmark of democracy. More notably, the initial contests or
elections that took place across new democracies have been described by
scholars of the transition paradigm as founding elections (O’Donnell and
Schmitter 1986). Bratton (1998) argues these elections were regarded as
founding elections since they marked for these countries or regions a tran-
sition from an extended period of authoritarian rule to fledging democratic
government. Thus, the event during this period is said to have given birth to
democracy considering that ordinary citizens across Africa, Latin America,
East Asia, and Central/Eastern Europe for the very first time were allowed to
vote and choose their governmental representatives.
The comparative literature on democratisation has focused on the effects
of elections on the overall democratisation process, with scholars such as
Lindberg (2006, 2009); Edgell et al. (2017); Teorell and Wahman (2017);
and Flores and Nooruddin (2016) contending regular elections helped to
promote democratisation. To be precise, the democratisation by elections
postulate that the mere holding of elections even if flawed can over time
lead to democracy, as each election is thought to advance democracy in a
process of small incremental steps or as Teorell and Hadenius (2009) put
it, a history of elections will over time pave the way for an eventual con-
solidation of democracy.30 However, some analysts have expressed doubt
about the role elections play in the development and consolidation of dem-
ocracy. Carothers (2002) has been very sceptical regarding the true value
of elections for democratisation. For example, in declaring the end of the
transition paradigm, he contends that elections are in and of themselves
largely insignificant to democratisation. Relatedly, Terry Karl (1986) warns
us about the fallacy of electoralism, namely that elections alone –how-
ever free and fair –do not or cannot guarantee a democracy but rather
other institutions such as a constitutional rule of law, independent legisla-
ture and judiciary, and civil society are a prerequisite for any functioning
democracy.
To avoid the charge of the electoral fallacy, many scholars shifted their
procedural emphasis from elections to broader issues of the rules laid down
for the functioning of the polity. Founding elections were a key but insuffi-
cient condition of democracy; that is, democracy would only be adjudged
consolidated after a series of elections,31 when institutionalisation of formal
and informal institutions had secured the “reproduction of the minimalist
procedure” and when all major political actors have accepted the demo-
cratic rules of the game. As mentioned by Millard (1999, p. 2), adopting a
proceduralist definition of democracy appeared to make comparative ana-
lysis easier but attempts to generalise about markedly different countries
also led to doubts as to how far the comparison could be taken. Moreover,
democracy as a theoretical concept has long been the subject of deliberation
and debate.
As contended by Ellis et al. (2006), the way in which democracy is defined
speaks not only of the nature of government, election, and power but also of
the nature of society, community, and humanity. Additionally, these scholars
hint that, in the last century, the concept of democracy has not only evolved
through many incarnations but, most importantly, it has been invoked as
the basis for a utopian society, a means to protect citizens from government,
a tool for the advancement of human development, a form of free-market
politics, and a way to ensure stability in a pluralist society. Even so, viewed
strictly as a tool or mechanism, representative democracy is defined only as
a system of government in which people participate in electoral politics in
order to appoint those who will act as their representatives and thus deter-
mine national policies. The act of voting provides all citizens with a direct
interest in the actions of their government and an opportunity to partici-
pate and to inform themselves and others. The results of the vote give those
elected the entitlement to govern or to oppose and oversee those who govern,
until the next election, and the responsibility to decide political issues on
behalf of their constituents. Voting then becomes the key form of interaction
between those elected and the ordinary citizen, as it provides the fundamental
building blocks for the operation of the rest of the democratic system (Ellis
et al., 2006, pp. 11–12).
Other theorists, however, contend that the standard form of the late
twentieth-century democracy has emerged as something of a paradox –uni-
versal suffrage ensuring political equality within the system of society that
emphasises individual rights and an economy that retains very real inequal-
ities of class, opportunity and affluence (Macpherson, 1977, pp. 78–92). This
form of democracy, they contend, is the very cause of the crisis exposed by
widespread voter alienation and falling voter turnout. Moreover, writers and
participants in democracy-building processes argue that rather than just being
an instrument of control or influence, politics must be viewed as a way of
living and social decision making. Democracy thus becomes a moral concept,
a way of life. Political participation also needs to be expressive, allowing a
citizen to feel like a participant of the process. In this view, democratic partici-
pation is conceived of as a vehicle to build both citizenship and community.
Substantial citizen participation in government decision-making –or rather,
a form of direct or deliberative democracy –is necessary to achieve a more
equitable and humane society (Macpherson, 1977, pp. 94–105). Clearly, dem-
ocracy is seen by some as a multi-layered and dynamic system with practices,
procedures, and attitudes entrenched in a complex interaction of political,
economic, social, and cultural dimensions; but this complexity makes com-
parison. Democracy consists of many qualities that emphasise the centrality
of the demos and the importance of mechanisms of political participation,
responsiveness, and accountability while stressing individual rights as part of
the substance of democracy, not just as necessary instruments of free elections
(Millard, 1999, p. 2).
Nevertheless, it is hard to disagree with the view of Ellis et al. (2006) that,
despite their differences, all theories of democracy maintain a place for voting
as a fundamental component. Elections remain the primary basis of public
influence within the representative democracy, provide a form of collective
decision-making as to who manages the affairs of government and who over-
sees them doing so, and provide some degree of popular control over elected
politicians. As a form of political participation, voting generally remains the
easiest to undertake and the one that most commonly engages a majority of
citizens (Teixeira 1987, pp. 3–4). From the numerous definitions of democracy,
I can conclude that democracy cannot and will not exist without the actual act
of voting. This takes us back to the role of founding elections. How do these
founding elections inform us about electoral participation in these regions?
O’Donnell and Schmitter (1986) stress the importance of founding elections
in either stimulating or depressing turnout, by arguing that under authori-
tarian rule pressure for political participation, most especially voting, would
build irresistibly overtime only to be discharged at the first elections after the
transition to democracy. These scholars, therefore, provide an explanation for
why turnout is expected to be extraordinarily high during these first elections,
but also why it tends to decline at the aggregate level thereafter, as the current
scholarship demonstrates. Building on these earlier works, recent scholarships
in the last decades on new democracies (see e.g. Kostelka, 2017; Kostadinova,
2003; Kostadinova and Powel, 2007; Pacek et al., 2009) shows that voter
turnout is strongly affected by the process of democratisation and democratic
consolidation. For instance, Kostelka (2017) demonstrates that, depending on
the type of democratic transition, voter turnout in the founding elections may
be strongly boosted and then declines back to the normal level. In sum, the
conventional wisdom suggested by these studies on why democratic consoli-
dation tend to depress electoral participation can be explained by the fact that
by the time democracy is consolidated voters have become disillusion with
democracy or are apathetic in the face of reduced electoral stakes.
Social capital theory

Having presented some of the findings from the democratic literature
pertaining to electoral turnout, it is also worth mentioning that a number of
scholars working in the social capital tradition have emphasised its import-
ance to building democratic norms and engagement. As we shall see in due
course, moreover, the concept of social capital has particular relevance to
the mobilisation model of electoral participation. Those working in this
tradition have generally emphasised the importance of a strong and active
civil society for the consolidation of democracy by providing a wide range
of empirical evidence that suggests that the quality of public life and per-
formance of its institutions are indeed influenced by norms and networks of
civic engagement. Interest in social capital was revived by Putnam’s ground-
breaking works (i.e. Making Democracies Work, 1993; and in Bowling Alone,
2000) in which he argues or claims that rich and dense associational networks
facilitate the underlying conditions of interpersonal trust, tolerance, and
cooperation, thus providing the social foundation for a vibrant democracy
(Norris, 2002). Putnam defines social capital as “connections among individ-
uals –social networks and the norms of reciprocity and trustworthiness that
arise from them”. As cited by Norris (2002), Putnam’s definition of social
capital can be understood both as a structural phenomenon, that is social
networks, and as a cultural phenomenon (i.e. social norms). In this book, my
aim is to understand how these social networks (i.e. associational networks
and memberships that are assumed to facilitate social trust) influence people’s
decisions whether to vote or not.
Theories of electoral participation: are newly consolidated

democracies different?
In this book, the core assumption is that classic Western-based explanation of
turnout might not be suitable for explaining vote choice in new democracies.
This assumption is supported since most of the theories of voter turnout were
developed mainly based on the experience of voters from advanced democ-
racies (i.e. United States, Canada, and Western Europe). Thus, it might be
that new democracies may differ from established democracies in terms of
individual-level turnout as substantial differences in the electoral context in
new democracies might condition how well classic models’ maps/fits into the
dynamics of vote choice. So, in this section, I move to show why and how new
democracies should or might differ. With the newness of elections, I propose
three distinct features that might distinguish the political and electoral con-
text of new democracies from those of older democracies of Western Europe
and North America: weak institutions (party identification and party system),
a fragile economy, ethnic and cultural diversity32.
First, beginning with the weak institution, new democracies (i.e. espe-
cially those in Africa, Asia, Eastern Europe, and Latin America) are generally
considered to have weak party identification or shallow roots (see Bustikova
and Zechmeister, 2017), which could explain by the fact that political parties
are all relatively new and have been subject to considerable change over a
short period. Birch (2018, p. 159) attributes the weak party identification in
Eastern European countries, which I expect to be true for other newer democ-
racies, to the fact most parties competing for the political fora or space were
created well after the majority of those eligible to participate have achieved
the age to vote. Thus, it is therefore not surprising that party identification
in these regions should be weak. Moreover, the weak parties and voters’ ties
can also be explained by the high rates of scepticism and declining trust in
parties in general.33 Consequently, because political parties do not have deep
roots in these regions, this may allow charismatic leaders to mobilise voters
out of nowhere (Nadeau et al., 2017, p. 5). Similarly, there is also an add-
itional factor that contributes to the weak ties between parties and voters in
these regions, one that has to do with the weak party system institutionalisa-
tion. Party systems or parties in new democracies do not always have deep
roots and are not fully institutionalised as such voters in these regions often
find it challenging to understand what contending parties stand for or what
they will do if elected in office. As contended by Bustikova and Zechmeister
(2017), the accumulation of electoral experience has not been enough to bring
about party system stability and institutionalisation. In the face of weak party
system institutionalisation, Carlin et al. (2015, p. 4) argue, politicians might
tend to activate socio-political cleavages, which may lead to some voters
connecting to these politicians based on group identities (particularly, class,
religion, or ethnicity).
Second, the fragility of the economy of new democracies might equally
make individual level-turnout differ. Although Eastern European countries,
East Asia, Latin America, and, to a lesser extent, sub-Saharan Africa have
made substantial progress in economic development and growth, most coun-
tries in these four geopolitical regions are still considered as the poorest in the
world, with modest income and growing levels of socioeconomic exclusion
or inequality when compared to those in older established democracies. For
example, in a cross-examination of countries’ gross national income (GNI)
per capita drawn from the World Bank data for 2017, I observe the GNI per
capita for most new democracies34 were less than $35,000 in comparison with
those of established democracies like the United States and/or Germany
whose GNI per capita equals $55,000. Relatedly, drawing on the Varieties
of Democracies (V-Dem) index of exclusion by socioeconomic group, the
data suggest socioeconomic exclusion35/inequality is quite a problem and
is increasing across the globe, particularly in emerging democracies of sub-
Saharan Africa and Latin America.
Lastly, there is considerable ethnic and linguistic diversity that could
make voters of new democracies look quite different when compared to
voters from older democracies in North America and or Western Europe.
To be precise, East Asia, Latin America, and Africa regions are not only
the largest in the world but are also the most heterogenous featuring diverse
societies, separated along ethnic lines, multiple languages, histories, class,
and culture. Scholars have often aggregated these different forms of social
cleavages under the collective term of ethnicity.36 That said, I argue that if
the notion of the clan or ethnic identities become exacerbated, it may lead
to a situation that candidates for national office in new democracies may
be elected in large part based on their ethnolinguistic and or regional ties.
Although these three distinct features or illustrations could make voters in
newly consolidated democracies look “different” when compared to those
of the global North (North America and Western Europe), I do not think
voters in new democracies would be fundamentally different from voters in
older democracies. Similarly, Bustikova and Zechmeister (2017) argue that
although new or young democracies may differ in fundamental ways, elect-
oral behaviour in these newer contexts also tend to share significant simi-
larities with those of established Western democracies. To be precise, and as
I will demonstrate shortly, voters in new democracies may tend to equally
respond or be influenced by many of the same factors that determine the
electoral behaviour of citizens in established democracies.
Electoral participation in Africa, Latin America, East Asia, and

post-communist European countries
Even though voting is essential in new democracies, the reasons why people
choose to vote in these countries and or regions are still poorly understood.
As we have observed, research on electoral participation at the individual
level is largely focused on Western established democracies. Nonetheless, a
few empirical studies have examined electoral participation within a single
region in new democracies, and it is useful to outline these in order to assess
what is known so far.
Beginning with the African case, Kuenzi and Lambright (2011) focus
on the question “Who Votes in Africa?” Drawing on evidence from the
Afrobarometer survey data to assess the determinants of voting for ten
African countries, they argue that agencies of mobilisation, most especially
identifying with a political party, are among the most important predictors of
voting. Individual attitudes, such as support for democracy, also influence the
individual decision of whether to vote. They also find that two institutional
variables, that is, type of electoral formula and concurrency of presidential
and legislative elections, have positive effects on voter turnout. Next, Bratton,
Bhavnani, and Chen (2012), also employing Afrobarometer (Round 3) survey
data for 16 African countries, suggest that would-be voters consider policy
performance, especially the government’s perceived handling of unemploy-
ment, inflation and income distribution. Moreover, their study highlights that
an adequate account of the intended vote choice in Africa also requires rec-
ognition that voters attempt to associate themselves with prospective winners
since they wish to gain material benefit and evade or escape retribution
after the election. Similarly, Tambe (2017a), relying on the Afrobarometer
survey data (Round 5) re-assessed the determinants of voting across eight
sub-Saharan African countries that are considered as liberal democracies (e.g.
Gahan, Benin, Botswana, Senegal, South Africa, Namibia, Cape Verde etc.),
with his result asserting people’s decision of whether to vote or not to vote
is affected by various political and sociological factors (i.e. electoral system,
partisan, age, and associational networks). Additionally, Bratton (2013) in
his book titled: Voting and Democratic Citizenship in Africa, describes and
analyses voting across a range of African regimes with the results suggesting
African voters are motivated by social identity, partisanship, and economic
performance when choosing their elected officials. Also, Isaksson (2014)
examines the role of individual resources in explaining political participation
drawing on the Afrobarometer data for 27,000 respondents across 20 African
countries, with her results showing the resource model does a poor job in
explaining political participation considering that resource-poor Africans
were significantly more likely to participate than their resource-rich. However,
although these are interesting studies, the authors sometimes draw inferences
from elections that were considered as clearly flawed in the eyes of national
and international observers, such as the case of Zimbabwe, Uganda, and/or
Togo, etc. Also, most of the countries referred to in these studies have a rating
of 3 and above on their political rights scale according to Freedom House
ratings, which casts doubt on the fairness of these elections and the credibility
of the electoral statistics such as turnout figures.
For the Latin America region, it is essential to point out that with regards
the return of democracy in Latin America in the 1980s (i.e. third wave), it is
appropriate to say Latin Americans’ experience with elections, political, and/or
voting behaviour now spans a period close to four decades.37 Notwithstanding
the importance of elections and or civic engagement to representative dem-
ocracy and democratic consolidation to the overall region, until recently (i.e.
partly because of the absence of cross-national public opinion survey) very
little is still known about individual-level turnout or how Latin Americans
decide their votes. Moreover, although the study of elections38 in this region
can be regarded as exciting and vibrate, most of what we know about turnout
in this region has been guided purely by aggregate-level data or analysis that
focuses on the country/institutional context. Carlin, Singer, and Zechmeister
(2015) and also Nadeau and colleagues (2017) note the relevant research on
individual-level behaviour in Latin America is rare when compared to system-
level aggregate studies (for an overview of these studies see: Fornos, Power, and
Grand, 2004; Lavezzolo, 2008; Ochoa, 1987; Perez-Linan, 2001; Kostadinova
and Power, 2004; Schraufnagel and Sgouraki, 2005; Benton, 2005; Schwindt-
Bayer, 2009; Johnson and Ryu, 2010). Equally, there have been some discrep-
ancies in the analysis of voting behaviour across the region with priority given
to single cases drawn mostly from Brazil or Mexico39 as opposed to regional
comparative studies.
Relatedly, and as pointed by Nadeau and colleagues (2017), large nations40
such as Brazil and Mexico have been studied more as observed by survey-
oriented studies (studies in Brazil, see Ames 2001; Baker, Ames, and Rennó,
2006; for Mexico, see Domínguez and McCann, 1996 and Moreno, 2003).
Thus, considering that cross- national studies of electoral behaviour and
or vote choice have not only been guided by aggregate-level studies and
single case studies dominated by large countries, some studies have already
examined the question of vote choice in this region. First, Carreras and
Castañeda- Angarita (2014) employing data from the AmericasBarometer
covering 17 Latin American countries evaluate the determinants of turnout
at the individual level through a series of logistic and hierarchical models,
with their results highlighting the importance of socio-demographic factors
(i.e. age and education) and mobilising networks (associational networks) as
strong predictors of electoral participation. Furthermore, Carlin, Singer, and
Zechmeister (2015), in their edited volume titled The Latin America Voter,
incorporate innovative statistical methods and an elite survey to show that
to a varying degree, citizens’ decision to participate across the different Latin
America countries is influenced by class, religion, and other demographic
factors. Their results also indicate the significance of adequately understanding
the context in which these elections take place to provide a clearer picture of
who the Latin American voter is. Relatedly, in a more recent book by Nadeau
and colleagues (2017), relying on the Michigan model, the authors examine
voting behaviour across 18 Latin American countries using survey data from
the AmericasBarometer. Their results suggest, like pretty much elsewhere,
voters in Latin America are influenced by long-term (i.e. social class, parti-
sanship, and political ideology) and short-term factors (e.g. economy, crime,
and corruption) in deciding who and how they cast their votes.
Turning to the East Asia regions, although research on voting behav-
iour is still developing, it is important to reference Fuh-Sheng Hseih and
David Newman’s edited volume How Asia Votes (2002). In this book, sev-
eral contributors attempt to improve our understanding of elections across
ten Asia- Pacific countries (Cambodia, China, Indonesia, Japan, South
Korea, Malaysia, Philippines, Singapore, Taiwan, and Thailand). This group
includes some very different regimes in terms of the freedom and fairness of
elections. Even so, although the results are mixed across the different coun-
tries, a number of factors have been pinpointed as important in influencing
people’s propensity to turn out, including age, associational networks, religi-
osity, ethnicity, and the competitiveness of elections. Also, Tambe’s (2016)
analysis of the “who votes” question across eight East Asian countries
(Japan, South Korea, Taiwan, Mongolia, Hong Kong, Philippians, Malaysia,
and Singapore) concludes institutional variables (i.e. closeness of elections,
parliamentarians, and concurrent elections) appear to have much stronger
effects than individual-level effects on the likelihood of voting across these
countries. Similarly, focusing on the case of Taiwan, Achen and Wang (2017)
find that ethnicity tends to be crucial in shaping voting decision compared to
the standard variables such as income or occupational status.
Finally, rounding up with the post-communist states, it is now 30 years
since the third wave of democratisation41 swept across the region of Central
and Eastern Europe, leading to the transition from communist regimes to
democratic regimes. Offé (1991) labelled this historical change of the 1990s as
a triple transition (i.e. transition from autocracy to democracy and centrally
planned market economy42). The speed at which the ruling East European
communist parties lost their hold on power and the eventual collapse of the
Soviet Union did make one observer describe this event as the end of history
(Fukuyama,43 1989; 1992). Also, Cek (2012, p. 1611) has described the democ-
ratisation process in Central and Eastern Europe as one of the greatest success
stories in the third wave of democratisation, given that many of the countries
have succeeded in establishing stable and durable democratic institutions and
well-functioning market economies. The immediate impact this change or
transition to democracy gave citizens in these areas for the first time a mean-
ingful opportunity to freely vote for their preferred leaders in parliamentary,
presidential, and local elections.
The first (i.e. founding elections) and subsequent elections prove fertile
ground for scholarly debate44 about the importance of elections to represen-
tative democracy, democratic consolidation and creating a profile of who
the post-communist voter was. Yet, only after five years, since the transition,
the attention of scholars quickly shifted to the decline of voter turnout in
Central and Eastern European countries, with Rose (1995) describing post-
communist voters as demobilised, while other studies reported a decline of
electoral participation (Pacek et al., 2009; Jasiewicz, 2007; Klingemann, 2007;
Kostadinova and Power, 2007; Barnes, 2006; Kostadinova, 2003; Rose and
Munro, 2003). Also, in a more recent study, Kostelka (2015) reconfirmed
post-communist countries as having experienced the largest decline of voter
turnout observe across democratic regimes since the Second World War. In
fact, from initial rates of 80 per cent during the first wave of open and competi-
tive elections, average turnout rates were reduced to 50–66 per cent in some of
these countries (Tambe, 2018b; Karp and Milazzo, 2015; Kostadinova, 2003;
Pacek, Pop-Eleches, and Tucker, 2009). To provide an answer to the variations
in voter turnout across countries in post- communist countries, existing
studies have relied heavily on aggregate data (e.g. van der Brug, Franklin,
and Toka, 2008; Evans, 2006; Evans and Whitefield, 1993; Pacek, 1994; Rose
and Mishler, 1998; Rose, Mishler, and Haerpfer, 1998; Tavits, 2005, 2008;
Tucker, 2001, 2006; Whitefield and Rohrschneider, 2009) with very few studies
having focused on electoral participation at the individual level. Nevertheless,
as Karp and Milazzo (2015, p. 97) argue, the dependence on aggregate data
makes it extremely challenging to establish what has led individuals in post-
communist countries to abstain from the political process. Thus, relying on
individual-level data, some studies have examined why people vote in new
democracies of post-communist states.
First, Orvista, Caplanova, and Hudson (2005), relying on survey data
obtained from Eurobarometer, examine the determinants of electoral partici-
pation in ten Central/Eastern European states. Their analysis focuses on both
referenda (about joining the EU and NATO) and national elections. Their
results suggest that intended electoral participation increases with age, income,
and education. Attitudinal variables such as confidence in the free market and
satisfaction with the general development of the country impact positively on
people’s intended decision to vote. In focusing on both referenda and national
elections, one wonders if the differences between these types of election might
affect people’s intended decision to vote in a systematic way. While referenda
can be considered as quasi-elections, in that they present people the oppor-
tunity of expressing their preferences, they are, nonetheless, different in the
sense that they focus just on one question or a small set of issues rather than
the question of who should represent and govern the people.
Second, Barnes (2006) examines the question of participation in some
Central and Eastern European countries, with results indicating individual
resources such as education, income, and other demographic factors such as
age tend to be associated with higher rates of participation across these coun-
tries. Third, Letki (2003) examines the issue of political participation in ten
East-Central European countries (i.e. including countries such as Belarus,
Russia, and Ukraine) across four types of political engagement (i.e. polit-
ical discussion, party membership, party support, and voting). Concerning
voting, her result reveals age, education, associational network, and income
as the most reliable predictors in explaining individual-level turnout. Lastly,
Bernhagen and Marsh (2007), drawing on much better data (i.e. Comparative
Study of Electoral Systems project and European Social Survey), investigate
the determinates of voting and protesting by comparing post-communist
Europeans states with Western European countries. Concerning individual-
level participation, the authors attest post-communist citizens’ decision to
vote tends to be driven by socioeconomic, resource, attitudes, and patterns of
civic engagement that determine electoral choice in the West.
Although the above-mentioned works each shed light on the nature of
electoral participation in these four regions, they suffer from a number of
shortcomings. First, most of the sub-Saharan Africa cases, East Asia, and
Eastern European cases are more appropriately classified as “electoral authori-
tarian” regimes rather than as free and fair democracies (e.g. Zimbabwe,
Togo, Uganda, Malaysia, Singapore, Belarus, and Russia etc.). Also, Carreras
and Castañeda-Angarita (2014) note the electoral process in Latin America is
still marred by irregularities (i.e. electoral institutions that have to supervise
elections are in most cases under the control of the executive) a phenomenon
that is quite common in Africa, East Asia, and a few post-communist states
such as Belarus and Russia. Gallego (2015, p. 23) suggests that the motives
and incentives for voting in elections in partial democracies and fully demo-
cratic countries may differ. This, therefore, implies the generalisations across
all electoral regimes may be challenging to sustain. Her line of argument is
empirically reinforced by Blaydes (2011) using the Egyptian case; for instance,
she reports that poorly or less educated Egyptians are twice as likely to vote
as the educated or affluent because it is cheaper to buy their votes. Thus, it
is important to re-examine the “who votes” question across these regions or
countries to only include countries that have consolidated their democra-
cies, and which have little risk of electoral malpractice such as vote-buying,
rigging, or intimidation.
Second, some of the survey data used by these studies (i.e. Eastern Europe,
Africa, and East Asia) were, in fact, gathered a decade ago. So, this does
not tell us how things might have developed in these regions since then with
respect to the factors influencing people’s propensity to turnout. Third, a
measurement issue that affects some of the previous research on sub-Saharan
Africa, Latin America and Central/Eastern Europe is that it deals solely
with the intention to vote, that is the likelihood of voting at the next election
(see Orvista et al., 2005; Bratton et al., 2012; Nadeau et al., 2017). However,
Schmitt and Mannheimer (1991) show that, in European elections, the inten-
tion to vote and actual voting do not always yield similar relationships with
other (explanatory) variables. In particular, the stated intention of voting is
likely to be even more prone to norms of social desirability than the reported
act of actually voting. Based on this, I would agree with Blais (2000), who
argues that it is more prudent to study reported voting than the mere inten-
tion to vote. Hence, the dependent variable45 in this book will be reported
turnout at the individual level, rather than declared intention to vote.
Alternative explanations of voting in new democracies

Thus far, I have outlined several features (i.e. weak institution, ethnic cleavages,
clientelism, and fragile economy) that might distinguish the electoral context
of new democracies, thus making them differ from older and economically
more developed democracies. While I have considered some of these issues in
more detail concerning the demographic model (i.e. ethnicity/racial diversity)
and the mobilisation model (clientelism and weak institutions), I proposed
an explanation of voting behaviour in newer democracies, would be incom-
plete without a group of variables that capture factors not predominately
considered in models of voting behaviour in established democracies. Thus,
in particular, I focus on three sets of indicators, corruption, authoritarianism,
and the economy, as these phenomena may somehow alter the logic of voting
when compared to older democracies of Western Europe.
First, beginning with corruption,46 the literature on new democracies pol-
itics is full of mentions of corruptions allegations by top government officials
and parties’ heads. Citing the cases from Eastern Europe, Kostadinova (2009)
argues corruption has emerged as a major problem, and it may well affect the
behaviour of all participants in the electoral process, including voters. Also,
not only is corruption widespread and problematic but, more recently, several
scholarships (Ades and Di Tella, 1996; Fishman and Svensson, 2000; Gupta
et al., 2002; Mauro, 1997; North, 1990; Podobnik et al., 2008) have shown
the potential effects of corruptions. Podobnik et al. (2008, p. 547) contend
that one of the areas most commonly affected by corruption is a country’s
economy as the government tends to lose significant parts of its revenue via
hidden bribes and also through the dissipation of resources that could other-
wise be used productively.
Relatedly, and more in line with the purpose of this book, another sig-
nificant effect of corruption is that it tends to erode democratic institutions
and the quality of democracy itself (Warren, 2004). So, the question that
can be posed is: how does corruption shapes electoral participation in newly
consolidated democracies? To answer this question, the academic debate so
far treats corruption as a factor that has direct and indirect effects on citizens
decision of whether or not to vote: first, drawing from the literature that sees
elections as an instrument through which voters reward or sanction politicians,
the first stream of research suggests an increase in corruption has a mobilising
effect that generates incentives for citizens to cast corrupt leaders or official
from office (see Bratton et al., 2005). To illuminate this point and relying on
experimental design and cross-national survey data from the Afrobarometer,
Inman and Andrews (2009) found support for the notion that perceptions of
governmental corruption increased voting among Senegalese voters.47
Second, another stream contends instead that corruptions discourage
or undermines the faith of voters in the democratic process, which there-
fore weakens their desire to participate (see Kostadinova, 2009 (for eastern
Europe); McCann and Dominguez, 1998 (Mexican case in Latin America),
and Stockemer et al., 2013 (i.e. democracies and less developing countries).
For instance, Kostadinova (2009, p. 693) argues in the context of post-
communist countries, the burden of market reforms was borne by ordinary
citizens for years, while the elite enriched themselves overnight resulting in
anger and frustration, which led to disengagement. Mindful of these two
streams of research concerning the role of corruption on electoral partici-
pation, in this book I explore the proposition further that an increase in
corruption will increase the incentives of voters to engage in politics as a way
to get rid of or punish the proprietors. In this context, corruption, therefore,
acts as a mobilising agent as revelations of corrupt or mischievous dealings by
a politician or party leaders to enrich themselves at the expense of the general
public are brought to light may, therefore, lead to agitation or anger by voters
to oust the leaders. Moreover, ample evidence from the Eastern European
region, especially in the Polish elections of 2007, suggests voters are more
likely to use this mechanism to remove ruling parties from power.
Second, I examine the impact of authoritarianism on vote choice.
Though most of the countries in Africa, Latin America, East Asia, and post-
communist Europe have now shifted towards democracy, this was not the case
before the mid-1970s or early 1990s where authoritarian leaders led most of
the countries. Previous studies have successfully demonstrated autocracies or
authoritarian regimes tend to leave an imprint on the political attitudes of the
citizens long after they have ceased to exist (see e.g. Neundorf, Gerschewski,
and Olar, 2019). To substantiate, Tun-jen Cheng and Yun-han Chu (2018)
contend that in South Korea and Taiwan, authoritarian nostalgia has refused
to fade away with a substantial segment of the population remembering the
old regimes as having delivered social stability, economic growth, and being
less susceptible to corruption and money politics that has characterised the
new regimes. Against this backdrop, I am therefore motivated by the question
to what extent authoritarianism influence the voting decision of ordinary citi-
zens across newly consolidated democracies of Africa, Latin America, East
Asia, and post-communist European countries.
If we can recall, under authoritarian rule, ordinary citizens were not
allowed to vote for their preferred leaders. Moreover, the regime will tend
to suspend the constitution, and close down political institutions such as
political parties, labour unions, and/ or other groups that channel mass
participation. Relatedly, it was common for the authoritarian leader to
govern through martial laws, and everything was done to limit the rights
and free speech of citizens. It is now more than three decades since most
of the countries in Africa, Latin America, East Asia, and post-communist
Europe introduced some form of liberalism, political rights, and multiparty
politics, yet I argue that recent events across most new democracies indicate
that the days of military rule, one-party system, and dictatorship is nothing
of the past. For instance, in post-communist Europe, Hungary has witnessed
the emergence of national leaders (e.g. Viktor Orbán) who have used or
sought to use unconstitutional tactics or populist rhetoric to limit or curtail
the rights or freedom of ordinary citizens (e.g. refugees and LBGT groups,
etc.) and to control media freedom. In East Asia, the democratic recession
was marked by the military coup in 2005 in Thailand and the executive deg-
radation in Philippian in 2007 (Tun-jen Cheng and Yun-han Chu, 2018).
Moving to sub-Saharan Africa, the introduction of multiparty politics saw
African presidents adopt presidential institutions that gave them opportun-
ities not just to personalise their power but to use unconstitutional tactics
to limit, weaken, or dwarf other branches of governments such as the legis-
lature and judiciary (see Bleck and Van de Walle, 2018). Rounding up with
the Latin America region, Bustikova and Zechmeister (2017) also show that
some countries across the region have experienced brief moments of authori-
tarianism and military intervention. For instance, in 1992, democratically
elected president Alberto Fujimori of Peru oversaw a “self-coup”, in which
he shut down Congress, while in 2009, the Honduran military removed the
sitting President Manuel Zelaya from office.
This development somehow confirms anti- democratic or autocratic
attitudes tend to last usually longer than the oppressive regimes in which
they were developed. Thus, it is important to ask: how would these new
authoritarian attitudes affect citizens’ voting decision? I argue voters who
value democratic rights in newly consolidated democracies may be keen to
vote to preserve democracy, that is, avoid the collapse of democracy, demo-
cratic ideals, and principles. That is, they may reason that everyone else will
go through the same cost-benefit calculus and will come to the conclusion
that it is rational to abstain. But when no one votes, democracy is threatened,
and they benefit from the existence of democracy or freedom and have a
long-term interest in its maintenance. As a consequence, individuals are more
willing to bear the cost of voting to ensure against the potential breakdown
of democracy. As Roberts (2008, p. 3) states, democratic participation is
believed to keep community life vital and public institutions accountable. It
resolves conflict through “a participatory process of ongoing, proximate self-
legislation and the creation of a political community capable of transforming
the dependent private individual into free citizens and the partial and private
interest into public goods”. This is an advantage that individuals may well
regard as very considerable, especially as they have had recent experience of
living under oppressive regimes. Based on this, I, therefore, hypothesise that
voters who value democracy or democratic rights in their countries would be
more likely to vote.
Third, considering the fragile economy, I argue the question of the
economy has value for understanding the electoral behaviour in newer dem-
ocracies considering their grinding poverty and or the volatile state of the
national economies. Thus, here, I focus on citizens’ economic wellbeing as
an essential subject that might influence their voting behaviour. My line of
argument is that through an evaluation of individual and country economic
situations, individuals may decide to vote. There has been an extensive litera-
ture that centres on economic reasons to explain the propensity of voters to
turn out at the polls. This “economic vote” (for an overview economic voting
in Western democracies, see Stegmaier and Lewis-Beck, 2013; Lewis-Beck
and Stegmaier, 2000; 2007; Duch, 2007; Hellwig, 2010) as it has come to be
called, may take different forms: it could be retrospective –that is, voters’
judge past economic performance and based on that assessment vote for or
against the political incumbent –or it could be prospective, where voters
act on the economic promises of candidates (Lewis-Beck and Stegmainer,
2008, p. 303). The main point to drive home is that at the core of this lit-
erature is the assumption that citizens vote for the incumbent government
if economic times are good. On the other hand, if voters judge that their
economic wellbeing and that of the government is bad, they simply vote
against such incumbents (Lewis-Beck and Stegmaier, 2000). The effect of
individuals considering their economic and government situations before
deciding to vote implies that elections will be won or lost on the nature of
the country’s economy. In principle, this could work in one of two ways.
The economic recession might incentivise people to turn out in the polls if
they are motivated to punish governments for poor economic performance.
Alternatively, they might be motivated to turn out where they approve of
the incumbent government’s record and wish to keep it in office. However,
economic voting has been extensively studied in the context of older dem-
ocracies, with most studies reporting statistically significant effects, whereby
voters tend to punish the government for poor economic performance and
rewarding it as well.
Research across low-income countries and or new democracies remains
scarce.48 However, Pacek (1994) is credited as one of the first to test the eco-
nomic voter framework in post-communist Eastern Europe. He examined
district data from the Polish presidential and legislative elections of 1990–
1991, finding that rising unemployment lowered the vote for the pro-reform
incumbent, as well as lowering voter turnout, which suggests that the former
of our alternative hypotheses is more likely to hold. Similarly, concerning
Latin America, studies by Gelineau and Singer (2015) but also Nadeau
et al. (2017) confirm the economy has a significant effect on Latin American
voters as voters believe politicians are responsible for the economic outcomes.
Rounding up with Sub-Saharan Africa, Bratton et al. (2012) suggest African
voters do engage in economic voting, with would-be voters more likely to
Table 2.5 Overview of the expectation of an alternative model on vote choice

Alternative model Corruption perception +

Authoritarianism +
Economic perception +
Europe).
consider policy performance and government handling of unemployment and

inflations. Based on this, I, therefore, propose that voters’ evaluations of the
economic performance of their countries are likely to be positively associated
with their decision to vote. This is best explained by the fact that individuals
who make a positive evaluation of the economy are more likely to turn out
and vote because they are motivated to keep the present government in place.
The variables to be tested are once again are presented in Table 2.5.
H5a: Increase in corruption will raise the incentives of voters to engage

in politics.
H5b: I expect voters who value good governance and democratic rights to
be more likely to vote.
H5c: I hypothesised a positive evaluation of the economy by an indi-
vidual increase his or her propensity to vote.
Summary and conclusion

I have now reviewed the literature pertaining to the decision to vote at the
individual level and derived a number of hypotheses that will be tested across
newly consolidated democracies to see how far the explanations vary by geo-
political region and political system. I have also noticed that there is a less
than extensive empirical literature on turnout in new democracies and that
it is in any case very limited in terms of what it takes into account. It is now
time to turn to a detailed explanation of the research design and methods that
this book will use.
Notes
1 For an overview of recent handbooks on voting, electoral behaviour or public
opinion see: Arzheimer, Evans, and Lewis-Beck (2017); Fisher et al. (2018).
2 For a good review of this literature, see Blais (2006).
3 For a meta-analysis of individual-level research on voter turnout: see Smets and
van Ham (2013); but also Evans (2004).
4 The majority of studies of voter turnout take the rational choice theory or Down’s
economic model as a starting point (e.g. see Blais 2000; Franklin 2004; Aldrich,
1993; Bufacchi, 2001).
5 It is a meta-theory in which most of the models are nested. For example, the
importance of mobilisation and/or institutions can be interpreted through the
lenses of the rational choice model.
6 As cited by Plutzer (2018, p. 69), the term “demographics” conjures up the idea
of personal characteristics whose categories are typically reported in national
censuses and government reports.
7 Nadeau et al. (2017, p. 27) note events of this nature are not likely to affect
everyone equally as they can leave more impressionable and long-standing effects
on some age groups than others.
8 As cited by Smet (2010, 2012), making a theoretical distinction between the life-
cycle and generational approach is easier than disentangling them empirically.
9 For an overview of the concept of waithood and political participation see: Dawson
(2014); Honwana (2012)
10 In this volume my interest is more on the gender gap related to electoral
participation.
11 However, in Latin America, the Catholic church has lost its monopolist positions
with significant rise of protestants and individuals with no religion.
12 Also, for a more recent and complete account of religiosity on vote choice in Latin
America see Boas and Smith (2015).
13 I thank one reviewer for pointing out this point during the initial review of this
book project.
14 However, it is equally possible that turnout among ethnic minorities would be
lower than among ethnic majorities, which may be partly explained by their
limited access to resources and skills.
15 In their meta-analysis of individual level turnout, Smet and van Ham (2013, p. 7)
show that this line of argument or hypothesis is updated as almost all the test and
studies find insignificant effects.
16 Bratton, Chu, and Lagos (2010) explain this by arguing social pressure to partici-
pate may be felt much more strongly in rural areas, which house more tightly knit
communities. In contrast, Kuenzi and Lambright (2011) advance the conditions
for political mobilisation through patronage networks are more favourable in rural
areas, as a recent study of electoral participation in Africa has demonstrated.
17 Jennings (1979) contends an individual’s socioeconomic status has a known posi-
tive relationship with most form of political participation.
18 Apart from age, no other socio- demographic variable has received as much
attention as education (see Burden 2009).
19 In the literature, high social status is defined as having a central social network
position in society
20 SES seems an important determinant of turnout in the context of Western democ-
racies amid a few mixed results. For an overview of these studies see Norris (2002);
Pierce (1995); Teorell et al. (2007); Topf (1995).
21 It can be argued that in Asia, poor voters are often tied into clientelistic networks
(which are very efficient in mobilising voters), while middle-class voters are often
disillusioned with politics and hence more likely to abstain.
22 This result is perhaps driven by a high proportion of older people in countries
such as Mongolia, Taiwan, and China who lack any kind of formal schooling but
who, nevertheless, are mobilised to become habitual voters.
23 Blaydes (2011) attributes the high rate of participation of those with low educa-
tion to vote buying.
24 See e.g. Karp et al. (2008).
25 I find it important to note Almond and Verba referred to Italy, Germany, and
Mexico of the 1950s/1960s
26 For instance, Franklin (1996, p. 219) has outlined a number of instrumental motiv-
ations that are largely determined by the institutional context in which elections
are held, such as the salience of elections, the use of compulsory voting, highly
proportional election systems, postal voting, weekend polling, electoral formula,
and effective number of parties.
27 For example, in their study of voter turnout at the aggregate in Latin America,
Fornos et al. (2004) do not detect any association between PR and turnout.
28 For individual-level evidence, see Perea, 2002 and Brockington, 2004.
29 Moreover, since scholars tend to view electoral choice as been contingent on other
freedoms such as: expression, assembly and association, the safeguarding of these
liberties also constitutes at least an implicit element of most definitions.
30 Studies in the context of sub-Saharan Africa and East Asia reveals holding
regular elections does not necessarily leads to democracy (e.g. see Bleck and van
de Walle 2018; Rodrigues Sanches, José Jaime Macuane, and Chipo Dendere 2020;
Morgenbesser and Pepinsky (2019).
31 See Svolik (2015).
32 Others such as clientelism, corruption, and authoritarism are discussed in more
detail along the core theoretical and alternative models or explanations of elect-
oral participation.
33 Pop-Eleches (2010); Rose and Mishler (1998) show anti-party sentiment has been
relatively high in Eastern Europe.
34 With a few exceptions like Gabon, Algeria and Mauritius, South Africa (Africa);
Brazil, Argentina (Latin America); South Korea (East Asia); Croatia, Czech
Republic, and Hungary (Eastern European).
35 Exclusion by socioeconomic group occurs when individuals are denied access to
services or participation in governed spaces based on their identity or belonging to
a particular group.
36 For detail of diversity and social cleavages in the Asia-Pacific region, see Reilly
(2006); Hyug Baeg Im (2004); Montinola (1999); Albritton (2004); Surin
Maisrikrod (2002).
37 However, it is important to note the time of democracy in Latin America and
citizens experience with elections and voting does vary across countries ranging
between 1980 and 1996 as this mark the breakdown of authoritarism in the region
except for Cuba. See Nohlen (2005, p. 9).
38 Pioneering studies of vote choice included those of: Dow’s (1998), which focuses
on the analysis of policy relevant voting in Chile; Domínguez and McCann’s
(1998) which dealt with a two-step model of vote choice in Mexico and, finally,
Mainwaring’s (1999), which focuses on the weak party attachment in Brazil.
39 Carlin, Singer, and Zechmeister (2015) cite analysts tended to focus either on
Brazil or Mexico because of the development and production vote- oriented
survey project. Also, for details on studies Brazil and Mexico see: Ames, Baker,
and Rennó (2008); Domínguez and Lawson (2003); Moreno (2003, 2009b); Telles
and Moreno (2013).
40 However, a select number of small states studies include those conducted in
Venezuela by Weyland (2003) and Nadeau, Belanger, and Didier (2013).
41 As noted by Kostelka (2014) but also Roberts (2018, p. 10) before 1989, Central
and Eastern European countries, which used to be known as the world harshest
dictatorship, now comprise countries that are ranked among the global demo-
cratic front-runners.
42 For a discussion of this regime change during this period see: Fagan and
Kopecký (2018)
43 Fakuyama further notes that human civilisation might be witnessing a final stage
of institutional evolution, represented by the universal spread of liberal democra-
cies. (Also, for more detail on this historical development see Herron (2009); Coffé
and van der Lippe (2010).
44 For details, see studies by Turner (1993), and Ishiyama (1996).
45 More details on the operationalisation of the dependent variable will be given in
Chapter 3.
46 Corruption is conventionally defined as the abuse of public office for private gain
(see e.g. Shleifer and Vishny, 1993)
47 Apart from voting, Monyake and Hough (2019) find that in sub-Saharan Africa,
people’s experience with corruption or bribery tends to increase the likelihood of
engaging in protest activities.
48 In a more recent of economic voting in major low-income regions, Gélineau (2013)
finds a significant relationship in the expected direction.
3
The comparative framework
This chapter deals with the research design and data upon which the entire
study rests. I begin with a discussion of the methodological issues and problems
associated with research on electoral studies, before proceeding to the next
section of the chapter that centres on the research method and design upon
which this book is based. Here, I describe the methods to be used and high-
light the methodological challenges associated with studies of electoral behav-
iour. This includes a discussion of case selection and the operationalisation
and measurement of our dependent and independent variables. But before
I move to these issues, it is essential to clarify one thing: why focus on the
newly consolidated democracies of sub-Saharan Africa, Latin America, East
Asia, and Central/Eastern Europe?
Comparative framework and parameters for comparing the regions

Why compare Africa, Latin America, East Asia, and post- communist
European states? There are some obvious similarities that might justify a cross-
regional comparison. First, the major commonality between them is that they
all experience the transition to democracy during the “third wave” of democ-
ratisation.1 In fact, for sub-Saharan Africa, post-communist Europe, and, to
a lesser extent, East Asia, I can argue the transition to democracy was almost
at the same time, which was in the early 1990s following the disintegration of
the Soviet Union and the end of the Cold War.2 Relatedly, all four regions can
be considered to have gone through the democratic transition for the same
reasons: international pressure for democratisation and the delegitimisation
of authoritarian rule. As argued by Levitsky and Way (2002) but also Bunce
(1999) and Verdery (1996), the collapse of the Soviet Union echoed cries for
the liberalisation of politics and economics, thus preventing countries from
embarking on a different path. This period, therefore, marked the first wave
of open and competitive elections in these states. After more than 20 years of
democratic experience and electoral participation, institutional and behav-
ioural patterns should be institutionalising, so it is the right time to research
the voting decisions in these regions, to see which factors are associated with
electoral participation.
60 Comparative framework
Second, Pietsch, Miller, and Karp (2015, p. 5) argue that East Asia and
Eastern Europe (i.e. which I expect to be true for the Africa and Latin
America regions) are frontline regions for the spread and consolidation
of democracy, as they both include a large number of new democracies
as well as electoral authoritarian regimes that may gradually become more
competitive. Additionally, Levitsky and Way (2010) and Boix (2011) con-
tend that the relatively high level of economic development makes these
regions a perfect context for democratic consolidation. But a decade ago
cross-regional comparison of voting behaviour across these four regions
would not have been possible because of the absence of survey research
from Africa, Latin America, and East Asia. So, comparative analyses are
possible now because of the recent development of public opinion surveys
in these regions.3
That said, the four regions are quite different in a number of ways, including
a history of communist regimes, the legacy of colonialism, rates of human
development, levels of economic development and different electoral and
party systems. This in itself is useful though since should I find differences
in our explorations of voting from one region to another, these differences
might prove fruitful as potential sources of explanation. On the other hand,
if the findings prove to be broadly similar across all of these regions, it would
suggest that one or more key underlying factors are driving the similarity of
outcome, despite the many differences between the regions.
To this extent, it would approximate the most different systems compara-
tive strategy. The most different systems design is a case selection technique
where cases are different on specified variables other than X1 and Y. That is,
cases are different in all respects except the variables of theoretical interest
(Gerring, 2007, pp. 90–215). In the most different system design, variations
on independent variables are prized, while variations on the outcomes are
eschewed. The researcher, therefore, tries to identify cases where just one inde-
pendent variable (X1), as well as the dependent variable (Y), covary, while all
other plausible factors (X2 a–d) show different values (Gerring, 2007, p. 139).
It is important to note this design is taken from Mill (1875, p. 541) and,
according to him, the most different system design is known as the method of
agreement. Mill defines the method of agreement as follows:
If two or more instances of the phenomenon under investigation have

only one circumstance in common, the circumstances in which alone all
the instances agree is the cause or effect of the given phenomenon.
The application of this method is straightforward in that if an investigator

wants to know the cause of a certain phenomenon, he or she should first
identify instances of the phenomenon and then attempt to determine which
circumstance invariably precedes its appearance. The circumstance that sat-
isfies this requirement is the cause (Ragin, 1987). The researcher, therefore,
aims to explain a circumstance that is present in all the cases of interest. If
Comparative framework 61
these instances have only one of several possible antecedent circumstances in
common, then this antecedent circumstance in which all the instances agree
is the cause or the effect of the phenomenon (Caramani, 2009, pp. 45–46).
Essentially, as Ragin (1987, p. 36) pointed out, the method of agreement is
a search for a pattern of invariance, since all instances of a phenomenon are
identified, and the investigator attempts to determine which of the possible
causal variables is constant across all instances. This approach does carry
certain well-known weaknesses. First, is the familiar problem of too many
variables and too few cases. However, this problem is not intrinsic to the
most different system design as it applies equally to the most similar systems
design.
Second, the problem of establishing equivalence is another well-known
difficulty of this approach. Equivalence can be understood as the degree to
which “under different conditions of observing and studying phenomena,
measurement operations yield measures of the same attribute” (Horn and
McArdle, 1992, p. 117). Survey-based research of the kind relied on this
study, covering many different nations and cultures (sub-Saharan Africa,
Latin America, East Asia, and Central/ Eastern Europe) encounters the
familiar problem of establishing whether responses to questions asked in
each country can be treated as functionally equivalent. For example, pre-
cise wording and coding for items such as those monitoring satisfaction with
democracy, political efficacy or party affiliation in Africa, Latin America,
East Asia, and Central/Eastern Europe may be different. To overcome this
problem, drawing on the work of Harkness et al. (2003),4 I have undertaken
the necessary steps to define, operationalise, and explain how the dependent
and independent variables that I propose to use in this book have been
measured.
Regardless of the aforementioned criticisms, the strongest utility of this
design and, of course, one that has stood the test of time as a comparative
research method is that the comparison of these different and diverse regions
facilitates theory-building and testing in a way that no other method is apt to.
This design is therefore well suited for my research question because of the
following: it enables me to see if there are differences between the established
and newer democracies as a whole, or whether there are differences between
the newer democracies as well with regards to the determinants of electoral
participation. In effect, it sets up a series of most different systems design
tests: for example, I might find that personal resources have a significant inde-
pendent impact on the decision to vote, even when the various cases we are
looking at differ widely in most respects, or that institutional factors such as
the electoral system do. And so on. To be precise, in general terms, I expect to
find one of the following patterns:
1. First, no significant differences between Africa, Latin America, East

Asia, and Central/Eastern Europe. This would affirm that the models that
have been developed in the oldest established democracies (i.e. Western
Europe) to explain why people choose to vote are “general” and have the
widest application among democracies.
2. Second, differences between newly consolidated democracies on the other
hand when compared to established democracies in the global north (i.e.
Western Europe) –suggesting that something different is happening in
newer democracies. This would then raise the interesting questions of
what and why?
3. Lastly, quite distinctive patterns in each of the four broad geo- pol-
itical regions. That is, each region is unique in some way: what might
explain these?
Selection of cases
One of the most durable conventions about what constitutes an appropriate
research strategy in comparative politics is the selection of cases (Geddes,
1990, p. 131). The cases in this volume consist of newly consolidated democ-
racies of sub-Saharan Africa, Latin America, East Asia, and Central/Eastern
European states. As previously mentioned in the introductory chapter, the
cases selected for empirical analysis in this book will be defined by adherence
to these three criteria: countries that have experienced the end of authoritarian
rule, countries that have experienced the installation of a democratic regime
and countries that are experiencing the consolidation of their democratic
regimes (i.e. democratic consolidation can be expected to take about 20 years,
that is, approximately five or six democratic elections). Beyond this, however,
I rely on Freedom House’s concept of “electoral democracy”. Specifically,
Freedom House’s rating judges which countries are considered democratic
and which are not. The evaluation of African, Latin America, East Asian,
and Central/Eastern European democratic experiences are made by using the
index of democratisation, which is based on an annual assessment of polit-
ical rights and civil liberties monitored by Freedom House. Freedom House
has been rating political rights and civil liberties in each country of the world
since the 1970s. It is the most systematic evaluation of the degree of democ-
racy, an evaluation whose validity is judged to be reputable (Bollen, 1993).
Freedom House’s index of democracy has been extensively used in cross-
national research (Burkhart and Lewis-Beck, 1994; Helliwell, 1994). Freedom
House gives every country a rating ranging from 1 to 7 on political rights and
civil liberties, a score of 1 corresponding to a maximum degree of freedom,
and a score of 7 to the minimum degree.5
My interest is directed mainly at the political rights scale, which focuses on
the fairness of elections. I construe a rating of 1 and 2 as reflecting a satisfac-
tory level of democracy; Freedom House itself collapses scores of 1 and 2 into
a single category indicating a free country (Gastil, 1979, p. 24). It should be
noted that in drawing on the Freedom House ratings, I do not make a decision
to include a case solely based on a single election or year, but rather look sys-
tematically at every national election covered by the surveys and then decide
if the country can be judged democratic overall. At the time of writing, out
of 54 countries in Africa, ten countries6 met these criteria. Concerning Latin
America, nine countries could also be classified as newly consolidated dem-
ocracies based on my three criteria mentioned above and Freedom House’s
ratings. Turning to East Asian countries, based on this criterion, out of the
18 East Asian countries, only four countries (Japan,7 South Korea, Mongolia,
and Taiwan) meet the requirement of being fully free and fair, while the
Philippines, Singapore, Indonesia, and Malaysia are only considered partly
free and as such are excluded from this study. Finally, rounding up with post-
communist states, I focus on six countries. Countries like Belarus, Russia,
Serbia, Albania, Kosovo and Bosnia, and Herzegovina were all eliminated
because they were not considered as free by Freedom House ratings, while
Croatia, Slovakia, and Latvia etc., although democratic, were excluded from
our study because of the absence of any survey for this country. The full set
of countries included in this study are reported in Table 3.1.
Also, Table 3.2 displays the political rights for these countries for a period
of 20 years (1999–2019) and the number of elections8 conducted since the
very first elections that were held following the third wave of democratisation.
Of the 26 countries from the four-geo political regions included in this book,
most of the countries (most notably those in post-communist Europe, Latin
America, and, to a lesser extent, those in Africa) have experienced no change
or stable trajectory in their political rights index and only Botswana,9 Poland,
Hungary,10 and Namibia experienced a slight decline.
Methodological challenges associated with studies of voting

behaviour
The principal ambition of this book is to test the theoretical expectations of
why people turn out on election day. However, it is crucial to address a signifi-
cant methodological challenge that is often associated with research on elect-
oral studies –the need to take into account more than one level of analysis.
In the last 50 years or so, electoral research has been spearhead by survey
research. As Curtice (2002, p. 162) points out, individual-level survey data is
often affected by the “context problem”. The difficulty here is that in order to
explain people’s propensity to vote or not to vote in elections, many studies
have tended to pay too little attention to the context in which an individual
vote is cast. In trying to understand or explain why people choose to vote or
not to vote, it is not good enough to analyse individual behaviour in isolation,
while ignoring what happens at the contextual, macro, or aggregate level. For
example, Curtice (2002) argues that at the macro-level or country level, many
factors could affect the voting decision of individual voters, including parties,
candidates, media reports, events, state of the economy, foreign wars, or just a
simple blunder. How do we get over this methodological issue? He, therefore,
proposes two important changes to the traditional individual-level approach
to election studies.
Table 3.1 Selected cases
Regions Countries
Sub-Saharan Africa Benin

