Leaders, Factions and The Game of Intra-Party Politics

Leaders, Factions and the Game of
Intra-Party Politics
The book provides a comprehensive view on the internal life of parties and
investigates the dynamics of intra-party politics in different party environments
to explain in which circumstances the party leader is more or less bound by the
wills of party factions.
Analyzing almost 500 intra-party documents from Italy, Germany and France,
it presents a theory of intra-party politics that illuminates internal decision-
making processes and sheds light on the outcomes of factional conflicts on the
allocation of payoffs within the party, on the risk of a party split and on the sur-
vival of the party leader. Using text analysis, the results show that consensual
dynamics can allow to preserve party unity and that directly elected leaders can
exploit their larger autonomy either to reward followers or to prevent splits.
This text will be of key interest to scholars and students of Party Politics,
Political Institutions, European Politics and more broadly to Comparative Pol-
itics, Political Theory and Text Analysis.
Andrea Ceron is Assistant Professor at the Department of Social and Political

Sciences, University of Milan, Italy.
Routledge Studies on Political Parties and Party Systems
Series Editors: Ingrid van Biezen
Leiden University, the Netherlands
and
Fernando Casal Bértoa
University of Nottingham, UK
This new series focuses on major issues affecting political parties in a broad sense.
It welcomes a wide-range of theoretical and methodological approaches on political
parties and party systems in Europe and beyond, including comparative works
examining regions outside of Europe. In particular, it aims to improve our present
understanding of these topics through the examination of the crisis of political
parties and challenges party organizations face in the contemporary world, the
increasing internal complexity of party organizations in terms of regulation,
funding, membership, the more frequent presence of party system change, and the
development of political parties and party systems in under-researched countries.
The Regulation of Post-Communist Party Politics

Edited by Fernando Casal Bértoa and Ingrid van Biezen
Party Systems in Young Democracies

Varieties of institutionalization in Sub-Saharan Africa
Edalina Rodrigues Sanches
Opposition Parties in European Legislatures

Conflict or Consensus?
Edited by Elisabetta De Giorgi and Gabriella Ilonszki
Party Members and their Importance in Non-EU Countries

A Comparative Analysis
Edited by Sergiu Gherghina, Alexandra Iancu and Sorina Soare
Party System Change, the European Crisis and the State of Democracy
Edited by Marco Lisi
Do Parties Still Represent?

An Analysis of the Representativeness of Political Parties in Western Democracies
Edited by Knut Heidar and Bram Wauters
For more information about this series, please visit: www.routledge.com/

Routledge-Studies-on-Political-Parties-and-Party-Systems/book-series/PPPS
Leaders, Factions and the
Game of Intra-Party Politics
Andrea Ceron
First published 2019
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2019 Andrea Ceron
The right of Andrea Ceron to be identified as author of this work has been
asserted by him in accordance with sections 77 and 78 of the Copyright,
Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or
utilized in any form or by any electronic, mechanical, or other means, now
known or hereafter invented, including photocopying and recording, or in
any information storage or retrieval system, without permission in writing
from the publishers.
Trademark notice: Product or corporate names may be trademarks or
registered trademarks, and are used only for identification and explanation
without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record has been requested for this book
ISBN: 978-1-138-55000-1 (hbk)

ISBN: 978-1-315-14709-3 (ebk)
Typeset in Times New Roman
by Wearset Ltd, Boldon, Tyne and Wear
To those who brought me to my first (2001) and latest
(2013) party congress
Contents
List of figures viii

List of tables ix
Preface x
Factions’ and parties’ acronyms and abbreviations xiii
1 The game of intra-party politics 1
2 Factional preferences in Italy, France and Germany 28
3 The determinants of factionalism 64
4 Bounded oligarchy: leaders, factions and the selection of

party platform 89
5 Portfolio allocation among party factions 109
6 The politics of fission: party splits in Italy, France and

Germany 135
7 Intra-party conflict and the survival of party leaders 159
8 Conclusion 177
Appendix 1: game–theoretic model: solutions through

backward induction 193
Appendix 2: detailed list of factions’ positions in Italy,

France and Germany 196
Index 210
Figures
1.1 Trends in the literature on intra-party politics relative to

studies on political parties 7
1.2 Leader selection and autonomy 12
1.3 Party Unity and Party Fission Game 14
2.1 Factions’ position on the left–right scale (Italy) 49
2.2 Weighted positions of Italian party factions in each party
congress 50
2.3 Discriminating power of words in Italy 51
2.4 Reliability of Wordfish estimates compared to hand-coding 53
2.5 Factions’ position on the left–right scale (France) 54
2.6 Factions’ position on the left–right scale (Germany) 55
4.1 Party position and weighted mean of factions’ position in Italy 97
4.2 Marginal effect of Gamsonian Agreement Position (GAP) on
Party Position as the new general elections approach 99
4.3 Party position and factions’ position in France and Germany 102
5.1 Marginal effect of Party Leader Faction on Weighted Share of
Ministers as the degree of Parliamentary Support for the
cabinet enlarges 117
5.2 Marginal effect of Share of Seats on Weighted Share of
Ministers as the Ideological Distance between one faction and
the median increases 118
5.3 Marginal effect of Share of Votes on Share of Ministers as
new general elections approach 121
5.4 Marginal effect of Share of Votes on Share of Ministers as the
Ideological Distance between one faction and the core of the
party increases 122
5.5 Factions’ votes and shares of ministers in Italy and France 128
6.1 Marginal effect of Ruling Party on Fission as Parliamentary
Support for the cabinet enlarges 143
7.1 Marginal effect of Internal Polarization and Overpaid Faction
on the survival of party leaders 169
Tables
1.1 Strategies played by the Leader and the Minority Faction and
potential outcomes based on the parameters of the Game 16
1.2 Parameters of the Game and expected effect on intra-party
dynamics 17
2.1 Italian data: list of parties, congresses and motions included in
the analysis 42
2.2 Correlation of Wordfish estimates when using pre-processing
strategies or not 58
3.1 Details on internal rules in Italy, France and Germany 71
3.2 The determinants of factionalism in Italy 77
3.3 The determinants of factionalism in France and Germany 79
4.1 OLS of party positions in Italy 98
4.2 OLS of party positions in France and Germany 103
5.1 Details on the Italian portfolio allocation data 114
5.2 Portfolio allocation among party factions in Italy 116
5.3 Portfolio allocation among party factions in France 120
6.1 Logit regression of faction breakaways in Italy 142
6.2 Logit regression of faction breakaways in France and
Germany 150
7.1 Determinants of party leader survival in Italy 168
7.2 Determinants of party leader survival (France) and
replacement (Germany) 171
8.1 The effect of factional policy preferences and internal rules in
Italy, France and Germany 177
8.2 Intra-party dynamics in parties ruled by autonomous or
constrained leaders 179
8.3 Parameters of the game: expected and observed effects on
intra-party dynamics 180
Preface
Everything started when I was a kid, and my father and grandfather brought me
with them to a local party congress meeting held in my town. The room was
crowded full of people and smoke. There were not enough chairs for everyone
and a lot of people were standing, raising their hands from time to time to cast a
vote. It was noisy. I saw people sharing ink-dirty newspapers that contained the
whole (long) texts of congress motions, passing them from hand to hand as pre-
cious goods. There were keynote speakers presenting each motion, common
members having their 15 minutes of fame to state their vote declaration. People
were chatting just outside the room. Others were bargaining behind closed doors.
Suddenly they came out with a creased sheet. That was the list of delegates
elected to the next stage, and a draft of the local party executive committee.
That’s where everything started.
I went to another party congress 12 years later. Everything was different.
Bigger room. Less smoke. Fewer speakers and pointless speeches. Boring cheap
rhetoric, but an enthusiastic audience. Shorter motions, on colorful high-quality
paper. Fewer decisions to be taken. I felt as if something had changed.
Now it’s exactly ten years from the first time I started to study party factions.
It was just a few weeks after the beginning of my PhD, in January 2009, when I
came up with the first general idea of investigating factional conflict and
cooperation inside parties. Three years later I was defending my PhD disserta-
tion and nowadays that idea is still at the core of the present book. A lot of water
has gone under the bridge in these ten years and my original project has grown
up so fast. More countries, more analyses, a refined game, new insights and a
wider time span, ranging from 1946 to the latest party congress held a short time
ago, in December 2018.
It has been a huge data collection effort indeed. I gathered almost 500 textual
documents, in several rounds, analyzing up to 3.6 million words. I went to librar-
ies hosted in places forgotten by God, to find pieces of paper forgotten by
humans. Sheets dating back to the 1940s almost crumbled in my hands. That’s
what research must look like. But time has passed, and the most recent waves of
data collection have been comparatively easier in the new digital era.
It has been a huge effort on my side. But I also want to thank all those who
contributed to this project in a way or another. I start with my family, blood
Preface xi
before faction. I also thank my (former and present) colleagues at the University
of Milan, too many to mention them all.
The thorough early draft of the project was presented, for the first time, at the
Research Workshop of Comparative Politics, Harvard University (Cambridge,
MA, February 2, 2011). All the smart comments that I received therein by the
discussant and the audience have been crucial in order to improve my work. In
the six months I spent there I also received precious comments from Stephen
Ansolabehere, Nahomi Ichino, James Snyder and Arthur Spirling. I want to
thank them for devoting time to read and discuss the present research with me.
Pieces of some chapters have been presented in conferences and workshops,
including the Conference of the Italian Society of Political Science (Palermo,
September 8–10, 2011), the Annual International Conference of the Political
Studies Association (Belfast, April 3–5, 2012), the Annual General Conference
of the European Political Science Association (Berlin, June 21–23, 2012), the
ECPR General Conference (Bordeaux, September 4–7, 2013), the MZES
Seminar (Mannheim, November 25, 2013) and the ECPR Research Sessions
(Essex, July 8–11, 2014). I thank the organizers, the discussants and the particip-
ants for their comments. I also thank all the reviewers that, year by year, con-
tributed to improve the quality of my work. Let me also thank, among others,
Kenneth Benoit, Daniela Giannetti, Chris Hanretty, Heike Klüver and Thomas
König, as well as all other scholars cited in the book. Additionally, I am grateful
to Thomas Bräuninger and Marc Debus for sharing with me their textual data on
German subgroups, and to Zac Greene for showing me how to access data on
French socialist factions. Finally, I want to thank all the colleagues of the Party
Congress Research Group for having revitalized my interest in factionalism and
intra-party dynamics.
Lately there is wide debate on the role of experts in politics. This concerns
several fields, including health and engineering. Conversely, in the field of
parties and political institutions everyone feels free to have his or her say. Politi-
cians, practitioners, analysts, journalists and people that you meet on the side-
walk often try to explain you the “game of politics,” ending up with predictions
that hardly ever come true. This book is also intended to show them that robust
political science theories can really be useful to predict political events. Indeed,
across the book the “game of intra-politics” will be successfully used to interpret
and anticipate the behavior of political actors in terms of policy line, payoff
allocation or party splits.
Apparently, the expertise of political scientists is not so much taken into
account, and electoral reforms or adjustments in party’s statutory provisions are
often drafted without or even against the advice of political science scholars.
That’s fine, that’s politics. But if you love experts, love them always, even in
this field, even when they say “no.” According to me, as this book will show, the
worst mistake a political leader can do is to surround himself with yes-men,
feeling that being the leader means always being right. The game–theoretic
model presented and tested throughout the book points exactly to this. Political
leaders should not act in isolation; they are not playing alone: intra-party politics
xii Preface
is a game involving (at least) two players and the preferences of factions con-
tribute to shaping the outcomes of the game too. Leaders should therefore con-
sider suggestions, criticism and disagreement. Dissenters might be right once in
a while.
This is not a book in defense of factions, but certainly in defense of disagree-
ments and, therefore, in a certain way, in defense of minority factions. Political
leaders should really listen to divergent viewpoints coming from internal minor-
ities. They should not necessarily address them, but these opinions should be
taken into account so that leaders can respond in the most appropriate way,
taking the right choice in the intra-party game. Such choice heavily depends on
the context, as this book shows.
To conclude, it is worth mentioning a statement by Veronica Roth’s book
Divergent (2011): “Faction before blood. More than family, our factions are
where we belong. Can that possibly be right?” That’s the question the book will
answer. Working together, leaders and faction can prevent damaging conflicts,
giving each other huge benefits and increasing the total amount of rewards avail-
able to the whole party, i.e., to their “family.” Accordingly, the book will show
what happens when leaders or factions put their own interest above all, regard-
less of what consequences this might have on the broader party community, in
terms of party unity, cohesion and splits.
Andrea Ceron, December 20, 2018
Factions’ and parties’ acronyms
and abbreviations
AfA Workers Association (Germany–SPD)

AfD Alternative for Germany
AGS Self-employed Association (Germany–SPD)
AN National Alliance (Italy)
CDA Christian Democratic Employees (Germany–CDU)
CDU Christian Democratic Union of Germany
CÉRÉS Center of Socialist Studies, Research and Education
(France–PS)
CSU Christian Social Union (Germany)
DC Italian Christian Democracy Party
DL Democracy is Freedom – The Daisy (Italy)
DL21 Forum of the Democratic Left 21 (Germany–SPD)
DS Democrats of the Left (Italy)
FdI Brothers of Italy
FDP Liberal Democratic Party (Germany)
FI Forza Italia
FK Frankfurt Circle (Germany–SPD)
FKA Union of Values/Liberalist–Conservative Movement
(Germany–CDU)
FLI Future and Freedom for Italy
FV Federation of the Greens (Italy)
IDV Italy of Values
LDP Liberal Democratic Party (Japan)
LN Northern League (Italy)
LREM La République En Marche! (France)
M5S Five Stars Movement (Italy)
MIT Middle Class and Business Association (Germany–CDU)
MoDem Democratic Movement (France/Italy–PD)
MSI Italian Social Movement
NB Berlin Network (Germany–SPD)
NCD New Centre–Right (Italy)
NPSI New Italian Socialist Party
PCF French Communist Party
xiv Acronyms and abbreviations
PCI Italian Communist Party
PCL Communist Workers Party (Italy)
PD Democratic Party (Italy)
PDA Action Party (Italy)
PDCI Party of Italian Communists
PDL People of Freedom (Italy)
PDS Democratic Party of the Left (Italy)/Party of Democratic
Socialism (Germany)
PL Parliamentary Left of the SPD parliamentary group
(Germany–SPD)
PLI Italian Liberal Party
PP People’s Party (Spain)
PPI Italian Popular Party
PRC Communist Refoundation Party (Italy)
PRI Italian Republican Party
PS French Socialist Party
PSDI Italian Socialist Democratic Party
PSI Italian Socialist Party
PSIUP Italian Socialist Party of Proletarian Unity
PSLI Italian Socialist Workers’ Party
PSOC Socialist Party (Italy)
PSOE Spanish Socialist Workers’ Party
RED Reformists and Democrats (Italy–PD)
RPR Rally for the Republic (France)
SA The Left–The Rainbow (Italy)
SC Critical Left (Italy)
SEL Left Ecology and Freedom (Italy)
SFIO French Section of the Workers’ International
SK Seeheim Circle (Germany–SPD)
SPD Social–Democratic Party of Germany
UDC Union of Christian and Center Democrats (Italy)
UDF Union for French Democracy
UKIP UK Independence Party
UMP Union for a Popular Movement (France)
UPyD Union, Progress and Democracy (Spain)
VERTS/EELV The Greens/Europe Ecology – The Greens (France)
WASG Labor and Social Justice – The Electoral Alternative (Germany)
1 The game of intra-party politics
Introduction
Many studies in the field of political science rely on the assumption that political
parties are unitary actors. Indeed, parties play a crucial role in representative
democracies. They interact together in the electoral market or in the parlia-
mentary arena, trying to maximize their own rewards through processes of
cooperation (e.g., building pre-electoral alliances or coalitions) and conflict (e.g.,
electoral competition or quarrels on everyday payoffs allocation). In a word,
parties matter: they affect government formation, portfolio allocation and policy-
making. Parties are also key elements in the process of political accountability
and can foster the responsiveness of the political system toward the demands of
citizens and voters.
In light of this, parties are often treated as unitary actors, where individual
members coalesce to reach common goals. However, the unitary actor assump-
tion is just a fictional representation of the reality. In fact, parties are usually
composed of members retaining similar but non-identical preferences. For this
reason, those members who retain shared preferences and common interests can
join together; they will create intra-party subgroups to boost their influence on
internal decision-making. These subgroups, that might have different shapes and
features, have been identified under the label of party factions.
From this perspective, the present book will relax the unitary actor assump-
tion, opening the black box of intra-party decision-making to provide new
insights on the internal life of political parties with an eye to inter-factional com-
petition. By doing that, it will shed light on how intra-party dynamics interact
with attributes of the party system and shape the nature of a political system.
Why factions?
Factions exist and many parties in many countries are factionalized. Sometimes
parties manage to restore their unity, finding an internal equilibrium that pushes
all subgroups to cooperate. Other times, these factional conflicts worsen and
produce public splits and party fissions. Indeed, real-world politics is filled with
examples of intra-party disagreements or factional fights.
Historically, between the nineteenth and the early twentieth century, two of
the oldest democracies, i.e., the United States and the United Kingdom, were
affected by factional rivalries leading to party splits. In the US, both the Demo-
cratic (1828) and the Republican Party were created by factions of the dissolved
Democratic–Republican Party. In the UK, both the Conservative Party (Tories)
and the Liberal Party were hit by the breakaways of dissident factions: the
Peelites faction split from the Tories in 1846 and dissenters of the Liberal
Unionist Party broke away from the Liberals in 1886.
At the beginning of the twentieth century, the enlargement of suffrage and the
entrance of socialist parties in the electoral arena fueled intra-party conflicts
further. Traditionally, in fact, socialist parties have been divided between an
ideologically orthodox faction supporting strong State intervention into the
economy, and a more liberal–democratic pro-market group of reformists. After
the October Revolution, a more radical Marxist faction formed. As a result, com-
munist parties were born as splits of the left-wing factions of the socialists.
Inside socialist parties, however, the internal conflict continued even after the
communist’s split.
In the post-war period, the socialists were still divided between moderate and
radical wings, struggling to reach a compromise. In some countries, such as
Germany, party unity was preserved. In France the socialists reached unity in 1969,
after a period of intense divisions. In other countries, the socialists split again. The
Italian Socialist Party (PSI) repeatedly split; in 1947 its anti-communist moderate
factions broke away to form the Italian Socialist Democratic Party (PSDI), while in
1964 the left-wing faction exited to create the Italian Socialist Party of Proletarian
Unity (PSIUP). In the UK the Labour Party split in 1981, when members of the
internal right-wing quit to form the Social Democratic Party, which allied with the
Liberal Party, revitalizing its electoral performance.
Factionalism and splits are not merely a thing of the past. Very recent polit-
ical events, all over the world, emphasize the importance of intra-party politics.
Just to mention a few, on the right-of-center camp one can think to Brexit, which
harmed the UK Conservative Party from 2016 to 2018, or to the EU migrants’
crisis, which weakened Angela Merkel, forcing her to quit the leadership of the
Christian Democratic Union (CDU) in 2018 after having been challenged by
emerging right-wing factions. On the left side, recent episodes of factionalism
took place in the context of welfare and labor market reforms, i.e., the
“loi Travail” (2016) in France and the “Jobs Act” (2014) in Italy, promoted
respectively by the dominant moderate faction of the French Socialist Party (PS)
and the Italian Democratic Party (PD) against the will of internal left-wings and
trade unions (Ceron & Negri 2016, 2017); intriguingly, this conflict mimics the
fight inside the Social–Democratic Party of Germany (SPD) around the “Agenda
2010” welfare reform.
To start with, the debate on Brexit is inherently linked with intra-party divi-
sions. In the United Kingdom the three main parties, Conservatives, Labour and
Liberal Democrats, have traditionally been divided into clubs, factions, wings or
tendencies and these clubs fought to take control of the party. Along this vein,
The game of intra-party politics 3
current UK politics highlights the existence of wide heterogeneity inside both
the Labour (the party were split in a vote on Brexit in 2018) and the Conser-
vative Party. Conservative Members of Parliament (MPs) were split in 2016,
during the referendum campaign, and they split again in autumn 2018, during
the negotiation, disagreeing on the Theresa May’s cabinet road to Brexit. These
divisions eventually led to the formation, in February 2019, of The Independent
Group, a pro-EU parliamentary group composed of 11 switchers coming from
the Labour Party (8) and the Conservative Party (3).
The good electoral performance of anti-establishment parties is now altering
the structure of European party systems, increasing political instability even in
countries such as the UK, Spain and Germany that for a long time have been
deemed stable and safe from political fragmentation.
Such rise of populist and Eurosceptic movements, however, is not a totally
new phenomenon. On the contrary, it was already partially rooted in intra-party
factionalism. The Eurosceptic wing of UK Conservatives, active well before the
rise of the UK Independence Party (UKIP), is a typical example of this. But
scholars report the existence of anti-EU factions even inside the German liberals
(Taggart 1998), many years before the advent of the Eurosceptic Alternative für
Deutschland (Alternative for Germany, AfD).
In France, “sovereignist” factions existed both inside the socialists and the
Gaullist parties. Since 1992, the left-wing factions of the PS, headed by Jean-
Pierre Chevènement and Jean-Luc Melénchon, strongly opposed the ratification
of European treaties. These factions, which later broke away from the PS, can be
considered as the seeds of the new anti-establishment party La France
Insoumise. Inside the Gaullists, a faction founded by Philippe Séguin and
Charles Pasqua was already active in the 1990s; this faction campaigned against
the Maastricht Treaty in the 1992 referendum and split in 1999, when it ran in
the European election separately, on a sovereignist platform. The heirs of this
faction continued to fight inside the Gaullist Party until 2007, when Nicolas
Dupont-Aignan’s Debout la République broke away to form a new right-wing
party, which in 2017 endorsed the far-right populist candidate, Marine Le Pen,
in the run-off of the Presidential election.
Moving back to the left camp, after the fall of Berlin’s wall and the decline of
ideologies, elements such as the rise of public debt as well as the fulfillment of
Maastricht criteria in the new era of welfare state retrenchment pushed socialist
parties to moderate their stances; this has exacerbated internal division between
neoliberal reformist factions and traditional pro-welfare leftist groups, which
tried to resist to such policy shift (Ceron et al. 2019; Marx & Schumacher 2013).
Without losing its unity, in the UK the Labour Party swung from the neolib-
eral platform enacted by Tony Blair’s faction to the more left-oriented policy
proposals set up by the new left-wing leadership of Jeremy Corbyn. In many
other cases, however, such factional strife between social–democrats and social–
liberals often ended up with party fissions that altered the structure of the party
system, weakening the socialist parties and opening the way to drastic political
changes. Examples can be found in several countries.
In Germany, the factional fight between Gerhard Schröder and Oskar
Lafontaine, inside the SPD, culminated with the exit (2005) of Lafontaine, who
disagreed with the neo-centrist Neu Mitte platform implemented by Schröder. In
response, Lafontaine promoted a new left party, Die Linke, which boosted the
electoral performance of the ancestor far-left Party of Democratic Socialism (PDS).
The rise of Die Linke was crucial to generate parliamentary gridlocks, forcing the
SPD to form Grosse Coalitions with the center-right CDU. Nowadays, the left-
wing of the SPD is still fighting with the right-wing Seeheimer Kreis, trying to dis-
mantle the Grosse Coalition to propose a new alternative left-wing alliance that
includes Die Linke. Remarkably, the internal conflict inside the SPD is reaching a
new peak: in the 2018 party congress, the former left-wing leader Andrea Nahles
(challenged by a more left-wing candidate) has been appointed with the second
worst score ever in the history of the SPD.
In France, part of the left-wing quit the PS in 2008 to create the left party,
Parti de Gauche, which later turned into La France Insoumise: its leader, Melén-
chon, ranked fourth in the 2017 Presidential election winning almost 20 percent
of the votes. The ideological divide between the right-wing and the left-wing of
the PS is also one of the causes of the party’s collapse in the 2017 elections. The
massive parliamentary conflict animated by internal left-wing frondeurs, against
the economic and welfare policies (particularly against the loi Macron) of the PS
cabinets led by Jean-Marc Ayrault, Manuel Valls and Bernard Cazeneuve, pro-
vided the basis for the exit of moderate politicians. As a result, Emmanuel
Macron launched the neo-centrist party La République En Marche! (LREM) and
won the 2017 Presidential election. After the elections, the PS (reduced to 7
percent of votes), was affected by further splits: Benoît Hamon (PS candidate in
that election) quit in July 2017, and another left-wing faction broke away in
October 2018.
In Italy, the conflict between liberal–democratic and social–democratic fac-
tions in center-left parties started well before the foundation of the PD (2007).
Later on, such strife reached a peak when the centrist Matteo Renzi became
party leader (2013); for the sake of enhancing policy cohesion, Renzi started to
neglect the requests of internal minorities and repeatedly imposed the whip.
Weakened by such internal disagreement and by a public strife that was visible
everyday on mass media and social media, the party broke up as left-wing fac-
tions quit in 2015 and 2017. In the following 2018 election, the PD obtained its
worst electoral score ever, winning only 18.76 percent of votes, while the
splinter groups got 3.4 percent. Intriguingly, looking beyond the PD, in the after-
math of the 2013 elections all the Italian parties have been affected by internal
conflicts and by repeated episodes of party fissions (Ceron 2017a, 2017b) that
have altered, once again, the structure of the party system.
The last example concerns the German CDU. This party was traditionally
characterized by the absence of organized factions, apart from loosely connected
tendencies in the 1960s (Dilling 2018). However, factionalism started to mani-
fest in recent years. As a reaction to the centrist shift in the party line promoted
by Merkel, particularly on immigration policy, the most conservative party
members joined forces to contest the Merkel leadership. Indeed, internal right-
wing factions started to form. In the mid of 2000s, conservative members formed
the Einstein-Connection (2007) and the Xantener Kreis; a few years later the
Berliner Kreis in der Union was founded (2012). In 2017, this latter faction con-
tributed to form the Liberalist–Conservative Movement “Union of Values”
(Freiheitlich-konservativer Aufbruch–die WerteUnion, FKA). As a reaction, a
moderate faction (Union der Mitte, Union of the Center) was born too, to support
the centrist policies of Merkel. Allergic to factionalism, the CDU leadership
refused to recognize these ideological factions as official CDU subgroups.1
This did not prevent an increase in ideological factionalism, as other conser-
vative factions organized to contest the party leadership race. Merkel attempted
to appease dissenters by appointing Jens Spahn, one of their leaders, in a minis-
terial position. This strategy was insufficient though. The mounting internal dis-
agreement, boosted also by the declining electoral performance of the CDU,
produced a first effect in September 2018 when the CDU/Christian Social Union
(CSU) Parliamentary Party Group replaced the incumbent party whip loyal to
Merkel. As a result, anticipating the possible outcome of the 2018 party con-
gress, Merkel decided to step back. This resulted in a tied leadership race that
highlighted the broad disagreement existing inside the CDU at this moment (see
Chapters 5 and 6).
All these recent events prove that factionalism has a visible and strong influ-
ence on everyday politics in several parties and countries. The CDU case, in this
regard, is particularly intriguing. It suggests that intra-party divisions can exist
even when they are not directly observable.
Indeed, in many parties, for most of the time “internal disagreements are
resolved before party positions are defined formally – as in party manifestos – or
behaviorally – as in legislative votes and speeches” because “The incentives for
parties to present a unified front in the wider political arena are strong” (Heller
2008: 2). In this sense “the display of party unity does not imply a lack of con-
flict inside the party; rather it simply shows that party members have agreed to a
party position that they all can support, or at least tolerate” (Heller 2008: 4).
In view of that, it is important to distinguish between unity and cohesion. In
the rest of the book I will stick to the following conceptualization: unity refers to
the idea of keeping the party together and implies that all factions remain inside
the party, so that one does not observe breakups, even if factions might still
retain different preferences and views on how the party line and strategy ought
to be. Unity can be observed in the presence of cohesion (when all factions agree
on the party line) or in the absence of it (when factions retain divergent views
and express their dissent inside the party). Conversely, the lack of unity suggests
that intra-party dissent produced party splits and fissions.
To the contrary, cohesion indicates that factions think and behave in a homog-
enous way, which is in agreement with the party line. Cohesion, then, is the lack
of dissent over the party line. Note, however, that cohesion can be reached
through cooperation (consensual agreement) or through enforced discipline
(even in the presence of heterogeneous views) or internalized loyalty.
To sum up, the simple fact that one does not hear any dissenting “voice” or
do not notice episodes of “exit” does not imply that disagreement and hetero-
geneity are utterly absent. In fact, conflict might have been resolved before it
produces visible negative consequences, or it might remain hidden because
internal dissent is too weak to openly defy the leadership.
Literature on intra-party politics and party factions

So far I have reported several examples from real-world politics showing that
factionalism is a widespread phenomenon. But what about academic studies?
Scholars argue that “most parties in the world have wings or tendencies”
(Chambers 2008: 304) so that “Factionalism is a fact of life within most political
parties” (Harmel et al. 1995: 7). Others are even more explicit and claim that
“Political parties are driven by the spirit of faction” (Heller 2008: 2) and
“factions are ubiquitous” (Dewan & Squintani 2016: 875). In view of that, the
academic literature has investigated the causes and the consequences of intra-
party divisions covering all the world’s regions, from Australia to Brazil, from
Mexico to Japan, from Germany to Israel and South Africa, not forgetting Italy,
the UK and the US.2
Apart from episodic earlier mentions, the study of party factions started in the
1960s and reached a peak in the 1970s. A rough query on Scopus about aca-
demic documents containing the words “factionalism,” “party faction/s,” and
“intraparty/intra-party” reveals that between 1970 and 1979 the number of pub-
lished studies (38) doubled compared to the previous decade (see the solid black
line in Figure 1.1). This represents almost 20 percent of all studies referring to
“political party/ies” at that time. In the 1980s and the 1990s other fundamental
contributions have been published, though only in the new millennium there is a
sharp growth (in absolute terms) of the academic publications related to such a
topic. Overall, from 1960 to 2018, almost 10 percent of all academic contribu-
tions on political parties (retrieved from such query) was devoted to intra-party
politics. The time series of the relative size of this subfield is stationary (dashed
line in Figure 1.1), suggesting that intra-party politics is a well-established area
of party politics, which consistently generates interest over time. Remarkably,
with respect to articles in journals, in the ten years between 2009 and 2018 the
number of published articles has increased by 180 percent compared to
the previous ten years. This is, by far, the steepest increase since the 1970s (see
the gray dropline in Figure 1.1).
If the attention on party factions has been revived in recent years, the usage of
such a term dates back in time. It was already employed in the eighteenth century
by intellectuals such as Burke, Hume and Madison as a forerunner for the
concept of party. It acquired its current meaning only later. In 1949, analyzing
primary elections of the US Southern Democrats, Key (1949) referred to factions
as intra-party groups.
The term and the definition of faction have long been debated in the literature.
Beside “factions” (Duverger 1954; Key 1949) some authors proposed alternative
Figure 1.1 Trends in the literature on intra-party politics relative to studies on political
parties.
terms: “tendencies” (Rose 1964), “fractions” (Sartori 1971, 1976), “currents”

(Belloni & Beller 1978) and “clusters” (Reiter 1981) as well as wings, rivalries
and nucleus.
Indeed, a wide number of studies have focused on the definition of factions,
providing classifications and taxonomies (Beller & Belloni 1978; Belloni &
Beller 1978; Bettcher 2005, Boucek 2009; Hine 1982; Key 1949; Nicholas 1965;
Panebianco 1988; Reiter 1981; Rose 1964; Sartori 1971, 1976; Seyd 1972;
Zariski 1960; Zincone 1972).
These studies classified factions according to their size, origin (parlia-
mentary arena or party body), level of organization (i.e., the presence of a fac-
tional press and headquarters), pervasiveness (local or national), aim
(ideology, clienteles, leadership support) and duration (longstanding groups or
temporary fluid aggregations). In addition, factions can be differentiated
according to the availability of an autonomous access to the resources (mem-
bership fee and fundraising) and links with collateral extra-party organizations
(i.e., trade unions: Ceron & Negri 2017).
Key (1949: 16) provided a first definition of faction as “any combination,
clique, or grouping of voters and political leaders who unite at a particular time
in support of a candidate.” Zariski (1960: 33) generalized this idea defining them
as “any intraparty combination, clique or grouping whose members share a sense
of common identity and common purpose and are organized collectively – as a
distinct bloc within the party – to achieve their goals.” Zariski, in particular,
pointed to the existence of several goals such as patronage, the pursuit of local
demands or interest groups’ requests, the promotion of values, or the influence
on party strategies and policies.
Later on, Nicholas (1965) focused on factions as “leader–follower” groups,
considering them as personal troops tied to their General. According to him, fac-
tions have defined tasks and hierarchy: members give support to the leader
during intra-party races, and they receive back jobs and money. Sartori (1971,
1976) defined factions as personalized power groups totally committed to power
manipulation. He distinguished between factions of principles/values (ideologic-
ally oriented) and factions of interests/convenience (office-seeking); however, he
emphasized the role of patronage claiming that factions of principles tend to
transform into factions of interest and maintain ideological elements only for
instrumental purposes.
Despite this, one should notice that it “Doesn’t really matter […] whether the
public expression of policy preference is sincere or not. What matters is that it is
costly to turn back from this.” As a consequence, “a politician whose private
desires deal only with getting into office for its own sake must become inextric-
ably associated with a particular policy position in order to fulfil these object-
ives” (Laver & Shepsle 1996: 249). As a result, even instrumental preferences
can become actual policy positions with real effects on politics.
In fact, other scholars pointed to the role of ideology (Rose 1964; Seyd 1972).
Rose (1964) distinguished between faction and tendency, though the existence
of common policy preferences was deemed important in both cases. In fact, a
tendency was defined as a disorganized group, built according to shared atti-
tudes, whereas a faction is a structured group (characterized by a clear ideology,
leadership, technical expertise and availability of resources) that aims to pursue
a range of policies through an organized activity.
To sum up, ideology matters and even authors that emphasize the role of
patronage also recognize that factions could be arrayed along the left–right scale
(Bettcher 2005; Golden & Chang 2001; Zuckerman 1979). For instance, Golden
and Chang (2001: 605) described Italian Christian Democracy (DC) factions as
“not ideologically distinct, nonetheless […] they can be arrayed ideologically”;
Bettcher, while stressing the primacy of patronage, admits that such factions are
similar to factions of principle as they “identified themselves along a left-right
spectrum” (Bettcher 2005: 350) and, more in general, all factions care (at least
partially) about policy.
With these features in mind, Beller and Belloni (1978: 448) tried to find a
common ground. They provided a general definition of factions as “any relat-
ively organized group that exists within the context of some other group and
which (as a political faction) competes with rivals for power advantages within
the larger group of which it is a part.” Thirty years later, Boucek (2009: 468)
summarized such debate providing a thorough definition of factionalism as “the
partitioning of a political party (or other organization and group) into subunits
which are more or less institutionalized and who engage in collective action in
order to achieve their members’ particular objectives.” Hereinafter, I will stick
to such more general definition of factions.
Beside this, the academic literature has focused on two large areas of study.
On the one hand, scholars considered factions as the dependent variable, trying
to assess the causes of factionalism. On the other, different streams of research
analyzed factionalism as an independent variable, in order to explain several out-
comes of the political system.
The first group of studies tries to assess the determinants of factionalism
taking into account a variety of elements such as party system competitiveness
(Boucek 2012; Sartori 1971, 1976; Zariski 1960), the amount of resources
available inside the party (Golden & Chang 2001; Zincone 1972; Zuckerman
1979), incentives provided by the electoral systems (Carey 2007; Carey &
Shugart 1995; Cox & Rosenbluth 1995; Cox et al. 1999; Katz 1986) or intra-
party rules (Rahat & Hazan 2001; Sartori 1971, 1976). Only recently scholars
highlighted the importance of policy preferences showing that they shape fac-
tional membership (Bernauer & Bräuninger 2009; Giannetti & Laver 2009;
Spirling & Quinn 2010).
Conversely other studies examine the impact of factionalism on politics and
policy. Concerning office, scholars showed the impact of factions on coalitions
and government bargaining (Giannetti & Benoit 2009; Laver & Shepsle 1990,
1996; Meyer 2012; Strøm 1994), portfolio allocation (Ceron 2014; Debus &
Bräuninger 2009; Leiserson 1968; Mershon 2001; Ono 2012), government dura-
tion, termination and reshuffle (Budge 1985; Chambers 2008; Druckman 1996)
or patronage and corruption (Golden & Chang 2001; Zuckerman 1979).
With respect to policy, studies investigated party change, party competi-
tion and the process of party position-taking (Budge et al. 2010; Caillaud &
Tirole 2002; Ceron 2012; Giannetti & Laver 2009; Levy 2004; Persico et al.
2011; Roemer 2001), as well as party cohesion and discipline in legislative
policy-making (Ceron 2015b; Giannetti & Laver 2009; Kam 2009; Spirling &
Quinn 2010).
The impact of factionalism on party unity and on the shape of the party
system (Kato 1998; Laver & Benoit 2003; Sartori 1976) has been scrutinized
too, with an eye to party switching (Desposato 2006; Heller & Mershon 2008)
and party fission (Ceron 2015a; Giannetti & Laver 2001; Kato 1998; Reed &
Scheiner 2003).
Such previous findings will be taken into account in the present study. In the
rest of the book I will explore all these areas of research, digging into the deter-
minants of factionalism, studying the impact of factions on the allocation of
policy and office payoffs, and combining these elements to understand the occur-
rence of party splits. In turn, these different topics will be investigated in light of
a new theory of party unity and party fission, which looks at intra-party politics
as the interplay between the party leader and minority factions.
Factional politics: a theory of party unity and party fission
Parties are voluntary associations composed of like-minded individuals that join
together to solve collective action and coordination problems (Aldrich 1995;
Cox & McCubbins 1993; Kiewiet & McCubbins 1991). It is well established
that politicians and political parties, as well as other political actors (including
factions), care to some extent about three different goals: policy, office and votes
(Müller & Strøm 1999; Strøm 1990). Although political actors can attach
different weights to these goals, politicians ideally try to achieve all three of
them also because they are, in a way or the other, interrelated.
By grouping into a single organization, rather than acting alone, politicians
and party members may extract greater policy payoffs in the parliamentary arena
through log-rolling and coordinated voting behavior, or get more votes in the
electoral market, where enforced cohesion increases the value of the party label
thereby enhancing prospects for re-election of party’s representatives (Snyder &
Ting 2002). Furthermore, acting as a united group also enhances the likelihood
of getting into office and the relative average reward linked with cabinet parti-
cipation (Bäck 2009).
However, the party is not a monolithic actor. Members may hold hetero-
geneous policy preferences, interests and personal career ambitions that do not
combine naturally with those of other members. For this reason, members with
most similar views have an incentive to gather together and create subgroups
and factions within a party. In light of this, I can consider the party as a coalition
of factions (Leiserson 1968).
Inside each party, factions are involved in a common effort to produce the
public good, i.e., party unity. As a matter of fact, unity can boost the electoral
performance of a party (McGann 2002; Snyder & Ting 2002), increasing the
expected share of policy payoffs that the party can get during everyday policy-
making. In addition, unity enhances party strength during the negotiation over
coalition formation (Bäck 2009; Baron 1998) and enhances the likelihood of
being involved in a coalition government (Ceron 2016) thereby increasing the
total amount of payoffs available to its members.
On the one hand, then, factions need to cooperate to reach an agreement to
foster the party. On the other, they compete against each other to take control of
the party and maximize their share of (policy, office and electoral) payoffs
derived from party membership and from having produced the public good. In
this way, intra-party politics swings between conflict and cooperation, with fac-
tions seeking a balance between the two (Maor 1997). To the extent that “polit-
ical parties are endogenous, then members of party factions may be seen to
belong for as long as it is rational to do so” (Laver & Kato 2001: 510). Accord-
ingly, inter-factional conflict and negotiation take place in the shadow of party
fission.
How to deal with such internal conflict? As far as we are dealing with the
production of a public good (party unity) we have to cope with the classic
collective action dilemma (Olson 1965). While repeated interaction between
factions can foster the emergence of cooperative norms, the party needs to select
one actor in charge of organizing the work of subgroups in order to produce such
public good (Cox & McCubbins 1993). Then, the solution is to select a party
leader that is the agent in charge of preserving unity (Kiewiet & McCubbins
1991), allocating payoffs and enhancing the party’s performance. By assigning
incentives, the leader could motivate party members and factions and get them
to collaborate, campaigning harder during the election (Carroll & Cox 2007) and
behaving cohesively in the aftermath, thereby providing more benefits to the
whole party. The leader will also be in charge of selecting the party platform and
exploits his3 charisma or any personal valence advantage in order to maximize
the party’s share of votes, seats, cabinet positions and policy outputs. In turn, the
leader has to make a decision on how to split these payoffs, trying to retain a
share of them as a reward for him4 and his followers,5 but always keeping in
mind the target of minimizing damaging internal conflicts with other subgroups/
factions.
Moreover, insofar as internal rules are not always the same, this process
should lead to different outcomes under different intra-party environments. In
fact, the degree of proportionality in portfolio allocation and a leader’s ability to
get benefit from his position may depend on his degree of autonomy from party
factions.
Several authors suggested that direct forms of internal democracy might allow
party leaders to manipulate the members for their own ends (Katz 2001; Katz &
Mair 1995; Kenig 2009; Marsh 1993; Michels 1915; Poguntke & Webb 2005;
Rahat et al. 2008). Leaders might formally empower members to overcome the
resistance of middle-level activists and dissenting factions. In addition, thanks to
elements such as name recognizability and the front runner effect (Kenig 2009),6
the direct election made by a wide and inclusive “selectorate” paves the way to a
“Bonapartist” form of intra-party democracy (Michels 1915) and to a less repre-
sentative party environment (Rahat et al. 2008).
The relevance of the leader selection process also emerges from an empirical
investigation of intra-party debates. Factions often complained of how the direct
election of leaders had altered the shape of intra-party dynamics. For instance,
after the implementation of direct election, in the 1984 DC congress, Vincenzo
Scotti, the runner-up party leader candidate, jointly supported by the moderate
faction Impegno Riformista and the left-wing Forze Nuove, warned against the
risk of transforming the DC into a plebiscitarian party ruled by charismatic
leaders, due to this new selection mechanism. Later on, in the 1989 congress,
Forze Nuove complained against plebiscitarianism, arguing that the direct elec-
tion had created a personalized leadership that attempts to disempower middle-
level activists by appealing directly to the grassroots.
Conversely, in the same congress a rival left-wing faction, Base, defended the
direct election as a solution to overcome the idea of party leader as a primus inter
pares, constantly bounded by factional veto powers. Such dichotomy between a
model of party ruled through factional agreements or led by a strong leader able to
promote the stakes of the party mainstream has long been a concern for party
factions: indeed, the DC faction Tavianei appealed for more leader autonomy in
spite of factional constraints already in its 1967 congress motion.
In recent years, a similar debate took place inside the PD too. For instance,
during his fight with internal minorities, the PD leader Renzi attacked dissenting
factions arguing that any attempt to criticize him or his policy choices was
disrespectful toward three million of PD members who appointed Renzi as party
leader. Later, in 2018, prominent PD politicians (including the pro-Renzi leader
candidate, Marco Minniti, which later stepped back for the sake of enhancing
party unity) warned against the risk that if no leader candidate will win more
than 50 percent of votes in the 2019 PD direct election, the new party leader will
be indirectly appointed by the party body and this will prove that the PD is
nothing more than a confederation of factions.7 Accordingly, I claim that the
degree of leader autonomy is affect by the leadership selection rules, i.e., by the
direct election and the size of the selectorate.
Figure 1.2 summarizes the two different scenarios. It brings together two ele-
ments indicating whether the party leader is elected directly by a wide selector-
ate (composed of members or activists), or indirectly by a restricted committee.
The dichotomy direct/indirect election overlaps with the inclusiveness of the
selectorate; in fact, when parties adopt direct election, they tend to expand the
selectorate. Conversely, indirect election is usually delegated to a small party
committee.
By looking at Figure 1.2, one can notice that in the first scenario, which is
typical of mass parties, the chain of delegation begins with party members,
which are entitled to elect their representatives at the national party congress.
The party congress, in turn, selects a small committee – the party executive –
that appoints the party leader.8 Here there is a strong emphasis on membership,
hierarchy and bureaucracy. The leader is the agent and should be responsive to
the will of his principal, which in this case is composed of high-ranking politi-
cians (party executive).
Figure 1.2 Leader selection and autonomy.

Conversely, the second scenario is more similar to the “cartel party” model,
with a stronger focus on the primacy of the party leader that bypasses middle-
level activists. In such a type of party, the selectorate of the leader has been
enlarged so that party members are directly entitled to cast a vote and select the
party leader;9 the leader, in turn, will select a team of followers that compose the
party executive and help him to carry out his tasks. On the one hand, the leader
should be directly responsive to party members (and somehow tied to the prom-
ises and the platform that have been presented directly to them). On the other
hand, the party executive becomes the agent of the leader (instead of his prin-
cipal). Accordingly, the chain of delegation is reversed: the leader can gain
wider control over party activists and becomes more autonomous from the wills
of party factions.
As a consequence, I distinguish between leaders directly elected (by party
members or delegates at party congress) and those elected indirectly in small
committees (smoke-filled rooms) where any single activist can easily alter the
equilibrium keeping the leader tied to factional vetoes.
Non-autonomous leaders selected in a small committee are nothing more than
the output of an inter-factional bargaining, appointed to implement an agreement
that has already been reached elsewhere. They retain no autonomy and should
simply stick to such deal offering a compromise to the minority faction and split-
ting the payoffs fairly. If the selection of the leader is the result of an inter-
factional compromise, factions retain a stronger veto power and any deviation
from the factions’ mandate can be seen as a treason that would lead to a leader
replacement.
In this context, even if the leader theoretically retains an agenda-setting power
(as in a classic “Ultimatum Game,” in which the leader has to split a total
amount of payoffs and can make a fair or unfair proposal that can be accepted or
rejected by the minority faction), his limited autonomy implies that one can
expect consensual dynamics leading to a more proportional allocation of payoffs.
Indeed, while the classic solution of the Ultimatum Game suggests that the
leader should propose an unfair agreement and the minority should accept it,
empirical evidence from real-world politics and experimental economics pro-
vided confirmation for an alternative solution in which payoffs are split fairly
(Güth 1982; Nowak et al. 2000).10
To sum up, indirect election that takes place in a small party committee domi-
nated by factions erases the autonomy of the leader. Factions’ vetoes become an
insurmountable constraint and, for the sake of enhancing his survival in office,
the leader will stick to the consensual inter-factional agreement avoiding any
conflict with internal factions.
What happens when leaders are directly elected by a wide selectorate and are
therefore more autonomous? I propose a game–theoretic model based on the
“Exit, Voice, and Loyalty” framework (Hirschman 1970) to shed light on
the interplay between the party leader and internal factions, and to investigate
the path leading either to party unity or to party splits.11 Figure 1.3 presents a
game of party unity and party fission, describing factional disputes over payoff
Figure 1.3 Party Unity and Party Fission Game.

Note
µ = value of party unity; α = minority’s size; ν = damage due to public voice; ε = exit cost; π = minority
payoffs after breakaway; ω = loss due to party fission. It has been assumed that: µ = 1; 0 < α < 0.5;
ε ∈ (0, 0.5); π ∈ (0, 0.5); ω ∈ (0, 0.5); ν ∈ (0, 0.5) and ν < α. F = minority faction; L = party leader.
Payoffs are indicated in square brackets and kept separated by a semicolon. Faction payoffs are indi-
cated first, followed by leader payoffs.
allocations under the threat of party breakup. Given that party leaders and fac-
tions are insiders, I argue that they have full access to information concerning
payoffs, strategies, moves and nodes of the game.12 I also assume common
knowledge between the actors. Hence I represent intra-party dynamics as a game
of perfect (and complete) information.
For simplicity, I consider only two actors, the party leader, L, tied to the
mainstream faction (composed of his followers) and a minority faction of dis-
senting members, F. The minority’s size α is the share of congress votes won by
the faction (a positive value below 0.5). The total amount of office payoffs to be
shared sum to µ,13 which is equal to one, and any strategy undermining party
unity imposes costs on the actors. L is in charge of allocating scarce resources
such as cabinet spoils (office payoffs), the party line (policy payoffs) and candi-
dacies (electoral payoffs) between the two factions.
As long as the leader receives the mandate directly from party members
(being appointed by a wide selectorate on the basis of a precise agenda of prom-
ises made to them), the leader will retain a larger degree of autonomy; further-
more, the leader will be stronger with respect to factions given that the direct
election emphasizes the added value of the leader’s personal charisma. Addition-
ally, the leader needs to reward his supporters to avoid being dismissed by the
mainstream faction (which is co-responsible for his selection).14
In such context, the leader will exploit his dominant position to retain all of
the benefits and propose an unfair deal to the minority. The minority may then
accept or use the voice option.15 If the faction complies, the gain will be zero
(as dissenters are excluded from the allocation of rewards), and the leader’s
payoffs will be equal to µ (outcome U1).16 If F uses voice, the entire party
incurs costs ν due to the public exposure of internal dissent. L can then
propose a new deal. The leader can either reverse the choice, offering a com-
promise (each faction will be rewarded on the basis of its strength), or use the
whip to force the minority to accept the original proposal. A compromise will
give the minority faction F α µ – ν, and retain (1 – α) µ – ν for L (outcome U2).
If L uses the whip option, the game reaches a final stage where the dissenters
either comply or leave the party. If they toe the line (outcome U3), they in
essence get the same payoffs as under the first stage outcome U1 (when F
plays “accept”) reduced by the cost ν, paid for displaying intra-party disagree-
ment to the public (Greene & Haber 2015; McGann 2002; Snyder & Ting
2002. See also Alesina & Cukierman 1990: 847).17 Conversely, if the minority
faction breaks away (outcome U4), F receives a payoff of π – ε, i.e., the benefit
obtained when creating a new party (or joining a different one) minus the cost
of leaving the current party. In the case of a split, L receives all the benefits
less the contribution of the minority faction that leaves the party. The leader
does not incur any cost for party disunity after a split, as the party becomes
more cohesive, but suffers a cost ω for the loss in strength and image caused
by the party breakup. This cost could be higher when new elections approach,
or in highly competitive party systems (Boucek 2010) particularly for ruling
parties.18 The leader’s final payoff will be (1 – α) µ – ω.
A number of results can be shown using backward induction (see Appendix
1). In the final stage, F faces the choice between “exit” and “accept.” For values
of ε larger than π the exit cost is too high if compared to the benefit of a break-
away. Unless the cost of party disunity ν is high as well, the minority has a non-
credible threat to split (i.e., for F the payoffs are higher inside the party as
π – ε < 0 – ν) and will always accept the whip rather than leave the party. This is
the second best outcome for L (who receives all of the payoffs minus the cost ν).
F knows that it would be better off by choosing “accept” at the first stage (due to
ν) and agrees to the unfair deal U1. This pattern resembles that of the “Dictator
Game,” where the responder has little choice but to accept the proposal.
On the other hand, when ε is larger than π but ν is high (so that π – ε > 0 – ν,
though ε > π), F retains a credible weak threat (in case it splits, the faction can
only minimize its loss).19 In this scenario, the outcome could be either U1 or U2
(Agreement). When the cost of party breakup ω is lower than the cost of party
disunity ν for the leader L, party fission will be less damaging than bargaining
for a compromise. As a consequence, F will accept the unfair deal (U1) to avoid
the negative payoff of a breakaway. Conversely, when ω is greater than ν, L will
focus on preserving unity at any cost to avoid the risk of party fission at the last
stage. This induces a compromise and a final outcome U2.
Finally, when ε is relatively small and π is large, F chooses “exit” at the last
stage and threatens to leave to enhance its bargaining power and obtain a larger
share. When ω is sufficiently large, L prefers to negotiate a compromise (U2).
However, if ω is lower than ν, L will seek cohesion even at the expense of party
unity and will refuse a new deal, using the whip against dissenters and generat-
ing a Breakaway outcome (U4).
Table 1.1 summarizes the possible outcomes of the game based on the rela-
tionship between the four parameters: π, representing the payoffs available to the
minority faction after the breakaway; ε, which is the exit cost; ν, is the cost due
to public voice; ω, which expresses the loss due to party fission. Concerning the
minority faction, I distinguish between a non-credible threat (when faction’s
payoffs is greater inside the party), a credible weak threat (by splitting away the
faction can only minimize its loss, compared to the worst option available inside
the party), and credible strong threat (the faction would rather break away
gaining positive payoffs, unless the leader overpays it with a larger amount of
revenues). Concerning the party leader, I distinguish between leaders interested
in pursuing unity (keeping the party together) or cohesion (enhancing the clarity
of the party label and the homogeneity of internal preferences).
Table 1.1 suggests that intra-party politics is a two-sided game and it high-
lights the relative power of the party leader (L) and the minority faction (F ). On
the one hand, some elements influence the faction’s choice. On the other, some
elements affect the leader’s reply. Each factor that alters the ratio between ω and
ν or between π and ε, leads to contrasting outputs increasing or decreasing the
expression of dissent, the fairness in payoffs allocation and the likelihood of
party fission.
Starting from the potential equilibriums discussed so far, and according to the
parameters of the game, I can formulate some general expectations related to the
shape of intra-party dynamics. First, based on this theoretical framework I can
draw some implications on the magnitude of factionalism (Chapter 3) to assess
Table 1.1 Strategies played by the Leader and the Minority Faction and potential out-
comes based on the parameters of the Game
Leader’s Attitude Minority Faction’s Threat to Break Away
Non-credible Credible but Weak Credible and Strong

(ε > π) and (π – ε < 0 – ν) (ε > π) and (π – ε > 0 – ν) (ε < π)
Focus on Unity U1: Unfair deal U2: Agreement U2: Agreement

(ω > ν) (Accept) (Voice; Compromise) (Voice;
Compromise)
Focus on U1: Unfair deal U1: Unfair deal U4: Breakaway
Cohesion (Accept) (Accept) (Voice, Exit; Whip)
(ω < ν)
what elements favor or inhibit the expression of internal dissent (voice). Second,
I can raise hypothesis on intra-party distributive dynamics concerning the alloca-
tion of policy and office payoffs. I will examine alternative patterns of inter-
factional bargaining, analyzing the selection of the party platform (policy
payoffs, Chapter 4) as well as portfolio allocation (office payoffs, Chapter 5). By
doing that, I can distinguish elements that promote fairness and consensual
dynamics (compromise) or boost the discretionality of the party leader, favoring
the compliance of the minority faction (accept). Finally, this theoretical discus-
sion provides insights on the occurrence of party fission (Chapter 6) highlighting
the determinants of factional breakaways (exit).
Per each area of intra-party dynamics (voice, compromise, exit), Table 1.2
reports the expected effect of an increase of each parameter of the game (ω, ν, π, ε),
along with a brief description of the empirical meaning of such parameter. My
theory also predicts different outcomes depending on leader selection mechanisms.
As such, I will raise some expectations concerning the outcomes in parties ruled by
autonomous (AL) or bounded leaders (BL). In addition, I will also focus on the
interaction between leader autonomy and party system competitiveness (AL*ω)
formulating hypotheses accordingly.
In detail, the value ε expresses the exit cost, which increases (for instance)
when the loyalty toward the party and its symbols is very strong, or in the pres-
ence of disproportional electoral rules that exert barriers to the entrance of new
parties in the party system.
The parameter π represents the net benefit available to minority factions
after the split. Due to membership costs (Snyder & Ting 2002), when party
heterogeneity is too large, fringe factions could benefit from a split, which
could increase the policy payoffs of the splinter groups. These factions will no
longer pay such membership cost (which is proportional to the distance from
the bulk of factions); furthermore, after the split the clarity of the party label
could be higher. The parameter π also accounts for the potential expected
increase in office payoffs, which is related to the leverage of each faction:
when the split increases the bargaining power of a faction beyond its size (or
beyond the reward obtained inside the party), the net benefit of its breakaway
raises. However, when factions are already overpaid within their own party
the value of such parameter shrinks.
Table 1.2 Parameters of the Game and expected effect on intra-party dynamics
Label Parameter Description Voice Compromise Exit
ε Exit cost: Loyalty and disproportional electoral systems – – –

π Net benefit after split: Heterogeneity and faction leverage + + +
ω Breakup cost: Party system competitiveness + + –
ν Voice cost: Intra-party democracy and discipline – – +
AL Autonomous leader Ø/– – +
AL*ω Autonomous leader in competitive contexts Ø/– + –
BL Bounded leader + + –
The third element, ω, refers to the cost of a split and can be related to party
system competitiveness: in competitive party systems any single split might alter
the balance of power between the government and the opposition (in the parlia-
mentary arena)20 or between rival parties (in the electoral market) raising the cost
of party breakup.
Finally, ν is the cost of voice that should increase when the intra-party level
of democracy is lower. When public voice is detrimental to the party, the
minority will pay a higher price for expressing dissent (due for instance to whip-
ping and sanctions); conversely the party leadership will suffer for the lack of
cohesion due to the public recognition of internal disputes.
Below I summarize the general expectations concerning factionalism, payoffs
allocation and fission.
Voice – Factionalism, Chapter 3. Elements that increase the exit costs (ε),
such as a strong party loyalty, should deter factions from disclosing their dissent
decreasing the degree of factionalism. In Table 1.1, when ε is larger than π voice
occurs only in one equilibrium out of four while for lower values of ε the
minority should always use the voice as a strategy to enhance its payoffs. In turn,
an increase in π boosts factionalism: the minority tries to exploit its stronger
leverage through the voice option, to bargain a better deal that might improve its
condition. In competitive party systems (ω), where each subgroup might alter
the balance of power between rival parties, the leverage of minorities increases
thereby making the use of voice more rewarding.21 Intuitively, I expect that an
increase in the cost of voice (ν) will reduce its usage (voice occurs in only one
equilibrium out of three when ν is large than ω). Finally, in parties with leaders
bounded by factions one should always observe voice. Factional bargaining in
fact is the method adopted to select the leader and to allocate payoffs. Accord-
ingly, each subgroup should try to affect this process by building its own faction.
Conversely one should observe slight dissent when leaders are autonomous
(voice appears in half of the equilibriums) although such distinction could be
less evident in competitive contexts.
Compromise – Payoffs allocation, Chapters 4 and 5. I consider policy and
office payoffs altogether, assuming that their allocation follows a common
dynamic. When the exit cost (ε) is larger the minority does not have any black-
mailing power on the party leadership. Therefore, I expect less consensual
dynamics (in three out of four equilibriums the final upshot should be to accept
the unfair proposal) and low compromise. For opposite reasons, when π is larger
than ε factions exploit their wide leverage and there could be room for a com-
promise. However, this is true only if party system competitiveness (ω) is strong
too. Competitive settings seem to be the primary determinant of factional
cooperation and promote a compromise between party factions. Accordingly, I
will focus mainly on this aspect in the related empirical chapters. Conversely,
the cost of voice (ν) pushes minorities to accept an unfair deal whenever they
know that the leadership is not willing to address their demands, and they are not
strong enough to be prepared to a breakaway. As a matter of fact, in parties ruled
by bounded leaders the compromise should be the norm: as discussed so far, the
leader is selected after an inter-factional negotiation and has to stick to such deal.
To the contrary, autonomous leaders try to enhance their own interests, reward-
ing their followers. This is no longer true, however, when party unity is an issue
at stake, i.e., in highly competitive political systems when a fission could
strongly harm the party.
Exit – Party fission, Chapter 6. Noticeably, the cost of exit (ε) should affect
the likelihood of a breakaway. Elements that boost existing costs contribute to
preserving party unity. In Table 1.2 there is no exit when ε is larger than π. Vice
versa, larger values of π confirms that the threat of a breakaway is strong and
credible; the minority will pick up the exit option unless the party leader is
prone to heed its requests. This will not happen in low competitive political
systems (e.g., in a dominant party system) where in fact one would expect more
exit, while an increase in competitiveness (ω) should decrease splits. A growing
cost of voice (ν) indicates the lack of room for internal dissent and brings the
leader to refuse any compromise trading unity for cohesion. As a consequence,
one should observe more exit for large values of ν, when the minority faction
retains a credible and strong threat. Finally, compared to their bounded counter-
parts, autonomous leaders should increase the likelihood of splits. However,
when facing competitive conditions, leaders could exploit their greater auto-
nomy to cater to minority’s needs in order to avoid damaging splits, thus pre-
serving party unity.
In the main empirical chapters (3, 4, 5 and 6), these general expectations are
used to develop and test more specific hypotheses, operationalizing the para-
meters of Table 1.2 according to each specific context and taking into account
also the findings of the existing literature. For instance, in Chapter 3, when
dealing with the causes of factionalism, wide attention will be paid to party
organization, statutes and internal rules. In Chapter 4, the analysis of party plat-
form will take into account the role of the electoral cycle as well as studies on
party competition. In Chapter 5, standard theories of portfolio allocation will be
considered and particular attention will be devoted to cabinet stability and
cabinet reshuffles. To conclude, the expectations reported in Table 1.2 will be
retrieved at the end of the book, in Chapter 8, to evaluate whether the theoretical
framework resisted to the empirical test of hypotheses.
Plan of the book

