Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
i
When theNerds Go Marching In
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
ii
Prototype Politics:Technology-
Intensive Campaigning and the Data of
Democracy
Daniel Kreiss
Taking Our Country Back:e Craing
of Networked Politics om Howard
Dean to Barack Obama
Daniel Kreiss
Media and Protest Logics in the Digital
Era:e Umbrella Movement in
Hong Kong
Francis L.F. Lee and Joseph M. Chan
Bits and Atoms:Information and
Communication Technology in Areas of
Limited Statehood
Steven Livingston and Gregor
Walter- Drop
Digital Feminist Activism:Girls and
Women Fight Back Against Rape
Culture
Kaitlynn Mendes, Jessica Ringrose,
and Jessalynn Keller
Digital Cities:e Internet and the
Geography of Opportunity
Karen Mossberger, Caroline J.
Tolbert, and William W. Franko
Revolution Stalled:e Political
Limits of the Internet in the
Post- Soviet Sphere
Sarah Oates
Disruptive Power:e Crisis of the State
in the Digital Age
Taylor Owen
Aective Publics:Sentiment,
Technology, and Politics
Zizi Papacharissi
e Citizen Marketer:Promoting
Political Opinion in the Social
Media Age
Joel Penney
China’s Digital Nationalism
Florian Schneider
Presidential Campaigning in the
Internet Age
Jennifer Stromer- Galley
News on the Internet:Information and
Citizenship in the 21st Century
David Tewksbury and Jason
Rienberg
e Internet and Political Protest in
Autocracies
Nils B. Weidmann and Espen
Geelmuyden Rød
e Civic Organization and the Digital
Citizen:Communicating Engagement in
a Networked Age
Chris Wells
Computational Proaganda:Political
Parties, Politicians, and Political
Manipulation on Social Media
Samuel Woolley and Philip
N. Howard
Networked Publics and Digital
Contention:e Politics of Everyday
Life in Tunisia
Mohamed Zayani
Oxford Studies inDigital Politics
Series Editor:Andrew Chadwick, Professor of Political Communication in the
Centre for Research in Communication and Culture and the Department of
Social Sciences, Loughborough University
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
1
iii
When theNerds Go
Marching In
HOW DIGITAL TECHNOLOGY MOVED FROM THE
MARGINS TO THE MAINSTREAM OF
POLITICAL CAMPAIGNS
RACHEL K.GIBSON
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
1
iv
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.
Published in the United States of America by Oxford University Press
198 Madison Avenue, NewYork, NY 10016, United States of America.
© Oxford University Press 2020
Some rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmied, in any form or by any means,
for commercial purposes, without the prior permission in writing of
Oxford University Press, or as expressly permied by law, by license or under
terms agreed with the appropriate reprographics rights organization.
is is an open access publication, available online and distributed under the terms of a
Creative Commons Aribution – Non Commercial – No Derivatives 4.0
International license (CC BY-NC-ND 4.0), a copy of which is available at
hp://creativecommons.org/licenses/by-nc-nd/4.0/.
Inquiries concerning reproduction outside the scope of this license
should be sent to the Rights Department, Oxford University Press, at the address above
Library of Congress Cataloging-in-Publication Data
Names: Gibson, Rachel, 1968– author.
Title: When the nerds go marching in : how digital technology moved from the margins
to the mainstream of political campaigns / Rachel K. Gibson.
Description: New York, NY : Oxford University Press, [2020] |
Series: Oxford studies in digital politics | Includes bibliographical references and index.
Identiers: LCCN 2020006842 (print) | LCCN 2020006843 (ebook) |
ISBN 9780195397789 (hardback) | ISBN 9780195397796 (paperback) |
ISBN 9780190949044 (epub) | ISBN 9780190949051 (online)
Subjects: LCSH: Internet in political campaigns—Cross-cultural studies.
Classication: LCC JF2112. C3 G53 2020 (print) | LCC JF2112 .C3 (ebook) |
DDC 324. 70285/4678—dc23
LC record available at hps://lccn.loc.gov/2020006842
LC ebook record available at hps://lccn.loc.gov/2020006843
9 8 7 6 5 4 3 2 1
Paperback printed by LSC Communications, United States of America
Hardback printed by Bridgeport National Bindery, Inc., United States of America
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
v
v
Contents
Acknowledgments vii
Introduction 1
1. The Four- Phase Model of Digital Campaign Development 7
2. A Review of the Literature:From Experimentation to
Mobilization 25
3. Digital Campaigning across Space:The Role of Technological,
Political, and Institutional Context 44
4. The Slow Burner:Digital Campaigning in the United Kingdom 72
5. The Early Bloomer:Digital Campaigning in Australia 107
6. The Late Bloomer:Digital Campaigning in France 134
7. The Trendsetter:Digital Campaigning in the United States 168
Conclusion:Digital Campaigns at the Crossroads 210
Appendices 225
Notes 269
References 285
Index 303
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
vi
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
vii
vii
Acknowledgments
is book was wrien over a number of years, as my very patient publisher
Angela Chpnako and series editor Andrew Chadwick can aest to. While its long
gestation period means that some of the insights it oers are somewhat “late to
market,” it has also given the book a richer historical perspective and hopefully
a longevity that will continue to serve scholarship in the future. In particular,
what has become apparent from taking this “longer view” is how truly humble
the origins of the modern digital campaign were. Only by seeing that initial ama-
teurism and exuberance can one fully appreciate the very rapid pace of advance
that has occurred in recent years. e “electronic billboards” of yesteryear that
won awards simply by having revolving ribbons now dwindle into uer insigni-
cance, compared with the digital behemoths that have taken over the heart of
campaign headquarters. Another unintended but useful consequence of its time
in preparation is the condence with which it can state its core conclusion. e
intertwining of digital technologies is now so pervasive within campaign opera-
tions that talk of “the” digital campaign as a stand- alone entity is anachronistic
to the point of sounding quaint.
In undertaking this project Ihave drawn on the inspiration and intellectual
stimulation of many colleagues— too many to name personally here, but Ithank
each and every one of them for the nuggets of wisdom they contributed to the
production of this volume. Perhaps Ican start where it all began and thank my
coauthor (of other works) and former Salford colleague and friend Steve Ward,
without whom Ithink Imight have actually missed the internet boat. at con-
versation in Crescent House canteen certainly bore a lot of academic fruit. Ihave
also had the great joy and privilege of regularly brainstorming ideas about digi-
tal political communication with my long- time coauthor (of other works) and
incredibly dear (and stylish) friend Andrea Rommele; Ilook forward to toast-
ing the book’s publication in our time- honored tradition. e roots of the book
itself took hold initially with the commencement of my Economic and Social
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
viii Acknowledgments
viii
Research Council (ESRC) professorial fellowship back in the early years of the
previous decade. rough that project (“e Internet, Electoral Politics and
Citizen Participation in Global Perspective,” RES- 051- 27- 0299), Iwas fortu-
nate to develop personal and professional relationships with a number of highly
inspiring scholars based at the University of Manchester and elsewhere. Iwas
particularly lucky in being able to persuade Marta Cunhill- Cantijoch to leave
the sunny shores of Barcelona for the griier and greyer landscape of the Oxford
Road corridor. Marta’s skills, patience, support, and dedication to that project
was immense, and Iowe her a huge debt of gratitude for the key role she played
in its successful completion. Ihave continued to learn from her as a colleague and
scholar and look forward to many years of doing so in the future. Others during
that time who have played a key role in shaping my thinking about digital cam-
paigns and the wider political context in which they are conducted include my
friends and coauthors (of other works)— John Aldrich, Bruce Bimber. Fabienne
Greet, Karolina Koc- Michalska, Jan Leighley, Ian McAllister, Sarah Oates, and
Misha Taylor- Robinson. Finally, Iowe a special thank you to my PhD students,
all of whom have prompted me to reect on and update my assumptions about
political change in the digital era, as well as my understanding of the best data
and methods for measuring our object of study. Aparticular shout- out goes to
Ros Southern for her help in coauthoring Chapter3 of this book and invaluable
support in studying the 2015 and 2017 UK elections. Alast word of heartfelt and
loving thanks must go to my husband and intellectual sparring partner Paul. e
homemade cappuccinos, copious cups of tea, and fresh pasta he prepared to help
sustain me during the long and particularly fraught nal days to nish this book,
and the hours of his life he gave to proofreading, will always be appreciated. e
walks we have shared through the woods to clear the cobwebs, never forgoen.
Rachel Gibson
Department of Politics and e Cathie Marsh Institute for Social Research,
University of Manchester
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
1
When theNerds Go Marching In. Rachel K.Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
Introduction
is book charts the increasing role and centrality of the internet within election
campaigns across established democracies, following its electoral debut in the
mid- 1990s. It does so by presenting a four- phase model of digital campaigns that
charts the movement of the technology from the margins to the mainstream of
party operations. Historically it reveals how the new medium shied from being
a mere novelty item to a basic necessity for any candidate or party now seriously
contemplating a run for political oce. It does so by combining a systematic
review of the extant literature with a range of secondary and original data sets
to present a 20- year overview of the evolution of internet- based electioneering.
rough extensive analysis of both large N and also more focused case studies of
the United Kingdom, the United States, France, and Australia, it reveals how the
four phases developed in dierent contexts, and highlights some of the reasons
for the varying evolutionary paerns that are observed.
While it is dicult to pinpoint exactly when the rst “cyberspace” campaign
was ocially launched, the general consensus is that the breakthrough moment,
at least in terms of public awareness, came during the 1992 US election cycle
(Bimber and Davis, 2003; Davis and Owen, 2008; Janda 2015). At the presiden-
tial level, it was Democrat nominee Bill Clinton who laid claim to this virtual
terra nova aer his sta uploaded a series of basic text les with biographical
information for voters to browse. Since that time, use of the internet in elections
has expanded dramatically in the United States and elsewhere.
As well as increasing in overall volume and visibility around the world, digital
campaigning has grown in stature and strategic importance over time. It is this
process of evolution and maturation that forms the focus of this book. In par-
ticular, it is argued that digital campaigning in established democracies has pro-
gressed through four main phases to date— experimentation, standardization,
community- building, and direct voter mobilization— although, as shall become
clear, not all countries have experienced the full cycle. Movement through these
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
2WHEN THE NERDS GO MARCHING IN
2
phases, we argue, has placed digital technology center stage in election manage-
ment and has changed the “art” of campaigning into something more of a sci-
ence. From being lile more than an aerthought for most parties 20years ago,
the technology now sits at the core of a nely tuned “get out the vote” (GOTV)
machine.
In addition to profoundly changing the way that campaigns operate, this
book argues that the growing use of digital technology is also transforming their
internal structure. While there were hopes that the new media would lead to the
devolution of power to members and supporters, who would exert a stronger
“co- production” role in running the campaign, more recent developments have
countered these expectations. From the mid- 2010s we have seen a new digi-
tally literate elite— comprising data analysts and soware engineers— start to
emerge at the apex of campaign organizations. ese new sta, some of whom
are imported directly from the tech world, exercise an increasing inuence over
key decision- making tasks.
us, the central conclusion of the book broadly endorses the long- standing
argument that the internet has “normalized” electoral politics by reinforcing the
power of the main parties and those who run them. However, it subverts and
reformulates that narrative to an extent. Rather than understanding this process
as one in which seasoned national “politicos” and consultants exert ever more
power over local- level underlings, instead we see a reconguration or changing
of the guard at the top as a new set of nontraditional apoliticos move in to run the
show. Such individuals are notable for their lack of experience of being in the
eld during elections, and concomitantly greater immersion in “big” data and
algorithms to remotely model and predict voter behavior. Furthermore, given
the resource implications of the new “data- driven” mode of digital campaigning
for political parties, the smaller players face an even greater challenge to compete
with their bigger rivals. is leads to the emergence of a new and even more
unbalanced communications playing eld— an environment we describe as one
of hypernormality.
In the chapters that follow, we locate these developments in the longer cycle
of digital campaigning. We begin with some conceptual groundwork. Chapter1
makes the case for the four- phase model of digital campaign development and
describes the key characteristics of each phase. It does so by seing out the two
main logics of equalization and normalization that have shaped the study of
e- campaigns to date, and shows how the rotation between them forms a four-
stage chronology of change. In particular, phases Iand II are characterized by
the move from conditions of greater openness and decentralization toward more
“normalized” or unequal conditions. Phases III and IV largely repeat this pat-
tern, with the former characterized by the swing back to more balanced and plu-
ralized digital elections, and the laer by more emphatic return to normality, or
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
3
Introduction 3
hypernormality (i.e., the rise of the new technocratic elites and further concentra-
tion of power in the party system). In addition to this cycle of power redistribu-
tion, each phase is also distinguishable according to a set of unique technical
and strategic aributes. e chapter describes each of these aributes and the
particular features of each phase in more detail.
Chapter2 presents a review of the empirical literature on the topic of digi-
tal campaigns. In particular, we show how the past two decades of scholar-
ship, when joined together, form a narrative of change that aligns closely to
the four- phase model set out in Chapter1. Specically, we report how most
early studies of digital campaigns, in a variety of seings, point to the period
of open and even naïve experimentation that is phase I.We then show how
subsequent analyses typically record a shi into more managerial approaches
to web campaigning in which major parties dominate and standardized
web communication tools emerge (phase II). Next we reveal how the nar-
rative changes again, returning (albeit briey) to adopt a more optimistic
tone that highlights the interactive and community- building potential of the
new web 2.0 tools (phase III). Finally, we document how the narrative of
more recent studies points toward a reassertion of normalization, as central
elites assert themselves and major parties dominate (phase IV). We argue,
however, that the new tools required to run these data- intensive campaigns,
along with the new skill set and apolitical background of those recruited to
deliver the campaign, have added a signicantly dierent twist to the earlier
reinforcement logic.
As well as providing a more “joined- up” historical narrative of the study of
digital campaigns, Chapter2’s review produces a more detailed picture of each
phase and a clearer idea of how the model applies “on the ground.” Finally,
Chapter2 oers some preliminary evidence as to where certain countries now
sit in the evolutionary cycle. In particular, we can see where the fourth and most
advanced mode of digital campaigning, which focuses on direct voter mobiliza-
tion, appears to be most developed.
Chapter3 follows up on the impressionistic conclusions of Chapter 2 to
take a more empirical and systematic look at where developments in digital
campaigning are most and least advanced globally. Using data from a recent
cross- national survey, the chapter compares rates of online voter mobilization
among 19 democracies and divides countries into four tiers of activity. ese
rankings are then systematically explored in a multilevel analysis that includes a
range of contextual and individual- level explanatory factors. e ndings reveal
a signicant degree of variance among countries’ levels of online mobilization,
with some usual suspects conrmed as leaders in this regard, as well as some
unexpected nations emerging as strong performers. Overall the analysis indi-
cates that political institutions and technological diusion maer in terms of
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
4WHEN THE NERDS GO MARCHING IN
4
driving forward this new mode of election contact and the wider cycle of digital
campaigning.
e last set of chapters— Chapters4 through 7— build on and extend the
insights from Chapter3 by taking an in- depth look at developments in digi-
tal campaigning in four countries that were included in the large N analysis of
Chapter3:the United Kingdom, Australia, France, and the United States. e
selection of cases is driven by pragmatic considerations to a large extent. Each of
the four nations scrutinized has featured prominently in the empirical literature
reviewed in Chapter2. As such, they provide the richer historical narrative nec-
essary to track changes in the supply and demand for digital campaigning over
time. From a theoretical perspective, however, these countries also provide an
interesting mix of the institutional characteristics that are likely, ceteris paribus,
to aect the pace of online campaign innovation and voter mobilization eorts.
First, in terms of their electoral systems, all four countries rely primarily on
single- member and plurality/ majority methods to elect their national politi-
cians. is is signicant in that such systems tend to produce a more personal-
ized or “candidate- centered” style of election campaigning (Carey and Shugart,
1995; Dalton etal, 2011). is is a factor that some scholars have linked with
more intensive and advanced forms of web campaigning (Ziel, 2015; Anstead
and Chadwick, 2009). France and the United States stand out further in that
they conduct regular presidential elections. ese are typically high- prole
mobilizing events, and one might expect that the greater levels of oine cam-
paign intensity they generate would spill over to the online environment and
spur on the use of digital tools by candidates. Finally, in terms of party system
size, there are also dierences between the cases that lead to some interesting
and mixed predictions about the pace of innovation. e United States consti-
tutes the “ideal” two- party environment, while Australia, the United Kingdom,
and France constitute a range of increasingly multiparty environments. While
a smaller party system may accelerate the process of adoption, as two well-
resourced parties compete for the “median” voter, a larger party system might
also stimulate change as the technology opens up new possibilities for smaller
players to compete.
Based on this very simple overview of our cases, therefore, one might expect
to nd the United States riding the crest of the digital wave, with France not
far behind. Australia and the United Kingdom would be more likely to bring
up the rear. is ordering may change, however, across the phases of develop-
ment. In particular, in the initial stages of experimentation when the technol-
ogy is cheaper and more accessible, a multiparty environment is likely to see a
faster pace of adoption. As resources become more important, however, systems
dominated by larger parties promote faster adaptation, particularly during the
transition into the more scientic and data- intensive phase IV methods.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
5
Introduction 5
e mix of comparative analysis and case studies that this book provides
allows us to examine these propositions in more depth. Specically, we can rst
systematically test the impact of regime- level characteristics and socio- structural
factors on the intensity of the later stages of digital campaigning across countries
in Chapter3. en, through the case studies we can look at the nature of prog-
ress within individual nations over time. is enables us to see how linear the
nature of developments are across the four phases and, if there are changes in
the pace of activity, whether this correlates with the characteristics of the party
system.
In addition to allowing for a closer look at the role of context in encouraging
or hampering the development of digital campaigns, the case studies develop
the analysis in several other ways. First, from a descriptive perspective, we can
describe the histories of digital campaign development in each nation in more
detail, and compare their progress using our four- phase model. Are their trajec-
tories more or less similar? What point has each country now reached within the
cycle? Second, as part of that narrative we can expose more clearly the role of
supply- side or organizational factors in the process of development. In particu-
lar, we are able to examine more closely the role that parties have played in driv-
ing incorporation of the new digital tools. Have the major parties consistently
led the way, or have minor parties played a key role? If so, which ones and when?
e third advantage the case studies provide is the opportunity to look in
greater depth at the movement across the four phases from “below” (i.e., from
perspective of the voters and party supporters). How have popular uses of the
technology changed during elections, and have they moved in step with elite
provision? Has the audience for the digital campaign become more activist
as the parties have shied to embrace more community- building initiatives?
Which parties have been more aligned with their supporters in this regard?
Finally, through close analysis of these four cases we can also dig deeper into the
drivers and impact of online voter mobilization examined in Chapter3. Who
is being reached by these new methods? New voters or the already mobilized?
What evidence exists to indicate that use of such tactics can make a dierence to
the election outcome in terms of shaping voters’ choices at the ballot box?
Before beginning our investigation of these questions, there is one last piece
of denitional groundwork that needs to be completed. e focus of this book
is on mapping and analyzing global trends in digital campaigning and how they
have culminated in the rise of a new campaign elite. It is important early on
that we specify what we mean when we refer to the “digital campaign.” What is
included and what is excluded? To date, a wide range of terms have been used
to describe this new form of electioneering, including labels such as “cyber,”
“internet,” “web,” “online,” and “e- campaigning.” W hile such terms do succeed in
denoting this distinctive new mode of campaigning, they also suer to varying
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
6WHEN THE NERDS GO MARCHING IN
6
degrees from being somewhat time- or technology- bound. Perhaps more impor-
tantly, on conceptual grounds these labels tend to reduce or narrow the foci of
study to a particular set of “objects” or platforms that parties have produced in
order to wage an electoral campaign. e emphasis is on a discrete and bounded
entity that is the online or web campaign— an outward interface or performative
construct that voters “see” and can engage with if they choose to.
By using the term “digital campaigning,” we deliberately seek to broaden the
concept out from a focus on the “front end” of operations to include the wider
range of the less visible activities, personnel, and infrastructure (both hardware
and soware) that lie beneath this outward exterior or “interface.” is more
expansive approach, we argue, is necessary to ensure that research on the topic
remains relevant and accurately reects how praxis has, and is, evolving. While
a vital part of a party’s digital campaign remains its outward- facing or pub-
lic “shopfront” (i.e., websites, email, and social media proles), this forms an
increasingly small tip of a much larger binary underpinning ecosystem. By using
the term “digital campaign” we thus encompass and document the changes in
this externally consumed component, as well the growing body of technological,
computing, data, and scientic expertise that underlies it.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
7
When theNerds Go Marching In. Rachel K. Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
1
The Four- Phase Model ofDigital
CampaignDevelopment
is chapter sets out the core argument of the book, which is that over the past
two decades digital technology has moved from the margins to the mainstream
of campaign operations, and in doing so has fundamentally changed how elec-
tions are fought and won. is process of transformation, we argue, can be bro-
ken down into four main phases of development:
Phase I :Experimentation
Phase II:Standardization and Professionalization
Phase III:Community Building and Activist Mobilization
Phase IV:Individual Voter Mobilization
In the sections that follow, we describe each of the phases in more detail, identify
their key dierences, and show how they cumulate in a transformative shi in
campaign practice. Before doing so, we reect more broadly on the structural
nature of the four- phase model, and how the evolution it describes compares
with earlier periods of media adoption by campaigns.
e idea of an initial phase of experimentation, followed by standardization
and professionalization in digital campaigning, would not seem to depart too
radically from an understanding of how parties have adapted to new commu-
nication technologies in the past, such as radio and television (Swaddle, 1988;
Selnow, 1998; Norris, 2001). We argue that the later phases of community build-
ing and individualized voter mobilization are more particular and indeed unique
to digital campaigns’ evolution. is distinctiveness is more immediately evident
in phase III. While the arrival of previous media may have enhanced activists’
voices and increased connectivity among the grassroots, the networked nature
of the internet and particularly social media exponentially increases the oppor-
tunities for this type of intra- organizational participation and communication.
For some it even leads to a new “co- production” or “citizen- initiated” model of
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
8WHEN THE NERDS GO MARCHING IN
8
campaigning that radically challenges (albeit briey) modern campaigns’ cen-
tralized management ( Johnson, 2011; Lilleker, 2013; Gibson, 2015).
With regard to phase IV, again the harnessing of new media to help mobilize
voters is hardly revolutionary in the annals of campaign development. However,
our argument here is that the level of precision and personalization that digital
tools and data bring to the process is far beyond what was possible for television
advertising and even direct mail. Voter targeting, as some scholars have argued,
moves from the “micro” to the “nano” level in terms of tailoring message con-
tent to the individual. is new scientic approach, along with the expertise this
brings to standard GOTV eorts, introduces a wholly new modus operandi to
mobilizing voters, and to the management of campaigns more generally.
e development of digital campaigning is also distinguished from earlier
periods of media adoption by its cyclical or rotational nature. Akey question
posed by scholars of internet campaigning since its inception has been whether
it has made electoral communication and party competition more “equalized”
or “normalized”? We explain these arguments in more detail in the following.
In brief, however, the former condition is one in which the networked and
decentralized structure of the internet is seen to promote the empowerment
of grassroots voices and to spread power to previously marginalized actors.
Normalization captures the response by established elites to constrain their
power loss, and to exploit the new medium to reinforce their dominance. e
ndings from this literature have shown that online election campaigning, in dif-
fering national contexts, swings back and forth between these two end states.
e rst such shi occurred as web 1.0 technologies entered the scene and
pushed practice in a more open direction. is was then followed quite rapidly
by a pushback by the major parties and a period of normalization. Asecond web
2.0 powered a move toward pluralization, which was again countered by an even
more emphatic reassertion of centralized elite control. is rotation between
these two power logics forms a key component of our four- phase model. eir
theoretical origins are discussed in greater detail in the next section of the chap-
ter. e overall four- phase model of change is summarized in Table1.1 later in
the chapter.
Theoretical Perspectives onthe Internet and
CampaignChange
While much of the evidence used to construct the four- phase model of digital
campaign development is inductive in nature and is based on a review of ndings
from the secondary literature, the dynamic element at its core draws on earlier
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
9
The Four-Phase Model 9
theoretical discussions about the future form of “e- democracy.” In particular, we
draw on the work of a group of scholars— the “e- pluralists”— who rst articu-
lated the two main power “logics” or end states of equalization and normaliza-
tion. ese arguments emerged from an early period of intense speculation and
visionary debate about the impact of the internet on modern political systems.
For a number of these early writers, the new wired world presented an oppor-
tunity to remake and revolutionize modern democracy. For some, the technol-
ogy would enable the return to older and purer forms of democracy in which
citizens could deliberate and make decisions without the interference of elected
ocials (Arterton, 1987; Schwartz and Oram, 1996; Dertouzos, 1997; Becker
and Slaton, 2000; Cli, 2000, 1997; Dahlberg, 2001). Communitarians such
as Howard Rheingold were particularly excited about how the internet would
“challenge existing political hierarchies” and enable a new system of “citizen-
based democracy” to emerge in which citizen power was the driving force
(2000:xxix). More libertarian- minded scholars were equally enthused, envi-
sioning how the internet would lead to a dismantling of the machinery of gover-
nance in favor of more direct forms of citizen rule (Budge, 1996; Dyson, 1997;
Grossman, 1995). Taken to its extreme, “being digital,” some argued, might well
lead to the entire “evaporation” of the nation- state (Negroponte, 1995).
Other scholars were equally radical in their predictions, although the out-
comes they foresaw were much less positive. Mark Poster (1995) and later Cass
Sunstein (2001) raised the specter of a post- democratic system dominated by
growing political division, isolation, and extremism. Similarly pessimistic sce-
narios were issued by those concerned about the growth of “Big Brother” style
government and the surveillance state. Under this new digital panopticon, dis-
sent would be silenced and citizen rights and liberties eroded (Barber, 1997,
1998; Street, 1997; Knopf, 1999; Akdeniz, 2000; Akdeniz etal., 2001; Lessig,
2000).1
While diering markedly in their understanding of what was to come next,
therefore, there was clearly broad agreement among these scholars that entry
into the internet age was likely to undermine the current system of represen-
tative democracy. Given this focus, it was not surprising that the debate made
virtually no reference to the future state of elections and political campaigns.
Such an exercise was essentially the intellectual equivalent of rearranging the
deck chairs on the Titanic. Amid these more apocalyptic scenarios, however,
there were a number scholars who took a more “institutionalist” approach to
understanding the impact of the new media and argued for a less technologically
determined “solution” to democracy’s current problems. For such writers, the
democratic state, while showing signs of wear and tear, was unlikely to be drasti-
cally reshaped or aened by the new wave of technological change. Instead, we
would see a process of more incremental reform and renewal.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
10 WHEN THE NERDS GO MARCHING IN
10
A “Third Way”? E- Pluralism and the“New
Jeffersonians”
Opting for this “third” way between democratic boom and bust were a num-
ber of largely US- based scholars who saw the internet as a pluralizing force that
would produce a more porous and open political system (Abramson etal., 1988;
Kapor, 1993; Browning, 1996; Corrado and Firestone, 1996; Bimber, 1998).
eir accounts were based primarily on what were seen as the unique structural
properties of the internet and its very low entry costs. Such factors, they argued,
rather than sweeping away current societal institutions, would actually put pres-
sure on them to become more responsive to citizen demands. Major players in
business, the traditional media, and government would inevitably face increas-
ing competition from smaller and previously marginalized groups in society.
is new interactive and decentralized communication network would open up
an entirely new space for mass production and consumption in which “size” no
longer maered.
e rapid proliferation of interests and increased competition arising from
the spread of internet use across society was likely to be particularly evident in
the political arena (Bonchek, 1995; Corrado and Firestone, 1996; Rash, 1997;
Bimber, 1998). According to Corrado and Firestone:
the most basic feature of this technology is that it will allow individuals
more easily to nd others who share their interests or views and com-
municate with them... this will mean the inevitable growth of new
political groups...many of which would transcend geographic or polit-
ical boundaries. (1996:12)
Similarly for Rash (1997), a likely consequence of the new “politics on the nets”
was “more groups” and “a greater number of views...helping [to] determine
policy” (178). Established organizations and institutions were also expected to
undergo major internal restructuring as existing management hierarchies came
under pressure from similar decentralist forces. According to Abramson etal.
(1988), the new media could help US political parties open up to more debate
and input from a wider range of “demographic, geographic, interest and issue
groups,” particularly during the nomination process (121).
Applying the logic of these “new Jeersonians” to the dynamics of the cam-
paign sphere, therefore, the most likely outcome was that elections would
become noisier and more crowded occasions in which smaller players would gain
greater prominence. It is also possible to envisage the soening, if not weaken-
ing, of the governance structures surrounding campaigns as a newly empowered
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
11
The Four-Phase Model 11
grassroots began to challenge the dominance of the professional consultants and
central elites who had held sway at the end of the twentieth century (Norris,
2000; Farrell and Webb, 2000).
While the establishment of this new “electronic commonwealth” was viewed
as a positive development by its proponents, they were ready to acknowledge its
potential downsides. Bimber (1998) in particular talked about the emergence of
a new form of “accelerated pluralism” in society. is intensied form of group-
centered politics, he argued, could lead to “a more fragmented polity dominated
by unstable issue publics” (156). For Rash (1997), such concerns were some-
what premature in that he worried more about the longer term sustainability of
the new liberalized polity. He argued that the resources required to keep pace
with technological developments would soon outstrip the capacity of smaller
groups and political “start- ups” to compete, leading to “the uneven spread of the
technology in tactical terms” (178).
Corrado and Firestone (1996) echoed these concerns, pointing out that a
fragmented media environment was likely to reduce electoral competition in the
longer term. e push toward developing ever more innovative niche- marketing
techniques would escalate the resources required to run an eective campaign
and price the smaller parties out of the electoral market. Internally, the new
trends would increase the “centralization of internal campaign organization,”
concentrating yet more power in the hands of senior political advisors and con-
sultants (108– 9).
us, rather than yielding to the advance of direct democracy or disappearing
under the weight of an Orwellian state, political campaigns, according to these
scholars, were likely to experience something of a renaissance as the internet era
took hold. Elections would become more open and accessible spaces, particu-
larly for previously minority and marginal players. is pluralistic environment,
however, was not likely to last very long. Time would see the major parties and
established elites reassert control and further reinforce their power base. is
predictive logic, as we shall see in the next chapter, has provided the interpreta-
tive framework for much of the subsequent study of digital campaigns. In partic-
ular, it has underpinned the notions of “equalization” and “normalization” that
have dominated the eld and informed its key conclusions.
Four Phases, Two PowerLogics
e model of campaign change set out in this book starts from the e- pluralists’
premise that the arrival of the internet prompts a shi toward more equalized
and pluralistic political conditions. is trend is then countered or normalized,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
12 WHEN THE NERDS GO MARCHING IN
12
as the established players begin to perceive the benets of the new technology
and deploy it to further their own interests. In the context of campaigns, this
means we expect the electoral system to become a more open and competitive
space as parties start to adopt tools like email and the web. is quickly changes,
however, as embedded political structures begin to see the benets of the new
tools, and preexisting power relationships re- emerge.
Taking this two- step process of campaign change, we make two further
theoretical tweaks that convert it into our four- phase model. First, rather than
viewing this process as a one- time transition from equalization to normaliza-
tion, we argue that it is more accurately regarded as a pendulum swing between
these two poles. With the benet of two decades of hindsight, it is clear that the
equalization to normalization cycle recurred following the entry of a new suite
of second- generation web tools around the turn of the millennium. As such, we
have at least four phases of development marking this ebb and ow to date, with
an initial period of web 1.0 equalization to normalization followed by a second
web 2.0– led cycle.
Our second “twist” to the e- pluralists’ logic is to extend it conceptually and
to specify more clearly where, how, and when the shi from equalization to nor-
malization takes place with regard to the key areas of campaign activity. In par-
ticular, we argue that movement between these two states occurs within three
distinct but interlinked domains of practice:
(i) systemic;
(ii) intra- organizational;
(iii) extra- organizational voter communication.
Equalization and normalization at the systemic level refer to increasing and
decreasing levels of inter- party competition. At the intra- organizational level,
the terms apply to the extent of power held by grassroots activists and supporters
versus central elites. Finally, with regard to voter communication, they describe
a situation in which campaigns either seek to exploit the participatory feature of
the medium and create a more inclusive dialogue with voters (equalization), or
one in which they revert to the familiar top- down broadcast model of informa-
tion dissemination (normalization). As the phases advance, we argue, the spaces
in which equalization and normalization take place, or at least are looked for,
extend from (i)to (iii). us in the move from phase Ito phase II, discussion
centers on the increase and decrease of inter- party competition. As we move on
to the later phases, the lens widens to examine how far the technology is promot-
ing grassroots empowerment versus centralized control, as well as boom- up
dialogue with voters or top- down targeted messaging. We discuss these trends
in more detail in the following.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
13
The Four-Phase Model 13
e arena in which equalizing or normalizing trends are most typically
observed, and the one that features most heavily in the e- pluralists’ accounts,
is the systemic level, and specically the intensity of inter- party competition.
Equalization occurs when the major and minor parties have a more balanced
or equal voice in terms of being heard by the electorate. Normalization sees the
reassertion of the larger parties’ communicative dominance. e second and
increasingly popular area of digital campaigning in which equalizing and normal-
izing tendencies are witnessed is at the intra- organizational level. Equalization
here refers to the use of the internet by internal actors to promote a redistribu-
tion of power downward and outward to the grassroots. Normalization is the
reversal of this tendency and the concentration of more power into the hands
of central elites. e third and nal arena of campaign activity in which these
terms are frequently invoked is that of voter communication. Both equalization
and normalization are understood in more relational terms in this context:the
former describing whether parties and candidates are exploiting the interactive
properties of the technology to allow for a more boom- up dialogue and “equal-
ized” conversation with the electorate; the laer denoting that these actors are
simply replicating the top- down approach used in the broadcast media and pro-
moting a more managed and elite- controlled style of communication.
Having broken down the notions of equalization and normalization into their
component parts and having specied more clearly what each means in prac-
tical terms, we can now join them back together to envision what each “state”
would look like if fully realized in a digital campaign. Equalization would see
increased competition between a large and diverse set of candidates who would
communicate their messages freely to voters, unfeered by centrally controlled
media editors. is opening up of the electoral system would be matched by an
internal decentralization of power as the lines between organizational elites and
grassroots blur and even potentially disappear. At the voter level, parties would
pursue an inclusive and genuinely interactive communication strategy. Citizens
would be given meaningful opportunities to engage in dialogue and discussion
with candidates over their policy stance and ideas, as well as with each another,
during the campaign.
By contrast, if normalization is fully realized, the level of electoral compe-
tition drastically decreases, as only the biggest and best- resourced parties can
survive. Internally, greater powers accrue to an elite group of highly technically
skilled operatives. ese new tech- gurus would take over running all aspects
of the campaign, with very lile input from activists on the ground and pos-
sibly from the candidates themselves and traditional party campaign managers.
Communication with voters and supporters would be highly regulated, and any
interactivity would be conducted in a controlled and instrumental manner that
is entirely concordant with campaign needs.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
14 WHEN THE NERDS GO MARCHING IN
14
Both of these scenarios are of course ideal types. Digital campaigning, as the
literature review that follows makes clear, has never reached a fully equalized or
normalized state. at said, what is evident from that review is that this binary
framework has formed a very useful and resilient heuristic for scholars in under-
standing the key shis and developments in digital campaigning over time.
Changes inthe Organizational Capacity and
Strategic Ends ofDigitalCampaigning
In addition to oscillating between these two ends of a power spectrum, the tech-
nological and strategic resources involved in running a digital campaign have
also changed and expanded over time. at expansion, we argue, can be similarly
understood as falling into four distinct phases. In the following we dene those
areas of expansion, rst with regard to the increases in the technological and
organizational capacity of parties, and second in terms of their strategic under-
standing of the medium.
TECHNOLOGICAL AND ORGANIZATIONALCAPACITY
On technical grounds there has clearly been a signicant growth in the sophis-
tication and functionality of the digital tools available to, and used by, cam-
paigners. More specically, in the rst decade of adoption, digital campaigns
were fought primarily with “web 1.0”– style tools such as home pages and email.
While initially usage was highly individualistic, as practice moved on, more
standardized tools such as “web in a box” and home page templates emerged.
e creation and spread of new web 2.0 and later social media technologies
reintroduced the possibilities for candidates to carve out a more individual-
ized presence. More signicantly, they increased the opportunities for more
interactive communication with voters and also among activists. e new
social networking platforms in particular oered the opportunity for indepen-
dent grassroots organizing and mobilization that formed the basis for a new
type of virtual community building. Innovations did not end here, however.
Perhaps the most signicant technical advances in digital technology, from the
campaign’s perspective, came in the second decade of the new millennium with
the move into the era of “big data,” cloud computing, and predictive modeling.
ese new tools signicantly increased the capacity of campaigns to mobilize
voters at an individual level, allowing them to amass and merge large quantities
of personal data and design an entirely new, scientically driven model of indi-
vidualized micro- targeting.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
15
The Four-Phase Model 15
ese increases in the technological capacity for conducting a digital cam-
paign have been accompanied by growth in both the size and prominence of
the organizational sub- team responsible for its conduct. In the earliest days, the
number of people involved in web- related operations in parties numbered less
than the ngers of one hand and were typically located in the IT department.
Within one election cycle, however, they were usually working as part of the com-
munications and marketing team. Following this merger, their next step up the
organizational hierarchy would be recognition as an independent self- contained
unit within the management structure. is promotion would give them a seat at
the “top table” of campaign management, rivaling that of other sub- teams such
as eld operations and fundraising. e next, and perhaps nal, advance in their
status comes with their elevation to overall control of operations. At this point,
the digital team assumes direct control over campaign decision- making. As the
size and range of expertise of the team expand, it becomes necessary to create
subdivisions. is sees the core preexisting team split o to form a new sub-
team dealing with the public interface side of the campaign (i.e., the website,
social media platforms, and email programs). In addition, new sub- teams are
formed to manage the vastly expanded data- collection and processing needs of
the campaign.
STRATEGICDEVELOPMENTS
Increases in the technical and organizational competence of parties to wage
digital campaigns are paralleled by an increased awareness, among those
in charge, of their strategic value and purpose. As internet use has spread
among the electorate, the opportunities it offers to widen parties’ support
base and convey a modernized image to the electorate have become more
apparent. This sharpened focus is evident in three main areas of campaigns’
strategic operations. First, those directing the campaign are much clearer
about the main purpose or primary goal of their efforts. Second, and relat-
edly, there is a stronger awareness of who the target audience(s) are. Finally,
there is a better understanding of what response or “action” is required from
those audiences.
Initially, when very few voters were online, there was no external pressure on
parties to do very much in any of these domains. Alaissez- faire strategy domi-
nated and there was very lile sense of purpose. ose parties and candidates
that did go online did so largely to avoid being labeled as technophobic or “out
of touch” with modern trends. is lack of vision persisted across much of the
rst decade of web campaigning. Site designers sought to improve the look and
feel of sites, but content lagged behind. ere were some limited aempts to
exploit the new medium to reach particular groups of voters, and sites took on
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
16 WHEN THE NERDS GO MARCHING IN
16
an increasingly niche- marketing role, with special pages being created for certain
segments of the electorate, such as young people and overseas voters. Journalists
were seen as a particularly important group to reach, with new “press centers”
being installed on most home pages. Aention shied to the transposing of mass
media content into the web environment. Sites became more like media hubs
or archives for storing press releases and campaign ads. Despite demonstrat-
ing a clearer understanding of the purpose of the web campaign, however, the
role of the audience was still largely passive and “reading” about the campaign
remained the main mode of voter response.
As audiences started to grow, parties and candidates started to see the
medium in a more proactive and dynamic light. is shi in thinking received
a signicant boost with the arrival of the new suite of participatory “web 2.0”
tools. Digital campaign content was no longer regarded as something for the
electorate to passively consume, but rather as a means for mobilizing support-
ers to help spread the campaign message, thereby reaching a wider electorate.
Audience engagement moves from a “reading” mode to one of “redistribution”
whereby supporters become nodes for spreading the parties’ digital content
through their own online networks.
Not content with mobilizing their base, however, campaigners become
increasingly eager to leverage the technology to make more direct and persuasive
contact with individual voters. e opportunities presented by the growth in the
size and quality of their databases make this increasingly possible. Undecided
voters become the main target audience of the digital campaign, with their pri-
mary mode of engagement now being to “receive” particular campaign content
and messages. ere remains an element of interactivity in the process, however,
in that there is an expectation that they will also return information to the cam-
paign by supplying their personal data.
Summarizing theFour Phases ofDigital
Campaigning:The MöbiusLoop
Drawing together the various elements in the process described in the preced-
ing sections produces a model of campaign change that is both linear or “pro-
gressive” in nature and cyclical. In visual terms, one might invoke the analogy
of the möbius strip, or loop, to capture this rotational movement. Movement
across the phases involves, on the one hand, an increase or “stepping up” of the
resources and the importance given to the digital campaign. is progression,
however, is accompanied by movement back and forth between two end points,
or poles— the rst being a condition of greater plurality and openness, and the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
17
The Four-Phase Model 17
second, one of power concentration and top- down elite control. e dening
features of each phase and the overall mode are captured in Table 1.1. Here we
describe each of the four phases, breaking each one down according to the same
eight features of the digital campaign, as shown in Table1.1.
Each of the phases is given a number and a more descriptive label. e former
indicates the chronological and progressive nature of the phases; for example,
phase Iis both earlier in the cycle and more “immature” than the later phases.
e laer is designed to capture the dominant ethos or rationale driving digital
campaigning among the key actors in what Foot and Schneider (2006) would
describe as the “electoral web sphere.”
Running down the le side of the table are the key criteria that we use to dif-
ferentiate the phases. We start with the dominant power logic associated with
each phase (i.e., equalization or normalization). is is disaggregated into the
three arenas of campaign activity highlighted earlier in which these logics are
most clearly observed— inter- party competition, intra- organizational commu-
nication, and external voter communication. e next set of criteria focus on
the capacity and resources associated with each phase of digital campaigning.
ese include the tools and technologies used to wage the digital campaign
and the structural location of the team established to deliver it. e nal set of
features used to mark out each phase centers on the strategic foci of the digital
campaign— namely, its primary goal or clearest objective, the target audience,
and desired and preferred mode of engagement or response from that audience.
While it is tempting to derive some causal implications from the table and
use it to specify the key drivers behind the shi into a new phase, its ambi-
tions are inevitably more descriptive and summative in nature. Technological
developments are clearly critical in enabling new digital strategies to emerge
(Bosea, 2018); however, organizations need to be able to “see” the value of
those innovations in light of their current priorities and goals. Despite its lack of
theory- building aspiration, Table 1.1 does aspire to have analytical merit. In par-
ticular, the möbius loop model of change that underlies it provides the organiz-
ing framework for the literature review that follows in Chapter2, and structures
each of the case studies we report in Chapters4 through 7.
In terms of the literature review, we show how the questions posed and con-
clusions reached by studies of digital campaigns over space and time broadly
follow the four phases of development. e model then becomes a template to
compare developments across four individual nations over a 20- year period.
e result is highly instructive, revealing an interesting paern of conformity
and variance across countries. Most nations appear to have followed a similar
pathway, but they have done so at a dierent pace. Some nations, such as the
United States, are consistently ahead of the curve. Others, like Australia, appear
to have seen a burst of initial momentum followed by a long period of stasis
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
18
Table1.1 Four Phases ofDigital Campaigning
Phase 1
Experimentation
(1992– 1996)
Phase 2
Standardization and
Professionalization
(1997– 2003)
Phase 3
Community Building and
Activist Mobilization
(2004– 2011)
Phase 4
Individual Voter
Mobilization
(2012 onward)
Dominant “Power” Logic Equalization Normalization Equalization Normalization
Power
Distribution
Inter- party
competition
Pluralistic Major party dominant Pluralistic Major party
dominant
Intra- organizational Localized— low scru-
tiny but limited candi-
date activity
Nationalized— local activ-
ity increases, templates
post hoc scrutiny
Localized— grass roots and
citizens as co- managers
Nationalized and
localized— Digital
teams coordinate
locally organized
groups
Communication
Mode
Top- down with
sporadic, random
interactivity
Top- down, information-
centered, static, controlled
feedback
Boom- up, action oriented,
dynamic, two- step
Top- down, micro-
targeted personalized
extractive
Capacity and
Strategy
Tool s Email and web 1.0 Email and web 1.0, RSS,
WPA, e- newsfeeds,
audio- visual
Web 2.0/ social media/ activ-
ist hubsites
Cloud computing,
big data, analytics
and experiments
Organizational location IT/ computing unit
or outsourced to a
volunteer
Campaign/ comms
team/ outsourced to
professionals
Independent unit operating
alongside other teams, seat at
head table
Digital team / direc-
tor driving strategy,
sub- teams formed
Primary goal “Me too” Transposition of oine
media content and
re- broadcasting
Mobilizing the base GOTV and personal
data extraction
Target Audience Anyone who happens
to visit
Media Journalists Supporters/ Activists/
Members
Undecided/ oating
voters
Voter Response Read Read Redistribute Receive and supply
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
19
Table1.1 Four Phases ofDigital Campaigning
Phase 1
Experimentation
(1992– 1996)
Phase 2
Standardization and
Professionalization
(1997– 2003)
Phase 3
Community Building and
Activist Mobilization
(2004– 2011)
Phase 4
Individual Voter
Mobilization
(2012 onward)
Dominant “Power” Logic Equalization Normalization Equalization Normalization
Power
Distribution
Inter- party
competition
Pluralistic Major party dominant Pluralistic Major party
dominant
Intra- organizational Localized— low scru-
tiny but limited candi-
date activity
Nationalized— local activ-
ity increases, templates
post hoc scrutiny
Localized— grass roots and
citizens as co- managers
Nationalized and
localized— Digital
teams coordinate
locally organized
groups
Communication
Mode
Top- down with
sporadic, random
interactivity
Top- down, information-
centered, static, controlled
feedback
Boom- up, action oriented,
dynamic, two- step
Top- down, micro-
targeted personalized
extractive
Capacity and
Strategy
Tool s Email and web 1.0 Email and web 1.0, RSS,
WPA, e- newsfeeds,
audio- visual
Web 2.0/ social media/ activ-
ist hubsites
Cloud computing,
big data, analytics
and experiments
Organizational location IT/ computing unit
or outsourced to a
volunteer
Campaign/ comms
team/ outsourced to
professionals
Independent unit operating
alongside other teams, seat at
head table
Digital team / direc-
tor driving strategy,
sub- teams formed
Primary goal “Me too” Transposition of oine
media content and
re- broadcasting
Mobilizing the base GOTV and personal
data extraction
Target Audience Anyone who happens
to visit
Media Journalists Supporters/ Activists/
Members
Undecided/ oating
voters
Voter Response Read Read Redistribute Receive and supply
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
20 WHEN THE NERDS GO MARCHING IN
20
and inertia. Still others, such as France, have followed a dierent and contrasting
path of early sluggishness followed by a rapid escalation into the laer phases of
activist and voter mobilization. Finally, the United Kingdom has taken what we
might label as the slow and steady approach, being neither an early bird nor a
late bloomer.
DATA SOURCESUMMARY
e reasons behind these dierent trajectories are explored through both a large
N comparative analysis (Chapter3) and a series of in- depth case studies using
original and secondary national data sets (Chapters4 through 7). Full details of
the data sources and measures used in each of the chapters are provided in the
appendices.
For the comparative analysis in Chapter3, we relied on the Comparative
Study of Electoral Systems (CSES) Module 4 data set. The CSES is a post-
election survey that is conducted by participating countries, typically as part
of their national election study. The surveys span a set time period. Module
4 was fielded during elections held between 2011 and 2015 and was selected
for this study given its focus on the theme of online mobilization. In par-
ticular, it included a battery of items that measured whether and in what way
respondents were contacted by political parties, and indirectly by friends
and family, during the election. Methods included SMS, email, and the web,
along with more traditional offline methods such as phone mail and in per-
son (see Appendix 3.1 for the full text of the survey items). At the time of
this analysis, Module 4 was in its second release, providing data on online
mobilization for 17 countries (specifically we used the Second Advance
Release [2.0], March 2015). Additional data was obtained for the United
Kingdom from the 2015 British Election Study (BES). The post- election
mail- back component of the study (fielded as part of the face- to- face cross-
section survey) included the Module 4 questions. This made it possible to
include the United Kingdom in the descriptive stage of the analysis, and to
benchmark its levels of online mobilization internationally, as well as in rela-
tion to our three other case studies.
In addition to the United Kingdom, three other countries were selected as
case studies for the book— Australia, France, and the United States. e sub-
stantive reasons for their selection and the light they shed on the role of national
context in shaping digital campaign developments are discussed in more depth
in subsequent chapters. ey were particularly useful for our analysis on empiri-
cal grounds since they each provided a series of relevant campaign data sets that
allowed for in- depth cross- national and over time comparison of changes in the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
21
The Four-Phase Model 21
supply and demand of digital campaigning.2 Specically, from the supply side,
national party and candidate websites in all four countries had been content ana-
lyzed using a standardized index applied during national elections held over the
period 2010– 2012. e index was designed specically to measure the extent to
which digital campaigns had started to shi into the more strategic third phase
of community building and activist mobilization. rough the data generated, it
was possible to not only compare the eorts made by parties within countries
to promote a more participatory and citizen- initiated model of campaigning,
but also to make some cross- country comparisons to assess which nations were
displaying the most eort in this regard, and how far into the four- phase cycle
they were. More details on the index and variable construction are provided in
Appendix 4.1.
On the demand side, a bespoke data set was produced in three of our four
cases (the United Kingdom, Australia, and France) that measured citizens’
level of interest and active involvement in national digital campaigns during
the same time period (2010– 2012) (Gibson, 2013). e same module of ques-
tions was elded either as part of a wider omnibus or a national election study.
e results are instructive in revealing the extent to which parties’ eorts to
mobilize activists and voters via their sites proved to be successful. e results
also allowed us to compare the levels and modes of public engagement with
the digital campaign across countries. While not directly comparable, simi-
lar questions were sourced from the 2012 American National Election Study
(ANES), which helped to extend the analysis to the case of the United States.
In addition to these directly comparable data sets, we used a wider range
of national election studies and other campaign surveys to measure change
over time in citizens’ engagement with digital campaigns across the four
cases. Selection of these additional data sources was determined by the extent
to which they provided variables that mapped onto the three key modes of
engagement associated with our four phases— namely, reading, redistri-
bution, and receiving. For the United Kingdom, we added data from the
2005“Campaigning in Cyberspace” Economic and Social Research Council
(ESRC) project and the 2015 British Election Study, also funded by the ESRC
(see Appendix 4.2 for full details). In Australia, we supplemented the analysis
with data from the 2013 Australian Election Study (see Appendix 5.1 for full
details). In the French case, we were able to use data for 2007 produced by
CEVIPOF Sciences Po, Paris, by the L’Institut Français d’Opinion Publique
(IFOP) (see Appendix 6.1 for full details). Finally, for the United States, we
combined the 2012 ANES data with ndings from the Pew “Internet and
American Life Project” campaign studies of 2004 and 2008 (see Appendix
7.1 for full details).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
22 WHEN THE NERDS GO MARCHING IN
22
CONTRIBUTION ANDLIMITATIONS
rough this combination of comparative and national analysis and blend of
primary and secondary supply- side and demand- side data sets, this book is able
to provide a unique overview of the evolution of digital campaigning across and
within countries. Specically, we can provide a “big picture” perspective on the
macro- level institutional, political, and socioeconomic factors that aect how
fast and far nations have moved through the four- phase cycle of digital cam-
paigning. We then drill down to the national level to show how richer histori-
cal and cultural factors have played a role in shaping the pace of innovation, as
well as inuencing organizations and individual voters. rough our case studies
we are able to show how parties, their supporters, and national electorates have
responded to the challenges and opportunities presented by the advent of digital
technologies. Furthermore, given the comparability of data available for these
four countries, we can assess the extent to which these ndings vary or hold up
across dierent contexts.
us, while clearly not covering all aspects of the move of digital technology
from the margins to the mainstream of campaigns globally, this book does oer
one of the most comprehensive insights into this process to date. at compre-
hensiveness lies in the length, breadth, and depth of the analysis that is under-
taken. Timewise, this book presents an overview of developments in digital
campaigns covering two decades. Breadthwise, our review of the literature pro-
vides a summary of international developments in digital campaigning to date,
which we then supplement with an original multilevel cross- national analysis of
the extent and contextual correlates of digital campaigning in recent elections.
Finally, from a depth perspective, we accompany these broader aggregate- level
ndings with a more nuanced picture of changes in the supply side and demand
side of digital campaigning within four individual countries. Using a series of
purpose- built data sets and national election studies, we chart the increasing
commitment of parties and candidates to undertaking online campaign activi-
ties and the response of voters. Although each of our case studies operates as a
stand- alone analysis, the consistent application of the four- phase model to parse
out and describe developments over time means it is possible to make some use-
ful comparisons of the pace of change and levels of popular engagement.
While highlighting the unique perspective and rich data that this book brings
to the study of digital campaigning, it is also important to acknowledge its
limitations— the most obvious of these being that a temporal model of digital
campaign development cannot fully describe the evolution of web campaigns
across all, or perhaps even the majority of, individual countries. As a generic
“one size ts all” framework, it loses nuance and accuracy when applied to real-
world cases. e phases serve as conceptual “phenotypes” that are designed to
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
23
The Four-Phase Model 23
distill and link together a series of critical technological, strategic, and organi-
zational dimensions of digital campaign practice. How these traits play out in
practice in terms of preferred platforms, the nature of the communities that are
built, and the micro- messaging that is delivered is context dependent.
Furthermore, while we expect the order or sequencing of the phases to hold
across countries (i.e., phase I will precede II and II precedes III), the dates
ascribed to them in Table 1.1 are meant to be indicative, not denitive. ese
time periods signify when a given phase is considered to be in a dominant or
ascendant position in broad global terms. e evolution of digital campaigning
in some nations may t quite neatly into the chronology specied, but this will
not be the case universally. Some countries that come later to digital campaign-
ing, for instance, may bypass the experimental phase and “leapfrog” directly into
the later, more sophisticated phases of use. Countries that share the same start-
ing point may display dierent rates of development. is variance is underlined
by our case studies, where we see how similarly advanced industrial democracies
dier in the length of time parties take to experiment and standardize their use
of the new tools, and when they begin to move into more strategic mobilizing
uses of the technology.
e case studies also serve to highlight a further important qualication to
the application of our model, which is that it does not apply in toto to most coun-
tries. Many countries have not yet entered, or are only on the cusp of phase IV.
We aempt to explain the reasons for this dierential development and pace of
change in the analysis of Chapter3. Here we show very clearly that there is con-
siderable variance in the levels of online voter mobilization across countries. We
go on to show how levels of technological and societal development, along with
the political and electoral incentive structures, have determined these paerns
of growth.
Finally and relatedly, while our framework suggests distinct phases, their
edges are— in reality— not as sharp as the model suggests, and there is inevitable
“real world” overlap and seepage. is porosity is perhaps most observable in the
technologies assigned to each phase. If we take the case of blogs or “web logs,”
for example, despite being part of the new “web 2.0” suite of tools, they actually
rst emerged in the late 1990s, well before the community- building and activist-
mobilization phase that we associate with social media. Similarly, while email
came to the fore during the experimental phase, it has continued on to power
some of the most precise scientic appeals of phase IV.
From a strategic perspective, the goals of activist and direct- voter mobiliza-
tion, which we equate with phases III and IV, also motivated campaigns in earlier
phases. Certainly, the candidacy of Jesse Ventura for the Minnesota governor-
ship in the late 1990s was seen as a highly eective example of political com-
munity building online (Hindman, 2005). Candidates and parties also made
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
24 WHEN THE NERDS GO MARCHING IN
24
aempts at online micro- targeting through tailored web pages, and encouraged
tactical voting through “vote swapping” websites before 2012. If we narrow the
lens to examine progression through the phases by individual parties, the bor-
ders between phases arguably become even more blurred. As our case studies
reveal, there are dierences in the pace of change and innovation between le-
and right- wing parties, and between the major and minor players. While some
may be ahead of the developmental curve, others lag behind.
Given these rather extensive qualiers, it is clear from the outset that map-
ping the chronology of a country’s pathway through the four phases and its cur-
rent point of development is something of an inexact science. at said, it is the
ambition of the book to try to do just that, for at least four major democracies
using varying methodologies. e next chapter begins that process by present-
ing impressionistic evidence culled from the empirical literature on the adop-
tion of digital campaigning at the national and international level. Chapter3
tackles this question using a more systematic and original approach that aims to
measure and explain levels of online electoral mobilization occurring in our four
countries, and elsewhere in comparable elections. is forms a proxy measure of
entry into phase IV of digital campaigning (i.e., direct voter mobilization).
e remaining four chapters take these top- line statistics and global ranking
of countries’ online mobilization levels and employ a mixed- method approach
to probe their dierences and similarities. is involves rst presenting a rich
historical narrative and overview of the particular factors in these countries that
have shaped their progress in digital campaigning. is qualitative analysis is sup-
plemented by a more standardized analysis of changes in web campaign outputs
over time, using the secondary literature and a web content analysis index. e
index is applied to national parties’ and candidates’ campaign sites during their
shi into the more strategic phase III mode of community building and activ-
ist mobilization. Finally, we present a comparable range of national survey data
that looks at the changing paerns of demand among the electorate for digital
campaigning, and the extent to which this ts with what we have learned in the
prior section about changes in its supply. rough this combination of sources,
we provide a unique, comprehensive comparative picture of the movement of
digital technology from the margins to the mainstream of campaign practice.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
25
When theNerds Go Marching In. Rachel K. Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
2
AReview oftheLiterature
From Experimentation to Mobilization
is chapter reviews the growing body of literature that has emerged on the sub-
ject of digital campaigns since their inception in the mid- 1990s. In particular,
it shows how, when viewed cumulatively and with the advantage of hindsight,
these studies form a historical narrative that identies four distinct phases of
development.
Phase I:Experimentation (1992– 1996)
...we just jumped and hoped that we would get down all right.
— Danish Liberal Party Web Campaign Manager (1997)1
An important, if somewhat obvious point to make in charting the history of
digital campaigning is that the use of computers in elections is not a new phe-
nomenon. While not documented in extensive detail, particularly from a com-
parative perspective, the process of diusion is generally seen as beginning in
US elections in the early 1960s. In particular, it was the Democratic Party’s
Simulatics project that pioneered the use of computer technology for voter tar-
geting (Janda, 2015; Issenberg, 2012; de Sola Pool etal. 1965; Burdick, 1964).
In the decades that followed, both major parties continued to use computers to
simulate and forecast electoral behavior and also to expedite their more mun-
dane information- processing tasks. It was really only in the mid- 1990s, and par-
ticularly following the development of the World Wide Web (WWW), however,
that computers started to be used externally for public communication.
According to most reports, the rst aempt at digital campaigning occurred
during the 1992 US congressional elections when Jerry Brown, the former gov-
ernor of California, experimented with sending email messages to his supporters
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
26 WHEN THE NERDS GO MARCHING IN
26
in his bid for a Senate seat (Janda, 2015). e rst ocial campaign websites
emerged a lile later, with a trickle of national parties launching home pages
during 1994. e following three years saw a urry of activity as parties fell like
dominoes into cyberspace (Gibson, 2004).
In terms of why they were there, the opening quote from a Danish party web
manager in describing his organization’s reasons for going online nicely captures
the curiosity- driven and experimental approach that most political actors took
toward the internet at this time. Certainly, the content and design of most of the
sites established during this period support the idea that the parties lacked a clear
rationale and purpose for their digital campaigns. Figures 2.1 and 2.2 reveal the
rather spartan look and feel of the home pages of some of the major European
le- wing parties at the time. e UK Labour Party site appears to be particu-
larly underdeveloped given that the next general election was just six months
away. eir right- wing counterparts were no more advanced, however. Some
years later, a leading UK politician and campaign director for the Conservatives
openly acknowledged the “homemade” feel of his own party’s site during these
early years.2
In the United States, a country that is widely seen as driving innovations in
campaign technology, the picture appeared to be surprisingly similar. Studies
of the web campaigns conducted in the 1996 US presidential election reported
Figure2.1. e UK Labour Party home page (November 1996).Source:Internet
Archive:hps:// web.archive.org/ web/ 19961109025623/ hp:// www.labour.org.uk/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
27
A Review of the Literature 27
that, with the notable and ironic exception of the septuagenarian Bob Dole, most
sites were of poor quality and showed very lile imagination (Epstein, 1996;
Stone, 1996; Hall, 1997; Reavy and Perlmuer, 1997; McKeown and Plowman,
1999). Moreover, like their European party counterparts, US candidates and
their managers were seen as lacking a basic understanding of the value of the
new medium. According to Stromer- Galley (2014), such eorts were lile more
than token gestures and were not designed to last beyond the current electoral
cycle (24). To support her contention, she pointed to the highly confusing array
of web addresses and URLs for the sites, most of which did not include the can-
didate’s name and so were unlikely to be recalled by voters very easily. Selnow’s
(1998) account of the 1996 campaign was similarly critical about the lack of
clear reasons candidates had for going online, leading him to label it the “me
too” web election. ose who launched a site, he argued, did so simply because
they “want to climb on board the bus with everyone else” (88). It was the fear of
being le behind, or perhaps more importantly of being seen to be le behind,
therefore, that formed the dominant incentive for web campaigning during this
period.
Details about the resources commied by parties to their web campaigns at
this time are sketchy to nonexistent for most countries. e evidence that does
exist comes largely from interviews with party ocials, and shows that invest-
ment in digital technology and teams was minimal. While Dole and some of the
Figure2.2. German Social Democratic Party (SPD) home page (December 1996).
Source:Internet Archive
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
28 WHEN THE NERDS GO MARCHING IN
28
other US candidates bucked the trend to employ professional web developers
(Stromer- Galley 2014), most parties used their IT sta and resources to support
their web eorts (Gibson and Ward, 1998; Ward and Gibson, 1998; Ho and
Lofgren, 1997). Among the smaller parties, operations were even more rudi-
mentary, usually comprising a single individual or “web master” who managed
the entire operation and who was oen an unpaid friend or family member.
One unintended but important consequence of this low- key approach to web
adoption by campaigns was that it opened the way for smaller parties to com-
pete. Based on their study of the UK’s “rst internet election” in 1997, Gibson
and Ward (1998) concluded that while radical change was not in evidence,
neither was it possible to conclude that “politics as usual” was carrying on in
cyberspace:
the smaller parties are indeed holding their own in terms of their web
sites’ appeal. Far from leaving the minor parties in the dust the internet
appears to be doing more to equalize the exposure of party ideas to the
electorate compared to other media. (22)
Scholars of extremist politics were also quick to point out the inherent advan-
tages that the net oered to smaller organizations on the ideological fringe, par-
ticularly those on the far right (Whine, 2000; Gerstenfeld etal., 2003). Such
parties, they argued, had a stronger incentive to establish a presence online,
given that it provided a new direct and unltered channel to reach existing and
new audiences, and to build up wider national and international organizational
networks (Ward etal., 2008b; Copsey, 2003).
e evidence in favor of equalization, however, centered largely on changes at
the systemic level and shis in inter- party competition during this early phase.
Any intra- organizational power redistribution stemming from the spread of
internet technology was minimal to nonexistent. is was due in large part to the
limited adoption of the new media at the lower echelons of the party and among
the grassroots. e earliest reported levels of adoption among candidates and
local parties in the United Kingdom from 1997 indicated that again only a tiny
minority— around 5 percent— were online (Auty and Nicholas, 1998; Ward
and Gibson, 1998). Even in the United States, where levels of state- and local-
level adoption were among the healthiest, candidates still had a patchy presence
online. An academic survey of campaign activity during the 1996 election cycle
reported that less than one in ve candidates in Senate, House, and gubernato-
rial races had a website (D’Alessio, 1997).
In terms of equalization in our third arena of interest— external voter
communication— there was also lile sign of any signicant change. Among the
citizen body, rates of internet adoption during this experimental phase were low,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
29
A Review of the Literature 29
with most countries struggling to report even 5percent of citizens as online in
1996.3 is clearly limited the capacity of parties to stimulate a more interactive
national campaign. e evidence from Figures2.1 and 2.2 certainly suggest the
major parties took very lile interest in oering opportunities for dialogue and
focused mostly on providing static, text- based content. Even Dole’s “beacon”
site (shown in Figure 2.3), which was praised for its stronger voter appeal, only
allowed visitors to interact with the site content, rather than directly with the
candidate or campaign in an interpersonal manner.
at said, there were some signs that parties recognized the potential of the
web to increase two- way communication with the public. Some sites presented
a range of email addresses that allowed users to target messages directly to par-
ticular branches, and even to individual people within the party organization.
e UK Liberal Democrats, for example, promoted the email address of its
party leader, Paddy Ashdown, on its home page. It is not clear how respon-
sive parties were to the emails they received. Anecdotal evidence suggested
that at least some major parties were treating them seriously. e Swedish
Social Democrats, for example, reportedly hired sta to help answer all emails
personally in the 1998 parliamentary elections. In addition, they hosted over
70 online Q&A sessions with politicians during the campaign (Gibson etal.,
2003b:19)
Figure2.3. Dole/ Kemp home page (November 1996).Source:Internet Archive:hp://
web.archive.org/ web/ 20160616160314/ hp:// www.dolekemp96.org/ main.htm
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
30 WHEN THE NERDS GO MARCHING IN
30
Overall, therefore, early digital campaign eorts were sporadic and tentative.
Parties and candidates were motivated to go online by a mix of curiosity and the
fear of being le behind. e resources devoted to their online activities were
typically limited, or in some cases nil. Activity, where it did occur, was concen-
trated largely at the national level, with local actors proving more circumspect
about investing their scarce resources in the new media. Some aempts at fos-
tering a more open dialogue with and among voters were in evidence; however,
the “web 1.0” environment largely encouraged parties to engage in a more top-
down information- centric campaign. Despite, or perhaps ironically because of
this reticence to invest in their web presence, the narrative of equalization at the
party- system level took hold. As the major parties in particular perceived lile
obvious strategic gains in this niche medium, the “bar” to entry was lowered and
the smaller players were able to maintain parity in terms of the quality of their
online oerings. e equalization that occurred was thus due more to default or
accident than design.
Phase II:Standardization and
Professionalization, 1997– 2003
...dragging the old media into the new.
— Phil Noble, Netpulse Special Report (2002:n.p.)
Early amateur forays into cyberspace were short- lived. e pressure of the elec-
toral cycle and increasing internet access among the electorate meant parties
soon started revamping and modernizing their online presence. While websites
and email remained the dominant tools of combat, parties (particularly the
larger ones) increasingly turned to commercial companies to design and man-
age their content and platforms. e result was convergence to a more visually
appealing graphical interface that was easier to navigate through the addition of
menu buons and home page icons. e upgrade and standardization in parties’
online oerings can be seen in Figures 2.4 and 2.5, which show the redesigned
UK Labour Party and German Social Democratic Party (SPD) home pages as
they appeared in the lead- up to national elections in 2001 and 2002, respectively.
As well as being more content rich than the versions shown in Figures2.1
and 2.2, the sites have a more professional look and feel that is aligned with the
party “brand.” Both sites feature a side or top menu bar, and buons linking out
to a press or media section and a series of options to join, shop, and donate to
the party. is increasing sophistication and stronger marketing appeal was also
a sign of how far control over the web campaign had shied out of parties’ tech-
nical support teams and into the hands of their PR and communications unit.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
31
Figure2.4. e UK Labour Party home page (May 2001).Source:Internet
Archive:hps:// web.archive.org/ web/ 20010520110505/ hp:// www.labour.org.uk/
Figure2.5. German Social Democratic Party (SPD) home page (October
2002).Source:Internet Archive:hps:// web.archive.org/ web/ 20021006111214/ hp:// www.spd.
de/ servlet/ PB/ menu/ 1009319/ index.html
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
32 WHEN THE NERDS GO MARCHING IN
32
In addition to qualitative changes in the style and structure of the web cam-
paign, phase II also sees a notable expansion in the overall volume or amount
of activity taking place. An audit by Norris (2001) at the turn of the millen-
nium reported that some type of online presence had become almost universal
across party systems in Europe and North America. At the subnational level,
adoption had also increased, although it was much patchier. As late as 2002 it
was reported that up to one- quarter of candidates competing for a US House
of Representative seat still lacked a web presence (Davis and Owen, 2008:96).
Furthermore, the level of commitment shown by those to maintaining their
web prole remained very low. Areport issued by Web analytics rm RightClick
in 2002 found that only around one- quarter of major party sites in House and
gubernatorial elections were actually updated on election day.4
Despite the growing investment in web campaigning, there remained a wide-
spread view that parties and candidates had still not developed a clear under-
standing of the advantages oered by the new medium (Davis etal., 2002).
ere was a more concerted focus on online fundraising, although this was argu-
ably prompted by changes in campaign nance regulations rather than any inter-
nal decision- making.5 e dominant approach among campaigners was to play it
safe by limiting interactivity and migrating oine content such as press releases
and television ads onto their sites. Such a lack of ambition became the focus
of increasing criticism from academic and industry observers. Phil Noble, the
editor of Netpulse, a popular online newsleer covering digital campaigns and
elections, was highly dismissive of the lack of imagination shown by candidates
in 2002. e eorts, he argued, amounted to lile more than “puing a new tele-
vision camera in front of a radio newsreader behind a microphone and calling
it TV” (Noble, 2002:n.p.). Looking back on this period, Bruce Bimber (2014)
echoed Noble’s core complaint, arguing that up until 2004 web campaigns in
the United States “had indeed been largely an amplied version of traditional
politics” (133).
Empirical analyses of web campaigning beyond the United States and the
United Kingdom in the late 1990s provided further support for the view that
campaigns lacked a clear goal. Coding schemes, designed to measure and com-
pare parties’ and candidates’ performance, were widely applied across national
elections. e results revealed a consistent picture of static and information-
heavy sites with limited opportunities for interaction (Roper, 1998; Trechsel
etal., 2003; Conway and Dorner, 2004). In their comparative overview of activ-
ity through to the mid noughties, Gibson and Ward (2009) found that “...an
astonishingly formulaic picture of adaptation” had emerged in internet cam-
paigning. ere was an almost compulsive focus on “information dissemination
and the migration of oine content to the online environment” and a pro-
found neglect “of the unique interactive features of the medium” (93). It is thus
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
33
A Review of the Literature 33
perhaps no surprise that it was around this time that the terms “virtual billboard”
(Sadow and James, 1999; Stromer- Galley, 2004), “electronic pamphlet,” and “e-
brochureware” (Foot and Schneider, 2002)started to appear in the academic
lexicon of digital campaign research. ese labels perfectly evoked the anodyne
and conservative approach taken to the web by political actors.
Application of the new coding schemes also served to dispel any notion
that the equalization of party competition detected during phase Iwas gaining
momentum. Subsequent studies of British parties’ online campaign sites during
the 1999 European elections and 2001 and 2005 general elections all pointed
to an increasing gulf between the online campaigns of the major and minor par-
ties (Gibson and Ward, 2000a; Ward and Gibson, 2003; Jackson, 2007). Other
studies across a range of established democratic polities, such as New Zealand,
Australia, Finland, Germany, and Canada, revealed a similar growing imbalance
in the quality of parties’ online eorts (Roper, 1998; Carlson and Djupsund,
2001; Gibson and Ward, 2002; Conway and Dorner, 2004; Schweitzer, 2005;
Gibson etal., 2008; Small, 2008; Ward etal., 2008a).
It was the work of Margolis etal. (1997, 1999)and partic ularly that of Margolis
and Resnick (2000), however, which proved most inuential in reseing expec-
tations for a more equalized future political system. eir analysis of the impact
of the internet on American parties, media, and business organizations was piv-
otal in claiming that a process of “normalization” was now underway. e “fron-
tier” spirit that had taken hold online in the early days, they lamented, was now
“fading fast...[as] political, economic, social and recreational life on the Net for
the mass public is increasing designed and guided by professionals” (Margolis
and Resnick, 2000:4).
Even when chinks in the normalization narrative did emerge, they were
almost always treated as exceptions to the rule. Studies of the Netherlands,
New Zealand, Italy, Japan, the United Kingdom, and Australia, for instance, all
pointed to the “surprisingly” strong performance of Green parties, particularly
with regard to the interactivity and participation opportunities provided on
their sites (Ward and Voerman, 1999; Newell, 2001; Tkach- Kawasaki, 2003).
Furthermore, the explanations provided for these deviant cases normally cen-
tered on their having a preexisting culture of decentralization and grassroots
empowerment. is meant that any innovation or “equalization” tendencies in
use of the medium were seen as a spontaneous or involuntary response by the
party in question, rather than as a deliberate strategic aempt to harness the
medium’s unique properties.
us, by the turn of the millennium, a new era of more standardized and pro-
fessionalized digital campaigning appeared to have taken hold across established
democracies. is second phase was marked by the growing reliance of parties
on professional consultants and increasing investment in their digital presence.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
34 WHEN THE NERDS GO MARCHING IN
34
e new commitment resulted in the gap between large and small parties widen-
ing, and an end to the talk about a likely leveling of the communications playing
eld that had emerged in phase I.Despite this increased commitment and grow-
ing consensus among parties that the internet maered, exactly how and why
remained something of a “black box.” Campaigns thus increasingly converged
on a model of sleekly designed, but largely static and generic, content. e focus
was primarily on playing it “safe” and translating existing content and appeals
from other media into the online space, rather than identifying and harnessing
its unique properties.
Phase III:Community Building and Activist
Mobilization (2004– 2008)
e Internet is tailor- made for populist, insurgent movement. Its roots in the open- source
ARPAnet, its hacker culture, and its decentralized, scaered architecture make it dicult for big,
establishment candidates, companies and media to gain control of it.
— Joe Trippi, e Revolution Will Not Be Televised (2004:102)
Findings of normalization in levels of online competition and the mode of
voter communication persisted for much of the first decade of the new mil-
lennia. Studies in a variety of national contexts repeatedly showed a deci-
sive gap opening up between the larger and smaller parties’ web presence
and a continuing emphasis on top- down communication (Vedel and Koc-
Michalska, 2007; Robles- Estrada etal., 2008; Strandberg, 2009; Schweitzer,
2008, 2011; Lilleker et al., 2011). Given the support that these findings
provided to the e- pluralists’ predictions of a return to politics as usual, and
the reinforcement of existing elites, one might expect the story of digital
campaigns to end here, at least with regard to any further major power
shifts. However, from the middle of the new decade, fresh evidence began
to emerge which indicated that the momentum toward equalization was not
entirely dead.
Speculation was triggered initially by the growing popularity of a suite of
more “user- friendly” tools that had developed in the aermath of the dot.com
collapse in 2001. Collectively labeled as “web 2.0” technologies (O’Reilly, 2005;
Anderson, 2007),6 the new soware was heralded as reviving the ethos of the
web as a dynamic, decentralized media that was open to all. Archetypal tools like
blogs and social networking sites such as Facebook were seen as revolutionary
in the extent to which they ceded power to ordinary users to create, share, and
promote new forms of content. e result was a new “architecture of participa-
tion” that promoted a radically dierent model of networked communication
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
35
A Review of the Literature 35
production and distribution beyond the reach of the traditional broadcasters
and press (Chadwick, 2008; Gueorguieva, 2008; Castells, 2009).
Excitement about the democratizing eects of web 2.0 traveled far beyond
the electoral sphere and indeed beyond the Western world (Lotan etal., 2011;
Howard and Hussain, 2013). In campaign terms, however, discussion quickly
seled again on questions of pluralization of party systems and opportunities for
more two- way voter communication. Platforms such as Facebook and Meetup
provided a multiplicity of cheap channels to recruit activists and engage with
those “hard to reach” sections of the electorate that had tuned out from main-
stream media and politics. Studies conducted across a wide range of national
contexts in Europe, Oceania, and Latin America served to conrm the appeal of
the new social web for the smaller parties in particular (Carlson and Strandberg,
2008; Gilmore, 2012; Gibson and McAllister 2011; Samuel- Azran etal., 2015;
Koc- Michalska et al. 2014; Hansen and Kosiara- Pederson, 2014; Dolezal,
2015).7
It was in the internal arena, however, where expectations of a shi toward
equalization now started to emerge perhaps most strongly. e newly devolved
ecology of communication meant that ordinary supporters had the opportu-
nity to self- organize and promote their preferred candidate without the need
for central resources or HQ intervention (Gibson, 2015; Lilleker and Jackson,
2010; Gibson etal., 2013; Vergeer etal., 2013). Although there had been some
notable aempts by candidates and parties to build up networks of activist sup-
port using the internet,8 it was in the US presidential election cycles of 2004 and
2008 where this co- production model came to fruition on a national scale. At
the forefront of this eort was Howard Dean, the lile known governor of one of
America’s smallest states, Vermont. Under the direction of his IT- savvy manager
Joe Trippi, Dean succeeded in building up a dense network of highly commied
digital volunteers, or “Deaniacs,” who catapulted him to early frontrunner status
in the Democratic primaries.
From the start of the campaign, Trippi had understood that the “open-
source...decentralized, scaered architecture” of the internet meant that it
is “tailor- made for a populist, insurgent movement” such as Dean’s (Trippi,
2004: 102). It was, however, Dean’s supporters’ use of social media and
particularly the Meetup.org platform to self- organize into local teams that
proved the critical factor in building his success. rough this interface the
campaign was able to tap into a huge swell of grassroots support and to con-
nect that online enthusiasm with oine action. Such an approach revived the
type of personalized electoral communication that Lazarsfeld etal. (1948)
had identied over half a century ago in their two- step ow model. For some,
it heralded a new mode of campaigning centered on “...a new form of politi-
cal community building” ( Johnson, 2007:140). Rather than following the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
36 WHEN THE NERDS GO MARCHING IN
36
typical “command and control” structure that had dominated campaigns in
the laer decades of the twentieth century, supporters and members were
now central actors in disseminating and creating the campaign message
(Trippi, 2004; Williams and Tedesco, 2006; Colville, 2008; Gueorguieva,
2008; Teachout and Streeter, 2008; Kreiss, 2009, 2012; Lilleker and Jackson,
2010; Gibson, 2015).
Where Dean led, others soon followed. Despite his early exit from the race,
Dean’s rapid rise had shown the value of social media as a means of activist
recruitment, and his techniques were quickly co- opted by his rivals (Stromer-
Galley, 2014:98). It was Republican George W.Bush’s digital team that proved
most adept in developing Dean’s legacy. eir “Personal Precinct” program
oered a new online nationwide volunteer management system that allowed
individuals to “join” the campaign and earn points by undertaking a range of
ocially recognized activities to help the candidate locally. e highest perform-
ers were then rewarded by a public listing on the campaign “leader board” pages
(Turk, 2012). Bush’s team thus understood, as had Dean’s advisors, the power
of the technology to engage and empower the grassroots and to build a sense of
ownership and community around the campaign. Unlike Dean, however, they
also understood how to incentivize and direct that energy toward the core goal
of winning the election. rough what was essentially an “in- house” version of
Meetup, they made it possible for central HQ sta to monitor and channel the
activities of the new recruits.
It was the perfection of this blended model of grassroots input and top-
down direction that defined the digital strategy of Illinois senator Barack
Obama in 2008 (Stromer- Galley, 2014:104). Guided by Blue State Digital,
the online agency formed out of the ashes of Dean’s implosion, the cam-
paign managed to effectively link Obama’s community organizing expe-
rience with the power of the new medium to deliver local activism at an
unprecedented national scale. Central to delivering this new form of net-
worked campaigning was MyBO, a purpose- built volunteer- management
tool designed by Chris Hughes, the cofounder of Facebook. Launched in
Spring 2007, MyBO was unique in merging the community element of a
social networking site with the functionality of Meetup. Registered mem-
bers could join groups and interact online through messaging and personal
blogs. They could also receive training and tools to allow them to engage in
more purposeful tasks such as canvassing voters, holding events, and rais-
ing funds online (Harfoush, 2009). The daily updates on volunteer activity
and success generated through MyBO were then used to update field opera-
tions and ensure a more efficient deployment of campaign resources. This
heightened the reach and prominence of the digital face of the campaign
and ensured that the team itself now moved up the organizational hierarchy
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
37
A Review of the Literature 37
from its position in phase II. Those managing the web side of operations
now secured a seat at the top table. This elevation brought them eyeball to
eyeball with established units like field, media relations, and fundraising in
the battle for resources.
While it reached its fullest expression in the Obama campaign of 2008,
the new “outsourced” model of digital campaigning was not confined to
North America. In South Korea several years earlier, Roh Moo- hyun, a little
known legislator, had launched a successful bid for the South Korean presi-
dency in 2003, fueled through his “Nosamo” network of online activists
(Kim, 2006; Qiu, 2008). Roh was even belatedly labeled the “Asian Howard
Dean” by observers, given the strong similarities in their style of digital
campaigning.
Beyond presidential systems, parties in parliamentary systems also appeared
to be waking up to the activist mobilization properties of the new social media
tools. Among party scholars, discussion focused on the possible emergence of
a new “cyber” or “networked” model of party, which featured a more decentral-
ized structure and empowered grassroots (Heidar and Saglie, 2003; Löfgren and
Smith, 2003; Marges 2006). While not necessarily gaining full expression in
the real world, key features of the new prototype were observed among certain
parties, particularly the smaller players. In Italy, for example, the Radical Party
was singled out for its experimental uses of the new technologies to recruit sup-
porters and develop a national public prole (Kies, 2004). Other countries wit-
nessed the formation of new virtual parties, such as Senator Online in Australia,
that were commied to using the internet to run an entirely new type of member-
driven organization (Chen, 2013). In Norway, Kalnes (2009) identied what he
termed the “e- ruptive” consequences of web 2.0 technologies for parties, which
involved a lowering of the threshold for participation among grassroots support-
ers and sympathizers.
Enthusiasm for the newly devolved model of campaigning also spread to
the major parties. Anumber of mainstream le- and right- wing players in the
United Kingdom, France, Australia, and Scandinavia launched copycat ver-
sions of MyBO following Obama’s victory in 2008. e sites— which typically
used the “My” prex (e.g., “MyConservatives” or “MyLiberals”)— were clearly
designed to emulate their US predecessor. Beyond their similarity in look and
feel, however, they shared the core purpose of recruiting non- members to the
cause. rough the sites, interested individuals were registered as supporters and
then were charged with a series of ocial tasks such as canvassing and organiz-
ing fundraisers and rallying events. e result was a new citizen- initiated model
of campaigning (CIC) that fundamentally challenged the traditional hierarchy
and boundaries of parties as member- based mass organizations (Lilleker and
Jackson, 2010; Gibson, 2015; Karlsen, 2013).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
38 WHEN THE NERDS GO MARCHING IN
38
Despite their interest in activist mobilization internally, the moves by major
parties into the web 2.0 space ultimately prompted a revival of claims of normal-
ization in systemic terms. Bespoke MyBO- type sites were typically beyond the
nancial reach of most smaller parties and independent candidates. In addition,
it was clear that although some powers were passed downward and outward to
supporters through these platforms, central monitoring was key to really making
the two- step ow model work. In practical terms, even if smaller parties were
making stronger eorts to campaign with “free” forms of social media, and were
found in some cases to receive a disproportionate electoral benet (Gibson
and McAllister, 2011; Gilmore, 2012, it was still the case that the major play-
ers gained the lions’ share of aention in terms of likes and follows (Larsson,
2017). Moreover, it was clear that despite Dean’s, Roh’s, and even Obama’s status
as “outsiders” within their own parties, they were very much part of the main-
stream political order.
Within campaign communication, the normalization narrative returned as
analyses of parties and candidates’ actual use of web 2.0 were undertaken. Hopes
for a richer and deeper dialogue with voters were dashed as studies consistently
reported a top- down approach being taken to the new platforms, and very few
genuine aempts at interaction with citizens (Graham etal., 2013; Murchison,
2015; Druckman etal. 2014). According to Stromer Galley (2014), most US
presidential candidates, including Obama and even Dean, had failed to fully
grasp the two- way communicative opportunities of the medium. Instead, they
had resorted to what she called a form of “controlled interactivity,” a form of two-
way communication that allowed for debate, but only on certain “safe” topics.
Subjects that might produce more meaningful debate and even disagreement
were studiously avoided.
us, while phase III did register a pushback against the long march toward
normalization of phase II, particularly at the organizational level, mounting evi-
dence suggested that this did not permeate across the layers of campaign activity.
It was clear that minor parties had not seen a reversal in their fortunes and genu-
ine dialogue had not yet broken out between elites and the masses. Campaigns
were, however, clearly doing more to enable supporters to talk among them-
selves, and to help them spread the word about a candidate or party within their
online networks. is cascaded approach to campaign communication helped
revive a two- step ow model of voter persuasion associated with an earlier, more
vibrant era of community- based campaigns (Lazarsfeld etal., 1948). In strategic
terms, the focus of digital campaigners had switched to mobilizing their base.
Supporters and activists were now the key targets for their eorts. Rather than
passively consuming content, as in the two previous phases, visitors to websites
were now being prompted to circulate and redistribute material through their
networks.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
39
A Review of the Literature 39
Phase IV:Voter Mobilization (2012– present)
e world of politics...is now done by Martians.
— Peggy Noonan, Wal l Street Journal (2011)9
e move into the fourth phase of digital campaigning constitutes perhaps its
biggest step forward to date. is is the point where digital technology nally
takes center stage and assumes responsibility for the core task of voter mobiliza-
tion. is is not to say that GOTV (get out the vote) objectives had been absent
in earlier phases. Scholars of campaigns and elections had been speculating for
some time on the voter targeting potential of the new medium (Norris, 2000;
Plasser, 2000; Plasser and Plasser, 2002; Farrell and Schmi- Beck, 2002).10
Phase II had seen several aempts by parties to exploit the “narrowcasting” capa-
bilities of the web by developing content for particular audiences, such as young
people, female voters, and journalists. Other more explicit vote- geing initia-
tives such as “vote- swapping” sites had also been promoted by smaller parties
to encourage the strategic exchange of ballots with ideological allies in a bid to
unseat or deny victory to a common “enemy.”11 Finally, phase III, as the previ-
ous section revealed, was dened in large part by digital campaigners’ eorts to
develop a new two- step or indirect model of voter mobilization through online
channels.