Botswana
Cape Verde
Ghana
Mauritius
Namibia
Sao Tome and Principe
South Africa
Latin America Argentina
Brazil
Chile
Costa Rica
El-Salvador
Panama
Peru
Guyana
Uruguay
East Asia Taiwan
Mongolia
South Korea
Post-communist Europe Czech Republic
Estonia
Hungary
Lithuania
Poland
Slovenia
Total number of cases 26
Note: Countries selected are based on the adherence to three criteria: countries that have
experienced the end of authoritarian rule; countries that have experienced the installation of
a democratic regime, and countries that are experiencing the consolidation of their democratic
regimes.
First, Curtice argues that there is a need to combine both micro-and

macro-level variables in order to understand why people vote or do not vote
in elections. The rationale here is that we cannot expect to understand elect-
oral behaviour by merely looking at it through the prism of the individual
voter; instead, we need also to measure the macro or aggregate context within
which electoral behaviour takes place. This, therefore, means that any study
trying to explain vote choice or voting intention must take into consideration
both the individual and country characteristics that are believed to influence
people’s propensity to turn out in elections. Second, this might affect our
understanding of causal relationships, for once context is brought in to the
analysis, researchers may no longer be interested simply in whether A causes
B. Rather, a researcher will be interested in whether the relationship between
Table 3.2 Political rights index for sub-Saharan Africa, Latin America, East Asia, and
post-communist states
Territory/region 2019 Average Average Average Average Number of

political 1999–2004 2005–2009 2010–2014 2015–2 019 elections
rights political political political political conducted*
index rights rights rights rights
index index index index
Sub-Saharan Africa (SSA)

Cape Verde 1 1 1 1 1 6
Benin 2 2.5 2 2 2 6
Ghana 1 2 1.2 1 1 7
Sao Tome and 2 2 2 2 2 6
Principe
Namibia 3 2 2 2 2.2 6
Botswana 3 2 2 2.4 2.4 6
South Africa 2 1 1.6 2 2 6
Mauritius 1 1 1 1 1 6
Latin America (LEC)
Argentina 2 2.2 2 2 2 9
Brazil 2 2.6 2 2 2 8
Chile 1 1.8 1 1 1 7
Costa Rica 2 2.6 2 2 2 18
Uruguay 1 1 1 1 1 7
El Salvador 2 2 2 2 2 8
Guyana 2 2 2 2 2 6
Panama 1 1 1 1.2 1.8 7
Peru 2 2.5 2 2 2 9
East Asia (EA)
South Korea 2 2 1 1.2 2 6
Mongolia 1 2 2 1.6 1 8
Taiwan 1 1.6 1.8 1 1 7
Post-communist European Countries (CEEC)
Poland 1 1 1 1 1.2 9
Lithuania 1 1.2 1 1 1 8
Czech Republic 1 1 1 1 1 9
Estonia 1 1 1 1 1 8
Slovenia 1 1 1 1 1 7
Hungary 3 1 1 1 2.6 8
Source: Freedom House. The index ranges from 1 (best) to 7 (worst) referring to the degree of
political rights in a country. The number of elections: These are based on the most important or
national election (i.e. for some countries these are parliamentary elections, while for others they
are presidential elections) that have been conducted since the first elections marking the third
wave and are derived from www.idea.int/data-tools/data/voter-turnout. * The number elections
relate to the current wave of democratisation across each region. I focus on the most important
national election. In Latin America, the focus was on presidential elections; in post-Communist
Europe, the focus was on legislative/parliamentary elections. In sub-Saharan Africa, the focus
was on presidential elections and parliamentary elections for South Africa and Botswana cases.
In East Asia, the focus was on presidential/or legislative election.
A and B varies according to the context, C. The method I, therefore, rely on
for this book must take account of the impact of context.
Data, measurement, and estimation strategy
Data
A decade ago, cross-regional comparison of voting behaviour would not
have been possible, as multinational efforts at generating comparable survey
projects are quite recent. (For an overview of cross-national survey projects see
de Vries, 2017; Heath et al., 2005; Kittilson 2007; Norris 2009). Among these
initiatives, I count the Comparative Study of Electoral Systems (CSES), the
World Values Survey (WVS) and the International Social Survey Programme
(ISSP). Unfortunately, I do not use any of these survey data because of the
following reasons.
First, beginning with the Comparative Study of Electoral System
(CSES),11 Module 1 (1996–2001) only included three countries from Latin
America (i.e. Peru, Mexico, and Chile), five countries from East Asia (i.e.
Hong Kong, Japan, South Korea, Taiwan, and Thailand), nine countries in
post-communist Europe (i.e. Ukraine, Russia, Slovenia, Romania, Poland,
Lithuania, Hungary, Czech Republic, and Belarus), with no sub-Saharan
Africa country. CSES’ latest Modules 4 and 5 only saw the addition of two
more countries in Latin America, one East Asia country, and two coun-
tries in sub-Saharan Africa (i.e. South Africa and Kenya). Not only are we
confronted with a limited number of cases, which makes it hard to draw any
kind of inferences across these regions but most of the countries included
(e.g., Thailand, Philippians, Hong Kong, Kenya, Belarus, Ukraine, Russia,
Serbia, Montenegro, Mexico, etc.) cannot be regarded as consolidated dem-
ocracies according to my logic of case selection outlined above (i.e. also see
Chapter 1).
Second, the World Value Survey (WVS)12 is another multinational survey
initiative that was launched in the 1980s and, so far, seven waves have been
conducted, which covers most of the countries from all the four regions.
However, the question about electoral participation is not ideal since it does
not ask directly whether respondents actually voted in the most recent national
elections, but rather asks “When elections take place, do you vote always, usu-
ally or never?”. I argue this question tends to capture the frequency of voting
rather than whether the respondent truly voted in the most recent national
election or not.
Third, the International Social Survey Programme (ISSP)13 is a cross-
national programme established in 1984 and has a common set of questions
for established social survey in Australia, Germany, Great Britain, and the
US. It now covers 42 countries and has collected data annually, through a
core questionnaire. Although, the survey capture attitudes towards economy,
gender, the legal system, socio-demographic as well as variables needed to
study voting behaviour, in terms of geographical coverage, the survey focuses
mainly on Western democracies. For instance, only two sub-Saharan African
and three Latin American countries are included (i.e. South Africa, Tunisia,
Chile, Mexico, and Venezuela).
Based on these shortcomings of the CSES, WVS, and ISSP data, I, there-
fore, rely on four regional datasets across the four regions, which incorporate
almost similar questions wordings for some variables and are therefore appro-
priate for comparing with each other. The Afrobarometer for sub-Saharan
African cases, AmericasBarommeter or Latin American Public Opinion
Project for Latin America sample, the AsianBarometer for East Asian cases,
and the European Social Survey for Central/Eastern European countries.14
These data thereby permit me to see how Africans, Latin Americans, East
Asians, and East European voters have behaved over recent decades with
regards to their decisions to vote or not to vote in elections.
First, the Afrobarometer is an independent, non-partisan research pro-
ject that measures social, political, and economic attitudes and behaviour in
Africa. The surveys have so far been conducted in 36 African countries and
are repeated on a regular cycle. They ask a standard set of questions so that
countries can be systematically compared since trends in public attitudes are
tracked over time. They are designed to generate samples that are nationally
representative across all citizens of voting age and has a sample size ranging
between 1200 and 2400. Second, the AmericasBarommeter15 conducted by
the Latin American Public Opinion Project is a periodic study of 34 countries
in the Western Hemisphere. It consists of a country survey based on national
probability samples of voting-aged adults. The principal objective of the
survey is to advance the accurate description and understanding of public
opinion and behaviour across the Americas. Third, the AsianBarometer is
an applied research program that aims to gauge public opinion on issues
such as political values, democracy, and governance across Asia. The survey
was launched in 2000, and so far four survey rounds have been completed
including 13 East Asian and five South Asian countries and the aim is a
sample size of 1200 per country. Finally, the European Social Survey (ESV) is
an academically driven cross-national survey that has been conducted across
Europe since its establishment in 2001. The survey measures the attitudes,
beliefs, and behaviour patterns of the diverse population in more than 30
nations.16
The beauty of using these four datasets is that most of the survey taps
questions pertaining to electoral or voting behaviour in the various regions.
Thus, these surveys are more appropriate than national election studies for
this book because they provide for the possibility of cross-regional com-
parison of people living across diverse political, economic, and social contexts.
As Curtice (2007) puts it, the advantage of a cross-national survey is that it
provides or enable researchers to assess the empirical generalisability of their
theoretical claims about the causes and consequences of political attitudes
and behaviour. Similarly, it permits one to examine the robustness of their
research findings by replicating their analyses using different independently
collected data sources.
Although the focus of this book is on newly consolidated democracies,
I nevertheless argue that it will be important to examine what is happening in
Western European democracies in order to establish a benchmark for com-
paring newly consolidated democracies. I therefore once again rely on Round
8 data from the European Social Survey, which was conducted in ten different
European countries: Austria, Belgium, Finland, Germany, Ireland, Italy,
Netherlands, Norway, United Kingdom, and Sweden. This offers one signifi-
cant advantage. First, as I have pointed above, it covers ten Western European
countries, and sets a baseline for comparing newly consolidated democracies
and old democracies (i.e. this allow me not only to determine whether but also
by how much new democracies differ from older democracies). Tables A1, A2,
A3, A4 (e.g. see Appendix 1) contains detailed information about the datasets.
First, Table A1 consists of eight sub-Saharan African countries for which
we are interested in exploring which factors determine an individual’s propen-
sity to vote in national elections. The Xs in the table show which Afrobarometer
survey was conducted in a country. This, therefore, means that the absence
of an X signifies that no survey was conducted in a country during that
round. From Table A1, we see that the survey is currently on its sixth round.17
Second, Table A2 shows the number of countries and their distribution in the
different round of the AmericasBarommeter that I am interested in studying.
The first set of the survey was conducted in 2004 in 11 countries; the second
took place in 2006 (consisting of 22 countries across the region); in 2008,
the survey grew and included 24 countries, and in 2010 and 2012, it included
26 countries. Finally, Round 6, the 2014 survey that this book focuses on,
consists of a total of 28 countries.18 Third, Table A3 shows the number of
countries and their distribution in the different wave of the AsianBarometer
that I am interested in studying. All of the cases were included across all the
four waves. Lastly, rounding up, with post-communist European countries,
Table A4 shows the number of countries and their distribution in the different
rounds of the European Social Survey, which I am interested in studying. In
summary, having outlined the datasets that this study will rely on, it is essen-
tial to note that I focus on the most recent rounds or waves (i.e. as it falls
approximately between five and six national elections conducted across these
countries) thus providing valuable perspective on voting behaviour across the
four geo-political regions.
Conceptualisation and operationalisation of variables19
Measuring electoral participation: dependent variable

To examine the determinants of voting at the individual level in newly
consolidated democracies in sub-Saharan Africa, Latin America, East Asia,
and post-communist European countries, it is vital to begin by clarifying
how the dependent variable is measured. A survey of the current litera-
ture presents three ways in which turnout on an individual level has been
operationalised: (1) the ex-post (report) method; and (2) the ex-ante (inten-
tion) method (Castillo, Miranda, Bonhomme, Cox, and Bascope, 2014;
Ekström and Östman, 2013; Hooghe and Wilkenfeld, 2008; Persson, 2014);
and (3) validated turnout data.
First, beginning with the ex-post method, turnout is most commonly
measured by using a post-election survey that asks respondents whether
they have voted in a previous election. However, this self-reported turnout
is often affected by problems such as recall bias and social desirability.
As cited by Smets and van Ham (2013, p. 347), the implication of this is
that reported turnout data tend to have an upward bias when compared to
actual turnout data.20 For example, using the American National Elections
Studies, Karp and Brockington (2005, p. 825) report the estimated difference
between actual turnout and reported turnout was around 20 percentage
points during the 1990s. The second type of measurement is the ex-ante
method or what is commonly known as the turnout intention.21 This is gen-
erally based on surveys that are held at some point before the elections, and
that ask individuals or respondents in national samples about their turnout
intention, either in the upcoming election or if elections were to be held the
next day. Scholars such as Brady (1999), Furnham and Gunter (1989), and
Norris (2004) argue that intention is a poor predictor of actual participa-
tion. In relation to this, previous studies by LaPiere (1934) provide evidence
for the fact that the link between intention and behaviour is quite weak.
The third form of measurement consists of the use of validated turnout
data, which are generally based on official voter records. However, its major
weakness is that a validated turnout is more challenging to obtain as voter
records are not kept or released in all countries. To sum up, in terms of
validity, validated turnout data is the most robust (although it is the most
difficult to obtain), while reported and intention turnout are prone to social
desirability (this is, even more, the case for intention voting as compared to
reported voting, as cited by Schmitt and Mannheimer, 1991). Thus, Blais
(2000) argues that it is more prudent to study reported voting than the mere
intention to vote.
Hence, the dependent variable in this book will be reported turnout at the
individual level rather than turnout intention or validated turnout. All said,
since my dependent variable is self-reported turnout,22 I, therefore, rely on a
series of question from the four surveys to operationalised electoral partici-
pation. (See Table B1, Appendix 2 for the exact question wordings across the
four survey data). All four datasets included variable asking respondents who
have attained voting age whether they voted in the previous election. I collapse
the dependent variable into a dichotomous variable according to whether a
voter voted in the last election or not, with a value of 1 if they did, and 0 if
they did not. Also, I omitted all respondents who were not eligible to vote at
the last election, with refusal, don’t know, and no answer responses coded as
missing values.23 Table 3.3 presents more details about our dependent vari-
able, including the official and self-reported turnout for each country across
the four regions, the election year and the date of the survey.
Worth recalling, the primary datasets I consider in my analyses of indi-
vidual voting decision are from the Afrobarometer, AmericasBarometer,
AsianBarometer, and the European Social Survey. First, beginning with the
Afrobarometer, the data I consider for analyses are based on Round 6, covering
a total of eight sub-Saharan African countries. The pooled dataset for these
countries consists of 11,986 respondents who were interviewed between 2014
and 2016. However, some 1017 respondents abstained, refused, or did not indi-
cate whether they voted in the most recent national election. Thus, the analysis
I present for the sub-Saharan Africa region is based on some 10,969 respondents
who reported their vote choice. Second, for the Latin America regions, the ana-
lysis is derived from the 2014 AmericasBarometer, covering a total of nine
countries, with a total sample size of 13,709 observations, with only a small
number (i.e. 107) having failed to report whether they voted in the most recent
national election or not. Thus, the analysis I present for the Latin America
region is based on some 13,602 respondents who reported their vote choice.
Third, for the AsianBarometer (i.e. East Asian cases) the data I consider for the
empirical analysis is based on Wave 4 covering a total of three East Asian coun-
tries that are deemed to have consolidated their democracy. The pooled dataset
for these countries consists of 4085 respondents who were an interviewed from
2014 to 2016. However, considering some respondents did not declare their vote
choice decision, the analysis I present for the East Asia countries is based on
some 3956 respondents who reported their vote choice when they survey was
conducted. Finally, rounding up with the post-communist states, the data I con-
sider for analysis are those from the 2016 European Social Survey covering a
total of six countries, with a sample size of 11,025.
As mentioned earlier, one shortcoming with survey data on self-reported
turnout is that social desirability bias leads to respondents overreporting their
self-participation (e.g. see Table 3.3 above reporting official and survey turnout
across the different regions).24 One might argue that over-reporting may,
in turn, bias the relationship between the variables of interest and turnout.
However, I find reassurance from a recent study by Persson and Solevid (2014,
p. 104), which established over-reporting does not affect the correlations with
other variables and probably does not bias analyses of causal effects on polit-
ical engagement or participation.
Measuring independent variables

In the previous chapter (i.e. see Chapter 2), the theoretical review of the litera-
ture on electoral participation resulted in a series of hypotheses that I will test
to explain people’s propensity to vote in newly consolidated democracies. This
section outlines the independent variables that are derived from each of these
theories/models and then specifies how these variables are operationalised.25
Table 3.3 Information about the dependent variable
Region/countries Survey yeara Election Official Self-

year turnoutb reported
Survey date/month turnoutc
Sub-Saharan Africa countries

Benin October 2014 2011 84 85
Botswana June 2014 2014 84 70
Cape Verde March/November 2014 2011 60 80
Ghana June 2014 2012 80 83
Sao Tome and November 2015 2011 74 82
Principe
Namibia August 2014 2009 98 80
Mauritius September 2014 2014 77 87
South Africa June 2015 2014 73 75
Latin America countries
Argentina 28 February to 22 March 2011 79 77
2014
Brazil 2 May to 28 May 2014 2010 78 77
Chile 16 April to 22 May 2014 2013 42 68
Costa Rica 4 March to 6 May 2014 2014 56 72
Uruguay 8 March to 23 April 2014 2014 89 85
El Salvador 28 March to 30 April 2014 60 75
2014
Guyana 4 June to 12 July 2014 2011 73 74
Peru 23 January to 8 February 2011 82 84
2014
Panama 13 March to 3 May 2014 2014 77 70
East Asia countries
South Korea October 2015–December 2012 75 85
2015
Taiwan June 2014–November 2012 74 83
2014
Mongolia June 2014–August 2014 2013 67 89
Central/Eastern European countries
Poland 07/11/16–22/02/17 2015 51 74
Lithuania 04/10/17–28/12/17 2016 51 58
Czech Republic 24/10/16–19/12/16 2013 59 58
Estonia 01/10/16–31/01/17 2015 64 72
Slovenia 21/09/16–11/01/17 2014 51 72
Hungary 14/05/17–16/09/17 2014 62 75
Western European countries
Austria 19/06/16–28/12/16 2013 75 84
Belgium 14/09/16–31/01/17 2014 89 92
Germany 23/08/16–26/03/17 2013 72 86
Finland 15/09/16–08/03/17 2015 67 83
United Kingdom 01/09/16–20/03/17 2015 66 76
Ireland 25/11/16–08/05/17 2016 65 76
Italy 11/09/17–19/11/17 2013 75 76
Netherlands 01/09/16–31/01/17 2012 75 82
(continued)
Table 3.3 Cont.
Region/countries Survey yeara Election Official Self-

year turnoutb reported
Survey date/month turnoutc
Norway 22/08/16–17/01/17 2013 78 88
Sweden 26/08/16–10/02/17 2014 86 93
Notes: a: Survey year for the different regions are as follows: 2014–2016 for the Africa cases; 2014
for the Latin America, 2014–2016 for East Asia and 2016 for the post-communist states/Western
European countries. b: Source of official turnout figures: International Institute for Democracy
and Electoral Assistance. c: Source of self-reported turnout figures: Afrobarometer (Round 6,
2016), AmericasBarometer (Wave 6, 2014), AsianBarometer (Wave 4, 2016) and European Social
Survey (Round 8, 2016).
Table B1 in Appendix 2 contains the exact question wordings of the inde-

pendent variables across the different survey data. While Table B2 in
Appendix 2 contains the variables description and coding scheme across the
four regions/surveys.
Socio-demographic variables
The main variables from this theoretical perspective include age, gender, reli-
gion, ethnicity/race, place of residence and education.
Demographic variables
First, beginning with demographic variables, age is included in the analysis.
Given that the relationship age and vote choice is non-linear, I find it effi-
cient to keep age as a continuous variable and entered it as a quadratic term.
I expect the variable coefficient to show a positive sign, as older people/citi-
zens should be more likely to vote.
Second, looking at gender, all four surveys have standard questions that
reflect the gender of respondents. For the sake of clarity and ensuring com-
parability across the regions, I recoded gender into a dummy variable, where
men are coded as 1 and women coded as 0 (i.e. women being our reference
category). Again, I expect the effect of the dummy variable to be positive,
indicating men are more likely to participate.
Third, looking at religion, across three regions, this variable is measured by
a question that asks about the frequency of attendance at religious services. For
example, how often do you personally engage in religious practices (Africa),
how often you practise religious services or rituals (East Asia), and how often
do you attend religious services apart from special occasions (post-communist
European)? Considering that I propose that voters who attend religious ser-
vices more frequently/or who are more religious should be more likely to vote,
to maintain comparability across the regions, I included a five-point scale for
church attendance. This variable runs from never or almost never attend a
religious service (0) to attending services more than once a week/every day (5).
However, for the Latin America regions, AmericasBarometer (Wave 6, 2014)26
I did not have any question measuring religion attendance. Thus to account
for the role of religion, I rely on a question that asks respondents: what is your
religion? I created a dichotomous variable, indicating whether a respondent
is Catholic (1) or not (0). Considering that the Latin America region is pre-
dominately Catholic, I, therefore, expect this variable to show a positive coef-
ficient, as Catholic voters should be more likely to vote.
Fourth, another demographic variable I consider is ethnic identity/race.
For the Africa region, I rely on the salience of ethnicity or the weight that
respondents or citizens place on ethnicity as the core identity, usually in com-
parison to other identities.27 The salience of ethnicity is measured by asking
respondents to choose between ethnic identity and national identity. In the
Afrobarometer the question reads as follows: Let us suppose that you had
to choose between being a (Ghanaian/etc.) and being a (respondent’s iden-
tity group), which of these two groups do you feel most strongly attached
to? The response options range across ethnic identity only, ethnic identity
more than national, national/ethnic identity equal, national identity more
than ethnic, and national identity only. To capture the salience of ethnicity
in Africa, I created a three-category variable that identifies those with Only
Ethnic Identity (coded 1), those with Equal Identity (coded 0.5) and those with
Only National Identity (coded 0). Based on the hypothesis, we would expect
voters with ethnic identity to vote more frequently.
For the East Asia cases, I rely on a question in the AsianBarometer that
taps or measure nominal ethnic identity and reads as follows: what is your
racial or ethnic background? Focusing on three East Asian countries (South
Korea, Taiwan, and Mongolia), I have a list of 14 groups across South Korea,
Taiwan, and Mongolia. I, therefore, classify these groups into two comparable
groups coded 1 for the largest ethnic group and 0 for the minority ethnic group.
The majority ethnic groups constituted 20 per cent or more of the national
population (i.e. Korean 33 per cent belonging to South Korea; Khalkh 27.7
per cent belonging to Mongolia and; Minan 24.6 per cent belong to Taiwan).
The minority ethnic group constitute 5 per cent or less of the population
(i.e. this included the following groups: Korean Chinese, Vietnamese, Kazak,
Hakka, Mainlander, Aboriginal, Vietnamese, Indonesian, Overseas Chinese,
Filipino, and others).
Turning to Latin America, relying on the AmericasBarometer, the question
that captures ethnicity reads as follows: do you consider yourself white, mes-
tizo, indigenous, black, mulatto, or of another race? To capture ethnicity,
I created a four-category variable where other ethnic groups are coded 0;
blacks/indigenous people are coded 0.5, mestizos/mulattos coded 1, and whites
coded as 2. Based on the theoretical discussion, I would expect white voters
to be more likely to participate. Rounding up with Central/Eastern European
countries to capture race or ethnicity, I include a dichotomous variable that
measured whether a voter belongs to a minority ethnic group coded 0 and 1
if the respondent is member of the largest ethnic group. Remember, I expect
that individuals from the largest ethnic group would tend to vote more.
Fifth, the last demographic variable I consider is the place of residence
(i.e. urban and rural divide). All four surveys have approximately the same
question words that reflect an individual place of residence. Thus, to account
for the difference in voting behaviour between urban and rural voters, I con-
struct a dummy variable for voter’s place of residence, with voters living in
urban or large cities coded 0 and those living in rural or small towns coded 1.
Socioeconomic variables
Drawing on our previous discussion (i.e. theoretical chapter), two standard
indicators that capture socioeconomic status are income and education.
Unfortunately, in this study, I do not consider respondents’ income as a
measure of voter’s socioeconomic status because of a large number of missing
values and, in addition, because of its problematic measurement in Africa.28
Thus, for the sake of ensuring comparability across the four geo-political
regions, I have therefore omitted income and focused more directly on edu-
cation. In all four surveys, citizens are asked about the highest level of edu-
cation they have completed or achieved. Thus, the different educational
systems and degrees allow me to create three educational categories: pri-
mary/low education (coded 0), secondary/medium education (coded 0.5),
and higher/tertiary education (coded 1). By recoding a respondent’s level of
education into these three categories, a higher value reflects a higher level of
education, which I would expect to be associated with a higher propensity
to vote in elections.
Mobilisation variables
For the mobilisation model, I consider three variables, which are all geared
towards testing this theoretical perspective. These variables include party
identification, civil society organisations/associational membership, and clien-
telism. First, beginning with party identification, I focus on a classic question
that looks at respondent’s closeness to political parties. In the Afrobarometer,
AmericasBarometer, AsianBarometer, and European Social Survey, party
identification is captured by a survey question that asks the respondents if
they feel close to any political party. I collapse the responses into a dummy
variable: no, not close to any party is coded as 0 (reference category), while
yes, feel close to a party is coded as 1. Once more, I expect the effect of the
dummy variable to be positive, indicating party identifiers are more likely to
participate.
Second, civil society organisation/ associational membership is also
included in our model. Concerning the African cases, this variable is measured
by a question that asks about membership in voluntary association or com-
munity group. I recoded this variable into three categories: 0 = Not a member,
0.5 = Inactive member, while Active member/Official leader29 was coded
1. For the Latin America cases, I combine the different question measuring
associational membership in four types of social groups, namely: religious
organisations, parent association, community improvement groups, and pol-
itical movements/parties. Consequently, I construct an additive scale of mem-
bership index that runs from no participation (0) to weekly participation in all
social groups (1). Turning our attention to the AsianBarometer, civil society
organisational membership is measured by a question that asks respondents
if respondents are members of any organisation. I, therefore, construct a
dummy variable with non-members (coded 0) and members (coded 1). For
post-communist countries, membership in voluntary or informal networks is
measured by evaluating a respondent’s membership in trade unions or similar
organisations. Thus, a code of 1 is assigned for voters who declared to be
members and 0 for otherwise. Here again, I expect the coefficient for asso-
ciational membership to be positive as those who declared to be members
would be more likely to participate.
Third, the last mobilisation variable that I examine is clientelism. I rely on
the surveys and include a measure of whether the respondent has received
any gift or promise from a political party to turn out. In the Afrobarometer,
the question measuring clientelism reads as follows: in your opinion, how
often do the following things occur in this country’s elections: Voters are
bribed (see Table B1 in appendix on the variable is coded). While in the
AmericasBarometer, the question measures clientelism through vote-buying
reads as follows: thinking about the last presidential elections, did someone
offer you something, like a favour, gift or any other benefit in return for your
vote or support? Based on these questions, I, therefore, created a dummy
variable with 1 assigned for voters who have been offered a gift in return for
their votes and 0 people with no experience (who never experience receiving
gifts/brides). I expect this variable to take a positive coefficient, suggesting
that voters who were offered gifts/brides would be more likely to participate
However, in the East Asian and post-communist cases, no question focuses
on the clientelism.
Political-psychological variables
A review of the existing literature on electoral participation pinpointed sev-
eral political-psychological variables that I intend to test concerning newly
consolidated democracies. These variables consist of political interest, polit-
ical trust, and political efficacy (internal/external efficacy).
First, starting with political interest, I measure this following the clas-
sical way of asking people how interested they are in politics or public
affairs (Gabriel and Van Deth, 1995, p. 395). In all four regions, political
interest is captured with an almost identical question. Accordingly, polit-
ical interest is measured in a four-point scale that runs from not interested
(0) to very interested (2). Second, turning to political trust, once again, I rely
on a standard survey question that tap respondents’ trust in several political
institutions (president, national assembly, political parties, national police,
etc.). Thus, for the sake of comparability across all regions, I created a polit-
ical trust index that runs from low/no trust to great/high trust.
Third, political efficacy is measured by items intended to capture internal
and external efficacy. Beginning with the Africa region, external efficacy
is measured by a question that taps individuals’ subjective assessments of
whether elections allow voters to remove from office leaders that people
do not want. The external efficacy variable within the African region is
therefore measured using a four-point scale that runs from 0 to 2, with a
low value meaning no or little efficacy and a high value indicating voter
is efficacious. While, regarding internal efficacy, this is measured by a
question which reads as follows: elections ensure voters’ views are reflected
and like the external efficacy indicator, it is measured with a four-point
scale running from 0 to 2. Next for the Latin America cases, I rely on the
question of whether respondents believe those who govern their country
are interested in what people like you think as a measure of external effi-
cacy. The variable is measured on a seven-point scale (1 = strongly dis-
agree to 7 = strongly agree) so that those who declared to strongly disagree
are considered as those with little or no efficacy. While those who strongly
agree are considered as efficacious individuals. Concerning internal effi-
cacy, this variable is measured by a question that taps respondents’ sub-
jective assessment of their understanding of the most important political
issues (i.e. measured on a seven-point scale).
In the East Asian region, external efficacy is measured by a question that
reads as follows: how well do you think the government responds to what
people want? I recoded the different responses into a four-point scale that
runs from not at all responsive (0) to very responsive (2), with a lower score
indicating a voter less efficacious and higher range meaning the voters is effi-
cacious. On the other hand, internal efficacy is captured by a question that
asks respondents if politics/government is so complicated that a person like
me can’t understand. This is equally coded into a four-point scale running
from 0 (strongly agrees politics is complicated) meaning the voter is less effi-
cacious to 2 (strongly disagrees politics is too complicated), which implies the
voter is more efficacious.
Lastly, in post- communist European countries, internal efficacy is
measured by the following questions: how confident are you in your ability to
participate in politics? I maintain the original categories but recoded: Not at
all confident30 coded as 0, A little confident coded 0.5, Quite confident coded
1, Very confident coded as 2 and Completely confident coded 3.31 Rounding
up with external efficacy, I yet again rely on the following question: and how
much would you say that the political system in your country allows people
like you to have an influence on politics? Recoded 0 = Not at all; 0.5 = Very
little; 1 = Some; 2 = A lot; 3 = A great deal.
Alternatives variables
To test an alternative model of vote choice in newly consolidated democracies,
I operationalise a handful of variables, including the economy, corruption,
and authoritarianism. First, the country’s economic variable is captured in
the African, Latin American, and East Asian regions as follows: looking
back, how do you rate the following compared to 12 months ago: economic
conditions in this country? Or, how would you describe the change in the
economic condition of our country over the last years? I, therefore, collapse
the different response options into three categories: worse (0), same (0.5), and
better (1). Turning to the Eastern European cases, this question is captured
as follows: how satisfied are you with the present state of the economy in-
country? This is measured in a ten-point scale from extremely dissatisfied with
the country’s economic situation (0) to extremely satisfied with one’s country
economic situation (10).32
Second, I include a corruption variable that taps people’s perception or
experience with corruption in the different regions. Beginning with sub-
Saharan Africa, corruption is measured based on an additive index of seven
similarly worded questions about the respondents’ perception of corruption
in their country. The question reads as follows: how many of the following
people do you think are involved in corruption, or haven’t you heard enough
about them to say: (a) office of the presidency, (b) members of parlia-
ment, (c) government officials, (d) local government councillors, (e) police,
(f) judges and magistrates, and (g) business executives. The response options
were: 0 = None, 1 = Some of them, 2 = Most of them, 3 = All of them.
Moving to the Latin America region, the question about corruption reads
as follows: is corruption among public officials very common? The response
options are 1 = very common, 2 = common, 3 = uncommon, 4 = very
common. Due to the small response rate across the four categories, I combine
the different categories and therefore created a dummy variable with 3 and 4
coded as 0 (corruption uncommon), while 1 and 2 were coded as 1 (corruption
common). For the East Asia cases, the question that measures corruption
reads as follows: Have you or anyone you know personally witnessed an act
of corruption or bribe-taking by a politician or government official in the past
year? 1 = Personally witnessed, 2 = Told about it by a family member who
personally witnessed, 3 = Told about it by a friend who personally witnessed,
4 = Personally never witnessed, 5 = No one I know has personally witnessed.
Relying on these different categories, I recoded people’s experience with
corruption into a dummy variable coded 1 = having witness corruption and
0 = never witness corruption. Rounding up with post-communist European
countries, our measure of corruption is derived from the Transparency
International corruption perceptions index (i.e. as no question in the survey
measure people’s experience/perception of corruption at the individual-level).
The index measured the perceived level of public/political corruption on a
scale of 0 (highly corrupt) to 100 (very clean).33
Third, our measure of authoritarianism is derived from the democratic
erosion event dataset34 that captures the symptoms of democratic erosion (i.e.
media repressions, curtailed of civil liberties, repression of the opposition,
reduction in judicial independence, etc.) across countries/regions over time.
Thus, relying on expert coding, the question measuring democratic erosion
asks: what is the overall erosion of a country’s democracy? Based on a four-
point scale the response ranges from 0 = No backsliding, and weak threat
of future backsliding; 1 = There are precursors to backsliding (e.g. the rise
of extremist parties, but erosion of democratic institutions has not yet taken
place); 2 = There is weak erosion of democratic institutions; 3 = There is
moderate erosion of democratic institutions; 4 = There is a severe erosion
of democratic institutions. Drawing on this data, I recoded authoritarianism
into a dummy variable where 0 and 1 are combined and coded as 0 (i.e. coun-
tries not experiencing democratic erosion). Note, for the countries we analyse
here no country was placed under the 4 category (i.e. 4 = there is a severe
erosion of democratic institutions). So, 2 and 3 are combined and coded as 1
(i.e. country experiencing democratic erosion).
Political and institutional variables

The theory underpinning this study proposes that the decision of people to par-
ticipate in electoral politics is both a function of an individual’s characteristics
and the political context in which the individuals find themselves. Drawing
from this argument, I, therefore, include several contextual or institutional
variables that are believed to influence a citizen’s propensity to vote: electoral
system, closeness of an election, concurrent elections, and compulsory voting.
First, beginning with the electoral system, this is measured by the least-
squares index35 (LSq), which measures disproportionality between the vote
distribution and the seat distribution (i.e. a low score indicates the electoral
system is highly proportional, while a high score signifies the system is highly
disproportional). We expected the coefficient of this variable to be negative
voters in the more disproportional system will be less likely to vote.
Second, I include in my analysis measures of whether legislative and
presidential elections are held concurrently. A dummy variable is created to
measure this variable. A code of 1 is attributed if the elections are concurrent
and 0 if the elections are not concurrent.
Third, I also include closeness of election as a contextual variable that will
affect people’s propensity to vote. The closeness of an election has generally
been considered likely to spur higher turnout among voters because parties
are more likely to mobilise people when the election is perceived to be a tight
race and because people see a higher likelihood of their vote making a diffe-
rence (Pérez-Liñán, 2001). Relying on earlier studies, I measure closeness of
election as the margin of victory for the winning candidate or over the runner-
up in presidential elections (Franklin and Hirczy de Miño, 1998), while for
parliamentary democracies, I measure closeness of election as the difference
in seat shares between the top two parties. For simplicity’s sake and to ensure
comparability across the four regions, I take a 6 per cent threshold as a diffe-
rence in the seat shares between the top two parties or between the runner-up
in presidential elections. That said, I coded the variable in such a way that a
winning margin of less 6% = 1, and a margin greater than 6% = 0.
Lastly, compulsory voting is coded as a dummy variable. With a code
of 0 = no compulsory election/ compulsory election without enforce-
ment; 1 = compulsory elections with enforcement. In conclusion, the
operationalisation of all variables included in the empirical analysis that
follows has now been explained in detail. The coding schemes that are
employed are based on the hypotheses formulated in the previous chapter and
the expected effects pertaining to citizens’ propensity to turn out in elections.
Empirical and estimation strategy

I conclude this chapter with a few words about the estimation strategy I adapt
to investigate my vote choice model. First, recall that my dependent variable
(voter turnout) is dichotomous as respondents were asked to declare whether
they have voted in the most recent national election in their country. This is
coded as 1 for voters who reported having voted and 0 if the respondent did
not vote. Because our dependent variable, voting, has only two values (“yes,
I voted” and “no, I did not vote”), I cannot conduct ordinary least square
(OLS) regression. An appropriate form of analysis is binary logistic regres-
sion, commonly known as logit analysis, which will allow us to estimate the
minimal effect of each variable on the dichotomous dependent variable while
controlling or holding the other independent variables constant. The foremost
advantage of using a multivariate technique, most especially binary logistic
regression, to explain people’s propensity to vote can be described as follows:
First, as a statistical model, Healey (2012, p. 406) argues that there are two
important reasons for utilising multivariate techniques: the need to simply gather
additional information about a specified bivariate relationship by observing
how the relationship between the dependent variable and an independent vari-
able is affected by the presence of a third, fourth, or fifth variable. Second,
the issue of establishing causation is deemed another reason that could jus-
tify the use of multivariate analysis. As Healey explains, multivariate statistical
techniques cannot prove that causal connections exist between variables; how-
ever, they can provide useful evidence in support of a causal argument, which
is considered an essential tool for testing and revising theory by controlling
for multiple potential sources of causation. Acton et al. (2009, p. 257) advance
similar reasons for using logistic regression. First, they maintain that an analyst
can employ logit regression simply because the goal may be to discover which
of a list of independent variables continues to exert a significant causal effect
upon the dependent variable once all of the potential candidate variables are
included in the same model. Second, the use of logit regression is also motivated
by the need to gauge the strength and direction of any interaction between the
independent variables in their effect on the outcome.
Second, I should equally recall that my interrogation of voting behav-
iour across newly consolidated democracies is structured according to the
main theoretical framework (i.e. socio-demographic, mobilisation, political-
psychology, alternatives explanations, and institutional factors). With this
ambition in mind, I, therefore, proceed as follows: first, I begin with a simple
baseline model that applies country dummy variables for each of the regions
(i.e. given that there are only a few elections, and country- level specific
variables cannot be properly tested) and the country election year or month
that helps to account for the different times the countries elections were
conducted in relation to the survey. Second, and most importantly, I add a
block-recursive or causal sequence specification technique in which I add to the
baseline model variables relating to our main theoretical framework building
step by step (i.e. first only the socio-demographic variables, second only the
effect of socio-demographics + the mobilisational variables, third only the
socio-demographic + mobilizational + political psychological variables and
fourth only the socio-demographic + mobilizational + political/psychological
variables + alternative model variables, etc.).
This estimating strategy allows me to estimate the contribution of each
of the four theoretical models to an overall accounting of individual-level
turnout in newly consolidated democracies. Third, and finally, in line with my
theoretical framework set up in Chapter 2, I introduce the contextual factors
that condition the role individual factors play in new democracies (i.e. all the
individual-level variables combined + the institutional/contextual variables).
Rounding up with the institutional or electoral context, permits me to assess
how the electoral context conditions the effects of each block of individual-
level predictors in my block recursive or causal sequence.
Notes
1 This period can be described to have ushered in an unprecedented political change
and reforms across these regions.
2 As cited by Brigevich (2008, p. 1) the collapse of the Soviet Union in the 1990s
discredited the viability of authoritarian regimes and one-party systems in Central
and Eastern Europe, Latin America, parts of Asia and sub-Saharan Africa, and
replacing them with governments determined to enact pro- democratic, liberal
reforms.
3 This book draws upon four different cross-national public surveys.
4 Harkness et al. (2003) outline two strategies to established equivalence by
adocating first whether constructs are, in fact, comparable and second whether the
instruments that are useful in one context are equally so in another.
5 https://freedomhouse.org/report/methodology-freedom-world-2018 [Accessed 18/
06/2019].
6 Tunisia is, however, excluded since data for this country were only collected three
years after Tunisia democratised. Also, although Senegal does qualify as a newly
consolidated democracies, it is, however, excluded as no question relating to voter
turnout was ask in the country.
7 I do not considered Japan as a new democracy and, therefore, my analyses in the
subsequent chapter will exclude Japan.
8 Here, I refer to only national elections that could be presidential or parliamentary
elections depending on the country.
9 Though Botswana is considered a stable democracy that has been dominated by
a single party since independence. The decline in its political rights index should
probably be down to tight control over media freedom and the discrimination of
migrants, refugees, and LGBT groups.
10 Hungary experienced a decline in its political right index due to increasing intimi-
dation of civil society groups and the opposition that has largely being orchestrated
since the Prime Minister Viktor Orbán’s Alliance of Young Democrats–Hungarian
Civic Union (Fidesz) took over power.
11 www.cses.org/datacenter/download.htm [Accessed 18/06/2019].
12 www.worldvaluessurvey.org/WVSContents.jsp [Accessed 18/06/2019].
13 http://w.issp.org/about-issp/ [Accessed 27/03/2019].
14 The four-survey data across the regions do not contain the same question wording
for all independent variables, and so they cannot be pooled into a single dataset
for analysis. Rather, I analyse each dataset separately, while trying to keep the
variables as similar as possible given that some variables have different questions
wordings.
15 The AmericasBarometer is coordinated by a team at Vanderbilt University, which
hosts the Latin American Public Opinion Project (LAPOP).
16 More technical information on the Afrobarometer, Latinobarometer,
AsianBarometer, and European Social Survey, can be obtained from their
websites: www.afrobarometer.org/survey-and-methods/samplingprinciples, www.
vanderbilt.edu/lapop/about-americasbarometer.php, www.asianbarometer.org/
survey/survey-countries, www.europeansocialsurvey.org/methodology/ess_meth-
odology/sampling. [Accessed 16/03/2019].
17 Important to note though, Round 7 (2016/2017) was completed and released
probably by the end 2020.
18 However, as previously specified in our case selection, we focus only on Latin
America countries that classify as free or as liberal democracies.
19 By summarising these materials here, it becomes possible to reduce the amount of
methodological justification offered in each of the individual chapter.
20 This problem arises as people tend to overestimate the turnout in surveys.
21 Turnout intention is likely to be even more prone to social desirability bias and is,
therefore, not often used.
22 In their meta-analysis of individual level research on voter turnout, Smets
and van Ham (2013) state that 82 per cent of the studies measured turnout as
reported turnout, 11 per cent as validated turnout and 7 per cent as turnout
intention.
23 The question wording from the four surveys is reported in Appendix 2 (Table
B1 and B2), while the description of variables is reported in Table B3 (i.e.
Appendix 2).
24 Wolfinger and Rosenstone (1980, p. 115) argue that the discrepancy between offi-
cial turnout and reported turnout in surveys can be explained by misreporting
of respondents who are reluctant to admit they did not vote because it is more
socially acceptable to be fulfilling one’s duty as a citizen by voting. Also, it may be
that the surveys are not randomly conducted and were therefore biased towards
voters. Curtice (2007, p. 902) has strongly criticised the survey (i.e. CSES survey)
for not implementing random sampling. Lastly, spoiled ballots may not have
been counted in the official totals. Finally, with respect to sub-Saharan Africa,
Isaksson (2010, p. 6) argues that the discrepancy between reported turnout in
survey and official/actual turnout percentages may arise because voting survey
questions simply ask respondents whether they voted in the last election or
not without taking into consideration whether the elections were concurrently
staged or not. For example, in a few cases like Ghana where the presidential and
parliamentary elections were held concurrently, the Afrobarometer does not tell
us with precision which of the elections the respondents voted for in the last
election.
25 The precise questions wording in the surveys across the four regions are presented
in Appendix 2 (e.g. Table B2).
26 This is because I have access only to the free or reduced version of the
AmericasBarometer data.
27 For an overview see; Bhavnani and Miodownik, 2009; Eiffert et al., 2010.
28 In the African region, in the Afrobarometer there is no question that captures
a respondent’s income, unfortunately. Previous studies (Bratton et al., 2005;
Isaksson, 2014; Kuenzi and Lambright, 2011; Tambe, 2017a), have simply used a
question in the Afrobarometer about access to food as a proxy for income (Over
the past years, how often, if ever, have you or your family gone without food?).
However, Maydom (2017) caution about the use of access to food (e.g. how often
respondents have gone without food) as a proxy to income as it might be capturing
other factors other than income. Relatedly, I would argue that trying to measure
voters’ socioeconomic status using access to food is very unconvincing given that
almost 80 per cent of individuals in these countries depend on agriculture or
farming for their basic subsistence. Moreover, in these countries it is a common
tendency for people to borrow from their neighbours if they cannot afford food
for themselves.
29 The response rate for both categories was small, thus enabling us to combine both.
30 “Not at all confident” in the sense of “having no confidence at all in your own
ability”.
31 A low value or score means little or efficacy, while a higher value means more
efficacy.
32 Since the mid-point of the scale lies between 5 and 6 and for the sake of compar-
ability across other regions, we, therefore, collapse the response options into three
categories: 0–4 on the scale was assigned as worse (0); 5–6 was coded as same (0.5);
while those in the top half (7 to 10 on the scale) was assigned better (1).
33 To ensure comparability with the other regions, the score was rescaled from 0 to 1
34 www.democratic-erosion.com/ [Accessed 13/04/2019].
35 For details see Gallagher (1991); Gallagher and Mitchell (2005).
4
Socio-demographics and the vote
Building on the very first studies of voting and or electoral behaviour (i.e.
Lazarsfeld, Berelson and Gaudet 1948; Berelson, Lazarsfeld and Mcphee,
1954; Campbell et al., 1960), scholarly literature on voting from older dem-
ocracies from the global North1 (i.e. Western Europe and North America)
gave enormous importance to voters’ socio- demographic or descriptive
characteristics in predicting individual-level turnout. Thus, in this chapter,
my examination of why people vote in newly consolidated democracies starts
by looking at the effect of six key socio-demographic indicators (e.g. age,
gender, ethnicity/race, religion, place of residence, and education) on vote
choice. Recall that while the primary interest lies in explaining the voting deci-
sion in newly consolidated democracies, I include Western European cases so
as to compare the experiences of people’s voting decisions in sub-Saharan
Africa, Latin America, East Asia, and Eastern European countries. That is,
are the socio-demographic factors that influence voters’ decisions to turn
out in elections in these newly consolidated democracies broadly similar to
those that we find in the established democracies of Western Europe? Having
identified the appropriate empirical indicators, I now move on to empirical
analysis, where I report bivariate and multivariate associations between the
independent variables (i.e. socio-demographics) and the dependent variable
(voter turnout).
Socio-demographics determinants of the vote: bivariate relationships

As stated above, this chapter begins with cross-regional bivariate correlations,
showing the relationships between voting and the socio-demographics inde-
pendent variables that have been identified from the previous theoretical
section as having an effect on people’s decision to vote or not to vote. I argue
that the motive for participating in electoral politics in these broad geograph-
ical regions may either be similar or differ and therefore generalisations that
may apply to a particular region may not apply to another. Thus, these simple
bivariate analyses will enable us to get an initial impression of whether or not
there are differences between the established and newer consolidated dem-
ocracies, or whether there are differences among the newer democracies as
Socio-demographics and the vote 85
well with regards to the determinants of electoral participation. Table 4.1
presents the bivariate relationships between turnout and the various socio-
demographic indicators across the different regions (i.e. Table C1, Appendix
3, reports results for descriptive statistics across the different geo-political
regions).
Table 4.1 shows the measures of association using Spearman’s rho,2 which
tells us about the direction of the relationship. Similarly, Spearman’s rho also
indicates the strength of the relationship between two variables, with a value
close to 1 indicating a strong relationship, while a value close to 0 signifies a
weak relationship. Note, the bivariate association that typifies our Western
European cases allows us to get a better understanding of a baseline against
which to compare newly consolidated democracies.
First, starting with the demographic’s indicators, a very strong age effect
that has been observed and documented in Western European democracies
(e.g. see Smets and van Ham, 2013) is also present in newly consolidated dem-
ocracies. A positive and strong association between age and voting is seen
across the four regions, as older citizens tend to participate at higher rates
than younger people. In sum, this initial bivariate result does suggest that
voting decision is largely a product of life-long processes and political experi-
ence/maturation.
Second, gender produced a much less convincing finding across the
different regions. The proposition regarding gender and voting was that men
would tend to participate at higher rates than women. Instead, the results
across all regions suggest this proposition is incorrect. What is uncovered is
that in the newer democracies of Eastern Europe and sub-Saharan Africa
(i.e. which is similar for Western European democracies) the relationship
between voting and gender is non-existent. In contrast, for Latin America
and the East Asia region, the results suggest women are at least more likely
to participate than men. Although these results are directly opposite to what
we have expected, they seem to be line with previous studies that suggest that
the gender gap in turnout has gradually disappeared (see e.g. Kostelka, Blais,
and Gidengil, 2019; Alexander and Coffé, 2018; Carreras, 2018; Coffé and
Bolzendahl, 2011; Childs, 2004; Inglehart and Norris, 2003).
Third, turning to religion, I find as expected a positive effect of religion on
electoral participation across all regions. However, the relationship is quite
weak in two of the newer democracies, sub-Saharan Africa and Latin America
(as in Western Europe), except for post-Communist European and East Asian
democracies. Overall, voters who frequently attend religious services or who
identified as Catholics (i.e. concerning Latin America) are more likely to turn
out. These initial bivariate results do portray that newer democracies resemble
or are similar to older democracies of Western Europe in these respects.
Fourth, a look at the effect of ethnicity/race on electoral participation
across the different geographical regions reveals more diverse findings. In new
democracies of Latin America and Eastern Europe (as in Western Europe),
whiter voters and those who belong to ethnic majorities are significantly more
86 Socio-demographics and the vote
Table 4.1 Socio-demographic variables and vote choice in newly consolidated democ-
racies: bivariate relationship
Regions Socio-demographic variables Spearman’s

rho
Western Europe Age 0.17**

Gender (1 = men) 0.01
Religion (2 = Everyday/once a week; 0 = Never) 0.05**
Nominal ethnic identity (Largest ethnic group = 1; 0.09**
Minority ethnic group = 0)
Residence (1 = rural) 0.04**
Education (Higher education = 1; Secondary 0.14**
education = 0.5; Primary education = 0)
Sub-Saharan Age 0.16**
Africa Gender (1 = men) 0.00
Religion (2 = More than once a day; 0 = Never) 0.06**
Salience of ethnicity (Only ethnic ID = 1; Equal -0.03**
national/ethnic ID = 0.5; Only national ID = 0)
Residence (rural = 1) 0.05**
Education (Higher education = 1; Secondary -0.05**
Latin America Age 0.33**
Gender (1 = men) -0.03**
Religion (Catholics = 1; Others = 0) 0.07**
Ethnicity/race (Whites = 2; Mestizo/Mulattos = 1; 0.02*
Blacks/Indigenous = 0.5; others = 0)
Residence (1 = rural) -0.01
Education (Higher education = 1; Secondary -0.01
East Asia Age 0.20**
Gender (1 = men) -0.04*
Religion (2 = Several times a day; 0 = Never) 0.11**
Nominal ethnic identity (Largest ethnic group = 1; -0.03
Residence (1 = rural) 0.03
Education (Higher education = 1; Secondary 0.01
Post-Communist Age 0.18**
Europe Gender (1 = men) 0.01
Religion (2 = Everyday/once a week; 0 = Never) 0.13**
Nominal ethnic identity (Largest ethnic group = 1; 0.05**
Residence (1 = rural) 0.00
Education (Higher education = 1; Secondary 0.15**
Notes: Bivariate results between voting and the socio-demographic variables across the regions
are derived from the European Social Survey, Afrobarometer, AmericasBarometer, and the
AsianBarometer. The observations are all weighted using the different surveys weights. Notes:
**Correlation is significant at the 0.01 level (2-tailed); *Correlation is significant at the 0.05 level
(2-tailed).
likely to turn out. For East Asia, the relationship between ethnicity and vote
choice is non-significant. Rounding up with sub-Saharan African democra-
cies, our data for the African cases show that there is a relationship between
ethnic identity and voting, but the result is surprising, revealing that voting in
Africa is not based on ethnic considerations since respondents or citizens with
ethnic identities were less likely to have voted when compared to those with
simple national identities. In the face of these findings, I, therefore, conclude
post-communist, and Latin America democracies do resemble older democ-
racies, but East Asia and African democracies are different.
Fifth, if we turn to the effect of an individual place of residence, the prop-
osition was that citizens living in rural areas would be more likely to partici-
pate. That said, looking at the preliminary bivariate results suggest this is true
only for African cases while for the remaining regions (Latin America, Eastern
Europe, and East Asia) the relationship between voting and whether an indi-
vidual resides in a rural or urban area is non-significant. Finally, concluding
with socioeconomic variables (i.e. education), while the result tells of a signifi-
cant relationship between voting and education in Eastern European democ-
racies (i.e. similar in Western Europe), for East Asia and Latin America our
data indicate a non-effect between education and turnout. Rounding up with
the Africa region, the results are even more astonishing given that those with a
higher level of education were less likely to vote.
So far, using bivariate statistics, I have now tested six different socio-
demographic groups of factors that I theorise will influence people’s decision
to vote across newly consolidated democracies in four geopolitical regions.
The bivariate analysis suggests some relationship (although very limited in
strength at least at the bivariate level) and quite a lot of agreement across the
regions but with some inconsistency too. Thus, to be more confident of this,
and to explore the sources of variation more rigorously, one needs to engage
with multivariate analysis of vote choice, including all six socio-demographic
characteristics.
Socio-demographics determinants of the vote: multivariate analysis

In the previous section, through a collection of bivariate statistics we saw that
certain socio-demographic factors tend to influence citizens’ decision to vote
in older as well as newly consolidated democracies and that while some of
these socio-demographic effects appear to be common in all regions, others
are more inconsistent. What is more, it is likely that some of these socio-
demographic variables that I have spelt out as influential in shaping people’s
decisions about whether or not to vote are somehow correlated with each
other3. It is, therefore, possible that the variables derived from the different
socio-demographic perspectives may overlap with each other. The desire to
eliminate the overlap and complex interrelatedness of these independent
variables requires multivariate analysis to enable one to produce estimates
that are statistically sound in predicting people’s decision to vote in newer
democracies. This section, therefore, constitutes a shift from bivariate statistics
to multivariate analysis. That said, following the estimation strategy I adopt
to evaluate why people vote in newly consolidated democracies through a
block-recursive or systematic stepwise-specification, Table 4.2 reports the first
block of our multivariate analysis which consists only of socio-demographic
variables and is expressed as follows:
Vote Choice = f (Socio − Demographics Characteristics )
Vote Choice = f ( age, gender, religion, residence, ethnicity,

and education ).
Table 4.2 reports the multivariate analyses in which all of the socio-
demographic variables are added simultaneously. This multivariate result is
remarkable because it will enable us to re-establish which of the variables will
continue to exert a positive and significant relationship with voting while con-
trolling for all of the others. The first model (Model 1) is the baseline model,
which includes a month’s variable (month since the last election) and country
dummies for each region.4 Thus, as a first step, I start by regressing vote
choice on months since the previous election and the country’s dichotomous
variables. This result shows that months since the last election has a negative
or no effect on vote choice in newer democracies of Eastern Europe, East
Asia, and Latin America except for sub-Saharan Africa (as similar in Western
European democracies). Next, the second column (Model 2, Table 4.2) shows
the socio-demographic variables of greater interest. This model presents the
standardised coefficient from the multivariate analysis of vote choice, with
standard errors in parentheses. In addition to the socio-demographic variables,
it includes the month variable and the country dummies for each region. Also,
the inclusion of the socio-demographic variables shows a marked improve-
ment in the model fit from the baseline model (i.e. Model 1). For example,
in Latin America, East Asia, and post-Communist European countries as
in Western European democracies the Nagelkerke pseudo R2 increases from
0.172, 0.136, and 0.146 respectively except for the sub-Saharan African region
that shows a low Nagelkerke pseudo R2 of 0.068. The logit models for each of
the geographical regions are interested in many regards:
First, starting with age, as expected, there is a highly significant overall
effect of age on voting in newly consolidated democracies of Africa, Latin
America, East Asia, and post-Communist European countries (as is gener-
ally true of established Western European democracies). Compared to young
people, older voters are significantly more likely to vote.5 These results do
not only strongly collaborate single country studies from the different regions
(Tambe, 2017a; Resnick and Casales, 2014; Kuenzi and Lambright, 2011;
Isaksson, 2010; 2014 (for Africa); Carrera and Castañeda-Angarita, 2014;
Seligson et al., 1995 (Latin America); Bernhagen and March 2007; Barnes,
2006 (post-communist Europe); Chung-li Wu and Tzu-Ping Liu, 2017; Mo,
Table 4.2 Logistic regression of socio-
demographic model and voting in newly
consolidated democracies
Regions/variables
Western Europe Model 1 Model 2
Months since last election 0.20*** (0.02) 0.19*** (0.02)