The next chapter presents the textual data that have been analyzed and used
throughout the empirical chapters. It explains the data collection process, with a
focus on the role of factional motions presented during party congress debates,
and briefly describes Wordfish, the technique of automated text analysis used to
estimate the policy positions of factions and subgroups in Italy, France and
Germany by analyzing a total amount of 3,599,740 words in these three lan-
guages. After having discussed the history and the characteristics of intra-party
politics in these three countries, this chapter summarizes the estimates related to
254 factions that competed in 83 congresses of 19 Italian parties (1946–2010),
along with the estimates of 151 party factions, concerning 39 congresses of five
French parties (1971–2016), and those of 33 subgroups belonging to the two
main German parties (1985–2018). It concludes by proving the validity and reli-
ability of the estimates, as well as the robustness of the results of textual analysis
to alternative pre-processing strategies.
Chapter 3 explains the sources of factionalism, focusing on statutes and intra-
party rules as the main explanatory variables. It analyzes the likelihood of
observing a contested party congress and examines what elements increase the
number of factions in Italy and France, or the fragmentation in the leadership
race in Germany. Results show that looming elections or strict internal rules
reduce the display of public disagreement during party congresses, while propor-
tional intra-party rules and internal polarization raise internal fragmentation.
Chapter 4 investigates the process of selection of the party platform in key
parliamentary debates or in electoral manifestos, to evaluate whether and to what
extent factions bind the party leader and limit the possibility of a policy shift
from one election to the following one. Indeed, party positions are linked with
factional preferences and factions bargain over the party platform following a
kind of proportionality rule. Overall, the party works as a “bounded oligarchy”:
the leader needs to consider the stances of all party factions for the sake of
enhancing party unity. Such effect is stronger when new general elections
approach and the need for party unity is heightened. However, in line with the
cartel party theory (Katz & Mair 1995), leaders who are directly selected by a
wide “selectorate” can get rid of factional ties.
While Chapter 4 investigates the allocation of policy payoffs, Chapter 5 con-
siders the other side of the coin, investigating office payoffs. Overall, spoils are
shared in proportion to the strength of each faction, in line with the prediction
of Gamson’s Law (Gamson 1961). Nevertheless, some important deviations
from this path are found. Rules that foster the autonomy of the party leader
grant him more discretion in allocation. The leader will use such discretion to
reward his followers or to ward off any credible and harmful threat to party
unity. In Italy such discretionality is used mainly to promote party unity, while
in France leaders tend to overpay their own faction. Nevertheless, when party
system competitiveness rises, i.e., in the case of looming elections, portfolio
allocation embraces criteria of perfect proportionality. Furthermore, powerful
minorities seem to be rewarded with a more than proportional return when they
are markedly dissatisfied in terms of policy payoffs: strategic portfolio alloca-
tion might balance out a lower amount of policy payoffs and becomes a strategy
to restrain minorities from playing exit, thus contributing to the preservation of
party unity.
Chapter 6 explores the determinants of factional breakaways and comple-
ments the results of Chapters 4 and 5, which have shown how factions bargain in
order to reach a compromise. Sometimes the cooperation fails and factional con-
flicts produce a party fission. I test a number of hypotheses related to my theoret-
ical framework. The results confirm the expectations and show that office, policy
and electoral motives influence the factions’ decision to break away. Other
elements, such as intra-party democracy, the electoral system and party system
competitiveness (in Italy) or the direct election of the party leader (in France),
alter the likelihood of a split by affecting the leaders’ interests in unity or cohe-
sion and therefore his attitudes toward compromising. In addition, based on the
model presented herein, I perform a postdiction to predict recent party fissions
using out-of-sample data related to the People of Freedom Party (PDL), in
2010–2011, and to the PD (in 2015 and 2017).
In turn, Chapter 7 takes into account the outcome of intra-party dynamics dis-
cussed in the previous chapters and evaluates how these affect the survival of the
party leader and the likelihood of a leader replacement. The results of the sur-
vival analysis confirm that the autonomy of the leader, and therefore his ability
to handle intra-party conflicts, depends on the internal mechanisms of leader’s
selection. As a consequence, the magnitude of the intra-party ideological conflict
or the adoption of a “winner-takes-all” style of portfolio allocation have a neg-
ative effect on the duration of the party leader, though only when the leader is
indirectly selected in a smoke-filled room by a small committee, as in this
context the leader will be merely an agent of the party executive with little or no
autonomy. Conversely, the leader’s style of payoffs allocation and the rise of
intra-party ideological conflict are no longer damaging when the internal rules
grant more autonomy to the party leader.
Finally, Chapter 8 summarizes the results, with an eye to the role played by
factions’ preferences and intra-party rules. I will evaluate whether the expecta-
tions derived from my theory have been empirically corroborated by the find-
ings. Additionally, I will build bridges for future researches on party congresses,
intra-party democracy and on intra-party conflict in the age of social media.
To conclude, two appendixes are provided. Appendix 1 describes the solution
of the game–theoretic model presented above, showing, through backward
induction, how stable equilibria can be reached. Appendix 2 reports the acro-
nyms and the full names of parties and factions for which abbreviations have
been used across the book. The datasets, replication materials and additional
supporting information will be made available on the author’s website: https://
andreaceron.com.
Notes
1 www.faz.net/aktuell/politik/inland/cdu-verweigert-neuen-vereinigungen-die-anerkennung-
15747209.html.
2 There are comparative analyses or single case studies explicitly referred to at least 62
countries: Australia, Austria, Argentina, Bangladesh, Belgium, Bolivia, Botswana,
Brazil, Canada, Chile, China, Colombia, Congo (DR), Costa Rica, Czech Republic,
Denmark, Estonia, France, Finland, Germany, Ghana, Greece, Honduras, Hungary,
Iceland, India, Indonesia, Ireland, Israel, Italy, Iraq, Japan, Kenya, Malaysia, Mexico,
Mongolia, the Netherlands, Nigeria, Norway, Pakistan, Peru, Poland, Portugal,
Romania, Russia, Slovakia, South Africa, South Korea, Spain, Sweden, Switzerland,
Taiwan, Tanzania, Thailand, Turkey, Ukraine, United Kingdom, United States,
Uruguay, Venezuela, Zambia and Zimbabwe. Additionally, there are studies related
to “factions” in the European Parliament.
3 For simplicity, I use the masculine wording when referring to party leaders given that,
empirically speaking, almost all the party leaders considered in the book are males.
4 This reasoning is analogous to the principal–agent paradigm: to solve the collective
action dilemma party factions should assign to the party leader (i.e., a political entre-
preneur) a remuneration that coincides with the surplus produced thanks to his contri-
bution (Alchian & Demsetz 1972).
5 Leaders face a kind of “Madison’s Dilemma”: “Agents who are placed in a position
of leadership […] can be expected to exploit it – to use the authority they have been
granted to advance their own interests” (Kiewiet & McCubbins 1991: 47).
6 The rationale is that “the less consistently and intensively involved the participant in
the candidate selection process, the more he or she will be swayed by name recogni-
tion and […] take cues from the highly visible central leadership” (Katz 2001: 291).
7 www.dire.it/18–11–2018/264168-minniti-se-nessuno-al-51-scacco-per-lintero-pd-
vincono-le-correnti/.
8 Examples of small committees are the National Council (e.g., inside the DC and PSI)
or the Central Committee (in communist parties), as well as the Director Committee,
for instance in the French PS; these are usually composed of a around 100 members,
in charge of controlling the leadership. Sometimes the leader is selected by even
smaller committees (usually composed of a dozen members), such as the Executive
Bureau/Committee in France or the Party Secretariat in Italy.
9 In some cases, leaders are directly elected by party members through the so-called
“party primaries.” More often, leaders are directly selected by the whole assembly of
party delegates during the national congress (composed of thousands of party activ-
ists). Notice, however, that the selection made by delegates is often just a formal
ritual: delegates usually ratify a choice that has already been made elsewhere, i.e.,
during the local party congresses, in which delegates are selected according to voters’
support for one or another candidate/faction.
10 Empirical evidence from Japan supports this idea: since 1980 the LDP president (in
charge of allocating office posts) has been elected through factional bargaining; as a
result, the fairness in portfolio allocation increased (Di Virgilio & Kato 2001).
11 The “Exit, Voice, and Loyalty” game has been applied to individual members facing
the choice between remaining in and leaving a party (Hirschman 1970). For specific
applications to party fission see Gehlbach 2006; Kato 1998.
12 This also helps us to understand why it is so crucial for factions (i.e., for minority fac-
tions) to get represented inside the party body (see also the discussion on the intra-
party electoral system in Chapter 3).
13 The greater the value of µ, the greater the sum of a party’s payoffs. However, to sim-
plify the analysis I assume that the amount of payoffs is fixed, so that µ is constant
and equal to 1 (i.e., the total amount of payoffs to be allocated).
14 In fact, the mainstream faction could decide to replace him with another leader who is
able to overpay the mainstream.
15 It could be argued that the game starts with a previous stage in which the leader can
decide whether to make a fair or unfair offer to the minority faction. However, I argue
that this game refers to a peculiar intra-party environment in which L has incentives
to make at least a somewhat unfair offer to F. Conversely, in the case of a fair alloca-
tion, the game would be similar to the context described above, with L bounded by
the wills of all factions. In this scenario, one would expect to observe voice (factions
have to show up to be considered in the deal, even though voice can be confined
behind a party’s closed doors, remaining hidden to the wide public) and compromise.
16 I assigned a zero payoff to the minority group. The experimental economics suggests
that L may offer F a non-zero share of payoffs. This share, however, remains lower
and unfair relative to F ’s strength, and the reasoning holds for any unfair payoff
allocation.
17 Arguably, heterogeneity could also strengthen the party in the electoral arena allow-
ing it to appeal to a broader public (Shepsle 1972; but see Alesina & Cukierman
1990). Nonetheless I assume that, overall, disunity is damaging for the party’s elect-
oral performance (Greene & Haber 2015).
18 This cost could be proportional to the strength of the splinter group (α) assuming that the
breakaway of a tiny faction does not damage the party too much. Note, however, that
even a small faction may be very powerful if it retains the power to make or break gov-
ernments (Boucek 2010), or to help get parliamentary representation (passing the thresh-
old) or not. For instance, in the Italian case, the split of the PSI in 1947 was crucial to
strengthen the role of the DC in the Italian party system. Similarly, the breakaways of
small factions (Democrazia Nazionale and Destra Popolare) from the Italian Social
Movement in 1977 were crucial to sustain the Giulio Andreotti III Cabinet; the split of
Cossuttiani faction from the Communist Refoundation Party (PRC) in 1998 sustained the
survival of the center-left ruling coalition until 2001. Finally, the breakaway of two small
factions from the PRC in 2006 and 2007 contributed to the fiasco of the electoral alliance
The Left–The Rainbow (SA) in the 2008 general election: these splinter groups created
two parties, Critical Left (SC) and the Communist Workers Party (PCL) that won respec-
tively 0.46 percent and 0.57 percent of votes, thwarting SA’s chances to pass the 4
percent threshold. Indeed, SA only won 3.08 percent of votes. To sum up, political
history is filled with examples of tiny splits that produced huge effects on policy-making,
coalition governments or on the structure of the party system.
19 Minimizing the loss could be a rational choice and, accordingly, the threat of split is
credible. Since the cost of breaking away still outweighs the benefit, however, this scen-
ario is weaker than cases where factions gain strictly positive payoffs after fission.
20 This is more damaging for ruling parties as they have much to lose by failing to keep
the support of a majority of MPs and by being voted out of office.
21 Notice that in this case, for tactical reasons, the party might try to hide internal con-
flict in order to present a cohesive front to voters. In this regard, a fictional image of
perfect cohesion might be displayed to the outside world, whereas conflict and
factionalism can grow behind the closed doors of party’s smoke-filled rooms. See
Chapters 3 and 8 for a broader discussion and empirical examples.
References
Alchian, A., & Demsetz, H. (1972). Production, Information Costs, and Economic Organ-
ization. American Economic Review, 62(5), 777–795.
Aldrich, J.H. (1995). Why Parties? The Origin and Transformation of Political Parties in
America. University of Chicago Press, Chicago IL.
Alesina, A., & Cukierman, A. (1990). The Politics of Ambiguity. The Quarterly Journal
of Economics, 105(4), 829–850.
Bäck, H. (2009). Intra-party Politics and Local Coalition Formation. In Giannetti, D.,
& Benoit, K. (eds) Intra-Party Politics and Coalition Government, 53–68,
Routledge, London.
Baron, D. (1998). Comparative Dynamics of Parliamentary Governments. American
Political Science Review, 92(3), 593–609.
Beller, D.C., & Belloni, F.P. (1978). Party and Faction. In: Belloni, D.C., & Beller, F.P.
(eds) Faction Politics, 417–450, ABC-Clio, Santa Barbara CA.
Belloni, F.P., & Beller D.C. (eds) (1978). Faction Politics: Political Parties and Faction-
alism in Comparative Perspectives. ABC-Clio, Santa Barbara CA.
Bernauer, J., & Bräuninger, T. (2009). Intra-party Preference Heterogeneity and Faction
Membership in the 15th German Bundestag: A Computational Text Analysis of Parlia-
mentary Speeches. German Politics, 18(3), 385–402.
Bettcher, K.E. (2005). Factions of Interest in Japan and Italy: The Organizational and
Motivational Dimensions of Factionalism. Party Politics, 11(3), 339–358.
Boucek, F. (2009). Rethinking Factionalism: Typologies, Intra-party Dynamics and Three
Faces of Factionalism. Party Politics, 15(4), 455–485.
Boucek, F. (2010). The Factional Politics of Dominant Parties: Evidence from Britain,
Italy and Japan. In Bogaards, M., & Boucek, F. (eds) (2010). Dominant Political
Parties and Democracy, 117–139, Routledge/ECPR Studies in European Political
Science, London.
Boucek, F. (2012). Factional Politics: How Dominant Parties Implode or Stabilize.
Palgrave Macmillan, Cham.
Budge, I. (1985). Party Factions and Government Reshuffles: A General Hypothesis
Tested against Data from 20 Post-war Democracies. European Journal of Political
Research, 13(3), 327–333.
Budge, I., Ezrow, L., & McDonald, M.D. (2010). Ideology, Party Factionalism and Policy
Change: An Integrated Dynamic Theory. British Journal of Political Science, 40(4),
781–804.
Caillaud, B., & Tirole, J. (2002). Parties and Political Intermediaries. The Quarterly
Journal of Economics, 117(4), 1453–1489.
Carey, J.M. (2007). Competing Principals, Political Institutions, and Party Unity in Legis-
lative Voting. American Journal of Political Science, 51(1), 92–107.
Carey, J.M., & Shugart, M.S. (1995). Incentives to Cultivate a Personal Vote: A Rank
Ordering of Electoral Formulas. Electoral Studies, 14(4), 417–439.
Carroll, R., & Cox, G.W. (2007). The Logic of Gamson’s Law: Pre-election Coalitions
and Portfolio Allocations. American Journal of Political Science, 51(2), 300–313.
Ceron, A. (2012). Bounded Oligarchy: How and When Factions Constrain Leaders in
Party Position-taking. Electoral Studies, 31(4), 689–701.
Ceron, A. (2014). Gamson Rule Not for All. Patterns of Portfolio Allocation among
Italian Party Factions. European Journal of Political Research, 53(1), 180–199.
Ceron, A. (2015a). The Politics of Fission: Analysis of Faction Breakaways among Italian
Parties (1946–2011). British Journal of Political Science, 45(1), 121–139.
Ceron, A. (2015b). Brave Rebels Stay Home. Assessing the Effect of Intraparty Ideological
Heterogeneity and Party Whip on Roll-Call Votes. Party Politics, 21(2), 246–258.
Ceron, A. (2016). Inter-factional Conflicts and Government Formation. Do Party Leaders
Sort Out Ideological Heterogeneity? Party Politics, 22(6), 797–808.
Ceron, A. (2017a). Intra-party Politics in 140 Characters. Party Politics, 23(1), 7–17.
Ceron, A. (2017b). Social Media and Political Accountability: Bridging the Gap between
Citizens and Politicians. Palgrave Macmillan, Cham.
Ceron, A., Curini, L., & Negri, F. (2019). Intra-party Politics and Interest Groups:
Missing Links in Explaining Government Effectiveness. Public Choice, doi: 10.1007/
s11127-019-00644-0.
Ceron, A., & Negri, F. (2016). The “Social Side” of Public Policies: Monitoring Online
Public Opinion and its Mobilization during the Policy Cycle. Policy & Internet, 8(2),
131–147.
Ceron, A., & Negri, F. (2017). Trade Unions and Political Parties in Italy (1946–2014):
Ideological Positions and Critical Junctures. South European Society & Politics, 22(4),
491–508.
Chambers, P. (2008). Factions, Parties, and the Durability of Parliaments, Coalitions and
Cabinets: The Case of Thailand (1979–2001). Party Politics, 14(3), 299–323.
Cox, G.W., & McCubbins, M.D. (1993). Legislative Leviathan: Party Government in the
House. University of California Press, Berkeley CA.
Cox, G.W., & Rosenbluth, F. (1995). Anatomy of a Split: The Liberal Democrats of
Japan. Electoral Studies, 14(4), 355–376.
Cox, G.W., Rosenbluth, F.M., & Thies, M.F. (1999). Electoral Reform and the Fate of
Factions: The Case of Japan’s Liberal Democratic Party. British Journal of Political
Science, 29(1), 33–56.
Debus, M., & Bräuninger, T. (2009). Intra-party Factions and Coalition Bargaining. In
Giannetti, D., & Benoit, K. (eds) Intra-Party Politics and Coalition Government,
121–145, Routledge, London.
Desposato, S.W. (2006). Parties for Rent? Ambition, Ideology and Party Switching in
Brazil’s Chamber of Deputies. American Journal of Political Science, 50(1), 62–80.
Dewan, T., & Squintani, F. (2016). In Defense of Factions. American Journal of Political
Science, 60(4), 860–881.
Dilling, M. (2018). In Defence of Moderate Factionalism. Paper presented at ECPR
General Conference, August 22–25, 2018, Hamburg.
Di Virgilio, A., & Kato, J. (2001). Factionalism, coalitions et fragmentation politique.
Qu’est-ce qui a vraiment changé dans le système partisan au Japon et en Italie dans la
décennie 1990. Revue Française de Science Politique, 51(4), 587–620.
Druckman, J.N. (1996). Party Factionalism and Cabinet Durability. Party Politics, 2(3),
397–407.
Duverger, M. (1954). Political Parties. John Wiley and Sons, New York.
Gamson, W.A. (1961). A Theory of Coalition Formation. American Sociological Review,
26(3), 373–382.
Giannetti, D., & Benoit, K. (eds) (2009). Intra-Party Politics and Coalition Government.
Routledge, London.
Giannetti, D., & Laver, M. (2001). Party System Dynamics and the Making and Breaking
of Italian Governments. Electoral Studies, 20(4), 529–553.
Giannetti, D., & Laver, M. (2009). Party Cohesion, Party Discipline, Party Factions in
Italy. In Giannetti, D., & Benoit, K. (eds) Intra-Party Politics and Coalition Govern-
ment, 146–168, Routledge, London.
Gehlbach, S. (2006). A Formal Model of Exit and Voice. Rationality and Society, 18(4),
395–418.
Golden, M.A., & Chang, E.C.C. (2001). Competitive Corruption: Factional Conflict and
Political Malfeasance in Postwar Italian Christian Democracy. World Politics, 53(4),
588–622.
Greene, Z., & Haber, M. (2015). The Consequences of Appearing Divided: An Analysis
of Party Evaluations and Vote Choice. Electoral Studies, 37(1), 15–27.
Güth, W., Schmittberger, R., & Schwarze, B. (1982). An Experimental Analysis of Ulti-
matum Bargaining. Journal of Economic Behavior & Organization, 3(4), 367–388.
Harmel, R., Heo, U.K., Tan, A., & Janda, K. (1995). Performance, Leadership, Factions
and Party Change: An Empirical Analysis. West European Politics, 18(1), 1–33.
Heller, W.B. (2008). Legislative Organization, Party Factions, and Party Policy Position:
Endogenous Adaptation to Exogenous Institutions. Unpublished manuscript.
Heller, W.B., & Mershon, C. (2008). Dealing in Discipline: Party Switching and Legis-
lative Voting in the Italian Chamber of Deputies, 1988–2000. American Journal of
Political Science, 52(4), 910–924.
Hine, D. (1982). Factionalism in West European Parties: A Framework for Analysis. West
European Politics, 5(1), 36–53.
Hirschman, A.O. (1970). Exit, Voice and Loyalty. Responses to Decline in Firms, Organ-
izations, and States. Harvard University Press, Cambridge MA.
Kam, C. (2009). Party Discipline and Parliamentary Government. Cambridge University
Press, Cambridge.
Kato, J. (1998). When the Party Breaks Up: Exit and Voice among Japanese Legislators.
American Political Science Review, 92(4), 857–870.
Katz, R.S. (1986). Intraparty Preference Voting. In Grofman, B., & Lijphart, A. (eds)
Electoral Laws and Their Political Consequences, 85–103, Agathon Press, New York.
Katz, R.S. (2001). The Problem of Candidate Selection and Models of Party Democracy.
Party Politics, 7(3), 277–296.
Katz, R.S., & Mair, P. (1995). Changing Models of Party Organization and Party Demo-
cracy: The Emergence of the Cartel Party. Party Politics, 1(1), 5–28.
Kenig, O. (2009). Democratization of Party Leadership Selection: Do Wider Selectorates
Produce More Competitive Contests? Electoral Studies, 28(2), 240–247.
Key, V.O. Jr. (1949). Southern Politics in State and Nation. Alfred A. Knopf, New York.
Kiewiet, D.R., & McCubbins, M.D. (1991). The Logic of Delegation: Congressional
Parties and the Appropriations Process. University of Chicago Press, Chicago IL.
Laver, M., & Benoit, K. (2003). The Evolution of Party System between Elections. Amer-
ican Journal of Political Science, 47(2), 215–233.
Laver, M., & Kato, J. (2001). Dynamic Approaches to Government Formation and the
Generic Instability of Decisive Structure in Japan. Electoral Studies, 20(4), 509–527.
Laver, M., & Shepsle, K.A. (1990). Government Coalitions and Intra-Party Politics.
British Journal of Political Science, 20(4), 489–507.
Laver, M., & Shepsle, K.A. (1996). Making and Breaking Governments. Cabinet and
Legislature in Parliamentary Democracies. Cambridge University Press, Cambridge.
Leiserson, M. (1968). Factions and Coalitions in One-Party Japan: An Interpretation
Based on the Theory of Games. American Political Science Review, 62(3), 770–787.
Levy, G. (2004). A Model of Political Parties. Journal of Economic Theory, 115(2),
250–277.
Maor, M. (1997). Political Parties and Party Systems. Comparative Approaches and the
British Experience. Routledge, London and New York.
Marsh, M. (1993). Introduction: Selecting the Party Leader. European Journal of
Political Research, 24(3), 229–231.
Marx, P., & Schumacher, G. (2013). Will to Power? Intra-party Conflict in Social Demo-
cratic Parties and the Choice for Neoliberal Policies in Germany, the Netherlands and
Spain (1980–2010). European Political Science Review, 5(1), 151–173.
McGann, A.J. (2002). The Advantages of Ideological Cohesion: A Model of Constitu-
ency Representation and Electoral Competition in Multi-Party Democracies. Journal of
Theoretical Politics, 14(1), 37–70.
Mershon, C. (2001). Contending Models of Portfolio Allocation and Office Payoffs to
Party Factions: Italy 1963–79. American Journal of Political Science, 45(2), 277–293.
Meyer, T.M. (2012). Dropping the Unitary Actor Assumption: The Impact of Intra-Party
Delegation on Coalition Governance. Journal of Theoretical Politics, 24(4), 485–506.
Michels, R. (1915). Political Parties: A Sociological Study of the Oligarchical Tend-
encies of Modern Democracy. Trans. P. Eden & P. Cedar, Free Press, New York.
Müller, W.C. & Strøm, K. (eds) (1999). Policy, Office, or Votes? How Political Parties in
Western Europe Make Hard Choices. Cambridge University Press, Cambridge.
Nicholas, R.W. (1965). Factions: A Comparative Analysis. In Banton, M. (ed.) Political
System and the Distribution of Power, 21–60, Tavistock Publications, London.
Nowak, M.A., Page, K.M., & Sigmund, K. (2000). Fairness versus Reason in the Ulti-
matum Game. Science, 289(5485), 1773–1775.
Olson, M. (1965). The Logic of Collective Action. Public Goods and the Theory of
Groups. Harvard University Press, Cambridge MA.
Ono, Y. (2012). Portfolio Allocation as Leadership Strategy: Bargaining among and
within Parties. American Journal of Political Science, 56(3), 553–567.
Panebianco, A. (1988). Political Parties. Organization and Power. Cambridge Press
University, Cambridge.
Persico, N., Pueblita, J.C.R., & Silverman, D. (2011). Factions and Political Competition.
Journal of Political Economy, 119(2), 242–288.
Poguntke, T., & Webb, P. (eds) (2005). The Presidentialization of Politics: A Study in
Comparative Politics. Oxford University Press, Oxford.
Rahat, G., & Hazan, R.Y. (2001). Candidate Selection Methods: An Analytical Frame-
work. Party Politics, 7(3), 297–322.
Rahat, G., Hazan, R.Y., & Katz, R.S. (2008). Democracy and Political Parties. On the
Uneasy Relationship between Participation, Competition and Representation. Party
Politics, 14(6), 663–683.
Reed, S.R., & Scheiner, E. (2003). Electoral Incentives and Policy Preferences: Mixed
Motives behind Party Defections in Japan. British Journal of Political Science, 33(3),
469–490.
Reiter, H.L. (1981). Intra-party Cleavages in the United States Today. Western Political
Quarterly, 34(2), 287–300.
Roemer, J.E. (2001). Political Competition. Harvard University Press, Cambridge MA.
Rose, R. (1964). Parties, Factions and Tendencies in Britain. Political Studies, 12(1), 33–46.
Sartori, G. (1971). Proporzionalismo, frazionismo e crisi dei partiti. Rivista Italiana di
Scienza Politica, 1(3), 629–655.
Sartori, G. (1976). Parties and Party System. A Framework for Analysis. Cambridge
University Press, New York.
Seyd, P. (1972). Factionalism within the Conservative Party: The Monday Club. Govern-
ment and Opposition, 7(4), 464–487.
Shepsle, K.A. (1972). The Strategy of Ambiguity: Uncertainty and Electoral Competi-
tion. American Political Science Review, 66(2), 555–568.
Snyder, J.M., & Ting, M.M. (2002). An Informational Rationale for Political Parties.
American Journal of Political Science, 46(1), 90–110.
Spirling, A., & Quinn, K.M. (2010). Identifying Intra-party Voting Blocs in the UK House of
Commons. Journal of the American Statistical Association, 105(490): 447–457.
Strøm, K. (1990). A Behavioral Theory of Competitive Political Parties. American
Journal of Political Science, 34(2), 565–598.
Strøm, K. (1994). The Presthus Debacle: Intra-Party Politics and Bargaining Failure in
Norway, American Political Science Review, 88(1), 112–127.
Taggart, P. (1998). A Touchstone of Dissent: Euroscepticism in Contemporary Western
European Party Systems. European Journal of Political Research, 33(3), 363–388.
Zariski, R. (1960). Party Factions and Comparative Politics: Some Preliminary Observa-
tions. Midwest Journal of Political Science, 4(1), 27–51.
Zincone, G. (1972). Accesso autonomo alle risorse: le determinanti del frazionismo.
Rivista Italiana di Scienza Politica, 2(1), 139–160.
Zuckerman, A.S. (1979). The Politics of Faction: Christian Democratic Rule in Italy.
Yale University Press, New Haven.
2 Factional preferences in Italy,
France and Germany
Factional politics: the Italian case

The life of Italian parties has always been characterized by a high level of
internal factionalism. As an example, inside the DC several subgroups were
active, including Dorotei, Nuove Cronache, Base, Forze Nuove and many others;
Sinistra Socialista and Autonomia coexisted and fought to prevail inside the PSI;
Presenza Liberale and Libertà Nuova animated the internal debates of the Italian
Liberal Party (PLI). Overall, in post-war Italy one could have found a number of
factions labeled as Rinnovamento, Riscossa, Iniziativa, Democrazia, Impegno
and Unità. These names, followed by an adjective (socialist, liberal, republican)
to identify their party affiliation, were equally spread across parties.
Factions have long been crucial for Italian politics. They were steadily men-
tioned in the everyday political chronicles and played a role in the legislative
arena. Many bill proposals were rejected due to factional vetoes. Many govern-
ments resigned due to intra-party disagreements and the fate of others heavily
depended on the outcome of party congresses (Giannetti 2010; Mershon 1994).
Factions were indeed able to promote or to end governments and political for-
mulas; it is not by chance that the formation of the first center-left government
coalition in the 1960s was postponed until DC’s and PSI’s factions that sup-
ported such an alliance became strong enough to overcome the resistance of
internal minorities. Indeed, the center-left Fanfani IV cabinet was appointed with
the decisive support of the PSI (February 1962) only after the Aldo Moro, leader
of a left-of-center DC faction, won the party congress (January 1962) gathering
support on a congress motion favorable to this center-left alliance.
Beside the DC, the PSI was factionalized as well. Since its foundation in the
nineteenth century the party has been composed of several subgroups, and its
far-left factions (communists and maximalists) were challenged by more
moderate reformists groups. In view of that, this party was repeatedly weakened
by the breakaways of moderate and radical wings. During the First Italian
Republic (1946–1994), almost all the relevant parties (including the
social–democrats, the liberals, the republicans and even the post-fascist Italian
Social Movement, MSI) were composed of organized factions (Lombardo 1972)
and were continuously affected by factional quarrels and compromises.
Factional preferences 29
Only the Italian Communist Party (PCI) was apparently cohesive and non-
factionalized. This was probably the consequence of strict and undemocratic
internal rules. The party was organized according to the principles of “demo-
cratic centralism,” combining strong party discipline in its external behavior
with (theoretically) wide freedom of thought in closed-door internal debates.
Notwithstanding this, at the end of the 1960s some intra-party groups formed,
though party rules did not allow them to stabilize and organize as stable and
recognized factions. Only one faction, called Manifesto, tried to create a stable
factional organization but it was expelled from the party in 1969; others factions,
i.e., the right-wing Miglioristi and the left-wing Cossuttiani, survived as latent
entities.
After 1994, a set of institutional changes led to the Italian Second Republic,
with a renewed and fluid party system characterized by huge party instability. In
such context factionalism, although softened, seemed to persist. All the main
actors that characterized the Second Republic are internally divided. Some of
them are openly factionalized while in others internal subgroups are more latent:
this happened, for instance, inside personal parties (Calise 2010) such as Forza
Italia (FI), the Northern League (LN) or Italy of the Values (IDV).
Inside the far-left Communist Refoundation Party (PRC) several Trotskyists
factions fought against the party mainstream. The right-wing National Alliance
(AN) was divided into three groups: a liberal–democratic faction, close to the
party leader Gianfranco Fini (Nuova Alleanza), was challenged by a group loyal
to the leader of the center-right coalition, Silvio Berlusconi (Destra Protagoni-
sta), and by a third far-right faction more sympathetic with pro-State positions
on the economic dimension (Destra Sociale). The center-left DS has been
affected, since its birth, by factional struggles promoted by an internal left (Cor-
rentone) and a right-wing group (Liberal); the mainstream faction, in turn, was
divided between Walter Veltroni’s partisans (Veltroniani) and the supporters of
Massimo D’Alema (Dalemiani).
Both the AN and the Democrats of the Left (DS) are no longer alive as they
merged with other groups to form new parties. However, their factions survived
and brought their pre-existing factional conflicts inside these new parties. The
DS merged with The Daisy (DL) into the PD. Two left-wing DS factions that
opposed this merger decided to break away. Despite this, the new party was
highly factionalized and the conflict between Veltroniani and Dalemiani con-
tinued inside it. The first PD leader, Veltroni, appointed in 2007 by a wide
margin and with the support of Dalemiani too, has been under attack by
D’Alema lieutenants since the beginning and was forced to resign quite soon. In
2009 the new party leader, Pierluigi Bersani, belonging to the Dalemiani faction
Riformisti E Democratici (Reformists and Democrats, RED), defeated the
incumbent Dario Franceschini (supported by Veltroni). After the congress,
Franceschini’s faction Area Democratica re-approached Bersani in order to
enhance party unity. Disappointed by this shift, Veltroni rallied his followers and
built a new organized liberal–democratic faction, Movimento Democratico
(Democratic Movement, MoDem), that overtly opposed the mainstream group.
30 Factional preferences
Right-wing factions managed to take revenge in the following party congress
(2013), when the moderate Matteo Renzi was appointed as the new party leader
(Ceron 2017).
Something similar happened to the AN. In the 2002 congress, factional con-
flict worsened and the party leader Fini changed the rules to preserve party unity.
Fini decreased the level of democracy (with a formal ban on factional grouping),
though at the same time he built intra-party routines based on compromise (e.g.,
all the factions retained ministers in the Berlusconi II Cabinet). Later on, a few
months before the fusion involving the AN and FI (which created the PDL in
2008–2009), a far-right splinter group coming from Destra Sociale (D-Destra)
broke away from the AN. In turn, the conflict between Fini’s partisans (Fare
Futuro/Finiani) and Berlusconi’s fans (Berluscones) that characterized the
internal life of the AN continued inside the PDL.
Inside the new party, however, reaching a compromise was harder. The agree-
ment set during the fusion assigned 70 percent of all spoils (in the party body
and public office) to former FI members; former AN factions had to divide the
remaining 30 percent and Fini’s followers only retained around 10 percent of all
posts.1 Contrary to FI, the AN was a heavily organized party, able to mobilize
members and activists. For this reason, Finiani subgroups were able to gain
more than 10 percent of approval among PDL members.2 As a consequence, Fini
demanded for a new congress in which each faction could have been weighted
according to its share of members’ votes. This was also a strategy to neutralize
the overriding power of Berlusconi, who was ruling the party alone. This fac-
tional strife was openly unveiled during a party meeting in April 2010, when
Fini claimed for more consensual intra-party dynamics. However, the main-
stream’s reply was to explicitly forbid any factionalism. The conflict worsened
in July when it became clear that public dissent was no longer accepted by the
leadership (as if the party was ruled through democratic centralism). The execu-
tive party branch lowered the degree of internal democracy stating that the leader
of the internal minority Fini was unfit to be a PDL member, given that his public
statements on several policy issues (justice, immigration, devolution, social
policy and civil rights) were considered to lay far away from the PDL platform
and outside the boundaries of its ideology. The executive party branch also
imposed sanctions on three members of Finiani. Such sanctions immediately
triggered the exit of the Finiani faction from the party.
This summary of factional struggles confirms that Italy can be considered a
benchmark to study the politics of faction (Zuckerman 1979), and the politics of
fission (Ceron 2015a), in terms of factional cooperation and conflict.
In this regard, the Italian case is particularly promising: the huge number of
relevant parties (around ten in each Legislature) and the high number of factions
(around three in each factionalized party) allow a wide number of observations
to be collected and a large-N statistical analysis to be performed.
Furthermore, the striking cabinet instability rate (60 governments in 64 years)
allows us to track reshaping in alliances, party policy positions and allocation of
office payoffs almost year by year; as such, the outcome of party congresses can
be put in relation with coalition bargaining: to give an example, the lapse of time
between one-party congress and the subsequent process of negotiation over
cabinet formation is very narrow, i.e., seven months on average. Analogously,
the wide number of general elections (17) held between 1946 and 2008 under
four different electoral systems provides a lot of information about electoral
payoffs and allows to test the effect of changes in the electoral laws.
On top of that, the Italian case can be considered as a kind of political labora-
tory suitable to generate and test hypotheses (Giannetti & Grofman 2011) thanks
to its multiple adjustments in electoral rules. In more detail, the electoral system
was a pure open list proportional representation (PR) from 1946 to 1992, except
in 1953 when a majority prize (two-thirds of seats) was theoretically granted to
the coalition able to gain a majority of votes. Elections were held under a mixed
system (75 percent plurality and 25 percent closed list PR) from 1994 to 2001.
Finally, in 2006, 2008 and 2013 the electoral system was a closed list PR with
thresholds and a majority prize for the coalition that won a plurality of votes;
this system, however, was replaced before the 2018 election by a new mixed-PR
system.
Interestingly, the Italian case is particularly suitable also due to the avail-
ability of a large amount of documents related to the internal life of Italian
parties and specifically to party factions (Ceron 2012). This element is crucial as
these data can be exploited to fill one of the major gaps in the literature on party
politics, i.e., the lack of information on the policy preferences of factions.
Although several authors pointed to the importance of collecting data about
factions’ preferences (Budge et al. 2010; Giannetti & Benoit 2009; Laver &
Benoit 2003), intra-party politics is often a sort of “invisible politics” that takes
place in the shadow (Sartori 1976). Consequently, finding information can be a
difficult task.
The present book contributes to filling this gap. Here I will present a dataset
related to the Italian case that covers almost 70 years of history, providing
information on intra-party rules and on the size and position of factions from
1946 to 2010. Overall, 254 motions presented by factions in 83 congresses of 18
parties will been analyzed.
Beside Italy, however, the book will also present data on intra-party politics
in France and Germany.
Beyond Italy: intra-party politics in France and Germany

For a long time, Italy has been considered as a deviant case in the realm of polit-
ical science, due to its extreme levels of cabinet instability and to the comprom-
ises required by oversized coalition governments, to its excessive political
fragmentation (with a large number of relevant parliamentary parties and fre-
quent reshuffles of the party system, particularly during the Second Republic),
and to the huge amount of factionalism, intra-party divisions and party splits or
parliamentary switching. Actually, these features have never been a peculiarity
of the Italian political system only. In fact, other countries shared similar traits.
Just to provide some examples, in Belgium and Finland one can find similar
levels of cabinet instability; in Denmark and the Netherlands strong party system
fragmentation and frequent party fissions (Ibenskas 2017) can be noticed; in
France too one can observe instability of the party system, with frequent changes
of party labels often linked with splits and mergers, as well as party switching,
which is also quite common in Greece and Switzerland (Volpi 2019).
More interestingly, recent real-world political events have shown that some
elements that were considered as peculiar to Italy can be now extended and
attributed to almost all the other European democracies.
Let us think, for instance, to the recent episodes of government instability or
to the long-lasting post-electoral negotiations involving rival parties in countries
such as Spain, Greece, Germany or the United Kingdom, which in the past were
considered as fairly stable or not used to coalition governments and political
compromises.
Analogously, the instability of the party system widens across Europe as con-
firmed by the recent splits in Spain and Greece, as well as by the rise of new
parties in many other countries. Along this vein, the topic of factionalism gained
relevance and intra-party divisions publicly emerged in Germany and France
(see below) but also in the United Kingdom, especially (but not solely) with
respect to the debate on “Brexit” (see the discussion in Chapter 1).
Precisely because Italy is not an outlier in terms of factionalism and it is not
the only country in which intra-party politics affects the political system, the
present book will compare the internal dynamics of political parties in two addi-
tional countries, i.e., France and Germany.
So far, I argued that Italy looks like a political laboratory (Giannetti &
Grofman 2011) and, due to its multiple institutional reforms, it represents a
quasi-experimental context that allows us to perform a comparison between very
different institutional settings. While holding constant other country-level fea-
tures, I can in fact assess whether my hypotheses hold both in the First Italian
Republic, characterized by a proportional electoral system and a relatively stable
party system (features that are similar, for instance, to the German case), and in
the Second Italian Republic, when the electoral system also includes majoritar-
ian elements and the party system becomes much more fluid (traits shared by the
French context too).
From this perspective, it is immediately clear why France and Germany can
be profitably compared with the Italian case. Nevertheless, other features typical
of these two countries provide additional reasons to include them in the analysis.
Let us start from France. With respect to factional politics, the French case
seems particularly suitable and it is perfectly comparable with the Italian one.
Indeed, intra-party politics in these two countries takes a very similar shape for a
number of reasons, including the fact that many parties in both countries hold con-
tested congresses to select the leadership and to set the party line, with several
intra-party factions that compete against each other by signing policy documents,
i.e., party congress motions or similar documents (sometimes shorter texts, such as
the profession de foi written by candidates running for the party leadership).
In this regard, as far as intra-party politics is concerned, the case of the Parti
Socialiste (French Socialist Party, PS) is particularly intriguing as its factional
structure and the shape of internal dynamics closely match those of many Italian
parties. The PS has been highly factionalized since its birth (Clift & McDaniel
2017; Ceron & Greene 2019). The ancestor of the PS, the French Section of the
Workers’ International (SFIO), was composed of competitive factions since 1907,
when the party leader Jean Jaurès institutionalized internal factionalism, establish-
ing that the party body had to be selected through PR of all factions, according to
the votes received by each factional motion. Although officially forbidden, factions
punctuated the history of the PS even in more recent years, starting from the
Epinay congress (1971). The party faced factional conflict and splits, as well as
attempts to cooperate producing syntheses (sincere or insincere) between the diver-
gent stances of all factions. To sum up, factions and factional leaders (ranging from
Jean Jaurès and Jules Guesde, to François Mitterrand and Michel Rocard, and
including leaders of internal minorities such as Jean-Pierre Chevènement or Jean-
Luc Mélenchon) have played a crucial role in the history of the PS.
Indeed, from 1971 until 2018, with the only exception of four congresses in
which there was no vote on congress motions and another three in which perfect
unity can be observed (a unique motion was presented and supported by all sub-
groups jointly), in the remaining 18 congresses (72 percent of times) intra-party
competition involved a minimum of two and a maximum of seven factions that
presented alternative congress motions.
Historically, the most long-lasting division inside the PS was the one between
the supporters of Mitterrand (mainstream) and the left-wing faction CÉRÉS (Centre
d’études, de recherches et d’éducation socialiste (Center for Socialist Studies,
Research and Education)), headed by Chevènement; however, at times, one noticed
either splits inside the mainstream or alliances between the mainstream and CÉRÉS
to form a majority. Furthermore, other factions closely tied to a prominent PS
leader played a role: there were small groups like the Mollétistes (partisans of Guy
Mollet) or the Poperénistes (Jean Poperen), as well as larger factions including
Rocardiens (Rocard), Fabiusiens (Laurent Fabius), or the supporters of Pierre
Mauroy. In recent years, the main cleavage involved the moderate supporters of
François Hollande, advocating more liberal–democratic positions, and the left-
wing faction Gauche Socialiste (affected by internal splits and mergers based on
the willingness to reach or not a compromise with the mainstream of the party).
Overall, the majority faction has often tried to synthesize different positions
to enlarge its support beyond the actual share of votes won in the party con-
gresses. Despite impressive episodes of internal conflict, as in 1990 and 2005
when highly divisive congresses were held, the PS party leaders have usually
attempted to foster unity (Ceron & Greene 2019). Even in the aftermath of the
extraordinary intra-party strife in 1990, rival factions agreed on ruling the party
jointly to restore unity.
The PS is not the only factionalized party though. On the left side of the
political spectrum, the congresses of the Green Party (Les Verts) were usually
contested by multiple factions that presented competing motions and there are
subgroups even inside the Parti Communiste Français (French Communist
Party, PCF ), particularly after the abolishment of democratic centralism when
public dissent became allowed (a pattern that is very similar to the case of the
PCI in Italy). The internal life of the Bayrou’s MoDem looks instead more
similar to that of Italian “personal parties,” with a more unitary environment in
which some divisions emerged when the congresses of the MoDem forefather
party (the Union for French Democracy, UDF ) decided to break the alliance with
the Gaullists and to adopt the new party name.
In turn, the Gaullists were internally factionalized too. Factions played a role
both inside the Union pour un Mouvement Populaire (Union for a Popular
Movement, PCF ) and in the ancestor party, Rally for the Republic (RPR). Active
since 1962, under different labels, the RPR officially recognized internal factions
only in 1989; internal disputes involved the mainstream faction (headed by
Jacques Chirac), some groups of liberal, moderate or reformer partisans (usually
pro-Europe, such as the Balladuriens/Conservateurs libéraux, followers of
Édouard Balladur), and a right-wing sovereignist Eurosceptic faction that gath-
ered the followers of Philippe Séguin (the Séguinistes/Gaullistes sociaux) and
Charles Pasqua (Gaullistes orthodoxe). As far as European matters are con-
cerned, in the 1990s the RPR split twice. Due to factional divisions the party did
not take a position on the 1992 referendum concerning the Maastricht Treaty:
Chirac and other party leaders campaigned in support of the Treaty, while
Séguin and Pasqua opposed it. Later, two Gaullists’ lists ran in the 1999 Euro-
pean election when the sovereignist movement led by Villiers and Pasqua (13.06
percent) won more votes than the official RPR list (12.82 percent) and broke
away after the election.
Factional conflicts continued inside the UMP, where the mainstream group
(headed by Nicolas Sarkozy and the Sarkozystes) was opposed by right-wing
factions, such as Debout la République (Nicolas Dupont-Aignan), that broke
away in 2007, or Droite Libre (Rachid Kaci). Internal divisions blew up even
more in 2012 when factionalism was officially authorized and, for the first time,
six organized factions contested the congress presenting competing motions.
This competition involved, among others, the right-wing La Droite Populaire,
the liberal–democratic France Moderne et Humaniste, as well as the “anti-
faction” faction La Boîte à Idées. These six subgroups, however, were also inter-
nally divided. In 2012 the leadership selection process generated an extremely
tied race between two candidates, Jean-François Copé and François Fillon, and
produced an impressive public fight (as intense as in the 1990 PS congress). The
first official results, showing a tiny margin of 98 votes in favor of Copé (50.03
percent versus 49.97 percent), were in fact rejected by Fillon who claimed to
have won the election (50.01 percent). When the party commission confirmed
the success of Copé (50.28 percent), Fillon immediately created a parliamentary
party group of UMP dissidents. Finally, three weeks later, the two candidates
reached an agreement to rule the party jointly and restore internal unity. Never-
theless, the divisions between the supporters of different UMP leaders, such as
Fillon, Copé, Sarkozy, Alain Juppé or Bruno Le Maire kept playing a role in the
internal life of the Gaullist Party.
Remarkably, as confirmed by this short summary of French factional politics,
France and Italy are also similar with respect to additional intra-party dynamics.
Concerning party splits and party fissions, in the post-war period, among Euro-
pean countries, Italy is the country with the highest number of splits, while
France ranks second, with a comparable number of party fissions that occurred
over time (one can find, respectively, no less than 17 splits in France and 23
main splits in Italy, along with additional minor events; see Ibenskas 2017).
Even when looking at party switching in the parliamentary arena, one notices
that Italy is the European country with the highest number of parliamentary
switches (Volpi 2019), followed by France that reaches almost the same levels
of switching: from 1946 to 2013 one can find an average of 7.88 percent
switches in the Italian Parliament and 7.58 percent in France. As such, intra-
party divisions are – at least in magnitude – very similar in these two countries
and produce wide effects on the party system as a whole.
Finally, in these two countries there is also a similar trend toward the direct
election of party leaders, often by means of (open) primary elections, particularly
in recent years. Accordingly, these changes in intra-party rules (see Chapter 3)
can be profitably put in relation to factional conflicts.
To sum up, for the French case I collected and analyzed 154 textual docu-
ments (congress motions or similar documents, such as the already mentioned
profession de foi) that allow the estimation of the policy positions of 151 party
factions related to 39 party congresses held by five French parties in 45 years,
between 1971 and 2016.
Turning to Germany, the selection of this case refers to other criteria. Fac-
tionalism and intra-party politics here are somehow different from the French
and Italian cases and the same applies to internal rules. To start with, parties tend
to present themselves as monolithic blocs and, apparently, they are less affected
by splits. Contrary to what has just been discussed with respect to Italy and
France, in Germany the average rate of switchers is six times lower and amounts
to 1.25 percent in the post-war period; analogously one can only account for
two-party fissions in the same amount of time.
Internal divisions tend to be hidden, especially during party congresses.
Factionalism manifests itself elsewhere, for instance, within the parliamentary
party group where the strength of each faction is measured according to the
number of MPs that decide to affiliate with one or another subgroup.
Accordingly, during the party congress there is no open competition between
rival factions to get a majority around a congress motion that outlines the party’s
ideological line. Such different dynamics usually end up in unitary congresses in
which the party leader is elected by overwhelming majorities if not by unan-
imity. For instance, when focusing on the two main German parties, the CDU
and the Social–Democratic Party of Germany (SPD), one notices that on average
party leaders won 90.6 percent of support. In the 41 congresses held by these
parties in the last 35 years, between 1983 and 2018, the party leader got a vote
share lower than 80 percent only five times. Remarkably this occurred three
times in the most recent years, i.e., in 2015 and 2018, for the SPD, and in 2018
for the CDU.
In the SPD, Sigmar Gabriel (2015) and Andrea Nahles (2018) were selected
respectively by the 74 percent and the 66.35 percent of delegates (the latter
represents the second worse results ever for a SPD leader).
In turn, the CDU approached its 2018 party congress showing a spike of intra-
party divisions (see the discussion in Chapters 5 and 6). The leadership of
Angela Merkel faced resolute internal disagreements. For the first time since
1973, the CDU/CSU parliamentary party group rejected the candidate proposed
by the CDU leader as the new party whip (i.e., the chief of party’s back-
benchers). By a narrow margin (52.7 percent), rebel MPs selected an alternative
candidate that pledged to be more independent from Merkel’s viewpoints and to
renew the party. This division is even more troublesome given that such a wide
number of rebel MPs (125) would be sufficient to defy the majority retained by
the government coalition between the CDU and the SPD.
Even more interestingly, for the first time in recent years, inside the CDU a
group of party activists has been building a truly organized party faction to chal-
lenge the moderate turn imposed by the leadership of Merkel and to restore a
more conservative ideology inside the party, changing the line accordingly. As a
consequence, in December 2018, the CDU held the first contested leadership
race since 1971. The outcome was extremely tied: the pro-Merkel candidate
Annegret Kramp-Karrenbauer was appointed with a very narrow margin (51.7
percent), raising doubts on the future unity of the party.
While, from many perspectives, the German case looks like a different scen-
ario, suitable to evaluate whether the theory presented in the book holds even in
a radically different intra-party setting, these recent events seem to suggest that
intra-party politics in Germany is changing a lot. Exactly like German politics
and the German party systems have been changing in the last decades, with an
increasing fragmentation, instability and more complicated post-electoral pro-
cesses of coalition building, one can expect a rise in intra-party conflicts and, in
a near future, it could be possible to observe stronger similarities between the
structure of factional politics in Germany compared to what already happens in
Italy and France.
So far, however, the German case still represents a further stringent test suit-
able to evaluate the implications of my theory in a rather different scenario.
Taking advantage of this, I investigate the logic of intra-party dynamics in more
depth, evaluating what happens to parties in which the internal conflict takes
other and different shapes rather than factionalism in party congresses. Accord-
ingly, I analyze the German case focusing on non-factional intra-party organiza-
tions or on particular factional subgroups called kreis (circles) or wings.
While in Germany middle-sized parties such as the liberals or the left are also
internally divided, to streamline the analysis I will focus on the two main
German parties, i.e., the CDU and SPD. These two parties differ from each other
in terms of the shape of internal subgroups. Inside the SPD, in fact, there exist
(formal or informal) organized factions. Although these factions do not really (or
not always) compete during the party congress, they are actually long-lasting,
stable and well-structured intra-party groups with their own ideology and their
own organization. This includes membership fees, hierarchies, meetings, maga-
zines, merchandising and updated websites containing information and docu-
ments drafted to express the viewpoints of the faction itself. These documents
are usually called position papers and contain ideas and contributions written to
affect the policy manifesto of the SPD.
These organized factions are really able to influence the direction of the party
(Debus & Bräuninger 2009). In this regard, factionalism inside the SPD repli-
cates a pattern that was already observed elsewhere. Its historical roots date back
to the 1970s. At that time, intra-party politics took the shape of a fierce internal
fight between ideologically rival factions and two main wings were created. In
1966 young left-oriented SPD members gave birth to the Frankfurter Kreis
(Frankfurt Circle, FK), which was particularly influential between 1982 and
1998. Conversely, moderate members retaining more liberal–democratic views
in 1974 coalesced into the right-wing Seeheimer Kreis (Seeheim Circle, SK),
which was particularly strong in the 1970s and after 1998 (Debus & Bräuninger
2009).
Nowadays this division still exists, though the intra-party complexity inside the
SPD has widened. On the left side, the Demokratische Linke 21 (Forum of the
Democratic Left 21, DL21)is the heir of the FK. Founded in 2000 to advocate a left
turn of the SPD, it is so structured that it has been defined as “a party inside the
party.”3 This faction was also headed by the current SPD leader Andrea Nahles. In
spite of her left-wing origins, in the 2018 congress Nahles was challenged by a
rival candidate, Simone Lange, who held more left-wing positions and after the
congress launched a new cross-party movement (Aufstehen/Stand Up) that aimed
to strengthen the left-wing stances in German politics.
Inside the SPD, the Parlamentarische Linke (Parliamentary Left of the SPD
parliamentary group, PL) is another left-wing faction; active since 1980 as the
heir of the Leverkusen Kreis (created in 1972), this group supports a Keynesian
approach to economic policies. The Parlamentarische Linke is the strongest
faction in the SPD parliamentary party group; in the new Bundestag elected in
2017 it represents 76 MPs out of 153 (49.67 percent) and its size was similar in
the previous parliament when the PL was still the first subgroup, supported by
86 MPs out of 193 (44.55 percent).
On the right side of the party, the conservative Seeheimer Kreis is still active
and in the previous Bundestag it was the second largest group with 70 MPs (36.26
percent) affiliated to it. For a long time, it could have been considered as the main-
stream SPD group, given that many prominent politicians and many SPD leaders
including Martin Schultz, Sigmar Gabriel, Thomas Oppermann, Peer Steinbrück
and Frank-Walter Steinmeier were considered close to the SK. This faction also
played a crucial role in supporting Gerhard Schröder’s Agenda 2010 welfare
reform. This reform was also supported by another moderate faction, the Netzwerk
Berlin (Berlin Network , NB). Created in 1999 by young party members, this
faction retains pragmatic and reformist positions. At the beginning factional affili-
ation to the NB was allowed on a generational basis only. Later this group
expanded its membership and it now accounts for around 25 percent of the parlia-
mentary group, though some of its members (including Gabriel, Steinbrück,
Steinmeier and Oppermann) are cross-affiliated to the SK faction.
While the internal life of the SPD is quite similar to that of French and Italian
parties, things are different inside the CDU. Until recent years, there were no
organized factions in this party (see the discussion in Chapter 1). Internal associ-
ations, however, exist and can play a role. The most important groups are the
lobbies of employees and employers (Debus & Bräuninger 2009).4 Founded in
1945, the Christlich-Demokratische Arbeitnehmerschaft (Christian Democratic
Employees, CDA) represents the interests of Christian–democratic workers and
labor groups; relying on the Catholic Social Doctrine, the CDA supports welfare
state issues and traditional family values. Conversely, the Mittelstandsvereini-
gung (Middle Class and Business Association, MIT) promotes the interests of
the middle class, as well as those of small and middle-sized companies or free-
lancers. Active since 1956, it is stronger and more influential than the CDA
(Debus & Bräuninger 2009). Although these two groups are not proper factions,
they still retain contrasting positions. Accordingly, they present their own posi-
tion papers and policy programs (Grundsatzprogramm) as well as enacting deci-
sions and resolutions (Beschluss) approved during the national conferences held
by the CDA and MIT. As such, I will exploit these documents to get some
insights on the different viewpoints existing inside the CDU and to provide a
measure of internal polarization over time.
Overall, by analyzing intra-party documents released by factions or intra-
party associations, I have been able to estimate the policy position of five sub-
groups at 33 different points in time; 20 of these estimates refer to subgroups
inside the CDU, while 13 are SPD factions. These data have been used to map,
year by year, the internal structure of these two parties over the last 33 years
(from 1985 to 2018).
Party congress, the arena where factional bargaining

takes place
To assess intra-party equilibriums and dynamics I need to gather data about the
factional structure of parties. In particular, I need to know the number, the
strength and, most importantly, the preferences of each faction. There are
different sources of information on factionalism; leadership strength can be
evaluated by looking at cabinets (based on the number and type of portfolios
received by each faction) or legislatures (based on parliamentary memberships,
i.e., the number of MPs affiliated to each faction), even though the most suitable
indicator is probably “the voting strength of individual factions on party
decisional bodies such as party congresses (using motion votes as indicators of
factional strength) or on party executives (using individual factions’ seat shares)”
Boucek (2009: 467).
Analogously, scholars can determine policy positions of subgroups looking at
a party body or analyzing their behavior in the legislative arena. Although some
authors showed that it is possible to estimate factional preferences through roll-
call votes (Rosenthal & Voeten 2004; Spirling & Quinn 2010), it has been
argued that “content analysis on texts drafted by […] intra-party groups seem to
be the best choice to identify their respective preferences, in particular if the
research question deals with changes of the positions of political actors over
time.” (Bäck et al. 2016; Benoit et al. 2009: 443). In a congress motion each
faction is (almost) completely free to present its view about the party platform;
in this context the discipline (Laver 1999) only plays a limited role in bounding
the expression of factions’ preferences so that texts should be less influenced,
compared to roll-call votes (Spirling & McLean 2007). In addition, not all votes
are roll-call votes, thus they do not provide a complete picture of voting behav-
ior. The decision to call a roll-call vote may also be strategic and biased accord-
ing to the issues or the group that request it (Carrubba et al. 2006; Giannetti &
Benoit 2009). Roll-call analysis is strictly related to voting behavior, which pro-
vides actors with a limited set of options: “yes,” “no” or “abstention,”5 while
there could be a variety of preferences behind a similar voting behavior: actors
may say “yes” or “no” for very different (and sometimes opposite) reasons. Text
analysis helps scholars to discover and highlight the actual positions of political
actors by providing a broader set of information about their preferences. For all
these reasons I will analyze factions’ positions through text analysis focusing on
party congresses.6
Scholars suggest referring to party congresses as a source of data (Boucek
2009; Giannetti & Benoit 2009; Giannetti & Laver 2009; Mershon 2001). Indeed
“Many gains in understanding intra-party politics may derive from analysis of
internal party organization, such as formal and informal factions” (Giannetti &
Benoit 2009: 235) in particular by examining “debates at party congresses
between different party factions” (Giannetti & Benoit 2009: 234). This is par-
ticularly true for Italy and France, but the party congress outcomes and the
debates held therein are also crucial to investigate intra-party politics in other
countries, including the UK and Germany (Greene & Haber 2016, 2017).
Internal debates can be more suitable if compared to parliamentary speeches, as
these can be subjected to party discipline and “may not reflect the true distribu-
tion of preferences” (Proksch & Slapin 2012: 520).
The party congress represents the key moment in a party’s internal life, like
general elections are the crucial starting point for representative democracy.
Indeed, “For factions, the closest approximation to that sort of competitive arena
is internal party elections, since it is there that factions organize teams of candid-
ates and appeal to people enjoying the right to vote for one team or another”
(Mershon 2001: 561). During the party congress those subgroups willing to
organize as a faction can present a list of candidates to the party body. Often this
list is attached to a “motion,” i.e., an omnibus policy document that aims to
shape party strategy and ideology. Motions set out factions’ “opposing views on
the ideological direction of the party” (Giannetti & Laver 2009: 154). After a
public debate, party congress’ delegates vote on the policy cornerstones and
establish the new party line that the leadership should pursue. They appoint
members in the party executive and the party body (committees in charge of
running the party); elect (directly or indirectly) the party leader; adjust and ratify
the party statute; set the new party line and strategy casting a vote on a motion
that commits the leadership to some general policy statements until the next con-
gress. In fact, “party delegates vote on the policy principals in an annual confer-
ence whereas the balance of power between faction, that is, the proportion of
votes that each faction receives, is translated into policy recommendations”
(Levy 2004: 251). During the competition for gaining members’ support, the
party splits in many subgroups. The party congress, however, is not only a
source of division. Instead, this is also the arena in which the party tries to reach
its unity through a compromise between internal factions. If the bargaining is not
successful, the party can more likely incur a breakup. Indeed, many fissions take
place during the congress or soon after.
During the party congress, politicians who want to distinguish themselves from
the party mainstream can exploit such visibility toward party members and voters
in order to show that they retain different views and policy positions. For these
reasons party congresses can be used systematically to map the factional structure
of parties (Giannetti & Laver 2009). Data drawn from congresses provide us with
objective criteria to analyze intra-party divisions. Information that flows from offi-
cial sources is the only way to overcome scholars’ personal beliefs about the actual
number of factions inside parties. In addition, these data often provide a measure of
the strength of each faction, based on the number of congress votes gained by each
list or the number of members elected in the party body (when data on members’
votes are lacking). Finally, party congress debates, based on textual documents
produced by the party leadership or by rival factions, also allow assessment (by
means of a content analysis) of the policy preferences of different subgroups; in
particular, recent developments in the field of quantitative text analysis can be
exploited to estimate factions’ ideal points by analyzing congress’ motions.
For these methodological reasons, and given the need to identify intra-party
subgroups according to objective criteria, in the Italian case (and in France too) I
consider as a faction each party subgroup that contests the congress presenting
its own list of candidates attached to a policy motion (Giannetti & Laver 2009).
Accordingly, I will be dealing with divided parties, excluding from the analysis
political actors such as the “personal parties,” as in these parties the leader exerts
a strong influence on members and activists (e.g., Raniolo 2006). Even when
these parties are internally polarized, such divisions might not evolve into organ-
ized factions given that those parties are, by definition, “parties without factions”
(Lombardo 1972). Indeed, they held congresses very rarely and these meetings
are almost never contested (being merely conventions devoted to celebrate
internal unity and to show the charisma of the party leadership).
Similarly, unitary congresses, i.e., congresses in which there is no competi-
tion between factions and the delegates approve a common unitary motion, are
excluded from the analysis. Sometimes minority factions might refuse to present
their motion and their list for various reasons; they might agree with the main-
stream or they do not feel strong enough to defy the leadership. Indeed, the
game–theoretic model proved that the existence of internal dissenters is not a
sufficient condition for the emergence of voice (i.e., minority motions). This
does not happen only in the presence of strong and autonomous leaders but also
when the party is ruled through inter-factional compromise.
Such agreements in fact could be reached either after the congress (if this
happens, one would observe internal conflict) or before. In the latter case, I
would not be able to measure intra-party division and conflict (even if it exists)
because disagreements have been already solved: the unitary congress motion
(as well as the division of party body’s spoils) expresses a common equilibrium
on which all subgroups agreed.7
Concerning France, the data collection process has been quite similar to Italy,
even though, to maximize the number of observations, I relaxed some of the cri-
teria discussed so far (see below for details). Conversely, in the German case, as
I already noticed, things are slightly different and I do not revert to party con-
gresses as a source of data on the positions and the strength of rival factions.
Nevertheless, the outcome of German party congresses will still be taken into
account. In the German case I resort to party congresses not as a source of data
to catch the main independent variables (heterogeneity, polarization and policy
positions) but as an arena that produces political outcomes; textual data coming
from external sources will be put in relation with such congressional outcomes.
This will be done, in particular, with respect to the leadership’s contest and dura-
tion and to manifesto documents drafted by the party.
Data collection on Italy, France and Germany