While these initiatives were clearly designed to increase parties’ and candi-
dates’ electoral support, they also faced some signicant challenges with regard
to their precision and inherent passivity as mobilization tools. Voters needed
either to nd the sites themselves or to be connected in some way with the
campaign through their online social networks in order to gain exposure to the
mobilizing messages. Connecting with that all- important pool of undecided vot-
ers that could help swing an election outcome was thus a maer of luck and
random exposure, rather than coordinated intent. In phase IV these uncertain-
ties and ineciencies are essentially removed from the process as a new set of
highly accurate “scientic” methods are introduced for pinpointing and mobiliz-
ing these latent pockets of support.
e shi in focus to more micro (or even what some termed “nano”) target-
ing of the electorate during phase IV is matched by an intensication of eorts
to “pull” more detailed information about individual voters into the campaign.
While websites had previously gathered personal details about visitors through
newsleer sign- ups and feedback forms, during the fourth phase, such requests
become more frequent and intrusive. Methods range from pop- up surveys and
“landing pages” that directly ask users about their voting intentions and geo-
graphical location, to demands for Facebook or Twier account information to
access certain content. Once collected, the new data are immediately added to
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
40 WHEN THE NERDS GO MARCHING IN
40
the parties’ voter les to be pored over and dissected by newly appointed teams
of computer and behavioral scientists. e results of these analyses are then fed
back into the micro- targeting process in order to produce yet more accurate
online and oine voter contacts.
e expertise and technical support required to deliver phase IV campaigning
produces a massive expansion in the budget and personnel allocated to digital.
is growth in sta numbers sees an increasing dierentiation and specialization
in the roles they perform. Long- standing “public” facing jobs such as home page
and email management remain core activities but are now joined by the new
“back- end” analytics and infrastructure development tasks. Beyond any restruc-
turing in size and shape, however, perhaps the biggest changes in the campaign
organization occur within the higher echelons of management and the culture of
decision- making. e importance of eld experience and intuition in campaign
planning are increasingly diminished as a new set of social, data, and information
scientists introduce a more data- driven approach to understanding and predict-
ing voter behavior, which forms the basis for key decisions in eld operations,
media advertising, and fundraising eorts (Issenberg, 2012; Nickerson and
Rogers, 2014; Anstead, 2017). e “Martians,” as Peggy Noonan describes this
new alien breed of campaign operatives who arrived on the US election scene in
2012, have indeed landed.
In structural terms, the central marker of phase IV is the rise of the digital
team to the apex of the organization. e director assumes the role of overall
campaign manager, ousting the more seasoned old- style “politicos” who were
typically drawn from the eld side of operations. At the same time as their power
increases, their identity as a discrete subunit within the campaign organization
becomes blurred. While they remain in charge of the “web” campaign writ large
(i.e., maintenance of the online interface and platforms to interact with voters),
the role of digital expertise now diuses out into other areas of operation. Digital
now plays a role in decisions over media advertising, GOTV eorts, and fund-
raising. e distinction between online and oine campaigning eectively now
disappears as the former subsumes the laer to deliver a new science of voter
mobilization.
Stepping back to prole the changes occurring during phase IV in the
three broader “power” domains of campaign activity identied in Table1.1 in
Chapter1, it is clear that the pendulum is once again on the move. At the sys-
temic level, the resources required to run such a highly data- intensive mode of
campaigning means that levels of inter- party competition return to a more nor-
malized state of aairs. e bigger parties are the only organizations that are real-
istically capable of developing and running these complex voter- management
systems. At the intra- organizational level, national- local relations show a similar
shi toward normalization as the new tech gurus take control over key decisions.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
41
A Review of the Literature 41
at said, however, the valuable new reach and resources created through the
local volunteer groups formed in phase III may lead them to resist full centraliza-
tion. One possible outcome of this tension is the emergence of a new “franchise”
style of party organization that comes to the fore, particularly during election
periods. Under this new model, central party sta retain control over the allo-
cation of resources and direct the campaign eort nationally. Meanwhile, local
groups gain considerable autonomy over tactics and implementation of those
plans, through their access to a new suite of purpose- built online resources.
Finally, the third “power” domain of elite- voter communication also sees the
shi back toward a more normalized top- down conditions. Two- step communi-
cation ow is still encouraged, but now it is more precisely controlled, targeted,
and monitored, using specialist apps and smart soware. Voters are understood
essentially as the recipients of these micro targeted messages, and interaction is
for the purposes of extraction of personal information, rather than the promo-
tion of dialogue and participatory engagement.
PHASE IV DIGITAL CAMPAIGNING INPRACTICE?
To date, phase IV digital campaigning has yet to emerge in most countries,
letalone operate as a norm. Given its reputation for electoral innovation it is
not surprising, however, that the United States is seen as a leading nation in this
regard. While there had been ongoing aempts by the Democratic National
Commiee (DNC) to institute a more “data- driven” style of campaigning
within the party over the preceding decade,12 it was the presidential election of
2012 and particularly the eorts of Barack Obama that rst really put the new
modus operandi of electioneering on the national and international news radar.
According to most observers, the campaign was a “game changer” for digital,
elevating it into a pivotal position within the wider decision- making hierarchy.
Post- election media reports contained extensive coverage of the radical overhaul
of campaign hardware and soware that took place and the new laser- like focus
on seeking out and mobilizing undecided voters.13
Academic experts were in agreement that the 2012 US election marked a
watershed moment, not just for digital campaigning, but for electioneering
practice more generally. For Vaccari (2013), 2012 signaled the ultimate “com-
modication” of online campaigning, the point when the practice nally became
aligned and subsumed into the main purpose of electioneering— voter mobili-
zation. Bruce Bimber (2014) agreed, arguing that
Obama’s digital media strategy...meant that there was no “online cam-
paign” or “Internet campaign” that stood in distinction to the “oine”
campaign. ere was a single campaign pursuing traditional campaign
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
42 WHEN THE NERDS GO MARCHING IN
42
goals in a way that aggressively exploited the communication environ-
ment associated with digital media. (134– 135)
In practical terms, one of the clearest indicators that the United States, and
specically the Democrats, had entered phase IV by 2012 could be seen in the
dramatic growth of the digital team, which increased almost threefold in size
compared to four years earlier.14 Along with this expansion in sta, there was an
emergence of new specialist sub- teams— tech and analytics. It was this develop-
ment and particularly the job advertisement to recruit employees for the laer
that prompted Peggy Noonan’s claims of an alien takeover of US politics.
e end result was a voter mobilization eort that achieved an unprecedented
level of individual granularity and precision— according at least to the post-
election confessions of those who helped to deliver it. Forecasts of the vote split
at the precinct level were reportedly accurate to within less than one percent-
age point, down to the town hamlet level in New Hampshire.15 e campaign’s
eorts at micro- targeting via email also reached a new and unprecedented level
of accuracy, according to Sasha Issenberg, allowing the Obama team to locate
individual Democrat voters amidst a previously unnavigable aggregate sea of red
Republican voters.16 Inevitably, questions have been raised since the election as
to the real eciency of Obama’s data- crunching machine and how far it actu-
ally secured his victory (Sides and Vavreck, 2013).17 Taking these criticisms on
board, however, it does seem that the US Democrats’ claims to be one of rst
parties to have entered phase IV are entirely legitimate. We probe these develop-
ments in greater depth in Chapter7, and return to take a closer look at how far
these practices became embedded post- 2012, in the concluding chapter.
Summary andConclusions
is chapter has reviewed the literature on digital campaigning since the eld
of study emerged in the mid- to late 1990s. It has shown that, cumulatively,
the work forms a narrative that identies four main phases of development—
experimentation, standardization and professionalization, community build-
ing and activist mobilization, and most recently, individual voter mobilization.
While the borders of the phases are uid, there are key changes in the physical
and technical infrastructure used to support and run the campaign and the level
of strategic thinking that, we argue, break down in a stepwise manner. At a more
abstract level, each phase can also be associated with a gravitational pull toward
one of two dominant power “logics”— equalization or normalization.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
43
A Review of the Literature 43
Whether all countries will cycle through all four phases and reach the point
of data- driven individualized voter mobilization is open to question. However,
what is evident is that if the phases are sequential and the process of adaptation
and innovation set out in Table1.1 continues, then countries are likely to end
up in an even more normalized condition, or what we are calling a state of hyper-
normalty. is is a campaign environment in which the major parties are the
only ones realistically capable of running a competitive campaign, in which cen-
tral elites control overall strategy, allowing some role for local activists to shape
local tactics, and communication with voters is personalized via algorithms
and any interaction is entirely instrumental, with the goal of extracting further
data. Akey twist on this process is that the intra- organizational concentration of
power does not lead to a reinforcing of conventional political consultants and
experienced politicos and activists within the campaign hierarchy. Instead, it is
bringing to the fore a new breed of apoliticos:individuals who are drawn from
the tech industry, and who have expertise in data management and analytics and
ideally a background in computing and behavioral sciences, but who lack signi-
cant exposure to elections and campaign experience.
Having now outlined the key phases in digital campaigns’ development, the
next chapter turns to examine the question of which countries have moved most
rapidly through the cycle, and what might help to explain the varying pace of
development that we observe.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
44
When theNerds Go Marching In. Rachel K. Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
3
Digital Campaigning acrossSpace
The Role of Technological, Political, and Institutional Context
(WITH ROSALYND SOUTHERN)
Chapter 1 set out the core theoretical argument of this book— that digital
campaigning has occurred in four main phases— with the most recent transi-
tion bringing the shi into a new state of hypernormality in which major party
dominance is secured, and a new power or apolitical elite sit at the heart of cam-
paign management. Chapter2 put some esh on the bones of this argument by
showing how the ndings from national studies of digital campaigning over time
align with the four- phase model, and provided some preliminary evidence about
where we are seeing the most advanced phase IV mode emerging. is chapter
switches the focus from examining developments in digital campaigning over
time to those occurring over space. How do nations compare in terms of their
location within the evolutionary cycle, and why do some countries appear to be
progressing faster than others?
Measuring how far and fast countries have progressed through the four- phase
cycle is of course not a straightforward task. Tracking the extent to which digi-
tal campaigns meet the criteria listed in Table1.1 in Chapter1 is dicult for
one country, letalone for multiple cases. Identifying a smaller set of “critical”
criteria on which national campaigns can be compared is thus a necessary rst
step in this process. We focus specically here on the demand- side characteris-
tics of each phase, given their greater amenability for purposes of international
calibration and comparison. Levels and modes of voter engagement with the
campaign can be measured relatively easily with relevant cross- national survey
data. Measures of the internal power distribution and primary goals of political
organizations, as well as specics about the location and size of the digital team,
are considerably harder to ascertain with any precision. Given that we are par-
ticularly interested in comparing rates of progress through our four- phase cycle,
we can further constrain the analysis to examine indicators of the shi into the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
45
Digital Campaigning across Space 45
demand side of phase IV of digital campaigning (i.e., the extent to which voters
have “received” online contact during recent national elections).
Who Has Entered Phase IV?
To map the extent of online contact1 occurring across countries, we make use
of a unique international data source— the Comparative Study of Electoral
Systems (CSES). e CSES is a post- election survey that is elded in national
elections over a xed time period, and includes a series of standard questions
about political aitudes and behavior.2 e study has a set of core questions and
a variable thematic component that changes over time. Module 4, which was
elded in elections between 2011 and 2015 in participating countries, focused
on the theme of mobilization and includes a baery of items that measure
whether respondents were contacted during the campaign by political parties
(direct) or by friends and family (indirect). e contact is also divided accord-
ing to whether it was online or oine. e online mode is split further into three
main types— email, SMS/ text messages, and web- based methods, including
social networks/ micro- blogs such as Facebook and Twier. Oine forms of
contact are also split into three types— face- to- face canvassing, phone, and mail
(see Appendix 3.1 for the full text of the survey items).
At the time of this analysis, module 4 of the CSES was in its second release
and included data for 17 countries. In addition, the 2015 British Election Study
(BES) had been released and included the CSES module which had been elded
in the post- election mail- back/ web completion survey component. is meant the
mobilization data were available for the case of the United Kingdom, although they
were not integrated into the CSES. As such, it was possible to include the United
Kingdom in the descriptive stage of the analysis. Its addition was important, since it
meant that all four case studies examined in subsequent chapters could be compared
to one another, and ranked internationally on their levels of digital mobilization.
Table 3.1 presents the basic frequencies for the dierent modes of contact
across the 17 countries included in the CSES data set, and for the extra case of
the United Kingdom.3 e gures are the percentage of those receiving a par-
ticular mode of contact as a proportion of the sample or population as a whole.
e results are reported by country and by year of election.
Columns three and four Table 3.1 report the main variables of interest for this
chapter, which are the proportion of citizens that received any online contact from
the party or informally through their networks. Column ve reports the total pro-
portion of the sample that received either direct or indirect online contact. For
comparison purposes, we report the frequencies of direct contact by the parties
through more traditional oine modes (i.e., face to face and mail/ telephone) in
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
46
Table3.1 Online and Oine Campaign Contact or “Receive Mode” byCountry (%ofPopulation)
Country Election Year Online Direct Online Indirect Total Online Oine Direct
F2F Mail/ Phone
Total Direct Sign- up Online
Australia 2013 8 8 14 4 56 59 5
Austria 2013 6 3 8 9 15 17 6
France 2012 5 6 10 2 6 10 7
Germany 2013 7 2 8 11 52 56 4
Greece 2012 24 16 31 25 28 37 6
Ireland 2011 2 2 4 43 17 49 5
Iceland 2013 15 10 23 15 23 32 12
Japan 2013 2 2 3 9 29 32 3
Mexico 2012 14 8 18 24 23 41 9
Montenegro 2012 4 2 5 15 8 21 7
New Zealand 2011 10 9 16 14 70 75 5
Poland 2011 2 2 4 4 4 7 3
Serbia 2012 9 3 11 15 26 30 1
Switzerland 2011 6 4 7 7 25 27 na
Taiwan 2012 24 7 27 8 35 50 11
ailand 2011 0.3 0.2 0.5 13 0.6 15 3
United States 2012 17 23 34 8 35 37 10
United Kingdom 2015 12 6 16 21 48 55 7
Sources:CSES Wave 4 Release 2.0; data are weighted using the “original demographic” weight, % is calculated with DK and Refused coded as “0” rather than missing to ensure
consistency across countries; UK gures are from the BES 2015 post- election self- completion mailback/ web survey (incl. CSES Wave 4 module). Data are weighted using
“wt_ combined_ CSES.”
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
47
Digital Campaigning across Space 47
columns six and seven. We again cumulate the preceding columns to report the
proportion of voters receiving any type direct contact from parties and candidates
(i.e., online or oine) in column eight. e rates of online sign- up by voters to
receive party and candidate alerts and e- news by country are reported in the nal
column. is variable is useful in terms of showing prior levels of interest in the
online campaign. More importantly, perhaps, it provides a more accurate picture
of the overall level of genuine or “net” online mobilization that is occurring within
a country. Given the lack of national email or mobile phone contact lists, it is di-
cult for parties to make unsolicited aempts to contact voters in the same way they
can do using oine methods. Online contacting is thus more likely to be reported
by those individuals who have already signed up to follow a party or candidate via
social media or for email/ SMS alerts. Once we take into account that prior sign-
up, it becomes easier to assess what proportion was unsolicited.
Viewed together, these variables paint an interesting picture of variance and
similarity in the amount of dierent types of contact occurring across countries.
Comparing the rates of oine to online contacting, we can see that the laer is
typically much less common than the former. is is particularly the case if we
look at the rates of mail and phone contact. e situation for face- to- face con-
tacting is more even, with most countries seeing similar rates of online and in-
person contact occurring. In a few nations, such as Taiwan and the United States,
rates of online contact far exceed those of face- to- face. However, in Ireland the
situation is reversed, with respondents reporting much higher rates of personal
contact compared with digital methods.
Comparing the rates of online contact reported in Table3.1, it is clear there is
considerable variance in the amount of online contacting that is occurring across
countries. Looking at rates of direct online contact by parties, the table shows that
almost one- quarter of the population in Greece and Taiwan received some kind of
digital message about their vote from the parties or candidates in the lead- up to the
election. is compares to a low of less than one percent in the case of ailand.
Iceland and the United States follow in reasonably close proximity to the two top-
performing nations, with 15percent and 17percent, respectively, of their electorates
having receiving ocial e- contact. e United Kingdom and Mexico rank somewhat
lower, but still report around one in ten of their population as having experienced
some type of digital mobilization by parties during a recent election campaign.
ere is a fairly sizable group of countries that fall into a middle band of direct
online contact, with between 4 and 10percent of the population receiving some
kind of e- stimuli from parties. Interestingly, when we look at the correspond-
ing rates of sign- up by voters to receive online contact from the parties, there
does not appear to be any direct parity, beyond the case of Austria. Typically,
online sign- up rates are lower than levels of direct online party contact. is is
interesting in that it suggests that, on average, parties are reaching out beyond
their base to voters who have not shown a prior interest in hearing from them.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
48 WHEN THE NERDS GO MARCHING IN
48
Most notable in this regard are the United States, Taiwan, and Greece, where the
total amount of online contact is at least double the rate of online sign- up. In ve
countries, however, the opposite holds true: there are more people reporting
that they had signed up for party updates than actually received contact. is is
most pronounced in ailand, where around 10 times more people reportedly
signed up to receive campaign information than were actually contacted online.
While the gap in most cases is small, the disparity does indicate there were sev-
eral countries in which parties were failing to fully exploit their own email lists.
Overall, Table3.1 reveals that levels of indirect online contact are typically
lower than direct forms. e exception here is the United States, where rates
of this type of informal online exchange actually exceed those of formal party
mobilization. While we do not know the content of that exchange, and particu-
larly whether it involved the passing on of campaign information, such ndings
suggest that the two- step ow model of digital mobilization is more established
in the United States than is the case elsewhere.
More generally, the frequencies reported in Table3.1 are important in con-
rming that countries vary in their intensity of online campaigning and in the
extent to which informal aempts at online persuasion occur in national elec-
tions. We use these ndings to group countries into levels of digital mobilization.
Specically, we divide the rates of contact into four tiers based on the maximum
and minimum rates of total online contact reported in the penultimate column
of Table3.1. e results of this ranking are shown in Table 3.2.
e top tier contains those countries in which at least 20percent of the
population received some kind of online electoral contact, either from the
parties themselves, or mediated through their social networks. Four countries
meet this criterion— the United States, Greece, Taiwan, and Iceland. Dropping
down a notch, we nd that both antipodean democracies— Australia and New
Zealand— make it into the second tier. as do Mexico, the United Kingdom, and
Serbia. France just edges over the 10percent threshold, based on levels of direct
and indirect online contact during its 2012 presidential election. e third tier,
which includes those countries where between 5 and 10percent of the popu-
lation have received some type of online contact, is quite heavily dominated
by the northern European democracies of Austria, Germany, and Switzerland.
Finally, the lowest rates of online contact are seen in tier four countries. is
includes a somewhat random mix of older and newer democracies from varying
geographic regions.
Before moving on to try to ascertain what might help to explain this ordering
of nations, we provide some additional insight into the rates of contact observed
by breaking down the online contact received across countries into its three
component parts— email, SMS, and web/ social network. e results are shown
in Table 3.3, and are interesting in that they point to a dierent paern of contact
within the top- tier nations.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
49
Digital Campaigning across Space 49
In particular, we nd that the vast majority of contact occurring in the case of
Greece and Taiwan is via SMS. At least one in ve voters in Greece and Taiwan
received some type of mobilizing text message from the parties or candidates.
Within the United States, the story is somewhat dierent. Here we nd that
email is the most common channel that candidates and parties use for online
mobilization, along with the web and social networking sites. Iceland is more of
a hybrid case, with both email and text messaging proving popular. is dieren-
tiation among countries in the channels utilized for dissemination of messages
across the top- performing countries is interesting in that it reveals that digital
campaigning at its most pervasive does not necessarily follow a “one size ts all”
model. Technological capacity and voters’ existing preferences for digital com-
munication are likely to play a role in shaping how those messages are distrib-
uted. Furthermore, the dierences observed raise questions about the relative
ecacy and power of email and social network contacting versus text messages.
Table3.2 Rates ofOnline Mobilization byCountry
TIER 1 (20% or more receive direct/ indirect online contact)
• United States
• Greece
• Taiwan
• Iceland
TIER 2 (From 10% up to 20% receive direct/ indirect online contact)
• Mexico
• New Zealand
• United Kingdom
• Australia
• Serb i a
• France
TIER 3 (From 5% up to 10% receive direct/ indirect online contact)
• Austria
• Ger many
• Switzerland
• Montenegro
TIER 4 (Less than 5% receive direct/ indirect online contact)
• Ireland
• Poland
• Japan
• ailand
Source:Based on ndings from “Total Online” contact column of Table3.1. CSES Wave 4 (2nd
Release); UK BES 2015 (see Table4.2 for further details)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
50 WHEN THE NERDS GO MARCHING IN
50
Explaining Levels ofOnline VoterMobilization
Having proled the variance in levels of online mobilization occurring across
several democracies during comparable recent national elections, the question
then arises of how we explain these ndings. Aquick eyeballing of the rank-
ings does not yield an immediate answer. While a majority of nations in tiers
one and two of Table3.2 report high rates of internet use among their popula-
tions, Greece does not.4 Institutionally, the cases are also quite varied. Tier one
contains both two- party and multiparty systems, as well a mix of proportional
Table3.3 Mode ofDirect and Indirect Online Contact or ‘receive mode’
byCountry (%ofPopulation)
Online Direct Contact
Mode
Online Indirect Contact
Mode
Country Year Email SMS Web/ SNS Email SMS Web/ SNS
Australia 2013 5 1 4 1 1 7
Austria 2013 3 2 3 1 1 3
France 2012 4 1 1 3 1 4
Germany 2013 5 0 4 1 1 1
Greece 2012 9 21 10 5 10 10
Ireland 2011 1 2 0 1 1 0
Iceland 2013 10 9 6 3 2 8
Japan 2013 1 1 1 0.4 2 0.4
Mexico 2012 6 10 7 2 4 4
Montenegro 2012 1 3 2 0.3 1 1
New Zealand 2011 7 1 5 3 1 7
Poland 2011 1 0.4 1 2 1 2
Serbia 2012 1 8 2 0.9 2 1
Switzerland 2011 5 1 2 2 0.3 2
Taiwan 2012 3 21 4 2 3 3
ailand 2011 0.1 0.1 0.3 0.1 0.1 0.1
U.S.A 2012 15 4 9 11 4 17
United Kingdom 2015 9 2 5 1 2 5
Sources:CSES Wave 4 (2nd Release); UK BES 2015 (see Table4.2 for further details)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
51
Digital Campaigning across Space 51
representation (PR) and plurality electoral systems. e campaign environment
also diers quite markedly among the top tier countries. Candidates and parties
in the United States and Taiwan enjoy quite liberal regimes in terms of controls
on their advertising and expenditure. Greece and Iceland, by contrast, impose
stricter spending limits on parties and oer state- subsidized rather than paid
media airtime.5
To help explain the variance in online contacting observed in Tables3.1 and
3.2 and, by proxy, those factors most conducive to entry into phase IV digital
campaigning, we adopt a more systematic approach. Specically, we t a multi-
level regression model to the CSES data set, which means the units of analysis—
voters— are treated as individuals (level 1)nested within groups, in this case,
countries (level 2). To build our explanatory model, we combine the insights
from two relevant, but largely disconnected bodies of comparative literature.
e rst includes studies that have aempted to explain variance in digital cam-
paigning across countries from a supply- side perspective. e second is a larger
and more established corpus of work that has focused on explaining rates of
voter mobilization and reported party contact among the electorates of dierent
nations.
COMPARATIVE STUDIES OFDIGITALCAMPAIGNS
e literature examining dierences in the extent and style of web campaigning
across countries is limited, but increasing. e rst major study was conducted
by Norris (2001), at the turn of the millennium. Taking a large N approach, she
compared the presence and content of party websites across 179 countries in
mid- 2000. Her explanatory model included a range of party characteristics, along
with a number of aggregate indicators that measured countries’ levels of socio-
economic, human, and technological development. Her ndings were impor-
tant in showing that societal factors and particularly technological advancement
helped explain how active and particularly interactive parties were online.
e increase in party websites aer the millennium meant that subsequent
comparative research focused more on the interesting questions of how parties
campaigned online, rather than simply explaining rates of adoption. Aention
shied from the broader socioeconomic environment to the political context
and the institutional conguration of the polity. Anstead and Chadwick (2009)6
were among the rst to look these structural factors, identifying the party sys-
tem as one of the main causes of dierences in web campaigning across coun-
tries. rough a small N comparison of the United Kingdom and the United
States, they argued that the more “stratarchical” or decentralized mode of party
organization in the laer, along with its more liberal campaign nance rules, had
permied a much faster rate of diusion and innovation of digital technology
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
52 WHEN THE NERDS GO MARCHING IN
52
during elections. Together, these factors generated an environment in which
candidates functioned as political entrepreneurs, building “start- up” organiza-
tions for each election and drawing in a massive number of volunteers and dona-
tions. e internet was the ideal campaign medium to use in this context, given
its fast, networked, and personalized qualities.
e view that institutions maered for innovation in web campaigning was
supported by Davis etal. (2008) in their editorial overview of the ndings from
online elections in 12 dierent countries. Looking across the assembled cases,
they concluded that the presidential candidate- centered systems of the United
States and Chile provided the best environment for internet campaigning to
develop. By contrast, party- centered parliamentary democracies were less open
to the new media and the more personalized political appeals it promoted (Davis
etal., 2008). e authors also speculated on whether the more commercialized
and consumer- driven media systems in the United States and Chile encouraged
heavier use of the internet by both candidates and voters. e medium might
appeal as a more direct and less “noisy” channel for both providing and accessing
political information.
Despite the persuasive case that these impressionistic and small N accounts
put forward to show dierences in web campaigning across countries, larger
cross- national investigation did not support their conclusions. Studies of both
the 2004 and 2009 European Parliament (EP) elections found that macro- level
political factors explained very lile of the variance observed between politi-
cal campaign sites and particularly use of web 1.0 versus web 2.0 features (Foot
et al., 2007; Vergeer et al., 2012). Subsequent analysis of party elites’ views
on web campaigning from 12 EU member states conrmed a lack of national
dierences in terms of what they saw as its value and purpose. e only slight
dierence that emerged was that campaigners in the newer democracies were
somewhat more likely to favor Facebook compared to those reporting a longer
electoral experience (Lilleker etal., 2014).
Outside the European Union, Vaccari’s (2013) seven- nation analysis of
parties’ and candidates’ web and email use over a four- year period (2006–
2010) found that systemic factors had a stronger impact. In particular, he
found that PR and higher voter turnout were linked to the production of richer
web campaign content. Conversely, lower rates of turnout and voter trust were
associated with greater email responsiveness by parties. ese dierences, he
argued, showed that where parties enjoyed higher rates of popular support,
they were more likely to focus on producing a quality product for their visi-
tors. However, where they were viewed more negatively, they tended to deploy
resources in a more proactive manner to encourage familiarity and participa-
tion. Despite the greater prominence of macro- level variables in explaining the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
53
Digital Campaigning across Space 53
paerns of web campaigning in Vaccari’s analysis, his overall conclusion was
that they remained secondary in signicance to party- level eects. e “selec-
tive adoption of specic digital applications by political actors,” he argued,
remains “rooted mostly in organizational rather than systemic considerations”
(Vaccari, 2013:115).
e lack of variance in web campaigning revealed by these studies is perhaps
not too surprising given the uniformity in practice that previous chapters have
indicated was developing at this point. e early noughties were a time when
phase II was in the ascendant, and professionalization and standardization were
the modus operandi for web campaign managers. Added to this is the method-
ological consideration that most of these studies were conducted during EU par-
liamentary elections, which are widely seen as less important or “second order”
events (Reif and Schmi, 1980). Parties were arguably less likely to invest time
and resources into developing their campaign presence compared with “rst
order” national elections. Notably, the one study that did actually analyze web
and email use in national election campaigns found a stronger eect for systemic
characteristics.
A nal, and possibly even more compelling, reason for the apparent lack of
country- level variance identied by these accounts is that all of them focused on
the supply side of the equation (i.e., websites and email content). Convergence
among party elites on a new set of campaign practices is, according to “conta-
gion theory,” not an unusual occurrence (Matland and Studlar, 1996). Indeed,
one might expect this type of diusion and mirroring to be more likely to occur
in response to a global technology like the internet. Had these studies looked
instead for variance in the demand side of activity (i.e., voters’ experience of the
digital campaign, which is the primary concern of this chapter), then it is likely
that a more nuanced and dierentiated picture would have emerged. Certainly,
this is a conclusion that is supported by the comparative literature on voter
mobilization more generally. It is thus to this body of work and its key ndings
that we now turn.
COMPARATIVE STUDIES OFVOTERMOBILIZATION
Eorts to explain campaign contact and voter mobilization, as one might expect,
preceded the arrival of the internet. Indeed, investigation of parties’ eorts to
“get out the vote” (GOTV) has a relatively long history in political science,
stretching back to the experimental work of Harold Gosnell in the 1920s (1927).
Until recently, however, much of the work focused on a single case— that of the
United States (Rosenstone and Hansen, 1993; Gerber and Green 2000, 2008).
e work also focused largely on contact as an independent variable, and its
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
54 WHEN THE NERDS GO MARCHING IN
54
eects on voters, rather than as a dependent variable and the drivers behind it.
Despite some signicant methodological dierences in approach, most studies
pointed to two very clear and consistent conclusions. First, contacting voters
is eective in increasing turnout. Second, face- to- face methods are much more
eective than other less personal tactics, such as direct mail or phone canvassing,
in geing voters to the polls.
Comparative work began in earnest in the early part of this century as
international data sources began to emerge. Having shown the impact of
contacting on turnout so convincingly in single- nation studies, questions
invariably arose as to whether it could help explain cross- national paerns
in voting behavior. Initiatives such as the Comparative National Election
Project (CNEP) study and module 2 of the CSES provided scholars with
the tools they needed to address these questions. eir ndings were impor-
tant, rst, in underscoring the key nding that contact maers for turnout.
However, they also broke new ground in demonstrating the extent of vari-
ance in the practice across countries (Magalhães, 2007; Karp etal., 2008;
Karp and Banducci, 2011; Dalton etal., 2011; Karp, 2012; Magalhães etal.,
2015). In one of the rst large N analyses of elections held during the 2001–
2004 period, Karp and Banducci (2007) reported a gap in contact rates of
over 40percent between the “top” performing nation, Ireland— where over
half (56.3percent) of respondents were contacted during the campaign—
and Spain, where just 5.8percent of individuals reported receiving any elec-
toral stimuli. e gures measured all forms of contact, since this earlier wave
of CSES did not dierentiate according to mode.
Such revelations switched the focus on contacting from an independent to
dependent variable, and explanatory models were developed to account for
the dierences observed. ese models included a range of institutional, cul-
tural, and socioeconomic variables similar to those that had been deployed
by the cross- national analyses of web campaign production discussed earlier.
e results revealed a much stronger role for system- level characteristics,
however, with citizen orientations and governance structures emerging as
key determinants of mobilization rates (Karp and Banducci, 2007; Karp etal.
2008; Dalton etal., 2011; Karp, 2012).7 Specically, countries with higher
rates of turnout, single member districts (SMD), or preferential voting sys-
tems and two large centrist parties, typically saw the highest rates of contact.
Such conditions, it was argued, increased GOTV eorts since it maximized
the likely “pay- o ” of such activities for the larger parties in terms of “poach-
ing” one another’s voters. Where there was a wider range of smaller parties
and more polarized electorates, the likelihood of such conversion was seen as
much lower.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
55
Digital Campaigning across Space 55
Toward anExplanatory Model
ofDigitalMobilization
Joining these two literatures together provides the basis for developing and test-
ing a comparative model of digital mobilization, which is the central goal of this
chapter. In particular, the ndings from the mobilization studies are central in
showing that party contact varies signicantly across countries, and that these
paerns are linked to structural as well as cultural aspects of the polity. We would
thus expect this to carry over into studies of digital contact. e comparative
analyses of elite adoption of digital tools revealed much less cross- national varia-
tion; however, where systemic eects were detected, their impact ran counter to
what had been observed in the mobilization studies. Specically, Vaccari (2013)
found that parties were signicantly more likely to invest in their digital cam-
paign content under proportional electoral rules rather than plurality or rst past
the post systems. Furthermore, cultural features such as higher levels of citizen
political engagement, which had been positively associated with parties’ oine
mobilization activities, appeared to have a less intuitive and even inverse rela-
tionship with parties’ use of some more interactive digital campaign tools. us,
while our explanatory model is based primarily on the theory and ndings from
the voter mobilization literature, we do consider the potential for dierences in
the direction of and impact of certain independent variables.
Following the approach taken in previous studies, we start by specifying the
“baseline” or capacity- related factors that are likely to determine the amount
of digital mobilization occurring in a country. is leads us to measure three
“standard” indicators of human, economic, and political development. Societies
with higher levels of human and economic capital are more likely to have the
infrastructure and levels of literacy required for modern GOTV activities to take
place. In addition, those countries with a longer track record of stable demo-
cratic rule and a history of free and fair elections will have parties that are more
experienced in running campaigns and mobilizing citizens. e fourth “base-
line” indicator we include is technological, and measures the extent of internet
use among the population. In order for citizens to receive online contact, they
need to have access to the medium.
Alongside these “necessary” conditions for digital mobilization to occur, a
number of politically relevant characteristics are expected to help increase its
frequency. First, the political culture of a country is seen as important, and
particularly whether a positive orientation exists toward governing authorities
and the system more generally. e expectation of the mobilization literature
is that a more politically engaged electorate will increase mobilization eorts
by the parties since they perceive a more receptive audience for their GOTV
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
56 WHEN THE NERDS GO MARCHING IN
56
messages. Indicators typically used to capture engagement are voter turnout
rates and levels of political trust and interest in society. Of course, a possible
counterargument that can be made here is that lower levels of engagement, and
particularly sudden declines in turnout, could spark increased eorts by parties
to get voters out to the polls. Such a response was supported by Vaccari’s (2013)
cross- national analysis of parties’ use of email during election campaigns. Levels
of existing political engagement are thus expected to have a signicant impact
on the extent of digital mobilization; however, the direction of the eect is le
unspecied.
Given the ndings from prior studies, regime characteristics are expected
to play a major role in determining levels of digital mobilization. Electoral sys-
tems are expected to be particularly relevant, with eects varying according to
whether they are based on preferential voting SMD, or PR methods in which
voters choose from a party list. Ceteris paribus, preferential systems are expected
to generate higher rates of contacting since they encourage, or even necessitate,
that individual candidates cultivate a personal vote and name recognition. e
more personalized mode of communication provided by websites and social
media proles arguably increases these incentives even further. at said, and
again based on the ndings from Vaccari’s analysis (2013), it may be that the
more equalized environment of PR increases the incentives for the smaller par-
ties to campaign online, given its relatively low cost. As such, while we expect
the electoral system to signicantly predict rates of digital mobilization, we are
agnostic about which type is most eective in doing so.
e electoral cycle, type of contest, and presence of compulsory voting are
also expected to aect the rates of digital campaign contact. Ahigher frequency
of elections is likely to produce a faster cycle of innovation and investment in
voter contacting overall, particularly in new methods. e type or level of elec-
tion being contested is also expected to maer. Presidential contests are likely
to increase rates of voter mobilization, over and above other types of national
elections. Such races have the highest status and prole for voters and typically
see the biggest turnout. Investment in mobilization is thus likely to yield a higher
return for parties. Again, given the opportunities that the digital medium pro-
vides for personalized campaigning, one might expect presidential elections to
see a particularly high investment in online mobilization. Finally, compulsory
voting is likely to increase all eorts to contact voters, including those using the
internet. Given that the odds of contacting someone who will actually vote are
obviously higher in countries where voting is mandatory, the potential benets
of voter mobilization eorts are correspondingly increased.
In terms of the party system eects, these are likely to be seen in two main
and interrelated dimensions— size and ideological dispersion or spread. Amul-
tiparty environment with a wide spread of ideologically opposing parties is, in
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
57
Digital Campaigning across Space 57
principle, likely to prompt high levels of campaign contact. However, as noted,
the ndings from comparative studies of traditional modes of voter mobiliza-
tion have consistently challenged this logic, nding that centrist two- party sys-
tems are in fact more likely to see higher rates of contacting. Given the empirical
evidence reported by Vaccari (2013), linking PR systems to higher web cam-
paign intensity, we retain the expectation of a positive relationship between the
number of parties and their level of ideological spread, and the extent of digital
voter mobilization. Athird and nal aspect of the party system that scholars of
comparative voter mobilization have argued is an important predictor of con-
tact rates is its overall strength. Extending the logic used to argue for preferen-
tial voting eects, the case is made that where party organizations are weak and
conversely candidates dominate, contact will increase given campaigns’ need to
cultivate a personal following and support base.
e nal “layer” of explanatory variables modeled here relates to the cam-
paign environment itself. Perhaps the most obvious and self- explanatory of
these is the competitiveness of the contest. e closer a race is, the more the
contestants are likely to try to galvanize support, and hence, the higher the levels
of voter mobilization will be. e rules governing campaign nance and expen-
diture are also expected to play an important role, with less regulated systems
expected to see a higher rate of voter contacting. Such systems provide par-
ties with more resources and freedom to undertake GOTV drives. Of course,
tighter spending restrictions might also make parties more eager to exploit the
new digital channels given their cheaper cost. As such, we may see an inverse
or negative relationship between nancial caps and online contacting in par-
ticular. e controls governing campaigns’ access to the media are also likely to
aect the level of voter outreach during an election— again in mixed ways. More
restrictive regimes (i.e., those where commercial advertising is prohibited and
state- subsidized media are the main channel for parties to communicate their
message) are expected to increase the incentives for parties to develop their own
direct channels to talk directly to voters. However, it is also possible that access
to a paid media environment could increase aention to digital contacting since
this allows for the cross- promotion and recirculation of the audiovisual material
produced for these other markets.
Beyond these “usual suspects,” a further layer of regulation in the area of data
protection and privacy is now arguably becoming necessary to include in studies
of voter mobilization and particularly those dealing with digital targeting of mes-
sages. To develop this point, we draw on Bimber’s (2014) study of the 2012 US
presidential election in which he noted that the level of personalized communi-
cation that occurred during this campaign was “not mirrored in other countries
at this point.” Akey reason for the disparity, he argued, was that “privacy regu-
lations prevent parties and candidates in many countries from engaging in the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
58 WHEN THE NERDS GO MARCHING IN
58
practices undertaken in the U.S., especially in 2012” (132). Eectively, American
candidates now enjoy an unmatched capacity to drill down and micro- target the
electorate through their extensive and increasingly rich voter les. e rapid
spread of smartphones and uptake of platforms like Facebook and Twier pro-
vide a new and very amenable conduit for this new type of individualized con-
tacting. e capacity to build and use these growing stores of data on citizens is
thus becoming a critical factor to consider in explaining cross- national paerns
of “receive” mode in any format, but particularly for digital messaging. Amea-
sure of the extent of countries’ data privacy and protection rules is therefore an
important new parameter to add to our explanatory model.
Table 3.4 presents a summary of the explanatory model of digital mobiliza-
tion developed to this point. Specically, it lists the independent variables we are
testing by category or layer of explanation, proceeding from the baseline prereq-
uisites through to the more xed and variable political characteristics. For each
variable, we indicate whether we expect a positive or negative impact on digital
mobilization based on our review of the extant literature. Where the literature
gives conicting or unclear expectations, we place a question mark next to the
variable.
TESTING THEMODEL
e core data set used in this analysis is the CSES module 4 (release 2.0). As well
as providing data on the level of receive mode across countries (i.e., our depen-
dent variable), CSES provides a range of individual- and macro- level variables
that map onto the explanatory model set out in Table 3.4. Where there are gaps
in the CSES, we turn to other international data sources. Full details of the data
sets used and variables operationalization are listed in Appendix 3.2.
DependentVariable
Since we are using multilevel modeling with voters (level 1), nested within coun-
tries (level 2), the dependent variable— receive— is measured at the individual
level. Specically, we combine responses to the CSES questions of whether a
voter received online political contact from a party or through the voter’s social
networks to create a new binary variable of “total online contact” (see Appendix
3.1 for full details of question wording of each component). is ensures that we
include the ocial contact coming directly from the parties, as well as the medi-
ated or two- step ow that is particularly associated with phase III.
IndependentVariables
Our model species both the systemic and individual- level characteristics
that are likely to determine receipt of online campaign contact. With regard
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
59
Digital Campaigning across Space 59
to individual- level variables, we are limited to the standard CSES socio-
demographic and political indicators. Based on the ndings from prior research
on comparative mobilization (reviewed earlier) we expect that age, sex, trade
union membership, education, income, and level of partisanship will be associ-
ated with receiving campaign contact. We add the online sign- up variable to this
prole. is variable is unique to module 4 and, as discussed earlier, is likely to
be highly correlated with receiving online contact. Including it on the le- hand
side of the equation, therefore, allows us to take into account this endogeneity,
and thereby generate a “cleaner” measure of the inuence of the other indepen-
dent predictors.
At the aggregate level we use a range of sources to measure the variables listed
in Table3.4 for the countries included in the analysis.8 We begin with the baseline
conditions, specied as levels of human, economic, and political development.
ey are measured using the UN index of Human Development scores, GDP
Table3.4 Explanatory Model ofComparative Digital Mobilization
Societal development
• Human development (+ve)
• Economic development (+ve)
• Political development/ institutionalization (+ve)
• Technological development (+ve)
Political culture
• Voter turnout (?)
• Political trust (?)
Electoral system
• Candidate centered (i.e., preferential SMD) (?)
• Frequency of elections (+ve)
• Presidential election (+ve)
• Compulsory voting (– ve)
Party system
• Size (+ve)
• Ideological polarization(+ve)
• Strong (– ve)
Campaign context
• Competitiveness of the election (+ve)
• Spending limits (?)
• Paid advertising (?)
• Subsidized/ state- provided media (?)
• Data protection laws (– ve)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
60 WHEN THE NERDS GO MARCHING IN
60
per capita at the time of the election in question, and age of democracy, respec-
tively. e fourth baseline variable— level of technological development— is
measured as the number of internet users per 100 of the population, and as the
natural log of this gure. e logarithmic transformation was applied, given that
the original values of variable were quite negatively skewed, with over 80percent
of countries having above 50percent of the population online. Aer the trans-
formation, the data were smoothed to reduce a disproportionate impact of the
lower values on the outcome variable.
To measure the political culture or democratic “health” of a country, we
used two aggregate measures of political engagement. e rst was aitudinal
and measured the proportion of respondents who said they were “very” or
“somewhat” interested in politics in each country. Since CSES data lacked an
appropriate question, the gures were calculated from a combination of the
World Values Survey (WVS) wave 6 (2010– 2014) and the European Social
Survey (ESS) round 7 (2014). Both sources were used to avoid the problem
of missing data since no single survey covered all of the countries included
in this analysis. e item on political interest was the most comparable and
relevant aitudinal measure available across the two surveys in terms of the
question stem and response categories. e second aggregate measure of
engagement was behavioral and measured average turnout in a country dur-
ing the post– World War II period. Given that we know some countries had
compulsory voting systems in place, we added this variable to the model as
a control.
e electoral system was operationalized as a binary variable based on
whether the country operated a PR system or a preferential SMD model for its
legislative or parliamentary elections. An indicator of the average number of
candidates per electoral district or district magnitude was included. is pro-
vided an additional test as to whether a personal or party vote was driving levels
of mobilization. Binary variables were used to indicate if the election in ques-
tion was for the presidency, and whether compulsory voting was in operation.
Finally, a measure of the number of months since the last national election was
included as a test the impact of a faster electoral cycle.
The fourth layer of explanation— the party system— was modeled using
three variables. The first of these was the size of the system, which was mea-
sured through the effective number of electoral parties (ENEP) as recorded
in the CSES macro file. Second, the level of ideological polarization of the
party system was measured using Dalton’s formula based on the average left-
right placement of parties in the CSES micro file. Finally, we included an
indicator of the strength of the party system, which was based on the extent
to which it was considered as party- or candidate- centered (Carey and
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
61
Digital Campaigning across Space 61
Shugart, 1995). For this we used scores on the index produced by Johnson
and Wallack (2012). See Appendix 3.2 for full details of the sources and
variable computation.
e h and nal bloc of variables measured the impact of the campaign
context and regulatory framework surrounding it. e competitiveness of the
campaign was measured by the margin of victory (in percentage terms) for
either the leading party or presidential candidate in the election under analy-
sis. Controls on spending and media advertising were taken from information
contained in the Institute for Democracy and Electoral Assistance (IDEA)
Political Finance Database. Specically, we coded countries according to
whether they imposed caps on election spending, and whether free or sub-
sidized media advertising was provided to parties and candidates. Finally, to
measure controls over parties’ use of citizen data in campaigns, a ve- point
additive index was created using the DLA Piper Data Protection Laws fact
book. is included indicators of whether the country has an ocial chief
information ocer or data commissioner, whether it mandates and enforces
data- protection laws, and how far it has regulated online privacy. For full
details of index construction, see Appendix 3.2.
Before undertaking the analysis, a number of adjustments were necessary to
ensure the modeling was as robust and reliable as possible. e rst and most
important of these was to enter the macro- level independent variables in a selec-
tive and sequential manner. Given the relatively small set of countries we had
available at the time of analysis, we could not test all the level 2 variables identied
in Table3.4 in a single model. We thus tested the impact of each bloc consecu-
tively and selected the most signicant predictive variables to include in a nal or
cumulative model.
A second major modication was to exclude Switzerland from the analysis
due to missing data. Specically, the online sign- up variable data, which we
considered a vital control variable, were not included in the Swiss study. is
reduced our overall country N to 16, which remains within the range of an
acceptable number of level 2 units.9
Because the outcome variable is binary, we used logistic multilevel mod-
eling. The analyses were performed with STATA 12, using maximum like-
lihood estimation10 and with fixed slopes and intercepts. We opted to fix
these parameters given that we had no a priori grounds to expect that the
impact of the independent variables (either individual or systemic) would
vary significantly in strength or direction across countries. Also, fixing
these parameters improved the efficiency and robustness of our model esti-
mation since it reduced the number of coefficients that it was necessary to
estimate.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
62 WHEN THE NERDS GO MARCHING IN
62
Findings:What Factors Are Driving Phase IV
ofDigital Campaigning?
In the following, we present the results of the analysis and draw out their implica-
tions for the drivers of digital campaigning. In particular, we start to build a pro-
le of the factors that seem most likely to facilitate a country’s entry into phase
IV. e rst set of results are reported in Table 3.5, where we show the ndings
for the ve separate explanatory blocs. Each model includes the same range of
individual- level predictors and dependent variable, but varies the explanatory
group or bloc of independent variables.
Looking rst at the individual- level factors, we can see a common paern
across all the models in terms of the demographic and political prole of the
online contactees. Gender is important, with men being more likely to be tar-
geted than women. Education has a consistently strong eect in that those with
the highest levels of education are up to 10 times more likely to receive online
contact than those with no qualications. Age is also important, with younger
people being signicantly more likely to receive political contact via digital
channels. is runs counter to the ndings from studies of oine and more tra-
ditional forms of mobilization, and suggests that parties may be engaging new
voters through digital channels. is “youth” eect is countered to a degree by
the fact that both union membership and partisanship are positively and signi-
cantly associated with receiving online contact. Both of these characteristics are
more typical of older voters. Finally, as one might expect, the online sign- up vari-
able has a very strong and positive coecient. Having signed up to receive party
messages makes someone around 5 times more likely to report online contact.
e fact that the other independent variables remain signicant aer controlling
for its eects, however, is important. It suggests that the parties are managing
to reach beyond their “usual suspects” (i.e., their existing online support base).
Model 1 presents the results for the baseline or societal development group of
variables. None of the variables is statistically signicant at the .05 level, although
the log of internet use comes close, hovering at just above the 0.10 level.11 Such
results suggest that the intensity of digital campaigning is not directly dependent
on the overall socioeconomic performance of a nation or the longevity of its
democratic experience. e extent of internet use within society and levels of
technological take- up within the wider populace, however, may have some eect
on the level of online contacting taking place.
Findings from the second explanatory bloc of political culture variables
(model 2) are more compelling, with one variable— political interest—
emerging as statistically signicant. Interestingly, its eect is negative, indicat-
ing that, in line with Vaccari’s (2013) conclusions about email responsiveness,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
63
Digital Campaigning across Space 63
Continued
Table3.5 Five Preliminary Logistic Regression Models Predicting Online
Contact
MODEL
(1)
Social
Development
(2)
Political
Culture
(3)
Electoral
System
(4)
Party System
(5)
Campaign
Context
Female 0.89
(0.04)***
0.89 (0.04)
***
0.89 (0.04)*** 0.90 (0.04)* 0.89
(0.04)***
Union Member 1.13 (0.07)* 1.13 (0.07)*** 1.14 (0.07)*** 1.16 (0.07)* 1.14
(0.07)*
No Quals (Ref)
Early
Childhood
0.73 (0.47) 0.73 (0.47) 0.73 (0.47) 0.24 (0.19) 0.73
(0.47)
Primary 2.78 (1.29)* 2.78 (1.30)** 2.78 (1.29)* 1.37 (0.72) 2.78
(1.29)*
Lower
Secondary
3.20
(1.48)***
3.12 (1.48)** 3.20 (1.48)** 1.14 (0.60) 3.20
(1.48)**
Upper
Secondary
4.47
(2.07)***
4.47 (2.43)*** 4.47 (2.07)*** 1.66 (0.86) 4.48
(2.07)***
Post-
Secondary
5.22
(4.42)***
5.23 (2.76)
***
5.25 (2.44) *** 1.99 (1.05) 5.23
(2.43) ***
Some Tertiary 5.92
(2.76)***
5.93 (3.29)
***
5.95 (2.77) *** 2.28 (1.20) 5.95
(2.77) ***
Bachelor’s 7.12
(3.30)***
7.12 (3.30)
***
7.13 (3.30) *** 2.74 (1.42)* 7.11
(3.29) ***
Master’s 9.25
(4.31)***
9.26 (4.31)
***
9.30 (4.33) *** 3.52 (1.83)* 9.26
(4.31) ***
Doctoral 9.73
(4.61)***
9.71 (4.61)
***
9.78 (4.64) *** 3.73 (1.98)* 9.69
(4.60) ***
Age 0.98
(0.00)***
0.98 (0.00)
***
0.98 (0.00)*** 0.98 (0.00)
***
0.98
(0.00)***
Online Sign- up 4.78
(0.30)***
4.77 (0.30)*** 4.77 (0.29)*** 4.83 (0.00)** 4.77
(0.29)***
Close to Party 1.68 (0.09) 1.69 (0.09)
***
1.69 (0.09)*** 1.71
(0.26)***
1.69
(0.09)***
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
64 WHEN THE NERDS GO MARCHING IN
64
Table3.5 Continued
MODEL
(1)
Social
Development
(2)
Political
Culture
(3)
Electoral
System
(4)
Party System
(5)
Campaign
Context
UN Dev Index 0.99 (0.00)
Log internet use 1.85
(0.80)~
GDP per cap 0.99 (1.27)
Age of Regime 1.00 (0.00)
Interest in pol 0.96 (0.02)*
Avg Turn 1.01 (0.02)
Comp Vote 1.06 (0.52)
Electoral system 0.48 (0.21)~
Months last elect 0.98 (0.01)
Pres Elect 5.30 (2.42)***
Comp Vote 1.21 (0.42)
District Mag 0.98 (0.05)
ENEP 1.00 (0.00)
Polarization 1.07 (0.26)
Party Cent 0.62 (0.17)
Party Cap Spend 0.75
(0.34)
Free Sub Med 1.49
(0.98)
Margin <2% (Ref)
2%– 3% 0.46
(0.32)
4%– 10% 0.75
(0.53)
>10% 0.15
(0.11)***
Data Privacy 0.92
(0.16)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
65
Digital Campaigning across Space 65
Table3.5 Continued
digital mobilization appears to occur more frequently in countries where citi-
zens are less positively disposed toward politics. is relationship could be
explained in one of two possible ways. First, one could argue that it indicates the
perceived weakness and lack of eectiveness of digital campaigns. Essentially,
where parties nd themselves facing a losing bale to persuade voters, they are
more likely to resort to cheaper online means. Amore positive interpretation,
however, is that where parties perceive more disconnection from politics among
the electorate, they invest more in newer methods to signal a shi from “poli-
tics as usual.” Either way, the ndings are somewhat surprising given the widely
accepted view among scholars of mobilization that it is generally geared toward
the more engaged and easily persuadable voters.
e ndings from model 3 provide some of the strongest evidence for macro-
level eects on the rates of digital mobilization. e variable indicating whether
the election in question was a presidential race emerges as one of the most sig-
nicant predictors of receiving online contact in any of the models tested. ese
higher prole and more personalized elections, as expected, provide a strong
stimulus to internet- based forms of voter contact. Beyond the nature of the
oce being contested, the only other coecient in the bloc that approaches
statistical signicance is the electoral system. Again, in line with the ndings
of Vaccari (2013), it appears that online campaign contact is intensied by PR
MODEL
(1)
Social
Development
(2)
Political
Culture
(3)
Electoral
System
(4)
Party System
(5)
Campaign
Context
N Individual
(Groups)
22,640 (16) 20,845 (16) 22,640 (16) 22,640 (15) 22,640
(16)
Constant 0.002
(0.00)***
0.090 (0.14) 0.063(0.05)*** 0.131 (0.17) 0.09
(0.08)***
Chi20.00 0.00 0.00 0.00 0.00
Random E Paras 0.92 (0.17) 0.90 (0.17) 0.74 (0.14) 0.98 (0.19) 0.77
(0.15)
Note:N of 16 is due to exclusion of Switzerland which did not include the sign- up question which
formed a key control variable. e N of 15 for the party system model is due to exclusion of Taiwan
which did not ask the Le- Right party placement question required for the calculation of the
polarization score.
Source:CSES Wave 4 (2nd Release); Signicance levels ***=≤.001, **=≤.01, * =≤.05; ~=≤.10
Dependent variable— Online contact (direct and indirect).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
66 WHEN THE NERDS GO MARCHING IN
66
systems. Even taking into account the weakness of the relationship, the nding
is still important in further supporting the idea that the dynamics driving online
contact dier from those shaping prior oine mobilizing eorts. Aempts at
voter persuasion through these more traditional methods were typically more
common within preferential voting systems.
e variables measuring the impact of the party system (model 4)appear
to be uniformly unimportant in predicting online contact. is is somewhat
surprising, particularly with regard to the size variable given the long- standing
association of internet technology with a more diverse and equalized party envi-
ronment. Given the ndings of model 3, such a nding could be seen as conrm-
ing that it is the incentives of the electoral system that maer most to stimulating
digital contact, rather than simply the number of parties competing. When
smaller parties have a more realistic prospect of gaining oce, they invest the
resources required into online voter outreach. e results from model 5, which
examines the impact of the campaign variables, are similarly non- compelling,
with one clear exception— the competitiveness of the race. If an election was
highly one- sided (i.e., the margin of victory was more than 10 percent), this
resulted in signicantly less online voter contact than in those races where the
gap was smaller than 10percent.
e fact that the regulatory environment has no apparent impact on rates
of digital contacting is somewhat surprising. e lack of impact of the rules
on spending and advertising is perhaps more easily explained, given the rela-
tively “immature” status of the internet as an election technology, and its con-
sequent immunity from regulations designed for an earlier media era. e lack
of any impact for the data privacy index is, however, less easy to account for,
and suggests that some revision and renement of the measure may be required.
In particular, as currently dened it does not capture the controls on political
organizations’ use of voters’ personal data, which are likely to dier from those
associated with commercial and nongovernmental entities. In addition, there
may be a gap between the ocial standards governing how individuals’ data
are handled, and the extent to which these rules are enforced and implemented.
is disjuncture is something we explore further in the case studies that follow.
Based on the preceding analyses, we extracted ve macro variables that are
signicant or of close signicance in explaining levels of online mobilization.
is includes the log of internet use, levels of political interest, whether the elec-
tion was presidential, the electoral system (preferential vs. PR) and margin of
victory. ese ve predictors were then carried forward to our nal multilevel
model. Table 3.6 presents the results of that consolidated analysis.
As we might expect, the individual variables continue to behave in the same
manner shown in the previous models. e macro eects also remain largely con-
sistent, although there are some changes in the strength of their eects, mostly in
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
67
Table3.6 Final Logistic Regression Model Predicting Online and Oine
Contact
Online Face to Face Mail/ Phone
Female 0.89 (0.04)*** 0.87 (0.04)*** 0.99 (0.02)
Union member 1.13 (0.07)* 1.34 (0.08)*** 1.33 (0.05)*
No Quals (Ref)
Early Childhood 0.73 (0.47) 1.32 (0.52) 1.38 (0.33)
Primary 2.78 (1.29) 1.72 (0.58) 2.14 (0.38)***
Lower Secondary 3.21 (1.48)** 1.56 (0.53) 2.33 (0.41)***
Upper Secondary 4.50 (2.07)*** 1.80 (0.60) 2.81 (0.49)***
Post- Secondary 5.25 (4.44)*** 1.77 (0.61) ~3.04 (0.55)***
Some Tertiary 5.99 (2.76)*** 1.85 (0.63)~3.41 (0.62)***
Bachelors 7.15 (3.31)*** 2.04 (0.68) * 3.29 (0.59)***
Masters 9.32 (4.34)*** 2.45 (0.85) *** 3.94 (0.72)***
Doctoral 9.73 (4.61)*** 2.38 (0.89) * 3.40 (0.67)***
Age 0.98 (0.00)*** 1.00 (0.00) 0.99 (0.00) ***
Sign- up 4.77 (0.30)*** 1.88 (0.14)*** 1.81 (0.11)**
Close to Party 1.69 (0.09)*** 1.64 (0.09)*** 1.62 (0.05)***
Pres Elect 5.20 (2.70) *** 5.15 (2.28) *** 12.84 (10.28)***
Margin <2% (Ref)
2%– 3% 0.26 (0.13)* 0.27 (0.13)*** 0.13 (0.11)*
4%– 10% 0.22 (0.13) ** 0.33 (0.16)*** 0.07 (0.07)***
>10% 0.23 (0.11) *** 3.34 (1.38)*** 1.35 (0.99)
Elec system 0.45 (0.20)~0.12 (0.05) *** 0.30 (0.21)~
Log Internet Use 1.64 (0.38)* 0.88 (0.17) 1.12 (0.39)
Interest in Pol 1.00 (0.02) 0.98 (0.01) 0.98 (0.02)
N Individual (Groups) 22640 (16) 22640 (16) 22640 (16)
Constant 0.01 (0.02)*** 0.333(0.47) 1.193 (2.98)
Chi20.00 0.00 0.00
Random E Paras 0.55 (0.11) 0.46 (0.09) 0.84 (0.16)
Source:CSES Wave 4 (2nd Release); Signicance levels ***=≤.001, **=≤.01, *=≤.05,=≤.10
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
68 WHEN THE NERDS GO MARCHING IN
68
an upward direction. Most notably variables measuring the closeness of the race
are now even more signicant and their eects are more ne- tuned. us, while
a very uncompetitive race still sees the least eort by parties to engage in digital
mobilization, an exceptionally close race (where there was less than a 2percent
margin of victory for the winner) does produce a signicantly higher investment
of resources. Interestingly, the increased and more nuanced eect of electoral
competitiveness is accompanied by a reduction in the impact of political inter-
est, which is now no longer signicant. e coecient for the log of internet use
is also boosted, and is now signicant at the .05 level. Similarly, the eect of a PR
electoral system increases, although it remains on the margins of signicance.
e main eect, however, still centers on the presence of a presidential election.
e coecient indicates that voters in these elections are ve times more likely
to be contacted online than voters in nonpresidential elections.
Summary andConclusions
e empirical aims of this chapter were twofold:rst, to identify where the
fourth phase of digital campaigning is emerging most strongly in global terms;
and second, to understand why some countries were more actively embrac-
ing this new style of campaigning than others. To investigate these questions,
we compared countries on the extent of internet- based contact that voters
reported receiving from parties and through their social networks during an
election campaign. For this purpose we used comparative survey data from
module 4 of the CSES. Receipt of targeted digital messages, as Chapter1 made
clear, is an important demand side indicator of phase IV “data- driven” cam-
paigning. e ndings largely endorsed the impressions generated through
the literature review in the previous chapter. Specically, the United States
emerged as a leading nation in terms of the frequency of online voter contact,
although this occurred in large part through social networks and indirectly,
rather than direct messaging from the parties and candidates. ere were other
strong performances by some of the more “wired” countries in Southeast Asia
and Northern Europe. By contrast, ailand, a young democracy with very
low technological capacity, had almost zero capacity for digital campaigning.
Among these more intuitive ndings, however, there were also a few notable
surprises. is included the location of several auent and established democ-
racies, such as Germany and Japan, toward the tail end of the distribution.
Conversely, the top- tier position of Greece proved somewhat puzzling given
its lack of prole in the wider literature on digital campaigning.
To help interpret the results, a systematic analysis of the individual and
contextual factors associated with higher rates of digital mobilization was
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
69
Digital Campaigning across Space 69
undertaken. e results of this exercise were revealing on a number of counts.
At the individual level, it was clear that there are certain types of voters who
are more likely to receive online contact during the campaign. is included,
perhaps most obviously, those who had already signed up to receive this type of
stimuli. Beyond these self- selection eects, however, factors such as being male,
younger, and more highly educated were also signicant in increasing the likeli-
hood of receiving online contact. Being aached to a political organization, such
as a party or a trade union, further increased the possibility of experiencing this
type of mobilization. With the exception of age, therefore, the results suggested
that parties were not reaching a new and more under- mobilized segment of the
population with their digital campaign eorts. Closer analysis of whether this
paern holds up for individual nations is undertaken in the country- specic
chapters that follow. At the system level, the results were signicant in showing
that higher rates of online contact are linked with the institutional conguration
and a certain level of technological prowess. Countries with very competitive
presidential elections and a more wired citizenry are typically most encouraging
of this new type of mobilization.
Such conclusions are of interest for this chapter. First, they help to shed some
light on the factors that encourage higher use of digital mobilization tactics by
parties, and thus by proxy, entry into phase IV campaigning. In addition, they
clearly challenge the ndings of earlier cross- national analyses of web cam-
paigning which had concluded that it was largely unaected by systemic factors.
Once we examine web campaigning from the voters’ experience or the demand
side, however, we can see that its intensity and preferred mode do vary across
countries, and that this variance is explicable by reference to the wider political
context in which it occurs. Finally, we also show that the systemic and individual-
level factors associated with online contact dier in subtle, but signicant ways
from those that associated with oine mobilization.
NEXTSTEPS
While the analysis presented in this chapter goes some way to explaining
advances in digital campaigning, and specically a country’s entry into phase
IV, it clearly faces limitations. From a purely methodological standpoint, it is
evident that the multilevel models tested here explain only a limited proportion
of the overall variance in the dependent variable.12 Furthermore, the sample size
in terms of number of level 2 units or countries included in our multilevel model
meant that it was not possible to test all the macro- level variables simultane-
ously. Subsequent releases of the CSES data will oer the opportunity to test a
more fully specied model.13 Even with a larger number of countries included,
however, the analysis is still likely to face the omied variable problem. e level
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
70 WHEN THE NERDS GO MARCHING IN
70
of internet use in a society, for example, though a useful proxy for the degree of
familiarity and openness of citizens to using the technology, does not fully cap-
ture the broader cultural propensity to use it for political purposes. is might
help to explain Greece’s very high rates of digital contact, despite having a lower
level of internet usage overall. Similarly, parties’ long- standing historical asso-
ciations with certain types of campaign media, such as direct mail or television,
may create a path dependency that accelerates or reduces their willingness to use
internet tools. Measuring and modeling these traits for a small number of coun-
tries is challenging; converting them into reliable and comparable indicators to
insert into larger N cross- national analysis is nigh on impossible.
Finally, and perhaps most importantly, the data provide only a snapshot of
current levels of digital contact across countries. To base our conclusions about
a country’s entry into phase IV and the extent of digital campaign innovation
occurring on this one measure in time is clearly placing a lot of pressure on one
aspect of what we have argued is a multidimensional phenomena.
In the chapters that follow, we aempt to address these gaps by providing a
more detailed historical account of digital campaign developments in four of
the nations examined here. Specically, we examine one country from the top
tier of mobilization— the United States— and three countries that occupy con-
secutively lower positions within tier two— Australia, the United Kingdom, and
France. In each case, we start by assessing how far each country’s ranking aligns
or conforms to the “ideal” macro conditions identied in this chapter. e corre-
spondence (or lack thereof) with our expectations provides the basis for a more
nuanced account of the conditions surrounding the growth of digital politics
in that particular country. In particular, we look for those factors that appear to
have played a role in constraining or accelerating the uptake of new media in
campaigns.
We then document developments in digital campaigning in more detail in
these nations over time, working forward to show the changing nature of inter-
net use by parties and candidates in successive elections. We do so through the
lens of our four- phase model. is provides both a rich understanding of key
developments in digital campaigning in each country, and a framework for com-
paring the pace and nature of that process of change. How far do developments
in these four countries align with the generic cycle outlined in Chapter1, and
where does each now sit, in evolutionary terms?
e case- study approach also allows us to take a closer look at changes in
the supply and the demand dimensions associated with each phase. In particu-
lar, it is possible to investigate the role of parties in driving the cycle. Are some
parties keener to embrace the new techniques and push into phase IV than
others? Does this vary with size, incumbency, or ideology, and is this the same
across countries? On the demand side, the case studies allow us to build a more
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
71
Digital Campaigning across Space 71
nuanced picture of the changing nature of voters’ engagement with the digital
campaign, based on the response modes outlined in Table1.1. We can measure
when citizens started to switch from simply reading about the campaign online,
to undertaking redistributive activities, and the extent to which they have been
targeted by digital campaign messages over time. Do these changes at the mass
level correspond to changes in the paerns of supply? At the party level, are dif-
ferences in approach to the medium reected “on the ground”? If leist parties
are more active in promoting online community- building, for example, are their
supporters more likely to engage in more redistributive activities?
In short, the case studies allow us to drill down below the “top- line” impres-
sion of Chapter2 that digital campaigning has evolved through four phases, to
see how well this holds up at the individual- nation level. We are also able to sup-
plement the quantitative ndings from Chapter3, identifying the factors driving
these developments, with a more nuanced qualitative understanding. Beyond
the institutional and technological context, what other less measurable factors
help to drive innovation in digital campaigning in a nation?
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
72
When theNerds Go Marching In. Rachel K. Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
4
The SlowBurner
Digital Campaigning in the United Kingdom
In this chapter, and the three that follow, we examine developments in digital
campaigning in more detail at the level of the individual nation. ese case stud-
ies serve several purposes. First, they allow us to trace developments in digital
campaigning in a “real world” context over time and thereby validate our four-
phase model of change. Second, they show how particular events, parties, and
politicians have played a role in driving that process of change. Finally, they allow
us to map and compare the responses of the electorate to developments in digital
campaigning. To what extent do changes in the demand side of the equation
map with supply? Do we see a progression in the mode of voter engagement
from “read” to “redistribute” and “receive,” as the four- phase model suggests? Is
this movement across the board, or is it concentrated among certain party sup-
porters? What evidence exists to suggest that online campaigning can make a
dierence to parties’ and candidates’ electoral fortunes?
The United Kingdom asa Context
forWebCampaigning
e United Kingdom provides a useful starting point to explore and validate our
four- phase model of web campaigning in more depth. As an established democ-
racy with a signicant majority of the population online, it meets the baseline tech-
nological requirements that Chapter3 showed were important in driving digital
campaigning. Figure 4.1 charts the growth in internet access within the United
Kingdom since 1997 and reveals how usage had reached a critical mass within
just two electoral cycles. Aer 2005, rates of use begin to level o, with almost
universal coverage achieved by 2015. As Figure 4.2 reveals, much of the growth
in recent years has occurred via broadband, with mobile broadband subscriptions
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
73
100
90
80
70
60
50
40
30
20
10
0
1997
7
34
70
85
89
2001 2005 2010 2015
Figure4.1. Growth of internet use in UK general elections, 1997– 2015 (%of population
using the internet).Sources:1997– 2010:World Bank, “Internet Users (per 100 people),” hp://
data.worldbank.org/ indicator/ IT.NET.USER.P2?page=1; 2015:Oce of National Statistics (ONS)
Statistical Bulletin, “Internet Users,” hp:// www.ons.gov.uk/ ons/ dcp171778_ 404497.pdf
1
16
31
38
44
90
0
10
20
30
40
50
60
70
80
90
100
2001 2005 2010 2015
Broadband (fixed line) Broadband (mobile)
Figure4.2. Growth in broadband subscriptions (per 100 inhabitants) in UK general
elections (2001– 2015).Source:OECD historical xed and mobile broadband penetration
subscriptions (per 100 inhabitants). Figures are from 4th Quarter (rounded to nearest %). hp://
www.oecd.org/ internet/ broadband/ 41551452.xls hps:// www.oecd.org/ sti/ broadband/ 1.5-
BBPenetrationHistorical- Data- 2016- 12.xls (Figures for mobile are available only from 2009 Q4
onward.)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
74 WHEN THE NERDS GO MARCHING IN
74
expanding dramatically aer 2010. By 2015, OECD (Organisation for Economic
Co- operation and Development) gures show the extent of use in the United
Kingdom was almost equivalent to one for every person in the population.
Despite having a supportive technological environment, as Table 3.2 in
Chapter3 reports, the United Kingdom is not in the top tier of nations when it
comes to rates of digital mobilization. is is not particularly surprising given the
ndings from the previous chapter. As a parliamentary democracy with a plurality-
based electoral system, the United Kingdom does not meet the structural criteria
that were associated with higher intensive rates of digital campaigning.
ere are other, more “localized” aspects of the UK electoral system that
arguably also dampen campaigners’ enthusiasm for the new media. Most
notably, there is very strong commitment among British parties to more tra-
ditional methods of voter outreach. Elections are typically dominated by “old
style” door- to- door canvassing, or “knocking up” the vote at the local level
(Whiteley and Seyd, 2002; Denver and Hands, 1997, 1992). is remained
the case even aer the advent of television and growing centralization in par-
ties’ management of the campaign. is long- standing aachment to more
personalized canvassing methods has meant that British elections are typi-
cally highly labour- intensive aairs. e role of technology has largely been
one of supporting the main “ground war” eort, rather than forming a voter
interface in its own right. Denver and Hands’s (1997) in- depth analysis of
local campaigning during the 1992 general election was highly instructive on
this point, showing how, even at the end of the twentieth century, these ai-
tudes persisted. According to their results from their survey of local electoral
agents, the main benet of personal computers was to take the “drudgery”
out of campaigning and perform the “routine tasks” such as preparing the lists
of voter names and addresses for the canvassing eort (51).
As this chapter goes on to reveal, this view of technology as a “supplemen-
tary” tool has undergone a substantial revision in the United Kingdom as par-
ties have moved more fully into the internet era. e historical commitment to
a more “hands- on” style of voter mobilization, however, is clearly likely to have
slowed the adoption of these newer, more remote forms of contact, such as
email, SMS, and social media messaging. In the sections that follow, we describe
that process of inertia, adoption, and growing enthusiasm in more detail.
Phase I:Experimentation (1994– 1997)
e United Kingdom was one of the rst countries where parties’ adaptation
to the internet during and outside of election periods became a subject for
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
75
The Slow Burner 75
systematic analysis (Gibson and Ward, 1998; Ward and Gibson 1998; Gibson
and Ward 2000b). According to these early accounts, the rst moves by parties
to establish websites took place in the mid- 1990s, with the period leading up to
the election of 1997 seeing a urry of adoption. By polling day, just over 30 par-
ties were recorded as having a site, up slightly from 28 the year before (Gibson
and Ward, 2000b:115– 16). e idea that an “internet election” was about to
occur was widely anticipated in the press, and by the parties themselves. Paddy
Ashdown, the leader of the Liberal Democrats, went so far as to proclaim that
the web was “the future of communications and information” and would help
to create a society that was more “democratic, open, and in which power is as
decentralised as possible.”1
Despite this initial optimism and excitement, a surprisingly large number
of minor parties resisted the move into cyberspace. Areview of parties’ online
presence during the campaign found that at least one- third of those competing
did not have a national site. Of those that had entered the fray, at least two par-
liamentary parties— Plaid Cymru and the Social and Democratic Labour Party
(SDLP)— failed to produce any election- specic pages on their sites (Ward and
Gibson, 1998). Even the two major parties, the Conservatives and Labour, failed
to create simple and easy to nd domain names.2
In terms of content, the sites were rudimentary, although there were some
signs of experimentation with interactivity, particularly among the smaller par-
ties. e Liberal Democrats, for example, provided a plethora of email addresses
across their home pages, including a personal appeal from leader Paddy Ashdown
for visitors to get in touch. For the most part, however, sites were used as static
“repositories” for information, and lile use was made of their two- way commu-
nication facility or capacity for real- time response to events (Ward and Gibson,
1998). e lack of content signaled a broader lack of understanding among the
parties about their reasons for being online. Indeed, according to close observers
of the election, the main motivation for launching a site was the fear of being le
behind (Jackson, 2007; Ward and Gibson, 1998:95).
Organizationally, the resources directed to the digital campaign were mini-
mal. Although a small number of parties engaged the services of an external web
company, the majority made use of existing sta who mixed it in with their other
duties. Surveys of party ocials at the time revealed that only one party— the
Northern Irish Democratic Unionists (DUP)— had conducted any analysis of
their audience using trac statistics (Gibson and Ward, 1998). At the local level,
activity was virtually nil. What did exist went largely unnoticed by central party
sta, with post- election interviews revealing the lack of any formal procedures
for reviewing or approving sub- national sites (Gibson and Ward, 1998).
One year after the 1997 election, a comprehensive survey of party activ-
ity revealed that sites had largely fallen into disuse, with some parties failing
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
76 WHEN THE NERDS GO MARCHING IN
76
to maintain any type of web presence post- election (Auty and Nicholas,
1998)At the local level, there continued to be a dearth of provision. A1998
survey of intra- party sites in the United Kingdom reported only around
5percent of local parties were online and that just 4percent of members
of Parliament (MPs) had an individual website (Gibson and Ward, 1999).
The advent of the 1999 European Parliamentary election stimulated a mod-
est burst of activity. However, the amateur approach persisted, with non-
intuitive URLs and broken or missing links proving commonplace. Both
Plaid Cymru and the SNP actually took their sites offline for significant
periods of time in the weeks and even days before the election for redevel-
opment purposes. Even among the major players, there were signs that the
medium was still seen as something of an afterthought. Labour, in particu-
lar, failed to list its website address in any of its offline campaign material
or election broadcasts. Given that the site was subsequently revamped and
relaunched in time for the party conference later that year, such an omission
may have also helped to avert attention from its rather underwhelming qual-
ity (Gibson and Ward, 2000a).
Phase II:Standardization and
Professionalization (2001 and 2005)
While 1997 may have been heralded as the rst “internet election” in chron-
ological terms, the election of 2001 was seen as the rst realistic opportunity
for the internet to have an impact on the campaign and the election outcome.3
Internet use (as shown in Figure4.1) had increased signicantly to around one-
third of the population, and there was a growing buzz around its power to appeal
to younger voters. ere was also a growing awareness among the parties of its
potential to help their campaigns in more marginal constituencies (Ward and
Gibson, 2003:193).
In practical terms, the election of 2001 saw a shi by parties toward a more
professionalized model of web campaigning. e three main parties all employed
external agencies to develop and maintain their sites, following a competitive
tendering process. Overall, it was estimated the parties spent around one million
pounds on their new media eorts (Crabtree 2001:8). Internally, the human
resources devoted to the e- campaign increased compared to 1997 levels, par-
ticularly for the two biggest parties. Labour and the Conservatives each had four
full- time sta working on their web campaigns, with several more temporary
personnel draed in from external agencies. In the case of the Conservatives, a
further 24 people were added to the core team. e Liberal Democrats reported
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
77
The Slow Burner 77
a much smaller scale deployment, with one full- time Internet campaign man-
ager managing two volunteers and two agency sta (Bowers- Brown and Gunter,
2002: 171). Beyond the three big parties, there was a renewed push by the
remaining stragglers to get online, and several of the more prominent minor par-
ties launched dedicated election sites, albeit several days aer the start of the
ocial campaign (Ward and Gibson, 2003).
Websites remained the centerpiece of the campaign, and email sign- up
facilities started to appear as a way for voters to keep in touch with party
news. Some parties also provided an option for voters to sign up to RSS feeds
and mobile access to their sites via wireless application protocol (WAP) and
personal digital assistant (PDA) technology. e appearance of sites was
upgraded and more closely integrated into the party’s brand and use of other
media and public relations channels (Auty and Cowen, 2001; Bowers- Brown
and Gunter, 2002). Domain names were streamlined and became more intui-
tive. e address www.conservatives.com was established as the primary URL
for the Tories’ home page. In a further sign of the mainstreaming of the digi-
tal medium, the major parties began to use their sites to criticize the opposi-
tion, with Labour focusing on Tory policy weakness, while the Conservatives
launched personal jibes against prominent Labour and Liberal Democrat poli-
ticians (Auty and Cowen, 2001:347).
Design- wise, a common template began to appear. is included an extensive
use of frames and menu bars to divide up the content and help visitors nd what
they were looking for. Figures 4.3 and 4.4 show the updated, more structured,
look of the Liberal Democrats’ and Conservatives’ sites as they headed into the
general election of 2001. In terms of content, journalists formed a key audience,
with most parties seing up some type of news or media center to host press
releases. ere were also sections containing proles of leaders, as well as policy
and conference information. Participatory opportunities were rare, and what
was oered tended to be of the “controlled” variety (e.g., feedback forms, games,
and search tools, rather than being interpersonal). One notable exception to
this was the Liberal Democrats’ live webcast interview with its leader, Charles
Kennedy, who participated in an online question- and- answer session (Auty and
Cowen, 2001; Gibson etal. 2003).
Despite the lack of opportunities for genuine interaction on parties’ home
pages during the election, there was growing evidence that they were beginning
to see the value of the new medium for supporter communication and mobi-
lization. Member- only areas were developed and “action” buons encouraged
people to sign up as volunteers or make a donation. Early aempts at indirect, or
two- step mobilization, were led by Labour, which developed a range of emails
and text messages for supporters to distribute. e vote- geing properties of
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
78
Figure4.3. e UK Liberal Democrats Party home page ( June 2001).Source:Wayback
Machine:hp:// web.archive.org/ web/ 20010601170822/ hp:// www.libdems.org.uk/
Figure4.4. e UK Conservative Party home page (May 2001).Source:Wayback
Machine:hp:// web.archive.org/ web/ 20010525234331/ hp:// www.conservatives.com/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
79
The Slow Burner 79
the medium were also now starting to be explored, albeit somewhat tentatively.
Labour went aer the youth vote, launching Ruup4it.org, an independent cam-
paign site that focused entirely on younger voters. e Liberal Democrats went
one step farther, making explicit pleas for tactical voting by their supporters
by providing links into vote- swapping sites. Asimple correlational analysis by
Gibson and Ward (2003) revealed a higher frequency of web campaigning in
the more competitive seats, suggesting that candidates were beginning to take
the medium more seriously as an election tool.
ese eorts were, in reality, a drop in the wider ocean of party indif-
ference. Local- level activity had increased since 1997, but use of the web
remained a minority pursuit. Only a quarter of siing MPs were found to
have a website in the lead- up to the election, and there remained a majority
of constituencies in which zero online campaigning took place. According to
the audit by Gibson and Ward (2003) of the 539 English constituencies, only
around one- quarter of candidates from one of the three main parties compet-
ing could lay claim to having some type of web presence, and of those that
did set up sites, a majority failed to update them at all during the campaign.
Perhaps even more telling was the fact that none of the national parties (with
the exception of Labour) managed to update their sites on election day itself
(Auty and Cowen, 2001).
More generally, it was clear that parties still lacked a rm understanding of
the benets of the internet as a campaign tool (Coleman, 2001a:680). If the
1997 election had been largely about staking a claim to turf in cyberspace,
the 2001 election centered on cultivating and improving their new piece of
virtual real estate— the primary aspiration or target being to bring it up to the
standards required for existing “broadcast” media channels. is view was
reinforced by post- election reports of a lack of “buy- in” at senior levels within
the Labour Party over the value of the web.4 Indeed, several years later, aer
leaving his post as chief media advisor to Tony Blair, Alastair Campbell freely
confessed to never having using the internet during his tenure at Number
10. Perhaps less surprisingly, Blair himself and Deputy Prime Minister John
Presco were also happy to disclose their “technophobic” tendencies to the
media.5
Four years later, a new wave of optimism arose that this would be “third time
lucky” for the UK parties in terms of geing an eective internet campaign o
the ground (Jackson, 2007). Expectations were high, given this was the rst
election in which a majority of voters were online (see Figure4.1). Also broad-
band access now exceeded dial- up for the rst time, which meant that voters
could enjoy a much faster and richer web experience.6 Among the parties, there
were encouraging signs of new interest and investment in the medium. Almost
all parties now had some type of online presence, and an increasing number
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
80 WHEN THE NERDS GO MARCHING IN
80
were making use of a new type of interactive digital platform— the weblog,
or blog (Francoli and Ward, 2008; Gibson etal., 2013). In what was perhaps
the strongest sign of a new strategic direction in their use of the medium, the
Conservatives launched their online recruitment tool, “Conservatives Direct,”
in the year before the election. e new platform was designed specically to
aract new activists and volunteers to help them target marginal seats (Ward
etal., 2008). Not to be outdone, Labour reported that 50,000 people had signed
up to their new national “supporters network” through their website since 2001.
is inux, they claimed, had helped them raise as much as £100,000 in email
donations since the last election.7
Despite these reasons to expect a signicant shi of gears by 2005 and entry
into a new more activist- oriented phase of digital campaigning, closer inspection
of the parties’ eorts concluded that the narrative of normalization persisted.
One study reported that less than a quarter (24%) of the 107 parties compet-
ing had actually launched an explicit e- campaign ( Jackson, 2007). Among
the parties that did make an eort, the verdict was lukewarm at best (Stanyer,
2007). For one seasoned observer of British e- democracy, the sites were lile
more than “dreary e- versions of the throwaway paper brochures of yesteryear”
(Coleman, 2005: 5). e emphasis remained on static “vertical communica-
tion,” with freshly updated news and press releases seen as a “centerpiece” (Ward
and Gibson, 2008:11).8 While some of the minor parties did gain special men-
tion for their more serious aempts to promote a two- way dialogue through
discussion boards, aempts to convert that discussion into action were thin on
the ground (Gibson and Ward, 2005; Jackson, 2007; Ward etal., 2008). Despite
the main parties’ bid to increase their online volunteer networks and email
lists, there was lile evidence that made any extensive use of them. e Liberal
Democrats were most active on this front, sending a total of 17 emails to voters
during the course of the campaign (Jackson, 2007:257). In what can be seen
as perhaps the most telling gap in performance, none of the parties launched a
YouTube channel, despite the platform having been established three months
earlier.
At the local level, although activity reportedly increased, the lack of commit-
ment and interest was still palpable. Estimates by Ward etal. (2008) were that,
at best, two- hs of candidates (37%) had developed some type of personal-
ized web presence, a 10percent rise from 2001. Among local parties, the picture
appeared to be even more dire. Work by the New Politics Network (NPN) in the
year prior to the election was damning in its conclusions about the state of local
parties’ online presence. According to the authors, “most local football fan clubs
have more sophisticated and beer maintained websites than their equivalent
local parties” (Runswick etal., 2004:10).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
81
The Slow Burner 81
Phase III:Community Building and Activist
Mobilization (2010)
e fourth time around, anticipation for an election breakthrough by the inter-
net again started to build.9 Facebook and Twier had now emerged on the
scene, joining YouTube as the most popular social media platforms among
voters. Facebook was seen as a particularly important arena for the parties to
occupy, given its estimated 23million UK users.10 e runaway success of tech-
savvy candidate Barack Obama in the US presidential election of 2008 served to
increase the pressure and expectations on UK parties to show their digital prow-
ess, particularly in terms of exploiting the mobilizing power of the new medium.
For their part, both major parties had taken steps to improve their credentials for
online public engagement since 2005. e Conservatives had funneled major
resources into a new interactive video channel centered on their new leader—
WebCameron— which was launched to great fanfare at their 2006 annual confer-
ence. As well as introducing the relatively unknown David Cameron to the wider
public, the site was also designed to signal his more open and accessible style of
leadership. is was also part of a wider push by the Tories to soen their public
image and rid them of the self- imposed label as the United Kingdom’s “nasty”
par ty.11 Labour opted for a less personalized “face” for their online engagement
strategy, launching a new virtual policy discussion platform— Let’s talk— in early
2006, shortly aer the election. As its name suggested, the site was designed to
initiate wider discussion between the party and the public about the party’s
future direction and policies.
In addition to these in- house eorts at promoting more voter interaction,
both major parties and most of the minor players started to make use of social
media well in advance of the 2010 election. e Greens proved to be the early
adopters of Facebook, seing up their pages in late 2007. e Conservatives
and Liberal Democrats followed suit in early 2008, while Labour brought up
the rear, launching its site in August 2008 (Bartle etal., 2013). Twier han-
dles were also set up from mid- 2007 onward, and here Labour was very much
at the forefront of the action. e party designated MP Kerry McCarthy as its
ocial “Twier Tsar,” her main mission being to guide and support candidates
and leaders in their use of the micro- blogging site (Newman, 2010:26). Several
ocial hashtags, such as #mobmonday and #labourdoorstep, were established
prior to the election to help coordinate supporters’ online and oine activities.
Even self- confessed techno- phobe Deputy Prime Minister John Presco joined
the social media fray, seing up a personal blogspot, Go4th, to support Labour’s
bid for a fourth term.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
82 WHEN THE NERDS GO MARCHING IN
82
Home pages also underwent signicant overhauls, with a much greater
emphasis given to content that promoted direct action rather than static
downward information dissemination. Calls to volunteer and donate featured
prominently, and signicant sums were spent revamping existing electronic
voter canvassing tools.12 e remodeling of the Conservatives’ main website,
Conservatives.com, alone was estimated to have cost around a quarter of a mil-
lion pounds (Crabtree, 2010). All the parties now made much greater use of
email lists compared with 2005. ere were claims that together the two main
parties had collected around 800,000 addresses. Messages were tested for their
eectiveness, and it was estimated that the parties raised one million pounds in
response to their online fundraising pleas (McGregor, 2010). e Conservatives
were regarded as particularly adept in this regard, with one email from William
Hague estimated to have generated £100,000 in one day (Newman, 2010:24).