Age 0.03*** (0.00)
Gender (1 = men) 0.07 (0.04)
Religion (2 = Everyday/once a week; 0 = Never) 0.12*** (0.01)
Nominal ethnic identity (Largest ethnic 0.96*** (0.09)
group = 1; Minority ethnic group = 0)
Residence (1 = Rural) 0.15*** (0.04)
Education 0.56*** (0.03)
Constant -2.96*** (0.38) -5.74*** (0.42)
Nagelkerke pseudo R2 0.044 0.146
% correctly classified 83% 83%
Number of observations/cases 18657/10 18332/10
Sub-Saharan Africa
Months since last election 0.01*** (0.00) 0.02*** (0.00)
Age 0.38*** (0.04)
Gender (1 = men) -0.02 (0.06)
Salience of ethnicity (1 = Only ethnic ID; Equal -0.24** (0.10)
national/ethnic ID = 0.5; Only national
ID = 0)
Residence (1 = Rural) 0.28*** (0.07)
Education -0.07 (0.12)
Constant 1.62*** (0.35) 0.45*** (0.16)
Latin America
Months since last election -0.05*** (0.01) -0.03*** (0.01)
Age 0.05*** (0.00)
Gender (1 = men) -0.12*** (0.05)
Religion (Catholics = 1; others 0) 0.23*** (0.05)
Ethnicity/race (Whites = 2; Mestizo/ -0.02 (0.04)
Mulattos = 1; Blacks/Indigenous = 0.5;
others = 0)
Residence (1 = Rural) 0.19*** (0.05)
Education 0.73*** (0.07)
Constant 3.18*** (0.33) -0.04 (0.37)
East Asia
Months since last election -0.04***(0.01) -0.09*** (0.01)
Age 0.06*** (0.00)
Gender -0.07 (0.09)
(continued)
Table 4.2 Cont.
Regions/variables
Western Europe Model 1 Model 2

Nominal ethnic ID (Largest ethnic group = 1; -0.15 (0.15)
Residence (1 = Rural) 0.25* (0.13)
Education 0.44*** (0.06)
Constant 2.65*** (0.24) 0.65 (0.36)
Post-communist Europe
Months since last election -0.01 (0.01) 0.00 (0.01)
Age 0.03*** (0.00)
Gender (1 = men) 0.22*** (0.05)
Nominal ethnic ID (Largest ethnic group = 1; 0.45*** (0.09)
Residence (1 = Rural) -0.05 (0.05)
Education 0.60*** (0.03)
Constant 1.15*** (0.14) -1.66*** (0.21)
Source: European Social Survey (Western and Eastern European countries); Afrobarometer (Sub-
Saharan Africa); AmericasBarometer (Latin America); AsianBarometer (East Asia). The data
are weighted using the weights provided for each survey data. Country dummies are included for
each region but not shown. For Western Europe ten country election dummies included with the
United Kingdom as the reference category; for Africa 08 included with South Africa as the refer-
ence category; for Latin America, 08 included with Brazil as the reference case; For East Asia, 03
included with South Korea as the reference category; For Eastern European countries, 06 country
dummies are included with Poland as the reference case. Significance: ***p < 0.001; **p < 0.01;
*p < 0.05. Entries are standardized logistic regression coefficients, standard errors in parentheses.
Brady and Ro, 1991; Park, 2002; Denton, 2016 (East Asia)) but even so con-
firm the assertion that in Western democracies, most of the tests between age
and turnout are successful/indicating a positive effect of age (Smets and van
Ham, 2013).
Second, concerning gender, our results show gender is not a significant pre-
dictor of turnout in new democracies of sub-Saharan Africa and East Asia
(as in Western European democracies) except for post-Communist European
states and Latin America. In post-Communist European democracies, men
are 24 percentage points more likely to vote than women. However, a closer
look at the parameter estimates across the other newer democracies tells of
negative sign, and when gender is found to be significant (e.g. Latin America),
it is usually women who turn to vote at higher rates than men. (for similar
results see e.g. Only the work of: Carreras and Castañeda-Angarita (2014) as
it focus just on Latin America).
Third, looking at religion, there is an overall significant and positive rela-
tionship (i.e. in the direction expected by our hypothesis) between being reli-
gious and voting in newly consolidated democracies of Africa, Latin America,
East Asia, and Eastern Europe (which is similar to Western European dem-
ocracies). Given this result is remarkably consistent across all newer democ-
racies, it, therefore, confirms Smets and van Ham’s (2013) meta-analyses in
Western established democracies that attendance of religious services is found
to affect individual turnout roughly half of the time, with most studies having
a success rate that lies between 50 and 57 per cent.
Fourth, looking at ethnicity/race, just like the bivariate analysis, the result
from the multivariate analyses suggest the findings are quite nuanced across
the different regions. First, the results between ethnicity and voting across
two newer democracies (East Europe and Africa) tells of a significant rela-
tionship that is similar to those of Western Europe. However, the parameters
estimate reveals that voters in Eastern Europe (similar to Western Europe)
who belong to the majority ethnic group tend to vote at higher rates than
those of ethnic minorities. In sub-Saharan Africa, the result is significant but
negative revealing that voting in Africa is not based on ethnic considerations
as citizens with ethnic identities were less likely to have voted when compared
to those with simple national identities. Concerning Latin America and East
Asia cases, surprisingly, our multivariate results disconfirm the bivariate
results in that the relationship between ethnicity/race is non-existent across
both regions.
Fifth, the effect of an individual place of residence (i.e. Urban vs Rural)
equally produces some mixed results. For example, the results for sub-Saharan
Africa, Latin America, and East Asia (i.e. which is similar to those of Western
European democracies) are in line with those of Bratton, Chu, and Lagos
(2010); Carreras and Castañeda-Angarita (2014); Resnick and Casale (2014)
indicating voters in rural areas are significantly more likely to turn out at the
polls compared to their urban dwellers. However, concerning newer democra-
cies of post-communist European countries, the results suggest the relation-
ship between individuals’ place of residence and whether they voted in the last
national election is non-significant.
Sixth and finally, concluding with the core socioeconomic variable (educa-
tion), as expected education has a significant and positive relationship with
vote choice in three newer democracies: Latin America, East Asia,6 and post-
communist European countries, with the exception being sub-Saharan Africa,
where the link appears to be reversed.7 To be precise, the positive and signifi-
cant effect of education on vote choice is more in line with those of Western
European democracies. From this, one can infer considerable support for
the fundamental proposition that a higher level of education does help to
reduce the cognitive cost of voting. These cognitive costs such as registration
procedures, the ballot structure, filling out forms and even meeting deadlines
may depress the participation rate of poorly educated citizens who may find it
hard to understand the bureaucratic hurdles of voting, while the participation
rates of citizens who are highly educated may not be affected as they tend to
understand voting procedures more easily (see, Gallego, 2015).
Having concluded with the multivariate analyses, and recall the central
objective lies in knowing how far Western generated theories could travel to
newer consolidated democracies of Africa, Latin America, East Asia, and
post-communist countries, Table 4.3 provides a tabular presentation of a the-
oretical matching of the socio-demographic model on vote choice based on
the results obtained from the regions.
Thus far, the multivariate analyses presented in Table 4.2 provide us with
the real test of which socio-demographic variables matter for vote choice in
newly consolidated democracies. Undoubtedly, no matter how exciting these
results are, I argued the pooled analyses for each of the regions presented
might tend to hide potentially essential differences between the different coun-
tries considered in the analyses. To be accurate, it can be possible that across
the different regions, individual countries may turn to vary with regards to
different socio-demographic characteristics. For example, in sub- Saharan
Africa and or Latin America, some countries might have a larger rural popu-
lation, while others may have a large urban population. Relatedly, the ethnic
composition in some countries/regions (i.e. East Asia and Latin America)
Table 4.3 Theoretical matching of socio-demographic model on vote choice –based

on the multivariate results
Models/ Expected Western Sub- Latin East Asia Post-

variables effect/ Europe Saharan America communist
direction Africa Europe
Socio-demographic model
Age + Applies*** Applies*** Applies*** Applies*** Applies***
Gender (male) + Does not Does not Applies** Does not Applies***
apply apply apply
Religion + Applies*** Applies*** Applies*** Applies*** Applies***
Ethnicity/race + Applies*** Applies** Does not Does not Applies***
apply apply
Residence (rural) + Applies*** Applies*** Applies*** Applies*** Does not
apply
Education + Applies*** Does not Applies*** Applies*** Applies***
apply
Notes: Theoretical matching are based on results from Afrobarometer, AmericasBarometer,

Asianbarometer, and European Social Survey. Applies***: This signifies that the relationship
between the dependent variable vote and the independent variable in question was statistically
significant and our hypothesis confirmed; Does not apply: This points to the fact that the rela-
tionship between vote and the independent variable under investigation was not statistically sig-
nificant and our hypothesis was rejected; Applies**: This shows that the relationship between
voted in the last election and the independent variable was statistically significant but not in the
expected direction, and therefore our hypothesis was rejected.
may be very distinct when compared to other countries within the same
area.8 Similarly, women may have been more emancipated in some regions
(i.e. Latin America) than in other regions, which therefore might explain why
they tend to vote more. That said, to effectively uncover some of the poten-
tial country differences across these four geopolitical regions that may have
hidden as a result of pooling survey data together, I proceed by running a
logistic regression for each of the individual countries across the four regions.
Doing this permits one to evaluate and compare whether the statistically sig-
nificant variables in the pooled model are also important predictors of voting
in the individual countries from the regions. Full details of individual country
results for each of the regions are reported in Appendix 3 (i.e. see Table C2).
Looking at the individual country results reported across the four regions
(sub-Saharan Africa, Latin America, East Asia, and Eastern European coun-
tries) that compare the role of each of the socio-demographic variables indi-
vidually, it is evident that an individual’s socio-demographic characteristics
play a more important role in certain countries than others in the different
regions.
First, country comparison for age indicates that older voters are, as
expected, more likely to vote compared to younger voters in 25 of the 26 newly
consolidated democracies across the four regions (i.e. Africa, Latin America,
East Asia, and Eastern European countries). This result strongly confirms age
is one of the strongest socio-demographic predictors of voter turnout.
Second, gender exerts no effect on the vote in 17 of the 26 countries
considered from the four regions. In sub-Saharan Africa, in one country
(Benin) men are more likely to vote compared to women and in two coun-
tries (Sao Tome and Principe and Botswana) women are more likely to
vote than men. Turning to Latin America, in three countries (Uruguay,
Guyana, and Brazil), women are more likely to vote than men. Concerning,
East Asia, in one country (Taiwan) women are modestly more likely to
vote than men. Finally, in Eastern European countries, in three countries
(the Czech Republic, Lithuania, and Poland) men are significantly more
likely to have reported voting than women. Overall, we can say that gender
does not have much of an effect on vote choice in newly consolidated
democracies.
Third, religion does not affect vote choice in 14 of the 26 countries from the
four regions. A more in-depth look at the individual country results from the
four regions reveals the following. In post-communist European states, religi-
osity exerts a statistically significant effect in all six countries (Czech Republic,
Lithuania, Poland, Estonia, Slovenia, and Hungary). In all six countries, the
result is strong, and all are in the expected direction. For East Asia, religion
exerts a significant and positive impact only in one country (Taiwan). For
sub-Saharan Africa, being religious is positively related to voting only in two
countries (Botswana and Ghana). Rounding up with Latin America cases,
remarkably being Catholics has a significant and positive effect on voting in
only three countries (El-Salvador, Costa Rica, and Chile).
Fourth, ethnicity/race also plays less of a consistent role for vote choice in
newly consolidated democracies. Ethnicity/race does not affect electoral par-
ticipation in 17 of the 26 countries across all four regions. The most exciting
results for the role of ethnicity were in Africa as it reveals voting across the
continent is not entirely based on ethnic consideration since voters with ethnic
identities were less likely to vote when compared to those with simple national
identities. However, in broder terms, this result may not be too surprising
given that in their meta-analyses of individual-level research on voter turnout,
Smets and van Ham (2013, p. 7) reported a modal failure for almost half of
the studies (90 studies) that included a race/ethnicity variable which implies
that a non-significant effect was found in most of the tests and studies with a
success rate of only 23 per cent.
Fifth, moving to the place of residence, we find that whether a voter lives
in a rural or urban area does not affect vote choice in 15 of the 26 countries
across the four regions. However, the result tends to vary across individual
regions. For example, in Africa, we uncover that in five of the eight coun-
tries (Benin, Sao Tome and Principe, Botswana, Ghana, and South Africa)
rural voters are more likely to vote as expected than urban dwellers. In the
remaining three countries, the results show a non-significant effect between
the place of residence and voting. For Latin America cases, three of the nine
countries (e.g. Panama, Argentina, and Chile) rural voters were more likely
to participate, except for this we found a non-significant relationship with
vote across the remaining six countries. Looking at East Asian countries, in
all three countries (South Korea, Mongolia, and Taiwan) the effect was insig-
nificant albeit being in the predicted direction of the supposed hypothesis.
Rounding up with Eastern European countries, in one of the six countries
(Lithuania) rural voters were more likely to turn out than urban dwellers.
Also, in two countries (Estonia and Poland) voters in urban areas were more
likely to have voted while in the remaining three countries the relationship was
non-significant.
Finally, education, just like age, also plays a more consistent role for vote
choice in newly consolidated democracies. As expected, voters with higher
levels of education were more likely to have voted in 16 of the 26 countries
across the four geo-political regions. To be precise, the positive role of edu-
cation on vote choice was powerful across three newer democracies (Latin
America, East Asia, and Eastern European countries) with the only exception
being sub-Saharan Africa where the effect of education was directly opposite
to what we have expected. An in-depth look reveals the following: in Latin
America, a significant and positive impact of education was evident in six of
the nine countries with the exception being Panama, Guyana, and Uruguay
where the impact of education on vote choice was non-significant but posi-
tive. Concerning East Asia and Eastern European countries, the positive and
significant effects of education were evidenced in all the countries. Finally,
coming back to sub-Saharan Africa, where the effect of education on vote
choice does not appear to be linearly related; still, one country (South Africa)
stood out as the exception (those with higher education were more likely to
vote). Except for this, across the remaining seven countries (Benin, Botswana,
Ghana, Namibia, Cape Verde, Sao Tome and Principe, and Mauritius), the
effect of education on voting was either non-significant, with voters with
lower levels of schooling showing a greater propensity to participate when
compared to those with higher education.
Conclusion
In this chapter, using bivariate and multivariate statistics, I have tested six
socio-demographics or descriptive characteristics on vote choice across four
geopolitical regions of the globe. Overall these results suggest quite a lot of
agreement across newly consolidated democracies but some inconsistency too.
To be specific, while the effect for some socio-demographics predictors of vote
choice across the four regions is more uniform (i.e. age, religion, education),
for others (gender, ethnicity/race, place of residence) there is less consistency.
Also, based on the region by region analysis derived from the pooled data, and
as reported in Table 4.3 (i.e. theoretical matching of the socio-demographic
model across newer democracies), one can make two inferences:
First, the analysis finds or suggests the transcendent importance of three
socio-demographic variables (age, religion, and education) on vote choice. To
be precise, the effect of age, religion, and education are not only very strong in
newer consolidated democracies but are largely similar in older democracies
of Western Europe. This would suggest that the socio-demographic models
that have been developed in the first of these types of countries (older democ-
racies) to explain why people choose to vote are general in many instances and
therefore have wide applications in democracies. Second, based on the value
of Nagelkerke pseudo R2 that shows the overall model fit when the socio-
demographic variables are added to the vote choice model, it would suffice to
say that socio-demographic model are strong determinants of vote choice in
Latin America, post-Communist European countries, and to a lesser extent
in East Asia democracies. In these three regions (as in Western democracies)
age, religiosity, and education are the most influential variables on electoral
participation. However, for sub-Saharan Africa, I cannot say that the socio-
demographic predictors are generally powerful determinants of vote choice.
Notes
1 See Lewis-Beck et al., (2008); Leighley and Nagler (2014) for a recent analysis on
how socio-demographics shapes individuals voting decisions.
2 Pearson’s P is not a proper correlation measure given that the main dependent vari-
able (voter turnout) is not continuous. In this context Spearman’s rho or Kendall
tau are more appropriate choices.
3 I thank an anonymous reviewer for pointing this point out as ethnicity tends to
interact with urban–rural residence in many countries while religiosity interacts
with gender.
4 Although I include country dummies across all regions, except for models 6 and 7,
we do not report estimated coefficient and standard errors in order to save space.
5 The odds ratios indicate that older voters, when compared to younger voters, are
3 per cent more likely to vote in Eastern and Western European democracies, 6
per cent more likely to vote in East Asian, 7 per cent more likely to vote in Latin
America and almost 46 per cent more likely to vote in sub-Saharan Africa.
6 This finding is in line with a more recent country study that focuses on Taiwan (e.g.
see Achen and Wang, 2017). However, it goes against previous studies by Tambe
(2016) and Bratton et al. (2010) that are more comparative (countries covered ranges
between 7 and 8) and includes both liberal and electoral authoritarian regimes.
7 See for example see Mattes and Mughogho (2010) for similar results that indicates
educated Africans are not the most likely to vote.
8 Although some of these contextual or country characteristics will be addressed
more fully in Chapter 8, it is still useful to highlight some of the difference here.
5
Mobilisation and the vote
The mobilisation model stipulates that to explain electoral participation,

one has to move outside the world of individuals to include unions, social
networks, and or political parties (Rosenstone and Hansen, 1993; Verba,
Schlozman, and Brady, 1995; Gray and Caul, 2000). Drawing on the role of
these mobilising agencies, in particular, this chapter considers the impact of
two standard variables that contrast voters with the non-voters: party iden-
tification and associational membership.1 To these two factors, I add clien-
telism,2 a third mobilisation variable, as its practice is quite common in new
democracies.
First, beginning with party identification, Langston (2007, p. 251) argues
party identification is essential in any democratic polity as it tells how well
voters are linked to parties. Wattenberg (1991) argues that if voters do not
hold party identification, then candidates and campaigns might become far
more important in explaining electoral outcomes. So far, there is now exten-
sive empirical literature reporting declining party memberships in Western
democracies (Van Biezen et al., 2012; Dalton, 2000). Yet, studies continue
to show partisan identification is still a strong predictor of voter behaviour
in older established democracies, as party identifiers are generally more
involved in electoral politics than non- identifiers (see, Lewis-
Beck et al.
(2008). Concerning new democracies, Mainwaring and Torcal (2006); Dalton,
Yun-han Chu, and Chull Shin (2007); Cheeseman, (2010) described parties
in new democracies as relatively new, organisational thin, and personalised
(i.e. with their creation, usually coincided with the rise of a new political
leader). Thus, in this chapter, the issue that concerns us is whether party iden-
tification serves a similar purpose in emerging democracies of sub-Saharan
Africa, Latin America, East Asia, and Eastern European countries as it does
in more established democracies. Second, evidence suggests the role of associ-
ational membership on vote choice in established democracies is rather weak
(see e.g. Smets and van Ham, 2013, p. 8). However, I must consider whether
a similar conclusion applies to newly consolidated democracies. Third,
I examine whether clientelism that is relatively common in emerging dem-
ocracies mobilises citizens to turn out on election day. This chapter proceeds
in two steps. The first section begins by tracking the trends or level of party
98 Mobilisation and the vote
identification in new democracies, and this is relevant considering the newness
of parties but their relative importance to the political system. The second
section investigates the impact of the mobilisation variables on vote choice via
bivariate and multivariate modelling.
Party identification in new democracies: levels and trends

As shown above, political parties are vital to contemporary democracies,
which therefore explains why Webb and White (2007) advocate for a more
stringent test for evaluating whether they fill the function that any stable
and effective democracy might expect of them. Given the role parties play
in national elections, it will not be entirely wrong to argue that people’s
attitudes towards political parties are not only vital in the assortment of their
feelings towards politics but also on how they tend to participate. That said,
though, political parties continue to remain very central to democratic pol-
itical systems and are thought to foster political participation. Yet, at the
turn of the 1970s, at least in the context of older established democracies
there was a debate about the alleged decline in party membership (see Webb,
1995). Thus, for our purpose, recall, the question that interests us is whether
a similar pattern in party membership decline applies to newly consolidated
democracies? Though one might question the validity of asking if parties in
transitional democracies are declining in any sense given that they have only
operated for a relatively short period. Still, I argue with new democracies
now having experience over 30 years of democratic experience, it might have
resulted in the institutionalisation of party systems, which does necessitate
us asking the same question about parties’ decline (see also Webb and White,
2007). Consequently, this section tracks the contemporary trend and levels of
party identification across the different geo-political regions. In qualitative
terms, this would enable us at least to see if there exists any variation in levels
of party identification across the different regions, but also to partly confirm
or reject the assertion of declining party identification that has been echoed
in the comparative literature.
Levels of party identification in new democracies

Table 5.1 compares party identification across the different geo- political
regions using the very first and the most recent cross-national survey that
forms the focus of this study.
First, I observe significant discrepancies in the frequencies of party iden-
tification in the different regions. Beginning with Western European coun-
tries, using the very first wave of European Social Survey (ESS, Round 1,
2002) consisting of 17 countries3 some 54 per cent reported feeling closer to a
particular party. Comparing the trends of party identification with the Round
8 of the European Social Survey conducted some 14 years later shows that
a total of 52 per cent of citizens reported being close to a political party.
Mobilisation and the vote 99
Table 5.1 Trends of the levels of party identification
Countries/regions Mean (% of party identification)
Earliest regional Most recent

survey regional survey
Sub-Saharan Africa
Benin N/A 0.41
Botswana 0.75 0.73
Cape Verde N/A 0.40
Ghana 0.67 0.66
Mauritius N/A 0.23
Namibia 0.71 0.79
Sao Tome and Principe N/A 0.63
South Africa 0.45 0.76
Latin America
Argentina 0.24 0.25
Brazil 0.25 0.22
Chile 0.21 0.12
Costa Rica 0.30 0.67
El-Salvador 0.40 0.44
Panama 0.32 0.45
Peru 0.19 0.19
Guyana 0.17 0.14
Uruguay 0.50 0.60
East Asia
Taiwan 0.54 0.48
Mongolia 0.89 0.81
South Korea 0.73 0.45
Czech Republic 0.47 0.36
Estonia N/A 0.40
Hungary 0.59 0.43
Lithuania N/A 0.26
Poland 0.29 0.30
Slovenia 0.36 0.35
Western Europe
Austria 0.55 0.47
Belgium 0.48 0.48
Germany 0.46 0.55
Finland 0.55 0.56
United Kingdom 0.51 0.54
Ireland 0.47 0.32
Italy 0.44 0.37
Netherlands 0.59 0.54
Norway 0.60 0.69
Sweden 0.68 0.64
Regional total
Sub-Saharan Africa 0.54 0.59
Latin America 0.31 0.35
(continued)
Table 5.1 Cont.
Countries/regions Mean (% of party identification)
Earliest regional Most recent

survey regional survey
East Asia 0.56 0.58
Post-communist Europe 0.42 0.36
Western Europe 0.54 0.52
Notes: Entries display the mean percentage of party identifiers in individual countries and overall
regional average. This is based on the earliest and most recent survey data across the different
regions. (i.e. Appendix 1 for information on earliest/most recent survey data). N/A: indicates the
country was a participating nation during the particular survey wave/round.
This result confirms party identification has been quite stable in established
Western democracies (i.e. amidst only a 2 per cent decline), which is not so
dramatic as proponents of party decline would want to believe. Second, for
post-Communist countries, using the same data and focusing on only four
countries (Czechia, Hungary, Poland, and Slovenia) included in the 2002
survey, our data suggest some 42 per cent of the citizenry reported being close
to a political party. Comparing the result with the 2016 survey (ESS, Round
8) indicates that only 36 per cent of citizens were revealed to be party identifiers
(i.e. 6 per cent decline party ID within 14 years). Overall, for post-communist
European countries, I can confirm party identification has remained relatively
low. Third, moving to sub-Saharan African, relying on the Afrobarometer
data (AB, Round 1) conducted between 1999 and 2001, across 12 countries4 54
per cent of Africans reported being party identifiers. Comparing party identi-
fication levels using a recent round of the Afrobaromter (Round 6, 2016) for
36 countries5 reports 59 per cent of Africans were declared to be members of
political parties. Fourth, looking at Latin America and relying on the second
wave of AmericasBarometer (LAPOP, 2008) for 16 Latin American coun-
tries6 reveals that 32 per cent of people were declared to be party identifiers.
Comparing this party identification level with a more recent survey conducted
in 2014 (Wave 6) reveals 36 per cent of respondents were declared to be iden-
tified with a political party, which is quite lower than what has been observed
in older established democracies of Western Europe. Rounding up with East
Asian cases, data from the first and most recent of the AsianBarometer survey
(i.e. Wave 1 and 4) conducted in 2001 and 2014 composed of 7 and 13 East
Asian countries suggest that 56 and 58 per cent of people report themselves
to be close to political parties.
Drawing from these party identification trends across all regions, we can
make three concluding remarks. First, the global trend in party identification
suggests that across the old and new democracies, the percentage of people
reporting ties to political parties is higher in Western Europe but surprising
too in sub-Saharan Africa, and East Asia. At the same time, it is lower in
post-communist Europe and Latin America. Second, while party identifica-
tion in Western Europe has been stable and declining for post-communist
Europe and Latin America, this has not been the case in other new democra-
cies, especially sub-Saharan Africa and East Asia. In these two regions, our
data suggest a surge in the number of people who report to be identified with
or close to political parties. Third, there seems to be a considerable variation
in levels of party identification across the different countries in each region.
That is, while some countries have experienced an increase in party identifiers,
others are experiencing a decline (i.e. this is even more evident across the Latin
America cases).
Party identification, associational membership, clientelism, and

voting: bivariate relationships
Table 5.2 displays simple bivariate relationships between vote choice and
the three mobilisation predictors, namely: party identification, associational
membership, and clientelism across the different regions.
First, beginning with party identification, the value of the Spearman’s rho
measure indicates a significant and strong association with vote choice in
all the regions except in sub-Saharan Africa where the association is weak.
Overall, the relationship is in the expected direction, thus validating the
expectation at least in the bivariate setting that, as in Western European dem-
ocracies, in new democracies those who are party identifiers tend to partici-
pate at a higher rate than non-party members. Second, a look at the effect
of associational membership on electoral participation reveals a significant
association between both variables across three regions notably sub-Saharan
Africa, Latin America, and post-Communist European countries (i.e. except
for East Asian democracies) as in Western European. Yet, the strength of the
association is quite weak in Latin America and sub-Saharan Africa except for
Central and Eastern European states. This result confirms our hypothesis that
members of associational or informal networks tend to participate at higher
rates than non-members. Overall, in the face of this finding, I, therefore, con-
clude that sub-Saharan African, Latin America, and Eastern European dem-
ocracies do resemble older democracies, but the East Asian democracies are
different.
The last of the mobilisation variables is clientelism. Due to data availability,
the effect of clientelism on voting was put to the test only in sub-Saharan
Africa and Latin America democracies. That said, regarding the impact of
clientelism on vote choice across these two regions, astonishingly the results
tend to be diverse. In the Latin America region, I found a modest (i.e. signifi-
cant at the 0.05 level) but weak relationship between clientelism and voting
while for sub-Saharan Africa, the association is non-significant and negative
signed. In the context of this bivariate finding, at least I can confirm Latin
American voters who are targeted for vote-buying are slightly more likely to
vote compared with voters in the African region. Even though these bivariate
Table 5.2 Party identification, associational membership, and clientelism: bivariate
relationship
Regions Mobilisation variables Spearman’s

rho
Western Europe Party identification (1 = party ties) 0.23**

Associational membership (1 = member currently; 0.16**
0.5 = member previously; 0 = not a member)
Clientelism N/A
Sub-Saharan Party identification (1 = party ties) 0.09**
Africa Associational membership (1 = active member/ 0.06**
official leader; 0.5 = inactive member; 0 = not a
member)
Clientelism (1 = receive gifts/bribes) -0.02
Latin America Party identification (1 = party ties) 0.17**
Associational membership 0.11**
Clientelism (1 = receive gifts/bribes) 0.02*
East Asia Party identification (1 = parties ties) 0.20**
Associational membership (1 = member; 0 = not a 0.01
member)
Clientelism N/A
Central and Party identification (1 = party ties) 0.34**
Eastern Associational membership (1 = member currently; 0.16**
Europe 0.5 = member previously; 0 = not a member)
Clientelism N/A
Source: Bivariate results between voting and the mobilisation variables across the regions are derived
from the European Social Survey, Afrobarometer, AmericasBarometer, and the AsianBarometer.
The observations are all weighted using the different surveys weights. Notes: **Correlation is sig-
nificant at the 0.01 level (2-tailed); *Correlation is significant at the 0.05 level (2-tailed).
relationships between the mobilisation predictors and electoral participa-

tion are quite interesting (i.e. most especially regarding party identification),
one cannot interpret them as causal relations. This, therefore, implies these
relationships need to be confirmed in the context of the multivariate environ-
ment, which I will examine in the next section.
General overview of the multivariate results

As reported in Table 5.3, here, multivariate analysis is used to verify the
patterns that emerged in the bivariate analysis. Accordingly, the model for
this second block in our recursive causal system can be expressed as follows:
Vote Choice = f (Socio-Demographics and Mobilisation )
Vote Choice = f ( age, gender, religion, residence, ethnicity, education +

party identification, associational membership,
and clientelism
Table 5.3 Multivariate models of voter turnout in newly consolidated democra-
cies: socio-demographic and mobilisation variables
Regions/variables
Western Europe Model I Model II Model III
Months since last election 0.20*** (0.02) 0.19*** (0.02) 0.15*** (0.02)
Age 0.03*** (0.00) 0.02*** (0.00)
Gender (1 = men) 0.07 (0.04) -0.02 (0.04)
Religion (5 = everyday/once a 0.12*** (0.01) 0.10*** (0.01)
week; 0 = never)
Nominal ethnic identity 0.96*** (0.09) 0.91*** (0.09)
(Largest ethnic group = 1;
Residence (1 = Rural) 0.15*** (0.04) 0.16*** (0.04)
Education 0.56*** (0.03) 0.49*** (0.03)
Party Identification (1 = party 1.13*** (0.05)
ties)
Associational membership 0.65*** (0.06)
(1 = member currently;
0.5 = member previously;
0 = not a member)
Clientelism N/A
Constant -2.96*** (0.38) -5.74*** (0.42) -5.11*** (0.44)
Nagelkerke pseudo R2 0.044 0.146 0.209
% correctly classified 83% 83% 84%
Number of observations/cases 18657/10 18332/10 17962/10
Sub-Saharan Africa
Months since last election 0.01*** (0.00) 0.02*** (0.00) 0.02*** (0.01)
Age 0.38*** (0.04) 0.35*** (0.05)
Gender (1 = men) -0.02 (0.06) -0.11 (0.07)
Religion (5 = more than once 0.08*** (0.03) 0.07** (0.03)
a day; 0 = never)
Salience of ethnicity (1 = only -0.24** (0.10) -0.18 (0.11)
ethnic ID; Equal national/
ethnic ID = 0.5; Only
national ID = 0)
Residence (1 = rural) 0.28*** (0.07) 0.20** (0.08)
Education -0.07 (0.12) 0.01 (0.13)
ties)
(1 = active member/official
leader; 0.5 = inactive
member; 0 = not a member)
Clientelism (1 = receive gifts/ -0.24** (0.09)
bribes)
Constant 1.62*** (0.35) 0.45*** (0.16) -0.43 (0.65)
(continued)
Table 5.3 Cont.
Regions/variables
Latin America
Months since last election -0.05*** (0.01) -0.03*** (0.01) -0.03*** (0.01)
Age 0.05*** (0.00) 0.05*** (0.00)
Gender (1 = men) -0.12** (0.05) -0.11** (0.05)
Religion (Catholics = 1; 0.23*** (0.05) 0.25*** (0.05)
others 0)
Ethnicity (Whites = 2; -0.02 (0.04) 0.01 (0.04)
Mestizo/Mulattos = 1;
Blacks/Indigenous = 0.5;
Others = 0)
Residence (1 = rural) 0.19*** (0.05) 0.15** (0.06)
Education 0.73*** (0.07) 0.68*** (0.07)
ties)
(1 = member; 0 = not a
member)
Clientelism (1 = receive gifts/ 0.18 (0.13)
bribes)
Constant 3.18*** (0.33) -0.04 (0.37) 0.21 (0.38)
East Asia
Months since last election -0.04*** (0.01) -0.09*** (0.01) -0.04*** (0.01)
Age 0.06*** (0.00) 0.06*** (0.00)
Gender -0.07 (0.09) -0.06 (0.09)
Religion (5 = several times a 0.12*** (0.04) 0.11** (0.04)
day; 0 = never)
Nominal Ethnic ID (Largest -0.15 (0.15) -0.08 (0.15)
ethnic group = 1; Minority
ethnic group = 0)
Residence (1 = Rural) 0.25* (0.13) 0.35** (0.14)
Education 0.44*** (0.06) 0.42*** (0.07)
Party identification (1 = party 1.05*** (0.11)
ties)
Associational membership -0.18 (0.10)
(1 = member; 0 = not a
member)
Clientelism N/A
Constant 2.65*** (0.24) -0.65 (0.36) -0.66** (0.39)
Months since last election -0.01 (0.01) 0.00 (0.01) -0.01 (0.01)
Age 0.03*** (0.00) 0.02*** (0.00)
Gender (1 = men) 0.22*** (0.05) 0.09* (0.05)
Table 5.3 Cont.
Regions/variables

Religion (5 = everyday/once a 0.16*** (0.02) 0.15*** (0.02)
week; 0 = never)
Nominal ethnic ID (Largest 0.45*** (0.09) 0.52*** (0.11)
ethnic group = 1; Minority
ethnic group = 0)
Residence (1 = rural) -0.05 (0.05) -0.00 (0.05)
Education 0.60*** (0.03) 0.53*** (0.03)
Party identification (1 = party 1.73*** (0.06)
ties)
0.5 = member previously;
0 = not a member)
Clientelism N/A
Constant 1.15*** (0.14) -1.66*** (0.21) -1.45*** (0.22)
Source: European Social Survey (Western and Eastern European countries); Afrobarometer (Sub-
Saharan Africa); AmericasBarometer (Latin America); AsianBarometer (East Asia). The data
are weighted using the weights provided for each survey data. Country dummies are included for
each region but not shown. For Western Europe, ten country election dummies are included with
the United Kingdom as the reference category; for Africa, 08 countries are included with South
Africa as the reference category; for Latin America, 03 countries are included with Brazil as the
reference case; for East Asia, 09 countries are included with South Korea as the reference case;
for Eastern European countries, 06 country dummies are included with Poland as the reference
case. Significance: ***p < 0.001; **p < 0.01; *p < 0.05. Entries are standardized logistic regres-
sion coefficients, standard errors in parentheses.
Recall that column 3; Model II takes into consideration only the effect of
the socio-demographic variables (age, gender, religion, ethnicity, residence,
and education) on voting across the four regions. That said, this socio-
demographic model is, of course, effectively the same as the one previously
described in Table 4.2 (see, Chapter 4). In column 4, Model III, I proceed by
adding mobilisation variables into the model that already consists of socio-
demographic variables (see, Table C3, Appendix 3, for descriptive results for
the mobilisation variables).
First, I examine the contribution of the three mobilising variables to the
explanatory power of these models, which is quite an improvement across
the different geo-political regions except for sub-Saharan African coun-
tries. In sub-Saharan Africa, the Nagelkerke pseudo R2 moves from 0.068
to 0.098, which is quite an improvement but it is still rather low compared
with other regions. In contrast, for three newer democracies (i.e. as in
Western European democracies) adding the mobilisation variables in Model
III moves the Nagelkerke pseudo R2 values substantially. To 0.213 in Latin
America, 0.175 in East Asia, and 0.273 in post-Communists European states,
with little changes in the values or estimates in the previous model, which
only included the block of socio-demographic predictors. In fact, for some
regions, Model III suggests the mobilisation variables mediate some of the
socio- demographic factors. For instance, for post- Communist European
states, adding the mobilisation variables reduced the impact of gender on
voting, as men are only now modestly (i.e. at the 0.05 level) more likely to
vote. For East Asian democracies, place of residence shows a more sub-
stantial influence on vote choice once I include the mobilisation variables.
Finally, for sub-Saharan Africa, except for age, we observe a decrease in the
effect estimate for the socio-demographic variables, notably religion and
one’s place of residence.
Second, I examine which of the three mobilisation indicators seems to
matter the most to vote choice in new democracies. Beginning with party
identification, in all newer democracies, I observe a statistically significant
effect of party identification on voting.7 This validates our expectation that
those who are party identifiers tend to participate at higher rates than non-
party members. Overall, this result is not only in line with that of Western
democracies (Smets and van Ham, 2013), and they are quite an interesting
finding given the weak roots that parties tend to have new democracies.
Also, a look at the effect of associational membership on electoral partici-
pation across the different geographical regions (i.e. new and older democra-
cies) reveals almost similar findings except for East Asian democracies. In new
democracies of sub-Saharan, Latin America, and post-communist European
countries as in Western European democracies, our data shows that there is
a statistically significant relationship between associational membership and
voting. In these regions, voters who declared themselves to be members of
informal networks and or unions tend to participate at higher rates than non-
members. However, for East Asian cases, this proposition is not supported
as the relationship between associational membership and voting is non-
significant and negatively signed.
Although this finding from the East Asia region seems inconsistent with the
conclusions of the other regions, it might not be too surprising given that there
is little agreement in the comparative literature on the types or dimensions
of social capital that matters for political engagement. For example, the
AsianBarometer asked respondents whether they are members of a group in 20
different areas. Thus, I argue it might be that not all associational memberships
may tend to be important for voting. To substantial, in their country study on
South Korea, Lee and Glasure (2007, pp. 109–111) show that while partici-
pation in some types of associations (i.e. trade unions) leads to higher rates
of political participation. In contrast, for other kinds of associations, such
as environmental groups, their findings reveal a reverse relationship with par-
ticipation as members were less likely to engage than non-members. Overall,
in the face of this finding, I, therefore, conclude that African, Latin America,
and Central/Eastern European democracies do resemble older democracies of
Western Europe, but East Asian democracies are different.
Finally, regarding clientelism, recall, this variable was subjected to empir-
ical testing only in sub-Saharan Africa and Latin America. As reported
in Table 5.3, the result across the two regions are quite revealing. In Latin
America, our data suggest a non- significant (i.e. although being in the
expected direction) relationship between clientelism and voting. This result
for the Latin American countries might not be too surprising given Carlin
et al.’s (2015) and Nadeau et al.’s (2017) studies that include a sample of 18
Latin American countries depict only a modest or limited relationship of the
impact of clientelism on vote choice/presidential vote.
For sub-Saharan Africa cases, the relationship between clientelism and
vote choice is significant and negatively signed, which is directly opposite to
what I have expected. First, for one reason, this result might be very incon-
clusive since our measure of clientelism or vote-buying in sub-Saharan Africa
does not capture precisely whether voters were offered material benefit or not
in exchange for this vote. But rather the question measuring vote-buying only
captures people’s perception of the present of vote-buying in their country
(i.e. in your opinion, how often do the following things occur in this country’s
elections: Voters are bribed). Second, even though we lack an appropriate
question for vote-buying, still this result might not be too surprising since evi-
dence convened by other scholars in the broader sub-Saharan African region
shows mixed results concerning the influence of vote buying and vote choice
(see Lindberg and Morrison, 2008; Young, 2009; Letsa, 2019) For instance,
in their country study of what shapes voting decision in Ghana, Lindberg
and Morrison (2008) concluded the effect of clientelism was relatively unim-
portant in influencing Ghanaians voting decision during the 2000 election.
Similarly, in a recent study of expressive voting in Cameroon, Letsa (2019,
p. 11) shows vote-buying does not necessarily predict voting.
Overall, irrespective of the lack of an appropriate measure for clientelism in
sub-Saharan African, the result in Latin American may as well be very incon-
clusive for the reason that the measures of clientelism refer to outright vote-
buying, which is typically illegal and underreported in the AmericasBarometer
survey. I consider the issue of clientelism in more detail towards the end of
this chapter, trying to assess its effect across all four regions. In sum, having
concluded with the logistic regression results for each of the geo-political
regions, Table 5.4 enables us to gauge how far the mobilisation model suits
newly consolidated democracies. In this regard, I can confirm the transcendent
importance of party identification and association memberships (albeit of a
stronger impact of party identification) in new democracies, with the result
very much mirroring those of older democracies of Western Europe.
Having concluded with the analyses of which mobilisation variables matter
for predicting individual voting propensity in the multivariate context, I now
Table 5.4 Theoretical matching of mobilisation model on vote choice: based on the
multivariate results
Models/ Expected Western Sub-Saharan Latin East Post-

variables effect/ Europe Africa America Asia communist
direction Europe
Mobilisation model
Party + Applies*** Applies*** Applies*** Applies*** Applies***
identification
Associational + Applies*** Applies*** Applies*** Does not Applies***
membership apply
Clientelism + N/A Applies* Does not N/A N/A
apply

AsianBarometer and European Social Survey. Applies***: This signifies that the relationship
assess the extent of cross-national variation in the effect of the mobilisation

model across individual countries in the four geopolitical regions, while con-
trolling for the influence of socio-demographic variables (i.e. see Appendix
3, Table C4, for the relevant results). First, starting with party identification,
individual country comparison strongly confirmed as expected that voters
who declared being party identifiers are more likely to vote compared to non-
party identifiers. In 22 countries of the 26 newly consolidated democracies
(e.g. six in sub-Saharan Africa, seven in Latin America, three in East Asia,
and six in Eastern Europe) party identification displays a statistically signifi-
cant relationship with vote choice. Thus, overall the results strongly confirm
party identification constitutes a rather important mobilisation variable for
voting behaviour in newly consolidated democracies.
Second, moving to associational networks, remember, the theoretical
expectation is that individuals who are involved in informal associations
would be more likely to participate. The individual country findings suggest
16 out of 26 newly consolidated democracies (four out of eight in Africa, nine
out of nine in Latin America, three out of six in Eastern Europe and none in
East Asia) shows a positive and significant relationship between associational
membership and voting.
Finally, I round-up with results associated with the last mobilisation vari-
able, which is clientelism. I have hypothesised that citizens that are targeted
by material benefit will be more likely to vote. However, recall, for the Latin
America region, the pooled results of the impact of clientelism on vote choice
are revealed to be non-significant, although being in the expected direction.
For sub-Saharan Africa, the result did indicate a negative but significant
relationship with electoral participation. The individual country results
across both regions are even more nuanced, yet a couple of things stand out.
(1) The expectation of the impact of clientelism on vote choice is met, in
statistical significance terms, in only two countries across both regions (e.g.
Mauritius and Peru). (2) In five of the nine countries in Latin America (i.e.
Panama, Uruguay, Costa Rica, Argentina, and Brazil) the effect of clien-
telism on vote choice is non-significant, albeit in the expected direction. (3) In
three sub-Saharan Africa countries (i.e. Benin, Ghana, and South Africa) the
relationship between clientelism and voting shows a significant but negative
(unexpected) relationship. (4) Across seven countries in both regions (i.e. 4
in Africa and 3 in Latin America), we observe a non-significant and negative
relationship between clientelism and vote choice.
Re-considering the impact of clientelism and voting in new

democracies
So far, in this chapter evaluating the effect of clientelism on vote choice in
newly consolidated democracies has been hampered by the absence of data
except for sub-Saharan Africa and Latin America. Yet, even testing the effect
of clientelism in sub-Saharan African and Latin America, the results appeared
to be contrary to what I have expected (see Table 5.3). I argue this result might
not be too surprising for two main reasons: first, in Africa and Latin America,
patrons, political parties, and politicians do not operate as political machines
as such vote-buying tends to be unmonitored (Kramon, 2017; Vicente, 2014;
Letsa, 2019; Guardado and Wantchekan, 2013; Lindberg, 2010; Bratton,
2008). The implication of this is that a voter might collect gifts or money and
then refrain from voting altogether.
Second, the measure of clientelism, particularly in Latin America, refers to
outright vote-buying, which is typically illegal and underreported in surveys.8
So, a better measure of clientelism would be one that also refers to implicit
exchanges between voters and politicians, which are not illegal. Thus, with
the need to evaluate the effect of clientelism on vote choice across all four
geo-political regions, I, therefore, rely upon a measure of clientelism from the
Varieties of Democracies data (V-Dem). V-Dem captures three dimensions
of clientelism: (1) Vote buying in elections,9 (2) Party linkages,10 and
(3) Particularistic or public goods.11 To be precise, particularistic spending
is narrowly targeted on a specific corporation, sector, social group, region,
party, or set of constituents.12 However, for the sake of simplicity and because
the scores across the three dimensions of clientelism are highly correlated,
V-Dem creates an additive index of clientelism index that measures to what
extent politics are based on clientelistic relationships. Clientelistic relations
now include the targeted, contingent distribution of resources (goods, ser-
vices, jobs, money, etc.) in exchange for political support.13 For our concern,
a particularly interesting question is whether the V-Dem measure of clien-
telism has a positive effect on vote choice in newer democracies. To examine
this, I build this argument in two ways: first, I explore the aggregate correlates
of clientelism across the different geo-political regions. Second, I then re-
evaluate the effect of clientelism on electoral participation by estimating
a multivariate model that takes into consideration the socio-demographic
and mobilisation variables (i.e. that now contain the V-Dem measure of
clientelism).
The scope of clientelism in new democracies

As a starting point, relying on a measure of clientelism from the V-Dem
dataset from 2011 through 2016,14 Table 5.5 summarises the average distribu-
tion of responses to the question: to what extent are politics based on clientel-
istic relationships? It is worth mentioning that the clientelism index is based
on an interval scale from low to high; so that a higher score equals more
clientelism.
One obvious conclusion from Table 5.5 is that clientelism tends to be
higher in new democracies, while regarding older established democracies of
Western Europe, clientelism is least present. However, if we focus only on
newly consolidated democracies, it is clear clientelism efforts are more evident
in sub-Saharan Africa, Latin America, followed by East Asia. Concerning
post-communist European countries, it tends to be quite low. A second con-
clusion is that there exist quite enough within-country variations with clien-
telism across the different regions. For instance, in sub-Saharan African, it
ranges between 0.23 in Botswana (lowest in the region) and 0.52 in Benin
(highest). Concerning Latin America, it ranges between 0.05 in Uruguay
(lowest in the region), which is quite comparable to those of older democ-
racies of Western Europe, and a high of 0.50 in Argentina and 0.55 in Peru.
Moving to East Asia, the lowest scores are South Korea (0.18) and the highest
is Mongolia (0.36). For post-communist regions, the country with the lowest
scores is Estonia (0.06), while Hungary has a higher score of 0.29. Rounding
up with the Western European countries, the scores of clientelism are very low
and comparable across all ten countries except for Ireland and Italy registering
a score of 0.13 and 0.25, respectively. In sum, these initial descriptive results
suggest that, in terms of a global comparison, the practice of clientelism is
quite present and common in newer democracies compare to older establish
democracies of Western Europe that stand out, generally showing low rates
of clientelistic efforts.
The correlates of clientelism and vote choice

Beyond assessing the scope of the phenomenon of clientelism, it is worth
investigating whether in countries characterised with a high rate of clien-
telistic relationships citizens would be more likely to vote or not. Here, my
focus is on the effect of clientelism and vote choice in newly consolidated
Table 5.5 Scope of clientelism practices in new democracies
Countries/regions V-Dem clientelism index
Mean
Sub-Saharan Africa
Benin 0.52
Botswana 0.23
Cape Verde 0.26
Ghana 0.50
Mauritius 0.33
Namibia 0.31
Sao Tome and Principe 0.45
South Africa 0.33
Latin America
Argentina 0.50
Brazil 0.35
Chile 0.10
Costa Rica 0.15
El-Salvador 0.46
Guyana 0.48
Panama 0.46
Peru 0.55
Uruguay 0.05
East Asia
Taiwan 0.29
Mongolia 0.36
South Korea 0.18
Czech Republic 0.14
Estonia 0.06
Hungary 0.29
Lithuania 0.16
Poland 0.12
Slovenia 0.09
Western Europe
Austria 0.07
Belgium 0.04
Germany 0.04
Finland 0.06
United Kingdom 0.06
Ireland 0.13
Italy 0.25
Netherlands 0.03
Norway 0.03
Sweden 0.03
Regions
Sub-Saharan Africa 0.37
(continued)
Table 5.5 Cont.
Countries/regions V-Dem clientelism index
Mean
Latin America 0.34
East Asia 0.28
Post-Communist Europe 0.15
Western Europe 0.07
Source: V-Dem data (2019), index of clientelism are based on 2011–2016. Index based on
three dimensions: election vote buying (C) (v2elvotbuy), particularistic or public goods
(C) (v2dlencmps) and party linkages (C) (v2psprlnks).
Table 5.6 The effect of clientelism on vote choice in new democracies
Model I Model II Model III
Sub-Saharan Africa
Clientelism index 3.49*** (0.26) 5.90*** (1.29) 7.73*** (1.41)
Country dummies Yes Yes Yes
Nagelkerke’s R2 0.029 0.068 0.104
Latin America
Nagelkerke’s R2 0.026 0.172 0.215
East Asia
Country Dummies Yes Yes Yes
Nagelkerke’s R2 0.009 0.136 0.175
Clientelism index 4.26 (0.39) -2.01 (0.79) 7.66 (5.08)
Nagelkerke’s R2 0.036 0.146 0.273
Notes: These results are based on a measure of clientelism that is derived from V-Dem data
(2019), the index of clientelism is based on 2011–2016. Significance: ***p < 0.001; **p < 0.01;
*p < 0.05. Entries are standardized logistic regression coefficients, standard errors in parentheses.
Model I has no Controls; Model II controls for socio-demographic variables; Model III controls
for socio-demographic and two mobilisation variables (i.e. party ID and associational networks).
The data are weighted using the weights provided for each survey data.
democracies.15 As reported in Table 5.6, I investigate the effect of clien-

telism on vote choice in three steps. First, I estimate a model (Model I) where
I regressed vote choice against the clientelism variable (i.e. without controlling
for the socio-demographic or the other mobilisation variables). Second, in the
next model (i.e. Model II) I examine how the result between clientelism and
vote choice change once controls from the socio-demographic variables are
included. Third, in the final model (Model III) I include additional control
variables that are drawn from the mobilisation variables (i.e. party ID and
associational membership).
As reported in Table 5.6, there is a statistically significant relationship
between clientelism and vote choice in three newer democracies except for
post-communist European democracies. This result depicts an increase in
the aggregate level of clientelistic relationships significantly increases the
individual-level voting decision across sub-Saharan Africa, Latin America,
and East Asia. However, a close look at the result across the regions where
clientelism has a significant effect reveals the impact of clientelism tends
to be more substantial in sub-Saharan Africa when compared to Latin
America and East Asia democracies. However, using the V-Dem aggregate
level measure of clientelism highlights the problem of ecological fallacies
(i.e. a better measure will have to capture clientelism at the individual-level,
but, of course, this depends on the availability of data). Overall, the result
suggests Africans and Latin Americans and East Asians are more likely
to vote because of clientelism networks. Still, more importantly, the result
attests to the fact that the mere existence of the phenomenon of clientelism
does not diminish the value of the socio-demographic or other mobilisation
variables.
Conclusion
In this chapter, I have discussed and tested the effect of three mobilisa-
tion variables on electoral participation: party identification, associational
networks, and clientelism. Having concluded with the multivariate ana-
lyses across the four regions, the results obtained are not only clear but
very interesting in several ways. The standout result is that party identifi-
cation plays a more dominant role in explaining voting behaviour in newly
consolidated democracies in sub-Saharan Africa, Latin America, East Asia,
and post-communist Europe as the pooled results revealed. Similarly, the
examination of the impact of party identification across individual-countries
re-attests the overriding effect of party identification as its relationship with
voting proves to be statistically significant and in the expected direction in
22 of the 26 countries in all four regions, which is a little surprising given
their limited experience with democracy, and the instability of their party
systems. Second, another mobilisation variable that seems to matter for vote
choice is associational membership, although not so dominant as the role
of party identification. To be precise, not only does the multivariate analysis
suggest the importance of associational membership on vote choice, but the
country by country analysis also reveals this variable tends to be essential
for electoral participation in 16 out of 26 newly consolidated democracies.
Finally, our examination of clientelism using a measure from Varieties of
Democracies (V-Dem data) indicates that even if the practice is relatively fre-
quent in emerging democracies and is generally used by politicians and pol-
itical entrepreneurs seeking to attract voters, the presence of clientelism does
not seem to diminish the value of other predictors of vote choice, notably the
socio-demographic and mobilisation variables.
Notes
1 However, the act of canvassing can be equally considered as a vital mobilisation
factor. Canvassing is seen as an act of requesting or inspiring people to partici-
pate, and this can definitely have a significant effect on people’s voting decisions.
According to scholars such as Gerber and Green (2000a, 2000b), the act of
canvassing and knocking on voters’ doors and reminding them to vote seems to
be a most effective tool in increasing citizens’ propensity to turn out in the polls.
Similarly, Johnston, Cutts, Pattie, and Fisher (2012), focusing on the British gen-
eral election in 2010, provide evidence that people who are contacted by a party
are more likely to vote than those who are not based on these studies, although
canvassing seems an important mobilisation factor. Yet, because of the absence of
data availability, I am unable to examine its impact on vote choice across newer
democracies.
2 In new democracies, it is not uncommon for politicians and or political
entrepreneurs to attempt to secure the popular vote or support of the voters by
offering gifts, money, or material benefits.
3 Countries surveyed include: Austria, Belgium, Switzerland, Germany, Denmark,
Spain, Finland, France, United Kingdom, Greece, Ireland, Italy, Luxembourg,
Netherlands, Norway, Portugal, and Sweden.
4 The countries included in the survey were: Botswana, Ghana, Lesotho, Malawi,
Mali, Namibia, Nigeria, South Africa, Tanzania, Uganda, Zambia, and
Zimbabwe.
5 The countries include: Algeria, Benin, Botswana, Burkina Faso, Burundi,
Cameroon, Cape Verde, Cote d’Ivoire, Egypt, Gabon, Ghana, Guinea, Kenya,
Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritius, Morocco, Mozambique,
Namibia, Niger, Nigeria, Sao Tome and Principe, Senegal, Sierra Leone, South
Africa, Sudan, Swaziland, Tanzania, Togo, Tunisia Uganda, Zambia, and
Zimbabwe.
6 The Latin America countries include: Bolivia, Mexico, Guatemala, El Salvador,
Honduras, Nicaragua, Costa Rica, Panama, Colombia, Peru, Paraguay, Chile,
Uruguay, Brazil, Argentina, and Dominican Republic.
7 For similar results see Carreras and Castañeda-Angarita (2014); Bratton, Chu,
and Lagos, (2010); Kuenzi and Lambright (2011).
8 For instance, in Latin America, our sample suggest only 4 per cent of the
respondents declared they had received bribes/gifts for their vote.
9 The question measuring vote buying reads as follows: in this national election, was
there evidence of vote and/or turnout buying?
10 A party-constituent linkage refers to the sort of good that the party offers in
exchange for political support and participation in party activities.
11 Considering the profile of social and infrastructural spending in the national
budget, how “particularistic” or “public goods” are most expenditures.
12 Such spending may be referred to as pork, clientelistic, or private goods.
13 Note the index is formed across all three dimensions by taking the reversed point
estimates, so that higher scores equal more clientelism.
14 This time period was chosen since it coincided very closely to the time the elections
and survey data were run across the different regions.
15 I did not conduct any analysis for Western European democracies as the rate of
clientelism is low, with very little variation across the countries.
6
and the vote
In their ground-breaking book, The American Voter Campbell, Converse,