For the Italian case, I gathered motions from the official proceedings related to
national congresses published by parties, and from parties’ official newspapers
or magazines.8 Motions related to the most recent congresses have been down-
loaded from the official parties’ websites.9 I tried to collect data on parties’ fac-
tional structures analyzing the whole population of party congresses held by
Italian parties (approximately 300). However, some of them (roughly 50 percent)
should be excluded for the methodological reasons discussed above. Unfortu-
nately, due to a lack of data sources I was not able to find information on the
whole population of contested congresses. After the collapse of the Italian party
system in 1992–1994, many of the existing parties disappeared and their archives
disappeared as well; moreover, issues of some parties’ newspapers are not
always available. I have been able to find information on approximately 50
percent of the total number of contested congresses. Overall I gathered 254
motions related to 83 congresses of 18 parties (see Appendix 2 for details on the
estimated policy positions). Almost 200 of them (nested in 64 congresses)
concern the First Republic while the remaining data (19 congresses with 56
motions) belong to the Second Republic. On average the number of motions per
congress is three. Table 2.1 shows more details about the motions gathered.
Table 2.1 Italian data: list of parties, congresses and motions included in the analysis
Party Total Contested Analyzed % on Total % on Contested Missing Motions (Total) Motions (Avg.)
AN 3 1 1 33 100 0 4 4
DC 18 13 11 61 85 2 41 3.73
DS 4 4 4 100 100 0 12 3
FV 18 16 2 11 13 14 6 3
MSI 18 7 5 28 71 2 20 4
NPSI 6 2 2 33 100 0 4 2
PCI 16 3 3 19 100 0 8 2.67
PD 3 3 1 33 33 2 3 3
PDA 2 2 1 50 50 1 3 3
PDCI 5 1 1 20 100 0 2 2
PLI 18 11 11 61 100 0 35 3.18
PRC 8 6 6 75 100 0 20 3.33
PRI 22 15 11 50 73 4 25 2.27
PSDI 24 21 9 38 43 12 25 2.77
PSI 24 12 12 50 100 0 38 3.17
PSIUP 4 1 1 25 100 0 3 3
PSOC 2 1 1 50 100 0 3 3
UDC 3 1 1 33 100 0 2 2
Total 198 120 83 42 69 37 254 3.06
For several parties (AN, PD, Action Party (PDA), Party of Italian Commu-
nists (PDCI), Socialist Party (PSOC), PSIUP, Union of Christian and Center
Democrats (UDC)) I found data concerning only one congress. Conversely the
maximum number of congresses analyzed pertains to the PSI (12 congresses)
and to the DC, PLI and PRI (with 11 congresses each). I found also nine PSDI
congresses and six related to the PRC.10
The present dataset contains more information on some parties than on others:
there are 38 motions nested in 12 PSI congresses but only two motions presented
in the unique contested congress held by the PDCI and UDC. This feature
however does not affect the analysis (see for instance Chapter 6). Column 7 in
Table 2.1 provides an estimate of the number of missing congresses, considering
only the contested (or presumably contested) ones. Among the parties included
in the analysis there are 37 missing cases whereas the (estimated) number of
contested congresses amounts to 120. The percentage of missing cases, then, is
around 30 percent.
With a few exceptions, the dataset covers the entire set of contested con-
gresses for a large number of parties. There are multiple missing congresses only
for two parties: PSDI and the Greens. The PSDI merged its archive with the PSI
(after 1966) but some data get lost; the Greens hold a National Assembly
approximately once a year and their internal life is particularly tangled so that it
was not possible to gather data on several old congresses. Interestingly, I
managed to account for the large part of internal debates within the DC, PLI,
Italian Republican Party (PRI) and PSI and I collected almost all the data on the
DS and PRC.
For the French case, I relied on online sources only. Several motions have
been downloaded directly from the official parties’ websites, sometimes these
documents were retrieved using the Web Archive and sometimes they were
indexed on Wikipedia. In turn, Wikipedia or the website France-Politique
(www.france-politique.fr/) also provided additional information on the outcome
of the congresses. Beside these sources, for the PS, the Socialists Archives
(www.archives-socialistes.fr/) were particularly useful as they provide a full
picture of the PS congresses from 1939 to 2005;11 they report speeches, motions
and any other information related to the congresses, referring directly to the offi-
cial party’s newspaper.
Overall 154 textual documents have been gathered to estimate the policy
positions of 151 party factions in 39 party congresses held between 1971 and
2016 (see Appendix 2 for details on the estimated policy positions). However,
with the exception of PS, data on other parties are related mostly to years
after 2000.
In one case (Fillon, 2012 UMP congress) two different documents have been
used in the analysis (the short profession de foi and a longer text outlining his
policy project). Furthermore, two textual documents related to the splits of the
PS have been included to anchor the results of the analysis: a statement released
by Marc Dolez and Jean-Luc Mélenchon that announced their choice to break
away from the PS and a policy document of the splinter group La Gauche
Moderne (2007), the movement created by Jean-Marie Bockel after his split.
Notice that, to maximize the number of observations, in the 2006 congress of the
UDF–MoDem, the speech of the leader of the dissidents has been used as a
proxy for the position of the internal minority.
Analogously, to maximize the number of cases in the dataset, I also included in
the analysis the unitary motions that have been presented in a few party congresses
(in detail, the 1984 congress of the Greens, the 2000 congress of the PCF, the 2008
and 2010 congresses of the MoDem, and the PS congresses held in 1976, 1981,
1987 and 1991), with the support of all intra-party factions and subgroups. By
doing that, and considering these documents as the factional compromise reached
before the congress, I managed to increase the number of observations comparing
such inter-factional agreement with the outcomes observed in terms of party policy
positions, payoffs allocation, party splits or leadership duration. For similar
reasons, information on contested party congresses has been considered, when
easily available, even if rival factions did not present competing motions or if these
motions were missing. As such, the total number of rows in the final dataset
increases up to 166 and this allows us to perform a more coherent and complete
test of hypotheses, especially with respect to the internal life of the PS.
I managed to gather motions related to 20 out of 24 PS congresses (83.3
percent) held between 1971 and 2015; I lack data on policy positions only for
the period 1992–1994, when the PS held three congresses in four years. I also
grasped the entire internal life of the UMP (analyzing all the four congresses
held in 2002, 2004, 2012 and 2014 and including also documents related to the
2016 primary election) and the MoDem (I have data on three congresses, from
2006 to 2010; though the first one is the last congress of the UDF, in which the
creation of the MoDem was approved). Unfortunately, I have no data on ancestor
Gaullist or centrist parties (notice that the Gaullist parties officially recognized
internal factions only in 1989 so the number of missing contested congresses is
very low). Conversely, there is only scant information on the internal life of the
PCF and the Greens. The PCF started to hold some contested congresses after
1994, when the rule of democratic centralism was abolished (see Chapter 3);
therefore, the number of missing contested congresses related to it is almost null.
For the German case, first and foremost I relied on documents collected and
analyzed by Debus and Bräuninger (2009) that have been kindly provided by the
authors. Furthermore, these documents were integrated with updated policy
papers issued in recent years by the main intra-party subgroup such as the CDA,
MIT, SK, DL21 or NB. These documents were retrieved from the official web-
sites of SPD factions and CDU associations.12 In some cases, multiple docu-
ments have been used.
While the book focuses only on CDU and SPD, in the textual analysis I also
included documents related to the factions of the Liberal Democratic Party
(FDP) or electoral manifestos of the Party of Democratic Socialism (PDS) that
are helpful to anchor the analysis (as in Debus & Bräuninger 2009), as well as
documents of associations of employees and employers within the SPD, though
these will not be considered in the empirical analyses.
Overall, 62 texts have been analyzed to estimate policy positions of 33 sub-
groups (see Appendix 2 for details on the estimated policy positions). These
estimates have been used to detect changes in factions’ positions and polariza-
tion inside the CDU and the SPD over time, from 1985 to 2018. Data have been
imputed in order to produce yearly estimates; the past estimate has been used
when there was no updated information on the policy preference of an internal
subgroup.
In the next section I illustrate how all these textual documents, concerning
Italy, France and Germany, have been analyzed in order to get factions’ policy
positions.
From words to fact(ion)s: estimating factions positions using

Wordfish
To extract factions’ policy positions from congress motions or policy papers I
need to apply some sort of text/content analysis technique. In political science,
one of the oldest methods of analysis consists in the hand-coding of textual
documents, following strategies similar to that of the Comparative Manifesto
Project (CMP). The CMP analyzes electoral manifestos presented by parties
through manual coding (Budge et al. 2001). For each manifesto, human coders
divide the text into a number of quasi-sentences and assign each of them to one
of the 56 pre-determined categories, according to a classification scheme. This
technique can be time consuming, particularly when the amount of texts is quite
large. Given the high number of texts gathered (almost 500), and their wide
length (overall, they contained 3,599,740 words) hand-coding strategies would
not be suitable.
To cope with such concerns, scholars started to develop semi-automated and
automated techniques of text analysis. Among the others it is worth mentioning
Wordscores (Laver et al. 2003) and Wordfish (Slapin & Proksch 2008; Proksch
& Slapin 2009). They provide quicker (but still reliable) means to determine the
ideal points of political actors. Accordingly, I will analyze data through Word-
fish, for the reasons explained below.
Wordfish is an automated scaling technique that produces estimates of policy
positions comparing several textual documents. This method was immediately
applied to different text sources (manifestos, speeches, statements and pledges)
measuring policy positions of parties and interest groups in several countries
(e.g., Klüver 2009; Proksch & Slapin 2009, 2010) and it is now largely used by
political science scholars.
Wordfish assumes that the words’ usage is informative of the policy position
of each textual document relative to others.13 It further assumes that, for each word,
the probability of appearing in a document is independent of the presence of other
words, therefore it considers each text as a vector of randomly distributed words.14
Looking at the relative frequencies of words contained in each text, Wordfish
allows the differences between alternative documents to be determined, scaling
them on a common latent dimension thereby measuring their relative distance and
policy position. Following previous studies on text analysis, Wordfish assumes that
words follow a Poisson distribution, which is suitable to match the skewed distri-
bution of words in texts. The Poisson is defined by one parameter only as the mean
and the variance of the distribution coincide. This allows the analysis to be simpli-
fied, decreasing the computational effort. The distribution then could be summa-
rized as follows: yijt ~ Poisson(λijt), where yijt measures how often the word j appears
in the document related to party (faction) i at time t. In the Wordfish analysis four
parameters are estimated: the document position on the latent scale, at time t, along
with its fixed effect (which accounts for the different length of the texts), and the
discrimination parameter (β) of each word, as well as its fixed effect (which is the
logged mean count of each word across all the documents). This latter parameter
allows for the higher or lower usage of certain words across different texts to be
controlled. Common words not associated with any political meaning (i.e., articles,
prepositions, etc.) appear frequently but are uniformly spread across documents.
These words retain a strong fixed effect while their discriminating power will be
close to zero. As such, the fixed effect ensures that rare words are properly taken
into account. Words that appear only in one document should carry a theoretically
infinite discriminating power as they would be sufficient to identify that document;
to deal with this, Wordfish decreases the β values of very rare words.
The two main parameters of interests, however, are the policy positions of
textual documents (with mean 0 and standard deviation equal to 1) and the
words’ discrimination parameter.
Starting from textual documents of factions and subgroups, Wordfish will
then produce the estimates of factions’ policy positions. They are scaled on a
single latent dimension whose meaning has to be interpreted ex-post according
to the political connotation of the texts submitted to the analysis.
For instance, if documents are related to a single issue (e.g., environmental
policy) this topic will define the latent policy continuum. To the contrary, when
documents outlining the policy views of a political actor on a variety of issues
are considered (as for party manifestos or faction motions) the latent dimension
should be interpreted as a more general left–right scale.15
The other main parameter, β, is the word discrimination parameter. It measures
how much each word is helpful in classifying documents; it also corresponds to the
word’s “policy position” along the latent scale, and allows whether the word stands
on the left or on the right of the continuum to be evaluated. Words that appear with
a high frequency across all documents have a discriminating power close to zero
while words that appear only in a few documents are highly helpful to distinguish
between documents that are on opposite sides of the scale. These words will retain
a higher absolute value of β and will be located on the extremes of the latent scale.
β values are important also because they allow a first diagnostic of the results to be
performed, by comparing the estimated policy positions of words to evaluate
whether these match their actual usage in real-world political language.
Wordfish is particularly suitable for the present analysis. It allows with large
amounts of textual data to be dealt with and produces replicable estimates of
documents positions (Klüver 2009; Slapin & Proksch 2008).
It does not require a reference text to anchor the documents and does not
require the building of dictionaries that define the political meaning of words; in
fact, it analyzes all the words contained in each document (without the need to
remove some of them) assigning a substantial political meaning to each word
(i.e., the discrimination parameter) without any ex-ante judgment.
Wordfish is also suitable to analyze time series, assuming that word usage is
constant over time (things can be different if language changes over time, see
below). Documents related to the same political actor at different points in time
are considered to be independent with respect to each other so that the position
of one document at time t does not affect its placement at time t + 1. Its position
over time will be similar only if its word usage does not change. Conversely, if
the content varies over time, the estimated position of that political actor will
change too.
However, Wordfish can require some pre-processing steps in order to improve
the quality of the estimates. Such pre-processing strategies can involve the
choice to stem words, to remove very common or peculiar terms (“stopwords”)
as well as rare ones (reducing the sparsity of the documents–terms matrix) or to
consider only words that are used repeatedly over time. These quite arbitrary
decisions are ultimately left to the researchers (Greene et al. 2016), even though
such choices can somehow affect the estimates. This aspect will be largely dis-
cussed and addressed at the end of this chapter.
Policy positions of party factions in Italy

Wordfish requires the assumption that word usage has remained relatively con-
stant over time. Italian data though range from 1946 to 2010; textual documents
span from the end of the Second World War, through the Cold War, the Protests
of 1968, the fall of Berlin Wall and the collapse of communism, until Tangen-
topoli and the advent of the Italian Second Republic.
For this reason I want to control for any possible modification in word
usage. To make sure that the research fits the requirement of language
stability I split documents into two datasets, choosing 1989 as the break point.
Far from producing the “end of history” (Fukuyama 1992), the downfall of
communism certainly produced changes that affect several issues, such as
foreign policy (due to modification in the structure of international system) or
economic policy (the Soviet Marxist left extreme of the state–market policy
scale vanished).
These changes have reasonably altered the political meaning of some words;
this effect could be even more consistent in the Italian case, as the PCI was quite
strong in the electoral and parliamentary arena and the debate on communism
was a relevant issue of the political agenda.
Indeed, center-right parties (the DC, PLI and MSI) often emphasized their
anti-communist position. This strategy though became useless at the end of the
1980s; for instance, during the MSI congresses factions debated the need to
define the party as something else other than a mere barrier against communism.
The Fall of the Berlin Wall was the first factor that contributed to altering the
Italian party system at the beginning of the 1990s, in particular with respect to
the renewal of the PCI (Giannetti & Mulé 2006). Many scholars argued that the
Italian transition began the day after the fall of the Berlin Wall: as long as Italian
politics was dominated by the cleavage of communism/anticommunism, the end
of such international strife led to adjustments in party competition and redefined
the players of such competition (Bull & Rhodes 1997; Pasquino 2000).
It prompted the transformation of the PCI into the Democratic Party of the
Left (PDS), and favored the electoral success of the LN in 1992, as moderate
voters were no longer “forced” to support classic anti-communist parties such as
the DC and its allies (Pasquino 2000).
The Italian political language and the dynamics of political communication
changed indeed, leading to the rise of new styles of language and communica-
tion. Some relevant changes concerning the lexicon, tone and content of mes-
sages occurred (Ceron 2015c); luckily, until 2010, in intra-party debates
ideological concepts such as “left” and “right” were still relevant (becoming
even more frequent after 1989), and there was neither a strong decline in the use
of traditional ideological terms nor a wider usage of populist words (Ceron
2015c). The language, in sum, remained largely ideological and suitable to
perform Wordfish analyses. Nevertheless, to account for some stylistic changes,
I retain the year 1989 as the breaking point.
I performed two separate Wordfish analyses. In both analyses the documents
cover the whole spectrum of the left–right continuum. In the first period
(1946–1988) I analyze motions of factions belonging to the left-wing PSIUP as
well as documents related to congresses held by the far-right post-fascist MSI.
Similarly, for the second period (1989–2010) data include texts belonging to the
far-left PRC, born from a fusion between the left-wing heirs of the PSIUP and
the left-wing heirs of the PCI, and motions of the AN (heir of the MSI). Due to
the lower number of center-right texts available in the second period, I included
in the analysis few party manifestos of center-right parties.16
Overall, the textual documents included in the analysis contain 1,718,412
words, with a number of unique words equal to 33,629 for the first period and to
35,391 for the second.
Figure 2.1 plots the policy positions of all the Italian factions. The latent
dimension can be interpreted as a left–right scale. Different symbols for each
party family are used. Post-fascist party factions (MSI and AN; denoted by a
square) are located on the far-right. The liberal–democratic family (PLI; hollow
square) is on the center-right. Christian democrats (DC and UDC; plus) are
located in the center during the first period, and move to the center-right in the
second period. Social–democratic and non-Marxist left party factions (PSDI,
PDA, PRI, DS, PD, New Italian Socialist Party (NPSI) and the Greens; triangle)
appear on the center-left in the first period but shift toward the center in the
1980s. Finally, socialists and communists (PCI, PSI, PSIUP, PRC and PDCI;
hollow triangle) are located on the left-wing. On the whole, party factions
belonging to different party families are properly located and there are few
Figure 2.1 Factions’ position on the left–right scale (Italy).

Note
First and second time period displayed together; factions clustered by party family.
overlaps: for instance, socialist and social–democratic factions are almost never
to the right of Christian–democratic or liberal–democratic factions.
Figure 2.2 confirms this idea; it shows the weighted mean of all the factions’
positions in each party congress. For simplicity, only the labels of the main
parties are displayed; left-of-center parties are displayed in black, right-of-center
parties are in gray. The picture confirms that the DC holds centrist positions, the
liberals and the post-fascists are on the center-right, communists, socialists and
social–democrats stand mainly on the center-left (with the far-left PSIUP being
the most extreme-left party before 1989). In the second period, the communists
PRC and PDCI are on the extreme-left; the AN and UDC (partners of Berlusco-
ni’s center-right coalition) are on the right; the DS, the PD and the Greens lay
somehow in the middle between these two extremes. The PD seems very close
to center-right parties, but this seems coherent with the recent history of this
party, which formed government coalition with the PDL and with splinter groups
of center-right parties (additionally, in the recent legislatures there have been
many switches from PD to centrist/center-right parties and vice versa).
There are some changes in the position of parties over time. The DC is closer
to the center-left in the 1960s, when the center-left formula was launched and
the PSI formed coalition governments with the DC. Overall, after the 1970s
and more deeply during the 1980s, many parties (particularly the PSI, PSDI, PLI
Figure 2.2 Weighted positions of Italian party factions in each party congress.
Note
First and second time period displayed together; main parties only.
and PRI) are affected by a rightward drift. This shift, however, is consistent with
the changes in their economic position after the 1973 oil crisis, and with the
growing need for economic policies able to drop inflation and cut government
spending and public debt.
Interestingly, the moderate shift of the PSI over time is consistent with the
expectations. The PSI moves to the left after 1947, when moderate factions exit
from the party; later, it starts to converge toward the center after 1959, when the
moderate faction Autonomia won the congress – this drift is even strengthened in
the 1970s (due to the breakaway of left-wing factions) and after the 1978 and 1981
congresses, when the reformist faction headed by Bettino Craxi took control of the
party (Ceron & Negri 2017). Analogously, the moderation of the heirs of the com-
munist party (PCI–PDS–DS–PD) is in line with the scholarly literature and the
empirical evidence (Ceron & Negri 2017; Giannetti & Mulé 2006).
Italian case: validity and reliability of the estimates

Let us turn to discuss the validity and the reliability of these estimates. Both in
the first and in the second period the length of the texts and the number of unique
words considered was large enough to obtain accurate estimates. On average, the
length of each text is equal to 5,403 words. There is, however, a wide variation,
ranging from a minimum of 70 words (one motion in the 1946 PDA congress),
to a maximum of 32,849 (PRC mainstream in 2002). Roughly 50 percent of
documents contain more than 3,400 words while only 37 (14 percent, mostly
related to PRI congresses) are below the threshold of 1,000 words.
To evaluate whether the Wordfish analysis caught the actual differences in the
use of political language across parties, let us take a look at the values of
the discrimination parameter ( β ). Figure 2.3 displays the β parameters of each
word and highlights, as an example, the label of some words related to the first and
second period. In the first period words such as “god,” “motherland” and “family”
(dio, patria and famiglia), which are traditionally associated with conservative
values, appear on the right side of the scale. On the left there are words like “class”
(classe), “solidarity” (solidarietà), “nationalization” (nazionalizzazione).17 In the
second period “no-global,” “collective bargaining” (concertazione)18 and “peace”
(pace) stands on the left, while “meritocratic” (meritocratici), “soldiers” (soldati)
and “devolution” (devoluzione) are on the right.
On both sides of the scale there are words that refer to different issues (e.g.,
economy, civil rights and foreign policy). For instance, concerns about “unem-
ployment” (disoccupazione) and “redistribution” (redistribuzione) are attributed
to left-wing parties, while “disinflation” (disinﬂazione) and “privatization”
Figure 2.3 Discriminating power of words in Italy.

Note
Beta coefficients of words’ estimates; first and second period displayed together.
(privatizzazioni) stand on the right. Concerning civil rights, the left emphasizes
“laity” (laicità) and “homosexuality” (omosessualità), while the right criticizes
the “Islamization” (islamizzazione) and emphasizes “birth-rate” (natalità) and
crime (criminalità). The same applies to other issues, such as foreign policy:
“peace” and “disarmament” (disarmo) are on the left, “gulag” and “soldiers” are
on the right. These examples confirm that β values are coherent with the actual
meaning of words in the Italian political language, therefore they appear on the
proper side of the left–right scale.
Let us move to discuss the estimates of factions’ positions to see if left-wing
and right-wing factions are properly located inside each party. To do that, per
each party congress, I aligned factions from left to right according to the liter-
ature (e.g., Giannetti & Laver 2009; Mershon 2001) and newspaper reports.
These expected ordinal positions have been compared with the observed rank of
factions according to Wordfish: the polychoric correlation is strong and signi-
ficant (0.68).19 Furthermore, there are clear differences in the policy preferences
of rival factions: in the 84 percent of cases (70 congresses out of 83), at least two
factions have positions that are statistically different from each other at the 90
percent level of confidence.
Factions seem to be properly located on the whole left–right scale too. There
is a strong and positive correlation between the estimated position of the median
faction, in each party congress, and several external measures of party positions
drawn from expert surveys (r = 0.8) or CMP data (r = 0.7). In both time periods
the mean position of factions belonging to the same party is coherent with the
expected placement of that party.
Finally, to assess the reliability, I hand-coded a subset of motions to compare
them with the estimates of Wordfish analysis. I selected a subsample of 49
motions, 38 related to DC factions, three to the PSI, two to the DS and six to the
PD.20 These motions have been analyzed by means of hand-coding, following
the coding scheme provided by Ceron et al. (2019). Wordfish results are com-
pared with the estimates of a left–right ratio scale related to economic issues; the
correlation is strong and positive (0.59), suggesting that Wordfish estimates are
reliable. Figure 2.4 plots these estimates together with a fitted regression line
(with 95 percent confidence interval).
Policy positions of party factions in France

For the French case a single Wordfish analysis that includes all the textual docu-
ments has been performed. Words have been stemmed, even though this choice
does not affect the results (see below). Overall, the documents included in the
French analysis contain 1,331,186 words, with a number of 39,231 unique words
(22,033 after stemming).
The policy positions of French factions are displayed in Figure 2.5. On the
left, there are communists (mean: –0.34; standard deviation: 0.65) and socialist
party factions (mean: –0.51; standard deviation: 1.04) as well as the Greens,
whose position matches that of the PS both in the 1980s and after 2000, though
Figure 2.4 Reliability of Wordfish estimates compared to hand-coding.

Note
Comparison based on a sample of 49 DC, DS, PD and PSI motions; the black line indicates the linear
fit with 95% confidence interval.
they look overall moderate due to unbalanced data (mean: 0.53; standard deviation:
0.50). On the right, there are UMP factions (mean: 1.02; standard deviation: 0.35)
as well as documents related to the MoDem Party (mean: 0.80; standard deviation:
0.13). Overall, the position of the PS remains quite constant until 1990. Later, the
PS shifts toward the center-right, especially after 2000, when the party started to be
led by Hollande and its liberal–democratic faction. The PS looks rather cohesive in
the 1970s and in the 1980s, when it held some unitary congresses and even the most
relevant left-wing minority, CÉRÉS, was recurrently involved in the coalition of
factions in charge of ruling the party. This unity ends up in 1990 and it is not by
chance that party splits involving the PS have been quite frequent since 1990.
Remarkably, when looking at the data, one can observe a huge internal disagree-
ment concerning the PS in 1990, when a highly conflictual party congress was held,
exacerbating internal divisions. Analogously, the estimates display a wide disagree-
ment also in 2005, when the PS held another very divisive congress (Ceron &
Greene 2019). Interestingly, when excluding unitary congresses, the correlation
between such measure of polarization inside the PS and alternative sources of data
based on the PS congress speeches (Greene & Haber 2016) is positive, as expected,
though not markedly strong (0.32). Notwithstanding this, the alignment of factions
Figure 2.5 Factions’ position on the left–right scale (France).
seems in line with the expectations. Inside the PS the left-wing CÉRÉS always
stands to the left of the mainstream faction Mitterrandistes; the average position of
CÉRÉS (–1.42) is lower (more left-oriented) compared to the mainstream faction
(–1.29) and this gap is statistically significant at the 95 percent level of confidence.
Analogously, inside the PCF the position of the mainstream (0.08) is statistically
more moderate compared to left-wing minority factions (–0.38). It is also worth
noticing that one of the most right-wing factions inside the PS is La Gauche
Moderne. This faction, headed by Bockel, split (2005) to create a movement that
promoted liberal–democratic views and merged inside the Gaullist UMP; interest-
ingly, La Gauche Moderne (0.93) is closer to the position of the PS mainstream
faction in 2015 (0.97), which was mostly composed of members that left the PS to
support Emmanuel Macron and his centrist movement La République En Marche.
Indeed, the faction promoted by Manuel Valls, who represented the right-wing of
the PS in the 2017 primary elections and later joined LREM, was considered in
continuity with La Gauche Moderne. Looking inside the center-right camp, one can
notice that the position of the mainstream MoDem faction is to the left, compared
to members (Hervé Morin) who wanted to preserve the alliance with the Gaullists.
Concerning the UMP, one of the most right-wing positions is that of Debout la
République (2004), the faction headed by Nicolas Dupont-Aignan, which later
broke away to create a right-wing sovereignist party (Debout la France); similarly,
the right-wing faction La Droite Forte is also located on the right end of the scale.
Policy positions of party subgroups in Germany
For the German case, I split data and performed two different Wordfish analyses
considering 2007 as the breaking point. In the German case, the textual docu-
ments included in the analysis contain overall 550,142 words (with 27,296
unique words related to the first period and 26,177 related to the second). The
policy positions of German subgroups are displayed in Figure 2.6 in black (SPD)
or gray (CDU).
In the first period, the MIT is consistently the most right-wing group (mean:
1.12; standard deviation: 0.31) and this placement is in line with the position of
the association of entrepreneurs within the conservative CDU. Conversely, the
workers’ association (CDA) is located more in the middle of the policy space
(mean: –0.11; standard deviation: 0.49), and its documents look quite left-
oriented in some years, particularly in the early 1990s (for a similar result: Debus
& Bräuninger 2009). Finally, SPD factions tend to be more on the left. The
Frankfurter Kreis holds the most left-wing position and overall the left-wing
factions (FK and DL21) are indeed more on the left (mean: –0.49; standard devi-
ation: 0.96) compared to the moderate Seeheimer Kreis, which has been located
as centrist in three documents out of four (mean: –0.003; standard deviation:
0.73). In the second period the alignment of factions remains the same, though
Figure 2.6 Factions’ position on the left–right scale (Germany).

Note
First and second time period displayed together; SPD factions in black, CDU subgroups in gray.
the reciprocal distances shrink. The MIT (mean: 0.78; standard deviation: 0.06)
stands on the right compared to the CDA (mean: 0.44; standard deviation: 0.07),
then, in a centrist position, one can find the two moderate groups within the SPD,
i.e., Netzwerk Berlin and SK (mean: 0.31; standard deviation: 0.04); finally, the
leftist DL21 remains the most left-wing subgroup among those considered here
(mean: 0.09; standard deviation: 0.04). After 2007, according to the Wordfish
estimates, the most left-wing group is the SPD workers’ association (Arbeitsge-
meinschaft für Arbeitnehmerfragen (Workers Association, AfA)), which has
strong ties with the German trade unions (that in turn supported some demon-
strations against the Agenda 2010 platform of welfare and labor market reforms).
Surprisingly, in the first period the position of the AfA was instead more
moderate and this group was located, on average, halfway between the SK (the
right-wing of the SPD) and the MIT (the most right-wing CDU subgroup), close
to the group of self-employed inside the SPD (Arbeitsgemeinschaft Selbststän-
dige (Self-Employed Association, AGS)). However, as discussed above, I do not
consider the AfA or the AGS in the rest of the book.
Admittedly, while in the Italian and French cases the Wordfish analysis pro-
duced estimates that appeared immediately valid and not particularly affected by
temporal trends, pre-processing strategies or the inclusion/exclusion of some
documents, in the German case the interpretation of the output is not completely
straightforward. In particular, splitting the analysis in two groups, choosing 2007
as the dividing line, was crucial to obtain estimates that are not aligned on a tem-
poral basis.
One possible reason is connected with the heterogeneous nature of the intra-
party subgroups considered and the textual documents included in the analysis.
In fact, not all the documents were similar to each other. There are differences in
the structure of the text when comparing position papers or factional contribu-
tions to the party’s manifesto with records of conferences held by different intra-
party organizations (factions or associations) as well as with the practical
resolutions voted and adopted during these conferences. The substance of some
documents could also be problematic given that the content of generalist texts,
which address a variety of topics, is intrinsically different from that of texts
focusing on a single policy field (or subfield), which address only very specific
aspects of it.
In principle, one should try to consistently select documents based on their
style and content. For the German case, though, I had to relax such criteria in
order to expand the analysis, gathering more data. Notwithstanding this, the
number of documents collected is still quite low compared to Italy and France,
and this can be another reason to explain why this textual analysis was tricky.
Interestingly, an analysis performed by applying another automated technique
of text analysis (Wordscore) on a subsample of these textual documents also
reports similar lights and shades (Debus & Bräuninger 2009). That study prop-
erly locates the CDU subgroups (CDA and MIT), while the position of the SPD
subgroups was in part unusual, particularly with respect to the AfA, which
appeared to be unexpectedly more moderate than the core of the SPD, while the
right-wing Seeheimer Kreis was estimated to the left of the party, from 1987 to
1998 (Debus & Bräuninger 2009: 137). Nevertheless, this comes as good news
in terms of reliability. I compared the estimates of my Wordfish analysis (con-
sidering also documents that will not be used in the rest of the book, such as the
PDS manifestos or documents of the AfA and the FDP) with the estimates pro-
vided by Debus and Bräuninger (2009). The correlation, concerning 49 docu-
ments, is high and positive (0.80) and it gets even stronger (0.88) when leaving
aside the AfA, AGS, FDP and PDS, to focus only on the estimates of subgroups
considered in the course of the book. For all these reasons, I consider the data
robust enough to profitably investigate intra-party dynamics inside the SPD and
the CDU.
Text analysis: robustness checks, insights and best practices

In light of the ongoing debate in the field of text analysis about pre-processing
strategies and best practices (Greene et al. 2016), I conclude this chapter by
evaluating the robustness of the different Wordfish analyses presented above. By
doing that, I will provide some insights concerning the actual relevance of pre-
processing choices in different contexts and across different languages. This test
will also be useful to evaluate the robustness of the estimates that will be used in
the empirical analyses to be discussed in the next chapters.
It has been argued that Wordfish can be run on the entire set of words included
in textual documents, although Proksch and Slapin (2009) recommend to use a
subsample of words. This can be particularly crucial if the language does not
remain stable over time. Furthermore, words that are used very rarely could be
removed too (Proksch & Slapin 2009) to facilitate comparisons over time. This
strategy can be done by reducing the number of cells with zero-word frequency
in the document–term matrix. Finally, stemming is another pre-processing
strategy that is often used by researchers (Greene et al. 2016). It consists in redu-
cing complex forms of words to their simplest root, to better connect substan-
tively similar words; as such, words with similar meanings but different endings
are considered as a unique stem given that these words usually convey similar
information about the latent position of a textual document.
Based on this, I have replicated the analyses on Italy, France and Germany by
adopting alternative pre-processing strategies compared to those discussed
above. Table 2.2 summarizes the results of such comparison, reporting the
correlation between the estimates obtained by applying pre-processing strategies
and the estimates produced when no pre-processing has been adopted.
To sum up, the following strategies have been tested. First, I compared docu-
ments without pre-processing and documents in which stemming was performed.
Second, to account for any possible change of language over time, I replicated
the analyses by considering only words that consistently appear both before and
after a certain date (see below for details on each country). Third, I reduced spar-
sity, lowering the size of the document–term matrix, by removing terms that
occurred in fewer than 1 percent (low reduction) or 20 percent (strong reduction)
Table 2.2 Correlation of Wordfish estimates when using pre-processing strategies or not
Wordfish Analysis Italy 1st Italy 2nd France Germany 1st Germany 2nd
Stemming 0.96 0.99 0.99 0.99 0.97

No temporal break 0.93 0.81 0.88 0.28 0.60
Words consistent over time 0.96 0.97 0.95
Reduce sparsity 1% (low) 0.90 0.98 0.98 0.99
Reduce sparsity 20% (strong) 0.74 0.93 0.95 0.93
Note
The baseline for the comparison is the analysis ran without any pre-processing.
of the documents. Finally, in light of the discussion on temporal trends, which

pushed me to split the analyses on Italy and Germany, I want to evaluate what
happens when no temporal break is applied, running a unique analysis in these
two contexts. Conversely, in the French case, in which a single analysis on the
whole period was run, I will assess what happens when imposing a temporal
break to split data, analyzing a subsample of documents related to a shorter lapse
of time (i.e., using only post-1989 data and getting rid of pre-1989 motions that
are exclusively related to the PS).
In the French case, for brevity, the analysis has been replicated on a sub-
sample of 132 documents excluding the motions of the PCF that have been gath-
ered later; notice that comparing the estimates with or without stemming on the
whole set of motions (153) produces the same results. With respect to France,
the scenario in which I only considered the same words over time has been done
by using the year 1992 as a breaking point (given that there are data from 1971
to 1991 and then there is a gap until 1997). Accordingly, I run this analysis com-
paring word usage only for words that existed both before and after 1992.
For the Italian case, to control for the consistency of the language over time, I
replicated the analysis on pre-1988 data only. In fact, the time period related to
the first analysis was quite long and could have been affected by changes in lan-
guage that should only play a limited role in the subsequent shorter period (post-
1989). I evaluate the robustness in two different ways. First, by selecting words
that appear both before and after 1988; second, by selecting words that were
already in use before 1972. In both cases the results are similar.
Finally, with respect to Germany, the test about sparsity has been done, for sim-
plicity, on a single analysis that includes textual documents of both time periods.
Analogously, the effect of consistent word usage over time has been done running
a single analysis and considering only words that were used both before and after
2007 (which is the breaking point considered in the main analysis).
The results suggest that, overall, adopting alternative pre-processing solutions
do not really make the difference with respect to my data. The correlations dis-
played in Table 2.2 are almost always equal to 0.9 or above.
In more detail, stemming words or not, the results remain very similar with
correlations close to 1. Reducing the sparsity of the document–term matrix, in
turn, does not really make the difference and the size of the reduction does not
affect the results too. When reducing sparsity, in seven cases out of eight there is
a strong correlation (above 0.9) between these estimates and the original non
pre-processed data. The unique exception is represented by Italy, in the first
period (1946–1988), though only if a strong reduction of the document–term
matrix is applied. Here the correlation remains strong (0.74) but it falls below
the level of 0.9.
Turning to the consistency of word usage, I do not find strong differences
when considering only words that have been continuously used over time. Once
again, the estimates of these robustness checks are still almost perfectly correl-
ated (between 0.95 and 0.97) with the original results. Although from a theoret-
ical point of view it seems quite reasonable to perceive a change in language
over time, from an empirical point of view, the Wordfish estimates hold the same
and are not affected by removing words that are solely used in shorter time
periods.
Results are slightly different only when contrasting analyses in which data
are split based on a temporal threshold, instead of running a single time series
analysis on the whole amount of texts. This can be taken as a signal that lan-
guage is indeed changing over time (for additional details and some empirical
evidence: Ceron 2015c), and such change can partially affect the estimates.
However, for Italy (first period: 0.93) and France (data after 1989: 0.88) the
differences are quite minimal; the estimates are strongly correlated with those
obtained without any pre-processing. In the German case time seems to matter
more. Indeed, when running a single analysis, the results are still correlated,
but to a lower extent (0.60) with post-2007 data, and the correlation with
pre-2007 data is even lower (0.28). This empirical check highlights the
importance of splitting data in the German case, as discussed in the previous
section.
In the end, the estimates seem quite stable no matter the pre-processing strat-
egies adopted. This is strikingly evident with respect to Italian and French data,
i.e., those that will be more extensively used to perform large-N statistical ana-
lyses in the course of the book. Based on this, I can be confident about the
robustness of the empirical findings presented in the next chapters.
Notes
1 They held roughly 8 percent of posts in party body: around 13 seats out of 172 in the
board and three out of 36 in the executive branch. In the Senate 10 out of 146 MPs
were loyal to Fini while in the House their percentage was slightly higher: at least 34
MPs belonged to Finiani (around 12 percent). Such figures, compared to the share of
other former AN factions, are not far from the actual relative size of Finiani inside
the AN.
2 Finiani faction was organized through a net of associations and foundations, linked
with and somehow coordinated by Fini’s think-tank FareFuturo. Some of these sub-
groups strongly opposed the PDL leadership (i.e., Generazione Italia, led by Fini’s
lieutenant, Italo Bocchino, and Libertiamo, led by a former Radical party leader
Benedetto Della Vedova); others, such as Area Nazionale or Spazio Aperto (created
by Silvano Moffa) were more willing to negotiate with the PDL leadership (indeed
Moffa left the PDL in July 2010 but re-entered in December).
3 “Zur Sonne, zur Einheit,” Der Spiegel November 2010, p. 28.
4 These associations exist even inside the SPD, however, when analyzing this party I
consistently focus on factional subgroups only.
5 Due to parliamentary rules, in the Italian context the voting behavior includes the
absence as a fourth alternative. When strategic, in fact, the absence allows a policy
position to be expressed (Ceron 2015b).
6 It could be argued that talk (and language) is cheap. However, Giannetti and Laver
(2009) found a relationship between politician speeches during party congress and
their consequential political behavior in the parliamentary arena. Therefore, “if the
politician speaks and behaves in a consistent manner, […] may well become associ-
ated with a particular position on a particular dimension” (Laver & Shepsle
1996: 248).
7 Notice that different subgroups might decide, for instrumental reasons, to contest the
congress together, presenting a common list and a common motion. However, for
both theoretical and practical reasons, I do not investigate the process leading to such
inter-factional alliance and I measure their common position and their common
strength assuming that inside the party they will bargain as a single subgroup.
8 Parties’ newspapers are the followings: Avanti! (PSI); La Giustizia, L’Umanità,
Critica Sociale (PSDI); L’Unità (PCI and DS); La Voce Repubblicana (PRI); Il
Secolo d’Italia (MSI).
9 Parties’ websites were the followings: Greens (www.verdi.it), NPSI (www.socialisti.
net), PD (http://beta.partitodemocratico.it), PDCI (www.comunisti-italiani.it), PRC
(www.rifondazione.it), PSOC (www.partitosocialista.it), UDC (www.udc-italia.it);
for AN (www.alleanzanazionale.it) and DS (www.dsonline.it), I relied mostly on the
Internet Archive (www.archive.org).
10 In 1998 the PRC mainstream split and during a party body meeting four factions com-
peted to take control over the party in order to set the new line; each faction presented
a motion voted by members of the party National Political Committee. Accordingly, I
considered it as if it was a real congress.
11 http://62.210.214.184/cg-ps/ladocps.php.
12 For the SPD see: www.parlamentarische-linke.de; www.forum-dl21.de; www.
seeheimer-kreis.de; www.netzwerkberlin.de. For the CDU: www.cda-bund.de or
www.mit-bund.de/.
13 It could be argued that single words are uninformative about policy positions. On the
contrary, words matter. For example, during the 2009 PD congress Bersani’s oppon-
ents criticized him arguing that his victory would have transformed the PD into a
social–democratic party. The PD president Rosy Bindi (a Bersani’s ally) answered
claiming that “Bersani’s motion does not contain the word ‘social-democracy’ ” (La
Repubblica, August 15, 2009, translation mine).
14 Wordfish estimates are robust to its main assumptions (Slapin & Proksch 2008).
15 Multidimensional analyses can be performed by dividing these documents into single
pieces of texts related to different topics and running distinct analyses on each topic.
16 Excluding these documents leads to very similar results.
17 Interestingly, the position of words used by party members to refer to each other is
also coherent: “comrades” (compagni) used by socialists and communists, stands on
the left; “friends” (amici) used inside the DC, is located on the center; “companions”
(camerati), used by the MSI is on the right.
18 See Ceron & Negri 2018.
19 Factions location on the internal left–right scale usually tends to be steady, though
scholars report a few of exceptions related to the DC: Nuove Cronache (tied to Amin-
tore Fanfani) moved from left to right over time; Andreotti’s faction (Primavera)
temporarily shifted from right to left in 1980, and back to the right afterwards
(Bettcher 2005; Boucek 2010); the left-wing Forze Nuove moved to the right in the
early 1980s, when it firmly opposed the idea of a “Historic Compromise” (i.e., a coali-
tion government between the DC and PCI). The estimates successfully caught these
three shifts.
20 For the PD I also included more recent motions, see Chapter 6.
References
Bäck, H., Debus, M., & Müller, W.C. (2016). Intra-party Diversity and Ministerial Selec-
tion in Coalition Governments. Public Choice, 166(3–4), 355–378.
Benoit, K., Bräuninger, T., & Debus, M. (2009). Challenges for Estimating Policy Prefer-
ences: Announcing an Open Access Archive of Political Documents. German Politics,
18(3), 441–454.
Science, London.
781–804.
Budge, I., Klingemann, H.-D., Volkens, A., Bara, J., & Tanenbaum, E. (2001). Mapping
Policy Preferences: Estimates for Parties, Electors, and Governments 1945–1998.
Oxford University Press, Oxford.
Bull, M., & Rhodes, M. (1997). Crisis and Transition in Italian Politics. Frank Class,
London.
Calise, M. (2010). Il partito personale. I due corpi del leader. Laterza, Roma-Bari.
Carrubba, C., Gabel, M., Murrah, L., Clough, R., Montgomery, E., & Schambach, R.
(2006). Off the Record: Unrecorded Legislative Votes Selection Bias and Roll-Call
Vote Analysis. British Journal of Political Science, 36(4), 691–704.
Ceron, A. (2015b). Brave Rebels Stay Home. Assessing the Effect of Intraparty Ideo-
logical Heterogeneity and Party Whip on Roll-call Votes. Party Politics, 21(2),
246–258.
Ceron, A. (2015c). Changing Politics, Changing Language. The Effect of Institutional
and Communicative Changes on Political Language Measured through Content Ana-
lysis of Italian Intra-party Debates. Journal of Language & Politics, 14(4), 528–551.
Ceron, A. (2017). Social Media and Political Accountability: Bridging the Gap between
s11127-019-00644-0.
Ceron, A., & Greene, Z. (2019). Verba volant, scripta manent? Intraparty Conferences
and Issue Salience in France. Party Politics, doi: 10.1177/1354068819836034.
491–508.
Ceron, A., & Negri, F. (2018). March Divided, Fight United? Trade Union Cohesion and
Government Appeal for Concertation. West European Politics, 41(1), 218–239.
Clift, B., & McDaniel, S. (2017). Is This Crisis of French Socialism Different? Hollande,
the Rise of Macron, and the Reconfiguration of the Left in the 2017 Presidential and
Parliamentary Elections. Modern & Contemporary France, 25(4), 403–415.
Fukuyama, F. (1992). The End of History and the Last Man. Free Press, New York.
Giannetti, D. (2010). Secret Voting in the Italian Parliament. Paper Presented at the
Rationalité et Sciences Sociales Colloque, Collège de France, Paris, June 3–4.
Giannetti, D., & Benoit, K. (eds) (2009). Intra-Party Politics and Coalition Government.
Routledge, London.
Giannetti, D., & Grofman, B. (eds) (2011). A Natural Experiment on Electoral Law
Reform: Evaluating the Long Run Consequences of 1990s Electoral Reform in Italy
and Japan. Springer, New York.
Giannetti, D., & Mulé, R. (2006). The Democratici di Sinistra: In Search of a New Iden-
tity. South European Society and Politics, 11(3), 457–475.
Greene, Z., Ceron, A., Schumacher, G., & Fazekas, Z. (2016). The Nuts and Bolts of
Automated Text Analysis. Comparing Different Document Pre-Processing Techniques
in Four Countries. Open Science Framework. November 1. osf.io/ghxj8.
Greene, Z., & Haber, M. (2016). Leadership Competition and Disagreement at Party
National Congresses. British Journal of Political Science, 46(3), 611–632.
Greene, Z., & Haber, M. (2017). Maintaining Partisan Ties: Preference Divergence and
Partisan Collaboration in Western Europe. Party Politics, 23(1), 30–42.
Ibenskas, R. (2017). Electoral Competition after Party Splits. Political Science Research
and Methods, forthcoming.
Klüver, H. (2009). Measuring Interest Group Influence Using Quantitative Text Analysis.
European Union Politics, 10(4), 535–549.
Laver, M. (1999). Divided Parties, Divided Government. Legislative Studies Quarterly,
24(1), 5–29.
Laver, M., & Benoit, K. (2003). The Evolution of Party System between Elections.
Laver, M., Benoit, K., & Garry, J. (2003). Extracting Policy Positions from Political
Texts Using Words as Data. American Political Science Review, 97(2), 311–331.
Laver, M., & Shepsle, K.A. (1996). Making and Breaking Governments. Cabinet and
Legislature in Parliamentary Democracies. Cambridge University Press, Cambridge.
250–277.
Lombardo, A. (1972). Dal proporzionalismo intrapartitico al frazionismo eterodiretto.
Mershon, C. (1994). Expectations and Informal Rules in Coalition Formation. Com-
parative Political Studies, 27(1), 40–79.
Mershon, C. (2001). Party Faction and the Coalition Government: Portfolio Allocation in
Italian Christian Democracy. Electoral Studies, 20(4), 555–580.
Pasquino, G. (2000). La transizione a parole. Il Mulino, Bologna.
Proksch, S.-O., & Slapin, J.B. (2009). How to Avoid Pitfalls in Statistical Analysis of
Political Texts: The Case of Germany. German Politics, 18(3), 323–344.
Proksch, S.-O., & Slapin, J.B. (2010). Position Taking in European Parliament Speeches.
Proksch, S.-O., & Slapin, J.B. (2012). Institutional Foundations of Legislative Speeches.
Raniolo, F. (2006). Forza Italia: A Leader with a Party. South European Society and
Politics, 11(3), 439–455.
Rosenthal, H., & Voeten, E. (2004). Analyzing Roll Calls with Perfect Spatial Voting:
France 1946–1958. American Journal of Political Science, 48(3), 620–632.
Slapin, J.B., & Proksch, S.-O. (2008). A Scaling Model for Estimating Time-Series Party
Positions from Texts. American Journal of Political Science, 52(3), 705–722.
Spirling, A., & McLean, I. (2007). UK OC OK? Interpreting Optimal Classification
Scores for the U.K. House of Commons. Political Analysis, 15(1), 85–96.
Spirling, A., & Quinn, K.M. (2010). Identifying Intra-party Voting Blocs in the UK
House of Commons. Journal of the American Statistical Association, 105(490),
447–457.
Volpi, E. (2019). Ideology and Party Switching: A Comparison of 12 West European
Countries. Parliamentary Affairs, 72(1), 1–20.
Yale University Press, New Haven CT.
3 The determinants of factionalism
Introduction
The present chapter investigates the determinants of factionalism. Several hypo-
theses drawn from the existing literature and connected with the theoretical
framework discussed in Chapter 1 will be discussed and tested. By doing that, I
evaluate what elements contribute to increasing or decreasing the degree of fac-
tionalism inside parties. Particular attention will be devoted to the organizational
structure of the party to assess how alternative intra-party rules can affect fac-
tionalism. Beside this, the impact of intra-party policy heterogeneity will also be
tested to check whether intra-party fragmentation is stronger when internal sub-
groups retain different views. If this is the case, one can argue that many factions
are policy oriented and should be considered as “factions of principle”; other-
wise, there would be support for the idea that factionalism has mostly to do with
office payoffs and career motives and in this case one should consider them
“factions of interest” (Sartori 1976).
Based on data availability in each country, three different types of analysis
will be performed. For Italy and France, I investigate the number of factions.
Arguably, the only objective criterion to count factions is to measure how many
subgroups contest the congress (presenting a motion). This is possible only if the
congress is contested as one would not observe any measurable value when
intra-party dynamics lead to a unitary debate.
Given that the number of factions can be observed only when the congress is
contested, to deal with any potential selection bias, I will also investigate what
elements impact on the likelihood of observing unitary or contested congresses.
Focusing on the latter, I will examine a range of variables to detect the determi-
nants of factionalism.
Concerning Germany, however, I consider the number of factions as given
for the reasons discussed in the previous chapter. Consequently, another indi-
cator of intra-party fragmentation will be used. I will examine whether, during
the party congress, the leadership race is contested or not. By doing that, I can
assess what elements affect the share of votes won by the party leader. In fact, a
higher support would be a signal of enhanced unity and can indicate the
existence of a weak internal fragmentation (for a similar operationalization see
Emanuele et al. 2018).
The determinants of factionalism 65
The determinants of factionalism: literature and hypotheses
Many studies investigated the causes of factionalism focusing on the role played
by rules and by attributes of the party and the party system. No study, however,
provided a comprehensive explanation of the phenomenon considering the role
of intra-party heterogeneous policy preferences along with that of internal rules
and statutes.
In this section I review the main expectations suggested by existing studies,
formulating additional hypotheses linked with my theoretical framework. Two
sets of hypotheses will be raised. The first one (“A”) is devoted to explaining the
causes of intra-party fragmentation. The second (“B”) focuses on the likelihood
of observing a contested congress and, in this regard, specific attention will be
paid to internal rules in order to assess whether different modes of party organ-
ization lead to different patterns of factional conflict (i.e., contested or uncon-
tested congresses) and to the emergence of a selection bias in the data.
The debate on the sources of factionalism was opened by Sartori (1971) who
focused on intra-party polarization and on the impact of internal electoral rules on
the number of factions (for an analytical review see Ceron 2011). To start with,
Sartori (1976) hypothesized that inside polarized parties the number of subgroups
should be greater due to the co-existence of divergent policy preferences. Other
scholars, in turn, hypothesized that the heterogeneity of the rank-and-file might
increase the number of factions (Bogaards & Boucek 2010). In Chapter 1, I
claimed that internal polarization affects the ratio between π and ε, due to the cost
of membership: in polarized parties, in fact, factions should activate and pick up
the voice option in order to extract a wider amount of payoffs. Accordingly, a
greater internal polarization should trigger a growth in factionalism.
Hypothesis 3.1A (H3.1A): Intra-party polarization increases the number of

factions.
Moving to the role of internal rules, I want to assess whether they affect the
degree of factionalism, but also the likelihood of observing contested congresses.
Indeed, rules can be crucial not only to decrease fragmentation, but also to hide
any dissent. To account for this, one can focus on party statutes that “contain,
roughly, three major elements: a set of prohibitions, the organizational structure,
and the electoral arrangements” (Sartori 1976: 85).
Starting from the latter, Sartori (1971, 1976: 86) applied Duverger’s Laws to
intra-party politics claiming that PR is a sufficient (but not unique) condition for
the growth of intra-party fragmentation: PR “does become a sufficient cause of the
multiplication of fractions.” Under disproportional intra-party electoral rules (i.e.,
majoritarian systems, PR with high thresholds and rules that establish a majority
prize for the first ranked group) factions should face the incentive to merge together
in order to take advantage of the electoral system and to increase their size (i.e.,
their strength) inside the party body. Accordingly, one would expect a greater
number of factions in parties with purely proportional internal rules.
Due to disproportionality, for minority factions it becomes difficult to get
represented inside the party body. As a consequence, they might even refuse to
openly defy the party mainstream. Small factions can avoid contesting the con-
gress and opt for a “bandwagoning” strategy. By adopting a cooperative behav-
ior, they aim to reinforce consensual internal dynamics hoping to receive (as a
reward) a decent share of seats in the party body.
Going back to the game–theoretic model, it could be argued that the cost of
voice (ν) outweighs ω, in parties that adopt disproportional rules. If this is the
case, one should observe low or no voice, i.e., a low rate of contested congresses
and little factionalism. Accordingly, I raise the following hypotheses.
Hypothesis 3.2A (H3.2A): Disproportional internal electoral rules decrease the

number of factions.
Hypothesis 3.2B (H3.2B): Disproportional internal electoral rules decrease the

likelihood that the congress will be contested.
Apart from the internal electoral system, party statutes include a wide range
of additional intra-party rules as well as formal (or informal) attributes related to
the party organization. All these rules might carry an explanatory power too. For
instance, statutes might prescribe an explicit ban on the existence of organized
factions. Arguably, such explicit prohibition should increase the cost of voice
(non-compliant subgroups can be subjected to sanctions) decreasing the number
of subgroups that contest the congress, thereby leading to a unitary congress.
Indeed, such a ban could be adopted inside parties whose leaders are focused on
enhancing cohesion (ν is greater than ω) precisely to reduce factionalism. In
view of that I hypothesize what follows.
Hypothesis 3.3A (H3.3A): An explicit ban on factions should decrease their

number.
Hypothesis 3.3B (H3.3B): An explicit ban on factions should decrease the likeli-
hood that the congress will be contested.
Another element deals directly with party organization. Some parties, espe-
cially those based on Marxist ideology, are organized according to the principles
of “democratic centralism.” This is something different from the mere absence
of organized factions. Under democratic centralism, party members are theor-
etically free to discuss and support different views about party strategy and ideo-
logy. However, dissent must be expressed only inside the party. Once the party
body sets the line, this choice must be enacted by all members (dissenters
included).
This very strict rule is therefore likely to suppress any open and public
dissent. The public expression of dissent jeopardizes the idea of a cohesive party
(granted by democratic centralism) weakening the leadership, which in turn will
impose cohesion through discipline (Ceron 2015b), raising the voice cost for dis-
senters too. In case of violation, in fact, democratic centralism prescribes the
adoption of severe sanctions and any organized faction will be expelled from the
party (Ceron 2015a). Accordingly, in congresses held under this rule one would
expect a complete absence of dissent due to the huge value of ν.
Hypothesis 3.4B (H3.4B): In parties organized through “democratic centralism”

the likelihood of a contested congress is lower.
Beside these rules, some formal and informal settings can also impinge on the
strength and the autonomy of party leaders. If leaders are directly elected by a
wide selectorate (e.g., party members), their autonomy can be boosted and they
become strong enough to unify the whole party behind their shoulders. To the
contrary, leaders indirectly appointed in smoke-filled rooms will be hostage to
party factions. In a small committee, in fact, the vote of any single activist might
be decisive to appoint one or another leader. Consequently, this represents an
incentive for small subgroups of activists to activate and organize as a new
faction, in order to become pivotal for the internal equilibrium and to extract
larger benefits. If this is the case, one can expect an increase in factionalism,
whereas in parties with more autonomous leaders the number of faction should
be lower; the leader’s strength can also decrease the likelihood of observing con-
tested congresses.
Hypothesis 3.5A (H3.5A): Direct election by a wide selectorate (which fosters

leader’s autonomy) should decrease the number of factions.
Hypothesis 3.5B (H3.5B): Direct election by a wide selectorate (which fosters

leader’s autonomy) should decrease the likelihood of a contested congress.
The most extreme scenario in which leader’s autonomy reaches its peak, is
the case of personal parties (Calise 2010). Inside these parties, the leader is often
self-enthroned (Kenig 2009) and the party heavily depends on his charisma, up
to the point that the leader’s name can be included in the party label and the
leader exerts a strong impact on members and activists (Raniolo 2006). In this
context, one should not find any observable disagreement. In fact, charismatic
leaders boost party loyalty, which increases the exit costs (ε) thereby making the
voice option pointless. As a consequence, despite any possible existence of
internal heterogeneity, the party congress could be uncontested.
Hypothesis 3.6B (H3.6B): In “personal parties” the likelihood of a contested

congress is lower.
So far, I have only discussed the role of heterogeneous policy preferences and
intra-party rules. The literature, however, has investigated other possible causes
of intra-party fractionalization. One of these refers to the availability of resources
(Golden & Chang 2001; Zincone 1972; Zuckerman 1979), namely money, clien-
teles and patronage. Zincone (1972) highlighted the importance of an auto-
nomous access to resources that is crucial to build, organize and keep alive a
faction. Organized factions have their own headquarters, collect membership
fees and print their own newspapers and magazines. This allows them to survive
and grow. Indeed, many party factions in Italy, France and Germany used to
print magazines to promote themselves, to express their views and to inform
members about party’s and faction’s meetings and decisions. Factional maga-
zines were so important that one strategy to decrease factionalism was precisely
to forbid printing them. Often the faction’s and the magazine’s name coincided.1
While financial resources matter, it is not easy to find measures concerning
the financial autonomy of intra-party subgroups. However, if parties allocate
resources between activists and factions, one can reasonably assume that parties
with a higher budget can provide more resources to internal subgroups. The
greater the party revenues, the higher the number of factions inside it. Analo-
gously, given that party members contribute by paying a membership fee, one
can expect that parties with a greater number of members and supporters will
have more financial resources and will be more factionalized. To sum up, the
availability of financial resources will be measured using the party size (i.e.,
votes share) proxy for it.
Pasquino (1972) pointed out that the elections represent, at any level, a crucial
opportunity to boost the faction, as long as gaining seats constitutes a mech-
anism to multiply the resources available to the faction itself. This reasoning also
applies to the public funding of parties, which is often distributed proportionally
to the share of votes won by parties in general elections, or proportionally to the
parties’ ability to mobilize members and sympathizers. Therefore, the number of
factions should be higher in largest parties.2
Notice that this is not only related to financial resources. In fact, the internal
structure of large parties is more complex, leading to an increase in factionalism
(König 2006). Intuitively, large parties also attract a wider number of members
and these members might retain more widespread policy preferences, leading to
an increase in terms of polarization and fragmentation (something that will be
controlled for when testing H3.1A).
Hypothesis 3.7A (H3.7A): The number of factions increases along with party
size and with the availability of resources within the party.
Several authors discussed the link between factionalism and electoral systems.
In Japan, there is evidence suggesting that changes in the electoral system altered
the factional structure of political parties (Cox & Rosenbluth 1993; Cox et al.
1999). In the Italian case scholars focused on the role of preference voting (Katz
1986; Katz & Bardi 1980; Pasquino 1972) arguing that it fosters intra-party
competition.
Preference voting allows factions to overcome any leadership constraint and,
if factions are strong enough from an electoral point of view, their members can
get elected in spite of any leader’s attempt to impose discipline. By providing
factions with an (almost) unmediated access to parliamentary seats (and to the
related resources), the preference voting grants them the opportunity to flourish,
producing a growth in internal fragmentation. Preference voting decreases the
price of ν and, as a consequence, should boost the use of voice in intra-party
dynamics.
The role of the electoral system can be tested by taking advantage of the
Italian case and comparing Italy before and after the electoral reforms experi-
enced since the 1990s. These changes allow a comparison between systems
based on open list PR, before 1994, and those adopted afterwards to be per-
formed (Carey & Shugart 1995). Since 1994 Italy has experienced two different
electoral rules: a mixed system (1994–2001) with a 25 percent closed list PR
quota and the remaining 75 percent of seats assigned via “first-past-the-post,”
and a closed list PR with thresholds and majority prize (in 2006 and 2008). Fur-
thermore, between 1994 and 2001 the selection of candidates elected under plu-
rality districts has been centralized, being transferred from the regional to the
national level (Di Virgilio & Kato 2001). Both electoral systems are based on a
closed list and promote the idea of a centralized representation. Conversely, open
list PR favors the dispersion of authority increasing the number of players inside
the party. When the district magnitude is very high, such as during the First
Republic, intra-party competition for preference votes fosters the incentive to
break the monolithic party structure (Carey & Shugart 1995; Shugart 2001) so
that candidates try to exploit their factional linkage in order to get elected. To
the contrary, under closed list systems, or when the candidate selection pro-
cedure is highly centralized, party elites retain a greater control over candidacies,
decreasing the relevance of factions as well as their number (Carey 2007; Cox
et al. 1999; Pasquino 1972).
Hypothesis 3.8A (H3.8A): Closed list PR and centralized candidate selection

decrease the number of factions.
A further hypothesis concerns the availability of office payoffs. Sartori (1976)

discriminates between factions of principles (focused on policies) and factions of
interests (focused on office spoils). Arguably, office payoffs are larger in ruling
parties. Therefore, the number of factions of interests can rapidly increase when
parties are in office as these subgroups can benefit from the payoffs linked with
cabinet membership (Ceron 2014). Keeping factions of principles constant, if
new subgroups are born to take advantage of office payoffs, one can expect to
observe a wider number of factions inside ruling parties.
Hypothesis 3.9A (H3.9A): The number of factions is greater for parties in office.
To conclude, the last argument is related to the impact of the electoral cycle.
Several scholars argue that intra-party division can produce negative con-
sequences on a party’s electoral performance. For instance, Greene and Haber
(2015), show that intra-party disagreement negatively affects voters’ perceptions
of parties and, consequently, it damages their electoral support. As a con-
sequence, when new general elections approach, the cost of voice (ν) could
increase. Given that divisions will hurt the party, any dissent could be more
easily sanctioned by the party leadership. This could halt the use of voice, and
rival factions should stop voicing their dissent, opting for a truce. Conversely,
when elections are not on the horizon, parties can feel free to divide and to pub-
licly display internal dissent and disagreement. As such, one can expect to
observe contested congresses.
Hypothesis 3.10B (H3.10B): Looming elections should decrease the likelihood

of a contested congress.
The role of rules: party organization in Italy, France and

Germany
The previous chapter mapped the intra-party preferences of factions and sub-
groups inside Italian, French and German political parties. Parties retain different
rules, and these rules can give rise to alternative party environments. As dis-
cussed in Chapter 1, these different organizational settings can shape intra-party
dynamics (promoting more consensual or less consensual relationships) and can
affect the degree of representativeness of internal minorities. In this section, I
describe different intra-party rules and styles of organization. Focusing on party
statutes, I provide details concerning the four main rules mentioned above: pro-
portional/disproportional internal electoral system, factional ban, democratic
centralism and direct election of party leader (leader autonomy).
Each rule can grant to internal minorities stronger or lower opportunities in
terms of freedom and representativeness. The lower the degree of representa-
tiveness, the greater the advantage retained by the party leader and the party
mainstream.3 Highly disproportional electoral rules, a strong ban on factions,
the existence of democratic centralism or the direct election of the party
leader, are features that generate low (L) representativeness. Conversely a
high (H) representativeness is granted by proportional electoral rules and indi-
rect selection of the leader (by a small party executive agency), as well as by
the lack of democratic centralism and factional ban. In turn, some rules might
lie in between producing a medium (M) levels of representativeness; for
instance, the internal electoral system can combine some proportional ele-
ments with more majoritarian rules, but also the ban on faction can be more or
less strict.
Table 3.1 reports, per each party in each country, the percentage of congresses
(considered in this book) held under each rule, distinguishing them on the basis
of their degree of representativeness (Low, Medium, High). The total number of
cases reported in this table is equal to 83 for Italy, 41 for France and 40 for
Germany. Rules vary across parties and countries, but often also over time
within the same party.
Table 3.1 Details on internal rules in Italy, France and Germany
Degree of Representativeness of Each Party Rule
Intra-Party Factions Ban Democratic Direct