Arguably, the most visible sign of UK parties’ entry into phase III was their
launch of several “home- grown” versions of Obama’s social networking site
MyBarackObama.com, otherwise known as MyBO, in the year before the elec-
tion. As with the original, sign- up was quick and simple, requiring just an email
address, postal code, and password. While such ease of access would not be any-
thing particularly unusual for US voters, it marked a much greater departure in
practice for the UK electorate. British parties, like many of their European coun-
terparts, operated a system of formal membership. e arrival of these virtual
networks meant that now anyone with a modicum of interest and an internet
connection could become ocially aliated with the party, and undertake cam-
paign activities on its behalf.
Among the parties, Labour made the rst foray into virtual community build-
ing and activist mobilization with the launch of LabourSpace shortly aer the
2005 election. is was an online platform that was open to both members and
non- members and allowed people to sign up and then nd ways to take action
to help the party promote its priority policy areas. is was followed by more
issue- specic platforms, such as Eds Pledge which focused on environmental
policy, and Back the Ban which sought to prohibit fox- hunting. ese eorts
culminated in 2009 in the conversion of their members- only internal social
network— Membersnet— into a public platform that anyone could join. Having
the existing infrastructure in place clearly made the roll- out of a MyBO- like site
a fast and largely painless technical exercise. Sign- o on the decision to do so,
however, needed to come from the top, and was given aer a series of trips by
leading party ocials to the United States during 2008, particularly visits to the
Democrats’ campaign headquarters (Straw, 2010; Anstead and Straw, 2009).
Despite this trigger, the switch did not bring about any radical changes in the
look and feel of the site, which retained its name and commitment to the Labour
“brand,” as Figure 4.5 reveals.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
83
The Slow Burner 83
e Conservatives’ move into MyBO territory marked more of a new direc-
tion for the party. e launch of MyConservatives.com in September 2009 con-
stituted the most explicit eort by any party to copy the Obama original, both
in name and overall visual design. is “foreign” or imported quality, combined
with criticisms that its rollout had been rushed in time for the election, prompted
accusations that it was largely a promotional stunt (Ridge- Newman, 2014:32).
Its designers, however, were eager to stress that the site formed part of a deeper
commitment within the party to use digital tools to broaden its appeal and
widen its activist base. e “Online Communities” team that led the initiative,
according to Ridge- Newman (2014), saw the site (and WebCameron) as sym-
bolizing a new spirit of “cyber toryism” that was designed to give “...a greater
and more independent voice. ..[to].. .ordinary Conservative participants”
(Ridge- Newman, 2014:115). While MyConservatives.com had shallower roots
than Membersnet, it was clear that it formed part of a wider push by right- wing
activists to use online tools to exert inuence on Conservative party leaders.
Sites like Iain Dale’s Blog Spot and ConservativeHome had become increasingly
popular as fora for the grassroots to congregate and debate Conservative party
policy (Gibson etal., 2013).
In addition to the eorts of the two main parties to embrace the new oppor-
tunities for online activism, there were also some signs of phase III among the
minor players. e most notable of these came from the Liberal Democrats, who
Figure4.5. Membersnet (May 2010).Source:Author’s archive.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
84 WHEN THE NERDS GO MARCHING IN
84
launched their supporter hub, LibDem ACT, shortly before the 2010 election, as
shown in Figure 4.6.
According to party headquarters, the goal of the site was to build a “Liberal
Democrat supporting community...that extends beyond the formal boundary
of party membership... and is open to non- members as well as members.”13
Take- up among their supporter base appeared to be relatively healthy, with
over 300 groups formed by the end of the campaign. Its success was, however,
somewhat overshadowed by the popularity of an unocial Facebook supporter
group that called itself Rage Against the Machine. Established in the aermath of
Nick Clegg’s stunning success in the rst televised leaders’ debate, the group’s
avowed purpose was to rage against the United Kingdom’s unfair electoral
machine, which prevented smaller parties from gaining a proportional share of
parliamentary seats. e group became the focal point for absorbing the wave
of interested citizens who wanted to promote the Liberal Democrat cause. e
group aracted in excess of 160,000 members at its peak. Such numbers were
notably well beyond those who had joined groups on the LibDemACT site14 and
far exceeded the party’s ocial membership gures (Newman, 2010:28).
Among the fringe players, the far right British National Party (BNP) proved
most active in developing a MyBO- style social networking site where members
and non- members could congregate and plan activities. As was the case for
the other parties, their bespoke platform, forum.bnp.org.uk, involved a simple
Figure4.6. LibDemACT (May 2010).Source:Author’s archive.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
85
The Slow Burner 85
registration process. Once signed in, users could join groups and local cam-
paigns as well as interact on various discussion fora. Although no overall gure
was reported on the total numbers registered, group sizes indicated that levels
of membership were equal to, if not in excess of, those seen on the main parties’
sites.15
Other parties that lacked the resource or time to develop a bespoke social
network relied more on open platforms like Facebook and free soware to
coordinate and mobilize their supporters and volunteers. e Greens were
notably active in this regard, which was perhaps no surprise given they were
the rst party to use Facebook to create events and promote their campaigns.
e return on their early investment, however, did appear to be somewhat dis-
appointing for them if one looks at the nal tally of likes during the election.
e Greens received just under 8,000 likes by election day, which was around
one- tenth of the number received by the major parties and signicantly lower
than their far right rivals, the BNP (Lilleker and Jackson, 2011:133). Of the
other, more prominent minor parties, the biggest surprise was perhaps the UK
Independence Party, UKIP, which placed very lile emphasis on mobilizing its
base and community building online. e main site had relatively few calls to
action and its presence on Facebook was virtually non- existent (Lilleker and
Jackson, 2011:141).
A more systematic eort to compare these aempts at online community
development by the parties was conducted by the author during the 2010 election
(2015). Using an index designed to measure the phenomenon of citizen- initiated
campaigning (CIC), she ranked the online activist platforms developed by each of
the main parties as well as the Liberal Democrats and the BNP in the 2010 cam-
paign. CIC was dened as “a practice...in which digitally registered supporters
who are not necessarily members make use of online tools created by the party or
candidate team to campaign both online and oine on its behalf” (5). e par-
ticular activities making up CIC were allocated across four action areas:
• community building;
• mobilizing internal resources (i.e., funds and other volunteers);
• mobilizing external resources (i.e., the electorate, “geing out the vote”);
• message development and distribution.
For each function, a simple additive index was constructed that captured
whether key features enabling this particular activity were present, scored as “1,”
or absent, scored as “0.” If the features were missing on the CIC site, but available
on the home page, they were counted toward the overall totals since the goal was
to measure a party’s overall CIC eort. e ndings of the analysis and key fea-
tures for the three main parties and the BNP are reproduced in Table 4.1. Afull
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
86 WHEN THE NERDS GO MARCHING IN
86
Table4.1 UK Parties’ Citizen- Initiated Campaigning Scores, 2010 General
Election
Liberal Dems
(LibDemAct)
Conservatives
(My Cons)
Labour
(Membersnet)
BNP*
Community Building
Prole
Photo
Biography
Why joined
Set up/ join groups
Set up blog
Set up Wiki
Email/ message system
Externally promote prole
Subtotal (additive 0– 8)
√
√
√
√
√
√
√
√
8
√
√
—
—
—
—
√
—
3
√
√
√
√
√
—
√
√
7
√
√
—
√
—
—
√
—
4
Resource Generation
Personal fundraising
Promote membership
Sign up as local organizer
Sign up as candidate
Organize/ add event
Vote leaders to aend events
Subtotal (additive 0– 6)
—
√
—
√
√
—
3
√
—
—
—
—
—
1
√
√
—
—
√
—
3
—
—
—
—
—
—
0
Voter Mobilization
GOTV oine
Access phonebank
Sign up for f2f canvassing
Sign up to discuss with
network.
Leaets download
Externally promote event
GOTV online
Send email
Post to Facebook
Post to Twier
GOTV phone app
Email forward to editor
Start e- petition
Subtotal (additive 0– 11)
—
√
—
√
√
—
—
—
—
√
—
4
√
√
—
—
√
√
√
√
√
√
—
8
—
—
√
—
√
√
√
√
√
√
—
7
—
—
—
na
—
—
—
—
—
√
—
1
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
87
The Slow Burner 87
Table4.1 Continued
description of the features and indicators used to measure them are provided in
Appendix 4.1.
Perhaps the most striking nding to emerge from the table is the conrma-
tion that it was the mainstream le that led the push toward this new phase in
web campaigning. e overall standardized scores for Labour and the Liberal
Democrats sites, reported at the boom of the table, reveal that a majority of
the key features of CIC were present on the sites, while the majority were in fact
missing for the Conservatives. Beneath these top- line ndings, the community-
building and message- dissemination aspects of CIC emerge as the particular
areas of strength for Labour and the Liberal Democrats. e Conservatives,
by contrast, put most of their eort into mobilizing voters, outperforming all
Liberal Dems
(LibDemAct)
Conservatives
(My Cons)
Labour
(Membersnet)
BNP*
Message Production
Message creation
Policy email forward/
customize
Poster/ leaet create/
customize
Policy input/ feedback
Message distribution
Web banners/ ads download
Posters/ leaets download
Email/ share policy docs
Newsfeed to website
Share blog posts externally
Link to SNS prole
Link to Twier account
Import email contacts
Subtotal (additive 0– 11)
—
√
—
√
√
√
√
—
√
√
√
8
—
—
—
√
√
—
—
√
—
—
√
4
√
—
√
√
—
√
√
√
√
√
√
9
—
—
—
√
√
—
√
√
—
√
—
5
Overall Score (0– 36) 23 16 28 10
Standardized Score (0– 100) 65 41 71 26
√=feature present on campaign site ;— =feature not present; na=not accessible.
Note: e British National Party (BNP) site was internal to their home pages not as a separate/
independent platform. Standardized scores are calculated by transforming each sub- index into a 0– 100
range and then averaging the scores. See Appendix 4.1 for details of variable denitions.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
88 WHEN THE NERDS GO MARCHING IN
88
other parties on this dimension. e more instrumental orientation of the site
and its emphasis on “results” were signaled very openly by the producers of
MyConservatives.com. Upon entry, users were presented with a message inform-
ing them that the main purpose of the platform was to increase the party’s elec-
toral support, rather than promote internal dialogue and discussion.
A further signicant nding to emerge from Table4.1 is the mixed evidence it
provides in support of the normalization thesis. e BNP, as the most marginal
party, clearly had the weakest site. However, the Liberal Democrats were very
serious contenders in the CIC stakes. Furthermore, beneath the overall perfor-
mance measure, the BNP’s emphasis on community building and message co-
production was stronger than that of the Conservatives and closer to the scores
of the mainstream le.
Based on the empirical evidence available regarding impact, it seemed the
le’s investment paid o in terms of the impact of the sites. According to gures
supplied by Labour, 35,000 people had registered with Membersnet by the end
of 2010 (Newman, 2010:25), a gure that had doubled by the end of following
year.16 Estimates given by the Tories of sign- up rates to MyConservatives pointed
to a much lower gure of around 10,000 registered users by the end of the cam-
paign (Ridge- Newman, 2014:30). e low take- up was no doubt linked to a
lack of adoption at the constituency level. Reports were that a signicant minor-
ity of local parties (over 200)were not actually making use of it during the cam-
paign to sign up extra supporters. Given the relatively short period of time the
facility had been operational, however, it was, as Ridge- Newman (2014) argued,
perhaps somewhat premature to consider “cyber- toryism” to be an entire failure.
Given their greater success of their online supporter hub, it was not surprising
that Labour also claimed victory in their broader goal of indirect voter mobiliza-
tion. According to the party’s own research, it managed to triple the number of
face- to- face contacts since 2005, yielding a grand total of 450,000 door- to- door
visits. Independent post- election analyses questioned this ebullience, however,
suggesting a more modest success rate for Membersnet than Labour had claimed.
Analyses of self- reported contact in the 2005 and 2010 British Election Study
(BES), in particular, failed to show a signicant spike in Labour’s level outreach
in the laer election (Gibson, 2015; Aldrich etal., 2016).
e Conservatives were less forthcoming in estimating the precise voter
impact of MyConservatives, reporting only that 390 constituency- level cam-
paigns were active users of the site (Ridge- Newman, 2014:19). Academic work
by Fisher etal. (2011) was generally positive about the impact of their data-
driven targeting eorts, arguing that the technology— and particularly the new
electoral database MERLIN— had given them an edge in the more marginal
seats. e extent to which that success could be aributed to the data gained
through MyConservatives was also questioned, however, given the criticisms
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
89
The Slow Burner 89
that emerged later over its lack of integration into MERLIN (Ridge- Newman,
2014:34).
Despite the likely over- claiming on each side, there was a growing consen-
sus among journalists and researchers that a divide was now emerging between
two main parties in their approach to digital campaigning. e Labour Party was
essentially seen as pioneering the phase III mode by focusing on activating its
grassroots, promoting two- step ow of communication, and indirect mobiliza-
tion (Crabtree, 2010; Newman, 2010; Gibson etal., 2010; Painter, 2010). In
doing so, its digital practice was seen as aligning with the party ’s w ider strategy for
2010, which was to make this a “word of mouth” election (Newman, 2010:25).
Sue MacMillan, the party’s new media director, had even gone on record during
the election to say that her main objective had been to build online community,
as a means of generating more oine activity (McGregor, 2010:36).
By contrast, the Conservatives were regarded as having adopted a much
more utilitarian and market- oriented digital strategy, which was designed to
catch the oating voter (Newman, 2010; Painter, 2010; Gibson etal., 2010).
e party invested heavily in advertising on popular platforms such as Facebook
and Google and tools for optimizing search engine visibility. According to Nic
Newman (2010), one of the architects of the BBC online news service, the
party devoted more resources to its online marketing eort than the other par-
ties combined. e switch over to MERLIN meant that they also upgraded their
“in- house” targeting capacity. Although these initiatives did show that the party
regarded direct voter mobilization as its priority, whether this equated with a
leap into the laser- like precision and scientic approach of phase IV– style digital
campaigning was heavily doubted (Ridge- Newman, 2014; Crabtree, 2010)
In organizational terms, it did seem that the major parties at least had pro-
gressed beyond the conguration associated with phase II. e numbers of full-
time personnel that both were prepared to dedicate to their internet operations
had increased signicantly to those seen in 2005. By 2010, the Conservatives’
core team had expanded to eight, while Labour had recruited seven full- time
sta to its new media oce (Gibson etal., 2010). Beyond this, the minor parties
clearly struggled to commit similar resources, although the Liberal Democrats’
increase to three full- time digital sta was arguably a bigger growth in relative
terms.17 As well as expanding in size, the digital teams also increased in promi-
nence and status within the campaign hierarchy. e specialized new media
units were now largely recognized as independent sub- teams who provided
direct input into the top tier of campaign planning.18
At the local level, activity rose again, and this time, the increase was substan-
tial. Results from a post- election survey of constituency agents for the two main
parties and the Liberal Democrats by Fisher etal. (2011) found that websites
were now deemed to be at least as important for the campaign as leaets and
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
90 WHEN THE NERDS GO MARCHING IN
90
mainstream media coverage. Southern and Ward’s (2011) comprehensive audit
of constituency web campaigning conrmed these ndings, revealing that for
the rst time a majority of candidates from all three parties had an independent
web presence (i.e., more than simply pages on national party sites). is wider
coverage did not extend to the minor parties, however, with UKIP, the Greens,
and the BNP candidates on average more likely to not have a personal website
than to operate one. at said, Southern and Ward (2011) also measured par-
ties’ social media presence and found a stronger parity in provision, especially
with regard to the use of blogs.
e commitment to genuinely interactive uses of these new tools, however,
remained limited to a very small minority of candidates. It was a minor party,
the Liberal Democrats, who performed best in this regard, with up to one- third
of its candidates reportedly having engaged in dialogue with their followers on
Twier (Southern and Ward, 2011). e existence of a partisan divide in aen-
tion to social media during the election is supported by the ndings of Fisher
etal. (2011). In a post- election analysis of survey data from local constituency
agents, they found that respondents from major parties saw more value in web-
sites compared with social media platforms and newer contact modes such as
texting. While not conclusive, such ndings suggest that 2010 saw some evi-
dence of an equalization between the parties in terms of their use of the newest
campaign tools.
Overall, therefore, it did appear that the 2010 digital campaign saw some sin-
cere eorts by parties to harness the community building and activist mobili-
zation activities associated with phase III. is push was led largely by parties
on the center- le, who were among the rst and most active to develop social
networking platforms and explicitly recognize the power of the two- step ow
model of online to oine communication. e right appeared more focused on
the next stage of direct voter mobilization. How far each approach paid o is, as
the evidence presented earlier has revealed, somewhat unclear. In nal section
of this chapter we return to this question through a more systematic analysis of
voters’ response to the digital campaigns over several elections, including that
of 2010.
Phase IV Moves toIndividual Voter
Mobilization (2015)
Before turning to examine evidence about paerns in the demand side of UK
digital campaigns, we conclude this review of the supply side of web campaigning
with some observations on the 2015 election. If 2010 saw the implementation
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
91
The Slow Burner 91
of a phase III– style redistributive approach to digital campaigning, what signs of
phase IV can be detected ve years later?
e divergence in the digital strategy of parties in 2010 does not appear to
recur in 2015, with all the main parties demonstrating a clear focus on mobi-
lizing the electorate, and targeting oating voters through websites and social
media advertising. Digital teams increased in size, and key operatives in the
Obama campaigns— Jim Messina and David Axelrod— were hired by the
Conservatives and Labour, respectively, as strategic advisors. ere was a new
emphasis on squeezing out voters’ personal information and online contact
details to help build bigger email lists and populate networked databases.19 For
the Conservatives, eorts centered on expanding and strengthening their exist-
ing in- house resource, MERLIN, which was seen to have at best a patchy record
of success since the last election.20 Labour, by contrast, made use of the com-
mercially available campaign soware Nation Builder, which was customized
and integrated with their new and improved in- house voter management sys-
tem, Contact Creator. It was the Liberal Democrats, however, who took perhaps
the biggest step forward into big data campaigning, recruiting the services of US
campaign tech specialists, the Voter Activation Network (VAN), to help them
build up their voter lists and more eectively target their messages.21
In addition to their internal soware and hardware revamps, parties were also
now very eager to exploit social media channels. e Conservatives, in particu-
lar, maintained and intensied their use of Facebook for promotion of their mes-
sage. Media reports estimated that the party had spent up to one million pounds
on advertising on the platform in the year leading up to the election.22 In addi-
tion, the paern of spending conrmed vote maximization to be the key goal of
the digital campaign, with highly competitive seats such as Newark receiving a
much larger chunk of the budget than places such as Clacton, where the party
was unlikely to win (Fisher etal. 2015).23
Judged simply by the number of likes received, it would seem that the Tories’
focus on Facebook paid o. eir party pages and Cameron’s own site each hov-
ered around the region of one million, an exponential rate of growth compared
with 2010. By contrast, Labour and Ed Miliband struggled to reach the half
million mark.24 e story on Twier was rather dierent, however. While the
overall numbers of followers were lower than on Facebook, Labour managed
to maintain a clear lead over the Conservatives during the campaign, peaking
at around 220,000 by election day. is total was around a third greater than
the Conservatives achieved, who had around 150,000 followers by the close
of polls. Although it is of course dicult to prove any causal relationship from
these numbers, the surprise victory of the Conservatives does show it to be a
more accurate barometer of the electorate’s sympathies.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
92 WHEN THE NERDS GO MARCHING IN
92
In place of online community building, the parties now concentrated more
explicitly on presenting their home pages as a “one stop” shop for direct mobili-
zation and data extraction from visitors. Akey innovation in this regard was the
addition of landing pages that preceded entry into the main sites. ese pop- up
pages confronted visitors with a short survey designed to nd out whether they
would be voting, and if so, whom they would vote for. Additional requests then
appeared, asking for further information about where the visitors lived, what
their views were on various party policies, and for their email address. In the
case of Labour, the survey questions continued for up to 10 further pages and
included appeals to donate, volunteer, and share the site URL with friends. W hile
it was possible to skip the survey, the buon enabling the user to go straight to
the main site was typically not prominently displayed on the sites. Although ini-
tially the practice appeared on the two major parties’ sites and also on UKIP’ s
home page, by the close of the campaign it had spread, with both the Greens and
the Liberal Democrats adopting this more interrogative approach. Figures 4.7
and 4.8 show the landing pages for the two major parties during the campaign.
Elsewhere, appeals for visitors to supply parties and candidates with their
Facebook and Twier credentials proliferated. Some parties even required the
entry of an individual’s social media account details in order to access certain
sections of the site. e Conservatives, for example, made their “share the facts”
feature and also parts of their manifesto available only aer linkage of Facebook
Figure4.7. e UK Labour Party landing page (May 2015).Source:Author’s archive.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
93
The Slow Burner 93
or Twier proles. is more intrusive and interrogative nature of the election
sites in 2015 was underscored further by the new prominence given to parties’
data privacy policies, which were now regularly referred to as users browsed
through their contents.
is more aggressive colonization of social media by the major parties at the
national level was replicated and reinforced at the local level. Work by Southern
and Lee (2015) showed that while levels of online constituency activity remained
fairly comparable to 2010 among both the major and minor parties, with around
9 in 10 candidates having some type of web presence, the preferred mode of
campaigning had shied quite substantially. In particular, the reliance on social
media among major party candidates was now much higher, while their use of
individual home pages had dropped signicantly. Furthermore, levels of interac-
tion occurring on social media platforms had actually increased. e majority of
candidates on Twier were found to have used it to engage with voters at some
point during the campaign. According to Southern and Lee (2015), this com-
pared with less than one in ve during 2010.
Such changes were interesting on a number of grounds. First, the growing
dominance of the major parties on social media platforms suggests a tilt back
toward normalization at the systemic level, with the bigger players now claim-
ing terrain in the social networking landscape. Second, while it might be tempt-
ing to see the increased interactivity as a sign that candidates had become more
Figure4.8. e UK Conservative Party landing page (May 2015).Source:Author’s archive.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
94 WHEN THE NERDS GO MARCHING IN
94
commied to engage with voters more meaningfully in their social spaces, the
actual nature of the interactions taking place were not known. Candidates’
increased proclivity to respond to voters’ concerns may simply have embedded
the “controlled interactivity” mode identied by Stromer- Galley (2014) and
others. It might also be reective of the shi by parties toward the instrumental
“extractive” mode of voter engagement synonymous with phase IV.
Overall, therefore, much of the evidence gleaned from the 2015 general elec-
tion suggests that phase IV campaigning was at a latent stage. Although digital
teams had expanded signicantly in their size and centrality, they still formed a
fairly small cog in the campaign machine, and certainly had not moved into the
nerve center of operations. ere was lile evidence of their internal dieren-
tiation into specialist sub- teams, such as the data analytics and soware engi-
neering units that had emerged under Obama in the United States. Post- election
evaluations of the parties’ eorts to engage with “big data” campaigning, while
not entirely dismissive, were largely unconvinced that they had the resources or
indeed ambition to engage in the level of scientic micro- targeting that had been
seen in the United States.25 According to mainstream media reports, the parties
still saw social media tools in static rather than dynamic terms. ey were more
likely to use them as tools for documenting and reporting on the election, rather
than “weaponizing” them to target voters.26
Based on developments in the supply side of digital campaign, therefore, it
does seem that digital campaigning in the United Kingdom had moved through
the four phases set out in Chapter1. An initial burst of experimentation had
been followed by a lengthier period of standardization. Parties had converged
on a more professionalized and static approach to their web “shopfronts,” but
lacked a clear strategic end or target audience in mind. Aer 2005, there was
a shi into phase III– type activities as the parties invested in building digital
supporter networks to help revitalize their base and reconnect with members.
is trend was most pronounced among the mainstream le. Among the main-
stream right, greater priority was given to the vote- maximizing aspect of the
new medium and the combining of a more inclusive “broadcast” approach with
aempts to make micro appeals to voters on social media. is early push by
the Conservatives into phase IV– style campaigning was further accelerated in
2015 as targeting of the general public via social media advertising intensied.
However, across the board, parties showed more interest in building up their
stores of information on individual voters in order to engage in a more precise
campaign of direct mobilization.
In the section that follows, we switch to examine this process of evolution
from the demand side of the situation. To what extent did the progress on
the supply side resonate among voters in terms of their experience of digital
campaigning?
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
95
The Slow Burner 95
The View fromBelow:Mapping Digital Campaign
Cycles inthe UKElectorate
e analysis now shis to look at how the public has responded to parties’ and
candidates’ digital campaign eorts. Can the paerns of change and innovation
recorded in elites’ use of the technology be detected in the electorate’s experi-
ence of web campaigning? Have UK voters moved on from simply reading web
campaign content, to redistributing it, and increasingly now to receiving it in the
form of digital messages? Finally, and more specically, are the parties’ strate-
gies reected “on the ground”? To what extent are the paerns of innovation
and “leadership” identied among the elite actors in the earlier sections actually
detectable among their activists and voters?
READING, REDISTRIBUTING, AND RECEIVING THEWEB
CAMPAIGN OVERTIME
To examine the demand for web campaigning over time, we make use of
three main data sources covering the 2005, 2010, and 2015 elections. For
2005 and 2010, we use two “bespoke” survey data sets that were produced as
part of grant- funded projects on which the author served as a principal or co-
investigator. ese surveys were designed to measure citizens’ use of online
technologies during the campaign and thus included a comparatively rich set
of items that allow us to measure the three main modes of engagement—
read, redistribute, and receive.27 For 2015, we rely on two surveys produced
from the British Election Study. e rst was wave 5 of the internet panel
run by YouGov. is was elded during the campaign and included a wide
variety of items about individual political uses of the internet that could be
used as measures of the read and redistribute variables. e second was the
post- election mail- back component of the random probability survey, which
included the CSES module on online contact. e CSES module included
measures of both direct and indirect “receive” mode— which was not the case
for the YouGov survey. It was also the basis for the results reported in the
United Kingdom in Chapter3.
Table 4.2 presents the results of this initial mapping exercise. Specically,
it shows the extent to which the three modes of voter engagement with e-
campaigns have occurred in UK elections since 2005. For reference purposes,
we present the proportion of the electorate that had internet access in each elec-
tion. is is done in order to contextualize the results, by providing the total size
of the population that potentially could have undertaken each mode of activity
in any given year.28
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
96
Table4.2 Voter Engagement inUK Digital Campaigns, 2005– 2015
Mode of Engagement
Year Internet use Voter as Audience— READ Voter as Activist— REDISTRIBUTE Voter as Target— RECEIVE
Online news Camp. sites Sign- up/ Download Share/ Exchange Party (Direct) F&F (Indirect)
2005
N=1,937
53%
N=1,033
15.5
(28.9)
1.8
(3.4)
0.8
(1.4)
1.5
(2.8)
2.7
(5.1)
1.7
(3.2)
READ 15.5 (28.9) REDISTRIBUTE 0.6 (1.2) RECEIVE 4.2 (8.1)
2010
N=1,960
75%
N=1476
28.7
(38.1)
15.0
(20.6)
6.3
(8.4)
8.5
(11)
2.6
(3.4)
15.1
(20.1)
READ 31.5 (41.9) REDISTRIBUTE 3.0 (4.0) RECEIVE 16.1 (21.4)
2015
N=30,725
86%*51.1
(59.4)
15.3
(17.8)
2.1
(2.4)
12.8
(14.9)
11.8
(13.6)
5.5
(6.4)
READ:51.5 (59.9) REDISTRIBUTE:1.7 (2.0) RECEIVE:16.3 (18.9)
* Internet use gures are from the Oce of National Statistics (ONS) Statistical bulletin “Internet users:2015,” hps:// www.ons.gov.uk/ businessindustryandtrade/ itandin-
ternetindustry/ bulletins/ internetusers/ 2015. Estimates are based on UK adults (aged 16– 24) who used the internet over the rst quarter of the year January– March. e 2015
YouGov panel study was an online survey; thus it was not possible to calculate internet use for the population as a whole. e CSES module was conducted oine but did not
include an internet use question. All population estimates for read, redistribute, and receive were calculated using the ONS gure of 86% as the baseline. See note 28 for further
details of the calculation.
Figures are % and reported for overall sample and internet users in parentheses.
Sources:2005 and 2010 were modules in national face- to- face omnibus surveys conducted by NOP (2005) and BMRB (2010); 2015 wave 5 (campaign wave) of the British
Election Study (BES) Internet Panel Study conducted by YouGov used to calculate read and redistribute. Mail- back CSES component of pre- and post- election random prob-
ability survey used to calculate receive. Survey weights applied. See Appendix 4.2 for further details of surveys and variable denitions.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
97
The Slow Burner 97
In presenting and describing these results, there are a number of method-
ological caveats that rst need to be issued. e most obvious of these is that
the specic measures used to create the variables of interest inevitably diered
across surveys. Our approach to maximizing their comparability was to rst
specify a generic set of activities associated with each mode, and then identify
those variables in each survey that most closely captured them. us, to measure
engagement in the rst category— reading— we focused on two sets of activi-
ties. e rst was use of the internet to access election news and information
in general, and the second was the more specic act of visiting an ocial web
campaign site. e variables that most closely corresponded to these activities
were used to create the aggregated variable “read.”
For redistribution, we also combined two types of online activities; however,
we imposed the restriction that both had to be performed, rather than either one,
for an individual to have engaged in this response mode. e rst indicator was
whether a respondent had signed up to assist a party or campaign or to receive
news or information from them. e second was whether the respondent had
shared or exchanged any election information with others. Finally, the measure
of receive follows the approach of Chapter3 in that it measures (where pos-
sible) the receipt of campaign information from parties and candidates (direct)
and through more informal means, for example, one’s social networks (indirect).
Full details of the surveys measures used to capture each mode, along with the
demographic and political aributes used in the tables that follow, are provided
in Appendix 4.2.
Bearing in mind these caveats about comparability, a clear nding to emerge
from Table 4.2 is that the proportion of the electorate engaging in all three
modes of engagement has increased over time. is holds for the population
as a whole, which one might expect given that overall levels of internet usage
have increased. More signicantly, however, it is also true for the internet- using
population (shown in parentheses), which controls for that increased access.
Across the modes, it is clear that the more passive mode of reading (i.e., paying
aention to the online campaign) is the dominant mode of engagement over
all three elections. It has also been among the fastest growing, having almost
doubled between campaigns. By contrast, redistribution and receive, while they
have also grown over time, are still engaged by only a minority of the electorate.
is is particularly the case for redistribute, which is the least engaged activity
of the three. It does appear, however, that the sharing content component of this
variable has increased signicantly across the 10- year period.
Looking from election to election, there does appear to have been a signi-
cant jump in all types of engagement between 2005 and 2010, with some modes
more than trebling in frequency in this period. is step change conrms our
understanding that web campaigns were a low priority for the parties until 2010.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
98 WHEN THE NERDS GO MARCHING IN
98
At this point, we see a noticeable expansion in voters’ modes of engagement
with the web campaign, and particularly a rise in the amount of redistributive or
phase III– type activity occurring. e numbers accessing campaign sites, sign-
ing up for email news, and using party- provided tools increased quite substan-
tially, although they still do not come close to being a majority activity among
the electorate. e rise in levels of interest in the web campaign is perhaps one of
the most notable features of the 2010 election, with around one in seven voters
claiming to have visited an ocial candidate- or party- produced site, compared
with only 2percent in 2005. In addition, there was a similarly sized jump in the
numbers receiving online messages about the election from friends and family.
ese paerns of increased engagement are supportive of the notion that 2010
saw a shi toward a more community- building and citizen- initiated model of
web campaigning.
By 2015, the gures reveal an interesting if somewhat uneven paern of
growth since 2010. Read and redistribute modes are measured using wave 5
of the YouGov BES Internet panel which was conducted during the campaign.
is study has a much larger N than any of the other surveys; however, the sam-
ple is drawn from internet users only. Given this lack of a baseline to calculate
proportions for the voting population as a whole, we substituted the Oce for
National Statistics (ONS) gure of internet use among the UK adult population
during the rst quarter of 2015 as our baseline (86percent). e rst and most
obvious conclusion to draw from these results is that the proportions of people
who were reading anything about the election online rose substantially again in
2015. However, this growth is concentrated among those reading general news
sources, rather than among those looking at ocial campaign material, which
appears to have plateaued. Equally notable is the drop in levels of redistribution
of political party content.
Closer inspection reveals that much of this fall is due to the drop in sign- up lev-
els in 2015, which was around three times smaller than in 2010. While this may
reect the decreased popularity of the ocial party- produced CIC platforms in
2015, it is more likely to be a methodological artifact, and due to the more con-
servative measure of sign- up used in the BES survey. Previous measures, taken
from bespoke ESRC project data sets, had equated a number of “soer” actions,
such as signing up for ocial e- news bulletins, downloading political organiza-
tions’ material, or following them on Twier or Facebook as measures of “sign-
up.” e measure used in 2015, however, expressly asked whether an individual
had signed up/ registered to help a party or candidate with their campaign.29 is
is likely to have depressed the total amount of sign- up being reported. Certainly,
the other more comparable measure of redistribution— sharing political con-
tent online— did not appear to have undergone a similar decline. Indeed, by
2015 this type of activity had increased by around a third, from 8.5percent of
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
99
The Slow Burner 99
the population to just under 13percent. It is possible, therefore, that we would
not have seen such a sharp drop in redistribution in 2015 if a less restricted mea-
sure of sign- up had been available.
Finally, the gures for receive point to an increase over time in overall lev-
els of contact, with around one in six of the population reporting some kind of
online contact about their vote in 2015. However, closer inspection of the ratios
of direct to indirect mode reveal a more mixed paern of rise and fall. In particu-
lar, it would seem that the number of voters receiving ocial party messages has
increased over time, and that this activity saw a particularly marked expansion in
2015. Just over one in 10 voters reported receiving some kind of online contact
from a party or candidate during this election. While the overall extent of receive
remains low relative to read, therefore, there does seem some tentative evidence
from the voters’ side that the United Kingdom was moving closer toward digital
campaigning focused on direct voter mobilization.
e gures for indirect contact reveal a less linear paern, with 2010 seeing a
peak in messages from friends and family— rising to just over 15percent of vot-
ers. is then dropped to around one in 20 by 2015. is reversal of fortunes for
indirect contact is rather surprising given the growth in content sharing noted
earlier in 2015. Also, while it might be the case that the parties were shiing to
concentrate on direct contact in 2015, this growth would not necessarily lead to
a decline of the two- step indirect mode associated with phase III. If anything,
the laer should also continue to rise as the two work in tandem. As with the
redistribute variable, suspicion falls on methodological inconsistency across the
surveys, and the use of a more conservative indicator of indirect contact in the
2015 study. For 2010, the available item simply asked whether respondents had
received any campaign- related messages online from people they knew. In 2015,
however, the item was somewhat more specic, asking whether someone they
knew had tried to persuade them to vote for a particular party or candidate.
READING, REDISTRIBUTION, AND RECEIVING BYPARTY
Having mapped trends in UK voters’ response to digital campaigns over time,
we turn to look at the partisan aliation of those involved in each mode, to see
how far this conforms to the supply- side picture presented earlier. Table 4.3
compares the intensity of engagement in each mode of activity across election
by party identication and for those declaring no party aliation.
In 2005, the question referred to the party that an individual “tended to sup-
port,” while in 2010 and 2015 it measured identication more explicitly, or the
party that a respondent felt closest to. e gures are calculated for internet
users only, which allows for greater comparability of the gures over time and
across the parties. e rst column of the table reports the levels of internet use
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
100
Table4.3 Party Supporters Engagement inUK Digital Campaigns, 2005– 2015
(Internet Users Only)
Mode of Engagement
Election Year and
Party
Internet
Access
READ REDISTRIBUTE RECEIVE
Direct
(Party)
Indirect
(Friends
& Family)
2005
Labour (N=809) 49.3 31.6 1.0 4.9 3.0
Cons (N=343) 53.5 29.3 1.0 4.3 2.7
LibDem (N=285) 59.3 29.1 1.8 3.7 5.5
Other (N=107) 59.2 35.7 3.7 5.6 0
DK/ No Party
(N=240)
59.2 21.5 1.4 5.6 2.8
2010
Labour (N=308) 72.1 41.2 3.9 3.6 25.0
Cons (N=323) 74.3 39.3 3.1 3.7 13.6
Lib Dems (N=231) 79.7 48.5 7.4 4.3 27.3
Other (N=118) 77.6 23.9 6.8 4.2 18.8
DK/ No Party
(N=417)
76.5 32.5 1.9 2.2 17.0
2015
Labour (N=5,934) —
*58.2 2.6 15.9 9.1
Cons (N=5,342) — 62.4 1.1 13.4 4.8
Lib Dems (N=1,720) — 64.8 2.4 13.1 3.4
Other (N=1,891) — 67.6 4.8 15.4 6.4
DK/ No Party
(N=5,227)
— 50.1 0.3 8.3 6.2
N refers to the full sample and the internet access gures are the % of that total who reported being
online. e gures for Read, Redistribute, and Receive are % of the online party identiers who engaged
in these activities.
* Internet access by party was not available for 2015 due to the YouGov survey having been con-
ducted online (i.e., it excluded non- internet users). e CSES module was conducted oine but did not
include a basic internet use question. e proportions for internet users were calculated using the ONS
estimate of 86% of UK adult population accessing the internet in rst quarter of 2015. See Table4.2 note
a for source information.
Sources:see Table 4.2 and Appendix 4.2 for full details of surveys and variable denitions. Survey
weights applied. e 2015 Receive gures are from the CSES module, sample Ns were Labour 489,
Cons 516, Lib Dems 106, Other 223, DK/ No Party 219.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
101
The Slow Burner 101
for each group of identiers. To simplify the presentation, we use the summary
versions of the read and redistribute variables presented in Table4.2. For receive,
however, we maintain the distinction between contact as party based, which is
seen as a marker of phase IV digital campaigning, and indirect, which occurs via
friends and family, and which is seen as more of an indicator of the type of two-
step ow communication associated with phase III.
e results are interesting in that they show a fairly similar level of internet
use among the parties. Labour supporters are slightly less likely to be online,
and the Liberal Democrats and the minor parties are among the most wired.
If we turn to our aention to particular modes of activity, we see an interesting
picture of both continuity and change time over time. As one might expect, read
is the dominant mode of engagement among all party supporters across all elec-
tions. Dierences across parties are again not particularly pronounced, although
there does seem to be slightly stronger propensity among Liberal Democrat and
minor party supporters to pay aention to the campaign online, particularly in
the most recent election.
Turning to redistribution, the overall gures conrm that parties’ eorts to
build online communities and recruit non- member activists into their campaign
operations have generally reaped very small rewards. Even using the more gener-
ous measures from 2005 and 2010, it is still the case that no party had more than
around 8percent of their supporters engaging in the type of citizen- initiated
campaigning that sites like Membersnet and MyConservatives were promoting.
at said, there was an upswing in activism between 2005 and 2010 (when
more comparable measures were used). is increase oers some support to the
argument presented in the supply- side analysis that the general election of 2010
marked a shi by the parties into the third phase of digital campaigning.
If we compare the levels of redistribution across the parties, we see some
interesting and unexpected results. First, looking at the three main parties, it
seems that both Labour and the Liberal Democrats enjoy a small but increasing
advantage over the Conservatives. us, the argument made earlier of a stronger
performance by the mainstream le in mobilizing its base and pushing digital
campaigning into phase III would seem to garner some support. at said, it
would seem that it is the fringe players who are the best at mobilizing their sup-
porters to redistribute their campaign content. While their greater success on
this front may be explained in part by the fact that a higher proportion of their
supporters are online, it may also be due to their earlier experiences in using
open source and “free to use” social media platforms. Certainly, the ndings pro-
vide corroborating evidence of the eectiveness of spaces such as Facebook and
Twier as tools for activist coordination. Based on these ndings, they appear to
have rivaled, if not actually outperformed, the more expensive party- produced
platforms in 2010.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
102 WHEN THE NERDS GO MARCHING IN
102
e last two columns of Table 4.3 examine the changes in receive mode
among party supporters over time. From these results, we can see that rates
of contact are typically higher than among the public as a whole, as we might
expect.30 However, they are still conned to a minority of individuals. Overall,
most party supporters in the United Kingdom do not receive digital messages
directly from campaigns during elections. Among those that do, it seems that
there is a relatively balanced distribution across all parties, with the minor play-
ers again holding a slight edge over their more prominent rivals. e distribution
of indirect contact over time is less evenly balanced. rough to 2010, Liberal
Democrat and Labour supporters were most likely to receive some type of
online election stimuli from people whom they knew. By 2015, however, Labour
moves to the fore, with almost one in 10 of their supporters reporting an experi-
ence of indirect mobilization during the campaign. By contrast, Conservative
supporters are among the least likely to have received this type of contact in
any election. Coupled with their weaker performance in the direct mobilization
stakes, these ndings cast some doubt on claims for a Tory dominance in its lev-
els of digital voter outreach by 2015. ey do, however, provide support for the
view that mainstream le- wing parties were the most active digital campaigners
and were particularly important in promoting the phase III era of two- step com-
munication and mobilization of the base.
In a nal step, we drill down into the gures for receive mode in order to
more fully assess the success of parties in mobilizing voters and see if this has
changed over time. In particular, we prole the demographics and political char-
acteristics of those who have received both forms of online contact over time.
Do we see a widening of the population receiving these messages, beyond the
highly engaged partisan? Are the targets for parties’ online messages beginning
to resemble those pursued by oine tactics, indicating the mainstreaming of
online contact by campaigns? Furthermore, to what extent is the receipt of these
messages associated with a greater propensity to actually turn out?
Table 4.4 presents distribution of sex, age, education, and vote- choice char-
acteristics for those individuals who reported receiving various types of politi-
cal contact during the campaign. e percentages should be read column wise
within each category. us, in 2005, 56percent of those receiving direct contact
were male, and in 2015 this had fallen to 53percent. Similarly, the proportion of
18– 34- year- olds receiving direct contact had dropped by 2015 from 46percent
to 40percent, while those over 55 had increased from 14 to just under a third of
those reporting such contact. To help highlight the particularities of the segment
of the population receiving online contact, we compare their demographics to
the sample as a whole, and in 2010 and 2015 (given additional data availability),
to those receiving oine contact. Oine contact includes all modes that are not
internet based (i.e., face- to- face, telephone, and mail).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
103
Continued
Table4.4 Socio- Demographic and Political Correlates of“Receive” Mode inUK General Elections, 2005– 2015
2005 2010 2015
Online
Direct
Online
Indirect
All Online
Direct
Online
Indirect
Oine
Direct
All Online
Direct
Online
Indirect
Oine
Direct
All
Sex
Male 56 59 48 67 57 46 49 53 53 48 51
Female 44 41 52 31 43 54 52 47 47 52 49
Age
18– 34 46 39 29 36 46 26 29 40 63 27 29
35– 54 40 46 36 44 38 37 36 31 29 35 34
55+ 14 15 35 20 17 38 36 29 8 39 37
Education
16 yrs or less 26 19 57 31 23 46 46 24 12 35 39
17– 18years 21 13 19 16 17 20 19 24 38 20 18
19years+ 42 56 20 47 43 28 29 53 50 45 44
Student 11 13 4 6 17 7 7 — — — —
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
104
Table4.4 Continued
2005 2010 2015
Online
Direct
Online
Indirect
All Online
Direct
Online
Indirect
Oine
Direct
All Online
Direct
Online
Indirect
Oine
Direct
All
Vote
Cons —
*— — 36 22 32 31 40 21 41 33
Labour — — — 31 39 29 32 33 52 31 32
LibDem — — — 28 29 27 25 7 8 10 7
Other — — — 6 8 10 10 19 20 18 14
DK/ Not sure — — — 0 2 1 2 1 0 0 14
Figures should be read as column % within each demographic / political variable. ey show the proportion of individuals w ithin each of the groups that reported online and oine
contact by parties. e “All” column reports the frequencies of the demographics within the sample as a whole. Cells may exceed 100% due to rounding.
* Vote choice variable missing in the 2005 survey.
Sources:See Table4.2 and Appendix 4.2 for full details of surveys and variable denitions. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
105
The Slow Burner 105
e ndings are interesting in that they indicate a broadening of the target audi-
ence for direct online contact between 2005 and 2015. e evidence shows that
recipients were slightly less likely to be male and were somewhat older by 2015.
e educational bias appears to intensify, however, with those from a higher edu-
cational background receiving a greater proportion of online contact from parties
over time. In terms of changes in the demographic audience for indirect contact, the
gures suggest a similar reduction in the gender bias. e gures for age and educa-
tion, however, reveal that early divides have increased markedly over time. By 2015,
almost two- thirds of those receiving some type of online indirect contact during
elections were in the youngest age group, compared with just under half in 2005.
Educationally the prole is much more biased toward those with tertiary education
compared with 2005, even more so than for direct contact. Acomparison of the
demographic prole of those receiving either type of online contact with the recip-
ients of oine contact is very instructive, in that the laer typically corresponds
much more closely to the sample distribution on all counts. Such ndings suggest
that online contact has become the channel for dierent types of voter mobiliza-
tion, one that reaches into a new, younger, and more interconnected elite.
When we look at the relationship between receiving online contact and voting,
the story seems to divide into two narratives. e paern of vote choice among
those who received direct contact in both 2010 and 2015 looks relatively similar
to the population as a whole, and to those who received oine contact, although it
is notable that the Conservatives were more popular among those who were con-
tacted online by campaigns in both elections. Such ndings suggest that, at best,
this type of stimulus had a reinforcing eect on vote choice, and perhaps gave a
slight advantage to the Tories. e voting preferences of those who received indi-
rect online contact, however, are quite dierent, with the Labour Party receiving a
much larger and disproportionate share of the votes among those contacted in this
way in both 2010 and 2015. Just over half (52percent) of those who received this
type of contact voted Labour in 2015, while less than quarter of those receiving
indirect messaging voted Conservative. e disparity is quite striking and becomes
even more so when one compares it with the voting paerns of the population as
a whole, as well as those received other types of contact. While of course it is di-
cult to conclude any causal eect of indirect contact on vote choice based on these
ndings, they are certainly suggestive that this type of two- step contact beneted
Labour much more than their right- wing rivals.
Conclusions
e arc of digital campaign development in the United Kingdom appears to t
quite well to the model set out Chapter1. An initial period of experimentation
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
106 WHEN THE NERDS GO MARCHING IN
106
prior to the 1997 election was then followed by an elongated period of standard-
ization and professionalization in which parties lacked an obvious strategy for
their web campaigns, other than talking “down” to the voters through websites.
e voters responded in turn, by engaging largely in passive “reading” of the
campaigns through to 2010.
Signicant declines in membership levels and the growing popularity of
social media prompted parties to explore more seriously the interactive poten-
tial of the internet and to build up their stocks of online supporters. By 2010, a
new understanding had emerged of the internet as a medium for two- step ow
and indirect communication with voters. Messages owed out from the party
through its activists and then out to the wider electorate. is new strategy of
online mobilization appeared to be most prominent among the mainstream
le, and several of the minor parties. From the data available, it does appear that
the approach paid dividends, with Labour clearly generating substantially more
electoral support from those contacted in this networked manner.
In 2015, the commitment to this more informal method of online mobiliza-
tion continued, although the parties began to pay more aention to developing
methods for directly targeting a wider swath of voters with digital ads. ese
eorts, however, appeared to fall somewhat short of expectations in that they
reached only a small minority of the public. Taking such limitations into account,
it does appear that the Conservatives were the most obvious beneciary of these
methods at the ballot box. Should the current slow but steady pace of digital
innovation continue in the United Kingdom, therefore, we would expect to see
the parties’ eorts at online voter mobilization to show a signicant advance,
in both scope and precision, in coming elections. Whether the advantage that
accrues to Labour and the smaller parties from indirect forms of online contact,
and to the Conservatives from direct contact, is maintained is an area for future
investigation.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
107
When theNerds Go Marching In. Rachel K. Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
5
The EarlyBloomer
Digital Campaigning in Australia
is chapter presents the second of our four case studies and focuses on
developments in digital campaigning in Australia. Following the structure of
the previous chapter, we start by comparing Australia’s position in Table 3.2
(in Chapter3) to where it should be, based on what we now know about the
institutional and technological environment that is most conducive to digital
campaigning. If the reality falls below or above our expectations, why is this the
case? What additional features of the Australian campaign landscape need to
be considered when explaining the parties’ prowess? e second section of the
chapter embarks on our core task of mapping the history of web campaigning
among Australian parties, and investigating how far that evolution follows the
four- phase model set out in Chapter1. How experimental and equalized was use
of the internet in the early days? Has there been a narrowing by Australian par-
ties to focus their web campaigns on activists and then on swing voters in more
recent election cycles? As part of this review, we identify the key actors in this
process. Which parties have been most instrumental in driving these changes
along? Finally, we turn the lens on voters’ involvement in the e- campaign. Has
the electorate changed in how they consume and respond to parties’ eorts over
time, and do these paerns align with changes we have observed at the supply
side? Which parties’ supporters are most active and mobilized in redistributing
digital content? Who is most successful in reaching their voters online, and to
what extent are parties now nding a new audience for their message?
Australia asa Context forDigitalCampaigning
Australia is a particularly interesting and helpful case to examine at this juncture
of the book. Like the United Kingdom, Australia is a parliamentary democracy
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
108 WHEN THE NERDS GO MARCHING IN
108
and does not conform to the model of competitive presidential elections that
the analysis of Chapter3 identied as prompting more active digital campaign-
ing. Australia also shares a long history of preferential voting in elections to
its lower House, another system- level trait associated with a lower intensity of
web campaigning. Unlike the United Kingdom, however, Australia operates a
more proportional system for elections to its second chamber, the Senate, where
smaller parties are regularly represented. Finally, internet adoption among
the Australian public is one of the most rapid and widespread in global terms.
Figure 5.1 shows how take- up rates in Australia got o to a very quick start and
expanded rapidly.
By the late 1990s, around one- third of the citizenry were online, and by 2001
this had risen to over half of the population. By contrast, the United Kingdom
only reached this point by the middle of the decade (see Chapter4, Figure4.1),
placing Australia at least one electoral cycle ahead in terms of the reach of the
technology among voters. e gures for broadband access conrm this paern
of accelerated growth in internet usage in Australia, particularly with regard to
mobile access. Figures from the OECD, reported in Figure 5.2, show that while
the number of xed subscriptions remained relatively static aer 2010, the num-
ber of dedicated mobile subscriptions grew signicantly. By 2013, the number of
accounts exceeded one per head of the population. is was a rate not matched
in the United Kingdom, even by 2015.
3
31
53
61
70 76
83
0
10
20
30
40
50
60
70
80
90
100
1996 1998 2001 2004 2007 2010 2013
Figure5.1. Growth of internet use in Australian federal elections, 1996– 2013 (%of
population using the internet).Sources:1996– 2001; 2007– 2013:World Bank, “Internet Users
(per 100 people),” hp:// data.worldbank.org/ indicator/ IT.NET.USER.P2?page=1 Internet; 2004:G.
Byrne, L.Staehr, S.Spencer, and A.Jenkins, “Current Internet Use in Australia:A Closer Look at the
Digital Divide,” Proceedings 17th Australasian Conference Information Systems, 2006, hp://
unpan1.un.org/ intradoc/ groups/ public/ documents/ apcity/ unpan046583.pdf
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
109
The Early Bloomer 109
Given this more favorable institutional and technological context, we would
expect Australia to be ahead of the United Kingdom in its rates of online voter
contacting but to fall somewhat short of the levels seen among presidential
democracies such as the United States and Taiwan. Aquick review of Table3.2
(in Chapter3) shows that this expectation is partially supported. Fourteen per-
cent of the Australian electorate reported receiving some kind of online mobi-
lization in 2013, a level that places it squarely in the middle of tier two nations.
It is, however, situated slightly below the United Kingdom among the tier two
countries. One of the main reasons for this lower ranking is no doubt due to the
fact that the CSES data are based on reported contact during the 2015 UK gen-
eral election, which was two years aer the Australian data were collected. As we
shall see from the following discussion, however, timing of the data collection is
only one factor to consider when explaining the lag in Australian parties’ online
campaigning.
DATABASES, E- DEMOCRACY, AND THETYRANNY
OFDISTANCE
A quick glance at the political communication landscape of Australia reveals
that, for the most part, it oers an appealing environment for the growth of
online campaigning. Australian parties are typically among the most advanced
in the use of electronic and computerized voter- management tools. is
has largely taken the form of investment in voter databases and direct mail
8
23 24 27
47
113
0
20
40
60
80
100
120
2004 2007 2010 2013
Broadband (fixed line) Broadband (mobile)
Figure5.2. Growth in broadband subscriptions (per 100 inhabitants) in Australian
federal elections, 2004– 2013.Sources:OECD historical xed and mobile broadband penetration
subscriptions (per 100 inhabitants). Figures are from 4th Quarter (rounded to nearest %). hp://
www.oecd.org/ internet/ broadband/ 41551452.xls and hps:// www.oecd.org/ sti/ broadband/ 1.5-
BBPenetrationHistorical- Data- 2016- 12.xls (Figures for mobile available only from 2009 Q4 onward.)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
110 WHEN THE NERDS GO MARCHING IN
110
campaigns (Young, 2010). Although Australia operates a very restrictive data
privacy regime,1 political parties and government agencies are exempt from its
provisions.2 is exemption, combined with their privileged access to a well-
maintained electoral roll (a product of compulsory voting), has meant the two
main parties in particular have been able to build and maintain very accurate
voter records.3 Since the 1980s and the advent of mass computing, they have
increasingly exploited their advantage in this regard by supplementing the
basic electoral roll information they hold with electronic data from local rep-
resentatives’ oces, and commercial and marketing lists (Howard and Kreiss
2009).4 e end result has been the development of two highly sophisticated
databases— Electrac and Feedback!— that have been used extensively by Labor
and the Liberals, respectively, to conduct large- scale direct mail campaigns (van
Onselen and Errington, 2004a; van Onselen and Errington, 2004b; van Onselen
and van Onselen, 2008; Mills, 2014).
Australian parties’ interest in online campaigning has been spurred on fur-
ther by the federal government’s commitment to making information and com-
munications technology (ICT) an integral part of the political process. One
of the earliest markers of this interest came with the opening of the new “state
of the art” Parliament House in 1988. e building was one of the most mod-
ern and technologically advanced of its kind in the world. e upper chamber,
the Australian Senate, was very eager to show its e- democracy credentials and
became one of rst legislative bodies to grant e- petitions the same status as those
signed by hand. It also pioneered the recognition of electronic submissions for
commiee hearings. By the late 1990s, these initiatives had started to take eect,
and the new petitioning method proved increasingly popular among citizen
groups lobbying for policy change (Magarey, 1999). State governments soon
followed suit, with both the Victoria and Queensland parliaments commission-
ing high- prole ocial inquiries into how digital technology could be used to
improve the governing process.
is growing interest and expertise in wiring up the political process led
external experts on the subject, such as Steven Cli, to declare Australia one of
leading nations in the global e- democracy movement (Cli, 2002). While some
analysts have since questioned the extent of genuine citizen empowerment that
these initiatives delivered (Bishop and Anderson, 2004; Flew, 2005; Chen etal.,
2007), Australia’s status as a leader in e- government and online service provision
is largely uncontested (West, 2008). In 2014, the United Nations’ global ranking
of countries on its e- government development index (EGDI), for example, put
Australia in second place, with only South Korea scoring higher.5
A nal spur to parties’ enthusiasm for digital campaigning in Australia is
the geographic circumstances in which it occurs. In principle, the sheer size
and sparsely populated nature of many constituencies means that standard
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
111
The Early Bloomer 111
canvassing methods are very costly and impractical to implement beyond the
inner and outer metropolitan areas. New methods that help campaigners over-
come this “tyranny of distance” problem would thus be likely to hold consider-
able appeal. While these advantages took some time to lter through, given the
digital divide that initially emerged between rural and urban Australians, they
did become increasingly obvious. Successive government made extensive eorts
to roll- out out internet and broadband access to the ‘bush’, which meant that
by 2010 the gap in usage had all but disappeared.6 Given that a signicant pro-
portion (44percent) of the constituencies classied by the Australian Electoral
Commission (AEC) as rural or provincial were also classied as marginal in
2013,7 the incentives for the parties to commit signicant resources to online
voter contacting in these more remote areas clearly only strengthened over time.
Phase I:Experimentation (1994– 1997)
Given the benets and encouragement that Australian parties faced in adopt-
ing the new medium, we would expect them to make relatively rapid progress
through the developmental cycle. is certainly appeared to be the case in
the early years, although activity came largely from one side of the political
spectrum— the le- wing Australian Labor Party (ALP). According to a state-
ment published on its website in 2000, the ALP was one of the rst parties
to set up an online presence worldwide. Having unveiled a prototype at its
annual conference in September 1994, the ocial site was formally launched
in July 1995.8 In a clear sign of the maturity and understanding the ALP
brought to the principles of net communication, the party gave it the intui-
tive and memorable site address of www.alp.org.au, which it has retained to
this day.
Elsewhere, uptake was more sluggish. e Liberals, the main right- wing
party, were notably slower o the block, launching their site almost one year
later, in 1995. In contrast to the ALP, and as a way of conrming their discomfort
with the new medium, they generated the highly unmemorable and non- user-
friendly URL for the site:www.adfa.oz.au~adm/ liberal.html. By 1996, most of
the minor parties had followed suit, including the le- wing Democrats and the
Liberal Party’s smaller, rurally based coalition partner, the Nationals. Among
the Greens, activity followed the more federalized structure of the party, with
three of the more prominent state Green parties (West Australia, Tasmania, and
Queensland) all launching sites in advance of the national party.9
While levels of investment varied, none of the parties was prepared to com-
mit extensive resources to start- up web eorts. Most relied on donated services
provided by members and party loyalists. e Liberals, for example, assigned
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
112 WHEN THE NERDS GO MARCHING IN
112
the task of seing up and running a site to an individual volunteer.10 e ALP
also relied on voluntary labor, but took a more coordinated approach and estab-
lished an interstate working group of members who collaborated for around a
year to develop the site.11 is early lead by Labor in online campaign prepara-
tion was reected in the media coverage of the 1996 election. e Australian, a
major national newspaper, reported that the ALP had, in the eyes of one expert
observer, produced “easily the best presented, organised and up- to- date site of all
the parties.”12 Content- wise, visitors could download an audio clip of the party
president and a “Howard Unplugged” buon presented viewers with a series of
policy contradictions aributed to the prime minister. Despite its interactive
gimmicks, however, the site’s primary focus was on downward communication
and one- way content dissemination. According to the Labour sources quoted
in e Australian piece, the main objective was providing the community with
“...another medium for obtaining information when, and as, they want it.”13
e level of sophistication evident in the ALP site can be seen in Figure 5.3.
e home page has an uncluered façade and a prominent and easy- to- use
menu. ere is a consistent use of font and the party logo throughout the site.
Its clean design and informative quality arguably give it the stature of a phase II
standardized and professionalized campaign site.
Figure5.3. e Australian Labor Party home page (October 1996).Source:Wayback
Machine:hp:// web.archive.org/ web/ 19961027214312/ hp:// www.alp.org.au
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
113
The Early Bloomer 113
e ALP site formed the exception to the rule, however. e Liberals’ site,
shown in Figure 5.4, was much more in the experimental vein, combining a
sparsity of content with a lack of navigational tools and heavy emphasis on text
rather than graphics.
e smaller parties’ oerings were similarly crude in design, although they
did not necessarily see this as a barrier to realizing any equalization rewards.
Post- election interviews with the “web masters” for the Greens, the Democrats,
and the Nationals showed a clear commitment and understanding of the lev-
eling power of the internet for them, and the new opportunities it oered for
extending their reach into their core electorates of young and also rural voters.14
Phase II:Standardization and
Professionalization (1998– 2007)
e lead- up to the federal election of 1998 saw added momentum take hold as
the ALP, in particular, demonstrated that it was taking the medium very seri-
ously. A dedicated “email response team” was established to ensure that any
queries from the public were quickly, and personally, responded to. e party
also experimented with live webcasting for their campaign launch in Brisbane
and dedicated a section of their site to video content. is early recognition of
the importance of visual content online and the rise of YouTube by Labor was
matched by their anticipation of the blogging craze. Party leader Kim Beazley
made daily diary entries on the site. Perhaps the clearest signal of the fast-
moving nature of developments, however, was the production of a number of
Figure5.4. e Australian Liberal Party home page (November 1996).Source:Wayback
Machine:hp:// web.archive.org/ web/ 19961109121927/ hp:// www.liberal.org.au/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
114 WHEN THE NERDS GO MARCHING IN
114
phase III– style mobilizing tools in the form of downloadable shareable banners
and e- postcards to allow volunteers to spread the campaign message.
ese eorts appeared to pay o in terms of aracting public aention.
According to the party’s own records, the average number of hits on their
national site jumped from around 11,000 in 1996 to 2 million by 1998,15 an
increase that “far outstripped the increase in internet connectivity.” Even if each
hit did not directly correspond to an individual voter, this rapid escalation of
interest in a country with 11million voters was impressive. For some in the
party, the new exposure and reach generated by the technology constituted a
game- changer, and meant that no major political party will enter an election in
the future “...without focusing on the Internet as an election tool.”16
e great leap forward, however, did not happen. Indeed, viewed in retro-
spect, 1998 appeared to be an early high watermark for digital campaigning in
Australia. Instead of moving swily onto activist mobilization and community
building, the parties, including the ALP, entered a prolonged period of stan-
dardization aer the election. e focus was on playing it safe, “broadcasting”
information, and avoiding any US- style overt eorts to recruit and mobilize sup-
porters online (Gibson and Ward, 2002). Given the highly fertile environment
for digital campaigning that Australia seemingly presented, this “freezing eect”
was particularly puzzling. Amore detailed examination of the Australian case
and a detour into developments in digital campaigning at the state level provide
some evidence to help explain the parties’ newfound restraint.
A HALT INTHEPROCEEDINGS
e surprise defeat of the right- wing governor of Victoria, Je Kenne, in 1999
ranks as probably the most humiliating and public implosion of a web campaign,
prior to the fall of Howard Dean in early 2004. Kenne declared his anity with
the internet early in his campaign, seing up a highly personalized website— Je.
com. Unfortunately, the site rapidly became a source of embarrassment for him,
as it was repeatedly lampooned by journalists, who touted it as a symbol of his
Melbourne- centric elitism, and remoteness from ordinary voters (Chen, 2013).
To add insult to injury, the spoof site Jeed.com, set up by a former sta member,
soon overtook the ocial site in popularity and remained a thorn in the gover-
nor’s side during the course of the campaign.17
Kenne’s highly visible virtual drubbing clearly dampened enthusiasm for
online campaigning in the lead- up to the 2001 federal election. Despite a few
headline- grabbing online stunts, such as the ALP’s Political Big Brother site which
allowed users to vote out unpopular Liberal front- bench politicians, the parties’
eorts were seen as disappointing, constituting lile more than a series of “elec-
tronic brochures aimed at wide audiences” (Chen, 2013:26). e results of a
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
115
The Early Bloomer 115
pre- election survey of party web managers conrmed this “safety rst” approach,
with information distribution being widely endorsed as the main benet of hav-
ing an online presence (Gibson and Ward, 2002). is lack of ambition at the
national level was replicated and magnied among state and local actors. Astudy
of the online presence of the two major parties and the Greens at the state level
in the months preceding the 2001 federal election revealed take- up to be very
patchy (Gibson and Ward, 2002). All three parties were missing one or more of
their branches online. ose that did maintain a presence followed the broad-
cast model of their federal counterparts and oered very few interactive features.
Candidates’ presence was also sporadic. Reports from the 2001 Australian can-
didate study (ACS) revealed that just under two- hs (37percent) of respon-
dents maintained a personal website during the election. Closer inspection
revealed that the overall gure masked a paern of normalization with just under
half of major party candidates online (49percent) compared to less than a third
of their smaller rivals (30percent) (Gibson and McAllister, 2006).
e election of 2004 saw lile advance on 2001 according to most observ-
ers, particularly among the major parties.18 According to Chen, the sites were
“largely passive in character with lile in the way of interactive components...”
(Chen, 2004:3). Among the minor parties, the narrative was more positive, with
the Greens and the Democrats both seen to have run “signicant” online cam-
paigns.19 is included a very “sticky” anti- government site, Democracy4Sale,
launched by the Greens, and the clever use of viral email to mobilize their activ-
ists.20 e ACS results for 2004 appeared to conrm this push by the minor
parties to exploit the medium with gures showing growth in the number of
minor party candidates online and a closing of the gap with the major parties.
Closer inspection of the evidence, however, revealed this growing parity was
actually the result of a slight drop in major party online presence, rather than a
minor party acceleration (Gibson and McAllister, 2006). As one close observer
of Australian e- politics put it, the truth of the maer was that the major par-
ties were still largely indierent to the value of the web and had yet to “see the
online environment as the site of signicant and meaningful electoral competi-
tion (Chen, 2005:120).
At rst sight, the 2007 election appeared to signal an end to the caution
and stasis that had held sway for the past decade. Almost three- quarters of the
population now reported access to the internet, and perhaps more importantly,
broadband use had quadrupled. is was also the rst election in which the new
social media platforms of Twier, Facebook, and YouTube featured. e laer
proved particularly popular among voters, leading some commentators to dub it
the “YouTube election” (Crook, 2007).21 Labor again led the way, expanding its
new media team into double digits and ensuring an active presence and healthy
following on popular social media spaces.22 Perhaps the most telling sign of the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
116 WHEN THE NERDS GO MARCHING IN
116
party’s new boldness for digital campaigning came with the launch of its new
leader’s website, which used the catchy self- titled URL Kevin07.com.23 With the
painful memories of Kenne now clearly banished, this highly personalized site
built for Kevin Rudd was, according to most observers, the “standout” site of the
election and its “real heavy hier,” aracting national and international media
aention. Reports from the election were that it had racked up almost half a mil-
lion unique visitors during the course of the campaign (Crook, 2007; Australian
Centre for Public Communication [ACPC], 2008; Miskin, 2008).24
As well as helping to introduce Rudd to the Australian public, the site was
designed to create a buzz, particularly among younger voters. According to one
insider, the site marked the parties’ shi away from the static model of web cam-
paigning pursued in earlier years to a more dynamic one focused on “building
networks of communities.”25 For Chen (2013), Labor’s eort marked a “...shi
in the role of digital media from a peripheral element of the overall campaign,
towards a more central role in the planning and execution of campaign strategy”
(28). It also demonstrated that Labor was now ready to return to the “riskier,”
but potentially more rewarding territory of “presidential” style web campaigning
that had been pioneered by Je Kenne in 1999.
Further signs of a pivot into a new phase appeared with the formation
of Australia’s rst virtual party, Senator Online. Organized and run entirely
through the internet, the party showed how digital technology could be used
as an infrastructure and tool for mobilizing activism. In addition, some of
the more established smaller parties were proving themselves as highly adept
at using social media as an organizing tool (Chen, 2008b; Chen and Walsh,
2010; Gibson and McAllister, 2011). Results from the 2007 ACS revealed that
although the major parties once again clearly held the advantage in terms of
running personal sites, Green candidates were much more adept at exploiting
web 2.0 technologies. Additional analysis by Gibson and McAllister (2011)
extended these ndings to show how adoption of such tactics also held divi-
dends at the ballot box. In particular, the authors found that Green candidates
who had campaigned using web 2.0 tools enjoyed a signicantly higher vote
share than major party candidates that had done so. While the causal mecha-
nism behind the Greens’ success remained unclear, the authors speculated it
was possibly due to their stronger online activist networks and higher rates of
indirect mobilization.
Challenging these seemingly transformative trends were several counter-
weights, however— the rst and most obvious of these being the performance
of the mainstream right. As in previous campaigns, the Liberals got o to a
slow and somewhat embarrassing start. e party faced widespread ridicule
as Prime Minister John Howard made his ill- fated debut on YouTube, greeting
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
117
The Early Bloomer 117
viewers to the 24- hour channel with a mistimed “Good evening.” Facebook
and Twier proved similarly challenging for the Howard and received scant
aention from the campaign in terms of content and updates (Crook, 2007).
In addition, with the exception of the Greens’ dynamic use of Facebook, the
national e- campaigns of the smaller parties were regarded as having failed to
build on the momentum gained in 2004 (ACPC, 2008). e widespread use
of YouTube by candidates and parties did lile to convince observers that a
genuinely more dynamic approach to e- campaigning was underway. Indeed,
the runaway popularity of the video channel among candidates, compared
with other web 2.0 tools during the campaign, was aributed to its similarity
with television and radio broadcasting, and capacity to aract mainstream
media coverage.26 According to Chen (2008a), most party and candidate
YouTube channels were not open to comment and mainly featured repur-
posed television advertisements. e nal nail in the con of the claim that
2007 was a step- change election, however, came with a post- election anal-
ysis of Labor’s eort, and particularly the Kevin07 site. Areport issued by
Australian Center for Public Communication (ACPC) questioned the extent
of genuine change that it had prompted in digital campaigning. Underneath
the shiny web 2.0 exterior, the authors argued, things were not very dierent
from 2004. Kevin07, they argued, was simply a “traditional controlled com-
munication campaign, albeit on a new platform” (ACPC, 2008:30)
While the election of 2007 restored some momentum to Australian web
campaigning, it did not prove to be the game changer that the pundits had
anticipated. Parties, with perhaps the exception of the Greens, still seemed
to remain locked into a standardized model of repackaged mainstream media
content, dressed up with eye- catching but ultimately supercial mechanisms
of engagement.
Phase III:Community Building and Activist
Mobilization (2010)
e digital campaign of 2010 was a much more muted aair than 2007. Kevin
Rudd, the subject of much of the previous web mania, had become a liability for
the ALP and was dispatched by an internal coup shortly before the campaign
began (Gibson and McAllister, 2011). e switch into Kevin10 thus never mate-
rialized. is lack of outward glitz, however, appears rather ironically to have
been a shield for some deeper and more signicant shis occurring within the
parties, in particular their new focus on digital technology as a community-
building and activist- mobilization resource.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
118 WHEN THE NERDS GO MARCHING IN
118
e most visible marker of this new approach came with the launch of Obama-
style MyBO hub sites by both main parties. According to its own press release,
Labor was rst out of the blocks, launching Labor Connect about a month before
polling day. In the statement, the party heralded the site as the “rst social media
platform purpose built for an Australian political party.” Perhaps even more tell-
ing, given the claims made by those behind the Kevin07 campaign, was that it
was presented as an exciting “rst step” for the ALP in using the new technology
to “strengthen ties to the community.” e site instantiated the citizen- initiated
campaign (CIC) logic that had emerged elsewhere by providing a platform for
the party to “form direct relationships” with members and supporters (i.e., non-
members) that would allow the laer, in particular, to increase their “collabo-
ration” in party aairs.27 Following the model of its US and UK counterparts,
Labor Connect oered similar opportunities for virtual community building,
resource generation, and message dissemination. Users were allowed to set up
and share their personal proles, to join or start groups and organize events, and
to share GOTV messages with their wider networks. Like UK Labour, the ALP
also split their CIC content across multiple platforms. Labor Connect served as
the main site and focused heavily on mobilizing activists. Eorts at harnessing
supporter input into policy formation were channeled through a separate site—
inkTank— which was seen as an “ideas incubator” and tool for crowdsourcing
opinion on future policy direction for the party.
e Liberal Party’s version, MyLiberal.com, also emerged shortly before the
election. Its launch was a much more low- key event than had been the case for
Labor Connect. Despite its more muted entry into the election arena, however,
its establishment was a sign that the Liberals were looking more seriously at
how digital technologies could help them organizationally and electorally. eir
decisive defeat in 2007 and widely derided web performance had prompted
a major internal review of their organizational strength and operational prac-
tices.28 e reform program that emerged focused on boosting membership lev-
els, and increasing their appeal among younger voters, given their aging support
base. Akey step to achieving this, according to the report, was through “a major
internet blitz.” e rst phase of this began in April 2008 with the launch of
an online platform that was designed to promote greater dialogue between the
party elders and younger members. e My.Liberal site formed the next phase of
the new digital program. As the name indicated, the site replicated the model of
CIC sites seen elsewhere. Registration was open to members and non- members,
and once logged in, users were able to set up a personal prole and were given a
personal dashboard that recorded their activity. Unlike Labor Connect, however,
the main focus of MyLiberal was on connecting supporters to local candidates
and campaigns, rather than oering a space for networking, discussion, and
community building.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
119
The Early Bloomer 119
LEVELS OFCITIZEN- INITIATEDCAMPAIGNING
To probe these impressions more systematically, the sites were subject to a
detailed content analysis using the CIC index that had been applied to the UK
parties’ hub sites in Chapter4. Sites were scored on the extent to which they
enabled supporters to engage in, or co- produce, four core campaign activities—
community building, resource generation, voter mobilization, and message pro-
duction. e results are reported in Table 5.1. ey make for interesting reading
in that they conrm the impression that Labor was more focused on promoting
community building on its site. Somewhat more surprisingly, however, Labor
also had a slight edge over the Liberals in promoting supporter involvement
in their GOTV eorts, although neither party was highly active in this regard.
Opportunities to help in mobilizing resources for the campaign were largely
absent from both sites, as were chances for crowdsourcing party messages. Use
of the sites for two- step communication was more prevalent, with multiple
prompts given to supporters to forward campaign emails and social media posts
to their networks.
Overall, the results of Table 5.1 conrm the ALP’s stronger commitment
to phase III digital campaign goals. Viewed in the light of the results from
Chapter4, however, it seems that neither party had a strong interest in using
its digital campaigns to promote activist mobilization and community building.
In the United Kingdom, at least two of the four campaign sites contained up to
two- thirds of the content of the CIC index. In Australia, neither party managed
to include more than half of the items on their sites. e presence of compulsory
voting may help explain the more spartan nature of the sites, particularly the
lower provision of incentives to help with GOTV activities. e complete failure
by both parties to use the sites to recruit local organizers to run events or help
to promote membership, however, is more surprising. It may be that the federal
nature of party organization in Australia plays a role here. Unlike in the United
Kingdom, party membership is managed at the state level (McAllister, 1991).
Any move toward a system of national online aliation might be seen as threat-
ening the authority and autonomy of the state parties.
While the smaller parties did not create ocial MyBO- style hub sites, they
did devote signicant time and resources to phase III– type initiatives in 2010.
e Greens again stood out in terms of their intensive use of social media and
email at the local level to mobilize activists and spread the party message. e
Victorian Greens were among the most active of the state parties in this regard,
focusing their eorts on Melbourne, where the local candidate Adam Bandt was
seen as the best hope for breaking the two- party stranglehold on lower house
seats. Bandt, who went on to win, was quick to point to the importance of social
media in securing his victory, particularly the organizing power it provided.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
120 WHEN THE NERDS GO MARCHING IN
120
Table5.1 Australian Parties Citizen- Initiated Campaigning Scores, 2010
Federal Election
MyLiberal.com Labor Connect
Community Building
Prole
Photo
Biography
Why joined
Set up/ join Groups
Set up blog
Set up Wiki
Email/ message system
Externally promote prole
Subtotal (additive 0– 8)
√
—
—
√
—
—
—
√
3
√
√
√
√
—
—
√
√
6
Resource Generation
Personal fundraising
Promote membership
Sign up as local organizer
Sign up as candidate
Organize/ add event
Vote leaders to aend events
Subtotal (additive 0– 6)
—
—
—
—
—
—
0
—
—
—
—
√
—
1
Voter Mobilization
GOTV oine
Access phonebank
Sign up for f2f canvassing
Sign up to discuss with social network
Leaets download
Externally promote event
GOTV online
Send email
Post to Facebook
Post to Twier
GOTV phone app
Email forward to editor
Start e- petition
Subtotal (additive 0– 11)
—
—
—
—
—
√
√
√
—
—
—
3
—
√
—
√
√
—
—
—
—
√
—
4
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
121
The Early Bloomer 121
Table5.1 Continued
Looking back on the election, he noted that the extensive use of tools like
Facebook “early on” in the campaign had “helped us continue engagement with
our volunteers and supporters” and to recruit help from outside of the constitu-
ency as well.29 More importantly, such an approach allowed the party to carve
out a new, more democratized approach to the online campaign, which, in the
words of another candidate, was “not about control and command” but about
harnessing grassroots momentum.30
Evidence from the 2010 ACS provided fresh evidence of the Greens’ apti-
tude for social media campaigning. Aclear majority (62percent) of candidates
reported having a social network prole, compared with less than half (48per-
cent and 49 percent, respectively) for the Coalition and Labor candidates.
Furthermore, and perhaps even more importantly, the Greens continued to be
the only party that managed to convert its social media activities into increased
support on election day (Gibson and McAllister, 2015). is increased promi-
nence and success on the web 2.0 campaign front was matched by a sharp
decline in web 1.0 competitiveness. Use of personal home pages dropped again
MyLiberal.com Labor Connect
Message Production
Message creation
Policy email forward/ customize
Poster/ leaet create/ customize
Policy input/ feedback
Message distribution
Web banners/ ads download
Posters/ leaets download
Email/ share policy docs
News feed to website
Share blog posts externally
Link to SNS prole
Link to Twier account
Import email contacts
Subtotal (additive 0– 11)
—
—
√
—
—
—
√
√
√
√
—
5
—
—
√
—
√
√
—
√
√
√
—
6
Overall Score (0– 36)
Standardized Score (0– 100)
12
33
17
47
√=feature present on campaign site;— =feature not present;
Standardized scores are calculated by transforming each sub- index into a 0– 100 range and then
averaging the scores. See Appendix 4.1 for details of variable denitions.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
122 WHEN THE NERDS GO MARCHING IN
122
among Green candidates to less than one- third (31percent). By contrast, candi-
dates from the two main parties exhibited a new enthusiasm for personal sites,
with 83percent of Liberal- Nationals and 69percent of Labor contenders report-
ing they had run one during the election. According to Gibson and McAllister
(2015), however, unlike the Greens, their eorts failed to deliver signicant
electoral benets.
One nal development in 2010 that signaled not only the arrival of phase
III– style digital campaigning in Australia, but also the lack of mainstream
party engagement with it, was the success of the online activist group GetUp!
Established in 2005 as a response to the Liberal Party’s sweep of both houses of
the federal parliament in 2004, GetUp! billed itself as a le of center, nonpartisan
movement dedicated to giving ordinary citizens a new way to challenge govern-
ment policy. Following the example of MoveOn.org in the United States and 38
Degrees in the United Kingdom, GetUp! pioneered the use of email and online
petitions to lobby for progressive policy changes. e group focused particularly
on areas of particular interest to younger voters, such as electoral registration
rules, same- sex marriage, and climate change. While they had intervened in the
2007 election to help sway support for favored candidates, the 2010 election saw
them adopt more direct tactics based around Obama’s Organizing for America
initiative. In particular, they used the web and emails to recruit volunteers into a
training program— Camp GetUp— which was designed to teach participants the
new arts of viral messaging and online to oine community building (Vromen
and Coleman, 2011, 2013). According to GetUp!’s own post- election report,
their eorts were highly successful. During the campaign, they estimated they
had signed up around 7,000 volunteers, who had then gone on to contact over
200,000 voters and enroll over 10,000 new voters.31 Comparable gures on the
impact of Labor Connect and MyLiberal were not available from the two main
parties. It is unlikely that their success exceeded that of GetUp! in terms of gen-
erating support. Indeed, assuming the numbers to be accurate, it would seem
that the key players driving Australian digital campaigning out of its long period
of standardization and into activist mobilization were in fact those at the fringes
of the party system.
Phase IV:Moves towardIndividual Voter
Moblization? (2013)
Despite the evidence indicating that Australia’s two main parties had fought
shy of embracing phase III digital campaigning in 2010, expectations that
2013 would see a breakthrough into phase IV surfaced early on. Figures on
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
123
The Early Bloomer 123
usage certainly helped to bolster expectations. As Figure5.2 (reported earlier)
revealed, between 2010 and 2013 rates of mobile broadband access doubled,
creating a situation where the number of individual subscriptions actually
exceeded the population as a whole.
Headlines began appearing at the start of the year posing the question, “Will
big data change the 2013 election?” Other news reports made explicit reference
to the recent US election, noting that the major parties’ importation of Obama-
style methods meant that “Australia’s rst big data election looms.”32
Campaign insiders also appeared to share in the expectation that 2013 would
see a step- change in campaigns’ use of technology. Stephen Mills, a former
speechwriter for Labor Prime Minister Bob Hawke, who became a historian
of Australian election campaigning, argued that 2013 would denitely see the
parties embracing the new data- driven era.33 e extent to which social media
proles and email addresses could enhance the precision and scale of their
micro- targeting eorts, he contended, would not be underestimated by senior
campaign sta (Mills, 2014). Practitioners were equally optimistic, arguing that
Australia’s use of compulsory voting, rather than slowing down the adoption
of the new mobilization tools, was in fact likely to speed it up.34 Knowing for
sure who is going to turn out, according to Michelle Levine, the director of Roy
Morgan Polling in Australia, removes much of the uncertainty from forecasting
models, and improves their robustness as methods for nding and persuading
latent or undecided supporters. With the fresh data sources now available from
online proles and internet- based interactions to add to their knowledge of the
electorate, the Australian parties were likely to be in an even stronger position to
pinpoint, and unlock, those more elusive pockets of support.
Not all observers were as convinced that the parties were ready to unveil
the new scientic model of voter mobilization that the US Democrats had pro-
moted a year earlier. In practical terms, the parties lacked the scale and type
of resources necessary to engage in the type of big data operations seen in the
United States. ere was also a residual question mark over the perceived utility
of web campaigning among senior party sta. is was particularly acute for the
Liberals, who still bore the scars from the fallout of the Kenne debacle. Beyond
the humiliation that his defeat meted out to the party, however, it also prompted
questions about the intrinsic value of the medium as a tool of voter persuasion.
Federal director Lynton Crosby and ocial pollster Mark Textor both openly
cautioned candidates against placing too much faith in the internet as a tool
for undecided voters. According to Textor, who went on to serve as a key elec-
tion strategist for the party in 2010 and 2013, the key take- away from Kenne’s
failure was not the perils of over- personalization on the web, but overrating the
internet as a campaign medium. Unlike television, radio, and direct mail, the
internet lacked an “intrusion mechanism,” leaving voters free to avoid political
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
124 WHEN THE NERDS GO MARCHING IN
124
messages.35 While the push power of the new media clearly increased substan-
tially once social network platforms emerged, and made sharing information
easier, the self- selection bias that Textor had identied also persisted, leading
to the emergence of new problems around “lter bubbles” and “echo chambers”
(Bruns, 2019; Quarociocchi etal., 2016).
In addition to lingering doubts about the eectiveness of the internet as a
vote- gathering medium, there was also concern that a stronger adherence to
data privacy in Australia would act as a brake on parties’ exploitation of the
new practices. Despite their privileged access to government- held data, it was
expected that politicians would want to avoid any perception among the voters
that they were “watching them.”36 Certainly, the state- level elections that took
place immediately prior to their federal counterparts had failed to show signs
of a breakthrough for the new tactics. Bruns and Higheld’s (2013) analysis of
the parties’ use of Twier in the 2012 Queensland state election concluded that
although the ALP maintained its advantage over the Liberals, there was lile
evidence they had used it strategically “to aect electoral outcomes in any direct
way.” ere was very lile evidence of any direct voter interaction or aempts
at persuasion by candidates through their twier feeds. Furthermore, any
exchanges that did occur took place largely among the “converted” (i.e., small
subset of highly engaged partisans or political “junkies”) (688).
Although much of the political talk and headlines focused on whether
Australia was ready for a US- style data- driven campaign, the real story of the
2013 digital campaign was the continuing success of GetUp! e online activist
organization maintained its eorts to build up its online community and engage
voters through two- step communication. According to the group’s annual
report, they had registered over half a million supporters (630,000), with over
9,000 of these counted as “core members” in terms of making regular dona-
tions. In 2013, they unveiled their “most ambitious election program” to date,
which included several new tactics, specically designed to reach less politically
engaged individuals. is included a “ground- breaking” Facebook app and a
“neighbour- to- neighbour” enrollment program. e new methods, combined
with their ongoing volunteer recruitment program, led to a signicant boost in
GetUp!’s outreach activities; they estimated yielded a total of 3million voters
contacted.37
Overall, the weight of opinion and evidence suggested that phase III digi-
tal campaigning was the dominant modus operandi in the 2013 federal election.
However, the main actors driving this more devolved model of electioneering
were not the usual suspects. It was the smaller parties and particularly nonparti-
san actors who were most commied to promoting the new style of supporter-
led digital activism. e bigger parties, by contrast, remained wedded to their
more static, standardized, and safer approach. e next section of the chapter
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
125
The Early Bloomer 125
re- examines these developments with evidence drawn from the voters and party
supporters. ere is a focus particularly on the paerns of engagement between
2010 and 2013. To what extent do activities among the electorate reect those
of elites and particularly among parties’ supporters?
To investigate these questions using the approach adopted in the previ-
ous chapter, we measure and compare levels of our three main modes of
engagement— read, redistribute, and receive— over time and across parties. We
then probe the rates of direct and indirect contact received by voters in more
depth. In particular we look at the demographic and political makeup of those
being reached by the digital campaign and see how this diers from the prole
of those contacted by more traditional methods. To what extent is digital cam-
paigning allowing the Australian parties to reach a new audience?
The View fromBelow:Mapping Web Campaigning
Cycles inthe AustralianElectorate
Table 5.2 reports the levels of read, redistribute, and receive mode based on
the ndings from the 2010 and 2013 Australian Election Studies (AES). e
data series starts in 2010, as this was the rst study to include the range of items
needed to measure the three main modes of engagement. While this lack of a
longer time series reduces the scope of our analysis in terms of tracking change
and stability among the electorate, we do gain signicant insight into the two
elections that, according to our supply- side analysis, saw the most change in
practice.
A rst glance at the ndings from Table 5.2 reveals a lack of any marked
change in public consumption of digital content. Such stability is not too sur-
prising given the short time period that is covered. Typically just less than half
of Australian voters sought out news and information on the campaign in 2010
and 2013. Comparing these ndings with those from elections in the United
Kingdom (Table4.2 in Chapter4), there appears to be quite a lot similarity in
the levels of general interest with the digital campaign.