Miller and Stokes (1960) posit one of the crucial ways for understanding
voting decisions is that it is based on voters’ psychological engagement and
or their orientation towards politics. Relatedly, a more recent work by Lewis-
Beck and colleagues that replicates the 1960 volume of Campbell et al.,
reconfirms voters’ behaviour tends to be guided by their psychological involve-
ment (Lewis-Beck et al., 2008, p. 92). Drawing from these landmark studies,
it, therefore, makes sense to expect that the stronger a person’s psychological
involvement in politics, the higher his/her propensity to turn out on election
day. This chapter, therefore, focuses on three standard political-psychological
variables notably: political interest, political efficacy, and political trust to
explain people’s decision of whether to vote or not in new democracies.
First, among the chief political-psychological predictors of turnout is pol-
itical interest. Blais and Daoust (2020, p. 7) content political interest is an
essential driver of vote choice because it tends to remains remarkably stable
over time. Consequently, it makes sense to expect that those who are interested
in politics are also more likely to vote, and those who have no interest are more
inclined to abstain. Second, another vital political-psychological indicator is
political efficacy,1 which is considered as “the feeling that individuals political
actions do have or can have, an impact upon the political process” (Campbell,
Gurin, and Miller, 1954, p. 187). Carlin et al., (2015, p. 39) cite scholars have
longed unpacked the role of political efficacy into internal and external effi-
cacy2 with both indicators expected to boost people’s voting decision. Yet, a
survey of the comparative literature reveals most research has tended to pri-
oritise external over internal efficacy (see e.g. Norris, 2004; Kittilson, Caul,
and Anderson, 2011). I argue both concepts are quite distinguishable, not
interchangeable; therefore, necessitating the need to evaluate their impact on
vote choice separately. Finally, the last psychological variable is political trust.
However, recent studies by van der Meer (2017) do suggest a decline of pol-
itical trust across established democracies, except for the Nordic countries
and Germany. Still, empirical evidence continues to show people who have
more confidence in the political system/institutions are more likely to engage
(Bélanger and Nadeau, 2005). In sum, the general hypotheses described for
Political-psychological variables 117
each of the three political-psychological variables lead to a set of expectations
with which the empirical results can be confronted in newer democracies. In
this chapter, I will examine the effect of the political-psychological variables
on vote choice in two ways. First drawing on the World Values Survey data,
I present some descriptive information on people’s political-psychological
orientation across the different regions. Second, using bivariate and multi-
variate analysis, I will examine the impact of political interest, political trust,
and political efficacy on the explanatory power of our vote choice model.
Underlying political-psychological orientations in new democracies

Figures 6.1 to 6.3 provide an overview of people’s political-psychological
orientation (i.e. here defined as political interest, political trust, and social/
interpersonal trust) across the different regions, based on the World Values
Survey (WVS) of 2010–2014.3 Although the countries covered are not in any
way a total representation of the different regions, it paints a picture of citi-
zens underlying psychological orientations.
First, looking at political interest and as revealed by Figure 6.1, if we com-
bine the categories very interested and somewhat interested, it is clear political
interest is highest in Western European democracies (50%) and lower for new
45
40
35
30
25
20
15
10
5
0
Western Europe Post-communist Sub-saharan Africa La n America East Asia
Europe
Very interested Somewhat interested Not very interested Not at all interested
Figure 6.1 Interest in politics

Source: World Values survey (Wave 6, 2010–2014); Western Europe samples: Germany
2013, Netherlands 2012, Spain 2011, Sweden 2011; post- communist Europe
samples: Armenia 2011, Belarus 2011, Estonia 2011, Romania 2012, Russia 2011,
Slovenia 2011, Ukraine 2011 and Poland 2012; sub-Saharan Africa samples: Algeria
2014, Egypt 2013, Ghana 2012, Libya 2014, Morocco 2011, Nigeria 2012, Rwanda 2012,
South Africa 2013, Tunisia 2013, Zimbabwe 2012; Latin America samples: Argentina
2013, Brazil 2014, Colombia 2012, Chile 2012, Ecuador 2013, Mexico 2012, Peru
2012, Uruguay 2011; East Asia samples: China 2013, Hong Kong 2014, Japan 2010,
Malaysia 2012, Philippines 2012, Singapore 2012, Taiwan 2012, Thailand 2013.
118 Political-psychological variables
democracies except for sub-Saharan Africa and East Asia democracies. In
the two regions, 51 and 50 per cent of sub-Saharan Africans and East Asians
declared themselves to be interested in politics, which is equal/comparable to
those of older democracies of Western Europe. In contrast, the percentage of
those who were revealed to be politically interested is comparatively lower in
Latin America and post-communist European states ranging from 30 to 37
per cent respectively.
Next, moving to political trust, defined here as confidence in government4
and as depicted by Figure 6.2, remarkably, political trust seems to be higher in
the East Asian region compared to Western European democracies. However,
it remains unclear what this high trust or confidence in government means for
the entire region. To be precise, in East Asia, the high rate in political trust
tends to be influenced by non-democracies and authoritarian regimes such as
China, Singapore, and or Thailand. In contrast, political confidence for the
region’s liberal systems tends to be low (i.e. also see Park, 2016). Furthermore,
political trust seems to be slightly higher for the sub-Saharan African region
and Latin America region (i.e. 42 and 40 per cent declared confidence in the
50
45
40
35
30
25
20
15
10
5
0
Western Europe Post-communist Sub-saharan Africa Lan America East Asia
Europe
Trust a great deal Trust quite a lot Not very much None at all
Figure 6.2 Trust in government

South-Africa 2013, Tunisia 2013, Zimbabwe 2012; Latin America samples: Argentina
government). Overall, trust in the government was revealed to be quite low
for post-communist European countries which is not too surprising as Welzel
and Dalton (2015) posit post-communist countries tend to be characterised
by a political culture of low trust.
Lastly, rounding up with social trust, overall, respondents in Western
European democracies tend to be more trustful of others when compared to
those in newer democracies. These figures are astonishing if we are to expect
that the level of trust or confidence in government and or political institutions
should reflect in part the general trusting nature of individuals (i.e. see Zmerli
and Newton, 2016). Thus, this would imply high political trust, most notably
in East Asia and sub-Saharan Africa, should be connected with more people
trusting strangers or their neighbours. But this is not the case as Figure 6.3
suggests interpersonal trust to be very low in these regions. So far, I have
taken steps to paint a picture of people’s psychological orientations across
the different geo-political regions. Yet, it remains unclear what they mean in
terms of people’s decision of whether or not to vote. This will be the focus of
the next section.
East Asia
Lan America
Sub-saharan Africa
Post-Communist Europe
Western Europe
0 10 20 30 40 50 60 70 80 90 100
Need to be very careful Most people can be trusted
Figure 6.3 Trust in people

South Africa 2013, Tunisia 2013, Zimbabwe 2012; Latin America samples: Argentina
Political-psychological variables and the vote: bivariate relationship

Here, I begin by examining the bivariate relationship between our four
political-psychological variables and vote choice across the different regions,
with the results presented in Table 6.1. It should be remembered that Table 6.1
shows the measures of association using Spearman’s rho, which tells us about
the strength and direction of the relationship between political interest, polit-
ical trust, external and internal efficacy, and vote choice.
First, beginning with political interest, I find a consistent pattern across
all four newly consolidated democracies (as in Western European democra-
cies), which supports the proposition that those who are politically interested
have a greater propensity to turn out in elections. Second, moving to polit-
ical trust, the result shows in newer democracies as in older democracies there
is a positive and significant relationship between political trust and voting.
Table 6.1 Political-psychological variables and electoral participation in new democ-

racies: bivariate associations
Regions Political-psychological variables Spearman’s rho
Western Europe Political interest 0.27**

Political trust 0.14**
External efficacy 0.15**
Internal efficacy 0.14**
Sub-Saharan Africa Political interest 0.08**

External efficacy 0.01
Latin America Political interest 0.15**

External efficacy 0.02*
East Asia Political interest 0.14**

External efficacy 0.02
Internal efficacy 0.01
Central and Eastern Europe
Central and Eastern Europe Political interest 0.36**
External efficacy 0.17**
Source: Bivariate results between voting and the political-psychological variables across the
regions are derived from the European Social Survey, Afrobarometer, AmericasBarometer, and
the AsianBarometer. The observations are all weighted using the different surveys weights. Notes:
(2-tailed).
However, the strength of the relationship is revealed to be quite weak in new
democracies of sub-Saharan Africa, Latin America and East Asia except for
post-communist European countries (as in Western European democracies).
Overall, given the significant and positive relationship across all regions, at
least in the bivariate context, I can tentatively conclude that new democra-
cies resemble Western European democracies. Moving to the last political-
psychological variables, external and internal efficacy, the proposition about
the effect of external and internal efficacy on voting produced varied findings.
The data from the two regions, Latin America and post-Communist states,
indicate that as in Western European democracies, the relationship between
electoral participation and external efficacy is positive and significant. In con-
trast, for sub-Saharan African and East Asian democracies, the relationship
between voting and external efficacy is non-significant. Rounding up with
internal efficacy, we observe a significant and positive relationship with vote
choice only in Latin America, sub-Saharan African, and post-communist
countries (i.e. as in Western European democracies) with the exception being
East Asia.
While these bivariate results tell us of the direction of the relationship
between political interest, political trust, external efficacy, and internal effi-
cacy and vote choice, to be more confident of these results, and to explore the
sources of variation more rigorously, one needs to engage with multivariate
analysis, which is the task of the next section.
Political-psychological variables and the vote: overview of the

multivariate results
In the previous section, the impact of the various political-psychological
variables on vote choice has been examined. Although these bivariate
explorations provide prima facie evidence that most of the political-
psychological variables (i.e. political interest, political trust, and political
efficacies) lead to increased levels of voter turnout, these relationships need
to be tested in a multivariate regression model for a better understanding of
vote choice in a cross-national/regional context. Thus, the model for this third
block of variables reads as follows:
Vote Choice = f (Socio-Demographics, Mobilisation

and Political -Psychological ).
Table 6.2 reports the logistic regression analysis for each of the geo-political
regions (i.e. Appendix 3, Table C5, provides summary statistics for the political-
psychological variables). Note, the model includes socio- demographic
and mobilisation variables to which I now add the political-psychological
indicators.
The first question to examine is the extent to which political-
psychological variables contribute to an overall explanation of people’s
newgenrtpdf
Table 6.2 Multivariate models of voter turnout in newly consolidated democracies: socio-demographic, mobilisation and political-psychology
122
variables
Regions/variables
Western Europe Model I Model II Model III Model IV
Months since last election 0.20*** (0.02) 0.19*** (0.02) 0.15*** (0.02) 0.14*** (0.02)
Age 0.03*** (0.00) 0.02*** (0.00) 0.02*** (0.00)
Gender (1 = men) 0.07 (0.04) -0.02 (0.04) -0.17*** (0.05)
Religion (2 = everyday/once a week; 0 = never) 0.12*** (0.01) 0.10*** (0.01) 0.08*** (0.01)
Nominal ethnic identity (Largest ethnic group = 1; Minority ethnic 0.96*** (0.09) 0.91*** (0.09) 0.89*** (0.10)
group = 0)
Residence (1 = rural) 0.15*** (0.04) 0.16*** (0.04) 0.20*** (0.05)
Education 0.56*** (0.03) 0.49*** (0.03) 0.31*** (0.03)
Party Identification (1 = party ties) 1.13*** (0.05) 0.85*** (0.05)
Associational membership (1 = member currently; 0.5 = member 0.65*** (0.06) 0.64*** (0.06)
previously; 0 = not a member)
Clientelism N/A N/A
Political interest 0.88*** (0.05)
Political trust 0.19*** (0.03)
Internal efficacy 0.09** (0.04)
External efficacy 0.11* (0.05)
Constant -2.96*** (0.38) -5.74*** (0.42) -5.11*** (0.44) -5.38*** (0.46)
Nagelkerke’s R2 0.044 0.146 0.209 0.244
% correctly classified 83% 83% 84% 85%
Number of observations/cases 18657/10 18332/10 17962/10 17318/10
Sub-Saharan Africa
Months since last election 0.01*** (0.00) 0.02*** (0.00) 0.02*** (0.01) 0.02*** (0.03)
Age 0.38*** (0.04) 0.35*** (0.05) 0.29*** (0.05)
Gender (1 = men) -0.02 (0.06) -0.11 (0.07) -0.16* (0.08)
Religion (2 = more than once a day; 0 = never) 0.08*** (0.03) 0.07** (0.03) 0.08*** (0.03)
Salience of ethnicity (1 = only ethnic ID; Equal national/ethnic -0.24** (0.10) -0.18 (0.11) -0.14 (0.12)
ID = 0.5; Only national ID = 0)
Residence (1 = rural) 0.28*** (0.07) 0.20** (0.08) 0.18** (0.08)
Education -0.07 (0.12) 0.01 (0.13) -0.09 (0.14)
Party identification (1 = party ties) 0.76*** (0.08) 0.61*** (0.09)
Associational membership (1 = active member/official leader; 0.56*** (0.11) 0.53*** (0.11)
0.5 = inactive member; 0 = not a member)
Clientelism (1 = receive gifts/bribes) -0.24** (0.09) -0.21* (0.11)
Political trust 0.08 (0.05)
Internal efficacy 0.06 (0.09)
External efficacy 0.05 (0.08)
Constant 1.62*** (0.35) 0.45*** (0.16) -0.43 (0.65) -0.17 (0.75)
Nagelkerke’s R2 0.029 0.068 0.098 0.102
Latin America
Months since last election -0.05*** (0.01) -0.03*** (0.01) -0.03*** (0.01) -0.03*** (0.01)
Age 0.05*** (0.00) 0.05*** (0.00) 0.05*** (0.00)
Gender (1 = men) -0.12** (0.05) -0.11** (0.05) -0.13*** (0.05)

Religion (Catholics = 1; others 0) 0.23*** (0.05) 0.25*** (0.05) 0.22*** (0.05)
Ethnicity/race (Whites = 2; Mestizo/Mulattos = 1; Blacks/ -0.02 (0.04) 0.01 (0.05) 0.05 (0.04)
Indigenous = 0.5; others = 0)
Residence (1 = rural) 0.19*** (0.05) 0.15** (0.06) 0.17** (0.07)
Education 0.73*** (0.07) 0.68*** (0.07) 0.59*** (0.07)
Associational membership (1 = member; 0 = not a member) 0.22*** (0.03) 0.18*** (0.03)
Clientelism (1 = receive gifts/bribes) 0.18 (0.14) 0.11 (0.13)
Political trust -0.05 (0.03)
Internal efficacy 0.02 (0.02)
External efficacy -0.01 (0.02)
Constant 3.18*** (0.33) -0.04 (0.37) 0.21 (0.38) -0.03*** (0.40)
Nagelkerke’s R2 0.026 0.172 0.213 0.224
(continued)
newgenrtpdf
Table 6.2 Cont.
124
Regions/variables
Western Europe Model I Model II Model III Model IV
East Asia
Months since last election -0.04*** (0.01) -0.09*** (0.01) -0.04*** (0.01) -0.04** (0.01)
Age 0.06*** (0.00) 0.06*** (0.00) 0.05*** (0.01)
Gender (1 = men) -0.07 (0.09) -0.06 (0.09) -0.08 (0.11)
Religion (2 = several times a day; 0 = never) 0.12*** (0.04) 0.11** (0.04) 0.11** (0.05)
Nominal Ethnic ID (Largest ethnic group = 1; Minority ethnic -0.15 (0.15) -0.08 (0.15) 0.02 (0.16)
group = 0)
Residence (1 = rural) 0.25* (0.13) 0.35** (0.14) 0.30** (0.15)
Education 0.44*** (0.06) 0.42*** (0.07) 0.31*** (0.07)
Associational membership (1 = member; 0 = not a member) -0.18 (0.24) -0.25** (0.11)
Clientelism N/A N/A
Political trust 0.04 (0.07)
Internal efficacy -0.05 (0.10)
External efficacy 0.01 (0.16)
Constant 2.65*** (0.24) -0.65 (0.36) -0.66** (0.39) -0.96** (0.43)
Nagelkerke’s R2 0.009 0.136 0.175 0.189
Months since last election -0.01 (0.00) 0.00 (0.01) -0.01 (0.01) -0.01 (0.01)
Age 0.03*** (0.00) 0.02*** (0.00) 0.02*** (0.00)
Gender (1 = men) 0.22*** (0.05) 0.09* (0.05) -0.08 (0.05)
Religion (1 = everyday/once a week; 0 = never) 0.16*** (0.02) 0.15*** (0.02) 0.13*** (0.02)
Nominal ethnic ID (Largest ethnic group = 1; Minority ethnic 0.45*** (0.09) 0.52*** (0.11) 0.44*** (0.12)
group = 0)
Residence (1 = rural) -0.05 (0.05) -0.00 (0.05) 0.03 (0.05)
Education 0.60*** (0.03) 0.53*** (0.06) 0.39*** (0.04)
Party Identification (1 = party ties) 1.73*** (0.06) 1.34****(0.06)
Associational membership (1 = member currently; 0.5 = member 0.54*** (0.09) 0.42*** (0.10)
previously; 0 = not a member)
Clientelism N/A N/A
Political trust 0.24*** (0.03)
Internal efficacy 0.20*** (0.06)
External efficacy 0.22** (0.07)
Constant 1.15*** (0.14) -1.66*** (0.21) -1.45*** (0.22) -1.63*** (0.25)
Nagelkerke’s R2 0.036 0.146 0.273 0.334

Source: European Social Survey (Western and Eastern European countries); Afrobarometer (Sub-Saharan Africa); AmericasBarometer (Latin America);
AsianBarometer (East Asia). The data are weighted using the weights provided for each survey data. Country dummies are included for each region but not
shown. For Western Europe, 10 country election dummies were included with the United Kingdom as the reference category; for Africa 08 countries were
included with South Africa as the reference category; for Latin America, 08 included with Brazil as the reference case; for East Asia, 03 countries were included
with South Korea as the reference category; for Eastern European countries, 06 country dummies are included with Poland as the reference case. Significance:
***p < 0.001; **p < 0.01; *p < 0.05. Entries are standardised logistic regression coefficients, standard errors in parentheses.
decision of whether or not to vote across the four regions. In other words,
what is the fit of the political-psychological model as a whole or collectively
to the survey data across the different regions? To answer this question,
I turn to the Nagelkerke pseudo R2, a measure showing the proportion
of variance in vote choice that can be assigned to each block of explana-
tory variables. As demonstrated by Nagelkerke pseudo R2, the proportion
of explained variance indicates a slight improvement in most instances,
although it equally tends to vary from one region to another. Relatedly,
given that the Nagelkerke pseudo R2 is also a pointer of whether the
political-psychological variable matters in some region and not in others,
we calculate the difference/change in the Nagelkerke pseudo R2 between the
political-psychological variables (i.e. Model IV) and that from the mobil-
isation variables (i.e. Model III). When this was done, we can confirm the
most substantial difference is found in new democracies of post-communist
European democracies where the addition of the political-psychological
variables increases the Nagelkerke pseudo R2 by 0.06.5 On the other hand,
the political-psychological variables do not explain much of the variation
in vote choice in new democracies of sub-Saharan Africa, Latin America,
and East Asian, where the difference in the Nagelkerke pseudo R2 only
moves by a small fraction/range.
But most importantly, what is clear is that some of the socio-
demographic variables tend to be mediated by the political-psychological
variables I include in Model IV. For example, regarding the influence of
gender on vote choice, by adding the political-psychological variables,
I confirm a couple of things. First, in post-communist European coun-
tries, the relationship between gender and vote choice that was signifi-
cant and positive through Model II and Model III indicating men were
more likely to participate is now non-significant with a negative sign.
Second, for Western European democracies and to a lesser extend sub-
Saharan African countries, the impact of gender on the vote that was non-
significant through Model II and Model II is now significant and negative.
Indicating women are far more likely to vote than men. Third, for Latin
America, we equally see an increase in the effect estimate on gender once
the political-psychological variables are included. Overall, these results are
quite impressive as it implies it is with regards the political-psychological
variables (i.e. interest, political trust, or political efficacy) where women
tend to lag behind men in participation. Consequently, if we keep the level
of the political-psychological variables constant for both genders, we see
women are more likely to vote than men.
Next, the second question to examine is which of the political-psychological
indicators seem to matter for electoral participation in newly consolidated
democracies. First, the impact of political interest on turnout is statistically
significant and positive in all four newer democracies. This result, therefore,
collaborates those of Blais and Daoust (2020); Aarts and Wessels, (2005)
who report similar findings on the impact of political interest on turnout in
established democracies.
Second, the indicator of political trust shows mixed results. Political trust
has a positive effect on electoral participation only in one newly consolidated
democracy of post-communist Europe, which is similar to those of older
democracies of Western Europe. To be precise, the odds of voting to non-
voting in Central/Eastern European countries increase by a factor of 1.275
for each unit of trust, keeping all the factors constant, which means that for
each unit increase in trust, the likelihood of voting increases by 28 per cent.
In sub-Saharan Africa, Latin America, and East Asia, the impact of political
trust on voting is non-significant.
Rounding up with internal and external efficacy, once again just as political
trust, both indicators of political efficacy show varied results. The positive
and significant effect of internal and external efficacy on electoral participa-
tion is observed only for Central/Eastern European countries (as in Western
Europe), which does reconfirm earlier results by Bernhagen and Marsh (2007).
Reflecting on the main results of this section for each of the regions enables
us to establish a theoretical matching (i.e. Table 6.3). Table 6.3 suggests the
impact of political interest on vote choice is clearly present across all four
newer democracies, but the effect of the remaining political-psychological
variables (political trust and political efficacies) is less pronounced. Thus,
overall, based on Table 6.3, as previously alluded to, we can tentatively con-
clude the political-psychological model/variables are very much important
for vote choice in post- communist European democracies as in Western
European democracies. But for other newer democracies, except for political
interest, their importance seems to be less relevant.
Having examined the overall performance of the vote choice model when
the third block of variables (political-psychological) are added, Table C6
(see, Appendix 3) shows how the political-psychological variables perform
in individual countries across the four regions. In particular, the impact of
political-psychological variables on vote choice also varies from one country
to another across the geo-political regions. However, the result that stands out
is that relating to political interest. As previously shown by the pooled data
across the different regions, the impact of political interest is the strongest
political-psychological variable across the individual countries. Political
interest is significant and positive in 16 of the 26 countries but uncovers a
large amount of variation between countries in the regions. For instance,
in post-Communists European countries and East Asian democracies, this
effect is significant and positive in all the countries we analysed. For Latin
America, in four of the countries, Chile, Costa Rica, Guyana, and Panama,
political interest has a positive effect on electoral participation. While in the
remaining five countries, Brazil, Uruguay, Peru, El-Salvador, and Argentina,
the impact of political interest is not statistically significant, although being
in the expected direction. Rounding up, with sub-Saharan Africa countries,
Table 6.3 Theoretical matching of political-psychological variables on vote choice:
based on the multivariate results
Models/ Expected Western Africa Latin East Asia Post-

variables effect/ Europe America communist
direction Europe
Political-psychological model
Political + Applies*** Applies*** Applies*** Applies*** Applies***
interest
Political + Applies*** Does not Does not Does not Applies***
trust apply apply apply
Internal + Applies*** Does not Does not Does not Applies***
efficacy apply apply apply
External + Applies*** Does not Does not Does not Applies***
efficacy apply apply apply

AsianBarometer, and European Social Survey. Applies***: This signifies that the relationship
those who declared themselves to be interested in politics are more likely to

participate in three of the eight countries (i.e. South Africa, Mauritius, and
Namibia); however, the impact of political interest is not significant in the
remaining countries.
Second, for the remaining political-psychological variables (e.g. political
trust, internal and external efficacy), their effect on vote choice across the
individual countries is revealed to be relatively weak or non-existent. These
results are, however, not too surprising given the lack of significance of these
variables in the pooled vote choice model (see Table 6.2). Yet, a closer look at
the country by country results suggests political trust, internal and external
efficacy seem to best capture vote choice in post-communist European states.
In particular, in post-communist Europe, the impact of political trust on
turnout is positive in five of the six countries. Internal efficacy has a positive
impact on voting in three of the six countries (i.e. Czech Republic, Hungary,
and Lithuania) while the impact of external efficacy is statistically significant
only in two of the countries (the Czech Republic and Lithuania). In sum,
given these results, it is clear the political-psychological model is revealed to
be more consistent in post-communist states (as in Western European democ-
racies) while East Asia, Africa, and Latin America democracies are proved to
resemble each other.
Conclusion
This chapter continues with our causal sequence of individual- level
explanations of vote choice in newer democracies by focusing on political-
psychological variables. It explores how political interest, political trust,
internal and external efficacy affect people’s voting decisions in four regions.
The overall contribution of this model to the vote choice model indicates it
is quite modest only in new democracies of post-communist European coun-
tries while for the other regions, sub-Saharan Africa, Latin America, and East
Asia, the general contribution of the model to vote choice is relatively weak or
tiny. To be precise, the surprise finding regarding the political-psychological
model that is tested is that all these variables except for political interest prove
to be non-significant for voting in sub-Saharan Africa, Latin America, and
East Asia. However, the clear message from this chapter is that among the
political-psychological variables, political interest constitutes the main driving
force behind people’s decision of whether to vote or not across all new dem-
ocracies. Consequently, the main implications of this result are that, given
that interest about politics positively inspires people to go to the polls, as indi-
viduals continue to gain political knowledge and experience with the further
consolidation of democracy across these emerging democracies, voters will
be more likely to hold their government, leaders, and policy-makers account-
able. This is positive for the overall quality of the democratic process in newer
democracies.
Notes
1 That is the feeling that political and social change is possible and that the individual
citizens can play a part in bringing about the change.
2 Internal efficacy refers to the belief that one can personally influence the political
process, whereas external efficacy refers to the belief that the political system is
responsive to one’s views (Balch 1974; Niemi, Craig and Mattei, 1991).
3 www.worldvaluessurvey.org/WVSDocumentationWV6.jsp.
4 Although I attempt to present a picture of people’s psychological orientation based
on political trust, as pointed out by Shi (2014), it might be argued whether political
trust has the same meaning in such diverse systems or regimes in the globe.
5 For established Western European democracies, the Nagelkerke pseudo R2 increases
by 0.035.
7
Alternative explanations and the vote
So far, to establish why people vote in the newly consolidated democracies, and
as illustrated from the previous empirical chapters (see e.g. Chapter 4, 5, and
6), I have relied heavily on theories of voting behaviour that were primarily
developed based on the experience of voters from older and economically
more developed democracies. However, considering the newness of elections
across emerging democracies, coupled with their cultural and historical leg-
acies, I contend new democracies could differ from established democracies
of the global North in terms of individual-level turnout.
To be precise, most countries in sub-Saharan Africa, Latin America,
East Asia, and post-Communist Europe continue to be plagued by wide-
spread corruption allegations, weak institutions, and socioeconomic
exclusion or inequality. Therefore, at the heart of this chapter is the
need to identify those factors that set newer democracies apart from the
established Western democracies. To do this, I examine what I deem to
be an alternative explanation1 of turnout that focuses on the impact of
corruption, authoritarianism, and the economy. In the first part of this
chapter, I explore and provide some descriptive and qualitative infor-
mation about the three core explanatory variables. In the second section
of the chapter, relying on inferential statistics, I estimate the effects of
corruption, authoritarianism, and the economy on vote choice across the
four areas with newer democracies.
Assessing the scale of corruption: any regional differences?

In sub-Saharan Africa, Latin America, East Asia, and post- Communist
European countries, it would not be an overstatement to say not a day goes
by without news or reports of corruption allegation and or the embezzlement
of public funds by public officials. Thus, as reported in Table 7.1, relying on
Transparency International Corruption Perception Index (CPI), I begin by
assessing the scale of national corruption across the countries that I analyse
in the different geo-political regions.
Alternative explanations and the vote 131
Table 7.1 Transparency International Corruption Perception Index
Regions/countries Corruption Perception Index
Sub-Saharan Africa
Benin 41
Botswana 61
Cape Verde 58
Ghana 41
Mauritius 52
Namibia 52
Sao Tome and Principe 46
South Africa 44
Latin America
Argentina 45
Brazil 35
Chile 67
Costa Rica 56
El-Salvador 34
Guyana 40
Panama 36
Peru 36
Uruguay 71
East Asia
Taiwan 65
Mongolia 35
South Korea 59
Czech Republic 56
Estonia 74
Hungary 44
Lithuania 60
Poland 58
Slovenia 60
Western Europe
Austria 77
Belgium 75
Germany 80
Finland 86
United Kingdom 77
Ireland 74
Italy 53
Netherlands 82
Norway 84
Sweden 85
Regional mean
Sub-Saharan Africa 49
Latin America 47
East Asia 53
Post-communist Europe 58
Western Europe 77
Source: Transparency International Corruption Perceptions Index for 35 countries, 2019

132 Alternative explanations and the vote
The Corruption Perception Index (CPI) places each country on a scale
ranging from 0 (highly corrupt) to 100 (very clean).2 First, of the countries
analysed here and assessed using the 2019 CPI, older established democracies
of Western Europe, notably Finland, Sweden, Netherlands, and Germany,
have the highest CPI rating of 80 and above while new democracies, par-
ticularly El-
Salvador, Brazil, and Mongolia, have the lowest CPI rating
ranging between 34 and 35. However, a close look at the figures indicates
that corruption is more prevalent in sub-Saharan Africa, followed by Latin
America than in any other global region.
Second, the cross-regional comparison suggests corruption perception
varies across countries in different geo-political regions. Thus, as cited
by Rose and Peiffer (2019), attempts at generalising about corruption in
countries grouped by geography or culture might be very misleading. To
be precise, Table 7.1, shows that the variations in corruption perception in
countries or national contexts within the different geo-political regions are
higher than the difference between the mean rating for each of the regions.
For instance, in Western Europe, regarded as the best performing region
(i.e. incidence of corruption is quite low), there is a 33 point difference
between the highest performing country, Finland, and the lowest performer,
Italy. Moving to East Asia, there is a spread of 30 points gap between the
highest performing country, Taiwan, and the lowest-performing country,
Mongolia. Strangely, this highlights what Wedeman (2002) refers to as
the “East Asian Paradox”, pervasive corruption parallels notable eco-
nomic growth.3 For sub-Saharan Africa, although the region is generally
considered as the most corrupt, there is a range of 20 points gap between
the CPI rating of Botswana, highest performer in the region, and Benin
or Ghana, bottom performers. Concerning Latin American countries, the
variations across the countries are once more persistent with a high of 71
points in Uruguay to a low of 34 and 35 points in El-Salvador and Brazil,
respectively. Finally, among post-communist European states, there is a
spread of 30 points between the highest-ranking country, Estonia, and the
lowest ranking country, Hungary. In sum, just as I have previously alluded,
Table 7.1 reconfirms the incidence of corruption is quite a problem in new
democracies, although it equally varies from one country to another. Thus,
it would be more interesting to see or examine how voters’ perception of
corruption in their countries influence their decision of whether or not to
turn out on election day.
Economic concerns and voters’ retrospective evaluations

As previously mentioned in the theoretical section (i.e. see Chapter 2), new
democracies have made tremendous economic and development progress
within a short period. Yet many of the countries, particularly those in sub-
Saharan Africa and Latin America, are still considered as regions with a
modest income, endemic poverty, and wide-spread socioeconomic inequality
and exclusion. To substantiate this point, descriptive evidence from the
different cross-national survey data analysed here reveals the most crit-
ical problem encountered by the citizens in sub-Saharan Africa, East Asia,
and post-Communists European are those related to the management of
the economy, unemployment, and/or income/wages issues except for Latin
American were crime registered as the most crucial problem, follow by those
relating to the economy.
For instance, across the eight sub-Saharan Africa countries, 34 per cent
of the respondents declared unemployment as the most crucial problem
(AB, 2016). For East Asia democracies, 21 per cent reported both man-
agement of the economy and unemployment as the most critical country
problems, followed by wages/incomes/salaries problems constituting 11 per
cent (AsianBarometer, 2014–2016). Concerning, Latin America, although
25 per cent of respondents declared the most crucial problems were those
related to crime. Yet the remaining problems have to do with unemploy-
ment and the economy standing at 10 and 11 per cent,4 respectively (The
AmericasBarometer, 2014). All said and done; given the importance of the
economy across these regions, I begin by looking at people’s perception
the economy and how it relates to vote choice. Thus, Table 7.2 presents
the distribution of respondents’ socio-tropic retrospective evaluation of
the performance of their national economy in the different geo-political
regions.
Although citizens’ assessment of their country’s economic condition does
vary from one region to another, what is clear in Table 7.2 is that most of the
respondents said the overall performance of their economy/or country eco-
nomic situation had become worse than it had been one year before. Thus,
with the economy being sufficiently important in new democracies, it would
be interesting to explore how voters’ evaluations of their country’s economic
condition help distinguish voters from non-voters.
Table 7.2 Voters’ socio-tropic retrospective evaluation of national economy in new

democracies
Sub-Saharan Latin East Asia Post-communist

Africa America Europe
Retrospective Worse 36.6 44 49 41.1

evaluation Same 32.8 40 29 34.4
Better 30.6 16 22 23.4
Notes: percentages are derived from the different cross-national survey data. European Social
Survey (Eastern European countries); Afrobarometer (Sub-Saharan Africa); AmericasBarometer
(Latin America); AsianBarometer (East Asia).
Authoritarian practices and tendencies in new democracies

With the introduction of democracy following the third wave, it would have
seemed possible to envisage this development would thoroughly sweep away
authoritarian tendencies and practices that characterised the period of mili-
tary rule, dictatorship, one-party states, and communism in sub-Saharan
Africa, Latin America, East Asia, and post-Communist European states.
However, this has not been the case considering many countries have soon
fallen into a recession. Or better still, there have been re-new resurgences of
authoritarian tendencies or what scholars now termed democratic erosion,
democratic backsliding, or as coined by Lührmann and Lindberg (2019), the
Third Wave of Autocratisation. Bermeo (2016) contends democratic back-
sliding denotes state-led debilitation or elimination of any of the political
institutions that sustain an existing democracy. Lust and Waldner (2015) posit
democratic backsliding is the deterioration or decline of qualities associated
with democratic governance within any regime. In fact, across sub-Saharan
Africa, Latin America, East Asia, and post-communist European states,
most countries have now witnessed the rise of dominant parties or executives
that have sought to control other branches of government, suppress oppos-
ition and undermine press/media freedom. To be precise, data from the
Democratic Erosion Data Event Dataset (2019)5 captures the symptoms and
precursors of democratic erosion across countries over time. Figures 7.1 and
7.2 confirm there have been resurgences of authoritarian tendencies in new
democracies.
Figure 7.1 represents the frequency of democratic erosion symptoms for
ten years across four regions (Africa, Asia, North and South America, and
Europe). Several things stand out: first, the frequency of democratic erosion
has been increasing over time in the four regions, and this has been more
60
50
40
30
20
10
0
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Africa Asia Americas Europe
Figure 7.1 Democratic erosion: number of symptoms by year and region

Source: Democratic Erosion Data Event Dataset (Gottlieb et al., 2019).
pronounced in Asia and surprisingly in Europe too, followed by Latin America
and sub-Saharan Africa. Second, while, it is not easy to detect, the high peak
of democratic erosion in the European region has been spearheaded mostly
by Central and Eastern European countries (i.e. Russia, Hungary, Kosovo,
Ukraine, Albania, Serbia, and Poland, etc.). Across Western European coun-
tries, democratic erosion seems to be low and less severe. Third, for the Asia
region, democratic erosion has risen over the past decades. Although the high
rate tends to be masked by countries such as China, Hong Kong, Thailand,
the Philippines, the severity of the erosion is equally welled documented across
East Asian liberal democracies such as those of Japan and South Korea and
Mongolia.6 Next, regarding the North and South America regions, it is essen-
tial to note that democratic backsliding has been more pronounced in coun-
tries such as Venezuela, Nicaragua, Haiti, El-Salvador, Brazil, and Mexico,
while democratic erosion has been less pronounced in others such as Uruguay,
Chile, and Costa Rica. Finally, rounding up with the Africa region, although
the data suggest democratic backsliding is quite low (i.e. compared to other
lobal regions) the region’s worse worse performers have been Kenya, Zambia,
and Botswana, etc.
Figure 7.2 shows the types of democratic erosion symptoms per region.
Here, the frequency of the kinds of erosion tends to be distinct across the
different regions. Yet, four signs stand out as the most dominant: media repres-
sion, curtailment of civil liberties, repression of the opposition, and reduction
of judicial independence. A closer look at each of the regions reveals that
media repression and curtailment of civil liberties is the dominating symptom
Revsion of the constuon
No confidence vote/decreased vote turnout
Weakened civil service/integrity instuons
Reducon in legislave oversight
Systemac reducon in elecon freedom/fairness
Reducon of judicial independence
Relaxaon of term limit
Repression of opposion
Suspension of the rules/constuon
Curtailed civil liberes
Media repression
0 10 20 30 40 50 60 70 80 90
Europe Americas Asia Africa
Figure 7.2 Democratic erosion: types of symptoms by region

Source: Democratic Erosion Data Event Dataset; symptoms are from 2000–2018.
of democratic erosion in Asia, North and South America, and Western Europe
(i.e. particularly across countries in East Asia, post-Communist Europe, and
Latin America countries). Similarly, we can confirm there also seem to be
more institutional attacks such as the reduction in electoral freedoms, and
judicial independence is more common in the relatively more developed states
in Europe.
Overall, as reported in Figures 7.1 and 7.2, we have seen that across
newer democracies, there has been a decline or erosion of the essential
quality of democracy. So the question posed at this juncture is: where
does democratic erosion lead to? I argue democratic erosion may lead to
a decline in citizens’ confidence in democracy and or the deconsolidation
of democracy. Bermeo (2016, p. 6) contend where democratic backsliding
includes rapid and radical change across a broad range of institutions, it
may lead to regimes that are explicitly authoritarian. Similarly, where back-
sliding takes the form of gradual changes across a more circumscribed set
of institutions, it is less likely to lead to an all-out regime change and more
likely to yield political systems that are ambiguously democratic or hybrid.
Putting this in context, a recent report from Varieties of Democracy pro-
ject (i.e. V-dem, 2018, p. 20) reveals four post- Communists European
countries, notably Hungary, Poland, Lithuania, and Slovakia, and two
countries from sub-Saharan Africa, South Africa and Mauritius, have
lost their status as liberal democracies to become electoral democracies.
In sum, I have explored and provided descriptive information on the per-
vasive of corruption and authoritarianism but also citizen’s evaluation of
their country’s economic condition. The next section examines how these
alternative mechanisms will impact on people’s voting decisions across the
different geo-political regions.
Corruption, authoritarianism, economy, and the vote

Here, our vote choice model is extended by investigating the impact of
corruption, authoritarianism, and the economy. Drawing from our recur-
sive casual model, the model I examine in this chapter includes the socio-
demographic, mobilisation, and political psychology from the previous
chapters. To these models/variables, I now add those indicators that pertain
to the alternative variables. In sum, the model for this fourth block in the
recursive casual system reads as follows:
Vote Choice = f (Socio-demographic, Mobilisation, Political-psychology,

and Alternative Variables).
However, before estimating the effects of all four blocks of variables on vote
choice, the bivariate relationship between the three alternative variables and
voter turnout is first assessed across all four newer democracies.
Corruption, authoritarianism, economy, and the vote –bivariate

relationships
Table 7.3 depicts the bivariate relationships between corruption, authori-
tarianism, and the economy and vote choice. Starting with corruption, the
bivariate evidence suggests in three regions (i.e. sub-Saharan Africa, East
Asia, and post-Communist Europe) voters’ perception of corruption tends to
decrease turnout. However, for Latin America, we observe a non-significant
relationship between corruption perception and voting. Still, the coefficient is
negative, which is in line with the results for the other three regions.
Next, looking at the impact of the state of the economy on voting, a topic
most citizens considered to be the main problem in their country, our data
reveal mixed findings across the four regions. On the one hand, for East Asia
and post-Communist democracies (i.e. as in Western European cases), we find
Table 7.3 Corruption, authoritarianism and the economy and voting in new democra-
cies: bivariate relationships
Region/alternative variables Spearman’s

rho
Western Europe
Corruption perception -0.04**
Authoritarianism (democratic erosion = 1; no democratic erosion = 0) N/A
Perception of the economy (better = 1; same = 0.5; worse =0) 0.07**
Sub-Saharan Africa
Authoritarianism (democratic erosion = 1; no democratic erosion = 0) -0.07**
Perception of the economy (better = 1; same = 0.5; worse = 0) 0.01
Latin America
Corruption perception -0.01
Authoritarianism (democratic erosion = 1; no democratic erosion = 0) 0.01
Perception of the economy (better = 1; same = 0.5; worse = 0) -0.02*
East Asia
Corruption perception -0.03*
Authoritarianism (democratic erosion = 1; no democratic erosion = 0) 0.05**
Perception of the Economy (better = 1; same = 0.5; worse = 0) 0.04**
Authoritarianism (democratic erosion = 1; no democratic erosion = 0) 0.12**
Perception of the economy (better = 1; same = 0.5; worse = 0) 0.09**
Source: Bivariate results between voting and the alternatives variables across the regions
(2-tailed).
a weak but significant relationship between voter’s perception of the economy
and voter turnout. Although, this association is quite weak, it is apparent
citizens who have a more positive evaluation of their country’s economic
conditions are more likely to participate than those who lack this positive
evaluation. However, the same cannot be said for new democracies of sub-
Saharan Africa and Latin America.
Finally, concluding with our last alternative variable, authoritarianism. It
is essential to note that because of data availability, I was unable to empirically
put this variable to test for the Western European countries.7 Notwithstanding
our inability to test this variable, the proposition vis-à-vis this variable was
that in countries experiencing a surge in authoritarian practices, voters would
be more likely to engage. Table 7.3 shows the results across the four regions
tend to be varied. For example, in East Asia and post-Communist European
cases, I can confirm voters are significantly more likely to vote in countries
experiencing the erosion of democratic principles. For sub-Saharan African
countries, I observe the opposite, in that a voter’s experience with authori-
tarianism or the erosion of democratic values significantly reduces the prob-
ability of voting. Concluding, with the Latin American region, I observe a
non-significant relationship between authoritarianism and voting. Although
our bivariate test of the impact of the alternative variables on vote choice in
new democracies does provide some clues about which factor is relevant or
not in which context/region, only a multivariate model would enable us to
be more confident of these results. Thus, the model that I build in the next
section takes into account all four blocks of variables (i.e. socio-demographic,
mobilisation, political-psychological, and alternative factors).
The vote choice: a multivariate model

Table 7.4 provides the detailed results of the logistic regression model with
voting as the dependent variable. (Table C7, Appendix 3, reports results
for descriptive statistics across the different geo- political regions.) Note,
following our causal estimation strategy column 3 to 5 (i.e. Model II–Model
IV) shows the effects/analysis of the socio-demographic, mobilisation, and
political-psychological variables already discussed in the previous chapters.
Thus, in this chapter, the focus is on Model V, which includes the effect of
socio-demographic, mobilisation, and the political-psychological variables, to
which I now add variables relating to the alternative mechanism.
First, I look at the Nagelkerke pseudo R2, since it gives an estimate of
the proportion of variations in the dependent variable that is explained or
accounted for by the variables in the analysis. To be precise, the Nagelkerke
pseudo R2 allows us to examine the extent to which corruption, authoritar-
ianism, and the economy contribute to the overall explanation of the vote
choice model across the different geo-political regions. In other words, what
is the fit of the alternative variables as a whole to the survey data across the
different regions? The simple answer to this question is that not very much
newgenrtpdf
Table 7.4 Multivariate models of voter turnout in newly consolidated democracies: socio-demographic, mobilisation, political-psychology
and alternative variables
Regions/variables
Western Europe Model I Model II Model III Model IV Model V
Months since last election 0.20*** (0.02) 0.19*** (0.02) 0.15*** (0.02) 0.14*** (0.02) 0.001 (0.00)
Age 0.03*** (0.00) 0.02*** (0.00) 0.02*** (0.00) 0.02*** (0.00)
Gender (1 = men) 0.07 (0.04) -0.02 (0.04) -0.17*** (0.05) -0.16*** (0.05)
Religion (1 = everyday/once a week; 0 = never) 0.12*** (0.01) 0.10*** (0.01) 0.08*** (0.01) 0.07*** (0.01)
Nominal ethnic identity (Largest ethnic group = 1; 0.96*** (0.09) 0.91*** (0.09) 0.89*** (0.10) 0.90*** (0.09)
Residence (1 = rural) 0.15*** (0.04) 0.16*** (0.04) 0.20*** (0.05) 0.23*** (0.05)
Education 0.56*** (0.03) 0.49*** (0.03) 0.31*** (0.03) 0.30*** (0.03)

Party identification (1 = party ties) 1.13*** (0.05) 0.85*** (0.05) 0.86*** (0.05)
Associational membership (1 = member currently; 0.65*** (0.06) 0.64*** (0.06) 0.72*** (0.06)
Clientelism N/A N/A N/A
Political interest 0.88*** (0.05) 0.85*** (0.05)
Political trust 0.19*** (0.03) 0.17*** (0.03)
Internal efficacy 0.09** (0.04) 0.09** (0.04)
External efficacy 0.11* (0.05) 0.06 (0.05)
Corruption perception 0.12 (0.07)
Authoritarianism N/A
Economic perception (better = 1; same = 0.5; 0.02 (0.06)
worse = 0)
Constant -2.96*** (0.38) -5.74*** (0.42) -5.11*** (0.44) -5.38***(0.46) -2.09*** (0.14)
Nagelkerke’s R2 0.044 0.146 0.209 0.244 0.221
% correctly classified 83% 83% 84% 85% 85%
Number of observations/cases 18657/10 18332/10 17962/10 17318/10 17187/10
(continued)
newgenrtpdf
Table 7.4 Cont.
140
Regions/variables

Sub-Saharan Africa
Months since last election 0.01*** (0.00) 0.02*** (0.00) 0.02*** (0.01) 0.02*** (0.03) 0.01*** (0.00)
Age 0.38*** (0.04) 0.35*** (0.05) 0.29*** (0.05) 0.31*** (0.05)
Gender (1 = men) -0.02 (0.06) -0.11 (0.07) -0.16* (0.08) -0.11 (0.09)
Religion (1 = more than once a day; 0 = never) 0.08*** (0.03) 0.07** (0.03) 0.08*** (0.03) 0.11*** (0.03)
Salience of ethnicity (1 = only ethnic ID; Equal -0.24** (0.10) -0.18 (0.11) -0.14 (0.12) -0.21 (0.12)
national/ethnic ID = 0.5; Only national ID = 0)
Residence (1 = rural) 0.28*** (0.07) 0.20** (0.08) 0.18** (0.08) 0.18** (0.09)
Education -0.07 (0.12) 0.01 (0.13) -0.09 (0.14) -0.20 (0.14)
Associational membership (1 = active member/ 0.56*** (0.11) 0.53*** (0.11) 0.46*** (0.12)
official leader; 0.5 = inactive member; 0 = not a
member)
Clientelism (1 = receive gifts/bribes) -0.24** (0.09) -0.21* (0.11) 0.06 (0.11)
Political trust 0.08 (0.05) -0.01 (0.05)
Internal efficacy 0.06 (0.09) 0.06 (0.09)
External efficacy 0.05 (0.08) 0.14 (0.09)
Corruption perception -0.09** (0.05)
Authoritarianism -0.35*** (0.09)
Economic Perception (better = 1; same = 0.5; -0.02 (0.11)
worse = 0)
Constant 1.62*** (0.35) 0.45*** (0.16) -0.43 (0.65) -0.17 (0.75) -0.22 (0.21)
Nagelkerke’s R2 0.029 0.068 0.098 0.102 0.080
Latin America
Months since last election -0.05*** (0.01) -0.03*** (0.01) -0.03*** (0.01) -0.03*** (0.01) -0.02*** (0.00)
Age 0.05*** (0.00) 0.05*** (0.00) 0.05*** (0.00) 0.05*** (0.00)
Gender (1 = men) -0.12** (0.05) -0.11** (0.05) -0.13*** (0.05) -0.15*** (0.05)
Religion (Catholics = 1; others 0) 0.23*** (0.05) 0.25*** (0.05) 0.22*** (0.05) 0.15*** (0.06)
Ethnicity/race (Whites = 2; Mestizo/Mulattos = 1; -0.02 (0.04) 0.01 (0.05) 0.05 (0.04) 0.03 (0.04)
Blacks/Indigenous = 0.5; others = 0)
Residence (1 = rural) 0.19*** (0.06) 0.15** (0.06) 0.17** (0.07) 0.15** (0.06)
Education 0.73*** (0.07) 0.68*** (0.07) 0.59*** (0.07) 0.55*** (0.07)
Associational membership (1 = member; 0 = not a 0.22*** (0.03) 0.18*** (0.03) 0.17*** (0.04)
member)
Clientelism (1 = receive gifts/bribes) 0.18 (0.14) 0.11 (0.13) 0.08 (0.12)
Political trust -0.05 (0.03) -0.05 (0.03)

Internal efficacy 0.02 (0.02) 0.03 (0.02)
External efficacy -0.01 (0.02) -0.01 (0.02)
Corruption perception -0.06 (0.06)
Authoritarianism -0.05 (0.06)
worse = 0)
Constant 3.18*** (0.33) -0.04 (0.37) 0.21 (0.38) -0.03*** (0.40) -1.98*** (0.15)
Nagelkerke’s R2 0.026 0.172 0.213 0.224 0.210
East Asia
Months since last election -0.04***(0.01) -0.09*** (0.01) -0.04*** (0.01) -0.04** (0.01) -0.03 (0.02)
Age 0.06*** (0.00) 0.06*** (0.00) 0.05*** (0.01) 0.05*** (0.01)
Gender (1 = men) -0.07 (0.09) -0.06 (0.09) -0.08 (0.11) -0.06 (0.10)
Religion (1 = several times a day; 0 = never) 0.12*** (0.04) 0.11** (0.04) 0.11** (0.05) 0.09** (0.04)
(continued)
newgenrtpdf
Table 7.4 Cont.
142
Regions/variables

Nominal ethnic ID (Largest ethnic group = 1; -0.15 (0.15) -0.08 (0.15) 0.02 (0.16) 0.03 (0.16)
Residence (1 = rural) 0.25* (0.13) 0.35** (0.14) 0.30** (0.15) 0.37** (0.15)
Education 0.44*** (0.06) 0.42*** (0.07) 0.31*** (0.07) 0.34*** (0.07)
Associational membership (1 = member; 0 = not a -0.18 (0.24) -0.25** (0.11) -0.21* (0.11)
member)
Political trust 0.04 (0.07) 0.04 (0.07)
Internal efficacy -0.05 (0.10) -0.06 (0.10)
External efficacy -0.01 (0.16) 0.02 (0.16)
Corruption perception -0.25** (0.13)
Authoritarianism -0.43 (0.25)
worse = 0)
Constant 2.65*** (0.24) -0.65 (0.36) -0.66** (0.39) -0.96** (0.43) -0.72*** (0.62)
Nagelkerke’s R2 0.009 0.136 0.175 0.189 0.195
Months since last election -0.01 (0.00) 0.00 (0.01) -0.01 (0.01) -0.01 (0.01) 0.01*** (0.00)
Age 0.03*** (0.00) 0.02*** (0.00) 0.02*** (0.00) 0.02*** (0.00)
Gender (1 = men) 0.22*** (0.05) 0.09* (0.05) -0.08 (0.05) -0.07 (0.05)
Religion (1 = everyday/once a week; 0 = Never) 0.16*** (0.02) 0.15*** (0.02) 0.13*** (0.02) 0.13*** (0.02)
Nominal ethnic ID (Largest ethnic group = 1; 0.45*** (0.09) 0.52*** (0.11) 0.44*** (0.12) 0.43*** (0.12)
Residence (1 = rural) -0.05 (0.05) -0.00 (0.05) 0.03 (0.05) 0.03 (0.05)
Education 0.60*** (0.03) 0.53*** (0.06) 0.39*** (0.04) 0.38*** (0.04)
Party identification (1 = party ties) 1.73*** (0.06) 1.34****(0.06) 1.35*** (0.06)
Associational membership (1 = member currently; 0.54*** (0.09) 0.42*** (0.10) 0.36*** (0.09)
Political trust 0.24*** (0.03) 0.20*** (0.03)
Internal efficacy 0.20*** (0.06) 0.18*** (0.06)
External efficacy 0.22** (0.07) 0.22** (0.07)
Corruption perception -0.49*** (0.08)
Authoritarianism 0.67*** (0.06)
Economic perception (better = 1; same = 0.5; 0.23** (0.07)
worse = 0)
Constant 1.15*** (0.14) -1.66*** (0.21) -1.45*** (0.22) -1.63*** (0.25) -2.39*** (0.18)