Electoral Rule Centralism Election
L M H L M H L H L H
Maj. Mix PR Yes Some No Yes No Yes No
Italy
AN – – 100 100 – – – 100 100 –
DC 55 – 45 55 – 45 – 100 45 55
DS – – 100 – – 100 – 100 100 –
FV – – 100 – – 100 – 100 100 –
MSI – – 100 – – 100 – 100 80 20
NPSI – – 100 – – 100 – 100 100 –
PCI 33 – 67 33 – 67 – 100 – 100
PD – 100 – – 100 – 100 100 –
PDA 100 – – – – 100 – 100 – 100
PDCI – 100 100 – – 100 – – 100
PLI 27 55 18 100 – – – 100 – 100
PRC – – 100 100 – – 100 – 100
PRI – 73 27 100 – – – 100 – 100
PS – – 100 – – 100 – 100 100 –
PSDI 10 45 45 – – 100 – 100 – 100
PSI – – 100 25 – 75 8 92 8 92
PSIUP – – 100 – – 100 – 100 – 100
UDC – – 100 – – 100 – 100 100 –
France
EELV – – 100 – – 100 – 100 50 50
MoDem – – 100 – 100 – 100 100 –
PCF 100 – – 100 – – – 100 – 100
PS – – 100 100 – – – 100 36 64
UMP 100 – – – – 100 – 100 100 –
Germany
CDU – – – – – – – 100 100 –
SPD – – – – – – – 100 95 5
Note
Percentage of congresses held under each rule.
Starting from the internal electoral system, it can be noticed that many parties
adopt proportional rules, usually a pure PR. In some cases, however, the PR is
combined with a majority prize (particularly for the Italian PLI, PRI and PSDI)
and a low threshold. Less representative rules, such as majoritarian systems or
PR with majority prize and high threshold of representation have seldom been
used and especially in the early post-war period. Sometimes, the statutes also
assign a certain number of posts to non-elected members belonging to the party
leadership and the party mainstream (so-called notabili). This happened inside
the DC, and partially also within the UMP in France.
Some parties changed this rule over time. For instance, the DC started with a
majoritarian system (leaving also room for notabili) but switched to PR in the
mid-1960s; however, to halt the proliferation of factions, in the 1980s it stepped
back introducing high thresholds (20 percent) and a majority prize, combined
with the novelty of the direct election of the party leader (see below).
The choice between a pure PR or a majoritarian system was usually deemed
crucial by internal minorities and the selection of the electoral system has often
been object of factional conflicts. Remarkably, left-wing DC factions have
always battled to get PR for the election of congress delegates (Boucek 2010;
Venditti 1981: 40; Zincone 1972). In 1954, the left-wing faction Forze Sociali
proposed a petition to introduce the PR; this petition, however, was defeated by
a small margin and the PR system was enacted only ten years later in 1964,
boosting internal fragmentation. At the end of the 1970s, to contain factionalism,
the electoral system was modified again. A threshold and a majority prize were
included. However, left-wing factions complained and fought to re-introduce a
pure PR (Venditti 1981), which would have allowed them to exert a stronger
influence over the party mainstream.
Interestingly, the internal electoral rule generated a heated debate also
during the (de facto) founding congress of the French Socialist Party held in
Epinay (1971). In that congress, three alternative options have been discussed.
The first option was to preserve the status quo adopted by the SFIO (ancestor
of the PS) since 1963 and until 1969. The status quo consisted in a full majori-
tarian system that safeguarded the rights of the minorities granting them some
posts. Alternatively, the party mainstream proposed a PR system combined
with a moderately strong threshold (10 percent or 15 percent) and a majority
prize (equal to the 25 percent of the votes won by the minority factions).
However, the left-wing minority faction CÉRÉS refused such a proposal and
supported a reform based on PR with a low threshold (5 percent). This latter
rule was approved by the party congress and it has never been changed for the
entire life of the PS. The PS system probably contributed to reinforcing the
existence of pluralistic views inside the PS. Remarkably, this choice was so
important that the party specified the validity of this norm in multiple sections
of the party statutes, including in an article devoted to discussing the repres-
entation of the internal minorities.
Moving to the factional ban, party statutes sometimes forbid the existence of
organized factions but also the existence of temporary tendencies (which were
instead allowed in the Italian PRC, heir of the PCI). In some cases, the ban was
implemented precisely as an attempt to lower the degree of factionalism,
although often with poor results.
In France, the MoDem allowed the creation of internal associations, which
were also officially recognized by the Gaullist RPR (ancestor of the UMP) since
1989. In turn, the UMP went even further. This party was in fact thought to be
structured in clubs and associations merged together. In view of that, the statutes
officially allowed the creation of factions (called “movements”), with the
purpose of representing the different stances existing inside the party. The
statutes provided them with the opportunity to contest the congress presenting
their own declaration of principles, i.e., motion (even though, to contain faction-
alism and preserve party unity, this rule was applied only in the 2012 congress).
Representation in the party body was granted to factions able to reach at least 10
percent of congress votes.
To the contrary, the PS explicitly forbids any organized tendency. Article 21
of the statute of the PS approved in Epinay (1971) stated that there is a full
freedom of discussion inside the party, though any organized faction would not
be tolerated. Intriguingly, the existence of the PS has always been affected by
factional conflict and, despite this formal ban, internal subgroups were de facto
allowed, for the sake of stimulating pluralism in internal debates.
Factionalism was forbidden inside the PCF too. Notice that the ban is consist-
ent with a mode of organization based on democratic centralism, which is typical
of Marxist parties, and inside the PCF it was effective only until democratic cen-
tralism was in place. In 1994 this setting was abolished and a few years later
PCF rival factions started to form and to compete, despite the formal existence
of a ban on organized tendencies.
With respect to democratic centralism, in the Italian context the PCI adopted
this structure for almost its entire life, and the statute was modified only at the
end of the 1980s precisely to allow inter-factional competition. Democratic cen-
tralism was also in place inside the PSI, from 1949 until 1957 during the
orthodox leadership of Morandi, when the party was still following Marxist prin-
ciples (and used to hold uncontested congresses). During the Second Republic
only one party, the communist PDCI, retained a similar organization, although
the People of Freedom, the LN and, later, also the Five Stars Movement (M5S)
seemed to replicate a similar structure, allowing internal dissent only until a
decision is taken and then forbidding and punishing any public disagreement.
The last rule reported in Table 3.1 concerns the modes of selection of the
party leader. As discussed in Chapter 1, I will distinguish between indirect elec-
tion demanded by a small party agency compared to direct election by party
members or congress delegates (when these delegates receive a clear mandate
based on the vote of members in local party branches congresses).
For a long time, the task of selecting party leaders was assigned to small party
committees, such as the National Council or the Central Committee. In many
parties and many countries, over the years, there has been a trend aiming to
increase the inclusiveness of the selection mechanism to involve party activists
and members, or sometimes even simple voters or unregistered sympathizers.
This trend is particularly evident in Italy and France.
In Italy, this process started at the end of the 1970s (Ceron & Negri 2017),
when the DC enlarged the selectorate of the leader (1976) to include the whole
audience of delegates at party congress. The same rule was also adopted by the
PSI in 1981. The heirs of the PCI implemented the direct election only after
1998 to select the DS leader, while a few years later, when the DS merged into
the PD, primary elections were introduced (involving party members but also
non-enrolled sympathizers).
In France this process took place in the 1990s as well. Congress delegates
selected the PS leader for the first time in 1993 (Rocard), while later on in 1997
direct election by party members was introduced. Analogously, in 1998 the RPR
switched to direct election and party members kept selecting the leader also in
the UMP.
As far as leader selection is concerned, the German case is interesting too.
Mainstream German parties, such as the CDU and the SPD, select their national
leaders during party congresses through the vote of delegates. However, German
parties usually avoid open leadership contests (Cross 2013). In fact, “party exec-
utives or smaller circles of party notables usually agreed upon a candidate who
was then presented to the party conference for coronation” (Astudillo & Detter-
beck 2018: 3). In this regard, the selection is indirect given that party members
have no voice in it. Only once, in 1993, the SPD tried to enlarge the selectorate,
introducing a consultative membership ballot for the selection of the national
party leader (Parteivorsitzender). After that episode, however, the SPD returned
to more traditional selection rules (Astudillo & Detterbeck 2018). With respect
to the other rules, the German case is less interesting for our purposes, given that
there is no open electoral competition between rival factions so it makes no
sense to focus on internal electoral rules and factional ban, even though the CDU
actually enacted a sort of ban in 2018, refusing to officially recognize recently
formed factions (the WerteUnion and the Union der Mitte; see the discussion in
Chapter 1).
While for France and Italy there is some variation in internal settings across
parties (but also within the same party), in Germany more similar rules exist.
Even if looking at the candidate selection (which is beyond the scope of this
book), in Germany there is very little variation even across parties. In terms of
inclusiveness and centralization, several scholars did not find differences
between alternative German parties (Bille 2001; Spies & Kaiser 2014).
To conclude, the rules described in this section will be used in the present
chapter to test hypotheses on the causes of factionalism. Later on, these rules
will be retrieved to test their direct or conditional effect on a number of out-
comes ranging from the allocation of payoffs to the occurrence of party splits
and leader replacements.
The determinants of factionalism in Italy

In the Italian case I will examine the number of factions that contest a congress
as a measure of factionalism, to assess its determinants. Overall, among the
parties included in this dataset, the average number of factions (NOF ) is equal to
3.08 (standard deviation: 1.14).4 There are no differences over decades or
between the First and the Second Republic. Analogously, differences between
parties are tiny. The number of factions inside the DC and MSI–AN is close to
four, while there are only two factions inside the UDC, PDCI and NPSI and
approximately two in the PRI (the mainstream and a minority group). In more
detail, in 40 percent of congresses there are exactly three factions, even though
some variation exists and in 35 percent of cases the competition involves only
two groups. Only a few times does one find four factions (13 percent) or more,
with a maximum of seven (DC 1969), while in three cases six factions contested
the congress (the PLI in 1974 and the MSI in 1987 and 1990). When applying to
intra-party politics the formula proposed by Laasko and Taagepera (1979) to
measure the effective number of parties one would notice a similar distribution.
The effective number of factions (ENF, see Boucek 2009) is indeed strongly and
positively correlated (0.80) with NOF. In detail, ENF ranges from 1.02 to 4.86,
with a mean of 2 (standard deviation: 0.76). In 30 percent of cases ENF is lower
than 1.5, while it is higher than 2.5 in 16 percent of congresses. It is unlikely to
find parties divided into many factions that retain approximately the same
strength. Nevertheless, even when the majority faction is quite strong, there is
still a considerable minority subgroup holding around 20 percent of congress
votes.
To test what elements affect the number of factions I built a dataset that
includes all the contested congresses of parties listed in Table 2.1 (Chapter 2) for
which I was able to find data; the total amounts to 83 cases. The main dependent
variable (y) is the NOF, which corresponds to the number of motions presented
by factions during the congress. I will test the previous hypotheses by means of
an ordinary least squares (OLS) regression. As far as there are multiple observa-
tions (congresses) nested within each party across time they might not be inde-
pendent. To cope with this issue observations have been clustered by party,
estimating clustered standard errors accordingly. Clustered standard errors, in
fact, are heteroskedastic and autocorrelation consistent (Rogers 1993).
The results of such OLS analysis could be biased if there is a problem with
sample selection. To assure unbiased estimates I ran a Heckman (1979) model.
Retrieving the parties listed in Table 2.1, for each of them I pooled together all
the congresses held during their entire life. This expanded dataset contains
information on both contested and uncontested congresses. Overall, 198 cases
have been considered (see Table 2.1, column 2). I applied the Heckman model
that estimates the parameters in a two-stage process. In the first stage, the model
assesses the effect of some variables on a dichotomous variable (z). The variable
z is a dummy that determines whether the main dependent variable y is observed
or not; it takes value 1 when the outcome is observed and 0 when it is not. In the
second stage, the model estimates the impact of another set of variables on the
expected value of y, conditional on y being observed. Whenever the error terms
of the two analyses (the first one on z and the second on y) are correlated (ρ ≠ 0)
a selection bias exists and this would confirm the need to revert to the Heckman
model.
In the first stage, I estimate the likelihood that the congress will be contested.
The variable z, Contested Congress, is a dummy variable equal to 1 when the
congress is contested and to 0 otherwise. In the dataset 120 congresses are con-
tested (60 percent of 198 total congresses). Four independent variables linked
with the intra-party rules described above are included in this stage: Dispropor-
tional Rule (H 3.2B) is a categorical variable equal to 0 when the internal
electoral rule is a pure PR, to 1 when the voting rule is a PR with majority prize
and low threshold (5 percent) and to 2 when the congress is held under a majori-
tarian system or under a PR with majority prize and high threshold of representa-
tion (20 percent);5 Factions Ban (H 3.3B) is equal to 1 when the statute contains
an explicit ban on the existence of organized factions, it is equal to 2 when tem-
porary tendencies are banned too, while it is equal to 0 when there is no explicit
ban; Democratic Centralism (H 3.4B) takes the value of 1 when the party organ-
ization follows this guideline and 0 otherwise; Direct Election (H 3.5B) is equal
to 1 when the party leader is directly elected by party congress delegates or party
members and it is equal to 0 when the leader is indirectly appointed by a small
committee. Beside these variables, I also control for the effect of the electoral
cycle through the variable Years Before the Election (YBE), which measures the
number of years remaining before the next general elections (H3.10B).
Then, in the second stage, I estimate the impact on NOF testing the hypo-
theses discussed above.6 The independent variables are the followings. Internal
Polarization (H3.1A), measured as the absolute distance between the two most
extreme factions in each party congress; Disproportional Rule (H3.2A); Factions
Ban (H3.3A); Direct Election (H3.5A); Party Size (H3.7A), i.e., the share of
votes won by the party in the last election before the congress (which accounts
for the amount of resources available inside the party);7 Open List (H3.8A), a
dummy variable that measures the impact of preference voting and takes the
value of 1 when the electoral system is an open list PR (from 1946 to 1992) and
the value of 0 when the electoral law includes a closed list PR (for the whole
amount of seats as in 2006 and 2008 or limited to the 25 percent quota assigned
through PR from 1994 to 2001); Party in Office (H3.9A), equal to 1 if the party
is in office and 0 otherwise. Finally, I also control for parties that are not in par-
liament through the variable Extra-parliamentary, which is equal to 1 when the
party is out of parliament and to 0 otherwise.
So far, 198 congresses have been considered. However, the total number of
congresses held by Italian parties between 1946 and 2010 is around 300. To
enlarge the dataset, I gathered further information from party statutes (concern-
ing the four variables related to party rules) and I assessed whether these missing
congresses were contested or uncontested. By doing that, I was able to extend
the dataset including more congresses and additional parties, providing a more
comprehensive description of the effects of party statutes on the likelihood of
observing unitary or contested congresses. This larger sample contains informa-
tion on 221 congresses. It includes parties that played an important role in the
Second Republic, such as the Italian Popular Party (PPI) and DL, heirs of the
DC, and the post-communist PDS. Some “personal parties” such as IDV, the LN
or FI and its heir the PDL (Calise 2010; Raniolo 2006) have been added too;8 to
account for them (H3.6B), I also include a further dummy variable Personal
Party, equal to 1 for FI, PDL, LN and IDV.9
Table 3.2 reports the results of the analysis. Model 1 shows the coefficients of
the OLS regression on the selected dataset; Model 2 presents the results of the
Heckman model on the whole dataset; Model 3 estimates the Heckman model on
Table 3.2 The determinants of factionalism in Italy
Model (I) (II) (III)
OLS Heckman MLE Heckman MLE
Outcome Equation DV: Number of Factions

Internal Polarization 1.777*** 1.804*** 1.802***
(0.294) (0.266) (0.264)
Disproportional Rule –0.472*** –0.407*** –0.394***
(0.103) (0.111) (0.109)
Factions Ban 0.039 0.014 0.001
(0.110) (0.111) (0.107)
Direct Election 0.390 0.448 0.465
(0.365) (0.324) (0.325)
Party Size 0.019** 0.019*** 0.019***
(0.008) (0.007) (0.007)
Open List 0.509 0.506* 0.507*
(0.304) (0.283) (0.281)
Party in Office –0.358 –0.349* –0.347*
(0.209) (0.206) (0.205)
Extra-parliamentary 0.037 0.012 0.012
(0.395) (0.347) (0.343)
Constant 1.827*** 2.037*** 2.048***
(0.244) (0.189) (0.190)
Selection Equation DV: Contested Congresses

Disproportional Rule – –0.279 –0.330
(0.216) (0.218)
Factions Ban – 0.125 0.184
(0.226) (0.209)
Democratic Centralism – –1.955*** –1.982***
(0.233) (0.230)
Direct Election – −0.336 –0.398
(0.460) (0.458)
Personal Party – – –5.914***
(0.387)
Years Before Election – 0.170** 0.167**
(0.076) (0.072)
Constant – 0.077 0.026
(0.386) (0.384)
ρ – –0.439*** –0.435***
(0.108) (0.111)
R2 0.41 – –
Log-pseudo likelihood – –194.20 –198.73
N (Congresses) 83 198 221
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. Clustered standard errors in parentheses.
the expanded dataset. The Heckman model has been estimated through
Maximum Likelihood; observations have been clustered by party and standard
errors are estimated on cluster to cope with the issue of the independence of
observations (repeated congresses within the same party).
Let us start from the results of the Heckman model on the selection equation.
The ρ is statistically different from zero meaning that there is sample selection
and the coefficients of the OLS could be biased. The results of the first stage
suggest that Democratic Centralism is the only variable related to party rules that
affect the outcome, decreasing the likelihood of observing a contested congress.
Beside it, YBE has an effect too. When YBE increases, the likelihood of a con-
tested congress grows, whereas in the case of looming elections contested con-
gresses become less likely.
In the extended dataset (Model 3), only one additional variable seems to
matter: Personal Party decreases the likelihood of a contested congress. To the
contrary, Disproportional Rule, Factions Ban and Direct Election are never
significant.
Turning to the results of the outcome equation, Internal Polarization has a
positive effect on NOF. In line with the expectation, the greater the range of
policy views inside the party, the higher the number of factions. With respect to
intra-party rules, Disproportional Rule is significant and decreases the number of
factions, while Factions Ban and Direct Election do not have effects.
Moving to the other hypotheses, the coefficient of Party Size is positive and
significant, suggesting that parties with more resources are more factionalized.
In turn, electoral systems based on Open List contribute to increasing NOF. Sur-
prisingly, Party in Office is negative and significant. In ruling parties, the number
of factions is lower, contradicting the idea that a larger share of spoils fosters
factionalism.
The determinants of factionalism in France and Germany

As a next step, the previous analysis will be replicated on French data. Due to
the limited number of observations, however, I rely on a simple OLS model to
test all the main hypotheses. Notice that, in the Italian case, Democratic Central-
ism was the main variable influencing the likelihood of a contested congress and
causing a selection bias in the sample. In France, Democratic Centralism affects
the outcome too given that there were not contested congresses in the PCF
before the abolishment of Democratic Centralism.
The dependent variable is the NOF, which on average is equal to 3.54
(standard deviation: 2.37). Interestingly, in the 34 percent of congresses there are
five or six factions. Factionalism reaches its peak in the congresses held by the
Greens in 2006 (eight factions), while seven factions contested the PS congress
in 1979 and 1990. In turn, NOF is highly correlated with NEF (0.85). The
independent variables are built exactly as in the Italian case, apart from two
minor differences: Disproportional Rule takes the value of 0 for PR with low
(5 percent) or null threshold and the value of 1 otherwise; Factions Ban is a
dummy variable equal to 1 if there is any type of ban on factions or tendencies.
Table 3.3 reports the results of the analysis.10 Standard errors have been clus-
tered on parties.
There are many similarities in the results related to France and Italy. To start
with, Internal Polarization has a positive and statistically significant effect on
NOF; in line with H3.1A, the results confirm that stronger polarization is associ-
ated with a higher level of factionalism. Party in Office is once again statistically
significant. In both countries, though, its effect is not in line with the hypothesis
on the availability of resources. YBE has a positive effect on NOF, suggesting
that in presence of looming elections lower fragmentation is displayed.
As far as rules are concerned, the results confirm that Factions Ban and Direct
Election have no effects in both countries.11 Contrary to the previous finding,
Disproportional Rule does not affect NOF. Notice, however, that the variation in
terms of internal electoral systems was greater in the Italian context while in
France cross-party differences tend to shrink. Furthermore, if excluding the three
MoDem congresses from the analysis, the effect of Disproportional Rule would
be negative and significant as expected. Finally, Party Size has no effects too.
Table 3.3 The determinants of factionalism in France and Germany
Dependent Variable: France Germany
Number of Factions Congress Fragmentation
Internal Polarization 2.305** 5.338**

(0.695) (2.527)
Disproportional Rule 0.090 –
(0.856)
Factions Ban 0.041 –
(1.320)
Direct Election 0.284 13.160***
(0.901) (2.296)
Party Size 0.003 –0.223
(0.043) (0.251)
Party in Office –1.669*** 3.530
(0.283) (2.971)
Years Before Election 0.508*** 3.366**
(0.101) (1.374)
SPD – 6.370**
(2.605)
Constant 1.389 4.084
(1.936) (8.689)
R2 0.43 0.36
N 35 40
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01.
In the German case, it is not possible to focus on the number of factions.
Accordingly, another indicator (still related to the party congress) will be used.
This refers to the votes casted therein, concerning the leadership selection
process. German parties tend to avoid open leadership contests (Cross 2013) and
usually select their leaders with high levels of support. In this regard, I argue that
any division concerning the selection of the party leader can be informative on
internal fragmentation. Rather than focusing on whether the congress is con-
tested by multiple candidates (which occurs only twice in the data), I consider
the votes share won by the main frontrunner candidate and I take any defection
as a sign of dissent and fragmentation.
The more the votes share won by the leader is far away from the unanimity
(100 percent), the more the party can be considered internally factionalized.
As such, I create a different dependent variable, Congress Fragmentation,
which is equal to the gap between the unanimity and the vote share actually
won by the party leader. Higher values indicate that the party leader was
elected with a lower support and suggest that the party is internally frag-
mented. Congress Fragmentation ranges from 0 (full unity, valued observe in
the SPD 2017 congress) to 37.4 (highest fragmentation, value reached by the
SPD in the 1995 congress). On average, Congress Fragmentation is equal to
10.20 (standard deviation: 8.73), though in two-thirds of cases its value is
lower than the mean.
The following independent variables have been included in the analysis.
Internal Polarization, Direct Election, Party Size, Party in Office and YBE have
been operationalized as described above. Direct Election takes the value of 1
only in one congress out of 40. Given that this is the only variable related to
intra-party rules suitable and available in the German case I decided to keep it
inside the model. Notice, however, that excluding this variable does not alter any
other result. The results of the analysis are displayed in the last column of Table
3.3. Data have been analyzed through an OLS. Party dummies are included in
the model. The dependent variable is theoretically bounded between 0 percent
and 100 percent, therefore a fractional logit (Papke & Wooldridge 1996) could
be better suitable to test the hypotheses. This choice, however, does not alter the
results and I stick to the OLS to simplify the content of Table 3.3.
Once again, the analysis confirms that Internal Polarization boosts the
fragmentation in intra-party elections and this holds true in all three countries. In
turn, YBE seems important as well and it has an impact on internal fragmen-
tation in Italy, France and Germany.12 When YBE increases, Congress Fragmen-
tation grows suggesting that the candidate proposed by the leadership is
somehow contested by party activists and delegates; conversely, when there are
looming elections, the fragmentation in the leadership vote drops and stronger
union can be observed.
The effect of Party Size and Party in Office is not statistically significant.
Finally, the only variable related to party rules seems to matter. Although related
to a single episode, the consultative Direct Election is associated with an increase
in fragmentation (notice that the sign of the coefficient is the same in Italy and
France with respect to NOF, even though in those countries, when more vari-
ation was available, this variable did not reach statistically significance).
Factionalism and voice: findings

In this chapter I focused on the causes of factionalism to determine what ele-
ments increase or decrease the number of factions and the likelihood of observ-
ing a contested congress. Three different analyses have been performed. In the
Italian case, thanks to data availability, I managed to run a Heckman model to
deal with any possible bias in sample selection. By doing that, I explain both the
likelihood of observing a contested congress and the magnitude of factionalism.
In the French case, the determinants of the number of factions were investigated,
while in the German case the focus was on the fragmentation that emerges in the
leadership race during the congress.
In all three countries a number of hypotheses have been tested, focusing in
particular on internal polarization (measured according to the data of Chapter 2),
on the electoral cycle (assessing the number of years before the next election),
and on an array of intra-party rules, particularly in Italy and France where more
variation in the party organizational settings can be found.
Starting from the latter, I tested whether intra-party rules make the difference
in terms of factionalism. The results reveal that some rules play a role while
others do not affect the expression or the magnitude of intra-party dissent.
First of all, Democratic Centralism (H3.4B) has a clear effect on intra-party
dynamics and this is the main attribute associated with a potential selection bias.
Such undemocratic rule prevents dissenters from contesting the congresses. This
come as no surprise given that democratic centralism, by definition, forbids any
public expression of dissent. For instance, in Italy the PCI had to change the
statute at the end of the 1980s to allow contested congresses and to involve
internal minorities in the decision-making process concerning the structural
changes of party ideology. Similarly, the PCF in France held contested con-
gresses only after 1994 when the democratic centralism was abolished.13
To the contrary, an explicit ban on the existence of organized factions has no
impact on factionalism both in Italy and France. Factions Ban has no effects
either on the number of factions (H3.3A) or on the likelihood of a contested con-
gress (H.3.3B). In Germany too, the party’s refusal to officially recognize new
subgroups in 2018 did not prevent new ideological factions from forming. The
ineffectiveness of the factional ban is not completely unexpected. First of all,
factions can take advantage of their unclear conceptualization to survive the ban.
Even the most heavily organized factions often do not call themselves
“faction.”14 Instead of faction, other names are usually used including “associ-
ation,” “foundation,” “area” and so on (Venditti 1981). Factions can thus hide
themselves behind the curtain of associational pluralism.
In addition, inside the party the policy-maker coincides with the policy-taker
(Zincone 1972). Party members might decide which rule to enact and which not.
It is more difficult to skip a stricter rule (e.g., Democratic Centralism) compared
to other rules linked with faint concepts, such as the ban on factions. Indeed, the
ban does not assure that factions dissolve and such prohibition was often
ignored: DC factions persisted despite the ban and the same happened in other
parties. To quote an Italian saying, “capons do not celebrate Christmas”; factions
often agreed on banning themselves in order to promote internal cohesion but,
following a prisoners’ dilemma logic, none of them was the first to demobilize,
so that such rule was hardly ever applied.
Direct Election does not affect factionalism (H3.5A; H3.5B). A directly
elected leader could be stronger and more autonomous from party factions, but
the opponents can still opt to challenge the leadership and contest the congress.
In fact, the only episode of direct election in Germany was associated with a
stronger fragmentation, though I did not find any systematic evidence in this
regard. In fact, from a theoretical point of view, the party can be factionalized
even in the presence of autonomous leaders, particularly when the leader is
focused on unity or the minorities have a strong and credible exit strategy (see
Chapter 1).
Based on my theoretical framework, the only scenario in which I expected
and found a leadership effect on factionalism is related to the variable Personal
Party (H3.6B). These parties represent the most extreme case of autonomous
leaders, given that the leader can be self-enthroned (Kenig 2009). Often the
leader is not formally voted by delegates but just elected by acclamation, which
can be in a sense an extreme form of direct election. Furthermore, the party
heavily depends on his charisma being strongly tied and loyal to him. In this
context, the party leader is usually focused on cohesion (ν) rather than unity (ω),
and this frustrates any attempt to use the voice option. Furthermore, due to the
leader’s charisma, in personal parties the exit costs (ε) increase, producing
negative effects on the public expression of dissent and on the likelihood of
contested congresses. This is in line with the empirical findings of the Heckman
model.
Notice that personal parties are not only less likely to hold contested con-
gresses but also often they do not celebrate congresses at all. For instance the
LN, that for ages has been firmly loyal to its leader Umberto Bossi, has not held
a national congress from 2002 until 2012, when Bossi leadership was overthrew
due to a corruption scandal and Roberto Maroni became the new party leader; FI
(Berlusconi’s party) celebrated only two (uncontested) congresses in 15 years;
IDV, a party created by the former attorney Antonio Di Pietro (famous after
Tangentopoli scandal), held its first congress in 2010, 12 years after its forma-
tion; despite the existence of some forms of internal dissent its minority faction
(called La Base, “rank-and-file”) failed to present its own motion and the con-
gress has been uncontested. These examples confirm that when leaders are
stronger, factions remain latent and the leader’s power produces unitary
dynamics (at least apparently).15
The last rule examined concerns the impact of the internal electoral system.
Disproportional Rule does not seem to affect the likelihood of a contested con-
gress (H3.2B). Conversely, in the Italian case it seems to affect factionalism
(H3.2A) by reducing the number of factions (in France this same effect holds
only if excluding the MoDem congresses). Indeed, disproportional rules reduce
the likelihood that small factions will be represented in the party body. Accord-
ingly, tiny groups have an incentive to merge in order to contest the congress
together, enhancing the opportunity to overcome the threshold or to win the
majority prize. Arguably, factions might feel free to divide soon after the con-
gress (and in some cases this happened indeed). Anyway, if factions contested
the congress together, signing a common motion, they start to be perceived as a
unitary actor and breaking the factional alliance might not be so easy. In fact, it
can be costly, at least in terms of credibility.
The role of the internal electoral system, confirmed by the empirical analysis,
is strongly emphasized when looking at real-world politics. As discussed earlier
in this chapter, inside the DC, in Italy, and inside the PS, in France, left-wing
non-mainstream factions heavily fought to modify the electoral rules in order to
get a PR. These minorities deemed such reform crucial to enhance their bargain-
ing power. In the case of the PS there is no counterfactual given that PR has
always been the rule. Conversely, inside the DC the PR heightened internal
fragmentation. In this regard, it seems that Disproportional Rule does succeed in
containing factionalism.
Going beyond formal rules, two other elements strongly affect factionalism in
all the three countries. First, the heterogeneity of policy preferences matters
(H3.1A). Internal Polarization increases the number of factions in Italy and
France, and fosters congress fragmentation in Germany. When the congress is
contested, divergent policy preferences lead to a higher number of factions that
run in the internal elections or decrease the votes share won by the party leader
proposed by the mainstream.
In turn, the electoral cycle seems important (H3.10B). YBE has a positive
effect on internal fragmentation in Italy, France and Germany. The higher the
number of years remaining before the next general election, the higher the incen-
tives to divide, express disagreement and cultivate conflictual dynamics. In the
Italian case, the likelihood of a contested congress is lower in presence of a
looming election. In France, this generates a decrease in the number of factions,
while in Germany looming elections reduce the Congress Fragmentation favor-
ing internal union and boosting the support for the mainstream leadership.
It could be argued that looming elections should raise the breakup cost, and
therefore one should expect higher factionalism. However, due to the cost of
voice, parties can try to avoid the display of public disagreement in such context.
Factionalism might increase, but only behind closed doors, so that voters do not
observe it (and behind those closed doors, leaders will more likely try to find a
compromise with these latent factions, unless the leader is focused more on
cohesion, see Chapters 4, 5 and 6). There are indeed some examples suggesting
that this can be the case and factionalism can be strong but hidden. For instance,
in the French PS, seven factions were active in the 1979 and 1990 congresses. A
few months later, in the following party congresses held just before the 1981
Presidential elections and the 1993 Legislative elections, these seven factions
apparently disappeared, and no division was shown. It is hard to believe that
perfect cohesion was reached in such short a time period; more likely, disagree-
ment was still there, though hidden and confined (for strategic electoral reasons)
behind the closed doors of the party headquarters.
Moving to the last hypotheses, in the Italian case the availability of resources
(H3.7A) increases factionalism, though in France and Germany no effect was
found. When it comes to H3.9A, Party in Office decreased NOF both in France
and Italy (contrary to the expectation). In fact, after controlling for rules, prefer-
ences and size, the total number of subgroups is lower in ruling parties.
Why did this happen? Party unity is a source of bargaining power that
increases the party’s appeal as a potential partner of a coalition government
(Ceron 2016). As such, less divided parties are more likely to get in office
(Warwick 1996). Once in office, they will face even more incentives to keep
their unity in order to exploit patronage and to improve government effective-
ness. The cost of division might increase because the life of ruling parties is
more under public scrutiny compared to opposition parties. Accordingly, and in
line with the theoretical model, in ruling parties the cost of voice (ν) might
increase, producing a decrease in factionalism (minorities will be less prone to
resort to the voice option). These elements seem to explain why the data show
that ruling parties are less factionalized.
Finally, the Italian data confirm that an Open List electoral system (H 3.8A)
provides incentives for factionalism. The open list PR grants a wider number of
factions the opportunity to gain parliamentary seats allowing them to flourish.
To conclude, the empirical analyses confirm that heterogeneous policy prefer-
ences are a source of factionalism. The electoral cycle matters too, boosting the
importance of party unity and creating incentives to hide disagreements. A
similar effect seems to happen to parties in office, which are less factionalized
due to the growing incentives for unity and cohesion.
Conversely, the importance of intra-party rules seems more questionable.
Formal rules play a role when they are strictly enacted. To some extent, dispro-
portional internal electoral systems seem to work according to the Duverger’s
Laws, containing factionalism.
Overall, less democratic party rules (i.e., democratic centralism) or less demo-
cratic party structures (e.g., those of personal parties), increase the cost of voice
and exert constraints on the expression of internal dissent. In fact, there is not
much to say about internal disagreement in these parties simply because it is
often impossible to observe it. Although heterogeneity of preferences might
exist, such party organization successfully hides internal struggles; here the con-
flict could probably be observed only when it erupts producing visible con-
sequences such as party splits. With respect to party organization, if a selection
bias exists, this is limited to parties organized according to such less democratic
internal rules. In view of that, caution is needed when discussing the generaliza-
bility of the results presented in the rest of the book to parties that are not char-
acterized by intra-party competition based on organized factions and congress
motions.
Notes
1 The first DC faction, Politiche Sociali, was named after the homonym review; the DC
left-wing faction Base published a periodical called exactly “rank-and-file”; within
the MSI the left-wing faction linked to Rauti changed its name from Linea Futura to
Spazio Nuovo and modified its magazine accordingly. In turn, the PCI subgroup
Manifesto (which was promptly charged with factionalism and expelled from the
party), took its name from the monthly review published in 1969. More recently, other
factions organized their own media, such as “Aprile,” the printed review linked with
the DS left-wing faction Correntone, or “Area” that expressed the views of the pro-
welfare right, Destra Sociale, within the AN. In recent years, the Web has decreased
the cost of publishing periodicals and Aprile soon became issued only via the internet
(Aprile on-line); similarly, the Finiani faction published online the review “Fare
Futuro Web Magazine.”
2 A number of studies have shown that large dominant parties are highly factionalized
(e.g., Bogaards & Boucek 2010).
3 I consider the direct election of party leader as a rule that favors the mainstream in
spite of internal minorities. See Rahat et al. (2008) for a similar argument on how an
inclusive selectorate might decrease the representativeness.
4 These factions could be defined as “factions of principles” because only factions that
contested the congress presenting a motion are considered here. Nonetheless, as
already discussed, all factions care to some extent about policy. Although the dataset
includes factions that Sartori (1976) would label as “factions of interests,” these are
also “faction of principles,” at least in part.
5 This operationalization is in line with Sartori (1976: 89).
6 To reach a convergence the Heckman model usually requires the inclusion of at least
one variable that affects z without affecting y. In the first stage I focus only on the role
of intra-party rules, then variables such as Party Size, Open List or Internal Polariza-
tion (which can be measured only when y is observed) have been excluded. In the
second stage, I decided to discard Democratic Centralism as well as YBE (among the
four variables included in the first stage). Indeed, based on my theoretical framework
I expect that Democratic Centralism has an impact only on the nature of the congress
itself; there are in fact only two contested congresses held under this rule and I do not
expect to find any particular effect of such variable on NOF.
7 Arguably, Party Size is only a proxy for party resources. Two alternative variables
could be used: the number of party members or the financial budget (i.e., the party
revenues). I gathered these data (although there is a lot of missing information). In the
dataset with 198 congresses, Party Size is positively correlated with membership
(0.85) and revenues (0.90), and these correlations are statistically significant. This
confirms that Party Size is a suitable proxy for the amount of financial resources.
8 Except one contested congress of the PPI (for which I did not find textual motions),
all the other congresses were not contested and therefore I do not consider them in the
rest of the book.
9 All the other variables remain the same as before, except Democratic Centralism that
now takes an intermediate value (0.5) for the LN and PDL as far as their party struc-
ture, although not officially based on Democratic Centralism, bears many resem-
blances with it, forbidding any public expression of dissent. Notice that this
operationalization does not affect the results.
10 The number of parameters is quite high if compared to the number of observations.
However, even when reducing the number of independent variables included in the
model all the results hold the same.
11 Interestingly, when interacting Direct Election with the number of years remaining
before the next Presidential elections, one would notice that autonomous leaders
reduce factionalism as new elections approach.
12 Notice that YBE would carry a positive effect on NOF also in the Italian case if this
variable was included in Model 1, Table 3.2. However, in Model 2 and 3, once YBE
is added to the selection equation, its effect on NOF vanishes. Nevertheless, to some
extent YBE impacts on factionalism by reducing the likelihood of a contested
congress.
13 Notice that democratic centralism, in theory, allows intra-party struggles (unless they
lead to factionalism). In fact, parties that adopt democratic centralism may hold a con-
tested congress (this happened twice: PDCI 2008 and PSI 1949).
14 In 2008, during the PD leadership of Veltroni, his main opponent, D’Alema, created
the association RED, and claimed that RED was not a faction but a resource for the
party. One year later Pierluigi Bersani, member of RED, was appointed as the new
party leader. In October 2011, during a meeting of the PD minority faction MoDem,
Veltroni, the head of this new subgroup, stated that the MoDem was not a faction
either. The claim was made by Renzi with respect to the Renziani faction.
15 Actually divisions can be observed, though these might remain latent so that politicians
claim that their parties are not factionalized. Within the LN a new tendency, Maroniani,
arose in 2011 gathering the followers of Roberto Maroni, who was the Minister of the
Interior at that time. After a failed leadership attempt to whip this faction, the Maroniani
managed to contest many local LN congresses presenting its own candidates in opposi-
tion to the party mainstream led by the founding father Bossi and informally labeled
Cerchio Magico (Magic Circle). Later on, the two recent congresses held by the LN in
2013 and 2017 have been contested by multiple candidates presenting rival platforms.
Matteo Salvini won both congresses against the candidates of the faction loyal to Bossi:
Bossi himself, in 2013, and Giovanni Fava, in 2017. Despite the absence of organized
factions, there were some divisions inside the PDL too: apart from the already men-
tioned case of Finiani, in 2010–2011 the followers of the former Minister of Economic
Development, Claudio Scajola, organized and built the Cristoforo Colombo foundation
that represented a kind of faction composed mainly by former DC politicians. In order to
force the Prime Minister Berlusconi to resign, a few MPs belonging to this faction
switched off from the PDL in autumn 2011 and created a small parliamentary group
affiliated with the new Italian Liberal Party. Inside the PDL there were also other associ-
ations and foundations like Riformismo E Libertà that includes former PSI politicians,
the conservative Magna Carta foundation, the network Rete Italia (tied to the Catholic
group Comunione e Liberazione), the right-wing Destra-PDL, Nuova Italia foundation,
as well as many others. Even in the latest legislatures, the PDL–FI faced a proliferation
of subgroups composed of the followers of Angelino Alfano (Alfaniani), Raffaele Fitto
(Fittiani) and Denis Verdini (Verdiniani), which split from the party in 2013 and 2015
(Ceron 2017a, 2017b); these factions, however, did not publicly manifest themselves in
a national congress.
References
Astudillo, J., & Detterbeck, K. (2018). Why, Sometimes, Primaries? Intraparty Democrat-
ization as a Default Selection Mechanism in German and Spanish Mainstream Parties.
Party Politics, https://doi.org/10.1177/1354068818795195.
Bille, L. (2001). Democratizing a Democratic Procedure: Myth or Reality?: Candidate
Selection in Western European Parties, 1960–1990. Party Politics, 7(3), 363–380.
Bogaards, M., & Boucek, F. (eds) (2010). Dominant Political Parties and Democracy:
Concepts, Measures, Cases and Comparisons. Routledge/ECPR Studies in European
Political Science, London.
Science, London.
Ceron, A. (2011). Correnti e frazionismo nei partiti politici italiani (1946–2010):
un’analisi quantitativa delle mozioni congressuali. Rivista Italiana di Scienza Politica,
41(2), 237–264.
246–258.
Ceron, A. (2017a). Intra-party Politics in 140 Characters. Party Politics, 23(1): 7–17.
491–508.
Cox, G.W., & Rosenbluth, F. (1993). The Electoral Fortunes of Legislative Factions in
Japan. American Political Science Review, 87(3), 577–589.
Science, 29(1), 33–56.
Cross W.P. (2013). Party Leadership and Intra-party Democracy. In Cross, W.P., & Katz,
R.S. (eds), The Challenges of Intra-Party Democracy, 100–116, Oxford University
Press, Oxford.
Emanuele, V., Vincenzo, E., Marino, B., & Martocchia Diodati, N. (2018). The Hidden
Side of Parties: The Effective Number of Factions in Western Europe (1965–2015) and
its Determinants. Paper presented at the Annual Meeting of the Party Congress
Research Group, Vienna, June 25–26, 2018
588–622.
Heckman, J.J. (1979). Sample Selection Bias as a Specification Error. Econometrica,
47(1), 153–161.
Katz, R.S., & Bardi, L. (1980). Preference Voting and Turnover in Italian Parliamentary
Elections. American Journal of Political Science, 24(1), 97–114.
König, T. (2006). The Scope for Policy Change after the 2005 Election: Veto Players and
Intra-party Decision-making. German Politics, 15(4), 520–532.
Laasko, M., & Taagepera, R. (1979). Effective Number of Parties: A Measure with
Application to West Europe. Comparative Political Studies, 12(1), 3–27.
Papke L.E., & Wooldridge J.M. (1996). Econometric Methods for Fractional Response
Variables with an Application to 401(K) Plan Participation Rates. Journal of Applied
Econometrics, 11(6), 619–632.
Pasquino, G. (1972). Le radici del frazionismo e il voto di preferenza. Rivista Italiana di
Politics, 14(6), 663–683.
Raniolo, F. (2006). Forza Italia: A Leader with a Party. South European Society and Pol-
itics, 11(3), 439–455.
Reiter, H.L. (1981). Intra-party Cleavages in the United States Today. Western Political
Quarterly, 34(2), 287–300.
Rogers, W.H. (1993). Regression Standard Errors in Clustered Samples. Stata Technica
Bulletin, 3(13), 19–23.
Sartori, G. (1971). Proporzionalismo, frazionismo e crisi dei partiti. Rivista Italiana di
Shugart, M.S. (2001). Electoral Efficiency and the Move to Mixed-member Systems.
Electoral Studies, 20(2), 173–193.
Spies, D.C., & Kaiser, A. (2014) Does the Mode of Candidate Selection Affect the Rep-
resentativeness of Parties? Party Politics, 20(4), 576–590.
Venditti, R. (1981). Il manuale Cencelli. Il prontuario della lottizzazione democristiana.
Un documento sulla gestione del potere. Editori Riuniti, Roma.
Warwick, P.V. (1996). Coalition Government Membership in West European Parlia-
mentary Democracies. British Journal of Political Science, 26(4), 471–499.
Yale University Press, New Haven.
4 Bounded oligarchy
Leaders, factions and the selection of
party platform
Introduction
In the previous chapter I investigated the display of public disagreement (voice)
while this chapter and the following one will be devoted to understanding intra-
party dynamics concerning the allocation of policy and office payoffs.
To start with, I consider the process of selection of the party platform looking
at how the leader sets the position of the party in key parliamentary debates (e.g.,
investiture debates) or in electoral manifestos (Ceron & Greene 2019).
Considering the process of party position-taking as the distribution of policy
payoffs within the party, I argue that the party position (publicly communicated
in manifestos or in highly visible legislative speeches) corresponds to the amount
of payoffs to be shared and each faction will try to affect this process minimiz-
ing the distance between the party platform and its ideal point.
By doing that I investigate intra-party dynamics focusing on how factions
bargain over policy payoffs. I assess when a “compromise” occurs and when not,
evaluating whether and to what extent factions bind the party leader’s choice of
party platform and limit the possibility to shift position from one election to the
following one.
Following the development of the spatial theory of voting (Downs 1957), a
large branch of literature started to address the question of how parties set their
positions. In a two-party competition along a single dimension, Downs’ theory
predicts the well-known result of convergence toward the median voter. This
result, however, relies on strict assumptions that hardly ever occur in real pol-
itics (e.g., Grofman 2004). Relaxing some of these assumptions leads to a very
different equilibrium, with parties moving away from the median voter’s ideal
point. For instance, in the presence of such phenomena as party identification,
voter abstention, or mass parties bound by their activists, the centripetal conver-
gence no longer holds.
Nonetheless, recent developments in the theory of voting provide new models
that foresee a centrifugal equilibrium, and they seem to find empirical confirm-
ation (e.g., Adams et al. 2005). In a multiparty system parties are often bound by
the preferences of their activists, factions, and members (Aldrich 1983; Strøm
1990; Ware 1992). Therefore, to avoid loss of votes, they should diverge from
90 The selection of party platform
the median voter in the direction of the rank-and-file. Vote-maximizing posi-
tions, however, tend to be more moderate when compared with those of party
members.
In more detail, I will try to assess whether factions have an impact on party
position (Aldrich 1983; McGann 2002; Strøm 1990; Ware 1992) or if, alterna-
tively, parties try to get rid of factional preferences in order to assume vote-
maximizing positions (Downs 1957). I will also investigate whether factions
bargain according to some sort of proportionality norm such as Gamson’s Law
(applied to the allocation of policy payoffs) or whether some actors (e.g., the
median faction) can extract larger revenues. By modeling party placement as the
result of inter-factional competition, I will highlight how and under what cir-
cumstances factions shape the party position, binding the leader in the choice of
the platform. Furthermore, I will assess how party organization affects this
outcome, analyzing whether different internal rules alter the equilibrium between
factions and leaders. In this way I will provide new insights about the linkage
between party leaders, members and activists.
Who sets the party position: literature and hypotheses

The wide literature on party preferences investigates how parties select their policy
position and focuses mainly on two facets. One stream of research based on party
competition examines the interaction among political parties to identify the exist-
ence of stable equilibriums. In this regard, the seminal Downs’ (1957) spatial
theory shows that in a two-party system such equilibrium consists of a centripetal
convergence of actors toward the median voter. This result, however, relies upon
multiple assumptions concerning party system features as well as motivations
behind the behavior of political actors. As a consequence, scholars have tried to
relax those assumptions taking into account some aspects of “real-world politics”
and provided more realistic models that better fit the empirical evidence.
For our purposes, I focus attention on those concerning multiparty systems
(such as Italy, France and Germany) and particularly on the “unified theory of
party competition” (Adams et al. 2005). This model argues that parties are
pushed away from the center, in the direction of those voters who feel identified
with them, producing a centrifugal equilibrium. Although parties maintain a link
with their partisans’ preferences, this model predicts that parties maximize votes,
adopting a slightly moderate position with respect to their followers.1
Another stream of research considers the selection of the party platform more
as the endogenous output of an inward competition within the party (Budge et al.
2010; Levy 2004), rather than the exogenous product of electoral competition.
Based on the idea that parties are somehow bounded by intra-party structure,
scholars investigate the effect of organization and internal rules, on one hand,
and that of divergent intra-party preferences, on the other. Both elements can
affect party position and its shifts over time.
The present chapter takes the cue from this second stream of research, though
the findings will also be useful for the literature on party competition. In more
The selection of party platform 91
detail, I will discuss alternative arguments that concern factional preferences,
intra-party organization, and the pressure for a (partial) convergence due to party
competition.
Arguably, parties are not unitary actors and in this regard it is straightforward
to expect that members and factions can play a role in shaping the structure of
the party as well as its policy preferences. Like any other political actor, intra-
party subgroups are interested in policy, office and votes (Müller & Strøm 1999).
The interests of one subgroup may coincide or not with those of other internal
actors so that intra-party competition occurs, generating conflict or cooperation,
and shaping the allocation of policy and office payoffs.
Let us start from the allocation of policy payoffs with reference to the selec-
tion of the party platform. It has been shown that policy preferences tend to
structure factional membership (Bernauer & Bräuninger 2009; Giannetti &
Laver 2009) and inside a party one can therefore observe different policy views.
When divergent preferences exist, the final party platform will depend on the
nature and strength of these preferences as well as on how internal organizations
and rules aggregate the preferences of members and factions. In divided parties
the internal equilibrium comes from a factional strife (Giannetti & Mulé 2006:
462; Levy 2004) and factions bargain to influence the party strategy according to
their strength (Laver & Shepsle 1990: 504).
Depending on the internal decision-making process, on the party’s organiza-
tional structure and on the model of party that is taken into account, the factional
competition might follow different paths leading to different outcomes (i.e.,
party platforms).
Parties can be (entirely or partially) bound by their factions, or alternatively,
they can be absolutely free to set their platform. In the first case party position
should lie somewhere inside the Pareto set of factions’ positions. According to
Levy (2004: 251), “parties can offer to voters any policy in the Pareto set of their
members.” To the contrary, “Parties cannot commit to offer any policy outside
the Pareto set but the party members can find mechanisms (such as bargaining)
that allow them to choose policies within the Pareto set” (in turn, this agreement
will be enforceable and stable). Any point inside this range, then, could be linked
with the final party platform. The actual location of the party along the left–right
continuum (or any more specific policy scale),2 however, will depend on the
bargaining power of each faction.
Such bargaining power can be related to the size of factions, measured as the
share of votes gained during the congress, or the share of seats inside party body
(but even as the share of MPs affiliated to a faction within the parliamentary
party group, which can be the most suitable measurement in the German case).
As Budge et al. (2010: 793) point out, “The stronger one faction is relative to the
others the more it overcomes resistance and carries its preferred policy further.”
Inside the Pareto set, two points carry a theoretically substantial meaning that
depends on the structure of intra-party dynamics.
To start with, for the sake of preserving unity, inter-factional competition may
generate a consensual environment where all faction preferences are somehow
taken into account (Budge et al. 2010; Giannetti & Laver 2009). At the level of
legislative party group Heller and Mershon (2009) hypothesized that the party
ideal point should be close to the average ideal point of legislative members.
Extending this idea to the party body, it can be argued that “the mechanisms in
which Western European parties reach internal compromise do mimic some
form of a weighted average of the ideal policies of their factions” (Levy 2004:
251). Accordingly, one would expect that policy payoffs (i.e., party position) are
distributed following a kind of Gamson rule (Gamson 1961; Warwick 2001) and
the party platform is consensually selected based on a marked degree of
proportionality.3
To sum up, if factions are able to affect the party’s decision-making accord-
ing to their reciprocal strength in the party body (due to the results of party con-
gress/convention), the final party position should be closely related to the
“Gamsonian agreement,” i.e., the mean of factions’ positions weighted by the
size of each faction. Such equilibrium would be the best predictor of party posi-
tion and would bound the leader in the choice of platform and strategy (remark-
ably, the strong emphasis on the link between party body and party strategy is
coherent with the concept of mass party).
Hypothesis 4.1A (H4.1A): The factions’ preferences affect party position pro-
portionally to the strength of each subgroup, so that a “Gamsonian agreement”
position is strongly related to the party’s ideal point.
It has been argued that when the electoral system provides incentives for
inter-factional competition, the common interests in partisan unity are insuffi-
cient to prevent internal dissension so that factions divide during the electoral
campaign (Morgenstern 2001).4 To the contrary, I claim that parties, to keep
their unity, cater to all factions according to the latter’s share of votes in the con-
gress body. Party unity is a key source of party strength, particularly in the elect-
oral arena (McGann 2002; Snyder & Ting 2002). Accordingly, the importance
of party unity should increase as general elections approach. In this regard,
scholars have stated that party manifestos take into consideration the concerns of
both mainstream and minority factions precisely to boost party unity (and cred-
ibility) before elections (e.g., Levy 2004). If this is the case, one should observe
a tighter link between party position and the “Gamsonian agreement” as the
legislative term comes to an end.
Hypothesis 4.1B (H4.1B): As time passes and the new general elections
approach, the effect of the “Gamsonian agreement” on party position should be
stronger.
Alternatively, inter-factional conflict might generate a “winner-takes-all”

competition, with the median faction (mainstream) free to set its ideal point as
the overall party position. The median faction can also suddenly change party
platform, selecting a point far away from the Pareto set for strategic or
contingent reasons, though overall the party position should be closer to the
mainstream’s ideal point.
The median faction, in fact, still has a commitment to party members, due to
the result of the party congress; strong deviations of its position from the ideo-
logical position of the bulk of party members could result in a loss of internal
consensus (for the mainstream) and in a change of the dominant faction in the
following party congress; minority factions could feel dissatisfied if the party is
ruled only according to the wills of the mainstream, and hence could exploit
their bargaining power by threatening to defect and to leave the party. Notice
that their bargaining power would be higher if the threat is credible and could
harm the party, resulting in a loss of votes, seats or strategic position. For all
these reasons, the mainstream will more likely select a platform located inside
the Pareto set.
To sum up, in a unidimensional policy space, the median voter theorem
applies, and the median faction’s position can be expected to emerge as the equi-
librium position.
To account for this, I want to check whether the party position is pushed in
the direction of the median faction, away from the inter-factional agreement
based on a proportionality norm.
Hypothesis 4.2 (H4.2): Party position is biased toward the ideal point of the
median faction.
Apart from these two clear-cut equilibriums, a third mixed pattern can arise if
parties are mildly bound by factions. The increasing personalization of politics
in modern democracies (Poguntke & Webb 2005) boosts the party leaders’
power in spite of the strength of factions. If the party leader, due to own
charisma or to the “dominant position” retained inside the party, becomes able to
set the new platform alone, one would expect to observe a deviation from the
inter-factional agreement in the direction of the leader’s interests. However,
party leaders are democratically elected; hence, the rank-and-file still retain an
influence over party position (Heller & Mershon 2009) given that the leader’s
first goal is to remain in charge. Nevertheless, as the agent delegated by party
members, the leader will have more discretionality in setting party position, and
will exploit such autonomy to get rid of middle-level activists and factions pro-
vided that members’ approval won’t be lost.
As far as the party leadership is concerned, it is worth taking into account that
different intra-party rules and organizational patterns affect internal dynamics
and provide the leader with greater or less power. This, in turn, affects the lead-
ership’s ability to deviate from the inter-factional agreement in the direction of
its preferred party position.
According to the cartel party model (Katz & Mair 1995), leaders want more
autonomy from activists’ factions in order to freely give rise to a cross-party
cartel (Katz 2001; Mair 1997). The leader aims to lower members’ stakes,
pushing the party toward a moderate position, thereby increasing both party
votes and the likelihood of being involved in government coalitions (Mair 1994;
Marsh 1993).
One strategy to enhance the leader’s autonomy is to empower ordinary party
members (or supporters), thus increasing the nominal level of intra-party demo-
cracy. Recent trends attest to an increase in the inclusiveness of party leader
selection mechanisms, and these trends are particularly evident among Italian
and French parties.
Historically based on a small committee (National Council or Central
Committee), the selection procedure has been extended to the whole audience of
delegates at party congress, and, recently, the selectorate has tended to include
all party members (and sometimes even unregistered supporters) through direct
election of the leader (by means of closed or open “primaries”). Italian mass
parties started to enlarge the selectorate at the end of the 1970s. In 1976 the elec-
tion of the DC leader was demanded of congress delegates instead of members
of the National Council; the PSI set the same rule in 1981, strengthening Craxi’s
personalized leadership (Ceron & Negri 2017).5
Something similar happened in France too, starting from the 1990s. For
instance, congress delegates of the PS directly appointed the party leader first in
1993 (Michel Rocard) and then in 1994 (Henri Emmanuelli). The rules changed
again in 1997 when, for the first time, the party leader was directly elected by
party members; Hollande was appointed by a wide majority (91.18 percent) and
started to move the party toward more moderate positions. Similarly, in 1998
Philip Séguin was re-elected leader of the Gaullist Party RPR by party members
through direct election with a wide support (95.07 percent), whereas only a few
months before, in 1997, he was nominated leader by the party committee (with a
lower amounts of votes: 78.85 percent).
Noticeably, the expanded inclusiveness of the leader’s selectorate does not
necessarily increase the rank-and-file’s control over the leadership but paradoxic-
ally can be a leader’s strategy to defang the base (Katz 2001; Mair 1994, 1997;
Marsh 1993; Poguntke & Webb 2005; Rahat el al. 2008). Inclusiveness produces
less tight contests and a decrease in competitiveness (Kenig 2009), complicating
the task of removing an unwanted leader.6 In addition, a directly elected party
leader retains a higher legitimacy and can get rid of middle-level activists and party
factions’ ties. As long as the leader’s power increases, the party should be less
bound by factions; the “Gamsonian agreement” would be less useful to predict the
party position, which instead would converge toward the center of the policy space
(in line with the idea suggested by the cartel party theory).
Hypothesis 4.3 (H4.3): When intra-party rules increase the autonomy of the
party leader, the “Gamsonian agreement” should be less (if at all) useful to
predict party position, which instead will be attracted by the center of policy
space.
Finally, if parties behave as catch-all actors (Kirchheimer 1966) they will

appeal to a broader public, seeking the support of groups that lie outside the
traditional party organization. As a consequence, the impact of factional con-
straints should decline so that party position would be independent from the
“Gamsonian agreement.” If this argument holds one would expect party posi-
tions to converge toward the center, thereby maximizing votes and resources,
regardless of the shape of intra-party preferences or the effect of party rules.
Hypothesis 4.4 (H4.4): Party position is independent from factions’ position

and, overall, will be attracted by the center of policy space no matter the shape
of intra-party rules.
The alternative expectations discussed so far represent roughly three different

models of party: mass, cartel and catch-all. Each expresses a fulfillment or a
response to the others (Katz 2001). I will compare them to assess under what
condition each one prevails, assessing whether parties are bound by internal fac-
tions (like in a mass party, H4.1) or are instead free to move (as suggested by the
catch-all model, H4.4). The third option lies somewhere in the between. In fact, the
rise of the cartel party (H4.3) is conditional on changes in party rules and leads to
different outputs depending on the shape of the intra-party environment.7
Comparing factions’ positions with the Italian legislative

speech database
To start with, I test these hypotheses on the Italian case, using data on 254
motions related to 83 congresses of 18 parties. I compare factional preferences
presented in Chapter 2 with the parties’ ideal points, employing a recently built
dataset, the Italian Legislative Speech Dataset (ILSD), which provides informa-
tion on the positions of Italian parties from 1946 to 2018 (Ceron et al. 2019;
Ceron & Negri 2017, 2018; Curini & Ceron 2013). These data are obtained by
hand-coding all the investiture debates of the Italian governments.8 Compared to
the CMP, ILSD retains some advantages. First of all, CMP does not distinguish
between positions of parties that run together in the election as a cartel but split
afterwards; second, it does not assess the positions of new parties that form
during the legislative term (i.e., due to party fission); finally, CMP provides only
one value per legislature, while ILSD tracks parties’ positions almost year by
year (due to the high level of government instability), increasing the number of
observations included in the analysis. This allows the positions of factions to be
compared with the first known value of party position expressed after the con-
gress (that is often temporally close to the congress); furthermore, it is possible
to compare the factional equilibrium reached during the party congress held at
time t with all the values of policy positions expressed by the party until the new
congress at time t + 1 is held. This allows changes to be modeled in intra-party
dynamics and bargaining over time; for instance, I can model the effect of party
fissions (occurring not immediately after the congress) over party position, and I
can assess whether bargaining dynamics vary as time passes (a feature that
cannot be tested in France and Germany).9
Party positions have been estimates applying the “vanilla method” (Gabel &
Huber 2000) to ILSD data. To do that, I ran a factor analysis on the percentage
of quasi-sentences extracting the first component; the standardized values of
party positions are located on the latent dimension that emerged from the
analysis.10 Vanilla estimates are strongly correlated with the traditional Left–
Right (RILE) scale measured on ILSD and with the results of the vanilla method
applied to CMP data.11
The two datasets that will be used in this section, i.e., ILSD and the database
on party factions, locate political actors on a traditional ideological left–right
scale; interestingly, both datasets cover the entire length of the continuum.
Indeed, in the first period the two most extreme parties according to the ILSD
Vanilla estimates are the PSIUP and the MSI; analogously, data on factional
motions concerning congresses of both the PSIUP and MSI are available and
Wordfish estimates locate these factions on the two endpoints of the scale. Sim-
ilarly, in the second period the two most radical parties according to the ILSD
are the PRC and MSI–AN and the factions of these two parties are, once again,
located on the wings of the left–right scale measured through Wordfish. As such,
I can claim that the left–right dimension is wholly covered by the data. However,
the range of the scales is different. To cope with this, I divided all the values for
the length of each continuum.
It is worth noticing that these two datasets come from two completely
different sources: the factions’ positions are extracted from debates that took
place in the party body, while the data about parties are related to speeches
delivered in the legislative arena. This can be an advantage because the two
measures are exogenous with respect to each other. One possible problem,
however, is related to the idea that the legislative party branch might retain a
different position compared with the party body. Anyway, given the strong
control usually exerted by Italian party bodies over their elected representatives
(e.g., Barnes 1977: 135), one can reasonably believe that the position of the
legislative party group would correspond to the actual party position.
Factional constraints on party positions in Italy