If we look at the extent of redistribution occurring, this is also similar to the
situation in the United Kingdom. Such activity was conned to a small minor-
ity of the population, and exhibited a paern of small but incremental growth.
In 2010, just over 2percent of internet users reported having both signed up to
receive ocial online campaign content and shared campaign content within
their networks. is increased slightly in 2013 to just over 4percent. Closer
inspection of the component activities of redistribution reveals that sharing con-
tent was typically more common than signing up for party or candidate updates.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
126
Table5.2 Voter Engagement inAustralian Digital Campaigns, 2010– 2013
Mode of Engagement
Year Internet use Voter as Audience— READ Voter as Activist— REDISTRIBUTE Voter as Target— RECEIVE
Online news Camp. sites Sign- up/ Download Share/ Exchange Direct (Party) Indirect (F&F)
2010
N=2,061
83%
N=1,697
41.6
(49.8)
7.8
(9.4)
3.5
(4.2)
8.6
(10.3)
1.9
(2.2)
—
—
READ 42.5 (50.8) REDISTRIBUTE 2.0 (2.4) RECEIVE (— )
2013
N=3,810
88%
N=3,353
46.1
(52.5)
12.1
(13.9)
6.8
(7.8)
10.0
(11.4)
7.9
(9.1)
7.6
(8.6)
READ 47.0 (53.7) REDISTRIBUTE 3.8 (4.4) RECEIVE 13.9 (15.9)
Figures are % and reported for overall sample and for internet users in parentheses
Sources:2010 and 2013 Australian Election Studies (AES); see Appendix 5.1 for further details of surveys and variable denitions. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
127
The Early Bloomer 127
Turning to levels of receiving, we do see some interesting paerns of change
across the two elections that bear correspondence to the preceding discussion of
elite- level activity. In particular, although we have no measure of indirect online
contact for 2010, it is clear that levels of direct contact did increase quite sharply
in 2013. is supports the view that Australian parties were becoming successful
in migrating their direct mail eorts online.
Table 5.3 presents an insight into the partisan distribution of those engaging
with and receiving digital campaign messages. is shows that all parties were
able to reach a large majority of their supporters with their online campaigns. It
Table5.3 Party Supporters’ Engagement inAustralian Digital Campaigns, 2010–
2013 (Internet Users Only)
Mode of Engagement
Election
Year & Party
Internet
Access READ REDISTRIBUTE
RECEIVE
Direct
(Party)
Indirect
(Friends &
Family)
2010
Liberal (N=746) 75.7 51.2 1.8 2.2 —
Labor (N=783) 80.2 49.8 2.3 2.9 —
Nationals (N=66) 75.8 38.0 6.0 2.0
Greens (N=121) 96.7 58.1 5.1 2.6 —
Other (N=59)
No Party (N=283)
89.8
89.0
56.6
56.0
3.4
2.0
7.5
0.8
—
2013
Liberals (N=1345) 86.9 51.5 2.1 8.7 8.0
Labor (N=1358) 86.2 56.2 4.6 7.8 8.3
Nationals (N=144) 77.9 46.8 0 6.4 7.3
Greens (N=236) 97.0 70.8 16.4 22.6 16.4
Other (N=165)
No Party (N=661)
86.6
93.7
58.9
53.2
17.1
2.3
17.8
6.6
13.2
8.2
N refers to the full sample, and the internet access gures are the % of that total who reported being
online. e gures for Read, Redistribute, and Receive are % of the online party identiers that engaged in
these activities.
Sources:Australian Election Study 2010 and 2013; see Appendix 5.1 for further details of surveys and
variable denitions. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
128 WHEN THE NERDS GO MARCHING IN
128
was the Greens, however, that occupied the leading position in this regard, with
almost universal uptake of the internet among their supporters. Furthermore,
Green partisans are among the most avid consumers of election news online.
Seventy- one percent of Green online partisans read news about digital cam-
paign in 2013, which is signicantly higher than for any of the other parties as
well as the online electorate more generally (see Table5.2).
e levels of redistribution across party supporters makes for even more
interesting reading in that again it is the minor parties that are most active in this
regard, and particularly so in 2013. is conrms the perception formed by the
preceding supply- side analysis of the lower level of commitment by the major
parties to virtual community building. In general, levels of engagement in this
more active type of sharing of party content hover between 2 and 5percent of
supporters for the two larger parties, a rate that is similar or even slightly below
the national average reported in Table5.2. Among the smaller parties, however,
the rate is considerably higher, with almost one in ve of Green supporters hav-
ing redistributed campaign content among their networks in 2013. e rate
for supporters of “other” minor parties is even higher. While we do not have
a measure of aliation to GetUp! in the AES, their origin as opposition to the
mainstream governing parties make it likely that their supporters are identifying
as “other.” is would help explain what appear to be very high rates of redistri-
bution among this group. Such a nding would also support GetUp!’s claims of
a signicant increase in their rates of voter contact during the 2013 campaign.
Similarly, for the Greens, the high level of involvement of their supporters in
spreading the word online during the campaign would help to explain Gibson
and McAllister’s (2011) nding of an apparent electoral advantage for the party
in its use of web 2.0 tools.
Turning to the partisan distribution of campaign messages, the nal two col-
umns of Table 5.3 report the extent to which supporters had received online
contact during the campaign, either directly from the parties or via their social
networks. e results show an increase in the rates of online contact across all
parties between the two elections; however, again it is the smaller parties that are
the most eective in geing their message out to their supporters. In particular,
by 2013, almost one- quarter of Green partisans reported receiving online con-
tact from the party. By contrast, less than 10percent of Labor or Liberal support-
ers received such communication. Rates of online indirect contact (i.e., political
messages coming from friends and family) were slightly less unevenly distrib-
uted between the major and minor parties, although the laer still enjoyed a
clear lead.
Looking into the partisan distribution of the receive mode across the two
years is also useful in unpacking and challenging the claims highlighted earlier
that 2013 heralded the entry of Australia into phase IV era of digital campaigning.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
129
The Early Bloomer 129
Certainly all parties increased their capacity to reach the supporters online over
the two elections. However, the biggest increase occurred among the least well-
resourced actors. is suggests that the spike in receive levels noted in Table5.2
was unlikely to have been driven by importation of the big data analytics model-
ing labs pioneered in the United States. Instead, it looks like the sharp increases
in voter contacting that occurred in 2013 resulted from the smaller parties, and
particularly the Greens, connecting more with their voters directly, and with
their core supporters, thereby intensifying their two- step communication ow
to the electorate.
e nal step of the voter- level analysis is presented in Table 5.4. is shows
the breakdown of the sociopolitical characteristics of those who received online
direct or indirect campaign contact, and compares them to those who experi-
enced oine modes (i.e., mail, face- to- face, or phone contact).
e results show similarity to those from the UK analysis reported in the pre-
vious chapter (see Table4.4). Both forms of online contact are more commonly
reported by those with higher levels of education and social status. Age- wise, the
paern appears to have changed for direct forms of contact, with younger voters
receiving less digital messaging from parties in 2013 compared to 2010. For indi-
rect contact, we don’t have corresponding gures for 2010; however, the rates
for 2013 show a strong bias toward the younger cohort, with just under half of
those receiving this type of contact being in the 18– 34 age bracket. Finally, with
regard to gender, there appears to be an increased tendency for online direct
contact to be received by men. By contrast, in 2013, women were more oen
the target of indirect online modes of campaign contact. Overall, the results sug-
gest that the parties are not widening their reach into the electorate via digital
methods of mobilization, at least insofar as their own eorts extend. ese are
received largely by the “usual suspects.” e two- step or mediated version, how-
ever, does appear to reach beyond the already mobilized, and extends parties’
reach, particularly with younger voters.
In terms of political characteristics, the results are intriguing in that they
reveal the online advantage enjoyed by the Greens in communicating with their
core supporters and activists, extending into the wider electorate. e nal row
of the table reveals that among the voters who received direct online contact, the
Greens’ support was three times greater than in the country at large. Labor, on
the other hand, and the Liberal- National coalition partners reaped no gain from
their e- campaign contacting eorts. Indeed, they both received less support
among those who were contacted by parties or candidates online than among
the electorate as a whole. Labor appeared to have a somewhat beer level of
support among voters that experienced online indirect contact, However, again
the Greens were clear winners in this regard. While these results do not oer
conclusive evidence that the Greens’ digital campaign increased their vote share,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
130
Table5.4 Socio- Demographic and Political Correlates of“Receive” Mode inAustralian Federal Elections, 2010– 2013
2010 2013
Online Direct Oine Direct All Online Direct Online Indirect Oine Direct All
Sex
Male 48 48 49 52 45 47 48
Female 53 52 51 48 55 53 52
Age
18– 34 56 24 27 29 47 25 26
35– 54 18 36 37 42 43 37 35
55+ 25 41 36 29 11 38 38
Education
Up to secondary 39 28 31 18 22 26 27
Diploma 21 37 37 28 23 33 34
Higher education 41 34 32 54 55 41 35
Social Class
Upper/ middle 72 56 54 66 72 57 53
Working 28 37 39 28 24 36 38
None 0 7 7 6 4 7 9
Vote
Lib- Nat Coalition 41 44 43 36 34 42 42
Labor 36 39 38 26 33 32 31
Greens 21 13 12 21 15 9 8
Other 3 2 2 13 13 12 12
Informal/ not voted 0 3 5 4 5 6 7
Figures should be read as column % within each demographic / political variable. ey show the proportion of individuals within each of the groups that reported online and
oine contact by parties. e “all” column reports the frequencies of the demographics within the sample as a whole. Cells may exceed 100% due to rounding.
Sources:Australian Election Study 2010 and 2013; see Table5.2 and Appendix 5.1 for further details of surveys and variable denitions. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
131
Table5.4 Socio- Demographic and Political Correlates of“Receive” Mode inAustralian Federal Elections, 2010– 2013
2010 2013
Online Direct Oine Direct All Online Direct Online Indirect Oine Direct All
Sex
Male 48 48 49 52 45 47 48
Female 53 52 51 48 55 53 52
Age
18– 34 56 24 27 29 47 25 26
35– 54 18 36 37 42 43 37 35
55+ 25 41 36 29 11 38 38
Education
Up to secondary 39 28 31 18 22 26 27
Diploma 21 37 37 28 23 33 34
Higher education 41 34 32 54 55 41 35
Social Class
Upper/ middle 72 56 54 66 72 57 53
Working 28 37 39 28 24 36 38
None 0 7 7 6 4 7 9
Vote
Lib- Nat Coalition 41 44 43 36 34 42 42
Labor 36 39 38 26 33 32 31
Greens 21 13 12 21 15 9 8
Other 3 2 2 13 13 12 12
Informal/ not voted 0 3 5 4 5 6 7
Figures should be read as column % within each demographic / political variable. ey show the proportion of individuals within each of the groups that reported online and
oine contact by parties. e “all” column reports the frequencies of the demographics within the sample as a whole. Cells may exceed 100% due to rounding.
Sources:Australian Election Study 2010 and 2013; see Table5.2 and Appendix 5.1 for further details of surveys and variable denitions. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
132 WHEN THE NERDS GO MARCHING IN
132
and self- selection eects cannot be ruled out, the disparity in the support they
received among those contacted online compared with other parties certainly
lends gives credibility to claims that they ran the most successful online mobili-
zation eort in 2013.
Summary andConclusions
Australia presents an interesting case to study when seeking to understand the
comparative evolution of digital campaigns for several reasons. First, it does
appear that developments have followed the four- phase model outlined in
Chapter1; however, progress was not necessarily as linear or steady in nature as
in the United Kingdom. Instead, advances have been more spasmodic, with an
initial burst of innovation and enthusiasm, led primarily by the mainstream le.
is was then followed a prolonged period of stasis and standardization as parties
retreated into the safety of static web content and managed interactivity. While
the causes of this retrenchment are clearly complex, one early event appears to
have been particularly inuential. e shocking defeat of the Liberal state gover-
nor Je Kenne, following a much- mocked web campaign, sent a clear warning
shot to the parties about the potential damage that a badly judged e- campaign
could do to their electoral fortunes. According to Peter Chen (2013), it took
almost a decade before Australian politicians were prepared to invest seriously
in digital campaigning again, especially in the more personalized “presidential”
style that seemed to drive innovation.
e frenzy generated by Kevin07, Labor’s 2007 online campaign eort, and
the launch of the two main parties MyBO- style social networking sites in 2010
suggested that Australian digital campaigning was back on track, and acceler-
ating toward phase IV. Closer investigation of these initiatives, however, raised
questions about the extent to which they really signaled any great leap forward.
Kevin07 was seen as an exemplar of “managed interactivity.” Certainly, the evi-
dence “from below” suggested that it was only by 2013 that digital campaigning
was succeeding in engaging activists and supporters. Furthermore, and perhaps
most signicantly, it appears from both supply- side and demand- side analyses
that these eorts were led primarily by the smaller parties and the nonpartisan
campaigners GetUp! Only a very small minority of major party supporters were
actually found to have engaged in any CIC- relevant activities.
e fact that the minor parties emerged as the strongest promoters of phase
III digital campaigning in Australia underscores its association with a swing
toward a state of greater equalization and the redistribution of power down
toward the grassroots. e prowess of the smaller players in promoting direct
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
133
The Early Bloomer 133
online mobilization in 2013 suggests this narrative continued, but also raises
something of a challenge to our ideas of phase IV being one of hypernormal-
ity and rise of the “machine- led” micro- targeting. Clearly, the smaller parties in
Australia had found a means of exploiting social media to communicate with
their core support base and directly with voters that exceeded the ambition or
skill of their larger counterparts. Whether they were able to maintain this advan-
tage in successive elections is an intriguing and important question for future
research to explore.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
134
When theNerds Go Marching In. Rachel K. Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
6
The LateBloomer
Digital Campaigning in France
Our third case study shis the analysis onto new political terrain and examines
developments in digital campaigning in the more personalized environment of a
French presidential election. As in the two previous chapters, we begin by exam-
ining France’s suitability as a context for digital campaigning. We do so using
the results of the cross- national analysis of Chapter3 as the framework for that
assessment. Where does France “sit” in comparison to other countries in terms
of its current rate of digital mobilization? How far does that position correspond
to what we might expect, given its institutional seing and level of technological
development? Finally, are there “local” factors, specic to the French context,
that might aect parties’ and candidates’ enthusiasm and capacity for digital
campaigning?
e chapter then shis the focus to examine developments in French digital
campaigning over time, and locates these recent trends in a wider historical per-
spective. Applying the framework set out in Chapter1, we trace parties’ adapta-
tion to the internet, and assess how far, and fast, they have progressed through
the four- phase evolutionary cycle. Looking rst at supply- side developments, we
use the extant literature to chart changes in the goals, tools, and organizational
infrastructure that French parties have used to wage their digital campaigns. Can
we see a similar transition to that which occurred in our earlier cases? Was an
initial experimental push online followed by a long period of stasis and profes-
sionalization, and then a burst of community building and activist mobilization,
as happened most notably in Australia? Or, have the advances in digital cam-
paigning been more incremental and linear, as in the United Kingdom? What
signs are there that the parties are now moving into phase IV and adopting a
more targeted and data- driven strategy of voter mobilization? Who is leading
that charge, and who is lagging behind? Again, do the paerns of activity seen
across the parties match with those observed in the previous two cases?
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
135
The Late Bloomer 135
e nal section of the chapter shis the lens to examine changes in the
demand side of digital campaigning. We look at how far, and in what ways, the
French public have responded to parties’ oerings over time. Has there been
a shi from voters passively “reading” about the digital campaign to actively
engaging with it and “redistributing” its contents across their networks? Do
these trends map onto the ebb and ow of elite activity documented in the ear-
lier section of the chapter? Do certain parties have a more active supporter base
than others? Finally, are the parties succeeding in reaching a new audience with
their online campaigns, and what evidence is there that it works, in terms of
inuencing their vote choices? To address this second set of voter- led questions,
we analyze national survey data collected during the presidential elections of
2007 and 2012.
France asa Context forDigitalCampaigning
As one of the world’s leading democracies, France is clearly an important coun-
try in which to examine developments in digital campaigning. France also forms
a particularly useful case for analysis at this point in the book, given that it
oers a “bridge” between the two countries just analyzed and the United States,
which follows. Like the United Kingdom and Australia, France operates a par-
liamentary system of government that relies on strong national parties, which, if
elected, implement a governmental program based on their campaign manifes-
tos. In common with the United States, however, France regularly elects a pow-
erful national chief executive who holds oce for ve years and can challenge
the wishes of majority party in parliament, if they are from the opposing side.
Since the mid- noughties, both of the two main parties also moved to adopt US-
style primaries to select their presidential candidates. is switch has enhanced
the personalized nature of the contest, and its prominence as a political event.
Given that the results of Chapter3 identied presidential elections as one of,
if not the main, driver behind higher rates of online voter mobilization, we might
thus expect France to assume a leading position, ahead of the United Kingdom
and Australia, in the intensity and maturity of its digital campaigning. Acheck on
the country rankings reported in Table3.2 of Chapter3, however, counters this
expectation. e table ordered countries according to their rates of online voter
mobilization, using data from Module 4 of the Comparative Study of Electoral
Systems (CSES). France is at the boom of tier two nations, below both the
United Kingdom and Australia. To help explain its weaker performance, we
turn rst to the ndings reported in Table3.6, which showed that several other
macro- level variables were signicant in predicting rates of online mobilization.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
136 WHEN THE NERDS GO MARCHING IN
136
ese included the presence of a proportional electoral system and a higher level
of internet use among the citizenry.
Based on these criteria, we begin to understand why France might not rank
as highly as one might expect in terms of the intensity of its digital campaign-
ing. French national elections use a two- round majoritarian system, which is one
of the least proportional methods for allocating votes to seats in international
terms. It is on the second count of internet use by the electorate, however, that
France most obviously struggles to provide a conducive environment for digi-
tal campaigning. Rates of internet use have traditionally been much lower than
is the case for most other advanced industrial nations. Indeed, from the mid-
1990s to the early part of the twenty- rst century, France consistently reported
one of the lowest internet adoption rates of any country in the developed world
(Villalba, 1999, 2003).
We can see this sluggishness in Figure 6.1, which reports the proportion of
the population that were online during presidential elections since 1997. Rates
of adoption over the time period are signicantly lower than is the case for any of
the other countries examined in this book (see Figures4.1, 5.1, and 7.1 for com-
parable statistics from the United Kingdom, Australia, and the United States).
at said, by 2012 France had clearly begun to make up for lost time, and adop-
tion levels had moved into line with those of other nations. Use of broadband
(shown in Figure 6.2) also appears to have grown at a rate comparable to that
seen elsewhere (see Figures4.2, 5.2, and 7.2), with more than half of the popula-
tion reporting xed access by 2012.
While explaining the prolonged failure of the internet to penetrate French
society is beyond the remit of this chapter, we can point to a number of likely
100
90
80
70
60
50
40
30
20
10
0
1997 2002
30
66
81
2007 2012
4
Figure6.1. Growth of internet use in French presidential elections, 1997– 2012 (%of
population using the internet).Source:World Bank, “Internet Users (per 100 people),” hp://
data.worldbank.org/ indicator/ IT.NET.USER.P2?page=1
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
137
The Late Bloomer 137
reasons for the inertia. First, on cultural grounds, some observers have pointed
to a strong distaste among French citizens toward the intrusion of this globalized
and linguistically anglicized media into their lives (Braen, 2005:519). Amore
obvious practical explanation for the lag in take- up, however, is that the public
already had extensive access to a “home- grown” version of the internet in the
shape of Minitel. Minitel was a government- funded national computer network
that was developed in the late 1970s by what eventually became French Telecom
(Breindl and Kuellmer, 2013; Kellerman, 2006). e service used the telephone
network to connect French homes to local terminals and provided access to
a wide range of government and commercial services. It proved very popular,
serving approximately 25million subscribers, or just under half of the popula-
tion, at its peak during the 1990s.1 Although usage declined thereaer, it was
only in 2012, with the onslaught of social media, that the government conceded
the internet had “won,” and Minitel was, nally, switched o.
Despite the rapid escalation of internet use in France in recent years, and par-
ticularly in broadband access, the presence of Minitel has clearly aected invest-
ment in both public and private web services. e impact on digital campaigning
was particularly signicant. According to Lilleker and Jackson (2011), the late
arrival of the web resulted in “French parties and candidates being an election
cycle behind their Anglo- Saxon counterparts” (57). We investigate the case for
this assertion in more detail in the next section of this chapter.
100
90
80
70
60
50
40
30
20
10 3
24
40
50
2002
Broadband (fixed line) Broadband (mobile)
2007 2012
0
Figure6.2. Growth in broadband subscriptions (per 100 inhabitants) in French
presidential elections, 2002– 2012.Source:OECD historical xed and mobile broadband
penetration subscriptions (per 100 inhabitants). Figures are from 4th Quarter. Available at hp://
www.oecd.org/ internet/ broadband/ 41551452.xls and hp:// www.oecd.org/ sti/ broadband/ 1.5-
BBPenetrationHistorical- Data- 2015- 06.xls (Figures for mobile available only from 2009 Q4 onward.)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
138 WHEN THE NERDS GO MARCHING IN
138
In addition to these more technical factors that have impeded web cam-
paigns’ progress in France, parties also face a series of regulatory controls on
their electoral communication, which are likely to have slowed the pace of inno-
vation. e rst, and most obvious, of these relates to the duration of the ocial
campaign period. French presidential campaigns are among the shortest among
established democracies. Candidates typically have just a couple of weeks before
the rst of the two rounds of voting to promote themselves and their policies to
the electorate. In addition, three months before the ballot, all forms of paid com-
mercial advertisements, through the press, via posters, phone calls, or by any
audiovisual means, are prohibited.2 ese pressures on the “air time” available
to them means that campaigns are arguably more likely to “default” to tried and
tested modes of contact, rather than investing in newer and unproven methods.3
As well as imposing strict limits on how long candidates can campaign, the
authorities also exert signicant control over who is heard, and what is said,
during this period. French law requires that all public service radio and televi-
sion channels (France Télévision, Radio France, and France Médias Monde)
give equal air time to all the presidential candidates and their supporters. e
rules are strictly enforced by the Conseil supérieur de l’audiovisuel (Supreme
Audiovisual Council; the CSA).4 In 2000, this rule was extended to include a
requirement that broadcasters also provide “equitable access” for political par-
ties that were not represented in Parliament (Vedel, 2005). Given that one of the
main benets of the new media over the “old,” according to the e- pluralists, was
that it provided a more open and “equalized” space for the expression of political
opinions, one can see how the French media rules might actually have reduced
the incentives for parties, particularly the minor players, to develop an online
presence.5
A nal regulatory factor to bear in mind when assessing the slower pace of
digital adoption by French campaigners is the extent of state control that has
typically been imposed on voter contacting, particularly more targeted modes,
both during and between elections. ese restrictions stem from an adherence
to the French Republic’s core principle of equal treatment of citizens before the
law. Public and private institutions are prohibited from gathering and using any
personally identifying information about citizens that could be used for pur-
poses of discrimination. is applies to a wide range of social characteristics,
such as race, religion, and ethnicity, as well as political views. ese principles
were rst enshrined into French law in 1978 in relation to electronic marketing,
when French authorities enacted the Information Technology, Data Files, and
Civil Liberty Law. e law imposed strict controls over the use of computerized
databases by public and private bodies. Akey provision was that any les con-
taining individuals’ names and personal information had to be registered with
the Commission Nationale Informatique et Libertés (CNIL). Furthermore, the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
139
The Late Bloomer 139
collection and preservation of these data required the express consent of those
mentioned in the le, and this was strictly enforced. In the case of voter registra-
tion data, access was highly restricted and made available only by request within
individual municipalities. Its use by parties was heavily circumscribed.
e upshot of these regulations has been that the French parties have faced
signicant legal and normative barriers to building up the type of large voter les
and email lists that have powered online campaigns elsewhere. Interestingly, the
introduction of the 2018 General Data Protection Regulation (GDPR),6 which
is seen as reinforcing and extending the French framework to other member
states,7 is expected by some observers to have a “chilling eect” on the devel-
opment of the new forms of data- driven campaigning emerging across North
America (Benne, 2016). Given the extent to which these laws have shaped
French campaign practice and voter expectations over several decades, it is thus
not too surprising that any chilling eects were felt here rst.
Of course, a key question that emerges from these observations is why the
national data privacy variable used in the multilevel analysis of Chapter3 failed
so emphatically to explain the variance in online contact across countries. If this
regulatory framework is important in accounting for parties’ use of new media
to contact voters, then it seems strange that it made no dierence to the pat-
terns of use we observed. Closer inspection of the index’s construction in light of
the French case suggests that some extension and revision to it may be required
in future analyses. In particular, the ve- point index covered only the presence
of regulatory agents and rules in a nation, rather than the extent of compliance
with those rules by political parties. Furthermore, it did not capture the broader
cultural views on the extent to which citizens’ personal information should be
available for political and commercial actors to use in their marketing and out-
reach. Such norms are clearly likely to inuence campaigns’ proclivity to engage
in direct online voter mobilization, and thus while dicult to quantify, some
aention should be paid to how they can be incorporated into the modeling of
the regulatory framework surrounding elections.
In the next section of this chapter, we map out the history of French digital
campaigning in more detail. Using a combination of secondary literature and
original ndings from the application of our CIC index, we track how campaigns
have changed at the supply side and how far this can be understood using the
lens of our four- phase model. How has the particular combination of contex-
tual factors highlighted earlier aected parties’ progress compared with other
countries? According to Lilleker and Jackson (2011), the slower pace of inter-
net adoption meant the French parties lagged at least one electoral cycle behind
their counterparts in other advanced democracies in their online electioneer-
ing activities. Is this indeed the case? How long did French campaigns linger in
the experimental phase before moving on to embrace a more standardized and
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
140 WHEN THE NERDS GO MARCHING IN
140
professionalized mode of operation? When and where did the rst moves into
phase III– style community building occur? Finally, which party, if any, has taken
the lead in importing the new more scientic approach to voter mobilization?
Once the problem of internet access was resolved, did the presence of presiden-
tial elections mean that the pace of development picked up? And, if so, who led
that charge?
Phase I:Experimentation (1994– 1997)
Although the time lag in mass internet use may have aected the pace of
change in digital campaigning in France, it does not appear to have delayed
its actual onset. Studies of the early years of French parties’ and candidates’
use of the web are limited, but suggest that movement online started in the
mid- 1990s. is is similar to the trends observed in the United Kingdom and
Australia (Villalba, 1999, 2003; Braen, 2005). Unlike the United Kingdom
and Australia, however, the initial entry into cyber- campaigning in France
did not come from one of the major parties. Instead, it was a smaller and then
very marginal party, the far right Front National (FN), that made the rst
foray into cyberspace in 1994. According to Braen (2005), the FN’s move
was part of a wider strategic agenda promoted by its leader, Jean Marie Le
Pen, to bypass what he saw as a highly biased mainstream media and reach
out to his supporters directly.
While Le Pen’s claims appear to challenge the argument advanced earlier that
the new media held less appeal in the French party system, due to the more bal-
anced nature of coverage provided by the “old” media, it was actually not until
2000 that the rules on “equitable access” for non- parliamentary parties— which
would include the FN— were introduced. Furthermore, according to most
accounts, Le Pen consistently portrayed the mainstream media as an enemy of
the party, and part of a wider corrupt political elite that was seeking to silence or
misrepresent its voice in French society. As such, the new media provided a very
welcome development for the party, in terms of oering a new means for com-
municating with supporters, and appealing directly to new voters (Stockemer,
2017; Quinn 2000).
e FN, however, were not alone in their early enthusiasm for the web. e
French Greens were also vocal supporters of the internet as a means of boost-
ing their political fortunes. Unlike the FN, however, the Greens concentrated
more on the internal democratizing potential of the new medium within the
party. Astatement posted on the inaugural home page overtly celebrated the
internet as a tool for empowering the grassroots. Although it is tempting to see
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
141
The Late Bloomer 141
the party as oering a highly prescient insight into the community- building and
activist- mobilizing power of the medium that we associate with phase III, the
main goal for the party at this time was seemingly just to prompt greater internal
debate and discussion— to generate what they saw as a “true place of exchange”
for members and supporters, rather than a platform to galvanize into action
(Villalba, 1999:10).
Elsewhere in the party system, the “me too” logic dominated. According to
Villalba (2003), most of the other French parties, wanted “simply to remain
visible in cyberspace” (125). The next 18months saw the familiar domino
effect in terms of digital campaign adoption. The National Assembly and
presidential elections of 1997 provided the main pressure points for par-
ties move online (Villalba, 1999, 2003). The signs of experimentation were
also widely in evidence in terms of the design and delivery of parties’ home
pages. There was a heavy emphasis on text- based content, as well as a ran-
dom and confusing array of fonts and colors. Some of this early handiwork
can be seen in Figure 6.3, which contains a screenshot of the Greens home
page from 1997. Despite having a well- developed strategic understanding
of the web, the party had clearly not yet managed to translate this into the
visual domain.
Figure6.3. French Greens home page (October 1997).Source:Wayback Machine:hp://
web.archive.org/ web/ 19961219160639/ hp:// www.verts.imaginet.fr/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
142 WHEN THE NERDS GO MARCHING IN
142
Phase II: Standardization and
Professionalization (1998– 2007)
Following this initial push online, the decade that followed saw a prolonged
period of standardization and stasis in web campaigning in France. While the
lull in proceedings followed the paern seen in the previous cases, the causes of
the inertia appeared to be rather dierent. Whereas in Australia the slowdown
was more of an internal party response to a very public early failure of web cam-
paigning, in France, it appeared to be due to external drivers, or the lack thereof.
e very slow uptake of email and the World Wide Web among citizens meant
that expectations about voters’ appetite for online campaigning remained low
(Villalba, 2003; Lilleker and Jackson, 2011; Greet, 2013).
1998– 1999
As was the case for parties elsewhere, the rst urry of anarchic website produc-
tion was followed by a period of extensive revamping and upgrading. According
to Villalba, the “proper development of the party sites” really began with the
regional elections of 1998 and the 1999 European Parliamentary (EP) elections
(1999:6). e EP elections in particular were seen as a turning point, with all
the major and key minor parties having established a home page by this point.
At the sub- national level, parties’ and candidates’ online presence also grew, and
national sites started to include links to regional branches.
Visually, sites had a more streamlined and less cluered appearance, and
greater use was made of hyperlinks to embed and organize content. One- quarter
of the parties competing in the EP elections reportedly had even gone to the
lengths of hiring a consultant to help them develop their sites (Villalba, 2003).
In keeping with the broadcast quality of phase II, interactive features such as
chat rooms and discussion forums were rare. Even the more automated forms of
interactivity, such as donation or joining facilities, were in limited supply. Some
of the smaller parties, particularly those on the le, however, did appear to chal-
lenge this trend. Most notably, the Greens and the French Communist party
(PCF) gained special mentions in post- election analyses for their promotion of
the participatory properties of the new medium (Villalba, 1999:Appendix 1,
Table1).
Despite these notable advances, pockets of amateurism and even explicit
disinterest in web campaigning persisted across the party system. According to
Villalba (2003), a minority of the parties competing in the EP elections (around
ve) still had no home page. Of those that had launched a site, most lacked an
intuitive and easy to remember URL and a small number (six) actually failed to
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
143
The Late Bloomer 143
reference to the campaign on their home pages, or through a dedicated election
site (see Villalba, 2003:Table 6.1, 128– 129). Even the FN, which was regarded
as something of a trailblazer in the experimental phase, ran into technical prob-
lems that meant their site was inaccessible for around a month prior to polling
day. Perhaps most telling, however, was the fact that almost all of the parties
failed to make any mention of the alliances that had formed to ght the election
on their home pages. For Villalba (2003), omission of this vital voter cue under-
scored the reality that most of them still failed to see the web campaign as part
of the “real campaign.”
2002
If 1998– 1999 marked a transitional or midway point between the early phases
of the digital campaign cycle, the 2002 presidential election saw French parties
move rmly into the second phase. As was the case for the United Kingdom a
year earlier, the mood was expectant, with leading politicians, including future
Socialist presidential contender Francois Hollande, pronouncing that the race
would see the internet take center stage (Villalba, 2003:127). All parties now
had a web presence (Greet, 2001; Villalba, 2003), and the French public, while
still not embracing the internet as widely as their counterparts in other democ-
racies, had experienced one of the fastest growth spurts, bringing access levels
to around a third of the French electorate (see Figure6.1). As well as bringing
the remaining stragglers into the internet era, 2002 ushered in a more stylish
and professional looking set of web campaign sites. Figures 6.4 and 6.5 show the
Socialist and Green parties’ home pages just before and aer the 2002 legislative
elections.
e gures show the parties as now adopting the more structured approach
to site design that is characteristic of the second phase. Content was divided and
layered according to sub- menus, and a more creative use was made of graphics,
hyperlinks, and multimedia features. e emphasis remained on the downward
dissemination of news and information, although there were aempts to oer it
in a more dynamic format. Reinforcing the reputation as internet pioneers, the
FN launched the rst partisan online news channel— Le Pen TV. Opportunities
for site interaction also increased, as parties oered sign- up facilities for e- news
and online membership, and almost all of the parties started to compile and use
email lists to reach out to their supporters (Villalba, 2003). Some aempts were
even made to use them for activation purposes. e FN again led the way here.
Following Le Pen’s surprisingly strong performance in the rst round of vot-
ing, the party added a feature to its site that allowed visitors to auto- generate an
email of support that they were then encouraged to send out manually, to their
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
144 WHEN THE NERDS GO MARCHING IN
144
Table6.1 French Presidential Candidates Citizen- Initiated Campaigning
Scores, 2012
Tous Hollande Comité de Soutien
2012
NS
Connect
Community Building
Prole
Photo
Biography
Why joined
Set up/ join groups
Set up blog
Set up Wiki
Email/ message system
Externally promote prole
Subtotal (additive 0– 8)
√
—
—
√
—
—
√
√
4
√
√
—
√
√
—
√
√
6
√
—
—
—
—
—
—
√
2
Resource Generation
Personal fundraising
Promote membership
Sign up as local organizer
Sign up as candidate
Organize/ add event
Vote leaders to aend events
Subtotal (additive 0– 5)
—
—
√
na
√
√
3
—
—
—
na
√
—
1
—
√
—
na
√
—
2
Voter Mobilization
GOTV oine
Access phone bank
Sign up for f2f canvassing
Sign up to discuss with social
network
Leaets download
Externally promote event
GOTV online
Send email
Post to FaceBook
Post to Twier
GOTV phone app
Email forward to editor
Start e- petition
Subtotal (additive 0– 11)
—
√
√
—
√
√
√
√
√
—
—
7
—
—
—
—
√
—
—
—
√
—
—
2
—
√
—
—
√
√
√
—
√
—
—
5
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
145
The Late Bloomer 145
networks. e FN also ran a series of webcast Q&A sessions with key candi-
dates, including one with Le Pen himself (Braen, 2005).
Although the FN clearly retained its status as an innovator in digital cam-
paigning, it formed something of an outlier. In- depth analysis by Villalba (2003)
of the 2002 campaign concluded that for the most part, “online partisan repre-
sentation...remained quite conventional.” Digital content was largely recycled
from other media channels and had “no impact...on the key functions” that the
parties carried out in the election (Villalba, 2003:135). is was particularly
evident in relation to voter interaction, which parties made very lile eort to
promote online. At best, he concluded, they opted for a form of “supervised par-
ticipation,” a notion that corresponds to that of “controlled interactivity” iden-
tied by Stromer- Galley (2014) as characteristic of parties and candidates in
this post- experimentation phase. Reinforcing this critique, Serfaty (2002) noted
Tous Hollande Comité de Soutien
2012
NS
Connect
Message Production
Message creation
Policy email forward/ customize
Poster/ leaet create/ customize
Policy input/ feedback
Message distribution
Web banners/ ads download
Posters/ leaets download
Email/ share policy docs
News feed to website
Share blog posts externally
Link to SNS prole
Link to Twier account
Import email contacts
Subtotal (additive 0– 11)
√
√
—
√
√
√
—
—
—
—
√
6
—
—
√
—
—
—
—
—
—
—
—
1
√
—
—
—
—
√
—
—
—
—
√
3
Overall Score (0– 35)a
Standardized Score (0– 100)
20
58
10
33
12
34
Standardized scores are calculated by transforming each sub- index into a 0- 100 range and then aver-
aging the scores. See Appendix 4.1 for details of variable denitions.
a e maximum raw score on the CIC index was 35 for French sites (1 point lower than for the
United Kingdom and Australia). is was due to dropping the “sign up as a candidate” variable from
the resource generation sub- index, as this was not applicable for the French presidential sites.
Table6.1 Continued
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
146
Figure6.4. e French Socialist Party home page (May 2002).Source:Wayback
Machine:hp:// web.archive.org/ web/ 20020523180652/ hp:// www.parti- socialiste.fr
Figure6.5. e French Greens home page (August 2002).Source:Wayback
Machine:hp:// web.archive.org/ web/ 20020523180652/ hp:// www.parti- socialiste.fr/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
147
The Late Bloomer 147
that the web in 2002 had served more as a “show room” for political elites than as
a genuine “chat room” for mass debate and engagement (149).
2007
e presidential election of 2007 saw a marked growth in the prominence of
web campaigning in France. Internet use had doubled since 2002, meaning that
for the rst time a majority of citizens were now online during a national elec-
tion. e intervening years had also seen a referendum held on the European
constitution in which internet technologies and particularly blogging tools had
been used extensively. is exploitation was particularly prominent by those on
the “no” side of the argument, who felt their views had been marginalized by the
mainstream media. Given that the constitution was ultimately rejected by the
public, and the “no” side won, the case for online campaigning clearly became
stronger as the presidential election approached (Maarek, 2015; Lilleker and
Jackson, 2011; Bousquet, 2009). Campaigns responded by taking a more
adventurous approach to the using the medium. Several candidates launched
their own online TV “news” channels, and there were several highly publicized
aempts by the parties to set up virtual headquarters in the animated fantasy
computer game Second Life.
Beneath the headline- grabbing initiatives, candidates also began to use the
technology to develop a more collaborative and citizen- centered (if not citizen-
initiated) model of campaign production. According to most observers, it was
Socialist candidate, Segolène Royal, who took the lead in promoting this co-
production approach. Like Howard Dean in the United States, Royal stressed
her “outsider” credentials and laid claim to the internet as a key weapon in her
bid for the nomination. Her candidacy gained a signicant boost following the
party’s decision to lower its annual joining fee to just 20 euros in the lead- up
to the primary election. e inux of new younger and tech- savvy members
helped to shi the balance of support in Royal’s direction (Pène, 2012; Vaccari,
2008a). Royal capitalized on her popularity with this new pool of supporters by
launching an innovative online discussion forum Désirs d’Avenir in 2006. e site
invited supporters to join and put their own “desires for the future” of France up
for debate within the Socialist Party (PS). It was quickly surrounded by a dense
network of independent bloggers, which became known as Ségoland. is link-
age of ocial and unocial digital platforms added a critical viral element to her
campaign communication, and ensured that she stayed closely connected with
her grassroots supporters.
While Royal’s tactics appeared to work in the short term and secured her
the nomination, they proved less eective in the general election. Despite her
stronger digital credentials at the outset, Royal did not manage to translate that
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
148 WHEN THE NERDS GO MARCHING IN
148
advantage into victory. One of the reasons for this, according to post- election
analysts, was her failure to sustain the open and participatory model of engage-
ment she had pioneered (Lilleker and Jackson, 2011). An increasing gap opened
up between her interactive online rhetoric and the actual practice of the cam-
paign. ere was a signicant “downsizing” of supporter engagement activities as
the election approached (Vaccari, 2008a). Post- election scrutiny of the content
of Royal’s online communication revealed that it had a similar, if less unapolo-
getic, “monologic” quality to that of her right- wing rival, Nicolas Sarkozy. Her
repeated use of the pronoun “we,” and the multiplicity of voices that surrounded
her online presence in the shape of Segoland, helped to create a strong sense of
grassroots involvement. However, in reality, genuine opportunities for active
involvement and co- production by her supporters were largely absent (Lilleker
and Malagón, 2010). e view that the digital campaign, and particularly Royal,
had failed to deliver on her early participatory promise was summed up neatly
by Darras (2008). His analysis of the internet campaigns of the candidates con-
cluded they had simply served to reinforce the “majeste” of the political class as
the “representatives of the people” (104).
Despite the buzz the 2007 French “Netcampagne” had generated, therefore,
the more considered verdict appeared to be that things had not really moved
on very much since 2002. For Vaccari (2008a), the 2007 election demonstrated
that France was “still at an intermediary stage” in its use of online campaigning
“...especially in terms of participation tools” (1). e internet was still a “minor
medium” in comparison with other communication channels (Darras, 2008,
104). e narrative of normalization also persisted in discussions about levels of
inter- party competition during the campaign. Several studies of the quality and
content of the minor party candidates’ eorts in 2007 concluded that they had
fallen even further behind their larger rivals since 2002 (Koc- Michalska etal.,
2014; Vaccari, 2008a; Vedel and Koc- Michalsaka, 2009).
Phase III: Community Building and Activist
Mobilization (2012)
While 2007 saw the presidential candidates dipping their toe into the pool of
online activism, the 2012 election saw the rst real drive to use the technology
to mobilize their supporter base. For regular observers of French online politics,
this election saw the rst truly “web 2.0” campaign (Giasson etal., 2014), and
represented “a substantial change in approach” from previous elections (Koc-
Michalska etal., 2014:226). Social media tools were widely used, and there was
a genuine shi from the “vertical” communication logic that had dominated
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
149
The Late Bloomer 149
internet use in earlier elections to a more “horizontal” style, which focused on
peer- to- peer interaction.
As in 2007, it was the mainstream le that took the lead. From 2010, Socialist
party candidate Francois Hollande and his team set about importing the tools and
techniques that had been developed and deployed so successfully by Obama in
2008.8 Under the direction of Vincent Feltesse, Blue State Digital was appointed
as campaign consultant. Feltesse also hired the services of a new French- based
online consultancy rm— Liégey, Muller Pons (LMP). Liégey, Muller Pons, as
the name suggests, was formed by three tech entrepreneurs. All three had stud-
ied in the United States and had observed at close quarters the Democrats’ 2008
campaign. On their return to France, they decided to set up a new rm with the
explicit mission of introducing Obama’s new digitally powered mass- canvassing
model to French parties and European party systems more generally.9
ese eorts formed a happy alliance with ongoing “renovation” activities
taking place within the PS following the election of Martine Aubry as First
secretary in November 2008. Under Aubry’s direction, the party had initiated
some fundamental reforms designed to import Obama- style practices into PS
campaign operations. is included the revival of its local- level organization and
eld activities and the establishment of a digital strategy department, headed
by Valerio Moa. According to Pène (2012), the new team consisted of at least
10 sta, making it by far the largest such unit among French parties at the time.
e equivalent unit in the right- wing Union for a Popular Movement (UMP)
consisted of just three people, while the smaller Greens dedicated one full- time
sta member to their online operations.
A key plank of the PS’s digital strategy was to build up an online supporter
network that would include both members and a larger group of interested indi-
viduals who stopped short of joining the party, but wanted to help the Socialist
cause. Following the MyBO model, La Coopérative Politique, or La Coopol, as it
became known, was launched in January 2010 for a total cost of around 300,000
euro.10 According to Benedict ieulin, the developer of the site and former advi-
sor to the Royal campaign in 2007, La Coopol was designed to be “a Facebook
for the le” in France. It formed a new “community- building platform” that gave
ordinary supporters the opportunity to aliate with the party and help during
elections, without taking out formal membership.11 According to the party, the
site proved to be very popular, with around 40,000 registered users aer one year
of operation. Forty percent of those registered were reportedly non- members.
Activity levels also appeared to be healthy, with around 2,500 groups on the site
during the rst year of its operation, although only a handful of them had more
than a thousand members.12
Hollande built on these eorts, investing heavily in his online presence and
reportedly employing a sta of 35 to run his digital campaign by election day.13
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
150 WHEN THE NERDS GO MARCHING IN
150
Furthermore, while ocial estimates of expenditure on the digital campaign
totaled just over half a million euros, informal estimates, provided by cam-
paign insiders, indicated that spending on online communications was closer
to two million euro, or 10percent of the entire budget (Koc- Michalska et al.,
2014:Pène, 2012). e centerpiece of this investment was Tous Hollande, an
online platform designed to recruit supporters who would then go on to power
the eld operations. While some reports likened it to MyBO, Obama’s 2008
social network site, Tous Hollande was distinctly less community oriented in its
content and appeal. Rather than seeking to build up a participatory Facebook
style space for group formation and discussion, however, as MyBO had done, the
main emphasis was on recruiting supporters to spread Hollande’s message and
help get voters to the polls.
e publicly stated goal of Tous Hollande was to sign up 150,000 volunteers
who would then go “porte à porte” to 5million households. To facilitate this
mass canvassing eort, visitors to the site were greeted by a landing page that
immediately called on them to “Agir pour le changement”— to act for change—
and presented them with a series of prominent invitations to donate, join a local
event, or become an “ambassador” for the campaign. Aer signing up, the new
recruits were encouraged to register and report their progress to campaign head-
quarters using TousHollande Terrain, a platform designed by the campaign team
to collate and monitor eld operations during the election (Liégey etal., 2013;
Pons, 2018). As the election approached, aempts to promote a two- step ow
model of voter mobilization via the site intensied. e volunteering option was
redesigned to target eorts on those districts where socialist sympathies were
high, but voter turnout low. e strategy was driven by a detailed analysis of
prior turnout across districts and a randomized eld experiment conducted at
the start of the campaign. is more scientic and data- driven approach to voter
mobilization was new to the French system, and constituted a direct aempt to
import and mirror the tactics that were being honed and perfected in the United
States (Pons, 2018).
By 2012, the mainstream le thus appeared to be fully commied to phase
III digital campaigning in terms of using the technology to mobilize their base.
ey also showed a growing interest in the use of phase IV– style tactics, and the
new science of individual voter mobilization. e PS was of course not alone in
shiing to adopt more strategic uses of the technology. e UMP, the Greens,
the Democratic Movement (or MoDEM), and the New Centre all established
partisan social- networking platforms by the end of 2009 (Pène, 2012).
e UMP’s Les créaturs du possible was designed to provide a new channel for
ordinary voters to become involved with the party, and act, as its name implied,
as the co- creators of political ideas and projects. Unlike its le- w ing counterpart,
La Coopol, the explicit goal of the developers was to bypass party “militants,” and
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
151
The Late Bloomer 151
create a new and bigger network of more moderate partisans, e site failed to
gain traction among the wider base of UMP supporters, however, and was dis-
mantled aer just one year of operation.14 According to the party’s own reports,
it had aracted only around 15,000 users since its launch, less than half the num-
ber reported for La Coopol. While the lack of compelling content was cited as a
key reason for its failure, the deliberate exclusion of party activists was also seen
as a major design aw. Acommon feature of La Coopol and Membersnet UK—
which were among the most successful examples of these activist hubsites— was
that they had started life as a members- only resource and had then expanded
to allow non- members to join. is mix of old and newer participants provided
both the critical mass and fresh momentum that were important to sustaining
these platforms in the longer term.
Undeterred by the UMP’s failure to build a new network of online sup-
port for their 2012 campaign, Sarkozy’s supporters launched a more person-
alized version prior to the election. In 2010 their Facebook page— Le Comité
de soutien— was established with the goal of providing a “support commiee”
for the President’s re- election bid. e page proved to be very popular and was
launched as an independent site in August 2011. In a bid to absorb some of the
momentum generated by Le Comité de soutien, the UMP established its own o-
cial version— NS Connect— creating a link to it from Sarkozy’s campaign home
page La FranceFort. Despite his rather late ocial entry on the digital campaign
scene, Sarkozy made up for lost time by investing heavily in his presence once
there. Expenditure records showed that around 5percent of his total budget, or
up to one million euros, was devoted to the online component of his re- election
campaign. is was double the amount he had spent in 2007, and exceeded by
some margin the expenditure recorded by other candidates, including Hollande
(Koc- Michalska etal., 2014).
Despite his greater investment and the initial traction gained by Le Comité de
soutien, closer inspection of Sarkozy’s ocial and unocial supporter platforms
with Tous Hollande revealed that the laer performed best. Using the CIC index
that was applied to party sites in the UK and Australian 2010 election campaigns
(results reported in Tables4.1 and 5.1), we compared the three sites on the range
of opportunities oered to supporters to engage in community- formation activi-
ties and co- production of the campaign. e results are reported in Table6.1.
e table reveals that Tous Hollande oered more of the CIC items than either
Le Comité de soutien or NS Connect. In total, Tous Hollande contained over half of
the items on the index, while the UMP’s ocial platform and the independent
Sarkozy site each delivered only slightly more than one- third. Acloser look at the
performance on each of the sub- indices reveals that Tous Hollande outperformed
Sarkozy’s ocial site, NS Connect, on all four areas of activity— community
building, resource generation, GOTV, and message dissemination— and
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
152 WHEN THE NERDS GO MARCHING IN
152
outperformed Comité de soutien on three of the four sub- indices. Aparticular
strength of Tous Hollande was its provision of opportunities for supporters to
help with message production and distribution (i.e., two- step communication).
Interestingly, the only area in which the right performed beer was that of com-
munity building. Le Comité de soutien outscores Tous Hollande by some margin
in terms of the opportunities it provides for supporters to interact and network
with one another. Although NS Connect is the least competitive of the three sites,
its strongest functional emphases are promoting voter- mobilization activities
and resource generation. Neither of the two pro- Sarkozy sites are strongly ori-
ented toward enlisting supporters’ help with message creation or distribution.
While the results of the CIC index admiedly provide only a snapshot of
each parties’ focus and priorities in the 2012 digital campaign, the ndings of
Table6.1 suggest that the PS and Hollande ran the most integrated digital cam-
paign in 2012. Tous Hollande acted as the main point of contact for the campaign
and provided both a community hub and a resource to galvanize and organize
local activists. e UMP’s campaign was more fragmented, comprising Sarkozy’s
ocial home page, France Forte, plus his party- run supporter site, NS Connect,
and the more organically driven grassroots Le Comité de soutien. is disaggre-
gation meant that the CIC tasks were distributed across platforms, making it
dicult for Sarkozy to develop the same critical mass and sense of a “joined- up”
community that Hollande inspired.
Viewed in comparative perspective, the ndings reported in Table6.1 show
some interesting similarities and contrasts to the ndings from previous chap-
ters. In all three countries, it is parties on the le that appear to be leading the
move into phase III campaigning. e overall score achieved by Tous Hollande is
actually slightly lower than that of UK Labour’s Membersnet in 2010, but slightly
higher than the ALP’s Labor Connect.15 Closer inspection of the results, broken
down by index, reveals some important additional nuances to this ranking, and
the varying functional emphases of sites across countries. In particular, we can
see that Hollande’s site promoted more of the activist mobilization elements of
CIC, while the Australian and UK sites placed more emphasis on community-
building activities. ere were no facilities, for example, on Tous Hollande for
supporters to start a blog or to provide a personal testimonial to explain why
they were supporting Hollande. ese were, however, prominent features on
both the Labour Party and ALP platforms.
Such dierences, while they may be reective of a more ruthless focus by the
PS and particularly Hollande’s team on winning, are also likely to be indicative
of the transitional nature of digital campaigning in France at this time, and the
shi toward more instrumentalist phase IV goals. ere was no doubt a desire by
Hollande’s digital team to build the type of community spirit that had inspired
the “Deaniacs” and the creators of MyBO in the United States, and their own
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
153
The Late Bloomer 153
party- based La Coopol. However, the idea of community building for its own
sake, rather than as a means to an end, was coming under question. e key
focus now was to realize the external or “real world” value of the community
built by this type of activist mobilization, by channeling these activists to take
oine action and GOTV.
While none of the candidates reported the sign- up gures for their CIC sites,
or total voter contacts made through their CIC platforms, the proxy statistics
that were available to gauge interest indicated they were of limited appeal to the
French public. According to a report from TechPresident, Hollande’s team col-
lected 650,000 addresses by the end of the primary election in October 2011,16
which equated to around 1.5percent of the voting- age population.17 It is perhaps
signicant that the gure was not updated by Feltesse during the course of the
campaign, and was clearly well below the 13million, or 6percent of US eligible
voters, that Obama had collected by the end of 2008.18
Subsequent analysis by Vincent Pons, who had been a key member of
Hollande’s digital advisory team, LMP, provided further evidence to suggest
that the CIC sites had failed to spark signicant public engagement with the
campaigns. Reviewing the results from an internal survey of almost 2,000 PS
canvassers who had gone “porte à porte” during the campaign, he found that
most of those who reported using Tous Hollande to support their eorts were
already party members. According to his study, only a small minority (just over
10percent) of those who signed up with Tous Hollande had never previously
been involved with a campaign. For Pons, this failure to draw in new activists via
the digital campaign platform sharply contrasted with Obama’s success in 2008
with MyBO, the eectiveness of which he had been able to observe at rst hand
during his work on the Democrats’ online campaign (Pons, 2018).
Beyond the production of CIC sites, 2012 also saw the expansion of candi-
dates’ presence across social media platforms. Use of Twier and Facebook was
ubiquitous, although in a nod to the early days of website experimentation, some
of the fringe candidates failed to set up a prole (Koc- Michalska etal., 2014).
As with the establishment of his personal supporter network NS Connect, the
launch of Sarkozy’s ocial proles on key social media platforms came aer that
of his main rivals. His Twier site went live on February 15, 2012, the day he
announced he was beginning his ocial re- election campaign. While the delay
in launching his prole to coincide with the announcement of his candidacy was
no doubt a deliberate move to maximize the impact of both initiatives, it under-
lined an ongoing ambivalence in his relationship with the new media. During
his presidency, Sarkozy had gained a reputation as an opponent of net freedoms
aer he spearheaded several policies that were designed to crack down on digital
piracy and copyright infringement (Sarkozy, 2011; Breindl and Kuellmer, 2013;
Nastasia, 2014).19
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
154 WHEN THE NERDS GO MARCHING IN
154
Despite the question marks that appeared to hover over Sarkozy’s commit-
ment to free speech on the internet, the strategically timed release of his Twier
prole appeared to work and it quickly gathered momentum, aracting over
40,000 followers aer its rst day of operation. By election day, this had increased
to just under 300,000. is was just slightly lower than the total achieved by
Hollande, who had opened his account over one year earlier. Sarkozy himself
wasted no time in using the channel, adopting what one critic called a “machine
gun” approach to his Twier feed. According to one report, Sarkozy (or his advi-
sors) sent around 50 tweets per day during the campaign. Hollande, in compari-
son, sent less than half that number. Both candidates stood in sharp contrast to
Obama, however, who had kept his daily rate of tweets well in the single digit
range. Quantity was also not a substitute for quality. e tweets that emanated
from both camps were seen as bland, lacking in interactivity, and were rarely sent
by either candidate personally.20
Performances on Facebook were stronger. Sarkozy in particular enjoyed
a commanding lead over Hollande, with over half a million Facebook likes by
election day. Akey element of his success was the addition of a timeline to his
prole. is was a purpose- built app developed by his digital team and installed
with special permission from Facebook. It was designed to promote interac-
tivity among users of the site by allowing his supporters to share their videos,
comments, and photos with other users and with the candidate. e timeline
feature gave Sarkozy a distinct edge over Hollande in terms of the reach of his
campaign, particularly to younger voters. By establishing a space for supporters
to congregate and exchange ideas on Facebook, Sarkozy was essentially taking
his campaign out to the people, rather than simply waiting for them to visit his
purpose- built community platform. Whether he was able to convert this online
advantage into “real world” votes was the question. Based on his following across
both main social media platforms, he was still only reaching around one percent
of the total electorate directly. Furthermore, a signicant proportion of those
people were likely to be supporters of his campaign already, having chosen to
join his network. Certainly, the fact that he lost to Hollande suggested that his
internet popularity was not enough to carry him over the line to victory. at
said, the indirect reach of his campaign, particularly through his large Facebook
following, should not be underestimated. We return to the impact of the candi-
dates’ indirect online mobilization on their electoral support in the nal section
of this chapter.
While the candidates’ investment in their CIC sites and social media proles
clearly increased, this was not the case for other more “standard” components
of their digital campaign production. Analysis of the presidential candidates’
main home pages in 2012 by Koc- Michalska etal. (2014) revealed that, despite
improvements in their technical quality since 2007, sites were actually less
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
155
The Late Bloomer 155
content rich than in the earlier election, particularly in terms of their interac-
tive opportunities. Controlling for an expansion in site features over time, the
authors found that the proportion oering users the opportunity to submit
comments had dropped from just under half in 2007 to less than a third in 2012.
e use of discussion facilities had disappeared entirely. Judged through the lens
of their conventional home pages, one may be tempted, therefore, to see French
parties’ digital campaigns as moving even further into managed or controlled
interactivity and one- sided communication. Given the wider context of plat-
form expansion, however, the results suggests a more nuanced story of substitu-
tion and transference of both resources and functionality.
One additional byproduct of the reduced investment by parties in their home
pages and increased use of social media was a boost in the online performance
of some of the smaller parties (Koc- Michalska etal., 2014). Despite the absence
of some of the fringe candidates on Twier and Facebook, several of the more
competitive minor parties saw the 2012 election as something of a watershed or
breakthrough moment in their eective use of the medium. e Greens, in par-
ticular, spent signicantly less on their online presence in 2012 compared with
2007,21 but regarded their digital campaign as much more successful. According
to the digital director, a key element of this success was the decision to divert
resources from standard web- marketing tools to social media and particularly
to the development of a network of “cultural creatives.” ese were volunteers
who helped design innovative content, such as animations and videos that were
designed to go viral, and build up a wider network of supporters. is approach
was arguably a more authentic version of message co- creation and distribution
than that practiced by either of the two major parties. It also appeared to be far
more cost- eective. e Greens’ digital campaign sta estimated that the party’s
cultural creative network helped generate an email list of 60,000 contacts and
donations of 150,000 euro.22
The View fromBelow:Mapping Web Campaigning
Cycles inthe FrenchElectorate
We now switch to examine the response of the French public to the online cam-
paigns of presidential candidates over time. In particular, we seek to establish the
extent to which voters have undertaken our three modes of engagement with
parties’ digital campaign content— reading, redistributing, and receiving—
during recent elections. How far do changes at the mass level correspond with
those we observed earlier at the elite level? As in our previous chapters, data to
measure citizens’ online political activities at the level of granularity we require
here are not available for all the elections included in the supply- side analysis.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
156 WHEN THE NERDS GO MARCHING IN
156
Our analysis is thus conned to the two recent presidential election campaigns
of 2007 and 2012.
In 2007, we can map the frequencies of two of our three modes of activity
among the electorate using a pre- election online survey conducted by the
Centre for Political Research, Sciences Po (CEVIPOF). While we can create
some measures of “read” and “redistribute” that are comparable to those used
in our other country studies, and to those available for 2012, we do not have
a measure of “receive” in 2007 (i.e., the rates of online contact of voters). In
addition, because it was an online survey, we are able to calculate estimates
of the modes of engagement only for internet users, not for the population
as a whole. In 2012, we use data from a translated version of the same “e-
campaign” module that was elded in the 2010 UK and Australian parliamen-
tary elections. is makes the gures for read, redistribute, and receive for
France in 2012 directly comparable to those reported for the United Kingdom
and Australia in those elections. e questions for 2012 were elded as part
of the post– French National Election Study (FNES), which is a probability
face- to- face sample of just over 2,000 adults and includes both internet users
and non- internet users. Further details of the surveys and variable construc-
tion are provided in Appendix 6.1.
Table 6.2 shows the basic frequencies of the three modes of engagement
between 2007 and 2012. Aquick glance at the ndings reveals that the numbers
engaging with digital campaigns were considerably higher in 2007 than in 2012,
and that this dierence holds even aer controlling for internet use (i.e., compar-
ing the gures in parentheses).23
is paern clearly challenges the ndings from the two previous chapters,
both of which had shown that the rates of popular engagement in online elec-
tions were increasing over time (see Tables 4.2 and 5.2). Closer scrutiny of
the data sets reveals a number of other signicant dierences in the samples.
In particular, we nd that respondents in the 2007 survey had considerably
higher levels of interest in the election than those surveyed in 2012.24 As numer-
ous authors have documented (Kaye and Johnson, 1999; Sanders etal., 2007;
Gibson and McAllister, 2008; Dillman, 2011), this type of skew is common
in online surveys and is a selection, rather than a mode, eect. Voluntary par-
ticipation in any survey tends to aract respondents who are more interested,
informed, and concerned about the survey topic, and who hold viewpoints that
are stronger and more extreme than those of other individuals (Dillman, 2007;
Cook etal., 2000; Wu & Weaver, 1997). is problem is exacerbated in online
surveys, where respondents are typically recruited on an opt- in basis rather than
via random selection. Given the potential bias this introduces to our analysis, we
focus more on the changes in the balance of engagement among respondents
within each survey over time, rather than the absolute levels of change.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
157
Table6.2 Voter Engagement inFrench Presidential Digital Campaigns, 2007– 2012
Mode of Engagement
Year Internet use Voter as Audience— READ Voter as Activist— REDISTRIBUTE Voter as Target— RECEIVE
Online news Candidate sites Sign- up/
Download
Share/ Exchange Direct
(Party)
Indirect (F&F)
2007
N=1,004
66% 44.9
(68.0)
29.5
(44.7)
3.6
(5.4)
17.2—
(26.0)
—
—
—
—
READ 45.5 (69.0) REDISTRIBUTE 2.2 (3.3) RECEIVE—
2012
N=2,005
77%
(1549)
32.9
(42.8)
16.1
(20.9)
4.8
(6.2)
10.6
(13.8)
4.6
(5.9)
6.1
(7.8)
READ 35.1 (45.6) REDISTRIBUTE 2.6 (3.4) RECEIVE 9.9 (12.7)
For 2007 the survey was online and it was not possible to calculate the population internet use. We use instead rate the 66% from 2007 reported in Figure6.1 is forms our baseline to
extrapolate the read, redistribute and receive gures. See note 23 for further details of the calculation. e 2012 survey included internet users and non- internet users; gures are reported
for the overall sample and internet users in parentheses.
Sources:2007 Internet panel “Observatoire de la Netcampagne” conducted by CEVIPOF/ IFOP, April 2007; 2012 French National Election Study conducted by Fondation Nationale
des Sciences Politiques/ TNS Sofres, May– June 2012; see Appendix 6.1 for survey further details of sur veys and variable denitions. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
158 WHEN THE NERDS GO MARCHING IN
158
Looked at through this lens, there does not appear to have been any major
shi in how voters engaged with parties’ web campaigns between 2007 and
2012. “Read” is the dominant form of engagement in both years, although inter-
est in the candidates’ sites vis- à- vis online news sources appeared to drop in the
2012 campaign. Sharing political content was a popular activity across both
years, exceeding sign- up to parties’ and candidates’ newsfeeds or social media
proles by some margin. Interestingly, the combination of the two (sign- up and
share), which measures the extent of redistribution of campaign content by sup-
porters, is the one mode that does appear to have grown in both absolute and
relative terms across the two elections. is suggests that the parties’ eorts to
encourage this type of activity in 2012, documented earlier, had been eective
in terms of cuing through and prompting a response.25
As the 2007 survey did not include measures of whether respondents were
contacted online by the parties, or by friends and family, it was not possible to
assess whether there had been any relative growth in the “receive” mode. e
overall proportion of the electorate receiving any type of online contact reported
in 2012 is around one in 10. We know from our comparative ranking of countries
in Table3.2 and the previous chapters’ ndings that this is somewhat lower than
for the most time- comparable elections in the United Kingdom (2010, 16%)
and Australia (2013, 14%). Furthermore, if we compare the number of voters
reporting direct online contact from parties to those who had signed up for such
messages, it is almost identical. is suggests that parties’ use of digital contact
was unlikely to be reaching new voters and thus having any genuinely mobiliz-
ing eects. We return to explore this conclusion in further detail in the following
when we investigate the sociopolitical prole of those receiving contact from
parties and friends and family.
Before looking in more depth at those engaged in “receive” mode in 2012,
we examine the changes in the partisan distribution in the dierent modes of
engagement between 2007 and 2012. is is helpful in probing below the top-
line gures presented in Table6.2 to see how eective the parties were in engag-
ing their supporters in the more active forms of online campaign activities. Table
6.3 reports the key ndings from this analysis. e columns report the total pro-
portion of those expressing particular partisan identication who had engaged
in a particular mode of activity. We report the gures for internet users in both
years to control for overall growth in internet use. As was the case in Table6.2,
given the dierences in the samples across the two years, we focus on changes in
the relative frequency of the dierent modes within the parties across elections,
rather than any change in the absolute levels of these activities.
e results show that, as was the case for voters in general, the dominant
mode of engagement with web campaigns among partisans is passive consump-
tion of election material, or “read” mode, and that this holds for both elections.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
159
Table6.3 Party Supporters’ Engagement inFrench Presidential Digital
Campaigns, 2007– 2012 (internet users only)
Mode of Engagement
Election Year
and Party
Internet
Access READ REDISTRIBUTE
RECEIVE
Direct
(Part y)
Indirect
(Friends &
Family)
2007
PS (N=260) — 78.1 4.2 — —
UMP (N=187) — 76.5 4.8 — —
UDF (N=113) — 77.9 1.8 —
Other (N=184) — 67.9 3.8 — —
No Party
(N=241)
— 50.4 1.6 — —
Average — 69.2 3.4 — —
2012
PS (N=497) 78.9 49.2 4.6 9.2 4.9
UMP (N=472) 68.7 49.1 1.2 6.5 7.1
MoDEM
(N=131)
87.0 61.4 2.6 2.6 17.5
Le Party
(N=124)
88.6 56.9 8.3 10.0 7.3
FN (N=155) 81.8 34.6 5.6 4.0 8.7
Other (N=164) 83.5 70.3 6.6 10.1 18.1
No Party
(N=445)
74.3 23.1 0 0.6 4.0
Average 77.2 45.7 3.3 6.0 7.6
For 2007 Internet access by party was not available due to the survey having been conducted online
(i.e., it excluded non- internet users). For 2012, N refers to the full sample, and the internet access gures
are the % of that total who reported being online. e gures for Read, Redistribute, and Receive are
% of the online party identiers that engaged in these activities. See note 24 for further discussion and
explanation of the apparent sharp drop in Read over the two years.
Sources:2007 Internet panel “Observatoire de la Netcampagne,” conducted by CEVIPOF/ IFOP,
April 2007; 2012 French National Election Study conducted by Fondation Nationale des Sciences
Politiques/ TNS Sofres, May– June 2012; see Appendix 6.1 for full variable denitions, sample size,
and survey details. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
160 WHEN THE NERDS GO MARCHING IN
160
When we compare across parties, however, see that interest in the web campaign
among minor party supporters was just slightly below that of the major parties’
adherents in 2007. By 2012, the situation appears to have reversed, and levels of
interest among supporters of the smaller parties had risen considerably above
average. e FN is the exception here in that “read” rates among its supporters
are lower than for other parties and similar to those without any party aliation.
Such disinterest among FN supporters is particularly surprising given the early
initiative shown by the party in using the medium and the strategic value placed
on it by the leadership. One possible explanation for the disparity, however, is
that FN supporters heeded the words of Le Pen and were more selective in the
information they accessed about the election online. ese were people who
were more likely to avoid the mainstream media and instead conned them-
selves to a more limited range of content that was supplied primarily, if not exclu-
sively, by the FN. We return to this point in the following discussion.
Table 6.3 also reveals that, similarly to the United Kingdom and Australia,
only a very small minority of supporters for any party were engaged in any redis-
tributive activities in any election. Somewhat surprisingly, given the widespread
excitement that had surrounded the PS candidate’s online presence, the rates of
engagement in this more active mode by supporters were quite similar across
the two main parties. is provides some credibility to the criticisms of Royal
that she lost that participatory impetus as the general election approached. By
2012, the situation had moved more in support of the supply- side narrative and
ndings from Table6.1, with the PS and Hollande taking the lead in terms of
their rates of redistribution compared to the mainstream right. While just under
5percent of PS partisans reported this type of CIC activity, this was true for only
one percent of UMP partisans.
Perhaps the most striking finding to emerge in Table6.3, however, is the
ability of the minor parties to encourage redistribution among their sup-
porters. This is particularly evident in 2012. Fringe players, along with more
prominent smaller parties on the far right and the far left, enjoyed higher
rates of redistribution among their supporters than either of the two major
parties. While this disparity may result from the smaller parties compen-
sating for their inability to engage in other, more expensive ways of mobi-
lizing their activists, it does suggest that the larger parties’ investment in
their dedicated CIC platforms did not pay off. The former’s exploitation of
cheaper social media tools seemed to have paid higher dividends. The find-
ings are particularly interesting in that they replicate those from Australia.
As revealed in Table5.3 in Chapter5, the Australian Greens also outper-
formed both the mainstream left and right parties in terms of the amount
of phase III co- production activities that their supporters engaged in during
the 2010 federal election.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
161
The Late Bloomer 161
Turning to look at the extent of online contact received by party supporters,
we can see that PS identiers were somewhat more likely than their UMP coun-
terparts to report having received some kind of digital messages from a cam-
paign. Almost 10percent of those who identied with the PS (9.2), and who
were online, received an SMS, email, or social network message from a party or
candidate. Given the gures reported do not specify the source of the contact, it
is not possible to know which party the messages were from and thus to assess
who was ultimately most eective in their online mobilization eorts.
In addition to reinforcing the dominance of the PS’s digital campaign in
2012, there is also evidence that the strong performance by the minor parties
continued, although this is not universal. Supporters of the Le Party (Parti de
Gauche, PG), a breakaway from the PS in 2009 that formed part of the “Le
Front” alliance in the rst round of the 2012 presidential elections, reported a
higher than average rate of direct online contact. is is also true for support-
ers of “other” parties (i.e., the smaller and more genuinely fringe organizations).
Partisan identiers with the center- right MoDEM and the far right FN, however,
reported much lower levels of digital contact by parties and candidates. Given
that MoDEM supporters also recorded lower redistribution rates compared
with other minor party adherents (as shown in Table6.3), the obvious conclu-
sion to draw is that the party was simply less active in reaching out to its activists
online.
e lower rate of receive among FN supporters, however, is less easily
explained by party failure, given that the levels of redistribution were compa-
rable to those seen in other parties. One explanation for this disparity could be
that the receive measure does not identify the source of the contact in terms of
the party it came from, but captures overall exposure to messages received from
any campaign. As such, FN supporters may be as engaged as other partisans in
receiving and sharing online campaign content, but also much beer at avoiding
online messages from parties other than the FN. is understanding of a greater
selective exposure by FN supporters to digital campaign contact aligns with the
ndings about their lower than average rate of reading about the campaign. eir
paerns of consumption and receipt of online election material are typically nar-
rower and more exclusive than is the case for other party identiers.
Levels of indirect online contact among party supporters (reported in the
nal column of Table 6.3) tell a somewhat dierent and arguably even more
interesting story than for direct mode, particularly for the major parties. e PS
performs worst on this measure in 2012, with only around 5percent of their
supporters saying that they had received some type of online mobilization mes-
sage from someone they know personally. By contrast, the UMP performs rela-
tively well, with just over 7percent of its supporters reporting some informal
online contact about their vote. Among the minor parties, the rates of indirect
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
162 WHEN THE NERDS GO MARCHING IN
162
contact are even higher than for the UMP, and in some cases twice the level seen
in the major parties. e jump into the lead on this measure by the MoDEM is
particularly surprising given that it had one of the lowest rates of direct contact
with its supporters. One possible explanation is that as a centrist party, its sup-
porters occupy more heterogeneous political networks than those on the le or
right. is median ideological position may mean they are more likely to receive
a wider range of crosscuing messages, compared to the average partisan.
A second explanation for its greater success in facilitating networked com-
munication among its support base, ironically, may stem from the party’s failure
to engage in this more directly and ocially. MoDEM supporters are essentially
taking active measures to ll the gap le by the party, in terms of maintaining
regular communication with its core support base. is idea is given some sup-
port by the relationship we observe between direct and indirect contact among
other parties’ supporters. Essentially, where party- directed communication is
higher, the rates of intra- supporter interaction appear to be lower, and vice versa.
ose parties that appear to be less eective or interested in engaging in direct
online contact with their supporters enjoy a higher rate of informal online con-
tacting among their grassroots.
Overall, it would seem that the evidence of Table6.3 aligns with the conclu-
sions from our supply- side analysis that the Socialists outperformed the UMP in
mobilizing their base online, and pushed the party system toward phase III of dig-
ital campaigning. However, it also indicates that this narrative needs extending to
recognize the important role played by the minor parties in these developments.
In particular, the claims by the smaller parties such as the Greens that they were
eectively exploiting social networking sites to generate CIC activity is borne out
by these ndings. Supporters of the smaller parties emerge as the most active in
digital campaigns. ey tend to consume more online political content, and are
more likely to spread it across their networks and to receive online communica-
tion from parties. Furthermore, their performance in this regard appears to have
strengthened over time. e proportions engaging with digital campaigns, par-
ticularly in more active and strategic ways, increased more among supporters of
some the minor parties in 2012 than was the case for the major parties.
In the nal step of this analysis, we take a closer look at those in the wider
electorate that did receive digital messages during the campaign. While it was
a limited pool of voters who found themselves in this position, the question of
who they were and whether those eorts produced any genuine mobilization are
interesting and important to explore. To do this, we report the key demographics
and vote choice of those voters who reported various types of campaign contact
in 2012. We report the gures only for 2012 given the lack of variables to mea-
sure “receive” in 2007. An initial look at the frequencies in Table 6.4 reveals that
those receiving online contact from the parties or via their personal networks
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
163
The Late Bloomer 163
were not strongly representative of the wider population on certain major socio-
demographic characteristics. In particular, those reporting such contact were
typically younger, more highly educated, and of higher income26 than the aver-
age sample respondent.
Table6.4 Socio- Demographic and Political Correlates of“Receive" Mode inthe
2012 French Presidential Election
Online Direct Online Indirect Oine Direct All
Sex
Male
Female
Age
18– 34
35– 54
55+
Education
Up to secondary
Baccalaureate
Higher education
Income
<1,000 euro
1,001– 2,000
2,001– 3,000
3,001– 4,000
>4,000
Vote
Hollande
Sarkozy
Bayrou
Mélenchon
Le Pen
Other
Null/ spoiled
50
50
31
36
33
21
20
59
6
35
26
18
15
44
23
3
13
10
7
0
54
46
46
34
20
16
26
58
6
26
28
16
24
20
17
14
19
16
13
2
46
54
28
24
47
35
15
50
11
24
19
17
20
40
27
3
9
13
9
0
48
52
26
34
40
47
20
33
9
36
28
12
12
27
26
9
11
18
6
3
Figures should be read as column % within each demographic / political variable. ey show the
proportion of individuals within each of the groups that reported online and oine contact by parties.
e “All” column reports the frequencies of the demographics within the sample as a whole. Cells may
exceed 100% due to rounding.
Source: 2012 French National Election Study conducted by Fondation Nationale des Sciences
Politiques/ TNS Sofres, May– June 2012; see Appendix 6.2 for further details of sur vey and variable
denitions. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
164 WHEN THE NERDS GO MARCHING IN
164
Interestingly, these biases are more pronounced among those receiving con-
tact through their online networks than from the parties. us, while circulation
of information through informal peer- to- peer communication may have helped
the smaller parties compete, it would also appear to have been conned to a small
set of elite citizens. e prole of those receiving oine contact from parties
(face- to- face, mail, and phone combined) corresponds more closely to the over-
all sample characteristics, particularly in terms of education and age. Notably,
however, beyond the educational bias, the lower age and average income levels
among those receiving direct online contact does suggest the new mode of o-
cial communication by candidates is reaching of a set of less engaged voters than
occurs through more traditional methods.
In terms of the extent to which the contact appeared to have a mobilizing
eect, the nal set of rows of the table reports the voting behavior of those who
were contacted via dierent modes, compared with the sample overall. e g-
ures show the proportion of those within each contact category who voted for
a particular candidate in the rst round of the presidential election. us, we
can see that the majority of those who reported receiving online messages from
a party or candidate (online direct) voted for Hollande (44%), while just less
than a quarter (23%) voted for Sarkozy. is is clearly a dierent paern of sup-
port than held for the electorate as a whole, particularly for Hollande, and is
repeated for oine forms of contact. Essentially, those who received some form
of oine, or online, prompt from an ocial campaign were much more likely to
have voted for the PS’s candidate.
Among the group who received online indirect contact, however, the story is
more mixed. Support for Hollande is much lower, and actually less than among
the electorate as a whole. e gap for Sarkozy is less stark, but in answer to our
earlier question about the possible ripple eect of his Facebook following, he
does not seem to have gained signicant electoral benets from a two- step mes-
sage ow among voters. Conversely, the minor parties do appear to have been
beneciaries of the online informal peer- to- peer communication that took place
during the election. e support for Mélenchon, the candidate of the Le party,
is almost twice as high among those who reported receiving some kind of online
political message from a friend or family member during the campaign, com-
pared to the electorate as a whole.
Drawing any causal conclusions from these data is dicult, given their cross-
sectional nature. e ndings regarding direct online contact, however, at least
conrm that Hollande’s digital team was the most successful in connecting with
those who ultimately went on to vote for the candidate. Furthermore, given the
disproportionately higher level of support that Hollande received from voters
contacted by oine methods, this provides at least some prima facie evidence
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
165
The Late Bloomer 165
that the party was successful in linking its online and traditional modes of voter
mobilization. When it comes to indirect online contact, however, the situation
appears to be reversed, with the support levels for the smaller parties being dis-
proportionately higher among the voters who received this type of peer- to- peer
contacting. Whether these informal cues mobilized additional support for non-
mainstream candidates in the rst round of presidential voting is dicult to tell
from these data alone. e younger age prole of those receiving indirect con-
tact, reported earlier, suggests a greater openness to peer group inuence. On
the other hand, the higher socioeconomic status of this group points to their
having higher levels of interest in the election, and being more likely to have
decided how they will vote.
Conclusions
is chapter has shown that although digital campaigning began at around the
same time in France as in other established democracies, it gained momentum
more slowly. In terms of our four- phase model, development eectively stalled
at the standardization stage for a period of almost a decade and a half. Between
1998 and 2012, elites made increasingly positive noises but largely hollow ges-
tures toward exploiting the interactive and mobilizing potential of the technol-
ogy. By 2012, however, the parties nally began to make up for lost time, and
the presidential election saw both major players launch themselves into the two-
step- style mobilization eorts associated with phase III. While the mainstream
le led the way in terms of the supply and demand for this new style of digital
campaigning, the smaller parties proved to be particularly adept in mobilizing
their base with the new tools. Using the cheaper infrastructure of social net-
working sites such as Facebook, they achieved twice as much reach into their
supporter base, compared with the PS.
e failure of the major parties to fully exploit the community- building
aspect of these new platforms, however, ironically may have signaled their focus
on the more instrumental and strategic aspects of the new web 2.0 tools. In com-
ing late to the digital campaign scene, they had an opportunity and incentive to
incorporate the best of international practice. e ndings drawn from both the
supply- and demand- side analyses point to the fact that the primary focus of the
major parties’ digital campaigns by 2012 was their external reach into wider elec-
torate. e parties, it seemed, had now nally grasped the power of social net-
works not only for connecting their activists, but also for organizing them, and
particularly for ensuring that they helped bring voters to the polls. As leading
scholars on French digital campaigns put it, the parties had at last moved from
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
166 WHEN THE NERDS GO MARCHING IN
166
viewing the technology as simply an “additional” election tool to an “integral”
multi- platform operation (Koc- Michalska etal., 2014).
Looked at in comparative perspective, the developments in France appear
to have followed a somewhat similar paern to that observed elsewhere. e
United Kingdom and Australia both experienced long periods of stasis and even
stagnation in the evolution of digital campaigning at the national and local level.
is was then followed by a concentrated burst of activity as the mainstream
le and several minor parties advanced onto to a phase III– style approach. As
in these other countries, the shi appears to be powered by dierent technol-
ogy. For the larger parties the emphasis is on bespoke portals and ocial party
hubs. For the smaller parties there is more of a reliance on free platforms such as
Facebook and open source soware such as NationBuilder.
While it is clear that digital campaigning as a practice has advanced rapidly
in recent years in France, the same acceleration is not evident among French
voters. Engagement with web campaigns, while it is growing, remains a minority
pursuit. Even by 2012, only around one- third of the electorate and less than half
of internet users were accessing news and information about the election online.
is is noticeably lower than was the case in both Australia and the United
Kingdom, where a majority of internet users were engaged in such activities in
proximate general elections. Despite this greater reticence, it seems the subset
of French voters who do get more involved with the campaign do so in a more
active manner than is seen elsewhere. Rates of redistribution among French vot-
ers in 2012 are comparable to, if not slightly higher than, those among British
voters three years later. Furthermore, it is again the mainstream le and minor
party supporters who are most willing to engage in this type of two- step activ-
ity. While such ndings suggest a growing strength on the le for digital cam-
paigning, whether it produces an advantage on election day is not clear. Based
on the evidence of this chapter, it would appear that the targeting and recircula-
tion of digital campaign content is not extending parties’ reach very far beyond
the “usual suspects.” However, among those they manage to connect with, the
Socialists do appear to enjoy a distinct electoral advantage over other parties,
specically in terms of their direct messaging of voters. By contrast, the minor
parties emerge as the main beneciaries of the more viral and networked forms
of digital messaging.
Whether this paern continues is clearly an interesting and open question.
e strong legal and cultural constraints on French parties’ use of voter data,
for both online and oine contacting, combined with the new EU- wide push
toward stronger enforcement of these restrictions under GDPR, suggests that
a more extensive rollout of phase IV– style data- driven micro- targeting tech-
niques is unlikely to gain a signicant head of steam for the foreseeable future. As
such, it may be the minor parties that assume the reins of innovation from here,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
167
The Late Bloomer 167
building on their prior record of success in promoting more indirect methods
of supporter and voter mobilization. Alternatively, following the normalization
line of argumentation, it may be that the major parties launch a “land grab” for
more terrain in this valuable new online sphere of informal inuence and persua-
sion. Rather than spending their greater resources on more targeted Facebook
advertisements and carefully craed email appeals, therefore, the mainstream
players will focus their energy and resources on manufacturing a greater supply
of “authentic” organic content, and investing in the mechanics to ensure its viral
spread.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
168
When theNerds Go Marching In. Rachel K. Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
7
TheTrendsetter
Digital Campaigning in the United States
A truth held to be virtually self- evident is that the United States leads the way
in digital campaigning. Although we have drawn on a range of countries’ experi-
ences to design the four- phase model presented in Chapter1, the United States
formed a critical case in building that framework. Its prominence in developing
this narrative stems in part from the fact that developments in the United States
have typically been the focal point of media and academic aention. e history
of digital campaigns’ evolution in the United States is thus among the most well-
documented in global terms. Interest spiked most notably aer the meteoric rise
(and fall) of Howard Dean in the 2004 presidential election, and again aer the
unexpected victory of Barack Obama in 2008. In each case, the candidates’ suc-
cess was widely regarded as the result of their innovative use of new web 2.0 tools
such as blogs, social networking platforms, and video- sharing sites. e promi-
nence of the United States as a leader in the digital campaign stakes, however,
also stems from the reality that parties around the world have explicitly sought to
import and copy the practices developed there. All three of the prior case stud-
ies provided evidence of parties having sought out American digital campaigns’
expertise. For some, this has involved dispatching “fact- nding” missions to the
United States to directly observe the candidates’ election preparations in situ.
Others have brought key personnel over to their campaign headquarters to pro-
vide strategic advice. In some cases, both approaches have been adopted.
Such transference of practice is, of course, nothing new. It maintains a tradi-
tion that began in earnest during the 1980s. Indeed, so slavish was the mimicry
of US techniques across democratic elections that the term “Americanization”
became shorthand to describe the changes occurring globally in campaigning
during the late twentieth century (Negrine and Papathanassopoulos, 1996).
ough the idea that US campaign methods could be directly transplanted into
other national contexts was subsequently questioned as a more adaptive process
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
169
The Trendsetter 169
of “modernization” was identied,1 the view of the United States as the engine
of electioneering innovation still resonates strongly within political communica-
tion scholarship (Negrine, 2015). In this chapter, we take a closer look at this
claim in relation to digital campaigning. How rapidly and in what ways have US
candidates and parties adapted to internet- based technologies? Is their reputa-
tion for global leadership in this domain truly warranted?
To do so, we proceed, as in previous chapters, by rst describing the extent
to which the United States oers an hospitable context for advances in digital
campaigning. How far does it correspond to the “ideal” context for online mobi-
lization revealed in the analysis of Chapter2? We then document developments
in the supply of digital campaigns over time, focusing particularly on the pace of
change seen in presidential elections. How far do advances in the tools, content,
and strategic focus of candidates’ eorts correspond to the four- phase model
outlined in Chapter1? How rapidly did progression through the evolutionary
cycle take, and where does the United States now sit in comparison to the coun-
tries examined in earlier chapters? Which party or parties have been instrumen-
tal in driving that progress? In the nal section of the chapter, we turn to look at
how US voters have responded to digital campaigns over time and the extent to
which any changes in demand have corresponded to those of supply.
The United States asa Context
forDigitalCampaigning
Based on the ndings of Chapter3, we know that US citizens experience some
of the highest rates of online mobilization worldwide. is prominence was not
particularly surprising given that their political environment features several of
the key traits that were found to be associated with more intense rates of digi-
tal contact. In particular, the United States enjoys regular presidential elections,
which are typically hard- fought and close- run aairs. Furthermore, although the
general election is decided by a combination of plurality voting and the “winner
take all” logic of the Electoral College, the much longer primary season that pre-
cedes it includes numerous intra- party contests that are fought under more pro-
portional methods. is means that US presidential elections can be seen, to an
extent, as oering the more open and “challenger friendly” type of environment
associated with PR, which the analysis of Chapter3 linked with more intensive
online campaigning.