Nagelkerke’s R2 0.036 0.146 0.273 0.334 0.334
AsianBarometer (East Asia). The data are weighted using the weights provided for each survey data. Significance: ***p < 0.001; **p < 0.01; *p < 0.05. Entries
are standardized logistic regression coefficients, standard errors in parentheses.
as the Nagelkerke pseudo R2 barely moves in some regions, while it reduces
in others. Second, I examine which of the socio-demographic, mobilisation,
and political-psychological indicators that were included in the third block of
our recursive casual system continue to matter for electoral participation even
after the alternative variables have now been included. The straight response
to this is that across the different regions, the introduction of the alternative
variables does not change the values of the socio-demographic, mobilisation,
or political, psychological variables very much, as the results remain mostly
unchanged.
Third, following up on the bivariate results (see Table 7.3), I probe
which of the alternative variables seems to matter for vote choice across the
different regions. As shown in Table 7.4, looking at the impact of corruption,
across across three newer democracies sub-Saharan Africa, East Asia, and
post-Communist European democracies (i.e. except for the Latin America
region8) the result reveals corruption has a significant and negative effect on
people's decision of whether to vote or not. To be precise, in sub-Saharan
Africa, East Asia, and post-Communist European democracies, this result
implies citizens’ perception of corruption in their countries tend to decrease
turnout significantly. These findings seem to corroborate those of Dahlberg
and Solevid (2016), Stockemer, LaMontagne, and Scruggs (2013), Sundström
and Stockemer (2015), Domínguez and McCann (1998), and Hooghe and
Quintelier (2014). All said and done, this finding gives weight to the fact that
corruption has a fundamentally different influence in newer democracies
when compared to established Western democracies.
Next, concerning authoritarianism, our data tells of a statistically sig-
nificant relationship between authoritarianism and vote choice only in
two newer democracies, notably sub-Saharan Africa and post-communist
European democracies. Yet, the direction of the coefficient does vary across
the two regions. On the one hand, the impact of authoritarianism on vote
choice is significant and positive for post-communist European countries
while it is significant and negative for sub-Saharan Africa. Consequently, we
can infer two things. (1) In post-communist European states, it is apparent
voters’ experience with the rise of authoritarian practices and or demo-
cratic erosion, which is characterised by the curtailed of civil liberties and
press freedom, leads to an increase in voter turnout. (2) In sub-Saharan
Africa, voters’ experience with democratic erosion or authoritarian tenden-
cies tends to decrease voters’ turnout. For the two remaining regions, East
Asia and Latin America, the effect between authoritarianism and voting is
non-significant.
Finally, looking at the economy, the expectation was that voters who had a
more positive evaluation of their country economy would be more likely to vote.
Our result was only confirmed in the post-communist European region. Apart
from this, in the remaining three regions notably sub-Saharan Africa, Latin
America and East Asia the result depicts a non-significant relationship with vote
choice. However, the coefficient across these regions was positively signed.
Conclusion
In this chapter, I proposed a framework to study voter participation that
focuses on the impact of corruption, authoritarianism, and the economy.
Overall, I find distinct patterns in the impact of these factors across the
different regions. On the one hand, my results indicate that the influence of
corruption on voting is quite similar in sub-Saharan Africa, East Asia, and
the post-communist European countries. In these three settings, perceptions
of corruption tend to significantly decrease turnout at the individual level.
However, the same cannot be said for the other two variables: the economy
and authoritarianism. Voters’ evaluations of the economy do not fully attain
significance across most of the regions except in post-communist European
countries, although the coefficient is positive, suggesting voters having a
positive evaluation of their national economy should be more likely to turn
out. The effect of authoritarianism on vote choice is significant only in sub-
Saharan Africa and post-communist European states but, even so, the result
across these regions reveals a mixed pattern. For sub-Saharan Africa, voters’
experience with authoritarian practices and undemocratic tendencies tends to
decrease turnout, whereas the opposite is true for post-communist countries.
Notes
1 Although, I am not the first person to examine the impact corruption, authoritar-
ianism, and the economy in explaining voting behaviour in new democracies. Yet,
except for Tambe (forthcoming), no cross-national study has been able to estimate
all these variables in a single model of vote choice in new democracies.
2 Rose and Peiffer (2019, p. 46) argue this categorisation permits for a greater degree
of differentiation than sorting countries into two categories, corrupt and high in
integrity.
3 Moreover, as Zakaria (1994) and Fukuyama (2001) argue, the so-called Asian
values tend to be conducive to corruption.
4 This was closely follow with corruption, with 8 per cent of the respondents
registering it as a crucial problem.
5 A cross-university collaboration that aims to help evaluate threats to democracy
by capturing the symptoms and precursors of democratic erosion across countries
over time www.democratic-erosion.com/event-dataset/project-summary/ [Accessed
01/03/2020].
6 However, democratic erosion has been so far less severe across East Asia liberal
democracies compared to those that are outright electoral authoritarianism or
authoritarian regimes.
7 In most Western European democracies there is either no democratic backsliding
and or weak threat of future backsliding. Similarly, although there might have been
some precursors to backsliding, such as the rise of extremist parties, erosion of
democratic institutions has not yet taken place.
8 In Latin America, corruption has no significant effect on whether people vote or
not. For a more recent similar result see Rose and Peiffer (2019, p. 134).
8
The electoral context and the vote
The previous chapters have focused predominantly on individual- level

factors that are relevant in shaping political behaviour. While accepting these
individual-level variables drawn from the socio-demographic, mobilisation,
political-psychological, and alternative models are central in explaining voting
behaviour, I argue we should also consider the role that political institutional
context has in shaping the decision of whether or not to participate. As pre-
viously mentioned, a substantial body of literature on electoral participation
already points to the fact that people’s decisions of whether to vote or not
depend not only on individual factors but also on the context in which indi-
viduals find themselves (see Gallego 2015; Singh 2010; Gordon and Segura
1997; Jackman and Miller 1995; Franklin 2004; Bühlmann 2006; Thomassen
2005; Van Egmond, De Graafi, and Van Der Eijk 1998; Van Egmond, 2003).
Gallego (2015, p. 194) suggests that individual (micro) and institutional
(macro) characteristics are not additive. But, instead, they interact with each
other, thereby affecting the decision of individuals to participate in electoral
politics. As Peters (2018) says, the two aspects of political life are, in practice,
closely linked and an understanding of one can help inform the knowledge of
the other. Peters (2018, pp. 42–46) further outlines two mechanisms through
which the institutional context can affect electoral participation. The first
mechanism sees institutions as an opportunity structure. Here, he posits that
the more open institutional structures are to the involvement of citizens, then
the greater encouragement they will provide for participation. At the same
time, the second mechanism sees institutions as an information structure.
Here, Peters argues that the extent to which institutions can reduce the infor-
mation costs of potential participants in the political process, the more likely
those citizens are to participate. In this context, and drawing from histor-
ical institutionalism that sees institutions as formal or informal procedures,
routines, norms and conventions embedded in the organizational structure of
the polity or political economy (Hall and Taylor, 1996)1 it, therefore, makes
sense to see institutions as formalised rules that could either motivate or
hinder individual attitudes/behaviour.
Consequently, this chapter seeks to look at the reciprocal influences of
political institutions and individuals in explaining electoral participation in
The electoral context and the vote 147
newer democracies. In short, in line with the theoretical framework established
earlier (e.g. see Chapter 2), this chapter scrutinises how contextual factors
such as disproportionality of the electoral system, the closeness of the
election, concurrent elections, and compulsory voting condition the role of
the individual-level factors in our vote choice model. The first section of this
chapter draws on aggregate voter turnout data to justify the importance of
combining individual-level and institutional approaches to explain individual-
level turnout. This section also presents some qualitative information on the
institutional variables to be subjected to empirical testing across the different
geo-political regions. In the second section of this chapter, using bivariate and
multivariate analysis, our vote choice model is extended by adding the last
block of variables (institutional variables) in our vote choice model.
The vote: why combine the individual and contextual approach?

In this study, it has been made clear from the onset that the best way to study
vote choice is obviously to employ individual-level data as their focus is on the
voters. However, to overcome an essential methodological challenge (i.e. con-
text problem) often associated with studies of voting behaviour, I have argued
it is vital to give due consideration to the context in which an individual vote
is cast. To elucidate, using the socio-demographic, mobilisation, political-
psychological, and alternative models, I have pinpointed several factors
that tend to determine electoral participation generally across democracies.
Although some of these individual-level variables have been shown to work in
similar ways across diverse geo-political regions (that is, in older democracies
of Western Europe as well as newly consolidated democracies of sub-Saharan
Africa, Latin America, East Asia, and post-Communist European countries),
an important question to ask is: does this imply we now know for a certainty
what determines electoral participation or vote choice? Well, as I will show in
the proceeding section, this is not quite the case.
Voter turnout trends in new democracies

Table 8.1 shows the trends in voter turnout2 between the first election and the
most recent national elections held in each country that I analyse across the
different geo-political regions. Note for new democracies the first elections are
considered as those conducted following the third wave of democratisation. In
contrast, for established democracies of Western Europe, although elections
have long been conducted here, for the sake of comparability I considered
elections that took place around the 1980s.
The figures suggest voter turnout tends to vary substantially across elections
in the different countries/regions. Moreover, using a strategy I adopted earlier
(see Tambe, 2018b, pp. 3–5) that maps turnout trends based on the following:
(1) countries that have a high voter turnout and turnout remains high;
(2) countries that have a high turnout and turnout later declines; (3) countries
148 The electoral context and the vote
Table 8.1 Trends in voter turnout in new democracies
Regions Countries First election Most recent

election
Year Turnout Year Turnout

(%) (%)
Sub-Saharan Benin 1991 64 2016 66

Africa Botswana 1989 68 2019 84
Cape Verde 1991 75 2016 66
Ghana 1992 50 2016 68
Mauritius 1991 84 2019 77
Namibia 1994 74 2019 61
Sao Tome and Principe 1991 76 2018 80
South Africa 1994 86 2019 66
Latin America Argentina 1983 83 2019 81
Brazil 1989 88 2018 79
Chile 1989 94 2017 46
Costa Rica 1982 78 2018 65
El-Salvador 1989 54 2019 52
Guyana 1980 93 2020 72
Panama 1984 73 2019 73
Peru 1985 91 2016 80
Uruguay 1984 87 2019 90
East Asia Taiwan 1991 68 2020 74
post-communist Mongolia 1993 92 2017 60
European South Korea 1992 71 2020 66
Czech Republic 1990 96 2017 60
Estonia 1990 78 2019 63
Hungary 1990 65 2018 69
Lithuania 1990 71 2016 50
Poland 1989 62 2019 61
Slovenia 1992 85 2018 52
Western Europe Austria 1983 92 2019 75
Belgium 1981 94 2019 88
Germany 1980 88 2017 76
Finland 1983 75 2019 68
United Kingdom 1983 72 2019 67
Ireland 1981 76 2020 62
Italy 1983 89 2018 72
Netherlands 1981 87 2017 81
Norway 1981 81 2017 78
Sweden 1982 91 2018 87
Source: www.idea.int/data-tools/data/voter-turnout. Notes: Data for Central/Eastern European

countries based on parliamentary elections. While those for sub-Saharan Africa, East Asia, and
Latin America countries are based on countries’ presidential or parliamentary elections.
that have a low voter turnout and turnout remains low; (4) countries that have
a low voter turnout and turnout then increases; (5) countries that have a stable
level of turnout over time, enable us to make the following inferences. First,
Table 8.1 reveals, across the different geo-political regions, in some countries
mostly in Latin America and Western Europe (e.g. Argentina, Uruguay, Peru,
Sweden, Netherlands and Belgium) that have a high voter turnout in their first
election, turnout has remained high is most of the countries. Second, in some
countries that have a high turnout during their first elections (e.g. mostly post-
communist European states and Chile and Mongolia for Latin America and
East Asia), turnout has later declined. Third, in a few countries (i.e. Poland)
that have a low voter turnout, turnout has remained low. Fourth, in some
countries (Taiwan) that had a low voter turnout, turnout has later increased.
Fifth, for Western European democracies, although turnout as shown a slight
decline, yet most of the countries have a stable level of turnout overtime.
In general, Table 8.1 paints a picture of turnout variation across the different
regions, which therefore raises the question of whether these variations can be
in any way accounted for by changes in voters’ descriptive characteristics. That
is, could it be that the composition of the electorate in terms of their socio-
demographic characteristics (education levels) or political- psychological
factors (political interest) across these countries/regions might have changed
over time to explain the patterns depicted in Table 8.1? For instance, focusing
on post-communist European countries where the decline in voter turnout has
been more drastic, could it be that the high turnout rate in the 1990s can be
explained by the presence of highly educated and politically interested voters,
who, later on, disappear in elections that show lower turnout? However, this
should not be the case as studies have long suggested the impact of individ-
uals’ characteristics on vote choice tends to be rather stable over time. In his
study of the determinants of electoral participation in the Netherlands, Van
Egmond (2003) argues: “people tend to be rather stable in most of their indi-
vidual characteristics and behaviour. If they change, this is likely to be can-
celled out by another voter changing in another direction” (p. 5).
Thus, in this chapter, I argue to understand electoral behaviour fully,
it is essential to combine information on individual and institutional
characteristics. I further gain assurance in this line of argument given that some
authors like Coleman (1990) but also Carmines and Huckfeldt (1996) posit
that combining both levels (micro and macro) bring considerable benefits as it
offers explanations on how contextual effects translate into individual behav-
iour. Relatedly, uniting both approaches enables one to position individuals’
characteristics within the boundaries of the political context, which there-
fore helps displays the influences institutional factors can have on individual
behaviour. To illustrate this point, Van Egmond, De Graafi, and Van Der
Eijk (1998, p. 282) using political interest and closeness of elections show the
forms these influences can take. The authors posit during a competitiveness
election –contextual factors such as intensive media coverage may increase
political interest in the electorate and consequently aggregate turnout.
However, this relationship might equally operate in the opposite direction
as alluded to by Campbell (1960). Campbell argues extensive media coverage
of a close race might ensure that voters who are less interested in politics
become fully aware of the political situation, thereby reducing the impact of
political interest. Drawing from these two examples, it is clear the contextual
characteristics or electoral context might function in either impacting dir-
ectly or indirectly individual voting behaviour. Thus, placing the voters within
the context of the election gives me the leeway to test the hypothesis on the
influences of individual and institutional variables on vote choice. All said,
having outlined some of the merits for combining individual and contextual
factors to explain vote choice in new democracies, I round up this section
by presenting some qualitative information on the institutional variables that
will be subjected to empirical testing.
New democracies and their political-institutional characteristics

Recent reviews of macro-level voter turnout indicate contextual factors that
could condition voter behaviour are quite extensive (e.g. see Stockemer, 2017;
Cancela and Geys, 2016). Yet, they do provide an idea of which indicators are
more influential in explaining electoral turnout. In the theoretical section (i.e.
Chapter 2), the review of the comparative scholarship had enabled us to iden-
tify some of these factors: compulsory voting, concurrent elections, the dis-
proportionality of the electoral system, and closeness of the election, which
I intend to empirically put to the test in the context of newly consolidated
democracies.
Table 8.2 presents an overview of these contextual characteristics for each
of the countries across the different geo-political regions.
First, beginning with compulsory voting, recall I expected that people
would be more likely to participate where the electoral laws ensure voting
is not only compulsory but enforced with sanctions. As shown in Table 8.2,
in new democracies except for Latin America countries, none of the other
regions: sub- Saharan Africa, East Asia, and post- Communist European
countries have any law mandating compulsory voting. Thus, because of data
availability, the impact of compulsory voting on vote choice will be tested only
in the context of Latin America. A close look at the Latin America region
reveals in four countries, Argentina, Brazil, Uruguay, and Peru, compulsory
voting exists, and it is enforced with sanctions.3 In contrast, the remaining five
either have no compulsory voting laws, or compulsory voting operates but is
not enforced with sanctions. Accordingly, the impact of compulsory voting
on vote choice in Latin America will be compared to Western Europe (i.e.
even here just Belgium has mandatory voting laws).
Second, for concurrent elections, I have argued that turnout will be high in
countries where elections are concurrently held. Here, a concurrent election
is assumed to have taken place when presidential and legislative elections
or when national and local elections are being conducted on the same day.
Thus, this meant that our operationalisation of concurrent elections excludes
elections that are scheduled apart. Table 8.2 suggests national elections are
synchronous mostly in Latin America (i.e. in seven of the nine countries).
In sub-Saharan Africa, I find this to hold in two of the eight countries (i.e.
newgenrtpdf
Table 8.2 Countries and values of their political-institutional characteristics
Regions/ Year*** Compulsory Concurrent Disproportionality Closeness of

countries Voting elections the election
Benin 2011 No No 14.17 No

Botswana 2014 No No 16.91 No
Cape Verde 2011 No No 1.95 Yes
Ghana 2012 No Yes 5.76 Yes
Sub-Saharan Africa Mauritius 2014 No No 20.44 No
Namibia 2014 No Yes 0.90 No
Sao Tome and Principe 2014 No No 7.27 No
South Africa 2014 No No 0.37 No
Argentina 2011 Yes Yes**** 7.55 No
Brazil 2010 Yes Yes 2.50 Yes*
Chile 2013 No Yes 8.04 No

Costa Rica 2014 No Yes 6.23 No*
Latin America El-Salvador 2012 No No 3.36 Yes*
Guyana 2011 No No 0.81 No
Panama 2014 No Yes 7.87 No
Peru 2011 Yes yes 10.23 Yes*
Uruguay 2014 Yes yes 1.00 No
Taiwan 2012 No Yes 9.07 Yes
East Asia Mongolia 2012 No No 11.77 Yes
South Korea 2012 No No 7.15 Yes
Post-communist Europe Czech Republic 2013 No No 6.12 Yes
Estonia 2015 No No 2.34 Yes
Hungary 2014 No No 17.80 No
Lithuania 2016 No No 12.56 Yes
Poland 2015 No No 12.56 No
(continued)
newgenrtpdf
Table 8.2 Cont.
152
Regions/ Year*** Compulsory Concurrent Disproportionality Closeness of

countries Voting elections the election
Slovenia 2014 No No 6.57 No
Austria 2013 No No 3.31 Yes
Belgium 2014 Yes Yes 4.60 No
Germany 2013 No No 7.83 No
Finland 2015 No No 3.03 No
Western Europe United Kingdom 2015 No Yes 15.02 No**
Ireland 2016 No No 5.62 Yes
Italy 2013 No No 17.34 Yes
Netherlands 2012 No No 0.99 Yes
Norway 2013 No No 2.56 No
Sweden 2014 No Yes 2.65 No
Notes: The sources from which these institutional characteristics are calculated are: Africans Elections Database; Political Database of the Americas; Adam
Carr Elections Archives; Gallagher (2019). Election indices dataset; www.idea.int/data-tools/data/voter-turnout/compulsory-voting. * = for these countries
closeness of the election result derives from the second rounds of the presidential elections. ** = the gap in percentage between the Conservative and the Labour
party was 6.5. However, this was rounded as 7 thus not a close race. *** = this year reflects the calculation of the least square index of disproportionality. Where
in doubt consult Table 3.3. ****In Argentina, elections are partially concurrent with presidential elections. However, the country 2011 presidential and legisla-
tive elections were conducted on the same day.
Ghana and Namibia). For East Asia, of the three countries, elections are
concurrently held only in Taiwan. For post-communist Europe, concurrent
elections were non-existent in all six countries. Rounding up, with Western
European countries of the ten countries it was only in Sweden and the UK
where national and local elections tend to be held concurrently.
Third, for the electoral system, the expectation was that proportional
representation (PR) would enhance people’s propensity to vote. This prop-
osition is based on the fact that most cross-national studies, notably Blais
and Dobrzynska (1998) and van der Eijk and Franklin (1996), show that
disproportionality reduces turnout since the translation of votes into
seats tends to be unequal leading voters to abstain. Across the different
regions (old and new democracies), it is clear the electoral system across the
majority of Western European countries except that of the UK, Italy, and
Germany is predominately proportional. However, this pattern is entirely
the opposite when I focus on newer democracies.4 In new democracies,
except for a few cases in sub-Saharan Africa (e.g. South Africa and Cape
Verde) and Latin America (Guyana, Brazil, Uruguay, etc.) that tend out
to be a parliamentary or presidential system with a proportional represen-
tation system, the electoral system is highly disproportional as the figures
suggest.
Finally, our last institutional variable is the closeness of the election. The
expectation was that in elections where voters perceive it to be a close race
or competitive, turnout will be high. Recall, I operationalise closeness of
the election as the margin of victory for the winning candidate or over the
runner-up in presidential elections while, for parliamentary democracies,
I measure the closeness of election as the difference in seat shares between
the top two parties. For simplicity’s sake and to ensure comparability across
the different regions, I take a 6 per cent threshold as a difference in the seat
shares between the top two parties or between the runners-up in presidential
elections. That said, in Table 8.2, Yes is assigned when the winning margin
is 6 per cent or less (i.e. closed race) and No when the margin is greater than
6 per cent. Using this criterion, it appears East Asian elections were more
competitive in terms of the closeness of the results. This variable will not be
examined in this context because it is a constant for the election year/period
I analyse. On the other hand, across the different regions/countries, there is
a wide variation of the closeness of the election results. Having identified
the empirical indicators, the next section asks if these political-institutional
variables are related to electoral participation in newly consolidated
democracies.
The electoral context and the vote: bivariate relationships

Table 8.3 reports the bivariate tests between vote choice and the political-
institutional variables: disproportionality, the closeness of the election, con-
current elections and compulsory voting.
Table 8.3 The impact of political and institutional factors on voting: bivariate
relationship
Region/institutional variables Spearman’s rho
Western Europe
Least squares index of disproportionality (0 = perfect -0.09**
proportionality to 100 maximum disproportionality)
Closeness of election -0.08**
Concurrent elections 0.05**
Compulsory voting 0.07**
Sub-Saharan Africa
Least squares index of disproportionality (0 = perfect 0.03**
Closeness of election 0.01
Concurrent elections 0.01
Compulsory voting N/A
Latin America
Least squares index of disproportionality (0 = perfect -0.02
Closeness of election 0.05**
Concurrent elections 0.02*
Compulsory voting 0.11**
East Asia
Closeness of election N/A
Concurrent elections -0.05**
Closeness of election -0.12**
Concurrent elections N/A
Source: Bivariate results between voting and the Institutional variables across the regions
(2-tailed).
First, using Western European cases so as compare the experiences of

people’s voting decisions across newer democracies, it is evident that dis-
proportionality of electoral systems has a significant but negative effect on
electoral participation only in Western European democracies. This result
suggests quite clearly that the more proportional elections outcomes are, the
more inclined voters are to turn out. For new democracies, contrary to the
expectation, in sub-Saharan Africa, East Asia, and post-communist democra-
cies (i.e. except for Latin America), the relationship is significant and positive.
In this respect, western European democracies are quite different from new
democracies.
Second, moving to the closeness of the elections, note this variable is
not subjected to empirical testing in East Asia as it was a constant (i.e. see
Table 8.2). However, regarding the other regions, except for sub-Saharan
Africa, the result shows a significant relationship with voting in Latin America
and post-communist Europe (i.e. as in older democracies of Western Europe).
However, a closer look at the sign of the measure suggests the findings are
mixed. For instance, for the Latin American region, I find as expected that
those who perceive their country election to be a close race or competitive
are more likely to vote. In contrast, in post-communist Europe (as in Western
European democracies) the relationship runs in the opposite direction,
suggesting those who perceive elections as close or competitive are less likely
to have voted.
Third, I also examine the effect of concurrent elections on electoral partici-
pation. The proposition was that citizens would be more likely to participate
when elections are concurrently held. That said, looking at the preliminary
bivariate results suggest this is true only for Latin America countries (as in
Western European cases). While for East Asian democracies the result tells
of a significant effect but ran in an unexpected direction. Thus, there is no
consistency of finding regarding the concurrentness of the electoral contests
and voting.
Lastly, I also examine the effect of compulsory voting on vote choice.
Although this is a significant institutional variable, I was unable to put this
variable to test across all four regions empirically, in particular in sub-Saharan
Africa, East Asia, and post-communist countries as none of these regions or
countries had any compulsory voting. Thus, we were left with just the Latin
America region comparing the result with those of Western European democ-
racies (even here, compulsory voting is present only in Belgium). Our findings
confirm that in Latin America as in the older democracies of Western Europe,
citizens showed a higher propensity to vote when voting is compulsory and
enforced with sanctions. So far, these bivariate relationships of the impact of
political-institutional variables on electoral participation in new democracies
reveal little consistency. To confirm and or disprove these preliminary results,
I proceed by building a multivariate analysis of vote choice where the indi-
vidual and country-level determinants of electoral participation are modelled
simultaneously.
Individual and political-institutional determinants and the vote: a

multivariate model
So far, I have examined the question of who votes across four regions by
conducting a series of multivariate regressions that look at the effects of
individual- level variables drawn from socio- demographic, mobilisation,
political-psychological, and an alternative model. Nevertheless, as highlighted
above, to adequately explain people’s decision of whether to vote or not, it
imperative to place voters within the context in which this election takes place.
Thus, Table 8.4 presents our multivariate model (i.e. Appendix 3, Table C8,
provides summary statistics for the institutional variables) for block five, the
last in our recursive causal system, which reads as follows:
Vote Choice = f (Socio-demographic, Mobilisation, Political-psychology,

Alternative and Political-Institutional variables)
In Table 8.4 (Model VI, column 7), the political-institutional factors are now
added into a model that already consists of all the individual-level explana-
tory variables. Accordingly, it is of foremost importance to examine if the
explanatory power of each of the individual-level blocks of factors from the
previous chapter somehow vary once these political-institutional factors are
included. We observe two things though: first, the addition of the political-
institutional predictors increases the Nagelkerke’s pseudo R2 value only min-
imally in sub-Saharan Africa and Latin America (i.e. as in Western European
democracies) except for East Asia and post-communist European democ-
racies.5 In fact, in East Asia and post-Communist European countries, the
Nagelkerke’s pseudo R2 barely moves or changes, revealing the electoral con-
text variables tend to be less influential than the individual-level indicators.
Second, across the different regions, we observe most of the individual-level
estimates show little or no change when the contextual factors are added to
our vote choice model. To be precise, controlling for the political-institutional
effects on voting, the socio-demographic variables, namely age, education,
and religiosity, continue to retain a highly significant and positive impact on
electoral participation. Similarly, a look at the mobilisation variables suggests
party identification and associational membership continues to emerge as sig-
nificant predictors of voting at the individual level. Also, political interest
remains the most crucial political-psychological predictor of vote choice, even
when political-institutional variables are taken into account. Consequently,
considering the political-institutional variables that are entered in Model VI
(Table 8.4, column 7), although the absence of data availability has hidden the
empirical testing across all regions, it is evident the results are more nuanced
across the different geo-political regions.
First, the disproportionality of the electoral system has a significant and
negative effect only in Western European democracies while for new dem-
ocracies the relationship tends to be non-significant except for sub-Saharan
African democracies where it is positive and significant at the conventional
level. This result simply replicates those of the comparative literature and
accentuates the debate why disproportionality of the electoral system has a
negative effect on electoral participation in older advanced democracies but
not in new democracies (see Gallego, Roco, and Anduiza, 2012). Overall, the
result reveals individuals living in countries where the disproportionality of
the electoral system is close to zero (i.e. more proportional) are more likely to
newgenrtpdf
Table 8.4 Multivariate models of voter turnout in newly consolidated democracies: individual and political-institutional variables
Regions/variables
Western Europe Model I Model II Model III Model IV Model V Model VI
Months since last election 0.20*** (0.02) 0.19*** (0.02) 0.15*** (0.02) 0.14*** (0.02) 0.001 (0.00) 0.01*** (0.00)
Age 0.03*** (0.00) 0.02*** (0.00) 0.02*** (0.00) 0.02*** (0.00) 0.02*** (0.00)
Gender (1 = men) 0.07 (0.04) -0.02 (0.04) -0.17*** (0.05) -0.16*** (0.05) -0.15*** (0.05)
Religion (2 = everyday/once a week; 0.12*** (0.01) 0.10*** (0.01) 0.08*** (0.01) 0.07*** (0.01) 0.11*** (0.01)
0 = never)
Nominal ethnic identity (Largest 0.96*** (0.09) 0.91*** (0.09) 0.89*** (0.10) 0.90*** (0.09) 0.94*** (0.10)
ethnic group = 1; Minority ethnic
group = 0)
Residence (1 = rural) 0.15*** (0.04) 0.16*** (0.04) 0.20*** (0.05) 0.23*** (0.05) 0.21** (0.05)
Education 0.56*** (0.03) 0.49*** (0.03) 0.31*** (0.03) 0.30*** (0.03) 0.30*** (0.03)
Party identification (1 = party ties) 1.13*** (0.05) 0.85*** (0.05) 0.86*** (0.05) 0.85*** (0.05)

Associational membership 0.65*** (0.06) 0.64*** (0.06) 0.72*** (0.06) 0.64*** (0.06)
0.5 = member previously; 0 = not
a member)
Clientelism N/A N/A N/A N/A
Political interest 0.88*** (0.05) 0.85*** (0.05) 0.85*** (0.05)
Political trust 0.19*** (0.03) 0.17*** (0.03) 0.17*** (0.03)
Internal efficacy 0.09** (0.04) 0.09** (0.04) 0.14** (0.04)
External efficacy 0.11* (0.05) 0.06 (0.05) 0.09 (0.05)
Corruption perception 0.12 (0.07) 0.16 (0.09)
Authoritarianism N/A N/A
Economic Perception (better = 1; 0.02 (0.06) 0.00 (0.07)
same = 0.5; worse = 0)
Disproportionality -0.02*** (0.01)
(continued)
newgenrtpdf
Table 8.4 Cont.
158
Regions/variables

Closeness of election 0.12 (0.07)
Concurrent election 0.38*** (0.08)
Compulsory voting 0.97*** (0.12)
Constant -2.96*** -5.74*** -5.11*** -5.38***(0.46) -2.09*** (0.14) -2.62*** (0.17)
(0.38) (0.42) (0.44)
Nagelkerke pseudo R2 0.044 0.146 0.209 0.244 0.221 237
% correctly classified 83% 83% 84% 85% 85% 85%
Number of observations/cases 18657/10 18332/10 17962/10 17318/10 17187/10 17187/10
Sub-Saharan Africa
Months since last election 0.01*** (0.00) 0.02*** (0.00) 0.02*** (0.01) 0.02*** (0.03) 0.01*** (0.00) 0.02*** (0.01)
Age 0.38*** (0.04) 0.35*** (0.05) 0.29*** (0.05) 0.31*** (0.05) 0.31*** (0.06)
Gender (1 = men) -0.02 (0.06) -0.11 (0.07) -0.16* (0.08) -0.11 (0.09) -0.13 (0.08)
Religion (2 = more than once a day; 0.08*** (0.03) 0.07** (0.03) 0.08*** (0.03) 0.11*** (0.03) 0.09*** (0.03)
0 = never)
Salience of ethnicity (1 = only -0.24** (0.10) -0.18 (0.11) -0.14 (0.12) -0.21 (0.12) -0.16 (0.13)
ethnic ID; Equal national/ethnic
ID = 0.5; Only national ID = 0)
Residence (1 = rural) 0.28*** (0.07) 0.20** (0.08) 0.18** (0.08) 0.18** (0.09) 0.16** (0.09)
Education -0.07 (0.12) 0.01 (0.13) -0.09 (0.14) -0.20 (0.14) -0.10 (0.14)
(1 = active member/official leader;
0.5 = inactive member; 0 = not a
member)
Clientelism (1 = receive gifts/bribes) -0.24** (0.09) -0.21* (0.11) 0.06 (0.11) -0.21 (0.12)
Political trust 0.08 (0.05) -0.01 (0.05) 0.03 (0.05)
Internal efficacy 0.06 (0.09) 0.06 (0.09) 0.05 (0.10)
External efficacy 0.05 (0.08) 0.14 (0.09) 0.05 (0.10)
Corruption perception -0.09** (0.05) -0.09** (0.05)
Authoritarianism -0.35*** (0.09) -0.64** (0.28)
Economic Perception (better = 1; -0.02 (0.11) 0.08 (0.11)
Disproportionality 0.02* (0.01)
Closeness of election 1.26*** (0.30)
Concurrent election 0.23 (0.49)
Constant 1.62*** (0.35) 0.45*** (0.16) -0.43 (0.65) -0.17 (0.75) -0.22 (0.21) -0.14 (0.24)
Nagelkerke pseudo R2 0.029 0.068 0.098 0.102 0.080 0.098
Latin America

Months since last election -0.05*** -0.03*** -0.03*** -0.03*** (0.01) -0.02*** (0.00) -0.01*** (0.00)
(0.01) (0.01) (0.01)
Age 0.05*** (0.00) 0.05*** (0.00) 0.05*** (0.00) 0.05*** (0.00) 0.05*** (0.00)
Gender (1 = men) -0.12** (0.05) -0.11** (0.05) -0.13*** (0.05) -0.15*** (0.05) -0.13*** (0.05)
Religion (Catholics = 1; others 0) 0.23*** (0.05) 0.25*** (0.05) 0.22*** (0.05) 0.15*** (0.06) 0.21*** (0.06)
Ethnicity/race (Whites = 2; -0.02 (0.04) 0.01 (0.05) 0.05 (0.04) 0.03 (0.04) 0.04 (0.05)
Mestizo/Mulattos = 1; Blacks/
Indigenous = 0.5; others = 0)
Residence (1 = rural) 0.19*** (0.06) 0.15** (0.06) 0.17** (0.07) 0.15** (0.06) 0.22*** (0.06)
Education 0.73*** (0.07) 0.68*** (0.07) 0.59*** (0.07) 0.55*** (0.07) 0.63*** (0.08)
(1 = member; 0 = not a member)
Clientelism (1 = receive gifts/bribes) 0.18 (0.14) 0.11 (0.13) 0.08 (0.12) 0.07 (0.13)
(continued)
newgenrtpdf
Table 8.4 Cont.
160
Regions/variables

Political trust -0.05 (0.03) -0.05 (0.03) -0.06 (0.03)
Internal efficacy 0.02 (0.02) 0.03 (0.02) 0.02 (0.02)
External efficacy -0.01 (0.02) -0.01 (0.02) -0.01 (0.02)
Corruption perception -0.06 (0.06) -0.01 (0.06)
Authoritarianism -0.05 (0.06) -0.23** (0.08)
Economic Perception (better = 1; 0.04 (0.08) 0.01 (0.07)
Disproportionality 0.02 (0.01)
Closeness of election 0.03 (0.07)
Concurrent election 0.61 (0.13)
Compulsory voting 0.77*** (0.09)
Constant 3.18*** (0.33) -0.04 (0.37) 0.21 (0.38) -0.03*** (0.40) -1.98*** (0.15) -1.65*** (0.16)
East Asia
Months since last election -0.04***(0.01) -0.09*** -0.04*** -0.04** (0.01) -0.03 (0.02) -0.03 (0.01)
(0.01) (0.01)
Age 0.06*** (0.00) 0.06*** (0.00) 0.05*** (0.01) 0.05*** (0.01) 0.05*** (0.01)
Gender (1 = men) -0.07 (0.09) -0.06 (0.09) -0.08 (0.11) -0.06 (0.10) -0.06 (0.11)
Religion (2 = several times a day; 0.12*** (0.04) 0.11** (0.04) 0.11** (0.05) 0.09** (0.04) 0.09** (0.05)
0 = never)
Nominal ethnic ID (Largest ethnic -0.15 (0.15) -0.08 (0.15) 0.02 (0.16) 0.03 (0.16) 0.03 (0.17)
group = 1; Minority ethnic
group = 0)
Residence (1 = rural) 0.25* (0.13) 0.35** (0.14) 0.30** (0.15) 0.37** (0.15) 0.37** (0.16)
Education 0.44*** (0.06) 0.42*** (0.07) 0.31*** (0.07) 0.34*** (0.07) 0.34*** (0.08)
Associational membership -0.18 (0.24) -0.25** (0.11) -0.21* (0.11) -0.21* (0.11)
(1 = member; 0 = not a member)
Political trust 0.04 (0.07) 0.04 (0.07) 0.04 (0.07)
Internal efficacy -0.05 (0.10) -0.06 (0.10) -0.06 (0.11)
External efficacy -0.01 (0.16) 0.02 (0.16) 0.02 (0.17)
Corruption perception -0.25** (0.13) -0.25** (0.13)
Authoritarianism -0.43 (0.25) -0.02 (0.14)
Economic perception (better = 1; 0.01 (0.15) 0.01 (0.15)
Closeness of election N/A
Concurrent election N/A

Constant 2.65*** (0.24) -0.65 (0.36) -0.66** (0.39) -0.96** (0.43) -0.72*** (0.62) -2.33 (0.50)
Nagelkerke pseudo R2 0.009 0.136 0.175 0.189 0.195 0.195
Number of observations/cases 3956/03 3724/03 3635/03 3194/03 3089/03 3091
Months since last election -0.01 (0.00) 0.00 (0.01) -0.01 (0.01) -0.01 (0.01) 0.01*** (0.00) 0.01** (0.00)
Age 0.03*** (0.00) 0.02*** (0.00) 0.02*** (0.00) 0.02*** (0.00) 0.02*** (0.00)
Gender (1 = men) 0.22*** (0.05) 0.09* (0.05) -0.08 (0.05) -0.07 (0.05) -0.07 (0.05)
Religion (2 = everyday/once a week; 0.16*** (0.02) 0.15*** (0.02) 0.13*** (0.02) 0.13*** (0.02) 0.13*** (0.02)
0 = never)
Nominal ethnic ID (Largest ethnic 0.45*** (0.09) 0.52*** (0.11) 0.44*** (0.12) 0.43*** (0.12) 0.43*** (0.12)
group = 1; Minority ethnic
group = 0)
(continued)
newgenrtpdf
Table 8.4 Cont.
162
Regions/variables

Residence (1 = rural) -0.05 (0.05) -0.00 (0.05) 0.03 (0.05) 0.03 (0.05) 0.03 (0.05)
Education 0.60*** (0.03) 0.53*** (0.06) 0.39*** (0.04) 0.38*** (0.04) 0.38*** (0.04)
Party Identification (1 = party ties) 1.73*** (0.06) 1.34****(0.06) 1.35*** (0.06) 1.34*** (0.07)
0.5 = member previously; 0 = not
a member)
Political trust 0.24*** (0.03) 0.20*** (0.03) 0.20*** (0.03)
Internal efficacy 0.20*** (0.06) 0.18*** (0.06) 0.18** (0.06)
External efficacy 0.22** (0.07) 0.22** (0.07) 0.22** (0.07)
Corruption perception -0.49*** (0.08) -0.49*** (0.09)
Authoritarianism 0.67*** (0.06) 0.66*** (0.07)
Economic Perception (better = 1; 0.23** (0.07) 0.24** (0.08)
Closeness of election -0.54 (0.04)
Concurrent election N/A
Constant 1.15*** (0.14) -1.66*** -1.45*** -1.63*** (0.25) -2.39*** (0.18) 1.05*** (0.20)
(0.21) (0.22)
Number of observations/cases
AsianBarometer (East Asia). The data are weighted using the weights provided for each survey data. Significance: ***p < 0.001; **p < 0.01; *p < 0.05. Entries
are standardised logistic regression coefficients, standard errors in parentheses.
participate than individuals residing in countries where the electoral system is
highly disproportional.
Second, looking at the closeness of the election,6 note this was only
tested across three newer democracies, notably sub-Saharan Africa, Latin
America, and post-communist European states. Also, recall we expected to
find a significant and positive influence on voting. Our analyses across the
different regions show that closeness of the elections meets the significance
level and in the expected direction only in sub-Saharan Africa. This result
indicates that compared to African voters who live in countries where the
elections are not considered to be close or competitive, those who live in
countries where the elections are close are more likely to vote. In contrast,
for post-communist European countries and Latin America as in Western
European democracies, the impact of the closeness of the election on vote
choice is non-significant. While these results might seem counter-intuitive,
empirically it is, in fact, not such a rare finding (Blais 2006, p. 116; Bratton
et al., 2010). Also, in their meta-analysis of individual-level research on
voter turnout, Smets and Ham (2013) find that only 35 per cent of studies
relying on national-level measures of closeness find a relationship between
closeness and vote choice while none of the studies using district-level
measures of closeness found any link.
Third, for concurrent elections, issues relating to data availability in post-
communist European states and problems of multicollinearity in East Asia
prevented the testing of this variable in both regions. Thus, we were only able
to empirically investigate the impact of concurrent elections on vote choice in
sub-Saharan Africa and Latin American democracies (as in Western European
democracies). Still, we expected that holding of simultaneous elections would
show a positive and significant effect on voting. However, I can only con-
firm that holding concurrent elections does not have a significant effect on
electoral participation in new democracies of sub-Saharan Africa7 and Latin
America (see also Bratton et al., 2010). Lastly, compulsory voting reveals a
significant and positive relationship with electoral participation in new dem-
ocracies of Latin Americans. To be precise, the coefficient estimate suggests
Latin America are twice as likely to vote when elections are compulsory and
enforced with sanctions than when they are not.
Conclusion
It was the aim of this chapter to empirically test the individual and con-
textual variables that I have previously spelt out in our theoretical section
as crucial in influencing people’s decisions of whether or not to vote. At the
individual level, most of the explanatory variables proved significant in the
expected direction in terms of our hypotheses. The most important findings
are the importance of age, education, religion, political interest, party mem-
bership and associational membership. Overall, these largely confirm earlier
studies on electoral participation in Western Europe. Turning our attention to
the contextual variables, I have mostly been unable to find that institutional
factors make much difference to people’s decision to vote in elections across all
geo-political regions. To be precise, with regards to the four main contextual
variables that I have proposed as important in determining people propensity
to vote, I find little consistency; indeed, the results are quite nuanced.8 Yet, the
effect of compulsory voting on political behaviour stands out. To be precise,
voters in new democracies (Latin America) are more likely to turn out where
these elections are mandatory and enforced with sanction than where they are
not. With regards to the other variables, I find that the closeness of elections
tends to be influential in the new democracies of sub-Saharan Africa but not
for the other regions. In sum, this chapter concludes that while the political-
institutional context matters, it is far from decisive as most of the difference
in whether or not a voter turns out on election day in the different regions still
very much depends on individual-level factors.
Notes
1 Also, “institution” means everything from formal structures like a parliament to
very amorphous entities like social class, with other components of the socio-
political universe such as law and markets, also being described as institutions
Peters (2005, p. 29).
2 Note, these turnout figures are drawn only on elections where the most important
figure of the executive is being elected (i.e. national elections).
3 www.idea.int/data-tools/data/voter-turnout/compulsory-voting.
4 However, this result is not too surprising given the dominance of mixed, majority,
and or plurality systems operating in most of these countries.
5 This is not too surprising given that I have been unable to put all the institutional
variables to test in these two regions.
6 A number of interaction effects between closeness of election and political interest,
internal and external efficacy were test across new democracies of sub-Saharan
Africa and Latin America (as in Western European democracies) but proved to
non-significant.
7 We should note that this result is somewhat inconclusive for sub-Saharan Africa
given how few cases there have been of elections held concurrently (i.e. only Ghana
and Namibia having some kind of concurrent election in this dataset).
8 However, I can confirm to observe some of the most substantial relationships on
the impact of the political-institutional variables in Western European democracies.
9
A comparative overview
The core purpose of this study has been to explain what influences people
to vote at the individual level and to establish how well the models of voter
turnout developed in established democracies can travel to newly consolidated
democracies. In order to do this, I have relied on models and or theories that
were initially formulated based on the experience of voters in older established
democracies; this, of course, flags the importance of understanding how far
these Western- generated theories could travel to other contexts (i.e. new
democracies).
Having conducted the relevant empirical analysis, I am now in a position
to address the key question. Does the theoretical explanation of voting derived
from the experience of voters from older established democracies fit the emer-
ging or new democracies of sub-Saharan Africa, Latin America, East Asia,
and post-Communist Europe well? Drawing on the empirical results across the
different geo-political regions, this chapter looks systematically at whether the
findings reported in the previous five chapters (i.e. Chapters 4, 5, 6, 7, and 8)
are generally consistent with each other. This will enable us to gauge how far
Western-derived models suit developing or emerging democracies. All said,
this chapter summarises what we have learned concerning the predictors of
voter turnout in newly consolidated democracies and the validity of our alter-
native explanation to vote choice.
Electoral participation in new democracies: are the models

consistent?
So far, our prediction of vote choice in new democracies has focused on
two sets of factors. First, individual-level variables that are drawn from the
socio-demographic, mobilisation, political- psychological, and alternative
explanations. Second, I have equally highlighted the importance of political-
institutional contexts such as disproportionality, concurrent elections, com-
pulsory voting, and closeness of elections. Thus, in this section, I pose the
following questions: How much consistency do we find across the different
regions? Do the established models work equally well across all of these global
166 A comparative overview
regions, or only in some of them? And which explanatory variables seem to be
most important?
As previously mentioned, asking these questions is important for the
following reasons: First, because they enable us to identify which of the
following scenarios is most applicable: (1) whether there is something dis-
tinctive about voting at the individual level in new democracies generally when
compared to established democracies; (2) whether there are no differences
between Western Europe, Central/ Eastern Europe, sub- Saharan Africa,
Latin America, and East Asia, but only general models of the widest applica-
tion; (3) whether new democracies in post-Communist Europe, sub-Saharan
Africa, Latin America, and East Asia generally resemble each other but differ
from the older democracies of Western Europe; (4) or lastly, whether each
region is unique in some way.
Second, the question is important as it enables us to pinpoint whether
some models are generally of little use, while others perform very strongly. In
the section that follows, my task is geared towards providing answers to these
questions by looking at each of these questions. To be precise, to achieve these
objectives, I compare and contrast the findings across the different regions
(i.e. newly consolidated democracies and with those from older established
democracies, which were our base group) by presenting the major effects for
each region side by side in a summary table. Table 9.1 does this by reporting
the direction and significance of each of the individual and contextual effects
for each of the geo-political regions.
How much consistency do we find across the four regions?

In an attempt to provide an answer to this question, and by relying on
Table 9.1 presented above, I proceed by systematically cross-checking each of
the five main models across the different global regions. This will be done by
summarising the expectation and findings for each of the individual-level and
political-institutional models or explanations. All said, when a given variable
is significant in two or three of the geo-political regions and is in the expected
direction of the hypothesis, this provides strong support for our conclusion.
Beginning with the first block of variables in our recursive causal system (see
Chapter 4), I examine the impact of six socio-demographic variables, not-
ably age, gender, religion, ethnicity/race, place of residence, and education
on vote choice. First, as illustrated in Table 9.1, age proves to be consistent
across all new democracies; as is generally true of established Western democ-
racies, older citizens are far more likely to vote than younger ones. Moreover,
the individual country results across the different regions showed that the
impact of age was positive and significant in 25 of the 26 countries of sub-
Saharan Africa, Latin America, East Asia, and post-communist European
newgenrtpdf
Table 9.1 Summary of empirical findings
Models Variables Expected effect Newly consolidated democracies Older established

across all democracies
regions
Sub-Saharan Latin America East Asia Post-communist Western Europe
Africa Europe
Socio- Age + +, Significant*** +, Significant*** +, Significant*** +, Significant*** +, Significant***

demographic Gender (male) + -, Insignificant -, Significant*** -, Insignificant -, Insignificant -, Significant***
Religion + +, Significant*** +, Significant*** +, Significant** +, Significant*** +, Significant**
Ethnicity/race + -, Insignificant +, Insignificant +, Insignificant +, Significant*** +, Significant***
Place of residence + +, Significant** +, Significant** +, Significant** +, Insignificant +, Significant***
(rural)
Education + -, Insignificant +, Significant*** +, Significant*** +, Significant*** +, Significant***
Mobilisation Party identification + +, Significant*** +, Significant*** +, Significant*** +, Significant*** +, Significant***
Associational + +, Significant*** +, Significant*** -, Insignificant* +, Significant*** +, Significant***
membership
Clientelism + -, Insignificant +, Insignificant N/A N/A N/A
Political- Political interest + +, Significant*** +, Significant*** +, Significant*** +, Significant*** +, Significant***
psychological Political trust + +, Insignificant -, Insignificant +, Insignificant +, Significant*** +, Significant***
Internal efficacy + +, Insignificant +, Insignificant -, Insignificant +, Significant** +, Significant**
External efficacy + +, Insignificant -, Insignificant +, Insignificant +, Significant** +, Insignificant
A comparative overview 167

Alternative Corruption perception + -, Significant** -, Insignificant -, Significant** -, Significant*** -, Insignificant
explanations Economic perception + +, Insignificant +, Insignificant +, Insignificant +, Significant** +, Insignificant
Authoritarianism + -, Significant*** -, Significant** -, Insignificant +, Significant*** N/A
Political/ Disproportionality - +, Significant* +, Insignificant +, Insignificant +, Insignificant -, Significant***
institutional Closeness of election + +, Significant*** +, Insignificant N/A -, Insignificant +, Insignificant
Concurrent elections + +, Insignificant +, Insignificant N/A N/A +, Significant***
Compulsory voting + N/A +, Significant*** N/A N/A +, Significant***
Significance: *** p<.001, ** p<.01, * p<.05. Direction: + = positive relationship, given the coding used. –= negative relationship, given coding used. Bold
variables and results across the different regions indicates instances for which it has not been possible to use identical question wording across all four regions
(so there is a chance that any differences might simply be artefactual rather than substantive).
democracies. These results, therefore, strongly reconfirm the transcendent
importance of age observed in the pooled data across four newer democ-
racies. One implication of this result is that it undermines the demand by
some advocates to reduce the voting age to 16, particularly with respect to
sub-Saharan Africa (Juma, 2011). As shown by our results across the four
geo-political regions (i.e. sub-Saharan Africa, Latin America, East Asia, and
Eastern Europe), such a decision to lower the voting age will not necessarily
result in a higher participation rate of young people in electoral politics as
these young individuals are revealed to be least likely to vote when I consider
the socio-demographic variables separately, but more importantly, their reluc-
tance to turnout is evident even when individual-level variables are combined
with country-level variables.
Second, turning to gender, the most striking factor is that it is consist-
ently negatively signed across all newer democracies as in older democracies
of Western Europe. To be precise, a closer look at the significance suggests
the relationship between gender and voting is non-significant across three
newer democracies, sub-Saharan Africa, East Asia, and post-Communist
Europe. However, when gender is found to be significant as in Latin America
or Western European democracies, it is usually women who turn out to
vote at higher rates than men. This result, therefore, confirms research that
continues to show the gender gap in electoral participation has gradually
declined or disappeared. Research on the emergence of a “gender gener-
ation gap” in voting across established democracies shows how the political
behaviour of women has been changing across the generations; this study
suggests that it agrees across countries and geo-political regions as well.
Could this be in part because the changes that have driven the emerging
“gender-generation gap” in Western countries have occurred at the same
rate elsewhere? Moreover, these gender-related differences might not just
affect electoral participation but equally impact non-electoral participation
(i.e. signing petition, boycotting, and demonstration). Therefore, revisiting
the global gender differences in political participation will enable us to fully
establish and be clearer if men or women tend to engage differently in polit-
ical activity other than voting.
Third, looking at religion, it is clear this variable seems to best capture
the effect of socio-demographic on vote choice in new democracies (as in
established democracies of Western Europe). To be precise, the data suggest
church attendance and religious affiliation (for Latin America) shows clear
consistency across all regions judging from the statistically significant rela-
tionship and has the expected sign (positive) in all regions. However, the
country by country results suggest the impact of religion was only significant
in 46 per cent of the cases (i.e. 12 out of 26 countries). A closer look at this
individual country result shows the impact of religion tends to vary signifi-
cantly from region to region, albeit a strong and more consistent effect of
religion in post-communist European states than it is compared to the other
global regions.
Fourth, the impact of ethnicity/race variable tends to vary significantly from
one region to another. Although this nuanced result is somehow unavoidable
since it has not been possible to use identical question wording to tap ethni-
city in one way or another, it is evident that the effect of ethnicity is positive
and significant in post-communist European states as in Western European
democracies. This implies that those who identify as belonging to an ethnic
majority in European democracies (i.e. Eastern and Western Europe) are more
likely to vote. Concerning Latin America and East Asian democracies, the
effect of ethnicity on vote choice is non-significant but points to the expected
direction (i.e. positive sign). In summary, with sub-Saharan Africa democra-
cies, ethnicity falls short of the theoretical expectations. The implication of
the non-significant relationship between ethnic identity and voting, particu-
larly in sub-Saharan Africa, suggests that elections there are much more than
mere ethnic censuses.
Fifth, an individual place of residence and how it affects vote choice
reveals a more systematic or consistent effect across the new democracies. The
data shows that the impact of one’s place of residence (i.e. rural residence) is
statistically significant and has the expected positive sign in three newer dem-
ocracies (e.g. sub-Saharan Africa, East Asia, and Latin America) as in older
democracies of Western Europe. However, in post-communist countries, the
effect of place of residence on vote choice is non-significant.
Finally, the last socio-demographic factor, education, a widely supported
notion in the existing literature suggests that well-educated individuals are much
more likely to turn out in elections than poorly educated citizens. However,
the effect of education on voting shows some consistency only across three
regions, particularly Latin America, East Asia, and post-Communist Europe
(i.e. as in Western European democracies). In Latin America, Eastern Europe,
and East Asia, our results show that turnout is highest among those with
higher levels of education. Still, in sub-Saharan Africa, it is non-significant
and with a negative sign1. All said, in Latin America, Eastern Europe, and
East Asia, all of these effects are statistically significant, so education seems
to matter everywhere.
Still, sub-Saharan Africa stands out from the other regions in this respect.
Notwithstanding the non-significant relationship between voting and educa-
tion in Africa, this result implies two things: (1) within sub-Saharan Africa,
low material status or SES does not pose a barrier to political engagement,
which is an unexpectedly positive finding given that inequality in political par-
ticipation is often documented to affect people’s socioeconomic standings;
(2) at the macro level, what this result shows is that democratic participation
is just as likely to be established and consolidated in poor and underdeveloped
countries (i.e. sub-Saharan Africa) as in the rich and developed countries (see
Przeworski et al. 2000).
Overall, the non-significant relationship between voting and education in
sub-Saharan Africa echoes Gallego (2015), who rebuffs the general assertion
that highly educated citizens turn out to vote at higher rates everywhere by
proposing this differs across countries and is dependent on a set of contextual
features.2 Thus, it will be therefore important for future studies to test whether
the relationship between education and voting depends on regime type or on
other country-level features such as voter registration. Overall, it is fair to say
the socio-demographic resource factors are broadly important for electoral
participation across the global regions, notwithstanding a few differences. In
particular, age, religion, education, and, to some extent, place residence all
serve to increase it.
If we turn to the mobilisation variables that were included in the second block
(see Chapter 5), the effect of three variables has been tested: party identifi-
cation, associational membership, and clientelism. The results obtained are
better illustrated in Table 9.1 and somewhat suggest the mobilisation variables
seem to be important and perform consistently well across most of the newer
democracies. Of the three mobilisation factors, two variables that exert the
most systematic and consistent effect on vote choice are party identification
and associational membership.
First, the party identification variable displays the largest effect; it has
the expected sign (i.e. positive) in all newly consolidated democracies as in
Western European democracies and is statistically significant. Relatedly, the
country by country results reconfirm the pooled results as the impact of
party identification on vote choice is significant in 85 per cent of the cases
(i.e. 22 out of 26 countries in the four regions). Second, an examination of
associational membership shows in three regions sub-Saharan Africa, Latin
America, and post-Communist Europe except for East Asia, the impact of
associational membership on vote choice is statistically significant and posi-
tively signed. Similarly, as earlier indicated in the country by country results,
the impact of associational membership on electoral participation is signifi-
cant in 62 per cent of the cases (i.e. 16 of the 26 countries) across the four geo-
political regions. In these countries and or regions, what is clear is that those
who identified as members of trade unions or informal networks are more
likely to turn out on election day than non-members. Overall, the fact that
these mobilisation variables, particularly party identification, largely mirror
those of advanced Western democracies suggests that even in more patri-
monial and clientelist environments like those of sub-Saharan Africa and or
Latin America, political parties do have a major role in ensuring the engage-
ment of the citizenry in electoral politics. Thus, as cited by Webb and White
(2007), this seems to confirm that one of the key functions that any stable
and effective democracy might expect of parties –fostering political partici-
pation –is largely fulfilled in emerging democracies. But the non-significant
relationship between clientelism and voting in sub-Saharan Africa and Latin
America suggests that elections there are much more than mere clientelist
appeals or censuses.
The political-psychological variables that constitute the third block of our
recursive causal system (i.e. see Chapter 6) examine the impact of four core
variables, notably political interest, political trust, internal and external effi-
cacy on vote choice. First, here, what is evident is that political interest is
the most important political-psychological variable guiding voting in new
democracies as in Western European democracies. To be precise, as Table 9.1
reveals, political interest shows that its effect with electoral participation is
statistically significant and positively signed across all the global regions.
Moreover, though the country by country results did reveal the impact of
political interest on electoral participation tends to vary from one country
to another in the different regions, the clear message from the country by
country result is that the impact of political interest was statistically sig-
nificant across 62 per cent of the cases in all four regions (i.e. 16 of the 26
countries).
Second, turning to the other political-psychological variables (e.g. polit-
ical trust, internal, and external efficacy), the surprising finding regarding
these variables that are tested is that all these variables prove to be non-
significant for voting in three regions, sub-Saharan Africa, Latin America,
and East Asia. However, they do produce some level of consistency with
the other regions, notably Eastern Europe and Western Europe, by produ-
cing a number of parameters that run in the expected direction. Besides,
looking beyond the direction of the relationship, the political-psychological
model is revealed to be more consistent in post-Communist states as in
Western European democracies. In contrast, sub-Saharan Africa, Latin
America, and East Asia democracies prove to be unique. Thus, the most
appropriate conclusion would seem to be that political trust, internal and
external efficacy factors do not work well in new democracies, especially
those of sub-Saharan Africa, Latin America, and East Asia. How far this
might be down to the lack of identical question wording across the regions
in assessing the effects of these political-psychological variables, notably
internal and external efficacy on electoral participation, is impossible to
determine. Still, we certainly cannot regard them as important (for similar
results see, Carlin et al., (2015, p.47) for the Latin America region, Kuenzi
and Lambright (2011, p. 787) for sub- Saharan Africa, and Tambe (2016)
for East Asia.).
Alternative variables
Chapter 7 focused on an alternative explanation of electoral participation,
which constituted the fourth block to our casual recursive system. Among
them, I did consider three variables, notably: corruption, authoritarianism,
and the economy. As shown in Table 9.1, except for the effect of corruption
on vote choice, there appears to be little evidence of consistency as the effects
tend to vary from one region to another. Corruption has a similar impact
on vote choice across three new democracies except for the Latin America
region. In these three regions, namely sub-Saharan Africa, East Asia, and
post-Communist European democracies, the effect of corruption on voting is
significant and negatively signed. This result, therefore, suggests that in new
democracies, voters’ perception of corruption tends to significantly decrease
turnout.
Next, regarding the effect of the economy on vote choice across the
different geo-political regions, Table 9.1 suggests our hypothesis is confirmed
only in post-communist European countries. Here, what is clear is that voters
who have a positive evaluation of their country’s economy are more likely
to participate. Apart from this, across the remaining three newer democra-
cies, sub-Saharan Africa, East Asia, and Latin America, I can only observe
some consistency in the direction of the effect of people’s perception of their
country’s economy. Still, even then, it is not always significant.3
Finally, I examine the effect of authoritarianism on electoral participation.
In both the theoretical and empirical chapters, I have shown a majority of
new democracies are now experiencing democratic erosion characterised by
the rise in media repression, curtailment of civil liberties, and the suppression
of oppositions. Relying on the Democratic Erosion Event dataset (DEED,
version 3), I have sought to examine how these authoritarian practices or
tendencies impact on vote choice. As shown in Table 9.1, the overall picture
depicts authoritarianism has a distinct effect on electoral participation across
new democracies. First, in post-communist Europe, democratic erosion, or
authoritarian tendencies is positive and significantly related to voter turnout.
Thus, the simple implication is that people will be more willing to engage in
voting and cast out such leaders as a way to protect democracies. Second, in
two newer democracies (i.e. Saharan Africa and Latin American countries)
democratic erosion or the subvention of democratic principles is negatively
signed, although statistically significant in sub-Saharan Africa. Given these
results, we can partially conclude any further increase with authoritarian
practices might pose a severe threat to electoral politics in these regions (e.g.
as voters might be less likely to participate).
Political-institutional variables
The last block of our causal recursive model (see Chapter 8) centres on the
political-institutional variables, particularly disproportionality, the closeness
of the election, concurrent elections, and compulsory voting. A picture of
the summary findings (see Table 9.1) suggests the impact of the political-
institutional model is rather mixed across the different regions. In some
respects, some institutional variables are more crucial in sub-Saharan Africa
and Latin America, and this is true of compulsory voting and the closeness
of elections. For instance, in Latin America, voters are more likely to engage
when elections are compulsory and enforced with sanctions. In contrast, in
sub-Saharan Africa, individuals are more likely to participate when elections
are considered closed or competitive. However, our inability to empirically
test all the institutional variables hampers the analysis of institutional effects,
and, overall, it is hard to identify any clear patterns here. What I can say is
that this analysis does suggest that institutions are no guarantee of higher
turnout, but there is a clear need for more context-specific analysis across
and within countries. Detailed studies of particular cases should help in this
regard.
Do established models work equally well across all regions?