Let us move to testing the hypotheses discussed above using Italian data. The
dependent variable is Party Position (PP), estimated applying the vanilla method
to ILSD. Right-wing parties retain positive values, while PP is negative for
parties located on the extreme-left of the scale. The independent variables are
the following. Gamsonian Agreement Position (GAP) is the mean of the fac-
tion’s position weighted by the share of seats in the party body of each faction
per each party congress; positive values indicate that the weighted average of
factions is on the right; the more GAP is on the right, the more PP is expected to
be there (and vice versa). Years Before Elections (YBE) is the number of years
remaining before the next general elections. It is analyzed in interaction with
GAP: as YBE shrinks, the marginal effect of GAP on PP should increase (and
vice versa). The variable Median Faction records the influence of the median
faction above and beyond its contribution to GAP; it has been measured as the
distance between the median faction and the weighted mean of all factions
(which corresponds to the inter-factional agreement).12 The ability of the median
faction to attract PP toward its ideal point is caught by a positive significant
value of this variable. Direct Election is a dummy variable that accounts for the
autonomy of the party leader; it takes the value of 1 when the leader is directly
elected (by delegates during party congress or by party members or supporters
through “primary election”), and the value of 0 when the party leader is indi-
rectly nominated by a less inclusive selectorate (i.e., a small committee, such as
the party’s National Council or Central Committee).13 This variable is tested in
interaction with GAP; the expected effect of GAP on PP should be lower or not
statistically different from zero in parties that directly select the leader. Finally,
Parliamentary Centre of Gravity is a proxy for the center of the political space
and corresponds to the mean of the parties’ positions weighted by their share of
seats in Parliament; if parties are attracted by the political center, one should find
a positive significant value of this variable.
Figure 4.1 provides a first visual analysis of my general claims. For each
party congress, the picture shows the value of GAP along with all the PP values
measured until the next congress. Parties in which the leader is directly elected
(black) are differentiated from parties that elected their leader in smoke-filled
Figure 4.1 Party position and weighted mean of factions’ position in Italy.
Note
The dotted lines express the fitted value of the regression when the party leader is directly (black) or
indirectly (gray) elected.
rooms (gray). Accordingly, two dotted lines with the fitted value of the regres-
sion for the two contexts are plotted.
At first sight, factions do matter. There is a positive relationship between
GAP and PP, attesting that the parties’ ideal point is not independent from the
factions’ preferences. However, there are also differences between parties
whose leaders are stronger or weaker. The gray line is shallower than the black
one, indicating that (everything else being equal) leaders who retain a direct
legitimacy can profit from a larger autonomy and are able to get rid of the fac-
tions’ influence (in fact, the slope of the gray line is not even statistically
different from zero).
I investigate the hypotheses in more depth by means of an OLS regression on
a sample of 65 observations. In Model 1, the baseline model, I regress PP (using
the first known value after the congress) on the inter-factional agreement (GAP).
In Model 2 I add the variables Median Faction and Parliamentary Centre of
Gravity. In Model 3 I include the two interactions concerning the conditional
effect of GAP on Direct Election and YBE. In the fourth model I replicate the
third model, but increasing the number of cases to take into account all the
values of PP that are available until the next congress. By doing that, I’m able to
analyze up to 172 cases. Given that observations are nested within parties they
might not be independent. To assure unbiased results of the analysis, observa-
tions are clustered by party, providing standard errors by cluster.14 Table 4.1
reports the results.15
Table 4.1 OLS of party positions in Italy
DV: Party Position (ILSD) (I) (II) (III) (IV)
Gamsonian Agreement Position (GAP) 0.392*** 0.410*** 0.678*** 0.544***

(0.106) (0.128) (0.134) (0.092)
Median Faction – –0.191 –0.131 –0.257
(0.310) (0.293) (0.176)
Parliamentary Centre of Gravity – –0.223 0.002 0.306*
(0.324) (0.346) (0.155)
Direct Election – – –0.020 –0.014
(0.036) (0.031)
Interaction GAP X Direct Election – – –0.291** –0.282**
(0.132) (0.109)
Years Before Election (YBE) – – 0.019*** 0.005
(0.005) (0.004)
Interaction GAP X YBE – – –0.072** –0.054***
(0.025) (0.020)
Constant 0.025 0.023 –0.008 0.022
(0.018) (0.017) (0.015) (0.017)
R2 0.50 0.51 0.64 0.54
N 65 65 65 172
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. Clustered Standard errors in
parentheses.
The GAP coefficient is always significant: factions exert an effect on their
parties’ placement, and this is confirmed after including some additional vari-
ables. On the contrary, Median Faction is never significant; there is no dispro-
portionate effect of the median faction. In each party the median faction affects
the platform only to the extent of its strength in the party body, like any other
faction. The political center seems to have no effect too. Parliamentary Centre of
Gravity is not significant in Models 2 and 3, whereas it appears to have an
impact on party position only in Model 4.16 Both interactions are significant and
retain the expected sign. In line with the expectations, Direct Election attenuates
the effect of GAP on PP, and this effect is quite relevant.17 For instance, in
Model 4, the impact of GAP ceases to be significant for parties that select their
leader through a wide selectorate.18 Finally, when general elections come closer,
inter-factional bargaining becomes more consensual, and indeed the impact of
GAP on PP increases. The effect of these interactions is summarized by
Figure 4.2 (based on Model 4). The picture draws the marginal effect of GAP on
PP as the new general elections approach when leaders are indirectly (left panel)
or directly (right panel) elected.
When leaders are selected by party committees (Figure 4.2, left panel) they
are more bound by the will of factions; factional preferences increasingly shape
Figure 4.2 Marginal effect of Gamsonian Agreement Position (GAP) on Party Position
as the new general elections approach.
Note
Difference between parties with leaders indirectly (left panel) or directly (right panel) elected; the 95
percent confidence interval of the estimates is displayed.
party position as new elections approach. On the contrary, when intra-party rules
grant more autonomy to the leader (Figure 4.2, right panel) the effect of GAP on
PP is no longer significant (the confidence interval of the marginal effect crosses
the zero line), unless looming elections increase the need for party unity, pushing
the leader to partially cater to the factions.
These results can be interpreted referring back to theories outlining the exist-
ence of different models of party. To start with, there is no evidence for the idea
of parties as catch-all actors: as long as GAP is always significant the idea that
PPs are independent from factional preferences can be rejected. In line with the
model of mass party, consensual dynamics, more than a winner-takes-all logic,
seem to drive the internal life of parties. Accordingly, any disproportionate
advantage retained by the median faction above and beyond its share of votes
does not emerge in the Italian context. However, the extent of such a link
between factional preferences and PP is conditional upon the shape of intra-
party rules and the features of the party system. On one hand, when new elec-
tions come closer, parties tend to set their platform with an increasing degree of
proportionality.19 On the other hand, the degree of the leader’s autonomy plays
a role. Although factions exert constraints on their parties, when the internal
organization promotes leader’s autonomy the party seems no longer bounded
by activists and its leader retains more discretionality in setting the platform.
This evidence is coherent with the model of party suggested by the “cartel
party” theory.
With respect to the literature on party competition, these results suggest that
in a multiparty system with parties tied by activists there is no convergence
toward the center (and this is in line with recent theories). The political center
exerts an attraction only when party leaders are free enough from factional
constraints to be able to set party position on their own. Once again, this feature
is in line with the cartel party theory and highlights the ability of autonomous
leaders to partially adopt moderate stances, moving the party toward more vote-
maximizing positions.
Factional constraints on party positions in France and

Germany
Going beyond Italy, I double check these findings using data from France and
Germany. In both cases, however, the number of observations is quite limited if
compared to the Italian case. On the one hand, this is due to the lower amount of
data available on party factions; on the other, the lack of annual data on the
policy positions of parties also reduces the size of the dataset. In fact, I will
match the estimates of factions’ position presented in Chapter 2 with CMP data,
using the traditional and widely employed RILE scale. Focusing on the RILE
implies that I measure the effect of factions’ position always in the election year,
and this limits the ability to assess how their impact varies across the electoral
cycle. For these reasons, I will restrict the analysis to the main hypothesis testing
the effect of key independent variables.
For the French case, I evaluate the impact of faction’s positions on party posi-
tion, also in interaction with leader’s autonomy, on a sample of 31 observations.
The dependent variable Party Position (PP) corresponds to the value on the RILE
scale of the party manifesto closest to the party congress. The effect of factions
(H4.1A) will be operationalized using their weighted mean, i.e., through the
Gamsonian inter-factional agreement described above (GAP). Finally, I include
the interaction between GAP and Direct Election, which is equal to 1 when the
leader is directly elected by delegates during party congress or by party
members, and equal to 0 otherwise.
For the German case the number of observations is even lower. I only account
for 18 party manifestos issued after the party congress, matching them with the
policy positions of subgroups. There is not enough variation in intra-party rules
to account for the autonomy of party leaders, as was done so far, therefore I only
provide a general and speculative discussion related to H4.3. Furthermore, there
are no detailed data on the size of factions, consequently I do not account for any
weighted mean of factions (as in H4.1A). However, I still assess whether the
party position is independent or not from the position of internal subgroups,
testing a concept that is intrinsically stated in the catch-all party hypothesis
(H4.4). I also evaluate whether a single group can influence the party position
more than others, particularly focusing on the mainstream group of each party,
which can be either a pre-defined group or the group resulting from the relative
strength of all factions (as in H4.2). To do that, I create the variable Mainstream
Faction, equal to the estimated position of the party mainstream. Following the
discussion made in Chapter 2, for the CDU I always consider the MIT as the
mainstream group in the last 35 years (Debus & Bräuninger 2009). For the SPD,
I consider the FK as the mainstream faction from the 1980s until 1998, since
then (and up to 2005), I take the SK as the new party mainstream (Debus &
Bräuninger 2009). In 2009, however, I apply the mean value of the right-wing
SK and the left-wing DL21 to account for the fact that in the years leading to the
2009 election there was a huge internal instability inside the party; repeated
changes in the leadership occurred and members of both factions alternate in the
role of party leader, suggesting that both subgroups played a role at that time.
Finally, after 2011 I consider the DL21 as the mainstream group, given that left-
wing factions took control of the party, reaching a relative majority of affiliates
in the parliamentary party group (see Chapter 2).20
Figure 4.3 compares the party position on the RILE with the estimates of fac-
tions’ positions and it also provides the linear fit between these two measures
(with a 95 percent confidence interval). For Germany (right panel), the position
of the mainstream faction is reported (accounting for the changes in the party
mainstream over time). For France (left panel), the weighted position of all fac-
tions has been plotted, distinguishing between directly (black) or indirectly
(gray) elected leaders.
Once again, there is a positive relationship between the position of parties
and factions, though such relationship is stronger in parties with indirectly
elected leaders (as in the Italian case). Conversely, the linear fit gets flatter when
Figure 4.3 Party position and factions’ position in France and Germany.
Note
Parties with directly elected (black) or indirectly elected leaders (gray) are kept separated. The
dashed lines express the fitted value of the related regressions. The 95 percent confidence interval of
the estimates is displayed (dotted lines).
focusing on directly elected leaders, as they can exploit their greater autonomy
to adjust the party platform going beyond the wills of internal factions.
Starting from this preliminary scrutiny, Table 4.2 provides the results of the
statistical analysis for France (Model 1) and Germany (Model 2). Party dummies
have been included in all the models.
The analysis on France displays exactly the same finding already observed in
the Italian case. The coefficient of GAP is positive and markedly significant sug-
gesting that the weighted position of all factions affects the party platform. This
effect, however, is attenuated in presence of autonomous party leaders. The
interaction between GAP and Direct Election is significant; in line with my
hypothesis (H4.3), the effect of GAP is huge and significant though only if
leaders are indirectly selected; conversely, when Direct Election is equal to 1,
the effect of GAP is negative and not statistically significant at the 95 percent
level of confidence. Accordingly, the French data confirm the findings related to
Italy and provide further support for the theoretical framework outlined in
Chapter 1.
Turning to Germany, the coefficient of the variable Mainstream Faction is
positive and significant. The position of the mainstream subgroup inside each
Table 4.2 OLS of party positions in France and Germany
DV: Party Position (CMP/RILE) France Germany
Gamsonian Agreement Position (GAP) 14.660*** –

(3.006)
Mainstream Faction – 9.366**
(4.158)
Direct Election –7.149* –
(3.317)
Interaction GAP X Direct Election –20.588*** –
(3.137)
Constant –10.179** 3.117
(3.791) (5.920)
R2 0.62 0.64
N 31 18
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. Party dummies embedded.
party does affect the final party platform. As such, the party position is not inde-
pendent from the ideal points of internal organizations (disconfirming the catch-
all party thesis outlined in H4.4). Even in this context then intra-party politics
matters. Interestingly, when excluding from the analysis the unique case of a
wider and direct selectorate, concerning the election of the SPD leader in 1993
(Astudillo & Detterbeck 2018; see also Chapter 3), the coefficient gets larger.
This could imply that a direct selection dampens the relevance of factions and
would be in line with the argument suggesting that factional preferences matter
less when leaders are more autonomous due to direct legitimization.
Overall, a first inquiry focused on the impact of single party subgroups sug-
gested that indeed the MIT is able to influence the position of the CDU, while the
CDA seems less relevant. Inside the SPD, the moderate SK faction seemed in turn
more important for the party platform compared to the internal left. However, when
considering SK always as the mainstream group in the SPD, the relationship dis-
played in Table 4.2 becomes less robust; the effect would be flatter and the coeffi-
cient is no longer significant at the 95 percent level of confidence (it only remains
significant at the 90 percent level). This can imply that, when their power is strong
enough, left-wing internal subgroups can still influence the policy program of the
party and can resist pressure for a vote-maximizing policy moderation that goes too
far away from what they would consider acceptable.
To sum up, while in the CDU the association of employers plays a crucial
role in anchoring the party to center-right positions, in the SPD the equilibrium
is more affected by the factional conflict and the reciprocal strength of each sub-
group can influence the final party platform issued before general elections.
This highlights the importance of factionalism inside this party. This result
has implications for the ongoing debate about the willingness of the SPD to
support either moderate cabinets, jointly with the CDU, or to promote a new
political alliance open to left-wing parties and movements.
The fact that factions and subgroups matter in German politics even now-
adays (let us consider for instance the discussion about the SPD mainstream after
2005) advocates the theoretical framework discussed in Chapter 1. Indeed, in the
German case the leadership selection is still demanded to the party congress; the
traditionally huge levels of consensus reached when appointing the leader can
suggest the existence of consensual dynamics that aim to preserve unity and to
recognize the role of factions in shaping the party platform.
Setting the party platform: findings

Going beyond the assumption of party as a unitary actor, this chapter scrutinized
the process of position-taking within the party using data from the legislative
(Italy) or the electoral arena (France and Germany). Under the idea that the
selection of the party position is a proxy for the allocation of policy payoffs
inside the party, I investigated whether and how factions bound party leaders in
choosing the platform.
I compared contrasting theories that propose alternative models of party,
trying to assess whether the party is run as an oligarchy whose leader is com-
pletely in control of party choices (Michels 1915) or, alternatively, the impact of
factional preferences prevails.
For this purpose, data on factional preferences have been matched with the
outcome of 65 and 172 processes of position-taking emerging from Italian
parliamentary speeches. These data have been used to test a wide set of hypo-
theses; such findings have been double-checked with additional analyses on bar-
gaining dynamics leading to the formulation of 31 party manifestos in France
and 18 in Germany.
Overall, in all three countries there is confirmation of the idea that PPs are
linked to the preferences of factions, discarding a model of catch-all parties com-
pletely free to move in the policy space. A proportional allocation of policy
payoffs seems to be the norm and, for the sake of preserving party unity, the
party platform is selected taking into account the stakes of all factions, propor-
tionally to their strength in the party body.
Due to the pressure exerted by factions, the party seems to work as a
“bounded oligarchy” (Ceron 2012). The impact of factional constraints is greater
when general elections approach and the electoral payoff of party unity pushes
the pursuit of a compromise among all party subgroups. In line with the cartel
party theory, however, the autonomy of leaders increases when they are directly
elected by a wider selectorate, in spite of the factions’ desires. This finding is
patently robust in Italy and France, but also in Germany the evidence seems in
line with this argument. Indeed, autonomous party leaders can exploit such auto-
nomy to set a more vote-maximizing position, promoting a partial convergence
toward the center. Although carried out in a non-strategic setting, these results
provide support for the most recent developments of Downsian theories that
predict a centrifugal equilibrium with partial convergence. Finally, this chapter
shows that direct election of party leaders (e.g., through “primaries”) might
defang the activists. Instead of promoting members’ control of party elites and
the emergence of a widely shared platform, intra-party democracy seems to
support the leader’s autonomy, making the party slip toward a “Bonapartist” and
oligarchic organization.
Notes
1 Other models predict similar results; see Budge et al. (2010) for a review.
2 It could be argued that intra-party politics involves more than one dimension. We
know, however, that parties work as logrollers and reduce complex issues into a low-
dimensional space (e.g., Levy 2004); if parties reduce the complexity on a single
dimensional space they need to compromise on that.
3 As is well known, the Gamson (1961) hypothesis is essentially a proportionality
rule stating that cabinet portfolios are allocated to each party (or faction: Ceron
2014) proportionally to their contribution to the ruling coalition (i.e., their share of
seats); see Chapter 5. Warwick (2001) extended this reasoning to policy payoffs,
showing that “coalition policy corresponds with the weighted mean position of the
parties in government, with the parties’ seat share constituting the weights”
(Warwick 2001: 1215).
4 This can happen also in parliamentary votes, when the electoral system or the intra-
party candidate selection process do not allow strong discipline to be enforced (Ceron
2015).
5 See Fabbrini (1994) and Calise (2005) on the link between leaders’ personalized
power and their direct election by party delegates.
6 With regard to candidate selection, inclusiveness also produces lower levels of repres-
entation (Rahat et al. 2008).
7 Contrary to the mass and the catch-all models, the cartel party “whether understood as
a synthesis/thesis in a dialectic process or as the next stage in a cyclical process […]
does not represent a steady state” (Katz 2001: 282).
8 For each debate, the authors selected and codified the speech released by the party
leader (or by a relevant representative) plus the Prime Minister’s programmatic
speech. The method adopted to codify speeches was similar to that employed by the
well-known CMP project to analyze the contents of party electoral programs (Budge
et al. 2001). In each legislative speech the authors identified the number of quasi-
sentences and assigned each of them to a number of pre-established categories that
form the classification scheme. To take account of the Italian political context the ori-
ginal 56 categories of the CMP dataset were increased to 68. The dataset contains the
percentage of the total text of legislative speech that deals with these categories. The
dataset has been constantly updated until June 2018 and also includes information on
a few additional debates over crucial confidence votes (i.e., the debate related to Prodi
I vote of no-confidence; two debates faced in 2010 by Berlusconi IV cabinet and one
involving the Letta cabinet in 2013).
9 When one or more factions broke away from the party during the period between one
congress and the next, I rearranged the result of factional agreement measuring the
new balance of power among the remaining subgroups. Cases in which, due to party
fissions, there was only one faction left have been excluded; however, including these
data does not affect the main results.
10 Given that motions have been divided in two time periods before running Wordfish,
the same approach was used in the factor analysis: ILSD data have been divided in
two subsamples, running two separate analyses. The results of these two factor
analyses are highly correlated (0.87), with parties’ positions as they emerge from the
analysis ran on the whole dataset.
11 The correlation between vanilla estimates and the RILE scale measured on ILSD is
strong and significant (0.85). With respect to data included in this analysis, the estim-
ates of parties’ positions measured by applying the vanilla method to ILSD and CMP
are positively correlated (0.61). Using CMP data instead of ILSD does not alter the
main results.
12 This operationalization is made to overcome problems of correlation between the
median and the Gamsonian agreement. Indeed, in the Italian case these two positions
are highly correlated (0.99). See Warwick (2001: 1223) for an analogous argument.
13 I do not distinguish between leaders elected by delegates or through open/closed
primaries because the latter context concerns only one congress (PD 2009). In addi-
tion, when leaders are elected by delegates, party members often know which leader
is attached to each factional motion they vote for (see Chapter 3).
14 In Model 4 there are repeated observations within each congress. Therefore, I pro-
vided standard errors clustered by congress. Clustered standard errors are heteroske-
dastic and autocorrelation consistent (Rogers 1993).
15 Given the uncertainty attached to the estimates of factions’ positions, the “simulation-
extrapolation” procedure (SIMEX) suggested by Benoit et al. (2009) to deal with
measurement error could be used. In this regard, using the upper and lower bounds of
Wordfish’s estimates to assess the actual value of GAP does not alter the main
findings.
16 When adding to Model 4 an interaction between this variable and Direct Election, the
effect of Parliamentary Centre of Gravity becomes significant only in the case of
direct election of leaders.
17 This effect is still significant even after controlling for the policy preferences of the
party leader’s faction. When the party is ruled by leaders not linked to the median
faction one would expect less consensual dynamics, and indeed the effect of GAP on
PP is lower. However, this effect is not always significant throughout the analyses.
18 The marginal effect of GAP on PP is 0.406 (0.061) and it is statistically significant at
the 99 percent level when Direct Election equals 0, while this effect is no longer signi-
ficant when leaders are autonomous: 0.124 (0.093). Standard errors in brackets.
19 This holds true regardless of the changes in the Italian electoral rules.
20 Below I will compare this scenario with the more straightforward case in which the
SK (traditionally supportive of the party leadership) is considered as the only main-
stream group in the SPD.
References
Adams, J., Merrill, S., & Grofman, B. (2005). A Unified Theory of Party Competition.
Cambridge University Press, Cambridge.
Aldrich, J. (1983). A Downsian Spatial Model with Party Activism. American Political
Science Review, 77(4), 974–990.
Barnes, S.H. (1977). Representation in Italy. Institutionalised Tradition and Electoral
Choice. The University of Chicago Press, Chicago and London.
Benoit, K., Laver, M., & Mikhaylov, S. (2009). Treating Words as Data with Error:
Uncertainty in Text Statements of Policy Positions. American Journal of Political
Science, 53(2), 495–513.
Bernauer, J., & Bräuninger, T. (2009). Intra-Party Preference Heterogeneity and Faction
781–804.
Budge, I., Klingemann, H.-D., Volkens, A., Bara, J., & Tanenbaum, E. (eds) (2001).
Mapping Policy Preferences. Oxford University Press, Oxford.
Calise, M. (2005). Presidentialization, Italian Style. In Poguntke, T., & Webb, P. (eds),
The Presidentialization of Politics: A Study in Comparative Politics, 88–106, Oxford
University Press, Oxford.
Ceron, A. (2015). Brave Rebels Stay Home. Assessing the Effect of Intraparty Ideological
Heterogeneity and Party Whip on Roll-call Votes. Party Politics, 21(2), 246–258.
s11127-019-00644-0.
491–508.
Curini, L., & Ceron, A. (2013). Parties’ Influence during Government Policy Negoti-
ations: Parliamentary Dynamics and Spatial Advantages in the First Italian Republic.
Journal of Legislative Studies, 19(4), 429–449.
Debus, M., & Bräuninger, T. (2009). Intra-Party Factions and Coalition Bargaining. In
Giannetti, D., & Benoit, K. (eds), Intra-Party Politics and Coalition Government,
121–145, Routledge, New York.
Downs, A. (1957). An Economic Theory of Democracy. Harper, New York.
Fabbrini, S. (1994). Personalization as Americanization? The Rise and Fall of Leader-
Dominated Governmental Strategies in Western Europe in The Eighties. American
Studies International, 32(2), 51–65.
Gabel, M., & Huber, J. (2000). Putting Parties in Their Place. American Journal of Polit-
ical Sciences, 44(1), 94–103.
Gamson, W. (1961). A Theory of Coalition Formation. American Sociological Review,
26(3), 373–382.
Italy. In Giannetti, D., & Benoit, K. (eds), Intra-Party Politics and Coalition Govern-
ment, 146–168, Routledge, New York.
Giannetti, D., & Mulé, R. (2006). The Democratici di Sinistra: In Search of a New Iden-
tity. South European Society & Politics, 11(3), 457–475.
Grofman, B. (2004). Downs and Two-Party Convergence. Annual Review of Political
Science, 7(1), 25–46.
Heller, W.B., & Mershon, C. (2009). Legislator Preferences, Party Desire: The Impact of
Party Switching on Legislative Party Position. In Heller, W.B., & Mershon, C. (eds),
Political Parties and Legislative Party Switching, 173–200, Palgrave Macmillan,
New York.
Katz, R.S., & Mair, P. (1995). Changing Models of Party Organization and Party Demo-
cracy: The Emergence of the Cartel Party. Party Politics, 1(1), 5–28.
Kirchheimer, O. (1966). The Transformation of the Western European Party System. In
LaPalombara, J., & Weiner, M. (eds), Political Parties and Political Development,
3–42, Princeton University Press, Princeton.
Laver, M., & Shepsle, K.A. (1990). Government Coalitions and Intraparty Politics.
250–277.
Mair, P. (1994). Party Organization: From Civil Society to the State. In Katz, R.S., &
Mair, P. (eds), How Parties Organize: Change and Adaptation in Party Organizations
in Western Democracies, 1–22, Sage, London.
Mair, P. (1997). Party System Change: Approaches and Interpretations. Oxford Univer-
sity Press, New York.
Marsh, M. (1993). Introduction: Selecting the Party Leader. European Journal of Polit-
ical Research, 24(3), 229–231.
McGann, A.J. (2002). The Advantages of Ideological Cohesion: A Model of Constitu-
ency Representation and Electoral Competition in Multi-Party Democracies. Journal of
Theoretical Politics, 14(1), 37–70.
encies of Modern Democracy. Free Press, New York.
Morgenstern, S. (2001). Organized Factions and Disorganized Parties. Electoral Incen-
tives in Uruguay. Party Politics, 7(2), 235–256.
Müller, W.C., & Strøm, K. (eds) (1999). Policy, Office, or Votes? How Political Parties
in Western Europe Make Hard Choices. Cambridge University Press, Cambridge.
Politics, 14(6), 663–683.
Rogers, W.H. (1993). Regression Standard Errors in Clustered Samples. Stata Technical
Bulletin, 3(13), 19–23.
Ware, A. (1992). Activist-Leader Relations and the Structure of Political Parties:
“Exchange” Models and Vote-Seeking Behaviour in Parties. British Journal of Polit-
ical Science, 22(1), 71–92.
Warwick, P.V. (2001). Coalition Policy in Parliamentary Democracies: Who Gets How
Much and Why. Comparative Political Studies, 34(10), 1212–1236.
5 Portfolio allocation among party
factions
Introduction
The previous chapter focused on the allocation of policy payoffs. To do that, I
investigated how parties select their policy platform. That chapter shed light on
inter-factional competition and on the role of the party leader in reaching an
agreement, rewarding his followers while preserving party unity at the same
time. In this chapter, I investigate related dynamics concerning the intra-party
bargaining around office payoffs to evaluate what conditions produce a fair
“compromise” or an unfair allocation of posts.
Political actors deal with two contrasting pressure forces when they engage a
negotiation toward the making of a government. On the one hand, each actor aims
to reach a compromise to form a new cabinet that should provide the public good to
all its members. On the other, cabinet partners compete to prevail and maximize
their own amount of private goods (ministers) in the bargaining process. Such
dichotomy exists not only between parties, but also within them, as internal factions
strive to obtain the largest share of office payoffs. Indeed, portfolio allocation
dynamics have been largely analyzed by political scientists to understand the pat-
terns of cooperation and competition both between and within parties.
Party factions can play a role in everyday policy-making (e.g., Ceron 2015b)
and in cabinet formation (e.g., Ceron 2016); furthermore, factions can employ
cabinet reshuffle as a device to negotiate a new distribution of portfolios
(Giannetti 2010). Consequently, it is important to investigate intra-party distribu-
tive dynamics by analyzing how rival factions reach an agreement to allocate
office payoffs (ministerial positions).
Modeling the party as a coalition of factions (Leiserson 1968; Panebianco
1988) that needs to split the total amount of cabinet ministers available, I will
examine two contrasting theories of portfolio allocation (Gamson’s Law versus
bargaining theories). Based on my game–theoretic model, I will sketch the shape
of distributive conflicts under different intra-party environments.
Portfolio allocation: Gamson Rule versus bargaining theories

The discussion between alternative theories of portfolio allocation (Gamson’s
Law versus non-cooperative bargaining theories) has been carried out mainly
110 Portfolio allocation among party factions
through the observation of how payoffs between parties are distributed, whereas
intra-party portfolio allocation has had very little relevance. Only a few, though
important, studies addressed the issue. These studies mainly investigated office
allotment in Italy or Japan with regards to two dominant parties: the Italian
Christian Democracy (Mershon 2001a, 2001b) and the Japanese Liberal Demo-
cratic Party (Adachi & Watanabe 2008; Bouissou 2001; Leiserson 1968; Ono
2012; Wada & Schofield 1996), or both of them (Kato & Mershon 2006).
Gamson’s (1961) seminal work opened the debate on how government coali-
tions hand out office payoffs to their members and argued that payoffs are dis-
tributed in proportion to the amount of resources provided by each member of
the coalition (Gamson 1961: 376). That is, in multiparty governments, cabinet
portfolios are allocated to parties according to their share of seats, which is
measured as a percentage of the total seats held by parties that support the ruling
coalition. Subsequent studies found empirical support for this thesis (Browne &
Franklin 1973; Browne & Frendreis 1980; Carroll & Cox 2007; Laver et al.
2011; Warwick & Druckman 2006). The empirical evidence, in fact, leads
scholars to refer to such a rule as Gamson’s Law. Laver (1998: 7) pointed out
that Gamson’s Law attains “one of the highest non-trivial R-squared figures in
political science.”
Nonetheless, those who criticize Gamson’s Law claim that this is merely an
empirical regularity with a need for sensible theoretical foundation (Warwick &
Druckman 2006: 660). In view of that, Carrol and Cox (2007) attempted to fill
this lack by developing a theoretical argument that underpins the adoption of a
proportionality criterion in portfolio allocation among the members of pre-
electoral coalitions. In particular, “By agreeing ex ante to a more Gamsonian
division of office spoils ex post, the coalition can motivate its members to
campaign harder, thus conferring external benefits on all” (Carroll & Cox
2007: 301).
Following this idea, I consider the party as a coalition that is composed of
several subgroups. By establishing an internal rule of thumb that assigns propor-
tional payoffs to each faction, the party pushes all its subgroups to work together;
as such, the party maximizes both its strength and its bargaining power.1 Propor-
tional allocation, in addition, allows the softening of the opposition of rival fac-
tions, keeping the party together. With this in mind, I can assess whether
Gamson’s Law applies to intra-party portfolio allocation.
Hypothesis 5.1 (H5.1): The faction share of portfolios should correspond with
its share of seats in party body.
Mershon (2001a) successfully tested this hypothesis with a study on the

Italian DC.2 Besides confirming Gamson’s hypothesis, Mershon’s analysis
revealed the existence of “variations across groups of governments […] and
across types of factions that are masked by the overall pattern” (Mershon 2001a:
573). These deviations are related to changes in party rule (i.e., the direct elec-
tion of party leader), to the strategic position of the faction (being it the median
Portfolio allocation among party factions 111
faction or one of the wings), and to party system features. Below I will account
for these and other conditional effects or alternative explanations.
Actually, the Gamson hypothesis is not the only expectation suggested by the
literature on portfolio allocation. In opposition to Gamson’s rule another branch
of the literature raised an alternative hypothesis based on non-cooperative bar-
gaining theories (Ansolabehere et al. 2005; Baron & Ferejohn 1989; Fréchette
et al. 2005; Morelli 1999; Snyder et al. 2005). These scholars argue that the
“formateur party,” that is the first party in charge of building a coalition, can
exploit its strategic role “to pivot between alternative minimal-winning coali-
tions” (Carroll & Cox 2007: 300) in order to obtain a share of posts greater than
its share of seats. This idea can be applied to intra-party politics and tested
against Gamson’s Law.
To do that, I adapt the concept of formateur party to fit the inter-factional
context. Following recent studies (Ono 2012), I assume that the party leader is
the agent in charge of allocating payoffs. Like in a setter game the leader can
propose to the factions one among several possible distributions of payoffs that
might be accepted or rejected. This idea can be modeled through the classic
“Ultimatum Game,” in which one actor (the party leader) has to split a total
amount of payoffs, making a proposal that can be accepted or rejected by the
minority group. If the minority refuses, the payoffs of both actors shrink to zero.
The leader might thus offer a fair deal (granting to the minority a quota of bene-
fits proportional to its strength) or, otherwise, present an unfair sharing (overpay-
ing his own faction). The minority faction may either accept or refuse the
proposal.
Before anything, the party leader wishes to keep his status, but also to reward
his followers and to enhance their strength within the party. Moving first, the
leader can exploit such discretionality to grant to his faction a more than propor-
tional reward. Thus, the faction the leader belongs to can theoretically take
advantage of its close connection to the leader, and might as well be considered
as a “formateur faction.” As a matter of fact, the game–theoretical solution of
this game seems to be that the leader proposes the unequal distribution and the
minority accepts the deal.3 Given the credible outcomes of the game, both actors
maximize their payoffs. As a consequence, the leader’s share of cabinet posts
should increase above the “Gamsonian” prediction.4
Hypothesis 5.2 (H5.2): The party leader’s faction should gain a share of port-
folios that is more than proportional to its seats share.
A theory of inter-factional bargaining: hypotheses

The idea of a setter game is related to what was discussed in Chapter 1. Therein
I presented a theory of intra-party politics in which I considered the party as a
coalition of factions that have to split policy and office payoffs (Leiserson 1968).
Factions are involved in a common effort to produce a public good, i.e., party
unity, which enhances party strength during the negotiation over coalition
formation (Bäck 2009) and increases the likelihood of being involved in a coali-
tion government (Ceron 2016).
Accordingly, factions need to cooperate, as reaching an agreement would
increase the total amount of payoffs available to the party. At the same time,
each faction is in competition with the others and they can fight on the distribu-
tion of these benefits. These internal dynamics are affected by intra-party rules
and one can expect to observe different outcomes under different intra-party
settings.
To start with, the degree of proportionality in portfolio allocation and leader’s
ability to extract a larger share of payoffs may depend on the leader’s degree of
autonomy from party factions, which in turn is a function of internal rules.
Non-autonomous leaders indirectly selected in a small committee are nothing
more than the output of an inter-factional bargaining, appointed to implement an
agreement that has already been reached elsewhere. They retain no autonomy
and will simply stick to such a deal offering a compromise to the minority
faction and splitting the payoffs fairly. Accordingly, the allocation of spoils
should be more proportional in this case (as suggested in H5.1).
To the contrary, the “formateur faction” could gain a more than proportional
amount of cabinet posts when the leader is autonomous from factional con-
straints i.e., when there is a direct election.
Some studies provided preliminary evidence in line with this idea. Mershon
(2001a, 2001b) showed that the direct election of DC party leaders lowered the
degree of proportionality in portfolio allocation. Kato and Mershon (2006)
pointed out that contrasting modes of leadership selection led to different pat-
terns of allocation: in the Japanese Liberal Democratic Party (LDP) the party
leader was elected, until the 1980s, by the party congress and in such scenario
portfolio allocation followed winner-takes-all dynamics; afterwards, the rules
changed and the LDP leader became selected after inter-factional negotiation.
These dynamics forced the leader to comply with the inter-factional deal,
increasing the fairness in portfolio allocation (Di Virgilio & Kato 2001).
Being autonomous from inter-factional agreements, the leader can decide
how to assign benefits alone. Autonomous leaders are less likely to be chal-
lenged or removed from their charge (Kenig 2009; Michels 1915; Rahat et al.
2008; see also Chapter 7) particularly when they are popular among voters or
when their faction is stronger within the party: in this sense leaders can assign
office payoffs with a huge degree of discretionality (Ono 2012). For instance, the
leader can exploit such discretionality to reward the followers, for on their
support the power of the leader is established.
Hypothesis 5.3 (H5.3): The degree of proportionality in portfolio allocation

should be lower among parties whose leaders are more autonomous.
Hypothesis 5.4 (H5.4): The party leader’s faction should gain a share of port-
folios that is more than proportional to its seats share when the party leader is
more autonomous.
Furthermore, according to the theoretical model, I claim that party system
competitiveness alters the leader’s attitude toward compromise. In a parlia-
mentary arena with a low level of competition, where the ruling coalition retains
a safe margin over the opposition (ω is low), the leaders of ruling parties will not
fear the risk of losing office due to potential defections of minority factions.
Accordingly, they will try to overpay their followers whereas under highly com-
petitive conditions the mainstream faction should be less overpaid.
Hypothesis 5.5 (H5.5): The party leader faction should be increasingly overpaid
as the party system competitiveness shrinks.
The blackmail power of a given faction might also alter the patterns of intra-
party bargaining. Intra-party minorities, in fact, might increase their bargaining
power with a threat to defect, that is leaving the party. The mainstream faction,
then, should cater to the minority as much as possible in order to avoid a fission
that could damage the party (Boucek 2010; Hirschman 1970). In this regard,
Mutlu-Eren (2015) highlights how party leaders use pork-barrel politics as a
strategy to keep their party together. Analogously, Wada and Schofield (1996)
argued that the shift toward proportional allocation in the Japanese LDP has
been fostered by factional leaders’ threat of exiting from the party.
Along this line, I hypothesize that minority factions will be overpaid,
receiving a share of ministers greater than their size, though only under two con-
ditions: the threat to defect must be credible and considerably harmful to the
party. In order to be harmful, the threat must involve at least a medium-sized
minority faction. In order to be credible, those factions should gain only a
limited amount of policy payoffs from party membership. In fact, factions incur
a cost for party membership (Snyder & Ting 2002) and this cost grows when the
distance between factions’ and parties’ positions enlarges (high π); in this vein,
it can be argued that the probability of a split (and the leverage of the minority)
increases with such distance (Ceron 2015a; Reed & Scheiner 2003; see also
Chapter 6). If this is the case, factions whose ideal point is far from the bulk of
party members should be overpaid in the allotment of cabinet posts.5
Hypothesis 5.6 (H5.6): Minority factions’ share of portfolios should increase

along with their size and their distance from the median faction (i.e., the likeli-
hood of their breakaway).
Inter-factional portfolio allocation in Italy

To analyze factional bargaining and portfolio allocation, I focus on party con-
gresses combining the data displayed in Chapter 2 with information on factional
affiliation of ministers. Considering as a faction any subgroup that contests the
congress presenting its own motion and looking at the motions presented during
the congress, I ascribe factional affiliation to party ministers according to their
signature on one of the competing documents (for a similar approach: Giannetti
& Laver 2009; Mershon 2001b). In a few cases factional affiliation has been
inferred looking at newspaper reports, ministers’ biography or sources of data
provided by the literature (Venditti 1981).
Per each party congress, the unit of analysis is the faction share of portfolios
within any single government that was in office until the following party con-
gress (for a similar operationalization: Mershon 2001b: 285–286). This allows
the number of observations to be increased and makes the analysis comparable
with other studies in the field of portfolio allocation.
By doing that, the analysis covers ministerial allocation within 42 different
Italian governments considering nine different parties. Such dataset includes
information on 275 factions (see Table 5.1) belonging to almost all parties that
held ministerial posts during the First Republic (DC, PLI, PRI, PSDI, PSI) and
to a few factionalized parties that were in office during the Second Republic
(AN, DS, NPSI, PRC).6
Portfolio allocation is a typical example of compositional data (Honaker et al.
2002): the share of ministers assigned to each party sums to one (100 percent)
and therefore the errors are correlated. To address this concern, I follow previous
studies (Carroll & Cox 2007; Fréchette et al. 2005) and drop one observation per
each intra-party bargaining, running the analysis on a subsample.7
Furthermore, given that the dependent variable is a proportion (the share of
ministers assigned to each faction) using the OLS regression might not produce
accurate estimates. This happens because the values of the dependent variable
are bounded between 0 (the faction does not hold any minister) and 1 (the whole
share of party ministers is assigned to one faction only). As a consequence, the
assumptions required by the OLS might not hold; there could be heteroskedas-
ticity or the errors might be not normally distributed (Wooldridge 2002). More-
over, the predicted values might fall outside the unit interval. I will thus analyze
such fractional response dependent variable by means of a fractional logit model,
which is suitable to handle these issues (Papke & Wooldridge 1996).8 In addi-
tion, there are repeated observations within each party congress and these are
Table 5.1 Details on the Italian portfolio allocation data
Party Congresses Governments N
AN 1 2 7
DC 10 29 121
DS 3 2 9
NPSI 1 1 2
PLI 5 10 26
PRC 1 1 5
PRI 7 10 23
PSDI 8 14 35
PSI 7 14 47
Total 43 42 275
nested by government; to cope with this, fixed effects by congress have been
included and observations have been clustered on governments, providing
standard errors accordingly.
The dependent variable is the Weighted Share of Ministers: the ratio of
cabinet posts (weighted) belonging to faction i over the total number of
(weighted) ministers assigned to party j within government k. The weights have
been applied using the portfolio ratings measured by Warwick and Druckman
(2005: 39–40) that allow to assess the importance of each cabinet position.9
Notice that the findings hold the same even when analyzing the unweighted
share of ministers as a dependent variable.
The independent variables are the following. Share of Seats (SofS): the per-
centage of seats retained in party body (i.e., the National Council or the Central
Committee).10 Party Leader Faction (PLF ): a dummy variable that identifies the
faction to whom the party leader belongs to (value 1). Direct Election: equal to 1
when the party leader is elected during the party congress directly by party dele-
gates or party members (more autonomous leader) while it takes value 0 when
the leader is indirectly appointed by a small committee (less autonomous leader).
Parliamentary Support: expresses the degree of party system competitiveness
measured according to the size of parties that support the cabinet. Ideological
Distance: corresponds to the squared distance between the position of faction i
and that of the median faction within the party. To investigate the effect of each
variable on portfolio allocation some interaction terms are also included. In par-
ticular, I focused on the interactions of SofS with Direct Election and Ideological
Distance, to attest whether under particular conditions the impact of faction size
changes. In this way I can better assess if Gamson’s Law actually shapes intra-
party bargaining. Finally, I also included the interactions between Party Leader
Faction and two other variables: Direct Election and Parliamentary Support.
Table 5.2 provides the results of the analysis. In Model 1 I test Gamson’s
Law (H5.1) against non-cooperative Bargaining Theory (H5.2).11 In Model 2
I add the interactions between those two variables and Direct Election testing
H5.3 and H5.4. In Models 3 and 4 I test the impact of party system competit-
iveness (H5.5) and the leverage of minority according to their share of policy
payoffs (H5.6).12
To start with, Share of Seats is always significant confirming that ministers
are distributed in proportion to faction size (H5.1).13 This does not implies a
pattern of perfect proportionality though. In fact, in Model 1, the coefficient of
Party Leader Faction is significant and positive as well, even controlling for
Share of Seats.14 This confirms, at least in part, the idea behind bargaining the-
ories (H5.2) suggesting that party leaders are able to exploit their strategic role
to overpay their own faction with a share of payoffs larger than the actual size.
The conditional marginal effect reports an advantage between 0.6 and 9.8
percent for the leader’s faction.
So far, only standard distributive theories have been discussed. Moving to my
argument, in Model 2 I evaluate the impact of Direct Election finding that con-
trasting rules for leadership selection lead to different patterns of portfolio
Table 5.2 Portfolio allocation among party factions in Italy
DV: Weighted Share of Ministers (1) (2) (3) (4)
Share of Seats 6.183*** 7.193*** 6.920*** 6.696***

(0.655) (0.890) (0.744) (0.738)
Party Leader Faction 0.485* 0.274 –0.792 –0.758
(0.262) (0.401) (0.507) (0.504)
Direct Election 1.469*** 1.472*** 1.510***
(0.464) (0.508) (0.475)
SofS X Direct Election –2.040* –1.695** –1.728**
(1.143) (0.784) (0.785)
PLF X Direct Election 0.253 – –
(0.481)
Parliamentary Support – –0.411 –0.462
(0.277) (0.293)
PLF X Parliamentary Support – 2.291** 2.361**
(1.015) (1.023)
Ideological Distance – – –0.476
(0.458)
SofS X Ideological Distance – – 3.049*
(1.567)
Constant –2.600*** –3.783*** –3.573*** –3.522***
(0.225) (0.423) (0.474) (0.448)
Log-likelihood –49.919 –49.702 –49.544 –49.444
N 189 189 189 189
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. Clustered standard errors in parentheses
allocation.15 In line with H5.3, the degree of proportionality is indeed lower

among parties ruled by autonomous leaders who are free to allocate payoffs
according to their wishes.16 However, the party leader faction does not automati-
cally gain from this shift as the interaction between Party Leader Faction and
Direct Election is not significant (against H5.4). Who does benefit from a
growing leader’s discretionality? Models 3 and 4 provide an answer, testing the
theoretical implications suggested by my theory.
I contend that the level of party system competitiveness impinges on intra-party
distributive dynamics. Under conditions of low competitiveness (when the govern-
ment retains a safe and wide margin over the opposition) party splits do not jeop-
ardize government stability. As a consequence, the party leader is free to overpay
his own faction even if this unfair allocation of portfolios may disappoint internal
minorities leading them to break away. Conversely, in highly competitive parlia-
mentary arenas the leader will be focused on keeping party unity insofar as a break-
away could jeopardize government duration and the party’s share of payoffs.17 The
interaction between Parliamentary Support and Party Leader Faction is positive and
significant and this empirical result provides evidence for H5.5. In Figure 5.1 I
display the advantage exploited by the mainstream faction, conditional on the
balance of power between government and opposition. When the Parliamentary
Figure 5.1 Marginal effect of Party Leader Faction on Weighted Share of Ministers as
the degree of Parliamentary Support for the cabinet enlarges.
Support is very low the mainstream cannot be overpaid. Conversely when the size
of the ruling majority increases the leader starts to take profit of his strategic role
granting to his faction a more than proportional share of spoils.18
To conclude, Model 4 tests the interaction between the minority’s Share of
Seats and its Ideological Distance from the bulk of party members (which
expresses the blackmail power of non-mainstream factions, as suggested in
H5.6). Figure 5.2 reports the marginal effect (along with a 90 percent confidence
interval) for two differently sized non-mainstream factions, which retain respec-
tively 10 (dashed line) and 40 percent (solid line) of seats in party body.
It is noticeable that the smaller faction is strongly underpaid: a 10 percent
increase in size is rewarded by only an additional 4 percent of the share of minis-
ters. On the contrary, the powerful minority receives a more than proportional
return in office payoffs. Moreover, as the distance between this subgroup and the
median faction grows, such benefit enlarges.
The growing distance between the ideal point of a non-mainstream subgroup
and the median faction raises the cost of party membership for that group (Snyder
& Ting 2002). Ceteris paribus its total amount of payoffs will be lower and the
faction will face a growing incentive to break away leaving the party. As a result,
minority factions can try to increase their amount of payoffs threatening to split
from the party. This blackmail strategy, however, will be successful only if the
Figure 5.2 Marginal effect of Share of Seats on Weighted Share of Ministers as the Ideo-
logical Distance between one faction and the median increases.
Note
Comparison between a small (10 percent of seats in party body, dashed line) and a strong minority
faction (40 percent of seats, solid line); the 90 percent confidence interval of the estimates is
displayed.
threat is credible and harmful. In fact, the party leader will take into consideration
requests from a minority faction only in such situations. The empirical analysis
seems to confirm this pattern. To keep party unity, the leader caters to minority fac-
tions with a share of seats greater than their actual size. When the likelihood of
their breakaway increases these factions receive a growing amount of cabinet posts
as an incentive not to leave the party. Even so, this happens only when minority
groups are strong, while small and weak factions do not carry any real blackmail
power and therefore their stakes are completely ignored. In sum, through portfolio
allocation the party leader compensates to those minority factions whose threat of
exit is credible and harmful for the party. In order to provide them with a reason to
stay inside the party, their lower amount of policy payoffs is balanced through a
more than proportional share of office payoffs (Ceron 2014; Warwick 1998).
Inter-factional portfolio allocation in France and Germany

The previous findings are in line with the theoretical framework. To evaluate the
robustness of these results, however, I replicate the analysis on French data. As
discussed in the previous chapters, due to data availability the number of obser-
vations is lower than in the Italian case. Accordingly, only the main hypotheses
will be tested.
Following what has been done so far, I analyze inter-factional bargaining
using data on intra-party policy preferences related to parties in office (see
Chapter 2). Factional affiliation of ministers has been evaluated by looking at the
signatures of competing factional motions or at ministers’ biographies.
Overall, I account for ministerial allocation in 17 French governments, from
1981 (Mauroy II) to 2016 (Cazeneuve), analyzing data on 94 party factions
belonging to the PS (78) or the UMP (16). In four cases (Ayrault I, Fillon I,
Mauroy I and Rocard I), the same government was formed twice, before and
after the legislative elections (i.e., the new cabinet was formed only a few weeks
later than the previous one). I excluded these four cases from the analysis con-
sidering these twin cabinets as a unique inter-factional bargaining. However,
duplicating these observations does not alter the results.
The dependent variable is Share of Ministers: i.e., the ratio of cabinet posts
belonging to faction i over the total number of ministers assigned to party j in
government k. For simplicity, the ministers have not been weighted according to
their relative importance, given that in the Italian case using the weighted or the
unweighted measure produced the same findings.
The independent variables are the following. Share of Votes (SofV): the per-
centage of votes won by each faction in the party congress (H5.1). Party Leader
Faction: a dummy variable equal to 1 for the faction to whom the party leader
belongs (H5.2). Direct Election: equal to 1 when the party leader is directly
elected during the party congress by party delegates or party members (more
autonomous leader). Ideological Distance: corresponds to the absolute distance
between the position of faction i and the weighted mean of all party factions.
Years Before Election: measures the number of years remaining before the next
general elections.
This latter variable accounts for party system competitiveness. In H5.5 I
hypothesized that when the level of competition shrinks (ω is low), the leaders
of ruling parties will not fear the risk of losing office and the incentives to pursue
party unity through compromises is lower. In the French context, the semi-
presidential system is combined with a two-round majority electoral system,
which tends to produce safer parliamentary majorities. In this regard, the number
of years remaining before the next election can be a better proxy for the level of
party system competitiveness, if compared to the margin of the ruling coalition,
and one can argue that competition grows as new elections approach. Indeed, in
the case of looming elections, any potential defections of minority factions can
damage the party (ω is high), raising the leader’s incentive to focus on unity and
to avoid overpaying its faction.
To test the hypotheses that lay at the core of my theory, I include in the ana-
lysis some interaction terms. In particular, I test the interaction between Party
Leader Faction and Direct Election (H5.4), and the interactions of Share of Seats
with YBE (H5.5) and Ideological Distance (H5.6).
Data have been analyzed by means of a fractional logit model (Papke &
Wooldridge 1996). To deal with compositional data (Honaker et al. 2002) one
random observation per each intra-party bargaining process has been dropped
from the sample. Given that there are repeated observations (factions) nested by
government, I clustered them on governments, providing standard errors
accordingly.
Table 5.3 provides the results of the analysis. Model 1 tests Gamson’s Law
(H5.1) against non-cooperative Bargaining Theory (H5.2). Model 2 assesses the
interaction between Party Leader Faction and Direct Election (H5.4), and the
impact of party system competitiveness (H5.5). Model 3 evaluates the moder-
ating effect of Ideological Distance (H5.6).
To start with, the findings of Model 1 in Table 5.3 are very similar to the results
displayed in Table 5.2 (Model 1): Share of Votes is positive and statistically signi-
ficant confirming that faction size matters in portfolio allocation (H5.1). However,
there is no perfect proportionality in France as well. For a one-unit increase in
Share of Votes, the expected growth in the Share of Ministers amounts to 0.71 (the
99 percent confidence interval ranges from 0.52 to 0.90). In turn, in Model 1, the
coefficient of Party Leader Faction is significant and positive as well suggesting
that party leaders can exploit their setter role to extract larger benefits, in line with
the bargaining theories (H5.2). Interestingly, Model 2 highlights that the leader’s
advantage holds only in the case of Direct Election, when the leader is more
Table 5.3 Portfolio allocation among party factions in France
DV: Share of Ministers (1) (2) (3)
Share of Votes 10.975*** 16.914*** 9.796***

(1.392) (2.311) (1.642)
Party Leader Faction 0.572* –0.056 –0.022
(0.317) (0.341) (0.353)
Direct Election –1.830*** –1.826***
(0.497) (0.608)
PLF X Direct Election 2.457*** 2.485***
(0.777) (0.949)
Years Before Election 0.471*** –
(0.147)
SofV X YBE –1.846*** –
(0.595)
Ideological Distance – –0.634
(0.826)
SoV X Ideological Distance – 8.300
(8.562)
Constant –4.441*** –5.461*** –3.725***
(0.403) (0.593) (0.374)
Log-likelihood –11.754 –10.748 –10.951
N 70 70 68
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. Clustered standard errors in parentheses.
autonomous. Therefore, even in the French case different modes of leadership
selection lead to different patterns of portfolio allocation. Differently from the
Italian case, however, French party leaders seem to exploit their discretionality to
overpay their faction (this provides support for H5.4). Such discrepancy might be
explained by the fact that, as argued above, the French electoral system produces
safer majorities: compared to Italy, in France the overall level of competitiveness
remains lower; as a consequence, the party leader, free from risks, can fully exploit
his autonomy to reward his followers.
However, when digging in more depth into the effect of party system competit-
iveness one notices that in the case of looming elections the degree of proportional-
ity in portfolio allocation increases a lot. When elections are approaching, for the
sake of preserving internal unity, party leaders tend to adopt criteria of perfect pro-
portionality. The expected marginal effect of a one-unit increase in the Share of
Votes leads to an increase in the Share of Ministers that is statistically close to 1 in
the electoral year or in the year before (see Figure 5.3).
Once again, this is in line with H5.5 and provides evidence for the idea that
when party system competitiveness rises (ω is high), the importance of party unity
grows and this is an incentive to adopt less conflictual dynamics (as observed also
in Chapters 3 and 4; see also: Ceron 2012). Indeed, leader’s incentive to focus on
unity gets stronger, generating a proportional allocation of spoils.
Figure 5.3 Marginal effect of Share of Votes on Share of Ministers as new general elec-
tions approach.
Note
95 percent confidence interval is displayed.
Finally, the French analysis provides further support also for H5.6. In
Figure 5.4 I report the results of the interaction between the minority’s Share of
Votes and its Ideological Distance from the bulk of party members. The figure
displays the marginal effect of Share of Seats (with a 95 percent confidence
interval) for two differently sized non-mainstream factions that won respectively
10 (dashed line) and 35 percent (solid line) of votes in the party congress. As in
the Italian case, the smaller faction is underpaid: a 10 percent increase in size is
rewarded by only an additional 4 percent of ministers. Conversely, the powerful
minority receives a more than proportional return. Furthermore, any increase in
the distance between this subgroup and the weighted mean position of all fac-
tions (i.e., any increase in the likelihood of its breakaway) generates more bene-
fits for both factions, but such gain is much wider and statistically significant
only for the strong minority, which can exploit its stronger blackmail power to
get a larger compensation in terms of office payoffs (to balance the lower amount
of policy payoffs that this faction receives from party membership).
To conclude, let us discuss portfolio allocation in Germany. Here I will not
provide any statistical analysis to compare the size of party factions with their
share of portfolios. In fact, I only have scant information (or no information at
Figure 5.4 Marginal effect of Share of Votes on Share of Ministers as the Ideological
Distance between one faction and the core of the party increases.
Note
Comparison between a small (10 percent of votes in party congress, dashed line) and a strong
minority faction (35 percent of votes, solid line); the 95 percent confidence interval of the estimates
is displayed.
all) on the size of German subgroups; in addition, it is harder to collect informa-
tion on factional affiliation of ministers given that I cannot rely on their signa-
tures on congress motions. In view of that, the patterns of portfolio allocation
will be examined combining a secondary data reading with qualitative evidence
from current German politics.
First of all, Debus and Bräuninger (2009) investigated the link between fac-
tionalism and coalition bargaining in a case study concerning six processes of
coalition government formation in Germany, from 1987 to 2005. They investi-
gated the conditions under which members of intra-party factions that hold
divergent preferences with respect to the core of the party are appointed to office.
To do that, they gathered data on the factional affiliation of German ministers
(and junior ministers) concerning the Ministry of Labor and that of Justice. This
information was combined with data on the policy preferences of factions and
intra-party subgroups (see the related discussion in Chapter 2).
Despite the fact that party leaders are members of (or at least responsible to)
the party core, which includes the median party member, Debus and Bräuninger
(2009) argued that one should observe, more often than not, that ministers from
extreme intra-party groups are appointed to office; in fact, party leaders can
appoint these politicians as a strategy to increase the bargaining leverage of their
own party (with respect to the coalition partner) and to maximize the policy
benefits of the party core (for a different interpretation concerning Grosse Coali-
tions in Italy: Ceron 2017).
Indeed, the results of their analysis suggest that the allocation of ministers
and junior ministers from extreme intra-party groups can be used to increase the
policy distance between the ministerial agenda setter and the coalition partner
and hence to foster the bargaining leverage of the party.
Taking the cue from this, I re-examine the data reported in that study, expand-
ing them in order to investigate the allocation of ministers to party factions in
German parties. By doing that, it can be assessed whether portfolio allocation is
also related to intra-party concerns. First, I can evaluate if the process is affected
by the need to represent the stances of all the different intra-party subgroups;
this somewhat refers to the idea of a fair or egalitarian, if not proportional (H5.1;
Gamson 1961), allocation of spoils. Conversely, I can check whether the party
mainstream (i.e., the party core) is overrepresented in ministerial allocation.
Finally, I will also discuss whether allocation choices can be influenced by party
leader’s concerns about party unity, following the argument that led to H5.6.
To start with, I investigate intra-party portfolio allocation in the SPD, over
the last 20 years, from 1998 to 2018. I collected data on the factional affiliation
of all SPD ministers in five governments. Two are red–green coalitions, involv-
ing the SPD and the Greens (cabinets Schröder I and Schröder II), while the
other three are Grosse Coalitions between the SPD and the CDU/CSU (cabinets
Merkel I, Merkel III and Merkel IV).
Overall, 50 ministers have been considered, including those who replaced
their colleagues during the same cabinet. By looking at their biographies (Debus
& Bräuninger 2009), available on Wikipedia, I evaluated whether they belong to
any of the SPD factions. Ministers have been classified in four categories: minis-
ters belonging to the SK (or clearly leaning to the right-wing of the party, even if
not publicly or officially affiliated to the Seeheimer Kreis); ministers belonging
to the left-wing, without separating those affiliated to the PL or DL21 (unaffili-
ated left-leaning ministers are considered too); ministers belonging to the NB;
ministers not clearly affiliated to any subgroup, which can be considered as rep-
resentative of the party core or close to the party mainstream. I detected a clear
leaning for 24 ministers (48 percent).
At first sight, one can argue that the party core, namely the mainstream group, is
widely represented given that 52 percent of ministers did not have a clear factional
leaning. However, it is worth pointing out that all the SPD factions have been
represented over time. In fact, I found left-wing ministers, such as Heidemarie
Wieczorek-Zeul, Andrea Nahles or Katarina Barley, and right-wing ones, includ-
ing Steinmeier or the SK’s Ulla Schmidt as well as conservatives like Bodo
Hombach, Wolfgang Clement and Olaf Scholtz. In turn, I also accounted for a
number of ministers considered close to the newest faction, i.e., the Berlin
Network; among them Kurt Bodewig, Steinbrück and Gabriel (notice that the latter
two, as well as Steinmeier, can be considered affiliated both to the NB and SK).
Interestingly, over the entire period, ministerial affiliations seem to mirror the
average strength of the three groups from 1998 on. On the whole, there are ten
ministers belonging to the SK (41.7 percent), nine belonging to the left (37.5
percent) and five coming from the NB (20.8 percent). When looking at temporal
trends, one notices that the left-wing is much more represented in the most
recent years, when it is actually stronger than it used to be. Net of unaffiliated
ministers, in the Merkel III and Merkel IV cabinets, approximately 50–60
percent of SPD portfolios can be considered assigned to internal left-wing politi-
cians. This share is in line with the 45–50 percent of MPs affiliated to the Parlia-
mentary Left in the 2013 and 2017 Bundestag. In terms of portfolio, the
Seeheimer Kreis was stronger from 1998 (Schröder I) until 2009 (Merkel I),
when the right-wing of the SPD was indeed the winner of intra-party competi-
tion, while it lost relevance in the most recent years, losing ministers too.
Arguably, this picture can be in line with the argument of Debus and
Bräuninger (2009), suggesting that in coalition government parties appoint
ministers belonging to the faction that is ideologically distant from the rival
partner. Indeed, more right-wing SPD ministers were appointed during the
red–green coalition. Conversely, more left-wing members got office in the
Merkel’s cabinet, with the important exception of Merkel I though; in this
cabinet, in fact, there was only one left-wing SPD minister versus seven right-
leaning SPD ministers.
In this regard, such distribution of ministers across SPD factions seems to be
in line with the relative strength of these rival factions and with the changes in
their reciprocal power over time. Although I cannot provide evidence of a
perfect proportionality, it seems that proportionality norms might somewhat
apply, so that a fair allocation emerges and all factions tend to be involved in the
distribution of office payoffs (this can be partially in line with H5.1).
Finally, I argue that portfolio allocation inside the SPD can be affected by the
need to preserve party unity too. In the most conflictual moments (temporally
close to the only two episodes of contested SPD leadership races), factions that
strongly disagreed with the party leadership and with the party line were repres-
ented in the cabinet.
As an example, the right-wing SPD Prime Minister Schröder appointed one
minister belonging to the Parliamentary Left (Wieczorek-Zeul). Furthermore, in
1998 he also involved his main left-leaning internal opponent, Oskar Lafontaine,
even though Lafontaine resigned soon (1999) and split from the SPD (2005) to
form an electoral alliance with the PDS (see the discussion in Chapters 1 and 6).
Nevertheless, Schröder’s strategy was probably aimed at keeping the party
together in a delicate period, in which the SPD was moving toward the center of
the political space, promoting liberal–democratic reforms of the labor market
and cutting welfare state provisions.
Analogously, in recent years, the SPD is losing votes; in the 2017 general
elections its advantage versus the Greens and Die Linke has been narrower
than it used to be and its electoral performance is worsening up to the point
that in 2018 the Greens have been challenging the role of the SPD as the first
party of the center-left in some Länder elections (i.e., Bavaria and Assia) and
in the nationwide opinion polls. Furthermore, left-wing factions started to
heavily criticize the Grosse Coalition with the CDU, questioning this strategic
choice; indeed, the party mainstream struggled to approve the formation of
the Merkel IV cabinet: in the 2017 intra-party referendum only 66 percent of
SPD members voted in favor of the Grosse Coalition, signaling that internal
division was huge and left-wing factions were almost able to influence the
party’s strategy. In this scenario, more left-wing politicians have been
appointed to office.
In view of that, it seems that when party system competitiveness increases or
when unity is at stake (i.e., there are large dissatisfied factions that might split,
weakening the party) portfolio allocation can become a strategy to keep the party
together even in the German case. In sum, this overview of ministerial allocation
in the SPD can support H5.6.
With respect to the CDU, I rely on Debus and Bräuninger’s (2009) findings
related to the last three Helmut Kohl cabinets (1987–1998) and on my scrutiny
of the Merkel IV cabinet. In a nutshell, that study shows that even if the MIT
was the most influential group inside the CDU, members of the left-leaning CDA
were repeatedly appointed to office (especially as ministers or junior ministers of
Labor). In this sense, all the subgroups were represented in the cabinet.
Despite not being divided into organized factions, the CDU in recent years is
experiencing high intra-party division. The centrist policies implemented by
Angela Merkel, particularly with respect to immigration policy, has fostered
internal disagreement. Two or three main party factions seem to form. One (Union
of the Center) is gathering around the new CDU Party leader, Annegret Kramp-
Karrenbauer, to continue the centrist policies pursued by Merkel. Conversely,
conservative party factions are forming around the right-wing CDU politicians Jens
Spahn and Friedrich Merz (the two challengers of Kramp-Karrenbauer in the 2018
leadership race).
Furthermore, an organized faction was already formed: in March 2017, conser-
vative party members created the Liberalist–Conservative Movement Freiheitlich-
konservativer Aufbruch – die WerteUnion (Union of Values, FKA). This tiny
conservative faction (only 1,000 affiliates out of 420,000 CDU party members)
aims to take back the CDU from Merkel’s centrist position, restoring the traditional
and conservative roots of the party. They launched a conservative manifesto, trying
to create a stronger network among conservative and economic liberal forces
within the CDU in order to challenge the leadership of Angela Merkel and run in
the 2018 congress. For this purpose, they organized as a faction well before Mer-
kel’s decision to step back from renomination (taken in October 2018, after the
electoral defeat in Assia). Although the threat of this specific faction can appear
weak, there have been other signs of strong disagreement about Merkel’s leader-
ship. For instance, in September 2018 the CDU/CSU Parliamentary Party Group
(PPG) voted against the party whip loyal to Merkel (the incumbent Volker Kauder)
and the MPs appointed instead a different PPG chair: Ralph Brinkhaus. Elected
with a narrow margin (125 votes against 112), the new chair stated that the PPG
should be more independent from Merkel’s viewpoints.
Excluding the CSU MPs, one can estimate that at least 40–45 percent of CDU
MPs voted against the leader’s line. This episode suggests that internal disagree-
ment is big enough and internal conservative subgroups are dissatisfied enough
(in terms of policy payoffs) to raise concerns about party unity.
What is more, party system competitiveness is increasing and the CDU now has
a strong challenger on the right-wing side of the policy space. Founded by former
CDU members (Dilling 2018), the Alternative for Germany (AfD) is rapidly
gaining support, boosting the need to preserve the unity of the CDU/CSU.
In this regard, it comes as no surprise that in the Merkel IV cabinet several
right-wing politicians have been appointed to office. Among others, Merkel
appointed the young Jens Spahn, one of her most outspoken critics. The conser-
vative Spahn wants to return the party to its right-wing roots, as an attempt to
regain the support of former CDU voters that are now leaning to the AfD.
Indeed, Spahn heavily criticized the immigration policy of Merkel’s cabinets and
has complained against the alleged “social-democratization” of the CDU enacted
by Merkel’s faction. The allocation of a ministerial position to Spahn has been
seen as a tactical move in order to win back delegates’ support in the early 2018
special party conference, after the threat of a right-wing rebellion against Mer-
kel’s leadership. Furthermore, Merkel also attempted to appease rebels and dis-
senters in order to promote unity, reducing the risks of split in case of the
creation of a new right-wing party faction loyal to Spahn.
On top of that, the outcome of the 2018 CDU congress revealed that internal
disagreement on Merkel’s line is even broader. In fact, this was the first con-
tested CDU leadership election since 1971, with three candidates and two
ballots. In the first ballot, Annegret Kramp-Karrenbauer (Merkel’s favorite
candidate) won a plurality (45 percent), whereas the two candidates of the
conservative factions, Friedrich Merz (39.2 percent) and Spahn (15.7 percent),
cumulatively won 55 percent of congress votes. In the run-off, Kramp-
Karrenbauer managed to become party leader though with a tiny margin (51.7
percent). The defeat after such a tied race pushed several conservative members,
in disagreement with the final outcome, to quit the party; despite the leadership’s
appeals for party unity, there are rumors about a potential party fission. Soon
after the congress, Alexander Mitsch, leader of the WerteUnion, highlighted that
the possibility of a breakaway should not be ruled out; accordingly, as proof of
good will in order to restore unity, he suggested that the new party leadership
should appoint prominent conservative members including Merz o Carsten Lin-
nemann (the leader of the MIT) as ministers in the Merkel cabinet.19 Indeed,
such a reshuffle would be a sign that the CDU leadership is taking into account
the changing relative strengths of its internal subgroups and the huge rise in size
of conservative CDU factions. This example, once again, points to the validity
of the theoretical framework discussed in Chapter 1, and the requests of more
portfolios made by the new powerful CDU minority factions are indeed in line
with the hypotheses presented above.
To sum up, even inside the CDU the stakes of all the different intra-party sub-
groups seem to be taken into account in portfolio allocation (as suggested in
H5.1). In line with H5.6, this attitude is reinforced in the presence of heightened
party systems competitiveness, especially now as the CDU is more and more
internally divided and externally challenged by the right-wing AfD.
Portfolio allocation: findings