Finally, the United States also provides a technological environment that,
according to the analysis of Chapter 3, supports growth and innovation in
digital campaigning. Americans have always been seen as “early adopters” of
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
170 WHEN THE NERDS GO MARCHING IN
170
new communication tools, with rates of television and radio take- up having
been among the highest and fastest growing in the world (Comstock, 1978;
Kellerman, 1999). As Figures 7.1 and 7.2 reveal, the United States has main-
tained its reputation with regard to internet adoption rates. Close to one in ve
of the US population (16percent) were online for the rst web election in 1996.
is is a substantially higher proportion of the electorate than in any of the
100
90
80
70
60
50
40
30
20
10
1996
16
43
65
74
79
2000 2004 2008 2012
0
Figure7.1. Growth of internet use in US presidential elections, 1996– 2012 (%of
population using the internet).Source:World Bank, “Internet Users (per 100 people),” hp://
data.worldbank.org/ indicator/ IT.NET.USER.P2?page=1
90
80
70
60
50
40
30
20
10
0
2000 2004
Broadband (fixed line)Broadband (mobile)
2008 2012
2
13
26 29
82
Figure7.2. Growth in broadband subscriptions (per 100 inhabitants) in US presidential
elections. 2000 and 2012.Source:OECD historical xed and mobile broadband penetration
subscriptions (per 100 inhabitants). Figures are from 4th Quarter. hp:// www.oecd.org/ internet/
broadband/ 41551452.xls and hp:// www.oecd.org/ sti/ broadband/ 1.5- BBPenetrationHistorical-
Data- 2015- 06.xls (Figures for mobile are available from 2009 Q4 onward.)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
171
The Trendsetter 171
countries examined in previous chapters. e United Kingdom comes closest,
with just one in 10 voters reportedly online during the 1997 election. If we look
at rates of xed line broadband access, US levels are also ahead of or comparable
to those of other countries examined in earlier chapters. Notably, mobile sub-
scriptions have surged by 2012 and put the United States close to the top of the
international leader board among OECD nations.
As well as meeting these more generic macro- or system- level criteria for
higher rates of digital campaigning, there are other aspects of the US political
context that help to boost levels of activity and innovation among parties and
candidates. In particular, the constitutional guarantee of freedom of expression
has, according to some observers, created a much more liberal culture of infor-
mation sharing and access to voter records than is seen elsewhere. According to
Benne (2016):
Any understanding of the US context has to begin with the overwhelm-
ing inuence of the First Amendment on the communication of political
speech and the raising of money to facilitate that communication. (262)
Such circumstances mean that any aempt to regulate the ow of personal infor-
mation within society has to confront “very powerful [legal] arguments” against
limiting that ow. For Benne (2016), the Help America Vote Act (HAVA),
passed in 2002 in the wake of the irregularities that emerged in the 2000 presi-
dential election, was particularly instrumental in helping parties build up their
enormous voter databases. e act eectively mandated states to create compre-
hensive computerized voter registration lists. ese lists, in combination with
the culture of tolerance toward electoral proling, set the parties up in a posi-
tion where, as one campaign manager later boasted, they were able to “measure
everything.”2
e importance of this more open environment for digital contacting by par-
ties contrasts quite strikingly with the more closed context of France, discussed
in the previous chapter. It also helps to further account for the failure of the data
privacy index, which was our proxy measure for state regulatory controls on par-
ties’ use of voter data, to achieve signicance in the analysis of Chapter3. As
noted in our discussion of the French case, the index only measures the formal
requirements covering public and private organizations’ use of personal data.
It does not account for the broader culture of enforcement surrounding those
requirements. While this was likely to have resulted in the signicant underesti-
mation of the constraints that parties faced in pursuing new strategies for online
voter outreach in France, in the United States the opposite arguably holds. Here
parties operate with greater license since they do so in the context of a “con-
stitutional framework that favours the almost unfeered ow of personal data
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
172 WHEN THE NERDS GO MARCHING IN
172
for campaigns” (Benne, 2016:262). Requirements for data protection ocers,
consent, and criminal sanctions for breeches of privacy notwithstanding, the
reality is that US campaigners at the state level have an almost unrestricted abil-
ity to gather, store, link, and interrogate increasingly vast quantities of personal
voter information that are now available to them.
is preliminary overview has conrmed the United States’ status as a lead-
ing nation in the digital campaign stakes, and has helped shed some light on the
factors that can explain why that might be the case. We turn now to examine
that reputation in greater detail, looking at developments over time through the
lens of our four- phase model. How much faster did the US parties progress dur-
ing the early phases of digital campaigning compared with those in the United
Kingdom, Australia, and France to achieve their current rates of GOTV and
direct voter mobilization? Which parties were most inuential in pushing prog-
ress along? At the supply side, was it the case that US electorates were always
more demanding of digital content, or has that happened later as the parties’
oerings have improved? In partisan terms, who is most likely to get involved in
digital campaigns, and to what extent are these eorts and more direct targeting
by the parties bearing any fruit in terms of securing additional electoral support?
Phases Iand II:Experimentation,
Standardization, and Professionalization
(1996– 2000)
As discussed in Chapter1, it is now widely accepted that the rst public use of
the internet in an election campaign occurred in the United States when for-
mer California Governor Jerry Brown emailed supporters in his bid for the US
Senate (Janda, 2015). At the presidential level, it was Bill Clinton who took the
honors, posting the content of his speeches and radio transcripts to a publicly
accessible URL. Given that the pool of internet users at that time was extremely
small and access required a modicum of computer skills, the site’s impact was
eectively next to nil. However, according to Davis and Owen (2008), Clinton’s
eorts constituted “the genesis of online campaigning” (95). Beyond the major
parties, the smaller players also displayed a keen interest in using the new tech-
nology, with the Libertarian party being the rst to launch a national party web-
site (Margolis etal., 1997).
Although internet campaigning made its debut in 1992, it was the presiden-
tial election of 1996 that formed its real anno domini (Klinenberg and Perrin,
2000; Bimber and Davis, 2003). is was the year in which the rst national
campaign websites were launched and candidates for the presidency began to
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
173
The Trendsetter 173
wake up to the importance of having an internet presence. For Stromer- Galley
(2014), it was the moment at which digital campaigning became embedded as a
communication “genre” (38). Despite its instantiation as a tool of electoral com-
bat, however, most candidates, even those at the national level, appeared to be
unclear as to the point of their web campaign. Indeed, it was the US 1996 presi-
dential election that prompted Selnow (1998) to introduce the “me too” label
to capture the lack of purpose and ambition characteristic of sites at that time:
Like kids who challenge each other to go rst, once the rst dare-
devil successfully crosses the stream or climbs the tree, the others feel
obliged to follow. And so it was for some campaigns with Websites in
1996. (89)
Candidates’ uncertainty about the reason for having a website was matched by
their ineptitude in helping voters to nd them online. At least two candidates
neglected to include their names in their website address, and the addresses that
were chosen indicated that the sites were viewed simply as temporary xtures
rather than long- term investments (Stromer- Galley, 2014:23). Email was used
in a limited manner with e- news updates available through a laborious subscrip-
tion process that entailed nding the address and sending an email with a blank
subject line. Rather surprisingly, given his initial enthusiasm for the medium in
1992, Bill Clinton proved to be one of the weakest proponents of web campaign-
ing in 1996. According to Stromer- Galley (2014), Clinton and his team judged
the web to be a low priority, preferring instead to rely on his White House pres-
ence until a few months prior to the general election. e address of the campaign
site that was eventually launched did not actually include either the president’s
or vice president’s names directly. Instead, the designers chose a highly unmem-
orable abbreviated version in the form of www.cg96.org. e overall commit-
ment of resources to their web presence was minimal and they relied, at least
initially, on volunteer labor, rather than paid sta. e web operatives who were
recruited were nested within the IT and tech team, a move that underscored
their perceived lack of relevance to the strategic aims of the campaign. Below the
presidential level, signs of disinterest were even more palpable, with rates of web
uptake among the two main parties’ congressional candidates estimated to be
between 16 and 19percent (D’Alessio,1997; Bimber and Davis, 2003).
One exception that emerged in 1996 to challenge this narrative of experi-
mental amateurism was the unexpectedly strong performance of Republican
candidate Bob Dole, who was by far the oldest contender in the eld. Despite
his advanced years, Dole gained widespread acclaim for the sophistication of
his online presence, the centerpiece of which was a state- of- the- art website. As
has already been seen in Figure2.3 (Chapter2), not only did the site have an
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
174 WHEN THE NERDS GO MARCHING IN
174
intuitive URL, www.dolekemp96.org, making it easy to nd, it was well designed,
with a user- friendly menu bar to help navigate the contents, high- quality graph-
ics, and dynamic content. Aparticularly noteworthy innovation was the pres-
ence of an interactive map that allowed users to follow Dole around the country
by clicking on regional and city markers (see Figure 7.3).
Despite the apparent incongruity of septuagenarian Bob Dole running a
cuing- edge new media operation, his eorts showed an early recognition by
mainstream political actors in the United States of the potential value of digital
campaigning. e person who was charged with running Dole’s online campaign,
Robert Arena, proved to be something of an early visionary (Stromer- Galley,
2014). Shortly aer his appointment, Arena circulated what could be regarded
as a personal manifesto for changing campaigns through use of the internet. His
“New Media Blueprint” argued that the key value of the web lay in its power
to draw in and mobilize supporters to help promote the candidate. Such ideas
went far beyond current practice and foreshadowed the two- step activism and
citizen- initiated campaigning (CIC) that was to follow a decade later with the
rise of Dean and then Obama. Under Arena’s stewardship, Dole even embraced
an early version of CIC by providing options for volunteers to design their own
campaign poster, send a personalized e- postcard to friends in support of Dole,
and to download promotional tools such as screen savers and printable leaets
Figure7.3. Dole/ Kemp 1996 site— interactive pages (November 1996).Source:Wayback
Machine:hp:// web.archive.org/ web/ 20160616195030/ hp:// www.dolekemp96.org/ interactive/
interactive.html
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
175
The Trendsetter 175
to distribute. Although these innovations have since been overtaken and forgot-
ten in the excitement generated by the arrival of social- networking tools, for
Stromer- Galley (2014) Arena’s work in this regard was seminal and constituted
“the blueprint for digital media strategy for political campaigns” (35).
THE 2000 PRESIDENTIALELECTION
Following the 1996 election, full entry into phase II was swi. e 2000 election
cycle saw both the Democratic and Republican parties step up their internet pres-
ence signicantly. According to Farmer and Fender (2005), both the Democratic
National Commiee (DNC) and the Republican National Commiee (RNC)
developed a national internet campaign strategy for the forthcoming elections.
e DNC reportedly commied around one million dollars to establishing a
new “Democrat Internet Center” (48). e RNC, in turn, increased its web team
to 15, making it ve times larger than the comparable unit at the DNC (Janda,
2015:25). For the presidential race itself, all the main candidates had an online
presence by the spring of 1999 (i.e., well in advance of the primary season)
(Benoit and Benoit, 2000). Domain names were rationalized, budgets increased,
and web campaign directors appointed with a clear line of communication to the
campaign manager (Stromer- Galley, 2014). e amateur experimentation that
had characterized most home pages in 1996 had largely disappeared. In its place
were sleeker, more professionalized looking sites that displayed much greater
uniformity in style and content (Hansen and Benoit, 2005). Visitors were now
typically greeted with a logo and easy- to- navigate menu bar that provided access
to a range of biographical information, news and press releases, issue positions,
and information on how to get involved in the campaign.
Improvements in site quality corresponded to a growing awareness among
designers of their key audience. Journalists and the mainstream media more
generally were now seen as one of the main consumers of online campaign con-
tent (Bimber and Davis, 2003; Ku, Kaid, and Pfau, 2003). is growing recog-
nition led to a more controlled and conservative approach to content creation
and an embrace of the “broadcast” mode of television and radio. Sites followed
a top- down “one to many” or point- to- mass model of voter communication.
Home pages became lile more than static vehicles to store and re- present mate-
rial designed for other channels, such as press releases or TV advertisements.
Notably, it was around this time that the terms “electronic brochure- ware” and
“online billboard” started to emerge to describe candidates’ and parties’ web
eorts (Kamarck, 1999, 2002; Gibson, 2012; Stromer- Galley, 2014). is more
polished, but static and unadventurous approach to site creation meant that
any conclusions drawn about an equalization in major and minor players’ per-
formance was largely illusory, and of limited signicance (Benoit and Hansen,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
176 WHEN THE NERDS GO MARCHING IN
176
2005; Gibson etal., 2003b). Challengers and those on the fringes might be pro-
viding equivalent content to that oered by incumbents and those with more
resources; however, the benchmark for quality was arguably low enough that it
made parity in provision almost inevitable.
Looking at congressional races, it was clear that diusion in web campaign-
ing had increased signicantly since 1996. Figures reported by Kamarck (2002)
estimated that 90percent of all Senate and gubernatorial major party candidates
now had some type of web presence. Her ndings for House of Representative
races showed a more modest but signicant rate of growth. Athird of candi-
dates were found to be online in 1998, and two- thirds by 2000. Acloser look at
the quality of the sites that were produced, however, made it clear that simply
having a web presence did not necessarily mean that candidates were strongly
commied to using the medium. According to most accounts of campaign sites
around this time, content was limited, with lile aention given to updating it,
or providing opportunities for user interaction (Leiter, 1995; Davis and Owen,
1998; Faucheaux, 1998; Selnow, 1998). Among the candidates that did make
an eort, normalization appeared to be the order of the day, with resources and
incumbency found to be the key drivers of online campaign intensity (Herrnson
etal., 2007).
By the turn of the millennium, therefore, web campaigning in the United
States appears to have moved squarely into phase II. As in 1996, however,
there were glimpses of future trends that emerged from unexpected quarters.
One of the most prominent and eective of these early net pioneers was the
Independent candidate for the Minnesota governorship, former wrestler Jesse
Ventura. Ventura stunned his rivals and most observers when he fought and won
the 1998 gubernatorial election aer running a highly eective online grassroots
campaign (Hindman, 2005; Davis etal., 2009). Christened “Jessenet,” the net-
work of online activists that built up around his candidacy through email lists
and discussion groups formed an early prototype of the much larger national
communities that Dean and then Obama created using digital technology.
In the presidential election itself there were some ickers of a more dynamic
and mobilizing approach to the medium as some of the lesser known candidates
trialed some new methods to encourage volunteers to sign up as local orga-
nizers. Republican Steve Forbes was perhaps the most notable on this front,
using his website to recruit “e- precinct leaders” to help run his campaign eorts
locally. According to his team, the program was a huge success, with over 5,000
e- precincts established and 30,000 cyber- volunteers signed up (Bimber and
Davis, 2003:40; Stromer- Galley, 2014:66). Forbes’s ideas were picked up later
in the election by his Republican rival George W.Bush, and on the Democratic
side by both New Jersey Senator Bill Bradley and Vice President Al Gore. Both
the RNC and DNC also moved to set up facilities for supporters to sign up as
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
177
The Trendsetter 177
“e- leaders” on their home pages. For Michael Turk, an internet strategist who
worked on Bush’s 2004 e- campaign, these initiatives, although rudimentary,
“laid the framework for the sophisticated tools that would come in later cycles”
(Turk, 2012:50).
Looking even further ahead, while none of the virtual artillery and infrastruc-
ture associated with the fourth phase of digital campaigning made an appear-
ance in this election cycle, there were some notable aempts to use the medium
to directly inuence voter decision- making. One of the most prominent of these
eorts came from the Greens, who launched a website designed to encour-
age vote- swapping between their supporters and Democrats in a bid to unseat
Republicans in marginal seats. Although it lacked the laser- like precision and
mobilizing power associated with phase IV– style campaigning, the initiative did
indicate that the parties— and particularly the minor players— were beginning
to grasp the strategic power of the web for oine electoral gains. As noted in
Chapter4, similar eorts were made by the Liberal Democrats in the UK general
election a year later, with a site being designed explicitly to encourage tactical
voting between their supporters and Labour voters, in order to unseat unpopu-
lar Conservative MPs.
Phase III:Community Building and Activist
Mobilization (2004– 2008)
Despite the growing awareness among campaigners of the practical value of the
web, particularly in terms of its organizing and resource generating potential, it
was not until the presidential election of 2004 that its value as a mainstream elec-
tion tool began to be recognized.
CITIZEN- DRIVEN CAMPAIGNING (2004)
A major impetus behind this shi in thinking stemmed from changes to the
legal framework governing elections and specically the passage of the 2002
campaign nance law. e new legislation imposed much weaker controls on
internet- based communication compared with other more established elec-
tronic modes such as television and radio. In particular, the web remained one of
the few areas of campaign activity where parties could spend their large reserves
of so- called so money.3
It was Howard Dean’s bid for the Democratic presidential nomination in 2003,
however, that provided the real “breakthrough” moment for internet campaign-
ing not only within the United States, but also globally. In the relatively short
period of time that his election bid survived, Dean’s clever use of the internet was
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
178 WHEN THE NERDS GO MARCHING IN
178
seen as propelling him from relative obscurity to front- runner position. In mate-
rial terms, he proved more successful in raising online funds than any previous
contender. By the time he bowed out of the race in early 2004, he had raised just
over 50million dollars, with half of that total coming through internet sources
(Hindman, 2005; Kreiss, 2012).4 In human capital, he had registered half a mil-
lion supporters online and prompted 70,000 volunteers in over 600 cities to set
up local support groups (Hindman, 2005; Stromer- Galley, 2014).
Critical to Dean’s successful exploitation of the medium was his appoint-
ment of Joe Trippi as his campaign manager. Trippi, a veteran campaigner at
the national and state level, also had an understanding of technology start-
ups and saw digital as the key to raising the profile of the then little- known
governor from Vermont. The new manager also had an eye for talent and
recruited a large group of volunteers and staffers with IT expertise. One of
these new hires was Joe Rospars, who went on to develop Barack Obama’s
e- strategy four years later, and to cofound Blue State Digital— one of the ear-
liest and best- known online campaign consultancy firms. Trippi’s core team
hit double digits in size (15) by the end of the campaign and, perhaps even
more importantly, moved up in the organizational hierarchy to take a seat at
the top table (Stromer- Galley, 2014). This more elevated status meant that
the internet started to play a stronger role in the day- to- day running of the
campaign and the conduct of major tasks such as fundraising and volunteer
recruitment (Hindman, 2005). For Vaccari (2008b), it was this widening
remit that constituted the real “game changing” aspect of Dean’s e- campaign.
It marked the point when the technology moved from being considered as
simply a tool for service delivery and became part of the “backbone” of
operations. For seasoned watchers of US campaigns like Larry Sabato, the
events of 2003 showed the medium was now “finally living up to its promise”
and shaping up as “one of the primary vehicles for both organization and
coverage from now on.”5 Simon Rosenberg, president of the centrist New
Democrat Network, went even further in recognizing the significance of
Dean’s campaign, noting that he was to the internet what “JFK was to televi-
sion, and Goldwater and McGovern were to direct mail.”6
Despite its headline- grabbing quality, the technical infrastructure employed
by the Dean camp was surprisingly basic, relying essentially on a blog, a web-
site with a dynamic fundraising baseball bat, and the free third- party soware
Meetup.org. rough the clever interweaving of these components, however,
Dean was able to carry out the core campaign tasks of engaging and informing
his activists, raising money, and reaching out to a wider support base. In simple
terms, his online buzz carried over to oine action, and it was this integration,
or merger of “mousepads and shoe leather,” as Trippi termed it, that proved to be
the real key to his success.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
179
The Trendsetter 179
Dean’s great strength, unfortunately, also proved to be his Achilles’ heel. His
decentralized use of the technology was vital in establishing a national network
of support; however, it also meant that local groups operated largely autono-
mously. Group leaders were self- appointed and maintained only limited contact
with central sta (Kreiss, 2012; Wolf, 2004). is lack of coordination among
Dean’s ground troops hampered the campaign’s GOTV operations, particu-
larly in Iowa, where it led to an inux of around 3,000 out- of- state “Deaniacs”
who alienated, rather than appealed to, voters (Kreiss, 2012). is laissez- faire
approach extended to the campaign’s day- to- day online communication with
supporters. Analyses of Dean’s blog revealed a surprising lack of dialogue with
users. Instead, rather like Segolène Royal, the le- wing candidate who competed
for the French presidency in 2007, Dean was regarded as promoting a “paraso-
cial” form of interactivity— the aim being to convey an image or sense of enjoy-
ing a face- to- face relationship with one’s supporters, while not actually engaging
in any direct and sustained manner with them (Stromer- Galley, 2014; Kreiss,
2012). is combination of a fragmented online infrastructure and the lack of
continuous and authentic communication with his base no doubt contributed
to Dean’s loss in Iowa. However, his problems arguably highlighted the deeper
fundamental tension that existed for campaigns at the time in trying to blend
the new more participatory version of the internet that had been unleashed with
web 2.0 tools with the practical demands of a serious bid for presidential oce.
Ultimately, it seems that the technology “won” and overshadowed the campaign
and the candidate.7 Or, as Joe Trippi (2004) put it, in his post- mortem on the
Dean campaign, “we were not using the Internet. It was using us” (103).
As one might expect, developments in web campaigning did not end with
Dean’s exit from the race. John Kerry and George W.Bush both went on to build
up sizable digital support teams that assumed an organizational importance
equivalent to that of the more established units such as eld, media, and fund-
raising (Stromer- Galley, 2014:98). Both candidates also easily surpassed Dean’s
online donation record and generated email lists that counted subscribers in the
millions, rather than the hundreds of thousands. It was the Republicans, how-
ever, that really picked up and ran with the networked model of campaigning
that Trippi had developed. e release of the Personal Precinct platform allowed
Bush to capture the energy and enthusiasm of his “netroots” in a similar way
that Dean had done through his Meetups. e site allowed volunteers to sign
up directly with the campaign as local organizers and undertake a series of tasks
to support Bush. Arange of online support services and resources were made
available to help them in carrying out these activities. is included access to
phone records and canvassing sheets that volunteers could download and use in
a new GOTV “walk and phone” experiment. Incentives to join the scheme were
further increased by the addition of a “Campaign Leader Board” to the site. is
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
180 WHEN THE NERDS GO MARCHING IN
180
important innovation allowed the campaign to publicly recognize and reward
their most successful recruits (Milkis and Rhodes, 2009; Vaccari, 2008b).
In building up their partisan network, the Bush campaign clearly drew inspira-
tion from Trippi’s model of recruiting volunteers and outsourcing key campaign
functions to them. By channeling these eorts through the Personal Precinct
soware, however, the party also added a layer of central coordination and con-
trol that meant it could vet and monitor its local operatives. e end result was
the creation of a new pool of online volunteers who were explicitly aligned with
the central aim of geing Bush elected. While the Republicans’ victory in 2004
clearly cannot be aributed wholly, or even predominantly, to its use of digital
tools, this innovative mix of top- down management and promotion of local tac-
tical autonomy was clearly a part of its winning strategy. It was this new style of
“controlled decentralism” in digital campaigning that, as we shall see in the fol-
lowing, formed the prototype, if not the inspiration for, developments in 2008.
CITIZEN- INITIATED CAMPAIGNING2008
By 2008, it was clear to both candidates and their managers that the internet
could unlock valuable new resources in terms of grassroots support; however,
this process required careful management and direction. In phase III terminol-
ogy, if 2004 was all about online community building, 2008 saw aention switch
to the activist- mobilization side of the equation and how to transfer that energy
into practical gains. Central in driving forward this new agenda were a group of
former Dean staers who had regrouped following his defeat to establish the
online consultancy rm Blue State Digital (BSD). Hired in 2006 by Democratic
hopeful, Illinois Senator Barack Obama, BSD was not only fully cognizant of
value of the medium as a means of empowering activists, but also aware of the
need to stay focused on the primary goal of winning the election. eir approach
thus followed the Bush model of controlled decentralism, and centered on a
“citizen- initiated” rather than “citizen- driven” model of digital campaigning
(Gibson, 2015). To deliver on this goal, they developed a new suite of centrally
provided tools that supporters could use to “meet up” à la Dean. However, criti-
cally, they were then directed to undertake specic actions for the campaign.
At the heart of these eorts was Mybarackobama.com, or MyBO, a volunteer
management platform that was launched just aer Obama announced he was
running for oce in February 2007. Built at an estimated cost of two million
dollars with input from Facebook cofounder Chris Hughes (Clayton, 2010;
Towner and Dulio, 2011), MyBO provided an ocial space for social network-
ing among supporters. Aer entering an email address, users could set up a pro-
le, launch a blog, and join or start a group to interact with like- minded others
(see Figure 7.4).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
181
The Trendsetter 181
Beyond the community spirit it inculcated, however, there was a steely focus
on ensuring that this enthusiasm translated into practical assistance. Volunteers
were given targets, ocial training in canvassing, and how to access and update
the campaigns’ voter les. eir performance was measured through quantita-
tive indicators that registered the frequency of their logins to the site and how
many online and oine contacts they had made (Levenshus, 2010). While this
combination of “top- down and p2p boom- up organizing” as noted earlier, had
already been trialed by the Republicans toward the laer stage of the 2004 elec-
tion cycle, BSD added a scale and intensity to the process that transformed it
into a highly eective mainstream campaign tool (Castells, 2009; Lilleker and
Jackson, 2010; Gueorguieva, 2008; Montero, 2009; Kalnes, 2009; Johnson,
2011; Kreiss, 2012; Karlsen, 2013; Stromer- Galley, 2014).
Organizationally, the ascent and expansion of the digital team within the
campaign hierarchy also continued. Nowhere was this more evident than in
Obama’s headquarters. While almost all candidates appointed a digital direc-
tor or chief internet strategist (Stromer- Galley, 2014:108) Obama’s director of
new media, Joe Rospars, was by far the most prominent and inuential, enjoying
a direct line of communication with campaign manager David Ploue, and an
unprecedented level of input to core decisions on eld and nance operations
(Levenshus, 2010). His team was also the biggest, peaking at 81 full- time sta by
the general election, a gure that doubled if one included the volunteers working
Figure7.4. MyBO.com (March 2007).Source:Author’s archive.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
182 WHEN THE NERDS GO MARCHING IN
182
in central and regional oces.8 e expansion in size promoted an increas-
ing internal dierentiation of skills and expertise. Specialized sub- teams were
formed, each of which had responsibilities for distinct areas of activity, such as
site design, online organizing, video production, and email messaging. Afurther
quota of around 25– 30 sta were kept on “stand- by” and were draed in to sup-
port local eld oces as needed (Levenshus, 2010).
Publicly, the numbers reinforced the view of 2008 as a transformational year
in US digital campaigning. Obama’s email list swelled to 13 million, almost
doubling the numbers achieved by George W.Bush in 2004. e number of
e- volunteers recruited through MyBO and other platforms reportedly reached
5million, well in excess of the 1.6million that had signed up to support John
Kerry (Cogburn and Espinoza- Vasquez, 2011; Costa, 2009; Milkis and Rhodes,
2009; Vaccari, 2008b). It was the growth in online fundraising, however, that
provided the clearest indicator of the tectonic changes occurring in the appeal
of e- campaigning. At the close of the race, media reports revealed that Obama
had secured contributions of up to half a billion dollars from up to 3million
online donors.9 is extraordinary feat meant that for the rst time ever, the
Democratic nominee for the US presidency could opt out of receiving public
funding and the spending limits that acceptance of this money entailed. Equally
signicantly, his Republican rival, John McCain, was not at liberty to do the same.
While much of the money that Obama raised went into television advertising,
especially in the closing days of the campaign,10 the crucial role that Rospars and
his team had played in generating these vast new stores of cash through their cre-
ative email program was undeniable. According to established journalists, like
Antonio Vargas at the Washington Post, “No other major campaign this cycle put
technology and the Internet at the heart of its operation at this scale.”11
In addition to running a highly successful ocial digital campaign, BSD also
ensured that Obama had a strong presence across the range of new web 2.0 sites
that had emerged since 2004 (Baumgartner and Morris, 2010). By election day,
Obama’s following on Facebook alone numbered in the tens of millions, far
exceeding the total achieved by John McCain. Despite the increased opportuni-
ties for voter outreach and interaction that these new platforms provided, use of
them by the campaign to involve and mobilize supporters was surprisingly lim-
ited (Towner and Dulio, 2011). Given that such reticence was unlikely the result
of a lack of resources, or indeed of ambition among the digital team, speculation
centered on the operational independence of the new spaces and their immunity
from campaign oversight (Gueorguieva, 2008). e type of devolved organiza-
tion structures that these social media tools oered opened up the possibility
for another Dean- like disintegration. Certainly the low tolerance of BSD for
supporter- led initiatives were soon in evidence as Joe Rospars swooped in take
over a very popular, but unocial, pro- Obama MySpace page shortly aer the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
183
The Trendsetter 183
Chicago senator entered the race. e site, set up by Californian Joe Anthony,
was transferred over to the campaign by MySpace at Rospars’s request and with-
out the owner’s consent (Castells, 2009; Levenshus, 2010). While Rospars
later allowed those who had been disturbed by his action to vent their concerns
through the ocial website, the incident demonstrated the level of “controlled
decentralism” that occupied the heart of the Democrats’ digital strategy. Joe
Anthony was arguably a ower that Joe Trippi and Howard Dean would have
allowed and even encouraged to bloom.
Despite the fact that Obama was a major party candidate, he was not
expected to win the nomination. While it may be an exaggeration to see his vic-
tory as keeping the pluralists’ hopes for equalization alive, the online campaign
of 2008 did help to generate a surge in support for outsider candidates. In par-
ticular, two Republican hopefuls— Ron Paul, a Congressman from Texas, and
Mike Huckabee, the former governor of Arkansas— succeeded in mobilizing an
unexpectedly large cache of votes and money through their supporters’ creative
use of the internet. Huckabee’s team mined the blogosphere to create “Huck’s
Army” and pulled o a surprise early victory in Iowa. e “Paulites” pioneered
the “money bomb,” which was a simple but highly eective fundraising tech-
nique that involved seing a date of some historical or political signicance, and
associating it with a monetary target for online donations. e date was then
intensively promoted by the campaign as a deadline for donations.12 e results
were astounding. Paul need a record total of 6million dollars and 24,940 new
donors in just one day, making him the leading recipient of online money by the
end of 2007.13
Rather like Dean, however, both candidates ultimately failed to sustain their
initial momentum. Paul in particular appeared to fall prey to the same centrifugal
forces that beset his Democratic predecessor. His supporters became increas-
ingly militant and active, but in a manner that occurred “primarily out of the eld
of vision of the campaign.” eir use of disruptive tactics to bombard sites with
supportive messages for Paul and aacks on opponents produced a backlash of
negative publicity, and led to a series of high- prole blockings of “Paulites” from
popular blog sites (Stromer- Galley, 2014:114).14
Further evidence supporting the narrative of the 2008 digital campaign as
a “challengers’ market” emerged with the distinctly lackluster performance of
the two preferred major- party candidates. Despite making promising starts,
neither Hillary Clinton nor John McCain were seen as having the vision and
commitment to “hardwire” the technology into their campaign as Obama
had done. Clinton launched her campaign on YouTube, talking about her
desire to use the web to enter into a “conversation” with the American public.
McCain also saw an early boost in his fortunes with the release of his “Joe the
Plumber” video that lambasted Obama’s tax policies and quickly went viral.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
184 WHEN THE NERDS GO MARCHING IN
184
Clinton soon relapsed into “old- fashioned broadcast” mode, however, relying
on the web for top- down “image management” through the production of
slick YouTube videos while supporters’ voices were “relegated to secondary
status” (Stromer- Galley, 2014:130). McCain similarly lost momentum, fail-
ing to capitalize on the success of his video with prompts for supporters to
act or engage with the campaign (Stromer- Galley, 2014:131). His answer to
MyBO, McCain Space, never came close to approximating the former’s popu-
larity and success.
Failure to engage with phase III– style digital campaigning by other main-
stream actors below the national level was also widely in evidence. Studies of
Senate races in 2004 revealed that take- up in general had essentially plateaued
since 2000, with a quarter of contenders failing to launch a site. e decit was
particularly noticeable among third- party and Independent candidates, indi-
cating that any national level swing toward greater equality had failed to trickle
down to state and local actors (Conners, 2004). Among those candidates who
had established any type of web presence, the use of community- building and
activist- mobi lization tools was scarce. Only around one- quarter of candidate sites
featured blogs, and even fewer (16percent) linked out to Meetup.org (Conners,
2004). Other studies of the 2004 digital campaign by Druckman etal. (2007)
and Foot and Schneider (2006) that extended the analysis to include House and
gubernatorial races conrmed this picture of stasis and standardization. Most
candidates were found to have used the medium in a broadcasting manner to
“inform” voters rather than engage them. Foot and Schneider (2006) did raise
the possibility that the next election cycle might see an incorporation of some
of the more “mobilizing” features that national- level campaigns had introduced.
Subsequent analysis by Druckman etal. (2014) of the 2008 election cycle, how-
ever, quashed any such hopes. eir analysis, which focused particularly on the
use of interactive web 2.0 technologies by congressional candidates, concluded
that they were used “to a much lesser extent” than expected. Furthermore, while
the authors accepted that a certain degree of inertia might explain some of the
gaps in provision, they also found evidence that in many cases it was a conscious
choice by front- runner candidates to avoid tools that reduced their capacity for
message control.
As this wider angle lens on developments in US digital campaigns reveals,
Obama was clearly more of an outlier than the norm in 2008. at said, it is
dicult to deny the transformative impact of his campaign. Aside from the con-
trolling tendencies shown toward MySpace, there was a genuine sense that BSD
had introduced “new forms of collaboration” into election practice (Montero,
2009:135; Stromer- Galley, 2014). For Johnson (2011), Obama’s campaign pre-
sented a “new model” of online communication, one where “citizen input [was]
encouraged and fostered” (26). e level of access to the inner workings of the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
185
The Trendsetter 185
campaign and the amount of voter intelligence that was provided to ordinary
supporters through tools like MyBO was unprecedented. e technology was
instrumental in giving individuals the sense that they were on an “equal foot-
ing with employees, so they felt like part of the team” (Levenshus, 2010:328).
Furthermore, although the new tools and platforms were key to delivering this
feeling of inclusivity and involvement, the bigger shi was arguably in the mind-
set and philosophy that BSD brought to the project. Even for Dean and Trippi,
the focus had been on the technology and cherry picking those tasks that were
most easily transferred online. Websites with dynamic baseball bats were seen
as smart fundraising devices, meetups provided the venues for activist organiz-
ing, and the blog acted as a national noticeboard. e key switch that Obama
introduced— and perhaps the dening moment in the evolution of web cam-
paigning to date— was to reverse this logic and start with the campaign activi-
ties that maered most, and then understand how the internet could be used to
enhance them.
Obama’s core strengths in community organizing and activist mobilization
provided the foundation and driving ethos behind the development of MyBO
and the wider program of CIC and two- step voter contacting. is interaction
and reciprocity of real- world and virtual networks delivered a new model for
running campaigns that, according to Castells (2009), “reprogrammed” existing
communication structures. Horizontal and vertical linkages were recongured
to create an organization that was both “local and global, interactive and central-
izing at the same time” (394). us while Obama’s core campaign objectives
or philosophy remained in place, the internet brought an entirely new level of
intensity and eciency to his activity. It was, to misquote Mr. Spock, commu-
nity organizing, but just not as we know it.
Phase IV:Individual Voter Mobilization (2012)
If 2008 had marked a “step change” in the role and importance of the internet
in the conduct of a national election campaign (Turk, 2012), developments in
2012 constituted a quantum leap. e election of Obama in 2008 had clearly sig-
naled to future contenders for oce that digital technology needed to be at the
heart of their operations. However, despite the moves by Blue State Digital to
“hardwire” the internet into the inner workings of the campaign, there was still
a sense that it formed a secondary channel for the core tasks of message com-
munication and voter persuasion. According to David Ploue, Obama’s cam-
paign manager in 2008, these activities were best le to the mainstream media,
with the internet taking on the vital, but supplementary, jobs of “organizing and
fundraising.”15
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
186 WHEN THE NERDS GO MARCHING IN
186
Obama’s historic victory in 2008 brought an emphatic end to this compart-
mentalized or “analog” style of thinking about the digital campaign.16 Over
the next four years, the DNC and Obama’s re- election vehicle— Obama For
America (OFA)— invested heavily in increasing their digital arsenal and par-
ticularly in advancing its GOTV capabilities. In doing so, they built on the grow-
ing competence in data- driven campaigning that had emerged among le- wing
organizations since turn of the millennium.17 is new approach and “upskill-
ing” had received a signicant boost with appointment of Howard Dean in 2005
as DNC chairman (Issenberg, 2012). Having learned from his own experience
of running for the presidency, Dean implemented an extensive program of
reform of state party infrastructure. is included an overhaul and standardiza-
tion of their IT systems and voter records. Aparticular focus of his “50 state
policy” was the creation of a comprehensive national voter le that would be
shared by candidates, and continuously updated across electoral cycles (Kreiss
and Welch, 2015).
Dean’s work gave the Democrats and other le- leaning organizations an impor-
tant head start in entering the new era of data- intensive campaigning. Democrat-
supporting data and soware vendors such as Catalist and NGP- VAN were quick
to recognize the new market for their services, in terms of helping candidates to
access, and make full use of, the new streamlined voter les. Beyond the parties,
progressive nonprot organizations and labor unions were also eager to exploit the
new technologies and data sources, and improve the accuracy and eciency of their
GOTV methods and eld operations. ere was also a growing interest among
these groups in the work of academics and particularly that of social scientists Alan
Gerber and Don Green, who were seen as pioneering a new behavioral science of
voter mobilization.18 is combination of technological and scientic advances in
voter mobilization eorts led to the formation of the Analyst Institute in 2007 in
Washington, DC. is provided an important forum for practitioners and activists
on the le, along with academic analysts, to share and exchange the latest thinking
about how to increase the public impact of their campaigns.19
Field operations also underwent a major rethink at this time within leist
circles, as Democratic activists like Zak Exley, who had been instrumental in the
rise of MoveOn.org, started to push for beer training of sta and volunteers in
the use of new digital techniques. ese eorts culminated in the foundation of
the New Organizing Institute (NOI) in 2005. Based in Washington, DC, the
core activities of the NOI centered on running a series of “boot camps” in online
organizing that would draw in progressive political activists around the country.
Although its program focused on capacity building in the use of the new tools,
its key message was about the value of digital as means of enhancing and scaling
up “tried and tested” methods of community organizing, rather than replacing
them.20
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
187
The Trendsetter 187
It was in this rapidly developing context of commercial, scientic, and orga-
nizational innovation on the le of US politics that Obama’s 2012 re- election
campaign was hatched. Drawing on the rich pool of digital skills, resources,
and infrastructure that now surrounded the Democratic Party, his team intro-
duced an entirely new style of campaign in 2012. Computational management
practices now dominated activities in a manner hitherto unseen (Kreiss and
Welch, 2015:13). Although Obama still retained the services of a traditional
“all- purpose” campaign manager with his appointment of Jim Messina, Messina
was clearly on board and on message with the new “data science”– led approach,
declaring early in the campaign that his goal was to “measure everything.”21 is
o- repeated phrase came to encapsulate the new zeitgeist of US campaign man-
agement. Indeed, it now made lile sense to talk about the web or e- campaign as
a stand- alone entity within the wider organization.
One very visible sign of these shiing priorities was the dramatic increase
in the size of Obama’s digital team. While 2008 was seen as a record- breaking
year for sta recruitment, 2012 dwarfed these eorts, with the numbers hired
trebling to over 300 by election day. Teddy Go, a BSD employee and the direc-
tor of Obama’s 2008 online campaign in the baleground states, was promoted
to the top spot.22 As well as presiding over a much larger team of operatives,
Go also had to recruit and manage an increasingly diversied and specialized
workforce, which now included up to 50 data analysts and a similar number of
soware engineers.23 For most onlookers, the new arrivals elicited a mixture of
curiosity and amusement. Media headlines talked about the “nerds” that had
gone “marching in” to Obama’s headquarters.24 Some seasoned observers were
less sanguine, however, likening their entry into the political arena to an “alien”
invasion. Peggy Noonan, a former speechwriter for Ronald Reagan, was par-
ticularly critical of the new appointments. Having read an advertisement for
a vacancy in Obama’s analytics team, she used her column in e Wall Street
Journal to declare that politics was now done by “martians.”25
e signicant expansion in digital personnel was accompanied by their
division into specialist teams. At the center was team Digital, managed by Joe
Rospars (now Obama’s chief digital strategist) who provided continuity from
the previous election and delivered the outward facing parts of the campaign
(i.e., email programs, web design, and online fundraising). In addition, two new
smaller support teams were established— team Analytics and team Tech. While
the former was led by Dan Wagner, who had prior campaign experience; the
laer was championed by Harper Reed, who was a newcomer to the political
scene. Reed’s background was in the nance industry, where he had focused on
building soware to enhance commercial collaboration. His appointment as
chief technology ocer (CTO) reshaped the structure of the campaign, in that
his sub- team acted more like an independent “internet startup” within the wider
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
188 WHEN THE NERDS GO MARCHING IN
188
OFA organization, delivering a range of IT and data services that were “unique
in the history of presidential politics.”26
With the core teams in place by May 2011, the work now began in earnest
to design a “world class” digital infrastructure that would bring a new scientic
approach to voter mobilization.27 Avital part of delivering that new model of
electoral engineering involved reecting on the experiences of 2008 and, in
particular, learning from the mistakes that BSD had made. Aention quickly
seled on the Houdini project, which was seen as a major GOTV investment
that had not delivered. e main goal of Houdini had been to “revolutionize”
the eciency of the “ground game” by allowing for the real- time tracking of
voter turnout by local poll watchers. It had suered a spectacular fail on election
day, however, as the hotline set up to transfer the local data collapsed under the
weight of demand. Volunteers were le to text or call in voter codes to the local
oce, which were then entered manually.28 Reviewing the problems, team Tech
argued that a fundamental upgrade and reform of internal IT management sys-
tems was required to prevent the recurrence of these problems. As a result, most
data services were migrated to Amazon Cloud services, and then dispersed and
replicated in servers across the country. Regular checks on the system’s robust-
ness were carried out during the course of the campaign, through a series of
simulated network outages and surges in demand.29
A second major area for improvement identied by Go and his team was the
suboptimal level of soware and data integration that had existed between teams
and applications in 2008.30 For Clint Ecker, a senior soware engineer in OFA,
[o] ne of the biggest problems in the last campaign was that you had all
these people who are out in the eld, who are volunteering, who start
building their own versions of these rogue tools to do the same thing
over and over again. Every eld oce assembled its own patchwork of
tools using spreadsheets or a hacked Web application to track opera-
tions. ey communicated over Google groups or simple e- mail lists.
is meant it was hard to keep everyone on the same page.31
In response to these problems, team Tech created Narwhal. Siing at the apex
of the cloud, this “whale” of a platform was designed to host and integrate the
campaign’s collection of soware and data, allowing multiple users to download,
upload, and modify resources on an ongoing basis and in “real time.” is con-
stant updating of records and integration of data sources provided the ideal envi-
ronment for application of the new modeling tools developed by team Analytics
to identify priority voters and forecast local outcomes.32
As well as identifying and dealing with some of the problems encountered in
2008, team Digital was also eager to enhance and build on the initiatives that had
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
189
The Trendsetter 189
worked well for them in 2008. In particular, the indirect or “two- step ow ” model
of persuasion or CIC that MyBO had engineered was seen as ripe for expansion.
e end result was Dashboard, a new online platform that was designed to maxi-
mize the eciency of the campaign’s volunteer management program and its
canvassing eorts. While it retained some of the community- building ethos of
MyBO, the main aim of Dashboard was to sign up and move activists into GOTV
operations as quickly as possible. New registrants were allocated to a local neigh-
borhood team, typically within 72 hours aer rst contact, and then dispatched
into the eld. Armed with mobile devices, they then entered the results of their
contact eorts and any vote commitments they received into their own personal
Dashboards. e connection into Narwhal ensured that these data were used
to update into existing voter records, ready for interrogation by team Analytics.
In addition to upgrading and streamlining the CIC platform, the Digital team
took steps to improve the accuracy and precision of the mobilizing activities it
generated. Two applications in particular helped with this. e rst, “Call Tool,”
focused on ne- tuning the phone canvassing eorts of volunteers to maximize
their contacts with undecided voters in baleground states. e second, a tar-
geted sharing app developed for Facebook, was aimed at extending and sharp-
ening the peer- to- peer exchange of campaign content among supporters. Once
downloaded, it allowed the campaign to identify persuadable individuals or
“priority voters” among their supporters’ online networks. e supporter would
then receive specic prompts designed by the campaign that they could use to
target selected friends to vote for Obama. According to the campaign, this inno-
vation resulted in 5million additional campaign contacts among the half a mil-
lion supporters who downloaded the app. Of those 5million contacts, one- h
reportedly acted on the prompts they received.33
Of course, the advances made in 2012 were extended beyond these initia-
tives. Team Digital was credited with creating the “most sophisticated email fun-
draising program ever.”34 According to FEC reports, Obama’s war chest swelled
from $500million in 2008 to $690million in 2012. Awhopping $504million of
that total was reportedly generated through digital channels.35 Asignicant por-
tion of team Digital’s success was aributed to their launch of “Quick Donate,”
which, as its name suggests, brought the ease of one- click purchasing to cam-
paign donation. According to one post- election report, Quick Donate alone
raised $115million, $75million of which, it was claimed, would otherwise not
have been raised.36 Other major breakthroughs included the production of a
new “micro- listening” soware— Dreamcatcher— which was designed to decode
social media chaer and identify potential Obama supporters.37
On the Republican side, there were eorts to launch similar programs.
MyMi was Romney’s version of Dashboard, and Orca, a GOTV program, was
presented as a superior version of Narwhal. It was clear from the outset, however,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
190 WHEN THE NERDS GO MARCHING IN
190
that Romney’s campaign faced an uphill struggle to compete. Not only was he
coming from behind in terms of building from scratch the digital infrastructure
that Obama had had years to create, those in charge of it had to ght hard to
siphon resources away from other areas of the campaign. Despite being publicly
upbeat about their prospects, Romney’s digital media strategist, Zac Moat, did
not enjoy the central position that Teddy Go held within the Obama organiza-
tion (Stromer- Galley, 2014:167). In terms of the new areas of investment by
the Democrats— tech infrastructure and voter analytics— Romney simply did
not have the nancial muscle or the time to match the pace and scale of Obama’s
commitments. Instead, much of his tech R&D was outsourced to commercial
providers.
Making a virtue out of necessity, Moat openly derided Obama’s investment
in voter modeling as an ego- driven exercise in “vanity metrics” (Stromer- Galley,
2014:167). e limitations of the Republicans’ digital operations were laid bare
on election day, however, as its “star” performer, Orca, collapsed in spectacular
style. Far from delivering the killer blow to Narwhal, the soware proved entirely
unt for its purpose, having been rushed out of the lab just as the polls opened,
without any beta testing or training for users.38
Just as the failure of Houdini had proved to be a touchstone for the Democrats
in reforming their digital strategy, the crash of Orca served as a sy mbol of just how
far behind the Republicans had fallen in the digital campaign stakes. Although
concerns had been voiced since their loss to Obama in 2008,39 the election of
2012, according to one insider, provided a “rude awakening” to the acute de-
cit that the party now faced.40 is perception was conrmed by the RNC’s
post- election report, the “Growth and Opportunity Project,” which sought to
diagnose and improve on the failings of 2012. According to the authors, a key
element of the Democrats’ success lay in the “clear edge” they now held with
regard to technological innovation, and particularly the extent to which they
were able to link this with their eld operations. By “marrying grassroots poli-
tics with technology and analytics, they successfully contacted, persuaded and
turned out their margin of victory.” Going further, the report argued that while
some of this superiority stemmed from the Democrats’ greater investment and
resources in the technology, an equally critical factor was the more open “culture
of data and learning” that had developed across the party during the past decade.
is collaborative and collectivist ethos, they concluded, was something that the
Republican Party had found more dicult to cultivate, given its more individu-
alistic outlook and hierarchical structure.41
Such a frank admission of the deep problems the Republicans were now
facing in running an eective digital campaign was signicant on a number of
grounds. First, it revealed how central the technology had become for parties in
the bale to mobilize votes. Second, in the context of this book, it adds weight
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
191
The Trendsetter 191
to the ndings from earlier chapters about the critical role of le- wing parties in
pushing the cycle of change in digital campaigning, particularly with the regard
to entry into the third community- building phase. e extent of these dier-
ences between the two main parties, and how they compare to the performances
of parties in the United Kingdom, Australia, and France is shown more clearly
in Table 7.1.
Table 7.1 reports the results from the application of the CIC index to the
two candidates’ online organizing platforms in 2012— Dashboard and MyMi.42
is exercise reveals that on all counts, the Democrats’ site out- performed the
Republicans’. is is particularly the case with regard to voter mobilization
actions and community building. On the laer front, Romney provided mini-
mal space for supporters to meet and greet one another. Obama, by contrast,
remained somewhat more in tune with the spirit of MyBo and its social network-
ing origins.
Despite its stronger performance, the index also reveals that Dashboard did
not fully exploit all the available opportunities for supporter mobilization.
Taking the site’s performance as a whole, we nd that only around 60percent of
the potential CIC content (as measured by the index) was provided to the end-
user. If we compare these results with those produced by application of the index
in other countries’ (see Tables4.1, 5.1, and 6.1), we can see that US national
campaigns were not necessarily the most advanced in using digital platforms to
mobilize their base. In France, for example, in the same year, Tous Hollande, the
online activist hub of French socialist candidate Francois Hollande, achieved
a similar overall score to Dashboard (see Table6.1). Furthermore, if one looks
more closely at the specic competencies of the two sites, it appears that Tous
Hollande actually performed beer in the areas of GOTV and message distribu-
tion than Dashboard. It was only because of the laer’s stronger emphasis on
community building that it came out ahead.
Of course, it is possible and indeed quite likely that had the same exercise
been carried out in 2008, we might have reached a dierent conclusion, at least
on the comparative performance of Obama’s site. MyBO was arguably the high
watermark for CIC in the United States and indeed globally. Dashboard’s more
spartan quality may have been due to a deliberate paring back of the site’s inter-
active qualities, in a bid to focus supporters’ energy on carrying out more exter-
nally relevant voter mobilization tasks.
Closer inspection of other aspects of the Obama’s 2012 digital operation sug-
gests a broader shi was occurring toward phase IV. In particular, digital sta
now formed the nerve center of operations, designing and applying a range of
specialized tools to seek out and mobilize new pockets of electoral support. In
power logic terms, normalization appears to have risen to the fore. e equal-
izing momentum that Howard Dean had briey injected into proceedings in
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
192 WHEN THE NERDS GO MARCHING IN
192
Table7.1 US Presidential Candidates Citizen- Initiated Campaigning
Scores, 2012
Dashboard MyMi
Community Building
Prole
Photo
Biography
Why joined
Set up/ join groups
Set up blog
Set up Wiki
Email/ msg system
Externally promote prole
Subtotal (additive 0– 8)
√
√
√
√
—
√
√
√
7
√
√
—
—
—
—
—
√
3
Resource Generation
Personal fundraising
Promote membership
Sign up as local organizer
Sign up as candidate
Organize/ add event
Vote leaders to aend events
Subtotal (additive 0– 4)
√
na
√
na
√
—
3
√
na
√
na
—
—
2
Voter Mobilization
GOTV oine
Access phonebank
Sign up for f2f canvassing
Sign up to discuss with f&f
Leaets download
Externally promote event
GOTV online
Send email
Post to Facebook
Post to Twier
GOTV phone app
Email forward to editor
Start/ promote e- petition
Subtotal (additive 0– 11)
√
√
—
—
√
—
√
—
√
—
—
5
√
—
—
—
—
—
—
—
√
—
—
2
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
193
The Trendsetter 193
2004, and that Obama had sustained to a degree in 2008, was now fading fast.
At the systemic level, competition had reached perhaps its lowest level. As the
Republican’s post- mortem on their 2012 electoral performance had concluded,
digital campaigning (and indeed campaigning more generally) was now domi-
nated by the Democrats to a degree hitherto unseen in US politics.
Internally, the level of top- down management and oversight under the new
elites intensied. e centralized system of soware and app development
under Narwhal was highly ecient but, according to some insiders, it had
a stiing eect on the creativity and entrepreneurial spirit that had given the
Democrats their edge over the Republicans in the rst place. For some campaign
analysts, the dominance of Narwhal had “killed an important ecosystem for bub-
bling up innovation.”43 e increasing convergence of online and oine strategy
also meant that the key decisions about where, who and how to mobilize were
le to a small inner team of data scientists, or “the cave,” as a number of media
Dashboard MyMi
Message Production
Message creation
Policy email fwd/ customize
Poster/ leaet create/ customize
Policy input/ feedback
Message distribution
Web banners/ ads d- load
Posters/ leaets d- load
Email/ share policy docs
News feed to website
Share blog posts externally
Link to SNS prole
Link to Twier account
Import email contacts
Subtotal (additive 0– 11)
—
—
—
√
—
—
—
—
√
√
√
4
—
—
—
√
—
—
—
—
—
—
√
2
Overall Score (0– 34)*
Standardized Score (0– 100)
19
61
9
32
Standardized scores are calculated by transforming each sub- index into a 0– 100 range and then aver-
aging the scores. See Appendix 4.1 for details of variable denitions.
* e maximum raw score on the CIC index was 34 for US sites (2 points lower than for the
United Kingdom and Australia and 1 point lower than for France). is was due to dropping both
the “sign up as a candidate” and “promote party membership” variables from the resource generation
sub- index as they were not applicable for the US presidential sites.
Table7.1 Continued
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
194 WHEN THE NERDS GO MARCHING IN
194
observers had christened them.44 While there was still some scope for local activ-
ists to exert tactical autonomy through tools like Dashboard, their empowerment
was born more of expediency and a desire to maximize the mobilizing reach of
the campaign, rather than to promote internal democracy. Indeed, with the help
of Narwhal, volunteers’ performance could now be measured, monitored, and
managed in real time, allowing headquarters to control and allocate resources in
an even more precise and centralized manner than ever before.
The View fromBelow:Mapping Digital Campaign
Cycles inthe USElectorate
e second half of this chapter re- examines these trends in digital campaign-
ing from the voters’ perspective. To what extent are the changes and dierences
identied in US parties’ use of digital technology reected within the electorate
and among partisans? To what extent do we see a growing emphasis on more
active modes of consumption, particularly during the Dean campaign and the
rst Obama campaign? Has this been concentrated primarily in Democratic sup-
porters, or have Republican supporters kept more in step with their rivals at the
mass level? Did 2012 see a signicant growth in the numbers of voters receiving
digital contact from the parties, or have campaigns simply become more expert
at reaching those all- important swing voters? We approach the task, as in previ-
ous chapters, by mapping the levels and changes in our three main modes of
voter engagement— read, redistribute, and receive— among US citizens with
a range of survey data. As well as tracking changes at the demand side of US
politics, we also compare the results to ndings from the previous chapters to
compare levels and particularly paerns of change in consumption over time.
READING, REDISTRIBUTING, AND RECEIVING THEWEB
CAMPAIGN OVERTIME
To examine the demand for web campaigning among the US electorate over
time, we make use of three data sets collected during the 2004, 2008, and 2012
presidential elections. e 2004 and 2008 data sets were produced by the Pew
Research Center as part of their “Internet and American Life” time series (now
the Internet, Science, and Tech program). For 2012, we make use of the ANES
pre/ post- election survey that included the CSES module of items on online
contact reported in Chapter2. e switch to ANES data in 2012 provides con-
sistency in the gures used in the earlier comparative analysis, but was also neces-
sitated by a lack of comparable Pew data for that year.45 All of the surveys were
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
195
The Trendsetter 195
conducted post- election, and drew respondents from oine sampling frames.
is made it possible to compare levels of e- campaign activity among the online
population to the population as a whole. Survey mode diered between Pew and
ANES in that the former relied on telephone random digit dialing (RDD), while
the laer used a combination of web and face- to- face methods (see Appendix
7.1 for full details).
Given that the Pew studies were designed primarily for measuring citizens’
use of online technologies during the campaign, they contained a much wider
range of items to measure the main categories of engagement than the ANES.
ese dierences inevitably introduce some measurement error into the analy-
sis, which means that, as in earlier chapters, we focus more on comparing change
in the relative prominence of our three modes of engagement over time, rather
than in their absolute levels. is is particularly the case between 2008 to 2012,
when the switch from Pew to ANES data occurred. Table 7.2 reports the top- line
ndings in terms of the frequencies of the three modes of voter engagement in
the digital campaign— read, redistribute, and receive (and their component or
sub- measure)— across the three presidential elections from 2004 to 2012.
As one would expect, the appetite for digital politics among the electorate
has grown over time. What is perhaps more interesting, however, is the extent to
which it compares with the levels of aention and growth seen in the countries
examined in earlier chapters. Essentially, the US public exhibits a much stronger
initial appetite for online campaign material than its counterparts in the United
Kingdom, Australia, and France, and has largely maintained that lead over time.
Starting in 2004, we can see that up to one- third of the population had “read”
something about the campaign online, which is a level that is typically achieved
in other countries at least one full election cycle later. e evidence from France
and the United Kingdom (see Tables6.2 and 4.2), for example, shows that these
countries reached similar levels among their publics by as late as 2010 and 2012,
respectively. Within this picture, it is also notable that interest in the ocial cam-
paign sites is much higher among US voters at an early stage, compared with
levels seen elsewhere. e ndings from Australia in Chapter5 (Table5.2), for
example, show that even by 2013, just over one in 10 of the population reported
visiting the site of a party or candidate. is was a level of interest that had been
reached among the US population around a decade earlier.
e gures for the two other modes of e- campaign engagement— “redis-
tribute” and “receive”— show an even more pronounced gap in activity levels
among US voters and those in the three other nations studied so far. As with
“read,” the disparity is sustained over time, with “receive” rates being at least
double those observed in comparable elections elsewhere. While it is tempting
to interpret these dierences as evidence of an American exceptionalism in web
campaigning, the ndings from the large N analysis of Chapter3 would appear
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
196
Table7.2 Voter Engagement inUS Presidential Digital Campaigns, 2004– 2012
Mode of Engagement
Year Internet use Voter as Audience— READ Voter as Activist— REDISTRIBUTE Voter as Target— RECEIVE
Online
news
Candidate sites Sign- up/ Download Share/ Exchange Party
(Direct)
F&F (Indirect)
2004
N=2,200
60% 34.0
(57.2)
12.9%
(21.8)
6.6
(11.1)
12.2
(14.2)
—
—
—
—
READ 35.4 (59.5) REDISTRIBUTE 3.2 (5.3) RECEIVE 13.9 (25.9)—
2008
N=2,254
73% 42.9
(58.7)
25.3
(34.6)
11.6
(16.0)
24.1
(33.0)
26.2
(35.8)
35.6
(48.8)
READ 46.4 (63.5) REDISTRIBUTE 7.2 (9.9) RECEIVE 43.1 (58.9)
2012
N=5622
89%a49.1
(55.1)
12.9
(14.5)
9.9
(11.1)
18.6
(20.8)
16.2
(18.2)
14.3
(16.0)
READ 51.2 (57.8) REDISTRIBUTE 5.3 (6.2) RECEIVE 25.7 (28.8)b
a e 2012 ANES full sample was recruited via random probability methods but was conducted using two modes:face to face (N=2,056) and the web (N=3,860). e face
to face (f2f) mode included the internet use question and was used to calculate the gure of 89.2% internet use in the population as a whole (N=1,762, “weight_ f2f ” applied). e
estimates of the modes of engagement for internet users are calculated from the combined sample of internet users in the f2f survey (1,762) and the full web survey sample (3860),
i.e., total nal N of 5,622“weight_ full” was applied. e estimates of the population or full sample engaging in each mode was derived by multiplying the % reported for internet
user by .892, which re- estimated them on a 0– 100 scale.
b Figures for “receive” mode dier from those reported for the United States in Chapter3, Table3.1. is is due to the fact that we are using a weighted version of the full
ANES sample (i.e., web and f2f respondents) for our estimates in this chapter, while the gures reported for the United States in Chapter3 are from the CSES module 4 data
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
197
set (with “original demographic” weight applied). e CSES US sample is smaller than the full ANES version, as it only included responses to the f2f survey (N=2,056).
While in the case of online direct or party contact the gures are very similar (17% in Table3.1 and 16.2% here), we do see clearer dierence in the gures reported for receiv-
ing indirect or friends and family contact and total online contact received. In the CSES and Table3.1 these gures were rounded to 23% and 34%, respectively; in the ANES
sample (i.e., f2f and web samples combined) the equivalent gures are 14% and 26%. While this still puts the United States in tier one of Table3.2 and among the top- most
nations for digital mobilization, the lower estimate, particularly for indirect (friends and family) online contact, warranted closer inspection. Breaking down the ANES results
by survey mode, we can see the gures from the f2f sample were comparable to those reported in the CSES (which we would expect), while the web sample estimates dif-
fer. Specically, the gures for the ANES f2f sample for direct, indirect, and total receive are 16%, 23%, and 32%, which correspond closely to the CSES gures reported in
Table3.1, and are dierences that could plausibly be accounted for by weighting dierences. e same estimates from the web sample are 18%, 12% ,and 26%. Given that the
web sample was twice as large as the f2f, when the two were combined the former will have a stronger eect on the overall estimates, thereby pulling them downward. We do
not know precisely why the web sample gures for indirect receive mode were lower than those produced in the f2f; however, the two samples did dier on key demographics
prior to weighting, and thus application of a combined weight may have had a dierential impact on estimates of other variables.
Figures are % and reported for overall sample and for internet users in parentheses.
Sources:2004 and 2008— Pew Internet and American Life Project data sets; 2012— ANES 2012 Time Series Study (2 wave pre- / post- election). Reported Ns are unweighted.
Estimates of % internet use and modes of engagement are based on weighted data. Survey weights supplied. See Appendix 7.1 for further details of surveys and variable
denitions.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
198 WHEN THE NERDS GO MARCHING IN
198
to challenge that narrative, or at least to moderate it. Specically, the results in
Table3.1 showed that there were similar or even higher levels of digital contact
or “receive” mode reported by voters in several other countries, beyond the case
studies examined in our earlier chapters. Whether these higher levels of engage-
ment extend back in time and across all modes for these nations is not possible
to determine from the CSES data set given the limited number of questions
in the module overall and its cross- sectional design. e fact that comparable
gures were obtained on at least one measure of activity for the United States,
however, would suggest that while the US population is clearly the most political
active citizenry of those studied in depth in this book, they are not necessarily
the most active worldwide.
As well as allowing us to compare the United States to our other countries,
Table 7.2 also permits a closer examination of changes in US voters’ engage-
ment with digital campaigns over time. As was the case for the other coun-
tries examined in earlier chapters, “read” is the dominant mode of engagement
across all years and increases over time. By 2012, just over half of the population
(51.2percent) reported accessing some type of election- related content, com-
pared with just over a third (35.4percent) in 2004. Within this overall paern
of growth, however, we can also discern that 2008 was actually the peak year for
most activities. Visits to ocial campaign sites in particular fell quite noticeably
in 2012 from their 2008 high point. Furthermore, although levels of overall con-
sumption of online campaign news grew among the population as a whole over
the eight- year time period, by 2012 there was a very slight dip in interest levels if
one controls for access and looks only at internet users. According to the table,
59.5percent of internet users “read” something about the campaign in 2004,
versus 57.8percent who did so in 2012.
It is important to note that, as with the ndings from the French case, the
change in levels of citizen engagement between elections, and here particu-
larly the drop between 2008 to 2012, may have a methodological explanation.
Essentially, this corresponds with the switch from Pew to ANES data sets. e
dierences in sample recruitment and mode between the two studies may have
led to an over and/ or under- representation of more politically active respon-
dents. Pew relies primarily on telephone RDD, while the ANES is conducted
using a combination of face- to- face and internet methods. Furthermore, there
are dierences in the measures used by each study, with some of the ANES items
being more narrowly worded than their Pew equivalents. e item used to mea-
sure sharing in the 2012 ANES, for example— which forms a critical component
of “redistribute”— focused specically on whether a respondent had sent politi-
cal messages via social media. e 2004 and 2008 Pew questions asked more
generally about whether respondents had shared various types of campaign
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
199
The Trendsetter 199
content online (see Appendix 7.1 for more details on mode and the specic
wording of questions across the surveys).
Even if we accept that at least part of the explanation for the trends observed
in Table 7.2 may be sample- related, the idea that 2008 formed a blip or high
point in the levels of public interest and active involvement in the digital cam-
paign clearly corresponds with the evidence from the earlier supply- side analy-
sis. e election of 2008 was Obama’s rst campaign for the presidency, the year
of MyBO and the slogan of “Yes We Can.” It also generated the 13million– strong
email list that OFA had later relied on as the basis for Narwhal and the 2012
digital campaign. e nding that the voters’ desire to engage in CIC activities
had waned somewhat by 2012 would thus appear entirely plausible. Indeed, it
conrms some of the earlier speculation that his digital team had switched to
designing tools like Dashboard that were more focused on GOTV and resource-
generating activities, rather than promoting more participatory and community-
boosting practices.
While declines in read and redistribute modes by 2012 appear to correlate with
changes in elite behavior, the dramatic drop in receive in 2012 is, at rst glance,
more of a puzzle. Given the massive resources that the Democrats dedicated to
improving their micro- targeting precision that year, and the increased popular-
ity of social networks such as Facebook and Twier that actively encourage con-
tent sharing, one might have expected the rates of formal and informal contact
to have risen in this election. Although methodological dierences between the
survey instruments and sampling methods cannot be discounted in explaining
the disparity, the idea that a natural decline occurred in the frequency of digi-
tal contact between the two elections is not entirely implausible. Certainly, the
lower levels of enthusiasm and momentum that surrounded Obama’s candidacy
in 2012 may have reduced the level of peer- to- peer mobilization. Furthermore,
in terms of direct contact, the overall amount of party contacting does not neces-
sarily provide an indication of its eectiveness. Indeed, if Obama’s digital teams
had signicantly improved the quality and precision of their eorts by 2012,
then a decrease in levels might follow due to their increased accuracy in nding
those voters that really maered to the outcome. We will return to this point
when we look the prole of those contacted online in 2012.
Having examined the paerns of engagement with digital campaigns among
the US electorate over time, we now break this broader picture down to look at
the distribution across partisan and Independent voters. Table 7.3 contains the
key ndings on this question. e table reports the growth in internet use among
Republican and Democrat identiers as well as for Independents, and compares
their rates of reading, redistributing , and receiving digital campaign content over
the three elections. e rst column shows, as we might expect, that internet
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
200
Table7.3 Party Supporters inUS Presidential Digital Campaigns, 2007– 2012
(Internet Users Only)
Mode of Engagement
Election Year and
Party
Internet
Access
READ
REDISTRIBUTE
RECEIVE
Direct
(Party)
Indirect
(Friends &
Family)
2004
Republican
(N=678)
66.1 63.2 4.8 22.8
Democrat
(N=731)
56.2 62.4 8.0 39.4
Independent
(N=561)
63.5 59.5 4.2 19.4
2008
Republican
(N=599)
83.8 66.6 7.1 36.6 55.0
Democrat
(N=780)
70.2 66.7 16.5 40.0 52.7
Independent
(N=634)
73.2 62.5 7.2 34.7 47.0
2012
Republican
(N=1,389)
92.1* 57.9 6.9 19.9 17.3
Democrat
(N=2,363)
86.6 58.0 7.8 19.9 14.1
Independent
(N=1,845)
89.3 58.8 3.7 15.8 16.5
Estimates of % internet access and mode of engagement survey are based on weighted data. Survey
weights supplied. See Appendix 7.1 for further details of surveys and variable denitions.
* e gures for % internet access by party in 2012 are based on face to face (f2f) sample which
included the internet use question (weight_ f applied). All other estimates and N reported for 2012
ANES are based on the full sample of internet users in the f2f and web survey respondents (N=5,622)
Sources: 2004 and 2008— Pew Internet and American Life Project data sets; 2012— ANES 2012
Time Series Study (2 wave pre- / post- election). Reported Ns are unweighted and refer to the full
sample. Internet access gures are the % of that total that reported being online. e gures for Read,
Redistribute, and Receive are % of the online party identiers that engaged in these activities.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
201
The Trendsetter 201
access has increased over time for all groups. e rates of growth are somewhat
lower for Democrat supporters than for the other two groups. e dierence is
not particularly surprising given the generally lower socioeconomic status of the
party’s support base. Once we control for levels of internet use, a more balanced
picture of partisan usage emerges. is is particularly evident with regard to
reading about the campaign. Here we see that the gap between Democrats and
Republicans closes almost entirely, with both sets of partisans displaying similar
levels of interest in accessing news and ocial information about the candidates.
As we move to examine the more active and targeted modes of voter engage-
ment among party supporters, however, a more interesting paern of dierence
starts to re- emerge.
Overall, the redistribution of candidate web content is, not surprisingly,
more common among partisans than Independents, and when compared to the
electorate as a whole (as seen in Table7.2). Furthermore, until 2012 it was the
Democrats that held the advantage in this area of supporter activity. e election
of 2008 saw a signicant boost in the numbers of Democrat identiers engaging
in this new style of citizen- initiated style of campaigning. is increase matches
the narrative presented earlier in the chapter about the stronger promotion of
phase III digital campaigning by the le, and particularly under Obama’s lead-
ership. e subsequent sharp fall in the level of redistributive activity among
Democrats in 2012 conrms the idea that 2008 constituted something of
an “all time high” in terms of their capacity to actively engage their supporter
base online. While the drop, as noted earlier, may result from the changes in
survey method in 2012, the fact that the same decline does not happen among
Republican supporters suggests there was a genuine rise and fall of enthusiasm
for this type of activity among le- wing partisans.
Looking nally at the rates of receiving online contact among partisans
and Independents, unlike the ndings for redistribution, the peak in 2008 and
subsequent decline in 2012 are universal across both sets of party supporters.
What is particularly notable here, however, and again is somewhat at odds with
our supply- side narrative, is the relative parity across the two parties in terms
of direct contact received by voters. Despite the Democrats’ widely accepted
greater repower in this area, it does not seem that this translated into their sup-
porters actually receiving a signicantly higher level of contact and communi-
cation from the campaign during the election. Furthermore, in both 2008 and
2012, the Republicans actually appear to have a slight competitive advantage
with regard to informal contacting. Republican supporters report the highest
“hit rate” in terms of receiving mobilization cues from their online social net-
works. Such ndings clearly raise further questions about the presumed superi-
ority of the Democrats’ digital mobilization eorts, particularly in 2012. While
it is dicult to account for the unexpectedly active strong peer- to- peer online
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
202 WHEN THE NERDS GO MARCHING IN
202
contacting by Republicans in these elections, it is possible that it stemmed from
a realization among the grassroots of the weakness of their ocial party machin-
ery. e much- publicized gap between McCain and then Romney and Obama
in terms of the strength of the laer’s digital operations may have had the eect
of stimulating their supporters to make the most of the online tools available to
them, and to go into digital bale on their candidates’ behalf.
e nal section of this chapter examines the question of the impact of digital
mobilization in more depth. In particular, we look at who in the wider elector-
ate received direct and indirect online contact. Is there any evidence that these
newer modes of contact are reaching a dierent and under- mobilized audience?
To what extent might they be aecting vote choice? Table 7.4 helps to address
these questions by comparing the socio- demographic prole of respondents
who reported receiving dierent types of online contact (direct and indirect) in
presidential elections from 2004 to 2012 with those who received more tradi-
tional types of “oine” contact. e table also reports the vote choice for each of
the dierent modes of contact, compared with the population as a whole.
e gures are interesting in that they reveal a picture of both change and
continuity in the type of citizens who are contacted online during US presiden-
tial elections. In particular, there does not appear to be a stronger gender bias in
rates of online contact, even in the earlier days of presidential campaigns. e
age prole of those contacted, however, does show a clear bias toward younger
people. While this skew does reduce over time for direct contact from the par-
ties, it actually becomes more prominent for indirect forms of online contact.
By 2012, older voters (i.e., those over 50)are more likely to be contacted online
by parties than those under 50. is switch is quite marked over only three elec-
tions and indicates that whether by intent or accident, parties were improving
their ability to reach older and more “reliable” voters with their online messag-
ing. By contrast, rates of indirect online persuasion, which are largely out of their
control, are noticeably more concentrated among younger voters.
e association of higher educational aainment and income with all types
of digital contact is very strong and present across all years. College- educated
individuals are much more likely to receive a digital message from a political
party or from someone in their social network, compared to those who failed to
graduate high school. Finally, in racial terms, we do see evidence of a bias in favor
of white voters in earlier years. By 2012, this has reduced somewhat, particularly
among black voters, where we see contact levels showing a closer parity to their
numbers within the population as a whole.
When we turn to examine the voting prole of those experiencing online con-
tact, an interesting paern emerges. As the nal two rows of the table reveal, until
2012, voters who experienced digital contact of any type (i.e., either from parties
or their social networks) were more likely to report voting Democrat than the
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
203
Continued
Table7.4 Socio- Demographic Correlates for“Receive” Mode overTime
2004 2008 2012
Online
(total)
Oine
(total)
Total
Sample
Online
Direct
Online
Indirect
Oine
Direct
(Mail)
Total
Sample
Online
Direct
Online
Indirect
Oine
Direct
Total
Sample
Sex
Male 49 48 48 53 47.5 48 49 49 45.5 47 48
Female 51 52 52 47 52.5 52 51 51 54.5 53 52
Age
17– 29 23 16 21 22 27 18 20 11 24 9 21
30– 49 43 39 40 47 43 36 37 29 37 29 32
50– 64 25 26 23 24 24 27 26 35 24 34 28
65+ 9 19 17 8 7 18 17 24 14 29 19
Education
Less than HS 3 11 14 3 3 9 13 6 5 8 10
HS graduate 20 33 36 23 24 35 36 21 21 26 30
Some college 30 25 24 29 29 24 23 29 32 29 30
College 47 31 26 45 43 31 37 43 42 36 29
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
204
Table7.4 Continued
2004 2008 2012
Online
(total)
Oine
(total)
Total
Sample
Online
Direct
Online
Indirect
Oine
Direct
(Mail)
Total
Sample
Online
Direct
Online
Indirect
Oine
Direct
Total
Sample
Income ($)
<20,000 11 17 21 7 8 16 20 12 14 16 21
20,000– 49,999 31 39 39 26 27 32 33 23 26 27 29
50,000– 74,999 19 18 17 19 23 19 17 21 20 22 19
75,000– 100,000 19 13 13 21 18 14 13 15 15 15 12
>100,000 21 13 11 27 24 19 17 28 26 22 19
Race
White 81 78 73 75 74 75 71 75 77 77 71
Black 5 9 11 9 10 10 11 13 9 12 12
Hispanic 6 8 10 11 12 10 12 7 9 6 11
Other 8 5 5 5 5 5 6 5 5 5 6
Vote (President)
Republican 48 52 52 38 43 43 42 46 52 50 46
Democrat 52 48 48 62 57 57 58 54 48 50 54
Figures should be read as column % within each demographic / political variable. ey show the proportion of individuals within each of the groups that reported online and
oine contact by parties. “Oine Direct” for 2004 and 2012 includes mail, phone, and face to face; for 2008 it includes only mail. e “All” column reports characteristics within the
sample as a whole. Vote is calculated based on votes for Democrat and Republican candidates, excluding independent/ other/ non- voters. Cells may exceed 100% due to rounding.
Sources:see Table7.2 and Appendix 7.1 for f ull details of surveys and variable denitions. Survey weights applied.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
205
The Trendsetter 205
population as a whole. is was particularly the case in 2008, when 62percent
of those who were contacted online by a party, and who cast a vote for one of the
two main candidates, voted for Obama. is gure was about 4percent higher
than among the electorate as whole (if we restrict it to those who voted either
for the Democratic or Republican ticket). By contrast, just 38percent of those
who had been contacted directly by a party online and then went on to vote for
a major party candidate supported Romney. is suggests that the Democrats’
digital campaign was particularly eective in 2008. Further evidence to support
this view can be seen if we look at the level of support that Obama received in
2008 among those who had experienced non- digital or oine types of contact.
Of those reporting contact from parties by mail, 57percent said they supported
Obama, a proportion that was actually slightly below the level of support that
the Democrats received within the electorate as a whole.
In 2012, the situation had clearly changed, however, with the Democrats’
advantage among voters contacted online eectively disappearing. Obama’s sup-
port levels among those who had received digital messages from a party were
no dierent than among the population as a whole. Of course, it is possible that
without the new forms of contacting, Obama’s vote would have been signicantly
lower, that is, digital messaging served to shore up and reinforce existing levels
of support rather than mobilizing new voters. e situation appeared to be simi-
lar for the Republicans in 2012 in that they did not see any increase in support
among those voters who received some type of ocial contact from parties that
year. For indirect online contacting, however, the story is dierent. Essentially,
among those voters who reported receiving mobilizing messages from their
friends and family networks, there was a distinct preference for Romney over
Obama, compared with the electorate as a whole. Overall, 52percent of those
who had received some kind of online prompt from within their social networks
and who cast a vote for one of the two main parties supported Romney. is was
6percent more than the 46percent of voters who supported the Republican
ticket over the Democratic alternative in the wider electorate.
e ndings are interesting on several levels. First, they show that despite the
heavy investments made by the Democrats in 2008, and particularly in 2012,
they are not necessarily the big winners when it comes to partisan or voter
mobilization. ere does appear to be a stronger degree of netroots mobiliza-
tion occurring within the Republican support base and their voters that balances
out, and even possibly counters, the more coordinated and centrally managed
digital power of the le. While Romney is clearly a major party candidate, the
self- confessed and objective disparity in his digital weaponry compared to the
Democrats does allow for some comparison to the Australian and French situa-
tion, where minor parties enjoyed an advantage in the levels of indirect or two-
step online contact during recent elections, compared with their major party
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
206 WHEN THE NERDS GO MARCHING IN
206
rivals. is idea of indirect online voter mobilization being the weapon of the
weak is thus given some further credibility by these results.
Second, the changing demographics of the recipients of online contact over
time indicates that US parties have made a conscious eort to rene and redi-
rect their digital strategy to ensure that they connect with those who are more
likely to vote. is is particularly the case with regard to age. e sharp drop in
the numbers of young people being contacted by both new and old methods
in 2012 suggests that parties were more successful in targeting their appeals to
those segments of the electorate where they were most likely to yield a return on
their eorts at the ballot box.
Finally, together these ndings also permit some insight into the question
posed earlier of whether the apparent decline in online contact recorded in 2012,
and particularly the drop observed among Democrats, was actually the result of an
increased precision and accuracy in the methods used— the “less is more” argu-
ment. e changing prole in recipients of parties’ digital messaging and increased
emphasis on older voters and those with a higher socioeconomic status suggests
that parties had adopted a more targeted approach, and were managing to direct
their messages at those most likely to turn out. e fact that neither party really
saw a signicant reward in terms of those contacted being more likely to support
them indicates that if their eorts had any impact, it was largely a reinforcing one.
at said, the party that appeared to gain most from their digital campaign in 2012
was the Republicans, with most of those gains coming through their supporter
networks rather than ocial campaign channels.
Based on this evidence, therefore, we would have to conclude that the new
digital and data- intensive mode of campaigning pioneered by Obama in 2012
failed to live up to expectations. is book is not unique in reaching this conclu-
sion. Anumber of post- election analyses of Obama’s campaign have reached a
similar verdict. An extensive report on the Democrats’ 2012 digital campaign,
“Inside the Cave,” produced by engage, a digital communications agency founded
by Republican strategist Patrick Runi, insisted that there were internal con-
icts and communication failures between the various digital sub- teams that
hampered their mobilization eorts. In particular, the reliance on an in- house
tech team of Harper Reed to build the tools from scratch
meant the tools wouldn’t be ready on day one, as they were with
My.BarackObama.com in 2008.... Field Director Jeremy Bird was later
open about the fact that they didn’t get technology tools to volunteers
early enough.46
Other accounts published by those on the le of the political spectrum echoed
the view that the Obama campaign team had faced signicant teething problems
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
207
The Trendsetter 207
in rolling out some of its more vaunted tools. e initial prototype of Dashboard
had reportedly aracted harsh criticism and even scorn from the eld operatives
as unt for its purpose. e reaction was strong enough to prompt threats to
cease using it if it was not improved. Although changes were subsequently made
that rendered the platform more user- friendly, tensions between tech and eld
proved to be a recurrent theme during the course of the campaign, as did conict
between the three sub- teams that Go headed. e resulting situation clearly
created a degree of ineciency and delay to the rollout of key apps and tools that
went largely unreported in the lead- up to election day.47
On the academic front, an article published a year aer Obama’s return to
the White House by the editor of the Harvard Political Review raised further
questions about the extent to which the Democrats’ win in 2012 was aribut-
able to the tech- savvy quality of their campaign. In his piece, entitled “Just How
Good Was the Obama Campaign?” Frank Mace identied an emerging body of
research that challenged the established view that it was Obama’s data- driven
methods which had won the day.48 Pointing particularly to the work of Sides and
Vavreck (2013) in their widely discussed book e Gamble, he reported how
serious doubt had been cast on the eectiveness of Obama’s data- crunching
eorts and particularly whether his use of micro- targeted advertising had deliv-
ered victory. e authors followed up on their study of the election with a more
explicit critique of the “moneyball” approach to campaigning that Obama had
pioneered and particularly the idea that it was his use of “big data” that had really
secured him victory. While they did not query the enhanced precision that his
new scientic approach to modeling voter behavior delivered, they did pose the
question of how much such techniques really maered for the nal outcome.
For the authors:
An election is a one- time, sudden- death contest. e election- year
economy and many other things were out of Obama’s and Romney’s
control. Moneyball can make a campaign more ecient, but cannot
always help the campaign win.49
Summary andConclusion
e story to emerge from this chapter is one of conformity and deviation from the
cycle of digital campaign development set out in Chapter1. In terms of confor-
mity, the progression in US candidates’ use of the technology at the national level
corresponds broadly to trends observed among parties and candidates in elec-
tions elsewhere. However, the pace and scale of change render it something of an
outlier. Compared to the path followed in other major established democracies,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
208 WHEN THE NERDS GO MARCHING IN
208
American campaigners have moved much further and faster through the four-
phase cycle outlined in Chapter1. Indeed, it is questionable whether the United
States ever really witnessed the type of amateur experimentation associated with
phase Ithat was widely manifest in the United Kingdom, Australia, and France
during their rst online elections. e websites that debuted in the presidential
election of 1996 displayed a degree of sophistication that took parties in other
countries at least one further electoral cycle to reach.
A further similarity that this analysis has revealed between the United States
and the previous case studies is the key role played by the mainstream political
le in pushing forward the digital campaign cycle. While it was the Republicans
who gained early plaudits for their online site design and fundraising eorts, it
was the Democrats who took the reins in the post- millennial period to exploit
the organizing and vote- geing potential of the new media. While Howard
Dean’s community- building activities captured the imagination of the media
and the public, it was Obama’s online campaign of 2008 that really moved the
technology to the center stage of elections. Key to this transition was Blue State
Digital, his technology advisors, who understood how to channel the grassroots
energy Dean had generated into strategically important online and oine activi-
ties. Basing their approach on Obama’s existing strengths in local organizing,
they designed a two- step ow model of digital mobilization on a scale hitherto
unseen.
By 2012, the focus had shied from designing a “two- step” to perfecting a
“one- step” model that centered on directly targeting those all- important unde-
cided voters with personalized appeals. is shi was powered by an unprec-
edented political tech operation that, for many, transformed the nature of
campaigning itself. Instead of eld- informed collective intuition and guesswork
being the mainstay of campaign strategy, managers now relied on data analysts
and highly skilled technicians to make their key decisions. e “art” of cam-
paigning was now transformed into something much closer to a science.50
us, by 2012 it would seem that entry into phase IV was well and truly
underway in the United States, at least looking at developments from the supply
side. e evidence from below, however, is rather less convincing. Overall, there
does appear to be some synchronicity in the paerns of mass engagement and
elite innovation. e most passive mode of citizen involvement in the digital
campaign— “read”— had gradually been supplemented by more redistributive
phase III– style activities, particularly among Democrat partisans. is peaked in
2008, as one would expect, given the strong embrace of CIC by Barack Obama’s
team. Campaigns have also increased their levels of digital communication with
their partisans and voters since 2004, with the mainstream le again assuming
an initial advantage in this regard. Until 2008, those who received some kind of
online persuasion during the election, either directly from the party or via their
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
209
The Trendsetter 209
personal networks, were more likely to vote Democrat. By 2012, however, they
were clearly struggling to retain any mobilizing benets from the technology.
Despite having increased their reach into those segments of the electorate most
likely to turn out, those contacted online by the campaign were no more likely
to support Obama than those who received no contact. Exposure to indirect
contact was actually associated with greater support for Romney.
e results leave us with the possibility that Democrats’ success in digital
mobilization in 2008 was something of an anomaly, and that by 2012 diminish-
ing marginal returns were seing in. Of course, only time and future research
can fully address this question. Notwithstanding any lull that may have occurred
in the forward march of digital campaigns in 2012, it was clear that at least one
party in the United States managed to meet several of the core criteria associated
with phase IV digital campaigning. In terms of its strategic goals, dened target
audience, tool development, and internal power conguration, the Democrats’
digital campaign aligns very closely with the criteria dening phase IV cam-
paigning in the nal column of Table1.1 in Chapter1, and more so than any of
the other parties examined in this book. No activity within the campaign— be it
fundraising, advertising, eld, or opinion and opposition research— was beyond
the inuence of Teddy Go and his team. As noted, it was perhaps only in the
area of voter engagement that they appeared to fail to hit the mark. While they
appeared to be increasingly ecient at extracting data from voters, it appears at
least from the broad analysis of this chapter that they had not worked out how
best to optimize that information, and convert it into squeezing out new sup-
port from the electorate. In the concluding chapter of the book, we speculate on
the limits and future of data- driven decision- making in campaigns and what, if
anything, lies beyond phase IV.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
210
When theNerds Go Marching In. Rachel K. Gibson, Oxford University Press (2020). © Oxford University Press.
DOI: 10.1093/oso/9780195397789.001.0001.