Having concluded the first question, where our task was to show how much
consistency we find across the different regions, the next issue is to estab-
lish if the models works equally well across all the global regions, or only in
some of them? Answering this question is challenging because some variables
seem to be very important in predicting people’s propensity to turn out while
others do not. Notwithstanding these difficulties, I can confirm that the socio-
demographic and mobilisation variables seem to work equally well across the
four geo-political regions. Similarly, the alternative explanations seem to work
quite well across the global regions, notably those of sub-Saharan Africa and
post-communist European states. The political-psychological model generally
seems to work well across new democracies of Eastern Europe (as in older
democracies of Western Europe), but has little impact in sub-Saharan, Latin
America, and East Asian democracies. Finally, the political- institutional
model is harder to assess in general terms because of the inconsistency of
findings. But there is some evidence of the effect of compulsory voting and
closeness of the election in newer democracies of Latin America and sub-
Saharan Africa.
However, as previously noted, an important caveat here is that it is quite
challenging to bring a number of institutional variables, such as concurrent
elections and compulsory voting to any kind of empirical testing. For example,
concerning compulsory voting, among our countries/regions of interests,
only Latin America exhibits or implements compulsory voting. To be precise,
in Latin America, among the nine countries included only four (Argentina,
Brazil, Peru, and Uruguay) have compulsory voting that is enforced with
sanctions. Also, concerning a concurrent election, in Eastern Europe, this
variable could not be tested in any way due to the absence of elections that
are concurrently held. Similarly, looking at the electoral system, almost all
of the countries of interest in East Asia had a mixed electoral system. Even
though I sought to overcome this problem by focusing on the disproportion-
ality, at times, there was a very small amount of variations across the different
regions. Given these scenarios, it is therefore difficult to compare these systems
to see which institutional factors do matter in predicting a citizen’s decision
of whether or not to vote. That said, we can state quite unequivocally that
all these conclusions pertaining to the political-institutional variables can be
no more than preliminary, waiting for replication of our study on a larger
number of countries and a more extended time period.
Which explanatory variables are most “consistently effective”

Having tested the various types of effect on the decision to vote in new
democracies, a number of conclusions can be reached with regard to which
variables we find to be most important across the four geo-political regions.
First, beginning with the Socio-Demographic variables age, religion, place
of residence, and education are confirmed to consistently significantly affect
most of the regions. Second, the mobilisational variables that matter a great
deal are associational membership and party identification, which signifi-
cantly increase the chances of turning out to vote. Third, of the political-
psychological attitudes, political interest makes a significant difference to
people’s propensity of whether or not to vote in elections. Lastly, rounding
up with the alternative explanations, corruption and authoritarianism make
a significant difference to people’s propensity of whether or not to vote in
elections. In a nutshell, putting together all the five theoretical frameworks
that I have empirically tested in Chapters 4, 5, 6, 7, and 8, our findings
reveal some similarities across new democracies especially with regards to
socio-demographic and mobilisation and a lesser extent the alternative and
political-psychological factors. The alternative model seems to matter more
in some regions compared to others, and the impact of political-institutional
factors is somewhat mixed and intrinsically harder to assess. Overall, this
suggests that two models (socio-demographic, mobilisation factors) referred
to that have been developed in the older democracies of Western Europe
and the US to explain people’s decision to vote are of general applicability
in most instances and help us understand the propensity to participate in
elections across in the democratic world. The political-psychological, alter-
native, and institutional explanations are either unconvincing or have more
limited application in some regions than others. An important implication
of this is that political organisations, whether they are political parties or
non-electoral institutions such as trade unions or churches, corruption,
and authoritarianism continue to play an important role across all types
of society in developing people’s political knowledge and identities, and in
inspiring them to take part in the democratic process. They do this in inter-
action with features of social structure such as class or religious background
and help to foster political interest.
Notes
1 Suggesting at least that those with the highest level of education are least likely to
participate.
2 For a review of this book, see Tambe (2017b).
3 However, it might be that the non- significant relationship observed here may
be probably explained by the fact very few of the countries’ analysed across the
different regions have undergone severe economic problems or hardship at least as
depicted by the misery index measuring economic difficulties. I would expect the
impact of the economy to be stronger in countries/regions facing more severe eco-
nomic problems.
Conclusion
Campbell, Converse, Miller, and Stokes (1960, p. 3) contend that voting

provides the modern state with a way of connecting the actions of the gov-
ernment with the preferences of the citizenry. It is therefore natural to ask the
question of why and how people vote. In the first instance, because of their
extensive history with democracy and competitive elections, this question
has attracted considerable attention from scholars of older democracies in
Western Europe and North America. However, with the transition of dem-
ocracies to other parts of the globe following the third wave of democra-
tisation a handful of countries notably, those in Latin America, East Asia,
sub-Saharan Africa, and post- Communist Europe started having regular
multiparty elections. Thus, we have seen an enormous literature emerge on
different aspects of democracy and elections as scholars began to examine the
question of who votes in this new context.
Therefore, with over 30 years of democratic experience and electoral par-
ticipation, and from a comparative perspective, I began this study by exam-
ining the question of why people vote in the new democracies of sub-Saharan
Africa, Latin America, East Asia, and Central and Eastern European coun-
tries. Relatedly, I also address the question of how well models or theories
of electoral participation that were initially developed in established democ-
racies “travel” to new democracies. As a useful start, I began by relying on
the explanation of political behaviour that focuses on two broad perspectives
(individual and contextual level), while also incorporating individual-level
explanations that take into account particular features of emerging democ-
racies. Having examined in detail the factors that influence people’s decisions
to turn out on election day across the four broad geo-political regions, sub-
Saharan Africa, Latin America, East Asia, and post-Communist Europe, it
is time to outline some of the limitations of what I have been able to do here,
assess the implications of the results for the wider research field of political
participation, and then round things up by suggesting some directions for
future research.
Conclusion 177
Limitation and practical implications of the main findings

Every study has its caveats, and in this section, I will address some of the
limitations of this book. First, with respect to the contextual level variables,
there are many other factors that I have been unable to address here. Although
comparable survey datasets have provided an opportunity to study people’s
motivations to engage in electoral politics, nevertheless, they cannot address
every conceivable influence in the political, economic, and social environ-
ments that might cause individuals to act in different ways across regional and
national contexts. In order to illustrate this point, Gallego (2015, p. 196) argues
that the Nordic countries tend to have a comprehensive and universalistic wel-
fare state, strong unions, PR electoral systems, and low-income inequality.
This implies if we, therefore, uncover that citizens in these countries are more
likely to vote than citizens from Eastern Europe, Latin America, African,
or East Asia, we may suspect that there is something in this Nordic context
that fosters participation. The question that arises is how does the researcher
tackle such a problem? Gallego contends that researchers usually employ a
regression framework in which the behaviour of interest is regressed against
the preferred contextual feature to specify which contextual characteristics
are considered important. This is, of course, exactly what I have done in this
book. Nevertheless, the shortcoming of this approach is that because the
potential range of differences between countries and regions are so many,
the number of potentially omitted variables are enormous. Much of the time,
therefore, we are unable to tell with precision what it is exactly about the con-
text that affects participation.
Second, in order to explain people’s propensity to turn out in elections
at the individual level, it is important to stress that I tested this with self-
reported turnout as a measure of participation. The limitation of relying on
self-reporting is that it is often clouded with too much over-reporting by indi-
viduals who claimed to have voted while they actually did not do so. That
said, we have no reason to expect over-reporting to vary much by social group,
country, or region, and therefore it should not bias our estimates. Third, the
findings in this book are restricted only to a group of countries that have
succeeded in consolidating their democracies and that have little risk of elect-
oral malpractice such as vote rigging or intimidation. As such, my results
should not be taken to explain why people vote in undemocratic or authori-
tarian regimes. This study is strictly limited to functioning democracies in
which elections are widely regarded as free and fair.
Fourth, our inability to rely on a comprehensive survey incorporating iden-
tical question wordings across all the regions means that there is a chance
that some of the differences or results obtained across the different regions
might simply be artefactual rather than substantive. As far as possible, how-
ever, I have tried to identify and use functionally equivalent variables where
this has been a problem. Finally, because of the need for theory testing (i.e. to
see if European-generated theories and or models of voting behaviour can be
178 Conclusion
usefully exported to other contexts) I have implemented a very strict criteria
on which countries are included or excluded in this book. This has limited
case selection to just a handful of countries across the different geo-political
regions (e.g. eight in sub-Sharan Africa, nine in Latin America, three in East
Asia, and six in Eastern Europe). Thus, it should be obvious that all these
conclusions can be no more than preliminary, waiting for replication of our
study on a larger number of countries and a longer time period. Nevertheless,
the range of cases on which this book draws is considerable enough for us to
be confident of the broad thrust of the findings.
Regardless of these limitations, it is important to ask what implications
the assessment of electoral participation in newly consolidated democracies
carries for the overall understanding of political behaviour more generally.
First, the results obtained in this book point to the fact that the determinants
of voting cannot be predicted or explained by a single theoretical model or
by a single set of variables. In fact, across the four regions, evidence suggests
that various individual and country-level factors play a role in influencing
citizen’s propensity to turnout, albeit to varying degrees of significance. My
analysis, therefore, illustrates the importance of providing some theoretical
breadth when considering what drives electoral participation, while avoiding
the temptation of a “kitchen sink” approach.
Second, some of the results have practical implications for the conduct of
future elections in these regions. In an age of widespread declining turnout,
it would seem that countries could seek to do a number of things that would
weaken or reverse such a trend. Thus, the pursuit of greater socioeconomic
equality would logically seem to be a factor that could have such an effect (at
least in each of the regions except sub-Saharan Africa), and providing wider
access to higher levels of education should have a similar effect in each region.
An active civil society in which as many citizens as possible are engaged in
organisations such as trade unions and churches is another factor that can
be expected to keep electoral participation at a maximum. Institutional
arrangements such as the concurrent staging of elections would help to boost
turnout and reduce the financial costs of running elections separately. Finally,
it is incumbent on politicians everywhere to avoid actions and behaviour that
might undermine trust in them and the wider political system. None of these
recommendations is necessarily achieved easily, of course –but they do seem
to be the logical implications of the research findings uncovered in this study.
Studying electoral participation in new democracies and

beyond: some directions for future research
In light of our findings, what can be suggested about the future direction of
research on political and electoral participation? First, on a methodological
level, in our final models (i.e. Model VI) we observed a significant improve-
ment in the estimates of some individual-level factors when the contextual
level factors were added, which serves to reemphasise the need for research to
Conclusion 179
incorporate both individual-level and contextual variables in theoretical and
methodological strategies when studying political behaviour. Second, while
I have covered a good deal of ground in attempting to explain electoral par-
ticipation across new democracies, there is scope to extend this approach as
more countries consolidate their democracies so as to develop an even more
truly general explanation of turnout at the individual level. Third, the con-
firmation in this study across all four regions that age is an important factor
for voter turnout carries implications for future research and possible polit-
ical action. The book adds further confirmation of the world-wide tendency
of young people to abstain from voting. Since young people are generally
considered “future democrats”, this is a worrying finding, especially in light
of Africa’s current youth bulge and the prospects of this increasing in the
next 20–30 years which therefore points to the continuing and particular need
for research that aims to understand the reasons for political disengagement
among young citizens and what steps might be taken to counteract it. (see
Holbein and Hillygus, 2020).
Fourth, as previously mentioned above, one of the limitations of this book
is that we were unable to make a direct comparison between some of the
models we ran in the various regions. In part, this is because some variables
exist for some countries/region, but not in others; but also, even when we had
similar variables, the exact question wordings are sometimes a little different.
This being said, as far as future direction of research is concerned, this book
shows there is a need to work with a comprehensive set of survey data that
includes a multitude of countries from around the globe in order to examine
comparatively what affects people’s decision of whether to vote or not to vote.
The utility of such data would be that the survey questions and wordings will
be uniform across the regions. The Comparative Survey for Electoral System
(i.e. CSES) is already moving in this direction, but there is a need for it to
incorporate more countries from the global South. Also, The World Values
Survey does incorporate a wide range of countries, including both established
and emerging democracies, from across the globe; however, the question
about electoral participation is not ideal since it does not ask directly whether
respondents actually voted in recent national elections, but rather asks “When
elections take place, do you vote always, usually or never?”
Fifth, one avenue for future research will be the re-examination of some
of the classic accounts of the role of gender in voting. As observed in the
literature, the traditional assertion is that men tend to participate more than
women, but our findings indicate this is simply not the case. Rather, the gender
gap is fast closing-up in both established and emerging democracies. Is this
because of the growing inclination of younger women to participate or is it a
cross-generational phenomenon?
Sixth, since the findings of this book have categorically shown that an
individual’s decision to vote is influenced by both personal demographic and
national-level contextual characteristics, one agenda for future research dealing
with electoral participation at the individual level will be to consider these two
180 Conclusion
levels not just through regular multivariate modelling but with multilevel level
modelling as the number of cases increases. Again, this depends on the avail-
ability of appropriate data, of course. Finally, and relatedly, future research
might make better use of localised contextual data. In this book, I have relied
heavily on what might be thought of as the Michigan as opposed to Columbia
school of thought in explaining electoral participation at the individual level.
In relying on the Michigan school, our explanation for people’s decision of
whether or not to vote has been determined by the degree to which voters
possess different characteristics and resources such as age, gender, or different
kinds of socioeconomic resources such as education, income, or occupational
status, as well as cultural attitudinal factors such as political interest, efficacy,
and political trust. In relying on the Michigan school that has its roots on
the works of Campbell et al. (1960), I have followed in the tradition of using
survey data of random samples of individuals in national context. However,
the weakness in using these survey data is that individuals are usually assumed
to be independent of each other in the statistical analyses that are then carried
out on the samples.
However, an alternative for future research would be focus more on the
Columbia school’s approach, which is based on the view that politics and
more specifically people’s decision of whether to or not vote can be better
understood as a social phenomenon (i.e. people’s decisions to vote are always
taken in a social context). Thus, a good avenue for future research would be to
examine social contexts such as family, peer groups, school, residential area,
and workplace to see how the interactions that occur within these places influ-
ence people’s decisions to vote in elections. Important questions that could be
posed then are: how do parents/other family members influence an individual’s
electoral participation? Or do friends affect how one think and acts in the
political arena? To answer these questions, the standard approach of using
survey data needs to be complemented using population-wide register data.
Appendix 1: Detailed description of datasets
Table A1 Afrobarometer: sub-Saharan African countries
Country Round 1 Round 2 Round 3 Round 4 Round 5 Round

1999–2001 2002–2003 2005–2006 2008–2009 2011–2013 2014–2016
Benin – – × × × ×
Botswana × × × × × ×
Cape Verde × × × × ×
Ghana × × × × × ×
Sao Tome and × × × × × ×
Principe
Mauritius – – – – × ×
Namibia × × × × × ×
South Africa × × × × × ×
Source: Afrobarometer survey data (www.afrobarometer.org/data).
Table A2 AmericasBarometer: Latin American countries
Round 1 Round 2 Round 3 Round 4 Round 5 Round 6 Round 7

2004 2006 2008 2010 2012 2014 2016/2017
Argentina × × × × × × ×
Brazil × × × × × × ×
Chile × × × × × × ×
Uruguay × × × × × × ×
Costa Rica × × × × × × ×
El Salvador × × × × × × ×
Guyana × × × × × × ×
Panama × × × × × × ×
Source: AmericasBarometer data (www.vanderbilt.edu/lapop/data-access.php).
Table A3 AsianBarometer: East Asian countries
Country Wave 1 Wave 2 Wave 3 Wave 4

2001–2003 2005–2008 2010–2012 2014–2016
Taiwan × × × ×
Mongolia × × × ×
South Korea × × × ×
Source: AsianBarometer survey data (www.asianbarometer.org/data).

newgenrtpdf
182
Appendices
Table A4 European Social Survey: post-communist European countries
Country Round 1 Round 2 Round 3 Round 4 Round 5 Round 6 Round 7 Round 8

2002 2014 2006 2008 2010 2012 2014 2016
Czech Rep. × × – × × × × ×
Estonia – × × × × × ×
Hungary × × × × × × × ×
Lithuania – × × × × × ×
Poland × × × × × × ×
Slovenia × × × × × × ×
Source: European Social Survey (www.europeansocialsurvey.org/data/). Round 9 has already been initiated but data not yet released.
newgenrtpdf
Appendix 2: Questions wording from the four survey data
Table B1 Questions wording for the dependent and independent variables
Africa Latin America East Asia Central and Eastern

Afrobaometer, AmericasBarometer AsianBarometer European Countries
Round 6, 2016. Round 6, 2014 Wave 4, 2016 European Social Survey,
Round 8, 2016
Dependent variable
Voting Understanding that some people Did you vote in the last In talking to people about Some people do not
were unable to vote in the most presidential elections elections, we often find vote nowadays for one
recent national election in of (year of last that a lot of people reason or another.
[20xx], which of the following presidential elections)? were not able to vote Did you vote in the
statements is true for you? 1 = voted, 2 = did not because they were away last national election?
0 = You were not registered vote, 888888 = Don’t from home, they were Here the respondents
to vote, 1 = You voted in the know, 988888 = No sick or they just did not were invited to answer
elections, 2 = You decided not answer have time. How about with a simple Yes or
to vote, 3 = You could not find you? Did you vote in No response.
the polling station, 4 = You the most recent national
were prevented from voting, election, parliamentary
5 = You did not have time to or presidential? 1 = Yes,
vote, 6 = You did not vote 2 = No, 0 = not yet
because you could not find your eligible to vote, 7 = do
name in the voters’ register, not understand the
7 = Did not vote for some other question, 8 = can’t choose,
Appendices 183
reason, 8 = You were too young 9 = decline to answer.
to vote, 9 = Don’t know/can’t
remember.
(continued)
Table B1 Cont.
184
Appendices
Round 8, 2016
Socio-demographics variables
Age How old are you? 18–105, 998– How old are you? 18–105 Actual age ranging from How old are you? 18–87,
999, -1 17–108
Gender Respondent’s gender: 1 = male, Respondent’s Respondent’s Respondent’s
2 = female gender: 1 = male, gender: 1 = male, sex: 1 = male,
2 = female 2 = female 2 = female
Religion People practise their religion What is your religion? About how often do you How often do you attend
in different ways. Aside from (1) Catholic, practise religious services religious services
weddings and funerals, how (2) Protestant or or rituals these days? apart from special
often do you personally engage Protestant non- Several times a day [01], occasions?
in religious practices like prayer, Evangelical, (3) Non- Once a day [02], Several Every day [01], More
reading a religious book, or Christian Eastern times a week [03], Once a than once a week [02],
attending a religious service or religion (Islam, week [04] Once a month Once a week [03], At
a meeting of a religious group? Buddhism, Hinduism, [05], Only during festivals least once a month
Religious practice: 0 = Never, Taoism, Confucianism, (or several times a year) [04], Only on special
1 = A few times a year, etc.). [06], Less often [07] and holy days [05], Less
2 = About once a month, Practically never [08] often [06], Never [07].
3 = About once a week, 4 = A
few times a week, 5 = About
once a day, 6 = More than
once a day, 7 = Respondent has
no religion, 9 = Don’t know
[DNR], -1 = Missing.
newgenrtpdf
Ethnicity/race Let us suppose that you had Do you consider yourselfWhat is your racial or ethnic Do you belong to a
to choose between being white, mestizo, background? 301 = Korea, minority ethnic group
a (Ghanaian/Kenyan/etc.) indigenous, black, 302 = Korea Chinese, in [country]? Yes
and being a (respondent’s mulatto, or of another 303 = Vietnamese, 1, No 2
identity group), which of race? [If respondent 501 = Khalkh, (Refusal) 7 (Don’t know)
these two groups do you feel says Afro-country, 502 = Kazak, 8.
most strongly attached to? mark (4) Black] 503 = others,
The response options range (1) White, (2) Mestizo, 701 = Hakka, 702 = Min-
across: 1 = Ethnic identity only, (3) Indigenous, nan, 703 = Mainlander,
2 = ethnic identity more than (4) Black 704 = Aboriginal,
national, 3 = national/ethnic 705 = Others,
identity equal, 4 = national 706 = Vietnamese,
identity more than ethnic and 707 = Indonesian,
5 = national identity only. 708 = Overseas Chinese,
709 = Filipino.
Place of Urban or rural primary sampling Which of the following Size of place: (1) National Which place best
residence unit: 1 = Urban, 2 = Rural. levels within the capital (metropolitan area) describes the area
country the respondent (2) Large city, (3) Medium where you live?
lives? 1 = Rural, city, (4) Small city, A big city [1]‌, The
2 = Urban. (5) Rural area. suburbs or outskirts
of a big city [2],
A town or a small city
[3], A country village
[4], A farm or home in
the countryside [5].
Education What is your highest level of How many years of What is the highest level of What is the highest
Appendices 185
education? 0 = No formal schooling have you education that individuals level of education
schooling, 1 = Informal completed? The have completed or you have successfully
schooling only (including response ranges achieved: 1 = No formal completed? Ranges
Koranic schooling), 2 = Some from: primary, education, 2 = Incomplete from: Not completed
primary schooling, 3 = Primary secondary, university, primary/elementary, ISCED level 1 = 000, I
school completed, post-secondary not 3 = Complete
university).
(continued)
Table B1 Cont.
186
Appendices
Round 8, 2016
4 = Intermediate school or Some primary/elementary, ISCED, completed
secondary school /high school, 4 = Incomplete secondary/ primary education
5 = Secondary school /high high school: technical = 113 to ISCED
school completed, 6 = Post- and vocational type, 6, Doctoral
secondary qualifications, 5 = Complete secondary/ degree = 800.
other than university e.g. a high school: technical and
diploma or degree from a vocational, 6 = incomplete
polytechnic or college, 7 = Some secondary/high school,
university, 8 = University 7 = Complete secondary/
completed, 9 = Post-graduate, high school, 8 = Some
99 = Don’t know [Do not university, 9 = University
read], 98 = Refused to answer, education completed,
-1 = Missing 10 = Post-graduate degree.
Party Do you feel close to any Do you currently identify Among the political parties Is there a particular
identification particular political party? with a political party? listed here, which party political party you feel
0 = No (not close to any (1) Yes, (2) No. if any do you feel closest closer to than all the
party), 1 = Yes, (feels close to a (888888) Don’t know, to? Party A = 01, Party other parties? 1 = Yes
party), 8 = Refused to answer, (988888) No answer. B = 02, Party C = 03 etc. 1, 2 = No 2, (Refusal)
9 = Don’t know, -1 = Missing. Don’t feel close to any 7, (Don’t know) 8.
party = 99.
newgenrtpdf
Associational Let’s turn to your role in the I am going to read you Question asks whether Are you or have you
membership community. Now I am going a list of groups and respondents are ever been a member
to read out a list of groups organizations. Please members of any formal of a trade union or
that people join or attend. tell me if you attend organisations: Political similar organisation?
For each one, could you tell meetings of these parties, Residential and If Yes, is that currently
me whether you are an official organisations at least community associations, or previously? Yes,
leader, an active member, an once a week, once or Religious groups, Sports/ currently [1]‌, Yes,
inactive member, or not a twice a month, once recreational clubs, previously [2], No [3].
member: Member of religious or twice a year, or Culture organisations,
group, Member of voluntary never. Meetings of any Charities, Public interest
association or community religious organization? groups, Labour unions,
group: 0 = Not a member, Meetings of a Farmer unions or
1 = Inactive member, 2 = Active parents’ association agricultural associations,
member, 3 = Official leader, at school? Meetings Professional organisations,
9 = Don’t know, 98 = Refused of a community, Business associations,
to answer, -1 = Missing. improvement Parent–Teacher
committee or Associations or PTA
association? Meetings producer cooperatives,
of a political Consumer cooperatives,
party or political organisations and Student
organisation? Meetings associations etc. 1 = Yes
of associations or and 2 = No, 8 = Can’t
groups of women or choose, 9 = Decline to
homemakers? answer.
Clientelism In your opinion, how often do And thinking about No question in survey. No question in survey.
the following things occur in the last presidential
Appendices 187
this country’s elections: Voters elections, did someone
are bribed: 0 = Never, offer you something,
1 = Sometimes, 2 = Often, like a favour, gift, or
3 = Always 9 = Don’t know, any other benefit in
98 = Refused to answer, return for your vote or
-1 = Missing. support? (1) Yes, (2) No,
(88) DK, (98) DA.
(continued)
newgenrtpdf
Table B1 Cont.
188
Appendices
Round 8, 2016
Political interest How interested would you say you How much interest do you How interested would you How interested would
are in public affairs? 0 = Not have in politics: a lot, say you are in politics? you say you are
at all interested, 1 = Not very some, little or none? Respondents had to in politics: very
interested, 2 = Somewhat (1) A lot, (2) Some, decide between four interested (1), quite
interested, 3 = Very interested, (3) Little, (4) None, categories: 1 = Not at all interested (2), hardly
9 = Don’t know, 98 = Refused (888888) Don’t know interested, 2 = Not very interested, (3) or, not
to answer, -1 = Missing. [DON’T READ], interested, 3 = Somewhat at all interested (4),
(988888) No answer interested and 4 = Very (Refusal) 7, (Don’t
[DON’T READ]. interested, 8 = Can’t choose, know) 8.
9 = Decline to answer.
Political trust How much do you trust each of the To what extent do you I am going to name a number Please tell me on a
following, or haven’t you heard trust the following: the of institutions. For each, score of 0–10 how
enough about them to say: Trust National Congress? please tell me how much much you personally
president, Trust parliament/ National Police? trust do you have in them? trust each of the
national assembly, Trust national Political parties? The president, courts, institutions I read out.
electoral commission, Trust the President/Prime national government, 0 means you do not
ruling party, Trust opposition Minister? 1 to 7 scale, political parties, parliament, trust an institution
political parties, Trust police, where 1 is “Not at all” civil service, military and at all, and 10 means
Trust army, Trust courts of and 7 is “A lot” police, local government, you have complete
law: 0 = Not at all, 1 = Just a and the election trust: Parliament,
little, 2 = Somewhat, 3 = A lot, commission. 1 = A great legal system, police,
9 = Don’t know/Haven’t heard deal of trust, 2 = Quite a politician, political
enough, 98 = Refused to answer, lot of trust, 3 = Not very parties.
-1 = Missing. much trust, 4 = None at all,
7 = Do not understand the
question, 8 = Can’t choose.
External efficacy Think about how elections work Those who govern this How well do you think the And how much would
in practice in this country. country are interested government responds to you say that the
How well do elections: Enable in what people like you what people want? Very political system in
voters to remove from office think. How much do responsive 1, Largely [country] allows
leaders who do not do what the you agree or disagree responsive 2, Not very people like you to
people want. 0 = Not at all well, with this statement? responsive 3, Not have an influence on
1 = Not very well, 2 = Well, You feel that you responsive at all 4. politics? Not at all 1,
3 = Very well, 9 = Don’t understand the most Very little 2, Some 3,
know, 98 = Refused to answer, important political A lot 4, A great deal 5.
-1 = Missing. issues of this country.
How much do you
agree or disagree with
this statement? Where
1 means “strongly
disagree” and 7 means
“strongly agree”.
A number in between
1 and 7 represents an
intermediate score.
Internal efficacy Think about how elections work Sometimes politics and And how confident
in practice in this country. How government seem so are you in your own
well do elections ensure voters’ complicated that a person ability to participate
views are reflected: 0 = Not like me can’t really in politics? 1 = Not
at all well, 1 = Not very understand what is going at all confident,
well, 2 = Well, 3 = Very well, on: 1 = Strongly agree, 2 = A little confident,
9 = Don’t know, 98 = Refused 2 = Agree, 3 = Disagree, 3 = Quite confident,
Appendices 189
to answer. 4 = Strongly disagree, 4 = Very confident,
7 = Don’t understand 5 = Completely
the question, 8 = Can’t confident,
choose, 9 = Decline to 7 = (Refusal),
answer. 8 = (Don’t know).
(continued)
newgenrtpdf
Table B1 Cont.
190
Appendices
Round 8, 2016
Alternative variables
Country Looking back, how do you rate Do you think that the How would you describe the How satisfied with
economy economic conditions in this country’s current change in the economic present state of
country compared to 12 months economic situation is condition of your country economy in country?
ago? 1 = Much worse, better than, the same over the last year? Responses ranges from
2 = Worse, 3 = Same, 4 = Better, as or worse than it 1 = Much better, 2 = A 0–10. 0 = Extremely
5 = Much better, 9 = Don’t was 12 months ago? little better, 3 = About dissatisfied and
know, 98 = Refused to answer, (1) Better, (2) Same, the same, 4 = A little 10 = Extremely
-1 = Missing. (3) Worse, (888888) worse, 5 = Much worse, satisfied.
Don’t know, (988888) 7 = Do not understand the
No answer. question, 8 = Can’t choose,
9 = Decline to answer.
Corruption How many of the following Is corruption among Have you or anyone you know Based on the
people do you think are public officials personally witnessed an Transparency
involved in corruption, or very common? act of corruption or bribe- International
haven’t you heard enough about 1 = Very common, taking by a politician or Corruption
them to say: (a) Office of the 2 = Common, government official in the Perceptions Index.
presidency, (b) Members of 3 = Uncommon, past year? 1 = Personally The index measured
parliament, (c) Government 4 = Very common. witnessed, 2 = Told about the perceived level
officials, (d) Local government it by a family member of public/political
councillors, (e) Police, who personally witnessed, corruption on a scale
(f) Judges and magistrates, 3 = Told about it by a of 0 (highly corrupt)
and (g) Business executives. friend who personally to 100 (very clean).
The response options witnessed, 4 = Personally
were: 0 = None, 1 = Some of never witnessed, 5 = No
them, 2 = Most of them, 3 = All one I know has personally
of them. witnessed.
Authoritarianism What the overall erosion of a What the overall What the overall erosion of What the overall
country’s democracy on the erosion of a country’s a country’s democracy erosion of a country’s
same five-point scale identified democracy on the on the same five-point democracy on the
above: –0: No backsliding, same five-point scale scale identified above: – same five-point scale
and weak threat of future identified above: – 0: No backsliding, and identified above: –
backsliding. –1: There are 0: No backsliding, weak threat of future 0: No backsliding, and
precursors to backsliding, e.g. and weak threat of backsliding. –1: There are weak threat of future
the rise of extremist parties, future backsliding. – precursors to backsliding, backsliding. – 1: There
but erosion of democratic 1: There are precursors e.g. the rise of extremist are precursors to
institutions has not yet taken to backsliding, e.g. parties, but erosion of backsliding, e.g. the
place. –2: There is weak erosion the rise of extremist democratic institutions rise of extremist
of democratic institutions, parties, but erosion has not yet taken place. – parties, but erosion
perhaps the institutions being of democratic 2: There is weak erosion of democratic
eroded are not critical for the institutions has not of democratic institutions, institutions has not
functioning of democracy. – yet taken place. – perhaps the institutions yet taken place. –
3: There is moderate erosion 2: There is weak being eroded are not 2: There is weak
of democratic institutions. – erosion of democratic critical for the functioning erosion of democratic
4: There is severe erosion of institutions, perhaps of democracy. –3: There institutions, perhaps
democratic institutions; it is the institutions being is moderate erosion of the institutions being
unclear whether democracy will eroded are not critical democratic institutions. – eroded are not critical
recover. for the functioning 4: There is severe erosion for the functioning
of democracy. – of democratic institutions; of democracy. –
3: There is moderate it is unclear whether 3: There is moderate
erosion of democratic democracy will recover. erosion of democratic
institutions. – 4: There institutions. – 4: There
is severe erosion of is severe erosion
Appendices 191
democratic institutions; of democratic
it is unclear whether institutions; it is
democracy will recover. unclear whether
democracy will
recover.
(continued)
newgenrtpdf
Table B1 Cont.
192
Appendices
Round 8, 2016
Electoral system Disproportionality (least square Disproportionality (least Disproportionality (least Disproportionality (least
index) the calculation is square index) the square index) the square index) the
either based on Gallagher calculation is either calculation is either calculation is either
(2019) Election indices or based on Gallagher based on Gallagher based on Gallagher
self-calculated. (2019) Election indices (2019) Election indices or (2019) Election indices
or self-calculated. self-calculated. or self-calculated.
Concurrent Dummy coding: 1 = Concurrent Dummy coding: 1 = Dummy coding: 1 = Dummy coding: 1 =
elections elections, 0 = Otherwise. Concurrent elections, Concurrent elections, 0 = Concurrent elections,
0 = Otherwise. Otherwise. 0 = Otherwise.
Closeness of I measure closeness of election I measure closeness of I measure closeness of I measure closeness of
elections as the margin of victory election as the margin election as the margin election as the margin
for the winning candidate of victory for the of victory for the of victory for the
or over the runner-up in winning candidate winning candidate or winning candidate
presidential elections. While for or over the runner- over the runner-up in or over the runner-
parliamentary democracies, we up in presidential presidential elections. up in presidential
measure closeness of election elections. While While for parliamentary elections. While
as the difference in seat shares for parliamentary democracies, we measure for parliamentary
between the top two parties. democracies, we closeness of election as the democracies, we
That said, I coded the variable measure closeness difference in seat shares measure closeness
in such a way that a winning of election as the between the top two of election as the
margin of less 6% = 1, and a difference in seat shares parties. That said, I coded difference in seat
margin greater than 6% = 0. between the top two the variable in such a way shares between the top
that a winning margin of two parties. That said,
less 6% = 1, and a margin I coded the variable
greater than 6% = 0.
parties. That said, in such a way that a
I coded the variable winning margin of less
in such a way that a 6% = 1, and a margin
winning margin of less greater than 6% = 0.
6% = 1, and a margin
greater than 6% = 0.
Compulsory 0 = No compulsory election/ 0 = No compulsory 0 = No compulsory election/ 0 = No compulsory
voting compulsory election without election/compulsory compulsory election election/compulsory
enforcement, 1 = Compulsory election without without enforcement, election without
elections with enforcement. enforcement, 1 = Compulsory elections enforcement,
1 = Compulsory with enforcement. 1 = Compulsory
elections with elections with
enforcement. enforcement.
Appendices 193
newgenrtpdf
Table B2 Variable descriptions and coding schemes across the four regions and surveys
194
Variables/models Africa Latin America East Asia Post-communist European
Appendices
Afrobarometer, AmericansBarometer AsianBarometer countries
Round 6, 2016 Round 6, 2014 Wave 4, 2016 European Social Survey,
Round 8, 2016
Dependent variable
Electoral Dummy 1 = Yes, voted, 0 = No 1 = Yes, voted, 0 = No 1 = Yes, voted, 0 = No 1 = Yes, voted, 0 = No did
participation/ did not vote. did not vote. did not vote. not vote.
voting Those who were too Those who were too Those who were too Those who were too young
young to vote, young to vote, blank young to vote, blank to vote, blank votes, and
blank votes, and votes, and abstention votes, and abstention abstention responses are
abstention responses responses are coded as responses are coded as coded as missing values.
are coded as missing missing values. missing values.
values.
Months since the Scale The number of The number of months The number of months The number of months
last election months from the from the last national from the last national from the last national
last national election election date to the election date to the election date to the
date to the month month of the survey month of the survey month of the survey
of the survey date. date. date. date.
Age Scale Age of respondents in Age of respondents in Age of respondents in Age of respondents in
years. years. years. years.
Gender Dummy 1 = Male and 1 = Male and 0 = Female. 1 = Male and 0 = Female. 1 = Male and 0 = Female.
0 = Female.
Religion Scale/ Respondents indicate What is your Respondents indicate Respondents indicate
dummy how often they religion: 1 = Catholic, how often they attend how often they
attend religious 0 = Other. religious services or attend religious
practices: 0 = Never/ rituals: 0 = Never/less services: 0 = Never,
few times a year, often, 2 = Once a year/ 2 = less often, 3 = Only
2 = About once a several times a year, on special holidays,
month, 3 = Once 3 = Once a month/ 4 = Once a month,
a week/few times a several times a month, 5 = Once a week/
week, 4 = Once a 4 = Once a week/ everyday
day, 5 = More than several times a week,
once a day. 5 = Once a day/several
times a day.
Ethnicity/race Scale/ Salience of Others ethnic groups = 0, Nominal ethnic Nominal ethnic
dummy ethnicity: 1 = Only Blacks/Indigenous identity: 0 = minority identity: 0 = Minority
ethnic ID, people = 0.5, Mestizos/ ethnic group, ethnic group,
0.5 = Equal ethnic/ Mulattos = 1, 1 = majority ethnic 1 = Majority ethnic
national ID, Whites = 2. group. group.
0 = Only national
ID.
Place of residence Dummy 0 = Respondents living 1 = Respondents living 0 = Respondents living 0 = Respondents living
in medium /large in medium /large in medium /large in medium /large
cities or urban areas, cities or urban areas, cities or urban areas, cities or urban areas,
1 = Respondents 0 = Respondents living 1 = Respondents living 1 = respondents living
living small/rural small/rural areas. small/rural areas. small/rural areas.
areas.
Level of education 3-point 0 = Primary education, 0 = Primary education, 0 = Primary education, 0 = Primary education,
Appendices 195
scale 0.5 = Secondary 0.5 = Secondary 0.5 = Secondary 0.5 = Secondary
education, education, 1 = Higher education, 1 = Higher education, 1 = Higher
1 = Higher education. education. education.
education.
(continued)
newgenrtpdf
Table B2 Cont.
196
Appendices
Round 8, 2016
Party Dummy 1 = Closed parties ties, 1 = Closed parties ties, 1 = Closed parties ties, 1 = Closed parties ties,
identification 0 = No party ties. 0 = No party ties 0 = No party ties. 0 = No party ties
Associational Scale/ 0 = Not a member, Never = 0, 0.5 = Once/ 0 = Not a member, 0 = Not a member,
membership dummy 0.5 = Inactive twice a year, 1 = Once/ 1 = Member. 0.5 = Member
member, 1 = Active twice a month, previously, 1 = Member
member/official 2 = Once a week. currently.
leader.
Clientelism Dummy Respondents identify Respondents identify as N/A N/A
as having receive having receive gift for
gift for their their vote: 1 = Receive
vote: 1 = Receive a gift, 0 = Did not
a gift, 0 = Did not receive any gift/present.
receive any gift/
present).
Political interest Scale 0 = Not interested, 0 = None/no interest, 0 = Not interested, 0 = Not at all interested,
0.5 = Not very 0.5 = Little interest, 0.5 = Not very 0.5 = Hardly interested,
interested, 1 = some interest, interested, 1 = Quite interested,
1 = Somewhat 2 = A lot of interest. 1 = Somewhat 2 = Very interested.
interested, 2 = Very interested, 2 = Very
interested. interested.
Political trust Scale 0 = No trust, Respondents say to 0 = None at all, A 10-point scale that
0.5 = Little trust, what extent they 0.5 = Not very much measured respondents
1 = Some trust, trust political/other trust, 1 = Quite a lot of trust in society (0 to 10).
2 = A lot of trust. institutions (1 to 7). trust, 2 = A great deal
of trust.
External efficacy Scale Respondent indicates Respondents indicate Government leaders Respondent indicates
how well elections whether those who implement what voters how much the political
enable voters to govern their country want (0 = Not at all system in country allows
removed leaders are interested in what responsive, 0.5 = Not people to influence
from office (0 = Not people like you think very responsive, politics: 0 = Not at
at all well, 0.5 = Not (1 to 7). 1 = Largely responsive, all, 0.5 = Very little,
very well, 1 = Well, 2 = Very responsive). 1 = Some, 2 = A lot,
2 = Very well). 3 = A great deal.
Internal efficacy Scale Respondents indicate Respondents’ subjective Respondents see politics/ Respondents indicate
how well the feel assessment of their government as how confident they
elections allow understanding of complicated that a are in their ability to
voters view to be the most important person like me can’t participate in politics?
reflected. (0 = Not political issues (i.e. really understand 0 = Not at all confident,
at all well, 0.5 = Not measured on a 7-point what is going on 0.5 = A little confident,
very well, 1 = Well, scale). (0 = Strongly 1 = Quite confident,
2 = Very well.) agree, 0.5 = Agree, 2 = Very confident,
1 = disagree, 3 = Completely
2 = Strongly disagree). confident.
Alternative explanatory variables
Country economy Scale Sociotropic Sociotropic retrospective Sociotropic retrospective Sociotropic retrospective
retrospective economic perception economic perception economic perception
economic perception [0 = worse, 0.5 = Same, [0 = Worse, 0.5 = Same, [0 = Worse, 0.5 = Same,
[0 = worse, 1 = Better]. 1 = Better] 1 = Better].
0.5 = Same,
1 = Better]
Appendices 197
(continued)
newgenrtpdf
Table B2 Cont.
198
Appendices
Round 8, 2016
Corruption Scale/ Respondent indicates Respondents indicate Respondents indicate Transparency Corruption
perception dummy how many people if corruption among if they or anyone Index. The index
they think are public officials is have personally measured the perceived
involved in very common? witnessed an act of level of public/political
corruption, or 1 = Very common, corruption or bribe- corruption on a scale of
haven’t heard 0 = Uncommon. taking by a politician 0 (highly corrupt) to 100
enough about them or government (very clean). Scores are
to say: 0 = None, official in the past rescaled from 0 –1.
1 = Some of them, year. 1 = Witness
2 = Most of them, corruption, 0 = Never
3 = All of them. witness corruption.
Authoritarianism Dummy What is the What is the overall What is the overall What is the overall
overall erosion erosion of a erosion of a erosion of a country’s
of a country’s country’s democracy? country’s democracy? democracy? 1 = Country
democracy? 1 = Country 1 = Country experiences democratic
1 = Country experiences democratic experiences democratic erosion, 0 = Country
experiences erosion, 0 = Country erosion, 0 = Country never experiences
democratic erosion, never experiences never experiences democratic erosion.
0 = Country democratic erosion. democratic erosion.
never experiences
democratic erosion.
Electoral system Scale Least squares index of Least squares index of Least squares index of Least squares index of
disproportionality disproportionality disproportionality disproportionality
(0 = Perfect (0 = Perfect (0 = Perfect (0 = Perfect
proportionality proportionality proportionality proportionality
to 100 maximum to 100 maximum to 100 maximum to 100 maximum
disproportionality) disproportionality) disproportionality) disproportionality)
Concurrent Dummy 1 = Concurrent 1 = Concurrent elections, 1 = Concurrent elections, 1 = Concurrent elections,
elections elections, 0 = Otherwise 0 = Otherwise 0 = Otherwise
0 = Otherwise
Closeness of Dummy 1 = Closed/competitive 1 = Closed/competitive 1 = Closed/competitive 1 = Closed/competitive
elections elections, 0 = Not elections, 0 = Not closed. elections, 0 = Not elections, 0 = Not closed.
closed. closed.
Compulsory voting Dummy N/A 0 = No compulsory N/A 0 = No compulsory
election/compulsory election/compulsory
election without election without
enforcement, enforcement,
1 = Compulsory 1 = Compulsory
elections with elections with
enforcement. enforcement.
Appendices 199
200 Appendices
Appendix 3: Summary descriptive statistics of variables and
individual country –results
Table C1 Descriptive statistics for dependent and socio-demographic variables
Region/variable Mean Standard Minimum Maximum

deviation
Western Europe
Voting 0.82 0.37 0 1
Age 48.4 18.5 15 100
Gender 0.49 0.49 0 1
Religiosity 2.08 1.80 0 5
Place of residence 0.54 0.49 0 1
Ethnicity/race 0.94 0.22 0 1
Education 0.84 0.80 0 2
Sub-Saharan Africa
Voting 0.72 0.39 0 1
Age 40.4 54.2 18 105
gender 0.49 0.50 0 1
Education 0.21 0.30 0 2
Latin America
Voting 0.76 0.43 0 1
Age 41.5 16.6 16 99
gender 0.48 0.49 0 1
Education 0.55 0.38 0 2
East Asia
Voting 0.83 0.35 0 1
Age 43.5 15.4 18 93
Gender 0.48 0.49 0 1
Education 1.07 0.79 0 2
Eastern Europe
Voting 0.67 0.47 0 1
Age 47.6 18.5 15 97
gender 0.45 0.49 0 1
Education 0.68 0.78 0 2
Notes: Descriptives statistics across the different regions are drawn from the Afrobarometer,
AmericasBarometer, AsianBarometer, and the European Social Survey. The observations are all
weighted using the different surveys weights.
Appendices 201
Table C2 The impact of socio-
demographic variables on vote choice: individual
country results
Regions/countries Age Male Religion Ethnicity Rural Education
Sub-Saharan Africa
Benin 0.35** 0.46*** 0.06 -0.67*** 0.51*** -0.88**
(0.16) (0.18) (0.06) (0.24) (0.18) (0.39)
Botswana 1.01*** -0.50** 0.15*** -0.28 0.39** -0.39
(0.19) (0.20) (0.05) (0.33) (0.22) (0.35)
Cape Verde 0.48** -0.18 0.06 -0.43* -0.11 0.17
(0.14) (0.16) (0.07) (0.36) (0.19) (0.25)
Ghana 0.39** -0.14 0.07** 0.22 0.23** 0.27
(0.10) (0.12) (0.04) (0.17) (0.12) (0.23)
Mauritius 0.54*** 0.26 0.03 -0.34 0.01 -0.15
(0.19) (0.20) (0.08) (0.33) (0.21) (0.30)
Namibia 0.95*** 0.10 0.05 -0.11 0.11 -0.10
(0.17) (0.18) (0.09) (0.28) (0.18) (0.30)
Sao Tome and 0.12 -0.50** 0.05 -0.41 0.50* -0.29
Principe (0.22) (0.02) (0.10) (0.47) (0.30) (0.43)
South Africa 0.64*** -0.09 0.05 -0.24* 0.27** 0.47**
(0.14) (0.11) (0.04) (0.14) (0.12) (0.26)
Latin America
Argentina 0.05*** 0.05 0.15 -0.05 0.35* 1.34***
(0.01) (0.13) (0.14) (0.12) (0.20) (0.21)
Brazil 0.08*** -0.23** 0.23 0.17 -0.24 0.93***
(0.01) (0.14) (0.14) (0.20) (0.20) (0.26)
Chile 0.05*** -0.06 0.35*** 0.06 0.76*** 1.03***
(0.01) (0.12) (0.13) (0.11) (0.22) (0.21)
Costa Rica 0.02*** -0.19 0.41*** 0.56*** 0.13 0.85***
(0.00) ((0.12) (0.12) (0.11) (0.12) (0.17)
El-Salvador 0.03*** 0.08 0.39*** -0.24** 0.01 0.74***
(0.01) (0.13) (0.13) (0.12) (0.14) (0.18)
Guyana 0.05*** -0.24** 0.16 -0.54** -0.03 0.35
(0.01) (0.12) (0.25) (0.15) (0.13) (0.22)
Panama 0.10*** -0.16 0.06 -0.03 0.39*** 0.14
(0.01) (0.13) (0.14) (0.12) (0.15) (0.22)
Peru 0.08*** -0.18 0.06 -0.02 0.08 0.65**
(0.01) (0.16) (0.19) (0.15) (0.05) (0.25)
Uruguay 0.11*** -0.39** 0.29 -0.18 -0.36 0.30
(0.01) (0.16) (0.19) (0.16) (0.34) (0.24)
East Asia
Taiwan 0.05*** -0.24* 0.16*** -0.26 0.20 0.43***
(0.01) (0.14) (0.05) (0.22) (0.19) (0.09)
Mongolia 0.06*** -0.08 0.14 -0.37 0.48 0.32**
(0.01) (0.22) (0.13) (0.35) (0.32) (0.016)
South Korea 0.08*** 0.13 -0.00 0.49 0.06 0.45***
(0.01) (0.16) (0.09) (0.24) (0.27) (0.15)
Czech Republic 0.03*** 0.46*** 0.13*** 0.70** 0.02 0.68***
(0.00) (0.09) (0.06) (0.30) (0.10) (0.07)
(continued)
202 Appendices
Table C2 Cont.
Regions/countries Age Male Religion Ethnicity Rural Education