Empirical evidence supports the idea of a proportional allocation of portfolios
among party factions. Overall, faction size matters and cabinet spoils are distrib-
uted according to the strength of each faction. Nevertheless, there are relevant
deviations from the pattern of proportionality under particular conditions.
Accordingly, it can be argued that Gamson’s rule does not necessarily apply to
all intra-party actors.
First of all, a significant advantage exploited by the party leader faction can
be observed; even controlling for faction’s size this effect still holds. Acting as
the proposer in a setter game, the party leader is able to reward his followers
with an amount of payoffs greater than their strength in the party body (this also
allows a reduction in agency costs, minimizing the distance between ministers
and their principal, i.e., the party leader: Bäck et al. 2016). This evidence is in
line with the bargaining theories that predict larger payoffs for the “formateur”
in spite of proportional allocation.
These results are summarized in Figure 5.5, which displays, for Italy and
France, the linear fit between the share of votes and the share of ministers for the
faction of the party leader (black) or for non-mainstream factions (gray). Indeed,
the picture reveals a relationship between factions’ votes and ministers. The
leader’s faction, however, seems able to extract a larger reward, as can be
noticed by the gap between the black and gray lines.
Figure 5.5 Factions’ votes and shares of ministers in Italy and France.
Note
Party leader (black) or non-mainstream factions (gray) are kept separated. The solid lines express the
fitted value of the related regressions. The 95 percent confidence interval of the estimates is dis-
played (dashed lines).
On top of that, statute rules seem to affect inter-factional bargaining. The

direct election of party leaders decreases the degree of proportionality in port-
folio allocation. To provide the public good (party unity and party strength),
the leader identifies and provides selective incentives that ensure the mobil-
ization of all factions. More autonomous leaders can exploit their role either
to provide more rewards to their followers or to keep party unity answering to
the requests made by internal minorities. In the Italian case, the findings
suggest that this second effect is prevailing. Besides rewarding their sup-
porters, party leaders address the requests of powerful minorities (whose
leverage is wider: high π) in order to solve all the credible threats to party
unity. In this regard, portfolio allocation becomes a strategy to ward off
potential harmful threats and to balance the total amount of returns for those
factions whose policy payoffs are lower, preserving party unity and party
strength.
This aspect is even more relevant in competitive parliamentary arenas, where
the margin of ruling parties is lower and any single split might alter the balance
of power between government and opposition. On the contrary, when the cabinet
retains a wide parliamentary support, party leaders will be able to exert a larger
degree of discretionality in portfolio allocation, being able to enhance their own
stakes and the payoffs of the mainstream faction.
This general framework seems to be confirmed also when looking at inter-
factional portfolio allocation in France and Germany. The French case strongly
corroborates the findings of the analysis on Italian factions. In France too there
is evidence suggesting that faction size matters in portfolio allocation (H5.1).
Analogously, one can also notice an advantage for the party leader’s faction
(H5.2) limited to directly elected leaders that are more autonomous from fac-
tional constraints.
While in Italy leader’s discretionality is used mainly to promote party unity,
French leaders (facilitated by secure parliamentary majorities produced by the
majoritarian electoral system) tend to overpay their own faction (H5.4). None-
theless, when party system competitiveness rises (ω is high), i.e., in the case of
looming elections, portfolio allocation embraces criteria of perfect proportional-
ity (this supports H5.5). In turn, like in the Italian case, powerful minorities seem
to be rewarded with a more than proportional return when they are markedly dis-
satisfied in terms of policy payoffs (H5.6).
In Germany, inside the SPD and the CDU all factions tend to be involved in
portfolio allocation. Apparently, faction size matters too, at least when con-
sidering the SPD. In line with Gamson’s hypothesis on proportional allocation
of spoils (H5.1), it seems that SPD factions get more ministers when they are
stronger. Nevertheless, in both parties the party core retains the majority of
portfolios.
In turn, German parties also reward powerful minorities in order to preserve
party unity, particularly when parties are deeply internally divided and could be
potentially hit by a split (H5.6). Party leaders appointed to office representatives
of internal minorities dissatisfied by the policy course of the party; this strategic
choice is clearly evident in the case of increasing party system competitiveness,
i.e., when the SPD and the CDU are challenged in the electoral arena by rival
parties located on their wings (the Greens and Die Linke or the AfD).
To sum up, the analysis of inter-factional conflict and cooperation on port-
folio allocation shows that each subgroup tries to maximizes its own payoffs,
balancing policy and office payoffs altogether. For instance, party leaders
attempt to reward the mainstream faction (that contributed to their election)
while minority factions may threaten to defect in order to bargain a greater sum
of payoffs. However, both the mainstream and the minority factions experience
the advantages of cooperation. In fact, autonomous party leaders might give up a
quota of their own payoffs in order to keep party unity, thereby increasing the
total amount available to all factions in the long run. In this regard, although one
might not always find patterns of perfect proportionality, there are reasons to
believe that parties are ruled according to consensual dynamics that reward each
faction according to its actual bargaining power, whose measurement encom-
passes both its size, its level of policy payoffs and its strategic position within
the party, but also the potential damage in the case of a split. This latter feature
will be largely addressed in the next chapter.
Notes
1 This might concern both the electoral campaign (particularly when factions compete
for preference voting) and everyday party politics that also involves fundraising and
membership drive, which are linked directly to factional power and party strength.
2 Analyzing data on 145 DC factions Mershon (2001a) found that a one-unit increase in
the share of seats held in the DC National Council raises the faction’s share of minis-
ters by 0.78; when considering data on junior ministers, she analyzed 74 factions and
found a perfect proportionality: per each 1-point increase in seats the faction’s share
of undersecretaries increased by 1.05.
3 Nonetheless, empirical evidence from experimental economics about the Ultimatum
Game sends us back to H5.1, providing confirmation for an alternative theoretical
solution according to which the payoffs will be split fairly (e.g., Güth et al. 1982).
4 Boucek (2010: 127) highlights that majoritarian arrangements within the LDP pushed
intra-party politics toward a bipolarized model of competition. In such context, even
if minority groups were not excluded from payoffs allocation, the mainstream factions
(which supported the elected party leader) won the lion’s share of portfolios (Boucek
2010; Di Virgilio & Kato 2001).
5 Warwick (1998) suggested a similar reasoning for coalition governments: a greater
distance between one party and the cabinet’s ideal point requires a higher attribution
of cabinet posts to bring the party inside the coalition.
6 In this regard, this dataset is larger if compared to the single party study provided by
Mershon, which analyzed portfolio allocation among 163 DC factions in 20 Italian
governments.
7 I could not revert to the solution proposed by Honaker et al. (2002) because data
related to each intra-party bargaining process do not refer to the same set of actors.
8 Other authors instead adopted the beta maximum likelihood estimation (Ono 2012).
However, given that 40 percent of factions retain a share of payoffs that is either 0
or 1 the fractional logit ends up to be more appropriate (Papke & Wooldridge
1996: 620).
9 Following the idea that different parties might attach a different importance to the
same cabinet position, another dependent variable that takes into account a party-
specific weight per each ministry can be proposed. Such an approach has been
developed by Bäck et al. (2011) who used CMP data to evaluate party salience on
each issue. Here I measured the party weight through the content analysis of parlia-
mentary debates (Ceron et al. 2019), to estimate the salience that each party assigns to
a particular policy area linked with a specific cabinet portfolio. When using this party-
specific dependent variable, which is strongly correlated (0.99) with the Weighted
Share of Ministers, the results hold the same.
10 When this value was unavailable, the share of votes in party congress has been used.
11 I also controlled whether these two theories only apply when the party system is
stable or fluid. In fact, party system fluidity may loosen the link between each intra-
party subgroup and the party as a whole. In this context, portfolio allocation may
become a non-repeated game, thus decreasing the incentives for factional cooperation.
To account for this, I tested the interactions between Share of Seats and Party Leader
Faction with the dummy variable Party System Fluidity (which is equal to 1 in the
Italian Second Republic, when the party system was indeed more fluid). The results
suggest that the leader seems to extract higher payoffs only in a fluid system, although
there is no difference when comparing the marginal effect of Share of Seats in the
First or the Second Republic: in both contexts the Gamson rule seems to work.
12 Testing the same models on a subsample that includes only “formateur” or “non-
formateur” parties wields the same substantive results.
13 If including in Model 1 a variable that measures the power index of each faction
within the party body, this covariate will not yield any significant effect. I also
controlled whether the proportionality norm increases over time. In fact, scholars
argue that repeated interactions between factions might lead them to discover the
advantages of cooperation, fostering proportional distributive dynamics (Boucek
2010); studies found that LDP factions developed proportionality norms over time
(Bouissou 2001). However, when testing the interaction between Share of Seats and
party age no differences were found in the implementation of proportional allocation
between older or younger parties. This might happen because, even if repeated inter-
action fosters cooperation, in older parties the effect of party loyalty is stronger and it
enhances the exit costs (ε) so that minority factions will not use the voice option to
criticize a not fully proportional allocation of payoffs and can accept an unfair distri-
bution (in line with my theory, see Chapter 1). To sum up, these two effects might
counterbalance each other.
14 It could be argued that this result is inflated by the presence of parties that hold a low
number of cabinet posts; here the minister(s) would be more likely assigned to the
mainstream faction at the expense of the minority. However, when excluding these
parties from the analysis the results hold and the advantage of the leader’s faction
seems even strengthened. The effect of Party Leader Faction remains significant even
after controlling for a dummy variable that distinguishes the median faction, as in a
unidimensional space the median faction can retain a stronger bargaining position and
can obtain larger payoffs (Bouissou 2001: 584; Mershon 2001a, 2001b). The effect of
this latter variable is not significant though. Notice that the median faction and the
party leader faction often coincide. Furthermore, apart from a few exceptions, party
leaders tend to belong to the median faction. This allows potential agency losses that
could be higher when leaders’ preferences are extreme to be minimized (Kiewiet &
McCubbins 1991).
15 With respect to intra-party rules, the effect of disproportional internal electoral
systems has also been tested (see Chapter 3). However, the results indicate that the
internal electoral system plays a role only by transforming votes (won by factions
during the congress) into seats (inside the party body). No further “winner-takes-all”
dynamics were found beyond this mechanical effect. As such, what really matters in
terms of rules is the direct election of the party leader.
16 This is in line with Mershon’s claim (2001a: 573): “Once direct elections for the DC
secretary general were instituted […] the norm of proportionality exerted a rather
weak influence on the distribution of ministerial posts among DC factions.”
17 Notice that one does not necessarily expect an increase in proportionality given that
some factions can exploit their blackmail power to extract a share of payoffs greater
than their strength.
18 This result also holds when considering the Italian First Republic only. Moreover, I
did not find significant differences between autonomous or constrained leaders, even
though the mainstream faction could hope to receive a slightly larger bonus when the
leader’s autonomy is stronger.
19 www.augsburger-allgemeine.de/politik/Konservativer-Unions-Fluegel-warnt-vor-
tiefer-Spaltung-der-CDU-id52901291.html.
References
Adachi, T., & Watanabe, Y. (2008). Ministerial Weights and Government Formation:
Estimation using a Bargaining Model. Journal of Law, Economics, and Organization,
24(10), 95–119.
Ansolabehere, S., Snyder Jr, J.M., Strauss, A.B., & Ting, M.M. (2005). Voting Weights
and Formateur Advantages in the Formation of Coalition Governments. American
Bäck, H. (2009). Intra-party Politics and Local Coalition Formation. In Giannetti, D., &
Benoit, K. (eds) Intra-Party Politics and Coalition Government, 53–68, Routledge,
London.
Bäck, H., Debus, M., & Dumont, P. (2011). Who Gets What in Coalition Governments?
Predictors of Portfolio Allocation in Parliamentary Democracies. European Journal of
Political Research, 50(4), 441–478.
Bäck, H., Debus, M., & Müller, W.C. (2016). Intra-party Diversity and Ministerial Selec-
tion in Coalition Governments. Public Choice, 166(3–4), 355–378.
Baron, D., & Ferejohn, J. (1989). Bargaining in Legislatures. American Political Science
Review, 83(4), 1181–1206.
Science, London.
Bouissou, J.-M. (2001). Party Factions and the Politics of Coalition: Japanese Politics
under the “System of 1955.” Electoral Studies, 20(4), 581–602.
Browne, E.C., & Franklin, M.N. (1973). Aspects of Coalition Payoffs in European Parlia-
mentary Democracies. American Political Science Review, 67(2), 453–469.
Browne, E.C., & Frendreis, J.P. (1980). Allocating Coalition Payoffs by Conventional
Norm: An Assessment of the Evidence from Cabinet Coalition Situations. American
246–258.
Ceron, A. (2017). Social Media and Political Accountability: Bridging the Gap between
s11127-019-00644-0.
Dilling, M. (2018). Two of the Same Kind? The Rise of the AfD and its Implications for
the CDU/CSU. German Politics and Society, 36(1), 84–104.
Fréchette, G., Kagel, J., & Morelli, M. (2005). Gamson’s Law versus Non-cooperative
Bargaining Theory. Games and Economic Behavior, 51(2), 365–390.
26(3), 373–382.
Rationalité et Sciences Sociales Colloque, Collège de France, Paris, June 3–4.
Güth, W., Schmittberger, R., & Schwarze, B. (1982). An Experimental Analysis of Ulti-
matum Bargaining. Journal of Economic Behavior & Organization, 3(4), 367–388.
Honaker, J., King, G., & Katz, J.N. (2002). A Fast, Easy, and Efficient Estimator for Mul-
tiparty Electoral Data. Political Analysis, 10(1), 84–100.
Kato, J., & Mershon, C. (2006). Internal Party Organization in the Italian Christian
Democrats and Japanese Liberal Democrats: Factional Competition for Office, Clien-
teles, and Corrupt Exchange. In Kawata, J. (eds), Comparing Political Corruption and
Clientelism, 77–114, Ashgate, Burlington VT.
Laver, M. (1998). Models of Government Formation. Annual Review of Political Science,
1(1), 1–25.
Laver, M., de Marchi, S., & Mutlu, H. (2011). Negotiation in Legislatures over Govern-
ment Formation. Public Choice, 147(3–4), 285–304.
Mershon, C. (2001a). Party Faction and the Coalition Government: Portfolio Allocation
in Italian Christian Democracy. Electoral Studies, 20(4), 555–580.
Mershon, C. (2001b). Contending Models of Portfolio Allocation and Office Payoffs to
Party Factions: Italy 1963–79. American Journal of Political Science, 45(2), 277–293.
encies of Modern Democracy. Trans. P. Eden & P. Cedar, Free Press, New York.
Morelli, M. (1999). Demand Competition and Policy Compromise in Legislative Bargain-
ing. American Political Science Review, 93(4), 809–820
Mutlu-Eren, H. (2015). Keeping the Party Together. Public Choice, 164(1–2), 117–133.
Ono, Y. (2012). Portfolio Allocation as Leadership Strategy: Bargaining among and
within Parties. American Journal of Political Science, 56(3), 553–567.
Panebianco, A. (1988). Political Parties. Organization and Power. Cambridge Press
University, Cambridge.
Papke L.E., & Wooldridge J.M. (1996). Econometric Methods for Fractional Response
Variables with an Application to 401(K) Plan Participation Rates. Journal of Applied
Econometrics, 11(6), 619–632.
Politics, 14(6), 663–683.
469–490.
Snyder, J.M., Ting, M.M., & Ansolabehere, S. (2005). Legislative Bargaining under
Weighted Voting. American Economic Review, 95(4), 981–1004.
Venditti, R. (1981). Il manuale Cencelli. Il prontuario della lottizzazione democristiana.
Un documento sulla gestione del potere. Editori Riuniti, Roma.
Wada, J., & Schofield, N. (1996). Bargaining in the Liberal Democratic Party of Japan. In
Schofield, N. (eds), Collective Decision-Making: Social Choice and Political Economy,
239–250, Kluwer Academic Publishers, Boston MA.
Warwick, P.V. (1998). Policy Distance and Parliamentary Government. Legislative
Studies Quarterly, 23(3), 319–345.
Warwick, P.V., & Druckman, J.N. (2005). The Missing Piece: Measuring Portfolio
Salience in Western European Parliamentary Democracies. European Journal of Polit-
ical Research, 44(1), 17–42.
Warwick, P.V., & Druckman, J.N. (2006). The Portfolio Allocation Paradox: An Investi-
gation into the Nature of a Very Strong but Puzzling Relationship. European Journal
of Political Research, 45(4), 635–665.
Wooldridge, J.M. (2002). Econometric Analysis of Cross Section and Panel Data. MIT
Press, Cambridge, MA.
6 The politics of fission: party splits
in Italy, France and Germany
Introduction
Parties are often treated as unitary actors, where individual members coalesce to
reach common goals. However, individuals or larger factions in parties must
overcome a variety of collective action problems in coordinating (Aldrich 1995;
Cox & McCubbins 1993; Kiewiet & McCubbins 1991). Many scholars have
looked at these issues in the context of party formation, but few have considered
how factional dynamics can make the unitary actor assumption untenable after
the initial formation (Ceron 2015; Ibenskas 2017; Ibenskas & Sikk 2017; Mair
1990; Mutlu-Eren 2015).
In fact, the literature has often considered parties as exogenous elements of
the political system, taking them as given. Far from being fixed and stable enti-
ties, parties are often involved in fusions or hit by fissions (Laver & Benoit
2003). Furthermore, Giannetti and Laver (2009: 146) highlight how, “In the real
political world […] it is often difficult to discuss the making and breaking of
parties without referring to factions or groupings of some shape and form.”
Accordingly, in this chapter parties will be considered as the output of inter-
factional bargaining to analyze the determinants of party fission.
Although fissions are a kind of rare event, they occur more often than one
might expect. Party splits are important sources of party system change in
several countries. Factionalism and internal disagreement can be sources of party
breakup worldwide. As an example, Mair (1990) recorded 34 major party
fissions occurring in Western European countries between 1945 and 1987. More
recently, Ibenskas (2017) provided data on 25 European countries in the post-
war period and found no less than 241 splits, with a peak in the 1990s (both in
Eastern and Western Europe), and a trend toward an increasing number of
fissions in the latest years (particularly in Western countries).
Indeed, even some party systems that had long been considered stable have
now experienced important splits in recent years. Let us consider, for instance,
Germany (the SPD split in 2005, see below; additionally, CDU members, includ-
ing the former MIT leader Gerd Robanus, contributed to the foundation of the
AfD in 2012: Dilling 2018), Spain (the centrist Union, Progress and Democracy,
UPyD was founded in 2007 by former politicians of the Spanish Socialist
136 The politics of fission
Workers’ Party, PSOE; the right-wing VOX, which obtained outstanding results
in the 2018 Andalusian regional elections, was founded in 2013 by splinter
members of the Partido Popular (People’s Party, PP; Gillespie et al. 1995), or
(a few more years ago) the UK.
Nevertheless, splits are more endemic in some (unstable) political systems,
including Italy, Japan (Cox & Rosenbluth 1995; Kato 1998, Reed & Scheiner
2003), France (Evans 2003) or Eastern European countries (Ibenskas & Sikk
2017), but also the Netherlands (Ibenskas 2017). In such contexts, fissions are an
ever-existing threat and an opportunity to reshape payoffs.
If parties are considered as endogenous organizations, one can argue that
“members of party factions may be seen to belong for as long as it is rational to
do so” (Laver & Kato 2001: 510).
Based on this, in my theoretical framework, I consider factions to be rational
actors coordinating their members’ behavior in order to maximize their own
share of payoffs. As a consequence, their decision to exit or adhere to the party
is based on a rational choice involving the estimated policy, office and electoral
payoffs. The interplay between factions and the party leader matters too, as the
leader can alter his responses to minority requests, depending on his interest in
either party unity (keeping the party together) or cohesion (enhancing agreement
on the party line).
By doing so, I will take on board some of the findings discussed in Chapters 4
and 5, related to the allocation of policy and office payoffs. This will allow an
evaluation of whether, as hypothesized, more consensual dynamics reduce the
likelihood of a party’s split. In line with the arguments proposed so far, the effect
of internal dynamics will be assessed in different organizational settings (con-
sidering changes in rules both inside the party and in the political system). In
turn, the impact of different levels of party system competitiveness will be taken
into account.
Party unity and party fission: theory and hypotheses

The game–theoretic model outlined in Chapter 1 highlights the relative power of
the party leader and the minority faction, showing how their interplay affects
intra-party distributive dynamics (i.e., portfolio allocation and party change) and,
consequently, the occurrence of party splits.
In the game, the likelihood of party fission is determined by two key ele-
ments, (1) the bargaining power of the minority, given by its ability to make a
credible threat, and (2) the party leader’s relative interest in unity or cohesion.
When the minority has a non-credible threat, the party leader has no incentive to
pursue a compromise as party unity is not threatened. Potential splinter groups
can exploit their bargaining power to demand a fair deal only when they have a
credible threat. In turn, the leader will accommodate the minority’s requests only
if a breakaway will damage the party more than internal dissent.
First of all, the choice to stay or to leave depends on the factions’ share of
payoffs. Factions compare the rewards available to them inside the party with
The politics of fission 137
the payoffs that they could obtain by splitting away. In view of that, fair and
consensual intra-party dynamics should promote party unity: when the leader
grants to factions a quota of returns proportional to their strength, the probability
to improve such quota elsewhere would be lower and potential defectors will
face fewer temptations.
Conversely when payoffs are inadequate (for instance, the party position lies
too far away from the faction’s ideal point or the subgroup is underpaid in terms
of cabinet spoils or candidacies) the likelihood of a party breakup rises.
Besides the benefits, however, factions have to also weigh up the costs of
their choice. The decision to break away, in fact, is subjected to constraints pro-
vided by intra-party rules and by the attributes of the party system (electoral
rules and competitiveness), which may increase (or decrease) the exit cost (and
the net gain of a split).
In presence of less democratic party rules, dissatisfied factions cannot
improve their condition by voicing from inside and the voice option could even
be denied. This happens, for instance, when the candidate selection process is
strongly centralized in the hands of the leadership.
Electoral rules matter too. When the electoral system provides high barriers
to entry (due to disproportionality, district magnitude, thresholds of representa-
tion and so on) the breakaways are no longer rewarding and party breakup
should become less likely.
The game of party unity and party fission involves party leaders too. In fact,
party leaders are in charge of allocating payoffs to different party factions and
are also responsible for preserving internal unity (which is a public good for the
partisan community). Leaders (according to their degree of autonomy) can
choose to follow a logic of fairness, to reward their supporters or alternatively to
allocate payoffs in order to preserve party unity heeding to the requests of
powerful potential splinter groups.
For the sake of preserving unity, leaders will answer the requests of the
minority as long as they feel that a fission will damage the party. Conversely, if
leaders feel that a split will not hurt the party because the splinter group is too
small or because the split can increase the cohesion of the group and the clarity
of the party label, the minority’s requests will be neglected. In this regard, com-
petitive electoral conditions that raise the need for party unity can certainly play
a role.
To sum up, the minority is more likely to split when the payoffs are greater
outside the party than inside (because of the unfair distribution) and exit costs
are low (for instance, if there is no strong loyalty or the electoral system does not
create barriers for new actors). Conversely, the leader will pander to the minority
when party unity is an issue at stake and fission would be a huge loss for the
party, for example if a ruling party has a narrow margin over the opposition. A
split is more likely when ideological disharmony and internal voice damage the
image of the party. For example, dominant or ruling parties rarely tolerate
dissent when they have a wide parliamentary margin. In such cases, party leaders
will focus more on cohesion than unity.
These implications suggest a number of specific testable hypotheses related to
the parameters of the game–theoretic model. I start by considering attributes that
affect the balance between the cost of exit (ε) and the net benefit of a breakaway
(π) for potential splinter groups to determine when a minority can gain from a
breakaway and hence has a credible threat to split off from the party.
Any element increasing the share of policy, office and electoral payoffs avail-
able to the splinter group after breaking away (π) will make a split more likely.
To start with, shared policy preferences structure both party membership
(Aldrich 1995; Snyder & Ting 2002) and factional affiliation (Bernauer &
Bräuninger 2009; Giannetti & Laver 2009). In turn, policy payoffs can be linked
with party splits too.1 In more detail, “divergent factional preferences […] create
splitting pressure” (Boucek 2009: 473).
Party members (and factions) incur a cost from party membership (Snyder &
Ting 2002). The cost is large when factions have ideal points far from the bulk
of party members and hence fewer payoffs from policy. A breakaway may thus
allow fringe factions to minimize membership cost and establish a party plat-
form with larger policy payoffs (higher π). Assuming that party position corres-
ponds to the weighted mean of all factions (Ceron 2012; Levy 2004; see also
Chapter 4), I propose the following hypothesis.
Hypothesis 6.1 (H6.1): A faction’s breakaway is more likely the greater the dis-
tance between that faction and the party position.2
Factions are of course also interested in office payoffs and career rewards.
They will consider their share of payoffs within the party and any potential gains
after a breakaway. Hence, “overpaid” factions with shares of office payoffs
greater than their vote share should be less willing to split.3 Conversely, “under-
paid” factions have larger expected payoffs (π) from defecting, and can be more
inclined to leave the party (Giannetti & Laver 2001; Laver & Kato 2001; Wada
& Schofield 1996).
Hypothesis 6.2 (H6.2): Higher/lower office payoffs relative to faction size

decreases/increases the likelihood of that faction’s breakaway.
Factions also consider electoral rewards, e.g., the number of parliamentary

seats they can win in future elections. More disproportional electoral rules pre-
serve party systems and act as barriers to new parties. They increase the exit
costs (ε) for dissidents4 and limit their ability to extract additional policy and
office payoffs in the long run, thus decreasing π. These effects should be greater
under single-member plurality systems, but can also apply under PR.
Hypothesis 6.3 (H6.3): A more disproportional electoral system decreases the

likelihood of a breakaway.
Candidate selection (the way MPs are selected) is another distributional issue
affecting party unity. The estimated costs and benefits of a split for potential
splinter groups depend on their chances of gaining seats. The party leader tends
to retain control over candidate selection under closed list PR and centralized
selection procedures (Carey 2007; Cox et al. 1999), and is thus able to exclude
dissenting factions. Minority factions that defy the leader fear being excluded
from the party list in retaliation. Thus, the limited room for dissent should
decrease exit costs (ε). Conversely, open list PR provides factions with access to
parliamentary seats through preference voting (Katz 1986; see also Chapter 3).
This institutionalizes factionalism and decreases the likelihood of splits.
Hypothesis 6.4 (H6.4): Closed list PR and centralized candidate selection

increase the likelihood of a breakaway.
Party loyalty affects the balance between the costs (ε) and the benefits (π) of a
breakaway. More specifically, partisan ties to symbols such as logos, labels
(Kertzer 1996) and the “logic of appropriateness” (Andeweg & Thomassen
2010; Kam 2009) internalized through participation might dissuade members
from leaving (Gehlbach 2006; Hirschman 1970; Kato 1998). This is particularly
relevant for older parties where party loyalty is well established and increases
the exit cost (ε).
Hypothesis 6.5 (H6.5): The loyalty effect decreases the likelihood of party
fissions in older parties.
In addition to features that provide minority factions with incentives or disin-

centives to break away, the model also suggests that leaders can have different
attitudes to party unity.5 Leaders may seek to preserve unity at any cost (when
ω > ν), or alternatively seek cohesion over a clear party line, with a willingness
to eliminate internal challengers at any cost (ω < ν).
Leaders focused on unity will accommodate potential splinter groups to
decrease the likelihood of party fission. At the same time, if internal dissent is
too damaging for the party, leaders will ignore the minority and make it more
inclined to leave.
Leadership’s attitude is related to internal democracy. Parties with intransi-
gent rules, for instance parties organized according to “democratic centralism”
(e.g., Marxist parties) typically do not tolerate “dissent.”6 Any public expression
of internal disagreement could weaken the party in the eyes of its voters. Under
“democratic centralism,” party members are free to discuss party strategy and
ideology, but can only disagree within the party. Once the party has established
a position, dissenting members must submit. In this context, the public expres-
sion of dissent is costly (higher ν) and the benefits of party unity lower. Thus,
the party leader would rather use the whip against the minority than seek a
compromise.
Hypothesis 6.6 (H6.6): The likelihood of a breakup increases in parties ruled

through “democratic centralism.”7
The cost of internal division (ν) is higher for ruling parties, which face greater
public scrutiny than the opposition. The competitiveness of the party system and
the margin of the ruling coalition affect the value of unity (ω) and shape the
balance between the costs of dissent and splits. When ruling parties have narrow
majorities, any split could jeopardize government stability. Leaders will be more
concerned with party unity given the higher cost of a breakup when the party
may lose office (large ω). A leader will tolerate internal divisions provided they
do not threaten unity. This gives party minorities bargaining leverage to extract
substantial shares within the party. Conversely, if the degree of competitiveness
is low (small ω) and the dominant coalition has a safe parliamentary margin
there is less need for party unity. Splits no longer threaten government stability
and party leaders may be less concerned over party fission. As such, leaders will
refuse to accommodate minorities, use the whip and conduct the intra-party
game to the breaking point (as in U4, see Chapter 1). In other words, leaders
trade unity for cohesion in order to improve government effectiveness and estab-
lish party authority.8
Hypothesis 6.7 (H6.7): Fission is less likely among ruling parties when the gov-
ernment has a narrow parliamentary margin and becomes more likely as the
margin widens.9
Party splits in Italy: analysis and results

As usual, the hypotheses will be tested starting from the Italian case. In both the
First (1946–1993) and the Second Republic (1994–present), Italian parties faced
bitter internal conflicts, often leading to splits and reshuffles of the party system.
For instance, the Italian Socialist Party experienced several fissions that gener-
ated new rival parties. In 1921, well before the Republic, the Italian Communist
Party was formed. Later on, between 1947 and 1949 (in three congresses in a
row), five socialist factions reluctant to form an alliance with the PCI and claim-
ing for more autonomy from the Soviet bloc, broke away to create the Italian
Socialist Democratic Party (PSDI). Other splits involved the socialist area in the
post-war period, including the breakaway of the Socialist Left from the PSI in
1964 (which formed the PSIUP) and a new split between the PSDI and PSI in
1969 (after their merger in 1966).
Other parties split too. The right-wing MSI broke up in 1977, when the
moderate faction Destra Popolare left the party. In turn, the PCI split in 1991.
When the party dissolved to create the ideologically moderate PDS, left-wing
communist minorities resisted to such ideological shift and founded the PRC.
In the Second Republic there are several other fissions mostly related to the
formation of new parties (as merger of existing organizations). The creation of
the PD (2007) and PDL (2008) triggered the breakaway of the extreme wings
inside the DS and AN, which were reluctant to merge with more moderate allies.
On the left side, the PRC split in 1998, and the formation of a more moderate
communist party (PDCI) was crucial to preserve the parliamentary majority of
the center-left coalition. Later on, in 2009, a new left-wing party Left Ecology
and Freedom (SEL) was founded by splinter groups of several left-wing parties
including the PRC, PDCI and the Greens.
In the present analysis I account for 30 breakaways related to the 254 factions
considered over the period 1946–2011 (see Appendix 2 for details). Overall,
11.8 percent of factions broke away.10
I identify a breakaway in all cases where a faction, its leader, or the majority
of its members split off from the party body after presenting a motion during a
congress. I consider all party splits, irrespective of their size, as even minor splits
may change the equilibrium between parties in parliament and jeopardize the
electoral performance of a party (see the discussion in Chapter 1).
Although I measure fissions in the “Party in Central Office,” most of them
also have consequences within the “Party in Public Office” and the “Party on the
Ground.”
The binary dependent variable, Fission, indicates whether faction i leaves the
party. This is not such a rare event, as there are splits in 26.5 percent of con-
gresses. There are two peaks in the distribution of Fission over time, one at the
beginning of the First Italian Republic and another one after the 2006 general
elections, when the party system was particularly fluid. Still, nearly one-third of
the breakaways occurred from 1960 to 1989 even though the Italian party system
is considered relatively stable during this period.
On average, there is a split every two years. As such, party fissions and fac-
tional breakaways are not a rare event, but an ever-present threat posed to the
party leadership and an opportunity for minority factions to negotiate a new dis-
tribution of payoffs (Giannetti 2010; Giannetti & Laver 2001; Laver & Kato
2001; Mershon 1996).
The unit of analysis is the faction-per-cabinet. This allows a better assessment
of the impact of office payoffs and parliamentary support, which might vary
between party congresses.
Several independent variables are adopted. To test H6.1, I consider Ideo-
logical Distance, i.e., the absolute distance between faction i’s ideal point and
the weighted mean of all factions in each congress.11 To test H6.2 I consider the
impact of office payoffs through the variable Extra Share of Ministers, i.e., the
difference between faction i’s share of ministers and size. I assigned a value of 0
for all parties not in office. To test H6.3 and H6.4 on the role of the electoral
system I consider Disproportionality (H6.3), measured through the Gallagher
index, and Closed List (H6.4), a dummy variable flagging closed list PR or plu-
rality systems with centralized candidate selection, as opposed to open list PR.12
I test H6.5 through Party Age, the number of years elapsed since a party’s found-
ing, and H6.6 by a dummy variable for parties ruled by means of Democratic
Centralism.13 Finally, I test H6.7 through an interaction between Ruling Party,
indicating parties in office, and Parliamentary Support, i.e., the inter-electoral
party system competitiveness measured by the size of the ruling coalition.
The hypotheses are tested through logistic regression. Data are encoded by
cabinet. I report robust standard errors clustered by cabinet in each party
congress to avoid possible problems from non-independent observations or non-
constant variances.14 Table 6.1 presents the results. In Model 1, I only include
variables that affect the credibility of the minority faction threat (faction side). In
Model 2, I separately estimate the effect of elements concerning the leader’s atti-
tude toward internal dissent (leader side). Model 3 includes all the variables.
Considering the faction side as well the leader side improves the model fit.15
The results provide strong support for the theoretical argument. Factions con-
sider policy, office, as well as electoral payoffs. First, higher Ideological Dis-
tance increases the likelihood of Fission. The cost of membership increases with
the distance from the core of party members; accordingly, fringe factions have
incentive to leave as they can get larger policy payoffs after a split without suf-
fering large exit costs.16 Second, a higher value of Extra Share of Ministers con-
tributes to preserving party unity. Factions that are “overpaid,” i.e., retain a share
Table 6.1 Logit regression of faction breakaways in Italy
DV: Fission (1) (2) (3)
Faction Side
Ideological Distance 2.168*** – 2.480***
(0.701) (0.769)
Extra Share of Ministers –4.678* – –4.551*
(2.478) (2.664)
Disproportionality –0.398** – –0.378*
(0.169) (0.200)
Closed List 2.237*** – 2.584***
(0.706) (0.739)
Party Age –0.079*** – –0.071**
(0.028) (0.033)
Leader Side
Democratic Centralism – 2.680*** 2.453***
(0.628) (0.650)
Ruling Party – –6.720** –7.732***
(3.201) (2.875)
Parliamentary Support – –2.927 –7.580*
(3.592) (4.584)
Ruling Party X Parliamentary Support – 11.931** 14.580***
(5.610) (5.490)
Constant –1.463* –1.813 1.823
(0.877) (1.847) (2.726)
Log-pseudolikelihood –98.44 –117.37 –93.57
Area under ROC Curve 0.824 0.605 0.847
Correctly predicted (%) 83.42 94. 65 85.77
Sensitivity (%) 68.97 13.79 72.41
N 766 766 766
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. Standard errors clustered by cabinet
in each party congress are shown in parentheses. The percentage of correctly predicted outcomes and
the sensitivity (percentage of correctly predicted positive outcomes) have been measured according
to a 0.05 cut-off point, close to the actual outcome rate (0.04).
of office payoffs greater than their size, are less likely to break away as they
would hardly ever find better conditions outside the party (and vice versa). This
result confirms the idea, discussed in Chapter 5, that party leaders may use stra-
tegic portfolio allocation to cater to potential splinter groups in order to avoid
fissions. Factions jointly weigh policy and office payoffs as substitutes. Accord-
ingly, a greater than proportional share of office payoffs could counterbalance a
lower share of policy payoffs (Ceron 2014; Warwick 1998).
Third, electoral motivation also matters. Disproportionality and Closed List
are both significant and in line with the theory. Disproportional electoral systems
decrease the likelihood of a breakaway by increasing exit costs and lowering the
expected electoral payoffs of splinter groups.17 If the electoral system does not
guarantee re-election, dissenters prefer to remain within the party. Conversely,
closed list PR and centralized candidate selection methods give leaders more
power over the selection of MPs, restricting the access of minority factions to
candidacies and thus reducing the cost of leaving.18 Loyalty restrains minorities
from breaking away. Even after controlling for Extra Share of Ministers, which
accounts for any cooperative patterns that might emerge over time, the coeffi-
cient of Party Age is significant. Therefore, this positive effect of loyalty on
unity is due to the “logic of appropriateness” developed during party militancy,
beyond the “logic of consequentiality.”19
Figure 6.1 Marginal effect of Ruling Party on Fission as Parliamentary Support for the
cabinet enlarges.
Note
Closed List is kept fix to 1. The 90 percent confidence interval is displayed.
On the leader side, I find that Democratic Centralism strongly increases the
risk of a split, consistent with the claim that “voice” becomes more expensive
than a breakup when parties do not tolerate internal dissent. Thus, the leader will
not accommodate minority factions, forcing splinter groups to accept the party
line or break away.
Finally, as far as party system competitiveness is concerned, Figure 6.1 dis-
plays the marginal effect of the interaction between Ruling Party and Parlia-
mentary Support. When the margin of the ruling coalition is narrow, parties in
office will be concerned about preserving unity insofar as any breakaway might
threaten government stability. Thus, the leader will cater to potential splinter
groups to keep the party together. Accordingly, the marginal effect is negative
for ruling parties when Parliamentary Support is below 44 percent of the seats.
By contrast, party fissions are less damaging to parties in governments supported
by wide majorities as leaders are more willing to accept the risk of fission for the
sake of promoting government effectiveness through party cohesion. The mar-
ginal effect becomes positive and significant for parties in office when Parlia-
mentary Support exceeds 62 percent of the seats,20 i.e., when the party system
competitiveness is low.21
The splits of the PD and PDL: an out-of-sample prediction

So far, I have analyzed the determinants of fissions focusing on splits linked with
party congresses that occurred between 1946 and 2011. The dataset did not
include splits that took place later, or splits that were not related to factional
strife in party congresses.
Interestingly, from 2011 on, a wide number of fissions have reshaped the
Italian party system once again. To start with, in 2010–2011, the People of
Freedom (PDL) party split in the wake of internal strife between its main leader,
Berlusconi, and the former Speaker of the Lower Chamber, Gianfranco Fini. The
minority faction repeatedly requested a party congress, in order to measure the
reciprocal strength of internal subgroups and to define a new shared policy line.
However, the PDL leadership refused to announce a new congress and started to
whip rebels belonging to the minority. In response to sanctions imposed by the
core of the party, the faction led by Fini (Fare Futuro/Finiani) broke away and
founded Future and Freedom for Italy (FLI).
In the aftermath of the 2013 general elections, another wave of fissions
involving almost all the main parties took place (Ceron 2017b). None of them
was actually free from factionalism and both small and large parties were
(repeatedly) hit by splits or (collective) parliamentary switching.
After the split of Finiani, Berlusconi’s party was hit by another fission (2012)
before the elections, when former AN politicians including Giorgia Meloni split
to form the right-wing party Brothers of Italy (FdI). In turn, Forza Italia (FI)
broke again in autumn 2014 when the Minister of the Interior (Angelino Alfano)
criticized Berlusconi’s choice to withdraw the support to the Letta cabinet and
created a moderate party called New Center–Right (NCD). Two additional party
fissions hit FI in June and August 2015. Party members loyal to Raffaele Fitto
(Fittiani) and Denis Verdini (Verdiniani) quit.
Interestingly, these splinter groups never won a significant share of votes in the
following elections and only FdI obtained some seats. Nevertheless, these fissions
go hand in hand with the shrinking electoral performance of the PDL/FI. Appar-
ently, the occurrence of splits could have damaged the electoral appeal of the party
signaling a weakness and reducing the valence endowment perceived by voters.
Inside the center-right camp, the Northern League was affected by a party
fission too. Despite having long been considered stable and homogenous, intra-
party divisions produced a breakup. MPs tied to Flavio Tosi, mayor of Verona
and former deputy party leader, switched off from the LN and Tosi himself was
expelled from the party. So far, this represents only a minor split if compared to
the more intense rivalries involving the supporters of the historic party leader,
Umberto Bossi, and the subgroups loyal to Roberto Maroni or Matteo Salvini. In
the LN party loyalty has probably made the difference, preventing larger splits;
in fact, the LN is the oldest Italian party, and leaving it after 30 years of mili-
tancy is an uneasy choice that carries high exit costs.
Conversely, the new and young party created in 2012 by the former Prime
Minister Mario Monti, Civic Choice, soon split in two. One of these two groups
rejoined the PD and the other merged with a splinter group of a small centrist
party to create a new moderate party.
The Five Stars Movement (M5S) repeatedly decided to expel some rebels
from the party. This party is organized according to strict internal rules that only
allow for the expression of dissent in private, while inhibiting any public dis-
agreement. Although activists are entitled to vote online on the party’s policies
and strategy, the leadership has the last word. Those who publicly criticize the
party line can leave the party on their own or be sent off. In view of that, in
January 2015, a large number of MPs broke away to create an autonomous
parliamentary group.
Remarkably, the outcome of multiple struggles inside the M5S is in line with
the theoretical framework. I argue that parties with intransigent rules do not tol-
erate dissent because internal disagreement could weaken them. Indeed, the M5S
leadership has been unwilling to bargain with dissenters (particularly when the
leadership was exerted by the party’s founder Beppe Grillo), producing several
exits and episodes of parliamentary switching.
Concerning the left, a moderate splinter group of SEL broke away to join PD
(2014) and support the Renzi cabinet. Finally, PD has been highly factionalized
for a long time. In 2012, a few years after its formation, there were around 11
organized factions inside it (Ceron 2017a). At that time, the party was still proud
of its heterogeneity, which was taken as a signal of a dynamic and democratic
internal life. PD factions frequently expressed their heavy disagreements in
public, quarrelling over different topics on social networking sites, as well as on
television and newspapers. Such factional bitterness, however, did not produce
splits until the leader of the minority, Matteo Renzi, won the congress and
became the new party leader.
Soon after, divisions inside the PD became sharper, particularly after the
labor market reform and the school reform promoted by the Renzi cabinet. In
response, a few MPs, including some factional leaders like Stefano Fassina and
Pippo Civati, switched off (Ceron & Negri 2016). Later, in 2017, the internal
left-wing faction, headed by the former party leader Bersani, split from PD to
create its own party.
In light of these events, I will double check whether my theoretical argument
can be applied to explain (or to predict) the occurrence of such fissions. In this
regard, I will focus on the two most relevant splits not included in the dataset.
First, I consider the split of Finiani from the PDL (unrelated to a party congress
debate), which contributed to the dissolution of Berlusconi’s cabinet in late 2011
(though the decisive role was played, in a second step, by individual rebels not
belonging to the Finiani faction).
Second, I analyze the breakaway of two left-wing factions from PD in 2015
and 2017, namely Civati’s Possibile and Sinistra PD, headed by Bersani; to do
that I will estimate the distance of each faction from the party mainstream using
textual data related to the motions presented in the 2013 PD congress (when both
minority factions raced against Renzi).
For these two cases, I will perform an out-of-sample prediction (or post-
diction, even though the model and the data used here were already collected or
available before these events). For this purpose, I apply Model 3 presented in
Table 6.1, using the coefficients of the parameters estimated therein to see
whether that model, based on data collected before 2011, could have been able
to predict the occurrence of these two splits.
All in all, I will also discuss whether the logic behind my game–theoretic
model is indeed useful to understand or anticipate current political events,
including the fight between the PD leader Renzi and the new minority faction
headed by Andrea Orlando concerning the selection of candidates in the 2018
general election.
To start with, according to my model, the likelihood of a split of the Finiani
from the PDL amounted to 73 percent. This estimate, though, does not con-
sider any policy distance. In absence of rival policy documents related to party
congresses it is not possible to account for the Ideological Distance between
this group and the party mainstream. However, assuming that the distance
between Finiani and the core of the PDL is equal to a faction’s average Ideo-
logical Distance from the party mainstream (0.24), the probability of a split
would grow by 8 points. Actually, the policy disagreement was probably much
larger than this, given that the minority faction complained about almost every
policy choice made by the party leadership. In view of that, assigning a wider
value of Ideological Distance (for instance, equal to 0.50) would raise the
probability of breakaway up to 88 percent (a 15-points increase compared to
the first estimate).
Nevertheless, the PDL split is mainly triggered by two other elements: the
large margin of the ruling coalition, which made the contribution of this
minority faction unnecessary, and the party’s attitudes to deny internal dissent.
The probability of a fission would have been 20 points lower in the case of a
tiny majority.
Anyway, if one assumes that this party is organized through a kind of demo-
cratic centralism, any dissenting faction that will show up is likely to break away
as the party leader would not be willing to cater to it by providing a share of
payoffs large enough to avoid a split.
What would happen inside a party that allows internal dissent? To answer this
question, I investigate the recent splits of the PD. After the poor results obtained
in the 2013 election, the internal equilibrium was less defined.
At that time, the closed list system could have favored breakaways, but the
strikingly high level of disproportionality that emerged in the election frightened
potential splinter groups.22 Furthermore, the adoption of primary elections is a
unifying factor (Hortala-Vallve & Mueller 2015) for PD.
The coalition government between the PD and PDL, however, was loosening
the ties between PD leaders and the rank-and-file, fostering the protest of young
“Occupy PD” activists (Ceron 2016). Despite its untenable policy heterogeneity,
consensual and conciliatory intra-party dynamics that take the stakes of all party
factions into consideration were the only way to keep the party together and
avoid harmful splits.
In this regard, the outcome of the PD congress held in autumn 2013 was
crucial in influencing the likelihood of a breakaway. When Matteo Renzi, a party
leader inclined to make clear-cut choices, won the congress, the fragile internal
equilibrium was put to the test.23
Renzi, leader of the liberal–democratic internal minority, was appointed through
direct election by a wide margin (67.6 percent), and his right-wing faction Renziani
(jointly with Enrico Morando’s Liberal/LibertàEguale and Dario Francheschini’s
AreaDem) obtained a safe majority in the PD National Assembly.
This change in the dominant faction was particularly damaging for the former
leadership, headed by Bersani; their faction Sinistra PD only won 18.2 percent
of votes. Finally, the faction Possibile, created by Civati (a former member of
Renziani) won the remaining 14.2 percent.
At the beginning, Renzi was inclined to change the line and the strategy of
the PD, getting rid of factional constraints. However, in terms of portfolio alloca-
tion, a norm of perfect proportionality was adopted during his cabinet formation
(2014): Sinistra PD received two out of 11 PD ministers (18.2 percent) and Pos-
sibile received one minister too (9.1 percent).
As far as policies are concerned, however, Renzi was not willing to bargain
with the minority. He always refused to modify the electoral law introducing
preference voting, as requested by the minority; all the major economic reforms
were proposed and approved without negotiating with the internal left-wing or
trade unions (Ceron & Negri 2017, 2018).
It is not by chance that, according to a Wordfish analysis of factional motions
related to the 2013 PD congress (which extends what has been done in Chapter
2), the distance between Sinistra PD and the party mainstream was huge (0.55),
being in the upper 10 percent of the distribution of Ideological Distance.
In the end, the two minority factions split in 2015 and 2017, though a small
part of Sinistra PD decided to remain inside. The model predicts the split of the
main minority group, Sinistra PD. Indeed, the likelihood of its breakaway is
equal to 51 percent. This outcome is certainly affected by the wide policy
disagreement existing between this social–democratic faction and the liberal–
democratic party line set by the new dominant faction.
With respect to Possibile the model predicts a 30 percent chance of break-
away, which raises to 38 percent if one considers that the only minister belong-
ing to this faction retired from office a few months before the split. Although
lower than 50 percent, this value is still ten times higher than the probability of a
split for a random faction in the sample (which is used as a cut-off point to
estimate the goodness of fit in Table 6.1). As such, the prediction emphasizes
that a breakup was very likely to occur.
Interestingly, both factions split soon after two reforms of the electoral rules.
In particular, the final electoral law to be used in the 2018 elections was still
based on a full closed list system (without any preference voting, which instead
could have allowed dissenters to remain inside the party); furthermore, the lack
of any majority prize indicated that the expected level of disproportionality in
the following elections could have been lower if compared to the 2013 election.
Indeed, the Gallagher index in 2018 is 4.5 (against the peak of 17.3 reached in
2013). In this scenario, with closed list and low disproportionality, there are
fewer incentives for factions to remain inside, and the expectation of lower bar-
riers to the entry in the political market triggered an increase in party fissions, as
well as the split of Sinistra PD.
Remarkably, if the expected level of disproportionality had not changed, the
likelihood of a split of Sinistra PD would have been close to zero; even more so
if preference voting was introduced, following the request of the PD minority.
In this regard, the split of the PD was the outcome of Renzi’s approach to
make clear-cut choices in spite of the will of the internal minorities. His attitude
to pursue cohesion even at the expense of unity emerges in light of the huge
ideological polarization that affected the PD, as confirmed by this text analysis.
Faced with internal heterogeneity, however, Renzi reacted by avoiding any
policy negotiation with the left-wing rebels; furthermore, he imposed an elect-
oral rule that, instead of leaving room for factionalism and internal dissent,
incentivized minority factions to split and create new parties.
To conclude, what was the destiny of the small part of the minority faction
that decided to remain inside the PD? Andrea Orlando, Minister of Justice in the
Renzi and Gentiloni cabinets, refused to break away and contested the 2017 con-
gress becoming (temporarily) the new leader of the internal left.
Looking at the game discussed in Chapter 1, and in light of the dynamics
described above, it can be argued that the PD leader Renzi was focused on cohe-
sion rather than unity. In this scenario, any minority faction can either leave the
party or accept an unfair deal. If the exit was the most rational outcome for many
dissenters, any rebel still inside the party can only comply with the leadership or
should be prepared to face the whip.
Notwithstanding this, a few weeks before the 2018 general election Orlando
complained about the unfair distribution of candidacies. Orlando claimed that
the minority faction was heavily underrepresented in the party list (8 percent of
posts despite a 20 percent of congress votes), and he launched an ultimatum to
the leadership, demanding a fairer allocation of posts.
Unsurprisingly, the leadership refused to compromise. Once again, this
outcome is in line with the expectations generated by the theoretical model.
Despite looming elections, in fact, Renzi was still stressing cohesion more than
unity. Based on my theory, I could have predicted both the unfair allocation of
candidacies and the refusal of any compromise by an autonomous party leader,
who was focused on cohesion (ω < ν) and centralized the candidate selection
process in his hands.
Accordingly, Orlando’s loyalty was not going to be rewarded, given that a
few weeks before the elections any other exit option was already vanished and
his threat to break away was not credible nor strong (ε > π and π – ε < 0 – ν).
Party splits in France and Germany

The evidence provided so far highlights how my game–theoretic model perfectly
fits the data on factionalism and party splits in Italy. Moving one step forward, I
replicate the analysis on France.
The dependent variable, Fission, is built as described above. In France I
account for nine breakaways related to the 151 factions considered in this
study, which are the unit of analysis (see Appendix 2 for details). Overall, 6
percent of factions break away. Among the others, I examine the breakaway
of Morin and the UDF rebels, unwilling to form the MoDem, who created the
party Nouveau Centre, as well as the split of Dupont-Aignan’s faction (the
right-wing Debout la République) from the UMP. The split of CÉRÉS from
the PS, which presented its last comprehensive motion in 1990, has been taken
into account too. This left-wing faction decided to exit due to its disagreement
about the war on Iraq in 1991. The minister Chevènement, leader of CÉRÉS,
resigned from office, and the faction rebelled during the parliamentary vote on
the war, despite the threat of sanctions declared by the party leadership (which
indeed denied the use of the PS label to the rebels: Verrier 1999); accordingly,
in 1992–1993 the faction split and launched a new party, the Mouvement des
Citoyennes.24 Concerning the PS, I also consider the breakaway of Bockel
(who created La Gauche Moderne) after the 2005 congress and the split of the
left-wing faction headed by Melénchon and Dolez, who created the Parti de
Gauche in 2008.
I retain most of the independent variables used in the analysis on Italy: Ideo-
logical Distance (H6.1), Extra Share of Ministers (H6.2), Disproportionality
(H6.3), as well as Party Age (H6.5) and Ruling Party. I also include Direct Elec-
tion (equal to 1 for leaders that are directly elected by congress delegates or
party members), and control for the number of years remaining before the next
Presidential election using the variable Years Before Election.
The results of logistic regression are displayed in Table 6.2. Observations
have been clustered by parties, providing standard errors accordingly. In Model
1, I only include the two main variables, Ideological Distance and Extra Share of
Ministers; in Model 2 all the other variables have been added.
In line with H6.1 and H6.2, the results confirm that policy and office payoffs
matter. Splits are more likely when the faction is ideologically distant from the
party line; conversely, factions that are “overpaid,” i.e., hold a share of ministers
larger than their size, are less willing to split. The coefficients of Party Age and
Disproportionality retain the expected negative sign, even though they do not
reach statistically significant. Finally, splits occur more often when leaders are
directly elected. This indicates that, in the presence of bounded leaders, French
parties manage to preserve their unity more easily.
Notice that in the present analysis I did not consider the splits related to the
collapse of the PS that occurred in 2017–2018 because only some politicians
belonging to the factions that contested the 2015 congress split.
Table 6.2 Logit regression of faction breakaways in France and Germany
DV: Fission France (I) France (II) Germany (III)
Ideological Distance 0.784** 1.046** –