Conclusion
Digital Campaigns at the Crossroads
The central premise of this book is that over the past two decades we have
seen digital technology move from the margins to the mainstream of politi-
cal campaigning within Western democracies. This largely uncontroversial
observation is accompanied by three more contestable claims that have been
developed during the course of the analysis. The first of these is that this
progression can be broken down into four main phases:(1) experimenta-
tion; (2)standardization and professionalization; (3)community building
and activist mobilization; and finally, (4)direct voter mobilization. Each
phase is defined by a distinctive configuration of tools, goals, and organiza-
tional resources and a proximity to one of two ends of a power continuum—
equalization or normalization. While most nations have not advanced
through all four phases, this trajectory, we argue, provides a useful heuristic
for understanding past, current, and future developments in digital cam-
paigning in established democracies.
e second main contention of the book is that the position of countries
in this evolutionary cycle diers, and that this variance can be explained by a
combination of system- level traits, country- specic factors, and individual- party
characteristics. Afull understanding of the drivers of digital campaign develop-
ments thus requires large N study, as well as detailed country- specic analyses.
Chapter3 delivered on the rst of these demands by testing the impact of vari-
ous regime- level characteristics on levels of online campaigning in 18 nations.
e results of that analysis revealed that a critical mass of internet users among
the electorate and competitive presidential elections were among the most sig-
nicant factors in predicting higher rates of digital mobilization. Subsequent
chapters presented a more detailed picture of developments in digital cam-
paigning at the national level. Taking four established democracies, we showed
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
211
Conclusion 211
how distinctive cultural and political norms had aected parties’ abilities and
incentives to incorporate the new technology into their campaign armory. ese
chapters were also able to reveal the pivotal role played by certain parties in trig-
gering the shi into a new phase. While there was no “one size ts all” model of
adoption, there did appear to be a notable tendency for mainstream leist par-
ties, and some of the more prominent minor parties, to emerge as the catalysts
for change. Furthermore, in Australia there also appeared to be a prominent role
for non- party actors.
e third, and possibly most contentious, claim of the book is that the “main-
streaming” of digital technology in elections is fostering the growth of a new
type of campaign operative— the apolitico— and the emergence of a new con-
dition of hypernormality. Full entry into phase IV– style digital campaigning
is marked by the rise to power of a new organizational elite that are recruited
from outside the regular party political channels. Drawn from academia and
industry, these data scientists and soware engineers increasingly challenge and
ultimately “trump” the traditional role of eld experience in the critical decision-
making of the campaign. e emergence and domination of this new group of
digital experts can be seen in one sense as a continuation and culmination of the
normalization logic rst articulated by the e- pluralists and developed further by
Margolis and Resnick (2000) and later Howard (2006). Control over decision-
making is increasingly centralized in the hands of an inner team of specialists.
At the systemic level, competition decreases, as only the bigger parties have suf-
cient resources to engage in this new digitally intensive electoral warfare. ere
is, however, a subversion of this narrative in that this new “scientic” elite are not
autonomous actors. ey are heavily, if not exclusively, reliant upon computer
algorithms and statistical modeling to formulate campaign strategy. Machine,
rather than human learning now determines decision- making at the higher ech-
elons of party organization. Voters are, in turn, viewed as lile more than remote
and manipulable data points. eir input is sought mainly to help improve the
accuracy of the campaign’s forecasts and its targeting activities. We characterize
this condition as one of hypernormality in that power is concentrated and cen-
tralized in the hands of a few key players, to an exponential degree. Voter com-
munication shis from a state of managed or “controlled interactivity” to one
of full automation, and internal decision- making is eectively depoliticized. We
return to further develop and defend these arguments later in the chapter when
we reect on what comes next for digital campaigns.
All three of our central claims have been developed and tested using a variety
of methods and data. In the following, we summarize each one in more detail
and evaluate the credibility of the evidence gathered to support them.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
212 WHEN THE NERDS GO MARCHING IN
212
Claim 1:The Four Phases ofDigital
CampaignDevelopment
Beginning with claim 1, our argument for a four- phase model of digital campaign
development is based on the insights of the e- pluralists, or new Jeersonians, who
theorized that politics was likely to become more uid in the internet era, at
least initially. Parties would compete on a more even footing during elections
and become more porous structures that were open to greater grassroots input
and interaction with voters. Following this period of “equalization,” a counter-
tendency would emerge to “normalize” the situation. is would see the larger,
well- resourced, and hierarchically structured parties move in to dominate the
new cyber- space arena, leaving lile room for minor players, activists, or wider
citizen input.
is initial pendulum swing was then extended to include a new cycle of
equalization and normalization, as changes in the technological environment
and growth of use among voters and supporters ushered in a new set of capabili-
ties and goals for digital campaigns. is ultimately produced a four- phase model
of development that was broken down into eight dimensions of party operation.
Progress through the phases occurred in both linear and cyclical terms. e move
from phase Ito phase IV of digital campaigning is marked rst by a monotonic
growth in campaigns’ capacity and incentives to engage in this new form of elec-
tioneering. is expansion is, however, underpinned by the rotation between
two competing models of campaign operation. e rst takes a radical boom-
up approach to the task and seeks to open the process up to grassroots voices
and less powerful actors. e second follows a more conservative and top- down
logic that essentially reinforces the existing imbalanced power relations. e full
model of change was summarized in Table1.1 in Chapter1.
e applicability of the four- phase model to “real world” cases was then tested
in a variety of ways. We began with a retrospective review of the literature. is
exercise revealed how the key questions posed and conclusions drawn about
digital campaigning could be “ploed” onto the four stages set out in Table1.1.
is impressionistic evidence was then supplemented with a series of in- depth
case studies that traced historical developments in the supply and demand for
digital campaigning in four major democratic nations. ese studies conrmed
that the four- phase model of change provided a meaningful framework for map-
ping developments within individual countries. ey also revealed some impor-
tant dierences in the pace and nature of those changes.
Specically, while all four countries appeared to have entered phase III at
the time of writing (i.e., parties were engaging in signicant online community-
building activities), there was a distinct hierarchy in their progress. American
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
213
Conclusion 213
parties and candidates had moved most quickly through the sequence. Campaigns
at the national level had spent relatively lile time in the early phases, advancing
to phase III within 10years of the internet’s arrival. As of 2012, they were show-
ing clear signs of having entered phase IV. In contrast, Australia and France both
appeared to have moved more slowly through the cycle, although their paerns
of development and the reasons for this comparative inertia diered.
Australia, as the title of the chapter describing developments there indicates,
was something of an “early bloomer” among nations. An initial surge of enthu-
siasm for the new technology by campaigners quickly subsided into an elon-
gated period of standardization and professionalization. Parties and politicians
appeared to be almost “frozen” in their approach to the new medium. Signs of
engagement with the community- building properties of the web that emerged
under Kevin Rudd and Labor in 2007 were initially lauded, but ultimately dis-
missed by critics as “window dressing.” It was thus only aer the start of the next
decade that a genuine focus on activist mobilization really began to take hold
among Australian campaigners.
Developments in France were also slower than in the United States, but took
a rather dierent trajectory. From the outset, the pace of change was glacial, and
continued to be so for much of the rst decade of the internet’s existence. Where
innovations emerged, they occurred largely among fringe actors on the far right
and le. Major parties did not appear to take the medium seriously as an election-
eering tool. Even up until the 2007 presidential election, candidates were being
accused of paying largely lip service to its mobilizing power. By 2012, however,
the parties had begun to make up for lost time. In particular, the major le- wing
presidential candidate Francois Hollande made a concerted eort to import and
exploit the techniques introduced by his US counterpart, Barack Obama. is
direct injection of American expertise catapulted his campaign into a hybrid mix
of community building and direct voter mobilization that pushed it onto the
center stage of digital campaign innovation globally.
Nestled in between these cases of “arrested development” and the fast- paced
adoption of the United States is that of the United Kingdom. Progression by the
British parties through the cycle appeared to be the most steady and incremental
of the four cases examined. Early experimentation in the 1997 election cycle was
followed by a period of sustained professionalization that lasted for the next two
election cycles. Signicant forays into community building and activist mobi-
lization followed in the campaign of 2010. ese innovations were developed
further in 2015 and were supplemented by the introduction of phase IV– style
data- intensive and micro- targeted techniques. ese eorts, according to more
recent reports, became even more widespread during the 2016 Brexit referen-
dum campaign.1
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
214 WHEN THE NERDS GO MARCHING IN
214
Claim 2:The Common and Unique Drivers
ofDigital CampaignChange
Observing the similarities and dierences in the paerns of digital campaign
development in these four countries leads us to the second main conclusion of
the book, namely that a country’s progression through the cycle is driven by a
range of both common and unique factors. On the former front, the analysis of
Chapter3 was important in highlighting the role of regime- level features in this
process. Starting from the premise that higher levels of direct and indirect voter
contact are indicative of entry into the later phases of digital campaigning, we
compared the frequency of both modes during recent elections in 18 countries.
e results produced a four- tier ranking of countries according to their digital
campaign intensity. Asystematic analysis of these rankings revealed that com-
petitive presidential elections and a critical mass of internet users were key driv-
ers in the push toward phase III– and IV– style digital campaigning.
Armed with this “baseline” information, the case studies probed how far indi-
vidual countries’ experience matched with the expectations drawn from results
of the comparative analysis. Given that two of these cases were presidential sys-
tems (the United States and France), and two were parliamentary (Australia
and the United Kingdom), we began with an expectation of how our four cases
would line up in terms of the rapidity of their progression through the four-
phase cycle:the former two being ahead of the laer two nations. e results
of this more focused over- time analysis were useful in conrming the United
States as a leading nation, both in terms of the pace of advances made in digital
campaigning at the supply side (i.e., among parties and candidates, and in terms
of the levels of engagement with those eorts by the wider public). According to
Table3.2, the United States sat at the top of tier one countries in terms of the pro-
portion of voters contacted online. Just over one- third (34%) of the American
public had reported receiving some type of digital prompt about their vote dur-
ing the 2012 presidential election campaign. is higher intensity was matched
by the rapid pace of innovation observed at both the elite and mass level in terms
of take- up and use of the new tools during campaigns. Given that other coun-
tries not selected as case studies for this book displayed similar or higher levels
of online contact to the United States in recent comparable elections, as noted
in Chapter3, we do not see our evidence as supporting the notion of US excep-
tionalism in terms of digital electioneering. However, our ndings, we argue, do
give credence to the view that US parties occupy the “bleeding” edge of cam-
paign modernization.
e ndings for the other countries aligned less closely to expectations. In
particular, France was found to be at the boom of tier two countries in terms
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
215
Conclusion 215
of digital mobilization rates, below both the United Kingdom and Australia, as
well as Serbia and Mexico. is weaker than anticipated performance was under-
pinned, and in part explained, by its slower progress in moving through the four-
phase cycle and the more modest level of enthusiasm shown by French voters
for reading and redistributing digital campaign content, at least in the most
recent comparable elections. Closer investigation of the possible causes of this
inertia identied a complex set of “suspects.” Chief among them was the technol-
ogy itself. Despite levels of internet use achieving an equivalence in France to
those seen in other advanced economies, this parity had only recently emerged.
Tracing the history of internet diusion in France revealed a distinctly dierent
paern from that observed in our three other case studies. In particular, take- up
was much slower among the public. While this might reect a lack of interest or
degree of technophobia on the part of the French, review of the wider commu-
nication environment challenged such an explanation. For several years prior to
the arrival of the internet, French citizens had had access to, and made extensive
use of, a rival computerized communication network in the from of Minitel. e
presence of this alternative platform clearly diverted any demand for internet
services during the rst decade of their existence.
Even aer Minitel had been dismantled, there remained other regulatory
aspects of the communication environment that helped slowed the progress of
internet campaigning in France. Specically, the rules ensuring plurality in news
coverage by public broadcasters gives smaller parties a more prominent voice
in the mainstream media coverage of election campaigns than in most other
democracies. is more equalized exposure meant that one of the main benets
of the internet as a campaign medium was signicant reduced. Afurther and
arguably even stronger restraint on French parties’ use of email and social media
to communicate with voters was the long- standing prohibition on their ability
to collect citizens’ personal data for electoral- targeting purposes. is type of
segmentation of the public opened the door to discrimination, and thus vio-
lated the principle of egalité under the law enshrined in the French constitution.
ese controls are in direct contrast to the situation in the United States, where
constitutional protections on freedom of speech are frequently invoked to sup-
port campaigns’ rights to build up and rigorously exploit extensive voter les.
e ndings from the two “parliamentary” cases conformed more closely
to expectations. Australia and the United Kingdom are both located in tier two
of the league of digitally mobilized nations (see Table3.2), reporting “healthy,”
rather than “stand- out” rates of online voter mobilization. e slightly lower
ranking for Australia was seen as largely due to the timing of the data collection,
which occurred two years prior to that for the United Kingdom. Certainly all
things being equal, the Australian context appeared to provide one of the most
fertile environments for the growth of internet campaigning. Australia had a
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
216 WHEN THE NERDS GO MARCHING IN
216
strong reputation internationally for promoting e- government and e- democracy
initiatives. Australian parties had extensive experience in using electronic data-
bases and micro- targeting methods such as direct mail. Finally, although its
geography did present digital campaigners initially with a logistical challenge to
reach the large swath of rural voters lacking internet connections, the response of
successive federal governments to “roll out” broadband to the bush transformed
this barrier into an incentive. Once they were online, these remote electorates
became much more accessible through the new methods of communication
compared with resource- and labor- intensive methods of doorstep canvassing
and leaeting.
Against these strong incentives for parties to seriously engage in digital cam-
paigning, however, the gure of Je Kenne loomed large. Just as no individual
can be seen as solely responsible for the successful take- o of internet cam-
paigning in a country, neither can any one person be held entirely accountable
for its failure. at said, however, the former Victorian state governor comes
perhaps closest to acquiring the laer status. Kenne’s high prole and highly
personalized web campaign became the symbol of his inability to relate to the
concerns of ordinary voters. e medium, along with the governor, became a
target for ridicule. e ensuing criticism of Kenne’s campaign ensured that
Australian politicians and parties avoided exing their tech- savvy credentials
for several elections aerward. While this episode served to underscore the
power of “events,” in addition to institutional frameworks and technology, in
determining the speed of parties’ take- up of the technology, it also provided an
interesting insight into how advances occurred in the vacuum of major party
innovation. Elsewhere it had been the mainstream le parties that acted as the
main catalysts for change, particularly in the later phases. It was notable that
in Australia this role fell to the smaller parties and also non- party campaign
actors. Judged in particular by the activities of their supporter base, it was the
Greens and GetUp! that led the push into online community building and
activist mobilization.
While this mix of comparative and country- specic analyses provides a cred-
ible basis for identifying the key drivers of digital campaigning, it clearly does
not “close the book” on the subject. With regard to the large N study, an obvi-
ous improvement would be to add more cases. e full CSES module 4 data set
that includes 38 countries is now available for analysis, and study has already
commenced on the question of online mobilization.2 Extension and revision of
our explanatory model should also be considered in future analysis. In particu-
lar, while we concluded that current data- protection rules had no eect on the
amount of online contact voters received, as our later case studies have shown,
this may be a result of measurement error. Closer scrutiny of the French and
US experiences of digital campaigning, in particular, identied a deeper set of
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
217
Conclusion 217
constitutional and cultural norms that have aected parties’ willingness and
capacity to use citizens’ personal data for campaign purposes. While it is dicult
to develop accurate and comparable measures of these latent cultural norms and
aitudes toward micro- targeting, it should be possible to design indicators that
beer capture their outcomes or results. In particular, more aention could be
given to constructing indices that measure the extent of successful enforcement
of these national and international frameworks, and levels of compliance among
political organizations.
In terms of small N analysis, further strategically selected case studies could be
included to road- test and conrm the relevance of the four- phase model beyond
the countries examined here. How well does the pathway we describe “travel”
to examine developments in other established democracies? Which paern of
development is most commonly observed? e late blooming of France, the
early bloomer enthusiasm of Australia, the trend- seing pace seen in the United
States, or, the slow burn shown by the United Kingdom? Are there alternative
trajectories of growth that we can add to this list, and what other features of the
campaign and communication environment are relevant to understanding those
paerns of development? Looking even further aeld, can the model be used
to map changes over time in some of the newer democracies? Does their more
limited experience in running election campaigns mean that progression stalls at
the earlier stages? Or does this lack of a “path dependency” in campaign practice
make such countries more adept and open to deploying the tools in innovative
ways? Perhaps their newness means they will “leapfrog” over the experimenta-
tion and standardization phases seen in the established democracies, and move
directly to embrace the strategic goals of phase III and IV?
Claim 3:The Rise ofthe Party Apoliticos and
theShift toHypernormality
e evidence in support of our third and most provocative claim— that the
mainstreaming of digital technology is producing a new apolitical elite and con-
ditions of hypernormality— is the most speculative. To date, much of the litera-
ture on digital campaigning (including this book) has focused on charting the
outward and more visible changes that the new technologies are promoting.
Aention has thus centered on developments in the external “facing” aspects of
the campaign (i.e., the establishment and content of websites, social media pro-
les, apps, and measurable voter responses). While this has produced a range of
interesting and important ndings, it is an approach that concentrates arguably
on the “low- hanging fruit” of digital campaign change. e more challenging
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
218 WHEN THE NERDS GO MARCHING IN
218
and pressing task for researchers is now to detect and measure the changes that
are occurring in the internal power dynamics of organizations.
Bearing these caveats in mind, we argue that our analyses point to the emer-
gence of a new and potentially concerning trend— the depoliticization of cam-
paign management. is is a process whereby algorithms, articial intelligence,
and machine learning, along with those who have the skills to understand them,
assume a stronger role in deciding campaign strategy, and in so doing, demote
the role of eld experience and expertise generated through long- term exposure
to voters. Put in a wider political context, such developments can be seen as fur-
ther steps in the “hollowing out” of democracy, a process that Peter Mair (2013)
so eloquently described, and argued against, in his nal book, Ruling the Void.
Although Mair’s focus was on the rising tide of anti- political sentiment emerg-
ing at the heart of government, which he saw as driven by the inux of “neutral”
experts and technocrats who eschewed the cut and thrust of ideological conict,
there is a natural extension of his argument to the “scientication” of campaign
management and party decision- making documented in this book.
While it is not the claim of this book that any country has as yet, or indeed
will, reach a state of hypernormality, it is argued that a shi toward this state
is now detectable in some democracies, and that the march of the apoliticos
into campaign headquarters, or the “nerds,” as the book’s title more colloqui-
ally terms them, has begun.3 Richer qualitative study of parties, campaigns,
and the external personnel they employ or structurally embed is now needed
to explore and test the veracity of this contention. How are the recruitment
criteria for campaign staers and managers changing? What are the critical
skills and resources required to deliver modern campaigns? What happens
when data and gut instincts collide? While this work is underway in the
United States in studies of the rise of prototype politics by Kreiss (2016)
and the growth of data- driven campaigning in the United Kingdom (Anstead,
2017), we need a more systematic picture of the extent to which these new
approaches are taking hold in democracies elsewhere, beyond these “usual
suspects.”
Future Directions:Toward a Phase V?
e evidence that is cumulated and analyzed in this book, we argue, presents a
persuasive case for accepting our four- phase model as a framework for mapping,
comparing and forecasting the progression of digital campaigning across coun-
tries. However, even if one accepts the four- phase model as set out, this inevi-
tably begs the question of what comes next. Despite the extensive nature of the
empirical analysis undertaken in this book, we had to conclude our investigation
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
219
Conclusion 219
at some point. Ending our analysis in 2015 has meant we have been precluded
from integrating several major digital campaign developments that have since
emerged. To what extent do these developments t within our existing frame-
work? Or, do they suggest a challenge to it and a possible extension into phase V?
e development of digital campaigns obviously did not end in 2015, which
marks the nal election studied here. According to accounts of the 2016 US
presidential race, the role of digital and data analytics became even more cen-
tral to the Democrats’ eorts. Hillary Clinton allegedly refused to make a move
without reference to her head of analytics, Elan Kriegal.4 e view that 2016
simply saw the intensication and perfecting of the scientic method practiced
by Obama in 2012 is, however, challenged to an extent by the rise of Donald
Trump. e eventual Republican nominee went on record to reject the “measure
everything” approach pioneered by his le- wing rival. In one widely publicized
interview, he was reported as saying he had always considered such methods to
be “overrated...Obama got the votes much more so than his data processing
machine. And Ithink the same is true with me.”5 Trump in contrast, took to
Twier to engage in more direct and spontaneous digital communication with
his supporters and to rebut the stories put out by what he later described as the
“fake news media” industry.
e idea that the Trump campaign opted out of harnessing data analytics
expertise has since been largely debunked, most notably by Trump’s own digital
director, Brad Parscale, who, in several interviews, made clear the vast scale of
the campaign’s investment in social media advertising, particularly Facebook.
Employees from the company, he noted, were “embedded inside our oces” in
order to explain how to use the platform to target voters.6 Subsequently pub-
lished academic research has conrmed this new type of personnel “sharing” was
taking place between the campaigns and major tech companies on both sides
of the political spectrum at this time (Kriess and McGregor, 2018). Spending
reports issued by the Federal Election Commission (FEC) for the nal weeks
of campaign, however, supported Republican claims of a Facebook advertising
blitz. Figures recorded for the critical “home stretch” period, leading into elec-
tion day, revealed that Trump outspent Clinton almost two to one on digital
advertising ($29million, as compared with $16million).7
While Parscale’s testimony, and the FEC spending gures, underscored the
importation of data- driven techniques for the Trump campaign, it was the hiring
of Cambridge Analytica that perhaps most visibly, and controversially, signaled
their adoption of phase IV– style practices (Gonzalez, 2017). e mission of the
company, in the words of CEO Alexander Nix, was to use its unique psychologi-
cal proling tool to “determine the personality of every single adult in the United
States of America,” and based on these data, to identify the millions of voters
who are most open to being persuaded to support Trump.8 While the rm was
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
220 WHEN THE NERDS GO MARCHING IN
220
crystal clear about its ambitions to mine voters’ personal data in order to deliver
a Republican win, it proved much less transparent about the methods used to
reach this goal. is lack of clarity and rising concerns about potential breaches
of voter privacy led to a series of extensive post- election investigations of the
company’s tactics, and the platform that had been used to deliver those tailored
messages— tech giant Facebook.9
is new and more negative twist in the evolution of phase IV campaigning
gained further traction as reports emerged about how social media and particu-
larly Facebook were being used by the Trump team to target a range of so- called
dark ads to certain groups of voters. Although selective political marketing and
negative advertising are standard practice in election campaigns, the untrace-
ability of these digital ads earned them the label of “dark,” and meant that cam-
paigns could avoid scrutiny of their more controversial and contested messages.
In the case of Trump, reports surfaced of racially targeted dark ads that were
explicitly designed to deter and demobilize Clinton’s support among the African
American voters.10
e view that digital campaigns had now taken on a new impetus in terms
of promoting the “dark arts” of voter manipulation and suppression gained per-
haps its strongest support aer the publication of the report by the US Senate
Select Commiee on Intelligence (SSCI) on foreign interference in the 2016
election. e report revealed a paern of coordinated and forceful “weaponiza-
tion” of social media to spread misinformation and increase divisions in the US
electorate during the presidential campaign (Howard etal., 2018).11 While the
campaign had been waged by foreign actors, most notably the Russian Internet
Research Agency (I), in partisan terms the content had been more favor-
able to the Republicans and its clear purpose had been to help secure a Trump
victory.12
As in 2012, claims about the accuracy and success of digital tools in mobiliz-
ing, and now, demobilizing US voters have been subsequently scrutinized and
challenged by academic analysts. According to Baldwin- Philippi (2017), while
Facebook ads may have been the sine qua non of the election, there was no evi-
dence of any “great leap forward” in their reach and eectiveness:
while the [Trump] campaign’s Facebook ad strategy was productive it
was ultimately similar in quality and more extensive in quantity than
that of the Clinton campaign, both of which were largely an extension
of the Obama 2012 campaign. (630)
In addition, the extent to which people were actually aected by deliberate dis-
information campaigns or the inadvertent circulation of “fake news” stories on
Facebook (and other social media platforms) has also been called into question.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
221
Conclusion 221
Based on survey data, election Twier data, and the publicly released I posts
from the SSCI inquiry, a number of studies showed that deliberately manufac-
tured false content was consumed by only a small segment of already “decided”
voters. Furthermore, the crudity and inauthenticity of the content of most of the
automated messages meant that their power to persuade was likely to be very
limited (Allco and Gentzkow, 2017; Boyd etal., 2018; Grinberg etal., 2019;
Salamanos etal., 2019).
Despite the rising doubts over its eectiveness, it is clear that this switch
of phase IV campaigning to embrace more opaque and manipulative tactics
on social media has spread to elections beyond the United States. e UK
Brexit Referendum in 2016 prompted several formal investigations of cam-
paigns’ use of online targeting tactics, and specically whether the services of
Cambridge Analytica had been used by those involved in the “leave” side of
the debate.13 Elsewhere, elections in Mexico, Brazil, and India have sparked
concerns that voters are at risk of manipulation by automated and deliber-
ate aempts at misinformation through targeted selective exposure to social
media and, in particular, Facebook advertising (Arnaudo, 2018; Bashyakarla,
2018; Glowacki etal., 2018).14 As in the case of the US 2016 election, while
some of these activities have been linked with the parties and candidates, there
is also evidence of involvement by non- party actors and foreign governments.
e term “digital astroturng” has now sprung up as a means of describing this
new phenomenon, whereby domestic or foreign actors manufacture activity
on the internet to mimic authentic grassroots activism, thereby building false
momentum in support of a particular agenda or candidate (Keller etal., 2020;
Kovic etal., 2018).
As well as creating new challenges for governments in terms of how to regu-
late these new practices and raising doubts about the legitimacy of election out-
comes, this shi in the direction of data- dr iven campaigning, and particularly the
rise of these new actors, poses a signicant new threat for political parties’ con-
tinuing relevance (Domme and Temple, 2018). Given talk of parties ongoing
“decline,” the growth of these new digitally enabled networks and satellite move-
ments, whether real or not, clearly has the capacity for siphoning away existing
and new sources of support (Andeweg and Farrell, 2017; Scarrow, 2014). Rather
than the end result being the further “hollowing out” and increased depolitici-
zation of campaigns, however, the outcome, at least in the short term, may be
a more intense period of repoliticization, as digital channels are used to spread
disinformation and polarize electorates.
As such, data- driven campaigning looks set to continue, but we may see a split
in the direction of travel. One version sees the continued mainstreaming and
renement of these scientic methods within the connes of conventional cam-
paign practice. Asecond variant sees the deployment of these techniques toward
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
222 WHEN THE NERDS GO MARCHING IN
222
more subversive and anti- systemic ends, with disinformation, demobilization,
and division of voters being the driving goals (Gibson and Römmele, 2020).
While we could stop our horizon- scanning exercise at this point, there would
appear to be both theoretical and empirical grounds for considering a third tra-
jectory for digital campaigning to take, based on post- 2015 developments. Here,
we draw on the recent work of Benne etal. (2018), particularly his aempts
to identify a new “connective’ ” style of political party, and earlier theorizing by
Andrew Chadwick (2007) about the rise of “organizational hybridity” in the
digital age. Both authors highlighted the growth of a new style of political orga-
nization that was particularly dependent on the internet, and which blends the
electorally oriented goals of parties with the structural uidity of social move-
ments. According to Benne, new European populist parties such as Píratar in
Iceland, Alternativet in Denmark, and Podemos in Spain are among the most
obvious incarnations of this new organizational model— their dening feature
being that internet technology sits at the heart of their organization and provides
their “central operating system.”
is immersion of party structures within the digital environment is perhaps
the ultimate expression of how the technology has moved from the margins to
the mainstream. However, lacking the intensive analytics capabilities and data-
base infrastructure of the bigger parties, these groups promote an alternative,
more organic, and boom- up approach to social media targeting. Messages are
seeded in supporter networks, and people, rather than machine algorithms,
power their dispersal. As a result, their campaigns can prove to be somewhat
chaotic and uncoordinated in their actions. is weakness can also be regarded
as a strength, however, in that it helps to cultivate an image of authenticity and
spontaneity, which in turn increases their appeal among those voters who are
turned o by the major parties’ switch to a more robotic or machine- led style.
Despite being born of necessity, therefore, this connective approach can be seen
to oer something of a pushback against some of the more machine- like tenden-
cies of phase IV.
Viewed together, these trends suggest that, rather than a wholesale shi into
a singular new phase V era, the future of digital campaigning is more likely to
involve the bifurcation and possibly trifurcation of current practice. On the one
hand, we can envisage an intensication of the scientic mode of voter mobili-
zation as campaign teams place more resources and eort into expanding their
databases and improving their predictive models of voter behavior. Alternatively,
we may also see the subversion of that scientic model, whereby campaigners
begin to understand how it can be “beer” used to demobilize voters and spread
misinformation. Finally, a third and more reactionary trend may erupt among
the smaller and newer parties that lack the resources necessary for either inten-
sication or subversion of the scientic model. ese actors weave the internet
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
223
Conclusion 223
into their infrastructure of organizational DNA to develop a more connective
and organic model of micro- targeting voters that relies on digital tools, but is
driven by human actors and real social networks.
Which one, if any, of these three trends will win out and redene the cur-
rent phase, or dene any new phase, is not clear at this point. However, what
is apparent is that it maers greatly which one does. e rst two scenarios, in
particular, present signicant challenges to the future healthy functioning of
democracy. Intensication of the scientic model threatens to further depo-
liticize and hollow out parties, and shrink the public sphere. Campaigning will
increasingly focus on the easily persuadable “perceived” electorate (Hersh,
2015)leaving the disengaged and “unperceived” electorate out of scope and
under- mobilized. Current participation and knowledge gaps will thus inevitably
widen as byproducts of the drive to achieve ever more precise levels of micro-
targeting. Intensication of the subversive mode of data- driven campaigning
clearly carries a more overt and direct threat to the democratic process, as par-
ties are increasingly sidelined, and “enemy” agents use digital technologies and
social media platforms to destabilize, demobilize, and divide electorates.
Perhaps the best hope for the future of democracy lies with the growth of the
connective model of digital campaigning and the shi toward increased organi-
zational hybridity. Here there is a blend of technological and human commu-
nication that eschews the worst excesses of the scientic and subversive modes
of digital campaigning. Amore diverse range of non- usual suspects are reached
with authentic campaign messages, and there is no deliberate aempt to manip-
ulate or mislead voters with false and divisive messages. Arguably, such methods
promote a return to the “grassroots in cyberspace” (Bonchek 1995)approach to
campaigning, or what has since been termed the “netroots” model of political
organization and communication (Armstrong, 2006; Feld and Wilcox, 2008).
e fact that this is essentially the strategy of the weak rather than the strong,
however, suggests that it is unlikely to become the dominant force in future cam-
paigns. Whichever mode does emerge as the primary modus operandi from this
point, what is clear is that this movement of digital technology from the margins
to the mainstream of campaigning is of profound signicance, not just for the
future of political parties and elections, but for the longer term health and sus-
tainability of the wider democratic project.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
224
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
225
225
APPENDIX 3.1
Tables3.1 and 3.3, Measures
ofCampaignContact
Data source: e Comparative Study of Electoral Systems (www.cses.org).
CSES Module 4 Second Advance Release [2.0]. March 20, 2015 Version. Final
release available at hps:// cses.org/ data- download/ module- 4- 2011- 2016/
Online Direct Contact (Binary)
1 = yes to the following question:“During the campaign, did a party or can-
didate contact you in person or by any other means?” AND to any of the subse-
quent response options:“Did they contact you by text- message or SMS? By email?
rough a social network site or other web- based methods?
Else = 0 DK, and refused also coded to zero to ensure consistency across
countries.
Online Indirect Contact (Binary)
1=yes to the following question:“During the campaign, did a iend, family mem-
ber or other acquaintance try to persuade you to vote for a particular party or can-
didate?” AND yes to any of the subsequent response options:“Did they try to
persuade you by text- message or SMS? By email? rough a social network site or
other web- based methods?
Total Online Contact (Binary)
1=yes to Online Direct Contact by SMS, email, or social network OR yes to
Online Indirect Contact by SMS, email, or social network.
Oine Direct Contact (Face to Face) (Binary)
1=“During the campaign, did a party or candidate contact you in person or by any
other means?” AND yes to “Did they contact you in person, face- to- face?”
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
226 Appendix 3.1
226
Oine Direct Contact (Mail/ Phone) (Binary)
1=“During the campaign, did a party or candidate contact you in person or by any
other means?” AND any of the subsequent response options:“Did they contact
you by mail? By phone?”
Total Direct Contact (Binary)
1=yes to Online Direct Contact by SMS, email, or social network OR yes to
Oine Direct Contact in person, by mail, or phone.
Sign- up Online:(Binary)
1=yes to the following question “Prior to, or during the campaign, did you use the
internet or your mobile phone to sign up for information or alerts (e.g., e- newsleers,
text messages, RSS, or blog feed) om a party or candidate?”
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
227
227
APPENDIX 3.2
Tables3.5 and 3.6, Dependent and
IndependentVariables
Primary data source:e Comparative Study of Electoral Systems (www.cses.
org). CSES Module 4 Second Advance Release [2.0]. March 20, 2015 Version.
Final release available at hp:// www.cses.org/ datacenter/ module4/ module4.
htm
Dependent Variable
CSES Total Online Contact (Binary)
1 =yes to Online Direct Contact by SMS, email, or social network OR yes
to Online Indirect Contact by SMS, email, or social network. Else = 0 (see
Appendix 3.1 for full question wording).
Independent Variables (Individual Level)
CSES Gender
0=Male; 1=Female
CSES Union Membership
0=Non- member; 1=Member
CSES Education (ISCED Codes)
No qualications (ref category) 1— primary; 2— lower secondary; 3— upper
secondary; 4— post- secondary non- tertiary; 5— short- cycle tertiary; 6—
bachelor or equivalent; 7— master or equivalent; 8— doctoral or equivalent
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
228 Appendix 3.2
228
CSES Close to a Party
“Do you feel close to one party? Or closer to one party than all the others?”
0=No; 1=Yes.
CSES Age
Measurement varies by country but typically based on individuals’ self- reported
age in years at the time of the survey.
CSES Le– Right Party- Placement
“W here would you place Par ty A, B, C, etc., on this scale?” 0=“Le”; 10=“Right.”
CSES Sign- up Online
“Prior to, or during the campaign, did you use the internet or your mobile phone to
sign up for information or alerts (e.g., e- newsleers, text messages, RSS, or blog feed)
om a party or candidate?”.
0=No; 1=Yes
Independent Variables (System Level)
Human Development
United Nations Human Development Index. CSES macro le. Variable D5083
Economic Development
Gross domestic product per capita. CSES macro le. Variable D5080_ 1
Political Development
Age of current regime. CSES macro le. Variable D5052.
Internet Use
International Telecommunication Union (ITU:number of internet users and
natural log of internet users per 100 inhabitants. CSES macro le Variable, D5095.
Interest in Politics
% of population=“very interested” and “interested” in politics calculated from
two comparative sources:
World Values Survey, Wave 6 (2010– 2014) Australia, Japan, Mexico, New
Zealand, Taiwan, ailand, and the United States
How interested would you say you are in politics? Are you:1 Very interested; 2
Somewhat interested; 3 Not very interested; 4 Not at all interested?
European Social Survey. Round 7 (2014) Austria, France, Germany, Greece,
Ireland, Iceland, Poland, Servia, Switzerland
How interested would you say you are in politics? Are you:1 very interested; 2 quite
interested; 3 hardly interested; 4 not at all interested? DK
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 3.2 229
229
Average Voter Turnout (Post– World War II)
Based on recent national elections (ranging from minimum of 8 in Germany to
maximum of 32 in Switzerland— national referenda included).
Source:IFES Electionguide; www.Electionguide.org
Compulsory Voting (Binary)
CSES macro le Variable D5044 Original variable coded 1– 4. Recoded as
binary:strong enforcement and weak enforcement=1; no sanctions=0
Electoral System
Single- Member District/ Majority/ Plurality/ Mixed=1; PR=0
e International Institute for Democracy and Electoral Assistance hp://
www.idea.int/
Frequency of Elections
Months since last national election CSES Macro le. Variable D5055
Presidential Election (Binary)
1=Presidential Election; 0=Legislative/ Parliamentary
Source:IFES Electionguide www.Electionguide.org
Average District Magnitude
Average number of candidates per district. Calculated for elections to lowest tier
chamber. CSES macro le Variable D5064
Eective Number of Electoral Parties
Based on formula of Laakso and Taagepera (1979) formula: ENEP=1∑v2i
where vi is the percent of votes obtained by the ith party. CSES macro le.
Variable D5103
Ideological Polarization (0– 10)
is was calculated according to Dalton’s Polarization index (Dalton, 2008). e
index is computed using an average of respondents’ individual placement of par-
ties on the le– right scale of parties in the micro CSES le (see the preceding).
e Polarization index is measured as the following:PI=SQRT{∑(party vote
share i) *([party L/ R score i— party system average L/ R score]/ 5)2}, where
i represents individual parties. is index is comparable to a measure of the
standard deviation of a distribution and is similar to the statistics used by other
scholars. It has a value of 0 when all parties occupy the same position on the
Le– Right scale and 10 when all the parties are split between the two extremes
of the scale. For further information, see Dalton (2008) . Data were missing for
Taiwan, L– R placement question was not asked.
Party- Centered System (Binary)
0=Party centered; 1=Candidate centered
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
230 Appendix 3.2
230
Calculated from the rankings supplied in the Johnson and Wallack Electoral
Systems Dataset, which covers 180 nations over the period 1975– 2005.
Specically, we used the DOM_ NK variable, which orders electoral systems
on a 1– 13- point ordinal scale based on the incentives provide to cultivate a per-
sonal vote. e measure is based on the original a to m ranking by Carey and
Shugart (1995). e DOM_ NK is calculated for elections to most populous
electoral tier within the legislature. Acountry with a DOM_ NK of 1 would
have a tier with the lowest possible rank of personal vote incentives, and that tier
would account for the majority of the members in the assembly. Conversely, a
country with a DOM_ NK of 13 would have a tier with the highest possible
rank of personal vote incentives. e rankings were recoded into a binary vari-
able with scores of 1– 8=1“party- centered” and 9– 13=1“candidate- centered.”
is resulted in Australia, France, Japan, New Zealand, Taiwan, ailand, and
the United States=candidate- centered, and Austria, Germany, Greece, Ireland,
Mexico, Poland, Serbia, and Switzerland=party- centered. For further details,
see Johnson and Wallack (2012). Data available at hps:// dataverse.harvard.
edu/ le.xhtml;jsessionid=6c796b432626437805a53e5fe02e?leId=2409085
&version=RELEASED&version=.0
Free or Subsidized Media Available (Binary)
1=Free or subsidized media available to parties or candidates for campaign-
ing; Source:e International Institute for Democracy and Electoral Assistance
(IDEA) Political Finance Database
hp:// www.idea.int/ data- tools/ data/ political- nance- database
Cap on Party Spending (Binary)
1=there is a cap on campaign spending by candidates or parties.
Source: e International Institute for Democracy and Electoral Assistance
IDEA) Political Finance Database:
hp:// www.idea.int/ data- tools/ data/ political- nance- database
National Vote Margin
Dierence in national % vote share between top two parties if election in ques-
tion was parliamentary/ legislative election and between top two candidates if a
presidential election.
Source:IFES Electionguide:www.Electionguide.org.
Note:Data for ailand 2011 not available; used Adam Carr’s Psephos data-
base:www.psephos.adam- carr.net/ Adam Carr
Data Privacy Index (0– 5)
Additive index based on presence of any of the following 5 features:
1) Has a data/ information commissioner or equivalent
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 3.2 231
231
2) Has a legal requirement to appoint a data protection ocer to certain
organizations
3) Can enforce breech in data protection by criminal sanctions
4) Has specic laws pertaining to online privacy
5) Requires data subjects’ consent to pass data on to a third party.
Source:DLA Piper Data Protection Laws fact book, hp:// dlapiperdataprotec-
tion.com/ #handbook/ world- map- section
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
232
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
233
233
APPENDIX 4.1
Table4.1, Citizen- Initiated Campaign
IndexConstruction
Community Building (0– 8 Additive Index)
Set up prole— Users can establish a personal page or prole within the site fol-
lowing the logic of a social network site such as Facebook. e contents of this
can include:
Photo— a personalized image or photo;
Biography— a short statement about themselves, their general interests, family
life, hobbies, etc.;
Why joined— a more political statement about their interest in the party and why
they want to help the campaign;
Set up/ join groups— a facility to start or get involved with a sub- community
of other members within the site to support the candidate or party, based on
a shared interest or identity. Examples could include gay and lesbian groups,
African American, trade unionists, environmentalists.
Set up blog— a facility to establish a personal blog within the site on which users
can post their thoughts and responses to the campaign, comment on and follow
other blogs, and be followed by other users;
Set up Wiki— a facility whereby a group of users can set up a collective work
space to share, write, and archive documents relating to policy or other maers
of interest;
Email/ msg system— an internal messaging system through which users can send
private messages to each other;
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
234 Appendix 4.1
234
Externally promote prole— Users can publicize their membership in the site
externally through a generic “share” buon, or are given an explicit option to
link their proles to their Facebook or Twier accounts.
Resource Generation (0– 6 Additive Index)
Personal fundraising— Users can download soware or can set up a publicly
accessible site within their account that allows them to solicit and receive dona-
tions of money for the party or candidate.
Promote membership— Users can send out appeals via email or their Facebook /
Twier accounts inviting others to join the party.
Sign up as local organizer— Users can complete a form (online or download,
print, and post) or are invited to send an email to the campaign HQ oering to
act as a local organizer, neighborhood or team leader.
Sign up as candidate— Users can complete a form (online or download, print,
and post) or are invited to send an email to the campaign HQ signaling their
interest in becoming a candidate in the future.
Organize/ add event— Users can complete a form (online or download, print, and
post) or are invited to send an email to the campaign HQ oering to organize/
host an event that will help to raise funds, or recruit volunteers for the campaign.
Vote leaders to aend events— Users can sign a petition or are invited to send an
email to the campaign HQ to “vote” on where the candidate or party leaders
should visit during the course of the campaign.
Voter Mobilization (0– 11 Additive Index)
Get Out e Vote (GOTV) oine— Users are given opportunities to mobilize
and remind voters in person, on the phone, or by posters to turn out for the
candidate or party on election day.
Access phonebank— Users can sign up to make GOTV phone calls to voters. In its
“ideal type” this will entail their being given access to download phone records, a
prepared script, and instructions on how to start calling voters from their homes.
Sign up for f2f canvassing— Users can sign up to start canvassing voters by visiting
them in their homes. In its “ideal type” this will entail their being given access
to download a list of likely voters’ names and addresses, a prepared script, and
street plans of where they need to go.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 4.1 235
235
Sign up to discuss with iends and family— Users are invited to make an online
pledge that they will contact a certain number of close associates before election
day to remind them to vote for the candidate/ party.
Leaets download— Users are given options to download pdfs of yers and pro-
motional material to distribute to voters that will publicize and promote voting
for the party/ candidate.
Externally promote event— Users are encouraged to publicize events they are
aending or organizing via the site externally to their networks through a generic
“share” buon, email, or directly posting to Facebook and Twier.
GOTV online— Users are given opportunities to mobilize and remind voters
through online communication tools to turn out for the candidate or party on
election day.
Send email— Users are given a template email that they can edit and send out to
their contacts reminding them to vote and promoting the candidate/ party.
Post to Facebook— Users are invited to post a message to their Facebook pro-
file reminding those in their network to vote and promoting the candidate/
par ty.
Post to Twier— Users are invited to post a message to their Twier feed remind-
ing those in their network to vote and promoting the candidate/ party.
Smart phone app— Users can download a custom- made smart phone application
that will allow them to send a SMS to their contacts, reminding them to vote and
promoting the candidate/ party.
Email forward to a newspaper editor— Users are given a template mail that they
can edit and send on to an editor of a local or national newspaper for publication
that is supportive of the candidate/ party’s message.
Start e- petition— Users are given the tools to set up an e- petition on a cause or
issue of importance to the candidate/ party during the campaign.
Message Production (0– 11 Additive Index)
Message creation
Policy email fwd/ customize— Users are oered a template email on party policy
and are invited to edit it and develop the party’s message by oering their per-
sonalized view and sending it to their contacts.
Poster/ leaet create/ customize— Users are oered tools to create a campaign
poster or leaet or to edit a template that they can send to their contacts online
or print and display oine.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
236 Appendix 4.1
236
Policy input— Users are given the opportunity to make comments on current
policy and oer ideas and suggestions to develop policy proposals of the party
via a special policy forum or “ideas” thinktank. ere is an explicit commitment
to consider the ideas with a national policy making body.
Message distribution
Web banners/ ads download— Users can download promotional items from the
site such as logos and banners that can be added to their own blogs, social net-
work proles, or other types of online presence
Posters/ leaets download— Users can download promotional items that can be
printed and displayed in their window or car. In contrast to the GOTV leaets,
these are items that individuals use to publicly express their own support for the
party, while the GOTV documents are more instrumental and are designed to
be distributed to get others to turn out to vote.
Email/ share policy docs— Users can click on share or forward buons to send
out policy documents such as the manifesto to those in their social network or
email contact lists.
Newsfeed to website— Users can set up a newsfeed from the site to their own
online presence so that RSS and news updates from the party are automatically
displayed on their blog or webpage.
Share blog posts externally— Users can click on share or forward buons to send
out party blog posts such as the manifesto to those in their social network or
email contact lists.
Link to SNS prole— Users can set up a link from the site to their own Facebook
or SNS account so that updates from the party are automatically posted to their
prole.
Link to Twier account— Users can set up a link from the site to their own
Twier account so that updates from the party are automatically posted to their
twier feed.
Import email contacts— Users are given the option to import their email address
book into their online prole so that they can easily send out messages to their
contacts from the party.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
237
237
APPENDIX 4.2
Tables4.2– 4.4, Data Sources and
VariableDefinitions
e results reported in Tables4.2, 4.3, and 4.4 use data from four surveys cover-
ing the 2005, 2010, and 2015 general elections.
2005
For 2005 we used data from the ESRC- funded project “Campaigning in
Cyberspace: e 2005/ 6 General Election Online,” RES- 000- 22- 1284. e
questions were elded as part of an NOP omnibus post- election survey of a
stratied sample of 1,937 British adults aged 18years or older. Quotas for age
and working status within sex were applied following a one- stage ACORN and
region stratication. e data were weighted to ensure that demographic proles
matched those for all adults in Great Britain age 18 or older. Interviews were
conducted face to face between May 12 and 17, 2005.
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
All questions were measured as binary 1=yes have done; else=0. e questions
used to measure the three main types of engagement were as follows:
Read
E- news:“How much of your news and information about the election did you
get from the Internet?” Alot/ some/ a lile recoded=1; none=0.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
238 Appendix 4.2
238
Campaign sites:“Which, if any, of the following websites did you visit to get
election information or news:A political party site OR a candidate site?”
Redistribute
Sign up/ Use Party Tools:“Which if any online election- related activities have
you participated in?” Volunteered online to help a party or candidate OR
Downloaded election material OR subscribed to receive email newsleer from
parties or candidates=1.
Share/ Exchange:“Did you send any emails about the 2005 general elec-
tion to any of the following:Family/ friends or work colleagues (people you
know)?”
Redistribute:Sign up/ Use Party Tools =1; AND Share/ Exchange=1.
Receive
Receive Online Direct:“Did you receive any emails about the 2005 general elec-
tion from any of the following? Parties OR Candidates?”
Receive Online Indirect:“Did you receive any emails about the 2005 general
election from any of the following:Family/ friends or work colleagues (people
you know)?”
2010
For 2010 we used data from the ESRC- funded project “e Internet, Electoral
Politics and Citizen Participation in Global Perspective,” RES- 051- 27- 0299.
e questions were elded as part of a BMRB omnibus post- election survey
of 1,960 British adults aged 18years or older. Quotas for age and working sta-
tus within sex were applied following a one- stage ACORN and region strati-
cation. e data were weighted to ensure that demographic proles matched
those for all adults in Great Britain age 18 or older. e data are available from
the UK Data Service: Gibson, Rachel (2013), e Internet, Electoral Politics
and Citizen Participation in Global Perspective [Data Collection]. Colchester,
Essex: Economic and Social Research Council. 10.5255/ UKDA- SN- 850856
File name “Original_ les_ UK_ 2010_ BMRB_ F2F_ post_ election_ survey_
w_ newspaper_ readership.sav”; available at hp:// reshare.ukdataservice.
ac.uk/ 850856/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 4.2 239
239
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
All questions were measured as binary 1=yes have done; else=0. e questions
used to measure the three main types of engagement were as follows:
Read
E- news:“Please could you tell me, whether you have done any of the following
activities in relation to ocial parties or candidates online:Read or accessed any
mainstream news websites or news blogs to get information about the campaign
(e.g., BBC news online, e Guardian online, etc.) OR Viewed or accessed vid-
eos with unocial political or election related content?”
Campaign sites:“Please could you tell me, whether you have done any of the
following activities in relation to ocial parties or candidates online:Read or
accessed any party or candidate produced campaign sites (home pages, ocial
Facebook prole, ocial Youtube channel, etc.)?”
Redistribute
Signup / Use Party Tools:“Please could you tell me, whether you have done
any of the following activities in relation to ocial parties or candidates
online:Used any online tools to help parties and candidates in their cam-
paign (e.g., sent OR forwarded email or texts promoting a party, set up or got
involved in a campaign meeting or event through party sites or Facebook or
Twier, reposted party logos or material on your own site or prole, helped
them design a poster/ ad, downloaded leaets or posters to promote the party
oine, etc.)?”
Share/ Exchange:“Which, if any, of the following activities did you do online
during the election campaign over the last month:Posted comments of a
political nature on a blog or a wall of a social networking site (either yours or
someone else’s) OR Joined or started a political or election related group on a
social networking site (e.g., Facebook, MySpace, etc.) OR Forwarded uno-
cial campaign content (links to video, news stories, jokes, etc.) to friends, fam-
ily or colleagues via email, SMS, Twier, or through your Facebook network
OR Embedded or reposted unocial campaign content (links to video, news
stories, jokes, etc.) on your own online pages (i.e., a social networking prole,
blog or home page)?”
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
240 Appendix 4.2
240
Receive
Receive Online Direct:“In the course of the recent election, did anyone from a polit-
ical party, campaign, or political organization contact you to ask about how you
were planning to vote through any of the following methods:Online or internet-
based contact (i.e., through email or any internet/ web- related technology)?”
Receive Online Indirect: “In the course of the recent election, did you receive
any campaign- related political messages or content through the internet from a
friend, a member of your family, or someone at work?”
Receive Oine Direct: “In the course of the recent election, did anyone from a
political party, campaign, or political organization contact you to ask about how
you were planning to vote through any of the following methods:By telephone,
mail, or in person and face to face?”
2015
For 2015 we used two British Election Study surveys. e rst survey was the
Wave 5 of the internet panel study conducted by YouGov, which was used for mea-
suring the read, redistribution, and receive modes in Tables 4.2 and 4.3. e N
was 30,725 adults aged 18years and older and included internet users only. e
data were weighted using the “core weight.” e second survey was the mail- back
component of the post- election face- to- face survey, which contained the CSES
module on mobilization and was used to measure the receive mode in Table4.2,
4.3, and 4.4 by parties and from friends and family. e sample was 1,567 and
included internet users and non- internet users. e data were weighted with the
wt_ combined_ CSES— combined CSES weight. e decision to use two dier-
ent data sources was prompted by the fact that no single source provided measures
of all three modes. e YouGov panel allowed us to measure the read and redis-
tribute modes of engagement reported in Tables4.2 and 4.3. e CSES module
provided the best measure of receive in that it allowed us to dierentiate direct and
indirect forms of this mode, which is an important distinction to retain. Finally, the
CSES data were used to measure the receive mode in the United Kingdom for the
comparative analysis of Chapter3 and reported in Table3.2. us, on consistency
grounds it made sense to use this measure of receive throughout the book.
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
All questions were measured as binary 1=yes have done; else=0. e questions
used to measure the three main types of engagement were as follows:
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 4.2 241
241
Read
E- news:“On an average weekday, how much time (if any) do you spend following
news about politics or current aairs from the Internet (not including online news -
papers)?” All responses from less than half an hour to over 2 hours=1; None=0.
Campaign sites:“In the last four weeks, have you read news or found informa-
tion about the upcoming general election or politics more generally that was
tweeted by any of the following people or organizations:Candidates or Parties
on Twier OR Candidates or Parties on Facebook?
Other than on Twier and Facebook, have you visited the website of a candidate
or party in the last 4 weeks?”
Redistribute
Sign up/ Use Party Tools:“Have you signed up or ocially registered online to
help a party or candidate in their campaign?”
Share/ Exchange:“In the last four weeks have you personally posted or shared
any political content online, e.g., through Facebook OR Twier OR email OR
instant messaging or another website/ platform?”
Receive
Receive Online Direct:“During the campaign did a party or candidate contact
you in person or by any other means?” If Yes:“Did they contact you by:text
message/ SMS OR email OR a social network site or other Web- based method?”
Receive Online Indirect: “During the campaign did a friend, family member,
neighbor, work colleague, or other acquaintance try to persuade you to vote for
a particular party or candidate?” If Yes:“Did they try to persuade you in any of
the following ways:text message/ SMS OR email OR a social network site or
other Web- based method?”
Receive Oine Direct: “During the campaign did a party or candidate contact
you in person or by any other means?” If Yes:“Did they try to persuade you in
person, face- to- face; by mail; by phone?”
Socio- Demographic and PoliticalVariables
PARTYID
2005:“Regardless of whether you voted in the election or not, generally speak-
ing, which political party, if any, do you tend to support?” Labour; Conservative/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
242 Appendix 4.2
242
Tory; Liberal Democrat; “Other” combines Scoish Nationalist Party (SNP),
Plaid Cymru (PC), Green Party, United Kingdom Independence Party (UKIP);
No particular party; DK; Refused.
2010:“In general, do you think of yourself as a lile closer to one of the par-
ties than the others?” If Yes: “Please can you tell me which party?” Labour;
Conservative; Liberal Democrat; “Other” combines Scoish Nationalist Party
(SNP), Plaid Cymru (PC), Green Party, United Kingdom Independence Party
(UKIP), BNP, and “other,” DK/ No party.
2015: (YouGov Wave 5 Online panel) “Generally speaking, do you
think of yourself as Labour, Conservative, Liberal Democrat, (Scottish
National/ Plaid Cymru) [in Scotland/ Wales] or what?” None/ No;
Labour; Conservative; Liberal Democrat; “Other” combines Scottish
Nationalist Party (SNP), Plaid Cymru (PC), Green Party, United Kingdom
Independence Party (UKIP), BNP, and “other”; None/ No and Don’t know;
Refused.
2015:(BES CSES module) “W hich party do you feel closest to?” Conservatives;
Labour; Liberal Democrats; “Other” combines UKIP; e Green Party; SNP;
Plaid Cymru; and “other”; Don’t know; Refused/ No answer.
SEX
2005 and 2010: Survey Company supplied data, questions not listed in survey
documentation. Response categories— 1 Male; 2 Female.
2015: (BES CSES module) Are you:Male; Female?
2015: (YouGov Wave 5 Online panel) “Are you male or female?” 1 Male; 2
Female.
AGE
2005 and 2010: Survey Company supplied data as categories, questions not
listed in survey documentation.
2015: (BES CSES module) “What was your age last birthday?” Enter years.
2015:(YouGov Wave 5 Online panel) “What is your age?”
EDUCATION
2005 and 2010:Terminal age of education
2015: (BES CSES module) “What is the highest qualication you have?”
Recoded to correspond to 3 categories of 2005 and 2010:
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 4.2 243
243
No qualication; GCSE A*– C, CSE grade 1; O level grade A– C; Scoish
Standard grades, Ordinary band; GCSE D– G, CSE grades 2– 5, O level D– E;
City & Guilds level 2, NVQ/ SVQ 2 and equi; City & Guilds level 1; NVQ/ SVQ
1 and equiv; Clerical and commercial qualications; Recognized trade appren-
ticeship; Youth training certicate, skill seekers=16years or less:
A level or equivalent; Scoish Higher or equivalent; ONC/ OND, City &
Guilds level 3,
NVQ/ SVQ 3.=17– 18years:
Postgraduate degree, rst degree, Univ/ poly diploma, Teaching qualica-
tion, Nursing qualication, HNC/ HND, City&Guilds level 4, NVQ/ SVQ 4/
5=19years +
(Note:Category of “Other technical, professional or higher qualication”
was coded as missing given the uncertainty over which level it corresponded to,
N=45.
VOTE
2005:Vote in 2005 not asked.
2010:Two- stage lter question:
“Whenever there is an election, some people decide that they have good rea-
sons not to vote, other people want to vote but are unable to, and some people
vote. inking of the recent general election on May 6th, which of the following
statements best describes you? Yes, Ivoted; No- Did not vote; No- Not eligible
to vote; Don’t know.”
If Yes:“Please can you tell me, which party you voted for in the general elec-
tion?” Labour; Conservative; Liberal Democrat; “Other” combines SNP, PC
Green Party, United Kingdom Independence Party (UKIP), British National
Party (BNP), Respect and “other”; Refused coded as missing.
2015:(BES CSES module) “Talking with people about the general election
on May 7th, we have found that a lot of people didn’t manage to vote. How about
you, did you manage to vote in the general election?”
If Yes: “Which party did you vote for in the general election?” Labour;
Conservative; Liberal Democrat; “Other” combines Scoish Nationalist Party
(SNP), Plaid Cymru (PC), Green Party, United Kingdom Independence Party
(UKIP), BNP, and “other”; Refused coded missing.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
244
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
245
245
APPENDIX 5.1
Tables5.2– 5.4, Data Sources and
VariableDefinitions
e results reported in Tables5.2, 5.3, and 5.4 use data from two surveys cover-
ing the 2010 and 2013 Australian Federal elections.
2010
For 2010 we used data from the ESRC- funded project “e Internet, Electoral
Politics and Citizen Participation in Global Perspective.” e questions were
elded as a module within the 2010 Australian Election Study (AES) which was
conducted by the Australian National University and the Social Research Centre
(SRC). Version 2.0 of the survey was used. e primary mode was a postal sur-
vey with a secondary online completion option. Respondents were recruited in
two waves. e rst wave were recruited by random sampling methods (strati-
ed by state) and sent the survey/ link to the online survey via mail; to correct
an age bias (under- representation of young people) in the mail- back survey, a
second “top up” wave of online respondents were recruited by telephone (using
a combination of re- contact lists from previous Australian National University
phone- based projects) and through the MyOpinions online panel database.
Fieldwork for wave 1 took place from August 23, 2010, to November 24, 2010.
Fieldwork for wave 2 took place from January 25, 2011, to February 7, 2011. e
response rate for wave 1 was 42.5percent (calculated aer removing out of scope
from the sample, i.e., deceased, incapable, return to sender); for wave 2 respon-
dents recruited by phone, the online survey completion rate was 37.3percent
and for the MyOpinions panel 8.1percent. Total of 2,214 surveys completed,
376 (16.9percent) completed online.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
246 Appendix 5.1
246
Weight2 supplied by SRC. Based on gender (national 18+ gender distribu-
tion), age (actual enrolled population), State/ Territory (actual enrolled popula-
tion), 2010 voting behavior (based on voting data provided by the ANU). For
full details of the survey including technical report and data download see
hps:// australianelectionstudy.org/ voter- studies/
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
All questions were measured as binary 1=yes have done; else=0. e questions
used to measure the three main types of engagement were as follows:
Read
E- news:“Did you follow the election campaign news on the internet?” Agood
deal/ some/ not much recoded=1.
Campaign sites:“During the 2010 election campaign, did you read or access
party or candidate campaign sites (e.g,. home pages, ocial Facebook proles,
ocial YouTube channels)?”
Redistribute
Signup / Use Party Tools:During the 2010 election campaign, did you do any of
the following activities online:Signed up to receive information from a party
or candidate and/ or registered as a follower/ friend/ supporter OR Used online
tools to help parties and candidates (e.g., forwarded/ shared/ reposted campaign
information)?”
Share/ Exchange: “During the 2010 election campaign, did you do any of the
following activities online: Posted comments on a blog, twier feed, or wall
of a social network site (either yours or someone else’s) OR Shared unocial
campaign content (e.g., links to videos, news stories) with others via email,
Facebook, twier OR Reposted unocial campaign content (e.g.. blog posts,
links to videos) on your own pages (Facebook or twier prole, blog)?”
Receive
Receive Online Direct:“During the election campaign, did a candidate or anyone
from a political party contact you to persuade you to vote for them?” If Yes:“By
email or through the web?”
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 5.1 247
247
Receive Oine Direct:During the election campaign, did a candidate or anyone
from a political party contact you to persuade you to vote for them?” If Yes:“By
telephone or mail or face- to- face?”
2013
For 2013 we used the 2013 Australian Election Study (AES) conducted by
the Australian National University and the Social Research Centre (SRC).
Respondents were recruited by random sampling methods (stratied by state).
Mode was a postal survey with online completion option; eldwork took place
from 6th September 6, 2013, to January 6, 2014. Total of 3,955 surveys com-
pleted, 576 (14.6percent) completed online. e response rate was 34.2percent
(calculated aer removing out of scope from the sample, i.e,. deceased, incapa-
ble, return to sender). Weight supplied by SRC. Based on:Sex, Age, and State
(based on AEC enrolment data for the 2013 election) and party vote (based on
AEC nal election vote tallies).
For full details of the survey, including technical report and data download, see
hps:// australianelectionstudy.org/ voter- studies/
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
All questions were measured as binary 1=yes have done; else=0. e questions
used to measure the three main types of engagement were as follows:
Read
E- news:“Did you follow the election campaign news on the internet?” Agood
deal/ some/ not much recoded=1.
Campaign sites:“During the 2013 election campaign, did you read or access any
of the following:Party or candidate campaign sites (e.g., home pages, blogs, o-
cial Facebook proles, ocial YouTube channels)?”
Redistribute
Signup / Use Party Tools:“During the 2010 election campaign, did you do any
of the following activities online:Signed up to receive information from a party
or candidate and/ or registered as their follower/ friend/ supporter on Twier or
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
248 Appendix 5.1
248
Facebook OR Used online tools to promote parties and candidates (e.g., shared,
posted or reposted ocial campaign information on a blog, Twier feed or
social network prole)?”
Share/ Exchange: “During the 2010 election campaign, did you do any of the
following activities online:Shared, posted, or reposted any non- ocial content
(e.g., links to videos, news stories, jokes) on a blog, Twier feed, or social net-
work prole?”
Receive
Receive Online Direct:“During the election campaign, did a candidate or anyone
from a political party contact you to persuade you to vote for them?” If Yes:“By
text message OR SMS by email OR by social network site or other web- based
method?”
Receive Online Indirect: “During the campaign, did a friend, family member,
neighbor, work colleague, or other acquaintance try to persuade you to vote for
a particular party or candidate?” If Yes:“By text message OR SMS by email OR
by social network site or other web- based method?”
Receive Oine:“During the election campaign, did a candidate or anyone from
a political party contact you to persuade you to vote for them?” If Yes:“By tele-
phone OR by mail OR face- to- face?”
Socio- Demographic and PoliticalVariables
PARTYID
2010 and 2013: “Generally speaking, do you usually think of yourself as Liberal,
Labor, National or what?”
Liberal; Labor National (Country) Party; Greens; Other party (please specify);
No party.
SEX
2010 and 2013:“Are you male or female?” Male=1; Female=2.
AGE
2010 and 2013:“In what year were you born?”
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 5.1 249
249
EDUCATION
2010 and 2013:“Have you obtained a trade qualication, a degree or a diploma, or
any other qualication since leaving school? What is your highest qualication?”
Up to Secondary:No qualication since leaving school;
Diploma:Associate Diploma; Trade qualication; Non- trade qualication;
Higher Education:Postgraduate Degree or Postgraduate Diploma; Bachelor
Degree, Undergraduate Diploma.
SOCIALCLASS
2010 and 2013:“Which social class would you say you belong to?”
Upper class OR Middle class; Working class; None.
VOTE
2010:“In the last federal election in August 2010, when Labor was led by Julia
Gillard and the Liberals by Tony Abbo, which party got your rst preference
then in the House of Representatives election?” Liberal Party; Labor Party
(ALP); National (Country) Party; Greens; Other (please specify party below);
Voted Informal/ Did not vote.
Note that in Table 5.4 Liberal and National votes are combined as
Coalition vote.
2013:“In the federal election for the House of Representatives on Saturday
September 7, which party did you vote for rst in the House of Representatives?”
Liberal Party; Labor Party (ALP); National (Country) Party; Greens; Other
(please specify party below); Voted Informal/ Did not vote.
Note:in table5.4 Liberal and National vote are combined as Coalition vote
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
250
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
251
251
APPENDIX 6.1
Tables6.2– 6.4, Data Sources and
VariableDefinitions
For results reported in Tables6.2, 6.3, and 6.4, use two main data sources cor-
responding to the 2007 and 2012 presidential elections.
2007
For 2007 the data were made available by Professor ierry Vedel and the
Centre for Political Research CEVIPOF at Sciences Po, Paris, France. e sur-
vey was conducted by the French market research company L’Institut Français
d’Opinion Publique (IFOP) using an Internet panel, and it was part of their
“Observatoire de la Netcampagne.” e survey was elded in April 2007. e
overall sample was 1,004 and consisted of French adult internet users aged
18years or older. Quotas for sex, age, PCS, and région*agglo were applied to
sample recruitment. PCS refers to the nine- category “Occupational and Socio-
occupational Categories” developed by French National Institute of Statistics
and Economic Studies (INSEE). Region *agglo refers to selection using a com-
bination of region and local community size, i.e., to ensure the sample is matched
to the dispersion of the French population. e data were weighted by IFOP to
ensure that demographic proles matched those for all adults in France age 18
or older.
READ AND REDISTRIBUTE VARIABLECONSTRUCTION
All questions were measured as binary 1=yes have done; else=0. e questions
used to measure the three main types of engagement were as follows:
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
252 Appendix 6.1
252
Read
E- news:“Au cours de la campagne présidentielle, vous est- il arrivé souvent,
de temps en temps:Rechercher des informations sur actualité politique
sur Internet?” (During the presidential campaign did you often, from time
to time, rarely or never: Search for information on political news on the
Internet)
Oen, time to time, and rarely=1; never=0.
Campaign sites:“Au cours de la campagne présidentielle, vous est- il arrivé
souvent, de temps en temps:Visiter le site d’un candidat à l’élection présidenti-
elle?” (Visit the sites of presidential election candidates)
Redistribute
Sign up/ Use Party Tools:Was measured in two stages.
Stage 1— a lter questions was used:
“Quelles sont vos trois principales sources d’information politique sur
Internet?” (What are your three main sources of political information on the
Internet)
Les sites d’information des chaînes de télévision (TV news channels)
Les sites d’information des radios (Radio news channels)
Les sites d’information de la presse écrite (Newspapers, print media)
Les portails d’information généralistes (General information portals)
Les sites ou les blogs de personnalités politiques (Sites or blogs of political
gures)
Les sites ou les blogs de journalistes politiques (Sites or blogs of political
journalists)
Les sites ou les blogs de citoyens (Sites or blogs of citizens)
Les sites de formations ou de partis politiques (Political party sites)
Les forums de discussion politique (Political discussion foruma)
Autres (Other).
Stage 2— All those reporting political party sites as one of their three main
sources of information online group were selected and the frequencies reported
for this group on whether they:
“Télécharger des argumentaires politiques, des textes ou des tracts poli-
tiques?” (Downloaded political arguments, texts or political leaets)
Share/ Exchange:“Transférer à des proches des informations sur la campagne
présidentielle par email?” (Transferred information on the presidential cam-
paign by email to relatives)
Redistribute = all those selected in stage 1 who had downloaded content
AND shared the content.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 6.1 253
253
2012
For 2012 we used data from the ESRC- funded project “e Internet, Electoral
Politics and Citizen Participation in Global Perspective.” e questions were
included as a module within the French National Election Study (FNES) con-
ducted by the Fondation Nationale des Sciences Politiques. e eldwork was
carried out by TNS Sofres using a stratied three- stage probability sample
(region, household, individual) and was conducted aer the second round of
the election. e overall sample was of 2,014 French citizens aged 18years or
older. e data was weighted to ensure that demographic proles matched to
the population on socio- demographics (sex, age, and occupation). Overseas ter-
ritories are not included. Interviews were conducted face- to- face between May
10 and June 9, 2012. For further details on the study design, see the CSES Macro
report at hps:// cses.org/ datacenter/ module4/ macro/ F_ 2012_ Macro.pdf
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
All variables were created in an identical manner as for the UK 2010 election,
using the BMRB survey. Refer to Appendix 4.1, Election 2010, for the English
translations. All questions were measured as binary 1=yes have done; else=0.
e questions used to measure the three main types of engagement were as
follows:
Read
E- news: “Lu ou accédé à des sites d’actualité généralistes ou à des blogs sur
l’actualité pour avoir des informations sur la campagne?”
Campaign sites:“Est- ce que vous avez lu ou accédé à un site Internet d’un parti
ou d’un candidat?”
Redistribute
Sign up/ Use Party Tools:“Est- ce que vous vous êtes inscrits comme soutien, ami
ou ‘follower’ d’un parti ou d’un candidat sur leur site Internet ou sur les réseaux
sociaux OR utilisé des moyens permis par Internet pour aider un parti ou un
candidat dans leur campagne?”
Share/ Exchange:“Rejoint ou créé un groupe sur la politique ou les élections sur
les réseaux sociaux? OR Posté des commentaires avec un contenu politique sur
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
254 Appendix 6.1
254
un blog ou le mur d’un réseau social OR Fait suivre du matériel de campagne
non ocielle (liens avec des vidéos, actualités, blagues, etc.) à certaines de
vos connaissances OR Inclus ou reposté du contenu de campagne non o-
cielle (liens avec des vidéos, actualités, blagues, etc.) sur vos propres pages
Internet?”
Receive
Receive Online Direct:“Au cours de la campagne pour l’élection présiden-
tielle, candidat a- t- il pris contact avec vous, que ce soit en personne ou
par d’autres moyens? Vous ont- ils contacté par SMS? par courrier électro-
nique; par l’intermédiaire de réseaux sociaux ou toute autre méthode sur
Internet?”
Receive Online Indirect: “Pendant la campagne, est- ce que l’un de vos amis, un
member de votre famille, un voisin, un collègue ou une autre de vos connais-
sances a essayé de vous convaincre de voter pour un candidat particulier? Vous
ont- ils contacté par SMS? par courrier électronique; par l’intermédiaire de
réseaux sociaux ou toute autre méthode sur Internet?”
Receive Oine Direct:“Au cours de la campagne pour l’élection présidentielle,
un parti ou un candidat a- t- il pris contact avec vous, que ce soit en personne
ou par d’autres moyens? Vous ont- ils contacté en personne, en face- à- face?; par
courier?; par téléphone?”
Socio- Demographic and PoliticalVariables
PARTYID
2007: “De quelle formation politique vous sentez- vous le plus proche ou
le moins éloigné?” (Which political party do you feel closest to or the least
distant from?); Parti socialiste (PS); Union for Popular Movement (UMP);
Union for French Democracy (UDF); “Other” combines Front National,
Lue ouvrière, Ligue communiste révolutionnaire, Parti communiste,
Mouvement pour la France, Une autre formation politique (other); “Aucune”
(none/ no party).
2012:A lter question was used. e rst part asked “Existence d’un parti ou
mouvement poiltique proche?” (In general, is there a party or a political move-
ment that seems more close to you than others?) If No:“Existence d’un parti
moins eloigne que d’autres?” (Is there nevertheless a political party that you feel
less distant to than others).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 6.1 255
255
Options were— Parti Socialist; L’UMP; MoDem; Parti de gauche; Front
National; “Other” combines Lue ouvrière ou NPA (Nouveau parti anticapitali-
ste), Parti communiste, Europe Ecologie— Les Verts, Nouveau centre, Other;
DK/ Refused
SEX
Interviewer instructed to code the gender of the respondent. 1 = Male;
2=Female.
AGE
“Can you tell me your date of birth?” Month and Year.
EDUCATION
2012: “Quel est le plus haut niveau d’études ou de formation que vous avez
terminé?” (What is the highest level of education or training you have
completed?)
Up to Secondary =
1. Non scolarisé ou école primaire non achevée
2. Ecole primaire uniquement
3. Certicat d’études primaires
4. Scolarité suivie de la 6ème à la 3ème
5. Brevet élémentaire, Brevet d’études du premier cycle, Brevet des collèges
6. Scolarité suivie de la seconde à la terminale
7. CAP, BEP, examen de n d’apprentissage artisanal
8. Diplôme d’aide- soignante, auxiliaire de puériculture, aide médico-
pédagogique, aide à domicile.
(No schooling or primary school not completed; only primary school, pri-
mary school certicate; schooling from 6th to 9th grade; elementary certicate;
Certicate of Primary Studies, GCSE; schooling from 10th to 12th grade; CAP,
BEP, examination for cra apprenticeship; diploma nursing auxiliary, auxiliary
childcare, medical and educational assistance, home help)
Baccalaureate =
9. Baccalauréat professionnel, Brevet de technicien
10. Baccalauréat technologique, Baccalauréat de technicien, BEA, BEC,
BEI, BES
11. Baccalauréat général, Brevet supérieur
12. Diplôme de la capacité en droit, Diplôme d’accès aux études
universitaires (DAEU)
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
256 Appendix 6.1
256
13. Diplôme de moniteur éducateur, Educateur technique spécialisé, Brevet
professionnel
(Professional baccalaureate; technical certicate; technical baccalaureate;
Bachelor of Technician; BEA, BEC, BEI, BES; General baccalaureate; higher
certicate; degree in law; degree of access to university (DAEU); diploma moni-
tor educator; educator specialized technical; professional certicate)
Higher Education =
14. Diplôme universitaire du premier cycle (DEUG), Classes
préparatoires aux grandes écoles
15. Diplôme universitaire de technologie (DUT), Brevet de technicien
supérieur (BTS)
16. Certicat d’aptitude pédagogique (instituteur), Diplôme d’éducateur
spécialisé, Diplôme
d’assistante sociale, Diplôme paramédical (laboratin, inrmier,
17. Licence professionnelle
18. Licence
19. Diplôme d’école d’ingénieur
20. DESS, Master deuxième année professionnel
21. Diplômes professionnels supérieurs divers (notaire, architecte,
journaliste,...)
22. Diplôme des grandes écoles
23. Maîtrise, CAPES, CRPE (professeur des écoles)
24. DEA, DES, Master deuxième année recherche, Agrégation
25. Doctorat en médecine ou équivalents (dentaire, pharmacie, etc.)
26. Doctorat
27. Autres
(Undergraduate degree (DEUG); classes préparatoires aux grandes écoles;
University Diploma in Technology (DUT); higher technician certicate
(BTS); teaching certicate (teacher); special education diploma; degree in
social work; paramedical Diploma (lab, nurses, etc.), Professional Bachelor,
Bachelor, Engineering school diploma; DESS; Second year professional master;
Various professional degrees (notary, architect, journalist, etc.), Grandes Ecoles
diploma; Maîtrise; CAPES, CRPE(school professor); DEA, DES; Second year
of reseearch master, aggregation; Doctorate in medicine (dentistry, pharmacy,
etc.); Doctorate; Others)
Refused, DK/ NA=missing.
INCOME
2012:“Si vous additionnez toutes les sources de revenus de votre
foyer, quelle lere correspond le mieux au revenu net mensuel de votre foyer?
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 6.1 257
257
Pour répondre, veuillez m’indiquer la lere qui correspond à votre réponse?”
(If you add up all sources of income of your household, which of the follow-
ing best corresponds to the net monthly income of your household?
1. H— Moins de 7,500 euros
2. U— De 7,500 à 15,000 euros
3. A— De 15,001 à 30,000 euros
4. N— De 30,001 à 50,000 euros
5. E— De 50,001 à 75 000 euros
6. Z— De 75,001 à 150,000 euros
7. P— De 150,001 à 300,000 euros
8. B— De 300,001 à 450,000 euros
9. S— De 450,001 à 750,000 euros 10. R— Plus de 750,001 euros
(Original 10 categories recoded to ve to ensure sucient N for subsequent
crosstabs)
Recoded:1=Less than 1,000 euros per month; 2=From 1,001 to 2,000 euros
per month; 2,001 to 3,000 euros per month; 3=From 3,001 to 4,000 euros per
month; 4=More than 4,001 euros per month; Refused/ Don’t know=missing;
Null=no income (not reported in table, N=11).
VOTE
2012: Measured as a two- stage lter question: “Beaucoup d’électeurs n’ont
pas voté au premier tour de l’élection présidentielle. Vous- même, pouvez- vous
me dire si vous avez voté au premier tour de l’élection présidentielle le 22 avril
dernier?” (Many voters did not vote in the rst round of the presidential elec-
tions. Can you tell me if you voted in the rst round of presidential elections on
April 22?) If Yes:“Pour qui avez- vous voté?” (For whom did you vote?) François
Hollande; François Bayrou; Nicolas Sarkozy Jean- Luc Mélenchon; Marine Le
Pen; “Other” combines Nathalie Arthaud, Philippe Poutou, Eva Joly, Nicolas
Dupont- Aignan, Jacques Cheminade. Ref, DK=missing.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
258
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
259
259
APPENDIX 7.1
Tables7.2– 7.4, Data Sources and
VariableDefinitions
e results reported in Tables7.2, 7.3, and 7.4 are from three main data sources
corresponding to the 2004, 2008, and 2012 presidential elections.
2004
For 2004 we used data from the Pew Internet and American Life Project—
November 2004 Daily Tracking Survey. e survey was conducted using standard
list- assisted random digit dialing (RDD) methodology with a nationally repre-
sentative sample of 2,200 adults living in continental United States telephone
households. Up to 10 aempts were made to contact every sampled telephone
number. e nal response rate for this survey was 30.6percent. e interviews
were conducted in English by Princeton Data Source, LLC, from November 4
to November 22, 2004. Statistical results are weighted to correct known demo-
graphic discrepancies. Further details can be found at hps:// w ww.pewresearch.
org/ internet/ 2005/ 03/ 06/ the- internet- and- campaign- 2004/ Data available at
hp:// www.pewinternet.org/ dataset/ postelection- 2004- tracking- survey/
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
Responses were coded as binary (done/ not done) unless otherwise stated.
Don’t know/ Refused coded as missing.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
260 Appendix 7.1
260
Read
E- news:“Please tell me if you ever do any of the following when you go online.
Do you ever:Look for news or information about politics and the campaign?”
Campaign sites:“Now thinking about some campaign websites, do you ever go
onto to (a)the Kerry/ Edwards OR (b)the Bush/ Cheney campaign website to
get news or information about the 2004 elections?”
Redistribute
Sign up/ Use Party Tools:“Have you ever signed up to receive email newslet-
ters or other online alerts containing the latest news about politics or the
election?” OR
“During this year’s election, did you happen to sign up ONLINE for any
VOLUNTEER activities related to the campaign— like helping to organize a
rally, register voters, or get people to the polls on election day— or did you not
sign up online for any volunteer activities?”
Share/ Exchange:“Have you sent emails about the 2004 campaign to groups of
family or friends who are part of an email list or online discussion group?”
Receive
Receive Online: In the past two months, have you received EMAIL urging you to
vote for a particular presidential candidate?” If Yes:“Was that urging you to vote
for Bush, for Kerry, some other candidate, or multiple candidates?” Response
categories were mutually exclusive and thus cumulated to calculate overall
frequencies.
Receive Oine: “In the past two months, have you received mail, or telephone
calls, or been visited at home by someone urging you to vote for a particular
presidential candidate?” Responses summed to=total oine contact.
2008
For 2008 we used data from the Pew Internet and American Life Project ‘e
Internet’s Role in Campaign 2008’ November 2008 Post Election Tracking
Survey. e survey was conducted using standard list- assisted random digit dial-
ing (RDD) methodology with a nationally representative sample of 2,254 adults
living in continental United States telephone households. Up to 10 aempts were
made to contact every sampled telephone number. e nal response rate for this
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 7.1 261
261
survey was 23percent. e interviews were conducted in English by Princeton
Survey Research Associates International from November 20 to December 4,
2008. Statistical results are weighted to correct known demographic discrepan-
cies. Further details can be found at hps:// www.pewresearch.org/ internet/
2009/ 04/ 15/ the- internets- role- in- campaign- 2008/ . Data available at hp://
www.pewinternet.org/ dataset/ november- 2008- post- election/
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
Responses were coded as binary (done/ not done) unless otherwise stated.
Don’t know/ Refused coded as missing.
Read
E- news:“Please tell me if you ever use the internet to do any of the following
things? Do you ever:Look for news or information about the 2008 campaign?”
Campaign sites:“Now thinking about some campaign websites, did you ever go
to the (a)Obama/ Biden OR (b)McCain/ Palin campaign website to get news
or information about the 2008 elections?”
Redistribute
Sign up/ Use Party Tools:“ere are many dierent activities related to the cam-
paign and the elections that a person might do on the internet. I’m going to read
a list of things you may or may not have done online in the past year related to
the campaign and the elections. Did you:Sign up online to receive updates about
the campaign or the elections OR Sign up ONLINE for any VOLUNTEER
activities related to the campaign— like helping to organize a rally, register vot-
ers, or get people to the polls? OR Signed up as a ‘friend’ of any of candidates on
a social networking site?”
Share/ Exchange:As above, “Did you:Share photos, videos or audio les online
that relate to the campaign or the elections OR Forward someone else’s political
commentary or writing to others OR Forward someone else’s political audio or
video recordings to others?”
Receive
Receive Online Direct: “inking about this year’s presidential election, people
have been communicating with each other and with the political campaigns
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
262 Appendix 7.1
262
in many ways, to talk about issues or where the campaign sends. What about?
Over the past several months, how oen did you: Receive EMAIL / TEXT
MESSAGES from a candidate or political party?”
Receive Online Indirect: As above, “Send or Receive EMAIL/ TEXT MESSAGES
with friends, family members, or others about the campaign?”
Receive Oine Direct: As above, “Receive MAIL from a candidate or party?”
2012
For 2012 we used data from the 2012 ANES Time Series Study (ANES, 2014).
e study was a two- wave pre- and post- election survey and consisted of two
probability- based samples: (1) Face to face (F2F), N= 2,054; (2) Internet
panel, N=3,860. Respondents were recruited separately, and in both cases the
sample universe were US eligible voters. e F2F sample was recruited using
an address- based, stratied, multistage cluster sample in 125 census tracts. e
internet sample was drawn from panel members of GfK Knowledge Networks.
Fieldwork began in September 2012 and concluded in January 2013. Pre- election
interviews were conducted with study respondents during the two months prior
to the 2012 elections and were followed by post- election re- interviewing begin-
ning November 7, 2012.
READ, REDISTRIBUTE, AND RECEIVE
VARIABLECONSTRUCTION
Responses were coded as binary (done/ not done) unless otherwise stated.
Don’t know/ Refused coded as missing.
Read
Enews (Post- election):“Did you read, watch, or listen to any information about
the campaign for president on the internet?”
Campaign sites (Post- election):“Did you visit any presidential candidates’ web-
sites, or did you never do that?”
Redistribute
Sign up: “Prior to or during the campaign, did you use the internet or your
mobile phone to sign up for information or alerts from a party or candidate?”
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 7.1 263
263
Share/ Exchange (Post- election):“During the past four years, have you ever sent
a message on Facebook or Twier about a political issue, or have you not done
this in the past four years?”
Receive
Receive Online Direct: “During the election campaign, did a candidate or anyone
from a political party contact you to persuade you to vote for them?” If Yes:“By
text message OR SMS by email OR by social network site or other web- based
method?”
Receive Online Indirect: “During the campaign, did a friend, family member,
neighbor, work colleague, or other acquaintance try to persuade you to vote for
a particular party or candidate?” If Yes:“Did they try to persuade you by text or
SMS; by email; through a social network site or other web- based method?”
Receive Oine Direct: “During the campaign, did a party or candidate contact
you in person or by any other means?” If Yes:“Did they try to persuade you in
person, face- to- face; by mail; by phone?”
Socio- Semographic and PoliticalVariables
PARTYID
2004 and 2008: “In politics today, do you consider yourself a Republican,
Democrat, or Independent?” No party/ Not interested in politics; Other party;
Don’t know/ Refused.
2012: “Generally speaking, do you usually think of yourself as a Democrat, a
Republican, an Independent, or what?”
SEX
2004 and 2008: Interviewer Recorded:1=Male; 2=Female.
2012:Interviewer Recorded (F2F):1=Male; 2=Female.
Internet sample:“Are you male or female?”
AGE
2004 and 2008:“What is your age?” Recorded in years (97=97 or older); Don’t
know/ Refused (treated as missing data).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
264 Appendix 7.1
264
2012:“What is the month, day, and year of your birth?” Author recoding of
ANES summary variable dem_ agegrp_ iwdate_ x
EDUCATION
2004 and 2008:“What is the last grade or class you completed in school?”
Less than High school— None, or grades 1– 8/ High school incomplete
(grades 9– 11);
High school graduate— (grade 12 or GED certicate);
Some college— Technical, trade, or vocational school AFTER high school;
Some college, no four- year degree (includes associate degree);
College— graduate (BS, BA, or other four- year degree); post- graduate
training/ professional school aer college (toward a Masters/ PhD, law or
medical
school);
Don’t know/ Refused=Missing.
2012:“What is the highest level of school you have completed or the highest
degree you have received?”
Less than high school;
High school credential;
Some college— some post high school no bachelor’s degree;
College— Bachelor’s degree/ Graduate degree.
Author recoding of ANES summary variable dem_ edugroup_ x
INCOME
2004 and 2008: “Last year, that is in 2003, what was your total family
income from all sources, before taxes? Just stop me when Iget to the right
category...”
1 Less than $10,000
2 $10,000 to under $20,000
3 $20,000 to under $30,000
4 $30,000 to under $40,000
5 $40,000 to under $50,000
6 $50,000 to under $75,000
7 $75,000 to under $100,000
8 $100,000 or more
9 Don’t know/ Refused
Recoded:1 to 2 =<20,000; 3 to 5= 20,000– 49,999; 6= 50,000– 74,999;
7=75,000– 100,000; 8=>100,000; 9=missing.
2012: “Information about income is very important to understand how peo-
ple are doing nancially these days. Your answers are condential. Would you
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 7.1 265
265
please give your best guess? Please mark the answer that includes the income of
all members of your family living here in 2011 before taxes.”
Author recoding of ANES summary variable inc_ incgroup_ pre
RACE
2004 and 2008: Variable “racethn” used. is is a combination of two variables:
1. “Are you, yourself, of Hispanic or Latino origin or descent, such as Mexican,
Puerto Rican, Cuban, or some other Latin American background?”