Estonia 0.02*** -0.03 0.18*** 0.83*** -0.27** 0.77***
(0.00) (0.12) (0.08) (0.17) (0.12) (0.07)
Hungary 0.02*** 0.07 0.10*** 0.70** 0.08 0.38***
(0.00) (0.13) (0.04) (0.27) (0.12) (0.09)
Lithuania 0.03*** 0.42*** 0.29*** 0.03 0.20* 0.51***
(0.00) (0.10) (0.04) (0.18) (0.12) (0.06)
Poland 0.01*** 0.34*** 0.19*** 0.63 -0.46*** 0.59***
(0.00) (0.12) (0.04) (0.40) (0.13) (0.08)
Slovenia 0.03*** -0.15 0.14*** 0.10 0.07 0.65***
(0.00) (0.14) (0.04) (0.41) (0.15) (0.09)
Cell entries are multivariate logit coefficients; standard errors are in parentheses. All relationships
are Significant: *significant at 10%, **significant at 5%, ***significant at 1%. The observations
are all weighted using the different surveys weights.
Table C3 Descriptive statistics for mobilisation variables

deviation
Western Europe
Party identification 0.52 0.50 0.00 1.00
Associational membership 0.33 0.42 0.00 2.00
Sub-Saharan Africa
Clientelism 0.75 0.43 0.00 1.00
Latin America
Associational membership 0.00 1.00 -0.90 3.82
Clientelism 0.04 0.19 0.00 1.00
East Asia
Eastern Europe
Appendices 203
Table C4 The Impact of mobilisation variables on vote choice: individual-country
results
Regions/countries Party Associational Clientelism

identification membership
Sub-Saharan Africa
Benin 0.23 1.07*** (0.26) -0.63** (0.33)
(0.20)
Botswana 0.63*** 1.42*** (0.35) -0.35 (0.24)
(0.27)
Cape Verde 0.46 0.46 (0.72) -0.94 0.41
(0.52)
Ghana 0.93*** 0.08 (0.17) -0.42* (0.22)
(0.14)
Mauritius 1.35*** 0.91** (0.47) 0.67** (0.32)
(0.37)
Namibia 0.84*** 0.35 (0.30) -0.07 (0.20)
(0.24)
Sao Tome and Principe 0.89*** -0.05 (0.40) -0.07 (0.41)
(0.29)
South Africa 0.09*** 0.28* (0.16) -0.44*** (0.13)
(0.13)
Latin America
Argentina 0.68*** 0.24*** (0.11) 0.99 (0.62)
(0.18)
Brazil 0.41*** 0.16** (0.09) 0.18 (0.22)
(0.18)
Chile 1.23*** 0.25*** (0.08) -0.51 (0.48)
(0.25)
Costa Rica 1.89*** 0.15*** (0.08) 0.28 (0.46)
(0.14)
El-Salvador 0.59*** 0.45*** (0.09) -0.08 (0.33)
(0.13)
Guyana 0.87*** 0.35*** (0.08 -0.09 (0.46)
(0.24)
Panama 1.25*** 0.13*** 0.05) 0.11 (0.34)
(0.15)
Peru 0.01 0.32*** (0.12) 0.75** (0.42)
(0.24)
Uruguay 0.11 0.23** (0.12) 0.33 (0.52)
(0.18)
East Asia
Taiwan 1.36*** 0.01 (0.15) N/A
(0.17)
Mongolia 0.71*** -0.38 (0.23) N/A
(0.26)
South Korea 0.82*** -0.25 (0.19) N/A
(0.20)
(continued)
204 Appendices
Table C4 Cont.
Regions/countries Party Associational Clientelism

identification membership
Czech Republic 2.10*** 0.22 (0.20) N/A
(0.12)
Estonia 1.03*** 0.73*** (0.25) N/A
(0.13)
Hungary 2.32*** 0.39 (0.27) N/A
(0.18)
Lithuania 2.49*** 1.14*** (0.23) N/A
(0.17)
Poland 1.15*** 0.22 0.25 N/A
(0.16)
Slovenia 1.10*** 0.43** (0.21) N/A
(0.16)
Cell entries are multivariate logit coefficients; standard errors are in parentheses. All relationship
are Significant: *significant at 10%, **significant at 5%, ***significant at 1%. This model control
for the socio-demographic variables.
Appendices 205
Table C5 Descriptive statistics for political-psychological variables

deviation
Western Europe
Political interest 0.84 0.58 0 2
Political trust 0.21 0.95 -2.26 2.60
Internal efficacy 0.74 0.70 0 3
External efficacy 0.72 0.58 0 3
Sub-Saharan Africa
Political trust 0.007 0.96 -1.82 1.76
Latin America
Political trust 0.02 0.98 -1.94 2.62
East Asia
Political trust -0.53 0.79 -2.53 2.33
Eastern Europe
Political trust -0.21 0.98 -2.26 2.6
206 Appendices
Table C6 The impact of political-psychological variables on vote choice: individual
country results
Regions/countries Political Political Internal External

interest trust efficacy efficacy
Sub-Saharan Africa
Benin 0.01 0.26*** 0.18 0.01
(0.15) (0.11) (0.25) (0.23)
Botswana -0.24 0.17 0.50* -0.23
(0.18) (0.26) (0.31) (0.27)
Cape Verde 0.75 -0.12 0.62 0.00
(0.48) (0.39) (0.60) (0.49)
Ghana 0.03 -0.05 0.08 -0.04
(0.10) (0.07) (0.15) (0.14)
Mauritius 0.52** 0.14 -0.19 0.54*
(0.22) (0.17) (0.35) (0.35)
Namibia 0.96*** -0.01 0.09 -0.05
(0.17) (0.14) (0.24) (0.21)
Sao Tome and Principe 0.25 -0.07 -0.62 0.27
(0.24) (0.22) (0.39) (0.37)
South Africa 0.26*** 0.09 0.05 0.25*
(0.10) (0.08) (0.15) (0.14)
Latin America
Argentina -0.04 -0.01 -0.05 -0.06
(0.15) (0.09) (0.05) (0.05)
Brazil 0.12 -0.04 -0.006 0.01
(0.15) (0.08) (0.05 (0.04)
Chile 1.35*** 0.25*** 0.01 -0.05
(0.17) (0.09) (0.04) (0.04)
Costa Rica 1.10*** -0.21*** 0.01 -0.02
(0.16) (0.08) (0.05) (0.04)
El-Salvador 0.17 -0.01 0.09*** -0.00
(0.12) (0.08) (0.04) (0.04)
Guyana 0.27** -0.14 -0.03 0.06
(0.18) (0.10) (0.05) (0.05)
Panama 0.72*** 0.04 0.08** -0.06
(0.17) (0.09) (0.05) (0.05)
Peru 0.07 -0.12 -0.01 0.02
(020) (0.11) (0.06) (0.05)
Uruguay 0.21 -0.08 0.01 0.04
(0.17) (0.11) (0.05) (0.05)
East Asia
Taiwan 1.45*** -0.21 0.24 0.02
(0.45) (0.26) (0.50) (0.25)
Mongolia 1.10*** -0.28 0.05 -0.04
(0.37) (0.19) (0.26) (0.04)
South Korea 0.76*** 0.09 -0.01 0.24
(0.25) (0.13) (0.17) (0.32)
Czech Republic 1.38*** 0.15** 0.37** 0.26*
(0.19) (0.06) (0.16) (0.16)
Appendices 207
Table C6 Cont.
Regions/countries Political Political Internal External

interest trust efficacy efficacy
Estonia 0.25*** 0.25*** 0.23 -0.05
(0.16) (0.07) (0.15) (0.15)
Hungary 1.64*** 0.21** 0.41** -0.14
(0.22) (0.09) (0.17) (0.20)
Lithuania 1.36*** 0.21*** 0.63* 0.62***
(0.19) (0.08) (0.17) (0.17)
Poland 0.31*** 0.32*** 0.07 -0.02
(0.17) (0.08) (0.12) (0.16)
Slovenia 0.93*** 0.08 0.17 0.32
(0.21) (0.10) (0.16) (0.21)
Cell entries are multivariate logit coefficients; standard errors are in parentheses. All relationships
are Significant: *significant at 10%, **significant at 5%, ***significant at 1%. The model control
for both the socio-demographic and mobilisation variables.
Table C7 Descriptive statistics for alternative variables

deviation
Western Europe
Corruption 0.22 0.42 0.00 1.00
Authoritarianism N/A N/A N/A N/A
Economic perception 0.54 0.40 0.00 2.00
Sub-Saharan Africa
Corruption -0.09 0.96 -2.09 2.27
Authoritarianism 0.37 0.48 0.00 1.00
Latin America
Corruption 0.81 0.39 0.00 1.00
East Asia
Corruption 0.17 0.37 0.00 1.00
Eastern Europe
Corruption 0.81 0.38 0.00 1.00
208 Appendices
Table C8 Descriptive statistics for political-institutional variables

deviation
Western Europe
Disproportionality 6.93 5.39 0.99 17.34
Closeness of election 0.44 0.49 0.00 1.00
Concurrent election 0.25 0.44 0.00 1.00
Compulsory voting 0.08 0.28 0.00 1.00
Sub-Saharan Africa
Compulsory voting N/A N/A N/A N/A
Latin America
Compulsory voting 0.44 0.49 0.00 1.00
East Asia
Closeness of election N/A N/A N/A N/A
Eastern Europe
Concurrent election N/A N/A N/A N/A
References
Aarts, Kees, and Bernhard Wessels (2005). Electoral turnout. In Jacques Thomassen
(ed.), The European Voter. New York: Oxford University Press.
Abramson, P. R., and J. H. Aldrich (1982). The decline of electoral participation in
America. American Political Science Review 76, 502–521.
Achen, Christopher H. and T. Y. Wang (2017). The Taiwan Voter. Ann Arbor: The
University of Michigan Press.
Acton, Ciaran, Robert L. Miller, D. Fullerton, John Maltby, and Ciaran Acton (2009).
SPSS Statistics for Social Scientists (2nd ed.). Basingstoke: Palgrave Macmillan.
Ades, A., and R. Di Tella (1996). The causes and consequences of corruption: A review
of recent empirical contributions. Institute of Development Studies 27(2), 6–11.
Afrobarometer Data (2016). Round 6. Retrieved from www.afrobarometer.org.
Ahn, Chung-si and Jaung Hoon (2000). South Korea. In I. Marsh, J. Blondel and
T. Inoguchi, Democracy, Governance and Economic Performance. Tokyo, Japan,
New York, NY: United Nations University Press.
Ahuja, Amit, and Pradeep K. Chhibber (2012). Why the poor vote in India: “If I don’t
vote, i am dead to the state”. Studies in Comparative International Development
47(4), 389–410.
Albacete G. G. (2014). Young People’s Political Participation in Western Europe:
Continuity or Generational Change?. Basingstoke and New York: Palgrave
Macmillan.
Albritton, Robert B. (2004). Cambodia in 2003: On the road to democratic consolida-
tion. Asian Survey 44(1), 102–109.
Aldrich, J. (1993). Rational choice and turnout. American Journal of Political Science
37, 246–278.
Alexander, A. C., and H. Coffé (2018). Women’s political empowerment through
public opinion research: The citizen perspective. In A. C. Alexander, C.
Bolzendahl, and F. Jalalzai (eds), Measuring Women’s Political Empowerment
across the Globe: Strategies, Challenges and Future Research. Basingstoke: Palgrave
Macmillan.
Almond, Gabriel A., and Sidney Verba (1963). The Civic Culture: Political Attitudes
and Democracy in Five Nations. Newbury Park: Sage Publications.
Ames, Barry (2001). The Deadlock of Democracy in Brazil. Ann Arbor: University of
Michigan Press.
Ames, Barry, Andy Baker, and Lucio R. Rennó (2008). The quality of elections in
Brazil: Policy, performance, pageantry, or pork? In Peter Kingstone and Timothy
Power (eds), Democratic Brazil Revisited. Pittsburgh: University of Pittsburgh Press.
210 References
Anderson, C. J., and M. M. Singer 2008. The sensitive left and the impervious
right: multilevel models and the politics of inequality, ideology, and legitimacy in
Europe. Comparative Political Studies 41, 564–569.
Andreas Hadjar, and Michael Beck (2010). Who does not participate in elections in
Europe and why is this?. European Societies 12(4), 521–542.
Arzheimer, K., J. Evans, and M. S. Lewis-Beck (2017). The SAGE Handbook of elect-
oral behaviour (Vols. 1–2). London: Sage.
Baker, Andy, Barry Ames, and Lucio R. Rennó (2006). Social context and cam-
paign volatility in new democracies: Networks and neighborhoods in Brazil’s 2002
elections. American Journal of Political Science 50, 382–399.
Balch, George I. (1974). Multiple indicators in survey research: The concept “sense of
political efficacy”. Political Methodology 1, 1–43.
Barnes, S. H. (2006). The changing political participation of postcommunist citizens.
International Journal of Sociology 36, 2.
Barnes, T. D., and S. M. Burchard (2013). Engendering politics: The impact of
descriptive representation on women’s political engagement in sub-Saharan Africa.
Comparative Political Studies 46, 767–790.
Beauregard, Katrine (2014). Gender, political participation and electoral systems: A
cross-national analysis. European Journal of Political Research 53, 617–634.
Becker, R. (2004). Voter turnouts in East and West Germany. German Politics 13, 1–19.
Beer, S. H. (1974). Modern Political Development. New York: Random House.
Bélanger, E., and Nadeau, R. (2005). Political trust and the vote in multiparty
elections: The Canadian case. European Journal of Political Research 44, 121–146.
Bennett, Stephen E. (1986). Apathy in America, 1960–1984: Causes and Consequences
of Citizens’ Political Indifference. New York: Transnational Publisher, Inc.
Benton, Allyson (2005). Dissatisfied Democrats or retrospective voters?. Comparative
Political Studies 38, 417–442.
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.
Berman, B., D. Eyoh, and W. Kymlicka (eds) (2004). Ethnicity and Democracy in
Africa. Oxford: James Currey.
Bermeo, Nancy (2016). On democratic backsliding. Journal of Democracy 27(1), 5–19.
Bernhagen, P. and Michael Marsh (2007). Voting and protesting: Explaining citizen
participation in old and new European democracies. Democratization 14(1), 44–72.
Bhatti, Y., and K. M. Hansen (2012). The effect of generation and age on turnout
to the European parliament –how turnout will continue to decline in the future.
Electoral Studies 31(2), 269–272.
Bhavnani, R., and D. Miodownik (2009). Ethnic polarization, ethnic salience, and civil
war. Journal of Conflict Resolution 53(1), 30–49.
Birch, S. (2010). Perceptions of electoral fairness and voter turnout. Comparative
Political Studies 43, 1601–1622.
Birch, Sarah (2009). Full Participation: A Comparative Study of Compulsory Voting.
Manchester: Manchester University Press.
Birch, Sarah (2018). Elections and Electoral Participation. In Adam Fagan, and
Petr Kopecký (eds), The Routledge Handbook of East European Politics. London:
Routledge.
Black, J. (1991). Reforming the context of the voting process in Canada: Lessons
from other democracies. In H. Bakvis (ed.), Voter Turnout in Canada. Toronto:
Dundurn Press.
References 211
Blais, André (2000). To Vote or Not to Vote –The Merits and Limits of rational
Choice Theory. Pittsburgh: University of Pittsburgh Press. Choice Theory.
Pittsburgh: University of Pittsburgh Press.
Blais, A. (2006). What affects voter turnout? Annual Review of Political Science 9,
111–125.
Blais, A., and K. Aarts (2006). Electoral Systems and Turnout. Acta Politica 41, 180–
196. doi:10.1057/palgrave.ap.5500148.
Blais A. and J.-F. Daoust (2020). The Motivation to Vote. Vancouver: UBC Press.
Blais, A. and A. Dobrzynska (1998). Turnout in electoral democracies. European
Journal of Political Research 33, 239–261.
Blais, A. and K. Carty (1990). Does proportional representation foster voter turnout?
European Journal of Political Research 18, 167–181..
Blais, A., E. Gidengil, N. Nevitte, and R. Nadeau (2004). Where does turnout decline
come from? European Journal of Political Research 43(2), 221–236.
Blais, André, Elisabeth Gidengil, Richard Nadeau, and Neil Nevitte (2002). Anatomy
of a Liberal Victory: Making Sense of the Vote in the 2000 Canadian Election.
Peterview: Broadview Press.
Blaydes, l. (2011). Elections and Distributive Politics in Mubarak’s Egypt. Cambridge:
Cambridge University Press.
Bleck, J., and N. Van de Walle (2018). Electoral Politics in Africa since 1990: Continuity
in Change. Cambridge: Cambridge University Press.
Boas, Taylor, and Amy Erica Smith (2015). Religion and the Latin American voter. In
Ryan E. Carlin, Matthew M. Singer, and Elizabeth J. Zechmeister (eds), The Latin
American Voter. Ann Arbor: University Michigan Press.
Boix, C. (2011). Democracy, development and the international system. American
Political Science Review 105(4), 809–828.
Boix, Carles, Michael K. Miller, and Sebastian Rosato (2013). A complete data set
of political regimes, 1800–2007. Comparative Political Studies 46(12), 1523–1554.
Bollen, K. A. (1993). Liberal democracy: Validity and source biases in cross-national
measures. American Journal of Political Science 37, 1207–1230.
Booth, J. A., and M. A. Seligson (2008). Inequality and democracy in Latin
America: Individual and contextual effects of wealth on political participation.
In A. Krishna (ed.), Poverty, Participation and Democracy: A Global Perspective.
New York: Cambridge University Press.
Boulding, Carew, and David S. Brown. (2013). Do political parties matter for
turnout: Number of parties, electoral rules and local elections in Brazil and Bolivia.
Party Politics 21(3), 404–416.
Brader, T., J. A. Tucker, and D. Duell (2013). Which parties can lead opinion?
Experimental evidence on partisan cue taking in multiparty democracies.
Comparative Political Studies 46(11), November, 1485–1517.
Brady, H. (1999). Political participation. In J. P. Robinson, P. R. Shaver, and L. S.
Wrightsman (eds), Measures of Political Attitudes. San Diego, CA: Academic Press.
Brady, Henry E., Sidney Verba, and Kay Lehman Schlozman (1995). Beyond SES: A
resource model of political participation. The American Political Science Review
89(2), 271–294.
Bratton, M. (1999) Political participation in a new democracy –institutional consider-
ations from Zambia, Comparative Political Studies 32, 549–588.
Bratton, M. (1998). Second elections in Africa. Journal of Democracy 9(3), 51–66.
Bratton, M. (2003). Islam, democracy and public opinion in Africa. African Affairs
102, 493–501.
212 References
Bratton, M. (2008). Vote buying and violence in Nigerian election campaigns. Electoral
Studies, 27, 621–632.
Bratton, M. (2013). Voting and Democratic Citizenship in Africa. Boulder: Lynne
Rienner Publishers.
Bratton, M., R. Bhavnani, and C. Tse-Hsin (2012) Voting intentions in Africa: Ethnic,
economic or partisan? Commonwealth and Comparative Politics 50(1), 27–52.
Bratton, M., Y. Chu, and M. Lagos (2010) Who votes? Implications for new democra-
cies. Taiwan Journal of Democracy 6(1), 1–30.
Bratton, M., R. Mattes, and E. Gyimah-Boadi (2005). Public Opinion, Democracy and
Market Recovery in Africa. Cambridge: Cambridge University Press.
Braungart, R. G., and Braungart, M. M. (1986). Life course and generational politics.
Annual Review of Sociology 12, 205–231.
Brigevich, Anna (2008). The Post- Cold War Wave of Democratization: Regime
Transitions in Sub-Saharan Africa and Postcommunist States Compared. A thesis
submitted to the faculty of the University of North Carolina at Chapel Hill in
partial fulfillment of the requirements for the degree of Master of Arts in the
Department of Political Science.
Brockington, D. (2004). The paradox of proportional representation: The effect of
party systems and coalitions on individuals’ electoral participation. Political
Studies 52, 469–490.
Broughton, David and Hans-Martien ten Napel (2000). Religion and Mass Electoral
Behaviour in Europe. London: Routledge/ European Consortium of Political
Research.
Bufacchi, V. (2001). Voting, rationality and reputation. Political Studies 49, 714–729.
Bühlmann, M. and Freitag, M. (2006). Individual and contextual determinants of
electoral participation. Swiss Political Science Review 12, 13–48.
Bunce, Valerie (1999). The political economy of postsocialism. Slavic Review 58(4),
756–793.
Burden, B. C. (2009) The dynamic effects of education on voter turnout. Electoral
Studies 28(4), December, 540–549.
Burkhart, R. E., and M. Lewis Beck (1994). Comparative democracy: The economic
development thesis. American Political Science Review 88, 903–910.
Burnet, J. (2011). Women have found respect: Gender quotas, symbolic representation
and female empowerment in Rwanda. Politics & Gender 7, 303–334.
Burns, N., K. L. Schlozman, and S. Verba (2001). The Private Roots of Public
Action: Gender, Equality and Political Participation. Cambridge, MA: Harvard
University Press.
Bustikova, L., and E. Zechmeister (2017). Voting in new(er) democracies. In K.
Arzheimer, J. Evans and M. S. Lewis-Beck (eds), The SAGE Handbook of Electoral
Behaviour (Vol. 2). London: Sage.
Campbell, A. (1960). Surge and decline: A study of electoral change. Public Opinion
Quarterly 24, 397–418.
Campbell, R. (2006). Gender and Voting Behaviour in Britain. Colchester: ECPR Press.
Campbell, A., Converse, P. E., Miller, W. E., and Stokes, D. E. (1960). The American
Voter. New York, NY: Wiley.
Campbell, A., G. Gurin, W. E. Miller, and D. E. Stokes (1954). The Voter Decides.
Westport, CT: Greenwood Press.
Cancela, J. and B. Geys (2016). Explaining voter turnout: A meta-analysis of national
and subnational elections. Electoral Studies 42(2), 264–275.
References 213
Caramani, Daniele (2008). Comparative Politics. Oxford: Oxford University Press.
Caramani, Daniele (2009). Introduction to Comparative Method with Boolean algebra.
Quantitative Application in the Social Sciences. Thousand Oaks, CA: Sage
Publication.
Carlin, R. E. (2006). The decline of citizen participation in electoral politics in
postauthoritarian Chile. Democratization 13, 632–651.
Carlin, Ryan E., Matthew M. Singer, and E. J. Zechmeister (2015). The Latin American
Voter: Pursuing Representation and Accountability in Challenging Context. Ann
Arbor: The University of Michigan Press.
Carmines, E. G. and Huckfeldt, R. (1996). Political behavior. In R. E. Goodin and
H.-D. Klingemann (eds), A New Handbook of Political Science. Oxford: Oxford
University Press.
Carothers, Thomas (2002). The end of the transition paradigm. Journal of Democracy
13(1), 5–21.
Carreras, M. (2018). Why no gender gap in electoral participation? A civic duty explan-
ation. Electoral Studies 52, 36–45.
Carreras, M. and N. Castañeda-Angarita (2014). Who votes in Latin America? A test
of three theoretical perspectives. Comparative Political Studies 47(8), 1079–1104.
Carter, J. R. (1984). Early projections and voter turnout in the 1980 presidential
election. Public Choice 43, 195–202.
Castillo, J. C., D. Miranda, M. Bonhomme, C. Cox, and M. Bascopé (2014). Social
inequality and changes in students’ expected political participation in Chile.
Education, Citizenship and Social Justice 9, 140–156.
Ceka, Besir (2012). The perils of political competition: Explaining participation and
trust in political parties in Eastern Europe. Comparative Political Studies 46(12),
1610–1635.
Chang, Eric C., and Yun-han Chu (2006). Corruption and trust: Exceptionalism in
Asian democracies? Journal of Politics 68(2), 259–271.
Cheeseman, Nic (2010). African elections as vehicles for change. Journal of Democracy
21(4), 139–153.
Cheibub, José Antonio, Jennifer Gandhi, and James Raymond Vreeland (2010).
Democracy and dictatorship revisited. Public Choice 143(2–1), 67–101.
Cheng, Tun-jen, and Yun-han Chu (2018). Routledge Handbook of Democratization in
East Asia. Abingdon: Routledge.
Chesnut, R. Andrew (2003). Competitive Spirits Latin America’s New Religious
Economy. New York: Oxford University Press.
Childs, Sarah (2004). A British gender gap? Gender and political participation.
Political Quarterly 75(4), 422–424.
Chin, James Ung-Ho (2002). Malaysia: The Barisan National supremacy. In John
Fuh-Sheng Hsieh and David Newman (eds), How Asia Votes. New York: Chatham
House.
Chung-li Wu and Tzu-Ping Liu (2017). Political participation in Taiwan. In Achen,
Christopher H. and T. Y. Wang (eds), The Taiwan Voter. Ann Arbor: University of
Michigan Press.
Clarke, H. D., and A. C. Acock (1989). National elections and political attitudes.
British Journal of Political Science 19, 551–562.
Cleary, Matthew R., and Susan Carol Stokes (2006). Democracy and the Culture
of Skepticism: Political Trust in Argentina and Mexico. New York: Russell Sage
Foundation.
214 References
Coffé, H. and C. Bolzendahl (2011). Gender gaps in political participation across sub-
Saharan African nations. Social Indicators Research 102, 245–264.
Coffé, Hilde, and Tanja van der Lippe (2010). Citizenship norms in Eastern Europe.
Social Indicators Research 96(3), 479–496.
Cohen. M. L. (2017). Support for electoral democracy in the Americas. In M. J.
Cohen, N. Lupu, and E. J. Zechmeister (eds), The Political Culture of Democracy
in the Americas: A Comparative Study of Democracy and Governance. The
AmericasBarometer 2016/17. A Comparative Study of Democracy and Government.
Nashville, TN: LAPOP, Vanderbilt University.
Coleman, J. (1990). The Foundations of Social Theory. Cambridge, MA: Harvard
University Press.
Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Svend-
Erik Skaaning, Jan Teorell, David Altman, Michael Bernhard, Agnes Cornell, M.
Steven Fish, Haakon Gjerløw, Adam Glynn, Allen Hicken, Joshua Krusell, Anna
Lührmann, Kyle L. Marquardt, Kelly McMann, Valeriya Mechkova, Moa Olin,
Pamela Paxton, Daniel Pemstein, Brigitte Seim, Rachel Sigman, Jeffrey Staton,
Aksel Sundtröm, Eitan Tzelgov, Luca Uberti, Yi-ting Wang, Tore Wig, and Daniel
Ziblatt (2018). Democracy for All? V-Dem Annual Democracy Report.
Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Jan
Teorell, David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen
Hicken, Anna Luhrmann, Kyle L. Marquardt, Kelly McMann, Pamela Paxton,
Daniel Pemstein, Brigitte Seim, Rachel Sigman, Svend- Erik Skaaning, Jeffrey
Staton, Steven Wilson, Agnes Cornell, Nazifa Alizada, Lisa Gastaldi, Haakon
Gjerløw, Garry Hindle, Nina Ilchenko, Laura Maxwell, Valeriya Mechkova, Juraj
Medzihorsky, Johannes von Römer, Aksel Sundström, Eitan Tzelgov, Yi- ting
Wang, Tore Wig, and Daniel Ziblatt. (2020). V-Dem [Country–Year/Country–
Date] Dataset v10. Varieties of Democracy (V- Dem) Project. https:// doi.org/
10.23696/vdemds20.
Córdova, A. and G. Rangel (2017) Addressing the gender gap: The effect of compul-
sory voting on women’s electoral engagement. Comparative Political Studies 50(2),
264–290.
Cox, G. W., and M. C. Munger (1989). Closeness, expenditures and turnout in the
1982 US house elections. American Political Science Review 83(1), 217–230.
Cox, M. (2003). When trust matters: Explaining differences in voter turnout. Journal
of Common Market Studies 41(4), 757–770.
Craig, S. C., and M. A. Maggiotto (1982). Measuring political efficacy. Political
Methodology 8, 85–109.
Crewe, I. (1981). Electoral participation. In D. Butler, H. R. Penniman and A. Ranney
(eds), Democracy at the Polls: Comparative Study of Competitive National Elections.
Washington, DC: American Enterprise Institute.
Croke. K, G. Grossman, H. A. Larreguy, and J. Marshall (2016). Deliberate disengage-
ment: How education can decrease political participation in electoral authoritarian
regimes. American Political Science Review 110(3), 579–600.
Curtice, J. (2002). The state of election studies: Mid-life crisis or new youth? Electoral
Studies 21(2002), 161–168.
Curtice J. (2007). Comparative opinion surveys. In Russell J. Dalton and Hans-Dieter
Klingemann (eds), The Oxford Handbook of Political Behavior. Oxford: Oxford
University Press.
References 215
Curtis G. (1971). Election Campaigning Japanese Style. New York: Columbia
Univ. Press.
Cutler, Neal E., and Vern l. Bengtson (1974) Age and political alienation: Maturation,
generation and period effects. In R. J. Dalton and H. Klingemann (eds), Oxford
Handbook of Political Behavior. Oxford: Oxford University Press.
Dahl, Robert A. (1971). Polyarchy: Participation and Opposition. New Haven: New
Haven Press.
Dahlberg, S., and Maria Solevid (2016). Does corruption suppress voter turnout?,
Journal of Elections, Public Opinion and Parties 26(4), 489–510.
Dalton R. J. (1984). Cognitive mobilization and partisan dealignment in advanced
democracies. The Journal of Politics 46(1), 264–284.
Dalton, R. J. (2000). The decline of party identification. In R. J. Dalton and M. P.
Wattenberg (eds), Parties without Partisans: Political Change in Advanced Industrial
Democracies. Oxford and New York: Oxford University Press.
Dalton, Russell J., and Christopher J. Anderson (eds) (2011). Citizens, Context,
and Choice: How Context Shapes Citizens’ Electoral Choices. New York: Oxford
University Press.
Dalton, Russell J., David M. Farrell, and Ian McAllister (2011). Political Parties and
Democratic Linkage: How Parties Organize Democracy. Oxford: Oxford Scholarship
Online (published online Feb. 2015).
Dalton, R. J., and Hans-Dieter Klingemann (2009). The Oxford Handbook of Political
Behaviour. Oxford: Oxford University Press.
Dalton, R. J., D. C. Shin, and Y.- H. Chu (eds) (2008). Party Politics in East
Asia: Citizens, Elections, and Democratic Development. Colorado: Lynne Rienner
Publishers.
Dassonneville, R. (2017). Age and voting. In K. Arzheimer, J. Evans and M. S. Lewis-
Beck (eds), The SAGE Handbook of electoral behaviour (Vol. 2). London: Sage.
Dawson, H. (2014). Youth and political unrest: Waiting and envy in a South African
informal settlement. Journal of Southern African Studies 40, 861–882.
de Vries, C. (2017). Cross-national data sources: Opportunities and challenges. In K.
Arzheimer , J. Evans and M. S. Lewis-Beck (eds), The SAGE Handbook of Electoral
Behaviour (Vol. 2). London: Sage.
Denton, Ginger L. (2016). Political Participation in Asia: Typologies of Political
Behavior across Democratizing States. Singapore: World Scientific.
Denver, D., Christopher Carman, and Robert Johns (2012). Elections and Voters in
Britain. Basingstoke: Red Global Press.
Desposato, S., and B. Norrander (2009). The gender gap in Latin America: Contextual
and individual influences on gender and political participation. British Journal of
Political Science 39, 141–162.
Diamond, L. (1999). Developing Democracy: Toward Consolidation. Baltimore:
JHU Press.
Diamond, L. (2002) Consolidating democracies. In Lawrence LeDuc, Richard G.
Niemi, and Pippa Norris (eds), Comparing Democracies 2: New Challenges in the
Study of Elections and Voting. London: Sage.
Diwakar, Rekha (2008). Voter turnout in the India states: An empirical analysis.
Journal of Elections, Public Opinion and Parties 18(1), 75–100.
Domínguez, Jorge I., and Chappell Lawson (eds) (2003). Mexico’s Pivotal Democratic
Election: Candidates, Voters, and the Presidential Campaign of 2000. Stanford,
216 References
CA: Stanford University Press and the Center for U.S.-Mexican Studies, University
of California, San Diego.
Domínguez, Jorge I., and James A. McCann (1996). Democratizing Mexico: Public
Opinion and Electoral Choice. Baltimore: Johns Hopkins University Press.
Domínguez, Jorge I., and James A. McCann (1998). Mexicans react to electoral fraud
and political corruption: An assessment of public opinion and voting behavior.
Electoral Studies 17, 483–503.
Dow, Jay (1998). A spatial analysis of the 1989 Chilean presidential election. Electoral
Studies 17, 61–76.
Downs, A. (1957). An Economic Theory of Democracy. New York: Harper and Row.
Duch, Raymond M. (2007). Comparative studies of the economy and the vote. In
Carles Boix and Susan C. Stokes (eds), The Oxford Handbook of Comparative
Politics. Oxford: Oxford University Press.
Duverger, Maurice (1955). The Political Role of Women. New York: UNESCO.
Edgell, A. B., V. Mechkova, D. Altman, M. Bernhard, and S. I. Lindberg (2017). When
and where do elections matter? A global test of the democratization by elections
hypothesis, 1900–2010. Democratization 1–23.
Eiffert, B., E. Miguel, and D. Posner (2010). Political conflict and ethnic identification
in Africa. American Journal of Political Science 54(2), 494–510.
Ekström, M., and J. Östman (2013). Family talk, peer talk, and young people’s
civic orientation. European Journal of Communication 28, 294–308. doi: 10.1177/
0267323113475410.
Ellis, Andrew, Maria Gratschew, Joh H. Pammett, and Erin Thiessen et al. (2006).
Engaging the electorate: Initiative to promote voter turnout from around the
world. Stockholm: International IDEA.
Endersby, J. W., and J. T. Krieckhaus (2008). Turnout around the globe: The influ-
ence of electoral institutions on national voter participation, 1972–2000. Electoral
Studies 27, 601–610.
Evans, Geoffrey (2006). The social bases of political divisions in post-communist
Eastern Europe. Annual Review of Sociology 32, 245–270.
Evans, Geoffrey and Nan Dirk De Graaf (eds) (2013). Political Choice Matters.
Explaining the Strength of Class and Religious Cleavages in Cross- National
Perspective. Oxford: Oxford University Press.
Evans, Geoffrey, and Stephen Whiteﬁeld (1993). Identifying the bases of party compe-
tition in Eastern Europe. British Journal of Political Science 23(4), 521–548.
Evans, J. (2004). Voters and Voting –An Introduction. London: Sage Publications.
Fagan, A., and P. Kopecký (2017). The Routledge Handbook of East European Politics
(1st ed.). Milton: Taylor & Francis.
Farrell, David (2001). Electoral Systems: A Comparative Introduction. Houndmills,
Basingstoke: Palgrave Macmillan.
Ferree, K. (2006). Explaining South Africa’s racial census. Journal of Politics 68(4),
803–815.
Filer, J. E., and L. W. Kenny (1980). Voter turnout and the benefits of voting. Public
Choice 35, 575–585.
Fisher, J., D. Denver, and G. Hands (2006) Party membership and campaign activity in
Britain: the impact of electoral performance. Party Politics 12(4), 505–519.
Fisher, J., E. Fieldhouse, M. N. Franklin, R. Gibson, M. Cantijoch, and C. Wlezien
(2018). The Routledge Handbook of Elections, Voting Behavior and Public Opinion.
London: Routledge.
References 217
Fishman, R., and J. Svensson (2000). Are Corruption and Taxation Really Harmful
to Growth? Firm Level Evidence. Policy Research Working Paper Series No. 2485.
Washington, DC: World Bank.
Flanagan, Scott, Shinsaku Kohei, Ichiro Miyake, and Bradley M. Richardson (1991).
The Japanese Voter. New Haven, CT: Yale University Press.
Flanagan, Scott, and Bradley M. Richardson. (1977). Japanese Electoral
Behavior: Social Cleavages, Social Networks, and Partisanship. London: Sage.
Flores, T. E. and I. Nooruddin (2016). Elections in Hard Times: Building Stronger
Democracies in the 21st Century. Cambridge: Cambridge University Press.
Fornos, C. A., Power, T. J., and Garand, J. C. (2004). Explaining voter turnout in Latin
America, 1980 to 2000. Comparative Political Studies 37(8), 909–940.
Fox J. (1994). The difficult transition from clientelism to citizenship: Lessons from
Mexico. World Polit. 46(Jan.), 151–184.
Franklin, M. (1996). Electoral participation. In Lawrence LeDuc, Richard G. Niemi
and Pippa Norris (eds), Comparing Democracies: Elections and Voting in Global
Perspective. Thousand Oaks, California: Sage.
Franklin, M. (2002). The dynamics of electoral participation. In Lawrence LeDuc,
Richard G. Niemi and Pippa Norris (eds), Comparing Democracies 2: New
Challenges in the Study of Elections and Voting (2nd ed.). London; Thousand
Oaks: Sage.
Franklin, Mark N. (2004). Voter Turnout and the Dynamics of Electoral Competition
in Established Democracies since 1945. Cambridge, UK; New York: Cambridge
University Press.
Franklin, M., and W. P. Hirczy de Miño (1998). Separated powers, divided govern-
ment, and turnout in U.S. presidential elections. American Journal of Political
Science 42(1), 316–326.
Fukuyama, Francis (1989). The end of history? The National Interest 16, 3–18.
Fukuyama, Francis (2001). Asian values in the wake of the Asian crisis. In Farrukh
Iqbal and John-Il You (eds), Democracy, Market Economics and Development.
Washington, DC: World Bank.
Furnham, A., and B. Gunter (1989). The Anatomy of Adolescence. Young People’s
Social Attitudes in Britain. London: Routledge.
Gabriel, O. W. and J. W. van Deth (1995) Political interest. In J. W. van Deth and E.
Scarbrough (eds), The Impact of Values. Oxford: Oxford University Press.
Gallagher, M. (1991). Proportionality, disproportionality and electoral systems.
Electoral Studies 10, 33–51.
Gallagher, M. and Paul Mitchell (2005). The Politics of Electoral Systems.
Oxford: Oxford University Press.
Gallego, A. (2009). Where else does turnout decline come from? Education, age, gener-
ation and period effects in three European countries. Scandinavian Political Studies
32(1), 23–44.
Gallego, A. (2010). Understanding unequal turnout: Education and voting in com-
parative perspective. Electoral Studies 29, 239–248.
Gallego, A. (2015). Unequal Political Participation Worldwide. Cambridge: Cambridge
University Press.
Gallego, A., Rico, G., and Anduiza, E. (2012). Disproportionality and voter turnout
in new and old democracies. Electoral Studies 31(1), 159–169.
Gastil, R. D. (1979). Freedom in the World: Political Rights and Civil Liberties.
New York: Freedom House.
218 References
Geddes, Barbara (1990). How the cases you choose affect the answers you get: Selection
bias in comparative politics. Political Analysis 2(1), 131–150.
Geertz, Clifford (1963). Old Societies and New States. New York: Free Press.
Gélineau, François (2013). Electoral accountability in the developing world. Electoral
Studies 32(3), 418–424.
Gélineau, François, and Matthew Singer (2015). The economy and incum-
bent support in Latin America. In Ryan E. Carlin, Matthew M. Singer, and
Elizabeth J. Zechmeister (eds), The Latin American Voter. Ann Arbor: University
Michigan Press.
Gerber, A. S., and D. P. Green (2000a). The effects of canvassing, telephone calls, and
direct mail on voter turnout: A field experiment. American Political Science Review
94(3), 653–663.
Gerber, A. S., and D. P. Green (2000b). The effect of a nonpartisan get-out-the-vote
drive: An experimental study of leafleting. Journal of Politics 62(3), 846–857.
Gerring, John (2007). Case Study Research: Principles and Practices. Cambridge:
Geys, B. (2006). Voter turnout: A review of aggregate-level research. Electoral Studies
25, 637–663.
Gordon, S. B. and Segura, G. M. (1997). Cross-National Variations in the Political
Sophistication of Individuals: Capability or Choice? Journal of Politics 59,
126–147.
Gottlieb, Jessica, Robert Blair, Hannah Baron, Aries Arugay, Cameron Ballard-Rosa,
Don Davidson, Laura Gamboa, Guy Grossman, Shelby Grossman, Christina
Kulich- Vamvakas, Nancy Lapp, Jennifer McCoy, Amanda Robinson, Steven
Rosenzweig, Eric Royer, Cathy Lisa Schneider, Susan Stokes, and Megan Turnbull.
2019. Democratic Erosion Event Dataset v3. Democratic Erosion: A Cross-
University Collaboration.
GRAY, M., & CAUL, M. (2000). Declining Voter Turnout in Advanced Industrial
Democracies, 1950 to 1997: The Effects of Declining Group Mobilization.
Comparative Political Studies, 33(9), 1091–1122.
Graziano L. (1973). Patron-client relationships in Southern Italy. Eur. J. Polit. Res.
1(1), 3–34.
Guardado, Jenny R., and Leonard Wantchekon (2013). Do electoral handouts affect
voting behavior? Paper presented at the annual meeting of the Midwest Political
Science Association, Chicago, IL, April.
Gupta, Sanjeev, Hamid Davoodi, and Rosa Alonso-Teme (2002). Does corruption
affect income inequality and poverty? Economics of Governance 3(1), 23–45.
Guy, Peters B. (2005). Institutional Theory in Political Science: The “New
Institutionalism”. Gosport: Ashford Colour Press Ltd.
Hall, P. A. and R. C. R. Taylor (1996). Political science and the three new
institutionalisms. Political Studies 44(5), 936–957.
Hamilton, L. C. (2002). Statistics with Stata. College Station, TX: Stata Corporation.
Harder J. and J. A. Krosnick (2008). Why do people vote? A psychological analysis of
the causes of voter turnout. Journal of Social Issues 64(3), 525–549.
Harkness, J. A., F. J. R. Van de Vijver, and P. P. Mohler (2003). Cross-Cultural Survey
Methodology. New York: Wiley.
Harris, F. C. (1994). Something within: Religion as a mobilizer of African-American
political activism. Journal of Politics 56(1), February, 42–68.
Healey, Joseph F. (2012). Statistics: A Tool for Social Research (9th ed.), International
edition. Belmont, CA: Wadsworth Cengage Learning.
References 219
Heath, A., S. Fisher and S. Smith (2005). The globalization of public opinion research.
Annual Review of Political Science 8, 297–333.
Heath, A. F., R. Jowell, J. Curtice, G. A. Evans, J. Field, and S. Witherspoon
(1991). Understanding Political Change: The British Voter 1964– 1987. Oxford:
Pergamon Press.
Helliwell, J. F. (1994). Empirical linkage between democracy and economic growth.
Hellwig, Timothy (2010). Elections and the economy. In Lawrence LeDuc, Richard G.
Niemi, and Pippa Norris (eds), Comparing Democracies 3: Elections and Voting in
the 21st Century (3rd ed.). Thousand Oaks, CA: Sage.
Herron, Erik S. (2009). Elections and Democracy After Communism? Basingstoke:
Palgrave Macmillan.
Hicken, Allen (2011). Clientelism. Annual Review of Political Science 14(1), 289–310.
Hillygus, D. Sunshine (2005). The missing link: Exploring the relationship between
higher education and political engagement. Political Behavior 27(1), 25–47.
Hoffmann-Martinot, V. (1994). Voter turnout in French municipal elections. In L.
Lopez-Nieto (ed.), Local elections in Europe (13–42). Barcelona, Spain: Institut de
ciènces politiques I socials.
Holbein, John B. and D. Sunshine Hillygus (2020). Making Young Voters: Converting
Civic Attitudes into Civic Action. New York: Cambridge University Press.
Honwana, Alcinda (2012). The Time of Youth: Work, Social Change, and Politics in
Africa. Sterling, VA: Kumarian Press.
Hooghe, M., and B. Wilkenfeld (2008). The stability of political attitudes and
behaviors across adolescence and early adulthood. A comparison of survey data on
adolescents and young adults in eight countries. Journal of Youth and Adolescence
37, 155–167.
Hooghe, M., and E. Quintelier (2014). Political participation in European coun-
tries: The effect of authoritarian rule, corruption, lack of good governance and
economic downturn. Comparative European Politics 12, 209–232.
Horn, J. L., and J. J. McArdle (1992). A practical and theoretical guide to meas-
urement invariance in aging research. Experimental Aging Research 18(3),
117–144.
Horowitz, D. (1985). Ethnic Groups in Conflict. Berkeley: University of California
Press.
Horowitz, D. (1991). A Democratic South Africa? Constitutional Engineering in a
Divided Society. Berkeley: University of California Press.
Hsieh, John Fuh-Sheng, and David Newman (2002). How Asia Votes. New York:
Chatham House.
Hsieh, John Fuh- sheng, and David Newman (2002). Continuity and change in
Taiwan’s electoral politics. In J. Hsieh and D. Newman (eds), How Asia Votes.
New York: Seven Bridges.
Huntington, S. (1991). The Third Wave: Democratization in the Late Twentieth Century.
Norman, OK: University of Oklahoma Press.
Hyug Baeg, Im. (2004). Faltering democratic consolidation in South Korea: Democracy
at the end of the “Three Kims” era. Democratization 11(5), 187.
Inglehart, R. and P. Norris (2000). The developmental theory of the gender
gap: Women’s and men’s voting behavior in global perspective. International
Inglehart, R., and P. Norris (2003). Rising Tide –Gender Equality and Cultural Change
around the World. New York: Cambridge University Press.
220 References
Inman, Kris, and Josephine T. Andrews (2009). Corruption and political participation
in Africa. Evidence from survey and experimental research. Paper presented at the
Midwest Political Science Association, April 3–6, Chicago.
International Institute for Democracy and Electoral Assistance (2002). Voter Turnout
since 1945: A Global Report (2nd edition). Stockholm: International IDEA.
Isaksson, Ann-Sofie (2010). Political participation in Africa: Participatory inequal-
ities and the role of resources. Working papers in Economics, No. (462),
Department of Economics School of Business, Economics and Law at University
of Gothenburg.
Isaksson, Ann-Sofie (2014). Political participation in Africa: The role of individual
resources. Electoral Studies 32, 244–260.
Ishiyama, J. (1996). Electoral systems experimentation in the new Eastern Europe: The
single transferable vote and the additional member system in Estonia and Hungary.
East European Quarterly 29(4), 487–507.
Ishiyama, J. and Fox, C. (2006). What affects the strength of partisan identity in sub-
Saharan Africa. Politics and Policy 34(4), 748–773.
Jackman, R. W. (1987). Political Institutions and Voters Turnout in the Industrial
Democracies, American Political Science Review, 81(2), 405–424.
Jackman, R. W. and Miller, Ross M. A. (1995). Voter turnout in the industrial democ-
racies during the 1980s. Comparative Political Studies 27, 467–492.
Jankowski, Thomas B. and John M. Strate (1995). Modes of participation over the
adult life span. Political Behavior 17(1), 89–106.
Jasiewicz, K. (2007). Citizens and politics. In S. White, J. Batt, and P. G. Lewis (eds),
Developments in Central and East European Politics 4. Basingstoke: Palgrave
Macmillan.
Jennings, M. Kent (1979). Another look at the life cycle and political participation.
American Journal of Political Science 23(4), 755–771.
Johnson, Gregg, and Sooh-Rhee Ryu (2010). Repudiating or rewarding neoliberalism?
How broken campaign promises condition economic voting in Latin America.
Latin American Politics and Society 52, 1–24.
Johnston, R., D. Cutts, C. Pattie, and J. Fisher (2012) We’ve got them on the
list: Contacting, canvassing and voting in a British general election campaign.
Electoral Studies, 31(2), 317–329.
Jones, M. P. (2009). Gender quotas, electoral laws, and the election of women: Evidence
from the Latin American vanguard. Comparative Political Studies 42, 56–81.
Jones-Correa, M. A., and D. L. Leal (2001). Political participation: Does religion
matter? Political Research Quarterly 54(4), December, 751–770.
Juma, Calestous (2011). Why Africa needs to lower its voting age to 16. The Guardian,
9 February 2011.
Karahan, G. R., Coats, M., and Shugart, W. F. II (2006). Corrupt political jurisdictions
and voter participation. Public Choice 126(1), 87–196.
Karp, J. A., and S. A. Banducci (2008). Political efficacy and participation in twenty-
seven democracies. British Journal of Political Science 38, 311–334.
Karp, J. A., S. A. Banducci, and S. Bowler (2007). Getting out the vote: Party
mobilization in a comparative perspective. British Journal of Political Science 38,
91–112.
Karp, J. A., and D. Brockington, (2005). Social desirability and response validity: A
comparative analysis of overreporting voter turnout in five countries. Journal of
Politics 67, 825–840.
References 221
Karp, Jeffrey A., and Caitlin Milazzo (2015). Democratic scepticism and political
participation in Europe. Journal of Elections, Public Opinion and Parties 25(1),
97–110.
Kasara, Kimuli, and Pavithra Suryanarayan (2015). When do the rich vote less than
the poor and why? Explaining turnout inequality across the world. American
Journal of Political Science 59(3), 613–627.
Katz, R. (1997). Democracy and Elections. Oxford and New York: Oxford
University Press.
Keefer, P. 2007. Clientelism, credibility, and the policy choices of young democracies.
American Journal of Political Science 51(4), 804–821.
Kim, Seok-Eun (2005) The role of trust in the modern administrative state: An inte-
grative model. Administration and Society 37(5), 611–635.
Kinder, D. R. (2006). Politics and the life cycle. Science 312, 1905–1908.
King, A. (1969). Political parties in Western democracies. Polity 2, 112–141.
Kitschelt, H., K. A. Hawkins, J. P. Luna, G. Rosas, E. J. Zechmeister (2010). Latin
American Party Systems. Cambridge, UK: Cambridge University Press.
Kittilson, M. C. (2007). Research resources in comparative political behavior. In J.
R. Dalton and H. Klingemann (eds), Oxford Handbook of Political Behavior.
Kittilson, M. C., and C. J. Anderson (2011). Electoral supply and voter turnout.
In Russell J. Dalton and Christopher J. Anderson (eds), Citizens, Context,
and Choice: How Context Shapes Citizens’ Electoral Choices. Oxford: Oxford
University Press.
Kittilson, M. C., L. Schwindt-Bayer (2012). The gendered effects of electoral institutions.
Klesner, J. L. (2007). Social capital and political participation in Latin America:
Evidence from Argentina, Chile, Mexico, and Peru. Latin American Research
Review 42, 1–32.
Klingemann, H.- D. (2007). Turnout in European and national parliamentary
elections. In C. Lequesne, and MacDonagh-Pajerova, M. (eds), La Citoyennete ´
De ´mocratique Dans l’Europe Des Vingt-Sept. Paris: L’Harmattan.
Konzelmann, L., Wagner, C., and Rattinger, H. (2012). Turnout in Germany in the
course of time: Life Cycle and cohort effects on electoral turnout from 1953 to
2049. Electoral Studies 31(2), 250–261.
Kostadinova, Tatiana. (2003). Voter turnout dynamics in postcommunist Europe.
European Journal of Political Research 42(6), 741–759.
Kostadinova, Tatiana (2009). Abstain or rebel: Corruption perceptions and voting in
East European elections. Politics & Policy 37(4), 691–714.
Kostadinova, T. and Power, T. J. (2007). Does democratization depress participation?
Voter turnout in the Latin American and Eastern European transitional democra-
cies. Political Research Quarterly 60(3), 363–377.
Kostelka, F. (2014). The state of political participation in post-communist democ-
racies: Low but surprisingly little biased citizen engagement. Europe-Asia Studies
66(6), 945–968.
Kostelka, F. (2015). To mobilise and demobilise: The puzzling decline of voter turnout
in post-communist democracies. PhD dissertation, Centre d’études européennes,
Sciences Po, Paris.
Kostelka, F. (2017). Does democratic consolidation lead to a decline in voter turnout?
Global evidence since 1939. American Political Science Review 111(4), 653–667.
222 References
Kostelka, Filip, Blais A., and E. Gidengil (2019). Has the gender gap in voter turnout
really disappeared? West European Politics 42, 437–463.
Kramon, E. (2017). Money for Votes: The Causes and Consequences of Electoral
Clientelism in Africa. Cambridge: Cambridge University Press.
Krishna, A. (2002). Enhancing political participation in democracies: What is the role
of social capital? Comparative Political Studies 35(4), 437–460.
Krishna, A. (2008). Do poor people care less for democracy? Testing individual-level
assumptions with individual-level data. In A. Krishna (ed.), Poverty, Participation
and Democracy: A Global Perspective. New York: Cambridge University Press.
Kuenzi, M., and G. M. S. Lambright (2007). Voter turnout in Africa’s multiparty
regimes. Comparative Political Studies 40(6), 665–690.
Kuenzi, M., and G. M. S Lambright, (2011). Who votes in Africa? An examination of
electoral participation in 10 African countries. Party Politics 17, 767–799.
La Due Lake, Ronald, and Robert Huckfeldt (1998). Social capital, social networks,
and political participation. Political Psychology 19(3), 567–584.
Ladner, A. and H. Milner (1999). Do voters turn out more under proportional than
majoritarian systems? The evidence from Swiss communal elections. Electoral
Studies 18, 235–250.
Lane, Roberts E. (1959). Political Life: Why People Get Involved in Politics. Glencoe,
Illinois: The Free Press.
Langston, Joy (2007). Strong parties in a struggling party system: Mexico in the
democratic era. In Paul Webb, and Stephen White (eds), Political Parties in New
Democracy. Oxford: Oxford University Press.
Lanning, K. (2008). Democracy, voting, and disenfranchisement in the United
States: A social psychological perspective. Journal of Social Issues 64, 431–446.
LaPiere, R. T. (1934). Attitudes vs. actions. Social Forces 13, 230–237.
Lavezzolo, Sebastián (2008). Adversidad económica y participación electoral en
América Latina, 1980–2000. Revista Española de Ciencia Política 18, 67–93.
Lazarsfeld, Paul F., Bernard Berelson, and Hazel Gaudet. 1948. The People’s
Choice: How the Voter Makes Up His Mind in a Presidential Campaign (2nd ed.).
New York: Columbia University Press.
LeDue, L., Niemi Richard, G., and Norris Pippa (1996). Comparing Democracies:
Elections and Voting in Global Perspective. Thousand Oaks, CA: Sage.
Lee, A.-R., and Glasure, Y. U. (2007). Social Capital and Political Participation in
South Korea. Asian Affairs: An American Review 34, 101–118.
Lehoucq, F., and D. Wall (2004). Explaining voter turnout rates in new democra-
cies: Guatemala. Electoral Studies 23, 485–500.
Leighley, J. E. and J. Nagler (1992). Individual and systemic influences on turnout: Who
votes? Journal of Politics 54(3), 718–740.
Leighley, Jan E., and J. Nagler (2007). Unions, voter turnout, and class bias in the US
electorate, 1964–2004. Journal of Politics 69(2), 430–441.
Leighley, J. E. and J. Nagler (2014). Who Votes Now? Demograhpic, Issues, Inequality
and Turnout in the United States. Princeton, NJ: Princeton University Press.
Letki, N. (2003). Explaining political participation in East Central Europe: Social cap-
ital, democracy and the communist past. Nuffield College Politics Working Paper
2003-W2, Oxford: University of Oxford.
Letsa, N. (2019). Expressive voting in autocracies: A theory of non-economic partici-
pation with evidence from Cameroon. Perspectives on Politics 18(2), 439–453.
References 223
Lever, H. (1979). Ethnicity and voting patterns in South Africa. Political Studies 27(3),
458–468.
Levitsky, Steven and Lucan Way (2002). The rise of competitive authoritarianism.
Journal of Democracy 13(2), 51–65.
Levitsky, S. and L. A. Way (2010). Competitive Authoritarianism: Hybrid Regimes after
the Cold War. Cambridge: Cambridge University Press.
Lewis-Beck, William G. Jacoby, Helmut Norpoth, and Herbert F. Weisberg (2008).
The American Voter Revisited. Ann Arbor: University of Michigan, Press.
Lewis-Beck, M., and M. Stegmaier (2000). Economic determinants of election
outcomes. Annual Review of Political Science 3, 183–219.
Lewis-Beck, Michael. S., and Mary Stegmaier (2007). Economic models of voting.
In Russell Dalton and Hans-Dieter Klingemann (eds), The Oxford Handbook of
Political Behavior. Oxford: Oxford University Press.
Lewis-Beck, Michael S., and Mary Stegmaier (2008). The economic vote in transi-
tional democracies. Journal of Elections, Public Opinion and Parties 18(3), 303–323.
Lindberg, S. (2006a). Democracy and Elections in Africa. Baltimore, MD: Johns
Hopkins Univ. Press.
Lindberg, S. (2006b). The surprising significance of African elections. Journal of
Democracy 17(1), 139–151.
Lindberg, S. I. (2009). Democratization by Election: A New Mode of Transition.
Baltimore, MD: Johns Hopkins Univ. Press.
Lindberg, Staffan I. (2010). What accountability pressures do MPs in Africa face and
how do they respond? Evidence from Ghana. Journal of Modern African Studies
48(1), 117–142.
Linz, J. J., and Stepan, A. (1996). Problems of Democratic Transition and Consolidation.
Baltimore: Johns Hopkins University Press.
Lührmann, Anna and Staffan I. Lindberg (2019). A third wave of autocratization is
here: What is new about it?, Democratization 26(7), 1095–1113.
Lust, Ellen and David Waldner (2015). Unwelcome Change: Understanding,
Evaluating, and Extending Theories of Democratic Backsliding. Washington, DC:
USAID.
Lyne, M. 2007. Rethinking economics and institutions: the voter’s dilemma and demo-
cratic accountability. See Kitschelt and Wilkinson 2007c, 206–226.
Macpherson, C. B. (1977). The Life and Time of Liberal Democracy. Oxford: Oxford.
Mainwaring, Scott (1999). Rethinking Party Systems in the Third Wave of
Democratization: The Case of Brazil. Stanford, CA: Stanford University Press.
Mainwaring, Scott, and Mariano Torcal (2006). Party system institutionalization and
party system theory after the third wave of democratization. In Richard S. Katz
and William Crotty (eds), Handbook of Political Parties. London: Sage.
Michael, Marsh, Richard Sinnott, John Garry, and Fiachra Kennedy (2008).
The Irish Voter: The Nature of Electoral Competition in the Republic of Ireland.
Manchester: Manchester University Press.
Mattes. R., and D. Mughogho (2010). The Limited Impacts of Formal Education
on Democratic Citizenship in Africa. Higher Education Research and Advocacy
Network in Africa Paper. Cape Town: Centre for Higher Education Transformation.
Mauro, P. (1997). The effects of corruption on growth, investment and government
expenditure: A crosscountry analysis. In K. A. Elliott (ed.), Corruption and the
Global Economy. Washington, DC: Institute for International Economics.
224 References
Maydom, B. (2017). Migrant Remittances and Democracy. DPhil diss., University of
Oxford.
McAllister, Daniel J. (1995). Affect and cognition-based trust as foundations for
interpersonal cooperation in organizations. Academy of Management Journal
38(1), 24–59.
McCann, J. A., and Domínguez, J. I. (1998). Mexicans react to political fraud and
corruption: An assessment of public opinion and voting behavior. Electoral Studies
17, 483–503.
McLaughlin, E. (2008). Beyond the racial census: The political salience of ethno-
linguistic cleavages in South Africa. Comparative Political Studies 40(4),
435–456.
Melo, D. F., and D. Stockemer (2014). Age and political in Germany, France and the
UK: A comparative analysis. Comparative European Politics 12(1), 33–53.
Mill, J. S. (1848/1875). A System of Logic: Ratiocinative and Inductive: Being a Connected
View of the Principles of Evidence and the Method of Scientific Investigation, Vol 1
(9th ed.). London: Longmans, Green Reader, and Dyer.
Millard, Frances (1999). Polish Politics and Society. London; New York: Routledge.
Miller, A. H., P. Gurin, G. Gurin, and O. Malanchuk (1981). Group conscious-
ness and political participation. American Journal of Political Science 25(3),
494–511.
Miller, A. H. and O. Listhaug (1990). Political parties and confidence in government.
Miller, W. E. (1994). An organizational history of the intellectual origins of the
American national election studies. In J. J. A. Thomassen (ed.), The Intellectual
History of Election Studies. Special Issue of the European Journal for Political
Research. Dordrecht: Kluwer.
Miller, W. E., and J. Merrill Shanks (1996). The New American Voter. Cambridge:
Harvard University Press.
Mo, Jongryn, David Brady, and Jaehun Ro. (1991). Urbanization and voter turnout in
Korea: An update. Political Behavior 13, 21–31.
Monyake Moletsane (2018). Can the poor fight back? A multilevel analysis of
corruption, poverty and collective demand for accountability in Africa, Journal of
Contemporary African Studies 36(4), 514–531.
Montinola, Gabriella R. (1999). The Philippines in 1998: Opportunities amid crisis.
Asian Survey 39(1), 67.
Monyake, Moletsane, and Dan Hough (2019). Citizens, bribery and the propensity to
protest. Commonwealth & Comparative Politics 57(3), 282–302.
Moreno, Alejandro (2003). El Votante Mexicano: Democracia, Actitudes Políticas y
Conducta Electoral. Mexico City: Fondo de Cultura Económica.
Moreno, Alejandro. (2009). La Decisión Electoral: Votantes, Partidos, y Democracia en
México. Mexico City: Miguel Ángel Porrúa.
Morgenbesser, L., and Pepinsky, T. B. (2019). Elections as causes of democratiza-
tion: Southeast Asia in comparative perspective. Comparative Political Studies
52(1), 3–35.
Nadeau, Richard, Éric Bélanger, and Thomas Didier (2013). The Chávez vote and the
national economy in Venezuela. Electoral Studies Sept. 32(3), 482–488.
Nadeau, Richard, Éric Bélanger, Michael S. Lewis- Beck, Mathieu Turgeon and
François Gélineau (2017). Latin American Elections: Choice and Change. Ann
Arbor: The University of Michigan Press.
References 225
Nathan, Noah (2019). Does participation reinforce patronage? Policy preferences,
turnout, and class in urban Ghana. British Journal of Political Science 49(1),
229–255.
Neundorf, A., J. Gerschewski, and R.-G. Olar (2019). How do inclusionary and
exclusionary autocracies affect ordinary people?. Comparative Political Studies.
doi: 10.1177/0010414019858958.
Newman, David (2002). Hong Kong: A decade of change. In John Fuh-Sheng Hsieh
and David Newman (eds), How Asia Votes. New York: Chatham House,
Nichter, S. (2008). Vote buying or turnout buying? Machine politics and the secret
ballot. American Political Science Review 102(1), 19–31.
Nie, N. H., J. Junn, and K. Stehlik-Barry (1996). Education and Democratic Citizenship
in America. Chicago: University of Chicago Press.
Niemi, Richard G., Stephen C. Craig, and Franco Mattei (1991). Measuring internal
political efficacy in the 1988 national election study. American Political Science
Review 85, 1407–1413.
Nohlen, Dieter (2005). Elections in the Americas: A Data Handbook Volume 2: South
America. Oxford: Oxford University Press.
Norris, P. (2002). Democratic Phoenix: Reinventing Political Activism. New York:
Norris, P. (2003). Young people & political activism: From the politics of loyalty to the
politics of choice? Retrieved from http://ksghome.harvard.edu/~pnorris/Acrobat/
COE.pdf.
Norris, P. (2004, October). Young people and political activism: From the pol-
itics of loyalties to the politics of choice? Paper presented at the Conference for
Civic Engagement in the 21st century: Toward a Scholarly and Practical Agenda,
Columbia, South Carolina.
Norris, P. (2009). The globalization of comparative public opinion research. In T.
Landman and N. Robinson (eds), The Sage Handbook of Comparative Politics,
London: Sage.
Norris, Pippa (2002). Democratic Phoenix: Reinventing Political Activism. New York:
Norris, Pippa (2002). Social capital and the news media. Harvard International Journal
of Press/Politics 7 (2), 3–8.
Norris, Pippa (2004). Electoral Engineering: Voting Rules and Political Behavior.
New York: Cambridge University Press.
North, D. C. (1990). Institutions, Institutional Change, and Economic Performance:
Political Economy of Institutions and Decisions. Cambridge: Cambridge University
Press.
O’Donnell, G., and P. C. Schmitter (1986). Transitions from Authoritarian Rule:
Tentative Conclusions about Uncertain Democracies. Baltimore: Johns Hopkins
University Press.
O’Neill, Brenda (2003). Examining declining electoral turnout among Canada’s youth.
Electoral Insight 5, 2.
Ochoa, Enrique C. (1987). The rapid expansion of voter participation in Latin
America: Presidential elections, 1845–1986. In James W. Wilkie and David Lorey
(eds), Statistical Abstract of Latin America. Los Angeles: UCLA Latin American
Center Publications, University of California.
Offe, Claus (1991). Capitalism by democratic design? Democratic theory facing the
triple transition in East Central Europe. Social Research 58(4), 865–892.
226 References
Olorunsola, V. (1972). The Politics of Cultural Sub-nationalism in Africa. Garden City,
NJ: Doubleday.
Orvista, Marta, Caplanova Anetta, and Hudson John (2005). Intended electoral par-
ticipation in transitional countries. Economics of Governance 6(3), 211–228.
Pacek, Alexander (1994). Macroeconomic conditions and electoral politics in East
Central Europe. American Journal of Political Science 38(3), 723–744.
Pacek, Alexander C., Grigore Pop-Eleches, and Joshua A. Tucker (2009). Disenchanted
or discerning: Voter turnout in post-communist countries. The Journal of Politics
71(2), 473–491.
Park, C.-M. (2016). Political trust in the Asia-Pacific region. In S. Zmerli and T. W. G.
van der Meer (eds), Handbook on Political Trust. Cheltenham, UK: Edward Elgar
Publishing.
Park, Chan Wook. (2002). Elections in democratizing Korea. In John Fuh-Sheng
Hsieh and David Newman (eds), How Asia Votes. New York: Chatham House.
Pattie, C., and R. Johnston (2001). Losing the voters’ trust. British Journal of Politics
and International Relations 3, 191–222.
Perea, E. A. (2002) Individual characteristics, institutional incentives and electoral
abstention in Western Europe. European Journal of Political Research 41, 643–673.
Pérez-Liñán, A. (2001). Neoinsitutional accounts of voter turnout: Moving beyond
industrial democracies. Electoral Studies 20, 281–297.
Persson, Mikael (2015). Education and political participation. British Journal of
Political Science 45, 689–703.
Persson, M., Wass, H., and Oscarsson, H. (2013). The generational effect in turnout
in the Swedish general elections, 1960–2010. Scandinavian Political Studies 36(3),
249–269.
Persson, M., and M. Solevid (2014). Measuring political participation –Testing social
desirability bias in a web-survey experiment. International Journal of Public Opinion
Research 26, 98–112.
Persson, T., G. Tabellini, F. Trebbi (2003). Electoral rules and corruption. J. Eur. Econ.
Assoc. 1(4), 958–989.
Peter, Guy (2018). Institutions and voting behavior. In J. Fisher et al. (ed.), The
Routledge Handbook of Elections, Voting Behaviour and Public Opinion. Abingdon:
Routledge.
Pierce, Roy (1995). Choosing the Chief: Presidential Elections in France and the United
States. Ann Arbor: University of Michigan Press.
Pietsch, J., Michael Miller and Jeffrey A. Karp (2015). Public support for democ-
racy in transitional regimes. Journal of Elections, Public Opinion and Parties
25(1), 1–9.
Plutzer, E. (2002). Becoming a habitual voter: Inertia, resources, and growth in young
adulthood. American Political Science Review 96(1), 41–56.
Plutzer, E. (2018). Demographics and social bases of voter turnout. In J. Fisher,
E. Fieldhouse, M. N. Franklin, R. Gibson, M. Cantijoch, and C. Wlezien (eds),
The Routledge Handbook of Elections, Voting Behaviour and Public Opinion.
Abingdon: Routledge.
Podobnik, B., J. Shao, D. Njavro, P. C. Ivanov and H. E. Stanley (2008). Influence
of corruption on economic growth rate and foreign investment. The European
Physical Journal 63(3), 547–550.
Poole. K T., and Zeigler, L. H (1985). Women, Public Opinion, and Politics: The
Changing Political Attitudes of American Women. New York: Longman.
References 227
Pop-Eleches, Grigore (2010). Throwing out the bums: Protest voting and unorthodox
parties after communism. World Politics 62(2), 221–260.
Posner, D. (2005). Institutions and Ethnic Politics in Africa. New York: Cambridge
University Press.
Potgieter, E. (2013). Predictors of political participation in new democracies: A com-
parative study. MA Thesis, Stellenbosch University.
Powell, G. B. (1980). Voting turnout in thirty democracies. In. R. Rose (ed.), Electoral
Participation. Beverly Hills: Sage.
Powell, G. B. (1982). Contemporary Democracies: Participation, Stability and Violence.
Cambridge: Harvard University Press.
Przeworski, A. (1991). Democracy and the Market. Cambridge: Cambridge University
Press.
Przeworski, A., M. Alvarez, J. Cheibub, and F. Limongi (2000). Democracy and
Development: Political Institutions and Well- Being in the World. New York:
Putnam, R. D (1993). Making Democracy Work: Civic Traditions in Modern Italy.
Princeton, NJ: Princeton University Press.
Putnam, R. D. (1995). Bowling alone: America’s declining social capital. Journal of
Democracy 6(1), 65–78.
Putnam, R. D. (2000). Bowling Alone: The Collapse and Revival of American
Community. New York: Simon and Schuster.
Pye, Lucien (1966). Aspects of Political Development. Boston: Little Brown.
Ragin, C. (1987). The Comparative Method: Moving Beyond Qualitative and
Quantitative Strategies. Berkeley, CA; London: University of California Press.
Ragsdale, L. and J. G. Rusk (1993). Who are non-voters? American Journal of Political
Science 37, 721–746.
Reilly, Benjamin (2006). Democracy and Diversity: Political Engineering in the Asia-
Pacific. Oxford: Oxford University Press.
Remmer, K. L. (2010) Political scale and electoral turnout: Evidence from the less
industrialized world. Comparative Political Studies 43(3), 275–303.
Resnick, Danielle, and Daniela Casale (2014). Young populations in young democra-
cies: Generational voting behaviour in Sub-Saharan Africa. Democratization 21(6),
1172–1194.
Riker, William H. and Ordeshook, Peter C. (1968). A theory of the calculus of voting.
American Political Science Review 62, 25–42.
Roberts, Andrew (2018). Democracy and democratization in post-communist Europe.
In A. Fagan and P. Kopecký (eds), Routledge Handbook of East European Politics,
1st ed. London: Routledge.
Roberts, Nancy C. (2008). The Age of Direct Citizen Participation. New York: M.
E. Sharpe.
Rood, Steven (2002). Election as complicated and important events in the
Philippines. In John Fuh-Sheng Hsieh and David Newman (eds), How Asia Votes.
New York: Chatham House,
Rose, R. (1995). Mobilizing demobilized voters in post-communist societies. Party
Politics 1, 4.
Rose, R., and W. Mishler (1998). Negative and positive party identification in post-
communist countries. Electoral Studies 17(2), 217–234.
Rose, R. and N. Munro (2003). Elections and Parties in New European Democracies.
Washington, DC: CQ Press.
228 References
Rose, Richard, William Mishler, and Christian Haerpfer (1998). Democracy and Its
Alternatives: Understanding Post-communist Societies. Baltimore, MD: John
Hopkins University Press.
Rose, R. and C. Peiffer (2019). Bad Governance and Corruption: Political Corruption
and Governance. London: Palgrave Macmillan.
Rosenbaum, W. A (1975). Political Culture. New York: Praeger.
Rosenstone, Steven J., and John Mark Hansen (1993). Mobilization, Participation and
Democracy in America. New York: Macmillan.
Sanches, Edalina Rodrigues, José Jaime Macuane, and Chipo Dendere (2020).
Introduction. Cadernos de Estudos Africanos 38, 9–14.
Schedler, Andreas (1998). What is democratic consolidation? Journal of Democracy
9(2), 91–107.
Schlozman. K.L, N. Burns, and Sidney Verba (1994). Gender and the pathways to par-
ticipation: The role of resources. The Journal of Politics 56(4), 963–990.
Schmitt, Hermann and Renato Mannheimer (1991). About voting and non-voting in
the European elections in 1989. European Journal of Political Research 19, 31–54.
Schraufnagel, Scot, and Barbara Sgouraki (2005). Voter turnout in Central and South
America. The Latin Americanist 49, 39–69.
Schumpeter J. A. (1942). Capitalism, Socialism and Democracy. London and
New York: Routledge.
Schwindt-Bayer, L. (2009). Making quotas work: The effect of gender quota laws on
the election of women. Legislative Studies Quarterly 34, 5–28.
Scott, Ian (1996). Party politics and elections in transitional Hong Kong. Asian Journal
of Political Science 4(1), 130–152.
Segura, G., and S. Bowler (2012). The Future Is Ours: Minority Politics, Political
Behavior, and the Multiracial Era of American Politics. Thousand Oaks: Sage.
Seligson, M. A. (2002). Trouble in paradise? The erosion of system support in Costa
Rica, 1978–1999. Latin American Research Review 37, 160–185.
Seligson, Mitchell A., Annabelle Conroy, Ricardo Cordova Macias, Orlando J. Perez,
and Andrew J. Stein (1995). Who votes in Central America? A comparative ana-
lysis. In Mitchell A. Seligson and John A. Booth (eds), Elections and Democracy
in Central America, Revisited. Chapel Hill, NC: The University of North
Carolina Press.
Shachar, R., and B. Nalebuff (1999). Follow the leader: Theory and evidence on polit-
ical participation. American Economic Review 89(3), 525–547.
Sheerin, C. (2008). Political Efficacy and Youth Non- Voting. Saarbrucken: VDM
Verlag Dr. Muller.
Shi, T. (2014). The Cultural Logic of Politics in Mainland China and Taiwan. Cambridge,
UK: Cambridge University Press.
Shleifer, A. and R. W. Vishny (1993). Corruption. Quarterly Journal of Economics 108,
599–617.
Singer, M. (2009). Buying voters with dirty money: the relationship between clien-
telism and corruption. Presented at the Annual Meeting of the American Political
Science Association 3–6 September, Toronto.
Singh, Shane P. (2010). Contextual influences on the decision calculus: A cross-
national examination of proximity voting. Electoral Studies 29, 425–134.
Smets, K. (2010). A widening generational divide? Assessing the age gap in voter
turnout between younger and older citizens, PhD dissertation, European University
Institute.
References 229
Smets, K. (2012). A widening generational divide? The age gap in voter turnout
through time and space. Journal of Elections, Public Opinion and Parties 22(4),
407–430.
Smets, K. and Van Ham, C. (2013). The embarrassment of riches? A meta-analysis of
individual level research on voter turnout. Electoral Studies 32(2), 344–359.
Smith, A. E. (2009). Legitimate grievances: Preferences for democracy, system
support, and political participation in Bolivia. Latin American Research Review
44, 102–126.
Stegmaier, Mary, and Michael S. Lewis- Beck (2013). Economic voting. In Rick
Valelly (eds), Oxford Bibliographies in Political Science Online. New York: Oxford
University Press.
Stepan, Alfred (1990). On the tasks of a democratic opposition. Journal of Democracy
1: 41–9.
Stockemer, D. (2015). Turnout in developed and developing countries: Are the two
turnout functions different or the same? Political Science 67(1), 3–20.
Stockemer, Daniel (2015). District magnitude and electoral turnout: A macro-level
global analysis. Acta Politica 50(1), 82–100.
Stockemer, D. (2016). Is the turnout function in democracies and nondemocracies
alike or different?. Politics and Policy 44, 889–915.
Stockemer, D. (2017). What affects voter turnout? A review article/meta-analysis of
aggregate research. Government and Opposition 52(4), 698–722.
Stockemer, Daniel, Bernadette LaMontagne, and Lyle Scruggs (2013). Bribes and
ballots: The impact of corruption on voter turnout in democracies. International
Stokes S. C. (2005). Perverse accountability: A formal model of machine politics with
evidence from Argentina. American Political Science Review 99(3), 315–325.
Stolle, Dietlind and Marc Hooghe (2005). Inaccurate, exceptional, one-sided or irrele-
vant? The debate about the alleged decline of social capital and civic engagement in
Western societies. British Journal of Political Science 35, 149–167.
Strate John M., Charles J. Parrish, Charles D. Elder and Coit Ford (1989). Life span
civic development and voting participation. The American Political Science Review
83(2), 443–464.
Studlar, D. T., I. McAllister, and B. C. Hayes (1998). Explaining the gender gap in
voting: A cross-national analysis. Social Science Quarterly 79(4), 779–798.
Sundström, Aksel, and Daniel Stockemer (2015). Voter turnout in the European
regions: The impact of corruption perceptions. Electoral Studies 40, 158–169.
Surin, Maisrikrod (2002). Political reform and the new Thai electoral system. In
John Fuh-Sheng Hsieh and David Newman (eds), How Asia Votes. New York:
Chatham House.
Svolik, Milan W. (2015). Which democracies will last? Coups, Incumbent takeovers,
and the dynamic of democratic consolidation. British Journal of Political Science
45, 715–738.
Tambe, E. B. (2016). Who votes in East Asia? European Journal of East Asian Studies
15(2), 149–173.
Tambe, E. B. (2017a) Electoral participation in African democracies: The impact of
individual and contextual factors. Commonwealth & Comparative Politics 55(2),
119–140.
Tambe, E. B. (2017b). [Review] Aina Gallego (2015). Unequal Political Participation
Worldwide. Party Politics 23(5), 605–606.
230 References
Tambe, E. B. (2018a) Electoral participation in new democracies: applying existing
models of turnout to new democracies in Africa, East Asia and Post-Communist
Europe. Doctoral thesis, University of Sussex.
Tambe, E. B. (2018b). Who Does Not Vote and Why? Implication for New Democracies,
Elections –A Global Perspective. In Ryan M. Yonk (ed.), A Global Perspective.
London: IntechOpen.
Tambe, E. B. (forthcoming). Why do the poor and less educated vote more? Explaining
the reverse relationship between socio-economic status and turnout in Africa.
Paper presented at the African Politics Conference in 2020.
Tavits, Margit (2005). The development of stable party support: Electoral Dynamics
in post-communist Europe. American Journal of Political Science 49(2), 283–298.
Tavits, Margit (2008). Policy positions, issue importance, and party competition in
new democracies. Comparative Political Studies 41(1), 48–72.
Teixeira, R. A. (1984). Why Americans Don’t Vote: Turnout Decline in the United States
1960–1984. New York: Greenwood Press.
Teixeira, R. A. (1987), Why Americans Don’t Vote: Turnout Decline in the United States
1960–1984. New York: Greenwood Press.
Teixiera, R. A. (1992). The Disappearing American Voter. Washington: Brookings
Institute.
Telles, Helcimara, and Alejandro Moreno (eds) (2013). Comportamento Eleitoral e
Comunicação Política na América Latina. Belo Horizonte: Editora UFMG.
Teorell, J., and A. Hadenius (2009). Elections as levers of democracy: A global inquiry.
In S Lindberg (ed.), Democratization by Elections: A New Mode of Transition.
Johns Hopkins University Press.
Teorell, J., P. Sum, and M. Tobiasen (2007). Participation and political equality: An
assessment of large- scale democracy. In J. W. van Deth, J. R. Montero and
A. Westholm (eds), Citizenship and Involvement in European Democracies: A
Comparative Analysis. New York: Routledge.
Teorell, Jan, and Michael Wahman (2018). Institutional stepping stones for democ-
racy: How and why multipartyism enhances democratic change. Democratization
25(1), 78–97.
Terry, Lynn Karl (1986). Imposing consent: Electoralism and democratization in El-
Salvador. In Paul W. Drake and Eduardo Silva (eds), Elections and Democratization
in Latin America, 1980–1985. La Jolla, CA: University of California-San Diego,
Center for International Studies.
Thies, Michael F. (2002). Changing how the Japanese vote: The promise and pitfalls
of the 1994 electoral reform. In John Fuh-Sheng Hsieh and David Newman (eds),
How Asia Votes. New York: Chatham House.
Thomas, Craig W. (1998). Maintaining and restoring public trust in government
agencies and their employees. Administration & Society 30(2), 166–193.
Thomassen, J. (1994). Special issue, The intellectual history of election studies.
European Journal of Political Research 25 (3).
Thomassen, J. (2000). From comparable to comparative electoral research. In J. van
Deth, H. Rattinger, and E. Roller (eds), Die Republik auf dem Weg zur Normalita
¨t? Wahlverhalten und Politische Einstellungen nach Acht Jahren Einheit. Opladen:
Leske þ Budrich.
Thomassen, J. (ed.) (2005). The European Voter. Oxford: Oxford University Press.
Timpone, R. J. (1998). Structure, behavior, and voter turnout in the United States.
American Political Science Review 92(1), 145–158.
References 231
Tocqueville A. D. (1994). Democracy in America. New York: Knopf:
Topf, Richard (1995). Electoral participation. In H. Klingemann, and D. Fuchs (eds),
Citizens and the State. Oxford: Oxford University Press.
Tripp, A. M., A. Kang (2008). The global impact of quotas on the fast track to increased
female legislative representation. Comparative Political Studies 41, 338–361.
Tsai, Chia-hung (2001). Why do Taiwanese vote? Journal of Electoral Studies 8(2),
125–158.
Tucker, Joshua (2001). Economic conditions and the vote for incumbent parties in
Russia, Poland, Hungary, Slovakia and the Czech Republic from 1990–1996. Post-
Soviet Affairs 17(4), 309–331.
Tucker, Joshua (2006). Regional Economic Voting: Russia, Poland, Hungary, Slovakia
and the Czech Republic, 1990–1999. Cambridge: Cambridge University Press.
Turner, A. W. (1993). Postauthoritarian Elections. Comparative Political Studies 26, 3.
Uhlaner, C. J., B. E. Cain, and D. R. Kiewiet (1989). Political participation of ethnic
minorities in the 1980s. Political Behavior 11(3), 195–231.
Van Biezen, Ingrid, Peter Mair, and Thomas Poguntke (2012). Going, going … gone?
The decline of party membership in contemporary Europe. European Journal of
Political Research 51, 24–56.
van der Brug, Wouter, Mark Franklin, and Gabor Toka (2008). One electorate or
many? Differences in party preference formation between new and established
European democracies. Electoral Studies 27(4), 589–600.
van der Meer, Tom W. G. (2017). Political trust and the “crisis of democracy”. Oxford
Research Encyclopedia of Politics.
Van Deth, J. (1989). Interest in politics. In M. K. Jennings and J. van Deth (eds),
Continuities in Political Action. Berlin: De Gruyter.
Van Egmond, M. (2003). Rain falls on all of us (but some manage to get more wet
than others): Political context and electoral participation. Doctoral dissertation,
University of Amsterdam.
Van Egmond, De Graafi and Van Der Eijk (1998). Electoral participation in the
Netherlands: Individual and contextual influences. European Journal of Political
Research 34, 281–300.
Verba, S., and N. H. Nie (1972). Participation in America. New York: Harper &
Row.
Verba, S., N. H. Nie, and J.-O. Kim, (1978). Participation and Political Equality:
A Seven-nation Comparison. Chicago, IL: University of Chicago Press.
Verdery, Katherine (1996). What Was Socialism, and What Comes Next? Princeton,
NJ: Princeton University Press.
Warren. M. E. (2004). What does corruption mean in a democracy? American Journal
of Political Science 48(2), 328–343.
Wass, H. (2007). The effects of age, generation and period on turnout in Finland 1975–
2003. Electoral Studies 26(3), 648–659.
Wattenberg, Martin P. (1991). The Rise of Candidate-centered Politics: Presidential
Elections of the 1980s. Cambridge: Harvard University Press.
Webb, P. D. (1995). Are British Political Parties in Decline?. Party Politics 1,
299–322.
Webb, Paul, David Farrell, and Ian Holliday (eds) (2002). Political Parties in Advanced
Industrial Democracies. Oxford: Oxford University Press.
Webb, Paul, and Stephen White (2007). Political Parties in New Democracy.
232 References
Wedeman, Andrew (2002). Development and corruption: The East Asian paradox.
In Edmund Terence Gomez (ed.), Political Business in East Asia. London, UK:
Routledge.
Welzel, C., and Dalton, R. J. (2015). From allegiant to assertive citizens. In R. J. Dalton
and C. Welzel (eds), The Civic Culture Transformed: From Allegiant to Assertive
Citizens. Cambridge, UK: Cambridge University Press.
Weyland, K. (2003). Economic voting reconsidered: Crisis and charisma in the election
of Hugo Chávez. Comparative Political Studies 36(7), 822–848.
Whiteﬁeld, Stephen, and Robert Rohrschneider (2009). Understanding cleavages in
party systems issue position and issue salience in 13 post-communist democracies.
Comparative Political Studies 42(2), 280–313.
Wirls, D. (1986). Reinterpreting the gender gap. Public Opinion Quarterly 50, 316–330.
Wofinger, R. E (1965). The development and persistence of ethnic voting. The
American Political Science Review 59(4), 896–908.
Wolfinger, R. E. and Rosenstone, S. J. (1980). Who Votes? New Haven: Yale University
Press.
Yadav, Yogendra (2000). Understanding the second democratic upsurge: Trends of
Bahujan participation in electoral politics in the 1990s. In Francine Frankel, Zoya
Hasan, Rajeev Bhargava and Balveer Arora (eds), Transforming India: Social and
Political Dynamics of Democracy. Oxford: Oxford University Press.
Young, D. J. (2009). Is clientelism at work in African elections? A study of voting
behavior in Kenya and Zambia. Afrobarometer Working Paper 106.
Zakaria, Fareed (1994). A conversation with Lee Kwan Yew. Foreign Affairs 73,
109–127.
Zhengxu, Wang, and Leihua Ye (2017). Political participation in East Asia. In T.
Cheng, and Y. Chu (eds), Routledge Handbook of Democratization in East Asia.
London: Routledge.
Zmerli, S., and Newton, K. (2016). Objects of political and social trust: Scales and
hierarchies. In S. Zmerli and T. W. G. van der Meer (eds), Handbook on Political
Trust. Cheltenham, UK: Edward Elgar Publishing.
Index
Note: Page numbers in italics indicate figures and in bold indicate tables on the
corresponding pages.
Aarts, K. 126 retrospective evaluations in 132–133,