(0.388) (0.504)
Extra Share of Ministers –0.053* –0.061** –
(0.032) (0.028)
Disproportionality – –0.076 –
(0.103)
Party Age – –0.009 –
(0.038)
Ruling Party – –0.376 –
(1.406)
Direct Election – 1.899*** –
(0.548)
Years Before Election – –0.013 –
(0.402)
Internal Polarization – – 1.234**
(0.630)
Constant –3.000*** –2.940*** –5.273***
(0.323) (0.633) (0.692)
Log-pseudolikelihood –33.226 –24.656 –4.768
Area under ROC Curve 0.780 0.759 0.938
Correctly predicted (%) 69.33 61.72 73.13
Sensitivity (%) 55.56 85.71 100
N 150 128 67
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. Standard errors clustered on parties
in parentheses. The percentage of correctly predicted outcomes and the sensitivity (percentage of
correctly predicted positive outcomes) have been measured according to a 0.05 (France) and 0.015
(Germany) cut-off points, that are closer to the actual rate.
Since 2014, the PS has been affected by a harsh left-wing internal opposition
(the so-called Fronde), which exacerbated intra-party divisions. When Benoît
Hamon, one of the leaders of the left-wing faction, won the PS primary election
in January 2017, many moderate PS members (including Valls, who lost the
primary election) left the party to support Macron’s La République En Marche.
The situation was quite fluid though. Some prominent PS politicians, such as
Gérard Collomb and Jean-Yves Le Drian, joined the new government appointed
by the Head of State François Macron. However, the PS leadership declared that
those in office are excluded from the party. In turn, some elements of the left-
wing broke away from the PS in July 2017; after the electoral defeat, Hamon
and other former PS politicians including Aurélie Filippetti (one of the leaders of
the parliamentary Fronde) split to create a new political movement. Overall, the
results are confirmed even if I consider these episodes as either a breakaway of
right-wing factions or a split of left-wing factions. Later on, in October 2018,
another left-wing group (L’Union et l’Espoir), headed by Emmanuel Maurel and
Marie-Noëlle Lienemann, split to approach the Melénchon’s party La France
Insoumise.
All these party breakups fit the theoretical framework.25 Indeed, the internal
polarization inside the PS was so huge that the party managed to keep its unity
only until office payoffs allowed it to be preserved. When the left-wing can-
didate won the primary election, the right-wing felt that the new party line could
shift too much toward the left, leaving these moderate members unsatisfied.
Furthermore, in the new centrist cabinet appointed by Macron, moderate PS
members had the opportunity to extract a share of policy and office payoffs
larger than that provided by the PS alone. Attracted by these payoffs, several PS
politicians joined the cabinet and accepted the risk of being sent off from the
party. In search of a new identity, the leadership of the PS refused to reach a
compromise and left the door open for the exit.
In turn, part of the internal left-wing split when, after the bad electoral per-
formance, the right-wing was going to retake control of the PS, as Aubry and
Ayrault’s moderate faction was going to win the 2018 congress.
In France and Italy, party splits are more likely to be observed. Conversely, in
Germany there is only one episode of party fission with respect to the data
included in this study. This recent split occurred in 2005 when Lafontaine, the
former party leader close to the left-wing party factions of the SPD, split to join
a new political movement, Labor and Social Justice – The Electoral Alternative
(WASG). This party was created in 2005 by left-wing and union activists disen-
chanted with the ruling Red–Green Coalition, who criticized the neoliberal pol-
icies of this cabinet, and in particular Schröder’s Agenda 2010 reform of welfare
and the labor market. The WASG immediately formed an electoral alliance with
the far-left PDS, which took the name of Die Linke (The Left).
The low rate of party fissions in Germany suggests that, in such a highly com-
petitive party-centered political system, party unity is deemed crucial by voters
and political elites. Indeed, in Chapter 5 I argued that portfolio allocation tends
to follow consensual dynamics precisely to preserve unity.
In view of that, it comes as no surprise that the only split coincides with a
peak in the internal polarization of the SPD. This suggests, once again, that
policy heterogeneity matters. From Model 3 (Table 6.2) it can be noticed that a
higher level of intra-party polarization (i.e., absolute distance between the two
most extreme factions in each party congress) is likely to generate more splits
even in the German case. From this perspective, the recent quarrel around the
decision to start a new Grosse Coalition with the CDU is a signal that SPD needs
a compromise to restore its ideological unity, in order to avoid new splits that
could be particularly damaging in light of the recent poor electoral performance
of this party.
Analogously, the outcome of the 2018 CDU congress suggests that party
unity is becoming more and more precarious even inside this party. The tie
race, won with a narrow margin (35 out of 999 votes) by the moderate can-
didate Kramp-Karrenbauer (in continuity with the centrist party platform set
by Merkel) disappointed many conservative party members who quit the party
to protest against the outcome.26 Indeed, in the first ballot Kramp-Karrenbauer
only won a plurality of votes, while the two candidates belonging to conser-
vative factions jointly won 55 percent of delegates votes. This suggests that
the party is heavily polarized and the new leadership should deal with the risk
of a split. It is not by chance that Alexander Mitsch, leader of the small
conservative faction WerteUnion, claimed that the possibility of a breakaway
should not be ruled out.
In turn, soon after the congress, some disgruntled conservatives have already
threatened to form their own party.
In her victory speech, Kramp-Karrenbauer immediately called for party unity,
arguing that she would work with her main opponent, Friedrich Merz, and the
two camps should reunite to strengthen the party in the electoral arena.27
In light of some rumors, according to which Merz and other prominent
conservative politicians are considering the option of leaving the CDU, many
other appeals have been made (both by centrist politicians, such as the deputy
chair of the PPG, Andreas Jung and by the MIT leader Carsten Linnemann) to
ask Merz to remain on board.
In turn, Kramp-Karrenbauer’s choice to appoint the young Paul Ziemiak as
the new secretary general has been seen as a move to keep the party together,
given that Ziemiak is deemed an ally of Spahn, the third-ranked conservative
candidate. However, this choice has not been much appreciated and Ziemiak
only received 62 percent of votes, raising the dissent of conservative fellows.
In view of that, to heal divisions at the heart of her party, Kramp-Karrenbauer
needs to make a huge effort, by adjusting the policy line in order to take conser-
vative views into consideration or by offering more office rewards to the leaders
of the internal right-wing factions. For sure, these are two main options available
to the party leader, according to the theoretical model, and there are indeed the
two requests immediately raised by conservative factional leaders in order to
prevent the occurrence of a party fission.28
Party fission: findings
To sum up, all these analyses support the theoretical argument and indicate that
factions consider office and policy, as well as electoral payoffs. In all three coun-
tries, policy motives seem to drive the breaking of parties (Kemahlıoğlu & Sayarı
2017). Subgroups sharing common preferences are more likely to join together and
more likely to split when intra-party heterogeneity increases. Even factions that are
usually considered merely office-seeking seem to pay attention to policy payoffs.
Conversely, the strategic allocation of portfolios can counterbalance a lower
amount of policy payoffs, thus contributing to preserve party unity.
The fission of the PS in 2017 is in line with this theoretical argument, and con-
firms that factions consider policy and office payoffs altogether, weighting their
actual level of benefits with the expected gain they can obtain in the case of break-
away. In 2017, the lower amount of policy payoffs available to PS factions (due to
heightened internal polarization), along with the opportunity to get policy payoffs
in the case of exit, pushed factions to leave the party.
In the Italian case, one can also notice that other elements, such as party loyalty,
disproportional electoral systems and open access to candidacies, increase the exit
costs for dissenting factions and decrease the probability of a split.
Splinter groups, however, do not make choices in isolation. They must weigh
the leadership’s attitude to party unity when threatening to defect. Under some con-
ditions leaders may be willing to compromise to preserve party unity, under others
they may use the whip against dissenters to enhance party cohesion. In France,
overall, splits are more likely when leaders are autonomous, whereas consensual
dynamics enacted by bounded leaders seem able to preserve unity.
Consistent with my theory, the level of intra-party democracy and the inter-
electoral level of party system competitiveness affect the leader’s attitude. In Italy,
the results indicate that the cost of “voice” is higher in parties ruled through demo-
cratic centralism where leaders do not tolerate dissent, making minorities inclined
to break away. For ruling parties, the cost of a breakup increases with a smaller
margin over the opposition, raising the leader’s interest in a compromise to preserve
party unity. Conversely, leaders in ruling parties focus more on cohesion to enhance
government effectiveness as the parliamentary support widens.29 The lower cost of
a breakup declines below the cost of voice, and splits become more likely.
Indeed, such search for greater cohesion and the firm rejection of any internal
dissent can help explaining the fission of the PDL in 2010–2011. Analogously,
the party leader’s choice to focus on cohesion rather than unity (to reduce the
sharp internal polarization), jointly with the decision to introduce a new electoral
law that incentivizes factions to break away, can explain the split of the PD in
2015 and 2017.
My findings demonstrate that parties can be considered as minimum winning
coalitions of factions where all unnecessary subgroups are disregarded by the
party leader and must face the choice between compliance and exit.
To conclude, both intra-party dynamics (linked with fights or compromises)
and different attributes of the party system (e.g., electoral rules or party system
competitiveness) affect the propensity to split and provide incentives to pursue
party unity or strengthen party cohesion.
When parties are internally polarized, party system fragmentation can be con-
tained through consensual intra-party dynamics, disproportional electoral
systems, and preference voting or decentralized candidate selection procedures
that favor the institutionalization of party factions (e.g., primary elections).
These aspects are highly relevant for political elites that aim to simplify political
supply through party mergers. When a party merger is merely a “cold fusion”
process based on instrumental and strategic concerns and without any policy basis
(Martocchia Diodati 2017), the internal wings will be more likely to break away
and will undermine elites’ efforts to decrease the fragmentation.
To sum up, any reform that aims to foster party system stability should
include adjustments in the rules of the game to help keep factionalism under
control and preserve party unity.
Notes
1 Some authors tried to test such an idea by looking at party switching (Desposato
2006; Heller & Mershon 2008; McElroy 2009). Overall their findings seem to confirm
this argument even though some of those results are more ambiguous.
2 Using the median faction position as a proxy for party ideal point does not alter the
results.
3 Minority factions could be overpaid when their threat is credible and the party leader
is concerned about party unity (see below). Once overpaid, however, the greater
reward itself influences the likelihood of a breakaway.
4 The splinter group will face high start-up costs for creating a new party and establish
itself as a relevant actor.
5 Remember that the concept of unity (i.e., factions may possibly support different
views and strategies, but without leaving the party) is different from that of cohesion
(i.e., factions think and act in a cohesive manner, in agreement with the party line).
6 This is not only related to Marxist parties. For example, relevant Italian parties such
as the People of Freedom/Forza Italia and the Five Star Movement have strict internal
rules.
7 The party leader may also expel dissenters from the party. My theory suggests that
breakaways and expulsions stem from the same underlying process.
8 The cost of voice may also be higher when a faction position is far from the bulk of
party members. The wide array of internal preferences blurs the party label (Snyder &
Ting 2002) and the party will suffer credible commitment problems with voters and
allies. In this context the leader will be more focused on internal cohesion than unity.
A split may clarify a party label and enhance a party’s image. This logic yields the
same outcome suggested in H6.1.
9 This hypothesis sends back to the notion of minimum winning coalitions (Leiserson
1968; Riker 1962).
10 The rate of fissions over the total number of cases, however, shrinks to 4 percent
given that observations have been multiplied by cabinet.
11 Assigning a value of 0 to factions whose positions are not statistically different from
the median faction does not affect the results.
12 Italy has adopted two different electoral rules since 1994, a mixed system
(1994–2001) with 25 percent of seats assigned through closed list PR and the remain-
ing 75 percent via “first-past-the-post,” and a closed list PR with thresholds and a
majority prize (since 2006). Between 1994 and 2001, the selection of candidates in
single-member districts was strongly centralized, making this system similar to a
closed list with magnitude one. Thus, I can compare open list PR systems in use until
1993 to subsequent systems (Carey & Shugart 1995).
13 This applies to the PSI, in 1949, under the orthodox leadership of Rodolfo Morandi
and to the PDCI in 2008.
14 Clustered standard errors are heteroskedastic and autocorrelation consistent (Rogers
1993). Controlling for temporal dependence through temporal dummies or random
effects does not alter the results. A rare events logistic regression model generates similar
results. Approximately 25 percent of the observations are related to the Socialist family
(PSDI and PSI) and one-third of the breakaways involve these two parties. Including a
dummy variable that accounts for this political family does not alter the results. I also
controlled for the impact of the electoral cycle and party left–right position, but these two
variables were not significant and do not affect my findings.
15 When dealing with rare events, the area under the receiver–operating characteristic
(ROC) curve allows the model performance to be evaluated. This area represents the
probability that a randomly selected positive outcome (Fission) is correctly rated with a
higher predicted probability than a randomly selected negative observation. Such a stat-
istic is higher in Model 3 (0.847), than in Model 1 (0.824) and 2 (0.605); this indicates
that both the credibility of the minority faction threat and the leader’s attitude toward
internal dissent help in predicting party splits. Other measures of the goodness of fit
confirm this pattern, which is consistent with my theoretical model.
16 The coefficient for an interaction term between Ideological Distance and Ruling Party
is not significant. When testing the model on the subsample of parties in office, the
effect of Ideological Distance remains the same. Given that in the Italian context the
main reward for cabinet participation is linked to office payoffs (Mershon 1996), this
result supports the idea that the behavior of potential “office-seeking” factions
(factions of interests: Sartori 1976) is driven also by policy motivations.
17 This finding holds when using other measures of disproportionality and holds also
when the First and the Second Italian Republic are analyzed separately.
18 I also tested H6.4 through a variable Centralization, which expresses the degree of
centralization in candidate selection process according to the rules described in the
party statute. This variable ranges from 0 to 10, where 0 indicates a decentralized
selection process while the value of 10 is assigned to parties that adopt centralized
mechanisms. This variable displays a positive and significant coefficient proving that
centralized candidate selection increases the likelihood of a breakaway, consistent
with H6.4.
19 This claim is supported when including a control variable for changes in party labels
and symbols during the party congress (unconnected to policy adjustments that are
captured by the variable Ideological Distance). When the party logo changes, the like-
lihood of a breakaway increases while the impact of Party Age still holds.
20 The marginal effect of Parliamentary Support is positive and significant for any
Ruling Party, but increasing Parliamentary Support implies lower party system com-
petitiveness, which in turn reduces the leverage of potential splinter groups. This
decreases the probability of splits in parties out of office as there are no incentives for
defection when the ruling coalition has a safe margin. The marginal effect of Parlia-
mentary Support is negative and significant for parties not in office.
21 In separate models I tested the impact of party system competitiveness in interaction
with leader’s autonomy. I found that leader’s Direct Election (which fosters
autonomy) increases the likelihood of a breakaway when party system competit-
iveness declines. To do that, I relied on the variable Margin of Votes, i.e., a measure
of competitiveness that records the difference between the share of votes won by the
two main rival parties (that ranked first and second in the election). When the distance
between the two main competitors is low (7 percent or below), Direct Election has no
significant effects because in this context any vote matters, boosting the need to
preserve unity. Accordingly, no difference can be observed between parties ruled by
autonomous or constrained leaders. Conversely, when Margin of Votes grows (reach-
ing 7.5–17.5 percent) the competitiveness decreases and the cost of party breakup
shrinks; consequently, an autonomous leader will not heed to the requests of internal
minorities, leaving them unsatisfied and inclined to split.
22 The Gallagher index was equal to 17.3, the highest value ever reached in Italy, whereas
the average of this index was 6.84 in the Second Republic and to 2.97 in the First.
23 http://blogs.lse.ac.uk/europpblog/2013/07/09/factions-and-fissions-in-the-aftermath-of-
the-2013-elections-why-italian-politics-never-gets-boring-2/.
24 For analytic purposes, I consider the split of CÉRÉS “as if ” it took place after the
1990 congress. Notice that considering it as a consequence of the 1991 congress
(when the party did not allow to present real motions but only amendments to the
mainstream’s theses) does not alter any of the results.
25 If applying Model 2 discussed above to predict the split of Hamon or that of L’Union
et l’Espoir (using textual documents related to the 2017 primary election, for Hamon’s
faction or to the 2018 PS congress, for L’Union et l’Espoir) I would get a probability
of breakaway that is, respectively, more than four times or more than two times larger
than the average. For instance, the model predicts a probability of split equal to 17
percent for Hamon’s faction. Although this value is lower than 50 percent, it is much
larger than the average expected probability of split for a random faction (which is
equal to 4 percent).
26 www.telegraph.co.uk/news/2018/12/08/cdu-party-members-quit-protest-merkel-successor-
split-widens/.
27 www.telegraph.co.uk/news/2018/12/07/angela-merkel-appeals-cdu-stay-political-centre-
party-votes/.
28 www.augsburger-allgemeine.de/politik/Konservativer-Unions-Fluegel-warnt-vor-tiefer-
Spaltung-der-CDU-id52901291.html.
29 Several studies attest that they tend to be less divided (e.g., Carey 2007).
References
Aldrich, J.H. (1995). Why Parties? The Origin and Transformation of Political Parties in
America. University of Chicago Press, Chicago, IL.
Andeweg, R.B., & Thomassen, J. (2010). Pathways to Party Unity: Sanctions, Loyalty,
Homogeneity and Division of Labour in the Dutch Parliament. Party Politics, 17(5),
655–672.
Bernauer, J., & Bräuninger, T. (2009). Intra-party Preference Heterogeneity and Faction
Ceron, A. (2015). The Politics of Fission: Analysis of Faction Breakaways among Italian
Ceron, A. (2016). Competing Principals 2.0? The Impact of Facebook in the 2013 Selec-
tion of the Italian Head of State. Italian Political Science Review, 46(3), 313–333.
Ceron, A., & Negri, F. (2016). The “Social Side” of Public Policies: Monitoring Online
Public Opinion and its Mobilization during the Policy Cycle. Policy & Internet, 8(2),
131–147.
491–508.
Cox, G.W., & McCubbins, M.D. (1993). Legislative Leviathan: Party Government in the
House. University of California Press, Berkeley CA.
Cox, G.W., & Rosenbluth, F. (1995). Anatomy of a Split: The Liberal Democrats of
Japan. Electoral Studies, 14(4), 355–376.
Science, 29(1), 33–56.
Desposato, S.W. (2006). Parties for Rent? Ambition, Ideology and Party Switching in
Brazil’s Chamber of Deputies. American Journal of Political Science, 50(1), 62–80.
Science, 60(4), 860–881.
Dilling, M. (2018). In Defence of Moderate Factionalism. Paper Presented at ECPR
Evans, J.A.J. (ed.) (2003). The French Party System. Manchester University Press,
Manchester and New York.
Gehlbach, S. (2006). A Formal Model of Exit and Voice. Rationality and Society, 18(4),
395–418.
Rationalité et Sciences Sociales Colloque, Collège de France, Paris, 3–4 June.
Giannetti, D., & Laver, M. (2001). Party System Dynamics and the Making and Breaking
of Italian Governments. Electoral Studies, 20(4), 529–553.
Italy. In Giannetti, D., & Benoit, K. (eds) Intra- Party Politics and Coalition Govern-
Gillespie, R., Waller, M., & Lopez-Nieto, L. (eds) (1995). Factional Politics and Demo-
cratisation, Democratisation, 2(1).
Heller, W.B., & Mershon, C. (2008). Dealing in Discipline: Party Switching and Legis-
lative Voting in the Italian Chamber of Deputies, 1988–2000. American Journal of
Hortala-Vallve, R., & Mueller, H. (2015). Primaries: The Unifying Force. Public Choice,
163(3–4), 289–305.
Ibenskas, R., & Sikk A. (2017). Patterns of Party Change in Central and Eastern Europe,
1990–2015. Party Politics, 23(1), 43–54.
Kam, C. (2009). Party Discipline and Parliamentary Government. Cambridge University
Press, Cambridge.
Kato, J. (1998). When the Party Breaks Up: Exit and Voice among Japanese Legislators.
Kemahlıoğlu, Ö., & Sayarı, S. (2017). Defecting Alone or Splitting Together? Individual
and Collective Party Switching by Legislators. Public Choice, 171(1–2), 187–206.
Kertzer, D.I. (1996). Politics and Symbols: The Italian Communist Party and the Fall of
Communism. Yale University Press, New Haven CT.
Laver, M., & Benoit, K. (2003). The Evolution of Party System between Elections. Amer-
ican Journal of Political Science, 47(2), 215–233.
Laver, M., & Kato, J. (2001). Dynamic Approaches to Government Formation and the
Generic Instability of Decisive Structure in Japan. Electoral Studies, 20(4), 509–527.
250–277.
Mair, P. (1990). The Electoral Payoffs of Fission and Fusion. British Journal of Political
Science, 20(1), 131–142.
Martocchia Diodati, N. (2017). Passing Time, Changing Minds? The Determinants of
Faction Membership After Party Fusion. Government and Opposition, 1–22.
McElroy, G. (2009). Intra-Party Politics at the trans-national level. In Giannetti, D., &
Benoit, K. (eds) Intra- Party Politics and Coalition Government, 205–225, Routledge,
New York.
Mershon, C. (1996). The Costs of Coalition: Coalition Theories and Italian Governments.
Mutlu-Eren, H. (2015). Keeping the Party Together. Public Choice, 164(1–2), 117–133.
469–490.
Rogers, W.H. (1993). Regression Standard Errors in Clustered Samples. Stata Technical
Bulletin, (3)13, 19–23.
Riker, W. (1962). The Theory of Political Coalitions. Yale University Press, New
Haven CT.
Verrier, B. (1999). Chronique d’une rupture. De Socialisme et République au Mouvement
des citoyens. Politix, 12(1), 87–113.
Wada, J., & Schofield, N. (1996). Bargaining in the Liberal Democratic Party of Japan. In
Schofield, N. (ed.), Collective Decision- Making: Social Choice and Political Economy,
239–250, Kluwer Academic Publishers, Boston MA.
Warwick, P.V. (1998). Policy Distance and Parliamentary Government. Legislative
Studies Quarterly, 23(3), 3193–45.
7 Intra-party conﬂict and the
survival of party leaders
Introduction
In factionalized parties, factions and subgroups compete against each other to
take control of the party and maximize their share of payoffs. At the same time,
factions must cooperate to produce party unity, which is a public good for the
party’s community as it allows the total amount of payoffs available to the whole
party to be increased. In this scenario, the party leader is the agent responsible
for preserving unity (Kiewiet & McCubbins 1991) and for pushing all factions
to cooperate in order to produce the public good.
By assigning incentives, the leader could motivate party members and fac-
tions and get them to collaborate, campaigning harder during the election
(Carroll & Cox 2007) and behaving cohesively in the aftermath, thereby provid-
ing more benefits to the whole party.
The previous chapters widely discussed the role of the party leader as the
actor in charge of allocating payoffs to intra-party factions (Ceron 2014). In
view of that, the leader’s role is crucial, as he retains a potentially strong
power of allocation, but his task is a delicate one and he is in a fragile posi-
tion. On the one hand, the leader needs to reward his supporters to avoid being
dismissed. On the other hand, he has to choose whether to preserve party unity
or to enhance party cohesion, trying to minimize internal conflicts. To sum up,
the allocation of payoffs affects party unity (Ceron 2015a), but also the dura-
tion of the leadership.
So far, I have also examined how, depending on intra-party rules, different insti-
tutional intra-party settings exist in which the party leader can be more autonomous
or less autonomous from factional constrains (see Figure 1.2, Chapter 1).
In my theoretical framework I hypothesize that leaders can be more auto-
nomous, when they are directly elected by a wide selectorate, composed of party
members or delegates (who, in turn, are selected by members to represent them
during the party congress, with a specific mandate). Accordingly, autonomous
leaders can exploit their dominant position to reward their supporters (proposing
an unfair deal to the minority); otherwise, they can cater to any other subgroup
that contributes to produce the party’s public good and therefore is useful to
prolong the survival of the leader.
160 The survival of party leaders
What about non-autonomous leaders? When leaders are indirectly elected by a
small party committee, in which any single factional leader can alter the equilibri-
ums and dismiss the party leader, their autonomy shrinks. In this context, leaders
are nothing more than the outcome of an inter-factional agreement. Usually, as an
agent of party factions, they have to implement consensual arrangements that have
already been reached elsewhere. There is limited scope for deviations from such
intra-party equilibrium. Any discretional leader’s behavior, in fact, can produce
dissatisfaction and can be a cause of leader removal.
In view of that, what happens to non-autonomous leaders that deviate from
the consensual compromise or that are facing harsh internal conflict? The present
chapter will answer this question showing how the different degree of leader’s
autonomy can make the difference in terms of survival in office.
A “principal–agent” framework
The democratic polity, nowadays, is more and more often characterized by a
process that takes the name of “personalization” (e.g., Blondel & Thibault 2009)
or “presidentialization” (e.g., Poguntke & Webb 2005) of politics. Traditionally,
political parties have been organized according to a chain of responsiveness that
passes from members and activists to party leaders. Leaders, however, tend to
achieve increasing power by becoming the most relevant actor both in the elect-
oral and in the parliamentary arena.
In the electoral arena, the party leader is the pivotal actor in charge of setting
the party platform (Ceron 2012) and, thanks to his electoral appeal and valence
advantage, he is one of the main determinants of the voters’ electoral choice
(Bellucci et al. 2015; Bittner 2011). His position is so crucial that the leader of
the party that wins the election is usually appointed as the “formateur,” and,
should there be the need to form a coalition, he is involved with other party
leaders in the bargaining over government formation (see Chapter 5). Finally, in
the parliamentary arena, the party leader is also in charge of setting the line on
key parliamentary votes and he is deemed responsible for passing pieces of
legislation (Ceron 2015b). While the importance of party leaders became strik-
ingly evident in recent years, this process dates back to the 1970s and is rooted
in both institutional and organizational adjustments within the structure of the
party (Fabbrini 1994).
For many years, scholars paid attention to the duration of governments ana-
lyzing the survival rate of ministers and prime ministers (Diermeier & Stevenson
1999; Warwick 1994). Differently, studies on party leaderships focused more on
the selection of party leaders (e.g., Cross & Blais 2012; Katz 2001; Kenig 2009),
even though a growing branch of literature has started to investigate the determi-
nants of party leaders’ survival in office (Andrews & Jackman 2008; Bynander
& ’t Hart 2007, 2008; Ennser-Jedenastik & Müller 2015; Ennser-Jedenastik &
Schumacher 2015; Horiuchi et al. 2015). Leader survival, in fact, can affect the
party effectiveness in policy-making and is clearly related to the process of
responsiveness and accountability of democratic political systems.
The survival of party leaders 161
Despite this, there is one major difference between the general factors affect-
ing the survival of political leaders and the survival of the party leader. In inter-
party politics, not all politicians play for the same team. For this reason, it is
always possible to find potential rivals wanting to replace those in power. Fur-
thermore, in democratic systems, such opportunity always exists at the end of
the legislative term, when new elections are called.
Conversely, in intra-party politics, although the party embodies a coalition of
somehow heterogeneous political actors, these actors belong to the same team.
In some cases, even if the team is losing the match, the players recognize that
there is no point in changing the leader, nor is there a way to do that. This
happens, for instance, when potential rivals know that the leader is much more
popular among party members, or when the leader represents a perfect synthesis
between the different interests existing inside the party. Accordingly, even in
presence of a non-successful leader, one shall observe no change if there are no
challengers willing to defy him.
For these reasons, while earlier studies focused mainly on the performance of
the party leaders, i.e., their ability to win votes and get into government
(Andrews & Jackman 2008; Bynander & ’t Hart 2007), scholars have recently
suggested looking at the shape of the internal party environment (Horiuchi et al.
2015; Kenig 2009) and at the factional structure of the party (Bynander & ’t Hart
2008; Ennser-Jedenastik & Müller 2015).
Accordingly, this chapter analyzes the determinants of the duration of party
leaders, also taking into account the internal features of intra-party politics, such
as the degree of heterogeneity, the allocation of payoffs and the shape of intra-
party rules, which may favor the emergence of internal rivals or prevent them
from showing up and succeed in dismissing the leader.
Given that the object of study is the survival of leaders in (more or less)
democratic parties, I will focus on the chain of delegation and responsiveness to
shed light on the interplay between members, activists, factions and leaders
rather than on the size of the selectorate and that of the winning coalition, as
suggested by the selectorate theory (Bueno de Mesquita et al. 2003).
In this regard, one can distinguish between two different intra-party
environments (Chapter 1, Figure 1.2). In the first, which is typical of mass
parties, the chain of delegation begins with party members, which are entitled
to elect their representatives at the national party congress. The party con-
gress, in turn, selects a small committee – the party executive – that appoints
the party leader. The leader is the agent and should be responsive to the will
of his principal, which in this case is composed of high-ranking politicians
(party executive).
Conversely, the rise of the cartel party gave birth to another type of party, in
which the selectorate of the leader is enlarged. Party members are entitled to cast
a vote and select the party leader directly. Later, the leader will nominate the
party executive, appointing a team of followers that helps him to lead the party.
On the one hand, the leader should be directly responsive to party members (who
selected him according to the party line that he put forward). On the other hand,
the party executive becomes the agent of the leader (instead of his principal).
Accordingly, the chain of delegation is reversed: the leader can gain a wider
control over middle-level party activists and becomes more autonomous from
the wills of party factions.
Intra-party determinants of leader survival: hypotheses

What elements affect leadership duration? One can distinguish between two
kinds of factors that may help the leader to keep his office or damage his chance
of survival.
First of all, some factors are external to the party and are related to its per-
formance in the electoral arena or to its ability to get into office. Even if not all
parties are vote-maximizing actors, winning more votes enhances the bargaining
power of the party also in terms of policy payoffs and office spoils (Strøm 1990).
Consequently, the leader should not be removed from office. Conversely, an
electoral defeat can be taken as a signal that the valence traits of the leader are
no longer deemed appealing by voters. For the benefit of the party, the leader
could resign; otherwise, party members can remove him from office (Andrews &
Jackman, 2008; Bynander & ’t Hart, 2007; Ennser-Jedenastik & Müller 2015).
Analogously, party leaders can be dismissed (and replaced by politicians that
retain a stronger valence reputation) when they are involved in political scan-
dals. In fact, their public image becomes damaged, with negative consequences
on the future electoral performance of the party. In addition, office payoffs avail-
able to ruling parties can positively affect the survival of party leaders (Andrews
& Jackman, 2008; Ennser-Jedenastik & Müller 2015).
A second set of elements that should affect the duration of party leaders is
related to the internal life of the party. Let us consider the party as a coalition of
factions (Leiserson 1968) that retain similar but non-identical policy preferences.
Accordingly, a higher degree of ideological heterogeneity could be damaging for
the leader’s likelihood of survival.
In a less polarized party, each member will be closer to the party platform and
will gain a higher policy payoff (i.e., the cost of membership is, on average,
low). As a consequence, it is easier for the leader to make members, activists and
factions happier. On the contrary, in a more polarized party, there will be a
higher level of anger, as members will be on average less satisfied with the
internal policy equilibrium (and the cost of party membership will be higher).
In view of that, in a polarized party there will be a higher number of rivals
whose position is very distant from that of the leader; these unsatisfied members
will be more willing to defy the leader. They will try to change the internal equi-
librium by altering the balance of power between factions or by promising
changes in the structure of (policy and office) rewards, in order to win the
support of a number of activists large enough to dismiss the leader.
Hypothesis 7.1 (H7.1): The higher the polarization of the party, the shorter the
duration of its leader.
Nevertheless, by increasing the anger of party members and raising the
number of potential challengers, the degree of intra-party polarization could be a
necessary but non-sufficient condition to dismiss the leader. In fact, even in a
deeply polarized party, potential challengers may not have a credible opportunity
to succeed in their challenge, and therefore give up in order to avoid the cost
associated with a failure. The challenger can evaluate the likelihood of success
according to the shape of internal rules and decide whether to challenge the
leader or not. When the leader is indirectly selected in a small committee,
the leader is only the agent of high-ranking politicians. Any single member of
the party executive is potentially able to alter the equilibrium and, as a con-
sequence, internal polarization could be more damaging for the survival of the
leader (Musil & Bilgin 2016); indeed, potential challengers can easily reverse
the situation by catering to few high-ranking politicians.
Interestingly, the selectorate theory (Bueno de Mesquita et al. 2003) argues
that incumbent leaders have an advantage over potential challengers when the
winning coalition is small because the leader can skim a higher share of
resources for personal consumption, and these resources can be exploited to buy
the consent of the winning coalition. This happens since the challengers cannot
credibly commit to make a better offer. However, in a very small group (such as
the party executive) a potential rival can credibly commit to a different alloca-
tion of payoffs and leverage on unsatisfied members of the party executive to
alter the equilibrium if, once in power, the same amount of resources available
to the former party leader will be available to the challenger.
On the contrary, when the leader is directly selected by party members, he
could benefit from such direct mandate; thanks to this stronger legitimacy it
will be more difficult for potential challengers to defy him. In fact, any
reversal of the leader could be considered as a betrayal of the members’
mandate and the leader himself may say that his action is supported by the
consent of a wide number of members who voted him as the party leader. Just
to provide an example, in 2013, when Matteo Renzi (leader of the Italian PD)
presented a reform of the electoral system that was criticized by the minority
of the party (see Chapter 6), he replied arguing that the minority was infring-
ing upon the wills of three million PD members who selected Renzi as party
leader.
What is more, due to elements like name recognizability, intra-party direct
elections tend to be less competitive (Kenig 2009) and, therefore, only larger
shocks in the attitudes of party members may alter internal equilibriums if
compared to elections in a smoke-filled room where things can change even
after minimal shocks. Then, one would expect that party polarization is more
damaging for a leader who is indirectly selected. Notice that such expectation
represents an analogy with the difference between government stability
in parliamentary and presidential democracies. In parliamentary democracies,
where the prime minister is elected indirectly by the Parliament, the
polarization of the government has indeed a negative effect on survival
(Warwick 1994).
Hypothesis 7.2 (H7.2): The damaging effect of party polarization on the dura-
tion of the party leader is higher when the leader is selected indirectly in a small
committee.
Even so, the fate of the leader is not just a consequence of internal equilibri-
ums. In fact, the leader can take advantage of the distributive tasks linked with
his role and can act to reshape such equilibriums in order to maximize his
chances of survival.
It has been argued that the benefits coming from government participation can
extend the survival of the leader (Andrews & Jackman 2008). This, however,
should depend on how the leader allocates these office payoffs within the party. As
a consequence, being in office can have positive effects on the leader’s duration in
power when the allocation of spoils is overall satisfying for party members.
However, internal factions can also be strongly dissatisfied with the allocation of
office payoffs. Evidence suggests that this can affect government duration produc-
ing a cabinet reshuffle (Giannetti 2010). This could jeopardize the survival of the
party leader too. Once again, internal rules must be taken into account.
In the case of direct election, due to his stronger legitimacy, the leader is more
autonomous from factional constraints and can therefore allocate payoffs according
to his will. He can decide to preserve party unity, but he can also act to reward his
followers, given that their support is crucial to keep office.
While an unfair allocation of payoffs between party factions can upset high-
ranking politicians and produce a crisis in the party executive (and a removal of an
indirectly elected party leader), it hardly ever changes the mind of the thousands of
party members who voted for the party leader in a direct election, as many of them
do not receive direct payoffs from office. Accordingly, the leader may even decide
to skim a larger quantity of spoils, assigning them to his high-ranking followers
(who, in turn, can assign cascade benefits to low-ranking activists and party
members).
Conversely, a leader selected indirectly in a small committee is nothing more
than the output of inter-factional bargaining that is usually based on a proportional
norm of allocation (e.g., Gamson 1961; see also Chapter 5). As a consequence, the
leader retains no autonomy from factional ties. When the leader allocates payoffs,
he has to stick to the deal reached within the party executive. In this context, if the
leader tries to exploit his role to overpay the followers, his chance of survival
should decrease, as rival factions will coordinate to dismiss him from the office.
Hypothesis 7.3 (H7.3): If the party leader overpays his faction with a more than
proportional share of office payoffs, there will be a damaging effect on the dura-
tion of a party leader who is selected indirectly in a small committee.
Leader survival in Italy

To start with, I test these hypotheses on the Italian case as usual. To create a
more coherent research design, the analysis is limited to the so-called First
Italian Republic (1946–1994), for the following reasons. During the First
Republic the Italian party system was less fluid than in the Second Republic
(1994–present); there was a defined number of long-lasting political parties that
were steadily relevant in the political life and that always get represented in the
Parliament. Furthermore, their internal life was articulated into internal debates
scheduled at regular intervals and characterized by well-defined rules and a
certain degree of leadership turnover. What is more, in the First Republic almost
every party was composed of relatively organized factions and internal conflicts
were openly observable.
In the Second Republic, instead, only a few parties match this description.
Several parties are mere “personal parties” (Calise 2010), tied to the image and
the political fortunes of their leaders. These parties do not have a structured
internal life. They adopt peculiar rules and leaders can even be “self-enthroned”
(Kenig 2009).1 Therefore they are not suitable to perform a survival analysis.
In light of this, the time period of this study begins with the Italian Liberation
Day (April 25, 1945), when all the political parties re-emerged after the Fascist
dictatorship, and ends on January 29, 1995, with the dissolution of the last party
belonging to the First Republic.
In this regard, the survival of 82 party leaders belonging to the seven main
Italian parties,2 which represented around 90 percent of voters, will be analyzed:
Christian Democracy (DC), the Italian Social Movement (MSI), the Italian Com-
munist Party (PCI), the Italian Liberal Party (PLI), the Italian Socialist Party
(PSI), the Italian Socialist Democratic Party (PSDI) and the Italian Republican
Party (PRI).
This time span is large enough to allow results of 13 national elections, port-
folio allocation in 54 governments and the outcome of 129 party congresses to
be considered. The average duration of party leaders, measured in days, is 1,405
(standard deviation: 1,654), that means around three years and ten months,
though with a considerable variation both within and across parties. This is in
line with other post-war democracies (Ennser-Jedenastik & Müller 2015; Laing
& ’t Hart 2011: 113).
Duration was on average lower for the PSDI (753 days) and the DC (1,044)
leaders if compared to the PCI (3,267) or the PRI (2,011). Still, even within the
same party, survival rates vary a lot. For instance, in the PSI one can observe
leaders lasting between 106 and 6,054 days, and the same applies to many other
parties. Overall, the PCI leader Palmiro Togliatti (6,810 days) was the longest
ruling leader, while the head of the PSDI, Ugo Guido Mondolfo, was in office
for only 35 days.
This information is combined with data on 189 factional motions related to
61 contested congresses held by Italian parties between 1946 and 1990. By doing
that, the estimates described in Chapter 2 will be used to create the independent
variable Internal Polarization (H7.1), which corresponds to the ideological range
of preferences inside each party congress.3 When the party congress is not con-
tested by any faction, the value of 0 has been assigned to the variable Internal
Polarization, to attest the lack of public disagreement.4 Due to some missing
data, in a few cases it was impossible to find information on the internal struc-
ture of the party, though overall I accounted for the level of disagreement in the
78 percent of the congresses held by the parties included in this analysis.
The impact of polarization can be conditional on intra-party rules (H7.2).
Accordingly, I also consider it in interaction with the variable Direct Election,
which takes the value of 1 when the party leader is directly elected by the party
congress or by party members during the local party congresses and the value
of 0 when he is indirectly selected by the party executive or by another small
party body.
To test H7.3, concerning portfolio allocation, I create the variable Overpaid
Faction that measures to what extent the faction of the party leader gets a share
of ministers more than proportional than its size (i.e., larger than a “Gamsonian”
allocation of ministers). This variable shows positive values when the faction of
the party leader is overpaid in terms of office payoffs and negative values when
it is underpaid. The value of 0 is assigned to parties not in government. To
compute the style of payoffs allocation and see how much it diverges from the
proportionality rule described by Gamson (1961), I ascribed factional affiliation
to party ministers according to their signature on one of the competing motions
presented at party congresses. After including this variable into the analysis, the
number of missing cases gets larger due to some non-contested congresses, in
which it is not possible to assess the factional membership of ministers. The
impact of Overpaid Faction should be conditional on internal rules; therefore, it
will be tested in interaction with Direct Election.
Besides internal elements related to factional competition, I will also consider
external elements related to inter-party competition and to exogenous political
scandals that may hamper the image of the leader and its valence reputation. The
following control variables will be taken into account: Electoral Performance,
which corresponds to the change in the share of votes of party i at election t,
compared to the previous election held at time t – 1; Ruling Party, which takes
the value of 1 when the party is in government and the value of 0 otherwise;
Statutory Length, which corresponds to the ordinary lifetime of the leadership
term, in years; Corruption Scandal, which takes the value of 1 if the party leader
has been involved in a corruption scandal covered by the mass media. This
allows to catch the impact of Tangentopoli (Ceron and Mainenti 2018; Della
Porta 2001), which involved the leaders of all the main parties including Renato
Altissimo (PLI), Bettino Craxi (PSI), Giorgio La Malfa (PRI) and Carlo Vizzini
(PSDI), along with the effect of other famous scandals, such as: “Golden Jails”
(involving Franco Nicolazzi, PSDI); “Irpiniagate” (involving Ciriaco De Mita,
DC); “Icomec”; the Masonic lodge “P2”; the “Petroleum scandal” (see Sberna &
Vannucci 2013: 573).
The hypotheses are tested by means of a survival analysis. I do not make any a
priori assumption about the functional form of the dependent variable, therefore the
semi-parametric Cox proportional hazards model will be used (Cox 1972).5
The main assumption of this model is that the hazard function of each
observation follows the same pattern over time (Box-Steffensmeier & Jones
2004: 132) and a violation of this assumption might produce biased coefficients
and decrease the power of significance tests (Box-Steffensmeier & Zorn 2001:
974). The test based on the Schoenfeld residuals confirms that the assumption
holds as none of the variables violate the proportional hazards assumption.
Given the presence of many time-varying covariates, the dataset contains
more than one row per each party leader. New observations are generated when
the same leader faces a new party congress, new general elections or when a new
cabinet is formed during his tenure in office. All these additional observations
have been considered as censored. The total number of row is therefore 695,
though due to missing data only 459 observations will be used in the analysis.
Observations concerning party leaders that depart from office due to death or
severe illness (Almirante, Berlinguer, Michelini, Togliatti) or party dissolution
(Costa, De Martino I, Del Turco, Ferri, Fini II, Martinazzoli, Occhetto, Tanassi
I) are considered as right-censored too.6
Table 7.1 presents the results of the analysis. Hazard ratios are reported and
standard errors have been clustered on party leaders. An independent variable
contributes to decreasing the likelihood of a removal when its hazard ratio is
lower than 1, while when this is higher than 1 the likelihood of a removal
increases.
The results provide evidence in line with the hypotheses. To start with, the
effect of Internal Polarization is always large and statistically significant (H7.1).
In Model 2, when such variable increases by one unit, the risk of a leadership
reversal increases by 99 percent.7
At the same time, as expected, Internal Polarization is not equally damaging
for all parties (see Figure 7.1).8 Its effect is conditional on internal rules as
shown by the coefficient of the interaction term between Internal Polarization
and Direct Election. For instance, in Model 4 when the value of Internal Polari-
zation increases by one standard deviation from its mean (with all other vari-
ables set constant at their means and the dummy variables set equal to 0) there is
an increase in the hazard ratio (+0.601; statistically significant at the 95 percent
level of confidence) only among parties that select their leader through indirect
election (H7.2). Conversely, the same increase in Internal Polarization is not
damaging when the party leader is directly selected.9
In line with my theory (H7.3), I also find that the effect of portfolio allocation
is conditional on party rules. Indeed, the leader’s attempt to overpay his faction
damages his survival only when the leader is indirectly selected: a one standard
deviation increase from the mean of Overpaid Faction (with all other variables
set constant at their means and the dummy variables set equal to 0) increases the
hazard ratio (+0.515) and this effect is statistically significant at the 95 percent
level of confidence. Conversely, the same increase is no longer damaging when
the leader is directly selected (see Figure 7.1).10 The results are also similar when
excluding opposition parties from the analysis.
In line with previous studies, the results confirm that a good electoral per-
formance increases the chances of survival (Andrews & Jackman 2008;
Bynander & ’t Hart 2007; Ennser-Jedenastik & Müller 2015). The statutory
Table 7.1 Determinants of party leader survival in Italy
Model (I) (II) (III) (IV) (V) (VI)
Internal Polarization 1.942** 1.994** 2.879*** 3.227*** 3.264*** 4.559***

(0.654) (0.640) (1.079) (1.310) (1.268) (2.151)
Direct Election – 1.064 2.174 2.262 1.241 3.065
(0.406) (1.060) (1.201) (0.616) (2.203)
Internal Polarization X Direct Election – – 0.279** 0.316* – 0.240*
(0.176) (0.219) (0.209)
Electoral Performance – 0.895* 0.887** 0.857** 0.867** 0.851**
(0.052) (0.052) (0.053) (0.059) (0.064)
Ruling Party – 1.748 1.740 1.349 – –
(0.664) (0.643) (0.579)
Overpaid Faction – – – – 74.497*** 83.503***
(122.481) (139.047)
Overpaid Faction X Direct Election – – – – 0.001** 0.000**
(0.003) (0.001)
Corruption Scandal – – – 11.736*** 24.378*** 25.374***
(8.517) (16.859) (17.516)
Statutory Length – – – 1.097 1.106 1.211
(0.227) (0.252) (0.300)
AIC 458.761 385.670 385.517 373.779 310.671 310.954
N (Leaders) 60 52 52 52 49 49
N (Spells) 459 430 430 430 391 391
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. Standard errors clustered on leaders in parentheses. The results refer to a Cox model. The Breslow
method is employed for handling ties.
Figure 7.1 Marginal effect of Internal Polarization and Overpaid Faction on the survival
of party leaders.
Note
Difference in Hazard Ratio with 95 percent confidence interval for directly or indirectly elected
leaders.
length of the leadership term has no effect (in line with Ennser-Jedenastik &
Müller 2015). Being in office, per se,11 has no effect as well. This latter result
contradicts previous research (Andrews & Jackman 2008) though it is in line
with more recent analyses showing that the effect of holding office is not equal
for all parties (Ennser-Jedenastik & Müller 2015). This latter point is also coher-
ent with H7.3. Finally, the effect of Corruption Scandal is remarkably strong and
goes in the expected direction.
Interestingly, while the effect of the internal determinants of leader’s survival
is conditional on intra-party rules, the external determinants are not moderated
by alternative intra-party institutional frameworks. If interacting Corruption
Scandal and Electoral Performance with intra-party rules, one would notice that
the effect of these two variables remains the same, regardless of whether the
party leader is directly or indirectly elected.12
Leader survival in France and Germany

To conclude, I investigate how factionalism affects the survival of party leaders
in France and the likelihood of their replacement in Germany. Starting from
France, I focus on the PS, i.e., the only party on which information is
consistently available over time. By doing that, I can continuously analyze
internal dynamics covering almost the entire life of the PS, from the end of the
Epinay congress (June 16, 1971) when François Mitterrand was appointed, until
the electoral defeats of 2017 leading to Jean-Christophe Cambadélis’ resignation
(he officially quit on September 30, 2017).
I will analyze the survival of ten PS party leaders. Their duration is on
average equal to 1,536 days (standard deviation: 1,330), with a maximum of
4,195 days spent in office by Hollande (1997–2008), and a minimum of 442
days for Rocard (1993–1994).
The hypotheses are tested using the variables discussed above. In particular,
Internal Polarization (H7.1) corresponds to the ideological range of preferences
inside each party congress (the value of 0 was assigned to non-contested con-
gress, i.e., in the 12 percent of the cases considered in this analysis). To test H7.2
I consider the impact of Internal Polarization conditional on Direct Election. To
test H7.3 I use the variable Overpaid Faction (in interaction with Direct Elec-
tion), which records the gap between the share of ministers of the party leader’s
faction and its size (this variable is equal to 0 for parties not in government).
Electoral Performance is included as a control.
Table 7.2 presents the results of the survival analysis performed by means of
a semi-parametric Cox proportional hazards model.13 Due to time-varying cov-
ariates, the dataset contains multiple rows per each party leader (related to the
occurrence of a new election, cabinet or party congress). Accordingly, standard
errors have been adjusted clustering observations on party leaders. The total
number of rows amounts to 67 (with ten failure events).
Once again, the results support the hypotheses. Model 1 shows that a one-unit
increase in Internal Polarization boosts the risk of a leadership reversal by 79
percent (in line with H7.1). Model 2, however, confirms that the effect of
Internal Polarization is statistically significant only when leaders are indirectly
selected. When Internal Polarization enlarges by one standard deviation from its
mean (with all other variables set constant at their means) there is an increase in
the hazard ratio (+0.559; statistically significant at the 95 percent level of confi-
dence) for parties that select their leader through indirect election (H7.2).14
Analogously, and in line with H7.3, the effect of Overpaid Faction holds only
when the leader is indirectly selected, increasing the hazard ratio and the risk of
a replacement. The leader’s attempt to overpay his faction damages his survival.
Model 3 shows that this effect is not statistically significant for directly elected
leaders.15 Finally, the analysis confirms that a good electoral performance boosts
the survival of party leaders.
In the German case it is impossible to perform a survival analysis given that I
do not have enough information on the exact timing of the changes in polarization
recorded in the dataset. Polarization, in fact, varies when new political documents
are issued by party subgroups and factions but the exact date is often missing.
To cope with this, I revert to investigating the likelihood of a replacement of
the party leader by the party congress. Per each party, the unit of analysis is the
year. Notice that even if party congresses are not scheduled every year, one can
Table 7.2 Determinants of party leader survival (France) and replacement (Germany)
Model France France France Germany

(I) (II) (III) (IV)
Internal Polarization 1.789** 2.048*** 1.462*** 1.264*

(0.485) (0.542) (0.203) (0.699)
Direct Election 0.445 1.091 1.496 1.301
(0.320) (0.759) (1.367) (1.600)
Internal Polarization X Direct Election – 0.426* – –
(0.218)
Electoral Performance 0.895*** 0.919** 0.854*** −0.042
(0.031) (0.035) (0.026) (0.100)
Overpaid Faction 1.071 – 1.330*** –
(0.048) (0.103)
Overpaid Faction X Direct Election – – 0.795*** –
(0.069)
Ruling Party – – – −0.522
(0.820)
Party Size – – – −0.075
(0.072)
SPD – – – 2.067**
(1.012)
Constant – – – −0.983
(2.893)
AIC 66.014 68.239 62.496 66.514
N (Leaders) 10 10 10 –
N (Spells) 67 67 67 –
N (Years) – – – 66
Notes
Significance (two tailed): * = p < 0.1; ** = p < 0.05; *** = p < 0.01. The results refer to a Cox model,
for France (Model I-III), and to a logistic regression of Replacement, for Germany (Model IV).
Standard errors clustered on leaders in parentheses (Model I-III).
assume that an extraordinary congress would be held if there is a huge internal

opposition against the party leader and there are challengers willing to remove
him. Analogously, even if a new party congress does not imply a new leadership
contest, I assume that a replacement can occur in any year if the internal dissent
is strong enough.
The dummy variable Replacement is the dependent variable. This variable is
equal to 1 in the year in which the incumbent party leader is replaced by a new
one. Overall, there are for 14 replacements (21 percent of cases), though most of
them refer to the SPD and only two involve the CDU.16
The independent variables are the following: Internal Polarization, Direct
Election, Electoral Performance, Ruling Party and Party Size. All of them have
been operationalized as described above, while Party Size corresponds to the
share of votes won by the party in the previous election. A party dummy is
included to account for the fact that observations are nested in parties. The
interaction between Internal Polarization and Direct Election is not tested, given
that there is a single episode of (partial) direct election, though I control for it.
The results of logistic regression are displayed in the last column of Table 7.2
(Model 4). They confirm that a higher degree of polarization increases the likeli-
hood of a replacement, thereby reducing the duration of the incumbent party
leader even in the German case. This finding, in turn, also seems empirically
confirmed by the evidence related to the recent quarrel blown up in the CDU
before the 2018 party congress, leading to the replacement of Angela Merkel.
Leaders survival: findings

The present analysis confirms the theoretical framework and reveals that the
magnitude of intra-party ideological conflict or the adoption of a “winner-takes-
all” style of portfolio allocation have a negative effect on the duration of the
party leader, though only when he is indirectly selected in a smoke-filled room
by a small committee. In line with the theoretical framework, I find evidence
suggesting that a leader indirectly selected in a smoke-filled room is merely an
“agent” of the party executive, with little or no autonomy.
In such a party environment, the magnitude of ideological conflicts or the
style of portfolio allocation among party factions will markedly affect his
chances of survival. In fact, as the internal polarization grows, the leader can
face more challengers willing to defy him by trying to reshuffle the internal
allocation of payoffs and in turn these challengers can be successful as they only
need to convince a few members of the party executive to remove the leader.
Conversely, when the leader is directly elected by party members he retains
higher legitimacy due to the direct mandate and it becomes harder for the party
executive to remove him (which actually becomes the “principal” of such a
small committee).
Similarly, overpaying his own faction can be damaging for the party leader
when indirect selection is in place: indeed, as the majority faction becomes over-
paid all the other factions will get more and more angry and will try to persuade
(just a few) followers of the leader (i.e., members of the party executive) that
may have been excluded from the allocation of spoils to switch and dismiss the
leader. This is no longer true in the context of direct election, when the leader
can even increase the payoffs assigned to his followers precisely to preserve the
degree of consent that party members have granted to him in the election.
Interestingly, these findings are perfectly similar in the Italian case (on the
First Republic) and in the French case (concerning the PS). The analysis on
Germany confirms that internal polarization is damaging for the survival of
leaders. This latter finding is in line with the latest events of German politics. It
is worth noticing that, in recent years, polarization has increased inside the CDU.
Indeed, in the early 2018 new conservative factions started to organize precisely
to contest the leadership of Merkel. More generally, there is a growing divide
between the moderate followers of Merkel, who supported the centrist policies
of her cabinets, and more conservative subgroups that want to restore the
traditionalist values of this party, shifting its policy position to the right of the
center. In light of this, my theory and the empirical findings suggest that such
rising polarization could have damaged the survival of Merkel, increasing the
chances of a leader replacement. This is what happened in the autumn of 2018,
when Merkel announced her decision to quit the leadership. In the following
party congress, held in December 2018, a new party leader was appointed (see
also the discussion in Chapters 5 and 6).
Beside intra-party attributes, other elements also affect the survival of party
leaders. Interestingly, in Italy the effect of corruption scandals is huge. On the
one hand, this comes as good news for the accountability of political systems
and suggests that, due to the parties’ interest in avoiding loss in their valence
endowment, political parties tend to “clean up” the leadership when a scandal
reaches the mass media. However, this can also be worrying as long as the
inquiries of the judiciary and the media coverage can be ideologically biased
(Ceron & Mainenti 2015). As such, these inquiries can ruin the career of politi-
cians that have the chance to prove their innocence only many years later.
Although political scandals have played a big role in Italian politics, Italy is by
no means the only democracy subjected to corruption scandals (e.g., Rudolph &
Däubler 2016) and therefore this result could be of interest for the wide literature
that investigates the relationship between scandals and votes (for a review:
Ceron & Mainenti 2018).
More generally, all these findings help to shed light on the chain of respon-
siveness and on the degree of accountability. The analysis has implications also
for scholars dealing with the selectorate theory (Bueno de Mesquita et al. 2003),
which was applied to investigate the duration of leaders in democratic political
systems.
While the selectorate theory is very useful to assess the differences between
democracy, autocracy and monarchy, my results point to a difference about how
the selectorate theory applies to different types of democratic political systems
(i.e., between a democratic state with a large selectorate and a large winning
coalition and a political party with a small selectorate and a small winning coali-
tion, which in turn is different from a non-democratic state with the same size of
selectorate and winning coalition).
Accordingly, I argue that the principal–agent literature and my game–theoretic
model based on exit, voice and loyalty could complement the selectorate theory,
allowing the differences between different democratic political institutions to be
described by paying attention to the direction of the chain of delegation and the rel-
ative autonomy of the leader from internal constraints, a topic that is also connected
to the phenomenon of increasing personalization of politics.
Notes
1 For this reason, the choice to exclude the Second Republic seems more conservative.
2 Leaders who were in office more than once (non-continuously) are considered as
different subjects. For instance, I distinguish De Gasperi I (DC leader between 1945
and 1946) from De Gasperi II (DC leader between 1953 and 1954).
3 The variable Internal Polarization is strongly correlated (0.9) with another variable
that also takes into consideration the strength of each faction, beside its position.
Using this latter variable as a measure of polarization does not alter the results.
However, the variable Internal Polarization seems more in line with the theoretical
framework discussed above.
4 Considering these observations as missing cases does not alter any of the findings.
5 The results are robust to the adoption of others functional forms such as the para-
metric Weibull method.
6 Party leaders who are dismissed from office after getting appointed to a government
position (minister or prime minister) are not considered as right-censored because
their inability to keep both positions can be considered as a signal of weakness.
Indeed, some party leaders like De Mita managed to maintain both positions for a
certain time. Notice that promotion to a ministerial position can be a polite way to
mask a dismissal (this happened, for instance, to the PLI leader Villabruna).
7 Internal Polarization does not have a quadratic effect.
8 All the marginal effects have been calculated following Brambor et al. (2006).
9 The difference in the hazard ratio (0.031) is not statistically different from zero at the
95 percent level of confidence.
10 The difference in the hazard ratio (–0.155) is not statistically different from zero at
the 95 percent level of confidence.
11 Even when interacting Internal Polarization with Ruling Party, being in office has no
effects.
12 This result is in line with Ennser-Jedenastik and Schumacher (2015).
13 The Schoenfeld residuals test confirms that the proportional hazards assumption is not
violated.
14 The difference in the hazard ratio (0.037) is not statistically different from zero at the
95 percent level of confidence for directly elected leaders.
15 A one standard deviation increase from the mean of Overpaid Faction, with all other
variables set constant at their means, produces an increase (+1.218) in the hazard ratio
that is not statistically different from zero (at the 95 percent level of confidence) when
Direct Election is equal to 1.
16 Including in the analysis the replacement of Merkel (December 2018) does not alter
the results.
References
Andrews, J.T., & Jackman, R.T. (2008). If Winning Isn’t Everything, Why Do They Keep
Score? Consequences of Electoral Performance for Party Leaders. British Journal of
Bellucci, P., Garzia, D., & Lewis-Beck, M. (2015). Issues and Leaders as Vote Determi-
nants: The Case of Italy. Party Politics, 21(2), 272–283.
Bittner, A. (2011). Platform or Personality? The Role of Party Leaders in Elections.
Blondel, J., & Thibault, J.-L. (2009). Political Leadership, Parties and Citizens: The Per-
sonalisation of Leadership. Routledge, London.
Box-Steffensmeier, J., & Jones, B. (2004). Event History Modeling in Political Science.
Cambridge University Press, Cambridge.
Box-Steffensmeier, J., & Zorn, C.J.W. (2001). Duration Models and Proportional Hazards
in Political Science. American Journal of Political Science, 45(4), 972–988.
Brambor, T., Clark, W.R., & Golder, M. (2006). Understanding Interaction Models:
Improving Empirical Analyses. Political Analysis, 14(1), 63–82.
Bueno de Mesquita, B., Smith, A., Siverson, R.M., & Morrow, J.D. (2003). The Logic of
Political Survival. MIT Press, Cambridge MA.
Bynander, F., & ’t Hart, P. (2007). The Politics of Party Leadership Survival and Succes-
sion: Australia in Comparative Perspective. Australian Journal of Political Science,
42(1), 47–72.
Bynander, F., & ’t Hart, P. (2008). The Art of Handing Over: Mis-managing Party
Leadership Successions. Government and Opposition, 43(3), 385–404.
246–258.
Ceron, A., & Mainenti, M. (2015). Toga Party: The Political Basis of Judicial Investiga-
tions against MPs in Italy (1983–2013). South European Society & Politics, 20(2),
223–242.
Ceron, A., & Mainenti, M. (2018). When Rotten Apples Spoil the Ballot: The Conditional
Effect of Corruption Charges on Parties Vote Share. International Political Science
Review, 39(2), 242–255.
Cox, D.R. (1972). Regression Models and Life-tables. Journal of the Royal Statistical
Society, 34(2), 187–220.
Cross, W., & Blais, A. (2012). Who Selects the Party Leader? Party Politics, 18(2),
127–150.
Della Porta, D. (2001). A Judges’ Revolution? Political Corruption and the Judiciary in
Italy. European Journal of Political Research, 39(1), 1–21.
Diermeier, D., & Stevenson, R.T. (1999). Cabinet Survival and Competing Risks.
Ennser-Jedenastik, L., & Müller, W.C. (2015). Intra-party Democracy, Political Perfor-
mance and the Survival of Party Leaders: Austria, 1945–2011. Party Politics, 21(6),
930–943.
Ennser-Jedenastik, L., & Schumacher, G. (2015). Why Some Leaders Die Hard and
Others Don’t: Party Goals, Party Institutions, and How They Interact. In Cross, W., &
Pilet, J.B. (eds), The Politics of Party Leadership: A Cross-National Perspective,
Fabbrini, S. (1994). Personalization as Americanization? The Rise and Fall of Leader-
Dominated Governmental Strategies in Western Europe in The Eighties. American
Studies International, 32(2), 51–65.
26(3), 373–382.
Rationalité et Sciences Sociales Colloque, Collège de France, Paris, 3–4 June.
Horiuchi, Y., Laing, M., & ’t Hart, P. (2015). Hard Acts to Follow: Predecessor Effects
on Party Leader Survival. Party Politics, 21(3), 357–366.
Laing, M., & ’t Hart, P. (2011). Seeking and Keeping the Hot Seat: A Comparative Ana-
lysis of Party Leader Successions. In ’t Hart, P., & Uhr, J. (eds) How Power Changes
Hands, 111–132, Palgrave Macmillan, Basingstoke.
Musil, P.A., & Bilgin, H.D. (2016). Types of Outcomes in Factional Rivalries: Lessons
from Non-democratic Parties in Turkey. International Political Science Review, 37(2),
166–183.
Rudolph, L., & Däubler, T. (2016). Holding Individual Representatives Accountable: The
Role of Electoral Systems. Journal of Politics, 78(3), 746–762.
Sberna, S., & Vannucci, A. (2013). It’s the Politics, Stupid! The Politicization of Anti-
corruption in Italy. Crime, Law, and Social Change, 60(5), 565–593.
Warwick, P.V. (1994). Government Survival in Parliamentary Democracies. Cambridge
University Press, Cambridge and New York.
8 Conclusion
Intra-party preferences and intra-party rules

The book has put forward a theory of party unity and party fission, to explain
inter-factional conflict and negotiation, in the presence of bounded or auto-
nomous leaders. I used text analysis to gather information on the policy prefer-
ences of intra-party factions and subgroups, and in five empirical chapters I
investigated intra-party dynamics concerning a variety of outcomes, i.e., faction-
alism, allocation of policy and office payoffs, party splits and leader survival.
Here I provide a theory-testing summary, reporting the main findings to evaluate
whether these are in line with the expectations.
To begin with, however, I investigated to what extent intra-party preferences
and intra-party rules affect internal equilibriums, in each country. For this
purpose, Table 8.1 distinguishes between policy preferences and rules (mostly
concerning the direct election of the party leader). Per each outcome discussed
in Chapters 3 through 7, I identified whether – according to my expectations –
factional policy preferences and internal party rules produced an effect (Y) or
not (N). When no relationship was tested, the space is left blank. I considered
intra-party rules to play a role even when they have a moderating effect on other
variables, e.g., on policy preferences.
Table 8.1 The effect of factional policy preferences and internal rules in Italy, France and
Germany
Ch. Outcome Italy France Germany
Policy Rules Policy Rules Policy Rules
3 Factionalism Y Y/N Y N Y N
4 Policy Payoffs Y Y Y Y Y
5 Office Payoffs Y Y Y Y
6 Party Fission Y Y Y Y Y
7 Leader Survival Y Y/N Y Y/N Y N
Note
Per each outcome investigated in the related chapter (Ch.) the table reports whether policy and rules
produced an effect (Y) or not (N).
178 Conclusion
On the whole, preferences matter a lot while rules have a more nuanced effect.
To start with, factional policy preferences affect all the domains of politics in all
the three countries. Heterogeneous preferences increase the number of factions and
the fragmentation in party congresses. With respect to policy payoffs, factional
policy views act as constraints in the selection of the party platform. Concerning
office payoffs, factions’ ideology is taken into consideration by party leaders both
in Italy and in France (though the results are more robust in the Italian case);
powerful factions located away from the core of the party get a compensation in
terms of office rewards to balance their lower benefits linked with policy payoffs.
It was not possible to test such a relationship in the German case though. Neverthe-
less, qualitative evidence suggests that German party leaders take policy disagree-
ments into account when allocating portfolios, to preserve party unity. Finally,
internal polarization increases the likelihood of party splits and damages the sur-
vival of party leaders, favoring their replacement.
Although some scholars argue that factions are mainly interested in office
payoffs and point their attention to spoils, clientelism and patronage (Bettcher
2005; Golden & Chang 2001; Sartori 1976; Zuckerman 1979), these findings
show that factional policy preferences affect all the different outcomes investi-
gated herein. It is worth noticing that policy views do not only influence party
positions and party unity (which is fair enough), but they also affect the alloca-
tion of payoffs or the selection of leaders (i.e., political careers). This evidence
proves that factions care about policy too (Müller & Strøm 1999).
Moving to the impact of intra-party rules, this has been widely examined in
Italy and France. In Germany there is a limited variation in internal rules; there-
fore, I only barely tested the effect of direct election taking advantage of a single
episode of consultative membership ballot for the selection of the SPD national
party leader in 1993 (Astudillo & Detterbeck 2018).
Intriguingly, the results point to a limited effect of rules in the area in
which one would expect a greater impact, i.e., on the internal fragmentation.
In fact, no rule wields effect in France, while in Germany the episode of direct
election is associated with higher, rather than lower, fragmentation. In Italy,
some settings (democratic centralism and the most extreme case of leader’s
autonomy, i.e., personal party) reduce the likelihood of a contested congress,
while a disproportional internal electoral system reduces the number of fac-
tions. However, other rules like the direct election or the explicit ban on fac-
tions have no effects at all. This is not completely surprising given that inside
parties the policy-maker coincides with the policy-taker (Zincone 1972).
Accordingly, factions and leaders can always find some informal ways to skip
(or change) unwanted formal rules. In view of that, it becomes crucial to go
beyond formal statutory provisions to account for customs and informal rou-
tines (Aylott & Bolin 2017; Pilet & Cross 2014).
To the contrary, the direct election clearly affects the allocation of policy and
office payoffs. On the one hand, directly elected leaders can exploit their direct
legitimacy to get rid of factional constraints when selecting the party platform
(though this is less likely to happen when new elections approach). Such
Conclusion 179
autonomy allows them to set more moderate positions that are suitable to maxi-
mize votes or build “cartels,” boosting a leaders’ chances of getting into office.
On the other hand, directly elected leaders tend to be more autonomous in
portfolio allocation. They are able to reward their followers by increasing the
share of ministers assigned to the mainstream faction (particularly in France) or
to preserve party unity by rewarding factions that are dissatisfied in terms of
policy payoffs (particularly in Italy).
Conversely, indirectly elected leaders tend to be associated with a sense of
fairness, as they assign payoffs proportionally to the strength of each faction;
such parties work as a bounded oligarchy (Ceron 2012) and can be more
responsive toward the preferences of middle-level activists, as these leaders are
constrained by the wills of all factions in setting the party platform.
Rules also influence the likelihood of party breakups. In France, the direct elec-
tion is associated with an increase in splits (as one would expect in the presence of
autonomous leaders). In Italy, the impact of direct election is less crucial, but it
plays a role conditional on party system competitiveness (measured as the gap in
the votes share of the two main rival parties): this suggests that in dominant party
systems more splits can occur when leaders are autonomous. Nevertheless, I also
found that strict internal rules, such as democratic centralism or the centralization
of candidate selection mechanisms, promotes party fission in the Italian case.
To conclude, with respect to the survival of party leaders, there is no effect of
direct election per se, although it has a conditional effect and reduces the
Table 8.2 Intra-party dynamics in parties ruled by autonomous or constrained leaders
Leaders Party Platform Portfolio Allocation Party Fission Leader Survival

(Policy Payoffs) (Office Payoffs)
Constrained Tied to Proportionality Preserve unity Hurt by polarization

factional Overpay followers (through and unfair allocation
preferences (in low consensual of payoffs (in
competitive dynamics) violation of inter-
contexts) factional agreement)
Autonomous Free from Lower Promote Survive to
factional ties proportionality breakaways (in polarization and
(particularly in low low competitive unfair allocation of
competitive contexts) payoffs
contexts)
Factional Overpay followers Prevent splits
preferences through
matter only portfolio
when elections allocation (in
approach competitive
Free to set Overpay potential contexts)
moderate splinter groups (to
positions (to preserve unity)
build cartels)
180 Conclusion
negative consequences of internal polarization or unfair payoffs allocation on the
leader’s chance of survival.
To sum up, internal rules matter to some extent. Among them, and in line
with my theoretical framework, the selection of the party leader seems the most
important attribute. The empirical findings confirm that the direct election by a
wide selectorate or the indirect election by a narrow party executive produce
different effects on the autonomy of the party leader, generating different intra-
party environments. In light of this, Table 8.2 condenses these findings and
briefly describes what alternative dynamics can be observed inside parties ruled
by autonomous or constrained leaders.
The game of intra-party politics: evidence and findings