2. “What is your race? Are you White, Black, Asian, or some other race?”
“Other” includes Asian or Pacic Islander, Mixed race, Native American/
American Indian, and “Other”; Don’t know/ Refused coded as missing.
2012:Author recoding of summary variable dem_ raceeth_ x.— White, non-
Hispanic; Black, non- Hispanic; Hispanic; Other (Asian, native Hawaiian or
other Pacif Islr, non- Hispanic, Native American or Alaska Native, non- Hispanic,
other non- Hispanic incl multiple races).
VOTE
2004 and 2008:Two- stage lter question:
“A lot of people have been telling us they didn’t get a chance to vote in the
elections this year on November 2.How about you— did things come up that
kept you from voting, or did you happen to vote?” Yes, voted; No, did not vote;
Don’t know/ Can’t remember/ Refused=missing.
2004:If Yes:“Did you vote for:the Republican ticket of George Bush and
Dick Cheney; the Democratic ticket of John Kerry and John Edwards?” “Other”
and Don’t know/ Refused coded as missing.
2008: If Yes:“Did you vote for:the Democratic ticket of Barack Obama and
Joe Biden, or the Republican ticket of John McCain and Sarah Palin?” “Other,”
Don’t know/ Refused coded as missing.
2012:“For whom did R vote for President? Barack Obama; Mi Romney?”
“Other” & did not vote=missing. Author recoding of ANES summary variable
Presvote2012_ x
Data SourcesBibliography
COMPARATIVE
Gibson, Rachel K.2013. e Internet, Electoral Politics and Citizen Participation in
Global Perspective. [data collection]. UK Data Service. SN:850856, hp:// doi.
org/ 10.5255/ UKDA- SN- 850856
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
266 Appendix 7.1
266
e Comparative Study of Electoral Systems (www.cses.org). CSES Module
4 e Comparative Study of Electoral Systems (www.cses.org). CSES Module 4
Second Advance Release [2.0]. March 20, 2015 Version. Final release available
at hp:// www.cses.org/ datacenter/ module4/ module4.htm
UNITEDSTATES
Pew Internet and American Life Project. 2004. Post- Election 2004 Tracking
Dataset (November). hps:// www.pewresearch.org/ internet/ dataset/
postelection- 2004- tracking- survey/
Pew Internet and American Life Project. 2008. Post- Election 2008 Tracking
Dataset (November). hps:// www.pewresearch.org/ internet/ dataset/
november- 2008- post- election/
ANES. 2014. User’s Guide and Codebook for the ANES 2012 Time Series
Study. Ann Arbor, MI, and Palo Alto, CA:University of Michigan and Stanford
University.
AUSTRALIA
McAllister, Ian, Clive Bean, Rachel Kay Gibson, and Juliet Pietsch. 2017.
“Australian Election Study, 2010”, doi:10.4225/ 87/ CYJNSM, ADA Dataverse,
V2, UNF:6:3iyzr2dBihOrVkbaf FkRZA==
Bean, Clive, Ian McAllister, Juliet Pietsch, and Rachel Kay Gibson. 2017.
“Australian Election Study, 2013,” doi:10.4225/ 87/ WDBBAS, ADA Dataverse,
V3, UNF:6:6gMySFLvbEH1ccG58om4Sg==
UNITEDKINGDOM
Ward, Stephen, and Rachel Gibson. 2005. Economic and Social Research
Council Funded Study. RES- 000- 22- 1284— Campaigning in Cyberspace:e
2005/ 6 General Election Online.
Fieldhouse, Ed, Jane Green, Geo Evans, Hermann Schmi, Cees van der
Eijk, Jonathan Mellon, and Chris Prosser. 2015. British Election Study Internet
Panel Wave 5.DOI:10.15127/ 1.293723
Fieldhouse, Ed, Jane Green, Geo Evans, Hermann Schmi, Cees van
der Eijk, Jonathan Mellon, and Chris Prosser. 2016. British Election Study,
2015:Face- to- Face Post- Election Survey. [data collection]. UK Data Service.
SN:7972, hp:// doi.org/ 10.5255/ UKDA- SN- 7972- 1.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Appendix 7.1 267
267
FRANCE
L’Institut Français d’Opinion Publique (IFOP) Survey on Internet population
(Observatoire de la Netcampagne), conducted 11.2006, 04.2007, 06.2009
e French Election Study 2012. Conducted by TNS- Sofres. Archived at the
Socio- political Data Centre (CDSP), Sciences Po/ CNRS. Available at hps://
cdsp.sciences- po.fr/ fr/ ressources- en- ligne/ ressource/ fr.cdsp.ddi.FES2012/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
268
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
269
269
Notes
Chapter 1
1. See Barber’s (1997) pessimistic prediction for example that “Le to markets, (and that is
where it is presently being le), it islikely to augment McWorld’s least worthy imperatives,
including surveillance over and manipulation of opinion, and the cultivation of articial
needs rooted in life- style “choices” unconnected to real economic, civic or spiritual needs.”
(225); and Barber’s (1998) outline of the “Pandora Scenario” whereby new technolo-
gies...give government instruments of indirect surveillance and control unlike any known
to traditional dictators.” (580).
2. ese data sets were generated as part of a broader cross- national research project led by the
author in the Economic and Social Research Council (ESRC) funded project “e inter-
net, electoral politics and citizen participation in global perspective,” RES- 051- 27- 0299.
Chapter 2
1. e quote appears in Ho and Lofgren (1997:6). is was a pilot study on the use and
strategy of Danish political parties and featured interviews with the “ICT- political spokes-
persons” from four parties, including the Liberals.
2. James Crabtree quotes Francis Maude, Conservative Party chairman during the early
noughties, as part of his study of the 2010 election; “David Cameron’s Bale to Connect,”
Wired Magazine, March 24, 2010, available at hp:// www.wired.co.uk/ magazine/ archive/
2010/ 04/ features/ david- camerons- bale- to- connect (accessed May 10, 2015).
3. According to World Bank Development Indicators, the average internet use among
member countries of the European Monetary Union (EMU) in 1996 was 2.9% and for
OECD nations 6.1%; hp:// databank.worldbank.org/ data/ reports.aspx?source=world-
development- indicators. e national statistics on growth of internet use in each of case
studies conrm this low starting point. See Figures4.1, 5.1, 6.1, and 7.1 for the United
Kingdom, Australia, France, and the United States, respectively.
4. “Ready on e- Day:A Web Analysis of the Major Campaigns in 2002,” Rightclick strategies,
available at hp:// www.rightclicks.com/ e- day/ e- day.pdf (accessed July 28, 2003).
5. Changes in the Federal Election Commission (FEC) rules about credit card donations
meant that online donations became eligible for match funding. See Bob von Sternberg,
“Cyber Campaign Is Geing Crowded,” Star Tribune, June 3, 1999:A6. Also Klotz (2004).
6. For a fuller denition and understanding of the relationship between web 1.0 and 2.0, see
Anderson (2007), especially pp.5– 6.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
270 Notes
270
7. ere are some studies that counter this general trend. Klinger (2013), for example, con-
cluded that the smaller players in Switzerland did not exploit social media particularly
strongly in the federal elections of 2011.
8. One of the most notable and successful of these eorts had come in the 1998 state- level
campaign by former wrestler Jesse Ventura. Ventura scored a surprise victory in the
Minnesota governorship that was credited to his skillful use of the internet and particularly
discussion forums that he used to generate a large network of young supporters, known
as “Jessenet.” Candidates in the 2000 presidential election, notably Steve Forbes and Pat
Buchanan, had also taken steps to build up e- precincts or “Buchanans’ brigade” (Stromer-
Galley, 2014:37).
9. Peggy Noonan, “ey’ve Lost at Lovin’ Feeling,” Wall Street Journal, July 30, 2011.
10. According to these authors, election campaigning had entered a new “professionalized” or
“postmodern” era by the late twentieth century. While this shi preceded the arrival of
the web, a key characteristic of the new era was its emphasis on careful message targeting
through direct mail methods and multi- channel TV, and a rejection of the “one size ts all”
mode that had dominated in the earlier “modern” era. e arrival of the internet and par-
ticularly tools like email was thus seen as aligning very well with this wider tendency toward
“narrowcasting” and micro- messaging of voters.
11. e US Green Party under Ralph Nader pioneered this strategy during the 2000 presi-
dential election. e party set up a platform that allowed Green and Democrat voters
in dierent districts to trade votes in a bid to unseat Republican incumbents. In the UK
general election of 2001, the Liberal Democrats copied the strategy, using their website to
encourage tactical voting among their own and Labour supporters in order to defeat Tory
candidates.
12. Sasha Issenberg, “A Vast Le- Wing Competency,” Slate Magazine, November 7, 2012,
available at hp:// www.slate.com/ articles/ news_ and_ politics/ victory_ lab/ 2012/ 11/
obama_ s_ victory_ how_ the_ democrats_ burned_ by_ karl_ rove_ became_ the_ party_
of.html (accessed April 8, 2013).
13. e reports are numerous, and Chapter 7 contains a more extensive review. In- depth
coverage can be found in articles by Sean Gallagher, “Built to Win:Deep Inside Obama’s
Campaign Tech,” Ars Technica, November 14, 2012, available at hp:// arstechnica.com/
information- technology/ 2012/ 11/ built- to- win- deep- inside- obamas- campaign- tech/
(accessed November 26, 2014); Christie Parsons and Kathleen Hennessey, “Obama
Campaign’s Investment in Data Crunching Paid O,” Los Angeles Times, November
13, 2012. hp:// articles.latimes.com/ 2012/ nov/ 13/ nation/ la- na- obama- analytics-
20121113 (accessed June 27, 2017).
14. “Inside the Cave:Obama’s Digital Campaign,” a report by Engage Research, available at
hp:// enga.ge/ dl/ Inside_ the_ Cave.pdf (accessed December 1, 2014).
15. According to one report, the campaign predicted the 50:50 split in the vote for Romney
and Obama in Dixville Notch in New Hampshire, a town with less than 100 residents
who have traditionally voted at midnight and provided the rst ocial results. See Christi
Parsons and Kathleen Hennessey, “Obama Campaign’s Investment in Data Crunching Paid
O,” Los Angeles Times, November 13, 2012.
16. Sasha Issenberg, “Obama’s White Whale,” Slate Magazine, February 15, 2012, available at
hp:// www.slate.com/ articles/ news_ and_ politics/ victor y_ lab/ 2012/ 02/ project_ nar-
whal_ how_ a_ top_ secret_ obama_ campaign_ program_ could_ change_ the_ 2012_ race_
.html (accessed July 12, 2013).
17. Frank Mace, “Just How Good Was the Obama Campaign,” Harvard Political Review,
November 29, 2013, available at hp:// harvardpolitics.com/ united- states/ just- good-
obama- campaign/ (accessed March 24, 2014). See also John Sides and Lynn Vavreck,
“Obama’s Not So Big Data Campaign,” Pacic Standard Magazine, January 21, 2014, avail-
able at hps:// psmag.com/ social- justice/ obamas- big- data- inconclusive- results- political-
campaigns- 72687 (accessed February 7, 2015).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Notes 271
271
Chapter 3
1. We use the term “online” rather than “digital contact” in this chapter since a key focus of the
analysis is to compare their frequency and distribution with nondigital forms of contact—
mail, leaets, phone, and in- person. When analyzed in conjunction with digital forms,
these other modes are typically referred to as “oine” and digital as “online.” To maintain
that ease of distinction and consistency with other studies, we adopt this prior terminology.
2. For further details see www.cses.org
3. Note that as of the time of writing, the gures for mail and phone dier from those later
reported in the nal version of the CSES data set, Module 4 Full Release, May 29, 2018
version. e discrepancy appears to be due to the fact that the mail and phone gures in the
full CSES are reported as the proportion of those having reported direct contact, i.e., a lter
has been applied. is does not seem to be the case, however, for the other forms of contact.
4. International Telecommunication Union (ITU) statistics for 2010 report that 44% of the
Greek population were online. is made it the only major Southern European democracy
to report less than 50percent of its population online, with Spain at 66percent and Portugal
and Italy at 54percent. See Excel le, “Percentage of Individuals using the Internet,” avail-
able at hps:// www.itu.int/ en/ ITU- D/ Statistics/ Pages/ stat/ default.aspx.
5. For details on countries’ electoral nance rules and media use during elections, see the
IDEA Political Finance Database, available at hp:// www.idea.int/ data- tools/ data/
political- nance- database.
6. Although the work cited was published in 2009, Anstead and Chadwick had produced an
earlier version of the chapter as a working paper in 2004.
7. is included a number of the key variables already tested in the web campaign models,
such as the level of democratic development, use of a single member/ preferential or pro-
portional voting system, the party system size, and the competitiveness of the race. It also
included some new structural features, such as the extent of ideological polarization within
the party system, the prominence of the election as a presidential or legislative race, the
district- level competitiveness of the race, and if compulsory voting was in place.
8. e multilevel model is tested on 16 and not all 18 countries listed in Table3.1. e United
Kingdom is not included in the nal cross- national analysis since at the point of this analy-
sis it was not part of the release 2.0 of the integrated Wave 4 CSES le. e results reported
here are taken from the 2015 British Election Study (BES) which elded the CSES mod-
ule as part of its mail- back component. Further details of the 2015 UK CSES survey and
sample are provided in Appendix 4.2. Switzerland is not included here, as they did not
include the question about sign- up in their survey and thus we were not able to use it as a
control variable in our model. Finally, the Party System model excluded Taiwan due to the
fact the Le- Right variable used to calculate its score on the Dalton polarization index was
not asked of respondents.
9. e general rule of thumb on the minimum N of level 2 units required for multilevel model-
ing varies, with some scholars arguing as few as 8, while other argue for up 100 (Stegmueller,
2013). Gelman (2006) has even claimed that Bayesian methods can produce unbiased esti-
mates of variance components with as few as 3 units at the highest level using a carefully
considered, weakly informative prior. However, Stegmueller (2013) concluded that multi-
level models using ML estimates that did not include cross- level interactions needed 15– 20
countries to produce estimates of macro eects with acceptable levels of bias.
10. e analysis was run with the xtmelogit command, which is a maximum likelihood esti-
mate (7 iterations).
11. Internet use as a simple percentage was not signicant in an earlier iteration of the model
and was removed from the nal specication.
12. Due to the multilevel structure of the model, summary statistics such as pseudo R- square
are not appropriate to calculate since they do not account for the variance in both levels of
analysis. Instead, intra- class correlation coecients (ICC) were calculated and compared
across models. e ICC is an inferential statistic designed for analysis of group data and
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
272 Notes
272
describes how much additional variance is explained by the addition of the level 2 variables.
None of the ICC coecients were higher than .23 for any of models tested (including the
nal version), which indicates the macro factors provided at best a modest contribution to
the explanatory model beyond the individual level variables.
13. See Conclusion, note 2, for details of subsequent studies that have used the full data set to
investigate online mobilization.
Chapter 4
1. James Temperton, “Elections, Politicians and the Internet:A History of Failure,” Wired,
April 22, 2015, available at hp:// www.wired.co.uk/ news/ archive/ 2015- 04/ 14/ digital-
general- election (accessed April 24, 2015).
2. e Conservatives’ address was www.conservative- party.org.uk and Labour occupied the
even less intuitive, albeit more aspirational address, www.labourwin97.org.uk (Bowers-
Brown and Gunter, 2002:167).
3. David Walker, “Electronic Election,” Guardian Online, May 10, 2001, available at hp://
www.theguardian.com/ technology/ 2001/ may/ 10/ internet.ukgeneralelection2001
(accessed April 4, 2015).
4. See McCarthy (2001) and Coleman and Hall, (2001)
5. ”Campbell Says He Never Used Net,” BBC Newsonline, February 21, 2006, available at
hp:// news.bbc.co.uk/ 1/ hi/ uk_ politics/ 4731764.stm (accessed May 5, 2015); “Blair
Discovers Joys of e- mailing,” BBC Newsonline, July 14, 2006, available at hp:// news.bbc.
co.uk/ 1/ hi/ uk_ politics/ 5180888.stm (accessed May 5, 2015)
6. Temperton, “Elections, Politicians and the Internet.”
7. Ma Carter, “Get Connected,” Progress Magazine, September 5, 2005, available at hp://
www.progressonline.org.uk/ 2005/ 08/ 25/ get- connected (accessed May 10, 2015).
8. See Cruddas and Harris (2006) who’s post- election Compass report criticized the “top-
down modus operandi” of the website.
9. Matt Warman, “General Election 2010: Never Underestimate the Power of the
Internet,” The Telegraph, April 7, 2010, available at http:// www.telegraph.co.uk/
news/ election- 2010/ 7561876/ General- Election- 2010- never- underestimate- the-
power- of- the- internet.html (accessed May 9, 2015); Jon Swaine, “General Election
2010:Facebook and Twitter to Have Unprecedented Impact,” The Telegraph, April
7, 2010, available at http:// www.telegraph.co.uk/ news/ election- 2010/ 7558703/
General- election- 2010- Facebook- and- Twitter- to- have- unprecedented- impact.html
(accessed August 9, 2015).
10. James Temperton, “Elections, Politicians and the Internet.”
11. e label “nasty party” entered the UK political lexicon in 2002 aer Teresa May used
it to refer to perceptions of the party at the Conservative Party conference that year.
See Michael White and Ann Perkins, “‘Nasty Party’ Warning to Tories,” e Guardian,
September 8, 2002, available at hps:// www.theguardian.com/ politics/ 2002/ oct/ 08/
uk.conservatives2002 (accessed June 1, 2016).
12. Labour was reported to have spent over £2million on developing Contact and Campaign
Creator and Phone Bank. ese were digital tools to help local- level activists coordinate
and increase their voter mobilization eorts. See Gibson etal. (2010:10) for details.
13. ABlog post on Lib Dem Voice by Central party staer, David Loxton. Posted November 9,
2009. Extracted by Mark Pack for author, November 14, 2015.
14. e largest group formed on LibDemACT was Lib Dem Youth with 312 members. Figures
compiled from site by author, May 1, 2010.
15. Lilleker and Jackson (2011:135) reported that the biggest groups ranged between 492
and 1,405 members, compared with 312 recorded by the author on May 1, 2010, for
LibDemACT
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Notes 273
273
16. According to “Refounding Labour,” Labour Party (2011), 59,000 members had logged in
during 2011, 3,300 events were created by over 400 CLPs, and over £100,000 donated to
local candidates through fundraising tools.
17. e estimate based on the author’s post- election interviews with party digital campaign
managers from all three main parties. Conrmed by email correspondence with Mark Pack,
May 9, 2015.
18. According to their LinkedIn proles, Rishi Saha, a Conservative candidate in 2005, became
head of new media for the Conservatives in January 2006, hps:// www.linkedin.com/
in/ rishisaha. Sue MacMillan became Labour’s director of new media in September 2007,
hps:// www.linkedin.com/ in/ suemacmillan
19. George Windsor and Shane Murphy, “Big Data and the 2015 UK General Election:Digital
Democracy or Digitally Divisive?,” Nesta blog, June 20, 2014, available at hps:// www.
nesta.org.uk/ blog/ big- data- and- the- 2015- uk- general- election- digital- democracy- or-
digitally- divisive/ (accessed July 6, 2018).
20. Roland Watson, “Merlin at Tory HQ in Disgrace,” e Times, March 25, 2013, avail-
able at hps:// www.thetimes.co.uk/ article/ merlin- at- tory- hq- in- disgrace- hr76ddt8vp5
(accessed March 25, 2013).
21. Iain Roberts, “ank You EARS, but the VAN Is Coming,” Liberal Democrat Voice, June 19,
2011, available at hps:// www.libdemvoice.org/ thank- you- ears- but- the- van- is- coming-
24486.html (accessed June 17, 2014).
22. Aisha Gani and Rowena Mason, “Tories Pumping £100,000 a Month into Facebook
Advertising,” e Guardian, February 6, 2015, available at hps:// www.theguardian.com/
politics/ 2015/ feb/ 06/ tories- pumping- facebook- advertising- email- ukip (accessed May
21, 2015).
23. Ross Hawkins, “Tories’ £100,000 a Month Facebook Bill,” BBC Newsonline, February 5,
2015, available at hps:// www.bbc.co.uk/ news/ uk- politics- 31141547 (accessed May
21, 2015).
24. James Temperton, “Elections, Politicians and the Internet.”
25. Anthony Ridge- Newman, “#GE2015— UK General Election and New Political
Communication,” Ridgenewman Blog, April 23, 2015, available at hp:// ridgenewman.
blogspot.co.uk/ 2015/ - 4/ ge2015- uk- general- election- 2015- and- new.html (accessed April
27, 2015).
26. Adriana Coppola, “Forget the Manifesto: Big Data Will Win Future Elections,” e
Guardian, May 6, 2015, available at hp:// www.theguardian.com/ media- network/ 2015/
may/ 06/ general- election- big- data- marketing- electioneering (accessed May 20, 2015)
27. ese were conducted for Economic and Social Research Council (ESRC) grant proj-
ects (Award Nos:RES- 000- 22- 1284; RES- 051- 27- 0299). e author was a principal or
co- investigator. ey include a 2005 post- election survey by NOP and a 2010 survey by
BMRB. All were conducted through face- to- face methods and included internet and non-
internet users in the sample. Full details of the survey and variable denitions are provided
in Appendix 4.2.
28. e estimates for the general population for 2015 read and redistribute were estimated
using the ONS gure of 86% internet access for that quarter i.e. the gure of 59.4% cal-
culated for read online news from the YouGov internet user only sample was multiplied
by 0.86 to derive the gure of 51.1, and similarly for redistribute gures. e reverse was
done for the receive gures given that we had the general population estimates but not the
internet user population. In this case the estimates were obtained by dividing the original
full sample estimates by 0.86.
29. ere were some questions about whether respondents used Facebook and Twier to get
information on candidates and parties on the YouGov panel survey; however, they were
not used to compute the sign- up variable since the question wording referred explicitly
to reading content, not actively signing up to receive updates. Given this, these variables
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
274 Notes
274
were used contribute to measuring the “read” mode, specically in terms of measuring
aention to campaign sites. e decision to exclude these items from sign- up inevitably
means that we are likely to have missed some of those who actively sign up to follow and
receive updates from the parties and thereby underestimate the numbers engaging in the
redistribute mode. When the Facebook and Twier items are included as part of the sign-
up variable, it increases to 15percent and redistribute (i.e., sign up and share) to 8.5per-
cent of internet users. In the mail- back component of the BES, where the CSES questions
were included, there was a measure of signing up to receive party information and news
which reported 7percent of the population as undertaking this activity, suggesting that the
Facebook/ Twier estimates are inated. Since there was no variable to measure sharing of
election news and information, however, it was not possible to compute the “redistribute”
from the F2F BES survey 2005 and 2010.
30. e percentages for party supporters engaged in receive mode in 2015 are likely to be
underestimates. Since there was no baseline measure of internet use for the BES respon-
dents who completed the CSES module, and this was used to calculate the receive gures
we used the ONS estimate of 86percent internet use among the UK adult population
reported in Table4.2 as the baseline for extrapolating the internet user % gures for each
party. Following the process detailed in note 28. As can be seen from 2005 and 2010, the
overall estimates of internet use for the population (53percent and 75percent, respec-
tively) are slightly lower than is typically the case for most party supporters, but not mark-
edly so. e gures for 2010 show a greater correspondence than 2005. us, we consider
use of the ONS estimate for the United Kingdom to be acceptable here to calculate the
internet user- only percentage within all parties.
Chapter 5
1. See Australian 1988 Commonwealth Privacy Act and the Privacy Amendment (Private
Sector) Act 2000. For further information see hps:// www.oaic.gov.au/ privacy- law/
privacy- act/ (accessed June 10, 2016).
2. See Australian Privacy Law and Practice (ALRC Report 108), available at hp:// www.alrc.
gov.au/ publications/ report- 108 (accessed June 10, 2016).
3. eir exemption from the privacy legislation allows the parties to exploit these data with-
out prior consent and also means they are not obliged to disclose the personal data that are
held (Howard and Kreiss, 2010).
4. For a discussion of the wealth of information held on voter preferences by the Australian
parties’ databases, see also Section 41, “Political Exemptions,” of the Australian Privacy
Law and Practice (ALRC Report 108), Section 41, available at hp:// www.alrc.gov.au/
publications/ 41.%20Political%20Exemption/ introduction (accessed June 10, 2016).
5. World E- Government Rankings, UN E- Government Survey, 2014, available at hps://
publicadministration.un.org/ egovkb/ Portals/ egovkb/ Documents/ un/ 2014- Survey/
Chapter1.pdf (accessed June 25, 2016).
6. e success of these policies can be seen from ocial gures which show a continual nar-
rowing of the gap between city and countryside dwellers’ use of the internet over time, par-
ticularly since 2010. Reports issued by the Australian Bureau of Statistics (ABS) show that
by 2013 the divide had almost disappeared, with 79percent of households located outside
of capital cities reporting access, compared with 85percent of households located within
capital cities.
7. e AEC denition of a marginal seat is where the leading party has received less than
56percent of the nal two- party preferred vote (i.e., aer preferences have been allocated)
in the election to the House of Representatives. Using the list of the 62 marginal seats in
2016 reported by the Hungton Post at hp:// www.hungtonpost.com.au/ 2016/ 05/
09/ the- 62- marginal- seats- that- will- decide- the- election/ and the corresponding ABS clas-
sication of the demographic status of the 150 electoral divisions (see hp:// www.aec.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Notes 275
275
gov.au/ proles), 27 or 44percent of these 62 marginals were identied as either rural or
provincial.
8. “e ALP and the Internet,” 2000, available at hp:alp.com.au/ about/ alpwww.html
(accessed July 10, 2000).
9. For full listing of party website URLs, see Ross Storey, “Internet’s Too Popular for Politics
to Ignore,” e Australian, February 13, 1996, p.57.
10. Post- election interviews by author with party web campaign ocers. Interview
period:October 26, 2000– November 3, 2000.
11. Storey, “Internet’s Too Popular for Politics to Ignore,” p.57.
12. Quote is from Mr. Geo Brooks, the Managing Director of Multimedia Creations, a com-
pany that ran a website (www.medianet.com.au) devoted to daily coverage of the 1996 elec-
tion. Reported by Ross Storey, “Internet's too popular for politics to ignore” e Australian,
February 13, 1996, p.57.
13. Ibid.
14. Post- election interviews by author with party web campaign ocers from Greens,
Democrats, and National Party. Interviews conducted October 26 and 27, 2000.
15. “e ALP and the Internet.”
16. Quote is from Bruce Hawker, former chief of sta to NSW Premier Bob Carr, 1999. Cited
in Gibson and Ward (2002):105.
17. Richard McGregor, “Politicians Cast the Net for Votes,” e Australian, November 3,
1999, p.12.
18. Lara Sinclair, “Campaign Websites Go the Way of Je aer Costly 99 Debacle,” Weekend
Australian, October 2, 2004, p.10.
19. Lara Sinclair and Sheen MacLean, “It’s a Party as Voters Blitzed,” e Australian, October 7,
2004, p.20.
20. Ben Oquist, “We Are All Americans Now,” Green, no.15 (Summer 2004– 2005); cited in
Australian Parliamentary Library Research Note, February 8, 2005, no.30, 2004– 2005,
available at hp:// apo.org.au/ system/ les/ 699/ apo- nid699- 3866.pdf (accessed July 21,
2018). Sinclair, “Campaign Websites Go the Way of Je aer Costly 99 Debacle,” p.10.
21. ”e YouTube Election,” Sydney Morning Herald, July 14, 2007, p.23.
22. Sally Jackson, “Labor to Increase Internet Dominion,” e Australian Media Section, August
27, 2007, p.33; ”Rush for Cyberspace Has Traps for New Players,” Sun Herald, August 12,
2007, p.31. Ben Doherty, “www.Kevin07.com.au:Rudd Ups the Ante in the Cyberspace
War,” e Age, August 8, 2007, p.8.
23. Sid Marris, “ALP Launches Kevin07,” e Australian, August 7, 2007; Misha Schubert,
“New Spin in Politics Web,” e Age, August 7, 2007, p.6.
24. Julian Lee, “New Media Ad Heaven for Kevin07,” Sydney Morning Herald, December 6,
2007, p.33.
25. Doherty, “www.Kevin07.com.au,” p.8.
26. ”How the Web Shaped the Australian Election,” stu.co.nz 23, November 2007, available
at hp:// www.stu.co.nz/ technology/ 132150/ How- the- web- shaped- the- Australian-
election (accessed January 2, 2008); Imre Salusinszky, “Slow Broadband Holds Back Web
Push,” Weekend Australian, October 2007, p.12
27. ”Labour Launches Social Media Platform,” July 15, 2010, Community Engine, available at
hps:// mumbrella.com.au/ labor- launches- social- media- platform- 29865 (accessed May
15, 2016).
28. Josh Gordon, “Libs Hope Answer to Woes Is Online,” Sydney Morning Herald, April 6,
2008, available at hp:// www.smh.com.au/ news/ national/ libs- hope- answer- to- woes- is-
online/ 2008/ 04/ 05/ 1207249537261.html (accessed May 14, 2008).
29. Interview with Author, October 12, 2010.
30. Farrah Tomazin, “Bare- Bones Greens Growing from Grassroots,” e Age, June 12,
2010, available at hps:// www.smh.com.au/ national/ barebones- greens- growing- from-
grassroots- 20100611- y3os.html (accessed July 21, 2018).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
276 Notes
276
31. GetUp!, “Post Election Report:Election 2010,” August 1, 2010, available at hp:// www.
GetUp!.org.au/ les/ about/ GetUp!_ annual_ report_ 2010.pdf (accessed April 2, 2011).
32. Jennifer Rayner, “Will Big Data Change the 2013 Election?,” e Drum online, January 8,
2013, available at hp:// www.abc.net.au/ news/ 2013- 01- 09/ rayner- - - big- data/ 4456834
(accessed May 5, 2015); Caitlin Fitzsimmons, “Lessons from the White House as Australia’s
First Big Data Election Looms,” February 14, 2013, Australian Financial Review, available
at hp:// www.afr.com/ it- pro/ lessons- from- the- white- house- as- australias- rst- big- data-
election- looms- 20130213- jyvfc (accessed May 5, 2015).
33. Stephen Mills, “Big Data Meets Doorknocking:The Political Contest’s New Frontier,”
The Conversation, July 10, 2014, available at http:// theconversation.com/ big- data-
meets- doorknocking- the- political- contests- new- frontier- 27822 (accessed May
12, 2015).
34. Michele Levine, director of Roy Morgan Polling, quoted in Fitzsimmons, “Lessons from
the White House.”
35. McGregor, “Politicians Cast the Net for Votes,” p.12.
36. Rayner, “Will Big Data Change the 2013 Election?”; Fitzsimmons, “Lessons from the
White House.”
37. “Annual GetUp! Report 1 July 2013– 30 June 2014,” 2013 Federal Election Section, avail-
able at hp:// annualreport2013- 14.getup.org.au/ #election (accessed May 27, 2015).
Chapter 6
1. Angelique Chrisas, “France Says Farewell to the Minitel— e Lile Box at Connected
a Country,” e Guardian, June 28, 2012, available at hp:// www.theguardian.com/ tech-
nology/ 2012/ jun/ 28/ minitel- france- says- farewell (accessed July 7, 2015)
2. For a good guide to the rules governing French campaigns, see the US Library of
Congress Law Library pages at hps:// www.loc.gov/ law/ help/ campaign- nance/ france.
php#campaigns
3. Length of ocial campaign period was not included in the CSES macro data le and
has not been used as an independent variable in previous cross- national models of voter
mobilization. Future iterations of the comparative analysis conducted in Chapter3 should
give consideration to testing the impact of this variable on rates of both oine and online
contact.
4. See report on 2012 election by the Oce for Democratic Institutions and Human Rights—
e OSCE/ ODIHR Needs Assessment Mission Report February 20– 22, 2012, available at
hps:// www.osce.org/ odihr/ elections/ 89000?download=true (accessed September 2,
2018). See also the OSCE report on “Political Advertising and Media Campaign during the
Pre- election Period:A Comparative Study,” available at hps:// www.osce.org/ mission- to-
montenegro/ 346631?download=true (accessed September 2, 2018).
5. It is worth noting that to date, neither print nor online media have been included under
the equal time rules. is omission clearly could increase the incentives for parties to
exploit these new channels during the election as an alternative means of reaching voters.
However, restrictions do apply to the purchase of online advertising (similarly to television
and radio), which means that candidates do not enjoy unfeered use of digital channels
to promote themselves during elections. Furthermore, although the lack of regulation of
parties’ free or “unpaid” speech online may increase its appeal, the wider legal and cultural
environment in France (which we also discuss later in the chapter) clearly discourages this
type of unsolicited, individualized contacting by political actors. ese norms, we argue,
would thus counter impulses by candidates to use their sites as platforms for direct self-
promotion. See the reports listed in note 3 for further information about control over par-
ties’ use of media in French presidential elections.
6. e EU General Data Protection Regulation Act (GDPR) was required to be enacted
in all member states as of May 2018. e new rules increase the rights of individuals to
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Notes 277
277
control their personal data and the obligations of organizations to ensure privacy and
consent in the collection of those data. e sanctions for noncompliance have also been
strengthened, particularly with regard to nes. Asummary of the key points and further
information is available from the UK Information Commissioners Oce (ICO) and can be
found here: hps:// ico.org.uk/ for- organizations/ guide- to- the- general- data- protection-
regulation- gdpr (accessed July 12, 2018).
7. Notably, the French government had already taken steps to increase the powers of CNIL in
2016 with the Digital Republic Bill; see hp:// www.gouvernement.fr/ en/ the- digital- bill.
See also hps:// www.huntonprivacyblog.com/ 2016/ 10/ 31/ entry- force- french- digital-
republic- bill/ . e new legislation pregured the arrival of GDPR by strengthening exist-
ing controls over use of French citizens’ data and reinforcing requirements for their explicit
consent to its collection and use. CNIL also released a secondary document in 2016 that
imposed more specic restrictions on political parties’ use of social media to contact voters.
More general information on data privacy regulation in France can be found in the DLA
Piper Global Data Protection Handbook at hp:// dlapiperdataprotection.com/ #hand-
book/ law- section/ c1_ FR (accessed June 2, 2016).
8. Steven Erlanger, “In France, Using Lessons from Obama Campaign,” NewYork Times, April
21, 2012, available hp:// www.nytimes.com/ 2012/ 04/ 22/ world/ europe/ in- france-
hollande- camp- tries- us- style- canvassing.html?_ r=0 (accessed July 21, 2015).
9. e company was the architect of Hollande’s 2012 online campaign. According to the
website at hp:// www.liegeymullerpons.fr (accessed June 12, 2015), they advertise them-
selves as “La première startup de stratégie électorale en Europe” with an explicit mission
of bringing the new “big data”– driven campaign tools to European campaigns. For a more
detailed account of their activities during the 2012 campaign see Liégey etal. (2013),
See also hp:// www.liegeymullerpons.fr/ references/ tous- hollande- 2012 (accessed June
28, 2018).
10. “La Coopol, le ‘réseau social’ du Parti socialiste,” nonction., available at hp:// www.non-
ction.fr/ article- 3417- la_ coopol_ le_ reseau_ social_ du_ parti_ socialiste.htm (accessed
July 21, 2015)
11. Oriane Ran, “Le PS lance la Coopol, son réseau social en ligne,” 20 Minutes, January
12, 2010, available at hp:// www.20minutes.fr/ politique/ 375770- 20100112- ps- lance-
coopol- reseau- social- ligne (accessed on July 22, 2015).
12. Rémi Duchemin, “La Coopol bouge encore,” Europe 1, August 26, 2011, available at hp://
www.europe1.fr/ politique/ la- coopol- bouge- encore- 687777 (accessed on July 22, 2015).
13. Antonella Napolitano, “Can an Obama- like Campaign Work in France?” Tech President,
March 21, 2012, available at hp:// techpresident.com/ news/ 21948/ can- obama-
campaign- work- france (accessed July 4, 2013).
14. Samuel Laurent, “La mort annoncée du réseau social de l’UMP,” Le Monde, December
28, 2010, available at hps:// www.lemonde.fr/ politique/ article/ 2010/ 12/ 28/ la- mort-
annoncee- du- reseau- social- de- l- ump_ 1458360_ 823448.html (accessed July 28, 2018).
15. On a standardized 0– 100 scale, Labour’s Membersnet scored 71, Tous Hollandee scored 58,
and Labor Connect scored 47.
16. Napolitano, “Can an Obama- like Campaign Work in France?”; ”Le PS peut désormais
compter sur une armée de sympathisants,” L’Obs, October 21, 2011, available at hps://
www.nouvelobs.com/ politique/ primaire- socialiste/ 20111019.OBS2859/ le- ps- peut-
desormais- compter- sur- une- armee- de- sympathisants.html (accessed July 4, 2013).
17. Based on IDEA gures for Voting Age Population in France in 2012, 51,979,508. See
hps:// www.idea.int/ data- tools/ question- countries- view/ 441/ 86/ ctr (accessed
September 3, 2018).
18. is email list obviously includes an unmeasured number of overseas and non- eligible US
residents. According to US Census reports, the total eligible voting population in 2008 was
206,072. is included those citizens over the age of 18; hps:// www.census.gov/ prod/
2014pubs/ p20- 573.pdf.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
278 Notes
278
19. According to media accounts, Sarkozy’s government spent millions seing up an agency
to implement the so- called HADOPI law or “Haute Autorité pour la Diusion des oeu-
vres et la Protection des droits d’auteur sur Internet,” which was designed to prevent ille-
gal le- sharing. e law ran afoul of the French Constitutional Council, however, in its
stipulation that those found guilty would have their internet access removed for one year.
Internet access, the Council argued, was a “human right.” For more coverage of the issues,
see Richard Wray, “French Anti- lesharing Law Overturned,” e Guardian, June 10, 2009,
available at hps:// www.theguardian.com/ technology/ 2009/ jun/ 10/ france- hadopi- law-
lesharing (accessed September 3, 2018).
20. ”Sarkozy and Hollande:Two Great Examples of How Not to “Do” Twier,” Myndset by
Minter Dial, June 12, 2015, available at hp:// myndset.com/ 2012/ 07/ 12/ sarkozy- and-
hollande- two- great- examples- of- how- not- to- do- twier/ (accessed on October 5, 2015).
”Hollande and Sarkozy:How Not to Use Social Media in Politics or Business,” Econsultancy,
November 8, 2012, available at hps:// econsultancy.com/ hollande- and- sarkozy- how- not-
to- use- social- media- in- politics- or- business/ (accessed October 4, 2015).
21. Estimates provided by Koc- Michalska etal. (2014) are that the Greens e- campaign budget
for 2012 was 25,000 euros, a h of the amount that it had spent in 2007.
22. Author interview with member of Greens’ digital team, June 12, 2012
23. e estimates for the general population for 2007 read and redistribute were derived
using the 66% internet access gure reported in gure6.1 using World Bank estimates
for that year. As in the UK for 2015 where an online sample was used to calculate read
and redistribute we multiplied the % engaging in an activity for online users by 0.66 to
approximate the general population gures. us 68% for internet users who read online
news about the campaign equates to 44.9% of the overall sample/ population. We apply
the same method to calculate the population estimates for candidate sites and the redis-
tribute variables.
24. e response categories dier, with four options in 2007 and ve in 2012. In addition, the
2007 question asks about interest in the presidential campaign, while the 2012 asks about
interest in politics generally. e results indicate that while the proportions with the high-
est interest were comparable across the two surveys (29percent and 28percent across the
respective years), those with only “a bit of interest” were signicantly smaller in number in
the 2012 survey (19percent vs. 46percent), as were those with no interest (5percent vs.
10percent).
25. ere may also be a methodological reason for its robustness. e measures used to calcu-
late the read mode of engagement in the two election studies were very similar. us, the
higher rate reported in 2007 is in line with the general paern of higher estimates of politi-
cal aitudes and behavior that were reported in that survey more generally. e measure
used for “redistribute” in 2007, however, was notably more conservative than that used in
2012 and thus likely underestimated numbers. Specically, the “redistribute” variable in
2012 was based on respondents saying yes to two items— whether they had “signed up to
help support a candidate/ party” AND had shared political content. In 2007 we used a two-
part lter question to measure redistribute. Specically, we rst selected those individuals
who said they had downloaded any political content AND also shared political material.
en, in order to make this a more precise measure of redistributing ocial party content,
we applied a lter that selected only those respondents who reported party sites as one of
their top three sources of information. is ltering process arguably introduced a more
stringent selection criteria than the measure used in 2012, and thus may have resulted in
suppressing the gures for redistribution.
26. Income was used rather than social class. Unlike the United Kingdom, France does not have
a commonly agreed hierarchy of class categories that are used by survey researchers. e
categories that do exist are very specic and extensive in nature. See Desrosières (2008)
Rose and Pevalin (2001) for further discussion of the complexity in constructing measures
of socioeconomic class in France.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Notes 279
279
Chapter 7
1. e notion of Americanization has been challenged by Negrine and Papathanassopoulos
(1996) and Plasser (2000), who have argued against the idea of a wholesale importation of
US techniques and the importance of local context in shaping a process of “moderni zation.”
For a good overview of the debates, see Farrell (2002).
2. e quote is aributed to Jim Messina, Obama’s campaign manager for his 2012 re- election
bid. See Dan Balz, “How the Obama Campaign Won the Race for Voter Data,” Washington
Post, July 23, 2013, available at hps:// www.washingtonpost.com/ politics/ how- the-
obama- campaign- won- the- race- for- voter- data/ 2013/ 07/ 28/ ad32c7b4- ee4e- 11e2- a1f9-
ea873b7e0424_ story.html?utm_ term=.796841dd4a0b (accessed March 3, 2018).
3. Phil Noble, “Internet and Campaign 2002 Analysis,” Special Report Politics Online,
November 4, 2002, available at hp:// netpulse.politicsonline.com/ soundo.asp?issue_
id=6.18 (accessed November 6, 2002).
4. J. Benson, “Dean Goes Bust:e $40 Million War Chest Is Gone— and so Is Campaign
Manager Joe Trippi. W hat Happened?” Salon, January 29, 2004, available at hp:// archive.
salon.com/ news/ feature/ 2004/ 01/ 29/ dean/ index.html (accessed March 12, 2007).
5. ”Howard Dean:Politics in Cyberspace,” NewsMax.com, July 5, 2003, available at hp://
www.newsmax.com/ archives/ articles/ 2003/ 7/ 4/ 161644.shtml (accessed July 20, 2003).
6. Quoted in E. J. Dionne Jr., “Dean’s Grass- Roots Cash Cow,” Washington Post, July 8,
2003, available at hps:// www.washingtonpost.com/ archive/ opinions/ 2003/ 07/ 08/
deans- grass- roots- cash- cow/ e4d4eaac- 53a0- 4c55- a0d8- 35e48db2c629/ ?utm_ term=.
a25271b5d9a9 (accessed July 20, 2003).
7. See Chris Suellentrop, “Peer to Peer Politics:Should Howard Dean be a Lile Bit Afraid
of the Internet?” Slate.com. July 14, 2003, available at hps:// slate.com/ news- and-
politics/ 2003/ 07/ should- howard- dean- be- afraid- of- the- internet.html (accessed June
28, 2007). Edward Cone “Election 2008:e Internet Campaign.” CIOInsight 6 August
6, 2007. Available at hps:// www.cioinsight.com/ print/ c/ a/ Trends/ Election- 2008-
e- Internet- Campaign, (accessed August 15, 2010). Marshall Ganz, who helped design
Obama’s eld organization, is reported as having said that Dean “understood how to use
the internet for the fundraising but not for the organizing” in Sarah Lai Stirland, “Obama’s
Secret Weapons:Internet Databases and Psychology,” blogwired, October 29, 2008, avail-
able at hp:// www.wired.com/ threatlevel/ 2008/ 10/ obamas- secret- weapon (accessed
November 12, 2008).
8. S. DiJulio and A. Wood, “Online Tactics & Success:An Examination of the Obama for
America New Media Campaign,” a report by the Wilburforce and Brainerd Foundation,
2008, available at hp:// www.wilburforce.org/ resources/ wilburforce- funded- reports/
online- tactics- and- success/ at_ download/ le (accessed July 31, 2016).
9. Jose Antonio Vargas, ”Obama Raised Half a Billion Online,” Washington Post, November
20, 2008, available at hp:// voices.washingtonpost.com/ 44/ 2008/ 11/ obama- raised- half-
a- billion- on.html (accessed February 26, 2018)
10. Mark Sweeney, “Barack Obama Buys 30- Minute TV Ad,” e Guardian, October 10, 2008,
available at hps:// www.theguardian.com/ media/ 2008/ oct/ 10/ barack- obama- tv- ad
(accessed February 26, 2018)
11. Vargas, “Obama Raised Half a Billion Online.”
12. Paul chose December 16, 2007, as the 234th anniversary of the Boston Tea Party, a date
of considerable importance to his libertarian- minded supporters. See Kenneth P. Vogel,
“‘Money Bomb’:Ron Paul Raises $6 Million in 24- Hour Period,” US A To d a y , December
17, 2007, available at hp:// usatoday30.usatoday.com/ news/ politics/ election2008/
2007- 12- 17- ronpaul- fundraising_ N.htm Accessed on August 1, 2016.
13. Kenneth P. Vogel, ‘Ron Paul Becomes $6 Million Man,” Politico Magazine, December
16, 2007, available at hps:// www.politico.com/ story/ 2007/ 12/ ron- paul- becomes- 6-
million- man- 007421 (accessed August 5, 2018); Andrew Malcom, Los Angeles Times,
February 1, 2008, “News Shocker:Ron Paul Was Biggest GOP Fundraiser Last Quarter,”
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
280 Notes
280
available at hp:// latimesblogs.latimes.com/ washington/ 2008/ 02/ news- shocker- ro.html
(accessed August 1, 2016).
14. Kate Phillips, “A Ban on Ron Paul Supporters,” e Caucus: e Politics and Government blog
of the NewYork Times, October 24, 2007, available at hp:// thecaucus.blogs.nytimes.com/
2007/ 10/ 24/ a- ban- on- ron- paul- supporters/ ?_ r=0 (accessed August 1, 2016).
15. Jose Antonio Vargas, “Meet the OPOs,” Washington Post, May 4, 2007, available at hp://
www.washingtonpost.com/ wp- dy n/ content/ article/ 2007/ 05/ 03/ AR2007050302546.
html (accessed February 28, 2018).
16. Vargas, “Meet the OPOs,” note 14.
17. Sasha Issenberg, “A Vast Le- Wing Competency,” Slate Magazine, November 7, 2012,
available at hp:// www.slate.com/ articles/ news_ and_ politics/ victory_ lab/ 2012/ 11/
obama_ s_ victory_ how_ the_ democrats_ burned_ by_ karl_ rove_ became_ the_ party_
of.html (accessed April 8, 2013).
18. e work of Alan Gerber and Don Green in New Haven in 1998 using eld experiments to
test and measure the precise impact of dierent methods of party contact on voter turnout
was particularly inuential here. See Chapter3 of Issenberg (2012) for a good historical
overview of the emergence of the new methods.
19. Benedict Carey, “Academic ‘Dream Team’ Helped Obama’s Eort,” New York Times,
November 12, 2012, available at hp:// www.nytimes.com/ 2012/ 11/ 13/ health/ dream-
team- of- behavioral- scientists- advised- obama- campaign.html (accessed March 3, 2018).
20. NOI was absorbed by Wellstone Action, a liberal- progressive organization that was
founded aer the death of Senator Paul Wellstone of Minnesota. Information on its mis-
sion and core activities is available at hp:// neworganizing.wellstone.org/ about/ hp://
neworganizing.wellstone.org/ about/ what- we- do/ (accessed March 4, 2016).
21. Balz, ‘How the Obama Campaign Won the Race for Voter Data.”
22. See Teddy Go, LinkedIn prole for 2007– 2008, hps:// www.linkedin.com/ in/ teddy-
go/ .
23. “Inside the Cave:Obama’s Digital Campaign,” a report by Engage Research, slide 10, avail-
able at hps:// enga.ge/ wp- content/ uploads/ 2018/ 01/ Inside_ the_ Cave- 1.pdf (accessed
December 1, 2014).
24. Alexis C. Madrigal, “When the Nerds Go Marching In,” e Atlantic, November 16, 2012,
available at hps:// www.theatlantic.com/ technology/ archive/ 2012/ 11/ when- the- nerds-
go- marching- in/ 265325/ (accessed March 4, 2018).
25. Peggy Noonan, “ey’ve Lost at Lovin’ Feeling,” Wall Street Journal, July 30, 2011.
26. Sean Gallagher, “Built to Win: Deep Inside Obama’s Campaign Tech,” Ars Technica,
November 14, 2012, available at hp:// arstechnica.com/ information- technology/ 2012/
11/ built- to- win- deep- inside- obamas- campaign- tech/ (accessed November 26, 2014).
27. Michael Scherer, “Inside the Secret World of the Data Crunchers Who Helped Obama
Win,” Time Magazine, November 7, 2012, available at hp:// swampland.time.com/ 2012/
11/ 07/ inside- the- secret- world- of- quants- and- data- crunchers- who- helped- obama- win
(accessed March 12, 2014).
28. Gallagher, “Built to Win.”
29. Derrick Harris, “How Obama’s Tech Team Helped Deliver the 2012 Election,” Gigaom,
November 12, 2012, available at hps:// gigaom.com/ 2012/ 11/ 12/ how- obamas- tech-
team- helped- deliver- the- 2012- election/ (accessed August 8, 2016).
30. Scherer, “Inside the Secret World of the Data Crunchers.”
31. Gallagher, “Built to Win.”
32. Issenberg, “A Vast Le- Wing Competency.” . Proof of their newfound accuracy was claimed
by the campaign aer the rst results came in on election night from the small hamlet of
Dixville Notch in New Hampshire. e vote showed an even split in support for Obama
and Romney, which, according to the analytics team, was exactly what their model had
predicted. Christi Parsons and Kathleen Hennessey, ‘Obama Campaign’s Investment
in Data Crunching Paid O,” LA Times, November 13, 2012, available at hp://
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Notes 281
281
articles.latimes.com/ 2012/ nov/ 13/ nation/ la- na- obama- analytics- 20121113 (accessed
December 3, 2012).
33. “Inside the Cave: Obama’s Digital Campaign,” a report by Engage Research, slide 87.
Available at hp:// enga.ge/ dl/ Inside_ the_ Cave.pdf (accessed December 1, 2014); Nick
Judd, “How Obama for America Made Its Facebook Friends into Eective Advocates,”
TechPresident, November 19, 2012, available at hp:// techpresident.com/ news/ 23159/
how- obama- america- made- its- facebook- friends- eective- advocates (accessed March
12, 2014).
34. Madrigal, “W hen the Nerds Go Marching In.”
35. Michael Scherer, ‘Exclusive:Obama’s 2012 Digital Fundraising Outperformed 2008,” Time
Magazine, November 15, 2012, available at hp:// swampland.time.com/ 2012/ 11/ 15/
exclusive- obamas- 2012- digital- fundraising- outperformed- 2008/ (accessed December 1,
2014); Tim Murphy, “Under the Hood of Team Obama’s Tech Operation,” Mother Jones,
November 15, 2012, available at hp:// www.motherjones.com/ politics/ 2012/ 11/ inside-
obama- campaign- tech- operation (accessed January 1, 2014).
36. Stromer- Galley (2014): 147; “Inside the Cave:Obama’s Digital Campaign,” a report by
Engage Research, slide 42, available at hps:// enga.ge/ wp- content/ uploads/ 2018/ 01/
Inside_ the_ Cave- 1.pdf (accessed December 1, 2014).
37. Murphy, “Under the Hood of Team Obama’s Tech Operation.”
38. Sean Gallagher, “Inside Team Romney’s Whale of an IT Meltdown,” Ars Tehnica, November
9, 2012, available at hps:// arstechnica.com/ information- technology/ 2012/ 11/ inside-
team- romneys- whale- of- an- it- meltdown/ (accessed March 5, 2018); Dashiell Benne,
“Romney’s Get Out the Vote Plan Was a Complete Disaster,” e Atlantic, November 9,
2012, available at hps:// www.theatlantic.com/ politics/ archive/ 2012/ 11/ romneys- get-
out- vote- plan- was- complete- disaster/ 321610/ (accessed March 5, 2018).
39. Directly aer their defeat in 2008 a coalition of prominent Republican online strategists
released a 10- point action program entitled “Rebuild the Party:A Plan for the Future.” e
program explicitly focused on how the party should exploit and harness new communica-
tion technologies to empower their grassroots supporters. See hp:// www.rebuildthepa-
rty.com/ plan (accessed January 9, 2009). In February 2009 the RNC organized a “Tech
Summit” for activists and strategists. See Emily Cadei, “Republicans Prep the Portals of
Online Organizing,” CQ Politics, February 17, 2009, available at hp:// www.cqpolitics.
com/ frame- templates/ print_ template.html (accessed April 4, 2009).
40. Blaise Hazelwood, former RNC political director and member of Romney’s targeting team,
quoted in Issenberg, “A Vast Le- Wing Competency.” .
41. omas B. Edsall, “e G.O.P’s Digital Makeover,” NewYork Times Opinionator Blog, April
3, 2013, available at hp:// opinionator.blogs.nytimes.com/ 2013/ 04/ 03/ the- g- o- ps-
digital- makeover/ ?_ r=0 (accessed April 4, 2013).
42. As in the French case, we adjusted the CIC index to take into account that it was a presi-
dential election by dropping the item in the resource mobilization sub- index measuring
whether the site encouraged people to stand as local candidates. Given that the US par-
ties do not oer formal membership, we also removed the item measuring if facilities were
available to help promote joining the party. is reduced the sub- index to a 0– 4 measure,
compared with 0– 5 in France and 0– 6 in Australia and the United Kingdom.
43. “Inside the Cave:Obama’s Digital Campaign,” a report by Engage Research, slide 68, avail-
able at hps:// enga.ge/ wp- content/ uploads/ 2018/ 01/ Inside_ the_ Cave- 1.pdf (accessed
December 1, 2014).
44. See “Inside the Cave:Obama’s Digital Campaign,” a report by Engage Research, available
at hps:// enga.ge/ wp- content/ uploads/ 2018/ 01/ Inside_ the_ Cave- 1.pdf (accessed
December 1, 2014).
45. Pew did release several data sets that examined voters’ use of internet technologies during
the 2012 election; however, none contained the range and direct comparability of items
available in the 2004 and 2008 studies. e data sets that were made available focused
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
282 Notes
282
more on particular uses, such as online video (hp:// www.pewinternet.org/ 2012/ 11/ 02/
online- political- videos- and- campaign- 2012/ ) and social networking (hp:// www.pewin-
ternet.org/ 2012/ 11/ 06/ social- media- and- voting/ ).
46. See “Inside the Cave:Obama’s Digital Campaign,” a report by engage Research, slide 55,
available at: hps:// enga.ge/ wp- content/ uploads/ 2018/ 01/ Inside_ the_ Cave- 1.pdf
(accessed December 1, 2014).
47. Madrigal, “W hen the Nerds Go Marching In.”
48. Frank Mace, “Just How Good Was the Obama Campaign?” Harvard Political Review,
November 29, 2013, available at hp:// harvardpolitics.com/ united- states/ just- good-
obama- campaign/ (accessed March 24, 2014).
49. See also John Sides and Lynn Vavreck, “Obama’s Not So Big Data Campaign,” Pacic
Standard Magazine, January 21, 2014, available at hps:// psmag.com/ social- justice/
obamas- big- data- inconclusive- results- political- campaigns- 72687 (accessed February
7, 2015).
50. Murphy, “Under the Hood of Team Obama’s Tech Operation.”
Conclusion
1. For investigations into and evidence of, the use of data- driven micro- targeting tech-
niques in the 2016 Brexit referendum, see Philip N.Howard and Bencde Kollanyi, “Bots,
#Strongerin, and #Brexit:Computational Propaganda during the UK- EU Referendum,”
2016, available at SSRN:hps:// ssrn.com/ abstract=2798311; see also extensive reports
by e Guardian on “e Cambridge Analytica Files,” available at hps:// www.theguard-
ian.com/ news/ 2018/ mar/ 26/ the- cambridge- analytica- les- the- story- so- far (accessed
December 2, 2018); and the 2018 report published by the UK Information Commissioners
Oce, “Investigation into the Use of Data Analytics in Political Campaigns,” available at
hps:// ico.org.uk/ media/ action- weve- taken/ 2259371/ investigation- into- data- analytics-
for- political- purposes- update.pdf (accessed December 2, 2018).
2. Further work by the author with colleagues on the drivers of digital campaign contact
using the full release of 38 countries provides a helpful update to our analysis. e study
by Magalhães et al. (2018) employs a simpler explanatory model that features fewer
macro- level variables. e analysis conrms the importance of internet use as a driver of
higher rates of online mobilization. Notably, the role of institutional context in the form
of single- member electoral districts is not found to be signicant in this larger N study.
While this discrepancy may be due to a sensitivity bias in our model based on the smaller
number of cases, it is also potentially due to the simpler explanatory model the authors
tested. Specically, the Magalhães etal. (2018) model did not test the eect of presidential
elections or their competitiveness. Furthermore, the dependent variable was measured in
a more nuanced manner, in that it was broken down into three separate modes— email,
SMS, and web- based/ social media. Also the focus was entirely on party- based contacting
or direct online mobilization, while in our analysis the dependent variable was based on
whether a respondent reported direct or indirect online contact.
3. Alexis C. Madrigal, “When the Nerds Go Marching In,” e Atlantic, November 16, 2012,
available at hps:// www.theatlantic.com/ technology/ archive/ 2012/ 11/ when- the- nerds-
go- marching- in/ 265325/ (accessed March 4, 2018)
4. Shane Goldmache,r “Hillary Clinton’s ‘Invisible Guiding Hand’:Meet the Lile- Known
Statistician behind the Democratic Nominee’s Most Important Strategic Decisions.”
Politico Magazine, September 7, 2016, available at hps:// www.politico.com/ magazine/
story/ 2016/ 09/ hillary- clinton- data- campaign- elan- kriegel- 214215 (accessed November
26, 2017).
5. Julie Pace and Jill Colvin, “AP Interview: Trump Says Big Rallies His Key Campaign
Weapon,” Associated Press, May 10, 2016, available at hps:// elections.ap.org/ content/
ap- interview- trump- says- big- rallies- his- key- campaign- weapon (accessed November 12,
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Notes 283
283
2018); Bernard Marr, “Could Big Data Be Donald Trump’s Achilles Heel?” Forbes Magazine,
June 17, 2016, available at hps:// www.forbes.com/ sites/ bernardmarr/ 2016/ 06/ 17/
could- big- data- be- donald- trumps- achilles- heel/ #7bee3bbe2952 (accessed November
12, 2017).
6. Holly Ellyat, “How I Helped Get Trump Elected:e President’s Digital Guru,” Tech
Trenders:A CNBC Special Report, November 8, 2017, available at hps:// www.cnbc.
com/ 2017/ 11/ 08/ how- i- helped- get- trump- elected- the- presidents- digital- guru- brad-
parscale.html (accessed November 12, 2017). Lois Becke, “Trump Digital Director Says
Facebook Helped Win the White House,” e Guardian, October 9, 2017, available at
hps:// www.theguardian.com/ technology/ 2017/ oct/ 08/ trump- digital- director- brad-
parscale- facebook- advertising (accessed November 11, 2019).
7. ”Donald Trump and Hillary Clinton’s Final Campaign Spending Revealed,” e Guardian,
December 9, 2016, available at hps:// www.theguardian.com/ us- news/ 2016/ dec/ 09/
trump- and- clintons- nal- campaign- spending- revealed (accessed November 10, 2019).
Issie Lapowsky, “What Facebook Isn’t Saying about Trump and Clinton’s Campaign Ads,”
Wired, February 28, 2018, available at hps:// www.wired.com/ story/ facebook- trump-
clinton- campaign- ad- cpms/ (accessed November 10, 2019).
8. Adam Lusher, “Cambridge Analytica: Who Are ey, and Did ey Really Help
Trump Win the White House?” e Independent, March 21, 2018, available at hps://
www.independent.co.uk/ news/ uk/ home- news/ cambridge- analytica- alexander- nix-
christopher- wylie- trump- brexit- election- who- data- white- house- a8267591.html (accessed
September 9, 2018); Michael Kranish, “Trump’s Plan for a Comeback Includes Building
a ‘Psychographic’ Prole of Every Voter,” Washington Post, October 27, 2016, available
at hps:// www.washingtonpost.com/ politics/ trumps- plan- for- a- comeback- includes-
building- a- psychographic- profile- of- every- voter/ 2016/ 10/ 27/ 9064a706- 9611- 11e6-
9b7c- 57290af48a49_ story.html?noredirect=on&utm_ term=.a81ec2f07f76 (accessed
September 9, 2018).
9. e UK news media led the way in these investigations. See the extensive reporting by
e Guardian and its “Cambridge Analytica Files,” available at hps:// www.theguardian.
com/ news/ series/ cambridge- analytica- les (accessed November 14, 2019); also see the
program archives of Channel 4 Television, “Data, Democracy and Dirty Tricks,” avail-
able at hps:// www.channel4.com/ news/ data- democracy- and- dirty- tricks- cambridge-
analytica- uncovered- investigation- expose (accessed November 14, 2019);
10. Issie Lapowsky, “e Real Trouble with Trump’s ‘Dark Post’ Facebook Ads,” Wired
Magazine, September 20, 2017, available at hps:// www.wired.com/ story/ trump- dark-
post- facebook- ads/ (accessed November 12, 2018); Lapowsky, “A Lot of People Are
Saying Trump’s New Data Team Is Shady,” Wired Magazine, August 15, 2016, available
at hps:// www.wired.com/ 2016/ 08/ trump- cambridge- analytica (accessed November
12, 2018); Joshua Green and Sasha Issenberg, “Inside the Trump Bunker, With Days to
Go,” Bloomberg News, October 27, 2016, available at hps:// www.bloomberg.com/ news/
articles/ 2016- 10- 27/ inside- the- trump- bunker- with- 12- days- to- go (accessed September
10, 2018).
Kevin Poulson, “Cambridge Analytica’s Real Role in Trump’s Dark Facebook
Campaign,” e Daily Beast, December 10, 2018, available at hps:// www.thedaily-
beast.com/ cambridge- analyticas- real- role- in- trumps- dark- facebook- campaign (accessed
December 12, 2018). Colin Lecher, “Trump Campaign Using Targeted Facebook Posts
to Discourage Black Americans from Voting,” e Verge, October 17, 2016, available
at hps:// www.theverge.com/ 2016/ 10/ 27/ 13434246/ donald- trump- targeted- dark-
facebook- ads- black- voters (accessed August 12, 2018). Craig Timberg,
11. “Final Report on Russian Interference in the 2016 US Election,” issued by Senate Select
Commiee on Intelligence, Vols. 1 and 2, released July and October 2019, respec-
tively, available at hps:// www.intelligence.senate.gov/ sites/ default/ les/ documents/
Report_ Volume1.pdf Volume 2 hps:// www.intelligence.senate.gov/ sites/ default/ les/
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
284 Notes
284
documents/ Report_ Volume2.pdf (accessed November 10, 2019. For a useful summary of
the key ndings, see the summary by the US House of Representatives, Permanent Select
Commiee, “Exposing Russia’s Eort to Sow Discord Online: e Internet Research
Agency and Advertisements,” =available at hps:// intelligence.house.gov/ social- media-
content (accessed November 8, 2019).
12. Elizabeth Dwoskin, Adam Entous, and Karoun Demirjian, “Russian Ads, Now Publicly
Released, Show Sophistication of Inuence Campaign,” Washington Post, November 1,
2017, available at hps:// www.washingtonpost.com/ business/ technology/ russian-
ads- now- publicly- released- show- sophistication- of- influence- campaign/ 2017/ 11/
01/ d26aead2- bf1b- 11e7- 8444- a0d4f04b89eb_ story.html?utm_ term=.a8bcbccc2ede
(accessed August 12, 2018).
13. See the report published by the UK Electoral Commission, “Report on an Investigation
in Respect of the Leave.EU Group Limited,” published May 11, 2018, available at hps://
www.electoralcommission.org.uk/ sites/ default/ les/ pdf_ le/ Report- on- Investigation-
Leave.EU.pdf (accessed November 8, 2018); the UK Information Commissioners Oce,
“Investigation into the Use of Data Analytics in Political Campaigns,” available at hps://
ico.org.uk/ media/ action- weve- taken/ 2260271/ investigation- into- the- use- of- data-
analytics- in- political- campaigns- nal- 20181105.pdf (accessed December 8, 2018); and
the House of Commons Digital, Culture, Media and Sport Commiee, “Disinformation
and ‘Fake News’:Final Report,” published February 18, 2019, available at hps:// publica-
tions.parliament.uk/ pa/ cm201719/ cmselect/ cmcumeds/ 1791/ 1791.pdf (accessed June
5, 2019).
14. For an in- depth account of the use of data- driven campaigning in the 2018 Mexican elec-
tion, see the report of Cedric Laurent. “How do Mexican political parties use your data to
inuence your vote?” Published June 29, 2018 available at hps:// sontusdatos.org/ 2018/
06/ 29/ partidos_ politicos_ datos_ inuenciar_ voto (accessed April 30, 2020).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
285
285
References
Abramson, Jerey B., Christopher F. Arterton, and Gary R. Orren. 1988. e Electronic
Commonwealth:e Impact of New Media Technologies on Democratic Politics. New York:
Basic Books.
Akdeniz, Yaman. 2000. “Policing the Internet:Regulation and Censorship.” In Rachel Gibson and
Stephen Ward (eds.), Reinvigorating Democracy? British Politics and the Internet. Burlington,
VT:Ashgate:169– 188.
Akdeniz, Yaman, Nick Taylor, and Clive Walker. 2001. “Regulation of Investigatory Powers
Act 2000 (1): Bigbrother.gov.uk: State Surveillance in the Age of Information and
Rights.” Criminal Law Review (February): 73– 90. hps:// papers.ssrn.com/ sol3/ papers.
cfm?abstract_ id=3181278
Aldrich, John H., Rachel K. Gibson, Marta Cantijoch, and Tobias Konitzer. 2016. “Geing Out
the Vote in the Social Media Era:Are Digital Tools Changing the Extent, Nature and Impact
of Party Contacting in Elections?” Party Politics 22(2):165– 178.
Alexander, Douglas, and Stella Creasy. 2006. Serving a Cause, Serving a Community:e Role of
Political Parties in Today’s Britain. London:Demos. Available for download at hp:// www.
demos.co.uk/ les/ Servingacause.pdf (accessed June 26, 2018).
Allco, Hunt, and Mahew Gentzkow. 2017. “Social Media and Fake News in the 2016 Election.”
Journal of Economic Perspectives 31(2):211– 236.
Anderson, Paul. 2007. What Is Web 2.0? Ideas, Technologies and Implications for Education.
Bristol: JISC. Available at hp:// www.jisc.ac.uk/ media/ documents/ techwatch/ tsw0701b.
pdf (accessed June 11, 2017).
Andeweg, Rudy, and David Farrell. 2017. “Legitimacy decline and party decline.” In Carolien van
Ham, Carolien, Jacque omassen, Jacques, Kees Aarts, Kees, and Rudy Andeweg (eds.),
Myth and Reality of the Legitimacy Crisis:Explaining Trends and Cross- National Dierences in
Established Democracies. Oxford:Oxford University Press, Oxford:76– 94.
Anstead, Nick. 2017. “Data- Driven Campaigning in the 2015 United Kingdom General Election.”
e International Journal of Press/ Politics 22(3):294– 313.
Anstead, Nick, and Andrew Chadwick. 2009. “Parties, Election Campaigning, and the
Internet:Toward a Comparative Institutional Approach.” In Andrew Chadwick and Philip N.
Howard (eds.), Routledge Handbook of Internet Politics. NewYork:Routledge:56– 71.
Anstead, Nick, and Will Straw (eds.). 2009. e Change We Need:What Britain Can Learn om
Obama’s Victory. London:Fabian Society.
Armstrong, Jerome., 2006. Crashing the Gate:Netroots, Grassroots, and the Rise of People- Powered
Politics. White River Junction, VT:Chelsea Green.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
286 References
286
Arnaudo, Dan. 2018. “Computational Propaganda in Brazil: Social Bots during Elections.” In
Samuel Woolley and Philip N. Howard (eds.), Working Paper 2017.8. Oxford, UK:Project
on Computational Propaganda. hps:// ora.ox.ac.uk/ objects/ uuid:e88de32c- baaa- 4835-
bb76- e00473457f46/ download_ file?file_ format=pdf&safe_ filename=Comprop- Brazil-
1.pdf&type_ of_ work=Working+paper. (accessed December 1, 2018).
Arterton, Christopher F.1987. Teledemocracy: Can Technology Protect Democracy? Newbury
Park:Sage Publications.
Australian Centre for Public Communication (ACPC). 2008. “E- Electioneering: Use of New
Media in the 2007 Australian Federal Election.” Sydney:University of Technology Sydney.
Available at hp:// www.acpc.uts.edu.au/ publications/ e- electioneering- research- report-
online.pdf (accessed September 22, 2009).
Auty, Caroline, and Alison Cowen. 2001. “Political Parties on the Net:4 Years Closer to Cyber
Utopia.” Aslib Proceedings 53(9):341– 352.
Auty, Caroline, and David Nicholas. 1998. “British Political Parties and eir Web Pages.” Aslib
Proceedings 50(10):283– 296.
Baldwin- Philippi, Jessica. 2017. “e Myths of Data- Driven Campaigning.” Political Communication
34(4):627– 633.
Barber, Benjamin. R. 1997. “e New Telecommunications Technology:Endless Frontier or the
End of Democracy?” Constellations 4:208– 228.
Barber, Benjamin R. 1998. “ree Scenarios for the Future of Technology and Strong Democracy.”
Political Science Quarterly 113(4):573– 589.
Bartle, Jamie, Sid Benne, Rutger Birnie, and Simon Wibberley. 2013. “Virtually Members:e
Facebook and Twier Followers of UK Political Parties.” ACentre for the Analysis of Social
Media (CASM) paper, Demos, London. Available at hps:// www.demos.co.uk/ les/
Virtually_ Members.pdf?1366125743 (accessed June 28, 2018).
Bashyakarla, Varoon. 2018. “India:Digital Platforms, Technologies and Data in the 2014 and
2019 Elections.” Areport published by Tactical Technology Collective’s Our Data Our Selves
project, Berlin. Available at hps:// ourdataourselves.tacticaltech.org/ posts/ overview- india/
(accessed December 1, 2018).
Baumgartner, Jody C., and Jonathon S. Morris. 2010. “Who Wants to Be My Friend? Obama,
Youth and Social Networks in the 2008 Campaign.” In John Allen Hendricks and Robert E.
Denton Jr. (eds.), Communicator in Chief:How Barack Obama Used New Media Technology to
Win the White House. Plymouth, UK:Lexington Books:51– 66.
Becker, eodore Lewis, and Christa Daryl Slaton. 2000. e Future of Teledemocracy. Westport,
CT:Praeger.
Benne, Colin J. 2016. “Voter Databases, Micro- Targeting and Data Protection Law:Can Political
Parties Campaign in Europe as ey Do in North America?” International Data Privacy Law
6(4):261– 275.
Benne, Lance W., Andrea Segerberg, and Curb B.Knüpfer 2018. “e democratic interface:tech-
nology, political organization, and diverging paerns of electoral representation.” Information,
Communication & Society, 21(11):1655– 1680.
Benne, W. Lance, and Michael Xenos. 2004. “Young Voters and the Web of Politics 2004:e
Youth Political Web Sphere Comes of Age.” CIRCLE working paper. Available at hp://
www.civicyouth.org/ PopUps/ WorkingPapers/ WP42BenneXenos.pdf (accessed June
26, 2018).
Benoit, William J., and Pamela J. Benoit. 2000. “e Virtual Campaign: Presidential Primary
Websites in Campaign 2000.” American Communication Journal 3(3):1– 22.
Bimber, B. 2014. “Digital Media in the Obama Campaigns of 2008 and 2012:Adaptation to the
Personalized Political Communication Environment.” Journal of Information Technology &
Politics 11(2):130– 150.
Bimber, Bruce. 1998. “e Internet and Political Transformation: Populism, Community and
Accelerated Pluralism.” Polity 31(1):133– 160.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
References 287
287
Bimber, Bruce, and Richard Davis. 2003. Campaigning Online: e Internet in US Elections.
Oxford:Oxford University Press.
Bishop, Patrick, and Lori Anderson. 2004. “e- Government to e- Democracy: High- Tech
‘Solutions’ to ‘No- Tech’ Problems?” In Peter Chen and Winsome Roberts (eds.), Proceedings
of the Australian Electronic Governance. Melbourne:Centre for Public Policy, University of
Melbourne:1– 22.
Bonchek, Mark S. 1995. “Grassroots in Cyberspace: Using Computer Networks to Facilitate
Political Participation.” Paper presented at the Annual Meeting of the Midwest Political
Science Association, Chicago, IL.
Bossea, Michael. 2018. “e Digital Architectures of Social Media: Comparing Political
Campaigning on Facebook, Twier, Instagram, and Snapchat in the 2016 US Election.”
Journalism & Mass Communication Quarterly 95(2):471– 496.
Bousquet, Franck. 2009. “Le net- campagne de 2007: Entre mobilisation et participa-
tion.” In P. Maarek (ed.), La communication politique de la présidentielle de 2007.
Paris:L’Harmaan:105– 120.
Bowers- Brown, J. and B. Gunter. 2002. “Political Parties’ Use of the Web during the 2001 General
Election.” Aslib Proceedings 54(3):166– 176.
Boyd, Ryan L., Alexander Spangher, Adam Fourney, Besmira Nushi, Gireeja Ranade, James
Pennebaker, and Eric Horvitz. “Characterizing the Internet Research Agency’s Social
Media Operations during the 2016 US Presidential Election using Linguistic Analyses.”
Available at le:// nask.man.ac.uk/ home$/ Downloads/ Boyd%20et%20al.%20(2018)%20-
%20Characterizing%20the%20I%20Whitepaper%20- %20Final%20(4).pdf (accessed
November 1, 2019).
Braen, L. Clare. 2005. “Online Zealotry: La France du peuple virtuel.” New Media & Society
7(4):517– 532.
Breindl, Yana, and Bjoern Kuellmer. 2013. “Internet Content Regulation in France and
Germany: Regulatory Paths, Actor Constellations, and Policies.” Journal of Information
Technology & Politics 10(4):369– 388.
Browning, Graeme. 1996. Electronic Democracy:Using the Internet to Transform American Politics.
Wilton, CT:PemPress.
Bruns, Axel. 2019. Are Filter Bubbles Real? :Cambridge, UK:Polity Press.
Bruns, Axel, and Tim Higheld. 2013. “Political Networks on Twier:Tweeting the Queensland
State Election.” Information, Communication & Society 16(5):667– 691.
Budge, Ian. 1996. e New Challenge of Direct Democracy. Cambridge, MA:Polity Press.
Burdick, Eugene. 1964. e 480. NewYork:McGraw- Hill.
Carey, John M., and Mahew S. Shugart. 1995. “Incentives to Cultivate a Personal Vote:A Rank
Ordering of Electoral Formulas.” Electoral Studies 14(4):417– 439.
Carlson, Tom, and Göran Djupsund. 2001. “Old Wine in New Boles? e 1999 Finnish Election
Campaign on the Internet.” Harvard International Journal of Press/ Politics 6(1):68– 87.
Carlson, Tom, and Kim Strandberg. 2008. “Riding the Web 2.0 Wave:Candidates on YouTube
in the 2007 Finnish National Elections.” Journal of Information Technology & Politics
5(2):159– 174.
Castells, Manuel. 2009. Communication Power. Oxford:Oxford University Press.
Chadwick, Andrew. 2008. “Web 2.0:New Challenges for the study of E- Democracy in an Era
of Informational Exuberance.” I/ S: A Journal of Law and Policy for the Information Society
5(1):9– 41.
Chadwick, Andrew. 2007. “Digital Network Repertoires and Organizational Hybridity.” Political
Communication 24(3):283– 301.
Chen, Peter J. 2013. Australian Politics in a Digital Age. Canberra: Australian National
University Press.
Chen, Peter J.2008a. “Australian Political Parties’ Use of YouTube 2007.” Communication, Politics
& Culture 41(1):114– 148.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
288 References
288
Chen, Peter J. 2008b. “Candidates’ New Media Use in the 2007 Australian National Election.”
In Franco Papandrea and Mark Armstrong (eds.), Record of the Communications Policy and
Research Forum 2008. Sydney:Network Insight Institute:62– 78. Available at hps:// web.
archive.org/ web/ 20120227075350/ hp:// www.networkinsight.org/ verve/ _ resources/
Record_ CPRF08.pdf (accessed June 26, 2018).
Chen, Peter J. 2004. “e- lection 2004? New Media and the Campaign.” Available at hp:// mem-
bers.optusnet.com.au/ pjchen/ Downloads/ e- lection%202004.pdf (accessed July 5, 2005).
Chen, Peter J. 2005. “e New Media:E- lection 2004?” In M. Simms, J. Warhurst, and R. Nile
(eds.), e Mortgage Nation: e 2004 Australian Election. Perth:API Network, Australia
Research Institute, Curtin University of Technology:117– 130.
Chen, Peter J., Rachel Gibson, Wainer Lusoli, and Stephen Ward. 2007. “Australian Governments
and Online Communication.” In Sally Young (ed.), Government Communication in Australia.
NewYork:Cambridge University Press:161– 180.
Chen, Peter J., and Lucas Walsh. 2010. “E- election 2007? Political Competition Online.” Australian
Cultural History 28(1):47– 54.
Clayton, Dewey M. 2010. e Presidential Campaign of Barack Obama:A Critical Analysis of a
Racially Transcendent Strategy. NewYork:Routledge.
Cli, S. 1997. “Building Citizen- Based Electronic Democracy Eorts.” Paper presented at “Internet
and Politics:e Modernization of Democracy through the Electronic Media.” Conference
of the Academy of the ird Millennium, Munich, Germany. Available at hp:// stevencli.
com/ building- citizen- based- electronic- democracy- eorts- by- steven- cli- 1997/ (accessed
June 26, 2018).
Cli, Steven. 2002. “e- Governance to e- Democracy: Progress in Australia and New Zealand
toward Information- Age Democracy.” Publicus.Net: Public Strategies for the Online World.
Available at www.publicus.net/ articles/ aunzedem.html (accessed January 14, 2007).
Cli, Steven. 2000. “e e- democracy e- book: Democracy is Online 2.0.” Available at hp://
www.publicus.net/ ebook/ edemebook.html (accessed June 26, 2018).
Cogburn, Derrick L., and Fatima K. Espinoza- Vasquez. 2011. “From Networked Nominee
to Networked Nation: Examining the Impact of Web 2.0 and Social Media on Political
Participation and Civic Engagement in the 2008 Obama Campaign.” Journal of Political
Marketing 10(1– 2):189– 213.
Coleman, Stephen. 2005. “It’s Interactivity Stupid!” In Stephen Ward and Stephen Coleman
(eds.), Spinning the Web:Online Campaigning in the 2005 General Election. London:Hansard
Society:5– 12.
Coleman, Stephen. 2001a. “Online Campaigning.” Parliamentary Aairs 54(4):679– 688.
Coleman, Stephen (ed.). 2001b. 2001 CyberSpace Odyssey: e Internet in the UK Election.
London:Hansard Society. Available at hps:// assets.contentful.com/ u1rlv vbs33ri/ 5khyKS-
6f YsAEgc8OU4IuGI/ 43c9b2a796ee89313f0d95db4297cb49/ Publication_ _ 2001- Cyber-
Space- Odyssey- e- internet- in- the- UK- Election- 2001.pdf (accessed October 6, 2017).
Coleman, Stephen, and Nicola Hall. 2001. “Spinning on the Web:E- campaigning and Beyond.”
In Stephen Coleman (ed.), 2001: CyberSpace Odyssey: e Internet in the UK Election.
London:Hansard Society:9– 21.
Colville, Robert. 2008. Politics, Policy and the Internet. London:Center for Policy Studies.
Comstock, George. 1978. “e Impact of Television on American Institutions.” Journal of
Communication 28(2):12– 28.
Conway, Mahew, and Dan Dorner. 2004. “An Evaluation of New Zealand Political Party Websites.”
Information Research 9(4). Available at hp:// informationr.net/ ir/ 9- 4/ paper196.html.
Cook, Colleen, Fred Heath, and Russel L. ompson. 2000. “A Meta- analysis of Response Rates
in Web- or Internet- Based Surveys.” Educational and Psychological Measurement 60:821– 836.
Copsey, Nigel. 2003. “Extremism on the Net:e Extreme Right and the Value of the Internet.”
In Rachel Gibson, Paul Nixon, and Stephen Ward (eds.), Political Parties and the Internet:Net
Gain? Abingdon, UK:Routledge:218– 234.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
References 289
289
Conners, Joan. 2005. “Meetup, blogs, and online involvement:US Senate campaign websites of
2004.” Paper presented at the Annual Meeting of the American Political Science Association,
Washington, D.C.
Costa, Pere- Oriol. 2009. “Barack Obama’s Use of the Internet Is Transforming Political
Communication.” Quaderns del CAC 33:35– 41.
Crabtree, James. 2010. “David Cameron’s Bale to Connect.” Wired Magazine. (March). Available
at hp:// www.wired.co.uk/ article/ david- camerons- bale- to- connect (accessed June
26, 2018).
Crabtree, James. 2001. “Whatever Happened to the E- lection? A Survey of Voter Aitudes
towards New Technology during the 2001 Election.” An iSociety Report. London: e
Industrial Society.
Crook, E. 2007. “e 2007 Australian Federal Election on the Internet.” Canberra: National
Library of Australia Sta Papers.
Cruddas, J., and J. Harris. 2006. “Fit for Purpose:A Programme for Party Renewal.” Compass.
Available online at hp:// www.compassonline.org.uk/ wp- content/ uploads/ 2013/ 05/ t_
for_ purpose.pdf (accessed June 26, 2018).
D’Alessio, D. 1997. “Use of the World Wide Web in the 1996 US Election.” Electoral Studies
16(4):489– 500.
Dahlberg, Lincoln. 2001. “Extending the Public Sphere through Cyberspace: e Case of
Minnesota E- Democracy.” First Monday 6(3). Available at hp:// rstmonday.org/ ojs/ index.
php/ fm/ article/ view/ 838/ 747 (accessed June 26, 2018).
Dalton, Russsell. J., 2008. e quantity and the quality of party systems:Party system polarization,
its measurement, and its consequences. Comparative Political Studies, 41(7):899– 920.
Dalton, Russell J., David Farrell, and Ian McAllister. 2011. Political Parties and Democratic
Linkage:How Parties Organize Democracy. Oxford:Oxford University Press.
Darras, Eric. 2008. “Free Journalism under Control: Election Coverage in France.” In Jesper
Strömbäck and Lynda Lee- Kaid (eds.), e Handbook of Election Coverage around the World.
Abingdon Oxon:Routledge:90– 108.
Davis, Richard, Jody C. Baumgartner, Peter L. Francia, and Jonathan S. Morris. 2009. “e Internet
in US Election Campaigns.” In Andrew Chadwick and Philip N. Howard (eds.), Routledge
Handbook of Internet Politics. NewYork:Routledge::25– 39.
Davis, Richard, and Diana Owen. 1998. New Media and American Politics. NewYork: Oxford
University Press
Davis, Richard, and Diana Owen. 2008. “United States: Internet and Elections.” In Stephen
Ward, Diana Owen, Richard Davis, and David Taras (eds.), Making a Dierence? Internet
Campaigning in Comparative Perspective. Plymouth, UK:Lexington Books:93– 112.
Davis, Steve, Larry Elin, and Grant Reehe. 2002. Click on Democracy:e Internet’s Power to Change
Political Apathy into Civic Action. Boulder, CO:Westview.
de Sola Pool, Ithiel, Robert P. Abelson, and Samuel L. Popkin. 1965. “A Postscript on the 1964
Election.” American Behavioral Scientist 8(9):39– 44.
Denver, David T., and Gordon Hands. 1997. Modern Constituency Electioneering:Local Campaigning
in the 1992 General Election. London:Frank Cass.
Denver, David T., and Gordon Hands. 1992. “Constituency Campaigning.” Parliamentary Aairs
45(4):528– 544.
Dertouzos, Michael. 1997. What Will Be:How the New Information Marketplace Will Change Our
Lives. San Francisco:HarperOne.
Desrosières, Alain. 2008. Pour une sociologie historique de la quantication:L’argument statistique 1.
Paris:Presses de l’École des Mines.
Dillman, D. A. 2011. Mail and Internet Surveys:e Tailored Design Method— 2007 Update with
New Internet, Visual, and Mixed- Mode Guide. NewYork:John Wiley & Sons.
Dolezal, Martin. 2015. “Online Campaigning by Austrian Political Candidates:Determinants of
Using Personal Websites, Facebook, and Twier.” Policy & Internet 7(1):103– 119.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
290 References
290
Domme, Katherine, and Luke Temple. 2018. “Digital Campaigning:e Rise of Facebook and
Satellite Campaigns.” Parliamentary Aairs 71(1), 189– 202, hps:// doi.org/ 10.1093/ pa/
gsx056.
Druckman, James N., Martin J. Kifer, and Michael Parkin. 2014. “US Congressional Campaign
Communications in an Internet Age.” Journal of Elections, Public Opinion & Parties
24(1):20– 44.
Druckman, J.N., Kifer, M.J., and Parkin, M., 2007. “e technological development of congres-
sional candidate web sites:How and why candidates use web innovations.” Social Science
Computer Review, 25(4):425– 442.
Dyson, Esther. 1997. Release 2.0:A Design for Living in the Digital Age. NewYork:Broadway Books.
Epstein, Edward. 1996. “Election ’96 Internet Style.” P C Wor ld (May):174– 180.
Farmer, Rick, and Rich Fender. 2005. “E- parties: Democratic and Republican State Parties in
2000.” Party Politics 11(1):47– 58.
Farrell, David M. 2002. “Campaign Modernization and the West European Party.” In Kurt
Richard Luther (ed.), Political Parties in the New Europe:Political and Analytical Challenges.
Oxford:Oxford University Press:63– 83.