Abramson, P. R. 34 133; voice choice model and 136
Achen, C. H. 48 American National Election Studies 69
Acton, C. 80 American Voter, The 116
Africa, electoral participation in Andrews, J. T. 52
46–47, 50, 51–55, 55; datasets on 181; Anetta, C. 49
descriptive statistics for alternative Asia, electoral participation in 48,
variables in 207; descriptive statistics 50, 51–55, 55; datasets on 181;
for mobilisation variables in 202; descriptive statistics for alternative
descriptive statistics for political- variables in 207; descriptive statistics
institutional variables in 208; for mobilisation variables in 202;
descriptive statistics for political- descriptive statistics for political-
psychological variables in 205; impact institutional variables in 208;
of mobilisation variables on vote descriptive statistics for political-
choice in 203; impact of political- psychological variables in 205; impact
psychological variables on vote choice of mobilisation variables on vote
in 206; impact of socio-demographic choice in 203; impact of political-
variables on vote choice in 201; psychological variables on vote choice
levels of party identification and in 206; impact of socio-demographic
98–101, 99–100; questions wording variables on vote choice in 201;
on 183–193; summary descriptive levels of party identification and
statistics of variables and 200; 98–101, 99–100; questions wording
variable descriptions and coding on 183–193; summary descriptive
schemes on 194–199; see also regional statistics of variables and 200;
comparisons variable descriptions and coding
age 15–17, 72, 85, 86, 88–90, schemes on 194–199; see also regional
89–90, 93 comparisons
Aldrich, J. H. 34 associational membership 101–102, 102
Almond, G. A. 28, 31–32, 35 authoritarian practices and tendencies
alternative explanations and the vote: 134–135, 134–136; bivariate
assessing the scale of corruption in relationships and 137, 137–138
130–132, 131; authoritarian practices
and tendencies in new democracies Banducci, S. A. 34
and 134–135, 134–136; bivariate Barnes, S. H. 50
relationships in 137, 137–138; Beauregard, K. 17
economic concerns and voters’ Beck, M. 33–34
234 Index
Becker, R. 34 variables in 171; socio-demographic
Beer, S. H. 32 variables in 166–170, 167
Bengtson, V. I. 16–17 causal sequence 80
Bennett, S. E. 2, 31 Chang, E. C. 35
Bermeo, N. 134 Cheeseman, N. 97
Bernhagen, P. 50, 127 Cheibub, J. A. 6
Bhavnani, R. 46 Cheng, T.-J. 52
Birch, S. 37, 38–39, 44 Chesnut, R. A. 19
bivariate relationships: in corruption, Chin, J. U.-H. 20
authoritarianism, economy, and Chu, Y.-H. 35, 52, 97
the vote 137, 137–138; electoral Civic Culture, The 31
context and the vote 153–155, 154; civic cultures 31–33
in party identification, association civic skills 22–23, 28
membership, clientelism, and voting civil society organisations 28–29
101–102, 102; between political- clientelism 29–30, 30, 101–102, 102;
psychological variables and the vote correlates of vote choice and 110–113,
120, 120–121; in voter turnout and 111–112; re-considering the impact of
socio-demographics 84–87, 86 voting and, in new democracies
Black, J. 37 109–110; scope of, in new
Blais, A. 15, 18, 37, 38, 39, 51, 69, democracies 110
116, 126 closeness of elections 39–40, 40
Blaydes, I. 25 Coleman, J. 149
block-recursive sequence 80 Comparative Study of Electoral System
Boix, C. 6, 60 (CSES) 66–68, 179
Brady, H. 33, 69 compulsory voting 37, 38–39
Bratton, M. 19, 24–25, 41, 46, 54–55 concurrent elections 39
Brockington, D. 69 context see electoral or political-
Broughton, D. 19 institutional context
Bühlmann, M. 16, 34, 39 Converse, P. E. 116, 176
Bunce, V. 59 corruption 130–132, 131, 137,
Burns, N. 18 137–138
Bustikova, L. 45–46, 53 Cox, G. W. 39, 40
Cox, M. 35
Campbell, A. 27, 34, 116, 149, 176, Craig, S. C. 34
180 Crewe, I. 37
Carlin, R. E. 45, 47, 48, 116 Croke, K. 25
Carmines, E. G. 149 Curtice, J. 63–64, 67
Carothers, T. 41 Cutler, N. E. 16–17
Carreras, M. 21, 47, 50
Carter, J. R. 39 Dalton, R. J. 97, 119
Carty, K. 37, 38 Daoust, J.-F. 116, 126
Castañeda-Angarita, N. 21, 47, 50 democracy: characteristics of electoral
causal recursive model: alternative 1–2; definition of 1; elections
variables in 171–172; consistency as central to 40–43; electoral
across four regions and 166; electoral participation research and 3–5; in
participation in new democracies new-democracies 5–6; substance of
and 165–166; established models 42–43; voting as necessary for 2–3
working across all regions and democratic consolidation 6
173–174; explanatory variables as democratic erosion 134–135,
most “consistently effective” and 134–136, 172
174; mobilisation variables in 170; democratisation process 40–43
political-institutional variables in Diamond, L. 1, 5–6
172–173; political-psychological Diwakar, R. 12
Index 235
Downs, A. 14 33–34; political interest and political
Duverger, M. 18 knowledge and 33; religion and 19–20;
research on 3–5, 12–13; social capital
economic concerns and the vote theory and 43–44; socio-demographic
132–133, 133, 137, 137–138 model of 15–21; socioeconomic
Edgell, A. B. 41 factors in 21–26, 25; theories of 13–15,
education 91–92, 94–95 44–46; voter turnout trends in new
efficacy, political 33–34 democracies 147–150, 148
electoral authoritarian regimes 60 electoral systems 38
electoral fallacy 41–42 Ellis, A. 42, 43
electoral or political-institutional end of history 48
context 146–147, 163–164; bivariate end of the transition paradigm 41
relationships in 153–155, 154; Estrada, J. 20
closeness of elections in 39–40, ethnicity or race 20–21, 73–74, 85–87,
40; compulsory voting in 38–39; 89–90, 91, 94
concurrent elections in 39; electoral Europe, electoral participation in
system in 38; institutional factors 48–50, 51–55, 55; datasets on 182;
in 38–40, 40; multivariate model of descriptive statistics for alternative
155–163, 157–162; new democracies variables in 207; descriptive statistics
and their political-institutional for mobilisation variables in 202;
characteristics and 150–153, 151–152; descriptive statistics for political-
survey of institutional explanations institutional variables in 208;
in 36–37; voter turnout trends in new descriptive statistics for political-
democracies and 147–150, 148; why psychological variables in 205;
combine individual and contextual impact of mobilisation variables
approach and 147 on vote choice in 204; impact of
electoral participation: in Africa, political-psychological variables on
Latin America, East Asia, and post- vote choice in 206–207; impact of
communist European countries, socio-demographic variables on vote
study of 46–51, 176; age and 15–17; choice in 201–202; levels of party
alternative explanations of voting identification and 98–101, 99–100;
in new democracies and 51–55, 55; questions wording on 183–193;
civic cultures and 31–33; civil society summary descriptive statistics
organisations and 28–29; clientelism of variables and 200; variable
and 29–30, 30; consistency of models descriptions and coding schemes
of new democracies’ 165–166; on 194–199; see also regional
contending theories on voter turnout comparisons
and 13; democratic theories, founding
elections, social capital, and electoral fallacy of electoralism 41
turnout in new democracies and Farrell, D. 38
40–43; direction for future research Filer, J. E. 39
on 178–180; electoral or political- Fisher, J. 28
institutional context in 36–40; Flores, T. E. 41
ethnicity or race and 20–21; gender founding elections 41, 43, 49
and 17–19; interest in study of 1–3; Freitag, M. 16, 34, 39
limitation and practical implications Fujimori, A. 53
of the main findings on 177–178; Furnham, A. 69
mobilisation model of 26–27; party
identification and 27–28; place Gallego, A. 23, 50, 146, 169
of residence and 21; political and Geertz, C. 32
social trust and 34–36, 36; political Gélineau, F. 54
culture/political-psychological model gender 17–19, 72, 85, 86, 89–90,
and 30–31; political efficacy and 90–91, 93
236 Index
Gidengil, E. 18 Latin America, electoral participation
Glasure, Y. U. 106 in 47–48, 50, 51–55, 55; datasets
Gunter, B. 69 on 181; descriptive statistics for
Gurin, G. 27 alternative variables in 207; descriptive
statistics for mobilisation variables
Hadenius, A. 41 in 202; descriptive statistics for
Hadjar, A. 33–34 political-institutional variables in
Hansen, J. M. 26 208; descriptive statistics for political-
Harder, J. 14, 28 psychological variables in 205; impact
Harkness, J. A. 61 of mobilisation variables on vote
Hayes, B. C. 18 choice in 203; impact of political-
Healey, J. F. 79 psychological variables on vote choice
Hicken, A. 29 in 206; impact of socio-demographic
Hillygus, D. S. 23 variables on vote choice in 201;
Hoffmann-Martinot, V. 21 levels of party identification and
How Asia Votes 48 98–101, 99–100; questions wording
Hseih, F.-S. 48 on 183–193; summary descriptive
Huckfeldt, R. 23, 149 statistics of variables and 200;
Huntington, S. 1, 5–6 variable descriptions and coding
schemes on 194–199; see also regional
income 24 comparisons
Inglehart, R. 18 Latin America Voter, The 48
Inman, K. 52 Lazarsfeld, P. F. 33
interest, political 33 Lee, A.-R. 106
International Social Survey Programme Levitsky, S. 59, 60
(ISSP) 66–68 liberal democracies 46
Isaksson, A.-S. 24, 46–47 life-cycle events 16–17
Lindberg, S. I. 41, 107, 134
Jackman, R. W. 37 Listhaug, O. 34
Jankowski, T. B. 16 Lührmann, A. 134
John, H. 49 Lust, E. 134
Johnston, R. 34 Lyne, M. 29
Junn, J. 23–24
Maggiotto, M. A. 34
Karp, J. A. 21, 34, 49, 60, 69 Mainwaring, S. 97
Kasara, K. 25 Mannheimer, R. 51
Keefer, P. 29 Marsh, M. 50, 127
Kenny, L. W. 39 McAllister, I. 18
King, A. 27 Michigan model 48
Kitschelt, H. 29 Milazzo, C. 49
knowledge, political 33 Mill, J. S. 60
Kostadinova, T. 51, 52 Millard, F. 41
Kostelka, F. 43, 49 Miller, A. H. 28, 34
Kramon, E. 29 Miller, M. K. 6, 60
Krosnick, J. A. 14, 28 Miller, W. E. 27, 116, 176
Kuenzi, M. 46, 171 Milner, H. 38
mobilisation model 26–27, 97–98,
Ladner, A. 38 113–114; associational membership,
La Due Lake, R. 23 clientelism, and voting in 101–102,
Lambright, G. M. S. 46, 171 102; correlates of clientelism and vote
Langston, J. 97 choice in 110–113, 111–112; general
Lanning, K. 2 overview of multivariate results on
LaPiere, R. T. 69 102–109, 103–105, 108; levels of party
Index 237
identification in new democracies and Ordeshook, P. C. 14, 15
98–101, 99–100; re-considering the orientations 32–33
impact of clientelism and voting in Orvista, M. 49
new democracies in 109–110; scope
of clientelism in new democracies and Pacek, A. C. 54
110; variables in 74–75, 170 Participation in America 22
mobilisation variables in regional participatory culture 31–32
comparisons 74–75 party identification 27–28, 97–98;
multivariate analysis: individual and associational membership, clientelism,
political-institutional determinants and voting and 101–102, 102; in new
and the vote 155–163, 157–162; democracies 98–101, 99–100
political-psychological variables and Pattie, C. 34
the vote 121–128, 122–125, 128; of Peiffer, C. 132
socio-demographic determinants People’s Choice, The 15
87–95, 89–90, 92; vote choice 138– Persson, M. 23, 29
144, 139–143; voter turnout 102–109, Peter, G. 146
103–105, 108 Pietsch, J. 60
Munger, M. C. 39, 40 place of residence 21, 74, 86, 87,
89–90, 91, 94
Nadeau, R. 18, 19, 47, 48, 54 Plutzer, E. 22
Nalebuff, B. 40 Podobnik, B. 51
Napel, H.-M. ten 19 political and institutional variables: in
Nevitte, N. 18 new democracies 150–153, 151–152; in
new democracies 5–6; age and voting regional comparisons 78–79
in 17; alternative explanations of political culture 30–31
voting in 51–55, 55; authoritarian political efficacy 33–34
practices and tendencies in 134–135, political interest 33
134–136; civil society organisations in political knowledge 33
29; clientelism in 29–30; consistency political-psychological variables 30–31,
of models of electoral participation 116–117, 129; bivariate relationship
in 165–166; democratic theories, between the vote and 120, 120–121;
founding elections, social capital, multivariate results on 121–128,
and electoral turnout in 40–43; 122–125, 128; in regional comparisons
education and voting in 25; electoral 75–77, 171; underlying new
participation in individual 46–51; democracies 117–119, 117–119
levels of party identification in political trust 34–36, 36
98–101, 99–100; political trust in Poole, K. T. 18
35; re-considering the impact of Posner, D. 20
clientelism and voting in Powell, G. B. 36–37
109–110; scope of clientelism in pre-adult socialisation model 23–24
110; socioeconomic factors in voter Putnam, R. D. 23, 34, 44
turnout in 24–25; and their political- Pye, L. W. 32
institutional characteristics 150–153,
151–152; theories of electoral race or ethnicity 20–21
participation and 44–46; underlying Ragin, C. 61
political-psychological orientations Ragsdale, L. 34
in 117–119, 117–119; voter turnout Ramos, F. 20
trends in 147–150, 148 reasons for electoral participation:
Newman, D. 27, 48 age 15–17; civic cultures 31–33;
Nichter, S. 30 civil society organisations 28–29;
Nie, N. H. 22, 23–24 clientelism 29–30, 30; ethnicity or race
Nooruddin, I. 41 20–21; gender 17–19; mobilisation
Norris, P. 18, 34, 44, 69 model 26–27; party identification
238 Index
27–28; place of residence 21; political social trust 34–36, 36
and social trust 34–36, 36; political socio-demographics: bivariate
culture/political-psychological relationships in 84–87, 86; in model
model 30–31; political efficacy of electoral participation 15–21;
33–34; political interest and political multivariate analysis of 87–95, 89–90,
knowledge 33; religion 19–20; 92; in regional comparisons 72–74,
socio-demographic model of 15–21; 166–170, 167
socioeconomic factors in 21–26, 25 socioeconomic variables: in regional
regional comparisons: alternative comparisons 74; in voter turnout
variables in 77–78, 171–172; 21–26, 25
comparative framework and sorting model 23–24
parameters for 59–62; Spearman’s rho 85
conceptualisation and Stehlik-Barry, K. 23–24
operationalisation of variables in Stepan, A. 6
68–70; consistency across regions Stokes, D. E. 27, 116, 176
and 166; data, measurement, and Stokes, S. C. 29
estimation strategy in 66–68; empirical Strate, J. M. 16
and estimation strategy in 79–80; Studlar, D. T. 18
established models working equally subject culture 31
across all regions and 173–174; Suryanarayan, P. 25
measuring independent variables Svolik, M. W. 5
in 70–72, 71–72; methodological
challenges associated with studies of Tabellini, F. 29
voting behaviour in 63–66, 64, 65; Tambe, E. B. 21, 25, 29, 46, 48, 171
mobilisation variables in 74–75, 170; Teixiera, R. A. 24
political and institutional variables in Teorell, J. 41
78–79; political-institutional variables Terry, L. K. 41
in 172–173; political-psychological Thies, M. F. 39
variables in 75–77, 171; scale of Third wave: Democratisation in the Late
corruption 130–132, 131; selection of Twentieth Century, The 5
cases for 62–63; socio-demographic Third Wave of Autocratisation 134
variables in 72–74, 166–170, 167 third wave of democratisation 1, 40–43,
religion 19–20, 72–73, 85, 86, 48–49, 134
89–90, 91, 93 Timpone, R. J. 35
Resource Model 22 Tocqueville, A. de 28
Riker, W. H. 14, 15 Torcal, M. 97
Roberts, N. C 2, 53 Transparency International Corruption
Rood, S. 20 Perceptions Index 130–132, 131
Rosato, S. 6 Trebbi, F. 29
Rose, R. 49, 132 triple transition 48
Rosenbaum, W. A. 32 trust, political and social 34–36, 36
Rosenstone, S. J. 24, 26 Tse-Hsin, C. 46
Rusk, J. G. 34 turnover rule 6
Schedler, A. 6 van der Meer, T. W. G. 116

Schmitt, H. 51 Van Egmond, M. 149
Schumpeter, J. A. 1 Van Ham, C. 23, 33, 69, 163
Shachar, R. 40 Verba, S. 22, 27, 28, 31–32, 35
Shin, D. C. 97 Verdery, K. 59
Singer, M. M. 29, 47, 48, 54 vote choice 110–113, 111–112;
Smets, K. 23, 33, 69, 163 multivariate model of 138–144,
social capital theory 43–44 139–143
social network centrality 23–24 Voting 15
Index 239
Voting and Democratic Citizenship in White, S. 98, 170
Africa 46 Who Votes? 24
Wirls, D. 17
Wahman, M. 41 Wolfinger, R. E. 20, 24
Waldner, D. 134 World Values Survey (WVS) 66–68,
Wang, T. Y. 48 117–119, 117–119, 179
Wattenberg, M. P. 97
Way, L. 59, 60 Yadav, Y. 24
Webb, P. 98, 170
Wedeman, A. 132 Zechmeister, E. 45–46, 47, 48, 53
Welzel, C. 119 Zeigler, L. H. 18
Wessels, B. 126 Zelaya, M. 53

Webpdf

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Webpdf

Uploaded by

Copyright:

Available Formats

Electoral Participation in Newly

Elvis Bisong Tambe is Senior Lecturer in the Department of Political Science

Health and Political Engagement

Theorizing Cultures of Political Violence in Times of Austerity

Interest Group Organisation in the European Union

Coalition Government as a Reflection of a Nation’s Politics and Society

Transparency and Secrecy in European Democracies

Climate Governance across the Globe

The Future of Political Leadership in the Digital Age

Electoral Participation in Newly Consolidated Democracies

Elvis Bisong Tambe

2 Why citizens vote  12

3 The comparative framework  59

4 Socio-​demographics and the vote  84

5 Mobilisation and the vote  97

6 Political-​psychological variables and the vote  116

7 Alternative explanations and the vote  130

8 The electoral context and the vote  146

9 A comparative overview  165

2.1 Overview of the expectation of the socio-​demographic model

6.1 Interest in politics  117

Despite my personal efforts in writing this study on electoral participation in

Purpose of the book

1. To see if there is something distinctive about voting in new democra-

▪ Countries that have experienced the end of authoritarian rule.

First, the countries referred to in this book as newly consolidated democ-

The plan of the book

Contending theories –​general features of the theories of

Theories of electoral participation

R = (B) (P) –​ C + D

Likelihood of voting = (Motivation to vote × Ability to vote) ÷ Difficulty

From this equation, it is clear that the decision of an individual of whether

There is considerable evidence to support the socioeconomic/​resource model

Table 2.1 Overview of the expectation of the socio-​demographic model on vote choice

Variables Expected effect for all regions

Civil society organisations

Variables Expected effect for all regions

Mobilisation Party identification +

back to their local communities to canvass local support. Thus, as pointed

H2a: There will be a positive association between party identification and

Political culture/​political-​psychological model

“Civic cultures” and political participation

Political interest and political knowledge

Political and social trust

Variables Expected effect for all regions

Political-​Psychological Political trust +

people’s decision to vote in Western democracies has so far enabled us to for-

The electoral or political-​institutional context

A survey of institutional explanations

The closeness of the election

Variables Expected effect for all regions

H4a: Disproportional electoral systems decrease the propensity to vote.

Democratic theories, founding elections, social capital, and electoral

Elections and the democratisation process: how

Social capital theory

Theories of electoral participation: are newly consolidated

Electoral participation in Africa, Latin America, East Asia, and

Alternative explanations of voting in new democracies

Variables Expected effect for all regions

Alternative model Corruption perception +

consider policy performance and government handling of unemployment and

H5a: Increase in corruption will raise the incentives of voters to engage

Summary and conclusion

2 Why citizens vote 12

3 The comparative framework 59

4 Socio-demographics and the vote 84

5 Mobilisation and the vote 97

6 Political-psychological variables and the vote 116

7 Alternative explanations and the vote 130

8 The electoral context and the vote 146

9 A comparative overview 165

2.1 Overview of the expectation of the socio-demographic model

6.1 Interest in politics 117

Contending theories –general features of the theories of

R = (B) (P) – C + D

Likelihood of voting = (Motivation to vote × Ability to vote) ÷ Difficulty

There is considerable evidence to support the socioeconomic/resource model

Table 2.1 Overview of the expectation of the socio-demographic model on vote choice

Political culture/political-psychological model

Political-Psychological Political trust +

The electoral or political-institutional context

Sub-Saharan Africa Benin

First, Curtice argues that there is a need to combine both micro-and

Territory/region 2019 Average Average Average Average Number of

Sub-Saharan Africa (SSA)

Region/countries Survey yeara Election Official Self-

Sub-Saharan Africa countries

Region/countries Survey yeara Election Official Self-

Socio-demographics determinants of the vote: bivariate relationships

Regions Socio-demographic variables Spearman’s

Socio-demographics determinants of the vote: multivariate analysis

Months since last election 0.20* (0.02) 0.19* (0.02)

Table 4.3 Theoretical matching of socio-demographic model on vote choice –based

Models/ Expected Western Sub- Latin East Asia Post-

Countries/regions Mean (% of party identification)

Countries/regions Mean (% of party identification)

Models/ Expected Western Sub-Saharan Latin East Post-

assess the extent of cross-national variation in the effect of the mobilisation

Re-considering the impact of clientelism and voting in new

Countries/regions V-Dem clientelism index

Countries/regions V-Dem clientelism index

Table 5.6 The effect of clientelism on vote choice in new democracies

In their ground-breaking book, The American Voter Campbell, Converse,

Underlying political-psychological orientations in new democracies

Figure 6.1 Interest in politics

Figure 6.2 Trust in government

Figure 6.3 Trust in people

Political-psychological variables and the vote: bivariate relationship

Table 6.1 Political-psychological variables and electoral participation in new democ-

Regions Political-psychological variables Spearman’s rho

Sub-Saharan Africa Political interest 0.08**