Going back to the game–theoretic model presented in Chapter 1, here I resume
the expectations generated by the theory and compare them with the empirical
results.1 By doing that, I can try to validate or deny the general claims derived
from the theory, assessing whether it resisted to several tests of hypotheses. In
addition, the practical implications of these findings will be discussed.
Table 8.3 retrieves the parameters related to the theoretical framework (see
Table 1.2, Chapter 1), and compares their expected (Exp.) and observed (Obs.)
effect on three outcomes of intra-party dynamics that emerge from the game: the
use of voice (factionalism, Chapter 3), the reaching of a compromise (in the
allocation of policy and office payoffs, Chapters 4 and 5) and the exit of minority
factions from the party (Chapter 6).
Voice – Factionalism, Chapter 3. The “Party Unity and Party Fission” game
and the theoretical framework related to autonomous or non-autonomous leaders
are mostly focused on distributive dynamics that might prevent splits. However,
the model also raises expectations about the expression of internal dissent. On
Table 8.3 Parameters of the game: Expected and observed effects on intra-party dynamics
Label Parameter Description Voice Compromise Exit
Exp Obs Exp Obs Exp Obs
ε Exit cost – – – Ø – Ø/–

π Trade-off + + + + + +
ω Breakup cost + – + + – –
ν Voice cost – – – Ø/– + +
AL Autonomous leader Ø/– Ø/– – – + Ø/+
AL*ω AL in competitive contexts Ø/– Ø/– + + – –
BL Bounded leader + Ø + + – –
Note
The table reports expected (Exp) and observed (Obs) effects. The minus (–) indicates a negative
effect of an increase in the value of the parameter on the outcome. The plus indicates a positive
effect. The null sign (Ø) indicates a null relationship.
Conclusion 181
the whole, these expectations have been confirmed by the empirical test of hypo-
theses performed in Chapter 3.
First of all, larger exit costs (ε) – for instance, due to strong party loyalty –
tend to contain factionalism. In personal parties, members are firmly tied to their
leader and nobody tries to contest the congress: internal minorities, if they exist,
will avoid challenging the leader and their only realistic option is the
compliance.
Larger values of π imply higher internal polarization. Inside polarized parties
a broader number of factions will be dissatisfied with the status quo. Polarization
increases the cost of membership and decreases the value of the party label.
Internal subgroups will be more likely to use the voice in order to advance their
policy views and to bargain for more favorable conditions by threatening to
defect. Indeed, polarization has a strong effect on internal fragmentation in all
three countries.
In line with the theory, an increase in ν reduces the use of voice. In Italy, it
can be noticed that disproportional internal electoral rules and centralized (closed
list) or strict internal settings (e.g., democratic centralism) can avoid the prolifer-
ation of factions; conversely preference voting represents an opportunity for the
persistence and proliferation of factions. The effect of democratic centralism
implicitly holds also in France (see the discussion on the PCF in Chapters 2 and
3), though here the internal disproportionality did not play a role (unless exclud-
ing the MoDem from the analysis).
While several studies consider factions as “power subgroups” that grows
inside ruling parties to exploit the benefits of patronage (Bettcher 2005; Golden
& Chang 2001; Sartori 1976; Zuckerman 1979), my theoretical model argues
that, for the sake of preserving cohesion and government effectiveness, parties in
office could be less factionalized. Interestingly, the empirical results outlined in
Chapter 3 seem to confirm this latter argument: both in Italy and France, the
number of factions decreases in ruling parties. The cost of internal division (ν)
can be higher for ruling parties (as they face greater public scrutiny) and, accord-
ingly, lower factionalism can be observed there.
Analogously, when new general elections approach, the level of internal
fragmentation decreases. While looming elections should increase the breakup
cost (ω) and therefore the degree of factionalism, they also raise the cost of voice
(ν). As a result, the whole party tries to avoid the display of public disagreement,
which can be damaging for the party’s electoral performance (Greene & Haber
2015). Accordingly, this produced fewer contested congresses, a lower number
of factions and low fragmentation in the leadership race. This does not imply
that factionalism has disappeared, but only that (for tactical reasons) it has been
hidden to the public in order to present a unified and cohesive front to the voters
(see below). In other terms, divisions are still there (and can even get larger) but
only behind the closed doors of parties’ smoke-filled rooms.
To conclude, although I claimed that direct election of party leaders should
decrease voice, no effect was found (apart from personal parties). This might
happen because direct election fosters a leader’s strength in front of middle-level
182 Conclusion
activists, so that a larger number of potential challengers will contest the leader-
ship (as in the German case), to take advantage of the higher autonomy. Interest-
ingly, in the French case I also found that – in line with the expectations – direct
election tends to reduce factionalism when new elections approach.
Compromise – Payoffs allocation, Chapters 4 and 5. In the process of
position-taking (Chapter 4), constrained leaders have to consider the policy pref-
erences of all factions, while autonomous leaders are free to set the party posi-
tion on their own (in fact, they tend to adopt platforms closer to the center of the
policy space, increasing the probability to build cartels and take part in coalition
governments). However, when new general elections approach the need for party
unity (ω) sharply increases. In this context, the leader will select the party plat-
form taking into account the stakes of all factions, opting for consensual
dynamics in order to reach a compromise. This also applies to autonomous
leaders: in the case of looming elections they will negotiate with internal minor-
ities. These findings support my theory and indicate that the need for party unity,
heightened in competitive contexts (wide ω), is the primary source of factional
cooperation and promotes the achievement of compromises between party
factions.
As far as portfolio allocation is concerned, most of the expectations are con-
firmed. There are a couple of exceptions though. The theoretical framework
argues that, as time passes, the price for infringed loyalty (ε) grows, reducing the
blackmail power of internal minorities and the need for a compromise. Con-
versely, other theories claim that repeated inter-factional interactions should
foster the emergence of cooperative norms, increasing proportionality in payoffs
allocation. In Chapter 5 a null effect was found, and this can signal that both
mechanisms are at work. This outcome does not seem to be in line with the
expectation, nevertheless I contend that the impact of exit costs is still there, as it
counterbalances the effect of repeated cooperation. Accordingly, this finding
does not utterly contradict the theory (see also the discussion in Chapter 6).
Analogously, an increase in ν (due to disproportional internal rules) does not
seem to have an effect on the fairness of portfolio allocation. Nevertheless, such
disproportional rules exert, per se, mechanical (and possibly also psychological)
unfair effects, reducing the size of minority factions compared to the party main-
stream. From this perspective, they are already at work to reduce consensual
dynamics of perfect proportionality. Beyond this point, however, any additional
“winner-takes-all” effect could be counterbalanced by an increase in π: if
minority factions are underrepresented, they could credibly threaten to break
away. This will foster their leverage, leading to an overall null effect on distribu-
tive dynamics.
The relevance of π finds additional support in Chapter 5. In Italy and France,
medium-sized minority factions, which are ideologically distant from the party
mainstream, are overpaid in terms of portfolios. Their lower amount of policy
payoffs (which boosts π and provides them with a credible exit threat) is
balanced by an increase in their share of office payoffs, though only when their
size is large enough to really pose a threat to the party (raising ω).
Conclusion 183
In turn, large values of ω enhance proportionality in portfolio allocation: in
France, when new elections approach, for the sake of preserving party unity the
leader tends to adopt a criteria of perfect proportionality; in low competitive
parliamentary arenas (low ω) the party leader tends to reward the mainstream
faction with a more than proportional share of ministers (see the results on the
Italian case).
The impact of direct election is in line with the theory too. The degree of pro-
portionality is lower in parties ruled by autonomous leaders; they try to overpay
their own faction and can do that more easily than bounded leaders, which would
be punished for any violation of the inter-factional agreement and will risk being
replaced. However, when party system competitiveness boosts the need for party
unity, even autonomous leaders step back to a fair allocation of payoffs or
choose to overpay factions that can credibly split and damage the party.
Exit – Party fission, Chapter 6. With respect to factional breakaways, the
expectations generated by the theoretical model have been successfully tested.
To start with, when ν grows one would expect more splits. Indeed, centralized
candidate selection processes (such as closed list PR or centralized decision-
making) and strict party discipline (i.e., democratic centralism) force dissenters
to exit from the party.
In turn, elements that raise the cost of exit (ε) tend to reduce the probability of
a fission. In Italy, the disproportionality of the electoral system acts as a barrier
to exit: for splinter groups it becomes more difficult to get parliamentary repres-
entation after the split. In addition, party loyalty seems to play a role, decreasing
the likelihood of breakaways in older parties.
When ε is larger than π dissenting members would remain inside the party
even if their share of payoffs is inadequate. Conversely, large values of π favor
the occurrence of party splits. Factions that are strongly underpaid can reason-
ably expect to gain more payoffs if leaving the party (net of the cost ε). Chapter
6 confirms this argument with respect to both policy and office payoffs.
As far as party system competitiveness is concerned, this attribute has an effect
conditional on the leader’s attitude toward unity or cohesion. Accordingly, when
the margin of the ruling coalition is narrow (high ω), parties in office will cater to
potential splinter groups in order to avoid splits. Conversely, party fissions can
occur more frequently in ruling parties when the margin of the ruling coalition is
safer; in this case the leadership is more willing to accept the risk of splits for the
sake of promoting government effectiveness through party cohesion. Interestingly,
in the Italian case the effect of party system competitiveness is also conditional on
the autonomy of the leader. Directly elected leaders can take advantage of non-
competitive party systems (small ω) to get rid of internal minorities and overpay
their supporters. Autonomous leaders trade unity for cohesion. As a result, they
boost the leverage of minorities (increasing π) and increase the probability of a
split. Finally, in France I also found that leader autonomy has an effect per se. In
the presence of direct election, the probability of a split raises. This indicates that,
overall, it is easier to protect party unity when leaders are bounded and stick to the
consensual inter-factional agreement that appeases all party factions.
184 Conclusion
To sum up, almost all the hypotheses are empirically supported by the data.
According to the theoretical framework I formulated 21 general expectations.
Overall, 18 of them hold true (there is a full support in 15 cases, and some
support in other three cases). In two cases no effect was found: exit costs do not
seem to affect the allocation of payoffs and there was no strong evidence sug-
gesting that factionalism grows in parties ruled by bounded leaders. Only in one
case is the evidence apparently contradicting the hypothesis. Indeed, if looming
elections raise the breakup cost one should observe a higher degree of factional-
ism, while the results indicate that factionalism decreases as new elections
approach. As discussed above, this might happen because parties tend to avoid
the display of public disagreement in such context. Factionalism might increase,
but only behind closed doors so that we are not able to observe such an
outcome.2 Nevertheless, it is worth noticing that direct election tends to reduce
factionalism when new elections approach, and this is in line with the expecta-
tions. To conclude, the theoretical framework has brilliantly survived the empiri-
cal tests performed throughout the chapters.
Implications for real-world politics

Far from any prescriptive intent, the results of this book provide some implica-
tions concerning real-world politics. Understanding intra-party dynamics, with
an eye to the interplay between party leaders and party factions in alternative
intra-party institutional settings, can be crucial for political analysts, advisors
and politicians that aim to reform the electoral system, build a new model of
party or transform the shape of intra-party and inter-party competition (e.g.,
Gauja 2017).
To start with, any attempt to reduce the number of parties must take into
account the risk of a growing factionalism. Electoral reforms that aim to reduce
political fragmentation should also consider the opportunity to design intra-party
rules able to contain intra-party heterogeneity, limiting the incentives for the
proliferation of factions; otherwise, the new party system will be composed of a
lower number of parties that are affected by broader internal divisions and this
could undermine the primary intent of the institutional reform.
At the same time, intra-party rules should be able to preserve party unity.
Adopting strict discipline might in fact contain factionalism and ideological het-
erogeneity, but at the expense of party unity. On top of that, when political elites
try to alter the shape of party competition, by means of party mergers, they must
be aware that such efforts could vanish due to party fissions. In view of that, any
plan to modify the structure of the party system can be successful only if polit-
ical system rules, intra-party rules and leaderships’ wills are not rowing against
each other.
The empirical findings disclose the potential impact of rules. To start with,
electoral systems based on preference voting undermine the cohesion of the
party and boost factionalism and internal competition. Conversely, PR systems
based on closed list and centralized candidate selection mechanisms could
Conclusion 185
contain the proliferation of factions; however, these institutions will increase the
likelihood of party splits, favoring the fragmentation and the instability of
the party system, unless other attributes of the political system emphasize the
importance of party unity, forcing leaders to cater to the wills of internal
minorities. Disproportional electoral rules (both inside the party and in the party
system) are effective in decreasing party fragmentation, while taking faction-
alism on a leash.
Rules are not always effective though. The analyses reveal that many formal
intra-party rules have no direct effect on the magnitude of factionalism. Never-
theless, some provisions included in party statutes can contribute to foster
loyalty, discipline and cohesion while containing the risk of splits.
In this regard, the direct election of the party leader plays a crucial role. It
grants a certain amount of autonomy to party leaders, which can exploit such
discretionality in a flexible way: they can adopt more consensual intra-party
dynamics when there is a huge need for party unity, or can use the whip when
party cohesion is at stake. Depending on the context (and on their aims) auto-
nomous leaders can prevent or promote party splits. In view of that, it is crucial
that party leaders stay focused on pursuing the party’s interest, getting rid of
internal minorities when this would benefit the party or introducing consensual
dynamics when this is needed.
If direct election makes party leaders more effective and decreases the black-
mail power of internal minorities, their stronger autonomy (as in any principal–
agent dilemma) might also have unwanted consequences. Charismatic leaders
can try to exploit their privileged status to transform the party into their personal
property; by appealing directly to party members, they can defang middle-level
activists, making the party slip toward a “Bonapartist” and “plebiscitarian”
organization, or generating a wave of splits and a party collapse.
It is worth noticing how the unwillingness to compromise with internal
minorities, combined with extreme levels of internal disagreement, have been
among the causes of the recent collapse of the PS in France, and the crisis of the
PD in Italy. In these parties, the behavior of the leadership was not successful in
restoring unity and in boosting the party’s electoral performance.
On the one hand, the conflictual choices taken by Macron, Valls or Hamon
(and the frondeurs) dramatically weakened the PS. It will be interesting to see
whether the party will be able to recover or not from this failure and in which
form. This already happened once in the history of the PS.
As a phoenix, it was able to rise from the ashes of its predecessor in the 1971
Epinay congress, increasing in a few years from 5 percent (1969) to 49 percent
(1974) of votes in the Presidential election.
In turn, the long-lasting and unresolved conflict between the Renzi leadership
and the PD left-wing minority damaged the party label and its credibility, trig-
gering the downfall of the party in the polls. Although Renzi resigned in 2017
(February), a few weeks later he ran again in the 2017 party congress and was
reappointed as leader (April). The poor performance of the party in the 2018
general election led him to resign for a second time and sparked rumors about
186 Conclusion
his willingness to leave the PD, following the example of Macron. Renzi repeat-
edly stated his aim to fight against intra-party factions using the “flamethrower.”3
Despite repeated waves of splits (see Chapter 6), new left-wing factions con-
tinued to form and resist his strategy; apparently the flamethrower was ineffec-
tive against such “phoenix” factions.
Indeed, the PD still remains highly factionalized. A new record has been
reached in the 2019 party congress, with seven candidates running for the leader-
ship (Nicola Zingaretti, Maurizio Martina, Francesco Boccia, Dario Corallo,
Maria Saladino, and the co-candidates Anna Ascani and Roberto Giachetti),
while the other three have decided to step back (Cesare Damiano, Matteo
Richetti and Marco Minniti).
Having acknowledged such fragmentation, in his newsletter published on
December 10, 2018, Renzi kept criticizing factionalism arguing that PD factions
are detrimental for the party; at the same time, he announced the will to carry on
his political activity reinforcing the Civic Committees (Comitati Civici),4 a non-
partisan organization that was launched during the ninth annual meeting of
Renziani faction (Leopolda 2018).5 History will tell us if these Civic Committees
are the first step toward a splinter party and if the PD will be affected by another
fission, this time promoted by its right-wing.
On the other hand, the behavior of Martin Schultz and Angela Merkel in
Germany looked different, as these two leaders (reluctantly or not) have put the
party’s interest above all, stepping back from their leadership position perman-
ently to help the party recover from shrinking electoral support and internal
polarization. It will be interesting to see whether the new SPD and CDU leader-
ships will be able to heal internal divisions, unifying the party through consen-
sual dynamic (a strategy typically used by the CDU to prevent splits: Dilling
2018) in order to regain the support of voters. In both camps, there will be no
easy solution to internal disagreements: factionalism and the threat of splits will
keep affecting the life of the SPD and the CDU for a while, as it never happened
before.
Having said that, intra-party dynamics are not the only deterministic outcome
of electoral systems and formal intra-party rules. On the contrary, the book sug-
gests that internal policy preferences affect any area of intra-party politics. Even
autonomous leaders (in different institutional settings) should acknowledge that
they are not playing alone. Indeed, intra-party politics is a game involving (at
least) two players and the policy preferences of internal factions contribute to
shaping the outcomes of the game too.
From this perspective, the book also suggests that party factions are not evil.
They are not just greedy political organizations focused on patronage and office
payoffs. On the contrary, factions retain their distinct ideology, preferences and
policy views. Such heterogeneous preferences are not merely a source of internal
conflict, but can even become a resource for the whole party, as long as they
push the leaders to take the right choice in the intra-party game (ignoring their
stakes or listening to them). If factionalism degenerates, producing a huge and
publicly observable disagreement, such behavior could damage the party and
Conclusion 187
strengthening internal cohesion would be a benefit. However, factions can also
produce benefits; by structuring dissent they can help to reduce the internal
polarization (Dewan & Squintani 2016). If factions contribute to developing the
party structure and to increasing the total amount of payoffs available inside it,
they should be rewarded according to their actual contribution: this would keep
internal conflicts under control, reinforcing the incentives for intra-party
cooperation (Ceron 2014).
Intra-party politics: directions for future research

The present book improves our knowledge of intra-party dynamics, providing
insights on a number of political topics. Such improvement certainly represents
a step forward in the understating of intra-party competition. Nevertheless, many
other steps should be taken and several research groups are already active in
moving this knowledge forward (for a review: Polk & Kölln 2017).
To start with, the present analysis could be extended to evaluate the impact of
factionalism and internal disagreement on a variety of topics: electoral perform-
ances (Greene & Haber 2015; Ibenskas 2017, Lehrer & Lin 2018), government
formation (Bäck et al. 2016b; Ceron 2016), cohesion in roll-call votes (Ceron
2015; Close & Gherghina 2018), everyday policy-making (Ceron et al. 2019;
Haber 2015; Marx & Schumacher 2013) as well as party formation and termina-
tion (Bolleyer et al. 2018).
The measurement of internal division can also be adjusted to study inter-
factional bargaining in a multidimensional policy space. This could be relevant
for parties that are internally divided along multiple lines (for instance, factions
might retain different positions on the economic dimension and on civil rights
issues), or to compare parties that are divided on some issues but stand united on
others (this can be particularly interesting for single issue parties such as the
Greens).
Future analyses on party factions could also take into account the role of col-
lateral organizations, such as trade unions and interests groups, to investigate the
relationship between these organizations and political parties or single factions
in parties (Allern & Bale 2017; Ceron & Negri 2017, 2018); similarly, factional-
ism can be examined from a multilevel perspective, to see how local party
branches affect internal dynamics (Bäck et al. 2016a; Debus & Navarrete 2018;
Verge & Gómez 2012).
Concerning intra-party rules, the analysis of party statutes performed in this
book can be improved to shed more light on the formal (and informal) degree of
intra-party democracy (Borz & Janda 2018) and intra-party balance of power
(Schumacher & Giger 2017). In this regard, several research projects have
already started to deepen the conceptualization and the measurement of intra-
party democracy (Rahat & Shapira 2017), providing original data on internal
rules. In this field, one of the latest developments is represented by the Political
Party Database Project (Poguntke et al. 2016), which provides data on party
resources, party decision processes and the outcome of decision-making
188 Conclusion
procedures in several countries. Beside rules, future research can also investigate
the party leadership, providing new data on party leaders and developing valid
measures of leader’s strength; studies on the selection of party leaders are a first
important step in this direction (see the Comparative Study of Party Leaders
project: Pilet & Cross 2014).
Concerning the policy preferences and the salience of different policy issues,
future studies can provide in-depth investigations into party congresses that
represent the arena in which party members can engage, propose new ideas and
vote on the party’s policies and leadership. By analyzing the speeches delivered
therein by individual politicians, it would be possible to compare the behavior of
backbenchers and frontbenchers, to shed light on the diffusion of information
inside parties and to understand internal decision-making processes. For this
purpose, in 2014 the Party Congress Research Group (PCRG) has been founded
by Greene and colleagues (Greene et al. 2016; Ceron & Greene 2019;
Kaltenegger & Müller 2018; Schumacher et al. 2018). So far, the PCRG project
already includes researchers from Austria, Germany, Israel, Italy, the
Netherlands, Scotland and Switzerland.6 The PCRG aims to collect speeches and
data on party congresses and conferences to evaluate the impact of such meet-
ings on intra-party decision-making. Party congresses can be scrutinized from an
informational perspective, considering these meetings as an opportunity in terms
of intra-party communication, to reduce the uncertainty about the distribution
and goals of intra-party actors and to provide a forum for the making and/or
legitimation of difficult decisions.
Along this vein, it would be interesting to map policy heterogeneity, factional
affiliation and the role of subgroups or internal collateral organization in parties
that usually hold unitary congresses or do not hold congresses at all, as well as
in non-factionalized parties. This can contribute to an evaluation of whether the
heterogeneity of preferences, although not publicly expressed, still contributes to
shaping the intra-party decisions in these types of parties or, conversely, dissent-
ing voices are not taken into account.
In light of this, beside party congress speeches and documents, other sources
of data can be profitably used to analyze intra-party politics. Among them,
parliamentary speeches (Bäck & Debus 2016) or elite and candidate surveys
(Steiner & Mader 2017).
To conclude, the recent rise of social media and social networking sites pro-
vides new sources of information on the policy preferences and the actual behav-
ior of factions and individual politicians. As far as the behavior of politicians on
social media will remain predominantly transparent, as it is nowadays, such data
can be suitable to study intra-party cohesion and disagreements (Boireau 2014;
Ceron 2017a, 2017b; Livne et al. 2011; Sältzer 2018; Temporão et al. 2018).
Furthermore, text analysis of online content can also be combined with other
approaches, such as network analysis of social media accounts and conversations
(Barberá 2015; Ecker 2017; von Nostitz et al. 2018).
To sum up, the recent developments of text analysis, web scraping and social
media analysis pave the way to new opportunities to opening the black box of
Conclusion 189
political parties and to investigating in more depth the internal dynamics. It goes
without saying that, in an era of declining party membership, disaffection, anti-
politics sentiment and raising support for populist parties often tied to charis-
matic leaders, one can fear that political parties are losing their grip, and their
relevance is shrinking. Nevertheless, if the representative democracy will survive
its actual crisis, there will still be the need for some sort of political parties. In
this case, parties will continue to be composed of members retaining similar but
non-identical preferences, they will be internally divided again, and they will
still have to organize meetings to resolve internal disagreements and take deci-
sions. Whether these meetings and these debates will be held in the party head-
quarters or maybe on an online social networking site does not really matter.
Parties will be there; factions will be there. These political actors will continue
to play a role in everyday politics and intra-party politics will retain its influence
on the party system, shaping the outcomes of the democratic polity.
Notes
1 Here I will also consider some results that have been discussed only in the text or in the
endnotes (without being displayed in tables or figures).
2 There is a lot of empirical evidence in this regard. As an example, in the French PS the
seven factions active in the 1979 congress disappeared a few months later, in the last
congress before the 1981 Presidential elections; the same happened to the seven PS
factions that took part in the highly divisive 1990 congress: none of them contested the
following congresses held in 1991 and 1992, before the 1993 Legislative elections.
Obviously, one can hardly ever believe that perfect cohesion was reached in such a
short time period; therefore, I claim that heterogeneity and disagreements were still
there, and factions still played the voice option, though hiding it to the wide public of
voters.
3 www.matteorenzi.it/enews-554-lunedi-10-dicembre-2018/.
4 www.matteorenzi.it/enews-554-lunedi-10-dicembre-2018/.
5 https://firenze.repubblica.it/cronaca/2018/10/21/news/leopolda_ultimo_giorno_matteo_
renzi-209547912/.
6 https://partycongressresearchgroup.wordpress.com/.
References
Allern, E.H., & Bale, T. (eds) (2017). Left-of-Centre Parties and Trade Unions in the
Twenty-First Century. Oxford University Press, Oxford.
Aylott, N., & Bolin, N. (2017). Managed Intra-party Democracy: Precursory Delegation
and Party Leader Selection. Party Politics, 23(1), 55–65.
Bäck, H., & Debus, M. (2016). Political Parties, Parliaments and Legislative Speech-
making. Palgrave Macmillan, Basingstoke.
Bäck, H., Debus, M., & Klüver, H. (2016a). Bicameralism, Intra-party Bargaining, and
the Formation of Party Policy Positions: Evidence from the German Federal System.
190 Conclusion
Bäck, H., Debus, M., & Müller, W.C. (2016b). Intra-party Diversity and Ministerial
Selection in Coalition Governments. Public Choice, 166(3–4), 355–378.
Barberá, P. (2015). Birds of the Same Feather Tweet Together: Bayesian Ideal Point
Estimation using Twitter Data. Political Analysis, 23(1), 76–91.
Boireau, M. (2014). Determining Political Stances from Twitter Timelines: The Belgian
Parliament Case. In Proceedings of the 2014 Conference on Electronic Governance
and Open Society: Challenges in Eurasia, 145–151.
Bolleyer, N., Ibenskas, R., & Bischoff, C. (2018). Perspectives on Political Party Death:
Theorizing and Testing Downsian and Sociological Rationales. European Political
Science Review, 1–17, doi: 10.1017/S1755773918000176.
Borz, G., & Janda, K. (2018). Contemporary Trends in Party Organization: Revisiting
Intra-party Democracy. Party Politics, 1–6, doi: 10.1177/1354068818754605
Ceron, A. (2015). Brave Rebels Stay Home. Assessing the Effect of Intraparty Ideological
Heterogeneity and Party Whip on Roll-call Votes. Party Politics, 21(2), 246–258.
s11127-019-00644-0.
491–508.
Close, C., & Gherghina, S. (2018). Towards a Better Understanding of Parliamentary
Unity. Parliamentary Affairs, gsx074, https://doi.org/10.1093/pa/gsx074.
Debus, M., & Navarrete, R.M. (2018). Do Regional Party Primaries Affect the Ideo-
logical Cohesion of Political Parties in Multilevel Systems? Evidence from Spain.
Party Politics, 1–13.
Science, 60(4), 860–881.
Dilling, M. (2018). In Defence of Moderate Factionalism. Paper Presented at ECPR
Ecker, A. (2017). Estimating Policy Positions Using Social Network Data: Crossvalidat-
ing Position Estimates of Political Parties and Individual Legislators in the Polish Par-
liament. Social Science Computer Review, 35(1), 53–67.
Gauja, A. (2017). Party Reform: The Causes, Challenges, and Consequences of Organ-
izational Change. Oxford University Press, Oxford.
Conclusion 191
588–622.
Greene, Z., Ceron, A., Schumacher, G., & Fazekas, Z. (2016). The Nuts and Bolts of
Automated Text Analysis. Comparing Different Document Pre-Processing Techniques
in Four Countries. Open Science Framework. November 1. osf.io/ghxj8.
Haber, M. (2015). The Legislative Consequences of Internal Conflict and Inter-party
Divisions. Research & Politics, https://doi.org/10.1177/2053168015589216.
Kaltenegger, M., & Müller, W.C. (2018). Party Congress Politics: Towards the Real
Story of Party Organizations. Paper Presented at the Annual Meeting of the Party Con-
gress Research Group, Vienna June 25–26, 2018.
Lehrer, R., & Lin, N. (2018). Everything to everyone? Not when you are internally
divided. Party Politics, doi: 10.1177/1354068818812222.
Livne, A., Simmons, M.P., Adar, E., & Adamic, L.A. (2011). The Party is Over Here:
Structure and Content in the 2010 Election. Association for the Advancement of Arti-
ficial Intelligence.
Marx, P., & Schumacher, G. (2013). Will to Power? Intra-party Conflict in Social Demo-
cratic Parties and the Choice for Neoliberal Policies in Germany, the Netherlands and
Spain (1980–2010). European Political Science Review, 5(1), 151–173.
Müller, W.C., & Strøm, K. (eds) (1999). Policy, Office, or Votes? How Political Parties
in Western Europe Make Hard Choices. Cambridge University Press, Cambridge.
Pilet, J.-B., & Cross, W.P. (eds) (2014) The Selection of Political Party Leaders in Con-
temporary Parliamentary Democracies. Routledge, Abingdon.
Poguntke, T., Scarrow, S.E., & Webb, P.D. (2016) Party Rules, Party Resources and the
Politics of Parliamentary Democracies: How Parties Organize in the 21st Century.
Polk, J., & Kölln, A.-K. (2017). The Lives of the Party: Contemporary Approaches to the
Study of Intraparty Politics in Europe. Party Politics, 23(1), 3–6.
Rahat, G., & Shapira, A. (2017). An Intra-party Democracy Index: Theory, Design and a
Demonstration, Parliamentary Affairs, 70(1), 84–110.
Sältzer, M. (2018). Ducking the Whip: Party Integration of German MPs on Twitter.
Paper Presented at the ECPR General Conference, Hamburg, August 22–25, 2018.
Schumacher, G., van der Velden, M., Hansen, D., & Kunst, S. (2018). A New Dataset of
Dutch and Danish Party Congress Speeches. Harvard Dataverse.
Schumacher, G., & Giger, N. (2017). Who Leads the Party? On Membership Size, Selec-
torates and Party Oligarchy. Political Studies, 65(1_suppl), 162–181.
Steiner, N.D., & Mader, M. (2017). Intra-party Heterogeneity in Policy Preferences and
its Effect on Issue Salience: Developing and Applying a Measure Based on elite Survey
Data. Party Politics, https://doi.org/10.1177/1354068817715553.
Temporão, M., Vande Kerckhove, C., Van der Linden, C., Dufresne, Y., & Hendrickx, J.
(2018). Ideological Scaling of Social Media Users: A Dynamic Lexicon Approach.
Political Analysis, 26(4), 457–473.
Verge, T., & Gómez, R. (2012). Factionalism in Multi-level Contexts: When Party
Organization Becomes a Device. Party Politics, 18(5), 667–685.
192 Conclusion
von Nostitz, F.-C., Banducci, S., Cioroianu, I., & Katz, G. (2018). Dynamic Cross-Party
Communities of Politicians. Paper Presented at the ECPR General Conference,
Hamburg, August 22–25, 2018.
Yale University Press, New Haven CT.
Appendix 1
Game–theoretic model: solutions through backward

induction
Depending on the different values of the various parameters (π, ε, ν, ω) the game
presents four alternative upshots (see Chapter 1): U1 (silent acceptance of the
unfair deal), U2 (compromise), U3 (compliance after the whip), U4 (breakaway).
However, only three outcomes (U1, U2, U4) are stable equilibria and they can
be found by backward induction.
1) The minority faction (F ) chooses to accept the unfair proposal bringing the
game to the outcome U1 under two scenarios:
a) when ε > π, provided that: π – ε < 0 – ν
Last stage: when ε > π, provided that: π – ε < 0 – ν, the minority’s reward after exit
is lower compared to the payoff gained when choosing comply (given that:
π – ε < 0 – ν). Then F will choose to comply.
Second stage: the payoff of the leader (L) for negotiating a compromise,
(1 – α)*µ – ν, is always lower than the reward of playing whip (µ – ν) as we know
that the minority will be prone to comply in the last stage. Assuming that µ is
equal to 1, for any value of α ε (0, 0.5) we observe that: 1 – α – ν < 1 – ν. In view of
that, L would always choose to whip the minority faction when the game reaches
this stage.
First stage: being aware that the expression of voice will lead the game to U3, in
the first stage F would rather play accept because its payoff in U1 (which is 0) is
higher than the gain available after compliance (0 – ν).
Therefore, when ε > π, provided that: π – ε < 0 – ν, the equilibrium is (accept,

comply; whip) and the payoffs will be (0; µ). Note that the ratio between ω and ν
then does not alter the dynamics of the game under this scenario so that the con-
ditions described above always lead to the equilibrium U1, no matter whether
ω > ν or vice versa.
194 Appendix 1
b) when ε > π and π – ε > 0 – ν, though only if ω < ν
Last stage: the minority’s payoff due to exit is greater than its reward after
comply insofar as π – ε > 0 – ν. Then F will always play exit.
Second stage: given that the minority would choose exit in the last stage, L faces
the choice between playing whip, which leads to the outcome U4, or com-
promise that provides the payoff displayed in U2. For any ω < ν we observe that
(1 – α)*µ – ω – (1 – α)*µ – ν; as a consequence the leader’s reward is larger when
playing whip, as this move leads to the outcome U4, instead of looking for a
compromise.
First stage: from this backward induction we discover that, at this stage, F has to
compare its payment in U1 (when it plays accept) and U4 (which will be the
consequence of choosing voice). Given that ε > π and therefore π – ε < 0, the
minority faction would rather play accept.
Therefore, when ε > π and π – ε > 0 – ν, provided that ω < ν, the equilibrium will be
(accept, exit; whip) and the payoffs are (0; µ).
2) The minority faction decides to voice and the leader will negotiate a com-
promise leading to the outcome U2 when π – ε > 0 – ν, tough only if ω > ν.
Last stage: Given that: π – ε > 0 – ν, playing exit guarantees to the minority a
reward larger than that provided by choosing comply. Accordingly, F will play
exit in the last stage.
Second stage: as long as the minority would choose exit in the last stage, L com-
pares the payoff of playing whip (which opens the route to the outcome U4) or
compromise that leads straightly to the outcome U2. For any ω > ν we notice that
(1 – α)*µ – ν > (1 – α)*µ – ω; given these payoffs, the leader is more willing to
negotiate a compromise instead of enacting the whip.
First stage: in the first stage F faces the choice between accept and voice, which
allows U2 to be reached. As we assumed that µ = 1, and ν < α, we can easily
verify that α*µ – ν > 0. In view of that, the minority faction finds it convenient to
play voice instead of accept.
Therefore, when π – ε > 0 – ν and ω > ν, the equilibrium is (voice, exit; com-
promise) and the payoffs are (α*µ – ν; (1 – α)*µ – ν).
3) The minority plays voice in the first stage, the leader’s reply is whip and, as a
consequence, the minority decides to exit in the last stage bringing the game to
the outcome U4, if and only if ε < π and ω < ν.
Appendix 1 195
Last stage: when ε < π, we observe that π – ε > 0 – ν. For that reason, the minority
would rather play exit, which gives back a larger payoff compared to comply.
Second stage: the leader has to choose between compromise (U2) and whip,
which ultimately leads to U4. Given that ω < ν, and accordingly
(1 – α)*µ – ω > (1 – α)*µ – ν, L discovers that it is more profitable to play whip
instead of compromise.
First stage: F has to choose between the payoff available in U1 and U4. As far as
ε < π, we notice that π – ε > 0 and consequently the minority will pick up the voice
option.
Therefore, when ε < π and ω < ν, the equilibrium is (voice, exit; whip) and the
payoffs are (π – ε; (1 – α)*µ – ω).
Appendix 2
Detailed list of factions’ positions in Italy,
France and Germany
The table below gives the estimates of the position of individual factions in each
party congress along the left–right scale (separately for Italy and France). The
positions of German subgroups are reported according to the year in which the
factional document was drafted. Negative values express left-wing positions;
positive values express right-wing positions. Notice that for Italy and Germany
the results refer to the two different Wordfish analyses that were run separately
in each country (see Chapter 2 for details). A star (*) indicates whether that
faction split from the party.
ITALY Left–Right Position

(Wordfish Estimates)
AN 2002
D-Destra* 1.459044
Destra Sociale 1.093735
Nuova Alleanza 1.054985
Destra Protagonista (Berluscones) 1.047826
DC 1954
Iniziativa Democratica –0.10293
Forze Sociali –0.25895
DC 1959
Primavera 0.044498
Centrismo Popolare –0.4425
Nuove Cronache –0.49352
Dorotei –0.56809
Base (Sinistra di Base) –0.67718
DC 1962
Centrismo Popolare 0.641803
Linea Moro-Fanfani 0.210654
DC 1964
Impegno Democratico (Dorotei) 0.316847
Centrismo Popolare –0.23121
Forze Nuove –0.32324
Appendix 2 197
DC 1967
Pontieri (Tavianei) 0.373126
Base –0.07593
Impegno Democratico (Dorotei) –0.13272
DC 1969
Forze Libere 1.16876
Ponte 0.593023
Impegno Democratico (Dorotei) 0.296835
Nuove Cronache 0.278839
Morotei 0.041232
Base e Forze Nuove –0.05006
Nuova Sinistra –0.072
DC 1980
Forze Nuove 0.366017
Iniziativa Popolare (Dorotei) 0.04033
Area Zac e Andreottiani –0.22031
Amici di Prandini –0.43351
DC 1982
PAF (Piccolo–Andreotti–Fanfani) 1.864879
Area De Mita (Base–AreaZac–others) 1.236874
Area Forlani 0.985618
DC 1984
Impegno Riformista (Scotti) e Forze Nuove 1.363929
De Mita (Base-others) 0.907559
DC 1986
Andreottiani 1.991661
De Mita (Base–others) 1.156328
DC 1989
Nuove Cronache 1.269407
Primavera 0.684833
Area del Confronto (Base) 0.089685
Azione Popolare (Dorotei) –0.1364
DS 2000
Veltroniani –0.34151
Sinistra Ds –0.42636
DS 2001
LibertàEguale (Liberal Ds) 0.043891
Riformisti (Fassino) –0.16876
Correntone (Sinistra Ds) –0.25832
DS 2005
Ecologisti Ds 0.541943
Riformisti (Fassino) 0.218535
Sinistra Ds per il Socialismo –0.22081
Correntone (Sinistra Ds) –0.22448
continued
198 Appendix 2

DS 2007
Riformisti (Fassino) 0.299215
Socialisti ed Europei* 0.031947
Correntone (Sinistra Ds)* –0.09264
FV 2008
Progetto Ecologista Federalista 0.984066
Ecologisti e Riformisti 0.903524
Futuro 0.153346
FV 2009
Futuro* 0.649128
Nuovi Verdi Nuovo Ulivo 0.56409
Ecologisti per la Costituente 0.320015
MSI 1965
Sinistra 1.098466
Spiritualisti 0.937569
Unità 0.508837
MSI 1977
Linea Futura 1.843722
Destra Popolare* 1.434709
Unità nella Chiarezza 1.181517
MSI 1979
Spazio Nuovo 1.751017
Continuare e Rinnovare 1.502459
MSI 1987
Andare Oltre 1.699732
Destra in Movimento 1.654327
Impegno Unitario 1.557026
Proposta Italia 1.466879
Nuove Prospettive 1.461168
Destra Italiana 1.349408
MSI 1990
Andare Oltre 0.73048
Destra in Movimento 0.652029
Nuove Prospettive 0.606023
Impegno Unitario 0.493075
Proposta Italia 0.464933
Destra Italiana 0.415289
NPSI 2003
Maggioranza (De Michelis) 0.468652
Socialismo e Libertà* 0.409608
NPSI 2005
Unità e Rinnovamento* 0.364814
Maggioranza (De Michelis) 0.121822
PCI 1989
Riformisti (Occhetto) –0.66631
Cossuttiani –1.24124
Appendix 2 199
PCI 1990
Riformisti (Occhetto) –0.93862
Rinnovamento (Ingrao) –1.04965
Cossuttiani –1.39972
PCI 1991
Antagonisti e Riformatori –0.83808
Per il Partito Democratico della Sinistra –1.03121
Per la Rifondazione Comunista* –1.12998
PD 2009
Vivi il PD/Cambia l’Italia (Marino) 0.895547
Area Democratica (Franceschini) 0.890285
Riformisti e Democratici (Bersani) 0.588933
PDA 1946
M aggioranza (Codignola) –0.5467
Democrazia Repubblicana* –0.9333
Autonomisti (Lombardi) –1.3212
PDCI 2008
Unire la Sinistra* –0.28469
Comunisti e Comuniste –0.99195
PLI 1966
Minoranza –0.05876
Maggioranza –0.06283
PLI 1969
Italia Liberale 1.449138
Presenza 0.954442
Libertà Nuova 0.497743
PLI 1971
Presenza 1.13387
Rinnovamento 1.098191
PLI 1973
Presenza 0.677456
PLI 1974
Concordia 0.7087
Unità Liberale 0.062677
Presenza –0.20775
PLI 1976
Autonomia Liberale 1.135356
Democrazia Liberale e Libertà Nuova 0.780794
continued
200 Appendix 2

PLI 1979
Democrazia Liberale 0.644463
PLI 1981
Nuove Iniziative 1.407666
PLI 1984
Nuove Iniziative 1.687732
PLI 1986
Nuova Democrazia Liberale 1.267091
Politica delle Libertà 1.242215
PLI 1988
Progetto Liberale 1.545205
Nuova Democrazia Liberale 1.419383
Politica delle Libertà 1.213491
PRC 1996
Bertinottiani e Cossuttiani –1.65048
Progetto Comunista –1.69118
PRC 1998
Pontieri* –0.73156
Cossuttiani* –1.12543
Bertinottiani –1.4339
PRC 1999
PRC 2002
PRC 2005
Essere Comunisti –1.30026
Sinistra Critica* –1.49916
Falce e Martello –1.60312
Progetto Comunista* –1.90193
PRC 2008
Pacifisti –0.9056
Rifondazione Per la Sinistra* –1.05695
Rifondazione in Movimento –1.28998
Falce e Martello –1.36915
L’Ernesto* –1.57714
Appendix 2 201
PRI 1950
PRI 1952
Sinistra Repubblicana –1.42395
PRI 1954
PRI 1956
PRI 1958
Unione Democratica –0.62423
PRI 1960
Unione Democratica –0.89322
PRI 1965
Destra 0.059085
PRI 1968
PRI 1984
Maggioranza 0.582068
Sinistra Repubblicana 0.400917
PRI 1987
Base 0.314384
PRI 1989
PSOC 2008
Unità Identità Autonomia e Sinistra Socialista 0.477024
Prima la Politica 0.247721
Un Nuovo Inizio 0.06582
PSDI (PSLI) 1948

Sinistra* –1.42855
continued
202 Appendix 2

PSDI 1957
Democrazia Socialista –0.93143
Autonomia Socialista* –1.07786
Fedeltà Socialista –1.16277
Unità Socialista* –1.22229
PSDI 1959
Autonomia e Unità Socialista –0.70292
Democrazia Socialista –0.78663
Rinnovamento Socialista –0.79571
Centrosinistra –0.93436
PSDI 1962
Iniziativa Socialdemocratica –0.4892
Centrosinistra –1.007
Rinnovamento Autonomia e Unità Socialista –1.00984
PSDI 1974
Democrazia Socialista 0.130791
Rinnovamento –0.02612
PSDI 1976
Socialismo Democratico 0.058287
Sinistra Socialdemocratica –0.56642
PSDI 1982
Socialismo Democratico 1.086165
Sinistra Socialdemocratica 0.777116
Sinistra Riformista 0.529498
PSDI 1987
Prospettiva Socialista Democratica* 1.078622
Iniziativa Socialdemocratica 0.458467
PSDI 1989
Autonomia Socialdemocratica 0.282067
Iniziativa Socialista –0.63268
PSI 1946
Iniziativa Socialista –1.22752
Critica Sociale –1.28284
Base (Sinistra) –1.40143
PSI 1947
Concentrazione Socialista (Critica Sociale)* –0.93963
Sinistra –1.25024
Iniziativa Socialista* –1.48961
PSI 1948 (January)
Autonomisti* –1.16378
Fronte –1.44416
Liste Separate –1.45709
Appendix 2 203
PSI 1948 (July)
Per il Socialismo* –1.15795
Riscossa Socialista –1.47544
Sinistra –1.5952
PSI 1949
Per il Socialismo –1.3833
Per il Partito e la Classe –1.42129
Sinistra –1.86065
PSI 1959
Alternativa Democratica –0.9693
Autonomia –1.35067
Sinistra –1.52821
PSI 1961
Alternativa Democratica –1.17569
Sinistra –1.38606
PSI 1963
Sinistra* –1.43178
PSI 1965
Sinistra Socialista –0.99181
PSI (PSU) 1968
Rinnovamento Socialista* –0.25559
Riscossa e Unità Socialista –0.65095
Impegno Socialista –0.71212
PSI 1978
Riformisti (Craxi) 0.326119
Sinistra per l’Alternativa –0.07566
Presenza Socialista –0.33449
Unità e Autonomia per l’Alternativa –0.58677
PSI 1981
Riformisti (Craxi) 0.470275
Sinistra per l’Alternativa* –0.01699
Presenza Socialista –0.26096
PSIUP 1972
Per la Confluenza nel PSI* –1.18405
Continuità e Rinnovamento Nuovo PSIUP* –1.20445
Per la Confluenza nel PCI –1.4575
UDC 2007
Popolari Liberali* 1.233613
continued
204 Appendix 2
FRANCE Left–Right Position

PCF 2000
Majorité (Hueistes) –1.0626
PCF 2003
Novateurs –1.07737
Gauche communiste, FNARC & Orthodoxes* –1.25931
Majorité (Hueistes–Refondateurs) –1.0381
PCF 2006
Colère et espoir* 0.458555
Lutte des classes –0.280265
Pour une union populaire 0.140888
Fièr(e)s d’être communiste –0.051976
La visée communiste 0.366115
Majorité 0.39806
PCF 2008
Renforcer le PCF (La Riposte) 0.060984
Faire vivre et renforcer le PCF 0.176937
Majorité 0.164906
PCF 2013
Renforcer le PCF (La Riposte) 0.120282
Ni abandon, ni effacement –0.14106
Unir les communistes 0.101503
Majorité 0.516215
PCF 2016
Pour un Front de gauche populaire et citoyen –1.00791
Pour une politique communiste –1.2103
Unir les communistes –1.14793
Reconstruisons le parti de classe –1.25251
Majorité 0.371071
PS 1971
CÉRÉS –1.4183
Chrétiens de gauche/Objectif 72 –1.1343
Mauroy–Deferre –1.28489
Mitterrandistes –1.388
Poperénistes –1.44215
Savary-Mollet –1.334
PS 1973
CÉRÉS –1.46083
Mitterrandistes (and others) –1.38769
Mollétistes (Fuzier) –1.32844
PS 1975
CÉRÉS –1.39497
Mitterrandistes (and others) –1.32311
Mollétistes (Fuzier) –1.34337
Notebart –1.46547
Appendix 2 205
PS 1976
Motion unique –1.09035
PS 1977
CÉRÉS –1.36599
PS 1979
CÉRÉS –1.39572
Defferre –1.36775
Dissidents CÉRÉS (Pierret) –1.14821
Féministes (Lhuillier) –1.40615
Mauroy –1.27501
Rocardiens –1.34915
PS 1981 (January)
Manifeste commun –1.29111
PS 1981 (October)
PS 1983
AGIRS (À Gauche pour l’Initiative, la Responsabilité –1.22715
et la Solidarité)
CÉRÉS –1.34016
Jospin-Mauroy-Rocardiens –1.29183
PS 1985
Jospin–Mauroy–CÉRÉS –1.20489
PS 1987
PS 1990
CÉRÉS (Socialisme et République)* –1.5748
Fabiusiens –1.10972
Gauche Socialiste (Nouvelle École Socialiste) 0.505271
Mauroy–Jospin –1.07615
SESAME/AGIRS –2.34406
PS 1991
Amendement Gauche Socialiste (not a full motion) –1.2273
Amendement Socialisme et République (not a full –1.06946
motion)
PS 1997
Gauche Socialiste –0.738471
Hollande (Réussir Ensemble) –0.586198
Réussir à Gauche (ex-poperénistes) –1.0371
continued
206 Appendix 2

PS 2000
Gauche Socialiste–ATTIKA –1.174
Démocratie & Égalité (Emmanuelli) –0.94992
Hollande 0.720801
PS 2003
Forces Militantes (Gauche) 0.671494
Hollande 0.795296
Nouveau Monde (Gauche) 0.743646
Nouveau Parti Socialiste/NPS (Gauche) 0.629974
Utopia 0.900831
PS 2005
La Gauche Moderne (Bockel)* 0.931406
Fabiusiens 0.879359
Hollande –0.128399
Nouveau Parti Socialiste/NPS 0.869725
Utopia 0.853817
PS 2008
Rénover Maintenant–Aubry–Fabiusiens 0.809195
La Ligne Claire (Collomb–Royal–Valls) 0.819124
NPS–Forces Militantes–Mélénchon(Gauche 0.605413
Socialiste)*
Delanoë–Hollande–Jospin–Rocardiens 0.764152
Pôle Écologique 0.654922
Utopia* 0.856111
PS 2012
Toulouse, mon congrès (Blanchard) 0.825197
Pour réussir le changement (Desir) 0.83638
Oser. Plus loin. Plus vite (Hessel) 0.973264
Maintenant la Gauche (Maurel) 0.918629
Question de principes (Méadel) 0.828007
PS 2015
Le Renouveau Socialiste 0.968476
(Cambadélis–Valls–Collomb–Aubry)
À Gauche (Maintenant la Gauche–Frondeurs–Hamon) 0.738208
Pacte Citoyenne Républicaine 0.755235
La Fabrique 0.866859
Cambadélis (profession de foi) 0.570797
Paul (profession de foi) 0.242541
UDF/MoDem 2006
Bayrou 0.705228
Nouveau Centre (Morin)* 0.939291
UDF/MoDem 2008
Bayrou 0.830217
UDF/MoDem 2010
Bayrou 0.668442
Appendix 2 207
UMP 2002
Droite Libre (Kaci) 0.812503
Juppé–Sarkozystes 0.924004
Ghazli 1.47105
Debout la République (Dupont–Aignan) 0.897561
UMP 2004
Forum des Républicains Sociaux (Boutin) 0.950649
Sarkozy–Sarkozystes 1.22537
Debout la République (Dupont-Aignan)* 1.353
UMP 2012
La Droite Sociale 0.977549
France Moderne et Humaniste 1.10771
La Boîte à Idées 0.812525
Gaullistes En Mouvement 0.988078
La Droite Forte 1.27656
La Droite Populaire 0.954009
Copé (profession de foi) 1.25617

Fillon (profession de foi) 1.04395
UMP 2014
Mariton 1.15157
Le Maire 0.912226
Sarkozy–Sarkozystes –0.015505
UMP 2016 (primaires)

Copé 1.34959
Juppé 1.50803
Morizet 1.18497
Poisson (Parti Chrétien-Démocrate) 1.18715
Le Maire* 0.431635
Fillon 1.31497
Sarkozy 0.219565
VERTS/EELV 1984
Motion unique/Texte d’orientation –1.36055
VERTS/EELV 2002
Désir Vert 0.740612
Ecolo 0.83268
Rénovons Maintenant 0.724516
Retrouver Convaincre la Société 0.774703
Singularités Verts 0.563103
Utopie 0.52885
VERTS/EELV 2004
Changer et Unir 0.703165
Verts Utiles 0.704087
Rassembler 0.746419
Regain Décidément 0.655522
Ecolo 0.868703
continued
208 Appendix 2

VERTS/EELV 2006
Ecologie Populaire 0.767969
Urgence Ecolo 0.812176
Verts et Forts 0.226415
Audaces! 0.764106
Ouvrir Reformer Rassembler 0.710098
Verts Utiles 0.404014
Espoir en Actes 0.822609
Alter-Ekolo –0.552531
VERTS/EELV 2016
L’écologie en commun 0.205255
Imprévu 0.652909
Europa 0.732062
Réinventer-Horizon2025 0.57114
TicTac 0.538998
GERMANY Left–Right Position

CDA (CDU)
1987 0.465234
1989 –0.264951
1993 –0.986282
1997 0.327365
2001 –0.125687
2002 –0.351412
2003 0.183221
2008 0.423225
2009 0.495879
2013 0.358412
2015 0.491924
MIT (CDU)
1985 0.952055
1989 1.0213
1993 1.27172
1997 0.723538
2003 1.63803
2005 1.12149
2009 0.781903
2011 0.841978
2014 0.715248
FK/DL21 (SPD)
1986 –1.20366
2001 –0.875678
2006 0.596826
2011 0.112804
2017 0.060901
Appendix 2 209
NB (SPD)
2003 –0.886002
2012 0.316248
SK (SPD)
1986 –1.07201
2003 0.150787
2005 0.348835
2006 0.558404
2015 0.26848
2017 0.343931
Index
backward induction 15, 21, 193 electoral cycle 19, 69, 76, 81–4, 100
Berlin: Network 37, 56, 124; Wall 3, 47–8 electoral performance 2–10, 69, 125,
Berlusconi, S. 29–30, 49, 82, 86n15, 141–52, 162–71, 181–7
105n8, 144–6 electoral system: French 119–21, 129;
Bersani, P. 29, 60n13, 86n14, 146–7 incentives of 9, 21, 92, 137–54, 178–86;
blackmail power 113, 117–8, 122, 131n17 intra-party 65–84; Italian 31–2, 163
bounded oligarchy 20, 104, 179 Epinay congress 33, 72–3, 170, 185
Brexit 2–3, 32 Eurosceptic 3, 34
exit cost 15–8, 67, 82, 131n13, 137–53,
cabinet instability 30–32 181–4
candidate selection 69, 74, 137–43, 149, exit, voice, and loyalty framework 13,
154, 179–84 22n11, 173
cartel party 13, 20, 93–5, 100, 104, 161
Central Committee 22n8, 73, 94, 97, 115 factions: affiliation 38, 113–4, 123, 138,
CÉRÉS 33, 53–4, 72, 149, 156n24 166, 188; ban on 30, 66-82, 178; of
Chevènement, P. 3, 33, 149 interests 8, 64, 69, 85n4; mainstream
closed list 31, 69, 76, 139–43, 147–8, 14–5, 29, 34, 54, 83, 101–3, 113–29,
181–4; see also electoral system 179–183; minority 9, 17, 54, 66, 72,
cohesion 4–10, 16–21, 66–7, 82–4, 92–3, 113, 113–29, 139–48, 180–2; of
136–59, 181–8 principles 8, 64, 69, 85n4
collective action dilemma 10 Finiani 30, 59n1–2, 144–6
Comparative Manifesto Project 45, 52, First Italian Republic 28, 41, 69, 114, 141,
95–100 165
congress motion 12, 28, 32–45, 84, 123 formateur 111–2, 127, 160
cooperation 1–10, 18–20, 91, 109, 129, Fronde 4, 151
182–7; cooperative norms 11, 182
corruption 9, 82, 166–9, 173 Gallagher index 141, 148, 156n22; see
cost of voice 18–9, 66, 70, 83–4, 153, 181 also electoral system
Craxi, B. 50, 94, 166 Gamson’s Law 20, 90, 109–11
Gauche Moderne 54, 149
De Mita, C. 166, 174n6 Gaullist parties 3, 34–5, 44, 54, 72, 94
democratic centralism 29–30, 34, 44, government duration 9, 116, 164
66–84, 139–53, 178–83 Grosse Coalition 4, 123–5, 152
direct election of leaders 11–3, 35, 67–74,
94, 147; effects 76–82, 97–104, 110–28, ideology 7–8, 36–9, 81, 178, 186
163–72, 177–85 inter-factional agreement 13, 44, 93–101,
Downs’ spatial theory of voting 89 112, 160, 183
Dupont-Aignan, N. 3, 34, 54, 149 internal dissent 6, 15–9, 70–3, 82–4,
Duverger’s Law 65, 84 136–53, 180
Index 211
internal polarization 38, 65, 76–83, 151–3, patronage 8–9, 68, 84, 178, 181, 186
163–72, 178–87 personal parties 29, 34, 40, 67–84, 165,
intra-party democracy 11, 21, 94, 105, 181
153, 187 policy preferences: heterogeneous 10,
intra-party divisions 2–6, 31–40, 145, 151; 65–7, 84, 178, 186; instrumental 8, 60n7
see also internal polarization populism 3, 48, 189
intra-party rules see party statute portfolio allocation 20–1, 109–29, 143–51,
165–72, 179–83
loyalty 17–8, 67, 131n13, 137–45, 149, position papers 37–8, 56
181–5 postdiction 21, 146
preference voting 68–9, 139–54, 181–4;
Macron, E. 4, 54, 151, 185–6, see also electoral system
Marxism 2, 47–8, 66, 73, 139 primary elections 6, 35, 54, 73, 147, 154
mass party 12, 89–95, 100, 161 principal–agent framework 22n4, 160,
Mélenchon, J.-L. 3–4, 33, 43, 149–51 173, 185
Members of Parliament 3, 35–8, 91, profession de foi 32, 35, 43
124–6, 138–46
Merkel, A. 2–5, 36, 123–7, 152, 172–3, rank-and-file 65, 82, 90–4, 147
186 Renzi, M. 4, 12, 30, 145–9, 163, 185–6
Mitterrand, F. 33, 54, 170 roll-call votes 39, 187
models of party 11, 91, 100, 184
Schröder, G. 4, 37, 123–5, 151
National Council 22n8, 73, 94, 97, 115, Schultz, M. 37, 186
130n2 Second Italian Republic 29–31, 41, 73–6,
non-cooperative bargaining theories 114, 140, 165
109–111 Seeheimer Kreis 4, 37, 55–7, 124
non-credible threat 15–6, 136 selectorate theory 161–3, 173
social media 4, 21, 145, 188–9
Pareto set 91–3 socialist party 2–3, 33, 52, 72, 140, 165
party competition 9, 48, 89–100, 184
Party Congress Research Group 188 Tangentopoli 47, 82, 166
party fission 1–21, 32–5, 95, 127, 135–53, Tendency 8, 72–9, 86n15
177–84 trade unions 2, 7, 56, 147, 187
party leader: autonomy 12, 17, 70, 183;
charisma 11–5, 40, 67, 82, 93, 18–9; Ultimatum Game 13, 111, 130n3
selection 11, 17, 74, 94; survival 13, 21, unitary congress 35, 40–1, 53, 66, 188
159–73, 177–80
party manifesto 5, 46–8, 92, 101–4 valence 11, 145, 160–6, 173; see also party
party membership 10, 113–22, 138, 162, leader: charisma
189
party statute 19, 40, 65, 81, 128, 185–7 whip 4–5, 15–18, 36, 126, 139–53, 185
party switching 9, 32, 35, 154n1 winner-takes-all 21, 92, 100, 112, 131n15,
party system competitiveness 9, 17–21, 182
113–29, 136–53, 155n21, 179–83 Wordfish 19, 45–59, 96, 105n10, 147

Leaders, Factions and The Game of Intra-Party Politics

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Leaders, Factions and The Game of Intra-Party Politics

Uploaded by

Copyright:

Available Formats

Leaders, Factions and the Game of

Andrea Ceron is Assistant Professor at the Department of Social and Political

The Regulation of Post-Communist Party Politics

Party Systems in Young Democracies

Opposition Parties in European Legislatures

Party Members and their Importance in Non-EU Countries

Do Parties Still Represent?

For more information about this series, please visit: www.routledge.com/

ISBN: 978-1-138-55000-1 (hbk)

List of figures viii

1 The game of intra-party politics 1

2 Factional preferences in Italy, France and Germany 28

3 The determinants of factionalism 64

4 Bounded oligarchy: leaders, factions and the selection of

5 Portfolio allocation among party factions 109

6 The politics of fission: party splits in Italy, France and

7 Intra-party conflict and the survival of party leaders 159

Appendix 1: game–theoretic model: solutions through

Appendix 2: detailed list of factions’ positions in Italy,

1.1 Trends in the literature on intra-party politics relative to

AfA Workers Association (Germany–SPD)

Literature on intra-party politics and party factions

terms: “tendencies” (Rose 1964), “fractions” (Sartori 1971, 1976), “currents”

Figure 1.2 Leader selection and autonomy.

Figure 1.3 Party Unity and Party Fission Game.

Leader’s Attitude Minority Faction’s Threat to Break Away

Non-credible Credible but Weak Credible and Strong

Focus on Unity U1: Unfair deal U2: Agreement U2: Agreement

Label Parameter Description Voice Compromise Exit

ε Exit cost: Loyalty and disproportional electoral systems – – –

Plan of the book

Factional politics: the Italian case

Beyond Italy: intra-party politics in France and Germany

Party congress, the arena where factional bargaining

Data collection on Italy, France and Germany

From words to fact(ion)s: estimating factions positions using

Policy positions of party factions in Italy

Figure 2.1 Factions’ position on the left–right scale (Italy).

Italian case: validity and reliability of the estimates

Figure 2.3 Discriminating power of words in Italy.

Policy positions of party factions in France

Figure 2.4 Reliability of Wordfish estimates compared to hand-coding.

Figure 2.5 Factions’ position on the left–right scale (France).

Figure 2.6 Factions’ position on the left–right scale (Germany).

Text analysis: robustness checks, insights and best practices

Stemming 0.96 0.99 0.99 0.99 0.97

of the documents. Finally, in light of the discussion on temporal trends, which

Hypothesis 3.1A (H3.1A): Intra-party polarization increases the number of

Hypothesis 3.2A (H3.2A): Disproportional internal electoral rules decrease the

Hypothesis 3.2B (H3.2B): Disproportional internal electoral rules decrease the

Hypothesis 3.3A (H3.3A): An explicit ban on factions should decrease their

Hypothesis 3.4B (H3.4B): In parties organized through “democratic centralism”

Hypothesis 3.5A (H3.5A): Direct election by a wide selectorate (which fosters

Hypothesis 3.5B (H3.5B): Direct election by a wide selectorate (which fosters

Hypothesis 3.6B (H3.6B): In “personal parties” the likelihood of a contested

Hypothesis 3.8A (H3.8A): Closed list PR and centralized candidate selection

A further hypothesis concerns the availability of office payoffs. Sartori (1976)

Hypothesis 3.10B (H3.10B): Looming elections should decrease the likelihood

The role of rules: party organization in Italy, France and

Degree of Representativeness of Each Party Rule

Intra-Party Factions Ban Democratic Direct

Maj. Mix PR Yes Some No Yes No Yes No

The determinants of factionalism in Italy

Internal Polarization 2.305 5.338

Gamsonian Agreement Position (GAP) 0.392* 0.410* 0.678* 0.544*

Share of Seats 6.183* 7.193* 6.920* 6.696*

Share of Votes 10.975* 16.914* 9.796***

Ideological Distance 0.784 1.046 –

Internal Polarization 1.942 1.994 2.879* 3.227* 3.264* 4.559*

Internal Polarization 1.789 2.048* 1.462*** 1.264*