Farrell, David M., and Rüdiger Schmi- Beck. 2002. Do Political Campaigns Maer? Campaign
Eects in Elections and Referendums. Abingdon, Oxon, UK:Routledge.
Farrell, David M., and Paul Webb. 2000. “Political Parties as Campaign Organizations.” In Russell
Dalton (ed.), Parties Without Partisans:Political Change in Advanced Industrial Democracies.
Oxford:Oxford University Press:102– 128.
Faucheaux, Ron. 1998. “How Campaigns Are Using the Internet: An Exclusive Nationwide
Survey.” Campaigns & Elections 19(9):22– 25.
Feld, Lowell, and Nate Wilcox. 2008. Netroots Rising:How a Citizen Army of Bloggers and Online
Activists Is Changing American Politics. Westport, CT:Greenwood.
Firestone, Charles M., and Anthony Corrado (eds.). 1996. Elections in Cyberspace:Toward a New
Era in American Politics. Aspen Institute Program on Communications and Society. American
Bar Association, Standing Commiee on Election Law. Washington, DC:Aspen Institute.
Fisher, Justin, David Cus, and Ed Fieldhouse. 2011. “Constituency Campaigning in 2010.” In
Dominic Wring, Roger Mortimore, and Simon Atkinson (eds.), Political Communication
in Britain:e Leader’s Debates, the Campaign and the Media in the 2010 General Election.
London:Palgrave Macmillan:198– 217.
Fisher, Justin, David Cus, Ed Fieldhouse, and Beine Roweiler. 2015. “Constituency
Campaigning at the 2015 General Election:Campaigning in a Multi- Party Context and a
Changing Environment for Voter Contact.” Paper presented at Campaigning and Voting in
Europe:New Challenges, New Perspectives. Paris. Available at hps:// bura.brunel.ac.uk/
bitstream/ 2438/ 11618/ 1/ Fulltext.pdf (accessed July 18, 2018).
Flew, Terry. 2005. “From e- government to Online Deliberative Democracy.” Queensland
University of Technology (QUT) eprints series. Brisbane, Australia. Available at hp://
eprints.qut.edu.au/ 39368 (accessed June 27, 2018).
Foot, K. A., & S. M. Schneider. 2002. “Online Action in Campaign 2000:An Exploratory analysis
of the U.S. Political Web Sphere.” Journal of Broadcasting and Electronic Media 46(2):222– 244.
Foot, Kirsten, and Steven M. Schneider. 2006. Web Campaigning. Cambridge, MA:MIT Press.
Foot, Kirsten A., Steven M. Schneider, and Meghan Dougherty. 2007. “Online Structure
for Political Action in the 2004 US Congressional Electoral Web Sphere.” In Randolph
Kluver, Nicholas Jankowski, Kirsten A. Foot, and Steven M. Schneider (eds.), e Internet
and International Elections: A Comparative Study of Web Campaigning. Abingdon, Oxon,
UK:Routledge:92– 104.
Foot, Kirsten A., Steven M. Schneider, Randolph Kluver, Michael Xenos, and Nicholas Jankowski.
2007. “Comparing Web Productions Practices across Electoral Spheres.” In Randolph
Kluver, Nicholas Jankowski, Kirsten A. Foot, and Steven M. Schneider (eds.), e Internet
and International Elections: A Comparative Study of Web Campaigning. Abingdon, Oxon,
UK:Routledge:243– 260.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
References 291
291
Francoli, Mary, and Stephen Ward. 2008. “21st Century Soapboxes? MPs and eir Blogs.”
Information Polity 13(1– 2):21– 39.
Gelman, Andrew. 2006. “Prior Distributions for Variance Parameters in Hierarchical Models.”
Bayesian Analysis 1(3):515– 534.
Gerber, Alan S., and Donald P. Green. 2000. “e Eects of Canvassing, Telephone Calls, and
Direct Mail on Voter Turnout: A Field Experiment.” American Political Science Review
94(3):653– 663.
Gerber, Alan S., and Donald P.Green. 2008. Get Out the Vote:How to Increase Voter Turnout (2nd
ed). Washington, DC:Brookings Institution Press.
Gerstenfeld, Phyllis, Diana Grant, and Chau Pu Chiang. 2003. “Hate Online:A Content Analysis
of Extremist Internet Sites.” Analyses of Social Issues and Public Policy 3:29– 44.
Giasson, ierry, Fabienne Greet, and Geneviève Chacon. 2014. “Digital Campaigning in a
Comparative Perspective:Online Strategies in the 2012 Elections in France and Québec.”
Paper presented at the Biennial Congress of the International Political Science Association.
Montreal. July.
Gibson, Rachel K. 2012. “From Brochureware to ‘MyBo’:An Overview of Online Elections and
Campaigning.” Politics 32(2):77– 84.
Gibson, Rachel K. 2015. “Party Change, Social Media and the Rise of ‘Citizen- Initiated’
Campaigning.” Party Politics 21(2):183– 197.
Gibson, Rachel. 2004. “Web Campaigning from a Global Perspective.” Asia- Pacic Review
11(1):95– 126.
Gibson, Rachel K., Kevin Gillan, Fabienne Greet, Ben Lee, and Stephen Ward. 2013. “Party
Organizational Change and ICTs:e Growth of a Virtual Grassroots?” New Media & Society
15(1):31– 51.
Gibson, Rachel K., Wainer Lusoli, and Stephen Ward. 2008. “Nationalizing and Normalizing the
Local? AComparative Analysis of Online Candidate Campaigning in Australia and Britain.”
Journal of Information Technology & Politics 4(4):15– 30.
Gibson, Rachel K., and Ian McAllister, 2008. “Designing Online Election Surveys:Lessons from
the 2004 Australian Election.” Journal of Elections, Public Opinion and Parties 18(4):387– 400.
Gibson, Rachel K., and Ian McAllister. 2011. “Do Online Election Campaigns Win Votes? e
2007 Australian ‘YouTube’ Election.” Political Communication 28(2):227– 244.
Gibson, Rachel K., and Ian McAllister. 2006. “Does Cyber‐Campaigning Win Votes? Online
Communication in the 2004 Australian Election.” Journal of Elections, Public Opinion and
Parties, 16(3):243– 263.
Gibson, Rachel K., and Ian McAllister. 2015. “Normalising or Equalising Party Competition?
Assessing the Impact of the Web on Election Campaigning.” Political Studies 63(3):529– 547.
Gibson, Rachel K., and Andrea Römmele. 2020. “Scientic and Subversive:e Two Faces of e
Fourth Era of Political Campaigning. New Media & Society 22(4):595– 610.
Gibson, Rachel K., and Stephen Ward. 2000a. “An Outsiders Medium? e European Elections
and UK Party Competition on the Internet.” In Philip Cowley, David Denver, Andrew
Russell, and Lisa Harrison (eds.), British Parties and Elections Review. London: Frank
Cass:173– 192.
Gibson, Rachel K., and Stephen Ward. 2000b. “British Party Activity in Cyberspace: New
Media, Same Impact?” In Rachel K. Gibson and Stephen Ward (eds.), Reinvigorating
Democracy:British Politics and the Internet. Aldershot, UK:Ashgate:107– 128.
Gibson, Rachel K., and Stephen Ward. 2009. “Parties in the Digital Age: A Review Article.”
Representation 45(1):87– 100.
Gibson, Rachel K., and Stephen Ward. 1999. “Party Democracy On- line: UK Parties and New
ICTs.” Information, Communication & Society 2(3):340– 367.
Gibson, Rachel K., and Stephen Ward. 1998. “UK Political Parties and the Internet:Politics as
Usual in the New Media.” Harvard International Journal of Press Politics 3(3):14– 38.
Gibson, Rachel K., and Stephen Ward. 2002. “Virtual Campaigning:Australian Parties and the
Impact of the Internet.” Australian Journal of Political Science 37(1):99– 129.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
292 References
292
Gibson, Rachel K., Stephen Ward, and Paul Nixon. 2003a. “Parties and the Internet:An Overv iew.”
In Rachel Gibson, Paul Nixon, and Stephen Ward (eds.), Political Parties and the Internet:Net
Gain? Abingdon, UK:Routledge:11– 38.
Gibson, Rachel K., Michael Margolis, David Resnick, and Stephen Ward. 2003b. “Election
Campaigning on the WWW in the USA and UK: A Comparative Analysis.” Party Politics
9(1):47– 75.
Gibson, Rachel K., Andy Williamson, and Stephen Ward (eds.). 2010. e Internet and the 2010
Election:Puing the Small “p” Back in Politics? London:Hansard Society. Available at hps://
assets.ctfassets.net/ u1rlvvbs33ri/ 1T8tpdVcsIkyyeuGw62mG6/ a735e387b5975f47d-
19b7d79006266db/ Publication_ _ e- internet- and- the- 2010- election- 2010.pdf (accessed
May 3, 2020).
Gilmore, Jason. 2012. “Ditching the Pack: Digital Media in the 2010 Brazilian Congressional
Campaigns. New Media & Society 14(4):617– 633.
Glowacki, Monika, Vidya Narayanan, Sam Maynard, Gustavo Hirsch, Bence Kollanyi, Lisa- Marie
Neudert, Philip Howard, omas Lederer, and Vlad Barash. 2018. “News and Political
Information Consumption in Mexico:Mapping the 2018 Mexican Presidential Election on
Twier and Facebook.” Computational Propaganda Data Memo, June 29. Available at hp://
comprop.oii.ox.ac.uk/ wp- content/ uploads/ sites/ 93/ 2018/ 06/ Mexico2018.pdf (accessed
September 1, 2018).
Gonzalez, R. J. 2017. “Hacking the Citizenry? Personality Proling, ‘Big Data,’ and the Election of
Donald Trump.” Anthropology Today 33(3):9– 12.
Gosnell, Harold F. 1927. Geing Out the Vote. Chicago:University of Chicago Press.
Graham, Todd, Marchel Broersma, Karin Hazelho, and Guido van’t Haar. 2013. “Between
Broadcasting Political Messages and Interacting with Voters: e Use of Twier dur-
ing e 2010 UK General Election Campaign.” Information, Communication & Society
16(5):692– 716.
Greet, Fabienne. 2001. “Les Partis politiques francais sur le Web.” In Dominique Andolfao,
Fabienne Greet, and Laurent Olivier (eds.), Les partis politiques: Quelles perspectives?
Paris:L’Harmaan (coll. Logiques politiques):161– 178.
Greet, Fabienne. 2013. “New Techniques, New Mobilizations?” In Paul Nixon, Rajash Rawal,
and Dan Mercea (eds.), Politics and the Internet in Comparative Context:Views om the Cloud.
NewYork:Routlege:59– 74.
Grinberg, N., Joseph, K., Lisa Friedland, Briony Swire- ompson, and David Lazer. 2019. “Fake
News on Twier During the 2016 US Presidential Election.” Science 363(6425):374– 378.
Grossman, Lawrence K. 1995. e Electronic Republic:Reshaping Democracy in the Information Age.
NewYork:Viking Penguin.
Gueorguieva, Vassia. 2008. “Voters, MySpace, and YouTube: e Impact of Alternative
Communication Channels on the 2006 Election Cycle and Beyond.” Social Science Computer
Review 26(3):288– 300.
Hall, Mark. 1997. “One to One Politics in Cyberspace.” Media Studies Journal 2:97– 103.
Hansen, Glenn J., and William L. Benoit. 2005. “Presidential Campaigning on the Web: e
Inuence of Candidate World Wide Web Sites in the 2000 General Election.” Southern Journal
of Communication 70(3):219– 228.
Hansen, Kasper M., and Karina Kosiara- Pedersen. 2014. “Cyber- campaigning in
Denmark: Application and Eects of Candidate Campaigning.” Journal of Information
Technology & Politics 11(2):206– 219.
Harfoush, Rahaf. 2009. Yes We Did! An Inside Look at How Social Media Built the Obama Brand.
Berkeley, CA:New Riders.
Heidar, Knut, and Jo Saglie. 2003. “Predestined Parties? Organizational Change in Norwegian
Political Parties.” Party Politics 9(2):219– 239.
Herrnson, Paul S., Atiya Kai Stokes- Brown, and Mahew Hindman. 2007. “Campaign Politics and
the Digital Divide:Constituency Characteristics, Strategic Considerations, and Candidate
Internet Use in State Legislative Elections.” Political Research Quarterly 60(1):31– 42.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
References 293
293
Hersh, Eitan. D.2015. Hacking the Electorate:How Campaigns Perceive Voters. NewYork:Cambridge
University Press.
Hindman, Mahew. 2005. “e Real Lessons of Howard Dean:Reections on the First Digital
Campaign.” Perspectives on Politics 3(1):121– 128.
Ho, Jens, and Karl Lofgren. 1997. “Political Parties and Computer Mediated Communication in
Denmark:General Use and Strategy.” Paper presented to the XIth Meeting of the Permanent
Study Group on Informatization in Public Administration, European Group of Public
Administration Annual Conference. Leuven, Belgium.
Howard, Philip N. 2006. New Media Campaigns and the Managed Citizen. NewYork:Cambridge
University Press.
Howard, Philip N., Bharath Ganesh, Doimitra Liotsiou, John Kelly, and Camille François.
2018. “e I, Social Media and Political Polarization in the United States, 2012– 2018.”
Oxford:Project on Computational Propaganda. Available at comprop.oii.ox.ac.uk.
Howard, Philip N., and Muzammil M. Hussain. 2013. Democracy’s Fourth Wave? Digital Media and
the Arab Spring. NewYork:Oxford University Press
Issenberg, Sasha. 2012. e Victory Lab:e Secret Science of Winning Campaigns. NewYork:Crown.
Jackson, Nigel. 2007. “Political Parties, the Internet and the 2005 General Election:ird Time
Lucky?” Internet Research 17(3):249– 271.
Janda, Kenneth. 2015. “Innovations in Information Technology in American Party Politics since
1960.” In Guy Lachappell and Philippe J. Maarek (eds.), Political Parties in the Digital Age:e
Impact of New Technologies in Politics. Berlin:Walter de Gruyter:9– 42.
Johnson, Dennis W. 2011. Campaigning in the Twenty- First Century:A Whole New Ballgame. Oxon,
UK:Routledge.
Johnson, Dennis W. 2007. No Place for Amateurs:How Political Consultants Are Reshaping American
Democracy, 2nd ed. Abingdon, Oxon:Routledge.
Johnson, Joel W., and Jessica S.Wallack, 2012, “Electoral Systems and the Personal Vote,” hps://
doi.org/ 10.7910/ DVN/ AMRXJA, Harvard Dataverse, V1.
Kalnes, Øvind. 2009. “Norwegian Parties and Web 2.0.” Journal of Information Technology &
Politics 6(3– 4):251– 266.
Kamarck, Elaine C. 1999. “Campaigning on the Internet in the Elections of 1998.” In Elaine C.
Kamarck and Joseph S. Nye Jr (eds.), Democracy.com? Governance in a Network World. Hollis,
NH:Hollis:99– 123.
Kamarck, Elaine C. 2002. “Political Campaigning on the Internet:Business as Usual?” In Elaine
C. Kamarck and Joseph S. Nye Jr. (eds.), Governance.com:Democracy in the Information Age.
Washington, DC:Brookings Institution Press:81– 103.
Kapor, Mitchell. 1993. “Where Is the Digital Highway Really Heading? e Case for a Jeersonian
Information Policy.” Wired Magazine (January 3). Available at hps:// www.wired.com/
1993/ 03/ kapor- on- nii/ (accessed June 29, 2018).
Karlsen, Rune. 2013. “Obama’s Online Success and European Party Organizations:Adoption
and Adaptation of US Online Practices in the Norwegian Labor Party.” Journal of Information
Technology & Politics 10(2):158– 170.
Karp, Jerey A. 2012. “Electoral Systems, Party Mobilisation and Political Engagement.” Australian
Journal of Political Science 47(1):71– 89.
Karp, Jerey A., and Susan Banducci. 2011. “e Inuence of Party and Electoral Systems on
Campaign Engagement.” In Russell J. Dalton and ChristopherJ. Anderson (eds.), Citizens,
Context, and Choice: How Context Shapes Citizens’ Electoral Choices. Oxford: Oxford
University Press:55– 78.
Karp, Jerey A., and Susan Banducci. 2007. “Party Mobilization and Political Participation in New
and Old Democracies.” Party Politics 13(2):217– 234.
Karp, Jerey A., Susan Banducci, and Shaun Bowler. 2008. “Geing Out the Vote: Party
Mobilization in a Comparative Perspective.” British Journal of Political Science
38(1):91– 112.
Katwala, Sunder, and Richard Brooks. 2005. “Come Together.” Fabian Review (Autumn):10– 15.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
294 References
294
Kaye, Barbara K., and omas J. Johnson. 1999. “Research Methodology: Taming the Cyber
Frontier: Techniques for Improving Online Surveys.” Social Science Computer Review
17(3):323– 337.
Keller, Franziska B., David Schoch, Sebastian Stier, and JungHwan Yang. 2020. “Political
Astroturng on Twier: How to Coordinate a Disinformation Campaign.” Political
Communication 37(2):256– 280.
Kellerman, Aharon. 2006. “Broadband Penetration and Its Implications:e Case of France.”
Networks and Communication Studies NETCOM 20(3– 4):237– 246.
Kellerman, Aharon. 1999. “Leading Nations in the Adoption of Communications Media, 1975 to
1995.” Urban Geography 20(4):377– 389.
Kies, Raphael. 2004. “e E- Democracy Strategy of Partito Radicale Online:A Case Study for
Scientic and Technological Option Assessment (STOA).” Annex Iin Alexander Trechsel,
Raphael Kies, Fernando Mendez, and Phillip C.Schmier, “Evaluation of the Use of New
Technologies in Order to Facilitate Democracy.” European Parliament working paper,
Scientic and Technological Options Assessment Series STOA 116 EN. 48– 57. Available at
hp:// orbilu.uni.lu/ bitstream/ 10993/ 21778/ 1/ IPOL- JOIN_ ET(2003)471583_ EN.pdf
(accessed June 26, 2018).
Kim, Yong- Ho. 2006. “Political Dynamics of Roh Moo- hyun’s Netizen Fan Club (Nosamo) and the
2002 Presidential Election in South Korea.” Paper presented at the XX IPSA World Congress
of Political Science, Fukuoka, Japan. Available at hp:// paperroom.ipsa.org/ papers/ paper_
5405.pdf (accessed July 9, 2018).
Klinenberg, Eric, and Andrew Perrin. 2000. “Symbolic Politics in the Information Age:e 1996
Republican Presidential Campaigns in Cyberspace.” Information, Communication & Society
3(1):17– 38.
Klinger, Ulrike, 2013. “Mastering the Art of Social Media: Swiss Parties, the 2011
National Election and Digital Challenges.” Information, Communication & Society
16(5):717– 736.
Klotz, Robert J. 2004. e Politics of Internet Communication. Lanham, MD:Rowman & Lileeld.
Knopf, Allegra. 1999. “Privacy and the Internet:Welcome to the Orwellian World.” University of
Florida Journal of Law & Public Policy 11:79– 100.
Koc- Michalska, Karolina, Rachel K. Gibson, and ierry Vedel. 2014. “Online Campaigning in
France, 2007– 2012:Political Actors and Citizens in the Aermath of the Web. 2.0 Evolution.”
Journal of Information Technology & Politics 11(2):220– 244.
Kovic, Marko, Adrian Raucheisch, Marc Sele, and Christian Caspar. 2018. “Digital Astroturng
in Politics:Denition, Typology, and Countermeasures.” Studies in Communication Sciences
18(1):69– 85. hps:// doi.org/ 10.24434/ j.scoms.2018.01.005
Kreiss, Daniel. 2009. “Developing the ‘Good Citizen’: Digital Artifacts, Peer Networks, and
Formal Organization during the 2003– 2004 Howard Dean Campaign.” Journal of Information
Technology & Politics 6(3– 4):281– 297.
Kreiss, Daniel. 2016. Prototype Politics: Technology- Intensive Campaigning and the Data of
Democracy. NewYork:Oxford University Press.
Kreiss, Daniel. 2012. Taking Our Country Back:e Craing of Networked Politics om Howard
Dean to Barack Obama. NewYork:Oxford University Press.
Kreiss, Daniel, and Philip Howard. 2010. “New Challenges to Political Privacy: Lessons from
the rst US Presidential Race in the Web 2.0 Era.” International Journal of Communication
4:1032– 1050.
Kreiss, Daniel, and Creighton Welch. 2015. “Strategic Communication in a Networked Age.” In
Victoria A. Farrar- Myers and Justin S. Vaughn (eds.), Controlling the Message:New Media in
American Political Campaigns. NewYork:NewYork University Press:13– 31.
Kreiss, Daniel, and Shannon C.McGregor. 2018. “Technology rms shape political communica-
tion:e work of Microso, Facebook, Twier, and Google with campaigns during the 2016
US presidential cycle.” Political Communication 35(2):155– 177.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
References 295
295
Ku, Gyotae, Lynda Lee Kaid, and Michael Pfau. 2003. “e Impact of Website Campaigning
on Traditional News Media and Public Information Processing.” Journalism & Mass
Communication Quarterly 80(3):528– 547.
Laakso, Marku, and Rein Taagepera. 1979. “Eective” number of parties:a measure with applica-
tion to West Europe. Comparative Political Studies 12(1):3– 27.
Labour NEC Party Development Taskforce. 2005. “Building a Party for the Future.” Labour Party
Annual Conference, Brighton (September 25– 29).
Labour Party. 2011. “Refounding Labour:A Part of the New Generation.” Available at hps://
debatsocialista.les.wordpress.com/ 2011/ 03/ refoundinglabour.pdf (accessed June
29, 2018).
Larsson, Anders O. 2017. “Going Viral? Comparing Parties on Social Media during the 2014
Swedish Election.” Convergence 23(2):117– 131.
Lazarsfeld, Paul F., Bernard R. Berelson, and Hazel Gaudet. 1948. e People’s Choice:How the
Voter Makes Up His Mind in a Presidential Campaign. NewYork:Duell, Sloan & Pierce.
Leiter, Lisa. 1995. “Goodbye BBQ and Buons; Hello Web and Home Page.” Insight on the News
11(37): 46– 48. October 2. Available at hps:// www.questia.com/ read/ 1G1- 17519009/
goodbye- bbq- and- buons- hello- web- and- home- page (accessed November 6, 1995).
Lessig, Lawrence. 1999 Code:And Other Laws of Cyberspace. NewYork:Basic Books.
Lessig, Lawrence. 2000. “e Death of Cyberspace.” Washington & Lee Law Review 5 (2):337– 347.
Levenshus, Abbey. 2010. “Online Relationship Management in a Presidential Campaign:A Case
Study of the Obama Campaign’s Management of Its Internet- Integrated Grassroots Eort.”
Journal of Public Relations Research 22(3):313– 335.
Liégey, Guillaume, Arthur Muller, and Vincent Pons. 2013. Porte à porte:Reconquérir la démocratie
sur le terrain. Paris:Calmann- Lévy.
Lilleker, Darren G. 2013. “Empowering the Citizens? Political Communication, Co-
production and the Harnessed Crowd.” In Richard Scullion, Roman Gerodimos, Richard
Jackson, and Darren Lilleker (eds.), e Media, Political Participation and Empowerment.
London:Routledge:40– 54.
Lilleker, Darren G., and Nigel A. Jackson. 2011. Elections 2.0:Comparing e- Campaigns in France,
Germany, Great Britain and the United States. London:Routledge.
Lilleker, Darren G., and Nigel A. Jackson. 2010. “Towards a More Participatory Style of Election
Campaigning:e Impact of Web 2.0 on the UK 2010 General Election.” Policy & Internet
2(3):69– 98.
Lilleker, Darren G., Karolina Koc- Michalska, Eva J. Schweitzer, Michal Jacunski, Nigel A. Jackson,
and ierry Vedel. 2011. “Informing, Engaging, Mobilizing or Interacting:Searching for a
European Model of Web Campaigning.” European Journal of Communication 26(3):195– 213.
Lilleker, Darren G., and Casilda Malagón. 2010. “Levels of Interactivity in the 2007 French
Presidential Candidates’ Websites.” European Journal of Communication 25(1):25– 42.
Lilleker, Darren G., Jens Tenscher, and Václav Štětka. 2014. “Towards Hypermedia Campaigning?
Perceptions of New Media’s Importance for Campaigning by Party Strategists in Comparative
Perspective.” Information, Communication & Society, 18 (7):747– 765.
Löfgren, Karl, and Colin Smith. 2003. “Political Parties and Democracy in the Information Age.”
In Rachel Gibson, Paul Nixon, and Stephen Ward (eds.), Political Parties and the Internet:Net
Gain? Abingdon, UK:Routledge:39– 52.
Lotan, Gilad, Erhard Grae, Mike Ananny, Devin Gane, Ian Pearce, and danah boyd. 2011. “e
Arab Spring— e Revolutions Were Tweeted:Information Flows during the 2011 Tunisian
and Egyptian Revolutions.” International Journal of Communication 5:1375– 1405.
Maarek, Philippe J. 2015. “Political Communication, Social Media and Social Networks in France.”
In Guy Lachappell and Philippe J. Maarek (eds.), Political Parties in the Digital Age:e Impact
of New Technologies in Politics. Oldenberg:De Gruyter:165– 180.
Magalhães, Pedro C. 2007. “Voting and Intermediation: Informational Biases and Electoral
Choices in Comparative Perspective.” In Richard Gunther, Hans- Jürgen Puhle, and José R.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
296 References
296
Montero (eds.), Democracy, Intermediation, and Voting in Four Continents. Oxford: Oxford
University Press:208– 254.
Magalhães, Pedro C., John H.Aldrich, and Rachel K.Gibson. 2018. “New forms of mobilization,
new people mobilized? Evidence from the Comparative Study of Electoral Systems.” Par ty
Politics.
Magalhães, Pedro C., Paolo Segai, and Tianjian Shi. 2015. “Mobilization, Informal Networks and
the Social Contexts of Turnout.” In Richard Gunther, Paul A. Beck, Pedro C. Magalhães, and
Alejandro Moreno (eds.), Voting in Old and New Democracies. NewYork:Routledge:64– 98.
Magarey, Kirsty 1999. “e Internet and Australia Parliamentary Democracy.” Parliamentary
Aairs 52(3):404– 427.
Mair, Peter. 2013. Ruling e Void:e Hollowing of Western Democracy. London:Verso Books.
Marges, H. 2006. e Cyber Party. In Richard. S. Katz and William Croy (eds) Handbook of
Party Politics. London:Sage:528– 35.
Margolis, Michael and David Resnick. 2000. Politics As Usual: e Cyberspace “Revolution”.
London:Sage.
Margolis, Michael, David Resnick, and Chin- chang Tu. 1997. “Campaigning on the Internet
Parties and Candidates on the World Wide Web in the 1996 Primary Season.” e Harvard
International Journal of Press/ Politics 2(1):59– 78.
Margolis, Michael, David Resnick, and Joel D. Wolfe. 1999. “Party Competition on the Internet
in the United States and Britain.” Harvard International Journal of Press/ Politics 4(4):24– 47.
Matland, Robert E., and Donley T. Studlar. 1996. “e Contagion of Women Candidates in
Single- Member District and Proportional Representation Systems: Canada and Norway.”
Journal of Politics 58(3):707– 733.
McAllister, Ian. 1991. “Party Adaptation and Factionalism within the Australian Party System.”
American Journal of Political Science 35(1):206– 227.
McCarthy, Kate. 2001. “View from the Parties— e Labour Party.” In Stephen Coleman (ed.),
2001:CyberSpace Odyssey:e Internet in the UK Election. London:Hansard Society:22–
25. Available at hps:// assets.contentful.com/ u1rlvvbs33ri/ 5khyKS6f YsAEgc8OU4IuGI/
43c9b2a796ee89313f0d95db4297cb49/ Publication_ _ 2001- Cyber- Space- Odyssey- e-
internet- in- the- UK- Election- 2001.pdf (accessed October 6, 2017).
McGregor, Matthew. 2010. “Click Here to Unsubscribe:The Use of Email in the 2010
Election.” In Rachel K. Gibson, Andy Williamson, and Stephen Ward. (eds.), The
Internet and the 2010 Election:Putting the Small “p” Back in Politics? London:Hansard
Society:33– 42.
McKeown, Carole Anne, and Kenneth D. Plowman. 1999. “Reaching Publics on the Web during
the 1996 Presidential Campaign.” Journal of Public Relations Research 11(4):321– 347.
Milkis, Sidney M., and Jesse H. Rhodes. 2009. “Barack Obama, the Democratic Party, and the
Future of the ‘New American Party System’ ”. e Forum 7(1). doi:hps:// doi- org.manches-
ter.idm.oclc.org/ 10.2202/ 1540- 8884.1297 (accessed June 29, 2018).
Mills, Stephen. 2014. e Professionals:Strategy, Money and the Rise of the Political Campaigner in
Australia. Collingwood, Australia:Black.
Miskin, Sarah. 2008. Politics and Technology in the 21st Century. Canberra:Australian Parliamentary
Li brary.
Montero, María. D. 2009. “Political e- Mobilisation and Participation in the Election Campaigns of
Segolene Royal (2007) and Barack Obama (2008).” Quaderns del CAC 33:26– 34. Available
at hps:// pdfs.semanticscholar.org/ 183f/ b6f1114c04a23a9dae663a1ec6e12663e5ba.pdf
(accessed June 27, 2016).
Murchison, Ashley. 2015. “Changing Communications? Political Parties and Web 2.0 in the 2011
New Zealand General Election.” In Guy Lachappell and Philippe. J. Maarek (eds.), Political
Parties in the Digital Age:e Impact of New Technologies in Politics. Berlin: De Gruyter
Oldenbourg:93– 110.
Nastasia, Sorin. 2014. “Europe.” In Kerric Harvey (ed.), Encyclopedia of Social Media and Politics,
Vol. 1. London:Sage:478– 481.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
References 297
297
Negrine, Ralph. 2015. “Americanization.” In Gianpietro Mazzoleni, Kevin G. Barnhurst, Ken’ichi
Ikeda, Hartmut Wessler, and Rousiley C.M. Maia (eds.), e International Encyclopedia of
Political Communication, Vol. 1. Hoboken, NJ:John Wiley & Sons:1– 5.
Negrine, Ralph, and Stylianos Papathanassopoulos. 1996. “e ‘Americanization’ of Political
Communication:A Critique.” Harvard International Journal of Press/ Politics 1(2):45– 62.
Negroponte, Nicholas. 1995. Being Digital. London:Coronet.
Newell, James L., 2001. Italian Political Parties on the Web. e Harvard International Journal of
Press/ Politics 6(4):60– 87.
Newman, N., 2010. # UKelection2010, Mainstream Media and the Role of the Internet:How Social
And Digital Media Aected the Business of Politics and Journalism. University of Oxford, Reuters
Institute for the Study of Journalism. Available at hps:// reutersinstitute.politics.ox.ac.uk/
sites/ default/ les/ research/ les/ Social%2520Media%2520and%2520the%2520Election.
pdf (accessed July 1, 2016).
Nickerson, David W., and Todd Rogers. 2014. “Political Campaigns and Big Data.” e Journal of
Economic Perspectives 28(2):51– 73.
Noble, Phil. 2002. “From Novelty to Necessity:No Breakthroughs but Broader, Faster, Quicker,
Smarter . . . and More to Come.” Politics Online. Available at hp:// www.politicsonline.com/
pol2000/ specialreports/ campaign- analysis 2002/ . (accessed November 24, 2003).
Norris, Pippa. 2001. Digital Divide: Civic Engagement, Information Poverty, and the Internet
Worldw i de. NewYork:Cambridge University Press.
Norris, Pippa. 2000. A Virtuous Circle: Political Communications in Postindustrial Societies.
NewYork:Cambridge University Press.
O’Reilly, Tim, 2005. What Is Web 2.0:Design Paerns and Business Models for the Next Generation
of Soware. Sebastopol, CA:O’Reilly Media. Available at hp:// www.oreilly.com/ pub/ a/
web2/ archive/ what- is- web- 20.html (accessed June 21, 2017).
Painter, Anthony. 2010. “A Lile More Conversation, ALile More Action:Orange’s Digital
Election Analysis.” Available at hps:// www.scribd.com/ doc/ 32756567/ Orange- s- Digital-
Election- Analysis (accessed July 15, 2018).
Pène, Clémence. 2012. “Digital Campaigning within the French Socialist Party: Innovation,
Importation Circulation.” Paper presented at the European Consortium for Political Research
(ECPR) Joint Sessions, Antwerp, Belgium.
Plasser, Fritz. 2000. “American Campaign Techniques Worldwide.” Harvard International Journal
of Press/ Politics 5(4):33– 54.
Plasser, Fritz, and Gunda Plasser. 2002. Global Political Campaigning: A Worldwide Analysis of
Campaign Professionals and eir Practices. Westport, CT:Greenwood.
Pons, Vincent. 2018. “Will a Five- Minute Discussion Change Your Mind? A Countrywide
Experiment on Voter Choice in France.” American Economic Review 108(6):1322– 1363.
Poster, Mark. 1995. “e Internet as a Public Sphere?” Wired 3(1):209. Avai lable at hps:// www.
wired.com/ 1995/ 11/ poster- if/ (accessed June 30, 2018).
Qiu, Jack Linchuan. 2008. “Mobile Civil Society in Asia:A Comparative Study of People Power II
and the Nosamo Movement.” Javnost- e Public 15(3):39– 58.
Quarociocchi, Walter, Antonio Scala, and Cass R. Sunstein. 2016. “Echo Chambers on
Facebook” (June 13). Available at SSRN:hps:// ssrn.com/ abstract=2795110 (accessed
November 12, 2019).
Qiu, J.L. 2008. Mobile civil society in Asia:A comparative study of People Power 2 and the
Nosamo movement. Javnost:e Public 15(5):5– 24.
Quinn, Adrian. 2000. “Tout est lié: e Front National and media conspiracy theories.” e
Sociological Review 48(2):112– 132.
Rash, Wayne. 1997. Politics on the Nets:Wiring the Political Process. NewYork:W. H.Freeman.
Reavy, Mahew M., and David D. Perlmuer. 1997. “Presidential Web Sites as Sources of
Information.” Electronic Journal of Communication 7(3). e Impact of the Internet on
Journalism, II [Online Journal]. Available at hp:// www.cios.org/ EJCPUBLIC/ 007/ 3/
007311.HTML (accessed June 29, 2018).
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
298 References
298
Reif, Karl Heinz, and Hermann Schmi. 1980. “Nine Second‐Order National Elections: A
Conceptual Framework for the Analysis of European Election Results.” European Journal of
Political Research 8(1):3– 44.
Rheingold, Howard. 2000. e Virtual Community:Homesteading on the Electronic Frontier (rev.
ed.). Cambridge, MA:MIT Press.
Ridge- Newman, Anthony. 2014. Cameron’s Conservatives and the Internet: Change, Culture and
Cyber Toryism. London:Palgrave Pivot.
Robinson, E. 2003. “Broadening Participation: inking Beyond Party Membership NPN
Recommendations.” In Broadening Participation: inking Beyond Party Membership New
Politics Network Policy Paper. London:New Politics Network: 9– 14. Available at hps://
static1.squarespace.com/ static/ 5525bbf5e4b026f75314c09b/ t/ 55e6dbb1e4b0c-
94c8af905cd/ 1441192881029/ broadeningparticipation.pdf (accessed June 30, 2018).
Robles- Estrada, Celestino, A. F. Orozco- Torres, and A. B. Solano- Navarro. 2008. “e Internet
and the Political Marketing:An Analysis of How the Political Parties and Candidates Used
the Internet during the 2006 Mexico Presidential Campaign.” IOS PRESS. Available at
hp:// hdl.handle.net/ 123456789/ 36181 (accessed March 21, 2009).
Roper, Juliet. 1998. “New Zealand Political Parties Online:e World Wide Web as a Tool for
Democratization or Political Marketing.” In Chris Toulouse and Tim Luke (eds.), e Politics
of Cyberspace. Abingdon, Oxon, UK:Routledge:49– 69.
Rose, David, and David J. Pevalin. 2001. “e National Statistics Socio- economic
Classication:Unifying Ocial and Sociological Approaches to the Conceptualisation and
Measurement of Social Class.” IISER Working Papers. Paper 2001– 4 Colchester:University
of Essex. Available at hps:// www.iser.essex.ac.uk/ les/ iser_ working_ papers/ 2001- 04.pdf
(accessed July 21, 2018).
Rosenstone, Steven J., and John Hansen. 1993. Mobilization, Participation, and Democracy in
America. NewYork:Macmillan.
Runswick, Alexandra, Peter E. Facey, Emily Robinson, David Denver, and Justin Fisher. 2004. Life
Support for Local Parties:An Analysis of the Decline of Local Political Parties and the Case for
State Support of Local Activism. London:New Politics Network. Available at hps:// static1.
squarespace.com/ static/ 5525bbf5e4b026f75314c09b/ t/ 55e6d821e4b048d3ac2e3f1c/
1441191969656/ Life+Support.pdf (accessed October 7, 2017).
Sadow, Jerey D., and Karen James. 1999. “Virtual Billboards? Candidate Web Sites and
Campaigning in 1998.” Paper presented at the Annual Meeting of the American Political
Science Association, Atlanta, GA.
Salamanos, Nikos, Michael J. Jensen, Xinlei He, Yang Chen, and Michael Sirivianos. 2019.
“On the Inuence of Twier Trolls during the 2016 US Presidential Election.” arXiv:
1910.00531v2
Samuel- Azran, Tal, Moran Yarchi, and Gadi Wolfsfeld. 2015. “Equalization versus
Normalization:Facebook and the 2013 Israeli Elections.” Social Media+ Society 15 1(2).
doi:10.1177/ 2056305115605861.
Sanders, David, Harold D. Clarke, Marianne C. Stewart, and Paul Whiteley. 2007. “Does Mode
Maer for Modeling Political Choice? Evidence from the 2005 British Election Study.”
Political Analysis 15(3):257– 285.
Sarkozy, Nicolas. 2011. “e State and the Internet:A Call for Collective Responsibility.” New
Perspectives Quarterly 28(3):7– 11.
Scarrow, S., 2014. Beyond Party Members: Changing Approaches to Partisan Mobilization.
Oxford:Oxford University Press.
Schneider, Steven M., and Kirsten A. Foot. 2006. “Web Campaigning by US Presidential Primary
Candidates in 2000 and 2004.” In Andrew Paul Williams and John C. Tedesco (eds.), e
Internet Election:Perspectives on the Web in Campaign 2004. Lanham, MD: Rowman and
Lileeld:21– 36.
Schwartz, Ed, and Andy Oram. 1996. Netactivism:How Citizens Use the Internet. Sebastopol,
CA:Songline Studios.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
References 299
299
Schweitzer, Eva J. 2005. “Election Campaigning Online:German Party Websites in the 2002
National Elections.” European Journal of Communication 20(3):327– 351.
Schweitzer, Eva J. 2008. “Innovation or Normalization in E- Campaigning? A Longitudinal
Content and Structural Analysis of German Party Websites in the 2002 and 2005 National
Elections.” European Journal of Communication 23(4):449– 470.
Schweitzer, Eva J. 2011. “Normalization 2.0:A Longitudinal Analysis of German Online Campaigns
in the National Elections 2002– 9.” European Journal of Communication 26(4):310– 327.
Selnow, Gary W. 1998. Electronic Whistle- Stops: e Impact of the Internet on American Politics.
Westport, CT:Praeger.
Serfaty, Viviane. 2002. L’Internet en politique: Des Etats- Unis à l’Europe. Strasbourg: Presses
Universitaires de Strasbourg.
Sides, John S., and Lynn Vavreck. 2013. e Gamble:Choice and Chance in the 2012 Presidential
Election. Princeton, NJ:Princeton University Press.
Small, Tamara A. 2008. “Equal Access, Unequal Success:Major and Minor Canadian Parties on
the Net.” Party Politics 14(1):51– 70.
Southern, Rosalynd, and Benjamin Lee. 2015. “An Assessment of Candidate Online Campaigns
at the 2015 UK General Election.” Paper presented at the Annual Meeting of the American
Political Science Association, San Francisco.
Southern, Rosalynd, and Stephen Ward. 2011. “Below e Radar? Online Campaigning at the
Local Level in the 2010 Election.” In Dominic Wring, Roger Mortimore, and S. Atkinson
(eds.), Political Communication in Britain:e Leader Debates, e Campaign and the Media in
the 2010 General Election. Basingstoke, UK:Palgrave:218– 240.
Stanyer, James. 2005. “Political Parties, the Internet and the 2005 General Election:From Web
Presence to e- Campaigning?” Journal of Marketing Management 21(9– 10):1049– 1065.
Stegmueller, Daniel. 2013. “How Many Countries for Multilevel Modeling? A Comparison of
Frequentist and Bayesian Approaches.” American Journal of Political Science 57(3):748– 761.
Stockemer, Daniel. 2017. e Front National in France:Continuity and Change under Jean- Marie Le
Pen and Marine Le Pen. NewYork:Springer.
Stone, Ben. 1996. “Politics ’96.” Internet World (November):44– 50.
Strandberg, Kim. 2009. “Online Campaigning:An Opening for the Outsiders? An Analysis of
Finnish Parliamentary Candidates’ Websites in the 2003 Election Campaign.” New Media &
Society 11(5):835– 854.
Straw, Will. 2010. “Yes We Did? What Labour Learned from Obama.” In Rachel K . Gibson, Andy
Williamson, and Stephen Ward (eds.), e Internet and the 2010 Election” Puing the Small “p”
Back in Politics? Hansard Society 43– 50.
Street, John. 1997. “Remote Control? Politics, Technology and ‘Electronic Democracy.” European
Journal of Communication 12(1):27– 42.
Stromer- Galley, Jennifer. 2004. “Interactivity- as- Product and Interactivity- as- Process.” e
Information Society 20(5):391– 394.
Stromer- Galley, Jennifer. 2014. Presidential Campaigning in the Internet Age. NewYork: Oxford
University Press.
Sunstein, Cass R. 2001. Republic.com. Princeton, NJ:Princeton University Press.
Swaddle, Kevin. 1988. “Hi‐Tech Elections:Technology and the Development of Electioneering
since 1945.” Contemporary Record 2(1):32– 35.
Swanson, David L., and Paolo Mancini (eds.). 1996 Politics, Media, and Modern Democracy:An
International Study of Innovations in Electoral Campaigning and eir Consequences. Westport,
CT:Greenwood.
Teachout, Zephyr, and omas Streeter. 2008. Mousepads, Shoe Leather, and Hope. Boulder,
CO:Paradigm.
Tkach- Kawasaki, Leslie M. 2003. “Politics@ Japan:Party Competition on the Internet in Japan.”
Party Politics 9(1):105– 123.
Towner, Terri L., and David A. Dulio. 2011. “e Web 2.0 Election:Does the Online Medium
Maer?” Journal of Political Marketing 10(1– 2):165– 188.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
300 References
300
Trechsel, Alexander H., Raphael Kies, Fernando Mendez, and Phillip C. Schmier. 2003.
“Evaluation of the Use of New Technologies in Order to Facilitate Democracy in Europe.”
European Parliament Working paper, Scientic and Technological Options Assessment
Series STOA 116 EN. Available at hp:// www.europarl.europa.eu/ thinktank/ en/ docu-
ment.html?reference=IPOL- JOIN_ ET(2003)471583 (accessed June 15, 2018).
Trippi, Joe. 2004. e Revolution Will Not Be Televised:Democracy, e Internet, and the Overthrow
of Everything. NewYork:Regan Books.
Turk, Michael. 2012. “Social and New Media:An Evolving Future.” In Richard J. Semiatin (ed.),
Campaigns on the Cuing Edge, 2nd ed. Washington, DC:Congressional Quarterly Press.
Vaccari, Cristian. 2013. Digital Politics in Western Democracies:A Comparative Study. Baltimore,
MD:John Hopkins University Press.
Vaccari, Cristian. 2008b. “From the Air to the Ground:e Internet in the 2004 US Presidential
Campaign.” New Media & Society 10(4):647– 665.
Vaccari, Cristian. 2008a. “Surng to the Elysee:e Internet in the 2007 French Elections.” French
Politics 6(1):1– 22.
van Onselen, Peter, and Wayne Errington. 2004a. “Electoral Databases:Big Brother or Democracy
Unbound?” Australian Journal of Political Science 39(2):349– 366.
van Onselen, A., and P. Van Onselen. 2008. “On Message or Out of Touch? Secure Web Sites and
Political Campaigning in Australia.” Australasian Political Studies Association 43(1):43– 58.
van Onselen, Peter, and Wayne Errington. 2004b. “Privacy Legislation, Political Party Exemption
and Constituent Databases.” Privacy Law Bulletin 1(3):43– 47.
Vedel, ierry. 2005. “Television across Europe:Regulation, Policy and Independence:France.”
In Television across Europe:Regulation, Policy and Independence. Monitoring Report, Open
Society Institute:637– 728. Available from Sciences Po HAL archive:hps:// hal- sciencespo.
archives- ouvertes.fr/ hal- 01053946/ document
Vedel, ierry, and Karolina Koc- Michalska. 2009. “e Internet and French Political
Communication in the Aermath of the 2007 Presidential Election.” Paper presented at the
5th European Consortium for Political Research (ECPR) General Conference, Potsdam.
Vedel, ierry, and Karolina Koc- Michalska. 2007. “Political Participation and the
Internet:Evidence from the 2007 French Presidential Election.” Paper presented at the 3rd
International Conference on e- Government (ICEG). Montreal, Canada.
Vergeer, Maurice, Liesbeth Hermans, and Carlos Cunha. 2012. “Web Campaigning in the 2009
European Parliament Elections: A Cross- National Comparative Analysis.” New Media &
Society 15(1):28– 148.
Vergeer, Maurice, Liesbeith Hermans, and Steven Sams. 2013. “Online Social Networks and
Micro- Blogging in Political Campaigning:e Exploration of a New Campaign Tool and a
New Campaign Style.” Party Politics 19(3):477– 501.
Villalba, Bruno 1999. “Internet and the European Elections Campaign of 1999 in France.” Paper
presented for the meeting of the COST A14 Working Group, Government and Democracy
in the Information Age (GADIA), Brussels, March 18– 19.
Villalba, Bruno. 2003. “Moving Towards an Evolution in Political Mediation? French Political
Parties and the New ICTs.” In Rachel K. Gibson, Paul Nixon, and Stephen Ward (eds.),
Political Parties and the Internet. London:Routledge:120– 138.
Vromen, Ariadne, and William Coleman. 2013. “Online Campaigning Organizations and
Storytelling Strategies:Getup! in Australia.” Policy & Internet 5(1):76– 100.
Vromen, Ariadne, and William Coleman. 2011. “Online Movement Mobilisation and Electoral
Politics:e Case Of Getup!” Communication, Politics & Culture 44(2):76– 94
Ward, Stephen, and Stephen Coleman (eds.). 2005. Spinning the Web:Online Campaigning in the
2005 General Election. London:Hansard Society.
Ward, Stephen, and Rachel K. Gibson. 1998. “ e First Internet Election? UK Political Parties and
Campaigning in Cyberspace.” In Ivor Crewe, Brian Gosschalk, and John Bartle (eds.), Political
Communications:e General Election Campaign of 1997. London:Frank Cass:93– 112.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
References 301
301
Ward, Stephen, and Rachel K. Gibson. 2003. “On- line and on Message? Candidate Websites
in the 2001 General Election.” e British Journal of Politics and International Relations
5(2):188– 205.
Ward, Stephen, Diana Owen, Richard Davis, and David Taras (eds.). 2008a. Making a Dierence:A
Comparative View of the Role of the Internet in Election Politics. Plymouth, UK:Lexington Books.
Ward, Stephen, Rachel Gibson, and Wainer Lusoli. 2008b. “e United Kingdom:Parties and
the 2005 Virtual Election Campaign— Not Quite Normal?” In Stephen Ward, Diana Owen,
Richard Davis, and David Taras (eds.), Making a Dierence:A Comparative View of the Role of
the Internet in Election Politics. Plymouth, UK:Lexington Books:133– 160.
Ward, Stephen, and Gerrit Voerman. 1999. “New Media and New Politics: Green Parties,
Intra- party Democracy and the Potential of the Internet (an Anglo- Dutch Comparison).”
Documentatiecentrum Nederlandse Politieke Partijen Jaarboek:192– 215.
West, Darrell. 2008. Improving Technology Utilization in Electronic Government around the World.
Washington, DC:Brookings Institution, Governance Studies.
Whine, Michael. 2000. “Far Right Extremists on the Internet.” In Douglas omas and Brian D.
Loader (eds.), Cybercrime:Law Enforcement, Security and Surveillance in the Information Age.
London:Routledge:234– 250.
Whiteley, Paul, and Patrick Seyd. 2002. High- Intensity Participation:e Dynamics of Party Activism
in Britain. Ann Arbor:University of Michigan Press.
Williams, Andrew Paul, and John C. Tedesco. 2006. e Internet Election:Perspectives on the Web in
Campaign 2004. Oxford:Rowman & Lileeld.
Williamson, Andy, Laura Miller, and Freddy Fallon. 2010. Behind the Digital Campaign: An
Exploration of the Use, Impact and Regulation of Digital Campaigning. London: Hansard
Society. Available at hps:// assets.ctfassets.net/ u1rlvvbs33ri/ 4toHB02uCsKm82Y-
eiumU/ 7e05cf1ee0534da6688b4b8712ca2470/ Publication_ _ Behind- The- Digital-
Campaign- 2010.pdf (accessed May 4, 2020).
Wolf, Gary. 2004. “How the Internet Invented Howard Dean.” Wired 12(1). Available at hp://
www.wired.com.manchester.idm.oclc.org/ wired/ archive/ 12.01/ dean.html (accessed
February 4, 2016).
Wu, Wei, and David Weaver. 1997. “Online Democracy or Online Demagoguery:Public Opinion
‘Polls’ on the Internet.” Harvard International Journal of Press/ Politics 2:71– 86.
Young, Sally. 2010. How Australia Decides:Election Reporting and the Media. Melbourne:Cambridge
University Press.
Ziel, omas. 2015. “Do Candidates Seek Personal Votes on the Internet? Constituency
Candidates in the 2009 German Federal Elections.” German Politics 24(4):435– 450.
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
302
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
303
303
Index
Tables and gures are indicated by t and f following the page number
For the benet of digital users, indexed terms that span two pages (e.g., 52– 53) may, on occasion, appear on only one
of those pages.
Abramson, Jerey, 10
activist motivation. See community building and
activist mobilization
age, data sources
Australia, 248
France, 255
United Kingdom, 242
United States, 263– 64
age, phase IV digital campaigning, 62, 63t
Americanization, 168– 69
American National Election Study (ANES, 2012),
21, 194– 95, 196t, 198– 99, 200t, 262
Analyst Institute, 186
analytics, 18t, See also specic campaigns
back- end, 40
modeling labs, 128– 29
Anstead, Nick, 51– 52
Anthony, Joe, 182– 83
apoliticos, 2, 43, 211, 217– 18
Ashdown Paddy, 29
astroturng, digital, 221
Australia, 107– 33, 212– 13
broadband subscriptions, growth, 108, 109f
candidate- centered election campaigning, 4
citizen- initiated campaigning, 118, 119– 22, 120t
community building and activist motivation
(2010), 117– 22
databases, Electrac and Feedback!, 109– 10
data source, 20– 21
Democracy4Sale, 115
e- democracy, 110
experimentation (1994- 1997), 111– 13,
112f, 113f
Greens, citizen- initiated campaigning, 119– 22
Greens, experimentation, 111, 113
Greens, standardization and professionalization,
114– 15, 116– 17
hypernormality, 132– 33
internet access and use, growth, 107– 8, 108f
Kevin07, 116– 17, 118, 132
Kevin10, 117
Labor Connect, 118
mobilization, 49t, 109
multiparty environments, 4
MyLiberal.com, 118
vs. other countries, 214, 215– 16, 217
phase IV, moves to direct voter mobilization
(2013), 122– 25
Political Big Brother, 114– 15
Senator Online, 37, 116
standardization and professionalization (1998-
2007), 113– 17, 120t
tyranny of distance, 110– 11
web campaign cycles, mapping electorate,
125– 32, 126t, 127t, 130t (see also web
campaign cycles, mapping electorate,
Australia)
web campaign cycles, mapping electorate,
Greens, 127– 32, 127t, 130t
Australia, data sources and variable denitions, 245
2010, 245– 47
2013, 247– 48
age, 248
education, 249
party ID, 248
sex, 248
social class, 249
socio- demographic and political
variables, 248– 49
vote, 249
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
304 Index
304
Australian Election Studies (AES), 21, 125, 126t,
128, 245, 247
Australian Labor Party (ALP), 111– 12
home page (1996), 112f, 112
Australian Liberal Party, home page (1996),
113f, 113
Axelrod, David, 91
back- end analy tics, 40
Back the Ban, 82
Baldwin- Philippi, Jessica, 220
Banducci, Susan, 54
Beazley, Kim, 113– 14
Benne, Colin, 171
Benne, Lance, 222
big data, 14. See also specic topics
billboard, online (virtual), 32– 33, 175– 76
Bimber, Bruce, 11, 32, 41– 42, 57– 58
Blue State Digital (BSD), 35– 36, 149
France, 149
Go, Teddy, 187, 188, 189– 90, 206– 7, 209
Rospars, Joe, 35– 37, 149, 178, 180, 184– 85,
188, 208
United States, 36– 37, 178, 180, 181, 182– 83,
184– 85, 188, 208
Braen, L. Clare, 140
Brexit, 221
British Election Survey (BES, 2015), 20, 21
broadband subscriptions, growth
Australia, 108, 109f
France, 136f, 136
United Kingdom, 72– 74, 73f
United States, 169– 71, 170f
Brown, Jerry, Senate race (1992), 25– 26, 172
Bruns, Axel, 124
Bush, George W., digital team and Personal
Precinct, 36
Cambridge Analytica, 219– 20, 221
“Campaigning in Cyberspace” Economic and Social
Research Council (ESR), 21
candidate- centered election campaigning, 4
case studies, 20. See also specic countries
approach, 70– 71
importance and value, 5
limitations, 23
Castells, Manuel, 185
Catalist, 186
CEVIPOF Sciences Po, 21
Chadwick, Andrew, 51– 52, 222
Chen, Peter, 115, 116– 17, 132
citizen- driven campaigning, United States
(2004), 177– 80
citizen- initiated campaigning (CIC), 7– 8, 37
Australia, 118, 119– 22, 120t
index construction, 233
index construction, community building, 233– 34
index construction, message promotion, 235– 36
index construction, resource generation, 234
index construction, voter mobilization, 234– 35
United Kingdom, 85– 88, 86t
United States (2008), 180– 85, 181f
Clegg, Nick, 84
Cli, Steven, 110
Clinton, Bill
1996 campaign, 172, 173
2012 campaign, 1, 172
Clinton, Hillary
2008 campaign, 183– 84
2016 campaign, 219
cloud computing, 14
coding schemes, new (late 1990s), 32– 33
Commission Nationale Informatique et Libertés
(CNIL), 138– 39
commonwealth, electronic, 11
communication, campaign, 38
controlled interactivity, 38
message promotion, citizen- initiated
campaigning, 235– 36
top- down, 34
two- way capabilities, 38
communication, electoral, 8
regulatory controls, France, 138– 39, 171– 72
regulatory controls, United States, 66
United States, 35– 36
communication, voter
elite- voter, 18t, 41
external, internet adoption, 28– 29, 29f
extra- organizational, equalization and
normalization, 12, 13
communitarians, 9
community building, citizen- initiated
campaigning, 233– 34
community building and activist mobilization, 7– 8
Australia (2010), 117– 22
France (2012), 144t, 148– 55
literature review (1997- 2003), 34– 38
summary, 7, 16– 24, 18t
United Kingdom (2010), 81– 90, 83f, 84f, 86t
United States (2004- 2008), 177– 85
United States (2004- 2008), citizen- driven
campaigning (2004), 177– 80
United States (2004- 2008), citizen- initiated
campaigning (2008), 180– 85, 181f
Comparative National Election Project (CNEP)
study, 54
Comparative Study of Electoral Systems (CSES)
Module 2, 54
Module 4, 20
competition, inter- party, 13
competitiveness, contest, voter mobilization, 57
compulsory voting, voter mobilization, 56
Conservative Party, UK, landing page (2015),
92, 93f
Conservatives.com, 82
Contact Creator, 91
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Index 305
305
contact measures, campaign, 225
contagion theory, 53
controlled interactivity, 38, 93– 94, 211
co- production, 7– 8
Corrado, Anthony, 10, 11
country. See also specic countries
comparisons, overall, 214– 17
rankings, 214
culture, political
phase IV digital campaigning, 62– 65, 63t
voter mobilization, 55– 56
Darras, Eric, 147– 48
Dashboard, 188– 89, 191– 93, 192t, 199
vs. MyMi, 191– 93, 192t
data analysts, 2
data- driven digital campaigning, 2, 40, 41, 68. See
also specic countries, campaigns, and topics
data privacy and protection
France, 138– 39, 171– 72
United States, 171– 72
United States, micro- targeting, 57– 58
data sources, bibliography, 265– 67
data sources and variable denitions, Australia, 245
2010, 245– 47
2013, 247– 48
age, 248
education, 249
party ID, 248
sex, 248
social class, 249
socio- demographic and political variables, 248– 49
vote, 249
data sources and variable denitions, France,
251, 254– 57
2007, 251– 52
2012, 253– 54
age, 255
education, 255– 56
income, 256– 57
party ID, 254– 55
sex, 255
socio- demographic and political
variables, 254– 57
vote, 257
data sources and variable denitions, United
Kingdom, 237
2005, 237– 38
2010, 238– 40
2015, 240– 41
age, 242
education, 242– 43
party ID, 241– 42
sex, 242
socio- demographic and political
variables, 241– 43
summary, 20– 21
vote, 243
data sources and variable denitions, United
States, 259– 65
2004, 259– 60
2008, 260– 62
2012, 262– 63
age, 263– 64
education, 264
income, 264– 65
party ID, 263
race, 265
sex, 263
socio- demographic and political
variables, 263– 65
vote, 265
Davis, Richard, 52, 172
Dean, Howard (2004), 35– 36, 168, 174– 75, 177– 80,
184– 85, 186, 191– 93, 194, 208
Democracy4Sale, 115
Democratic National Commiee (DNC), data-
driven campaigning (2012), 41
Democratic Party. See also specic campaigns
and topics
Simulmatics project, 25
Democrat Internet Center (2000 election), 175
Denver, David, 74
dependent variable, 58, 227. See also specic variables
digital astroturng, 221
digital campaigning
concept, 5– 6
data- driven, 2
denition, 5– 6
digitally literate elite, apolitico, 2
equalization and normalization, 8– 9
four- phase model, 1– 2 (see also
four- phase model)
get out the vote (GOTV), 1– 2
growth, 1– 2
history, Bill Clinton 2012 breakthrough, 1, 172
on internal structure, 1– 2
literature review, 1
in practice, literature review, 41– 42
digital campaigning, across space, 44– 71
comparing campaigns, critical criteria, 44– 45
demand- side characteristics, 44– 45
digital mobilization, explanatory model, 55– 61
model testing, 58– 61
model testing, core data set, 58, 59t, 227
model testing, dependent variable, 58, 227
model testing, independent variables,
58– 61, 227– 31
next steps, 69– 71
phase IV, entrants, 45– 49, 46t, 49t
phase IV, factors driving, 62– 68, 63t, 67t
voter mobilization, comparative studies, 50– 54
voter mobilization, comparative studies, digital
campaigns, 51– 53
voter mobilization, comparative studies, voter
mobilization, 53– 54
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
306 Index
306
digitally literate elite, 2
digital mobilization, explanatory model, 55– 61.
See also explanatory model, digital mobilization
testing, 58– 61
digital team. See also specic campaigns, countries, and
individuals
broader power domains, 18t, 40– 41
as organizational apex, 40
Dole, Bob (1996), presidential campaign, 26– 29,
29f, 173– 74, 174f
Dreamcatcher, 189
drivers, digital campaign change, 214– 17
Druckman, James, 184
e- brochureware, 32– 33, 175– 76
Ecker, Clint, 188
e- democracy, 8– 9, 110
Eds Pledge, 82
education, data sources
Australia, 249
France, 255– 56
United Kingdom, 242– 43
United States, 264
education, phase IV digital campaigning, 62, 63t
e- government development index (EGDI), 110
electoral cycle, voter mobilization, 56
electoral systems, voter mobilization, 56
Electrac, 109– 10
electronic commonwealth, 11
electronic pamphlet, 32– 33
elite- voter communication, 18t, 41
engagement, voter
Comparative studies, European Union, 51– 52
extractive, 93– 94
U.S. presidential digital campaign (2004- 2012),
195– 99, 196t
voter mobilization, 55– 56
e- pluralism, 8– 9, 10– 11, 212
e- precinct leaders (e- leaders), 176– 77
equalization, 8– 9, 11, 212
intra- organizational, 13
systemic, 13
equalization, literature review
external voter communication, internet adoption,
28– 29, 29f
systemic and inter- party competition, 28, 30
equalization and normalization
fully digital campaign, 13
inter- party competition, 13
intra- organizational, 12
systemic, 12, 13
voter communication,
extra- organizational, 12, 13
equalization to normalization
campaign activity areas, 12
web 1.0, 12
European Social Survey (ESS), 60
European Union (EU). See also France; United
Kingdom
voter mobilization, comparative studies, 51– 52
Exley, Zak, 186
experimentation, 7– 8
Australia (1994- 1997), 111– 13, 112f, 113f
France (1994- 1997), 140– 41, 141f
literature review (1992- 1996), 25– 30, 26f,
27f, 29f
summary, 7, 16– 24, 18t
United Kingdom (1994- 1997), 74– 76
United States (1996- 2000), 172– 77, 174f
United States (1996- 2000), 2000 presidential
election, 175– 77
explanatory model, digital mobilization, 55– 61, 59t
baseline indicators and necessary conditions, 55
cross- national variation, 55
electoral cycle, type of contest, and compulsory
voting presence, 56
party system eects, 56– 57
political culture, 55– 56
regimes and electoral systems, 56
explanatory model testing, digital
mobilization, 58– 61
dependent variable, 58, 227
independent variables, 58– 61, 227– 31
extra- organizational voter communication. See voter
communication, extra- organizational
Facebook, 35, 219– 20
fake news, 220– 21
Farmer, Rick, 175
Feedback!, 109– 10
Feltesse, Vincent, 149
Fender, Rich, 175
Firestone, Charles, 10, 11
Fisher, Justin, 39– 40, 88– 89
Foot, Kirsten, 17, 184
Forbes, Steve, 176– 77
forum.bnp.org.uk, 84– 85
four- phase model, 1– 2, 7– 24, 210, 212– 13. See also
specic phases and topics
application, real- world, 18t, 212
contributions and limitations, 22– 24
data source, summary, 20– 21
“one size ts all” framework, limitations, 22– 23
ordering and sequencing, across countries, 23
overlap and seepage, 23– 24
overview, 7
phases, summary, 7, 16– 24, 18t
phases, two power logics, 11– 14
progress, by country, 18t, 212– 13
strategic developments, 15– 16, 18t
summary, 7, 16– 24, 18t
technological and organizational capacity, 14– 15
theoretical perspectives, internet and campaign
change, 8– 9
“third way”: e- pluralism and “new Jeersonians,”
10– 11, 212
two power logics, 11– 14
France, 134– 67, 212– 13
broadband subscriptions, growth, 136f, 136
candidate- centered election campaigning, 4
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Index 307
307
Commission Nationale Informatique et
Libertés, 138– 39
community building, digital activism, and
mobilization (2012), 144t, 148– 55
context, digital campaigning, 135– 40, 136f, 137f
data source, 20– 21
experimentation (1994- 1997), 140– 41, 141f
French Greens, 140– 41, 141f, 142, 146f, 149,
150, 155, 162
French Socialist Party, 143, 146f, 147, 149,
162, 166
Front National and Le Pen, 140– 41, 142– 47,
158– 60, 159t, 161, 163t
General Data Protection Regulation, 139
Hollande, Francois and Tous Hollande, 143, 144t,
149– 50, 151– 54, 163t, 164– 65, 198– 99
internet use, growth, 136f, 136– 37
La Coopérative Politique (La Coopol), 149,
150– 51, 152– 53
Le Comité de Soutien, 144t, 151– 52
Liegey Muller Pons, 149
Minitel, 136– 37, 215
multiparty environments, 4
vs. other countries, 214– 15, 217
presidential elections, regular, 4
regulatory controls, electoral communication,
138– 39, 171– 72
Royal, Segolène, 147– 48, 149, 160, 179
Sarkozy, Nicolas, La FranceFort and, 151, 152
Sarkozy, Nicolas, NS Connect and, 144t, 147–
48, 151– 52, 153– 54, 163t, 164
sociodemographics and political correlates,
receive mode, 162– 63, 163t
standardization (1998- 1999), 142– 43
standardization (1998- 2007), 142– 48
standardization (2002), 143– 47, 146f
standardization (2007), 147– 48
Union for a Popular Movement, 149, 150– 51,
159t, 160, 161– 62
web campaign cycles, mapping electorate reading,
redistributing, and receiving, 155– 65, 157t, 159t
web campaign cycles, mapping electorate
receive mode, sociodemographic and political
correlates, 162– 63, 163t
France, data sources and variable denitions, 251
2007, 251– 52
2012, 253– 54
age, 255
education, 255– 56
income, 256– 57
party ID, 254– 55
sex, 255
socio- demographic and political
variables, 254– 57
vote, 257
FranceFort, 151, 152
Front National (FN), 140– 41, 142– 47, 158– 60,
159t, 161, 163t
fundraising, 234
Australia, 120t
data- driven approach, 40
early 2000s, 32
France, 144t
United Kingdom, 82, 86t
United States, 178, 179– 80, 182, 184– 85, 187– 88,
189, 192t, 208
future direction, 218– 23
Gamble, e (Sides and Vavreck), 207
gender, 62, 63t
gender, data sources
Australia, 248
France, 255
United Kingdom, 242
United States, 263
General Data Protection Regulation (GDPR),
France, 139, 166– 67
Gerber, Alan, 186
German Social Democratic Party (SPD)
home page (1996), 26, 27f
home page (2001, 2002), 30, 31f
get out the vote (GOTV), 1– 2
Australia, 119, 120t
comparative studies, country, 53– 54
Gibson, Rachel, 28, 32– 33, 77– 79, 116,
121– 22, 128
Go4th, 81
Go, Teddy, 187, 188, 189– 90, 206– 7, 209
Gosnell, Harold, 53– 54
Greece, high digital contact, 70
Green, Don, 186
Greens, 33
France, 140– 41, 141f, 142, 146f, 149, 150,
155, 162
United States, vote- swapping, 177
Greens, Australia, 160
citizen- initiated campaigning, 119– 22
experimentation, 111, 113
standardization and professionalization,
114– 15, 116– 17
web campaign cycles, mapping electorate, 127– 32,
127t, 130t
Hands, Gordon, 74
Help America Vote Act (HAVA), 171
Higheld, Tim, 124
Hollande, Francois (Tous Hollande), 143, 144t,
149– 50, 151– 54, 163t, 164– 65, 198– 99
Houdini project, 187– 88
House of Representatives, U.S., web presence
(2002), 32
Howard, John, 116– 17
Huckabee, Mike (2008), 183
Hughes, Chris, 35– 36
hypernormality, 2– 3, 43, 44, 132– 33, 211, 217– 18
Iain Dale’s Blog Spot, 83
income, data sources
France, 256– 57
United States, 264– 65
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
308 Index
308
independent variables, 58– 61, 227– 31. See also
specic variables
index construction, citizen- initiated
campaigning, 233
community building, 233– 34
message promotion, 235– 36
resource generation, 234
voter mobilization, 234– 35
individual voter mobilization. See voter mobilization
infrastructure development, 40
Institute for Democracy and Electoral Assistance
(IDEA), Political Finance Database., 61
institutions. See also specic types
web campaigning role, voter mobilization
comparisons, 52
interactivity, controlled, 38, 93– 94, 211
interface, 6
internet access and use
Australia, 107– 8, 108f
France, 136f, 136– 37
United Kingdom, 72– 74, 73f
United States, 169– 71, 170f
“Internet and American Life Project” campaign
studies, Pew (2004 & 2008), 21
Internet Research Agency (I), 220– 21
inter- party competition, 13
intra- organizational domain, 12, 13, 18t, 40– 41
Issenberg, Sasha, 42
Jackson, Nigel, 139– 40, 141
Jeersonians, new, 10– 11, 212
Johnson, Dennis, 184– 85
Johnson, J. W., 60– 61
Kalnes, Øvind, 37
Kamarck, Elaine, 176
Karp, Jerey, 54
Kenne, Je, 114– 15, 116
Kevin07, 116– 17, 118, 132
Kevin10, 117
Koc- Michalska, Karolina, 154– 55
Kreiss, Daniel, 218
Kriegal, Elan, 219
Labor Connect, 118
Labour Party, UK
home page (1996), 26f, 26
home page (2001, 2002), 30, 31f
landing page (2015), 92f, 92
LabourSpace, 82
La Coopérative Politique (La Coopol), 149,
150– 51, 152– 53
La FranceFort, 151, 152
Lazarsfeld, Paul, 35– 36
Le Comité de Soutien, 144t, 151– 52
Lee, Benjamin, 93
Le Pen, Jean Marie, 140, 142– 45, 158– 60, 163t
Let’s Talk , 81
LibDemACT, 83– 84, 84f
Liegey Muller Pons (LMP), 149
Lilleker, Darren, 139– 40, 141
limitations, 22– 24
L’Institut Français d’Opinion Publique (IFOP), 21
literature review, 25– 43
community building and activist motivation
(1997- 2003), 34– 38
digital campaigning in practice, 41– 42
experimentation (1992- 1996), 25– 30, 26f,
27f, 29f
standardization and professionalization (1997-
2003), 30– 34, 31f
voter mobilization, individual (2012+), 39– 42
Mair, Peter, Ruling the Void, 218
Margolis, Michael, 33, 211
McAllister, Ian, 116, 121– 22, 128
McCain, John (2008), 183– 84
McCarthy, Kerry, 81
media access, voter mobilization, 57
Meetup, 35, 178, 179– 80
Membersnet, 82, 83f, 88, 101
MERLIN, 88– 89
message promotion, citizen- initiated
campaigning, 235– 36
Messina, Jim, 91
micro- targeting, 14, 23– 24, 39– 40,
216– 17, 222– 23
Australia, 123, 132– 33, 215– 16
France, 166– 67
United Kingdom, 213
United States, 199, 207
United States, advantage, 57– 58
United States, Obama campaign (2012), 42, 94
Minitel, 136– 37, 215
mobilization, individual voter. See voter
mobilization
model, four- phase. See four- phase model
Moo- hyun, Roh, 37
MoveOn.org, 186
Mybarackobama.com (MyBO), 35– 36, 82,
180– 83, 181f
MyConservatives.com, 83, 88– 89, 101
MyLiberal.com, 118
MyMi, 191– 93, 192t
vs. Dashboard, 191– 93, 192t
nano- targeting, 39– 40
narrowcasting, 39
Narwhal, 188, 193– 94
Nation Builder, 91
N comparative analysis, 20
“new Jeersonians,” 10– 11, 212
Newman, Nic, 89
New Organizing Institute, 186
NGP- VAN, 186
Nix, Alexander, 219– 20
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Index 309
309
Noble, Phil, 32
Noonan, Peggy, 40, 42
normalization, 8– 9, 11, 212. See also equalization
and normalization
intra- organizational, 13
systemic, 13
Norris, Pippa, 32, 51
Norway, 37
Nosamo, 37
NS Connect, 144t, 151– 52, 153
Obama, Barack (2008), 36– 37, 168, 208– 9
Blue State Digital and Mybarackobama.com, 35– 36,
82, 180– 83, 181f, 184– 85, 188
citizen- initiated campaigning, 180– 85, 181f
Dashboard, 188– 89, 191– 93, 192t, 199
Dashboard vs. MyMi, 191– 93, 192t
digital campaign cycles, mapping electorate, 199,
200t, 201– 5, 206– 7
Dreamcatcher, 189
historic victory, on digital campaigns, 185– 86
Orca, 189– 90
Rospars, Joe, 178, 181– 83, 187– 88
voter mobilization, individual, 185– 86, 187– 88 ,
191– 93, 192t
Obama, Barack (2012)
data- driven campaigning, 41– 42
micro- targeting, 42
online (virtual) billboard, 32– 33, 175– 76
Orca, 189– 90
organization, campaign
capacity, 14– 15 (see also technological and
organizational capacity)
centralization, 11
hybridity, 222
structure, internal, 1– 2
Owen, Diana, 172
pamphlet, electronic, 32– 33
Parscale, Brad, 219– 20
pa rt y. See also specic parties and campaigns
strength, voter mobilization, 56– 57
system, phase IV digital campaigning, 63t, 66
party ID, data sources
Australia, 248
France, 254– 55
United Kingdom, 241– 42
United States, 263
Paul, Ron (2008), 183
Pène, Clémence, 149
Personal Precinct, 36, 179– 80
Pew “Internet and American Life Project” campaign
studies (2004 & 2008), 21
phase V, 218– 23
Ploue, David, 185
pluralism
accelerated, 11
e- pluralism, 8– 9, 10– 11, 212
Political Big Brother, 114– 15
political culture
phase IV, digital campaigning, 62– 65, 63t
voter mobilization, 55– 56
political interest, phase IV digital campaigning,
62– 65, 63t
political trust, voter mobilization, 55– 56
Poster, Mark, 9
power
intra- organizational redistribution, 28
two logics, 11– 14
predictive modeling, 14
Presco, John, 81
privacy and protection, data, micro- targeting
and, 57– 58
professionalization. See standardization and
professionalization
race, U.S. data sources, 265
Rage Against the Machine, 84
rankings, country, 214
Rash, Wayne, 10, 11
redistribution, 16
Reed, Harper, 187– 88, 206
regimes, voter mobilization, 56
regulatory environment, 63t, 66
Resnick, David, 33, 211
resource generation, citizen- initiated
campaigning, 234
Rheingold, Howard, 9
Ridge- Newman, Anthony, 83, 88
Rospars, Joe, 178, 181– 83, 187– 88
Blue State Digital, 35– 37, 149, 178, 180, 184– 85,
188, 208
Royal, Segolène, 147– 48, 149, 160, 179
Rudd, Kevin, 115– 16, 213
Runi, Pat, 206
Ruling the Void (Mair), 218
Russian campaign interference (2016 U.S.
election), 220
Sabato, Larry, 178
Sarkozy, Nicolas
La FranceFort, 151, 152
NS Connect, 144t, 147– 48, 151– 52, 153– 54,
163t, 164
Schneider, Steven, 17, 184
Selnow, Gary, 26– 27, 172– 73
Senator Online, 116
Serfaty, Viviane, 145– 47
sex, 62, 63t
sex, data sources
Australia, 248
France, 255
United Kingdom, 242
United States, 263
shopfront, 6
Sides, John, 207
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
310 Index
310
sign- up, online, 62, 63t
Simulmatics project, 25
social class, Australia data sources, 249
Social Democratic Party (SPD), Germany
home page (1996), 26, 27f
home page (2001, 2002), 30, 31f
social networking platforms, 14
socio- demographic and political variables, Australia
data sources, 248– 49
age, 248
education, 249
party ID, 248
sex, 248
social class, 249
vote, 249
socio- demographic and political variables, France
data sources, 254– 57
age, 255
education, 255– 56
income, 256– 57
party ID, 254– 55
sex, 255
vote, 257
socio- demographic and political variables, United
Kingdom data sources, 241–43
age, 242
education, 242– 43
party ID, 241– 42
sex, 242
vote, 243
socio- demographic and political variables, United
States data sources, 241– 43
age, 263– 64
education, 264
income, 264– 65
party ID, 263
race, 265
sex, 263
vote, 265
Southern, Rosalynd, 89– 90, 93
South Korea, Nosamo, 37
space, digital campaigning across, 44– 71. See also
digital campaigning, across space
standardization and professionalization, 7– 8
Australia (1998- 2007), 113– 17
France (1998- 1999), 142– 43
France (1998- 2007), 142– 48
France (2003), 143– 47, 146f
France (2007), 147– 48
literature review (1997- 2003), 30– 34, 31f
summary, 7, 16– 24, 18t
United Kingdom (2001 & 2005), 76– 80, 78f
United States (1996- 2000), 172– 77, 174f
United States (1996- 2000), 2000 presidential
election, 175– 77
Storey, Ross, 111– 12
strategic developments, 15– 16, 18t, See also specic
campaigns, individuals, and topics
Stromer- Galley, Jennifer, 26– 27, 38, 93– 94, 145– 47,
172– 73, 174– 75
structure. See also organization, campaign
infrastructure development, 40
internal campaign, 1– 2
Sunstein, Cass, 9
systemic domain, 12, 18t, 40– 41
comparative studies, voter mobilization, 52– 53
equalization and normalization, 13
team, digital. See also specic campaigns, countries,
and individuals
broader power domains, 18t, 40– 41
as organizational apex, 40
technological and organizational capacity, 14– 15
theoretical perspectives, internet and campaign
change, 8– 9
“third way,” 10– 11, 212
Tous Hollande, 144t, 149– 50, 151– 53, 191
Trippi, Joe, 34, 35– 36, 178– 80, 182– 83, 184– 85
trust, political, voter mobilization, 55– 56
Turk, Michael, 176– 77
two- party systems, centrist, voter
mobilization, 56– 57
type of contest, voter mobilization, 56
tyranny of distance, Australia, 110– 11
UK Labour Party. See Labour Party, UK
undecided voters, as main target, 16
Union for a Popular Movement (UMP), 149, 150– 51,
159t, 160, 161– 62
United Kingdom, 20, 72– 106, 213
Back the Ban, 82
broadband subscriptions, growth, 72– 74, 73f
candidate- centered election campaigning, 4
case studies, 20
citizen- initiated campaign, 85– 88, 86t
community building and activist motivation
(2010), 81– 90, 83f, 84f, 86t
Conservatives.com, 82
context, digital campaigning, 72– 74, 73f
context, web campaigning, 72– 74, 73f
data source, 20, 21
digital mobilization, 49t, 74
Eds Pledge, 82
experimentation (1994- 1997), 74– 76
rst internet election (1997), 28
forum.bnp.org.uk, 84– 85
Go4th, 81
Iain Dale’s Blog Spot, 83
internet adoption and access, 28, 72– 74, 73f
LabourSpace, 82
Let’s Talk , 81
LibDemACT, 83– 84, 84f
Membersnet, 82, 83f, 88, 101
MERLIN, 88– 89
multiparty environments, 4
MyConservatives.com, 83, 88– 89, 101
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
Index 311
311
vs. other countries, 214– 15, 217
phase IV moves to direct voter mobilization
(2015), 90– 94, 92f, 93f
Rage Against the Machine, 84
standardization and professionalization (2001 &
2005), 76– 80, 78f
vs. United States, voter mobilization, 51– 52
Voter Activation Network (2015), 91
WebCameron, 81, 83
web campaign cycles, mapping electorate,
95– 105, 96t
web campaign cycles, mapping electorate,
reading, redistributing, and receiving web
campaign, over time, 95– 99, 96t, 237
web campaign cycles, mapping electorate,
reading, redistributing, and receiving web
campaign - by party, 99– 105, 100t, 103t
United Kingdom, data sources and variable
denitions
2005, 237– 38
2010, 238– 40
2015, 240– 41
age, 242
education, 242– 43
party ID, 241– 42
sex, 242
socio- demographic and political
variables, 241– 43
vote, 243
United States, 168– 209, 212– 13. See also
Obama, Barack
1996 explosion, 172– 73
Blue State Digital, 36– 37, 178, 180, 185, 208
broadband subscriptions, growth, 169– 71, 170f
Brown, Jerry, Senate race (1992), 25– 26
Bush, George W. (2008), 179– 80
candidate- centered election campaigning, 4
Clinton, Bill (1992), 1, 172
Clinton, Bill (1996), 172, 173
Clinton, Hillary (2008), 183– 84
Clinton, Hillary (2016), 219
community building and activist mobilization
(2004- 2008), 177– 85
community building and activist mobilization
(2004- 2008), citizen- driven campaigning
(2004), 177– 80
community building and activist mobilization
(2004- 2008), citizen- initiated campaigning
(2008), 180– 85, 181f
context, digital campaigning, 169– 72, 170f
data source, 20– 21
Dean, Howard (2004), 35– 36, 168, 174– 75,
177– 80, 184– 85, 186, 191– 93, 194, 208
Democrat Internet Center (2000), 175
Dole, Bob (1996) and dolekemp96.org, 26– 29,
29f, 173– 75, 174f
experimentation, standardization, and
professionalization (1996- 2000), 172– 77, 174f
experimentation, standardization, and
professionalization (1996- 2000), 2000
presidential election, 175– 77
Forbes and e- precinct leaders, 176– 77
freedom of expression, 171
Greens and vote- swapping, 177
Help America Vote Act, 171
Houdini project, 187– 88, 190
Huckabee, Mike (2008), 183
internet access and use, 169– 71, 170f
Kerry, John (2008), 179– 80
McCain, John (2008), 183– 84
MoveOn.org, 186
New Organizing Institute, 186
Obama, Barack (see Obama, Barack)
vs. other countries, 214, 215, 217
party supporters, presidential digital campaign
(2007- 2012), 199– 201, 200t
Paul, Ron (2008), 183
Personal Precinct, 179– 80
phase IV, individual voter mobilization (2012),
185– 94, 192t
Presidential elections, candidate websites
(1996), 26– 27
presidential elections, regular, 4
reading, redistributing, and receiving web
campaign, 194– 207, 196t, 200t, 203t
Senate, House and gubernatorial campaign
websites (1996), 28
Senate races (2004), 184
sociodemographic correlates for receive mode,
over time, 202– 6, 203t
as trendseer, 168– 69
Trippi, Joe, 34, 35– 36, 178– 80,
182– 83, 184– 85
vs. United Kingdom, voter mobilization, 51– 52
Ventura, Jesse, 176
voter engagement, presidential digital campaign
(2004- 2012), 195– 99, 196t
web campaign cycles, mapping electorate,
194– 207
web campaign cycles, mapping electorate,
reading, redistributing, and receiving web
campaign, 194– 207, 196t, 200t, 203t
United States, data sources and variable
denitions, 259– 65
2004, 259– 60
2008, 260– 62
2012, 262– 63
age, 263– 64
education, 264
income, 264– 65
party ID, 263
race, 265
sex, 263
socio- demographic and political
variables, 263– 65
vote, 265
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
312 Index
312
Vaccari, Cristian, 41, 52– 53, 55– 57, 62– 66, 148, 178
Vavreck, Lynn, 207
Ventura, Jesse, 176
Villalba, Bruno, 140– 47
virtual billboard, 32– 33, 175– 76
vote, data sources
Australia, 249
France, 257
United Kingdom, 243
United States, 265
Voter Activation Network (VAN), UK (2015), 91
voter communication, extra- organizational, 12
equalization and normalization, 12, 13
voter engagement
extractive, 93– 94
mobilization, 55– 56
U.S. presidential digital campaign (2004- 2012),
195– 99, 196t
voter mobilization, Australia, 49t, 109, 122– 25
voter mobilization, data and studies, 7, 8
centrist two- party systems, 56– 57
citizen- initiated campaigning, index
construction, 234– 35
comparative studies, 50– 54
comparative studies, digital campaigns, 51– 53
comparative studies, systemic domain, 52– 53
comparative studies, voter mobilization, 53– 54
competitiveness, contest, 57
compulsory voting, 56
electoral cycle, 56
electoral systems, 56
entrants, 45– 49, 46t, 49t
entry into, full, 211
factors driving, 62– 68, 63t, 67t
index construction, 234– 35
institutions, web campaigning, 52
levels, 50– 54
literature review (2012+), 39– 42
media access, 57
party strength, 56– 57
political culture, 55– 56
political trust, 55– 56
regimes, 56
summary, 7, 16– 24, 18t
type of contest, 56
United Kingdom (2015), 90– 94, 92f, 93f
voter engagement, 55– 56
voter mobilization, United Kingdom, 90– 94, 92f,
93f, 95– 105
2015, 90– 94, 92f, 93f
reading, redistributing, and receiving web
campaign, by party, 99– 105, 100t
reading, redistributing, and receiving web
campaign, over time, 95– 99, 96t, 237
vs. United States, 51– 52
voter mobilization, United States, 185– 94, 192t
2012, 185– 94, 192t
Obama, Barack (2008), 185– 86, 187– 88 ,
191– 93, 192t
vs. United Kingdom, 51– 52
voter targeting, 7, 8. See also individual voter
mobilization
voter turnout rates, voter mobilization, 55– 56
vote- swapping sites, 39
Wagner, Dan, 187– 88
Wallack, J. S., 60– 61
Ward, Stephen, 28, 32– 33, 77– 79, 80, 89– 90
web 1.0 technologies, 8
equalization to normalization, 12
technological and organizational capacity, 14
web 2.0 technologies, 8. See also specic campaigns
and topics
democratizing eects, 34– 35
strategic developments, 16
technological and organizational capacity, 14
WebCameron, 81, 83
web campaign cycles, mapping electorate, Australia,
125– 32, 126t, 127t, 130t
distribution, online receivers’ sociopolitical
characteristic, 129– 32, 130t
distribution, partisan, 126t, 127– 28, 127t
distribution, partisan, 2- year receive mode,
127t, 128– 29
distribution, partisan, campaign messages,
127t, 128
Greens, 127– 32, 127t, 130t
reading, 125, 126t
redistribution, 125– 27, 126t
redistribution, party supporters, 126t, 128
web campaign cycles, mapping electorate, France
reading, redistributing, and receiving, 155– 65,
157t, 159t
receive mode, sociodemographic and political
correlates, 162– 63, 163t
web campaign cycles, mapping electorate, United
Kingdom, 95– 105, 96t
reading, redistributing, and receiving, by party,
99– 105, 100t, 103t
reading, redistributing, and receiving, over time,
95– 99, 96t, 237
web campaign cycles, mapping electorate,
United States
reading, redistributing, and receiving, 194– 207,
196t, 200t, 203t
United States, 194– 207
World Values Survey (W VS), 60
youth effect, phase IV digital campaigning, 62,
63t
Downloaded from https://academic.oup.com/book/33611 by guest on 27 November 2022
